Download DGfS 2016 Konstanz: Tagungsband

“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page a — #1
38. Jahrestagung
der Deutschen Gesellschaft
für Sprachwissenschaft
Sprachkonzil:
Theorie und Experiment
24.–26. Februar 2016
Universität Konstanz
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page b — #2
Umschlag/Plakat:
WWA-Grafik Universität Konstanz
Universitätsstraße 10
78464 Konstanz
Tel.: 07531 88-4317
E-mail: wwa-grafi[email protected]
Druck:
Hartmanndruck & Medien GmbH
Obere Gießwiesen 34
78247 Hilzingen
Tel.: 07731 8797-70
Fax.: 07731 8797-66
E-mail: [email protected]
Satz:
Dieser Tagungsband wurde mit XƎLATEX in den Schriften Linux Libertine und Sans PT gesetzt.
19. Januar 2016
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page i — #3
Grußwort der Organisatoren
Sehr geehrte Damen und Herren, liebe Kolleginnen und Kollegen,
wir freuen uns sehr, Sie zur 38. Jahrestagung der Deutschen Gesellschaft
ür Sprachwissenschaft an der Universität Konstanz begrüßen zu dürfen.
Nach 1999 findet die Jahrestagung zum zweiten Mal in Konstanz statt.
Das Jahr 2016 ist aus mindestens zwei Gründen ein besonderes Jahr ür
Konstanz. Unsere Universität wird 50 und das Konstanzer Konzil feiert
sein 600-jähriges Jubiläum. Konstanz wurde damals ausgewählt den größten Kongress des Mittelalters zu organisieren, unter anderem, weil es ausreichend Herbergen gab und die Speisen „nicht allzu teuer seien“. Diese
Gründe haben bei der Wahl von Konstanz ür die DGfS-Jahrestagung wahrscheinlich eher eine kleinere Rolle gespielt als die Tatsache, dass die Universität einen runden Geburtstag hat. Wir freuen uns, dass wir das Jubiläumsjahr durch die Ausrichtung der DGfS-Jahrestagung mit unseren Gästen
aus dem In- und Ausland feiern können.
Das Rahmenthema der Tagung haben wir bewusst vor diesem Hintergrund gewählt:
Sprakonzil: eorie und Experiment
Das Konzil verstehen wir als einen Ort der Zusammenkunft und des gemeinsamen zielgerichteten Austausches, das Sprachkonzil demnach als Ort
des Austausches über sprachwissenschaftliche Themen. „Theorie und Experiment“ soll natürlich nicht als Gegensatz verstanden werden. Für uns
sind theoretisch informiertes experimentelles Arbeiten und experimentell
untermauerte Theoriebildung zentrale Bausteine einer fundierten linguistischen Analyse von Sprache.
Mit dieser Herangehensweise war unser Fachbereich Sprachwissenschaft
(in Zusammenarbeit mit dem Fachbereich Informatik und Informations-
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page ii — #4
wissenschaft) in diesem Jahr erfolgreich bei der Beantragung einer Forschergruppe estions at the Interface (FOR 2111, Sprecherin: Miriam Butt),
die am 1. April 2016 starten wird. Nach einer Phase des Umbruchs und der
konstruktiven Neuorientierung ist es uns also gelungen, wieder Förderung
ür interessante Verbundforschung zu bekommen. Wir freuen uns alle darüber und darauf und möchten diese gute Nachricht an dieser Stelle gerne
mit Ihnen teilen.
Wir wünschen Ihnen allen eine interessante Tagung, einen anregenden
wissenschaftlichen Austausch und einen schönen und erfolgreichen Aufenthalt in Konstanz.
Mit herzlichen Grüßen
Nicole Dehé & Janet Grijzenhout
ii
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page iii — #5
Organisation
Federführend/Wissensalie Leitung
Nicole Dehé & Janet Grijzenhout
Organisatorise Leitung
Anna Ney, Veranstaltungsmanagement, Universität Konstanz
Organisationsteam
Tina Bögel, Miriam Butt, Simon Dold, Regine Eckardt, Constantin Freitag,
Annette Hautli-Janisz, Katharina Kaiser, Georg Kaiser, Achim Kleinmann,
Svenja Kornher, Tanja Kupisch, Alexandra Rehn, Janina Reinhardt,
Maribel Romero, Jana Schlegel, Gloria Sigwarth, Sebastian Sulger,
Andreas Trotzke, Yvonne Viesel, Carmen Widera, Daniela Wochner &
Irene Wolke.
iii
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page iv — #6
Danksagungen
Die Organisatorinnen und Organisatoren bedanken sich herzlich bei den
folgenden Sponsoren:
• Dr. August und Annelies Karst Stiftung (www.profil.uni-konstanz.de/
stiften-und-foerdern/dr-august-und-annelies-karst-stiftung)
• Universitätsgesellschaft Konstanz e.V. (www.ugk.uni-konstanz.de)
• Verein der Ehemaligen der Universität Konstanz (www.veuk.uni-konstanz.de)
• Walter de Gruyter Stiftung (www.walterdegruyter-stiftung.com)
• AThEME (www.mehrsprachigkeit.uni-konstanz.de/informationen/atheme)
• BRILL (www.brill.com)
• De Gruyter Mouton (www.degruyter.com)
• DUDEN Bibliographisches Institut GmbH (www.duden.de)
• Erich Schmidt Verlag GmbH & Co. KG (www.esv.info)
• Frank & Timme GmbH Verlag ür wissenschaftliche Literatur (www.franktimme.de)
• Franz Steiner Verlag (www.steiner-verlag.de)
• Helmut Buske Verlag GmbH (www.buske.de)
• ibidem-Verlag (www.ibidemverlag.de)
• IUDICIUM Verlag GmbH (www.iudicium.de)
• J.B. Metzler’sche Verlagsbuchhandlung und C. E. Poeschel Verlag
GmbH Stuttgart·Weimar (www.metzlerverlag.de)
• John Benjamins Publishing Company db (www.benjamins.com)
• LINCOM GmbH (www.lincom.eu)
iv
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page v — #7
• Missing Link (www.missing-link.de)
• Narr Francke Attempto Verlag GmbH + Co. KG (www.narr.de)
• Oxford University Press (www.oup.com)
• Peter Lang GmbH (www.peterlang.com)
• Stauffenburg Verlag Brigitte Narr GmbH Tübingen/Julius Groos Verlag Tübingen (www.stauffenburg.de)
• Universitätsverlag Winter GmbH (www.winter-verlag.de)
• utb GmbH (www.utb.de)
• Verlag C.H.Beck oHG (www.chbeck.de)
• Waxmann Verlag GmbH (www.waxmann.com)
Stand bei Redaktionsschluss
Die Lehramtsinitiative der DGfS bedankt sich bei dem Referat ür Gleichstellung und Familienörderung und der Binational School of Education,
Universität Konstanz.
Das Team des Tagungsbandes bedankt sich ganz herzlich bei Fabian
Heck und Matthias Schrinner, die uns den LATEX-Code des DGfS-Tagungsbandes 2015 zur Verügung gestellt haben. Der Code dieses Bandes wird, in
guter Open Source-Tradition, selbstverständlich auch allen Interessierten
weitergegeben. Zu diesem Zwecke einfach bei Constantin Freitag mailden
([email protected]).
v
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page vi — #8
Dr. August und
Annelies
Karst
Stiung
vi
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page vii — #9
Inhaltsverzeichnis
Grußwort der Organisatoren
I.
i
Informationen
1
Informationen zur Tagung
3
Anreise
9
Lageplan und Raumübersicht
13
Essen und Trinken
19
II. Programmübersicht und AG-Programme
33
Programmübersicht
35
AG Programme
37
III. Plenarvorträge
61
IV. Arbeitsgruppen und Abstracts
67
Arbeitsgruppe 1
Verb second in grammar and processing: its causes and its
consequences
69
vii
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page viii — #10
Inhaltsverzeichnis
Arbeitsgruppe 2
The syntax of argument structure: empirical advancements and
theoretical relevance
103
Arbeitsgruppe 3
Agentivity and event structure: Theoretical and experimental
approaches
125
Arbeitsgruppe 4
Sentence complexity at the boundary of grammatical theory
and processing: A special challenge for language acquisition
149
Arbeitsgruppe 5
The grammatical realization of polarity: Theoretical and
experimental approaches
177
Arbeitsgruppe 6
Computational Pragmatics
201
Arbeitsgruppe 7
Sign language agreement revisited: New theoretical and
experimental perspectives
217
Arbeitsgruppe 8
Gradienz im Spannungsfeld von empirischen Methoden und
Grammatiktheorie
247
Arbeitsgruppe 9
Geschriebene und gesprochene Sprache als Modalitäten eines
Sprachsystems
269
Arbeitsgruppe 10
Morphological effects on word order from a typological and a
diachronic perspective
291
viii
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page ix — #11
Inhaltsverzeichnis
Arbeitsgruppe 11
Indefinites between theory and language change
311
Arbeitsgruppe 12
Presuppositions in language acquisition
331
Arbeitsgruppe 13
Adjective order: Theory and experiment
341
V. Sektionenprogramm
361
Postersession der Sektion Computerlinguistik
363
Tutorium der Sektion Computerlinguistik
391
Doktorandenforum
393
Infotag der Lehramtsinitiative der DGfS
397
Tagung der Arbeitsgemeinschaft Linguistische Pragmatik
405
VI. Anhang
409
Notizen
411
Gesamtübersicht der Arbeitsgruppensitzungen
417
Personenverzeichnis
421
ix
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page x — #12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 1 — #13
Teil I.
Informationen
1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 2 — #14
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 3 — #15
i
Informationen zur Tagung
Veranstalter
Fachbereich Sprachwissenschaft, Universität Konstanz
Deutsche Gesellschaft ür Sprachwissenschaft (DGfS)
Wissenschaftliche Leitung
Prof. Dr. Nicole Dehé, Fachbereich Sprachwissenschaft
Prof. Dr. Janet Grijzenhout, Fachbereich Sprachwissenschaft
Organisatorische Leitung
Anna Ney, Veranstaltungsmanagement
Homepage
www.dgfs2016.uni-konstanz.de
Tagungsort
Universität Konstanz
Universitätsstraße 10
78464 Konstanz
www.uni-konstanz.de
Tagungsbüro: Anmeldung und Info-Point
Die Teilnehmerregistration erfolgt ab dem 23.2. im Tagungsbüro. Dort erhalten Sie Ihre Teilnahmeunterlagen und alle wichtigen Informationen zur
Tagung. Das Tagungsbüro ist während der gesamten Tagung besetzt und
dient als zentrale Anlaufstelle ür alle Fragen. Hier finden Sie außerdem das
Fundbüro und eine Ansprechperson ür Notälle (Erste Hilfe). Am 23. und
3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 4 — #16
i
Informationen zur Tagung
24.2. befindet sich das Tagungsbüro im Foyer auf der Ebene A5 (Haupteingang).
Ab dem 25.2. befindet sich das Tagungsbüro zentral im Foyer, in der Nähe der Fachausstellung und des Caterings.
Öffnungszeiten des Tagungsbüros/Garderobe
Dienstag, 23. 02. 2016, 8:30–10:30, 14:30–19:00 Uhr
Mittwoch, 24. 02. 2016, 8:00–19:00 Uhr
Donnerstag, 25. 02. 2016, 8:30–19:00 Uhr
Freitag, 26. 02. 2016, 08:30–15:00 Uhr
Telefon
Bei dringenden Fragen können Sie das Tagungsbüro während der oben genannten Zeiten unter folgender Telefonnummer erreichen:
(+49) 07531 88-5276
Internetnutzung an der Universität Konstanz – W-LAN
Um sich im ‚conference‘-WLAN anzumelden, klicken Sie auf das Funknetz
conference auf Ihrem Gerät und geben in der Anmeldemaske folgenden
Benutzernamen und das Kennwort ein:
Benutzername: JahrestagungDGFS
Kennwort: dgfs2016
Für Hochschulangehörige wird alternativ der Zugang über Eduroam angeboten (weitere Informationen unter www.eduroam.org). Bitte beachten
Sie, dass in der Mensa grundsätzlich kein W-LAN Empfang besteht.
Kopieren und drucken - Canon Druckcenter
Sollten Sie kurzfristig Unterlagen vor Ort ausdrucken wollen, erhalten Sie
hierür am Tagungsbüro eine Kopierkarte ür das Canon-Druckcenter (neben dem Haupteingang).
4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 5 — #17
Informationen zur Tagung
Erste Hilfe
Am Info-Point finden Sie stets eine Ansprechperson ür Notälle. Außerdem kann über das Konferenz-Personal jederzeit telefonisch Hilfe angefordert werden.
Garderobe
Während der Tagung besteht die Möglichkeit, einen Garderobenservice zu
nutzen. Wenden Sie sich hierzu bitte an das Tagungsbüro.
Catering
Während der Konferenz werden auf der Ebene A5 in den Pausen verschiedene Erfrischungen (Kaffee, Kaltgetränke und Snacks) angeboten.
In der Mensa auf Ebene K6 erhalten Sie täglich jeweils von 11:15–13:45
Uhr verschiedene Mittagsgerichte (Selbstzahler). Ein Informationsblatt zur
Orientierung in der Mensa liegt Ihren Tagungsunterlagen bei, die Sie am
ersten Tagungstag bei der Anmeldung erhalten.
Weitere Angebote:
• Snack- und Getränkeautomaten(Ebene K4)
• Campus Café (Ebene A5)
• seezeit Café in der Bibliothek (Ebene B4)
• Asia Bistro Arche (Ebene K4)
Barrierefreiheit
Die Räumlichkeiten der Konferenz sind barrierefrei.
Anmeldung
http://www.dgfs2016.uni-konstanz.de/anmeldung
5
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 6 — #18
i
Informationen zur Tagung
Tagungsgebühren
Anmeldung bis 23.1.2016
DGfS-Mitglieder mit Einkommen
DGfS-Mitglieder ohne Einkommen
Nicht-Mitglieder mit Einkommen
Nicht-Mitglieder ohne Einkommen
50€
35€
70€
40€
Anmeldung ab 24.1.2016
DGfS-Mitglieder mit Einkommen
DGfS-Mitglieder ohne Einkommen
Nicht-Mitglieder mit Einkommen
Nicht-Mitglieder ohne Einkommen
55€
40€
75€
45€
Bankverbindung
Universitätskasse Konstanz
BW-Bank Konstanz:
Konto Nr.: 7486501274
BLZ: 60050101
IBAN: DE92600501017486501274
BIC: SOLADEST
Referenz: Vorname Nachname DGfS 2016
Rahmenprogramm
Warming-up, Dienstag, 23. 02. 2016, ab 19:00 Uhr im Restaurant Brauhaus
(Selbstzahler)
Sektempfang, Mittwoch, 24. 02. 2016, 19:00 Uhr Rathaus der Stadt Konstanz
(kostenfrei)
Geselliger Abend am Donnerstag, 25. 02. 2016, 19:00 Uhr im Konzil Konstanz (38€)
6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 7 — #19
Informationen zur Tagung
Hotelreservierung
Wir haben über die Tourist Information Konstanz Zimmerkontingente in
Hotels reserviert. Sie finden diese unter:
http://www.konstanz-tourismus.de/themen/tagungen/dgfs-tagung
Geldautomat
Im Eingangsbereich der Universität, neben dem Campus Café, befindet sich
ein Geldautomat der Sparkasse.
Mobilfunknetz
Durch die Nähe zur Schweiz kann es vorkommen, dass bei automatischer
Netzwahl ein Schweizer Netzbetreiber ausgewählt wird, wodurch höhere
Kosten entstehen können. Dies kann vermieden werden, indem die automatische Netzwahl deaktiviert und der Netzbetreiber manuell ausgewählt
wird.
Fachausstellung
Bitte besuchen Sie auch die Fachausstellung der Verlage im Foyer auf A5!
Kontakt Organisatorische Leitung
Anna Ney
Universität Konstanz
Kommunikation und Marketing
Veranstaltungsmanagement
Universitätsstraße 10
78464 Konstanz
Telefon: +49 (0)7531 88-5276
E-Mail: [email protected]
7
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 8 — #20
Albrecht Schöne
Der Briefschreiber
Goethe
539 S., zahlr. Abb. Ln.
€ 29,95
ISBN 978-3-406-67603-1
„Von brillanter Klarheit und mitunter geradezu atemberaubend
zu lesen.“ Hubert Spiegel, Frankfurter Allgemeine Zeitung
„Was für ein Glück, diesen Albrecht Schöne zu haben.“
Gustav Seibt, Süddeutsche Zeitung
C.H.BECK
www.c h b ec k. d e
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 9 — #21
i
Anreise
Anreise nach Konstanz
Mit der Bahn
Zielbahnhof: Konstanz Bahnhof
Die Universität Konstanz bietet Ihnen in Kooperation mit der Deutschen
Bahn ein exklusives Angebot ür Ihre bequeme An- und Abreise zu Veranstaltungen der Universität Konstanz 2016 an. So wird ür Sie Reisezeit
ganz schnell zu Ihrer Zeit. Nutzen Sie Ihre Hin- und Rückfahrt einfach
zum Arbeiten, Lesen oder Entspannen. Für was Sie sich auch entscheiden,
Sie reisen in jedem Fall mit dem Veranstaltungsticket im Fernverkehr der
Deutschen Bahn mit 100% Ökostrom.
Der Preis ür Ihr Veranstaltungsticket zur bundesweiten Hin- und Rückfahrt beträgt:
Mit Zugbindung
2. Klasse
1. Klasse
99,– €
159,– €
Vollflexibel
2. Klasse
1. Klasse
139,– €
199,– €
Den Link zur Buchung Ihres DB-Veranstaltungstickets finden Sie auf der
Tagungswebsite unter:
www.dgfs2016.uni-konstanz.de/anreise-und-unterkunft.
Mit dem Auto
Von Stuttgart (180 km)
A 81 in Richtung Singen. Ab dem Kreuz Hegau ist Konstanz ausgeschildert.
In Konstanz folgen Sie den Wegweisern „Universität“.
9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 10 — #22
i
Anreise
Von München (220 km)
A 96 in Richtung Lindau. Bei der Ausfahrt Sigmarszell auf die B 31 in Richtung Friedrichshafen - Meersburg. Von Meersburg mit der Autoähre nach
Konstanz. In Konstanz folgen Sie den Wegweisern „Universität“.
Von Zürich (75 km)
Autobahn A1 Richtung St. Gallen, bei der Verzweigung Winterthur-Ost auf
die A7 Richtung Kreuzlingen/Konstanz. Nach der Grenze richten Sie sich
zunächst nach „Mainau“. Ausschilderung „Universität“ beachten.
Mit dem Flugzeug
Nächstgelegene Flughäfen
Flughafen Zürich (ZRH)
Bodensee-Airport Friedrichshafen (FDH)
Per Bahn vom Flughafen Zürich
Es fahren halbstündlich Züge nach Konstanz. Fahrzeit 1:06 (ohne Umstieg),
1:15 (1 Umstieg)
Montag bis Freitag: zwischen 5.15 und 23.18 Uhr
Samstag: zwischen 5.47 und 0.57 Uhr
Sonntag: zwischen 5.47 und 23.18 Uhr
Ihre individuelle Routenplanung mit den Schweizer Bundesbahnen:
www.sbb.ch
Mit öffentlichen Verkehrsmittel vom Bodensee-Airport
Mit der Bahn (IRE) über Radolfzell (1 Umstieg).
Mit der Bahn bis Friedrichshafen, Hafenbahnhof und weiter mit dem Katamaran nach Konstanz Hafen (1 Umstieg).
Für Ihre individuelle Anreise vom Bodensee-Airport: www.qixxit.de
10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 11 — #23
Anfahrt zur Universität
Anfahrt zur Universität
Die Universität Konstanz können Sie bequem mit dem Bus erreichen. Zusätzlich zu den Linienbussen bieten wir zu den Hauptzeiten zusätzlich kostenlose Shuttlebusse an. Besucher, die länger als eine Nacht in Konstanz
bleiben, zahlen 2€ Kurtaxe pro Person und Nacht und erhalten den Bodensee-Gästepass, dem ein Gutschein zur kostenlosen Linienbus-Nutzung beiliegt. Bitte fragen Sie gegebenenfalls in Ihrem Hotel nach. Weitere Informationen zum Bodensee-Gästepass unter: www.konstanz-tourismus.de/uebernachten/kurtaxe-gaestepass.html. Für die Fahrten mit den Shuttlebussen
benötigen Sie keinen Fahrschein.
Ein Stadt-, Lage- und Buslinienplan liegt Ihren Tagungsunterlagen bei,
die Sie am ersten Tagungstag bei der Anmeldung erhalten.
Linienbusse
Linie 9 (A und B)
Montag bis Freitag: Im 15-Minuten-Takt vom Zentrum direkt zum Haupteingang der Universität Einstieg z. B. an der Haltestelle Bahnhof – Sternenplatz – Zähringerplatz. Ausstieg an der Endhaltestelle Universität (Haupteingang).
www.stadtwerke-konstanz.de/fileadmin/content/download/bus/
Fahrplaene_2016/Linie_9ABx.pdf
Linien 4/13 und 13/4
Montag bis Freitag: ab 6:00 Uhr (Bahnho) im 30 Minuten-Takt (zwischen
15:30 Uhr und 18:30 Uhr im 15-Minuten-Takt)
Die Linien 4 /13 (Stadt Richtung Universität) und 13/4 (Universität Richtung Stadt) verbinden die Universität mit der Mainau und den Ortsteilen
Litzelstetten, Dingelsdorf, Oberdorf, Wallhausen und Dettingen bzw. Staad
in Richtung Innenstadt. Ausstieg an der Haltestelle Egg/Universität (Fußweg zum Haupteingang: etwa 10 Min.).
www.stadtwerke-konstanz.de/fileadmin/content/download/bus/
Fahrplaene_2016/Linie_4_13_13_4x.pdf
11
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 12 — #24
i
Anreise
Mit dem Taxi
Dornheim Telefon: +49 (0) 7531 5 77 77
Müller Telefon: +49 (0) 7531 454 96 97
Seeteufel Telefon: +49 (0) 7531 44 9 44
Mit dem Auto – Parken
In Konstanz folgen Sie den Wegweisern „Universität“. Wir empfehlen das
Parken im Parkhaus Süd (siehe Übersichtsplan). Von dort ist der Weg zum
Konferenzort ausgeschildert. Die Parkgebühren betragen 1,30 Euro pro Tag.
Diese können entweder bar (Geld muss passend eingeworfen werden) oder
per SMS bezahlt werden. Zur Bezahlung per SMS senden Sie einfach ihr
KFZ-Kennzeichen an die Kurzwahl-Nr. 83115. Mit Erhalt der BestätigungsSMS ist Ihr Tagesticket bezahlt.
12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 13 — #25
i
Lageplan und Raumübersicht
Die Konferenz findet auf dem Campus der Universität Konstanz statt. An
der Universität werden die Gebäude mit Buchstaben benannt. Die darauf
folgende Zahl bezeichnet die Ebene. Die letzten zwei Ziffern bezeichnen die
Raumnummer. Die Plenarvorträge finden im Audimax statt. Das Audimax
(A600 = Gebäude A, Ebene 6, Raum 00) befindet sich im Hauptgebäude A
der Universität. Die AGs tagen in Seminarräumen in den Gebäuden D, E,
G und F.
Die Mensa befindet sich im Gebäude K, Ebene 6.
Die Wege zu den Tagungsräumlichkeiten werden ausgeschildert.
13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 14 — #26
i
Lageplan und Raumübersicht
Übersichtsplan Universität Konstanz
14
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 15 — #27
Übersichtsplan Foyer A5
Übersichtsplan Foyer A5
Der folgende Übersichtsplan zeigt einen Ausschnitt der Ebene 5 des Hauptgebäudes A und des Foyers der Universität Konstanz. Auf dieser Ebene liegen die beiden Haupteingänge zur Universität. Im A-Gebäude finden die
Plenarvorträge statt. Zudem befinden sich hier die Anmeldung, das Tagungsbüro, das Catering, die Verlagsausstellung sowie die Posterausstellung CL.
15
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 16 — #28
i
Lageplan und Raumübersicht
Raumübersicht
23. Februar 2016
Tagungsbüro
A5-Foyer
Arbeitskreis Linguistische Pragmatik
G 201
CL-Tutorium
G 227
Doktorandenforum
F 425
Lehramtsinitiative
C 230, 421, 422, 423, 424, 427
24.– 26. Februar 2016
Plenarvorträge
Lisa Lai-Shen Cheng (Leiden
Universiteit)
A600
(Audimax)
Louise McNally (Universitat Pompeu
Fabra)
David Poeppel (New York
University)
Höskuldur Þráinsson (Háskóli
Íslands)
Arbeitsgruppe 1
Verb second in grammar and
processing: Its causes and its
consequences
G 309
Arbeitsgruppe 2
The syntax of argument structure:
Empirical advancements and
theoretical relevance
G 300
16
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 17 — #29
Raumübersicht
Arbeitsgruppe 3
Agentivity and event structure:
Theoretical and experimental
approaches
F 426
Arbeitsgruppe 4
Sentence complexity at the
boundary of grammatical theory and
processing: A special challenge for
language acquisition
G 530
Arbeitsgruppe 5
The grammatical realization of
polarity: Theoretical and
experimental approaches
F 425
Arbeitsgruppe 6
Computational pragmatics
F 420
Arbeitsgruppe 7
Sign language agreement revisited:
New theoretical and experimental
perspectives
D 406
Arbeitsgruppe 8
Gradienz im Spannungsfeld von
empirischen Methoden und
Grammatiktheorie
E 404
Arbeitsgruppe 9
Geschriebene und gesprochene
Sprache als Modalitäten eines
Sprachsystems
E 403
Arbeitsgruppe 10
Morphological effects on word order
from a typological and a diachronic
perspective
G 308
Arbeitsgruppe 11
Indefinites between theory and
language change
G 201
Arbeitsgruppe 12
Presuppositions in language
acquisition
E 402
17
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 18 — #30
i
Lageplan und Raumübersicht
Arbeitsgruppe 13
Adjective order: Theory and
experiment
E 402
Anmeldung
A5-Foyer unter der Empore
Tagungsbüro
A5-Foyer
Kaffeepausen
A5-Foyer
Gepäckaufbewahrung
A5-Foyer
Verlagsausstellung
A5-Foyer
DGfS-Mitgliederversammlung
A 701
Postersession CL
A5-Foyer
CL-Mitgliederversammlung
F 429
18
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 19 — #31
i
Essen und Trinken
Die folgende Übersicht ist ein Ausschnitt von Restaurants, Kneipen, Cafés, Weinstuben und Bars in Konstanz. Alle liegen im Stadtzentrum oder
zentrumsnah. Angegebene Preise beziehen sich auf Hauptgerichte.
Regionale & Internationale Küche
Name & Adresse
Lage
Preis
Barbarossa
Obermarkt 8-12, 78462 Konstanz
www.hotelbarbarossa.de
Zentrum
Restauration Bodan Bar Restaurant
Grill
Bodanstr. 4, 78462 Konstanz
www.restauration-bodan.de
Zentrum
10–24€
Brasserie Ignaz
Bahnhofplatz 5, 78462 Konstanz
www.brasserie-ignaz.de
Zentrum
10-21€
Brauhaus Johann Albrecht
Konradigasse 2, 78462 Konstanz
www.brauhaus-joh-albrecht.de
Zentrum
10-23€
Brigantinus
Reichenaustr. 15, 78467 Konstanz
www.brigantinus.de
Am Seerhein
15-40€
19
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 20 — #32
i
Essen und Trinken
Name & Adresse
Lage
Preis
Chez Léon
Zollernstraße 1, 78462 Konstanz
www.chezleon.de
Zentrum
10-36€
Constanzer Wirtshaus
Spanierstr. 3, 78467 Konstanz
www.constanzer-wirtshaus.de
Am Seerhein
10-25€
DELI
Bodanstrasse 1-3, 78462 Konstanz
www.deli-konstanz.de
Zentrum/LagoShoppingcenter
10-26€
Dischinger
Untere Laube 49, 78462 Konstanz
www.dischinger-kn.de
Zentrum
ab 14€
DOM
Brückengasse 1, 78462 Konstanz
www.dom-kon.com
Zentrum
8,20-30€
Dominikanerstube im Steigenberger
Inselhotel
Auf der Insel 1, 78462 Konstanz
www.konstanz.steigenberger.de
Zentrum
Elefanten
Salmannsweilergasse 32-34, 78462
Konstanz
Zentrum
Ess | bar
Bahnhofstr. 15, 78462 Konstanz
http://essbar-konstanz.de
Zentrum
20
ab 7,70€
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 21 — #33
Essen und Trinken
Name & Adresse
Lage
Preis
Eugens Bio Cafe-Restaurant
Münzgasse 1, 78462 Konstanz
www.eugens-bio.de
Zentrum
8,50-24€
Goldener Sternen
Bodanplatz 1, 78462 Konstanz
www.goldener-sternen-konstanz.de
Zentrum
Hafenhalle
Hafenstr. 10, 78462 Konstanz
www.hafenhalle.com
Zentrum/Hafen
8,50-26€
Hafenmeisterei
Hafenstraße 8, 78462 Konstanz
www.hafenmeisterei.de
Zentrum/Hafen
6-25€
Hollys
Reichenaustr. 19, 78467 Konstanz
www.hollys.de
Am Seerhein
10-18€
Hotel Halm und Maurischer Saal
Bahnhofplatz 6, 78462 Konstanz
www.hotel-halm-konstanz.de
Zentrum
10-22€
Konstanzer Bürgerstuben
Bahnhofplatz 7, 78462 Konstanz
www.konstanzer-buergerstuben.de
Zentrum
8-18€
Konzil-Gaststätten
Hafenstr. 2, 78462 Konstanz
www.konzil-konstanz.de
Zentrum
9,50-27€
Krone Restaurant Kaffeehaus
Brotlaube 2, 78462 Konstanz
www.krone-konstanz.de
Zentrum
10-21€
21
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 22 — #34
i
Essen und Trinken
Name & Adresse
Lage
Umami
Sigismundstr. 12, 78462 Konstanz
www.mato-konstanz.de
Zentrum
Münsterhof
Münsterplatz 3, 78462 Konstanz
www.muensterhof-konstanz.de
Zentrum
Rosie’s Pavillon am See
Stadtgarten, 78462 Konstanz
www.facebook.com/PavillonKonstanz
Zentrum/Am
See
Roter Gugelhan
Salmannsweilergasse 12, 78462
Konstanz
www.rotergugelhan.de
Zentrum
7-17€
Steg 4
Hafenstr. 8, 78462 Konstanz
www.steg4.de
Zentrum/Hafen
6-26€
Storikenescht
Döbelestr. 3, 78462 Konstanz
www.storik.de
Zentrum/Stadtteil 6-22€
Paradies
Suppengrün
Sigismundstr. 19, 78462 Konstanz
www.suppengruen.biz
Zentrum
ab 4,20€
Tolle Knolle
Bodanplatz 9, 78462 Konstanz
www.tolle-knolle.de
Zentrum
8-19€
22
Preis
8,30-23€
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 23 — #35
Essen und Trinken
Name & Adresse
Lage
Preis
TURM
Hussenstr. 66, 78462 Konstanz
www.cafe-turm.com
Zentrum
ab 8,50€
VIDA Store Konstanz – Bistrot-Café
Neugasse 20, 78462 Konstanz
www.eatdifferent.de
Zentrum
Wessenberg Café und Restaurant
Wessenbergstr. 41, 78462 Konstanz
www.wessenberg.eu
Zentrum
Zeitlos
St.-Stephansplatz 25, 78462 Konstanz
www.cafe-zeitlos.net
Zentrum
Zur Wendelgard
Inselgasse 5, 78462 Konstanz
www.zur-wendelgardkonstanz.xregional.de
Zentrum/
Stadtteil Niederburg
Zum Pfannkuchen
Hüetlinstr. 39, 78462 Konstanz
Zentrum/
Schweizer
Grenze
9-23€
Gehobene Küche
Name & Adresse
Lage
Preis
Friedrichs
Reichenaustr. 17, 78467 Konstanz
www.restaurant-friedrichs.de
Am Seerhein
14-36€
23
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 24 — #36
i
Essen und Trinken
Name & Adresse
Lage
Preis
Papageno
Hüetlinstraße 8a, 78462 Konstanz
www.restaurant-papageno.net
Zentrum/
Schweizer
Grenze
ab 30€
San Martino
Bruderturmgasse 3, 78462 Konstanz
www.san-martino.ne
Zentrum
ab 26€
Seerestaurant im Steigenberger
Inselhotel
Auf der Insel 1, 78462 Konstanz
www.konstanz.steigenberger.de
Zentrum
ab 28€
Name & Adresse
Lage
Preis
Bangkok
Brauneggerstr. 47, 78462 Konstanz
www.bangkok-konstanz.de
Zentrum/
Stadtteil Paradies
5–20€
Bo Dai Tei
Stadelhofgasse 1, 78462 Konstanz
www.bodaitei.de
Zentrum
3–25€
China Restaurant
Marktstätte. 30, 78462 Konstanz
Zentrum
Hanoi
Wallgutstr. 3, 78462 Konstanz
Zentrum/
Stadtteil Paradies
Asiatisch
24
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 25 — #37
Essen und Trinken
Name & Adresse
Lage
Preis
Karma
Sigismundstr. 14, 78462 Konstanz
www.karma-konstanz.de
Zentrum
Maharani
Konradigasse 1, 78462 Konstanz
www.maharani-konstanz.de
Zentrum
Mandarin
Bodanplatz 4, 78462 Konstanz
www.mandarin-konstanz.de
Zentrum
8–15€
Mayura
Am Fischmarkt 1, 78462 Konstanz
www.mayura-restaurant.de
Zentrum
5–20€
Ratsstube
Kanzleistraße 16 78462 Konstanz
Zentrum
3–8€
Sitara
Paradiesstraße 7, 78462 Konstanz
www.sitara-restaurant.de
Zentrum
6–20€
The Rambagh Palace
Brückengasse 1, 78462 Konstanz
www.inselmedia.de/rambagh-palace
Zentrum
25
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 26 — #38
i
Essen und Trinken
Griechisch/Türkisch
26
Name & Adresse
Lage
Preis
Akropolis
Obere Laube 55, 78462 Konstanz
www.akropolis-restaurant.de
Zentrum/
Stadtteil Paradies
Delphi
Brauneggerstr. 46, 78462 Konstanz
www.delphi-konstanz.de
Zentrum/
Stadtteil Paradies
Eumel
Hüetlinstraße 23, 78462 Konstanz
Zentrum/
Schweizer
Grenze
Radieschen
Hohenhausgasse 1, 78462 Konstanz
www.radieschen-konstanz.de
Zentrum
6–13€
Sedir
Hofhalde 11, 78462 Konstanz
www.cafe-restaurant-sedir.de
Zentrum
6–11€
Stephans Keller
St. Stephansplatz 41, 78462 Konstanz
www.stephanskeller.com
Zentrum
6–15€
Taverna Pan
Salmannsweilergasse 13, 78462
Konstanz
Zentrum
7–15€
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 27 — #39
Essen und Trinken
Italienisch
Name & Adresse
Lage
Preis
Vinothek+Osteria Cantina Rabaja
Kreuzlingerstr. 7, 78462 Konstanz
www.cantina-rabaja.de
Zentrum
14–31€
Casablanca
Marktstätte 15, 78462 Konstanz
www.casablanca-kn.de
Zentrum
7–28€
Don Alfredo
Hofhalde 7, 78462 Konstanz
www.restaurant-donalfredo.de
Zentrum
9–27€
Il Boccone
Bodanstr. 20-26, 78462 Konstanz
www.ilboccone.de
Zentrum
10-32€
L’Anima
Konzilstraße 3, 78462 Konstanz
Zentrum
La Grotta
Untere Laube 33, 78462 Konstanz
Zentrum/
Stadtteil Paradies
La Paesana Trattoria-Pizzeria
Hussenstraße 44, 78462 Konstanz
www.paesana.de
Zentrum
8–20€
La Piazza Ristorante-Pizzeria
Marktstätte 2, 78462 Konstanz
www.lapiazza-kn.de/main.htm
Zentrum
8–25€
L’Italiano
Bruderturmgasse 2, 78462 Konstanz
Zentrum
27
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 28 — #40
i
Essen und Trinken
Name & Adresse
Lage
Preis
Löhlinbad
Untere Laube 9, 78462 Konstanz
Zentrum/
Stadtteil Paradies
Pastis
Hohenhausgasse 14, 78462 Konstanz
Zentrum
Rigg’s Burger Restaurant
Bodanstraße 23, 78462 Konstanz
www.riggs-burger.com
Zentrum
8–16€
Pinocchio
Untere Laube 47, 78462 Konstanz
www.pinocchio-konstanz.de
Zentrum/
Stadtteil Paradies
11–24€
Spanisch/Mexikanisch
28
Name & Adresse
Lage
Costa del Sol
St. Johanngasse 9, 78462 Konstanz
Zentrum
La Bodega Tapasbar
Schreibergasse 40, 78462 Konstanz
Zentrum
Latinos
Am Fischmarkt, 78462 Konstanz
www.latinos-konstanz.de
Zentrum
Preis
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 29 — #41
Essen und Trinken
Steakhaus
Name & Adresse
Lage
Preis
Hexenküche Steakhaus
Bodanstr. 30, 78462 Konstanz
www.hexenkueche-steakhaus.de
Zentrum
9–32€
Name & Adresse
Lage
Preis
Pano Konstanz
Marktstätte 6, 78462 Konstanz
www.pano.coop
Zentrum
Rosgarten Café
Rosgartenstr. 9, 78462 Konstanz
www.rosgarten-cafe.de
Zentrum
Stadtkind Konstanz
Braunegger Str. 31, 78462 Konstanz
www.stadtkind-konstanz.de
Zentrum/
Stadtteil Paradies
Das Voglhaus Café
Wessenbergstr. 8, 78462 Konstanz
www.das-voglhaus.de
Zentrum
Cafés
10–15€
29
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 30 — #42
i
Essen und Trinken
Weinstuben
30
Name & Adresse
Lage
Weinstube Bürgertröpfle
Hüetlinstr. 13, 78462 Konstanz
Zentrum/Schweizer
Grenze
Weinkeller Franz Fritz Niederburg
Niederburggasse 7, 78462 Konstanz
www.weinhandlung-fritz.de
Zentrum/Stadtteil
Niederburg
Weinstube Hintertürle
Konradigasse 3, 78462 Konstanz
www.hintertürle.de
Zentrum/Stadtteil
Niederburg
Weinstube Küfer Fritz „Zum Pfohl“
Salmannsweilergasse 7-11, 78462
Konstanz
Zentrum
Weinstube Niederburg
Niederburggasse 7, 78462 Konstanz
www.weinhandlung-fritz.de
Zentrum/Stadtteil
Niederburg
Weinstube Weinglöckle
Inselgasse 13, 78462 Konstanz
Zentrum/Stadtteil
Niederburg
Weinstube Weinteufele
Konradigasse, 78462 Konstanz
Zentrum/Stadtteil
Niederburg
Weinstube Weinteufele
Konradigasse, 78462 Konstanz
Zentrum/Stadtteil
Niederburg
Weinstube Zum Guten Hirten
Zollernstr. 6-8, 78462 Konstanz
www.tamaras-weinstube.de
Zentrum
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 31 — #43
Essen und Trinken
Name & Adresse
Lage
Wein- und Bierstube Zum Salzbüchsle
Salmannsweilergasse 26,
78462 Konstanz
www.salzbuechsle.de
Zentrum
Weinstube Zur Steinernen Kugel
Hohenhausgasse 8, 78462 Konstanz
Zentrum
Bars und Pubs
Bar 107
Paradiesstr. 5, 78462 Konstanz
Zentrum
Bar Café Pfiff
Konzilstr. 1, 78462 Konstanz
Zentrum
Casba
Obere Laube 55, 78462 Konstanz
Zentrum/Stadtteil Paradies
die cocktailbar
St. Johanngasse 4, 78462 Konstanz
www.die-cocktailbar.de
Zentrum/Stadtteil
Niederburg
DOM Konstanz
Brückengasse 1, 78462 Konstanz
www.dom-konstanz.de
Zentrum/Stadtteil
Niederburg
Globetrotter Cocktailbar
Hüetlinstraße 14, 78462 Konstanz
www.globetrotter-bar.de
Zentrum/Schweizer
Grenze
Heimat Bar
Schreibergasse 2, 78462 Konstanz
www.heimatbar.de
Zentrum/Stadtteil
Niederburg
31
i
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 32 — #44
i
Essen und Trinken
32
K9
Hieronymusgasse 3, 78462 Konstanz
www.k9-kulturzentrum.de
Zentrum
Klimperkasten
Bodanstraße 40, 78462 Konstanz
Zentrum
Logans Pub
Zogelmannstr. 2, 78462 Konstanz
http://logans-pub.de
Zentrum
Old Mary’s Pub
Kreuzlingerstr. 19, 78462 Konstanz
Zentrum/Schweizer
Grenze
Shamrock
Bahnhofstraße 4, 78462 Konstanz
www.shamrock-konstanz.de
Zentrum
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 33 — #45
Teil II.
Programmübersicht und AG-Programme
33
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 34 — #46
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 35 — #47
Programm
Programmübersicht
Dienstag, 23. 02. 2016
8:45–18:00
10:00–17:00
9:00–18.30
15:00–18:45
ab 19:00
Tagung der Arbeitsgemeinschaft
Linguistische Pragmatik (ALP)
G 201
Computerlinguistik Tutorium
G 227
Doktorandenforum
F 425
Lehramtsinitiative
Plenum: C 230
Workshops: C 421, C 422, C 423, C 424, C 427
Warming-up im Brauhaus, Konradigasse
2, 78462 Konstanz
Mittwoch, 24. 02. 2016
8:00–9:00
Registrierung
Foyer (A5)
Begrüßung,
Plenarvortrag: Lisa Cheng (Leiden
University),
Verleihung des Wilhelm von
Humboldt-Preises
A 600
(Audimax)
Kaffeepause
Foyer (A5)
12:00–13:00
Plenarvortrag: Louise McNally
(Universität Pompeu Fabra)
A 600
(Audimax)
13:00–14:00
Mitgliederversammlung Sektion
Computerlinguistik
F 429
9:00–11:30
11:30
35
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 36 — #48
Programmübersicht
14:00–16:00
Programm
16:00
16:30–18:30
ab 19:00
Arbeitsgruppensitzungen
Gebäude D, E, F, G
Kaffeepause
Foyer (A5)
Arbeitsgruppensitzungen
Gebäude D, E, F, G
Sektempfang im Rathaus Konstanz,
Kanzleistraße 13/15
Donnerstag, 25. 02. 2016
9:00–11:00
11:00
Arbeitsgruppensitzungen
Gebäude C, D, E, F, G
Kaffeepause
Foyer (A5)
11:30–13:00
Arbeitsgruppensitzungen
Gebäude D, E, F, G
13:00–14:30
Postersession Sektion
Computerlinguistik und Mittagspause
Foyer (A5)
14:30–18:30
Mitgliederversammlung der DGfS
A 701
ab 19 Uhr
Geselliger Abend im Konzil Konstanz,
Hafenstraße 2, 78462 Konstanz
Freitag, 26. 02. 2016
9:00–11:00
11:00
11:30–14:00
36
Plenarvorträge:
David Poeppel (New York University)
Höskuldur Þráinsson (University of
Iceland)
A 600
(Audimax)
Kaffeepause
Foyer (A5)
Arbeitsgruppensitzungen
Gebäude D, E, F, G
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 37 — #49
Programm
AG Programme
AG 1
Verb second in grammar and processing: Its
causes and its consequences
Oliver Bott, Constantin Freitag & Fabian Schlotterbeck
Raum: G 309
Mittwoch, 24. 02. 2016
14:00–14:30
Oliver Bott, Constantin Freitag & Fabian Schlotterbeck
Introduction
14:30–15:00
Heike Wiese, Eva Wittenberg & Oliver Bunk
Variations on V2: e information-structural dynamics of
the le periphery in German
15:00–16:00
Sten Vikner
e derivation of V2 in Germanic main and embedded
clauses
16:00–16:30
Kaffepause
16:30–17:30
Jan Casalicchio & Federica Cognola
Relaxed V2 languages and their Le Periphery. Two cases
from Northern Italy
17:30–18:30
Bettelou Los & Ans van Kemenade
V2 in the history of English: why did it arise, why was it
lost, and what difference did it make?
37
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 38 — #50
AG Programme
Donnerstag, 25. 02. 2016
Programm
9:00–10:00
Markus Bader
e role of V2 in sentence comprehension and sentence
production
10:00–10:30
Peter de Swart und Geertje van Bergen
Effects of verbal information in the V2-position during
parsing: What eye movements reveal about prediction
(and integration)
10:30–11:00
Bettina Braun und Eva Smolka
Hör endlich auf/zu (‘Now stop/listen’)! – e Lexical
Representation and Semantic Activation of German
Particle Verbs
11:00–11:30
Kaffeepause
11:30–12:00
Isaac Gould
Modeling Verb Placement Errors in Swiss German
Children’s L1 Acquisition
12:00–12:30
Emanuela Sanfelici, Corinna Trabandt & Petra Schulz
On the nature of integrated V2 relative clauses
12:30–13:00
Sophie Repp
Semantic restrictions in verb-second vs. non-verb-second
wh-exclamatives
Freitag, 26. 02. 2016
11:30–12:00
Kajsa Djärv, Caroline Heycock & Hannah Rohde
Embedded V2, Factivity and Main Point of Uerance
12:00–12:30
Rebecca Woods
A Different Perspective on Embedded V2: Unifying
Embedded Root Phenomena
38
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 39 — #51
AG Programme
12:30–13:00
Nicholas Catasso
…Obwohl Nebensätze können doch auch assertiv sein: On
the disambiguating role of V2 in COMP-introduced
adverbial clauses
13:00–13:30
Thomas Roeper & Rebecca Woods
Separating Tense and Assertion: Evidence from
Embedded V2 and Child Language
13:30–14:00
Oliver Bott, Constantin Freitag & Fabian Schlotterbeck
Summary Discussion
AG 2
The syntax of argument structure: Empirical
advancements and theoretical relevance
Artemis Alexiadou & Elisabeth Verhoeven
Raum: G 300
Mittwoch, 24. 02. 2016
14:00–15:00
Maria Polinsky
e relationship between theoretical and empirical
syntax
15:00–16:00
Patrick Brandt & Petra Schumacher
Effects of repairing illegal argument structures
16:00–16:30
Kaffeepause
16:30–17:30
Nino Grillo, Berit Gehrke, Nils Hirsch, Caterina
Paolazzi & Andrea Santi
It’s all about verb-type: Passives are not inherently more
complex than actives
17:30–18:00
Patricia Irwin
Discourse and unaccusativity: antitative effects of a
structural phenomenon
39
Programm
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 40 — #52
AG Programme
18:00–18:30
Programm
Dmitry Ganenkov
Relativization in two morphologically ergative
languages: a corpus study
Donnerstag, 25. 02. 2016
9:00–10:00
Isabel Oltra-Massuet, V. Sharpe, K. Neophytou & Alec
Marantz
Syntactic priming as a test of argument structure: A
self-paced reading experiment
10:00–11:00
Linnaea Stockall, Christina Manouilidou, Laura
Gwilliams & Alec Marantz
Un/Re-packing argument and event structure restrictions
on prefixation: MEG evidence
11:00–11:30
Kaffeepause
11:30–12:30
Paul Kiparsky
On the syntax and argument structure of agent nouns
12:30–13:00
Tibor Kiss
Argument structure and reflexive binding
Freitag, 26. 02. 2016
11:30–12:30
Helen de Hoop
Grammar under pressure: the case of subject hun ’them’
in Dutch
12:30–13:00
Anna Czypionka & Carsten Eulitz
Case marking affects the processing of animacy with
simple verbs, but not particle verbs: An event-related
potential study
40
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 41 — #53
AG Programme
13:00–13:30
13:30–14:00
AG 3
Sandra Pappert, Michael Baumann & Thomas
Pechmann
e issue of lexical guidance in sentence production:
Evidence from structural priming experiments
Programm
Sabine Reuters, Sarah Verlage & Martina Penke
Animacy effects in German sentence production
Agentivity and event structure: Theoretical
and experimental approaches
Beatrice Primus & Markus Philipp
Raum: F 426
Mittwoch, 24. 02. 2016
14:00–14:30
Welcome / Introduction
14:30–15:00
Simon Kasper
e perceptual and sociocultural foundations of
agentivity in language
15:00–16:00
Franziska Kretzschmar, Svenja Lüll, Phillip Alday, Ina
Bornkessel- Schlesewsky & Matthias Schlesewsky
Actor prototypicality in the comprehension of
intransitive clauses in German
16:00–16:30
Kaffepause
16:30–17:00
Valentina Apresjan
Agentivity, control and semantic structure in Russian
Causatives
17:00–18:00
Fabienne Martin
On atypical agents
18:00–18:30
Tim Graf, Markus Philipp & Beatrice Primus
Agentivity and impersonal passives
41
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 42 — #54
AG Programme
Donnerstag, 25. 02. 2016
Programm
9:00–9:30
Gianina Iordachioaia
Agentivity and Eventivity in Psych Nominalizations
9:30–10:00
Vasiliki Koukoulioti & Stavroula Stavrakaki
e situation of aspect from the viewpoint of language
pathology: A comparison between stroke induced aphasia
and semantic dementia
10:00–11:00
Stefan Hinterwimmer
Experiencer Verbs as Indicators of Perspective-Taking
11:00–11:30
Kaffeepause
11:30–12:00
Marta Donazzan & Lucia M. Tovena
Agentive dispositions and causal responsibility: a case
study
12:00–13:00
Robert D. Van Valin Jr.
Agents, effectors and event structure
Freitag, 26. 02. 2016
11:30–12:00
Anke Lensch
Agentivity in Nominalizations of Phrasal Verbs. On
Passers-by and Winder-uppers
12:00–12:30
Ekaterina Gabrovska & Wilhelm Geuder
Agentivity and Force Exertion: the German Verb
”schlagen”
12:30–13:00
Hamida Demirdache, Angeliek van Hout, Jinhong Liu,
Fabienne Martin & Iris M. Strangmann
Testing the Agent Control Hypothesis with
non-culminating events. Experimental evidence from
Adult Dutch and Mandarin
13:00–13:30
Odelia Ahdout
Psych Nominalizations in Hebrew
42
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 43 — #55
AG Programme
13:30–14:00
Final discussion (optional)
Programm
AG 4
Sentence complexity at the boundary of
grammatical theory and processing: A special
challenge for language acquisition
Flavia Adani, Tom Fritzsche & Theodoros Marinis
Raum: G 530
Mittwoch, 24. 02. 2016
14:00–14:30
Flavia Adani, Tom Fritzsche & Theo Marinis
Introduction
14:30–15:30
Luigi Rizzi
Intervention effects in adult grammar and language
acquisition
15:30–16:00
Virginia Valian
What Two-Year-Olds Know; What Two-Year-Olds Say
16:00–16:30
Kaffeepause
16:30–17:00
Elena Pagliarini & Fabrizio Arosio
Processing of object clitics in Italian monolingual children
17:00–17:30
Rasha Zebib, Cornelia Hamann, Philippe Prévost, Lina
Abed Ibrahim & Laurice Tuller
Syntactic complexity, verbal working memory, and
executive function in bilingual children with and without
Specific Language Impairment: a sentence repetition
study in France and in Germany
17:30–18:30
Atty Schouwenaars, Esther Ruigendijk & Petra
Hendriks
Which questions do German children process in an
adult-like fashion?
43
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 44 — #56
AG Programme
Donnerstag, 25. 02. 2016
Programm
9:00–10:00
Shravan Vasishth
Complexity and Memory
10:00–11:00
Yair Haendler
Children’s processing of relative clauses depends on who
’they’ are
11:00–11:30
Kaffeepause
11:30–12:00
Lars Meyer
Processing versus Grammar of Syntactic Dependencies:
Neural Oscillations of Chunking, Storage, and Retrieval
12:00–12:30
Iya Khelm Price & Jeffrey Witzel
Misalignment of offline and online measures in Russian
relative clause processing
12:30–13:00
Irina A. Sekerina
Retrieval Interference in Relative Clause Aachment
Ambiguity: Cross-Linguistic Evidence
Freitag, 26. 02. 2016
11:30–12:00
Jill de Villiers & Tom Roeper
How representations determine stages of acquisition
12:00–13:00
Corinna Trabandt, Emanuela Sanfelici & Petra Schulz
What does semantic complexity mean for children? –
Insights from the acquisition of relative clauses in
German
13:00–13:30
Daniele Panizza & Karoliina Lohiniva
When pragmatics helps syntax: An eye tracking study on
scope ambiguity resolution in 4- to 5-year-old children
44
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 45 — #57
AG Programme
13:30–14:00
AG 5
Laura E. de Ruiter, Anna L. Theakston, Silke Brandt&
Elena V. M. Lieven
Temporal, causal and conditional sentences in English
child-directed speech
Programm
The grammatical realization of polarity.
Theoretical and experimental approaches
Christine Dimroth & Stefan Sudhoff
Raum: F 425
Mittwoch, 24. 02. 2016
14:00–14:30
Christine Dimroth & Stefan Sudhoff
e grammatical realization of polarity. Introductory
remarks
14:30–15:00
Beata Gyuris
On two types of polar interrogatives in Hungarian and
their interaction with inside and outside negation
15:00–16:00
Horst Lohnstein
Verum focus and contrast
16:00–16:30
Kaffeepause
16:30–17:00
Peter Öhl
Negative polarity, focus accent and the embedding of
interrogatives by veridical predicates
17:00–17:30
Julia Bacskai-Atkari
Complementisers as markers of negative polarity in
German comparatives
17:30–18:00
Unaisa Khir Eldeen
Can but be a negative polarity item?
45
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 46 — #58
AG Programme
18:00–18:30
Programm
Leah Roberts & Beatrice Szczepek-Reed
Establishing polarity contrasts in English: Evidence from
analyses of natural conversations
Donnerstag, 25. 02. 2016
9:00–10:00
Giuseppina Turco
Contrasting polarity: Verum focus and affirmative
particles in Germanic and Romance languages
10:00–11:00
Berry Claus, Marlijn Meijer, Sophie Repp & Manfred
Krifka
Polarity particles in response to negated antecedents: Two
groups of speakers for German ja and nein
11:00–11:30
Kaffeepause
11:30–12:00
Cecilia Andorno & Claudia Crocco
In search for verum focus marking in Italian: a
contribution from Map Tasks data
12:00–12:30
Heiko Seeliger & Sophie Repp
Rejections and rejecting questions: Declaratives with
clause-initial negation in Swedish
12:30–13:00
Anja Arnhold, Bettina Braun, Filippo Domaneschi &
Maribel Romero
Prosodic realization of Verum Focus in English polar
questions
Freitag, 26. 02. 2016
11:30–12:00
46
Dejan Matić, Irina Nikolaeva
Polarity focus across languages: Processes vs. things
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 47 — #59
AG Programme
12:00–13:00
Daniel Gutzmann, Katharina Hartmann, Lisa
Matthewson
Cross-linguistic evidence that verum ≠ focus
13:00–13:30
Davide Garassino, Daniel Jacob
Non Canonical Syntax and Polarity Focus in French,
Italian, and Spanish
13:30–14:00
Kyoko Sano
Assertion and Polarity in koso -e construction in Old
Japanese
Programm
Computational Pragmatics
AG 6
Anton Benz, Ralf Klabunde, Sebastian Reuße &
Jon Stevens
Raum: F 420
Mittwoch, 24. 02. 2016
14:00–15:00
Harry Bunt
Computational pragmatics revisited
15:00–16:00
Sabine Janzen & Wolfgang Maaß
Balancing dialogues with mixed motives
16:00–16:30
Kaffeepause
16:30–17:30
Martín Villalba & Alexander Koller
Interactive natural language generation in virtual
environments
17:30–18:30
Kees van Deemter
Computational models of choice in language production:
the case of reference
47
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 48 — #60
AG Programme
Donnerstag, 25. 02. 2016
Programm
9:00–10:00
Jon Stevens
e turnip question: A game-theoretic look at non-literal
answers
10:00–11:00
Michael Franke & Leon Bergen
Embedded scalars and reasoning about the QUD
11:00–11:30
Kaffeepause
11:30–12:30
Noah Goodman
Unusual uncertainty in language understanding:
Vagueness and accommodation
Freitag, 26. 02. 2016
11:30–12:00
Mark-Matthias Zymla
Pragmatic inferencing via abstract knowledge
representation in LFG
12:00–12:30
Florian Kuhn
Towards building a German legal decision corpus for
argumentation mining
12:30–13:00
Eva Horch
Article missing?
13:00–13:30
Simon Musgrave, Michael Haugh & Andrea Schalley
Looking for a good laugh: Using ontologies to access
pragmatic phenomena through spoken corpora
13:30–14:00
Schlussbemerkungen
48
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 49 — #61
AG Programme
AG 7
Sign language agreement revisited: New
theoretical and experimental perspectives
Barbara Hänel-Faulhaber, Annika Herrmann,
Christian Rathmann & Markus Steinbach
Programm
Raum: D 406
Mittwoch, 24. 02. 2016
14:30–15:00
Elena Benedicto
Classifiers as agreement … or not?
15:00–15:30
Adam Schembri
Against the agreement analysis of ’classifier’ morphemes
in sign languages
15:30–16:00
Svetlana Dachkovsky
e development of a RC marker from a deictic gesture in
Israeli Sign Language
16:00–16:30
Kaffepause
16:30–17:00
Lynn Y-S Hou
Pointing as seeds of directionality
17:00–17:30
Kearsy Cormier
Role shi is not agreement
17:30–18:30
Richard P. Meier
Pointing to the analysis of personal pronouns and
directional verbs in the acquisition and grammar of ASL
Donnerstag, 25. 02. 2016
9:00–9:30
Irit Meir
Explaining the special typological properties of Sign
Language verb agreement
49
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 50 — #62
AG Programme
Programm
9:30–10:00
Kearsy Cormier, Jordan Fenlon & Adam Schembri
”Agreement” verbs in sign languages: Are we missing the
point?
10:00–10:30
Carlo Geraci, Mirko Santoro, Lara Mantovan &
Valentina Aristodemo
Backward agreement is not so backward aer all: the role
of loci in the grammar of SL
10:30–11:00
Guilherme Lourenço
Regular and backward agreement verbs in Libras: a
Case-based derivation
11:00–11:30
Kaffeepause
11:30–12:00
Antonio Balvet & Brigitte Garcia
Agreement in Sign Languages, allow me to disagree
12:00–13:00
Josep Quer
A place for locative agreement in sign languages
Freitag, 26. 02. 2016
11:30–12:00
Julia Krebs, Dietmar Roehm & Ronnie Wilbur
Two agreement markers in Austrian Sign Language
(ÖGS)
12:00–12:30
Matic Pavlič
Verb-argument agreement and word order in SZJ
ditransitives
12:30–13:00
Brendan Costello
”Defective” agreeing verbs in LSE: an OT account
13:00–13:30
Jeremy Kuhn
Dependency marking in American Sign Language
13:30–14:00
Roland Pfau & Martin Salzmann
e order of Agree and Merge - evidence from sign
language agreement
50
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 51 — #63
AG Programme
AG 8
Gradienz im Spannungsfeld von empirischen
Methoden und Grammatiktheorie
Jana Häussler & Tom Juzek
Programm
Raum: E 404
Mittwoch, 24. 02. 2016
14:00–14:30
Jana Häussler & Tom Juzek
Introduction - Why gradience maers
14:30–15:00
Ankelien Schippers
Negative island violations: *?#how bad are(n’t) they
really?
15:00–16:00
Robert Külpmann & Vilma Symanczyk Joppe
Gradient Acceptability and Categorical Distinctions: the
Case of Imperative Constructions
16:00–16:30
Kaffepause
16:30–17:30
Jennifer Culbertson (invited speaker)
Competing grammars and the representation of subject
clitics in French
17:30–18:00
Gabi Danon
Choose your features: Lexical optionality and variation
in QNP agreement
18:00–18:30
Julie Franck, Garrett Smith & Whitney Tabor
A theory of agreement araction based on a continous
semantic representation space
Donnerstag, 25. 02. 2016
9:00–10:00
Emilia Ellsiepen
Problematic cases for weighted constraint models:
Subadditivity and cost-free violation
51
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 52 — #64
AG Programme
Programm
10:00–11:00
Marta Wierzba
Towards an efficient evaluation method of generative
theories using gradient data and regression
11:00–11:30
Kaffeepause
11:30–12:30
Antonella Sorace (invited speaker)
Gradience at interfaces
12:30–13:00
Frances Blanchette
A Gradient Acceptability Study of English Sentences with
Two Negatives
Freitag, 26. 02. 2016
11:30–12:00
Anne-Laure Besnard
asi-Modals in English
12:00–13:00
Joanna Nykiel
Ellipsis alternation
13:00–13:30
Chiyo Nishida
Dative clitic doubling variation in Spanish reverse psych
verb sentences: Syntax meets discourse-pragmatics and
semantics
13:30–14:00
Matthias Schrinner
Competing embedded clauses in German: Conflicts in
position of extraposed relative and argument clauses
52
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 53 — #65
AG Programme
AG 9
Geschriebene und gesprochene Sprache als
Modalitäten eines Sprachsystems
Martin Evertz & Frank Kirchhoff
Programm
Raum: E 403
Mittwoch, 24. 02. 2016
14:00–14:30
Martin Evertz & Frank Kirchhoff
Einührung und Überblick
14:30–15:00
Andreas Nolda
Script types: Definition and classification
15:00–15:30
Hartmut Günther
Schri- und Lautsprache – eine Übersetzungstheorie
15:30–16:00
Monika Budde
Modalitätsübergreifende und modalitätsspezifische
Strukturaspekte sprachlicher Äußerungen: Auf welchen
Beschreibungsebenen können sich modalitätsspezifische
Unterschiede manifestieren?
16:00–16:30
Kaffeepause
16:30–17:00
Karsten Schmidt
Das Wort als zentrale Einheit der Graphematik
17:00–17:30
Vilma Symanczyk Joppe
Nur ein Reflex der Morphosyntax? Das graphematische
Wort in Norm und Gebrauch
17:30–18:00
Fabian Renz
Expressive Intensitätspartikeln in gesprochener und
geschriebener Sprache
53
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 54 — #66
AG Programme
Donnerstag, 25. 02. 2016
Programm
9:30–10:00
Tilo Reißig
Verknüpfungsstrukturen listenmodaler Texte –
Schnistellenphänomene und modalitätsspezifische
Ausprägungen
10:00–10:30
Janina Kalbertodt, Beatrice Primus & Petra Schumacher
Right dislocations and aerthoughts: the effects of
punctuation and discourse structure on prosody
10:30–11:00
Ilka Huesmann & Frank Kirchhoff
Interpunktion und Intonation bei Interjektionen im
Deutschen
11:00–11:30
Kaffeepause
11:30–12:00
Katharina Turgay & Daniel Gutzmann
Zur (ortho)grafischen Markierung von sekundären
Inhalten: eine empirische Studie
12:00–12:30
Kristian Berg
Homophone und Heterographen
Freitag, 26. 02. 2016
11:30–12:00
Nana Fuhrhop
Silbenkerne im Spannungsfeld zwischen Laut und Schri
oder Silbenkerne als modalitätsübergreifende Einheit
12:00–12:30
Silke Hamann
One phonotactic restriction for reading and listening: e
case of the no geminate constraint in German
12:30–13:00
Sabine Wahl
Gesprochen, geschrieben, gesungen – Sprache in
Werbespots
13:00–13:30
Schlussbemerkungen
54
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 55 — #67
AG Programme
AG 10
Morphological effects on word order from a
typological and a diachronic perspective
Þórhallur Eyþórsson, Hans-Martin Gärtner &
Tonjes Veenstra
Programm
Raum: G 308
Mittwoch, 24. 02. 2016
14:00–15:00
Olaf Koeneman, Hedde Zeijlstra
e Rich Agreement Redux
15:00–16:00
Eric Fuß
Hand in hand or each on one’s own? On the connection
between morphological and syntactic change
16:00–16:30
Kaffeepause
16:30–17:30
Tonjes Veenstra
From rags to riches: the RAH from a creole perspective
17:30–18:30
Peter Slomanson
e contribution of contact linguistics to the Rich
Agreement debate
Donnerstag, 25. 02. 2016
9:00–10:00
Ásgrímur Angantýsson
V2 and verbal morphology in Övdalian
10:00–11:00
Hans-Martin Gärtner
On the Role of Verbal Mood in Licensing Dependent V2
Clauses
11:00–11:30
Kaffeepause
11:30–12:00
Heimir van der Feest Viðarsson
Re-challenging the RAH: Problematisation of structural
and social aspects in 19th-century Icelandic
55
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 56 — #68
AG Programme
12:00–13:00
Programm
John Sundquist & Caroline Heycock
Revisiting the RAH in Light of Diachronic Data from the
History of Danish
Freitag, 26. 02. 2016
11:30–12:00
Elyse Jamieson
Rich agreement in the Shetland dialect of Scots
12:00–13:00
Eric Haeberli, Tabea Ihsane
e Rich Agreement Hypothesis: Diachronic (lack o)
evidence from English
13:00–14:00
Thórhallur Eythórsson
’If It’s Tuesday, is Must Be Belgium’: Some alleged
syntax-morphology correlations re-examined
AG 11
Indefinites between theory and language
change
Chiara Gianollo, Klaus von Heusinger &
Svetlana Petrova
Raum: G 201
Mittwoch, 24. 02. 2016
14:00–15:00
Maria Aloni
Indefinites as fossils
15:00–16:00
Urtzi Etxeberria & Anastasia Giannakidou
Anti-specificity and the role of number: the case of
Spanish ’algún/algunos’
16:00–16:30
Kaffepause
56
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 57 — #69
AG Programme
16:30–17:30
Irene Franco, Olga Kellert, Guido Mensching & Cecilia
Poletto
On (negative) indefinites in Old Italian
17:30–18:30
Remus Gergel
Another route towards epistemic indefinites: A case for
VERUM?
Programm
Donnerstag, 25. 02. 2016
9:00–10:00
Ljudmila Geist
From indefinite NP to bare NP: why does the indefinite
article disappear?
10:00–11:00
Patrick G. Grosz
Scalar epistemic indefinites: a case study of ’weiß Go
w-’ in Present Day German
11:00–11:30
Kaffeepause
11:30–12:00
Ricardo Etxepare
From correlative protases to existential pronouns in
Basque
12:00–12:30
Amel Kallel & Pierre Larrivée
Strong polarity contexts and evolution of n-words
12:30–13:00
Moreno Mitrović
Indefinite polarisation and its scalar origin: evidence
from Japonic
Freitag, 26. 02. 2016
11:30–12:00
Rosemarie Lühr
Konstruktionen mit Indefinita in altindogermanischen
Sprachen
57
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 58 — #70
AG Programme
Programm
12:00–12:30
Andrei Sideltsev
Relative and indefinite pronouns: synchrony and
diachrony. e case of Hiite
12:30–13:00
Silvia Luraghi
Partitive case markers and indefiniteness: a diachronic
survey
13:00–14:00
Discussion
Presuppositions in language acquisition
AG 12
Anja Müller & Viola Schmitt
Raum: E 402
Mittwoch, 24. 02. 2016
14:00–15:00
Begrüßung
Kenneth Drozd (Invited speaker)
Cumulative universal quantification
15:00–15:30
Francesca Panzeri and Francesca Foppolo
e presuppositions of also and only: the view from
acquisition
15:30–16:00
Yi-ching Su
Only for children
16:00–16:30
Kaffeepause
16:30–17:00
Tom Roeper, Jennifer Rau
Children fail to repair presuppositions
17:00–17:30
Magda Oiry
How children deal with a contextually canceled
presupposition
58
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 59 — #71
AG Programme
17:30–18:00
Cory Bill, Jérémy Zehr, Lyn Tieu, Jacopo Romoli,
Stephen Crain & Florian Schwarz
On the acquisition of presupposition projection
18:00–18:30
Lilla Pintér
Exhaustivity of structural focus in Hungarian:
presupposition or implicature?
Programm
Adjective order: Theory and experiment
AG 13
Eva Wittenberg & Andreas Trotzke
Raum: E 402
Donnerstag, 25. 02. 2016
9:00 -9:30
Gregory Scontras, Judith Degen, Noah Goodman
Property subjectivity predicts adjective ordering
preferences
9:30–10:00
Sven Kotowski, Holden Härtl
Adjective order restrictions: e influence of
temporariness on prenominal word order
10:00–11:00
Guglielmo Cinque
Remarks on the order of adjectives cross-linguistically
11:00–11:30
Kaffeepause
11:30–12:00
Tine Breban, Kristin Davidse
A functional-cognitive analysis of the order of adjectival
modifiers in the English NP
12:00–12:30
Elnora ten Wolde
Linear vs hierarchical, two accounts ofpremodification in
the of-binominal nounphrase
59
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 60 — #72
AG Programme
12:30–13:00
Giuliana Giusti, Rossella Iovino
Free not-so-free adjectival word order in Latin
Programm
Freitag, 26. 02. 2016
11:30–12:00
Myrthe Wildeboer
An electro-encephalography study on Dutch-Papiamento
code-switching production
12:00–12:30
Claudia Turolla, Andrea Padovan & Ermenegildo
Bidese
Adjective orders in Cimbrian DP
12:30–13:00
Fryni Panayidou
Adjective ordering is not just semantics: A language
contact perspective
13:00–13:30
Melita Stavrou
Greek noun-adjective ordering revisited
13:30–14:00
Eva Wittenberg, AndreasTrotzke, Emily Morgan &
Roger Levy
Preferences in adjective order: Hierarchical and semantic
approaches reconciled
60
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 61 — #73
Teil III.
Plenarvorträge
61
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 62 — #74
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 63 — #75
Plenarvorträge
Causal wh & extra wh
Lisa Lai-Shen Cheng
Leiden University
[email protected]
Mittwoch, 24. 02. 2016, 9:30–11:30, Raum: A 600 (Audimax)
Cross-linguistically, it is common to see that the counterparts of how and
what can be used to express a causal reading. I take Tsai (2008) as a starting
point, and I argue, contra Tsai (2008) and Stepanov and Tsai (2008), that
the causal reading of how does not stem from its status as a sentential
operator, but instead from its dependence on modality. I examine data from
Mandarin and Cantonese and compare the causal reading of how with the
reading of how come, and argue that though the how-causal questions have
actuality entailment, they differ from how come questions in not having
factivity.
In addition to how, I also examine the causal reading of what. I argue that
the source of the causal reading differs from the source of the causal reading
of how. In relation to this, I discuss reason-applicatives and its implication
for causal what in languages like Dutch and German.
Semantic theory and computational experiments
Louise McNally
Universitat Pompeu Fabra
[email protected]
Mittwoch, 24. 02. 2016, 12:00–13:00, Raum: A 600 (Audimax)
In recent years there has been a very noticeable increase in the use of experiments with human subjects for the testing predictions made by linguistic
theories or analyses. In this talk, I focus on the insights that can be gained
through experiments involving computational modeling.
Though such experiments might be criticized for having even less ecological validity than controlled experiments with human subjects, I argue
63
PV
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 64 — #76
Plenarvorträge
that the very exercise of bringing together theory and concrete implementation can be extremely useful as a methodology for pushing research forward. I illustrate using examples from the recent, fruitful interaction of
theory and experiment in the area of distributional semantics.
Speech is special and language is structured
PV
David Poeppel
Max-Planck-Institute & New York University
[email protected]
Freitag, 26. 02. 2016, 9:00–10:00, Raum: A 600 (Audimax)
I discuss two new studies that focus on general questions about the cognitive science and neural implementation of speech and language. I come
to (currently) unpopular conclusions about both domains. Based on experiments using fMRI and exploiting the temporal statistics of speech, I argue
for the existence of a speech-specific processing stage that implicates a particular neuronal substrate that has the appropriate sensitivity and selectivity for speech. Based on a set of experiments using MEG, I discuss how temporal encoding can form the basis for more abstract, structural processing.
The results demonstrate that, during listening to connected speech, cortical activity of different time scales is entrained concurrently to track the
time course of linguistic structures at different hierarchical levels. Critically, entrainment to hierarchical linguistic structures is dissociated from
the neural encoding of acoustic cues and from processing statistical relations between words. These results demonstrate syntax-driven, internal
construction of hierarchical linguistic structure via entrainment of hierarchical cortical dynamics. The conclusions – that speech is special and language structure driven – provide new neurobiological provocations to the
prevailing view that speech perception is ‘mere’ hearing and that language
comprehension is ‘mere’ statistics.
64
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 65 — #77
Plenarvorträge
Incomplete acquisition and language attrition in different
settings
Höskuldur Þráinsson
University of Iceland
[email protected]
Freitag, 26. 02. 2016, 10:00–11:00, Raum: A 600 (Audimax)
PV
In recent years, there has been growing interest in research on so-called
heritage languages, i.e. languages that speakers typically acquire at home
as children, sometimes incompletely, but use only to a limited extent since
they are growing up and living in a society where another language is
dominant. This paper reports on ongoing research on North American Icelandic (NAmIce), a heritage language still spoken to some extent in certain
areas of Canada and the US but on the brink of extinction. The focus will be
on selected syntactic phenomena, including V2/V3, reflexives, case marking and processing of syntactically complex structures. Most of the data
on NAmIce were collected in recent field trips, using various elicitation
techniques.
Longitudinal data from extensive letter writing by speakers of NAmIce
will also be considered since they provide an interesting perspective: Some
of the writers started out as relatively perfect speakers (writers) of Icelandic but later show some signs of language attrition. The average age of
the NAmIce subjects interviewed on the field trips was about 77 years and
it turns out that in order to interpret their linguistic performance correctly
it is necessary to compare it to that of speakers of “Icelandic Icelandicˮ of
roughly the same age that were interviewed and tested using the same elicitation techniques. This way we have collected three different sets of data
which make it possible to distinguish, at least to some extent, incomplete
language acquisition, language attrition in a heritage language setting and
“natural” language attrition. This sheds an interesting light on “knowledge
of language: its nature, origin and use”.
65
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 66 — #78
New Text Books
How to do Linguistics with R
Data exploration and statistical analysis
Natalia Levshina
Université catholique de Louvain
This book provides a linguist with a statistical toolkit for exploration and
analysis of linguistic data. It employs R, a free software environment for
statistical computing, which is increasingly popular among linguists.
How to do Linguistics with R: Data exploration and statistical analysis is unique in
its scope, as it covers a wide range of classical and cutting-edge statistical
methods, including different flavours of regression analysis and ANOVA,
random forests and conditional inference trees, as well as specific linguistic
approaches, among which are Behavioural Profiles, Vector Space Models
and various measures of association between words and constructions.
The statistical topics are presented comprehensively, but without too much
technical detail, and illustrated with linguistic case studies that answer
non-trivial research questions. The book also demonstrates how to visualize
linguistic data with the help of attractive informative graphs, including the
popular ggplot2 system and Google visualization tools.
2015. xi, 432 pp. + index
Hb 978 90 272 1224 5 EUR 105.00 /e-inst 978 90 272 6845 7
Pb 978 90 272 1225 2 EUR 36.00 /e-priv 978 90 272 6845 7
/ 978 90 272 1224 5 USD 158.00
/ 978 90 272 1225 2 USD 54.00
ExpectedOctober2015
EUR 105.00
EUR 36.00
/ 978 90 272 6845 7 USD 158.00
/ 978 90 272 1225 2 USD 54.00
A Grammar of Mandarin
Jeroen Wiedenhof
Leiden University
A fascinating description of a global language, A Grammar of Mandarin combines broad perspectives with illuminating depth. Crammed with examples
from everyday conversations, it aims to let the language speak for itself.
The book opens with an overview of the language situation and a thorough
account of Mandarin speech sounds. Nine core chapters explore syntactic,
morphological and lexical dimensions. A final chapter traces the Chinese
character script from oracle-bone inscriptions to today’s digital pens.
This work will cater to language learners and linguistic specialists alike.
Easy reference is provided by more than eighty tables, figures, appendices,
and a glossary. The main text is enriched by sections in finer print, offering
further analysis and reflection. Example sentences are fully glossed, translated, and explained from diverse angles, with a keen eye for recent linguistic
change. This grammar, in short, reveals a Mandarin language in full swing.
2015. xxv, 477 pp.
Hb 978 90 272 1227 6
Pb 978 90 272 1228 3
ExpectedOctober2015
EUR 105.00 /e-inst 978 90 272 6775 7
EUR 36.00 /e-priv 978 90 272 6775 7
/ 978 90 272 1227 6 USD 158.00
/ 978 90 272 1228 3 USD 54.00
EUR 105.00
EUR 36.00
/ USD 158.00
/ USD 54.00
JOHN BENJAMINS PUBLISHING COMPANY
www.benjamins.com [email protected]
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 67 — #79
Teil IV.
Arbeitsgruppen und Abstracts
67
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 68 — #80
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 69 — #81
Arbeitsgruppe 1
Verb second in grammar and processing: its
causes and its consequences
Oliver Bott1 , Constantin Freitag2 & Fabian Schlotterbeck1
1
Universität Tübingen, 2 Universität Konstanz
AG1
oliver.bott@uni-tübingen, [email protected],
[email protected]
Raum: G 309
Workshop description
The verb second (V2) property seen in most Germanic but also in some
other Indoeuropean or even extra-Indoeuropean languages may be part of
a wider variational scenario in which par- ticular features (as encoded in
the finite verb) must be represented in the left clausal periphery (cf. Anderson, 1993). Although the V2 property has received much attention in the
syntactic literature, there is still dissent which functional projections/steps
of movement are involved in the derivation of V2 order or if it is even base
generated. Furthermore, it is still unclear if the V2 order is a purely structural linearization condition, or if it is tied to the semantic component in
narrow syntax, as proposed for German (Truckenbrodt, 2006). Every generalization must also consider the variation among V2 languages concerning
basic word order, clause types which exhibit V2 (main clause, embedded
clause, relative clause), and co-occurrence of V2 order and complementizers.
L1-acquisition research suggests that children acquire the V2-property
of German as a secondary step after having settled for head-final basic word
order (Clahsen and Muysken, 1986). For L2-acquisition it is reported that
69
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 70 — #82
AG 1: Verb second in grammar and processing
AG1
the acquisition process runs through a fixed order of structural hypotheses
which differ from L1-acquisition patterns. These findings suggest that the
V2 is a derived order.
Despite the consideration of V2 in theoretical work, the consequences of
V2 for sentence processing have not received much attention. For instance,
German has been mostly studied with respect to its underlying verb-final
order. But the early availability of the morphological and lexical verb information in V1/V2 may have important consequences for the parsing process.
For instance Knoeferle et al. (2005) showed that the verbal information is
immediately used in anticipating upcoming event participants. Other aspects of verb related interpretation processes, however, such as covert reconstruction of quantifiers (Bott and Schlotterbeck, 2015), thematic prominence effects (Scheepers et al., 2000), and NPI licensing (Freitag and Bayer,
2015) seem to be delayed to the right clause boundary, i. e. the supposed
base position of the finite verb.
The main questions we want to address are: What is the structural analysis of the V2 position in different clause types? Is V2 only a linearization
phe- nomenon, or is it tied to semantics/pragmatics? What is the role of
V2 verbal information in sentence processing? Which aspects of the interpretation are immediately triggered by the verb in V2 position and which
are assigned at its base position.
We invite submissions that present theoretical or empirical contributions based on language-specific, cross-linguistic, diachronic, or language
acquisition research on the empirical properties of phenomena that are
caused by, or correlate with the V2 property. Theoretical proposals should
make clear-cut predictions that allow for experimental falsification. Experimental approaches, on the other hand, should ad- dress the predictions of
theoretical implementations. By bringing together the above mentioned
lines of research, we hope to come one step closer to a deeper understanding of V2, and to find directions for future research.
70
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 71 — #83
AG 1, Raum G 309
Introduction
Oliver Bott1 , Constantin Freitag2 & Fabian Schlotterbeck1
1
Universität Tübingen, 2 Universität Konstanz
oliver.bott@uni-tübingen, [email protected],
[email protected]
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: G 309
Variations on V2: the information-structural dynamics of
the left periphery in German
AG1
Heike Wiese1 , Eva Wittenberg2 & Oliver Bunk1
1
Universität Potsdam, 2 University of California, San Diego
[email protected], [email protected], [email protected]
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: G 309
Modern German is usually regarded as a typical instance of a Germanic V2
language, with a strict V2 restriction for matrix declaratives that requires
exactly one position to be filled in the ‘forefield’ in front of the finite verb.
Deviations from V2 described in the literature are either separate constructions such as left dislocation or hanging topics (Frey 2005), specific exceptions such as sentences with irrelevance conditional and counterfactual adverbial clauses in the forefield (Axel 2004, d’Avis 2004), or patterns of only
putative multiple frontings that might be subsumed under V2 (Müller 2005
on multiple VP constituents in the forefield). Accordingly, examples such
as in (1) are usually constructed to illustrate ungrammatical linear orders,
and starred accordingly:
(1)
“*[Sobald/Wenn/Weil es aufheitert], wir können spazieren gehen.”
[Axel 2004:25]
However, findings from spoken language use outside formal standard German provide evidence for just such linearisations, cf. (2), including ones
with non-clausal adverbials (3):
(2)
WENN du mir nisch GLAUBST wir legen AUF [KiDKo, MuH3WT]
71
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 72 — #84
AG 1: Verb second in grammar and processing
(3)
HEUte ich werd meine zigaRETten mitbringen
MuH11MD]
[KiDKo,
This suggests that there might be systematic extensions of V2 in German, to
a more liberal forefield that can also accomodate V3. Evidence for this was
first reported from Kiezdeutsch, an urban dialect found in informal speech
among multilingual peer groups (Wiese 2009, 2013), and has subsequently
also been found in more monolingual contexts of German (Schalowski
2012, 2015), cf. (4) and (5) (utterances by speakers at a lecture series and
an annual DGfS conference, respectively):
AG1
(4)
wenn sie das GANze hören sie wissen genau …
[BSa-Sch 8]
(5)
und dann (-) die partikeln verÄNdern sich nicht
[BSa-Sch 24]
These findings point to a specific pattern which (a) is syntactically integrated into German as V3, rather than an allochthonous SVO construction (Wiese 2013, te Velde to appear; contra Auer 2013), indicated, e.g., by
the preservation of the verbal bracket, and (b) at the level of information
structure, allows both framesetters and topics to appear together in the left
periphery.
Our paper presents results from a cross-linguistic study that further explored such an information-structural motive for this pattern. We investigated whether speakers of German and English were more likely to place
verbs in a V3 position, after framesetter plus topic, if language-specific
grammatical restrictions were removed. In order to test this, we presented
speakers with a (non-verbal) comic sequence and asked them to describe
the final picture, which included a frame-setter (a time indicated on a clock)
and an animate or inanimate topic. Participants had to render the scene (a)
verbally and (b) in a semi-verbal set-up, using little plastic figures, wooden
clocks, and paper slips with written verbs.
Results indicate that verbal descriptions followed the typical standard
language patterns, German speakers displaying V2 with either the topic/
subject or the framesetter/adverbial in the forefield (i.e., Adv Vfin S or S Vfin
Adv), and English speakers displaying SVO with the topic/subject always
in front of the verb and sometimes preceded by the framesetter/adverbial
72
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 73 — #85
AG 1, Raum G 309
(i.e., SV Adv or Adv SV). In contrast to this, in the semi-verbal condition,
both German and English speakers also used an additional ordering option
pointing to V3, with both topic and framesetter presented before the verb,
and in both internal orders (ie., framesetter > topic, and topic > framesetter). We analyse these findings from the point of view of the interface
between syntax and information structure, and discuss their implications
for a syntactic account of German V3.
The derivation of V2 in Germanic main and embedded
clauses
AG1
Sten Vikner
Aarhus University
[email protected]
Mittwoch, 24. 02. 2016, 15:00-16:00, Raum: G 309
This talk will give an overview of the verb second (V2) phenomenon, as
found in both main and embedded clauses in the Germanic languages, and
it will also explore a particular derivation of (embedded) V2, in terms of a
cP/CP-distinction.
All the Germanic languages except modern English (but including e.g.
Old English) are V2, i.e. in all declarative main clauses and in all wh-questions, the finite verb is in the second position. regardless of whether the
first position is occupied by the subject or by some other constituent. This
can be extended to yes/no-questions, provided it is assumed that the first
position in such questions is empty (and such an assumption is supported
by the fact that it allows an account for Greenberg’s 1963:83 “Universal 11”,
cf. Vikner 2007). English only requires V2 in some main clauses: questions
and negative topicalisations.
As far as embedded clauses in the Germanic languages are concerned,
V2 is never obligatory, and although it is optionally possible in many embedded clauses, this is not the case for all types of embedded clauses, as
e.g. embedded questions never allow V2 (Julien 2007, Vikner 2001).
I will explore a particular derivation of (embedded) V2, in terms of a
cP/CP-distinction, which may be seen as a version of the CP-recursion
73
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 74 — #86
AG 1: Verb second in grammar and processing
AG1
analysis (de Haan & Weerman 1986, Vikner 1995 and many others). This
analysis will be compared to a fine-grained left periphery approach (Rizzi
1997 and many others).
The idea is that because embedded V2 clauses do not allow extraction,
whereas other types of CP-recursion clauses do (Christensen et al. 2013a,b),
CP-recursion in embedded V2 is assumed to be fundamentally different
from other kinds of CP-recursion, in that main clause V2 and embedded
V2 involve a CP (”big CP”), whereas other clausal projections above IP are
instances of cP (“little cP”).
Part of the talk builds on joint work with Ken Ramshøj Christensen and
Anne Mette Nyvad.
References: • Christensen, Ken Ramshøj, Johannes Kizach & Anne Mette Nyvad. 2013a. Escape from the island: Grammaticality and (reduced) acceptability of wh-island violations in
Danish, Journal of Psycholinguistic Research 42, 51-70. • Christensen, Ken Ramshøj, Johannes
Kizach & Anne Mette Nyvad. 2013b. The processing of syntactic islands – an fMRI study,
Journal of Neurolinguistics 26.2, 239-251. • Greenberg, Joseph. 1963. Some universals of grammar with particular reference to the order of meaningful elements in Joseph Greenberg (ed.):
Universals of Language, MIT Press, Cambridge MA. • deHaan, Germen & Fred Weerman.
1986. Finiteness and Verb Fronting in Frisian. In: Hubert Haider & Martin Prinzhom (eds.),
Verb Second Phenomena in Germanic Languages. Foris, Dordrecht, pp. 77-110. • Julien, Marit.
2007. Embedded V2 in Norwegian and Swedish. In: Working Papers in Scandinavian Syntax 80.
Lund: Lund University, 103-161. • Rizzi, Luigi. 1997. The fine structure of the left periphery.
In: Liliane Haegeman (ed.), Elements of Grammar. Dordrecht: Kluwer, 281-337. • Vikner, Sten:
1995, Verb Movement and Expletive Subjects in the Germanic Languages, New York: Oxford
University Press. • Vikner, Sten: 2001, Verb Movement Variation in Germanic and Optimality
eory, Habilitationsschrift, University of Tübingen. www.hum.au.dk/engelsk/engsv/papers/
viknhabi.pdf. • Vikner, Sten. 2007. .Teoretisk og komparativ syntaks. In Henrik Jørgensen &
Peter Widell (eds.), Det bedre argument – Festskri til Ole Togeby, 7. marts 2007, Wessel &
Huitfeld, Århus, pp. 469-480. www.hum.au.dk/engelsk/engsv/papers/vikn07a.pdf
74
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 75 — #87
AG 1, Raum G 309
Relaxed V2 languages and their Left Periphery. Two cases
from Northern Italy
Jan Casalicchio1 & Federica Cognola2
Università di Trento, 2 Università Ca’ Foscari di Venezia
1
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 16:30-17:30, Raum: G 309
Introduction: The aim of this talk is to contribute to our understanding
of the V2 phenomenon by comparing two varieties spoken in Northern
Italy (Trentino-South Tyrol): the Rhaeto-Romance variety of Badia (henceforth: R) and Mòcheno (M), a German dialect. Both have been independently claimed to be relaxed V2 languages, in which i) the finite verb and
one XP have to move to CP for EPP reasons (Roberts 2004, Holmberg 2015),
and ii) this movement coexists with V3/V4 word orders (cf. Rowley 2003
for M; Poletto 2002 for R, a.o.). By discussing a series of novel data on the
structure of the left periphery in R&M, we show that they pattern alike
in most contexts, and exhibit two properties which are incompatible with
previous accounts of R (cf. Poletto 2002) and of relaxed V2 languages in
general (Benincà 2006).
Properties of R&M: In previous accounts, the possibility of V3/V4 word
orders in relaxed V2 languages was ascribed to the presence of an articulated left periphery in these varieties, like in modern Italian (cf. Rizzi 1997,
Benincà 2001). However, R&M exhibit properties which are absent from
modern Romance and have mostly remained unnoticed so far. These are:
1. Relativized Minimality effects (RM, cf. Rizzi 2004) in fronted topics:
Two fronted topics can precede the verb; however, a given subject and a
given object cannot contemporary appear in the left periphery:
(1)
a.
a’.
(2)
Lucaj ala
mamak tik à=lj
cumprè n liber (R)
Luca to-the mum her has=he.cl bought a book
Der Luca en de mama hòt a puach kaft
(M)
the Luca to the mum has a book bought
a. *La mamaj l liberk, (lk) a(=laj)
cumprà inier
(R)
the mum the book (it) has(=she.cl) bought yesterday
75
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 76 — #88
AG 1: Verb second in grammar and processing
kaft
(M)
a’. *De mama s puach hòt gester
the mum the book has yesterday bought
AG1
We propose that the V3 order in (1) is due to the fact that the EPP feature is
not satisfied by one of the fronted arguments, but by a pro in Fin0 . Topics
move directly to TopicP and are not base-generated there. Subject and object cannot co-occur because they share the same featural marking (+topic,
-focus, -case, where case is to be understood as morphological case. This
means that the presence/absence of a case assigning P is distinctive for +/case in both R&M).
2. Restrictions on V3 with a fronted focalised argument: When a focalised element is fronted, strict V2 order is mandatory in R. Following
Poletto (2002), we propose that the focalised element moves to ForceP, the
highest projection of the clause in R. The EPP feature is on Force0 , and the
verb also moves to Force0 to satisfy it (see 3).
In M, just one topic can precede a fronted focus. In our proposal, this is
due to the fact that in M focalisation is an instance of (focalised) topicalisation: since M has only two TopicPs (3), the lower one is occupied by the
focus and one TopicP is available for topics. Moreover, in these sentences
there are superiority effects: the [+contrastive] topic moves first to FinP to
satisfy the EPP feature, creating the typical “bottle-neck” effect, which can
be circumvented only by a topic whose base-position is higher than the
base-position of the [+contrastive] XP.
3. Asymmetries between main declarative and interrogative clauses: In
wh-interrogatives, V3/V4 word orders are possible in both R&M, but with
one difference: in M we observe the same RM-effects observed in declarative clauses (2), while in R there are no restrictions on the simultaneous
fronting of subject and object. We claim that in this case the EPP-feature
in Fin0 is satisfied by the wh-element, and that in M the fronted topics are
moved from a lower position; in R, instead, topics are base-generated when
there is a wh-element.
To conclude: The base structure of the left periphery in both R&M is the
following:
(3)
76
[CP [Force[+EPP] [Topic [Topic [wh [Fin[+EPP] [ V ]]]]]]]]
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 77 — #89
AG 1, Raum G 309
NB: recall that the EPP-feature is usually in Fin0 , but in R it is in Force0
when a focus is fronted.
References: • Benincà (2006). A detailed map of the left periphery of Medieval Romance. In:
Zanuttini et al. (eds.), Crosslinguistic research in Syntax and Semantics. Washington, 53-86.
• Poletto (2002). The left periphery of a V2 Rhaetoromance dialect: a new perspective on V2
and V3. In: Barbiers et al. (eds.), Syntactic Microvariation. Amsterdam: Meertens, 214-242.
• Rizzi (2004). Locality and left periphery. In: Belletti (ed.), Structures and beyond. Oxford:
OUP, 223-251. • Rowley (2003). Liacht as de Sproch. Grammatica della lingua mòchena. Palù
del Fersina.
AG1
V2 in the history of English: why did it arise, why was it
lost, and what difference did it make?
Bettelou Los1 & Ans van Kemenade2
1
Edinburgh University, 2 Radboud University Nijmegen/CLS
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 17:30-18:30, Raum: G 309
Old English has been shown to have a version of V2 since van Kemenade
(1987). Subsequent work (Haeberli 2002, van Kemenade 2012) has arrived at
a consensus that Old English may move the finite verb to C, as do PresentDay Dutch and German, but also to an additional lower position F. The
Old English verb moves to C if the first constituent is a negative or whconstituent (or a temporal adverb þa/þonne, which is a separate development), while it is in F when the first constituent is an object or an adverbial.
Spec,FP is a position for pronouns and certain types of nominal subjects:
The examples in (1) demonstrates that movement to C may have been motivated originally to mark off focused constituents, and movement to F to
demarcate discourse links (ðuruh þæt gescead ana, Mid þam) and established referents (we) from new information (sælran þonne þa ungesceadwysan nytenu, an mæden).
77
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 78 — #90
AG 1: Verb second in grammar and processing
(1)
AG1
SpecCP
C
SpecFP
hwæt
selþ
he
what
sells
he
F
TP – VP - …
…
ðuruh þæt gescead
ana
we
synd
sælran
þonne
þa
ungesceadwysan nytenu
through that understanding alone
we
are
better than the unreasoning animals
Mid þam
wunode
an mæden
with those (people)
lived
a maiden
Movement to C survives to Present-Day English as syntactically motivated T-to-C movement, although structures as in (2a) have suffered some
competition from the stressed-focus it-cleft (cf. (2c), which emerged in the
15th Century (Ball 1991; Komen 2013).
(2)
a.
Only after a few minutes did I realize that everyone was staring
at me.
b. *Only after a few minutes, I realized that everyone was staring
at me.
c. It was only after a few minutes that I realized that everyone
was staring at me.
Movement to F, however, declined in the 15th Century (van Kemenade and
Westergaard 2012), for reasons that are not fully understood (e.g. Fischer
et al. 2000: 131-4). One clue to its demise is the decline of adverbial discourse links in first position as measured by the ratio of clause-initial PPs
containing demonstratives (Los & Dreschler 2012; Dreschler 2015; PérezGuerra 2005: 357). There is a loss in the ability to refer to the immediately
preceding discourse in elements of the D-system other than demonstratives, too, like then and there (Los & van Kemenade forthcoming). Firstposition demonstratives used independently, whether subjects, objects or
as complement of prepositions, function as topic shifters in Present-Day
78
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 79 — #91
AG 1, Raum G 309
Dutch and German (Bosch et al. 2003), as they did in Old English, but this
ability is lost, too, when demonstratives can no longer refer to singular
animate referents, the result of the loss of gender. These various losses in
referentiality affected the function of V to F as a demarcator of information
structure.
The role of V2 in sentence comprehension and sentence
production
Markus Bader
Institute of Linguistics, Goethe University Frankfurt
AG1
[email protected]
Donnerstag, 25. 02. 2016, 09:00-10:00, Raum: G 309
This talk will review various strands of evidence from acceptability and
production experiments and from corpus studies concerning the linearization of verb arguments in German. The order of arguments is relatively
free in German in the sense that besides the most common SO pattern sentences with OS order also occur with some regularity. This variability has
three distinct sources. First, an interplay of lexical-conceptual and grammatical factors determine a base order that constitutes the unmarked order
in the middlefield. Second, phrases can change position within the middlefield (so-called scrambling). Third, a single phrase has to be put into the
prefield. If the middlefield initial phrase is put into the prefield, this does
not cause a change in word order, but if any other phrase is put into the
prefield, word order deviates from the middlefield internal word order.
The main focus of my talk will lie on differences between word order
in the middlefield and word order when the prefield is involved. One part
of the evidence will concern effects of lexical accessibility and verb semantics. While word order in the middlefield is strongly affected by these
factors, word order involving the prefield is much less so. Several production experiments show that participants are very reluctant in producing
OS sentences with the object in the prefield. Instead, passivization is used
to bring an argument into the prefield when this is favored by lexicalconceptual considerations (animacy, verb semantics). A further strand of
79
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 80 — #92
AG 1: Verb second in grammar and processing
AG1
evidence pertains to the interaction between different referential means
(pronouns, d-pronouns, full NPs), word order, and syntactic position (middelfield vs. prefield). The evidence that will be discussed – mainly corpus
evidence and results from acceptability judgments – indicates both parallels and differences between putting an object pronoun into the first position within the middlefield and into the prefield. On the one hand, those
factors (animacy, verb semantics) that strongly favor the placement of an
object pronoun before the subject in the middlefield also allow the placement of the object pronoun in the prefield. On the other hand, OS order
within the middlefield always seems to be possible with an object pronoun
as long as the subject is not a pronoun itself, even if this is not particularly
favored. Putting the object pronoun into the prefield under the same circumstances leads to degraded acceptability, however. In this case, either
SO order is preferred or the replacement of the pronoun by a d-pronoun.
Effects of verbal information in the V2-position during
parsing: What eye movements reveal about prediction
(and integration)
Peter de Swart1 & Geertje van Bergen2
1
Centre for Language Studies, Radboud University Nijmegen, 2 Max
Planck Institute for Psycholinguistics, Nijmegen
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 10:00-10:30, Raum: G 309
Goal: We investigate how V2-verbal information affects the parsing of
transitive sentences. More specifically, we contrast sentences with a neutral auxiliary with sentences with a lexical verb in second position and
determine the effect on predictive processing of arguments, as witnessed
by eye movements.
Baground: In the past decade or so, evidence has accumulated that language users make use of verbal information to predict upcoming referents
(e.g. Altmann & Kamide, 1999; Knoeferle et al., 2005; a.m.o.). Previous research has mainly focused on the prediction of a single argument based
80
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 81 — #93
AG 1, Raum G 309
on the verb and one additional argument. We take this research one step
further by investigating the anticipation of both arguments in a transitive
sentence and how this is mediated by verbal information and animacy. To
this end we crucially exploit the V2-property, which makes it possible to
provide the verb before its arguments when an adverbial expression occupies the sentence-initial position (the net result being a V-initial parsing
situation).
Experiment: We recorded eye-gaze patterns of 87 native speakers of Dutch,
while they looked at a two-picture display containing an animate and inanimate character and listened to sentences with an XP-V-NP1-NP2-(Participle)PP structure, in which NP1 can only be interpreted as the subject. We manipulated two factors: (i) eb pe in V2-position: auxiliary (hee ‘has’)
or lexical; (ii) animac configaion of the NPs. This resulted in four
conditions:
1. XP-VAUX:hee -NP1SU:anim -NP2OBJ:inan -ParticipleVlex -PP
2. XP-VAUX:hee -NP1SU:inan -NP2OBJ:anim -ParticipleVlex -PP
3. XP-VLEX -NP1SU:anim -NP2OBJ:inan -PP
4. XP-VLEX -NP1SU:inan -NP2OBJ:anim -PP
Results: We found a significant effect of eb pe
in the window from verb
offset until onset of NP1+100ms,
in which information about
NP1 is critically not yet accessible. A lexical verb in
second position elicited a
larger proportion of fixations to the inanimate character in comparison to an
auxiliary verb (see Figure).
Also, the data suggests that
the later integration of arguments is delayed in sentences with an auxiliary verb in V2-position.
81
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 82 — #94
AG 1: Verb second in grammar and processing
AG1
Interpretation: Our data suggest a clear effect of V2-information on both
predictive processing and integration. The auxiliary verb hebben ‘have’
evokes an expectation for an animate character with a long lasting effect
on parsing (not shown in the Figure). Lexical verbs, on the other hand, anticipate an inanimate argument. We suggest that this is an effect of ‘VP’frequency. Even though NP1 was always the subject in our sentences, in
everyday speech a lexical verb in V2-position is most frequently followed
by an inanimate object (the subject being in sentence-initial position).
Implications: This study suggests a more important role for verbal information occurring in second position than hitherto assumed based on findings from earlier experiments on German (Scheepers et al., 2000; Bayer &
Bader, 2006). It also contrasts with the theoretical proposal that “the lexical
part [of the verb] is evaluated in its base position” (Bayer, 2008, p.3).
Hör endlich auf/zu (‘Now stop/listen’)! – The Lexical
Representation and Semantic Activation of German
Particle Verbs
Eva Smolka1 & Bettina Braun1
1
Department of Linguistics, University of Konstanz
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 10:30-11:00, Raum: G 309
Because German is a verb-second language with an SOV word order (e.g.,
Haider, 1985), particle verbs in German are decomposed whenever they
occur in finite forms. Since the particle, which complements the meaning
of the whole particle verb, must appear sentence final, it can be presented
many words after the stem, as the following example of the base hören
(‘hear’) demonstrates:
(1)
82
Der Junge hörte, nachdem er vergeblich um Eis gebettelt hatte, schließlich wieder auf /zu.
(L: ‘Finally, the boy stopped/listened after having unsuccessfully
cried for ice cream’).
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 83 — #95
AG 1, Raum G 309
Consequently, the meaning of the whole verb is fully understood only by
the end of the sentence, when the particle is encountered. It is possible
that German readers/listeners are therefore used to keeping more than
one possible meaning of the verb active upon encountering a verb stem.
The present study investigated (1) how particle verbs are stored and represented in lexical memory, (2) and whether the morphological activation of
the stem cascades through to the semantic level.
To this end, we conducted two experiments with cross-modal priming, a
paradigm that is typically used to tap into lexical processing (for a review
see Smolka et al., 2014). In Experiment 1, 21 participants heard isolated
complex verbs in isolation and made lexical decisions to visually presented
base verbs (e.g., fallen, ‘fall’), see (2). Auditory primes were (a) semantically transparent particle verbs (whose meaning can be constructed from
the meaning of particle and stem, e.g., hinfallen, ‘fall down’), (b) semantically opaque particle verbs (whose meaning cannot be constructed from
the meaning of the particle and the stem, e.g., auffallen, ‘attract attention’),
and (c) form-related controls (whose stem is phonologically related to the
target, e.g. ausfalten, ‘fold out’).
auditory prime: (a) hinfallen/(b) auffallen/(c) ausfalten – visual target: (2)
fallen/(3) Sturz
Relative to the form controls, semantically transparent and opaque particle verbs primed their base to the same extent. Equivalent morphological
priming effects replicate previous findings (Smolka et al., 2009, 2014) that
particle verbs undergo morphological decomposition regardless of their
meaning composition. We conclude that particle verbs are lexically represented via their stem.
Experiment 2 tested whether the morphological access to the stem influences also the semantic level. In Experiment 2, 47 participants heard the
same primes as in Experiment 1 (i.e., semantically transparent and opaque
particle verbs and form controls) and made lexical decisions to visually presented associations to the stem, see (3) – these associations were previously
collected in a web experiment, with 105 participants. Results showed that
neither semantically transparent nor opaque verbs induced priming to the
stem associations (relative to form controls). These findings indicate that
the previously observed morphological effects do not extend to a semantic
83
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 84 — #96
AG 1: Verb second in grammar and processing
AG1
level. That is, the previous decomposition of the stem does not activate its
meaning.
In sum, the present findings indicate that German particle verbs are lexically represented via their stem. This may be a characteristic of particle
verbs in a verb-second language, where particles are separated from the
stem in finite forms. Nevertheless, even though the stem represents the
lexical level, its meaning is not automatically cascaded to the semantic
level (or is inhibited by the meaning of the particle verb). This latter finding corresponds to previous findings (with prefix verbs) in languages other
than verb-second with SOV word order. We will discuss present models of
lexical representation and meaning representation with respect to particle
verbs in German.
References: • Haider, Hubert (1985). V-Second in German. In H. Haider & M. Prinzhorn
(Eds.), Verb second phenomena in Germanic languages (pp. 49-75). Dordrecht: Foris Publications. • Smolka, E., Komlósi, S., & Rösler, F. (2009). When semantics means less than morphology: e processing of German prefixed verbs. Language and Cognitive Processes, 24(3), 337-375.
• Smolka, E., Preller, K., & Eulitz, C. (2014). ‘Verstehen’ (‘understand’) primes ‘stehen’ (‘stand’):
Morphological structure overrides semantic compositionality in the lexical representation of German complex verbs. Journal of Memory and Language, 72, 16-36.
Modeling Verb Placement Errors in Swiss German
Children’s L1 Acquisition
Isaac Gould
The University of Kansas
[email protected]
Donnerstag, 25. 02. 2016, 11:30-12:00, Raum: G 309
e acquisition puzzle: Schönenburger (2001, 2008) describe a detailed
longitudinal study of the spontaneous productions of 2 Swiss German (SG)
children between ages 3;10-8;01. Early in the study, the children make systematic errors in the placement of the finite verb in embedded clauses.
In contrast to the adult grammar, which places the finite embedded verb
clause-finally in (1a), in the children’s productions of such an embedded
clause, the finite verb appears in a non-final position (1b).
84
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 85 — #97
AG 1, Raum G 309
(1)
a. Adult: [Complementizer S O V]…
b. #Wenn du trinksch öppis…(M: 4;01)
‘When you drink something…’
After a period of near ceiling error rates, the error rates gradually decline as
the children approach adult-like performance. The puzzle is: why do the SG
children make such errors, which differ from the input they hear? I present
a probabilistic learning model that arrives at the errors in a principled way:
the errors result from a grammar that is compatible with a majority of
the input, which is ambiguous and underdetermines the correct structural
analysis. The model also has the potential to capture variability in error
rates across German-learning children.
‘Outline of proposal’ Schönenburger proposes that the SG children raise
the verb to some head-initial phrase, whereas adults do not. Such verb raising would account for the errors, but it remains unclear what would cause
the errors in the first place. I build on this analysis in proposing that the
SG children initially misanalyse T as head-initial and raise the verb to this
position (2a). They then gradually reset the parameter to T-final, as in the
adult grammar (2b).
(2)
a. #T-initial: [CP dass [TP V+T [VP Subj [VP Obj V ] ] ] ]
b. ✓T-final: [CP dass [TP [VP Subj [VP Obj V ] ] V+T ] ]
The misanalysis crucially relies on the ambiguity of matrix clauses for Tinit/final. Matrix clauses are the primary source of German input (85%; cf.
Sakas 2003). Moreover, a majority of the grammars compatible with matrix
clauses are in fact T-initial, thus favoring T-initial. These grammars are
constrained in the model by the 5 basic parameters in (3), which determine
verb placement.
(3)
Parameters:
a. [±V-to-T]
b. [±T-to-C]
c. [V-init/fin]
d. [T-init/fin]
e. [C-init/fin]
85
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 86 — #98
AG 1: Verb second in grammar and processing
AG1
The majority of the schematic corpus used for the learning model strongly
favors T-initial. (The input types represent a diverse range of declarative
matrix and embedded clauses.) For example, given SVO input and a grammar that raises the verb to T but not to C, T must be head-initial in order
for the verb to precede the object. Further, although embedded clauses favor T-final, they are still compatible with T-initial given a non-verb raising
grammar.
I propose a probabilistic learning model implemented in Church (Goodman et al. 2008) that adopts a grammar of best fit, which at first is T-initial.
In such a learning model, each parameter value is associated with some
probability. When presented with some input, the model will sample a
grammar from these probabilities and reinforce them if the sampled grammar is compatible with the input (cf. Yang 2002). Early in the learning process, on average we expect the model to sample and reinforce T-initial more
than T-final. Further, there is robust unambiguous evidence (35.46% of the
input) for verb raising. Thus the model is pushed toward a non-target [+Vto-T, T-init] grammar, which results in the embedded clause production errors. The effect of unambiguous evidence is to push the learner even more
strongly toward [+V-to-T] than the ambiguous evidence pushes toward Tinitial. Once the model has learned [+V-to-T], then, a new grammar of best
fit emerges: embedded clauses can be taken as evidence for T-final. As embedded clauses are a relatively small proportion of the input, the switch to
T-final (and thus the adult grammar) is gradual.
Results and Discussion: 50 simulations were run, and in 13 (26%) we see
development like the SG children. As expected, we see nearly ceiling error
rates for [+V-to-T, T-init] grammars (striped segments), which are gradually displaced by the adult grammar. These non-target parameter settings
successfully account for verb placement errors in embedded clauses. Unexpectedly, the errors are partially due to a [–T-to-C] grammar. We thus
expect to see errors in matrix clauses with a verb third position verb. Indeed, such errors are attested in the SG corpus, though rare (4).
(4)
86
#[Nämlich] [ned alli Lüt] händ di gliiche Schtimm. (M: 4;11)
‘Not everybody has the same voice.’
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 87 — #99
AG 1, Raum G 309
Support for these results comes from Waldmann (2011) who reports such
errors are relatively frequent in Swedish before 3;6, but then become rare.
Gawlitzek-Maiwald et al. (1992) report similar errors in Standard German
around age 3;0. This is consistent with the SG corpus, which begins later
at 3;10. A further strength of the model is the possibility to capture variability in error rates across children learning a variety of German dialects.
Note that the simulations do not always have high error rates, and that
the input types in Figure 1 are found across German varieties. The results are also consistent, then, with reports of low (Clahsen 1982; Standard German), intermediate (Penner 1996; Bernese), and high error rates
(Gawlitzek-Maiwald et al. 1992; Standard German) of finite embedded verb
placement reported in the literature. The model thus mis-sets a parameter,
recovers, and sheds further light on SG acquisition.
On the nature of integrated V2 relative clauses
Emanuela Sanfelici1 , Corinna Trabandt2 & Petra Schulz3
Goethe Universität, Frankfurt am Main
1,2,3
Donnerstag, 25. 02. 2016, 12:00-12:30, Raum: G 309
This study investigates the structure of the so-called V2 relative clauses,
labeled iV2 structures (iV2) following Gärtner (2001a/b). In German relative
clauses (RCs), the verb usually occupies the final position (1a). However,
under specific conditions iV2 structures as in (1b) are licensed (cf. Brandt
1990, Gärtner 2001a/b, Zwart 2005).
(1)
a.
b.
Da sind zwei Frauen, die den Präsident getroffen haben
final
Da sind zwei Frauen, die haben den Präsident getroffen
‘Here there are two women that met the President.’
ViV2
iV2 structures are licensed under the following conditions: (a) the predicate in the main clause optimally is presentational/existential; (b) the antecedent must be indefinite and have wide scope; (c) the relative pronoun
in the iV2 has to be a d-pronoun; (d) the iV2 clause must be in sentence final
87
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 88 — #100
AG 1: Verb second in grammar and processing
AG1
position; (e) the iV2 must be prosodically integrated in the main clause. In
the literature iV2 structures are usually treated as main clauses linked by a
discourse paratactic head to CP1 , da sind zwei Frauen in (1b) (Gartner 2001
a/b, Endriss & Gärtner 2005). Here, we will rethink this conclusion and
argue that iV2 structures are an instance of embedded root phenomena,
which rephrases Gärtner’s (2001: 107) observation. Our claim in a nutshell
is that iV2 structures are a subtype of subordinate clauses, i.e. TopicP in
which the verb has moved to Fin0 and the demonstrative is a resumptive
topic pronoun (2). Building on Chung & Ladusaw’s (2004) proposal, we argue that IV2 are merged as adjunct in the specifier at the topmost v/VP
level, where they saturate the weak indefinite NP as in (2).
(2)
VP
VP
TopP
V’
NP
V
Under this analysis, a) the restrictive nature of iV2, b) their information
unity with the host sentence, c) the behavior of focus sensitive particles,
and d) the nature of the NP antecedent are derived as the result of the position where the iV2 originates. At the same time, this analysis accounts
for several problems of existing iV2 analyses (e.g., licensing nominal predicates, the [+REL] feature on the discourse head), by maintaining the advantage of Gärtner’s (2001a/b) proposal, which treats iV2 as main clauses.
The last piece of evidence in support of (2) comes from acquisition. In
our picture-supported delayed-imitation task, 3 to 5-year-old monolingual
German-speaking children repeated V-final RCs more often correctly than
iV2 structures at all ages, and changed iV2 structures into V-final relatives
significantly more often than the other way around. Whereas under a main
clause analysis (Gärtner 2001a/b) these findings are unexpected, our proposal predicts this pattern: children are expected to acquire iV2 structures,
a type of subordinate clause, later than the canonical subordinate configuration.
88
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 89 — #101
AG 1, Raum G 309
In sum, besides accounting for the core syntactic properties of iV2 structures, the analysis in (2) also offers a plausible explanation for our acquisition results.
Semantic restrictions in verb-second vs. non-verb-second
wh-exclamatives
Sohpie Repp
Humboldt-Universität zu Berlin
[email protected]
Donnerstag, 25. 02. 2016, 12:30-13:00, Raum: G 309
AG1
In some Germanic languages wh-exclamatives occur as verb-second (V2)
or as non-verb-second structures (non-V2). Recently it was shown that
V2/nonV2 in German and Dutch are not interchangeable. In Dutch, V2
underlies stricter semantic restrictions than non-V2, i.e. verb-final (V):
the noteworthiness evaluation exclamatives are often thought to express
can concern individuals and propositions in Vf but only individuals in V2
(Nouwen & Chernilovskaya 2013). For German, Vf has been shown to be
more restricted than V2 w.r.t. the realization of the exclamative accent on
the finite verb, which has been argued to be due to the availability of
(verum) focus alternatives (Driemel 2015). The present paper presents evidence that there are different semantic restrictions on German V2 vs. Vf
wh-exclamatives and argues that the two structures come with different
force operators. Like Dutch, German allows the full range of wh-words that
occur in questions also in exclamatives (Repp 2013). The evidence for different restrictions on V2/Vf comes from exclamatives with the wh-words wer/
wen (‘who/m’) and was (‘what’). Both wh-words may occur with the quantifier alles (‘all’), which in questions indicates exhaustiveness (Zimmermann
2007), and in exclamatives a large amount, i.e. a high degree: (1) illustrates
the semantic contribution of alles in Vf exclamatives with wen (‘whom’).
(2) shows that the corresponding V2 exclamative is only felicitous with
alles, i.e. is restricted to a high degree reading. Was is a multipurpose question word that can ask about entities, propositions, reasons (see 3a) and
89
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 90 — #102
AG 1: Verb second in grammar and processing
AG1
degrees. Was-exclamatives can express surprise at all these semantic objects, apart from reasons: (3b), which is string-identical to (3a), receives a
high degree reading (in V2 and V). Note that the verb in (3) is intransitive
so was cannot be the object of the verb. To explore the readings of was
and the co-occurrence of alles with was in V2 vs. Vf exclamatives, a corpus study was conducted (268 mil token sub-corpus of deWaC (Baroni et
al. 2009); was ür (‘what a’) was excluded; who(m) exclamatives did not occur sufficiently often). The corpus analysis revealed that was in any of the
above readings was combined with alles equally often in V2 and Vf (total
n=105; χ(1) = 2.143, p>.05). For was-exclamatives without alles, two types of
readings were explored: was as a direct object of a transitive verb, and was
combined with an intransitive verb (= degree was). The former readings
only occurred in Vf exclamatives. The latter occurred in V2 and in Vf. So,
again V2 but not Vf exclamatives can be shown to be subject to a degree
restriction. To account for the difference I propose that V2 and Vf exclamatives host different exclamative force operators (rather than e.g. question operators as in D’Avis 2001, Abels 2005), see (4)/(5). V2 only allows
degree readings and Vf allows degree, individual (and manner) readings.
Wh-phrases are set restrictors requiring their complement either to be a
set of entities, degrees or manners so that wh-structures either are individual, degree or manner properties yielding the appropriate semantic object
for the force operator in V2 vs. Vf.
(1)
a.
b.
(2)
a.
b.
(3)
90
a.
Wen der alles eingeladen hat!
whom he all invited
has
(surprise at high nb of (noteworthy) people)
Wen der eingeladen hat!
(surprise at noteworthy person/people)
Wen hat der alles eingeladen!
(surprise at high nb of (noteworthy) people)
⁇
Wen hat der eingeladen!
Was hast du (so) geweint?
what have you so wept
(question about reason of weeping)
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 91 — #103
AG 1, Raum G 309
b.
Was hast du geweint!
(surprise at degree of severity of weeping)
(4)
〚Excl-Deg〛= λD⟨d,t⟩ ∃d [the speaker finds λw.D(d)(w) surprising] =
V2
(5)
〚Excl〛= λP⟨d,t⟩ ∃x [the speaker finds λw.P(x)(w) surprising], for ⟨τ ⟩
= ⟨e⟩ and ⟨τ ⟩ = ⟨d⟩ and ⟨τ ⟩ = ⟨m⟩ (manner) = Vf
Embedded V2, Factivity and Main Point of Utterance
Kajsa Djärv1 , Caroline Heycock2 & Hannah Rohde3
1
The University of Pennsylvania, 2,3 The University of Edinburgh
AG1
[email protected], [email protected], [email protected]
Freitag, 26. 02. 2016, 11:30-12:00, Raum: G 309
Since the seminal work of Hooper & Thompson 1973, many researchers
have pursued the insight that V2, as a classic Main Clause Phenomenon
[MCP] is licensed in formally subordinate clauses to the extent that such
clauses are asserted. H&T categorised embedding predicates into 5 classes
largely according to whether their complement clauses could be interpreted
as asserted, a status which they took to be the converse of presupposed.
The class of verbs of communication such as say occupied one pole – allowing MCP freely in their complements – while factives such as be happy
that occupied the other. In an important update of this tradition, Simons
2007 has considerably sharpened H&T’s concept of assertion, proposing
that the crucial distinction is whether the subordinate clause contributes a
proposition that makes the utterance relevant; as a diagnostic, in a question/response sequence, “whatever proposition communicated by the response constitutes an answer (complete or partial) to the question is the
main point of the response.” Simons demonstrates that given this definition/diagnostic, even factive clauses may constitute the Main Point of Utterance MPU; hence, in such contexts, they should also allow V2.
In this talk we present the results of three experiments (one on Swedish
and two on English) that aimed to test empirically the claim that the possibility of V2 in an embedded clause (EV2) follows from whether or not
91
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 92 — #104
AG 1: Verb second in grammar and processing
AG1
the embedded clause constitutes the MPU (cf. Julien 2007, Jensen & Christensen 2013). In the first experiment, 104 L1 speakers of Swedish were
asked to judge the acceptability of question-response pairs where, following Simons 2007, the question was manipulated to vary the location of the
MPU in the response: in the main or the embedded clause. There were two
other independent variables: the classification of the embedding verb in
the response, and whether or not the embedded clause in the response exhibited V2.
We show that, on the one hand, the results support the claim that Swedish
EV2 is possible under semi-factive (discover/realize) and non-factive (think/
claim) clause-embedding predicates, but not under purely factive ones (be
happy/be surprised) (Wiklund et al. 2007). Strikingly, the judgments also
mirror the frequency difference between EV2 in the complements to epistemic vs. communicative non-factives (e.g. suppose vs. say) reported for
Danish corpus data in Jensen & Christensen 2013. However, the results
show no interaction between the effect of embedded V2 and embedded
MPU: that is, our data suggest, contra Julien 2007, Jensen & Christensen
2013, that the low acceptability/frequency of V2 under factives cannot be
explained by the twin hypotheses that MPU licenses EV2 and that factives
cannot embed MPU.
An alternative interpretation, preserving the idea that MPU licenses EV2,
would be that participants may have essentially ignored the MPU-licensing
questions when evaluating the acceptability of the responses. Under such
an account the low acceptability of EV2 under factives would have to follow
from the inability of speakers to interpret clauses in this immediate environment as the MPU. In order to investigate this possibility, two follow-up
experiments were conducted, this time with English speakers, where participants were presented with question-response pairs where the MPU of
the response was either in the main or the embedded clause, or the response did not address the question. The second variable was whether the
embedding predicate was factive or non-factive. In this experiment the participants were asked to judge whether the response was a direct or indirect
answer to the question, or did not answer it at all. In the conditions where
the response did not address the question, informants reliably scored the
responses low for directness, showing that at least here the participants
92
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 93 — #105
AG 1, Raum G 309
paid attention to the question, and that the question-response paradigm effectively manipulated MPU. Nevertheless, we found no effect of predicate
type (factive/non-factive) on judgments of how well the response answered
the question.
These follow-up experiments thus support Simons’ contention that speakers can interpret the complements to factives as the MPU. They therefore
also support our conclusion that the low rating for EV2 in factive contexts
in Swedish cannot be accounted for in purely pragmatic terms, but motivates instead a more narrowly semantico-syntactic explanation, such as
Haegeman’s 2013 intervention account.
AG1
A different perspective on embedded V2: Unifying
embedded root phenomena
Rebecca Woods
University of Huddersfield
[email protected]
Freitag, 26. 02. 2016, 12:00-12:30, Raum: G 309
The distribution and effects of embedded V2 (EV2) have been long debated.
The analysis attracting the most support in recent times in that EV2 is conditioned by and marks assertion; it only appears under matrix predicates
compatible with assertion, hence is blocked under negation and factives.
However, as Wiklund (2010) notes, there are reasons to doubt that this is
the whole story; some speakers allow EV2 under predicates which are not
typically assertive such as semifactives and negation and EV2 is not essential in order to include other root phenomena such as speech act adverbs
in the embedded clause.
This paper supports Wiklund (2010) by looking to unify Germanic EV2
with a parallel embedded root phenomenon in English: embedded questions with subject-auxiliary inversion (EIQs) such as (1).
(1)
I asked him please would he cook dinner for me
English, UK)
(North West
93
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 94 — #106
AG 1: Verb second in grammar and processing
AG1
EIQs are only available under interrogative predicates, including negated,
modalised and questioned factive predicates, but are parallel to EV2 in
being blocked under simple factives such as know and find out. EIQs are
clearly not asserted but also like EV2 as they are islands for extraction, cannot appear in sentence-initial position, and license other root phenomena
such as speech act adverbs. They also both disambiguate between competing perspectives: unlike unmarked embedded clauses, they are not ambiguous between reporting speaker and original speaker orientation. This will
be argued to be the key effect of embedded verb movement.
Interestingly, the perspective marked by embedded verb movement differs from language to language; a fact noted with respect to embedded imperatives by Kaufmann (2015). In English EIQs, subject-auxiliary inversion
gives rise to a quasi-quotational environment in which the perspective of
the original speaker (the matrix subject) takes precedence. The embedded
clause is clearly subordinate to the matrix clause as shown by indexicality and sequence of tense, as well as the (occasional) occurrence of the
complementiser under the right syntactic conditions. However, expressive
elements, speech act adverbs and discourse particles orient to the matrix
arguments, i.e. the original speakers. The original discourse is also privileged in terms of the availability of de re and de dicto readings (only the
latter are available, even if the reporting speaker has de re knowledge), and
the fact that the matrix subject is understood to have a close relationship
with or interest in the arguments of the embedded clause. The original discourse is also privileged in this way in the related Romance phenomenon
of recomplementation (the presence of multiple complementisers). These
facts also hold in English embedded imperatives, whose subject must be
the original addressee. Finally, the use of an EIQ presupposes that the EIQ
was a question-under-discussion in the original discourse context, whereas
use of an indirect question does not:
(2)
a.
b.
94
Everyone wanted to know was Jack coming to the party =
Jack’s coming was discussed
Everyone wanted to know if Jack was coming to the party =
does not entail a discussion
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 95 — #107
AG 1, Raum G 309
In EV2, by contrast, movement of the verb in the embedded clause promotes the perspective of the reporting speaker over that of the original
speaker; Wiklund (2010) claims that expressive elements in Swedish EV2
clauses orient solely to the reporting speaker, while the same elements in
non-V2 clauses can orient either to the reporting or original speaker. This
fits Kaufmann’s (2015) claims that the subject of embedded imperatives in
German must be the addressee in the reporting context.
It is proposed that these effects are brought about by the differences
in the syntax between clauses with verb movement/marked with root phenomena and non-marked clauses. It is claimed that marked clauses contain
extra structure, namely a nominalising Illocutionary Act (IA) head (cf. Potts
2002, Lahiri 2002) and a variable denoting the Centre of Evaluation (CoE)
– the coordinates of the relevant discourse and the relationship between
the relevant discourse participants. Evidence for the nominalising head includes the islandhood of the IAP and the fact that IAP clauses can directly
modify overt content nouns. There is also cross-linguistic evidence (from
Mupun, Frajzyngier 1985) for the overt spell-out of the CoE. In English, the
CoE encodes the original discourse; in Swedish and German, the reporting
discourse. This structure renders the EIQ/EV2 clause specific, picking it out
in the relevant discourse and leading to the interpretations outlined above.
(3)
I asked him [IAP [C of E] [IA n] [ForceP [Force would] [IP [DP he] [I ]
[VP [V cook] [DP dinner]]]]].
This analysis contributes to the wider discussion of the syntactisation of
perspectives and the embeddability of perspectives in language. It builds
on Cook’s (2014) work on Plains Cree to show that overt marking of perspective can be embedded, helping work towards a better understanding
of how languages and language families vary in this respect, and some of
the micro-differences involved.
95
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 96 — #108
AG 1: Verb second in grammar and processing
…Obwohl Nebensätze können doch auch assertiv sein: On
the disambiguating role of V2 in COMP-introduced
adverbial clauses
Nicholas Catasso
LMU Munich
[email protected]
Freitag, 26. 02. 2016, 12:30-13:00, Raum: G 309
AG1
In recent years, the literature on non-canonical V2 clauses introduced by
formally subordinating connectors in German has extensively focused on
the interaction between V-to-C movement and its pragmatic implications.
In particular, the relevant question concerns the function of V2 and the
conditions licensing its occurrence in adverbial clause structures in which
both arrangements would, in principle, be possible. Cf. (1):
(1)
a.
b.
Das reicht deshalb nicht, weil
das Programm des
dem suffices c.conn neg because the program of-the
Landes {ist} keine strukturelle Hilfe {ist}, sondern eine
country is no structural help is but-rather a
temporäre Unterstützung {ist}. (DLF, Nov. 25th , 2014)
temporary support
is
Find ich recht positiv, daß da irgendwie geholfen wird,
find I quite positive that there somehow helped is
diese Leute zu finden, obwohl es hat auch seine Nachteile
these people to find although it has also its drawbacks
hat. (AGD, Dec. 12th , 1974)
has
While the standard analysis implies a paratactic categorization of such constructs in light of their apparent illocutionary independence (cf. Antomo &
Steinbach 2010, Antomo 2012, Freywald 2014), it has also been pointed out
that the corresponding Vfin embedded clauses may allow for an assertive
potential (cf. Simons 2007, Holler 2008), although this hypothesis is still under debate. Building on syntactic-pragmatic evidence (licensing of assertive
96
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 97 — #109
AG 1, Raum G 309
modal particles, agreeing test, question tags, resumption of clause-internal
cataphoric connectors realized in the introducing predicate) in both Vfin
and V2 adverbial clauses, I argue for a hypotactic analysis of V2 causal weil
clauses and concessive obwohl and wobei clauses. Given that V2 may exclusively appear in certain types of COMP-introduced adverbial constructions
allowing for an assertive reading, I will make the following points:
a. V2 weil and obwohl/wobei clauses are hypotactically, not paratactically,
bound to their matrix predicate;
b. Their relative grade of integration into and dependency on the matrix
clause is by no means affected by the position of the verb, which amounts
to the assumption that the V2/Vfin arrangement is basically not sensitive to
Haegeman’s (2004, and much subsequent work) distinction between central and peripheral adverbials (vs. Freywald 2014);
c. The role of V2 in COMP-introduced adverbial clauses consists in disambiguating the assertive potential of the embedded clause.
References: • Antomo, Mailin / Steinbach, Markus (2010). Desintegration und Interpretation.
Weil-V2-Sätze an der Schnittstelle zwischen Syntax, Semantik und Pragmatik. Zeitschri ür
Sprachwissenscha 29: 1-37. • Antomo, Mailin (2012). Interpreting Embedded Verb Second.
Causal Modifiers in German. In Costantinescu, Cornelia et al. (Eds.), Proceedings of ConSOLE
XVII, 27-51. • Freywald, Ulrike (2014). Parataktische Konjunktionen. Zur Syntax und Pragmatik
der Satzverknüpfung im Deutschen - am Beispiel von obwohl, wobei, während, wogegen und dass.
PhD dissertation, Humboldt-Universität zu Berlin. • Haegeman, Liliane (2004). The Syntax of
Adverbial Clauses and ist Consequences for Topicalization. In Coene, Martine / de Cuyper,
Greet / d’Hulst, Yves (Eds.), Antwerp Papers in Linguistics 107 [= Current Studies in Comparative Romance Linguistics], 61-90. Antwerp: University of Antwerp. • Holler, Anke (2008). German Dependent Clauses from a Constraint-Based Perspective. In Fabricius-Hansen, Cathrine
/ Ramm, Wiebke (Eds.), ‘Subordination’ versus ‘Coordination’ in Sentence and Text: A Cross-Linguistic Perspective [= Studies in Language Companion Series 98], 187-216. Amsterdam: Benjamins. • Simons, Mandy (2007). Observations on Embedding Verbs, Evidentiality, and Presupposition. Lingua 117/6: 1034-1056.
97
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 98 — #110
AG 1: Verb second in grammar and processing
Separating Tense and Assertion: Evidence from Embedded
V2 and Child Language
Thomas Roeper1 & Rebecca Woods2
UMass Amherst, 2 University of Huddersfield
1
[email protected], [email protected]
Freitag, 26. 02. 2016, 13:00-13:30, Raum: G 309
AG1
We explore the claim that Tense and Assertion are separate projections in
the grammar of some Germanic languages (following Klein 1998, 2006 and
Duffield 2007). Our principal claim is that auxiliary inversion removes the
presupposition of truth. Our approach leads towards an eventual mapping
of the related effects of assertion, Verum Focus and point of view onto a
syntactically present (and separately available) Illocutionary Act Phrase in
both the matrix and the subordinate clause (Woods, to appear).
An asserted truth is the illocutionary force of a declarative with the
tensed verb in situ. Inversion lifts the presupposition inherent in a declarative when Tense moves to V2. In German, for example, there are three
positions for the tensed verb (assuming matrix V2 is between C and T in
Fin):
(1)
Tensed verb final (embedded clauses) – Presupposed; no assertion
a. Ich weiss warum er singen kann
(German)
(2)
V2 (matrix and embedded clauses) – Asserted; no presupposition
a. Ich kann nicht singen!
(German)
b. Maria sagte, er kann singen.
(German)
(3)
Auxiliary inversion – no assertion, no presupposition, therefore interrogative
a. I asked her could he sing (English dialects, Woods to appear)
There is much evidence for this: in standard embedded clauses, the verb
is left in situ and the clause is neutral as to illocutionary force. In embedded V2 clauses, the verb raises, the presupposition of its truth is dropped,
and the clause is treated as an assertion (Julien 2009, Steinbach & Antomo
98
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 99 — #111
AG 1, Raum G 309
2010, Wiklund 2010). Embedded V2 is not available under predicates which
independently induce a presupposition of their truth, such as factive complements:
(4)
a. *Ich weiss warum kann er singen/*Ich weiss warum er kann
singen
(German)
b. *I knew could he come tonight
(English dialects, Woods to
appear)
Furthermore, a fleeting but remarkable feature of English child language,
auxiliary doubling, shows evidence of the two positions available to the
tensed verb (e.g. “Is Tom is busy?”). Children use auxiliary doubling in the
place of more complex tag questions (e.g. “Tom is busy, isn’t he?”) or cleft
constructions (“Is it that Tom is busy?”) in order to maintain a presupposition about which a further question is asked:
(5)
Father: Do you want to go outside?
Child: No! (to friend:) Do you don’t want to go outside? (Child, 4;0,
Roeper 2014)
In this case, the child is not asking if their friend wants or doesn’t want
to go outside (i.e. a normal polar question), but whether the friend is in
agreement with a view that presupposes not going outside. We argue that
experimental studies (e.g. Rowland and Theakston 2009a,b) who elicit a
large proportion (around 40%) of auxiliary doubled questions by children
between 2;6 and 3;6 and claim them as incorrect polar questions actually
induce these structures due to an experimental setup biased towards these
kinds of semi-confirmation questions.
This talk provides an avenue of explanation for the presence of two tense
positions in German, which has been a neglected problem for decades. We
propose different semantic and discourse functions for each position: the
lower position is purely a [+Tense] position where the higher position is
also [+Assertion]. However, this distinction is masked in languages like
English in which both functions are conflated on the same head, T. Our approach makes predictions about Verum Focus: where it typically attaches
to Tense in English, it can only attach to the V2 position in German, so we
99
AG1
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 100 — #112
AG 1: Verb second in grammar and processing
predict that Verum Focus is only available on the complementiser and not
the tensed main verb in verb-final clauses, but it is available on the verb
in EV2: both predictions are borne out (Höhle 1992). Note that verb movement to the [+Assertion] position is the not only way to achieve an asserted
interpretation: the presence of discourse particles and speech act adverbs
in non-V2 clauses leads to an asserted interpretation in Swedish (Wiklund
2010) and in English, emphatic do-support forces an asserted interpretation when the verb remains low. The difference with German is that the
verb final position makes explicit – particularly in embedded clauses – the
separation of the two “Tense” positions, long a mystery.
AG1
References: • Duffield, N. (2007), Aspects of Vietnamese clausal structure, Linguistics 45.
• Höhle, T (1992) Über Verum-Fokus im Deutschen, Linguistische Berichte, Sonderheft 4/1991-92.
• Julien, M. (2009), Embedded clauses with main clause word order in Mainland Scandinavian,
Ms., University of Lund.. • Roeper, T. (2014) Strict interface principles, Language Sciences.
• Steinbach, M. and M. Antomo (2010) Desintegration und Interpretation, Zeitschr. ür Sprachwissenscha. • Wiklund, A.-L. (2010), In search of the force of dependent verb second, Nordic
Jnl of Linguistics 33(1). • Woods, R. (to appear) Embedded Inverted Questions as Embedded
Illocutionary Acts, WCCFL 33.
100
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 101 — #113
AG 1, Raum G 309
Summary Discussion
Oliver Bott1 , Constantin Freitag2 & Fabian Schlotterbeck1
1
Universität Tübingen, 2 Universität Konstanz
oliver.bott@uni-tübingen, [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:30-14:00, Raum: G 309
AG1
Alternate speakers
V2 in a sign language
Jóhannes Gísli Jónsson1 & Elísa Guðrún Brynjólfsdóttir1 , 1 University of Iceland
[email protected], [email protected]
101
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 102 — #114
A Grammar of Haro
Hirut Wolde-Mariam
Addis Ababa University
Haro is an endangered language spoken by less than 200 people
who live on the eastern shore of an island in Lake Abaya. Lake
Abaya is located in the southwestern part of Ethiopia.
Genetically, the language belongs to the Ometo linguistic group
of the Omotic language family within the Afro-Asiatic superfamily. This study provides description of the phonological,
morphological and syntactic structures of the language. The
structures of nouns, pronouns, verbs, adjectives, deictics,
numerals, simple sentences and complex sentences are
described and analyzed. Haro has a largely suffixal, transparent,
agglutinative morphology that allows concatenation of up to
four suffixes in a word stem. It is common for inflectional
categories to be expressed cumulatively by the use of
portmanteau morphemes in contrast to derivational categories
expressed by separate morphemes.
Haro is an interesting language from typological and
historical perspectives. For instance, unlike the situation with
related languages, the case system in Haro involves three core
cases, and employs 'differential case marking' that exempts
certain nouns from case marking. The three-way number
marking in nouns is also attested uncommon among the Ometo
languages. Haro exhibits an intricate system of focus marking
that affects the morpho-syntactic properties and categorization
of a verb.
ISBN 978 3 386288 666 1. Languages of the World/Materials
505. 248pp. EUR 72.80. 2015.
Das Konzept der Objektrelation und das Kontinuum
ihrer Varianten: Ein
muttersprachlicher Zugang
Hansjakob Seiler (ed.)
Hansjakob Seiler, Ioanna BerthoudPapandropoulou & Yoshiko Ono
Die vorliegende Studie ist ein Versuch, sich einer postulierten
Universalität des Begriffs der Relation „Objekt im Satzgefüge“
auf vergleichend-einzelsprachlichem Wege zu nähern. Die drei
nach Verfügbarkeit ausgewählten Einzelsprachen sind Deutsch,
Neugriechisch und Japanisch. Für sie sind, je Sprache, die drei
Co-Autoren als Muttersprachler zuständig.
Es ist bekannt, dass „Objekt“ in den Einzelsprachen nicht als
Monolith, sondern in mehreren Erscheinungsformen auftritt,
und dass es von Sprache zu Sprache nicht immer dieselben
Erscheinungsformen sind. Unsere Grundhypothese besagt, dass
sich, zunächst in einer Einzelsprache, die dort feststellbaren
Erscheinungsformen in einer kontinuierlichen Abfolge
zwischen zwei Polen ordnen lassen und dass dadurch erst ihr
Status als Varianten erwiesen werden kann. Wenn in einer
LE
LINCOM EUROPA
academic publications
prätheoretischen Annahme das Konzept „Objektrelation“
bestimmt werden kann als Betreffen einer EntitÄT, so ist der
eine Pol eben das Betreffen im Prinzip ein Vorgang (Verb), der
andere Pol die EntitÄT im Prinzip eine Wesenheit (Nomen).
Es geht jetzt um den Nachweis einer kontinuierlichen
Abfolge einzelsprachlicher Ausdrücke zwischen den zwei
Polen. Maßstab ist der Begriff der Prominenz: Prominenz der
EntitÄT nimmt zu von Pol1 („links“) nach Pol2 („rechts“) und
nimmt ab in entgegengesetzter Richtung. Für Betreffen gilt das
Umgekehrte. Zum Nachweisen der kontinuierlichen Abfolge ist
es nicht nötig, dass in einer Sprache alle „Stationen“ eines
Kontinuums vertreten sind. Partielle Kontinua sind
aussagekräftig, wenn sie semantisch bzw. pragmatisch dieselben
Eigenschaften der Promenienz: entweder „mehr nach Pol1“ oder
„mehr nach Pol2“ aufweisen. Mit in die Argumentation
einzubeziehen sind Beobachtungen des kookkurrenten
Verhaltens der Verben sowie der Subjekte im Satz. Hier gelten
oft feine semantische oder pragmatische Unterschiede, weshalb
das Wissen eines Muttersprachlers unabdingbar ist.
ISBN 978 3 86288 642 5 (Hardcover). LINCOM Studies in
Theoretical Linguistics 57. 153 S. EUR 84.80. 2015.
On Laryngealism
A Coursebook in the History
of a Science
Joe Voyles and Charles Barrack
University of Washington
This book is a much-needed refutation of the laryngeal theory of
Proto-Indo-European (PIE) phonology. It is intended for both
linguists and non-linguists, particularly scholars interested in the
history and logic of the sciences. Each chapter concludes with
exercises and a key with answers to the exercises.
Chapter 1 "Terminology and method." explains the
approach to both synchronic and historical linguistics taken in
this work. Chapter 2 "The Indo-European (IE) background."
gives a survey of the major IE languages as well as a summary of
reconstructed PIE phonology. Chapter 3 "Laryngealism."
presents a history of the development of laryngeal theory from
Saussure (1878) until contemporary times. Chapter 4 contains
two examples of laryngealist methodology, namely the
reconstruction of the original PIE vowel system as well as a
laryngealist explanation of the varying reflexes of the class-7
reduplicating verbs in Proto-Germanic. Both these laryngealist
accounts are negatively critiqued; and alternative explanations
are proposed. Chapter 5 "The case of Hittite." treats of the Hittite
evidence, often adduced in support of laryngealism. This
evidence is found to be specious. Chapter 6 "Logic and
laryngealism." places the laryngeal theory within the spectrum
of other false theories which have been proposed from time to
time in various sciences - such as the phlogiston theory in
chemistry. Hence laryngealism is shown to have been one of the
growing pains in the history of the science of linguistics.
ISBN 978 3 86288 651 7. LINCOM Coursebooks in
Linguistics 23. 130pp. EUR 52.80. 2015.
webshop: www.lincom-shop.eu
LINCOM GmbH
Hansjakobstr. 127a, D-81825 Muenchen
[email protected]
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 103 — #115
Arbeitsgruppe 2
The syntax of argument structure: empirical
advancements and theoretical relevance
Artemis Alexiadou1 & Elisabeth Verhoeven1
1
Humboldt-Universität zu Berlin
AG2
[email protected], [email protected]
Raum: G 300
Workshop description
The syntax of argument structure has been the focus of a number of empirical studies investigating phenomena such as the causative alternation
(Fadlon 2014), unaccusativity (Keller & Sorace 2003, Hirsch & Wagner 2011,
Irwin 2013, Verhoeven & Kügler 2014), ergativity (Longenbaugh & Polinsky 2013), argument hierarchies and argument realization (Bornkessel et
al. 2005, Lamers & de Swart (eds.) 2012), the dative alternation (Bresnan
et al. 2007), inherent vs. structural case (Jacobsen 2000, Bayer et al. 2001),
psych predicates (Lamers & de Hoop 2014, Verhoeven 2014, 2015), etc. Such
studies provide interesting but potentially controversial contributions to
linguistic theory: some discover gradience in the verbal lexicon that can
only be precisely measured with quantitative methods (see e.g., Keller &
Sorace 2003); others claim that properties attributed to verbal syntax are
an epiphenomenon of other layers of grammar (see e.g., Hirsch & Wagner
2011); yet other studies show reflexes of core properties of verbal syntax in
processing (see e.g. Polinsky et al. 2012 on ergativity).
The workshop will address the following issues:
• Does the progress in empirical methods promise theoretical advancements in the syntax of argument structure? In particular, do em-
103
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 104 — #116
AG 2: e syntax of argument structure
pirical data obtained through corpus or experimental methods confirm/reject/ validate evidence previously gained through linguistic
intuitions?
• A major shortcoming of experimental and corpus data is that they
contain artefacts associated with sources of variation that are external to the grammar. How can we distinguish between grammatically relevant information and grammar-external variation that is
involved in experimental or corpus data?
• Theoretical accounts make a distinction between core grammatical
properties and linguistic properties attributed to processing. How
can this distinction be established by experimental data?
AG2
This workshop brings together theoretical linguists, corpus linguists,
and psycholinguists that are interested in the syntax of argument structure and employ precise empirical methods in building theories thereof.
The discipline of experimental linguistics
Maria Polinsky
University of Maryland
[email protected]
Mittwoch, 24. 02. 2016, 14:00–15:00, Raum: G 300
Experimental work is now pervasive in linguistics, and one of the pressing
questions has to do with the need to combine theoretical and experimental
approaches to languages. In the present talk I evaluate those contexts in
which experimental work can, in my opinion, be truly useful, versus those
contexts where I believe such work it does not move the field forward.
The main conclusion is that theory and experimental work on language go
hand in hand, and experimental work constitutes yet another diagnostic
tool in the analysis of data. In thinking about experiments, it is important to draw a line between confirmatory studies, which help us choose
one theory over another, and exploratory studies, which are needed to
establish a general idea of the empirical landscape – before a solid theory
104
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 105 — #117
AG 2, Raum G 300
is built. Formal experiments are well suited to the needs of confirmatory
studies, which test competing theories, but it is less clear whether such
experiments are fit for exploratory work. In the context of confirmatory
studies, successful experimental work should be driven by theoretical questions. Worthwhile experiments build on theoretical knowledge and feed
back into existing linguistic theories, forcing us to revisit familiar concepts
and develop new theoretical principles. This unsurprising conclusion also
suggests an important corollary: that we may get the best overall results
by working in teams. A successful team needs a theoretical linguist, an
empirical expert who controls the subtleties of the language under investigation, and a good experimental linguist. Building on this premise, this
talk concludes by considering a number of readily available opportunities
for team building, including dissertation committees, community involvement, cross-disciplinary projects, and international collaborations.
Effects of repairing illegal argument structures
Patrick Brandt1 & Petra Schumacher2
IDS Mannheim, 2 University of Cologne
1
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: G 300
Recently, the idea that certain argument structure configurations violate
conditions of the interface and are therefore in need of repair before interpretation has lead to exciting new perspectives on some long standing
problems; e.g., Schäfer 2013 argues for passives of reflexive verbs that case
and binding conflicts are repaired syntactically by means of a special, last
resort agreement relation. Regarding the semantics/ pragmatics side of the
interface, Schumacher 2015 presents evidence from event-related brain potentials that the processing of argument structures involving ’privative’
adjectives like fake exerts a cost that is due plausibly to ‘shifting’ the reference of the full NP beyond the denotation of the head noun, cf. (2)
(2)
Geart is a tall Dutch professor. → Geart is a professor.
Geart is a fake professor. → Geart is not a professor.
105
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 106 — #118
AG 2: e syntax of argument structure
AG2
Schumacher’s experimental results suggest that the generation and ensuing repair of “illegal” argument structures may provide for particular
sound-meaning pairings that would not be available modulo particular
interface repair strategies. Further exploring this option theoretically as
well as experimentally, the talk discusses whether – and how – surprising
meaning aspects of certain argument structure configurations – in particular, modal interpretations of superficially reflexive mediopassives or of
excessive structures – can similarly be attributed to special repair operations at the interface that lead to an interpretive shift of certain meaning
components (cf. Brandt 2009).
(3)
Der Text liest sich nicht gut. (~man kann den Text nicht gut lesen)
the text reads SICH not well (~one can not read the text well)
(4)
Der Text ist zu lang. (~der Text ist länger als er sein soll)
the text is too long (~the text is longer than it should be)
From a theoretical perspective, the effects of repairs of prima facie unusable argument structures may provide an alternative to the postulation of
construction meaning (Goldberg 1995) or to the stipulation of empty structure that is typically held responsible for unexpected meaning aspects in
the generative camp (e.g., Bhatt 2006). We thus hope to contribute to making an empirical case against closing the analytical record on certain recalcitrant structures too early by retreating to descriptive as opposed to explanatory instruments in grammatical analysis. References: • Bhatt 2006: Covert
Modality in Non-finite Contexts. Berlin: de Gruyter. • Brandt 2009: Generische Möglichkeit
in Medialkonstruktionen. In: W. Abraham und E. Leiss: Modalität. Tübingen: Stauffenburg,
79-100. • Goldberg, A. 1995: Constructions. A Construction Grammar Approach to Argument
Structure. Chicago: University of Chicago Press. • Schäfer, F. 2013: Passives of reflexive verbs:
e repair of a Principle A violation. In: P. Brandt & E. Fuß (eds.): Repairs. The added value
of being wrong. Berlin: de Gruyter. • Schumacher, P. (2015): Processing vagueness: e online
comprehension of adjective-noun combinations. Paper presented at the workshop Gradability,
Scale Structure, and Vagueness: Experimental Perspectives. Center for Social Sciences and
Humanities, Madrid.
106
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 107 — #119
AG 2, Raum G 300
It’s all about verb-type: Passives are not inherently more
complex than actives
Nino Grillo1 , Berit Gehrke2 , Nils Hirsch1 , Caterina Paolazzi3 &
Andrea Santi3
1
Humboldt-Universität zu Berlin, 2 CNRS & Universität Stuttgart,
3
University College London
[email protected], [email protected], [email protected],
Mittwoch, 24. 02. 2016, 16:30–17:30, Raum: G 300
Complexity of passives has often been imputed to agent-first heuristics
(Bever 1970; Ferreira et al. 2003; Townsend & Bever 2001). In passives,
agent-first would result in an inaccurate sentence representation, revealed
by comprehension errors, unless the output is corrected by algorithmic
processes of the parser. These revision processes should impose higher processing costs (observable, e.g., as longer reading times) in passives than actives, where reanalysis is not required. In line with this view, offline studies
demonstrated that passives are harder to process than corresponding actives (Ferreira 2003). Online studies, however, do not provide supporting
evidence for higher processing costs in passives. On the contrary, the main
verb in passives is read at the same speed as in actives, if not faster (Carrithers 1989; Rohde 2003; Traxler et al. 2014). The asymmetry could be generated by several factors. English passives are often ambiguous between a
verbal and an adjectival interpretation. Disambiguation depends largely on
a combination of verb type and the presence of a by-phrase. The majority of
previous studies of verbal passives, however, did not control the properties
of the predicates selected in the experimental stimuli (which included perceptual and psych verbs, activities and accomplishments/achievements). In
the current study, we control for these variables by using: (1) only eventive
predicates which introduce a clear consequent state sub-event, (2) German
verbal passives, which, contrary to English, are unambiguously introduced
by the auxiliary wurde.
Method: : 34 native German-speakers participated in a self-paced reading
task contrasting actives and passives. Each of the 30 experimental and 60
filler sentences was followed by a comprehension question.
107
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 108 — #120
AG 2: e syntax of argument structure
AG2
Results: : There was no significant difference in Accuracy scores (83.5% passives vs. 86.1 actives) and Response Times in comprehension questions.
There was also no significant difference in Reading Times at the verb. Both
the offline and online results clearly indicate that passives are not inherently more complex than actives once certain properties of the verb are
controlled for and, consequently are problematic for agent-first heuristics.
Differences between present and previous findings (the uniformity of our
offline and online results) are rooted in the properties of the predicates used
across studies. Two classes of verbs commonly used in previous experiments are known not to freely participate in verbal passivization (perceptuals/object experiencers). We further show that unambiguous verbal passives of subject-experiencer predicates (another type of stative verb commonly used in the previous literature) are severely restricted with episodic
by-phrases and prefer generic ones, in both German and Italian. We take
the limited availability of verbal passivization with states (Gehrke & Grillo
2009) to be the source of the problem with previous experiments, and argue that alternative, frequency-based, accounts (Street & Dabrowska, 2006)
miss this important generalization. A follow-up study with stative predicates in unambiguous verbal passives is under way to obtain a clearer picture on both accounts.
Discourse and unaccusativity: Quantitative effects of a
structural phenomenon
Patricia Irwin
University of Pennsylvania
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:00, Raum: G 300
Introduction. This paper brings together syntactic analysis with corpus
results to argue that a subset of unaccusative VPs – those that denote simple motion (e.g., arrive, come in) share syntactic structure with existential BE sentences in English, and that other unaccusative VPs (roughly,
those that denote changes-of-state) do not share the relevant structure. In
108
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 109 — #121
AG 2, Raum G 300
this way our analysis follows theoretical work that argues for more than
one type of unaccusative VP (Kural 1996; Alexiadou and Anagnostopoulou
2004; Deal 2009, inter alia). We provide experimental support for our analysis in the form of a corpus experiment that tests the hypothesis that in
all-new sentences, only the relevant subset of unaccusative VPs serve the
same discourse function as existential BE sentences. Indefinite subjects
of unaccusative sentences. Although English allows indefinite subjects,
they occur with vanishing frequency (Prince, 1981). One explanation for
this might be a general processing preference such that old entities occur
before new entities; indeed, English usage generally conforms to givennew ordering. But we show that among intransitive sentences, violations
of given-new ordering occur with a coherent subset of predicates: unaccusatives that denote directed motion (unacc-simple motion). We argue
that these VPs share structure and meaning with existential BE sentences,
and that unacc-simple motion VPs have the discourse effect of establishing new discourse referents (dRefs) by the same means as existential BE
sentences (McNally, 1997). These properties are illustrated in the made-up
contrasts shown in (1), where # shows degraded felicity in dRef establishment.
(1)
Context: “We were sitting around the bar last night … ”
a.
There was a fancy lady next to me. She ordered a drink. existential BE
b.
A fancy lady waltzed in. She sat down next to me.
unacc-simple
c.
A fancy lady sneezed. #She sat down next to me. motion unergative
d. A glass broke. #It went into many pieces.
unacc-change of state
Syntactic analysis and corpus results. We present data from the Switchboard Corpus (Godfrey et al., 1992) that support the dRef-introducing properties illustrated in (1): the ratio of unaccusative (86%) to unergative (14%)
VPs with discourse-new subjects is significantly different (p < .001) from
the ratio of unaccusative to unergative sentences with subjects of any discourse status: in other words, given the frequency of unaccusatives in the
corpus with subjects of any information status, the higher frequency of
unaccusatives with discourse-new subjects is not the result of chance. Our
syntactic analysis extends McCloskey’s (2014) analysis of Irish existentials
109
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 110 — #122
AG 2: e syntax of argument structure
such that unacc-simple motion VPs and VPs with “unergative” roots that
establish dRefs (e.g., a lady walked in) have an existential predication involving contextually determined location as part of their structure and
meaning (Francez, 2007). In these VPs, an activity-denoting little-v selects
for a SC whose specifier is a PathP. We suggest that an integrated approach
to argument structure is important in the current theoretical landscape,
where what constrains acceptability is not the Theta Criterion but syntactic structures and available interpretations (Marantz, 2013).
AG2
Relativization in two morphologically ergative languages:
a corpus study
Dmitry Ganenkov
Institute of Linguistics, Moscow / Bamberg University
[email protected]
Mittwoch, 24. 02. 2016, 18:00–18:30, Raum: G 300
Lezgian and Avar are morphologically ergative languages from the NakhDaghestanian (East Caucasian) language family spoken in Daghestan, Russia. The present corpus study deals with frequency of relativization of the
core grammatical functions S, A, and P in these two languages. The main focus of this paper is on relativization of core arguments in clauses with twoplace (standard transitive, TR, and subject experiencer, SE) verbs. Three
types of data on relativization in Lezgian and Avar have been extracted
from corpora yielding 11 different datasets for each language: (i) a random
sample of 2000 relative clauses, (ii) a random sample of 150 core argument
(subject or object) relative clauses for each of the following five TR verbs:
‘write’, ‘put’, ‘build’, ‘eat’, ‘show’, and ‘throw’, (iii) a random sample of 150
core argument (subject or object) relative clauses for each of the following five SE verbs: ‘see’, ‘hear’, ‘find’, ‘love’, and ‘know’. Two main empirical findings of this corpus study are as follows. First, Lezgian and Avar
differ with respect to frequency of relativization of A and P. Avar, as preliminarily reported earlier by Polinsky et al. (2012), shows no preference
for relativization on A or P. Lezgian, by contrast, displays statistically significant preference for relativization on the absolutive P. Second, in Avar
110
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 111 — #123
AG 2, Raum G 300
relative clauses, SE verbs do not significantly differ from TR verbs. In Lezgian, however, there is such a difference – in contrast to standard A and
P, relativization of SE verbs does no provide statistical evidence for preference of the absolutive object over the dative subject, or vice versa. In
addition, the difference between TR and SE verbs seems to be gradient,
since some SE verbs are close to TR verbs with respect to relativization,
while others are radically different. These two points of divergence – (i)
object preference in Lezgian vs. no relativization preference in Avar, and
(ii) object preference with TR verbs vs. no preference with SE verbs in Lezgian – are an input for a theoretical analysis. In this paper, I argue that
neither proccessing-based accounts nor purely semantic/thematic explanations can predict the observed distribution. In particular, I argue against
the account proposed by Polinsky et al. (2012) who derive the absence of A
or P preference in Avar relative clauses from dissociation of, and competition between, grammatical function and case in morphologically ergative
languages where grammatical function (subject) works for the A preference, whereas morphological case (absolutive) works for the P preference.
I show that the observed differences in frequency correlate with data on
anaphor binding and conclude that both ultimately derive from different
structural position of core arguments at earlier steps of derivation. In the
end, I briefly discuss the results of this study in the context of current theoretical approaches to syntactic ergativity.
Syntactic priming as a test of argument structure: A
self-paced reading experiment
I. Oltra-Massuet1,2 , V. Sharpe3 , K. Neophytou2 & A. Marantz2,3
Universitat Rovira i Virgili, 2 New York University Abu Dhabi, 3 New York
University
1
[email protected], [email protected], [email protected], [email protected]
Donnerstag, 25. 02. 2016, 09:00–10:00, Raum: G 300
Using data from a structural priming experiment, we test two competing
theoretical approaches to argument structure, (i) Hale & Keyser’s (1993,
111
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 112 — #124
AG 2: e syntax of argument structure
2002) approach as developed in Mateu (2002), Acedo-Matellán (2010) and
Acedo-Matellán & Mateu (2011, 2013) [AM&M], and (ii) Marantz’s (2005,
2011) [M]. These theories attribute different structures to transitive structures like (2-6) and make different claims about the relationship between
transitive structures and unergatives like (1), thus making different predictions about priming relations between them.
CONDITIONS
AG2
NP
V
NP(/PP)
PP
(1) C1. Unergative
The dog
barked
in a quiet park
at night.
(2) C2. Cognate
The man
dozed
a restful doze
on the train.
(3) C3. Creation
The cook
baked
a carrot cake
with spelt flour.
(4) C4. Saddle/Shelve
The girl
saddled
a wild horse
in the farm.
(5) C5. Strong transitives
The athlete
ignored
a slight niggle
in his knee.
(6) C6. Spray/Load ‘with’
The worker
loaded
a rail wagon
with hay.
In AM&M theory unergatives (1) are analyzed as derived transitive configurations and pattern with cognate objects (2) as well as with verbs of
creation (3), thus predicting syntactic priming among these sentence types
but not between these sets and the remaining types (4)-(6). The latter are
assumed to select for a small clause type complement structure, and are
predicted to prime among them in this model. On the other hand, the M account does not predict structural priming between the unergatives (1) and
the surface transitives, nor between complex complement constructions (6)
and the other surface transitive sentences. However, M approach does predict some cases of priming that the AM&M theory does not; specifically,
M predicts priming between sets (2)-(3) and (4)-(5), which display distinct
underlying structures in the AM&M account. We run a self-paced reading
language comprehension study to 600 subjects over MTurk. 24 sentences
of each type were selected, to be read in 4 chunks (subject, verb, direct object/PP, PP), presented in 3 blocks of 48 in a randomized order. The large
number of subjects allows us to model the reading times at the direct object/first PP and at the second PP of the same sentences as a function of
the structure of the immediate preceding sentence, testing for structural
priming within and across sentence types.
112
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 113 — #125
AG 2, Raum G 300
References: • Acedo-Matellán, V. 2010. Argument Structure and the Syntax-Morphology Interface. A Case Study in Latin and Other Languages. UB, PhD Thesis. • Acedo-Matellán, V. &
Mateu, J. 2013. Satellite-framed Latin vs. verb-framed Romance: a syntactic approach. Probus
25, 227-265. • Hale, K. & Keyser, S. J. 1993. On argument structure and the lexical expression of
syntactic relations. The view from Building, 20, 53-109. • Hale, K. & Keyser, S. J. 2002. Prolegomenon to a theory of argument structure. MIT Press. • Mateu, J. 2002. Argument Structure.
Relational Construal at the Syntax-Semantics Interface. UAB, PhD Thesis. • Marantz, A. 2005.
Objects out of the lexicon: Objects as events. MIT, Ms. • Marantz, A. 2011. Syntactic approaches
to argument structure without incorporation. Talk presented at the Workshop Structuring the
argument, Structures Formelles du Langage UMR 7023 Paris 8/CNRS, Paris, 5-7 September.
Un/Re-packing argument and event structure restrictions
on prefixation: MEG evidence
Linnaea Stockall1 , Christina Manouilidou2 , Laura Gwilliams3 &
Alec Marantz3
1
University of London, 2 University of Patras, 3 New York University
[email protected], [email protected], [email protected],
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: G 300
We exploit selectional restrictions on verbal prefixation and the spatiotemporal resolution of MEG to investigate how native speakers process syntactic argument structure and event semantics in real time. We build on
linguistic work analyzing the distinct syntactic and semantic properties of
different prefixes, and recent MEG studies investigating the neural bases
of lexical processing. Re-, un- and out- vary in which kinds of vP they
attach to. Re- attaches low, directly to the affected nominal of a resultstateP: reopen a door = [[v open] [re[the door]]] ([1],[2]) and cannot attach to unergatives, ditransitives or psych-predicates, which do not contain the right number/type of arguments (*relaugh, *reput, *refear). Verbal un- also requires a change of state denoting vP, but requires that this
change be reversible, and involve a return to a ’normal’ state ([3]), suggesting a more direct relationship than re- with the lexical root and that
the restrictions involve conceptual event semantics: unbend a wire = [[un[v
113
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 114 — #126
AG 2: e syntax of argument structure
AG2
bend]] [a wire]] (*unflush, *unbuild). Out-prefixation also involves an internal argument, but rather than require that the vP independently ’supply’ the object, ’out-’ adds an internal argument to an otherwise unergative structure: John ran → John outran the bus = John [[v ran] [out [the
bus]]]. Out- cannot attach to obligatorily transitive or unaccusative verbal stems (*outmurder, *outfall). Previous work [5,6] finds that English,
Greek and Slovenian native speakers are faster and more accurate at detecting nonce prefixed (English) and suffixed (Greek and Slovenian) words
that violate argument/ event structure restrictions (arg.viol) than matched
nonce words that violate no restrictions, and faster and more accurate still
at detecting lexical category violating (cat.viol) affixation such as *reflat or
*καρεκλατής (karekla-tis/’chair-er’). MEG activity was recorded from 25
native English speakers as they read cat.viol and arg.viol prefixed words
and judged their acceptability. In Left-anteriorTemporalLobe (LaTL) we
found that for un- and out-, cat.viol items evoked more negative activity
than arg.viol items between 335-375ms (out-) and 365-440ms (un-), while
re-arg.viol items evoked more activity than re-cat.viol, in an earlier time
window (270-320ms). LaTL activity between 170-300ms has been associated with syntactic category entropy [7] and verb subcategorization entropy [8] for monomorphemic words, and thus our results suggest that (a)
arg.struc restrictions for re- are parsed as syntactic, and rapidly evaluated,
but (b) arg.struc restrictions on out- and un- are not. These later, opposite
direction effects appear in the time-window associated with root lexical
semantic processing [9], as expected for un-. Combining fine-grained linguistic analyses with fine-grained neuroimaging tools promises to not only
confirm key properties of well-studied argument and event structure phenomena, but also provide evidence to understand less studied ones.
References: 1 Marantz, 2007. Ms. NYU. 2 Alexiadou, Anagnostopolou & Lechner. 2014. Handout, York. 3 Horn, L. (2005). MIT Press. 4 Marantz, A. 2009. Handout, Stuttgart. 5 Manouilidou,
C. & Stockall, L. 2014. IJL 26:2 6 Manouilidou, Dolenc, Marvin, Pirtosek. (under review). CLP.
7 King, Linzen & Marantz. (in press). LI. 8 Linzen, Marantz & Pylkkänen. 2013. The Mental
Lexicon, 8:. 9 Fruchter & Marantz. 2015. B&L.143.
114
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 115 — #127
AG 2, Raum G 300
On the syntax and argument structure of agent nouns
Paul Kiparsky
Stanford University
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:30, Raum: G 300
The Functional Nominalization Thesis (Kornfilt & Whitman 2011) holds
that nominalizers head a nominal projection at some level of the V projection; the structure above and below them respectively determines their
nominal and verbal syntactic properties. Baker & Vinokurova 2009 use the
FNT to explain the lack of verbal agent nominalizations such as *the quickly
writer the leer: to have verbal properties they should be introduced above
v, but then they should not be just agentive, but should attach freely to
every kind of verb including unaccusatives. I argue that B&V’s version of
the FNT is incorrect, and (1) that agent and action nominals have verbal
syntax just in case they bear Tense/Aspect features, and (2) that the syntax
of nominalizers does not correlate with independent diagnostics of their
height in the V projection. Vedic ?-tar- (preaccenting) should be “high” by
B&V’s criteria: it forms agent nouns that assign structural case, take adverbs, and have strictly agentive argument structure:
(1)
íṣkartā
víhrutam púnaḥ (RV 8.1.12)
fixer-Nom wrong-Acc again
‘the maker right again (o) what has gone wrong’
But it is structurally low: it is always adjacent to the root, and never goes on
prefixed bases. It is inherently present/imperfective in that it only denotes
agents of ongoing eventualities. The tenseless accented agent nominalizer
-tár- has B&V’s “low” properties: it can be added to non-agentive/unaccusative verbs and forms nominals that take genitive objects and adjective
modifiers. But is structurally high: it is separable from the root by causative
and other V→V suffixes, and affixed to the whole verb base, including its
preverbs. Vedic nominalizers with nominal properties can’t be spelled out
115
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 116 — #128
AG 2: e syntax of argument structure
AG2
low and raised to their actual position, because the v head may contain
causative and other suffixes, and because of accent and tmesis, and nominalizers with verbal properties can’t be spelled out high and then lowered
to their actual position, for we can’t ensure at spellout that this position
is empty. The Finnish agent nominalizer -ja has a mix of “high” and “low”
properties that is also incompatible with the FNT. It attaches to any kind of
verb including unagentives / unaccusatives, yet does not assign structural
case and is compatible with Voice morphology. As expected on the present
proposal, agent nouns of this type consistently lack Tense/Aspect features.
My conclusion that the best predictor of verbal properties in agent nominalizations is Tense/Aspect is broadly compatible also with recent work on
action nominalizations: verbal gerunds have imperfective Aspect (Pustejovsky 1995, Alexiadou 2001, Alexiadou et al. 2010), whereas regular action nominalizations are aspectless nominal heads. It also relieves little v
of the functional overload it has acquired in recent syntactic work. In fact,
it opens the door to an account of the generalizations behind the FNT in a
lexicalist approach to nominalizations.
Argument structure and reflexive binding
Tibor Kiss
Sprachwissenschaftliches Institut, Ruhr-Universität Bochum
[email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: G 300
Examples like (1) show a surprising ambiguity: the reflexive contained in
the lower object NP can take the NP-internal specifier as well as the subject
as its antecedent.
(1)
Annaj betrachtete [NP [SPR Karinsi ] Bild von sichi/j ].
Similarly, we find grammatical examples in English that according to anyoneÕs Principle A of Binding Theory should be ungrammatical (cf. Kuno
(1987), Runner et al. (2003), and Runner (2007)).
(2)
116
Maryi isn’t interested in [anybody’s opinion about herselfi ].
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 117 — #129
AG 2, Raum G 300
Runner (2007) has suggested that the grammaticality of (2) can be accounted
for if the argument structure of the derived nominal is considered. Basically, he assumes with Grim- shaw 1990 (cf. also Borer 2013) the presence
of an argument structure in complex event nominals. The absence of argument structure suggests an analysis in which the reflexive is exempt
(Reinhart and Reuland 1993, Pollard and Sag 1994). This analysis is problematic for various reasons, a major one being that the phenomenon should
not be observable in languages without exempt reflexives, such as German
(cf. Kiss 2001, 2003, 2012). In this talk, we will present an extension of Kiss
(2012) in which anaphoric dependencies are not only introduced through an
argument structure, but also by the presence of a syntactic specifier. The
crucial difference is that the presence of an argument structure does not
only introduce a dependency, but also requires its local termination. The
presence of a specifier in (1) introduces an anaphoric dependency, but the
absence of an argument structure makes it possible that the dependency is
only eliminated later, i.e. upwards, in the structure. The analysis predicts
that derived nominals with an argument structure, such as (3), do not show
the ambiguity present in (1).
(3)
Mankej erzählte, dass erj [NP Heinesi Untersuchung gegen sichi/j ]
erwartete.
The analysis does not involve exemption, but deals with the ambiguity
by separating specifiers – as syntactic means– from argument structure,
presumably a manifestation of lexical semantics.The analysis is contrasted
with a sentence comprehension experiment, in which we wanted to find
out whether the assumed is dependent on the presence of argument structure. Preliminary results suggest that this might be the wrong track.
117
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 118 — #130
AG 2: e syntax of argument structure
Grammar under pressure: The case of subject hun ‘them’ in
Dutch
Helen de Hoop
Radboud University Nijmegen
[email protected]
Freitag, 26. 02. 2016, 11:30–12:30, Raum: G 300
AG2
The object pronoun hun ‘them’ in Dutch has been in use as a subject for
over 100 years now. Its use has spread over all parts of the Netherlands and
is found in speakers of all ages and social classes. At the same time, this
use of hun ‘them’ as a subject is widely disapproved of in the Netherlands,
and it generally prompts strong feelings of repugnance, especially in highly
educated native speakers of Dutch. A linguistic analysis of the use of hun
‘them’ as a subject (van Bergen et al. 2011) led to an enormous debate in the
media in the Netherlands, including even a letter submitted by the Minister
of Education to a national newspaper in which he declared that he would
never allow for this construction to become part of Dutch grammar. While
the grammar of an adult speaker is usually thought of as a fully symmetrical system in which sentences that can be interpreted by the grammar
will also be produced by that grammar and the other way around (Hendriks 2014), prescriptive rule violations such as the use of hun ‘them’ as
a subject in Dutch raise an interesting question, because many especially
highly educated speakers do not produce such constructions themselves
while they do understand them perfectly well. That is, native speakers of
Dutch all share the grammatical intuition that hun ‘them’ in a sentence like
Wat maken hun een vreselijk lawaai! ‘They are making a terrible noise!’ can
only refer to people (or animals), while its prescriptively correct counterpart ze ‘they’ in the same context could also refer to engines or air planes
(de Hoop 2013). Therefore, while some people do not produce hun ‘them’ as
a subject themselves, they do interpret such constructions correctly. This
raises the question whether grammatical norm violations are part of their
grammar or not. In order to address that question, I will report on an fMRI
experiment that we conducted in Nijmegen (joint work with Ferdy Hubers and Tineke Snijders) to examine the differences in processing between
118
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 119 — #131
AG 2, Raum G 300
grammatical norm violations on the one hand, and truly ungrammatical as
well as truly grammatical sentences on the other. The aim of my talk is to
shed light on how to model the interaction of grammatical principles and
factors of argument marking that go beyond the grammar proper, such as
sociological pressure reflected in grammatical norms.
References: • van Bergen, G., Stoop, W., Vogels, J. and de Hoop, H. (2011). Leve hun! Waarom
hun nog steeds hun zeggen. Nederlandse Taalkunde 16, 2-29. • Hendriks, P. (2014). Asymmetries between Language Production and Comprehension. Studies in Theoretical Psycholinguistics, Vol. 42. Springer, Dordrecht. • de Hoop, H. (2013). The rise of animacy-based differential subject marking in Dutch. In: Serzant, I.A. and Kulikov, L. (eds.), The Diachronic
Typology of Non-Canonical Subjects. John Benjamins, Amsterdam/Philadelphia.
Case marking affects the processing of animacy with
simple verbs, but not particle verbs: An event-related
potential study
Anna Czypionka1,2 & Carsten Eulitz1
1
Konstanz University, 2 Wrocław University
[email protected], [email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: G 300
In sentence comprehension, animacy contrasts are used in subject- objectdisambiguation: Sentences with two animate arguments are more difficult to process than sentences with animate subjects and inanimate objects
(Weckerly & Kutas 1999, Frisch & Schlesewsky 2001, Grewe et al. 2007). In
German, this animacy effect is modulated by verbal case-marking pattern:
Animacy effects are weaker for verbs assigning structural case (nomacc)
than for verbs assigning lexical case (nomda) (Czypionka 2014). However, it is still unclear if this modulation reflects the nonstandard syntactic
structure or nonstandard argument semantics of the nom-dat verbs (Blume
2000, Meinunger 2007, Grimm 2010). Another complication is that comprehension experiments on nomda verbs so far have used a mix of different verbs in their stimuli: simple verbs and particle verbs. Simple nom-dat
119
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 120 — #132
AG 2: e syntax of argument structure
AG2
verbs are assumed to have more complex syntactic structures than simple nomacc verbs (Bayer et al. 2001, but see Fanselow 2000), while particle verbs are more complex than the standard structure with both casemarking patterns. The nonstandard argument semantics are the same for
both simple and particle nomda verbs. We present the results of two ERP
studies on German sentence comprehension, comparing the interplay between the processing of argument animacy and case-marking for both simple and particle verbs. Simple accusative verbs show an effect of object animacy, reflected in negative deflections at right-anterior sites, and positive
deflections at left-posterior sites. Simple dative verbs did not show effects
of object animacy. Particle verbs, however, show only animacy effects, and
no modulations or main effects of caused by verbal case marking pattern.
Our findings suggest that the modulation of the object animacy effect for
simple nomda verbs reflects the build-up of a more complex syntactical
structure, in line with predictions from syntactic theory. We assume that
the semantic difference between nomacc and nomda verbs does not
contribute crucially to the case effects found for simple verbs – otherwise,
there should have been case effects and interactions of case and animacy
for particle verbs, too. Our findings also support syntactic accounts assuming more complex syntactic structures for simple nomda than simple
nomacc verbs.
The issue of lexical guidance in sentence production:
Evidence from structural priming experiments
Sandra Pappert1 , Michael Baumann1 & Thomas Pechmann2
1
Universität Bielefeld, 2 Universität Leipzig
[email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: G 300
Psycholinguistic accounts of sentence production differ in the role they
attribute to lexically represented information as, e.g., the argument structure of verbs. According to the hypothesis of strict incrementality, conceptual factors are the main determinants of linguistic encoding. In contrast,
120
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 121 — #133
AG 2, Raum G 300
lexicalist approaches claim that argument structure information triggers
the mapping of thematic roles onto syntactic functions. In a series of experiments that combined a structural priming manipulation with a sentence generation task, the impact of argument structure on formulation
was tested in German. Dative alternation (DA) primes (e.g., Der Rechtsanwalt schickte seinem Klienten den Vertrag / den Vertrag an seinen Klienten,
‘The lawyer sent his client the contract / the contract to his client’) were
found to prime the syntactic realisation of benefactive alternation (BA) targets (e.g., Die Sekretärin backte ihrem Chef einen Kuchen / einen Kuchen ür
ihren Chef, ‘The secretary baked a cake for her boss / her boss a cake.’)
and vice versa (Pappert & Pechmann 2013). The observation of priming
across argument structures disfavours a lexical account of structural persistence (Pickering & Branigan 1998). A more recent experiment combined
DA primes and BA primes (the latter now without reference to caused
possession, e.g., Der Schüler wischt dem Lehrer die Tafel / die Tafel ür den
Lehrer, ‘The pupil wipes *the teacher the blackboard / the blackboard for the
teacher’) with DA targets. There was significant priming, but the effect was
not modulated by the alternation type (DA vs. BA), either. Thus, there was
no evidence that differences in semantics or in the syntactic configuration
are an issue. The found effects most probably arise during conceptualisation or phrase structural realisation. Even though there is little evidence
for a necessary lexical involvement, DA priming is boosted by verb repetition (Pickering & Branigan 1998; Chang et al. 2015). Future experiments
will show whether this also holds for adjunct priming.
References: • Chang, Franklin & Baumann, Michael & Pappert, Sandra & Fitz, Hartmut. 2015.
Do lemmas speak German? A verb position effect in German structural priming. Cognitive
Science 39 (5). 1113–1130. • Pappert, Sandra & Pechmann, Thomas. 2013. Bidirectional structural priming across alternations: Evidence from the generation of dative and benefactive alternation structures in German. Language and Cognitive Processes 28 (9). 1303–1322. • Pickering, Martin J. & Branigan, Holly P. 1998. The representation of verbs: Evidence from syntactic
priming in language production. Journal of Memory and Language 39 (4). 633–651.
121
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 122 — #134
AG 2: e syntax of argument structure
Animacy effects in German sentence production
Sabine Reuters1 , Dr. Sarah Verlage1 & Prof. Dr. Martina Penke1
1
Universität zu Köln
[email protected], [email protected], [email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: G 300
AG2
Linguistic communication often takes place in contexts in which speakers
talk about events and entities they currently observe in the extralinguistic
visual world. In order to faciliate communication and convey an utterance
appropriately, speakers have to choose between diverse syntactic alternatives (Myachykov 2010: 53). The picture depicted below, for instance, can
be described by formulating a German active like Der Teufel trägt den Sack
(e devil is carrying the sack), but one can also choose a passive sentence
such as Der Sack wird von dem Teufel getragen (e sack is being carried by
the devil) or a topicalization like Den Sack trägt der Teufel (e sack [ACC]
is carrying the devil [NOM]) to describe the scenario adequately. In recent
decades, numerous linguistic research paradigms have proven that this arrangement of words in sentence production is by no means arbitrary, but
reflects the interaction of language and cognition in form of a linguistic
message (Bock 1982; Jackendoff 2002). More concretely, diverse crosslinguistic studies have shown that concepts which are placed higher on the
animacy hierarchy scale are chosen as sentential subject or in an earlier
clause position (Bock & Warren 1985; Bock et al. 1992; Prat Sala 1997; Van
Nice & Dietrich 2003).
We are going to present a study which was designed to find out how animacy determines the selection of a specific syntactic structure in – to this
122
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 123 — #135
AG 2, Raum G 300
date under-researched – German sentence production. We were specifically interested in the impact of patient animacy on the choice of a particular syntactic structure. For this purpose, we conducted a sentence production experiment with monolingual, unimpaired German participants who
were asked to describe simple black-and-white drawings depicting interactions between either i) an animate agent and an animate patient or ii)
an animate agent and an inanimate patient and which were designed to
elicit simple transitive sentences including an action verb in the form of an
active or passive clause or a topicalization. We predicted that pictures with
animate patients would lead to a significantly higher number of marked
passive and topicalized structures as well as to longer reaction times between stimulus presentation and speech onset. The results of our study
show that psycholinguistic experiments like ours can shed light on the nature of the relationship between conceptual factors and syntactic choices.
Alternate speakers
Speakers’ judgements on English unaccusativity diagnostics
James Baker, University of Cambridge
[email protected]
Assessing agentivity and eventivity in object-experiencer verbs:
the role of processing
Jeannique Darby, Universität Stuttgart (SFB 732)/Humboldt-Universität zu
Berlin
[email protected]
Groups of object experiencer (ObjExp) verbs in German —
empirically revisited
Nils Hirsch, Humboldt-Universität zu Berlin
[email protected]
123
AG2
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 124 — #136
AG 2: e syntax of argument structure
What is competence in txting? A corpus-based analysis of
Swiss Fren Text messages
Aurélia Robert-Tissot, University of Zurich
[email protected]
AG2
124
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 125 — #137
Arbeitsgruppe 3
Agentivity and event structure: Theoretical and
experimental approaches
Beatrice Primus1 , Markus Philipp1
1
University of Cologne
AG3
[email protected], [email protected]
Raum: F 426
Workshop description
This DGfS-AG aims to bring together theoretical and experimental accounts
of different concepts of agentivity and their influence on event interpretation. Events are usually identified by their spatio-temporal properties.
However, the way participants are engaged in an event also plays a major role (e.g. Eckardt 2002). While the impact of the semantic property of
an incremental theme (or patient) on event interpretation has been widely
discussed (following e.g. Krifka 1998), the influence of the (possibly contextually driven) semantic properties of the agent or causer role on event interpretation is still understudied, particularly from an experimental perspective. We also have a poor understanding of the impact of different agentivity features (or entailments, e.g Dowty 1991) on event interpretation. Additionally, a related open question is to what extent event interpretation
is semantically determined and/or pragmatically driven. Presentations on
the following and related topics may contribute towards a clarification of
the relationship between agentivity and event structure in this AG:
• features of agentivity (volitionality, causation, etc.) in event interpretation
125
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 126 — #138
AG 3: Agentivity and event structure
• agentivity and point-of-view aspect (e.g. the progressive)
• agentivity and non-culminating events
• agent-oriented adverbial modification
• agentivity and event interpretation in nominalizations
AG3
A primary focus lies on accounts that try to integrate experimental data
and theoretical considerations with respect to agentivity in event interpretation. The AG will serve as a discussion forum for researchers from
different fields such as theoretical syntax and semantics, psycho- and neurolinguistics at the syntax-semantics interface and experimental pragmatics.
References: • Eckardt, Regine. 2002. Event semantics. In: Fritz Hamm & T. Ede Zimmermann
(eds.). Semantics. Linguistische Berichte, Special Issue 10, 91-128. • Dowty, David R. 1991. Thematic proto-roles and argument selection. Language 67, 547-619. • Krifka, Manfred. 1998. The
origins of telicity. In: Susan Rothstein (ed.). Events and grammar. Dordrecht: Kluwer, 197–235
Welcome/Introduction
Beatrice Primus1 , Markus Philipp1
1
University of Cologne
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: F 426
126
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 127 — #139
AG 3, Raum F 426
The perceptual and sociocultural foundations of agentivity
in language
Simon Kasper
Philipps University of Marburg
kaspers@staff.uni-marburg.de
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: F 426
In the presentation I will develop the notion of agent as part of the theory
of Instruction Grammar (Kasper 2014, 2015). In Instruction Grammar the
agent category is doubly grounded in (a) perception and conceptualization,
and (b) in sociocultural habits of causal attribution. (a) Acknowledging the
close ties between language and action/perception (cf. Pecher & Zwaan
2005) I assume that an utterance is an instruction for conceptualization
which, in turn, can be characterized as simulated perception. A basic syntactic construction can be described as diagrammatically iconic vis-à-vis
the structure of an event concept. That means the sequential structure of
an event (e.g. Peter kicking a ball into a pond) is retinotopically reflected
in the topological structure of percepts and/or concepts and is then preferentially expressed syntactically in a diagrammatically iconic way (Peter
kicked the ball into the pond). As a result, the sequential (and causal) structure of the real-life event is preserved in the syntactic construction. Thus,
the agent – as a causer – is grounded in the way humans perceive and conceptualize events. (b) Taking the simulation rationale seriously, however,
raises a problem: The question of whether some activity is to be considered
“intentional” or “accidenta”l is highly significant for communication and
the coordination of our verbal and non-verbal sociocultural praxis. But the
information of whether an object of perception or conceptualization (Peter) acted intentionally or accidentally (kick the ball) can be argued to be
neither perceptual nor conceptual information. Philosophical and social
psychological action theories suggest that we should not look for intentions in the structures of perception and conceptualization but rather in
acquired attribution habits, i.e., the socioculturally variable criteria under
which members of a (speech) community explain “observed behavior in
order to arrive at a decision regarding the reason or cause for the behavior
127
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 128 — #140
AG 3: Agentivity and event structure
[…].” (Moskowitz 2005: 234). Thus, the agent is – as a responsible actor –
also grounded in variable sociocultural praxes of attribution which must
be imposed on the spatial perceptual/conceptual layouts of events.
References: • Kasper, S. 2014. Herleitung einer Instruktionsgrammatik. Zeitschri ür Germanistische Linguistik, 42-2, 253–306. • Kasper, S. 2015. Instruction Grammar. From Perception
via Grammar to Action. Berlin/ Boston: de Gruyter. • Moskowitz, G. B. 2005. Social cognition.
Understanding self and others. London/New York: Guilford Press.
AG3
Actor prototypicality in the comprehension of intransitive
clauses in German
Franziska Kretzschmar1 , Svenja Lüll1 , Phillip Alday2 ,
Ina Bornkessel-Schlesewsky2 & Matthias Schlesewsky2
1
Gutenberg University, Mainz, Germany, 2 University of South Australia,
Adelaide, Australia
[email protected]
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: F 426
Recent work in both theoretical-typological and experimental linguistics
has highlighted the actor-centredness of language (e.g., Riesberg & Primus
2015, Bornkessel-Schlesewsky & Schlesewsky 2014). The evidence stems
mostly from investigations on the role of noun semantics, particularly with
respect to animacy asymmetries, and the role of (special types o) monoand dianiie verbs. Little, however, is known about agency effects in
the comprehension of inaniie active clauses (but see Primus 2011
for passives). The present study specifically investigated the influence of
animacy and two groups of manner-of-motion verbs that either preferred
animate or inanimate actors (Levin & Rappaport 1992). This allowed us to
test the interplay of animacy and the two verb entailments volition and
autonomous motion, both of which are defining features for agency and,
hence, the prototypical actor concept (Dowty 1991; Primus 2006). To this
end, we fully crossed actor animacy and the two groups of manner-ofmotion verbs (cf. 1). In addition, we varied word order which enabled us to
128
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 129 — #141
AG 3, Raum F 426
investigate whether the point in time at which types of agency information
become available might influence sentence processing (1, 2a,b).
(1)
(2)
Ein Tourist/Frachter schwimmt/driet über den See, während …
‘A tourist/freighter swims/drifts over the lake, while …’
a.
b.
Es schwimmt/driet ein Tourist/Frachter über den See, während
…
‘There swims/drifts a tourist/freighter over the lake, while …’
Über den See schwimmt/driet ein Tourist/Frachter, während …
‘Over the lake swims/drifts a tourist/freighter, while …’
Experiment 1 monitored readers’ eye movements and found that mismatches between verb preference and noun animacy increased reading times at
the clause-final region containing a prepositional phrase (1, 2a), but not
with a clause-final actor (2b). Experiment 2 measured event-related potentials (ERPs) in response to the same postverbal actors (2a,b). We found
enhanced N400 amplitudes when the verb and noun animacy indicated a
non-prototypical actor. These effects were mainly driven by inanimate actors. These findings support the view that animacy is a strong indicator
of actor prototypicality, and extend its scope to inaniie clauses. Intriguingly, because animacy is more strongly related to volition, it thereby
trumps the autonomous motion feature provided by verb semantics. This
is the first demonstration that noun-based information is not only independent of but may also be more important than verb-based information
in shaping the prototypical actor concept.
References: • Bornkessel-Schlesewsky, I. & Schlesewsky, M. 2014. Competition in argument
interpretation: Evidence from the neurobiology of language. In MacWhinney, B., Malchukov,
A., & Moravcsik, E. (eds.), Competing motivations in grammar and usage, Oxford: Oxford University Press., 107-126. • Dowty, D. 1991. Thematic proto-roles and argument selection. Language 67. 547-619. • Levin, B. & Rappaport Hovav, M. 1992. The lexical semantics of verbs of
motion: The perspective from unaccusativity. In Roca, I. M. (ed.), ematic structure: its role
in grammar, Berlin: Walter de Gruyter, 247-269. • Primus, B. 2006. Mismatches in semantic-role hierarchies and the dimensions of role semantics. In Bornkessel, I., Schlesewsky, M., &
Comrie, B. (eds.), Semantic role universals and argument linking: theoretical, typological and
129
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 130 — #142
AG 3: Agentivity and event structure
psycholinguistic approaches, Berlin: Mouton de Gruyter, 53-87. • Primus, B. 2011. Animacy
and telicity: Semantic constraints on impersonal passives. Lingua 121. 80-99. • Riesberg, S. &
Primus, B. 2015. Agent prominence in symmetrical voice languages. STUF – Language Typology and Universals 68.4, 551–564.
Agentivity, control and semantic structure in Russian
Causatives
Valentina Apresjan
Higher School of Economics Moscow
AG3
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: F 426
The paper examines the correlation between the degree of agentivity in
Russian causatives and their semantic structure. It supports the view that
the distinction between agentivity vs. non-agentivity in causatives is an
important feature with various ramifications, and motivates the role of the
‘control’ feature in the event structure of the causatives.
Russian causatives differ with respect to their interaction with negation, which can affect either both elements in their event structure – ‘causing situation’ and ‘caused situation’ (wide scope) or only the second element – ‘causing situation’ (narrow scope). Generally, Russian non-agentive
causatives admit narrow scope negation, whereas agentive causatives favor wide scope negation, which is explained by the status of the component
‘causing situation’ in the semantic structures of the causatives. If it is in the
assertion, it gets negated; if it is in the presupposition, it projects. However,
why is ‘causing situation’ usually asserted in agentive causatives, but presupposed in non-agentive causatives? The hypothesis is that the status of
this component in causatives is contingent upon the degree of control that
the agent exercises over the ‘caused situation’: the more control, the more
likely the causing event will be asserted.
This hypothesis is confirmed by the data from Russian emotional causatives, which are largely non-agentive and presuppose the causing event.
However, certain emotional causatives, mostly denoting the causation of
130
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 131 — #143
AG 3, Raum F 426
fear, anger and amusement are agentive and involve intentionality and
control over the outcome. They allow wide scope reading:
(1)
On menja bol’še ne pugaet
He I.ACC more not scare.3SG
‘He doesn’t scare me anymore = [doesn’t make attempts to scare me
anymore]’
As corpus data show, the hierarchy of control in Russian emotional causatives, as manifested in their ability to be used in the imperative, corresponds to the frequency of their wide scope reading: the more control, the
more possibility of a wide scope reading and thus of the asserted status
of the ‘causing situation’ component. The logic is as follows: the greater is
the agent’s control over the result, the closer are the causal relations between the components ‘causing situation’ and the ‘caused situation’: no
result means no causing situation. Thus, those two components belong to
the same level of semantic structure – assertion. The lower is the agent’s
control over the result, the weaker are these relations: thus, the absence of
the result does not mean the absence of the causing situation. Thus, these
two components belong to different levels of semantic representation –
presupposition (‘causing situation’) and assertion (‘caused situation’).
On atypical agents
Fabienne Martin
University of Stuttgart
[email protected]
Mittwoch, 24. 02. 2016, 17:00–18:00, Raum: F 426
Natural languages offer many distinctive ways to convey weak, non prototypical agentivity on the part of a (human) subject. However, defining
agents that are not ‘full’ or ‘prototypical’ agents and differentiating them
from ‘pure’ (inanimate) causers is not always a trivial task, complicated by
the fact that what ‘weak agentivity’ exactly means differs from one language to another.
131
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 132 — #144
AG 3: Agentivity and event structure
AG3
The first goal of this paper is to offer a typology of agents that allows to
more precisely differentiate subtypes of atypical agents across languages,
on the basis of Searle’s distinction between prior intention and intention
in action. It rests on two hypotheses. The first one is that prior intention,
intention-in-action and (agent) control are three independent ingredients
of (strong/full) agentivity. The second hypothesis is that (agent) control
can be defined through the distribution of sine qua non conditions for the
realization of the event among the different participants to this event: given
a verb denoting an event property θ, we will say that an agent exerts a
full control on his action a if he fullfils the sufficient conditions for the
development of a in a θ event. An agent without full control on her action
a fullfils conditions which are causally necessary but causally insufficient
for a to develop into a θ-event.
The second goal of this paper is to shows how atypical agency affects
what Demirdache and Martin 2015 calls ‘zero-change of state’ non-culminating reading (see their ‘agent control hypothesis’). Our cross-linguistic
data suggest that among atypical agents, only one of the three subtypes
generally licences this non-culminating reading.
Agentivity and impersonal passives
Tim Graf1 , Markus Philipp1 & Beatrice Primus1
1
University of Cologne
[email protected]
Mittwoch, 24. 02. 2016, 18:00–18:30, Raum: F 426
In impersonal passives, the implicit argument is assumed to be restricted
to a volitional agent, and hence a human or at least a higher animate entity
in a number of languages including for example Dutch and German (e.g.
Siewierska 1984; Rapp 1997; Zifonun et al. 1997). However, corpus data
and a preliminary acceptability judgement test (Primus 2011) revealed that
this restriction might be too strong for German. Furthermore, in previous
research voice alternation phenomena like passives were tightly linked to
the assumption of hierarchies of grammatical functions and monolithic semantic roles (cf. Levin & Rappaport-Hovav 2005 for an overview).
132
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 133 — #145
AG 3, Raum F 426
However, these accounts cannot capture the acceptability cline found in
preliminary data (Primus 2011) why we assume that a multi-dimensional
account of semantic roles might be better suited to capture the data in terms
of an agentivity cline. To this end, we will present data of an acceptability
rating study and an ERP experiment testing impersonal passives by systematically varying proto-agent features – volition/control, motion and sentience – that are entailed by the verbal predicate (see table below). Following Dowty (1991) and Primus (1999) we discuss the hypothesis whether a
cluster concept of agentivity can explain the data more adequately. Finally,
the observed data pattern gives rise to theoretically explore the possibility
of decomposing agentivity features into finer grained bundles.
Verbclass
Example
controlled activity, human
Es wurde gearbeitet, weil…
‘there was working, because’
uncontrolled process, human
Es wurde geschwitzt, obwohl…
‘there was sweating, although’
uncontr. emotional state, human
Es wurde gebangt, weil…
‘there was fearing, because’
uncontrolled state, human
Es wurde geglänzt, obwohl…
‘there was glittering, although’
References: • Dowty, D. 1991. Thematic proto-roles and argument selection. Language 67.
547-619. • Levin, B. & Rappaport Hovav, M. 2005. Argument realization. Cambridge. • Primus,
B. 1999. Cases and thematic roles. Ergative, accusative and active. Tübingen Niemeyer. • Primus,
B. 2011. Animacy and telicity: Semantic constraints on impersonal passives. Lingua 121-1,
80–99. • Rapp, I. 1997. Partizipien und semantische Struktur. Tübingen Stauffenburg. • Siewierska, A. 1984. e passive. A comparative linguistic analysis. London: Croom Helm. • Zifonun,
G., L. Hoffmann, B. Strecker, J. Ballweg & U. Brausse. 1997. Grammatik der deutschen Sprache.
Berlin de Gruyter.
133
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 134 — #146
AG 3: Agentivity and event structure
Agentivity and eventivity in psych nominalizations
Gianina Iordachioaia
University of Stuttgart
[email protected]
Donnerstag, 25. 02. 2016, 09:00–09:30, Raum: F 426
AG3
Since Lako’s (1966) description of stative adjectives and verbs, agentivity
tests have been used to negatively diagnose stativity. Thus, agentivity implies eventivity/ non-stativity. In this paper, I explore one consequence of
this correlation, which explains the long-observed ’agent exclusivity effect’
in nominalizations from psych verbs illustrated in (1).
(1)
a.
b.
The clown/the situation humiliated the audience.
the clown’s/*the situation’s humiliation of the audience
Assuming a word formation model in which the ontological type of the root
interacts with the event template in which it appears (Rappaport Hovav
& Levin 1998, Alexiadou et al 2015), I argue that psych roots are stative
and psych nominals as in (1b) are ambiguous between eventive and stative
readings. I will show that the stative reading, which excludes agents, is
derived from the root, while the eventive reading can be imposed on the
stative root only by the presence of an agent. Agentivity thus forces an
eventive structure that is otherwise absent in psych nominals. To support
this analysis, I will compare the data in (1) to non-psych nominals, which do
allow non-agentive causers (the war in (2)), but only with eventive readings
(see Alexiadou et al 2013 for aspectual tests). I will argue that the contrast
between (1b) and (2) with the war lies in the ontological difference between
psych and non-psych roots, as well as the reduced verbal template structure
of nominalizations.
(2)
the teacher’s/the war’s/*adultery’s separation of Jim and Mary
The study shows that, although agents share properties with some nonagentive causers in terms of eventivity, agentivity is special in contributing
eventive structure that is not available with non-agentive causers.
134
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 135 — #147
AG 3, Raum F 426
References: • Alexiadou Artemis & Anagnostopoulou Elena & Florian Schäfer. 2015. External
arguments in transitivity alternations. Oxford: Oxford University Press. • Alexiadou Artemis
& Gianina Iordachioaia & Mariangeles Cano & Fabienne Martin & Florian Schäfer. 2013. The
realization of external arguments in nominalizations. Journal of Comparative Germanic Linguistics 16:2. 73-95. • Lakoff, George. 1966. Stative adjectives and verbs in English. In A.G. Oettinger (ed.), Mathematical Linguistics and Automatic Translation, Report NSF-17, 1-16. Cambridge, Mass.: The Computation Laboratory, Harvard University. • Rappaport Hovav Malka
& Beth Levin. 1998. Building verb meanings. In M. Butt, W. Geuder (eds.), e projection of
arguments: Lexical and compositional factors, 97-134. Stanford: CSLI Publications.
The situation of aspect from the viewpoint of language
pathology: A comparison between stroke induced aphasia
and semantic dementia
Vasiliki Koukoulioti1 & Stavroula Stavrakaki2
Goethe University of Frankfurt, 2 University of Thessaloniki
1
[email protected]
Donnerstag, 25. 02. 2016, 09:30–10:00, Raum: F 426
Situation aspect refers to the inherent temporal features of the event described by the verb (Vendler 1967), whereas viewpoint aspect refers to the
way an event is viewed and described by a speaker (Smith 1997). Smith
(1997) distinguishes between unmarked and marked combinations of the
two aspects, in the former situation and viewpoint aspect are in agreement,
in the latter there is a clash. This clash is resolved in unimpaired language
by means of a situation type shift with the viewpoint aspect overriding the
situation aspect.
There is evidence of markedness effect in aphasia. Russian patients with
aphasia after stroke (henceforth AaS) have difficulties producing marked
combinations in comparison to unmarked (Bastiaanse & Platonov 2015). In
the present study we explore this effect in Modern Greek comparing two
populations: AaS and patients with semantic dementia (SD). The participants performed a sentence completion task.
135
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 136 — #148
AG 3: Agentivity and event structure
AaS participants showed a markedness effect, whereas SD patients manifested such an effect only in imperfective aspect. We suggest that in AaS
the situation aspect overrides the viewpoint aspect, rather than the other
way round, due to grammatical deficits. The selective impairment in imperfective aspect in SD is related to the semantic deficit, which affects the
encoding of dynamic situations, whereas it leaves core components of verb
semantics intact.
References: • Bastiaanse, R. & Platonov, A. 2015. Argument structure and time reference in
agrammatic aphasia. In: R. G. de Almeida & C. Manouilidou (eds.). Cognitive Science Perspec-
AG3
tives on Verb Representation and Processing. Cham, Heidelberg, New York, Dordrecht, London:
Springer, 141–156. • Smith, C. S. 1997. The parameter of aspect (2nd ed.) (Studies in linguistics
and philosophy: Vol. 43). Dordrecht, London: Kluwer Academic. • Vendler, Z. 1967. Verbs and
Times. In Linguistics in philosophy, 97–121. London: Cornell Univ. Press.
Psych verbs as indicators of perspective taking
Stefan Hinterwimmer
University of Cologne
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: F 426
Sentences in narrative texts can sometimes be understood as expressing
the thoughts or feelings of salient protagonists different from the narrator
even though they are not the complements of propositional attitude verbs
whose subject is a noun phrase referring to the respective protagonist. This
phenomenon is known as Free Indirect Discourse (FID), and according to a
prominent line of analysis in formal semantics it involves the introduction
of a fictional context whose author is the protagonists to whom the respective thought is ascribed (Doron 1991, Schlenker 2004, Sharvit 2008, Eckardt
2014). In this talk I will take a close look at the role of psych verbs in licensing the introduction of such fictional contexts, and their interaction with
the larger context.
Consider the contrast between (1a) and (1b): While (1a) can easily be
understood as expressing a thought of Mary about John, (1b) sounds weird
136
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 137 — #149
AG 3, Raum F 426
in the context provided by (1). This is due to a conflict: On the one hand,
(1b) cannot plausibly be ascribed to Mary, but only to George. On the other
hand, only Mary has been made available as a perspective holder by (1), but
not George. Intuitively, this contrast is due to George’s being the stimulus
or theme of the state introduced by the verb hate in (1) and Mary’s being
the experiencer.
(1)
Mary hated George.
a. The dumb jerk always thought he knew everything better!
b. ⁇ The mean old bat always tried to make him look like an idiot!
With the additional context provided by (2), in contrast, (2b), which is identical to (1b), can quite naturally be understood as expressing a thought of
George about Mary. Intuitively, this is due to George’s functioning as the
discourse topic of the entire mini-text in virtue of having been introduced
text-initially. At the same time, (2a), which is identical to (1a), can still be
understood as expressing a thought of Mary about George.
(2)
George entered the restaurant. Mary was sitting at a table in the
corner with her best friend. Mary hated George.
a. The dumb jerk always thought he knew everything better!
b. The mean old bat always tried to make him look like an idiot!
In the talk I will propose an analysis which accounts for these and similar
observations by ascribing a privileged status to the experiencers of psych
verbs with respect to perspective taking insofar as they can (a) serve as the
authors of fictional contexts even in the absence of any further context,
and (b) remain available as perspective holders in the presence of globally
more prominent protagonists.
137
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 138 — #150
AG 3: Agentivity and event structure
Agentive dispositions and causal responsibility: A case
study
Marta Donazzan1 & Lucia M. Tovena2
1
University of Cologne, 2 Université Paris 7 – Denis Diderot
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: F 426
AG3
It has been noted in the literature that unaccusative predicates such as fall,
whose subject is not an agent, can be used in causative light-verb constructions requiring an agentive subject. However, when the predicate enters a
complex predication, as (1a,b) in English (Wierbizca 1982) and (2a,b) in Italian, the light-verb only accepts animate referents as subjects.
(1)
a.
b.
Mary/the apple fall from the tree.
Mary/*the apple had a fall.
(2)
a.
b.
Maria/la mela è caduta dall’albero
Maria/*la mela ha fatto una caduta.
In this talk, we will focus more specifically on Italian complex predicates
such as (2), whose argument is a deverbal noun. Event nouns ending by
[A]ta in Italian, such as caduta (fall) in (2), are deverbal nominalisations
where an inflectional suffix possibly is at the origin of the word-formation
pattern. In present day Italian, the suffix has become a productive derivational morpheme, which nevertheless carries semantic information related
to its original inflectional function. [A]ta nouns have been described as denoting discretised (Acquaviva 2005) and particularised events (Gaeta 2000)
and have been analysed in the context of complex predications (SamekLudovici 2003, Folli & Harley 2013). Part of these specific properties, we
claim, are due to the fact that [A]ta-nouns, despite the fact that they are
nominalisations and therefore have no syntactic external argument, have
the peculiarity of denoting events with a semantically active initiator. The
event noun imposes some specific constraints that show in light-verb constructions when the initiator is realized and must coincide with the external
argument of the light-verb. In this talk, we will show that these constraints
138
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 139 — #151
AG 3, Raum F 426
are at the origin of contrasts such as (2a) vs.(2b), which does not depend
on animacy, but rather on the possibility for the referent of the subject to
be causally responsible of the event denoted by the nominalised predicate.
We will propose that causal responsibility can be made dependent on the
properties of the referent of the subject, which we define as agentive dispositions towards an event (Donazzan & Tovena to appear). Dispositions
are properties ascribed to an entity that are perceived in the perspective of
a manifestation. Assume the association of a causal chain to an event-type.
Talking of agentive dispositions is a way of telescoping two pieces of information. On the one hand, agentive dispositions are properties ascribed
to an entity and, on the other hand, those properties are seen as the first element of a causal chain leading to a class of events. Disposition ascription
is a form of bridging between an entity and a class of event, which does
not carry any existential commitment on the instantiation of the class of
events.
Agents, effectors and event structure
Robert D. van Valin, Jr.
Heinrich Heine University Düsseldorf
[email protected]
Donnerstag, 25. 02. 2016, 12:00–13:00, Raum: F 426
Among the various thematic relations that have been posited, the notion
of ‘agent’ stands out as one of the most important and the most problematic. It is important due to its role as the default choice for subject in most
languages and its role in the syntax, e.g. in some languages the antecedent
of a reflexive must be an agent-like argument. It is problematic in that it
seems to be an inherent property of some verbs, e.g. murder, but an optional property of others, e.g. kill, which can take agent-canceling adverbials (accidentally, inadvertently) or inanimate subjects (e.g. A falling tree
branch killed John’s dog). From a cross-linguistic perspective, it is striking
that such variation is not found in all languages: in Japanese, for example,
aniie verbs disallow inanimate subjects and expressions that explicit
139
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 140 — #152
AG 3: Agentivity and event structure
AG3
contradict an agentive interpretation (Hasegawa 1996). Moreover, it is not
altogether clear how agentivity relates to event structure; for example, is
it the case that any event type can have an agentive argument?
This paper will argue for the following points:
1. In many languages, agentivity is best analyzed as being an implicature involving human or high animate arguments, rather than an inherent
lexical property of most verbs; this idea comes from the pioneering work
of Holisky (1987). This does not rule out there being inherently agentive
verbs in such languages, but they are the exception rather than the rule.
2. In some languages, agentivity is a lexical property of many verbs, and
the kind of variable interpretations found in the first type of language is
not found in them. What is exceptional in first group of languages is the
norm in them.
3. The basic notion operative in both kinds of language is ‘effector’, the
doer of the action; this notion is neutral with respect to agentivity and
animacy (Van Valin & Wilkins 1996). All agents are also effectors; agentivity is an overlay over the more basic effector role, which can be due to
implicature or lexical properties of the predicate.
4. Effector arguments are always associated with activity predicates (Van
Valin & Wilkins 1996), and therefore only those event types which have an
activity component can potentially have an agent argument. Of the twelve
event types proposed in Van Valin (2005, 2015), this rules out state and
change of state predicates (states, achievements and (process) accomplishments) from involving agent arguments.
References: • Hasegawa, Y. 1996. A study of Japanese clause linkage: the connective TE in
Japanese. Stanford: CSLI. • Holisky, D.A. 1987. The case of the inaniie subject in Tsova-Tush (Batsbi). Lingua 71.103-132. • Van Valin, R. 2005. Exploring the syntax-semantics interface. Cambridge: CUP. • Van Valin, R. 2015. Some issues involving (active) accomplishments.
To appear in the Proceedings of 2013 RRG Conference, Univ. of Freiburg. • Van Valin, R. & D.
Wilkins. 1996. The case for ‘effector’: Case roles, agents and agency revisited. In M. Shibatani
& S. Thompson, eds., Grammatical constructions, 289-322. Oxford: Oxford University Press.
140
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 141 — #153
AG 3, Raum F 426
Agentivity in nominalizations of phrasal verbs. On
passers-by and winder-uppers
Anke Lensch
Johannes Gutenberg-University Mainz
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: F 426
The agentive -er suffix is one of the most frequent derivational affixes in
English (Baayen & Lieber 1991). When attached to phrasal verbs, it produces highly interesting and theoretically challenging variants and this
word- formation pattern is not restricted to verbs proper but also extends to
phrasal verbs. There are two types of agentive nominalizations of phrasal
verbs with -er that need to be distinguished. The first type, example (1),
involves derivational marking of the phrasal verb exclusively on the verb.
(1)
“If one day this magazine were to publish a list of the 10 most heroic
runner-ups in sport history …” (e Guardian 2005)
The origins of this type in the Oxford English Dictionary reveal that this
is the older of the two patterns, attested since Middle English Times. The
second type, exemplified in (2) and (3), is characterized by a reduplicative
derivational process (McIntyre 2013:42), where double marking with -er
occurs on both elements, the verb and the particle.
(2)
“He is a great winder-upper. You learn to live with it.” (e Daily
Mail 1998)
(3)
“…served as a crowd warmer-upper, despite failing on her only attempt…” (e Guardian 1995)
Nominalizations of the second type will be shown to be semantically more
agentive and pragmatically more expressive than those belonging to the
first type. Thus, a passer-by happens to pass the scene of an accident or
crime by chance and not intentionally. By contrast, a winder-upper embodies the ability to unnerve other people. Both types share the property of being considered left-headed (cf. McIntyre 2013:42) and of denoting agents.
141
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 142 — #154
AG 3: Agentivity and event structure
Based on a large-scale corpus analysis, this paper argues that differences
in argument structure determine which type of derivational pattern, the
single or the double-marked form, is employed by the language users.
References: • Baayen, H. & Lieber, R. 1991. Productivity and English derivation: A corpus-based
study. Linguistics 29/5: 801-843. • Cappelle, B. 2010. Doubler-upper nouns: A Challenge for usage-based models of language? In: A. Onyster and S. Michel (eds.) Cognitive Perspectives on
Word-Formation. Berlin/New York: Mouton de Gruyter, 335-374. • Chapman, D. 2008. Fixer-uppers and passers-by: Nominalization of verb-particle constructions. In: v. Fitzmaurice,
S. M. & Minkova, D. (eds.) Studies in the History of Language IV. Berlin: Muton de Gruyter.
AG3
265-299. • McIntyre, A. 2013. English particle verbs as complex heads: Evidence from nominalization. In: Härtle, H. (ed.) Interfaces of Morphology. Berlin: Akademie Verlag. 41-57.
Agentivity and force exertion: The German verb “schlagen”
Ekaterina Gabrovska1 & Wilhelm Geuder1
1
University of Düsseldorf
[email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: F 426
In this talk, we investigate the meaning and constructional variation associated with the German verb “schlagen”. We analyse “schlagen” as forming
at least 4 variants that we refer to as the naccaie, the caaie
elaie, the (simple) aniie, and the (simple) oblie constructions. The latter three variants involve an agent argument, and combine it
with patient and/or locative target arguments in different syntactic constellations. One prominent effect that has already been noted in the literature is that direct objects of “schlagen” in the simple aniie construction seem largely confined to animates (e.g. Lundquist & Ramchand 2012).
Hence we find: [aniie] Der Bauer schlug den Esel (’The farmer beat
the donkey’), but not: Der Bauer schlug den Tisch (‘The farmer hit the table’);
rather: [oblie] Der Bauer schlug auf den Tisch (‘The farmer beat on the
table’). The aniie type is also marked by additional idiosyncrasies: (i)
It implies an instrument wielded by the agent, or the use of her/his hands,
142
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 143 — #155
AG 3, Raum F 426
and (ii) only the aniie construction undergoes a number of lexicalisation processes yielding V-NP collocations with specialised meanings.
We argue that the variants listed above differ in semantic complexity,
involving a basic concept that is force-related, but with additional conceptual layers of movement and of agency / intentionality that may be superimposed on it, also implying differences of complexity in the “agent” role
itself. In particular, the superficially simple aniie variant is actually
the semantically most complex one and incorporates all aspects present
in the other variants. This result is based on a fine-grained analysis of the
argument roles (building on Erteschik-Shir & Rapoport 2010, Vogel 2013)
and, mostly, on a corpus study on the patterning of adjectives functioning
as modifiers with “schlagen”. For instance, while the modifier “grausam”
(‘cruelly’) exclusively occurs with the aniie construction in our data
set, all modifiers found with the naccaie construction may recur
in the other types. Our modelling of the results will use ongoing work in
Frame theory, based on Goldschmidt et al. (2015).
References: • Erteschik-Shir, N. & Rapoport, T. 2010. Contact and Other Results. In: Rappaport-Hovav, M., Doron, E., Sichel, I. (Eds.). Lexical Semantics, Syntax, and Event Structure.
Oxford University Press, Oxford, 59-75. • Goldschmidt, A., Gabrovska, E., Gamerschlag, T. &
Petersen, W. 2015. Does the rain hit the window playfully? A frame-based analysis of German
hit-verbs. Paper presented at the 11th TbiLCC, September 21-26, Tbilisi, Georgia (to appear in
the proceedings). • Lundquist, B. & Ramchand, G. 2012. Contact, animacy, and affectedness
in Germanic. In: Ackema, P., Alcorn, R., and Heycock, C. (Eds.), Comparative Germanic Syntax: e state of the art. Philadelphia, PA, USA: John Benjamins Publishing Company, 223-248.
• Vogel, R. 2013 (in press.). Optimal Constructions. In: G. Legendre, M. Putnam, & E. Zaroukian
(Eds.). Studies in eoretical Linguistics. Advances in Optimality theoretic-syntax and semantics
Oxford: Oxford University Press.
143
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 144 — #156
AG 3: Agentivity and event structure
Testing the agent control hypothesis with non-culminating
events. Experimental evidence from adult Dutch and
Mandarin
Hamida Demirdache1 , Angeliek van Hout2 , Jinhong Liu3 ,
Fabienne Martin4 & Iris M. Strangmann2
1
Université de Nantes, 2 University of Groningen, 3 Guangzhou College
South China, 4 University of Stuttgart
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: F 426
AG3
We present empirical evidence from a comprehension study with native
speakers of Dutch and Mandarin Chinese for Demirdache & Martin’s (henceforth DM) Agent Control Hypothesis (ACH). According to the (weak version of the) ACH, it is easier to deny the whole Change of State (CoS) encoded by/ conventionally associated with a predicate when the subject’s
referent is a ‘full’ agent than when it is an (inanimate) causer. DM give
a.o. crosslinguistic evidence for the ACH from Mandarin monomorphemic
verbs versus bimorphemic CoS-verbs (V-V compounds). With the former,
but not the latter, the culmination of the described event can be denied.
Crucially, monomorphemic verbs conventionally associated with a CoS allow the denial of the whole CoS (DM’s ‘zero-CoS’ reading), but only when
the subject’s referent is a full Agent, not when it is a Causer. In contrast VV
compounds do not licence the zero-CoS reading with either kind of subject. Ten Dutch and 50 Mandarin native speakers judged sentences given
short movie clips showing events with either a whole CoS, as encoded by
the predicate, or no such CoS at all. The 2x2 design, varying Situation (Full
CoS versus Zero CoS) and Subject type (Agent versus Causer), tested eight
transitive CoS verbs. For Mandarin we tested two paradigms–one with
monomorphemic verbs and the other with V-V compounds–with two different groups of adults. Zero-CoS was rejected categorically in Dutch, as
well as in Mandarin, for V-V compounds. For Mandarin monomorphemic
verbs on the other hand, zero-CoS situations were occasionally accepted
and significantly more often for Agent than for Causer subjects. We take
the absence of an effect of Agent vs. Causer with Dutch particle verbs and
144
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 145 — #157
AG 3, Raum F 426
Mandarin V-V compounds to simply come from the fact that strong markers of telicity, such as Dutch particles or Mandarin V-V compounds (where
the second V encodes the result state), make a zero-CoS reading impossible
to begin with, independently of the nature of the subject. There was, however, an effect of Agent vs. Causer with monomorphemic verbs in Mandarin, thus providing novel experimental support for the ACH. The latter,
however, is sensitive to the grammar of culmination: the effect of agenthood on non-culmination can only show up with verbs and constructions
that are open to non-culminating readings. Participants interpreted (nonmodal) Dutch particle verbs and Mandarin resultative V-V compounds as
requiring culmination, thus confirming that subject type plays no role when
the occurrence of the CoS is lexically entailed. At the same time, the acceptance of zero-CoS situations for Mandarin monomorphemic verbs confirms
the role of agenthood, as predicted by the ACH, with culmination behaving as a cancellable implicature with Agents, but as an entailment with
Causers.
Psych nominalizations in Hebrew
Odelia Ahdout
The Hebrew University of Jerusalem
[email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: F 426
On the basis of contrasts between object experiencer (OE) verbs and their
nominalizations, it has been claimed that derived nominals have restrictions on the realization of the external argument which do not hold in
the corresponding verb (Rappaport, 1983; Grimshaw, 1990). While the verbal clause is usually ambiguous between a stative and a causative reading, the nominals tend to be stative (e article/my kid disappointed me;
my disappointment (*by my kid/the article)). Moreover, the nominal clause,
as opposed to the verbal clause which permits both agents and causers,
is restricted to agents only (e insult/my enemy humiliated me; *e insult’s/my enemy’s humiliation of me). The data on Hebrew, a language with
145
AG3
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 146 — #158
AG 3: Agentivity and event structure
rich verbal morphology, yields two findings: 1. two transitive verbal templates derive eventive Psych nominals (2); 2. Similar to previous findings
on Greek and Romanian (Alexiadou and Iordăchioaia, 2014), in Hebrew,
the restriction to agentive causers is shown to hold only in the transitive
nominalizations; intransitive (subject experiencer) nominalizations allow
causers realized via a causative preposition.
(1)
ha’alavat ha-saxkan al-yedey mevaker ha-tarbut/*ha-bikoret b-a-iton
nominal
‘The cultural critic’s/the newspaper critique’s insulting of the actor’.
(2)
ha-hitaxzevut (ha-xozeret) šel dani me-ha-sofer/me-ha-ben šelo SE nominals
‘Danni’s (repeated) disappointment from the writer/from his son’
AG3
OE
However, not all OE predicates derive eventive nominals: some roots
have either stative nominals, as in English, or none. A possible explanation
for this can be achieved by a closer inspection of the semantics of transitive Psych verbs, following the sub-classification in Pesetsky (1995), who
differentiates between causative verbs and T/SM (target/subject matter)
verbs. In the former, the cause of the mental state can be indirect, possibly
triggering some emotion that is actually directed towards a different object
(the T/SM), while the latter verbs denote a mental state directed directly towards a target. Crucially, the T/SM argument is never agentive, and as the
nominal clause is restricted to agents, T/SM verbs cannot produce eventive
nominals.
References: • Alexiadou, A., & Iordăchioaia, G. 2014. Causative Nominalizations: Implications for the Structure of Psych Verbs. A. Bachrach, I. Roy & L. Stockall (eds.) Structuring
the Argument. 119-140. John Benjamins. • Grimshaw, J. 1990. Argument Structure. MIT Press:
Cambridge, MA. • Pesetsky, D. 1995. Zero Syntax: Experiencers and Cascades. MIT Press. • Rappaport, M. 1983. On the nature of derived nominals. In B. Levin, M. Rappaport & A. Zaenen
(eds.) Papers in Lexical-Functional Grammar Indiana University Linguistics Club, p. 113-444.
146
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 147 — #159
AG 3, Raum F 426
Final discussion
Beatrice Primus1 , Markus Philipp1
1
University of Cologne
[email protected], [email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: F 426
AG3
147
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 148 — #160
Aus unserem Gesamtverzeichnis 2015-2016
Hagen Hirschmann
Nadio Giger
Modifikatoren im Deutschen
Generative Varietätengrammatik
Ihre Klassifizierung und varietätenspezifische
Verwendung
Studien zur deutschen Grammatik, Band 89
2015, 374 Seiten, kart.
ISBN 978-3-95809-540-3 € 64,–
Ulrike Freywald
Parataktische Konjunktionen
Zur Syntax und Pragmatik der Satzverknüpfungen im Deutschen – am Beispiel von obwohl, wobei, während,
wogegen und dass
Studien zur deutschen Grammatik, Band 90
Ende 2015, ca. 350 Seiten, kart.
ISBN 978-3-95809-541-0 € 64,–
am Beispiel der Nominativ-Akkusativ-Variation
im Schweizerhochdeutschen
[Stauffenburg Linguistik, Bd. 86]
2015, 371 Seiten, kart.
ISBN 978-3-95809-507-6 € 49,80
Dieser Band stellt mit seinen Erkenntnissen einen wichtigen Brückenschlag zwischen den Disziplinen der Generativen Grammatik und der Varietätenlinguistik dar und
behandelt folgende Kernfragen: Wie lässt sich Generative Grammatik als Theorie einer Varietätengrammatik
auffassen, die Grammatikkompetenz in Bezug auf die
Varietäten einer Einzelsprache erfasst? Und inwiefern
sind Variationsphänomene in Bezug auf die Kasus Nominativ und Akkusativ im Schweizerhochdeutschen ein
Beleg für eine solche generative Varietätengrammatik?
Das Buch zeigt, dass solche Variationsphänomene auf
unterschiedlicher Parametrisierung spezifischer Merkmale mit einzelvarietär verschiedener Geltung und somit auf Mikroparametrisierung basieren – und dass auch
ihretwegen das Konzept einer generativen Varietätengrammatik gerechtfertigt ist.
Ulrich Wandruszka
Christoph Schwarze / Leonel F. de Alencar
Sprache – linearisierte Struktur in Bewegung
Lexikalisch-funktionale Grammatik
Eine Einführung in die Mechanik der Sprache
auf der Basis der Kategorialgrammatik
Stauffenburg Einführungen, Band 31
2015, 247 Seiten, kart.
ISBN 978-3-95809-412-3 € 29,80
Alexander Lasch / Alexander Ziem (Hrsg.)
Konstruktionsgrammatik IV
Konstruktionen als soziale Konventionen und kognitive Routinen
Stauffenburg Linguistik, Bd. 76
2015, 330 Seiten, kart.
ISBN 978-3-86057-121-7 € 39,80
Eine Einführung am Beispiel des Französischen
mit computerlinguistischer Implementierung
Stauffenburg Einführungen, Band 30
Winter 2015
ISBN 978-3-95809-411-6
Susanne Günthner / Wolfgang Imo / Jörg Bücker (Hrsg.)
Konstruktionsgrammatik V
Konstruktionen im Spannungsfeld von
sequenziellen Mustern, kommunikativen
Gattungen und Textsorten
Stauffenburg Linguistik, Bd. 77
2015, 310 Seiten, kart.
ISBN 978-3-86057-122-4 € 39,80
Stauffenburg Verlag GmbH
Postfach 25 25 D-72015 Tübingen www.stauffenburg.de
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 149 — #161
Arbeitsgruppe 4
Sentence complexity at the boundary of
grammatical theory and processing: A special
challenge for language acquisition
Flavia Adani1 , Tom Fritzsche1 & Theodoros Marinis2
1
AG4
Universität Potsdam, 2 University of Reading
[email protected], [email protected], [email protected]
Raum: G 530
Workshop description
Complex sentences (e.g., relative clauses, wh-questions, passives, clefts,
extractions from weak islands) are considerably challenging for children
who acquire their first or second language but also for adults when they
are tested under time pressure. Attempts to explain the effects of sentence
complexity have developed, at least, along two directions. On one hand,
theoretical linguists have been aiming to clarify the nature of sentence
complexity, how it manifests itself within one language and across different languages and under which conditions the grammaticality of complex sentences is disrupted (e.g., Rizzi 2013). Grammatical theories of sentence complexity have also been used to interpret children’s non adult-like
performance on experimental tasks. On the other hand, psycholinguists
and cognitive scientists have assessed how individuals understand various
types of complex sentences either in real time (whilst they read or listen
to them) or off-line (after the sentence is completed). These results have
shown interesting differences and similarities across languages and populations and they have enriched our knowledge on how language interacts
149
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 150 — #162
AG 4: Sentence complexity at the boundary of grammatical theory …
with other cognitive abilities (e.g., Lewis et al. 2006). Recent attempts have
been made to establish a dialogue between the grammatical and processing accounts (e.g. Lewis & Phillips 2015) thereby reviving the interest in
the relation between grammar and mental processes.
References: • Lewis, Shevaun & Phillips, Colin. 2015. Aligning grammatical theories and language processing models. Journal of Psycholinguistic Research 44. 27–46. • Lewis, Richard L.,
Vasishth, Shravan & Van Dyke, Julie A. 2006. Computational principles of working memory
in sentence comprehension. Trends in Cognitive Sciences 10. 447–454. • Rizzi, Luigi. 2013. Locality. Lingua 130. 169–186.
AG4
Introduction
Flavia Adani1 , Tom Fritzsche1 & Theodoros Marinis2
1
Universität Potsdam, 2 University of Reading
[email protected], [email protected], [email protected]
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: G 530
Intervention effects in adult grammar and language
acquisition
Luigi Rizzi
Università di Siena & Université de Genève
[email protected]
Mittwoch, 24. 02. 2016, 14:30–15:30, Raum: G 530
Intervention is a key concept both in syntactic theory and in the experimental study of language acquisition and adult processing. Are intervention effects in these domains amenable to a unified formal approach? In
syntax, intervention is a central component of the theory of locality. In the
Relativized Minimality approach (Rizzi 1990), an antecedent-trace relation
is disrupted when an element intervenes which is of the same type as the
antecedent. But in the case of A’-constructions, the deviance of intervention configurations is typically graded:
150
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 151 — #163
AG 4, Raum G 530
(1)
⁇ Which car do you wonder if we could buy __?
(2)
* What do you wonder if we could buy __?
This raises an interesting challenge for grammatical theory, which normally deals with simple categorical distinctions. A possible approach can
be based on the idea that the typology of positions is defined in terms of
morphosyntactic features, and a more richly specified element can be more
easily extracted from a weak island than a less richly specified element
(fRM: featural Relativized Minimality, Starke 2001; Rizzi 2004). In acquisition, certain object A’-dependencies crossing an intervening subject are
notoriously difficult for language learners:
(3)
Show me the boy that the doctor hugs __
Friedmann, Belletti & Rizzi (2009) treated these difficulties as grammarbased, and arising from fRM. Later work underscored more specific aspects
of the grammatically selective nature of the effect: e.g., number mismatch
helps the Italian speaking child significantly more than gender mismatch
(Adani et al. 2010); and the role of gender mismatch varies across languages
(Belletti et al. 2012). These findings led to the conclusion that the relevant
properties are morphosyntactic features triggering movement in the particular grammar. A uniform hierarchy of distinctness in terms of the settheoretic relation in featural specification between the target and the intervener was assumed to hold, with different cut-off points for adults and
children:
(4)
identity > inclusion > intersection > disjunction
This unified approach raises various questions:
1. What causes the observed differences between child and adult systems?
2. The complexity of inclusion (3) for adults emerges in ways other than
marginality, (e.g., Gordon et al. 2004).
3. How do marginality and complexity relate in general?
151
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 152 — #164
AG 4: Sentence complexity at the boundary of grammatical theory …
4. Does intersection, relevant for the child system, play any role in the
graded judgments involving weak-island extraction in adults?
References: • Adani, Flavia, van der Lely, Heather K. J, Forgiarini, Matteo & Guasti, Maria
Teresa. 2010. Grammatical feature dissimilarities make relative clauses easier: A comprehension study with Italian children. Lingua 120 (9). 2148–2166. • Belletti, Adriana, Friedmann,
Naama, Brunato, Dominique & Rizzi, Luigi. 2012. Does gender make a difference? Comparing the effect of gender on children’s comprehension of relative clauses in Hebrew and Italian. Lingua 122(10). 1053–1069. • Friedmann, Naama, Belletti, Adriana & Rizzi, Luigi. 2009.
Relativized relatives: Types of intervention in the acquisition of A-bar dependencies. Lin-
AG4
gua 119(1). 67–88. • Gordon, Peter C., Hendrick, Randall & Johnson, Marcus. 2004. Effects
of noun phrase type on sentence complexity. Journal of Memory and Language 51. 97–114.
• Rizzi, Luigi. 2004. Locality and let periphery. In Belletti, Adriana (ed.), Structure and Beyond,
223–251. New York: Oxford University Press. • Rizzi, Luigi. 1990. Relativized Minimality. Cambridge, MA: MIT Press. • Starke, Michal. 2001. Move dissolves into Merge: A theory of locality.
Doctoral dissertation, University of Geneva, Geneva.
What two-year-olds know – What two-year-olds say
Virginia Valian
Hunter College, The City University of New York
[email protected]
Mittwoch, 24. 02. 2016, 15:30–16:00, Raum: G 530
Three English case studies – determiners, subjects, and the passive – illustrate a strategy for understanding the interplay between knowledge and
processing of a given syntactic structure. It is surprisingly difficult to find
basic aspects of syntax (as opposed to syntactic details) that two- or threeyear-olds do not represent. In contrast, it is easy to find problems that
the tyro speaker has with the executive functions (EFs) involved in speaking and listening: planning, maintaining working memory, integrating the
different components of the grammar, and performing the right speech
acts. Because beginning speakers‘ EFs are less well developed than mature
speakers‘, their performance falls short of their competence.
152
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 153 — #165
AG 4, Raum G 530
To demonstrate the interplay between competence and performance, I
propose a two-part strategy: a) use diagnostic syntactic tests that two- and
three-year-olds should pass if they have the hypothesized knowledge and
b) use non-syntactic tests that manipulate demands on EFs to detect consequences for children’s performance. If children pass the syntactic tests
and if EF demands are apparent, one can conclude that children mentally
represent the structure at issue and begin to understand the role of cognition in child processing. All three aspects of grammatical demonstrate
both syntactic knowledge and immature EFs.
The predominant result of immature EFs is errors of omission (Brown
1973): a failure to include an obligatory element, such as, in English, a determiner, a subject, or an inflection – because every morpheme counts.
Failures to lexicalize cannot be random if the child is to communicate successfully and will not be random if the child can make use of other parts
of language. Two methods the child can use to figure out what not to include in her utterances are a) to exploit information structure and exclude
low-information elements, such as determiners that are not essential for
meaning and b) to use already established prosodic structures to fit her
utterance into, resulting in a failure to include elements that do not fit
the prosodic template, such as initial pronominal subjects. She can eschew
structures where agents are not canonically represented as subjects, such
as the passive. As processing demands increase, so will the child’s reliance
on reduced structures. In the case of determiners, children pass a variety
of syntactic tests (Valian et al. 2009) and show evidence of controlled processing that is ameliorated with age (Zangl & Fernald 2007), as well as decreasing reliance on prosodic templates with age (Demuth 2014). Data for
null subjects and passives show a similar pattern.
References: • Brown, Roger. 1973. A First Language. Cambridge: Harvard University Press.
• Valian, Virginia, Solt, Stephanie & Stewart, John. 2009. Abstract categories or limited-scope formulae? The case of children’s determiners. Journal of Child Language 36. 743–778.
• Zangl, Renate & Fernald, Anne. 2007. Increasing flexibility in children’s online processing
of grammatical and nonce determiners in fluent speech. Language Learning and Development
3. 199–231. • Demuth, Katherine. 2014. Prosodic Licensing and the development of phonological and morphological representations. In Farris-Trimble, Ashley W. & Barlow, Jessica
153
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 154 — #166
AG 4: Sentence complexity at the boundary of grammatical theory …
A. (eds.), Perspectives on phonological theory and development: Essays in honor of Daniel A.
Dinnsen, 11–24. Amsterdam: John Benjamins.
Processing of object clitics in Italian monolingual children
Elena Pagliarini1 & Fabrizio Arosio1
Università degli Studi di Milano-Bicocca
1
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: G 530
AG4
We investigated the comprehension of 3rd singular object clitics (DOCs)
in an act-out task (AOT) and in a looking while listening task (LWLT) in
monolingual Italian speaking children aged 5 years. 25 children participated in the AOT, 13 children in the LWLT. We administrated a digit span
test as a measure of short term memory. We aimed at studying: (i) children’s use of the clitic morphology to identify its referent in discourse;
(ii) the impact of the antecedent syntactic function (subject, object, VPadjunct); (iii) the impact of memory on DOC comprehension and processing strategies.
In the AOT, children acted out short stories with two characters of different genders; stories included sentences containing a DOC matching in
gender with one of the characters. We tested 5 items per condition in a
2x4 design with gender and syntactic function of the antecedent as factors. In the LWLT, children looked at pictures on a pc screen representing events involving two characters of different gender. One appeared in
the left bottom corner of the picture, the other in the right bottom corner.
While looking, children listened to sentences containing DOCs matching
with one of the previously named characters and answered to comprehension questions. Looks on pictures were recorded by a camera. We used the
AOT design.
Analyses: repeated measure logistic regression on accuracy and on looks
on target every 1/33sec. Children were divided in two groups according to
their d-span scores. AOT: (i) subjects in sentences with adjuncts were best
antecedents while adjuncts were worst antecedents; (ii) high span children
154
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 155 — #167
AG 4, Raum G 530
had better comprehension than low span children; (iii) there was an interaction between d-span groups and antecedent syntactic function. LWLT: (i)
high span children had better accuracy at the comprehension question; (ii)
subject antecedents were looked more than non-subject ones and subjects
in sentences with adjuncts were mostly preferred; (iii) d-span modulated
looks on target. Data show an initial preference for subject antecedents
but the presence of an object in the previous sentence impacts on the effectiveness of this strategy. The impact is modulated by short term memory
and it’s not a recency effect. Data suggest that a subject-topicality strategy
competes with a parallelism of grammatical function strategy requiring
pronouns to be interpreted as anaphoric to constituents filling the same
grammatical function (Grober et al. 1978; Maratsos 1973; Sheldon 1974;
Wykes 1981).
References: • Grober, Ellen H., Beardsley, William & Caramazza, Alfonso. 1978. Parallel function strategy in pronoun assignment. Cognition 6(2). 117–133. • Maratsos, Michael P. 1973.
The effects of stress on the understanding of pronominal coreference in children. Journal of
Psycholinguistic Research 2(1). 1–8. • Sheldon, Amy. 1974. The role of parallel function in the
acquisition of relative clauses in English. Journal of Verbal Learning and Verbal Behavior 13(3).
272–281. • Wykes, Til. 1981. Inference and children’s comprehension of pronouns. Journal of
Experimental Child Psychology 32(2). 264–278.
155
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 156 — #168
AG 4: Sentence complexity at the boundary of grammatical theory …
Syntactic complexity, verbal working memory, and
executive function in bilingual children with and without
specific language Impairment: A sentence repetition study
in France and in Germany
Rasha Zebib1 , Cornelia Hamann2 , Philippe Prévost1 , Lina Abed Ibrahim2
& Laurice Tuller1
1
François Rabelais University, Tours, 2 Universität Oldenburg
[email protected], [email protected],
[email protected]
AG4
Mittwoch, 24. 02. 2016, 17:00–17:30, Raum: G 530
Children with SLI are particularly sensitive to aspects of syntax argued
to involve high degrees of complexity as measured by movement, intervention in A’-dependencies, and sentential embedding. Sentence repetition
(SR) tasks have been shown to be excellent indicators of SLI in children
(Conti-Ramsden et al. 2001). However, it is often assumed that SR may in
fact measure verbal working memory (WM) and other non-linguistic cognitive skills, suggesting that SLI could be a fundamentally domain general
deficit. It has also been assumed that bilingualism enhances some aspects
of non-linguistic cognition (Calvo & Bialystok 2014), and, though children
have been targeted by only few studies, enhanced non- linguistic cognition
could in turn boost their ability to process complex syntax.
We conducted a SR study in 5- to 8-year-old L2 children with and without
SLI whose goal was to develop tools for the identification of SLI in bilinguals. French and German SR tasks were constructed to target complex
constructions mentioned above (Marinis & Armon-Lotem 2015). We recruited 82 bilingual children in France/Germany who spoke either Arabic,
Portuguese or Turkish, and the country language. Twenty-three children
met criteria for SLI (standard exclusionary criteria and subnormal language
performance, ascertained via assessment of the children’s languages and
use of pathology cut-offs adapted for bilingualism (Thordardottir 2015).
The Bi-SLI group displayed significantly lower SR scores than the BiTD group. Bi-SLI/Bi-TD differences were also significant for Forward Digit
Span (FDS), Backward Digit Span (BDS), and Card Sorting (CS). Specific
156
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 157 — #169
AG 4, Raum G 530
effects of the degree of syntactic complexity were also found. Correlation
analyses showed that SR is linked to BDS (verbal complex WM) in the BiSLI group, but not in the Bi-TD group, suggesting that a minimal memory span may be required to perform well on SR, which is the case in the
Bi-TD group (explaining the absence of BDS correlations), children with
SLI may attempt to compensate their linguistic deficit by relying on their
WM, and/or good linguistic skills entail good WM skills. Finally, CS, which
measures shifting, emerges for SR in the Bi-TD group, suggesting that their
performance on a given item is mainly influenced by other SR items. These
analyses will be completed by a regression analysis once more Bi-SLI data
are collected. The complexity effects manifested in Bi-SLI and Bi-TD children appear to be the combined result of linguistic competence and extralinguistic cognitive capacities.
References: • Calvo, Alejandra & Bialystok, Ellen. 2014. Independent effects of bilingualism and socioeconomic status on language ability and executive functioning. Cognition 130,
278–288. • Conti-Ramsden, Gina, Botting, Nicola & Faragher, Brian. 2001. Psycholinguistics
markers for Specific Language Impairment. Journal of Child Psychology and Psychiatry 42.
741–748. • Marinis, Theodoros & Armon-Lotem, Sharon. 2015. Sentence repetition. In Armon-Lotem, Sharon, de Jong, Jan & Meir, Natalia (eds.), Assessing Multilingual Children.
Disentangling Bilingualism from Language Impairment, 95–122. Bristol: Multilingual Matters.
• Thordardottir, Elin. 2015. Proposed Diagnostic procedures for use in Bilingual and Cross-Linguistic Contexts. In Armon-Lotem, Sharon, de Jong, Jan & Meir, Natalia (eds.), Assessing multilingual children: Disentangling bilingualism from language impairment, 331–358. Bristol: Multilingual Matters.
157
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 158 — #170
AG 4: Sentence complexity at the boundary of grammatical theory …
Which questions do German children process in an
adult-like fashion?
Atty Schouwenaars1 , Esther Ruigendijk1 & Petra Hendriks2
Carl von Ossietzky Universität Oldenburg, 2 University of Groningen
1
[email protected], [email protected],
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:30, Raum: G 530
AG4
Many studies report that object questions, in which the object precedes the
subject, are difficult to comprehend for children due to a strong subjectfirst bias (i.a. Friedmann & Novogrodsky 2011). The ability to identify who
is doing what to whom in non-canonical sentences using case or verbagreement develops late (e.g. Dittmar et al. 2008; Metz et al. 2010; De Vincenzi et al. 1999).
We aim to find out (1) when and to what extent German children use
these cues in their interpretation of object which questions, and (2) whether
different cues lead to different processing in terms of eye-gaze patterns.
We present an Optimality Theory analysis that allows us to describe and
predict intermediate and final interpretations based on interacting constraints regarding word order, subject-hood, verb-agreement and case. Children’s non-adult-like interpretations are explained in terms of an incorrect
ranking of constraints. In addition, we also predict for adults and children
differences between initial and final interpretations based on incremental
optimization. For example, in object questions with ambiguous case morphology and in passive questions the initial interpretation will be driven
by word order (i.e., first NP interpreted as subject), as no other information
is available yet.
A picture selection task with eye-tracking was carried out to test German
children’s (age=7-10 years, N=36), and adults’ (n=30) comprehension of
subject, object and passive questions. The subject and object questions were
disambiguated by different morphosyntactic cues: agreement and case, agreement only, and case only. Offline data confirm that children and adults respond less correctly on object questions than on subject and passive questions (children: 86%, 98%, 98%, respectively). Children’s accuracy scores are
158
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 159 — #171
AG 4, Raum G 530
correlated with their digit span scores. Unlike the offline data, the online
data indicate differences between the cues: object questions disambiguated
later by agreement and case cues (on NP2) lead to incorrect initial interpretations in both groups, due to a subject-first bias, whereas object questions
disambiguated immediately by case (on the wh-constituent) do not.
Our study shows that German children are sensitive to both case and
verb agreement and do revise their incorrect initial interpretation of object
questions. However, this ability is acquired quite late in German (around
age 8-9 years), which is argued to be due to an incorrect initial constraint
ranking. Revising the initial ranking may depend on the availability and
salience of the linguistic cues as well as on cognitive resources such as
memory.
References: • De Vincenzi, Marica, Arduino, Lisa, Ciccarelli, Laura & Job, Remo.1999. Parsing strategies in children comprehension of interrogative sentences. In Bagnara, Sebastiano
(ed.), European Conference on Cognitive Science, 301–308. Istituto di Psicologia del CNR, Rome.
• Dittmar, Miriam, Abbot-Smith, Kirsten, Lieven, Elena & Tomasello, Michael. 2008. German
children’s comprehension of word order and case marking in causative sentences. Child Development 79(4). 1152–1167. • Friedmann, Naama & Novogrodsky, Rama. 2011. Which questions
are most difficult to understand? The comprehension of wh-questions in three subtypes of SLI.
Lingua 121(3). 367–382. • Metz, Marijke, van Hout, Angeliek & van der Lely, Heather. 2010.
Understanding who and which questions in five to nine-year-old Dutch children: The role of
number. New developments in the acquisition of Dutch, Groninger Arbeiten zur germanischen
Linguistik 51. 27–41.
Complexity and memory
Shravan Vasishth
Universität Potsdam
[email protected]
Donnerstag, 25. 02. 2016, 09:00–10:00, Raum: G 530
When listening to or reading a sentence, syntactic complexity often has an
impact on ease of comprehension. One of the reasons that syntactic complexity impacts comprehension may be that the working memory system
159
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 160 — #172
AG 4: Sentence complexity at the boundary of grammatical theory …
imposes constraints on what is remembered and which dependencies are
built. In this talk, I will present one perspective: the role of working memory processes in sentence comprehension according to the ACT-R cuebased retrieval model (Lewis & Vasishth 2005; Lewis et al. 2006; Vasishth
et al. 2008; Engelmann, Vasishth et al. 2013; Patil et al. 2015; Nicenboim et
al. 2015; Engelmann et al. submitted).
References: • Felix Engelmann, Shravan Vasishth, Ralf Engbert & Reinhold Kliegl. 2013. A
framework for modeling the interaction of syntactic processing and eye movement control. Topics in Cognitive Science 5(3). 452–474. • Felix Engelmann, Lena A. Jäger & Shravan
AG4
Vasishth. submitted. e determinants of retrieval interference in dependency resolution: Review and computational modeling. • Richard L. Lewis & Shravan Vasishth. 2005. An activation-based model of sentence processing as skilled memory retrieval. Cognitive Science 29.
1–45. • Richard L. Lewis, Shravan Vasishth & Julie Van Dyke. 2006. Computational principles of working memory in sentence comprehension. Trends in Cognitive Sciences 10(10).
447–454. • Bruno Nicenboim, Shravan Vasishth, Reinhold Kliegl, Carolina Gattei & Mariano
Sigman. 2015. Working memory differences in long distance dependency resolution. Frontiers
in Psychology 6(312). • Umesh Patil, Sandra Hanne, Frank Burchert, Ria De Bleser & Shravan
Vasishth. 2015. A computational evaluation of sentence comprehension deficits in aphasia.
Cognitive Science. • Shravan Vasishth, Sven Bruessow, Richard L. Lewis & Heiner Drenhaus.
2008. Processing Polarity: How the ungrammatical intrudes on the grammatical. Cognitive
Science 32(4). 685–712. • Shravan Vasishth & Richard L. Lewis. 2006. Argument-head distance
and processing complexity: Explaining both locality and antilocality effects. Language 82(4).
767–794.
Children’s processing of relative clauses depends on who
‘they’ are
Yair Haendler
Universität Potsdam
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: G 530
Friedmann, Belletti & Rizzi (2009; FBR) claim that difficulties with object
relative clauses (OR) arise when both the head (the noun the relative clause
160
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 161 — #173
AG 4, Raum G 530
modifies) and the embedded subject are full DPs. They found children were
relatively accurate on Hebrew ORs with an embedded unpronounced pronoun (Ha- sus she-pro mesarkim oto ‘The-horse that-pro are brushing him’,
meaning ‘The horse that someone is brushing’). This facilitation was explained as due to the fact that pro is not a full DP.
However, two alternative explanations exist. First, the grammatical feature (GF) Number is different on the two relevant DPs: the head (e horse)
is [+singular] and the embedded subject (pro) [+plural]. Facilitation in such
cases is expected, according to further developments of FBR ʼs account
(Rizzi 2013) and recent experimental findings (Adani et al. 2014). Second,
pro refers to some unspecified, arbitrary agent. Not relating to any specific
discourse referent is a property that might facilitate its processing. Indeed,
sentence processing is constrained by discourse accessibility (DA) characteristics of pronouns. For instance, first-person pronouns involve a DA
operation that is cognitively less demanding than third- person pronouns
(3pro): the former relate to their discourse referents directly; the latter indirectly (Warren & Gibson 2002).
We tested 5-year-oldsʼ comprehension of Hebrew ORs whose embedded
subject is either a regular 3pro (Ha-susim she-hem tofsim ‘The horses that
they are catching’) or an arbitrary pro (Ha-susim she-pro tofsim otam ‘Thehorses that-pro are catching them’). The head and the embedded subject
had the same GFs: [+third-person][+plural][+masculine]. We thus neutralized effects due to the embedded pronoun or to mismatch in GFs. Hence,
no difference is expected between the two OR types according to FBR. But
3pro relates to its discourse referent indirectly, whereas pro has no specific
discourse referent. Therefore, if DA matters ORs with pro should be easier
than ORs with 3pro.
Children were more accurate on ORs with pro than with 3pro, suggesting that processing was influenced by the cognitive mechanism underlying
the execution of a DA operation. Specifically, the DA characteristics of pro
determined a more accurate performance. They are therefore likely to have
played a role also in FBRʼs items, although an additional facilitating effect
of mismatch in the GF Number cannot be ruled out. The results are discussed in the light of research that examined the impact of pronouns ʼ DA
characteristics in sentence processing and highlighted its relation to memory capacity (Haendler et al. 2015).
161
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 162 — #174
AG 4: Sentence complexity at the boundary of grammatical theory …
References: • Adani, Flavia, Forgiarini, Matteo, Guasti, Maria Teresa & van der Lely, Heather
K. J. 2014. Number dissimilarities facilitate the comprehension of relative clauses in children
with (Grammatical) Specific Language Impairment. Journal of Child Language 41(4). 811–841.
• Friedmann, Naama, Belletti, Adriana & Rizzi, Luigi. 2009. Relativized relatives: Types of
intervention in the acquisition of A-bar dependencies. Lingua 119(1). 67–88. • Rizzi, Luigi.
2013. Locality. Lingua 130. 169–186. • Warren, Tessa & Gibson, Edward. 2002. The influence of
referential properties on sentence complexity. Cognition 85(1). 79–112. • Haendler, Yair, Kliegl,
Reinhold & Adani, Flavia. 2015. Discourse accessibility constraints in children’s processing of
object relative clauses. Frontiers in Psychology 6(860).
AG4
Processing versus grammar of syntactic dependencies:
Neural oscillations of chunking, storage, and retrieval
Lars Meyer
Max Planck Institute for Human Cognitive and Brain Sciences, Leipzig
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: G 530
This talk summarizes a series of electroencephalography studies on the
neurocognition of syntactic dependencies, focusing on chunking, storage,
and retrieval: To establish syntactic dependencies, speech is chunked into
syntactic phrases, which must be stored and retrieved for dependency establishment (Lewis et al. 2006).
In a study on attachment ambiguities, the phase of delta band oscillations
in cortical regions related to auditory attention was found to predict participants’ chunking of sentences’ individual words into syntactic phrases
(Meyer et al., submitted). In a study on filler-gap dependencies, alpha band
power, predicted by working memory span, increased during argument
storage (Meyer et al. 2013). In a third study on pronoun reference, embeddedness of an antecedent noun in a prior clause increased both retrieval
demands and theta band network synchronicity (Meyer et al. 2015).
In the talk, I will argue that these results reconcile the processing and
grammar accounts of syntactic complexity (Lewis & Phillips 2015): On the
one hand, the links between the delta band and auditory attention and
162
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 163 — #175
AG 4, Raum G 530
between the alpha band and verbal working memory suggest that domaingeneral cognitive mechanisms (Jensen & Mazaheri 2010; Schroeder & Lakatos 2009) account for syntactic complexity. On the other hand, the link
between the theta band and syntactic structure suggests that grammar accounts for syntactic complexit–contents of verbal working memory appear
hierarchically structured (cf. Caplan & Waters 2013; Dillon et al. 2014).
The current findings generate interesting hypotheses for language acquisition research: Infants’ ability to process syntactic dependencies depends on working memory span (Felser et al. 2003) and neuronal growth
in cortices involved in working memory (Fengler et al. submitted); yet, it
is unknown whether this reflects the emergence of cortical inhibition via
the alpha band. Furthermore, infants learn to chunk words into syntactic phrases around six years of age (Männel et al. 2013), but it is unclear
whether this reflects infants’ increase in auditory attention via changes in
cortical delta band oscillations.
References: • Caplan, David & Waters, Gloria. 2013. Memory mechanisms supporting syntactic comprehension. Psychonomic Bulletin and Review 20. 243–268. • Dillon, Brian W., Chow,
Wing-Yee, Wagers, Matthew, Guo, Taomei, Liu, Fengqin & Phillips, Colin. 2014. The structuresensitivity of memory access: evidence from Mandarin Chinese. Frontiers in Psychology 5.
1025. • Felser, Claudia, Clahsen, Harald & Münte, Thomas F. 2003. Storage and integration in
the processing of filler-gap dependencies: An ERP study of topicalization and wh-movement
in German. Brain and Language 87. 345–354. • Fengler, Anja, Meyer, Lars & Friederici, Angela
D. Submitted. How the Brain Attunes to Sentence Processing: Relating Behavior, Sturcture,
and Function. • Jensen, Ole & Mazaheri, Ali. 2010. Shaping functional architecture by oscillatory alpha activity: Gating by inhibition. Frontiers in Human Neuroscience 4. 186. • Lewis,
Richard L., Vasishth, Shravan & Van Dyke, Julie A. 2006. Computational principles of working memory in sentence comprehension. Trends in Cognitive Sciences 10. 447–454. • Lewis,
Shevaun, & Phillips, Colin. 2015. Aligning grammatical theories and language processing
models. Journal of Psycholinguistic Research 44. 27–46. • Männel, Claudia, Schipke, Christine S. & Friederici, Angela D. 2013. The role of pause as a prosodic boundary marker: Language ERP studies in German 3-and 6-year-olds. Developmental Cognitive Neuroscience 5.
86–94. • Meyer, Lars, Henry, Molly J., Schmuck, Noura, Gaston, Phoebe & Friederici, Angela D.
Submitted. Beyond Speech Processing: Cortical Delta-Band Oscillations Predict Formation of
Syntactic Phrases during Human Language Comprehension. • Meyer, Lars, Grigutsch, Maren,
163
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 164 — #176
AG 4: Sentence complexity at the boundary of grammatical theory …
Schmuck, Noura, Gaston, Phoebe & Friederici, Angela D. 2015. Frontal–posterior theta oscillations reflect memory retrieval during sentence comprehension. Cortex 71. 205–218. • Meyer,
Lars, Obleser, Jonas & Friederici, Angela D. 2013. Left parietal alpha enhancement during
working memory-intensive sentence processing. Cortex 49. 711–721. • Schroeder, Charles E.,
& Lakatos, Peter. 2009. Low-frequency neuronal oscillations as instruments of sensory selection. Trends in Neurosciences 32(1). 9–18.
Misalignment of offline and online measures in Russian
relative clause processing
AG4
Iya Khelm Price1 & Jeffrey Witzel1
1
University of Texas at Arlington
[email protected], jeff[email protected]
Donnerstag, 25. 02. 2016, 12:00–12:30, Raum: G 530
Recent attempts to establish how grammar and language processing could
be part of the same cognitive system (Lewis & Phillips 2015; Phillips &
Lewis 2013) have called for comparisons of offline and online responses
to the same input. Such comparisons show how these responses complement each other even when they are not perfectly aligned. In particular,
it has been suggested that while online responses show intermediate steps
in building grammatical representations, offline judgments reflect different
stages of computation in the same system.
The present study investigated this idea by examining the processing of
Russian relative clauses (RCs) with online (self-paced reading) and offline
(acceptability judgment) measures. The sentences of interest were subjectextracted RC (SRC) and object-extracted RC (ORC) sentences in which an
NP argument intervened between the modified noun and the RC verb. This
created a configuration in which the same number of NP arguments was
available for integration at RC verb, across the same linear distance, in both
SRCs and ORCs. Furthermore, the influence of structural expectations was
investigated by using different NP types–descriptive NPs and pronouns–
inside the embedded clause. An offline acceptability judgment experiment,
complemented by a corpus analysis, indicated that these NP types are associated with different word order frequencies/preferences.
164
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 165 — #177
AG 4, Raum G 530
Some indications of online processing difficulty patterned with the offline measures. Specifically, in sentences that were dispreferred in offline
judgments or less frequent in the corpus, longer reading times were revealed at the first unexpected word–the embedded-clause NP. Other online effects did not correspond to the offline measures. For example, ORCs
were rated higher and were more frequent than their SRC counterparts.
However, there were comparable integration costs for SRCs and ORCs
at/after the RC verb when distance and the types of integrated elements
were held constant. Moreover, although ORCs with descriptive NPs were
judged offline as highly acceptable, late-‐stage comprehension difficulty
was revealed for these sentences in particular. This indicates that similaritybased interference, combined with ORC structural processing difficulty,
also influences processes related to retrieving and assigning thematic roles
to NPs during RC processing. These results thus suggest that intermediate
steps in online structure building related to expectation-based processing
correspond to offline measures, whereas online processing disruptions and
comprehension difficulty that appear to relate to memory demands do not.
These differences between the online and offline results might be taken
to reflect different stages of computation in a single cognitive system for
language processing.
References: • Lewis, Shevaun, & Phillips, Colin. 2015. Aligning grammatical theories and
language processing models. Journal of Psycholinguistic Research 44. 27–46. • Phillips, Colin &
Lewis, Shevaun. 2013. Derivational order in syntax: Evidence and architectural consequences.
Studies in Linguistics 6. 11–47.
165
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 166 — #178
AG 4: Sentence complexity at the boundary of grammatical theory …
Retrieval interference in relative clause attachment
ambiguity: Cross-linguistic evidence
Irina A. Sekerina
College of Staten Island, The City University of New York
[email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: G 530
AG4
Cross-linguistic differences in RC attachment (Someone shot the servant
[NP1] of the actress [NP2] who was on the balcony) that divide languages
into low- and high-attaching ones (English vs. Bulgarian and Russian) have
been explained by either grammatical factors (lexical semantics, prosody,
syntax; Grillo & Costa 2014) or working memory (WM) constraints in monolinguals (Swets et al. 2007), L2 adults (Hopp 2014), and children (Felser et al.
2003). In the present study, we suggest that the effect of WM can be modulated by retrieval interference from NP2 regardless of age and languagespecific attachment preferences.
We report data from a cross-linguistic study conducted with adults (N=123)
and 4-to-6-year-old children (N=76) in English, Bulgarian, and Russian using semantically shallow sentences that describe geometric shapes. Participants viewed pairs of pictures (9 experimental items and 21 fillers) while
listening to a spoken sentence (e.g. What color is the tip of the triangle that
has an umbrella in the middle?) with appropriate prosody and answered
the questions by naming the color. Attachment site–N1 (the tip) or NP2
(the triangle)–was visually disambiguated or left ambiguous. We measured
participants‘ accuracy and attachment preferences.
In the ambiguous display condition, contrary to the expected high attachment preference in Russian (Fedorova & Yanovich 2006) and Bulgarian
(Sekerina et al. 2003), there was a stronger low (59.9%) than high (27.1%)
preference for all languages and groups, although it was weaker for children (45.4% vs. 32.2%). In the disambiguated display condition, participants
were at ceiling naming colors in low (91.5%), but not in high (67.7%) condition. Surprisingly, participants in all languages frequently made a specific error in the disambiguated high condition that we refer to as “WholeObject”: they referred to the correct picture but named the color of NP2
166
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 167 — #179
AG 4, Raum G 530
instead of NP1. The Bulgarian and English adults were as likely to make
this error as the children (22.0% vs. 25.5%).
We hypothesize that low preference in the ambiguous condition overrides the language-specific high attachment preference (Bulgarian and Russian) because of computational demands on WM. Weaker low preference in
children can be due to guessing. “Whole-object” errors in the disambiguating high condition for both adults and children point to retrieval interference from the intervening NP2 that participants erroneously select. Thus,
not only cognitive factors per se but also their interaction must be taken
into account before the psycholinguistic debate on universality of parsing
strategies in RC attachment ambiguity and their development is settled.
References: • Fedorova, Olga & Yanovich, Igor. 2006. Early preference in relative clause attachment: The effect of working memory differences. In Lavine, James Eric (ed.), Annual
Workshop on Formal Approaches to Slavic Linguistics 14: The Princeton Meeting 2005. Michigan Slavic Publications. • Felser, Claudia, Marinis, Theodore & Clahsen, Harald. 2003. Children’s processing of ambiguous sentences: A study of relative clause attachment. Language
Acquisition 11(3). 127–163. • Grillo, Nino & Costa, João. 2014. A novel argument for the universality of parsing principles. Cognition 133(1). 156–187. • Hopp, Holger. 2014. Working
memory effects in the L2 processing of ambiguous relative clauses. Language Acquisition
21(3). 250–278. • Sekerina, Irina A., Fernández, Eva M. & Petrova, Krassimira A. 2003. Relative Clause attachment in Bulgarian. In Arnaudova, Olga, Browne, Wayles, Rivero, Maria
Luisa & Stojanović, Danijela (eds.), The Proceedings of the 12th Annual Workshop on Formal
Approaches to Slavic Linguistics. The Ottawa Meeting 2003, 375-394. Michigan Slavic Publications. • Swets, Benjamin, Desmet, Timothy, Hambrick, David Z. & Ferreira, Fernanda. 2007.
The role of working memory in syntactic ambiguity resolution: A psychometric approach.
Journal of Experimental Psychology: General 136(1). 64–81.
167
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 168 — #180
AG 4: Sentence complexity at the boundary of grammatical theory …
How representations determine stages of acquisition
Jill de Villiers1 & Tom Roeper2
Smith College, Northampton, 2 University of Massachusetts Amherst
1
[email protected], [email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: G 530
AG4
Unlike parsing approaches (Omaki et al. 2014) to long distance wh-movement,
we argue that all stages of grammar reflect changes in the child‘s representations. The parsing approach emphasizes the similarity of children’s parsing preferences to adults’ in terms of a filler-gap strategy (First Resort). Yet
published evidence reveals that children prefer long over short-distance
movement–non-adult behavior, including long distance wh-extraction across
light verbs (1) and even allowing it with non- light verbs (2).
(1)
Where did he make the decision to wash?
(2)
How did he like the decision to play?
Children also move across semi-factives (know):
(3)
When did the boy know he hurt himsel?
Note that the range of verbs across is not restricted to the verb say, making
it implausible to argue that children treat in “said NP” the verb as a parenthetical/evidential. Furthermore, they move WH across quantificational
adverbs (4) and negatives (5).
(4)
Why did the Mom always say the boy fell down?
(5)
Why did the boy not say that he had an ice-cream?
In each case we have argued for an explanation in terms of increasing representational knowledge, not parsing. Movement over not reflects the absence of a NEGP, following Relativized Minimality (Rizzi 1990). The common phenomenon in which children up to age 6 or 7 treat a medial WH
as a real question (6) is best explained by grammatical changes at a deeper
level than parsing, following a preference for interpretation one phase at a
time, at the first Phase.
168
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 169 — #181
AG 4, Raum G 530
(6)
How did the boy say where he rode a horse?
Nevertheless it is not the case that all such non-adult preferences are for
long distance movement. In responding to adjunct questions with small
clauses (7) children show a greater bias than adults do for a short distance
answer (how-see).
(7)
How did the Mother see him riding the horse?
We explain this as a failure to treat the small clause as having Exceptional
Case Marking, reflected in the fact that small children say e.g. “help my
dress”, and thus aligning it with a nominal that blocks long distance movement for adults too:
(8)
How did the Mother see his riding the horse?
Under a representational account, we predict variety in when the long or
short distance readings will predominate dependent upon other factors
such as lexical subcategorization and feature changes (Labeling) on structural nodes, and we can show that these predict preferences.
References: • Abdulkarim, Lamya. 2001. Complex wh-questions and Universal Grammars: New
evidence from the acquisition of negative barriers. Unpublished PhD thesis, University of Massachusetts, Amherst. • de Villiers, Jill G., Roeper, Thomas & Vainikka, Anne. 1990. The acquisition of long distance rules. In Frazier, • Lyn & de Villiers, Jill G. (eds.), Language Processing and
Acquisition. Dordrecht: Kluwer. • de Villiers, Jill G. & Roeper, Thomas. 1995. Barriers, binding
and acquisition of the DP/NP distinction. Language Acquisition 4, 73–104. • de Villiers, Jill G.,
Roeper, Thomas, Bland-Stewart, Linda & Pearson, Barbara. 2008. Answering hard questions:
Wh-movement across dialects and disorder. Applied Psycholinguistics 29, 67–103. • de Villiers,
Jill G., de Villiers, Peter A. & Roeper, Thomas. 2010. Wh-questions: Moving beyond the first
Phase. Lingua 121(3). 352–366 • Omaki, Akira, Davidson-White, Imogen, Goro, Takuya, Lidz,
Jeffrey & Phillips, Colin. 2014. No fear of commitment: Children’s incremental interpretation
in English and Japanese wh-questions. Language Learning and Development 10(3). 206–233
• Philip, William & de Villiers, Jill G. 1992. Monotonicity and the acquisition of weak islands.
In Clark, Eve V. (ed.), Proceedings of the twenty-fourth Stanford Child Language Conference,
99–111. Stanford University CSLI. • Rizzi, Luigi. 1990. Relativized Minimality. Cambridge, MA:
169
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 170 — #182
AG 4: Sentence complexity at the boundary of grammatical theory …
MIT Press. • Roeper, Thomas & de Villiers, Jill G. 2011. The acquisition path for wh-questions.
In de Villiers, Jill G. & Roeper, Thomas (eds.), Handbook of Generative Approaches to Language
Acquisition. Dordrecht: Springer. • Roeper, Thomas & de Villiers, Jill G. 1991. Ordered decisions in the acquisition of wh-questions. In Weissenborn, Jürgen, Goodluck, Helen & Roeper,
Thomas (eds.), eoretical Issues in Language Development, 191–236. Hillsdale, NJ: Lawrence
Erlbaum. • Roeper, Thomas & de Villiers, Jill G. 1994. Lexical links in the wh-chain. In Lust,
Barbara, Hermon, Gabriella & Kornfilt, Jaklin (eds.), Syntactic eory and First Language Acquisition: Cross Linguistic Perspectives, Vol. 2. Hillsdale, NJ: Lawrence Erlbaum.
AG4
What does semantic complexity mean for children?
Insights from the acquisition of relative clauses in German
Corinna Trabandt1 , Emanuela Sanfelici1 & Petra Schulz1
Goethe Universität Frankfurt am Main
1
[email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 12:00–13:00, Raum: G 530
This study is the first to investigate the role of semantic complexity in the
acquisition of German restrictive (RRCs) and appositive relative clauses
(ARCs). Previous research on RCs focused on the subject-object asymmetry, using mainly RRCs (1a), but not ARCs (1b) or semantically ambiguous
RCs (2) (Adani et al. 2012; Friedmann et al. 2009; King & Just 1991; Friederici
1998).
(1)
a.
b.
(2)
Nimm die zweite Uhr, die gelb ist.
Take the second watch who yellow is
‘Take the second watch that/which is yellow.’
a. Restrictive reading: ‘Take the second of the yellow watches’
b. Appositive reading: ‘Count watches independent of color; take
the second’
170
e reporter that the senator aacked admied the error.
Peter, who the senator aacked, admied the error.
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 171 — #183
AG 4, Raum G 530
RRCs are intersective property-denoting modifiers attached to the NP-level
(Heim & Kratzer 1998). ARCs, attached high at the DP edge, add information about the referent of the DP they relate to. They may be semantically
more complex than RRCs since additional semantic operations are argued
to apply at LF (Del Gobbo 2003; Potts 2005). Additionally, parsing principles (Right Association, Kimball 1973; Late Closure, Frazier 1987) predict
that RRCs due to their low attachment are less complex in adult processing. It is an open question how semantic complexity influences acquisition:
Does their semantic complexity delay the acquisition of ARCs, similarly to
the delay of mastering intervention, favoring subject over object RCs?
To address this question, we designed two experiments (computer-based
truth-value judgment, TVJT; preference task, PT), testing 4-to-6-year old
monolingual German-speaking children (TVJT: N=64, PT: N=53) and 20
adults each. All participants mastered a pre-test assessing knowledge of
ordinal numbers. In both tasks, semantically ambiguous RCs as in (2) were
presented in two prosodic conditions (restrictive, appositive). In the TVJT,
the visual context provided only the target-interpretation as indicated by
prosody. Participants judged whether the object-choice of the robot matched
the pre-recorded RC. In the PT, participants had to select the appropriate
object, indicated by pre-recorded RCs. Here, both readings (restrictive and
appositive) were provided by the visual context.
The results of the TVJT demonstrate that restrictive and appositive interpretations are available already at age 4. The results of the PT showed a
clear preference for restrictive interpretations, even when the RC-prosody
was appositive. Our findings indicate that semantic complexity does not
delay acquisition of ARCs. Instead, it is reflected in a strong preference for
RRCs, which are semantically less complex and are the structure initially
assigned during processing.
References: • Adani, Flavia, Sehm, Marie & Zukowski, Andrea. 2012. How do German children and adults deal with their relatives. In Stavrakaki, Stavroula, Lalioti, Marina & Konstantinopoulou, Polyxeni (eds.), Advances in Language Acquisition, 14–22. Newcastle, UK:
Cambridge Scholars Publishing. • Del Gobbo, Francesca. 2003. Appositives at the interface.
Unpublished doctoral dissertation. University of California, Irvine. • Frazier, Lyn. 1987. Sentence processing: A tutorial review. In Coltheart, Max (ed.), Aention and Performance 12: The
171
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 172 — #184
AG 4: Sentence complexity at the boundary of grammatical theory …
Psychology of Reading, 559–586. Hillsdale, NJ: Lawrence Erlbaum Associates. • Friederici, Angela D. 1998. Diagnosis and reanalysis: Two processing aspects the brain may differentiate.
In Fodor, Janet Dean & Ferreira, Fernanda (eds.), Reanalysis in Sentence Processing, 177–200.
Dordrecht: Kluwer. • Friedmann, Naama, Belletti, Adriana & Rizzi, Luigi. 2009. Relativized relatives: Types of intervention in the acquisition of a-bar dependencies. Lingua 119(1), 67–88.
• Heim, Irene & Kratzer, Angelika. 1998. Semantics in Generative Grammar. Malden, MA:
Blackwell. • Kimball, John. 1973. Seven principles of surface structure parsing in natural language. Cognition 2(1). 15–47. • King, Jonathan & Just, Marcel Adam. 1991. Individual differences in syntactic processing: The role of working memory. Journal of Memory and Language
30(5), 580-602. • Potts, Christopher. 2005. e Logic of Conventional Implicatures. Oxford: Oxford University Press.
AG4
When pragmatics helps syntax: An eye tracking study on
scope ambiguity resolution in 4- to 5-year-old children
Daniele Panizza1 & Karoliina Lohiniva2
1
Georg-August-Universität Göttingen, 2 Université de Genève
[email protected], [email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: G 530
Sentences with two scope-taking operators, subject-position alle (‘all’) and
nicht (‘not’), have two readings.
(1)
Alle Piraten sind nicht auf das Schiff zurückgekehrt.
‘All of the pirates did not go back to the ship.’
Under the surface-scope reading of (1), no pirates went back, whereas under the inverse-scope reading, not all pirates did. According to previous
studies, children display a strong surface-scope preference (Musolino et al.
2000), although they may adopt an inverse-scope interpretation if semantically primed (Musolino & Lidz 2006) or pragmatically facilitated (Viau et al.
2010; Gualmini et al. 2008). This could be due to children’s lack of processing resources, preventing them from revising their initial parse (assumedly
surface-scope) (Musolino & Lidz 2006; Viau et al. 2010), but uncontrolled
prosody may also influence children’s choices (Büring 1997; Syrett et al.
172
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 173 — #185
AG 4, Raum G 530
2014) and elicit incongruent results (Conroy et al. 2009). Finally, children
can derive indirect scalar implicatures (‘not all’ → ‘some’) (Bill et al. to
appear), possibly facilitating access to inverse scope.
To further investigate children’s accessing and preference of the two
readings, we designed an experiment combining a semantic decision task
with eye movement recording. Test sentence prosody was controlled for.
45 German-speaking children (46.9-71.9 months, mean=61.6) and 50 adult
controls heard 16 pirate stories, acted out by two toy-actor groups. Each
group depicted a) a FALSE scenario where all pirates went back, b) a NONE
scenario where no pirates did, or c) a SOME scenario where only some pirates did. Critically, SOME is only compatible with inverse-scope, while
NONE is compatible with both interpretations. The subjects rewarded the
group that best followed instructions (the test sentence). The NONE-FALSE
condition shows access to any of the two interpretations, the SOME-FALSE
condition shows access to inverse-scope, and the NONE-SOME condition
shows preference and timing of choice with respect to the access conditions. Children provided significantly more correct responses in SOMEFALSE vs. NONE-FALSE and displayed a slight preference for SOME in
NONE-SOME. In contrast, eye movement analysis of participants with good
comprehension of both configurations shows a lower latency in shifting
of looks towards the correct scenario in NONE-FALSE, suggesting that although surface scope can be accessed faster, inverse scope boosts children’s
overall accuracy. Furthermore, children who chose SOME in NONE-SOME
showed an indicative looking preference two seconds earlier than those
who chose NONE.
These results speak against the processing-based hypothesis and suggest
that pragmatic inferences may facilitate children’s access to inverse scope.
In conclusion, we found that German-speaking children and adults access
both scopal interpretations readily, even without prosodic biases.
References: • Bill, Cory, Romoli, Jacopo, Schwarz, Florian & Crain, Stephen. To appear. Scalar
implicatures versus presuppositions: e view from acquisition. Topoi. • Büring, Daniel 1997.
The great scope inversion conspiracy. Linguistics and Philosophy 20(2). 175–194. • Conroy,
Anastasia, Lidz, Jeffrey & Musolino, Julien. 2009. The Fleeting Isomorphism Effect. Language
Acquisition 16(2). 106–117. • Gualmini, Andrea, Hulsey, Sarah, Hacquard, Valentine & Fox,
173
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 174 — #186
AG 4: Sentence complexity at the boundary of grammatical theory …
Danny. 2008. The question-answer requirement for scope assignment. Natural Language Semantics 16(3). 205–237. • Musolino, Julien & Lidz, Jeffrey. 2006. Why children aren‘t universally successful with quantification. Linguistics 44(4). 817–852. • Musolino, Julien, Crain,
Stephen & Thornton, Rosalind. 2000. Navigating negative quantificational space. Linguistics
38, 1–32. • Syrett, Kristen, Simon, Georgia & Nisula, Kirsten. 2014. Prosodic disambiguation
of scopally ambiguous quantificational sentences in a discourse context. Journal of Linguistics
50. 453–493. • Viau, Joshua, Lidz, Jeffrey & Musolino, Julien. 2010. Priming of abstract logical
representations in 4-year-olds. Language Acquisition 17(1). 26–50.
AG4
Temporal, causal and conditional sentences in English
child-directed speech
Laura E. de Ruiter1 , Anna L. Theakston1 , Silke Brandt2 , &
Elena V. M. Lieven1
1
University of Manchester, 2 Lancaster University
[email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: G 530
Young children appear to have difficulty in comprehending and producing
temporal, causal or conditional sentences. Between three and five years
they misinterpret sentences like “Before the girl jumped the gate, she patted the horse” to mean that the jumping occurred first (e.g., Clark 1971).
They also reverse cause and effect in causal sentences (e.g., Emerson 1979).
Different factors have been suggested to influence children’s performance,
such as iconicity (e.g., Clark 1971) or memory limitations (e.g., Blything
et al. 2015). It has also been claimed that main-subordinate clause orders
should be easier to process and to produce in general (Diessel 2005).
One potential factor that has received relatively little attention is the
language that children actually hear as they grow up. What are the properties of these complex sentences in child directed speech (CDS)? Do certain
clause orders occur more often than others? Information about the input
is necessary to discern the relative contribution of semantic, syntactic and
processing factors on the one hand and familiarity/frequency effects on the
other.
174
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 175 — #187
AG 4, Raum G 530
We extracted all occurrences of the four prepositions/connectives aer,
before, because and if in about 93 hours of CDS (N=2399) from two dense
(British) English corpora of parent-child interaction (Lieven et al. 2009),
starting at the third birthday and covering six weeks.
Preliminary analyses indicate that temporal terms are relatively infrequent (ca. 12%). Most ‘before’s’ and ‘aer’s’ occur in other constructions
(e.g., in phrasal verbs), only 4% occur in temporal clauses. For all four sentence types, there appear to be clear preferences in clause order: beforeand because-sentences appear primarily in main-subordinate order, while
aer- and if -sentences tend to appear in subordinate-main order. We also
found that between 44.8% (if -sentences) and 68.5% (before-sentences) contain some sort of additional syntactic complexity (e.g., being embedded in
other subordinate clauses). Furthermore, over 83% of all subjects are either
pronouns or null forms; definite noun phrases, which often feature in experiments, are subjects in only about 5% of all sentences. Finally, over 90%
of all clauses contain given referents.
If input frequencies had an impact on the acquisition of complex sentences, they would predict different developmental trajectories in cases
where they run counter to suggested processing or iconicity preferences.
We will present analyses of the complete data set and discuss these predictions in connection with existing findings and the implications for currently planned experiments.
References: • Blything, Liam, Davies, Robert & Cain, Kate. 2015. Young children’s comprehension of temporal relations in complex sentences: The influence of memory on performance.
Child Development. • Clark, Eve V. 1971. On the acquisition of the meaning of before and after.
Journal of Verbal Learning and Verbal Behavior 10(3). 266–275. • Diessel, Holger. 2005. Competing motivations for the ordering of main and adverbial clauses. Linguistics 43(3). 449–470.
• Emerson, Harriet F. 1979. Children’s comprehension of because in reversible and non-reversible sentences. Journal of Child Language 6(2). 279–300. • Lieven, Elena, Salomo, Dorothé,
& Tomasello, Michael. 2009. Two-year-old children‘s production of multiword utterances: A
usage-based analysis. Cognitive Linguistics 20(3). 481–507.
175
AG4
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 176 — #188
AG 4: Sentence complexity at the boundary of grammatical theory …
Alternate speakers
e acquisition of complex sentences in German ildren
with hearing impairment
Eva Wimmer1 , Martina Penke1 & Monika Rothweiler2 , 1 Universität zu Köln,
2
Universität Bremen
AG4
[email protected], [email protected], [email protected]
Differential demands in the acquisition, production and comprehension of passive sentences: Disentangling an apparent
paradox
João C. de Lima Júnior1 , Letícia M. S. Corrêa1 & Marina R. A. Augusto2 ,
Pontifícia Universidade Católica do Rio de Janeiro, 2 Universidade do Estado do Rio de Janeiro
1
[email protected], [email protected], [email protected]
176
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 177 — #189
Arbeitsgruppe 5
The grammatical realization of polarity:
Theoretical and experimental approaches
Christine Dimroth1 & Stefan Sudhoff2
1
Westälische Wilhelms-Universität Münster, 2 Universiteit Utrecht
AG5
[email protected], s.sudhoff@uu.nl
Raum: F 425
Workshop description
The expression of polarity contrast that is particularly prominent in languages like German and Dutch has recently been in the centre of empirical as well as theoretical investigations. In these languages, contrasts between statements with negative and positive polarity are marked with the
help of prosody (nuclear pitch accent on the finite verb or complementizer,
i.e. verum focus, cf. Höhle 1992) or assertive particles (wel/wohl; toch/doch;
schon) that also carry focal stress (contributions by Blühdorn and Sudhoff
in Lohnstein & Blühdorn 2012; Turco et al. 2014).
To date there is no consensus on the exact meaning contribution of these
devices or on the kind of contrast that is actually evoked. Possibilities under
discussion include assertion vs. non-assertion, polarity, illocution, and sentence mood. Other open questions concern the fate of the verum operator
in case it is not focused, the question how similar assertive particles and
verum focus really are, how comparable contexts are expressed in other
languages, what the specific parameters of the prosodic marking of verum
focus are, and how they relate to other kinds of prosodic focus marking.
With few exceptions, this vivid debate is not informed by empirical data.
177
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 178 — #190
AG 5: e grammatical realization of polarity
The workshop wants to bring together researchers from a theoretical
and an empirical orientation and to enhance our understanding of the phenomenon with the help of cross-linguistic comparisons. It mainly focuses
on (but is not restricted to) West-Germanic languages. We welcome contributions dealing with the syntactic, semantic, pragmatic and/or prosodic
aspects of the phenomenon.
References: • Blühdorn, Hardarik. 2012. Faktizität, Wahrheit, Erwünschtheit: Negation, Negationsfokus und „Verum“-Fokus im Deutschen. In Lohnstein, Horst & Blühdorn, Hardarik
(eds.). Wahrheit – Fokus – Negation. Linguistische Berichte, Sonderheft 18, 137 – 170. • Blüh-
AG5
dorn, Hardarik & Lohnstein, Horst. 2012. Verumfokus im Deutschen: Versuch einer Synthese.
In Lohnstein, Horst & Blühdorn, Hardarik (eds.). Wahrheit – Fokus – Negation. Linguistische
Berichte, Sonderheft 18, 171 – 261. • Dimroth, Christine; Andorno, Cecilia; Benazzo, Sandra &
Verhagen, Josje. 2010. Given claims about new topics. How Romance and Germanic speakers
link changed and maintained information in narrative discourse. Journal of Pragmatics. 42,
3328 – 3344. • Gutzmann, Daniel. 2012. Verum – Fokus – Verum-Fokus? Fokus-basierte und
lexikalische Ansätze. In Lohnstein, Horst & Blühdorn, Hardarik (eds.). Wahrheit – Fokus –
Negation. Linguistische Berichte, Sonderheft 18, 67 – 103. • Hogeweg, Lotte. 2009. The meaning and interpretation of the Dutch particle wel. Journal of Pragmatics. 41, 519–539. • Höhle,
Tilman. 1992. Über Verum-Fokus im Deutschen. In Jacobs, Joachim (ed.), Informationsstruktur
und Grammatik, 112 – 141. Opladen: Westdeutscher Verlag. • Klein, Wolfgang. 2006. On finiteness. In Van Geenhoven, Veerle (ed.), Semantics in Acquisition, 245 – 272. Dordrecht: Kluwer.
• Lohnstein, Horst. 2012. Verumfokus – Satzmodus – Wahrheit. In Lohnstein, Horst & Blühdorn, Hardarik (eds.). Wahrheit – Fokus – Negation. Linguistische Berichte, Sonderheft 18, 31
–66. • Sudhoff, Stefan. 2012. Negation der Negation – Verumfokus und die niederländische
Polaritätspartikel wel. In Lohnstein, Horst & Blühdorn, Hardarik (eds.). Wahrheit – Fokus –
Negation. Linguistische Berichte, Sonderheft 18, 105– 136. • Turco, Giuseppina; Braun, Bettina
& Dimroth, Christine. 2014. When contrasting polarity, Germans use intonation, the Dutch
particles. Journal of Pragmatics 62, 94 – 106. • Turco, Giuseppina; Dimroth, Christine & Braun,
Bettina. 2013. Intonational means to mark Verum Focus in German and French. Language and
Speech. 56, 460 – 490.
178
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 179 — #191
AG 5, Raum F 425
The grammatical realization of polarity: Introductory
remarks
Christine Dimroth1 & Stefan Sudhoff2
1
Westälische Wilhelms-Universität Münster, 2 Universiteit Utrecht
[email protected], s.sudhoff@uu.nl
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: F 425
On two types of polar interrogatives in Hungarian and
their interaction with inside and outside negation
AG5
Beáta Gyuris
RIL HAS Budapest
[email protected]
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: F 425
The aims of the paper are i) to argue for the validity of the distinction between “outside negation” (ON) vs. “inside negation” (IN) readings of negative interrogatives (cf. Ladd 1981) for Hungarian, a language with two root
interrogative form types; ii) to point out some contrasts in this language
that do not seem to be predicted by theories of ON-IN readings by Büring
& Gunlogson (2000), Romero & Han (2004), Reese (2007), and Krifka (to
appear); and iii) to propose explanations for them.
The phenomena discussed include the following. Intonationally marked
(rise-fall or /\-) interrogatives display both IN and ON readings, whereas
morphologically marked (-e-) interrogatives only have the latter. The range
of contexts where negative -e-interrogatives are felicitous are much more
restricted than those where /\-interrogatives on their ON-readings are available, calling the possibility of a uniform semantic account of ON-readings
into question. There is a difference in the range of contexts where negative -e-interrrogatives appear, depending on whether the interrogative
marker attaches to the verb or to the negative particle (in some nonstandard dialects). ON readings of negative /\-interrogatives can be expressed
179
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 180 — #192
AG 5: e grammatical realization of polarity
by structures string-identical to negated verum focus or negated VP structures, presenting a problem to analyses along the lines of Romero & Han
(2004).
References: • Büring, Daniel & Gunlogson, Christine. 2000. Aren’t positive and negative polar
questions the same? UCLA & UCSC. Unpublished manuscript. • Krifka, Manfred. to appear.
Negated polarity questions as denegations of assertions. In Kiefer, Ferenc & Lee, Chungmin
(eds.), Contrastiveness and scalar implicatures. Berlin: Springer. • Ladd, D. Robert. 1981. A
first look at the semantics and pragmatics of negative questions and tag questions. CLS 17.
164–171. • Reese, Brian. 2007. Bias in questions. Austin: University of Texas. (Doctoral disser-
AG5
tation.) • Romero, Maribel & Han, Chung-hye. 2004. On negative yes/no questions. Linguistics
and Philosophy 27. 609–658. • van Rooij, Robert & Šafářová, Marie. 2003. On polar questions.
SALT XIII. 292–309.
Verum focus and contrast
Horst Lohnstein
Bergische Universität Wuppertal
[email protected]
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: F 425
The term verum focus was porposed by Höhle (1992) to denote a phenomenon
that he characterized as emphasizing the expression of a propositions’s truth
as exemplified by the following example:
(1)
Karl HAT seine Großmutter besucht.
Carl has his grandmother visited
‘Carl DID has visited his grandmother.’
Paraphrase: ‘It is true that Carl has visited his grandmother.’
Even for interrogative and imperative clauses – which do not allow for
truth value assingment at all – similar paraphrases using the predicate
“true” are proposed by him by inserting an element VERUM into the clausal
structure.
In this talk, I will argue that in clausal structures there is no element
VERUM at all, neither in the syntactic nor in the semantic part. Rather, the
180
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 181 — #193
AG 5, Raum F 425
position is taken up that sentence moods are intensional functions which
map states of affairs to truth values along the lines of Frege’s (1892/1997)
distinction between sense and reference. The focus feature [+F] inducing
verum effects is located in the head position M0 of a mood phrase MP in
the sentential left periphery in German.
Embedded verum focus constructions realize verum focus in the position occupied by a complementizer as well as by relative or wh-pronoun.
A position shared by some scholars based on a proposal by Höhle is the
assumption that wh- and relative phrases as well as complementizers and
fronted finite verbs form a natural class and – moreover – occupy the same
left peripheral position. Contrary to this proposal, I will argue that in the
case of embedded wh- and relative clauses the focus feature [+F] is interpreted in the head position M0 at LF, but is realized at PF in the specifier
position SpM exactly in case the head M0 is phonetically empty. The respective features form a chain whose head is interpreted at PF and the foot
at LF – as the usual concept proposes.
Based on these assumptions, a theory is proposed which derives the relevant verum focus facts in a compositional way. It makes use of a systematic
interaction of the principles of focus interpretation and the regular constitution of sentence moods in German. Verum focus – under this perspective
– is a regular focus phenomenon in that it tries to reduce alternatives to
the originally given sentence mood function which forms a contrast to the
(verbal) behaviour or (presumed) beliefs of the discourse participants.
The proposed theory delivers the grammatical basis for the communicative functions verum focused utterances are assumend to perform in discourse situations. For instance, that a speaker attempts to downdate the
question under discussion (QUD) as proposed by Gutzmann & Castroviejo
(2011), Gutzmann (2012) and with similar ideas also by Lai (2011) an others.
Thus, verum focus is a grammatical instrument for the purpose of common ground managment (cf. Krifka 2008). Placed on the head position M0
of the sentence mood phrase MP, the theory answers the question why
verum focus is realized in the left position of clauses and derives its communicative effects from the interaction of regular grammatical means (cf.
Lohnstein to appear). Seen from this perspective, the phenomenon should
be labeled more adequately as focus on sentence mood.
181
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 182 — #194
AG 5: e grammatical realization of polarity
References: • Frege, Gottlob. 1892/1997. On sense and reference. In Ludlow, Peter (ed.), Readings in the Philosophy of Language. Cambridge, Mass.: The MIT Press. • Gutzmann, Daniel.
2012. Verum – Fokus – Verumfokus. In Lohnstein, Horst & Blühdorn, Hardarik (eds.), Wahrheit
– Fokus – Negation. Linguistische Berichte, Sonderheft 18, 6–114. • Gutzmann, Daniel & Castroviejo Miró, Elena. 2011. The Dimensions of Verum. In Bonami, Olivier & Cabredo Hofherr,
Patricia (ed.), Empirical Issues in Syntax and Semantics 8, 143–165. • Höhle, Tilman. 1992.
Über Verum-Fokus im Deutschen. In Jacobs, Joachim (ed.), Informationsstruktur und Grammatik. Linguistische Berichte, Sonderheft 4, 122–141. • Krifka, Manfred. 2008. Basic Notions of
Information Structure. In Krifka, Manfred; Féry, Caroline & Fanselow, Gisbert (eds.), Interdisciplinary Studies on Information Structure 6, 13–55. Potsdam. • Lai, Catherine. 2011. Update
Foregrounding Verum Focus, Prosody and Negative Polar Questions. Paper presented at the
AG5
85th Annual Meeting of the Linguistic Society of America (LSA). Pittburgh January 6-9. 2011.
• Lohnstein, Horst. to appear. Verum Focus. In Féry, Caroline & Ishihara, Shinichiro (eds.),
Handbook of Information Structure. Oxford: Oxford University Press.
Negative polarity, focus accent, and the embedding of
interrogatives by veridical predicates
Peter Öhl
Bergische Universität Wuppertal
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: F 425
Veridical predicates like convinced normally select that-clauses, whereas
certain non-veridical predicates like doubt or wonder select if-clauses:
(1)
a.
b.
Homer is convinced that 0 is a prime number.
Homer doubts/wonders if 0 is a prime number.
However, many of these veridical verbs embed if -clauses, if they are in
scope of a nonveridical operator like NEG (cf. Giannakidou 1998). This phenomenon has been termed unselected embedded questions in the literature
(Adger/Quer 2001; Öhl 2007):
(2)
182
Homer is not convinced that/if 0 is a prime number.
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 183 — #195
AG 5, Raum F 425
Speakers’ judgments show that the acceptance of if -clauses with veridical
predicates varies – both contextually and dependent on the speaker. With
most of these predicates, however, the average rate of acceptance is lower
without such an operator. This is indicated by ‘#’, below:
(3)
Homer is convinced that/#if 0 is a prime number.
Focus accent seems to be relevant for the speakers’ judgment: if the predicate is stressed, a that-clause is preferred also under negation (4a). If the
negation particle is stressed, there is an (admittedly not as strong) tendency
towards embedding an if -clause (4b). If the finite matrix-verb is stressed
(verum-focus; Höhle 1988; cf. Lohnstein to appear), both complemetisers
seem to be licensed the same way (4c).
(4)
a.
b.
c.
Homer is not coninced that/#if 0 is a prime number.
Homer is no convinced (#)that/if 0 is a prime number.
Homer i not convinced that/if 0 is a prime number.
We will defend the claim that, if an if -clause is selected by the verb or
licensed by a nonveridical operator, the acceptance is clear-cut. Otherwise, a nonveridical context must be logically accessible. A focussed NEG
takes immediate scope over the predication and forces nonveridicality (5a).
The focussed predicate, however, keeps its selectional properties. Thus, the
whole proposition is negated (5b).
(5)
a.
b.
[[ convinced’ [prime number’(0’) prime number’(0’)] ]] = 1
[[ convinced’ [prime number’(0’)] ]] = 0
Since verum-focus is associated with a syntactic position that may scope
over both kinds of structure, both readings are licensed.
References: • Adger, David & Quer, Josep. 2001. The syntax and semantics of unselected embedded questions. Language 77 (1). 107-133. • Giannakidou, Anastasia. 1998. Polarity sensitivity as (non)veridical dependency. Amsterdam, Philadelphia (PA): Benjamins. • Höhle, Tilman.
1988. Vorwort und Nachwort zu Verumfokus. Sprache und Pragmatik 5. 1–7. • Lohnstein,
Horst. to appear. Verum Focus. Féry, Caroline & Ishihara, Shinichiro (eds.): Handbook of Information Structure. Oxford: Oxford University Press. • Öhl, Peter. 2007. Unselected Embedded
183
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 184 — #196
AG 5: e grammatical realization of polarity
Interrogatives in German and English. S-Selection as Dependency Formation. Linguistische
Berichte 212. 403-437.
Complementisers as markers of negative polarity in
German comparatives
Julia Bacskai-Atkari
Universität Potsdam
[email protected]
Mittwoch, 24. 02. 2016, 17:00–17:30, Raum: F 425
AG5
I examine three types of German comparatives in terms of their synchronic
and diachronic relation: (I) positive polarity AS-clauses (Ralf ist so groß
wie Peter ‘Ralph is as tall as Peter’), (II) negative polarity THAN-clauses
(Ralf ist größer als Peter ‘Ralph is taller than Peter’), and (III) hypothetical
comparatives (my daughter is shouting as if she were at the dentist’s).
German has two basic ways of forming (III); the first is a biclausal construction involving wie wenn, where the wie-clause can be elliptical (meine
Tochter schreit, wie sie sreien würde, wenn sie beim Zahnarzt wäre ‘my
daughter is shouting, as she would be shouting if she were at the dentist’s’).
The wie-clause has positive polarity just as in (I), and the wenn-clause has
negative polarity, as a regular conditional clause containing a yes/no operator. The second way is a monoclausal construction involving als wenn
or als ob, with no full clausal variant (meine Tochter schreit, als wenn/ob
sie beim Zahnarzt wäre). Here both elements encode negative polarity.
I argue that the reanalysis from biclausal into monoclausal combination
is enabled if the comparative complementiser is extended from contexts
(I) to (II), since otherwise it would not be compatible with a negative polarity complementiser in the same clause. In turn, the combination also
allows the relative feature to be encoded only by the comparative complementiser: hence ob was preserved in (III) even though it is no longer
a conditional complementiser otherwise. The lower CP projection is still
necessary to host the yes/no operator, but if there is no overt operator, an
overt head must license the projection, which is satisfied either by ob or by
verb movement.
184
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 185 — #197
AG 5, Raum F 425
Can but be an NPI?
Unaisa Khir Eldeen
University of Essex
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:00, Raum: F 425
Utterances of the form not X but Y, as in (1), proved to be problematic for
many accounts. There are two approaches: the first considers but to be ambiguous (Anscombre & Ducrot 1977); the second monosemous (Blakemore
2002).
(1)
AG5
I didn’t go to Paris, but to London.
I will argue that there are two syntactically different buts corresponding to
different meanings. In (1), the use of but is constrained by the negation in
the first part of the utterance. If the order of the coordinated conjuncts is
reversed, but seems to be optional as in (2) and (3). However, this optionality is superficial as the meaning of (3) is different from that in (2).
(2)
I went to London, not to Paris.
(3)
I went to London, but not to Paris.
Semantically anomalous utterances such as (4) give greater evidence that
but used in (3) is a denial but that is different from the correction but in (1).
(4)
*Kim is a woman, but not a man.
I argue that correction but in (1) is syntactically different in that it is licensed by the negation in the first conjunct. The absence of negation in the
first conjunct in (2) shows that the correction but cannot be used. Therefore, I posit that correction but is a negative polarity item that is licensed
by negation in the clause preceding but, conditioning that what follows
but must be an alternative to the element that is focused by negation in
the first conjunct. For example, in (1), the focus of negation is on ‘to Paris’,
and what follows but is ‘to London’, which functions as an alternative to
‘to Paris’.
185
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 186 — #198
AG 5: e grammatical realization of polarity
References: • Anscombre, Jean-Claude & Ducrot, Oswald. 1977. Deux mais en français? Lingua, 43 (1), 23-40. • Blakemore, Diane 2002. Relevance and linguistic meaning: e semantics
and pragmatics ofdiscourse markers (Vol. 99). Cambridge University Press. • Gajewski, Jon
Robert. 2007. Neg-raising and polarity. Linguistics and Philosophy, 30 (3), 289-328.
Establishing polarity contrasts in English: Evidence from
analyses of natural conversations
Beatrice Szczepek-Reed1 & Leah Roberts1
1
University of York
AG5
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 18:00–18:30, Raum: F 425
Based on findings from recent research on German and Dutch using elicited
oral production techniques, it has been observed that these two Germanic
languages establish contrasts between positive and negative polarity via
prosodic emphasis on the finite verb (verum focus in German) or stressed
lexical particles (e.g., wel ‘indeed’, in Dutch) (e.g., Turco, Braun & Dimroth,
2014). In this talk, we will discuss our current research on how English
native speakers establish positive and negative polarity contrasts. English
is an interesting case to add to the cross-linguistic data set on the topic.
This is because it is another Germanic language that can both make use
of particles such as indeed, and of prosodic marking of the finite verb as
indicated in (1) below. However, unlike German and Dutch, English also
has available do-support for marking contrasts (2).
(1)
A: ‘You haven’t seen the film e Hobbit yet, have you?’
B: ‘I have seen it….’
(2)
A: ‘You didn’t see e Hobbit, right?’
B: ‘I did see it….’
English speakers can use do-support on its own, or with optional additional
prosodic highlighting on the inflected form of do, and with optional lexical
items such as indeed or actually. English therefore offers a wider range of
linguistic options which partially overlap with those available to German,
186
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 187 — #199
AG 5, Raum F 425
Dutch. In this project, we examine the functional contexts in which polarity
contrasts are produced, and the specific devices that are employed, using
an existing corpus of 13 hours of video recordings of naturally-occurring
native English dinner conversations. In sum, in this talk we will discuss the
range of interactional and linguistic contexts and devices used in English
speakers’ expression of polarity contrasts, and discuss similarities and differences to those of German and Dutch.
References: • Turco, Giuseppina; Braun, Bettina & Dimroth, Christine. 2014. When contrasting polarity, Germans use intonation, the Dutch particles. Journal of Pragmatics 62, 94-106.
Contrasting polarity: Verum focus and affirmative particles
in Germanic and Romance languages
Giuseppina Turco
Université Sorbonne Nouvelle
[email protected]
Donnerstag, 25. 02. 2016, 09:00–10:00, Raum: F 425
In this talk I am going to present an investigation on the expression of
(affirmative) polarity contrast (e.g. ‘the child DID cry’) from a typological
and acquisitional perspective. In order to test the relevance of this pragmatic category for the discourse flow, a dialogue task was carried out in
four different languages: Dutch, German, French and Italian. A first production study on Dutch and German showed that even two closely related
languages behave differently when it comes to signalling the same pragmatic function. Contrastive polarity was expressed by German speakers
using so-called verum focus (a high-falling pitch accent on the verb, ‘das
Kind HAT geweint’), whereas Dutch speakers mostly produced the particle wel (‘het kind heeft wel gehuild’). A second study investigated the same
contrast in French and Italian, which were predicted not to encode polarity contrast in a systematic way. Results showed an intonational contrast,
much like the verum focus of German, in one third of the cases. This contrast had not been documented before. Yet, compared to Dutch and German, contrastive polarity was expressed much less frequently in French
187
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 188 — #200
AG 5: e grammatical realization of polarity
and Italian, thereby suggesting that the information unit ‘polarity’ plays
a less relevant role for discourse coherence. Data from L2 acquisition production provided further support for these typological differences. Possible
implications for language processing are discussed in light of preliminary
findings suggesting that language-specific expressions for polarity influence speakers’ interpretation of polarity.
Polarity particles in response to negated antecedents: Two
groups of speakers for German ja and nein
AG5
Berry Claus1 , A. Marlijn Meijer1 , Sophie Repp1 & Manfred Krifka1
1
Humboldt-Universität zu Berlin
[email protected], [email protected], [email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: F 425
The present study attends to German polarity particles and focusses on
agreeing responses to antecedents with sentential negation. In this case,
German ja ‘yes’ and nein ‘no’ are not complementary (Blühdorn 2012: 386)
– rather, both, ja and nein, can be used (e.g. A: Jim schnarcht nicht. ‘Jim
doesn’t snore.’ B: Ja./Nein. [=He doesn’t snore]). Two recent accounts of
polarity particles, Roelofsen & Farkas (R&F, 2015) and Krifka (2013), make
predictions concerning preference patterns for ja and nein. In a nutshell,
R&F’s feature model predicts a general preference of nein over ja in agreeing responses to antecedents containing sentential negation. In contrast,
Krifka’s anaphor account predicts a default preference of nein over ja which
can be reversed in particular contexts. We tested the predictions in a series of experiments. Participants were presented with short dialogues and
judged the naturalness and suitability of the response in the given dialogue
and context. The results for agreeing responses to negated antecedents
were neither consistent with R&F’s nor with Krifka’s predictions: overall,
ja was rated significantly higher than nein with no contextual modulation.
However, a closer data inspection revealed differences among participants.
A majority of participants showed the unpredicted preference for ja over
nein (jagop). Yet, a notable minority displayed a preference for nein
188
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 189 — #201
AG 5, Raum F 425
over ja (neingop). In our preliminary account, we assume that these
two subgroups differ in the underlying response system: hale vs.
polai baed (Jones 1999). The ja-group tends towards a truth-value
strategy with ja signalling the truth of the antecedent and nein its falsity,
i.e. a polarity contrast between antecedent and response. The nein-group,
however, tends towards a polarity-based strategy with nein signalling a
negative response polarity and ja a positive one.
References: • Blühdorn, Hardarik. 2012. Negation im Deutschen. Tübingen: Narr Verlag. • Krifka,
Manfred. 2013. Response particles as propositional anaphors. Proceedings of SALT 23. 1-18.
• Jones, Bob Morris. 1999. e Welsh answering system. Berlin: Mouton de Gruyter. • Roelofsen,
Floris & Farkas, Donka. 2015. Polarity particle responses as a window onto the interpretation
of questions and assertions. Language 91. 359-414.
In search for “verum focus” marking in Italian: a
contribution from MapTasks data
Cecilia Andorno1 & Claudia Crocco2
1
Università degli Studi di Torino, 2 Universiteit Ghent
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: F 425
Despite extensive use of both prosody and constituent position to encode
information structure (Avesani & Vayra 2005; Bonvino 2005; Crocco 2013;
Scarano 2003), Italian does not seem to be a well-fit language for focusing
on polarity in isolation. Namely, both negation and auxiliaries behave as
clitics in Italian (Avesani & Vayra 2005; Swerts et al. 2002); particles in
post-finite verb position and verum focus marking are only exploited to
a minimal extent by Italian speakers - when compared with speakers of
Germanic languages - both in relating events in narratives (Dimroth et
al. 2010; Benazzo et al. 2012), and in comparing scenes in dialogic tasks
(Turco 2014). In the current paper, we extend the search for verum focus
marking in Italian to elicited data (MapTasks and Interviews), in which
speakers need to (positively) reply to negatively biased sentences (You did
189
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 190 — #202
AG 5: e grammatical realization of polarity
not learn German then? Yes I did learn German). The overall positional
and prosodic sentence organization of such replies have been taken into
account, in order to find out whether and how speakers exploit resources
from their own language, which both lacks a particle as German doch and
an auxiliary as English do. To evaluate the role of prosodic prominence in
verum focus marking, the relevant utterances have been auditory analyzed
and their pitch contours have been visually inspected using Praat.
References: • Avesani, Cinzia & Vayra, Mario. 2005. Accenting deaccenting and information
structure in Italian dialogues. Madison, Wisconsin (USA): Omnipress. • Benazzo, Sandra et
AG5
al. 2012. Perspective discursive et influence translinguistique: Exprimer le contraste d’entité en
français et en italien L2. LIA 3-2, 173-201. • Bonvino, Elisabetta. 2005. Le sujet postverbal: Une
étude sur l’italien parlé. Paris: Ophrys. • Crocco, Claudia. 2013. Is Italian Clitic Right Dislocation grammaticalised? A prosodic analysis of yes/no questions and statements. Lingua 133,
30-52. • Dimroth, Christine et al. 2010. Given claims about new topics. How Romance and
Germanic speakers link changed and maintained information in narrative discourse. Journal
of Pragmatics 42, 3328 – 3344. • Scarano, Antonietta. 2003. Les constructions de syntaxe segmentée: syntaxe, macrosyntaxe et articulation de l’information. In Scarano, Antonietta (ed.),
Macrosyntaxe et pragmatique. L’analyse linguistique de l’oral, 183-201. Roma: Bulzoni. • Swerts,
Marc et al. 2002. Prosodic marking of information status in Dutch and Italian: a comparative
analysis. Journal of Phonetics 30, 629 - 654. • Turco, Giuseppina. 2014. Contrasting opposite
polarity in Germanic and Romance languages: Verum focus and affirmative particles in native
speakers and advanced L2 learners. Nijmegen: Max Planck Institute for Psycholinguistics Series.
Rejections and rejecting questions: Declaratives with
clause-initial negation in Swedish
Heiko Seeliger1 & Sophie Repp1
1
Humboldt-Universität zu Berlin
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 12:00–12:30, Raum: F 425
Declarative questions (DQs) in English have been observed to be subject
to certain contextual restrictions: The proposition that is denoted by the
190
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 191 — #203
AG 5, Raum F 425
declarative must not contradict contextual evidence: for a DQ ?p the context must not imply –p, and for ?–p the context must not imply p (Gunlogson 2003). The present paper studies negative DQs in Swedish that are
not subject to this restriction but have been claimed to require the opposite
contextual licensing, i.e. for ?–p, the context must imply p. With Seeliger
(2015) we argue that in such contexts the negated utterance can express
(I ) a rejection (which is uncontroversial; cf. Østbø 2013) or (II ) a “rejecting
question” (RQ), which expresses the speaker’s incredulity w.r.t. p.
It has been claimed that the negative marker in Swedish RQs must front
to the clause-initial, pre-verbal position (= fronted negation; FN), or alternatively low negation must combine with the modal particle (MP) väl
’surely’ (Petersson 2008, Brandtler & Håkansson 2014). Without väl, low
negation can only occur in contexts providing evidence for –p, like English
DQs. We present evidence from an acceptability study (16x4 items, 26 participants) suggesting that contexts with weak contextual evidence for p indeed license FN-RQs, irrespective of the presence of a MP, which is required
for low negation. Furthermore, we show that rejection vs. RQ readings are
distinguished by prosody for string-identical FN-sentences (NegVSO). We
present evidence from a production experiment (32 items, 9 participants)
that suggests that speakers mark the difference between rejections vs. RQs
by raising the F0max and F0mean in RQs on verb and object. This intonational difference between rejections and RQs appears to be similar to the
difference that has previously been found between Swedish assertions and
questions in general (e.g. Gårding 1979).
To account for the difference between speech acts as well as between
rejections and negative assertions, we argue with Seeliger (2015) that FN
denotes the falm operator (Repp 2009), which imposes the contextual
restrictions that are not present for negative assertions or DQs. The speech
act difference is a result of falm occurring under different speech act
operators, namely Ae in rejections and Q.Ae in RQs.
References: • Brandtler, Johan & Håkansson, David. 2014. Not on the edge. e Journal of
Comparative Germanic Linguistics 17 (2). 97–128. • Gårding, Eva. 1979. Sentence intonation
in Swedish. Phonetica 36. 207–215. • Gunlogson, Christine. 2003. True to form. New York:
Routledge. • Østbø Munch, Christine. 2013. North Germanic negation. Tromsø: University of
191
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 192 — #204
AG 5: e grammatical realization of polarity
Tromsø. (Doctoral dissertation.) • Petersson, David. 2008. Inte, nog och visst i mittält och
fundament. In Josefsson, Gunlög (ed.), Nordlund, 111–153. Lund: University of Lund. • Repp,
Sophie. 2009. Negation in gapping. Oxford: Oxford University Press. • Seeliger, Heiko. 2015.
“Surely that’s not a negative declarative question?” – Polar discourses in Swedish, German
and English. Proceedings of Sinn und Bedeutung 19. 591–609.
Prosodic realization of verum focus in English polar
questions
AG5
Anja Arnhold1 , Bettina Braun1 , Filippo Domaneschi2 & Maribel Romero1
1
Universität Konstanz, 2 Università degli Studi di Genova
[email protected], [email protected], fi[email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: F 425
Research on Verum focus in German has argued that accentuation may be
realized on different locations – on the finite verb or the COMP(lementizer)
domain – leading to meaning differences (Höhle 1992). For English, in the
realm of polar questions, Verum is argued to be realized as accent on the finite auxiliary in COMP (Romero & Han 2004) or on the lexical verb (Asher
& Reese 2007). We investigate what realization is produced in English biased questions of a certain kind and discuss why this should be the case.
A psycholinguistic experiment manipulated English speakers’ prior bias
relative to the truth of a proposition, specifically negative bias, e.g., Sorry, I
don’t have the car and neutral bias I don’t know if I have the car, before presenting them with positive contextual evidence If you want, I can give you
a li this evening. Speakers then inquired about the proposition, choosing
one of several positive or negative polarity question forms. We recorded 42
participants who completed 5 items in both bias conditions. Here, we only
investigate polar questions preceded by really, e.g., Really? Do you have the
car?.
Participants used this question form in 177 utterances (100 in negative
and 77 in neutral bias condition). Three annotators independently coded
the presence of accents in these utterances; at least two of them agreed
192
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 193 — #205
AG 5, Raum F 425
for each judgement. These annotations indicated that, irrespective of condition, participants placed the nuclear accent on the last word of the sentence in 69%. Interestingly, nuclear accents appeared on the lexical verb
more often in the negative than in the neutral bias condition (10% vs. 5%),
despite comparable information structure. Additionally, participants produced pre-nuclear accents on 47% and 29% of the lexical verbs in negative and neutral condition, respectively. By contrast, accents appeared on
less than 3% of auxiliaries in either condition. Our findings are thus in line
with predictions by Asher & Reese (2007), but do not show the accentuation described by Romero & Han (2004), possibly due to a lack of direct
contradiction to a preceding statement or because of lexical marking, i.e.,
really.
References: • Asher, Nicholas & Reese, Brian. 2007. Intonation and discourse: Biased questions. Interdisciplinary Studies on Information Structure 8, 1–38. • Höhle, Tilman. 1992. Über
Verum-Fokus im Deutschen. Linguistische Berichte Sonderhefte 4, 112–141. • Romero, Maribel & Han, Chunghye. 2004. On negative yes/no questions. Linguistics and Philosophy 27(5),
609–658.
Polarity focus across languages: Processes vs. things
Dejan Matić1 & Irina Nikolaeva2
1
Universität Graz, 2 SOAS London
[email protected], [email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: F 425
Polarity focus is commonly defined as focus on the truth value of the proposition in the reference world – actual world in comparison to other possible
worlds. Intuitively, it seems to assert that a proposition is true, often (but
not necessarily) despite expectations to the contrary. The intuition that
there are utterances in all natural languages in which the truth value is
the object of the purported semantic operation of focusing has led to the
tacit assumption that polarity focus is a discrete linguistic category based
on a common denotation, defined as a kind of truth-value operator that is
193
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 194 — #206
AG 5: e grammatical realization of polarity
AG5
placed in a certain position in the clause and realised in different ways in
the syntax and phonology in different languages.
While we do not exclude the possibility that there are indeed languages
in which the communicative task of emphasising the truth value of a proposition is carried out by means of a truth value operator of sorts, we shall
argue in the spirit of Matić & Wedgwood (2013) and Matić & Nikolaeva
(2014) that this conception of polarity focus is a poster child of the logical
fallacy of reification of processes. In short, we will try to show, drawing
on data from a wide array of languages, that there is no empirical evidence
to the claim that the emphasis on the truth value has to be achieved via
any kind of discrete denotation. Rather, this emphasis can and often is an
interpretative effect derived through the process of pragmatic inference
with the help of more general pragmatic principles from quite different
source denotations, or via composite interpretation of clues from other areas of grammar. Without attempting to be exhaustive, we identify three
possible ways of deriving polarity focus interpretations: (a) underspecification (English, Thompson River Salish, Serbian/Croatian); (b) composite
interpretation (German, Even, Udihe); and (c) unrelated source denotations
(Tundra Yukaghir, possibly Vietnamese).
We conclude that there is no evidence that polarity focus is a universal
semantic and formal category valid across all languages. There is at best
a universal communicative intention, which languages may realise with
different semantic and formal devices. Polarity focus is thus to be understood as a process to achieve a certain communicative end, not as a thing, a
discrete entity in the grammatical and semantic repertoires of the world’s
languages.
References: • Matić, Dejan & Nikolaeva, Irina. 2014. Realis mood, focus, and existential closure in Tundra Yukaghir. Lingua 150, 202-231. • Matić, Dejan & Wedgwood, Daniel. 2013. The
meanings of focus: the significance of an interpretation-based category in cross-linguistic
analysis. Journal of Linguistics 49, 127-163.
194
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 195 — #207
AG 5, Raum F 425
Cross-linguistic evidence that verum ≠ focus
Daniel Gutzmann1 , Katharina Hartmann2 & Lisa Matthewson3
1
Universität zu Köln, 2 Universität Wien, 3 University of British Columbia
[email protected], [email protected], [email protected]
Freitag, 26. 02. 2016, 12:00–13:00, Raum: F 425
The accent pattern known as verum focus is commonly understood as an
ordinary alternative focus on the truth of a proposition. This standard view,
which we call the focus accent thesis (FAT), can be contrasted with the lexical operator thesis (LOT), according to which the accent pattern that looks
like focus in languages like German or English is actually not an instance
of focus marking, but realizes a lexical verum operator, whose function is
to relate the current proposition to a question under discussion. Although
it is hard to distinguish between the FAT and LOT on the basis of German
or English alone, a broader cross-linguistic perspective seems to favor the
LOT. Drawing from fieldwork on Tsimshianic (Gitksan) and Chadic (Bura,
South Marghi), we first show that in none of these languages is verum
realized in the same way ordinary alternative focus in marked in these
languages. This sheds doubt on the unity of verum and focus and hence
speaks against the FAT. Secondly, the FAT predicts that a language cannot have co-occuring verum and focus, if it does not allow multiple foci,
and that a language should allow them to co-occur if it allows for multiple
foci. Furthermore, the FAT leads one to expect that there are focus sensitive
operators that can associate with verum focus just like they do with ordinary constituent focus. Again, while it is hard to find counterexamples in
German or English, the data from our cross-linguistic investigation again
favors the LOT. In addition to these cross- linguistic arguments, we present
data from German and English that is hard to account for with the focus
semantics required by FAT, but can easily be accommodated by a lexical
verum operator.
195
AG5
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 196 — #208
AG 5: e grammatical realization of polarity
Non canonical syntax and polarity focus in French, Italian,
and Spanish
Davide Garassino1 & Daniel Jacob1
1
Universität Freiburg
[email protected], [email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: F 425
AG5
Unlike Germanic languages, in the Romance family the prosodic realization of polarity focus is strongly restricted (Turco 2014). Instead, we observe a wealth of formal means that involve other language domains; an
interesting role is played for instance by non canonical syntactic structures such as preposed infinitive (Bernini 2009) and non focal fronting
(Leonetti/Escandell-Vidal 2009).
In this paper, we aim at providing a fine-grained analysis of the contribution of marked syntax to the expression of polarity focus in the Romance
languages on the basis of authentic data (from the EUROPARL corpus) and
explaining the differences observed between French, Italian, and Spanish.
Non focal fronting, for example, is common in Spanish, but is very limited
in Italian and French. A first corpus survey has also revealed that Italian
and French prefer to use other syntactic structures such as dislocation:
(1)
Il faut rendre honneur au président Chirac […] qui a vaincu sur sa
vision de l’Europe, parce que, lui, il a une vision (Europarl, Dupuis,
12.12.2000)
‘We should honour President Chirac who conquered for his vision
of Europe - because he did have a vision’
The main concern of this paper is thus to understand “how comparable contexts are expressed in [different] languages” (cf. call for papers). Instead of
assuming an illocution type or a verum operator, we will describe the relevant contexts according to the estion Under Discussion model (Roberts
2012) and consider the pragmatic effects of the examined structures as instances of Common Ground Management (Krifka 2007).
196
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 197 — #209
AG 5, Raum F 425
References: • Bernini, Giuliano. 2009. Constructions with preposed infinitive: typological
and pragmatic notes. In Lunella Mereu (ed.), Information structure and its interfaces, 105–128.
Berlin: Mouton de Gruyter. • Krifka, Manfred. 2007. Basic notions of information structure. In
Féry, Caroline, Fanselow, Gisbert & Krifka, Manfred (eds.), e notions of information structure 6, 13–55. Potsdam: Universität Potsdam. • Leonetti, Manuel & Escandell-Vidal, Victoria.
2009. Fronting and verum focus in Spanish. In Dufter, Andreas & Jacob, Daniel (eds.), Focus
and background in Romance languages, 155–204. Amsterdam & Philadelphia: John Benjamins.
• Roberts, Craige. 2012 [1996]. Information structure in discourse: Towards an integrated formal theory of pragmatics. Semantics and Pragmatics 5, 1–69. • Turco, Giuseppina. 2014. Contrasting opposite polarity in Germanic and Romance languages: Verum Focus and affirmative
particles in native speakers and advanced L2 learners. Ph.D. Thesis. Max Planck Institute for
AG5
Psycholinguistics, Nijmegen.
Assertion and polarity in the koso –e construction in Old
Japanese
Kyoko Sano
University of Washington
[email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: F 425
Frellesvig (2010:255) states that the koso –e construction expresses “unique
identification” and “p koso q–e” means, “It is p (and only p) that is q”. Koso is
presumably a focus particle that attaches to a phrase/clause and co-occurs
with an exclamatory conjugation –e on the verb in the main clause. I propose that the conjugation form –e is a primitive form of subordinating conjunction that carries a hidden polarity, and the negative polarity in –e gives
rise to a concessive meaning ‘although/but’. What is interesting about this
construction is that the translation picks out one of the two contrasting
implications, depending on the polarity of –e determined by the context.
When –e carries the concessive meaning as equivalent to ’but’, it has a
negative implication that “p koso q–e” is not true, and otherwise it has a
positive implication that “p koso q–e” is true. The negative implication of
the koso –e is instantiated in a context such as (1), contrasted with the emphatic use as in (2). In (1), the speaker doesn’t believe that there will be a
197
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 198 — #210
AG 5: e grammatical realization of polarity
time when “the moon disappears.” I adopt Klein (2006) and claim that the
negative implication of the koso –e is associated with “non-assertiveness”:
the negatively implicated sentences are not asserted.
(1)
Tukwi no use-na-mu
pi koso, a-ga
Moon GEN disappear-PERF-SUPP.ADN day KOSO, I-GEN
kwopwi yama-m-e.
longing stop-SUPP-EXCL
“On the very day when the moon that shines in the broad heavens
ceases to be, my affection would have come to cessation.” (Manyoshu
12: 3004; Suga 1991: Part II, 364)
(2)
Wa-ga inoti ik-ye-ru
pi ni koso, mi-ma-ku
I-GEN life live-STAT-ADN day LOC KOSO, see-SUPP-NMNL
pori
sur-e.
want.INF do-EXCL
“I would like to see my dear while I am alive.” (Manyoshu 7: 2592;
Suga 1991: Part II, 282) (EXCL = exclamatory, SUPP = suppositional,
STAT = stative)
AG5
References: • Frellesvig, Bjarke. 2010. A history of the Japanese language. Cambridge: Cambridge University Press. • Klein, Wolfgang. 2006. On finiteness. In Van Geenhoven, Veerle
(ed.), Semantics in acquisition, 245-272. Dordrecht: Springer. • Kojima, Nohyuki; Kinoshita,
Masatoshi & Tōno, Haruyuki. 1994-1996. Manyoshu. Vol 6-9. Syogakkan. • Suga, Teruo. 1991.
e Man’Yo-Shu: A complete English Translation in 5-7 Rhymes, Part II. Kanda Institute of Foreign Languages. Kanda University of Institutional Studies.
198
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 199 — #211
Advancing the European Multilingual Experience (AThEME)
Im Jahr 2013 wurde das europäische Projekt „Advancing the European Multilingual
Experience“ initiiert. Ziel der 17 Partner aus 8 Ländern ist es, Mehrsprachigkeit in Europa zu
erforschen. Die Forschungsfelder reichen dabei von der Erhaltung der L1-Sprachen bis zu
klinischen oder neurowissenschaftlichen Themen. Ein essentieller Bestandteil dieser Projekte
ist die Veröffentlichung der Forschungsergebnisse. Diese sollen sowohl der Politik und
LehrerInnen als auch der interessierten Öffentlichkeit zugänglich gemacht werden. Diese
Transparenz soll insbesondere zu einem positiven Verständnis von Mehrsprachigkeit und dem
selbstverständlichen Gebrauch von mehreren Sprachen in Familie und Bildungseinrichtungen
führen.
Im Rahmen des europäischen Projekts werden die nationalen Partner an das Netzwerk
„Bilingualism Matters“ angeschlossen. Bilingualism Matters wurde an der Universität
Edinburgh von Prof. Antonella Sorace gegründet. Schon seit Jahren herrscht in diesem
Bereich ein großer Informationsbedarf, und auch eine verstärkte Zusammenarbeit von
WissenschaftlerInnen sowie Akteuren im Bereich der Mehrsprachigkeit an der Universität
Konstanz sowie der Stadt Konstanz ist angestrebt. Die Konstanzer Version von Bilingualism
Matters ist das Zentrum für Mehrsprachigkeit, das von Prof. Dr. Janet Grijzenhout und Dr.
Tanja Rinker 2014 gegründet wurde. Das Zentrum hat es sich zum Ziel gesetzt, an der
Universität Konstanz sowie in der Stadt Konstanz und darüber hinaus als zentrale Stelle im
Bereich Mehrsprachigkeit zu fungieren. Hierzu gehören bspw. die Einrichtung eines
interdisziplinären Forschungskolloquiums „Mehrsprachigkeit“ an der Universität Konstanz,
Veranstaltungen für Familien und Bildungseinrichtungen, Beratung von Behörden und
Institutionen.
Von 2015-2017 führt das Zentrum für Mehrsprachigkeit in Kooperation mit dem Staatl.
Schulamt Konstanz, den Städtischen Kindertagesstätten in Konstanz sowie dem Italienischen
Generalkonsulat Stuttgart das Transferprojekt „Mehrsprachigkeit in Kita und Schule“ durch.
Weitere Informationen unter www.mehrsprachigkeit.uni-konstanz.de
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 200 — #212
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 201 — #213
Arbeitsgruppe 6
Computational Pragmatics
Anton Benz1 , Ralf Klabunde2 , Sebastian Reuße2 &
Jon Stevens1
1
ZAS Berlin, 2 Ruhr-Universität Bochum
[email protected], [email protected], [email protected],
[email protected]
AG6
Raum: F 420
Workshop description
Computational pragmatics can be understood in two different senses. First,
it can be seen as a subfield of computational linguistics, in which it has a
longer tradition. Example phenomena addressed in this tradition are: computational models of implicature, dialogue act planning, discourse structuring, coreference resolution (Bunt & Black 2000, and others). Second, it
can refer to a rapidly growing field at the interface between linguistics, cognitive science and artificial intelligence. An example is the rational speech
act model (Frank & Goodman 2012) which uses Bayesian methods for modeling cognitive aspects of the interpretation of sentence fragments and implicatures. Computational pragmatics is of growing interest to linguistic
pragmatics, first, due to the availability of theories that are precise enough
to form the basis of NLP systems (e.g. game theoretic pragmatics, SDRT,
RST), and second, due to the additional opportunities which computational
pragmatics provides for advanced experimental testing of pragmatic theories. As such, it enhances theoretical, experimental and corpus-based approaches to pragmatics.
201
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 202 — #214
AG 6: Computational Pragmatics
In this workshop, we want to bring together researchers from both branches of computational linguistics, as well as linguists with an interest in
formal approaches to pragmatics. Topics of the workshop include, but are
not limited to, the following issues:
• implicature calculation and its implementation in NLP systems: interaction with information structure, discourse relations, dialogue
goals etc.
• computational models of experimental results and computational systems as a means for experimental research
• corpus annotation of pragmatic phenomena
AG6
References: • Bunt, H. & Black, W. 2000. The ABC of Computational Pragmatics. In: Bunt, H. &
W. Black (eds.) Abduction, Belief and Context in Dialogue: Studies in Computational Pragmatics.;
1–46. • Frank, M. C., & Goodman, N. D. (2012). Predicting pragmatic reasoning in language
games. Science, 336(6084), 998.
Computational pragmatics revisited
Harry Bunt
Tilburg University
[email protected]
Mittwoch, 24. 02. 2016, 14:00–15:00, Raum: F 420
As a sequel to Bunt and Black (2000), which presented a characterization of
the field of computational pragmatics and a survey of its main issues, this
paper discusses some of the most interesting developments in the field in
the last 15 years. Current research is dependent on large-scale annotated
corpora. The paper includes an overview of such corpora and accompanying software tools. Of the pragmatic phenomena that have received attention in such corpora, the use of dialogue acts in spoken interaction stands
out. Dialogue acts, which have become popular for modeling the use of
202
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 203 — #215
AG 6, Raum F 420
language as the performance of actions in context, are realized by ‘functional segments’ of communicative behavior; these may be discontinuous,
may overlap, and may contain parts contributed by different speakers.
Based on the DIT++ taxonomy of dialogue acts, the ISO 24617-2 standard
for dialogue act annotation has been defined, including the Dialogue Act
Markup Language DiAML, which supports the annotation of functional
segments with multiple communicative functions, type of semantic content, speaker and addressee(s), functional and feedback dependences, pragmatic qualifiers, and rhetorical relations. The context-update semantics of
DiAML accounts for inference relations among dialogue acts.
Computational pragmatics contributes to dealing with the fundamental
challenge of pragmatics to understand how language interacts with context
by providing computational models of interpretation, generation, inferencing and learning. What is still missing, however, is the use of powerful context models. Much of the work that takes context information into account
considers only the linguistic context, i.e. the preceding discourse. This is
virtually the only kind of context information that is available in corpora,
and therefore for applying machine learning techniques. As a result only
a fraction of the relevant context information is taken into consideration.
Ideally, dialogue and discourse corpora should include information from
richer context models including e.g. speaker and hearer beliefs, mutual beliefs, communicative goals, multimodal perceptual information, and social
relations. Manual addition of this information in corpora hardly seems feasible, in view of its complexity, therefore a challenge for computational
pragmatics is the development of new computational tools to make this
feasible.
203
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 204 — #216
AG 6: Computational Pragmatics
Balancing dialogues with mixed motives
Sabine Janzen1 & Wolfgang Maaß1
1
Saarland University
[email protected] [email protected]
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: F 420
AG6
Mixed motives represent a mixture of congruent, i.e., joint motives as well
as incongruent, partially conflictive motives of interlocutors in dialogues.
Motives refer to objectives or situations that interlocutors would like to accomplish in the sense of a motivational state. As mixed-motive dialogues
we describe all grades between collaborative dialogues with exclusively
congruent interlocutors’ motives, e.g., when solving a PC problem together,
and non-collaborative dialogues with purely incongruent motives of dialogue participants, e.g., in a pro/contra debate. Adopting the idea of mixedmotive games by Schelling, we consider these dialogues as situations in
which participants are faced with a conflict between their motives to cooperate and to compete with each other, e.g., in sales conversation, where
bargainers have to make concessions to establish a compromise agreement,
but at the same time, they must compete to achieve a good bargain. In
everyday life, interlocutors are able to solve this conflict between cooperation and competition with trade-offs between selfishness and fair play
for creating dialogues perceived as fair. Despite of the overall presence of
mixed-motive dialogues in everyday life, little attention has been given to
this topic in dialogue planning in contrast to scrutinized collaborative as
well as non-collaborative dialogues. Therefore, the support of these rarely
considered dialogue type by dialogue systems in real-world environments
is still a challenge.
Our objective is the investigation of dialogue systems that support mixedmotive dialogues between users and indirect, absent interlocutors, for instance customers and retailers in sales conversations. Adopted motives by
indirect interlocutors as well as anticipated motives by users constitute
mixed motives that are processed by the dialogue system when generating
answers to posed questions. Since complete satisfaction of all motives by
all interlocutors at any point in mixed-motive dialogues is not possible, we
204
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 205 — #217
AG 6, Raum F 420
draw the concept of satisficing by Herbert Simon (1956) for capturing the
idea of finding the best alternative available in the sense of a sufficient satisfaction of motives by all interlocutors. Therefore, satisficing answers are
planned that lead to mixed-motive dialogues perceived as fair by all interlocutors regarding the absolute and relative satisfaction of their motives.
Restricted to question-answering settings, our contribution is an approach
for satisficing answer planning in mixed-motive QA dialogues by means of
a game-theoretical equilibrium approach. Based on the proposed approach,
we implemented a text-based QA system that provides a sales assistant in
an online shopping scenario. The validity of the approach was evaluated in
an empirical end user study (n=120) with the QA system with promising
results.
Interactive natural language generation in virtual
environments
Martín Villalba 1 & Alexander Koller1
1
University of Potsdam
[email protected] [email protected]
Mittwoch, 24. 02. 2016, 16:30–17:30, Raum: F 420
In this talk, we will report on our research on interactive natural language
generation (NLG) in the context of situated dialogue systems. In situated
communication, the meaning of a sentence is always relative to the environment in which it is uttered, requiring us to model both in parallel.
Within this task, we are particularly interested in generating referring expressions (REs), i.e. of noun phrases that identify a given object effectively
within the scene.
More specifically, our research on situated NLG focuses on generating
instructions that help a human user solve a given task in a virtual 3D environment. This domain has the advantage of a technical complexity and reliability that is greatly reduced compared to situated communication in reallife environments. Furthermore, data collection and evaluation can be done
with experimental subjects that are recruited over the Internet. We will report on the GIVE Challenge, an NLG evaluation challenge organized by
205
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 206 — #218
AG 6: Computational Pragmatics
AG6
our group which is built on top of this idea. Since 2009 we have developed
a set of tools capable of recording, analyzing, and modeling user behavior around this challenge scenario, and have collected hundreds of hours
of interactions between NLG systems and experiment subjects, which we
can use to train and evaluate our systems. Crowdsourcing, the practice of
collecting data from participants all over the world, allows us today to test
new hypotheses in a cheap and efficient manner.
Next to this training and evaluation setting, we will also report on our
work on the interactive, situated generation of REs. Generating REs, reacting to misunderstandings and establishing common ground are some of
the pragmatic phenomena that we must take into account. We developed a
data-driven approach that allows us to generate the “best” RE for any given
situation. Unlike some earlier research, we take “best” to mean the RE that
maximizes the chance that the listener will understand the RE correctly. We
then exploit the interactivity of the environment by tracking the listener’s
behavior in the virtual environment in real time. We have implemented
a system that detects automatically whether the listener has understood
the RE correctly, and generates corrective feedback if a misunderstanding
occurred.
Computational models of choice in language production:
The case of reference
Kees van Deemter
University of Aberdeen
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:30, Raum: F 420
In this talk I will, firstly, summarise the state of the art of the Generation of
Referring Expressions, viewed as the construction of computational models of human reference production; in this first part of the talk, I will ask
what algorithms in this area are able to do well and what it is that they
still struggle to do. In the second part of the talk, I will argue that the most
difficult problems for the Generation of Referring Expressions arise from
206
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 207 — #219
AG 6, Raum F 420
situations in which reference is something other than the “simple” identification of a referent by means of knowledge that the speaker shares with
the hearer; I will give examples of these epistemically problematic situations and of the generation algorithms that try to address them. The talk
offers a sneak preview of my book “Computational Models of Referring: a
Study in Cognitive Science”, which is soon to appear with MIT Press.
The turnip question: A game-theoretic look at non-literal
answers
Jon Stevens
ZAS Berlin
AG6
[email protected]
Donnerstag, 25. 02. 2016, 9:00–10:00, Raum: F 420
Consider the problem of generating and intepreting non-literal utterances
in the context of (1), where “Rewe” and “Edeka” are supermarkets.
(1)
Q: Does Rewe sell turnips?
A: (a) Edeka sells turnips.
(b) ?Rewe sells carrots.
(c) #Rewe sells soap.
Intuitively, (1-a) is licensed by the presumption that the questioner/hearer
wants to buy turnips, and conveying that Edeka sells them would be helpful in accomplishing this goal. But why wouldn’t the hearer have simply
asked, “where can I get some turnips?’” A strategy for answering that whquestion by breaking it down into yes/no sub-questions (see Büring 2003)
makes sense if two conditions are met. First, the questioner expects the
answerer to supply a single candidate store, rather than an exhaustive list.
Second, the questioner has a preferred outcome: perhaps for reasons of
convenience or price, he/she would rather go to Rewe for turnips. Asking about Rewe first avoids an outcome where the questioner is led to a
sub-optimal supermarket. In this case, a helpful answerer does well to supply the alternative in (1-a), but only in the case where Rewe does not sell
207
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 208 — #220
AG 6: Computational Pragmatics
AG6
turnips. With this in mind, the hearer will draw the implicature from (1-a)
that Rewe does not sell turnips. A similar implicature can be drawn from
(1-b), but one gets the intuition that (1-a) is a better answer than (1-b). And
more clearly, (1-c) is downright infelicitous. This should fall out as a direct
consequence of how (un)likely it is that the alternatives supplied help accomplish the questioner’s goal. Recently, game theory has proven to be a
useful formal tool for modeling reasoning of this kind, and has begun to
be applied to problems of language generation in a computational setting
(Stevens et al., 2015). We propose a framework for developing methods
to solve generation/interpretation tasks in parallel using iterated gametheoretic reasoning over algorithms. A discourse situation is modeled as a
cooperative Bayesian game between two interlocutors, taking into account
their conversational and domain-level goals. The strategies are algorithms
for generating and interpreting/reacting to propositions. Starting with a
principled default speaker strategy, algorithms iteratively refined to better
achieve the players’ goals until fixed point has been reached, à la Franke
(2009). Pragmatic inferences are made based on conditions on algorithm
outputs. We illustrate our approach by applying it to (1).
Embedded scalars and reasoning about QUD
Michael Franke1 & Leon Bergen2
1
University of Tübingen, 2 MIT
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: F 420
The question whether and when pragmatic enrichments, like scalar implicatures, can occur in nonmatrix position is crucial for understanding pragmatic inferences and processing in general. Here, we would like to address
the associated disambiguation problem (c.f. Chemla & Singh, 2014): any
theory of implicature-like meaning enrichments should ideally specify, for
any sentence and context pair, which candidate readings are preferred, and
to what extent even dispreferred readings may be selected.
With this goal in mind, we turn to probabilistic computational pragmatics, which aims to bridge classical formal pragmatic theory and the de-
208
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 209 — #221
AG 6, Raum F 420
mands of empirical data analysis. In particular, we look at a joint-inference
model in which the listener infers, not only the most likely world state
that could have triggered the speaker’s utterance, but also the speaker’s
intended meaning, modeled here as a topic proposition (a special kind of
question under discussion). In keeping with previous probabilistic pragmatics models that build on Frank and Goodman (2012)’s rational speech
act model, we define a chain of naive listener R0, Gricean speaker S1 and
pragmatic interpreter R2, where each next component builds on the previous. The main innovation of this model is that the speaker’s choice of utterance depends on a choice of topic proposition which in turn depends on the
actual world state. Speakers are assumed to select topic propositions probabilistically, so that more informative (surprising) propositions are more
likely to be selected. Utterances should then make the to-be-communicated
topic proposition likely, given conventional semantic meaning. Listeners
then jointly infer world state and topic proposition based on the utterance.
We show how this joint-inference model makes appealing predictions
about complex sentences with scalar implicature triggers in line with recent empirical data about preference in disambiguation (Franke et al., 2015).
We also argue that the joint-inference model offers many possibilities for
linking model predictions to experimental conditions
References: • Chemla, Emmanuel & Raij Singh (2014). Remarks on the Experimental Turn
in the Study of Scalar Implicature (Part I & II). In: Language and Linguistics Compass 8.9,
pp. 373–386, 387–399. • Frank, Michael C. & Noah D. Goodman (2012). Predicting Pragmatic
Reasoning in Language Games. In: Science 336.6084, p. 998. • Franke, Michael et al. (2015). Embedded Scalars, Preferred Readings and Intonation: An Experimental Revisit. Under review,
Journal of Semantics.
209
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 210 — #222
AG 6: Computational Pragmatics
Unusual uncertainty in language understanding:
Vagueness and accommodation
Noah Goodman
Stanford University
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:30, Raum: F 420
AG6
Probabilistic models of human cognition have been widely successful at
capturing the ways that people represent and reason with uncertain knowledge. The Rational Speech Act framework uses probabilistic modeling tools
to formalize natural language understanding as social reasoning: literal
sentence meaning arises through probabilistic conditioning, and pragmatic
enrichment is the result of listeners reasoning about cooperative speakers.
I will consider how this framework provides a theory of the role of context in language understanding. In particular I will show that when uncertainty about the speaker is included in the pragmatic inference several of
the most subtle aspects of language emerge: vagueness (in scalar adjectives
and generics) and presupposition accommodation.
Pragmatic inferencing via abstract knowledge
representation in LFG
Mark-Matthias Zymla
University of Konstanz
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: F 420
We present an extension of LFG’s Abstract Knowledge Representation (AKR)
component that integrates a model of the Common Ground (CG) and allows for the calculation of pragmatic inferences. The system uses a rule
set based on Gunglogson’s (2002) discourse model. We illustrate our implementation with respect to a set of German discourse particles. These
particles arguably contribute information that is pertinent for the CG (e.g.,
Zimmerman 2011).
210
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 211 — #223
AG 6, Raum F 420
Our pragmatic parser for dialogues uses the existing AKR framework
built on top of LFG’s syntactic architecture (e.g., Bobrow et al. (2007) and
Crouch & King (2006)) within the XLE grammar development platform.
The platform integrates an XFR rewriting system that allows for packed
rewriting of XLE’s syntactic output. It produces semantic representations
that allow for Entailment & Contradiction detection (Bobrow et al. 2007).
We extend this component to produce a semantic/pragmatic representation that is dynamically updatable for pragmatic reasoning. Our system on
the one hand enriches AKRs with pragmatically relevant information, e.g.
speaker, speech time, state of information in discourse. On the other hand,
we modified the ECD system such that it determines discourse moves (conversational actions) and accordingly modifies the AKR that represents the
discourse.
To illustrate the system we use German discourse particles to demonstrate how grammatical information interacts with pragmatic information.
Concretely, our pragmatic parser interprets the meaning that the German
discourse particles ja, doch and wohl add to utterances in discourse like
structures. We show how dynamic pragmatic inferencing takes place within
the AKR system based on the information coming from the particles.
In sum, we present an extension of a meaning component that has been
used for information retrieval and reasoning in a Question&Answer system. Our extension provides a model of the CG and allows for dynamic
reasoning about the information in the CG. Furthermore, our system provides a treatment of German discourse particles that is computationally
elegant and linguistically well motivated.
References: • Asher, N. & A. Lascarides. 2003. Logics of Conversation. • Bobrow, D. G., B. Cheslow, C. Condoravdi, L. Karttunen, T. H. King, R. Nairn, & A. Zaenen. 2007. PARC’s bridge and
question answering system. GEAF 2007 Workshop. • Condoravdi, Cleo, D. Crouch, R. Stolle,
V. de Paiva, & D. G. Bobrow. 2003. Entailment, intensionality and text understanding. Human Language Technology Conference. • Crouch, D. und T. Holloway King. 2006. Semantics
via f-structure rewriting LFG06 Gunglogson, C. 2002. Declarative questions. SALT XII. • Zimmermann, M. 2011. Discourse particles. In P. Portner, C. Maienborn, & K. von Heusinger, edt.,
Semantics. HSK 33.2
211
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 212 — #224
AG 6: Computational Pragmatics
Towards building a German legal decision corpus for
argumentation mining
Florian Kuhn
IDS Mannheim
[email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: F 420
AG6
The work in progress presented here is a contribution to argumentation
mining in the German legal text domain. Focused in this abstract is the
building of a corpus of argumentative sequences and argumentation structures of German legal decisions that will later provide models of these layers for conditional random field-based sequence labelling and tree-kernels
for structure classification. Most related works are Mochales and Moens
2011 and Stab et al. 2014. However, there is no corpus of German legal decisions available and building a gold-standard corpus of this genre will be
an important addition to all related fields of research. The data collection
has been compiled from a free online service and consists of 100 private
law decisions. For pre-processing, a genre-specific sentence tokenizer has
been trained. The annotation framework chosen for the study is Webanno
(Yimam et al. 2013).
The study divides into two subtasks: The first step is the labelling of all
argumentative sequences in the justification section of a decision document on sentence level. The second annotation task is to enrich each of the
premises with structural information on its local argumentative elements
on word token level.
Besides being part of the argumentation mining study, the corpus will
deliver valuable information for discourse related studies in the German
legal domain and can contribute to comparative studies among different
argumentative text genres.
References: • R. Mochales & M. F. Moens. Argumentation mining. Artificial Intelligence and
Law, 19(1):1–22, 2011. • C. Stab, C. Kirschner, J. Eckle-Kohler, & I. Gurevych. Argumentation
mining in persuasive essays and scientific articles from the discourse structure perspective.
Frontiers and Connections between Argumentation eory and Natural Language Processing,
212
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 213 — #225
AG 6, Raum F 420
Bertinoro, Italy, 2014. • S. M. Yimam, I. Gurevych, R. E. de Castilho, & C. Biemann. Webanno:
A flexible, web-based and visually supported system for distributed annotations. In ACL (Conference System Demonstrations), pages 1–6, 2013.
Article missing?
Eva Horch
Saarland University
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: F 420
Every morning, while reading the newspaper, we are faced with a lot of different omissions, which we do not always even realize. We read headlines
as
(1)
a.
b.
Größte Dürre seit einem halben Jahrhundert
Kampfjet in Bayern abgestürzt (zeit.de)
In the first case, we consider an (def. or indef.) article is missing, as there
is an obligatory article before singular nouns in German. In 1 b., the structure additionally lacks some copula verb, like “(Ein) Kampfjet ist in Bayern
abgestürzt”.
These kinds of ellipsis are found not only in headlines. We claim that we
can get a profile of text types on the basis of their distribution of ellipses.
Hereto, we built a corpus containing more than 10 different text types (spoken and written language) to compare the patterns. A big challenge was the
right detection and annotation of the missing elements. How can we more
or less automatically find the missing article (<art>)?
(2)
a.
b.
Geh <art> Schritt zurück!
[pos=“VVIMP”] . #a:[pos!=“ART” & #a . [pos=“NN”]
Geh <art> großen Schritt zurück!
Since the STTS doesn’t distinguish between singular and plural nouns, the
query for patterns as in 2 a. gives no satisfying output. Cases like in 2(b)
213
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 214 — #226
AG 6: Computational Pragmatics
further challenge the task. Finally, we did the annotation by hand for having a reliable annotation. Secondly, I want to discuss the different forms
of article omissions in the light of Information Theory. A central claim is
that such omissions are a way to “densify” an utterance in order to reduce
redundancy. For this purpose we train Language Models on different text
types and calculate Information Density (i.e. -log2 P(w|c)) like in (3) for
headlines.
(3)
a.
b.
AG6
“Die Stadt hat wenig Chancen”
(Total #: 43)
-log2 P(Die | <s>) = -log2 P(0.040307) = 4.6328
-log2 P(Stadt | Die) = -log2 P(0.00440483) = 7.7953
“Stadt droht durch Erosion unterzugehen”
(Total #: 73)
-log2 P(Stadt | <s>) = -log2 P(0.000309407) = 11.6582
In (3), the ID of the noun without preceding article (11.6582) is higher and
hence much more “dense” than in the case of article realization (7.7953). The
ID of the article itself is quite low (4.6328). Thus, a puzzle I want to address
in the talk is, why the article sometimes is realized instead of its omission
– and vice versa. Furthermore, which role plays the Uniform Information
Hypothesis (e.g. Jaeger 2010) here?
A further aim of this (ongoing) work is to compare the values in different
text types. The claim of a certain “profile” for each text type should be reflected in different probability values, different ID profiles respectively. The
aim is to show first profiles and to discuss further possibilities in CompPrag
– since there are some other ellipsis on hold.
214
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 215 — #227
AG 6, Raum F 420
Looking for a good laugh: Using ontologies to access
pragmatic phenomena through spoken corpora
Simon Musgrave1 , Michael Haugh2 & Andrea C. Schalley2
Monash University, 2 Griffith University
1
[email protected], m.haugh@griffith.edu.au, a.schalley@griffith.edu.au
Freitag, 26. 02. 2016, 13:00–13:30, Raum: F 420
In this paper, we use the phenomenon of ‘embarrassed laughter’ as a case
study of one approach to corpus pragmatics. We construct a set of interlinked ontologies by comparing the transcription practice of various collections of data as summarised by Hepburn and Varney (2013), making explicit
the implied knowledge underlying those transcription practices about the
characteristics of laughter which have been treated as interactionally relevant. These ontologies allow us to see the essentially combinatorial nature
of certain pragmatic phenomena and therefore also allow us to develop
strategies for searching for relevant data. We then proceed to illustrate
how such search strategies can work with the example of ‘embarrassed
laughter’. Such laughter often occurs early in an interaction (especially
first encounters) and following long pauses. We can therefore establish a
set of search criteria (laughter AND (start of interaction OR long pause) to
try to find possible instances of this phenomenon in varied collections of
data such as those which form part of the Australian National Corpus. Our
approach acknowledges the complexity of the factors which may be relevant to the identification of any pragmatic phenomenon without relying
on the prior identification of instances in any specific dataset and has the
capability to generate candidate sets of examples across varied data sets
while relying on features which are annotated in standard practice. We
suggest that looking for clusters of features which characterize pragmatic
phenomena and organizing our knowledge of the features with ontologies
constitutes a very promising approach in the field of corpus pragmatics.
215
AG6
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 216 — #228
AG 6: Computational Pragmatics
References: • Hepburn, Alexa & Scott Varney. 2013. Beyond ((Laughter)): Some Notes on
Transcription. In Phillip Glenn & Elizabeth Holt (eds.), Studies of Laughter in Interaction,
25–38. London: Bloomsbury Academic. http://www.bloomsburycollections.com/book/studies
-of-laughter-in-interaction (16 August, 2015).
Schlussbemerkungen
AG6
Anton Benz1 , Ralf Klabunde2 , Sebastian Reuße2 & Jon Stevens1
ZAS Berlin, 2 Ruhr-Universität Bochum
1
[email protected], [email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: F 420
216
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 217 — #229
Arbeitsgruppe 7
Sign language agreement revisited: New
theoretical and experimental perspectives
Barbara Hänel-Faulhaber1 , Annika Herrmann2,3 ,
Christian Rathmann1 & Markus Steinbach2
1
Universität Hamburg, 2 Universität Göttingen, 3 Universität zu Köln
AG7
[email protected], [email protected],
[email protected], [email protected]
Raum: D 406
Workshop description
Research in the last 30 years has shown that agreement in sign languages
differs in interesting ways from agreement in spoken languages (LilloMartin/Meier 2011, Mathur/Rathmann 2012). In the literature, various phenomena such as verb agreement, classifier constructions, or role shift have
been discussed under the notion ‘agreement’. On the one hand, agreement
in sign language is subject to grammatical restrictions. On the other hand,
its gestural basis and typological uniformity have questioned the grammatical status of agreement in sign language. In each sign language, we find,
for instance, similar distinctions between verbs that are lexically specified
as non-inflectional (plain verbs) and verbs that show inflection (agreement
verbs). Likewise, the system of classifiers seems to be very similar across
sign languages.
Recent experimental studies initiated a controversial debate about the
grammatical status of agreement, modality-specific properties (use of space,
body as subject) and the way agreement is processed in sign languages as
217
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 218 — #230
AG 7: Sign language agreement revisited
AG7
opposed to spoken languages (Hänel-Faulhaber et al 2014). Recent corpusbased approaches suggest that agreement marking is not obligatory (de
Beuzeville et al 2009). Concerning the origin of agreement, three phenomena are relevant: (i) Sign languages seem to have the unique property to
grammaticalize gestural elements. (ii) Plain verbs may develop into agreement verbs over time. (iii) Some sign languages have developed specific
agreement markers to fill the agreement gap with plain verbs (Pfau/Steinbach 2011).
This workshop will expand our understanding on agreement in sign and
spoken languages through different experimental, corpus-based, and theoretical approaches and addresses both well-established researchers and
young researchers. Topics to be discussed at the workshop include
• Lexical, morphological, syntactic, and semantic properties of agreement in sign languages
• Typological variation of agreement and agreement in standardized
(‘old’) as well as young sign languages and ‘village sign languages’
• The formal analysis of different phenomena related to sign language
agreement
• The grammatical status of agreement (phonological & animacy restrictions, verb type, optionality)
• Grammaticalization of agreement at the interface between gesture
and sign language
• Modality-specific and modality-independent typological aspects
• New insights from experimental and acquisitional studies on sign
language agreement
• Corpus-based analyses of agreement phenomena in sign languages
• Agreement verbs, classifiers, and role shift in complex sentence constructions and discourse
218
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 219 — #231
AG 7, Raum D 406
References: • de Beuzeville, L. et al. 2009. The use of space with indicating verbs in Australian
Sign Language: A corpus-based investigation. Sign Language & Linguistics 12, 52-83. • Hänel-Faulhaber, B. et al. 2014. ERP correlates of German Sign Language processing in deaf native
signers.BMC Neuroscience 15, 1-11. • Lillo-Martin, D. & Meier, R.P. 2011. On the linguistic status of ‘agreement’ in sign languages. eoretical Linguistics 37, 95-142. • Pfau, R. & Steinbach,
M. 2011. Grammaticalization in sign languages. In: Heine, B. & Narrog, H. (eds.), Handbook of
grammaticalization, 681-693. • Mathur, G. & Rathmann, C. 2012. Verb agreement. In: Pfau, R.
et al. (eds.). Sign language, 136-157.
Classifiers as agreement … or not?
AG7
Elena Benedicto
Purdue University
[email protected]
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: D 406
1. Goals: This talk addresses the nature of Classifiers (CLS) as an agreement
phenomenon (or not) in SLs, and the nature of potential modality effects
(or not) when compared to spoken languages (SpLs). In particular, its goals
are: (i) to specify the common formal properties of agreement and CLS;
(ii) the distinguishing formal properties [2]; (iii) the formal representation
of those properties [3]; and to initiate a comparative analysis with similar
CLS phenomena in SpLs [4]. The main claim is that agreement and CLS
share the formal syntactic AGREE operation, both in SLs and SpLs.
2. CLS as agreement?: Mathur/Rathman 2012, following Corbett 2006, attribute four components to agreement: (i) a controller (the element that
determines agreement), (ii) target (the element whose form is determined
by agreement); (iii) domain (the syntactic environment where agreement
occurs), (iv) the features involved. I will contend that agreement and CLS
only share (iii) the domain for the syntactic operation, but (i), (ii) and (iv)
are crucially different. In agreement phenomena, the controller (i) is the
DP argument, whose features determine the morpho-syntactic realization
of those features in the verb. In CLS phenomena, on the other hand, the
CLS is the controller: it assigns featural content to the DP; evidence for
219
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 220 — #232
AG 7: Sign language agreement revisited
AG7
that comes from the fact that any given DP is compatible with a range of
CLS (with the subsequent change in meaning); if the DP was the site for
those features, the result of using different CLS would be ungrammatical,
contrary to fact. The target (ii) is the complementary phenomenon to (i): in
CLS the target of feature transfer is the DP; in agreement phenomena, the
target is the verb. The features (iv) in agreement are P/#, in CLS they are
[class]. In agreement, this feature transfer takes place in a Spec-Head configuration; the domain (iv) for CLS feature exchange is also the Spec-Head.
3. Formal Representation: We follow Benedicto/Brentari2004 in conceptualizing CLS as a syntactic head introducing the internal/external argument, but reinterpret their f2/f1 as a v head and voice head, respectively
(Hale/Kayser1993-2002, Harley1995, Kratzer1996). CLS are part of a complex v/voice-head; each contains [class] features and an (uninterpretable)
D-feature. This uD-feature is valued by the D in the DP merging in the Spec
of the v/voice-head; as part of this AGREE operation, the [class] features are
reciprocally transferred to the DP (Chomsky2005-2008). The formal operation AGREE is, we claim, common to both agreement and CLS phenomena:
the uD-feature is in T in verbal agreement, and in v/voice in CLS; CLS, however, have an extra transfer of the [class] feature from v/voice to DP.
4. CLS in P’orhépea and Washo: P’orhépecha has a CLS system on both
the internal argument and the location (mirroring the locative agreement
of SLs), whereas Washo has a similar split between body-part (unergative)
and whole entity (unacusative) CLS. These two languages illustrate that
CLSs in SpLs parallel those in SLs not only in the syntactic properties they
exhibit, but also in restricting them to a subset of verbs (contrary to agreement phenomena in SLs and SpLs).
(1)
220
tatruni kirai -nuk -sti
terunukwak -rhu
beani cl:rdi -cl:spk -P3 patiok -P
‘the beans are in the patio’
P’ohpecha
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 221 — #233
AG 7, Raum D 406
Against the agreement analysis of ‘classifier’ morphemes in
sign languages
Adam Schembri
La Trobe University, Australia
[email protected]
Mittwoch, 24. 02. 2016, 15:00–15:30, Raum: D 406
In this paper, I review the notion of Corbett’s (2006) canonical agreement
and its relevance for an understanding of classifier morphemes in sign languages. A number of scholars have proposed that classifier handshapes are
a type of noun class agreement morpheme, marking agreement between
the verb and its arguments (Supalla, 1982, Gluck & Pfau, 1998, Zwitserlood,
2003; Benedicto & Brentari, 2003). This proposal, as Zwitserlood (2012)
points out, has been perhaps more controversial than the related agreement
analysis of indicating verbs, and it certainly is not as widely accepted in the
sign language linguistics literature. Zwitserlood (2012) defends the agreement analysis of classifier morphemes in signed languages by discussing
possible objections to this account. For example, several researchers have
shown that the use of specific depicting verbs containing classifier morphemes are not obligatory and that the specific handshape morpheme used
in a motion verb may vary (e.g., Engberg-Pedersen, 1993; Schembri, 2001).
The obligatory presence of agreement markers and their consistent use
are features of canonical agreement according to Corbett (2006), although
as Zwitserlood (2012) correctly notes, they are not definitive: agreement
marking may be optional in some languages (Corbett, 2006). Agreement
marking also canonically occurs on all tokens of a specific lexical category
(e.g., on all verbs, or all adjectives), but in signed languages, classifier morphemes only appear in a subset of verbs. Zwitserlood (2012) argues, building on the original analysis by Supalla (1982), that classifier handshapes
are affixes that attach to a verb root. These verb roots have no phonological
specification for handshape, she claims, and this is the reason that classifiers only occur with this subcategory of verbs. The handshape as affix and
movement as verb root analysis is, however, also controversial (McDonald, 1982; Engberg-Pedersen, 1993: Schembri, 2001), and appears to reflect
221
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 222 — #234
AG 7: Sign language agreement revisited
AG7
an attempt “…at forcing traditional morphological classification onto languages that resist such classification” (Engberg-Pedersen, 1996: 89). Most
importantly, however, it is clear that current agreement analyses of classifier morphemes assume that some semantic or formal properties of the
verb’s argument(s) controls the choice of classifier morpheme on the verb.
It has already been established in the literature, however, that the formal
or semantic properties of the verb’s arguments alone do not determine the
choice of classifier in a number of sign languages. Engberg-Pedersen (1993)
proposes that the type of motion is relevant for the choice of some classifier
morphemes in Danish Sign Language, and Schembri (2001) provides further evidence in support of this claim for Auslan. Data collected from 25
native signers of Auslan using the Supalla (1982) Verbs of Motion Production task showed that in motion events involving human and other animate
referents, the 1 handshape was preferred for linear or turning movement
types (59% of all responses), while the 2 ‘legs’ handshape was preferred
for falling, jumping, and bouncing motion events (82% of all responses).
Similar interactions between the choice of classifier handshape and other
aspects of the motion event were found for other categories of classifier
handshape. This data, combined with the other aspects of non-canonical
aspects of sign language classifier systems in terms of agreement, appears
to undermine the case for analysing classifier handshapes as agreement
markers.
The development of a RC marker from a deictic gesture in
Israeli Sign Language
Svetlana Dachkovsky
University of Haifa, Israel
[email protected]
Mittwoch, 24. 02. 2016, 15:30–16:00, Raum: D 406
The development of a RC marker from a deictic gesture in Israeli Sign Language. Demonstratives provide an important link between gesture, discourse and grammar due to their communicative function to coordinate
222
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 223 — #235
AG 7, Raum D 406
the interlocutor’s focus of attention and information flow (Arbib 2012). This
underlies their frequent cross-linguistic development into a wide range of
pronouns and morphemes (Diessel 1999). The present study provides evidence for an explicit link between gesture and grammar by tracking diachronic development of a relative clause marker in Israeli Sign Language
(ISL) relative clauses, which starts as a pointing gesture, and is grammaticalized into a clitic connecting relative and main clauses and agreeing
with referent loci. Diachronic changes were inferred from the data collected from three generations of signers. The results reveal that the behavior of demonstratives in the data varied with the signers’ ages according to
four diagnostic criteria of grammaticalization (e.g., Hopper and Traugott
1993): increased frequency, syntactic and semantic change, and phonological reduction.
Figure 1: Older signer’s pointing Figure 2: Younger signer’s RC asgesture accompanied by eye gaze
similated to the height of the
INDEX, RIDE-BICYCLE, KITE
previous sign
HOLD
GIRL EAT-ICECREAM INDEX
‘(e boy) who is riding a bike is
SWING
holding a kite’.
‘e girl who is eating ice-cream is
swinging’.
References: • Arbib, M. (2012). How the brain got language: e Mirror System Hypothesis. Oxford: Oxford University Press • Diessel, H. (1999). Demonstratives. Form, function, and grammaticalization. Amsterdam: John Benjamins. • Hopper, P.J. & E. C. Traugott (1993). Grammaticalization. Cambridge: Cambridge University Press. • Labov, W. (1963). The social motivation
of a sound change. Word 19, 273-309.
223
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 224 — #236
AG 7: Sign language agreement revisited
Pointing as seeds of directionality
Lynn Y-S Hou
The University of Texas at Austin
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: D 406
AG7
Two theories propose how “verb agreement” or directionality originates
in sign languages based on experimental data. One, agreement markers
develop from grammaticalization of pointing gestures, evolving from locatives to pronouns (Pfau, 2011). Two, verbs start as uninflected forms that
move between signer and addressee, and gradually move toward different
spatial loci in the signing space once the language allows referential loci
for verbal arguments to be established (Meir et al., 2013).
I present one hour of spontaneous data from three users of an emerging village sign language, San Juan Quiahije Chatino Sign Language, that
bears directly on both theories. This language co-opts absolute pointing
gestures, which includes points to real-world and geographical locations
of referents. Data reveals 58 tokens of verb candidates for directionality:
gie (n=28), alk (n=7), ee (n=13), ph (n=10). Most tokens point to realworld locations of referents (1,2). They omit agents and prefer to mark the
goal/recipient even if it is first-person (2). They also seldom mark absent
agents and are never localized as arbitrary loci or nominal arguments in the
signing space (2). They can mark the residences of absent goals/recipients
(3).
(1)
IX:LOC[bag of lollipops] LOLLIPOP Ø.GIVE.2 NEG:WAG-1 IX:LOC
[bag of lollipops]
“The bag of lollipops, (she) is not going to give it to you.”
(2)
NAME-SIGN[RE] IX:PRO3[brown puppy] Ø.GIVE.1
“Regina gives it(=the brown puppy) to me.”
(3)
IX:PRO3[black puppy] NAME-SIGN[RE] Ø.GIVE.3
“Regina gives it(=the black puppy) to him(=the neighbor).”
224
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 225 — #237
AG 7, Raum D 406
Analysis of the locations of directional verbs reveals that the language favors deictic pointing for pronominal reference and does not deploy abstract
pointing. I propose that directionality buds from deictic pointing gestures
and can exhibit spatial modifications based on topographic space, resembling locative but not verb agreement. Pointing gestures enter the grammar
of a language, carrying pronominal, locative, and demonstrative functions.
The points subsequently evolve in varying degrees of conventionalization
with respect to handshape, movement, and syntactic distribution. When
points and nominal arguments are used to establish arbitrary loci for verbal arguments, thereby enabling anaphoric reference, the gestural roots of
directionality branch out to a more abstract system.
References: • Meir, I., Padden, C., Aronoff, M. & Sandler, W. (2013). Competing iconicities in
the structure of languages. Cognitive linguistics 24, 309–343. • Pfau, R. (2011). A point well
taken: On the typology and diachrony of pointing. In: D.J. Napoli & G. Mathur (eds.), Deaf
around the world: e impact of language. pp. 144–163. Oxford: Oxford University Press.
Role shift is not agreement
Kearsy Cormier
University College London
[email protected]
Mittwoch, 24. 02. 2016, 17:00–17:30, Raum: D 406
Sign languages are known to make use of a representational device where
one or more bodily articulators (including the head, face, eyegaze, arms
and torso) are used to represent the utterances, thoughts, feelings and/or
actions of one or more referents - known variably as constructed action
(Metzger, 1995) or role shi (Padden, 1986). The use of nonmanual markers
for this purpose has been considered by some to be instantiations of grammatical agreement. For example, Kegl (1995) described what she called a
role prominence marker – specifically a role prominence clitic. She proposed that these nonmanual features act as a subject clitic, that the subject
225
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 226 — #238
AG 7: Sign language agreement revisited
AG7
NP agrees with this clitic, and that is interpreted with role prominence
such that it marks the person from whose perspective the event is viewed.
More recently, Hermann and Steinbach (2012) have argued that nonmanual markers including eyegaze change, head position, body lean and/or
facial expression act as grammaticalised agreement markers. They argue
that “role shift does not agree with syntactic arguments but with higher
level semantic entities, namely the signer and the addressee of a reported
utterance” (p. 221). Additionally they argue that role shift is fundamentally
different from comparable uses of gesture and eyegaze for enactment in
non-signers.
Here I call upon typological work on agreement by Corbett (2006) to argue against role shift as agreement. Corbett (2006) argues that agreement
must involve systematic co-variance between a semantic or formal property of one element (the controller) and a formal property of another (the
target). In addition to a lack of systematicity in nonmanual marking, as
noted by Hermann and Steinbach, role shift “does not agree with syntactic
arguments”; therefore it simply cannot be agreement because the signer
and addressee of a reported utterance do not form part of the grammar of
sign languages.
Instead, I argue that constructed action is essentially demonstration, in
the sense of Clark and Gerrig (1990), and shares many features with nonmanual enactment used by non-signers used in face-to-face communication, as shown by work that documents the coordination of non-signers’
use of prosody, gesture and eyegaze along with syntactic resources when
producing reenactments and quotations (e.g., Sidnell, 2006). The most overt
cases of constructed action, where the signer uses both manual and nonmanual articulators to represent a referent, are the most similar to (also
overt) cases of constructed action used by non-signers (Cormier, Smith, &
Sevcikova, in press). Differences in signers and non-signers arise in less
overt constructed action which involve simultaneous coordination of lexical and partly-lexical material along with the non-lexical demonstration –
resulting in multimodal constructions in speakers but unimodal constructions in signers that can diverge in various ways. But, I argue, such differences do not on their own justify a grammatical/agreement analysis of
constructed action/role shift unique to sign languages.
226
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 227 — #239
AG 7, Raum D 406
References: • Clark, H. & Gerrig, R. J. (1990). Quotations as demonstrations. Language, 66(4),
764-805. • Corbett, G. (2006). Agreement. Cambridge: Cambridge University Press. • Cormier,
K., Smith, S. & Sevcikova, Z. (in press). Rethinking constructed action. Sign Language & Linguistics. • Hermann, A. & Steinbach, M. (2012). Quotation in sign languages. A visible context
shift. In: I. v. Alphen & I. Buchstaller (Eds.), otatives: Cross-linguistic and Cross-disciplinary
Perspectives (pp. 203-230). Amsterdam: John Benjamins. • Kegl, J. (1995). The Manifestation
and Grammatical Analysis of Clitics in American Sign Language. Chicago Linguistic Society,
31(2), 140-167. • Metzger, M. (1995). Constructed dialogue and constructed action in American
Sign Language. In C. Lucas (Ed.), Sociolinguistics in Deaf Communities (pp. 255-271). Washington, DC: Gallaudet University Press. • Padden, C. A. (1986). Verbs and role shifting in American Sign Language. In C. Padden (Ed.), Proceedings of the Fourth National Symposium on Sign
Language Research and Teaching (pp. 44-57). Silver Spring, MD: NAD. • Sidnell, J. (2006). Coordinating gesture, talk, and gaze in reenactments. Research on Language and Social Interaction,
39(4), 377-409.
Pointing to the analysis of personal pronouns and
directional verbs in the acquisition and grammar of ASL
Richard P. Meier
University of Texas at Austin
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:30, Raum: D 406
Pointing gestures and signs are ubiquitous and transparent, yet yield surprising insights for linguistics and psycholinguistics. English-speaking children with autism avoid personal pronouns in favor of names. When asked
to identify a picture of themselves or an experimenter, native-signing deaf
children with autism (n = 15) avoid the pointing signs me and o in favor
of names (Shield, Meier, & Tager-Flusberg 2015). This result suggests the
opacity of spoken pronouns cannot account for their avoidance by hearing
autistic children.
Meier (1990) argued that pointing signs in ASL are organized grammatically, with first- and nonfirst-person being distinguished. Cormier, Schembri, & Woll (2013) note that evidence for distinctive first-person singular
227
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 228 — #240
AG 7: Sign language agreement revisited
AG7
pronouns is scant. Lynn Hou and I have elicited first-person object verb
forms. Sixteen native or near-native Deaf signers participated in a translation task targeting 73 directional verbs, yielding 1116 contexts for firstperson object verb forms. In 72%, a first-person object form was produced.
We treat only the morphology of the elicited verb forms; we do not analyze
the syntax of our participants’ responses.
Following Liddell (2003), Hou and I consider the height of first-person
object forms; we find evidence for at least a two-way distinction between
chest-height and face-height directional verbs; some face-height verbs (e.g.,
ano) may lower to a default location at the center of the upper chest.
One class of directional verbs (e.g., ell, infom, hono, ignoe) have
initial contact on the face. These face-anchored verbs make final contact
on the upper chest in their first-person object forms. The verbs callb
phone, conince, and emind are unique in having first-person object
forms with final contact on the body at a location other than the center
of the upper chest. For the stimulus targeting conince, 10 participants
produced the two-handed symmetrical form of conince; all produced it
with final contact on the neck or collarbone. Lastly, certain verbs (each,
inie, offe) may be two-faced, in that different hand parts face first
and non-first objects. On such evidence, we argue that some first-person
object verb forms are listed in the lexicon of ASL. This supports the claim
that there is a first/non-first distinction in the grammar of ASL.
Explaining the special typological properties of sign
language verb agreement
Irit Meir
University of Haifa, Israel
[email protected]
Donnerstag, 25. 02. 2016, 9:00–9:30, Raum: D 406
It is well-known that sign language verb agreement (SLVAgr) is characterized by certain properties that distinguish it from verb agreement systems
in spoken languages (Aronoff et al 2005, Lillo-Martin & Meier 2011, Mathur
228
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 229 — #241
AG 7, Raum D 406
& Rathmann 2012 among many others). Here we will address the following properties: 1. Only in sign languages verb agreement is restricted to a
sub-set of the verbs in the lexicon. 2. This sub-set is basically defined on
semantic basis (verbs of transfer inflect for agreement). 3. SLVAgr is very
similar across different sign languages. 4. SLVAgr encodes both thematic
relations (source-goal) and syntactic roles (subject and (dative) object). 5.
Agreeing forms are not obligatory; even in ASL, signers often use the citation form or single-agreeing forms for double-agreeing verbs. There is
great variation in the use of fully agreeing forms among different singers,
including differences among generations.
By studying the different stages in the diachronic development of SLVAgr
in Israeli Sign Language (ISL) across three generations, these unique typological properties can be accounted for. The diachronic study shows that
agreement verbs developed from plain verbs, by re-analyzing first the final
location of verbs of transfer, the location away from the signer’s body, and
then the initial location, as morphemes encoding the referential index of
the verbs’ source/subject goal/object agruments (Meir 2012). (For TAKEtype verbs, the locations and the association of the thematic and grammatical roles are reversed). Crucially, what makes this re-analysis possible is
that verbs of transfer share a formational element: they all consist of a path
movement whose one end is near the signer’s body and the other is in the
signing space.
Based on these diachronic findings, properties 1-3 are attributed to the
gestural origins of sign languages, specifically the way in which the notion of transfer is expressed in gesture: the movement of the hands from
the signer outwards, towards an imaginary addressee (or inwards towards
the signer in TAKE-type verbs). It is the ‘loose’ end of verbs of transfer that
lends itself to re-analysis as an agreement marker. Verbs of transfer, then,
share a particular form and a meaning component (transfer from one possessor to another), and are subject to the same re-analysis. This accounts
for the fact that only verbs of transfer agree (at least in earlier stages of the
language), for the semantic basis of this class, and for the cross-linguistic
similarity, as the notion of transfer is expressed very similarly in various
manual-gestural systems. Property 4 is account for by the visuo-spatail
modality, and its ability to encode more information simultaneously. The
229
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 230 — #242
AG 7: Sign language agreement revisited
thematic tier is encoded by the movement of the verb and the syntactic
tier by the facing of the hands (Meir 1998). These mechanisms are separate,
yet they are expressed simultaneously. Finally, property 5 is attributed to
the young age of sign languages (cf. Aronoff et al 2005). SLVArg systems
did not have enough time to fully stabilize and grammaticize. Hence older
generations tend to use less fully agreeing forms, but even in the younger
signers’ system, verb agreement is not fully stable and obligatory, showing
that the obligatoriness usually associated with inflectional categories is a
result of continuous use and conventionalization.
AG7
References: • Aronoff, M., I. Meir & W. Sandler. 2005. The paradox of sign language morphology. Language, 81, 301- 344. • Lillo-Martin, D.& Meier, R. 2011. On the Linguistic Status
of ‘Agreement’ in Sign Languages. eoretical Linguistics 37, 95-141. • Mathur, G. & C. Rathmann. 2012. Verb agreement. In Pfau, R., Steinbach, M. & B. Woll (Eds.): Handbook on Sign
Language Linguistics. Mouton De Greuter, 136-157. • Meir, I. 1998. Syntactic-Semantic Interaction in Israeli Sign Language verbs: The Case of Backwards Verbs. Sign Language and
Linguistics 1, 3-33. • Meir, I. 2002. A cross-modality perspective on verb agreement. Natural
Language and Linguistic eory, 20, 413-450. • Meir, I. 2012. The evolution of verb classes and
verb agreement in signed languages. eoretical Linguistics, 38(1-2), 145 – 152.
“Agreement” verbs in sign languages: Are we missing the
point?
Adam Schembri1 , Jordan Fenlon2 & Kearsy Cormier3
1
La Trobe University, Australia, 2 University of Chicago, 3 University
College London
[email protected], [email protected], [email protected]
Donnerstag, 25. 02. 2016, 9:30–10:00, Raum: D 406
Indicating verbs (or ‘agreement’ verbs, see Padden, 1988) in sign languages
have been the subject of much debate. These verbs, such as GIVE in British
Sign Language, can be directed towards present referents or locations in
space associated with absent referents. Some scholars (e.g., Lillo-Martin &
Meier, 2011) have argued this modification is obligatory (at least for object
230
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 231 — #243
AG 7, Raum D 406
marking), is fundamentally the same as grammatical agreement in spoken languages, is often accompanied by grammatical non-manual markers
such as eye-gaze (Neidle et al., 2000), and typically make arbitrary use of
the signing space (e.g., Poizner et al., 1987; Emmorey et al., 1995; Barberà,
2014). Others (e.g., Corbett, 2006; Cysouw, 2011) propose that this modification is fundamentally different from both canonical and non-canonical
agreement, with Liddell (2000) claiming specifically that it represents a fusion of both morphemic and deictic gestural elements. Still others recognise
a role for deictic gesture while nevertheless maintaining grammatical person and/or agreement marking analysis (e.g., Mathur & Rathmann, 2010,
2012; Lillo-Martin & Meier, 2011). To move the debate forward, more data
are needed about how these verbs are structured and used.
Here we consider linguistic and social factors in the use of 1679 indicating verbs in the BSL Corpus conversation data (101 participants in 4
cities) (Schembri et al., 2013). These verbs were annotated in ELAN for linguistic factors including (1) whether the verbs’ directionality was modified or not, (2) participant roles for indicating verbs which are modified,
e.g. 1st -to-2nd person, 2nd -to-1st , 3rd -to-3rd , etc., (3) the path of the verbs’ directionality (body-sagittal, body-diagonal, or side-to-side) and (4) presence
of constructed action (i.e. whether non-manual enactment co-occurs with
the manual verb sign or not). We also consider other factors including presence or absence of verbal arguments, animacy of arguments, and whether
referents of these arguments were coreferential with the preceding clause.
Results reveal that modification of indicating verbs occurs for both subject (801/1228 = 65%) and object arguments (926/1450 = 64%), but is not
obligatorily for either. Furthermore, modification of verbs between two locations in space, often assumed to be prototypical for 3rd -to-3rd person,
is rare, occurring only 9 times in our data. (Examples of prototypical indicating verbs in the literature involve explicit establishment of reference
via pronouns/determiners and 3rd -to-3rd person modification (e.g. JOHN
POINTa MARY POINTb a ASKb “John asked Mary”)). We compare this with
the much higher incidence of side-to-side marking in 3rd -to-3rd person contexts found in elicited data. Additionally, both overt (marked via multiple
articulators) and subtle instances of constructed action (i.e. gaze towards
the location associated with the verb’s argument in space) consistently pre-
231
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 232 — #244
AG 7: Sign language agreement revisited
dicted modification of indicating verbs for patient arguments (p <0.001).
In this paper, we will discuss why these findings overall provide new
support for Liddell’s (2000, 2003) claim that indicating verbs do not constitute a grammatical agreement system (canonical or otherwise) but instead
are a typologically unique construction, involving a fusion of gestural and
morphemic elements, which is primarily used as a reference-tracking system in sign languages.
Backward agreement is not so backward after all: the role
of loci in the grammar of SL
AG7
Mirko Santoro1 , Lara Mantovan2 , Valentina Aristodemo1 & Carlo Geraci1
1
CRNS, Institut Jean-Nicod, Paris, 2 Ca’ Foscari University, Venice
[email protected], [email protected], [email protected],
[email protected]
Donnerstag, 25. 02. 2016, 10:00–10:30, Raum: D 406
Baground: The analysis of directionality as agreement (sensitivity to
number) as in Lillo-Martin & Meier (2011) leaves two open issues: 1) backward verbs, and 2) unaccusative verbs. The former exhibit an inverse pattern (apparent object-to-subject directionality, cf. (1)), the latter show sensitivity to number for their subject loci(cf. (2)). Examples are from Italian
Sign Language (LIS).
(1)
Ix-3 picture copypicture → ix-3
‘He copied the picture’
(2)
(Everybody) leave+++/arrive+++/etc. ‘Everybody left/arrived, etc.’
Goals: i) Show the effect of directionality in LIS; ii) Show that backward
verbs are unaccusative/ pseudopassive forward agreeing verbs; iii) Propose
a unified analysis in which (1) and (2) involve agreement/clitic incorporation; and iv) Frame the role of loci in the grammar of SL.
Data: Corpus and elicited data from LIS are presented. Corpus data show
the effect of verb type in licensing null subjects in LIS. Specifically, agreeing
and spatial verbs equally license null arguments more likely than plain
verbs (mixed-model analysis). Elicited data show that the (only) argument
232
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 233 — #245
AG 7, Raum D 406
of unaccusative verbs and the first locus of backward verbs are sensitive to
number (c.f. (2) and (3)), like object loci (cf. Lillo-Martin & Meier 2011).
(3)
aignmen cop+++ABC → ASSIGNMENT ‘The assignment is copied
from many sources’
Analysis: 1) Both unaccusative and backward verbs involve agreement.
2) Backward verbs are actually unaccusative/pseudopassive transitive forward agreeing verbs. Standard unaccusative verbs show agreement with
the underlying object (hence the surface subject is sensitive to number like
object loci as in (2)). The same happens with the first locus of backward
verbs, which show forward agreement between the underlying object and
a secondary object but crucially not with the external argument (i.e. the
agent –for a similar proposal see Meir 2002).
e big picture: The role of loci in the syntax of SL is discussed by looking
at its effects in weak crossover, sentential center embedding, comparatives,
and the spatial linearization process. All these facts will be used to help
disentangling the proper analysis between agreement and clitic incorporation of directional verbs (Nevins 2011). The fact that loci are active “beyond agreement” suggests that these should be analyzed as clitic elements,
at least in LIS. Whether the combination of loci and directional movement
may count as agreement is also a possibility.
Regular and backward agreement verbs in Libras: A
case-based derivation
Guilherme Lourenço
Federal University of Minas Gerais, Brazil
[email protected]
Donnerstag, 25. 02. 2016, 10:30–11:00, Raum: D 406
This work aims at presenting a syntactic derivation for regular agreement
verbs (RAV) and backward agreement verbs (BAV). The terminology of
the Generative Theory will be adopted. Assuming the Case-Dependency of
233
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 234 — #246
AG 7: Sign language agreement revisited
AG7
Agreement Parameter (Baker 2008), the proposal is that in Libras (Brazilian Sign Language) the agreement pattern of the verb will depends on the
Case received by the DPs during the syntactic derivation. Once there is no
Case marker in the language, we consider that the agreement relations will
indicate the Case borne by each nominal. Libras is a nominative language
(Quadros 1999). So, subjects are marked with nominative Case and objects
receive accusative Case. Considering RAVs, we can claim that the first slot,
which agrees with the subject, actually agrees with the nominative argument of the sentence. On the other hand, the second slot agrees with a
non-nominative argument, to wit the accusative argument. Thus far, the
claim is that RAVs show a nominative array of agreement, in which the
nominative subject grees with the ϕ-probe on T, externalized on the first
agreement slot. The accusative object agrees with the ϕ-probe on v (marked
on the second agreement slot of the verb). In BAVs, the first agreement slot
of the verb agrees with the object of the sentence, while the second agreement slot agrees whit the subject. Considering that the first slot agrees with
a nominative DP, the claim is that the object is marked with nominative
Case. So, now we have to identify the Case of the subject. By presenting
some tests proposed by Woolford (2006), we claim that the subject receives
inherent ergative Case. We also propose that in Libras there is a movement
of v to T. However, we claim that this v-to-T movement makes the in situ
object visible to T and available to receive nominative Case. This claim is
consistent with the notion of phase extension (Den Dikken, 2007). We also
propose this movement, because the subject enters in an Agree relation
with the ϕ-probe in v, which is only possible if v is higher in the tree than
the subject. Moreover, it is important to say that in Libras there is no Vto-T movement (Quadros, 1999). Finally we need to explain why it is the
ergative subject that moves to Spec,TP in order to satisfy the EPP requirement. Again, the explanation relies on the v-to-T movement. According to
Chomsky (2008), Miyagawa (2010) and others, what triggers the EPP is the
agreement between T and a DP. However, when v moves to T, the ϕ-probe
of v is moved along to that position. So, this new complex head (v+T) has
two ϕ-probes and, therefore, two Agree relations with two different DPs:
the subject and the object. Thus, both the subject and the object are available to move to Spec,TP to satisfy the EPP. The subject is the one who
moves because of the Minimal Link Condition.
234
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 235 — #247
AG 7, Raum D 406
References: • Baker, M. (2008). e syntax of agreement and concord. Cambridge: Cambridge
University Press. • Chomsky, N. (2008). On phases. In Robert Freidin, Carlos Otero, & Maria
Luisa Zubizarreta, eds., Foundational issues in linguistic theory. Cambridge, Mass.: MIT Press,
p. 133–166. • Den Dikken, M. (2007). Phase Extension. Contours of a Theory of the Role of Head
Movement in Phrasal Extraction. eoretical Linguistics 33, p. 1-41. • Miyagawa, S. (2010). Why
Agree? Why Move? Unifying Agreement-based and Discourse Configurational Languages.
MIT Press, Linguistic Inquiry Monograph 54. • Quadros, R. M. (1999) Phrase structure of Brazilian sign language. Tese de Doutorado. PUCRS. Porto Alegre. • Woolford, E. (2006). Lexical
Case, inherent Case, and argument structure. Linguistic Inquiry 37, p.111–130.
Agreement in Sign Languages, allow me to disagree
AG7
Antonio Balvet1 & Brigitte Garcia2
1
Univ. Lille, CNRS, 2 Univ. Paris 8
[email protected] [email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: D 406
In this presentation, we will argue in favor of an iconic and spatial approach to the description and formalization of a wide range of phenomena
in SLs, among which “agreement”. We will show that “agreement” in SLs
has to do mostly with “inflectional person marking” (Cysouw, 2011), and is
essentially devoted to thematic roles instantiation rather than true inflection. We will follow the semiological model (Cuxac, 2000; see also Sallandre
& Garcia 2013; Garcia & Sallandre 2014), which integrates Lexematic Units
(established signs), Transfer Units (e.g. classifier constructions/depicting
verbs) as well as agreeing/directional/indicating verbs into a unified iconic
view of SLs. By putting iconicity and spatial grammar at the forefront, and
by recognizing the influence of macro-structure (text structure, enunciation acts, signing strategy) on the realization of individual signs, the semiological model, in our view, accounts for “agreement” in SLs in a consistent
and economic way. Moreover, the semiological model accounts for “agreement” even in non conventional units (Transfer Units).
References: • Cuxac, C. (2000). La langue des signes française (LSF). Les voies de l’iconicité.
Faits de Langues, 15-16. Paris: Ophrys. • Cuxac, C., Sallandre, M-A. (2007). Iconicity and ar-
235
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 236 — #248
AG 7: Sign language agreement revisited
bitrariness in French Sign Language: Highly Iconic Structures, degenerated iconicity and diagrammatic iconicity. In Pizzuto, E., P. Pietrandrea, R. Simone (eds.): Verbal and Signed Languages: Comparing Structures, Constructs and Methodologies. Berlin: Mouton de Gruyter, pp.
13–33. • Cysouw, M. (2011). Very atypical agreement indeed. eoretical Linguistics, (3-4), pp.
153–160. • Garcia, B., Sallandre, M.-A. (2014). Reference resolution in French Sign Language. In
Cabredo Hofherr, P., & Zribi-Hertz, A., (eds.), Crosslinguistic studies on Noun Phrase structure
and reference. Syntax and semantics series, volume 39. Leiden: Brill, pp. 316–364. • Sallandre,
M-A. and Garcia, B. (2013). Epistemological issues in the semiological model for the annotation of sign language. In L. Meurant, L., Sinte, A., Van Herreweghe, M., & Vermeerbergen,
M., (eds.), Sign Language research, uses and practices, Crossing views on theoretical and applied
sign language linguistics (Sign Language and Deaf Communities), Berlin/Boston: Mouton De
AG7
Gruyter and Nijmegen: Ishara Press, pp. 159–177.
A place for locative agreement is sign languages
Josep Quer
CREA-Universitat Pompeu Fabra
[email protected]
Donnerstag, 25. 02. 2016, 12:00–13:00, Raum: D 406
The nature of verb agreement in sign languages is still controversial among
researchers (see, for instance, Lillo-Martin & Meier 2011 and commentaries, or Costello 2016 for a very recent take on this issue). For the defenders of its grammatical nature, a great deal of the discussion has revolved around the adequacy of Padden’s (1983/1988) seminal proposal of
verb classes in ASL. Despite its importance for subsequent research, the
strict tripartite division of sign language verbs into plain, (person) agreement and spatial verbs as distinct morphosyntactic categories has proven
empirically inadequate. In view of the existence of verbs that partake in
the two types of agreement, Quadros (1999) and Quadros & Quer (2008),
for example, propose to get rid of the separation between person agreement
verbs and spatial/locative agreement verbs, and to tackle them as a single
class: the “types” of agreement reduce to the type of feature a referential locus (and its linked argument) is associated with (person or locative feature),
236
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 237 — #249
AG 7, Raum D 406
and verbal roots enter one type or the other depending on their lexical semantics, the arguments they combine with and their morphophonological
makeup.
With few exceptions (e.g. Janis 1992, 1995), spatial (lexical) verbs and
spatial/locative agreement have received relatively little attention in the
aforementioned discussion, since spatial agreement seems to come “for
granted” in the signed modality. However, no explanation has been offered for the striking fact that spatial verbs agree with their locative arguments, and not with the (personal) subject when they cooccur, as in
MONTH ̂NEXT IX1 LONDONa NEW-YORKb aFLYb ‘Next month I’ll fly
from London to New York.’ Janis (1995: 219) simply states the generalization in her system for ASL agreement: “A nominal with locative case can
control agreement regardless of its G[rammatical] R[elation], S[emantic]
R[ole], or animacy features.”
In this talk I will offer a possible line of analysis of spatial/locative agreement in sign languages by drawing a link to the cases of locative agreement
that we find in some spoken languages, like those of the Bantu family. In
Chichewa, for example, locative NPs cannot only function as locative adjuncts, but they can also occur as grammatical subjects, as in (1) (Bresnan
& Kanerva 1989: 2), triggering class agreement on the verb (class 17 agreement in this case).
(1)
ku-mu-dzi ku-liu chi-tsîme
17-3-village 17:SU-be 7-well
‘In the village is a well.’
I explore the validity of the parallelism with such cases of locative agreement, as well as its limitations, in order to understand the common core
between the two language modalities.
237
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 238 — #250
AG 7: Sign language agreement revisited
Two agreement markers in Austrian Sign Language (ÖGS)
Julia Krebs1 , Ronnie Wilbur2 & Dietmar Roehm1
1
University of Salzburg, 2 Purdue University
[email protected], [email protected], [email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: D 406
AG7
In many sign languages, different forms of markers have been described
showing agreement in transitive constructions involving uninflected verbs
and even inflected agreeing verbs (e.g. Steinbach & Pfau 2007, Sapountzaki
2012). In ÖGS there are two agreement marker signs: AgrM-PERSON and
AgrM-MF.1 AgrM-PERSON, phonologically very similar to the German
Sign Language agreement marker (PAM), is produced with “baby-C”-handshape. AgrM-MF is signed with forward pointing middle finger. Both show
path movement from subject to object position, and facing towards the object. We conducted an online questionnaire which tested 1) the preferred
syntactic position of the two agreement markers (pre- vs. post-verbal), and
2) the possibility of combining AgrM-PERSON with different verbs (e.g.
regular/irregular agreeing verbs). The first experiment revealed that, although both agreement markers may appear before or after the verb, they
are slightly preferred in pre-verbal position. Prior reports on agreement
markers have indicated frequency of preverbal and sentence final position,
but not preference. The second experiment showed that AgrM-PERSON
may be combined with different verbs. In particular, whereas sentences
in which agreeing verbs alone mark agreement were preferred over structures in which an additional agreement marker appeared, double agreement seems to be acceptable in ÖGS. However, this statement can only be
made 1) for structures in which regular agreeing verbs co-occur with an
agreement marker that shows a phonologically reduced path movement
and 2) with respect to constructions in which AgrM-PERSON appears in
1 AgrM
is the abbreviation for agreement marker. The second part of the glosses refers to
the phonological form of both signs. In more detail, AgrM-PERSON is phonologically
very similar to the ÖGS sign PERSON (person) and AgrM-MF is produced by a forward
pointing middle f inger.
238
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 239 — #251
AG 7, Raum D 406
post-verbal position. The option of a reduced path on AgrM and the possibility of double agreement, of which both have been previously reported
for other sign languages, have not been described for ÖGS so far. Also interesting is that, in case of the basic sign order (SOV), plain verbs in ÖGS
can but do not have to be accompanied by an agreement marker, supporting previous claims that agreement markers are not simply ‘agreement gap
fillers’. This study does not only provide information about the syntactic
structure of ÖGS, but also about typological variation within the agreement
systems of sign languages.
References: • Sapountzaki, G. (2012). Agreement auxiliaries. In: Sign language. An international handbook, Pfau R., Steinbach M., & Woll, B. (eds.), 204-227. Berlin: Mouton de Gruyter.
• Steinbach, M., & Pfau, R. (2007). Grammaticalization of Auxiliaries in Sign Languages. In:
Visible Variation: Comparative Studies on Sign Language Structure, Perniss, P., Pfau, R., & Steinbach, M. (eds.), 303-339. Berlin: Mouton de Gruyter.
Verb-argument agreement and word order in SZJ
ditransitives
Matic Pavlič
Ca’ Foscari, Venice
[email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: D 406
There is not much work carried out on sign language (SL) ditransitives although ditransitive examples regularly appear in SL literature. From Padden (1988) on, the agreeing verb class is usually illustrated by a distransitive verb gie. Agreeing transitives verbs start in r-locus associated with
Subject and end in r-locus associated with Direct Object (Od ). Ditransitive
verbs are agreeing by default. They start in r-locus associated with a Subject and end in r-locus associated with an Indirect Object (Oi ). In many
SVO SL, ditransitives with non-classifier predicate display SVOd Oi . To my
knowledge, the only study that examines word order (WO) in ditransitives
with classifier predicate is Sze (2003). She reports that they display SOd VOi
239
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 240 — #252
AG 7: Sign language agreement revisited
AG7
WO in HKSL but she does not discuss their verb-argument agreement. On
a small corpus of SVO Slovenian Sign Language (SZJ) I repeat her results
and go on to link ditransitive WO to the ditransitive agreement pattern.
In (1), all three arguments (S, Od and Oi ) are aligned so that Oi is signed
in signing space (SS) on the ipsilateral side, Od is signed in neutral signing space (NSS) while the Oi is signed in SS on the contralateral side. A
ditransitive non-classifer predicate (1a) is signed in between the Subject
and Od ; it agrees with the Subject and Oi . A ditransitive classifer predicate
(1) is signed in between Od and Oi ; it agrees with Od and Oi . Od does not
seem to be included in manual verb-argument agreement of SZJ ditransitive non-classifer verbs. Note, however, that it signed more to the more
to the ending point of non-classifer predicates (and more to the starting
point of the classifer predicates). Now, observe an example of transitive
verb kick in (2).
It starts in r-locus associated with Subject and
ends in r-locus associated
with Od in front of the
signer; it agrees with the
Subject and Od . But then
a sign with pointing handshape and arc movement
is added. It connects Od rlocus with Oi r-locus and
functions as an overt Applicative head (it is not an auxiliary verb because
it only introduces applicative arguments and it is positioned in between
the verb and Oi ).
Finally, if the subject is firtst person, SZJ ditransitive verb such as kick
(3a) agrees with Od and Oi and is also placed in between these arguments.
240
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 241 — #253
AG 7, Raum D 406
This fact offers a new interpretation of SZJ backwards verbs. As noted by
De Quadros and Quer (2008), backwards verbs represent a curious case because in many SL they seem to be the only transitive verbs within the group
of otherwise ditransitive agreeing verbs. What if they are ditransitive? SZJ
data suggests that Subject and Oi of backwards verb have the same referent. Crucially, they agree with Od and Oi , starting at the former and ending
at the latter (3b).
AG7
References: • Padden, C. (1988). Interaction of Morphology and Syntax in American Sign Language. Ph. D. thesis, New York. • De Quadros, R. M. & J. Quer (2008). Back to back(wards) and
moving on: on agreement, auxiliaries and verb classes in sign languages. In R. M. De Quadros
(Ed.), Sign Languages: spinning and unraveling the past, present and future, Volume TISLR9
of eoretical Issues in Sign Language Research Conference: Florianopolis, Brazil, Petrópolis/RJ,
Brazil, pp. 530-551. Editora Arara Azul. • Sze, F. Y. B. (2003). Word order of Hong Kong Sign
Language. In A. E. Baker, B. van den Bogaerde, and O. A. Crasborn (Eds.), Cross-linguistic perspectives in Sign Language research. Selected papers from TISLR 2000, Hamburg, pp. 163-192.
Signum
“Defective” agreeing verbs in LSE: An OT account
Brendan Costello
Basque Center on Cognition, Brain and Language, San Sebastián
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: D 406
The ability of agreeing verbs to show marking for both arguments depends
on having unspecified location slots in the phonological matrix of the verb
(Padden 1983/1988). When a location slot is specified for a particular location, this gives rise to a conflict with the agreement morpheme. Often the
result is a defective agreement paradigm. For example, for the ASL verb
241
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 242 — #254
AG 7: Sign language agreement revisited
AG7
ee, the lexically specified location (near the eye) may block the appearance of subject agreement.
Some defective agreeing verbs of this type do show full agreement in
certain cases and this involves unusual agreement forms, often by means
of an additional movement that makes more location slots available. Such
a mechanism was first described for Israeli Sign Language (ISL) by Meir
(1998): a verb like ak (lexically specified at the mouth) starts at the subject
locus, moves to the lexically specified location, and then to the object locus.
However, such forms are only possible for first person object forms. In
contrast, Spanish Sign Language (LSE) shows an alternative strategy that
is available for the entire paradigm: a verb such as an (lexically specified
at the chin) moves from the lexically specified location to the subject locus
and from there to the object locus. Examples of such forms are shown in
(a-c).
Interestingly, first person object forms in LSE also show an alternative
that follows the ISL pattern (d). I offer an Optimality Theory (Prince &
Smolensky 1993) analysis to account for these facts based on constraints
developed for spoken language data. Critically, the notion of linearity plays
is central to explaining the differences between the data from these two
sign languages.
References: • Meir, Irit. 1998. ematic structure and verb agreement in Israeli Sign Language.
Ph.D. Thesis, Hebrew University of Jerusalem. • Padden, Carol A. 1983/1988. Interaction of
morphology and syntax in American Sign Language. New York: Garland. • Prince, A. & P.
242
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 243 — #255
AG 7, Raum D 406
Smolensky. 1993. Optimality eory. Technical Report 2. Rutgers Center for Cognitive Science,
Rutgers, NJ.
Dependency marking in American Sign Language
Jeremy Kuhn
CNRS, Institut Jean Nicod
[email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: D 406
In American Sign Language (and also RSL: [4]), a variety of dependency
constructions display agreement between a dependent plurality and its licensor. For example, in (1), the indefinite one or the adjective ame moves
in an arc-movement over the same area of space that was established by
the plural all bo. This inflection has a semantic effect: (1a) entails that a
plurality of books are distributed over the boys, one each; (1b) only allows
an ‘internal’ reading where the ‘sameness’ is distributed over the boys.
(1)
bo i-arc-a ead one-arc-a book. ‘The boys read one book each.’
(2)
all-a bo ead ame-arc-a book. ‘All the boys read the same book
as each other.’
With plural inflection on one or ame, agreement with the licensor is obligatory; the examples in (1) become ungrammatical if one or ame is signed
over a different area (e.g. locus b).
I provide a semantic analysis: the environments where arc-movement
is licensed are characterized by the introduction of a functional discourse
referent; agreement specifies the input of the function. Evidence for an
analysis in terms of functions comes from licensing conditions: arc-movement on one and ame is licensed exactly where a pronoun in English can
retrieve a functional antecedent, as in (5) ([5]). In particular, licensing is
not possible under none.
(3)
a.
all-a bo ead one-arc-a book. ‘All the boys read one book
(each).’
243
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 244 — #256
AG 7: Sign language agreement revisited
b. *bo i-arc-a, none ead one-arc-a book.
AG7
(4)
ha cla i-arc, …
’In that class, …
a. none den ead ame-neutral book. …no students read
the same book.
b. *none den ead ame-arc book.
(5)
a.
Every boy read {a/the same} book, and they each liked it. English
b. *No boys read {a/the same} book, but they each liked it.
Modulo the use of space, the sign language data exactly replicates patterns of dependency familiar from spoken language (‘dependent indefinites’: [1,3]; same: [2]). With spatial agreement, however, dependency is
made overt; sign language is able to disambiguate readings where spoken
language cannot. In particular, dependent indefinites in spoken language
(e.g. in Hungarian) are ambiguous when there are multiple potential licensors; in ASL, they are not.
(6)
A fiúk két-két könyvetadtak a lányoknak.
Hungarian
The boys two-two book give.3Pl the girls
‘The boys gave the girls two books {per boy OR per girl}.’
(7)
all-a bo gae all-b gil one-arc-b book.
‘All the boys gave all the girls one book per girl.’
References: • Balasu 2006. Distributive reduplication in Telugu. NELS. • Barker 2007. Parasitic
Scope. L&P. • Henderson 2014. Dependent indefinites and their post-suppositions. S&P. • Kimmelman 2015. Distribu- tive quantification in RSL. FEAST. • Nouwen 2003. Plural Pronominal
Anaphora in Context. LOT.
244
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 245 — #257
AG 7, Raum D 406
The order of Agree and Merge: Evidence from sign
language agreement
Roland Pfau1 , Martin Salzmann2
University of Amsterdam, 2 University of Leipzig
1
[email protected], [email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: D 406
I. SL agreement: Across sign languages (SLs), only a subset of verbs (agreement verbs, AVs) agrees by means of movement/orientation features while
others (plain verbs, PVs) do not. Meir (2002) offers a hybrid account suggesting that agreement by movement is thematic (Source → Goal), while
orientation marks syntactic object agreement. This allows for a unified
analysis of regular (RAVs) and backwards AVs (BAVs). Crucially, not the
verb agrees but rather a bound DIR-morpheme expressing transfer.
II. Against Meir’s account: (i) If DIR expresses abstract transfer, it is mysterious why certain orientation-only AVs don’t combine with DIR. (ii) Meir
postulates distinct DIRs for RAVs and BAVs → conceptual problem: since
roots do not combine freely with available DIR-morphemes, she has to assume lexical pre-specification for roots. (iii) agreement auxiliaries, as attested in some SLs, are void of lexical content and therefore cannot contain
DIR – hence this agreement has to be syntactic.
III. Non hybrid-approa: (i) RAVs carry the features [iv/iT] and move
via v to T, triggered by unvalued [uv/uT]-features. (ii) PVs do not carry
[iv/iT]-features; uFs on v/T remain unvalued, leading to a crash unless
a, which carries [iv/iT], is inserted. (iii) We propose that BAVs, which
challenge the syntactic approach, instantiate ergative agreement. Following Müller (2009), we assume that ergativity/accusativity is a property of
lexical items. Depending on the verb, merge of the internal argument may
precede Agree with the external argument (accusative pattern, cf. (1ai)), or
vice versa (ergative pattern: first subject agrees with v (1bii), then object
with T (1biii)). Agr on T is realized in the first slot of AVs and Agr on v in
the second. Thus, agreement on T is usually associated with the syntactic
subject, but with the object in the case of BAVs. In contrast, a, which
245
AG7
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 246 — #258
AG 7: Sign language agreement revisited
spells out a v+T complex, is lexically accusative. (iv) Agreement by orientation always follows an accusative pattern; it results from an independent Agree relationship between V and Obj. We submit that verbs showing
agr-by-orientation have an [uPerson] feature that is matched against the
[iPerson] feature on the object (2i). Since Agree does not involve all features, the object remains active for later Agree with v (2ii).
AG7
(1)
a.
(2)
Agreement by orientation
246
Agree before Merge → accusative b. Merge before Agree → ergative
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 247 — #259
Arbeitsgruppe 8
Gradienz im Spannungsfeld von empirischen
Methoden und Grammatiktheorie
Jana Häussler1 & Tom Juzek2
1
University of Wuppertal, 2 University of Oxford
AG8
[email protected], [email protected]
Raum: E 404
Arbeitsgruppenbeschreibung
In den vergangenen Jahren wurden Fragen der Datenbasis ür die linguistische Theoriebildung immer wieder intensiv diskutiert (vgl. Special Issue of Theoretical Linguistics, 33 (3), 2007, und Forum in der Zeitschrift ür
Sprachwissenschaft, 28 (1), 2009). Ziel dieser Arbeitsgruppe ist es, verschiedene Ansichten zum Verhältnis von experimentellen Methoden und syntaktischer Theoriebildung zusammenzubringen. Dabei liegt der Fokus auf
dem Phänomen der Gradienz. Traditionell basieren grammatische Theorien auf kategorialen (binären) Unterscheidungen und introspektiven Urteilen, die der/die untersuchende Linguist(in) selbst bereitstellt. Auf der anderen Seite lässt sich in empirischen Studien immer wieder Gradienz beobachten. Aus dieser Diskrepanz ergeben sich Fragen, wie die folgenden,
die in der AG diskutiert werden.
Ist Gradienz ein experimentelles Artefakt oder von linguistischer Relevanz?
• Falls letzteres zutrifft: Ist Gradienz ein Performanzeffekt, den es zu
reduzieren gilt, oder ist Gradienz
247
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 248 — #260
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
• Teil der Grammatik und sollte daher in grammatischen Modellen erfasst werden?
• Sind kategoriale Modelle und Gradienz miteinander vereinbar?
• Finden wir Gradienz auf allen Ebenen der Grammatik und hat sie
dabei jeweils die gleiche Relevanz?
• Sind introspektive Urteile von Linguisten unverzichtbar, da sie möglicherweise weniger anällig ür konfundierende Faktoren sind, oder
sind sie eher wenig reliabel und sollten daher durch empirische Daten gestützt werden?
AG8
• Wie lässt sich der Einfluss konfundierender Faktoren abschätzen und
minimieren?
Introduction: Why gradience matters
Jana Häussler1 & Tom Juzek2
University of Wuppertal, 2 University of Oxford
1
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: E 404
Negative island violations: *?# how bad are(n’t) they
really?
Ankelien Schippers
University of Oldenburg
[email protected]
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: E 404
We present the results of 2 acceptability judgment tasks investigating negative intervention in German. The point of departure is a contrast reported
in Rizzi (1992), showing that partial wh-movement (PM) is more sensitive
248
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 249 — #261
AG 8, Raum E 404
to negation than long-distance (LD) movement. He attributes this to the
non-referential status of the scope marker. However, various other factors
come into play as well: PM and LD movement are not equally acceptable for
German speakers (more acceptable for southern than northern speakers,
cf. Fanselow et al., 2005). Secondly, negating matrix predicates occurring
in PM and LD constructions is pragmatically problematic: negated questions require d-linking, but questions with verbs like glauben ‘believe’ (as in
Rizzi’s example) typically have a non d-linked reading of the complement
clause (cf. Dayal, 1994 and Horvath 1997). To investigate the individual contributions of these factors, we used both northern and southern speakers.
We also included conditions where it is not the scope-marker chain which
is interrupted by negation, but where negation is intervening in a different type of non-referential dependency, namely between a non-referential
how many NP phrase and its trace. This way we could investigate whether
the effects on acceptability where comparable in both situations. Finally,
we varied the position of negation by putting it either in the matrix or the
subordinate clause. Since all conditions involved movement of the embedded object wh-phrase, wh-movement was always crossing negation. Under
a purely structural account of wh-islands, the exact position of the negation
shouldn’t matter, as long as it is structurally intervening. Our results show
a number of interesting facts: first of all, we weren’t able to replicate the
original contrast of Rizzi (1992). Instead, matrix negation itself had a very
strong effect on acceptability. The effect of referentiality however was very
weak and not consistent across experiments. Regarding dialect, the relative
patterns of acceptability where the same for northern and southern speakers, although LD-movement was more acceptable for northern speakers.
We discuss the implications this has for the analysis of PM vs. LD movement and negative intervention and address the possible reasons for the
discrepancy between our finding and those of Rizzi.
References: • Dayal, Veneeta. 1994. Scope Marking as Indirect Wh-Dependency. Natural Language Semantics 2.137-170 • Fanselow, Gisbert., Kliegl, Reinhold. & Schlesewsky, Matthias.
2005. Syntactic variation in German wh-questions. Empirical investigations of weak crossover
violations and long wh- movement. Linguistic Variation Yearbook 5.37-63. • Horvath, Julia.
1997. The Status of ‘Wh-Expletives’ and the Partial Movement Construction of Hungarian.
249
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 250 — #262
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
Natural Language and Linguistic eory 15.509–572. • Rizzi, Luigi. (1992). Argument/Adjunct
(A)symmetries. In Broderick, K. (Ed.), Proceedings of North Eastern Linguistics Society 22, 365-382.
Amherst, MA: GLSA.
Gradient acceptability and categorical distinctions: The
case of imperative constructions
Robert Külpmann1 & Vilma Symanczyk Joppe1
University of Wuppertal
1
kuelpman@uni-wuppertal, [email protected]
AG8
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: E 404
Our talk shall deal with a family of imperative constructions we call IDCs:
disjunctions or conjunctions of an imperative and a declarative sentence.The
established model for the organization of this family (e.g. Kaufmann 2012,
Culicover/Jackendoff 1997), distinguishes two basic categories of IDCs. Type
I IDCs (1) are thought to represent independent speech acts. Their first conjunct is not semantically subordinated and shows the features of a plain
directive imperative. Type II IDCs (2), on the other hand, receive a conditional interpretation. Their first conjunct is semantically subordinated and
behaves like a conditional subclause.
Type I
(1)
Drück den Knopf und ich betätige den Regler.
“Press the button and I turn on the switch.”
Type II
(2)
Drück den Knopf, und wir werden alle sterben!
“Press the button, and we will all die!”
Type I and type II IDCs are typically associated with certain formal features
(cf. Table 1), either typical for plain imperatives (type I) or conditional subclauses (tape II):
250
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 251 — #263
AG 8, Raum E 404
Table 1: Properties of IoD/IaD subclasses (IMP: plain imperative, CS: conditional subclause)
IMP
Type I
Type II
CS
[- imp]
1st conjunct semantically subordinated
conditional interpretation
-
binding 2nd into 1st conjunct
insertion of NPIs possible
[+ imp]
-
1st conjunct: speech-act status
-
+
-
+
-
+
-
+
+
-
+
+
insertion of discourse particles possible
+
+
-
-
do-support possible
+
+
-
-
Based on a set of acceptability rating studies, we show that argument
omission (AO) in imperatives differs from AO in other sentence types (Külpmann/ Symanczyk Joppe 2015).
(3)
a.
b.
Nimm mal (den Korb)!
Take.imp pa the basket
Er nahm *(den Korb).
He took the basket.
A categorical model would predict that AO in type I, but not type II IDCs
should resemble AO in imperatives. We conducted several further studies
on AO in different types of IDCs, but could confirm this prediction only
to some extent. Our talk ends with a discussion whether the categorical
model of IDCs should be supplemented or replaced by a gradual one.
References: • Culicover, P. / R. Jackendoff (1997): Semantic subordination despite syntactic
coordination. Linguistic Inquiry 28, 195-217. • Kaufmann, M. (2012) [Schwager 2006]: Interpreting imperatives. Dordrecht: Springer. • Külpmann, R. / V. Symanczyk Joppe (2015): Argument omission between valency and construction. Evidence for sentence type effects from
acceptability rating studies. In: G. Jäger (ed.): Proceedings of the 6th Conference on antitative
Investigations in eoretical Linguistics. University of Tübingen.
251
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 252 — #264
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
Competing grammars and the representation of subject
clitics in French
Jennifer Culbertson
University of Edinburgh
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:30, Raum: E 404
AG8
The representation of subject clitics across the Romance languages has generated significant debate. In this talk I highlight the importance of quantitative data in the analysis of these elements in French. Specifically, I show
that distributional analysis from child-directed and adult-directed speech
corpora support distinct representational analyses of clitics which depend
on the register used. These distinctions, which correlate with structural
differences in the grammar in terms of the position of the clitic and use
of discontinuous negation, can be formalized in terms of grammar competition. I discuss the implications of this account for discrete versus probabilistic grammar formalisms, and discrete vs. continuous representations
of grammatical elements.
Choose your features: Lexical optionality and variation in
QNP agreement
Gabi Danon
Bar-Ilan University
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:00, Raum: E 404
Quantified noun phrases (QNPs) in subject position allow several agreement patterns on the predicate, with acceptability showing considerable
gradience. We present experimental findings from a study of QNP agreement in Hebrew and Russian in sentences like the following:
252
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 253 — #265
AG 8, Raum E 404
(1)
me-ha-kita
yada
/ yad‘a
/ yad‘u et
xeci
half.M.S of-the-class.F.S knew.M.S / knew.F.S / knew.PL OM
ha-tšuva.
the-answer
‘Half of the class knew the answer.’ (Hebrew)
In Hebrew, QNPs consisting of a Q bearing morphological gender &
number and a singular group N (as in (1)) may trigger 3 agreement patterns:
Q-agr (agreement involving Q’s morphological gender&number), N-agr
(agreement involving N’s morphological gender&number) or S-agr (plural agreement, matching the subject’s semantic number). A fourth pattern
found in Russian, default agreement, will not be discussed.
We argue that the findings support the model proposed in Danon (2013),
where subject agreement involves an abstract set of features (inde, following Wechsler & Zlatic 2003, henceforth WZ; see also Sauerland & Elbourne 2002) distinct from the morphologically-related gender and number (concod, following WZ). The predicate agrees with a QNP subject’s
INDEX; the QNP’s head gets its features valued either pre-syntactically
(lexically) or via agreement. The alternation in agreement is thus argued
to be the result of optionality in lexical feature specification: Q’s inde
may either be specified in the lexicon to match its concod (leading to
Q-agr), or Q may enter the derivation with unvalued inde, to be valued
via agreement with NP, leading to N-agr or S-agr. Deriving S-agr involves
a mismatch between N’s inde and its concod; N-agr with group nouns
involves an inde-semantics mismatch. We argue that gradience is linked
to these mismatches between inde and semantics or morphology. The
current study aims to test the predictions of this analysis.
Two acceptability judgment studies, one for each language, were performed using a 5-point scale. For QNPs with plural nouns, in both languages N-agr was rated significantly higher than Q-agr; this follows from
the hypothesis that assigning an independent inde to Q (leading to Qagr) is a marked option, as inde is associated with reference. With group
nouns, the ratings for both N-agr and S-agr were lower than those for N-agr
with plurals, in both languages. This is explained as a ‘penalty’ for either
an indeconcod mismatch (S-agr) or an inde-semantics mismatch (N-
253
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 254 — #266
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
agr); Hebrew seems to be neutral regarding the choice between these options, while Russian has a clear preference for the former. Q-agr with group
nouns, on the other hand, was rated similarly to Q-agr with plurals, as expected since N’s features play no role in deriving Q-agr. These results thus
support a lexical model of agreement alternations, where the source of the
alternation lies in the pre-syntactic specification of abstract agreement features; gradience in this model is due to lexical feature choice rather than to
syntax. Both languages ‘penalize’ cases involving a feature mismatch, but
they differ regarding how alternative mismatches are ranked.
AG8
A theory of agreement attraction based on a continuous
semantic representation space
Garrett Smith1 , Julie Franck2 & Whitney Tabor1
University of Connecticut, 2 University of Geneva
1
[email protected], [email protected], [email protected]
Mittwoch, 24. 02. 2016, 18:00–18:30, Raum: E 404
Speakers sometimes exhibit agreement attraction effects, in which a verb
agrees with a noun other than its nominal controller (e.g. rating e key to
the cabinets are. . . higher than e key to the cabinet are...) While previous
morpho-syntactic [1], conceptual [2], and mixed approaches [3] acknowledge gradience in attraction, none provide careful semantic analysis or a
theory of how particular semantic representations produce attraction. Collecting acceptability judgments (1–7 scale, 7 = best) and relating these to
semantic tests, we hypothesized that plural attraction rate should correlate
with semantic similarity to a grammatically plural form, thus grading into
grammaticality.
We identified five different classes: (i) Arbitrary Containments (e.g., a
box (N1) with apricots (N2)), (ii) Full Containments (a box of apricots), (iii)
Collections (a set. . . ), (iv) Measures (a lot. . . ), and Plural Quantifiers
(many apricots). Full Containments differ from Arbitrary Containments in
implying the container is full of its contents, often implying large quantity.
Collections also imply quantity, but differ from Containments in lacking
254
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 255 — #267
AG 8, Raum E 404
a physical container. Both Containments and Collections specify a spatial
arrangement of the contents. Measures and Quantifiers express quantity,
lack a physical container, and do not specify spatial arrangement. Thus,
in progressing across these classes, the meaning of the pre-N2 material
is increasingly like the meaning of a quantifier, predicting that attraction
should gradually turn into grammaticality.
Test sentences had the form A N1 (P N2) is/are in the NP. In Experiment 1 (Arbitrary vs. Full Containments), ratings were significantly higher
in the Verb Plural Full condition (b = 4.569, SE = 0.124, t = 36.85; mixed
model, dummy-coded factors) than the Verb Plural Arbitrary condition (b
= −0.510, SE = 0.094, t = −5.40), as predicted. Experiment 2 (Full Containments and remaining classes crossed with Verb Number (singu- lar/plural)
and Length (with/without P N2)) also supports our prediction via a threeway interaction (likelihood ratio test: χ2(3) = 16.86,p < .001). Post-hoc tests
confirmed that Short-condition acceptability increases from Full Containments to Collections, but Collections, Measures, and Quantifiers did not
differ, consistent with the prediction of increased acceptability for plural
verb agreement as N1 becomes more quantifier-like.
The Verb Plural, Long condition was elevated relative to Short in Containers and Collections, but not Measures and Quantifiers, supporting attraction for the former and grammaticality for the latter. As we predicted,
gradual semantic change shifted attraction into grammaticality. We conclude that linking semantic gradience to gradience in the morpho-syntactic
distribution provides a more predictive model of conceptually-driven agreement attraction, clarifying previous accounts with a more precise semantic
treatment. References
References: • [1] Franck et al. 2002. Lg. Cog. Proc. 17(4). • [2] Solomon & Pearlmutter. 2004.
Cog. Psych. 49. • [3] Eberhard et al. 2005. Psych. Rev. 112.
255
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 256 — #268
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
Problematic cases for weighted constraint models:
Subadditivity and cost-free violation
Emilia Ellsiepen
University of Frankfurt
[email protected]
Donnerstag, 25. 02. 2016, 09:00–10:00, Raum: E 404
AG8
In order to account for gradient judgment data, syntactic models in the
spirit of Harmonic grammar incorporate numerically weighted constraints,
e.g. the Decathlon model (DM, Featherston, 2005) and Linear Optimality
Theory (LOT, Keller, 2000). In the strict reading, LOT and DM assume that
constraint violations always apply and directly influence relative acceptability. This claim can be broken down into the following three predictions:
(1)
a.
b.
c.
Violating a constraint always results in a decrease in acceptability, even for ‘optimal’ structures (grammatical sentences)
Constraint violations are cumulative, therefore every additional
violation leads to a further decrease in acceptability (ungrammatical sentences)
Constraint violations always have the same effect, therefore
violating constraint C in different contexts results in a numerically equal decrease in acceptability
While Keller (2000) and others gathered evidence for the second prediction, the other two remain largely untested. Acceptability rating studies
carried out in our lab, however, identify problematic cases for those predictions.
The first issue is raised by a series of experiments using magnitude estimation (ME) to investigate word order preferences in the German middle field. In two instances, constraints that had a measurable effect in one
context did not result in a penalty when violated by optimal structures.
AniFirst, e.g., enabled us to account for the preference of indirect object
before subject in passive clauses as well as the canonical order for ditransitive clauses. When manipulating the animacy of an agentive subject in
256
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 257 — #269
AG 8, Raum E 404
transitive clauses, however, we did not detect a loss in acceptability for
the conflicting configuration, but instead an ameliorating effect on OS orders, if the animacy hierarchy was respected. This is evidence in favor of
a traditional model, where optimal structures are not expected to differ in
acceptability, whereas LOT and DM make wrong predictions.
A second problem is posed by quantitatively diverging penalties associated with the same constraint. Two experiments testing up to three constraint violations in the same sentence with different scales (continuous,
ordinal) indicate that constraint violations are cumulative in the sense of
(1-b), but not additive. We manipulated two categorical constraints, namely
number agreement and the position of the auxiliary inside a verb cluster,
in addition to the soft constraint S>O. Although each violation resulted in
a decrease in acceptability, this decrease was numerically smaller, if other
violations coincided. This suggests that there is no linear relationship between acceptability as measured by ME or on a Likert scale, and the harmony value that is used by Keller and Featherston, i.e. the sum of all individual violation costs. In summary, our findings challenge existing fully
quantified models of gradient acceptability.
Towards an efficient evaluation method of generative
theories using gradient data and regression
Marta Wierzba
University of Potsdam
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: E 404
With Linear Optimality Theory (LOT), Keller (2000) introduced a framework that allows to account for gradient acceptability of syntactic structures. Each factor that is relevant for the acceptability of a sentence is formulated as an Optimality Theoretical constraint associated with a numerical weight. The sum of weights of all constraints that a given sentence
violates then maps directly to its grammaticality, which is assumed to be
measurable directly in terms of acceptability. The severity of the constraint
is estimated by a statistical algorithm.
257
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 258 — #270
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
AG8
Like Standard Optimality Theory (Prince & Smolensky 1993), LOT also
involves a generating and an evaluating component. Keller provides a detailed proposal for the latter and leaves the former unspecified, except for
the statement that the generating component must be able to produce permutations of constituents. It is thus tempting to combine LOT with an explicit the- ory of grammatical structure-building, such as the Minimalist
Framework (Chomsky 1995 and further work), to attain a more comprehensive theory as well as a way to test specific Minimalist proposals empirically. However, some non-trivial issues arise with this attempt. One
of them is that it is a core assumption in Keller’s work that grammaticality is gradient. In contrast, in Minimalism (and more generally, in generative grammatical models), by design, there is a categorical distinction
between structures that can be derived by the set of structure-building operations/rules, and those that cannot be derived.
I want to argue that this issue can be resolved in the following way.
Sentences that are derivable by a Minimalist grammar and are therefore
grammatical by definition within that framework can still involve interpretative, prosodic, or processing-related problems to a vary- ing extent.
If a theory predicts such a problem for a specific type of structure, a systematic acceptability decrease should be found for it consistently across
the data. Under this view, it is possible to conceptualize a generative theory in terms of weighted constraints, and to make it compatible with an
LOT-inspired empirical evaluation: A constraint weight estimation algorithm (e.g. multiple regression) can be used to estimate the severity of the
constraint violation and to evaluate how consistently the acceptability decrease is really found where the theory predicts it to occur. In the talk, I will
show in detail how to apply the idea to a data set of gradient acceptability
judgments concerning word order variation in Czech.
In sum, by establishing a link between generative grammar and the LOT
framework, I hope to provide a powerful evaluation method of generative
linguistic theories which can be used for a broad range of approaches.
References: • Chomsky, Noam. 1995. e Minimalist Program. Cambridge, MA: MIT Press.
• Keller, Frank. 2000. Gradience in grammar. Experimental and Computational Aspects of Degrees of Grammaticality. Doctoral Dissertation, University of Edinburgh. • Prince, Alan, and
258
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 259 — #271
AG 8, Raum E 404
Paul Smolensky. 1993. Optimality eory: Constraint Interaction in Gener- ative Grammar. Rutgers University Center for Cognitive Science.
Gradience at interfaces
Antonella Sorace
University of Edinburgh
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:30, Raum: E 404
Much research on the syntax-lexicon interface has shown that notion of a
categorical distinction between unaccusative and unergative verbs is difficult to maintain: verb behaviour in syntactic diagnostics of split intransitivity systematically varies along a gradient hierarchy defined by the aspectual type of the verb and the context in which the verb appears (Cennamo & Sorace, 2007; Keller & Sorace, 2003; Legendre, 2007; Legendre &
Sorace, 2003; Sorace 2000, 2004, to appear). Based on online and offline experiments on native and non-native speakers, I will show that gradience
in the selection of perfective auxiliaries avere/haben (‘have’) and essere/sein
(‘be’) with intransitive verbs in Italian and German is sensitive to interactions between the event structure complexity of verbs and the cognitive
capacity of individual speakers to apply aspectual coercion. I will finally
compare these phenomena with data from the syntax-pragmatics interface, and especially pronominal reference (Sorace 2011), where one sees
similar interactions of linguistic and general cognitive factors.
259
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 260 — #272
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
A gradient acceptability study of English sentences with
two negatives
Frances Blanchette
Pennsylvania State University
[email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: E 404
AG8
English sentences with two negatives have two possible interpretations:
Negative Concord (NC) ((1a) and (2a)), with a single negative meaning,
and Double Negation (DN) ((1b) and (2b)) where each negative makes an
independent semantic contribution:
(1)
a.
b.
(2)
a.
b.
Object NC: Sam went to Lisa’s favorite store with her, but he
just stood there while she shopped.
He didn’t buy nothing in that store.
Object DN: Sam said he walked out of the store without buying
anything, but I know better.
He didn’t buy nothing in that store.
Subject NC: The teacher forgot to get the classroom ready for
the first day.
Nothing wasn’t ready before the students arrived.
Subject DN: The teacher worked all night to prepare the classroom for her students.
Nothing wasn’t ready before the students arrived.
Though widespread in North America (Wolfram and Christian 1974), English NC is heavily socially stigmatized. This study uses gradient acceptability to test the hypothesis that speakers who do not accept NC nevertheless have grammatical knowledge of it. English NC usage displays microsyntactic variation: Only a subset of Object NC (1a) users employ Subject
NC (2a) (Smith 2001). Furthermore, speakers who do not use NC nevertheless interpret strings like (1) out of the blue as NC and not DN (Coles-White
2004), but the same is not true for Subject NC. This gradient acceptability
study exploits these differences, predicting that though NC will be unac-
260
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 261 — #273
AG 8, Raum E 404
ceptable overall, Object NC should be significantly more acceptable than
Subject NC. Results showed that participants distinguished between unacceptable sentence types. A 2 (negative subject vs. negative object) by 2 (NC
vs. DN) ANOVA revealed a significant preference for negative objects (M
= 2.85) over negative subjects (M = 2.48) (F(1, 100) = 20.03, p = .001), and
a significant interaction in which items with a negative object were more
acceptable in NC contexts (1a, 2a) than in DN contexts (1b, 2b) (F(1, 100) =
14.74, p < .001). Participants thus preferred Object NC over Object DN, but
showed no preference for NC or DN with negative subjects. These effects,
which reveal syntactic knowledge of NC, would remain obscured if only
binary acceptability were considered.
References: • Coles-White, D’Jaris. 2004. Negative concord in child African American English: implications for specific language impairment. Journal of Speech, Language, and Hearing
Research 47: 212–222. • Smith, Jennifer. 2001. Negative Concord in the Old and New World:
Evidence from Scotland. Language Variation and Change 13: 109–143. • Wolfram, Walt & Ralph
Fasold. 1974. e study of social dialects in American English. New Jersey: Prentice Hall.
Quasi-modals in English
Anne-Laure Besnard
University of Nantes
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: E 404
The categorisation of items such as be able to, be likely to, or be supposed
to — usually described as belonging to a set of ‘quasi-modals’ — raises numerous issues in English and is an area where even reference grammars
suggest the existence of a cline. One major problem is that ‘[t]here is a
gradience between a semi-auxiliary such as be bound to and an occurrence
of the copula BE followed by an adjectival or participial construction such
as happy to or compelled to’ (Quirk et al. 1985: 144), which means that ‘[t]he
membership of the set is by no means clearcut, and is difficult to delimit
in a principled fashion’ (Collins 2009: 17). This is, however, what Westney
(1995) sets out to do. His overall study is descriptive and corpus-informed,
261
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 262 — #274
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
AG8
but when trying to single out quasi-modal expressions, he only works with
his own intuition-based responses to a number of syntactic and semantic
tests, and it is noteworthy that the results he obtains are actually not that
clear-cut. Based on the review of Westney’s conclusions and their comparison with authentic data taken from a 40-million-word corpus (e Independent 2009), I will try to show that a purely theoretical approach may
produce results contradicted by empirical analysis, while not necessarily
allowing the elimination of gradience. For example, able to supplementive
clauses, which are deemed ‘very odd or impossible’ by Westney (1995: 20),
were in fact identified in the aforementioned corpus – as in Able to command the services of the world’s top musicians, he played alongside Richard
ompson, Dave Gilmour and Eric Clapton – and in similar numbers to likely
to supplementive clauses. I will thus argue for the consideration of empirical data within theoretical work, while insisting that gradience is not
created by empirical studies but exists in language, and should as such be
accounted for by linguistic theory. I will demonstrate how this can be done
within the Theory of Predicative and Enunciative Operations (Culioli 1990)
via the representation of ‘quasi-modals’ and related structures as complex
markers operating and interacting in context.
References: • Collins, Peter. 2009. Modals and asi-Modals in English. Amsterdam-New
York: Rodopi. • Culioli, Antoine. 1990. Pour une linguistique de l’énonciation T.1. Gap: Ophrys.
• Quirk, Randolph, et al. 1985. A Comprehensive Grammar of the English Language. New York:
Longman. • Westney, Paul. 1995. Modals and Periphrastics in English. Tübingen: Niemeyer.
Ellipsis alternation
Joanna Nykiel
University of Silesia
[email protected]
Freitag, 26. 02. 2016, 12:00–13:00, Raum: E 404
I discuss here what I dub ellipsis alternation. This alternation refers to the
possibility of using stranded phrases (remnants) with prepositions or without them in elliptical constructions (A: I heard Pat yelling at someone. B: (At)
262
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 263 — #275
AG 8, Raum E 404
who?). The availability of ellipsis alternation in var- ious languages have
received considerable attention in theoretical linguistics (Merchant 2001,
2004, Stjepanovic 2008, Rodrigues et al. 2009, Sag and Nykiel 2011, Nykiel
2013, Leung 2014). The argument is that this alternation is freely available
in some languages, but is re- stricted to a specific environment in other languages: antecedents containing lexical NPs (as opposed to indefinite pronouns) as correlates for remnants. The availability of ellipsis alter- ation is
given syntactic motivation: those languages that allow ellipsis alternation
also allow preposition stranding (Merchant 2001, 2004). However, there is
empirical evidence that ellip- sis alternation is freely available independently of preposition stranding (Nykiel 2013, Leung 2014). I argue here
that the availability of ellipsis alternation doesn’t differ between languages,
but the distribution of remnants without prepositions vs. remnants with
prepositions does. Toward this purpose, I explore the nature of the contrast between lexical NPs and indefinite pronouns in English corpus data,
using mixed-effects regression modeling for statistical analy- sis. The data
reveal that English remnants without prepositions have higher frequencies if their antecedents contain lexical NPs than if they contain indefinite
pronouns, but the strength of this effect depends on the lexical-semantic
relationship between prepositions and verbs.
I connect these results to principles of efficient language processing of
Hawkins (2004, 2014) and to research on anaphora resolution by Ariel
(1990), and discuss their significance to existing cross-linguistic research
on ellipsis alternation. I then discuss the significance of this research with
respect to the theme of the gradience, particularly the fact that it provides
support for the view that (1) introspective judgments misrepresent the
space of grammatical possibility, and (2) competence grammar is shaped
by performance preferences and hence studying such preferences can help
motivate the design features of competence grammar.
263
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 264 — #276
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
Dative clitic doubling variation in Spanish reverse psych
verb sentences: Syntax meets discourse-pragmatics and
semantics
Chiyo Nishida
University of Texas at Austin
[email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: E 404
AG8
It is widely assumed that the canonical word order for Spanish sentences
with reverse psych verbs like gustar ‘to appeal’, importar ‘to matter’, etc, is
[IO-V-S] and dative clitic doubling (datCLD) is obligatory (A Pedro LE/*Ø
gusta la música ‘Pedro likes music.’). However, for sentences occurring in a
non-canonical [S-V-IO] order, datCLD becomes optional, as amply attested
in on-line corpus data (Ese tío LES/Ø gusta a las mujeres ‘That guy appeals
to (the) women’). In addition, datCLD shows sensitivities to certain referential properties. A strictly syntactic analysis like Cuervo (2010) offers no
account for these variations, since it assumes that the two constructions
are semantically equal and transformationally related.
This paper shows that the distribution of datCLD can be adequately explained by integrating various discourse-pragmatic and semantic factors
into analysis. First, we note that judging from the discourse context in
which they appear, the [IO-V-S] and [S-V-IO] constructions have different
information structures: [foc IO V S] and [[op S] [foc V IO]], respectively.
We postulate that these constructions are also distinct semantically, hence
derived independently from each other. In the former, the IO is an experiencer and an obligatory argument. The latter roughly means: “Subject has
a property P relative to IO”, where the IO specifies “to whom” – call it a
elaiie – the judgment made (S having P) holds. Note that this IO is
syntactically optional (cf. Las películas japonesas gustan. ‘Japanese movies
are appealing’). This semantic difference is the basis for the datCLD variation in two constructions. Jaeggli (1982) explains that CLD becomes obligatory when the dative marker a ‘to/for’ alone cannot assign a proper θ-role
to the IO; otherwise CLD is optional. An experiencer (or possessor, source,
264
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 265 — #277
AG 8, Raum E 404
etc.) IO falls into the first case and a relativizer (or recipient) IO into the
second.
Regarding the datCLD variation in [S-V-IO] sentences, preliminary statistical analysis results on some 760 tokens of psych verb sentences from
an on-line modern Spanish showed some referential effects on datCLD:
(1) Pronominal IOs require CLD and (2) referential, individual (=non collective), or singular IOs prefer CLD. These results strongly indicate that
distribution of datDAT in [S-V-IO] psych verb sentences also exhibit some
systematic patterns.
In conclusion, CLD in dative constructions is not governed by a single
factor but by a combination of various discourse pragmatic and semantic
factors interacting with morphosyntax.
Competing embedded clauses in German: Conflicts in
position of extraposed relative and argument clauses
Matthias Schrinner
University of Leipzig
[email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: E 404
Claim I investigate the serialization of sentence-final relative clauses (RC)
and argument clauses (AC) in German. Both possible serializations are to be
classified as grammatical while RC»AC provides the preferred strucure. In
addition, if the RC is extraposed out of a subject while the AC is object, both
serializations are equally adequate. The task of the analysis is to develop
a grammar that produces RC»AC and AC»RC-structures according to the
ratings given by the speakers.
Baground For the analysis of preferred structures, two general principles of grammar play a role: (1) Principle of End-Recursion: Final-embedding
structures are preferred over center- embedding structures (Miller & Chomsky 1963). (2) Principle of Proximity: There is proximity between governors
and dependents (Ferrer i Cancho 2006).
265
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 266 — #278
AG 8: Gradienz im Spannungsfeld von empirischen Methoden …
Investigation I developed a questionnaire with items exhibiting RC»AC
and AC»RC structures permutating the grammatical functions of the clauses
(subject/direct object and direct object/indirect object, see table).
AG8
RC antec.
SUB
DO
IO
AC is
DO
SUB
DO
RC»AC
3.25
2.78
3.17
gramm.
1.95
AC»RC
3.17
3.56
3.84
ungramm.
5.15
fillers
Analysis A high ranked constraint EXTRAP makes sure that all candidates with unextraposed clauses leave the competition early. All CPs under discussion are assumed to have a [+extrap]- feature that triggers extraposition together with the constraint. The resulting distance effects are
minimized by three constraints. The first, MA, leads to proximity between
RC and its antecedent. The second, AA, leads to argument adjacency between verb and AC. The third, *CRDEP, penalizes crossing dependencies
that could lead to unfavorable dependency constellations (cf. Ferrer i Cancho 2006).
(1)
Eapoiion (Eap)
CPs with a [+extrap]-feature are extraposed.
(2)
Modifie Adjacenc (MA)
Assign one * per linearly intervening constituent between a modifier and its licenser.
(3)
Agmen Adjacenc (AA)
Assign one * per linearly intervening constituent between an argument and its licenser.
(4)
*Coing Dependencie (*CDep)
Surface structures do not exhibit crossing dependencies
Roughly, Stochastic Optimality Theory claims that at every utterance,
a new competition is run and in every competition the constraints are
ranked anew. I present scaled values for the constraints such that for a
266
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 267 — #279
AG 8, Raum E 404
large number of competitions, the grammar will produce 50% AC»RC vs.
50% RC»AC candidates in AHA-Context and 33% AC»RC vs. 66% candidates
in other contexts, which equals to the proportion in which the structures
were rated with grades 1 and 2 by the speakers.
References: • Boersma & Hayes (2001) Empirical Tests of the Gradual Learning Algorithm. LI
32, 45–86. • Ferrer i Cancho (2006) Why do syntactic links not cross? Europhysics Leers 76,
1228–1234. • Haider (1994) Detached Clauses—e Later e Deeper. Working Paper. • Keller
(1994) Extraposition in HPSG. Verbmobil Report 30. • Miller & Chomsky (1963) Finitary models
of language users.
267
AG8
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 268 — #280
Linguistik bei J.B. Metzler
Die bewährte Einführung informiert über die linguistischen Kerngebiete, erläutert Grundbegriffe, illustriert
sie an Beispielen aus dem Deutschen und gibt einen
Einblick in die linguistische Theoriebildung. Kindlicher
Spracherwerb und Sprachwandel – zwei Gebiete, die
von großer Bedeutung für ein tieferes Verständnis
der menschlichen Sprache sind – werden in weiteren
Kapiteln vorgestellt. Mit Übungen, einem Glossar der
wichtigsten Fachtermini, einer weiterführenden
Schlussbibliographie und einem Sachregister.
Für die 3. Auflage wurde der Band umfassend überarbeitet und aktualisiert.
Jörg Meibauer / Ulrike Demske / Jochen Geilfuß-Wolfgang
Jürgen Pafel / Karl Heinz Ramers / Monika Rothweiler
Markus Steinbach
Einführung in die germanistische Linguistik
3., überarbeitete und aktualisierte Auflage 2015, XII, 369 Seiten, € 19,95
ISBN 978-3-476-02566-1
Sprachwissenschaft umfasst mehr als nur Grundlegendes
wie Laute, Wörter und Sätze. Daher greift das Lehrbuch
über die grammatischen Disziplinen hinaus zahlreiche
Themen auf, die das Interesse für das Fach Linguistik
wecken: sprachliche Interaktion, Variation und Wandel,
Sprachkontakt und Mehrsprachigkeit, Sprache und Kultur,
Ursprung der Sprache u. a. Das große Format und das
zweifarbige Layout erleichtern den Überblick.
Peter Auer (Hrsg.)
Sprachwissenschaft
Grammatik - Interaktion - Kognition
2013, IX, 465 Seiten, 71 farb. Abb., 66 farb. Tabellen, € 29,95
ISBN 978-3-476-02365-0
[email protected]
www.metzlerverlag.de
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 269 — #281
Arbeitsgruppe 9
Geschriebene und gesprochene Sprache als
Modalitäten eines Sprachsystems
Martin Evertz1 & Frank Kirchhoff1
1
Universität zu Köln
AG9
[email protected], frank.kirchhoff@uni-koeln.de
Raum: E 403
Arbeitsgruppenbeschreibung
Die traditionelle Annahme, dass die Schriftsprache direkt von der gesprochenen Sprache abgeleitet ist, hat sich in jüngerer Forschung als problematisch erwiesen (vgl. ür einen Überblick Dürscheid 2012). Primus (2003)
spricht sich gegen eine direkte Ableitung aus und postuliert ein Schnittstellenmodell, in dem gesprochene und geschriebene Sprache Modalitäten eines übergreifenden Sprachsystems sind. Die geschriebene Sprache
korrespondiert in diesem Modell mit Subsystemen der Sprache (Phonologie, Morphologie, Syntax, etc.) und bildet mit der gesprochenen Sprache
Schnittstellen.
In der Literatur ist bereits eine modalitätsübergreifende Phonologie im
Begriff, sich zu etablieren (vgl. Domahs & Primus 2015). In dieser AG soll
untersucht werden, welche neuen Einblicke in Gemeinsamkeiten und Unterschiede zwischen Laut- und Schriftsprache ein solcher modalitätsübergreifender Ansatz gewährt und inwiefern dieser Ansatz darüber hinaus
auch ür andere Teilgebiete der Linguistik fruchtbar ist.
Diese AG richtet sich an SchriftsystemforscherInnen und an SprachwissenschaftlerInnen aus anderen Teilgebieten der Linguistik, die sich mit der
269
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 270 — #282
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Schnittstelle zur Schrift aus empirischer, theoretischer, historischer oder
didaktischer Sicht auseinandersetzen.
Zu den Fragen, die in dieser AG untersucht werden, gehören u.a.:
• Welche sprachlichen Einheiten kommen in beiden Modalitäten vor
und wie unterscheiden sie sich modalitätsspezifisch?
• Welche Schnittstellen-Phänomene können beobachtet werden?
• Welche Untersuchungsmethoden (Experimente, Korpora, Datenbanken) bieten sich ür modalitätsübergreifende Ansätze an?
• Wie kann ein modalitätsübergreifendes Modell didaktisch genutzt
werden?
AG9
Literatur: • Domahs, Ulrike & Beatrice Primus. 2015. Laut – Gebärde – Buchstabe. In Ekkehard Felder & Andreas Gardt (eds.), Sprache und Wissen. 125–142. Berlin/ New York: de
Gruyter. • Dürscheid, Christa. 2012. Einührung in die Schrilinguistik. Grundlagen und eorien. 4. überarb. und akt. Aufl. Göttingen: UTB. • Primus, Beatrice. 2003. Zum Silbenbegriff in
der Schrift-, Laut- und Gebärdensprache – Versuch einer mediumunabhängigen Fundierung.
Zeitschri ür Sprachwissenscha 22. 3–55.
Einführung und Überblick
Martin Evertz1 & Frank Kirchhoff1
1
Universität zu Köln
[email protected], frank.kirchhoff@uni-koeln.de
Mittwoch, 24. 02. 2016, 14:00–14:30, Raum: E 403
270
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 271 — #283
AG 9, Raum E 403
Script types: Definition and classification
Andreas Nolda
University of Szeged
[email protected]
Mittwoch, 24. 02. 2016, 14:30–15:00, Raum: E 403
The typological classification of writing systems and their scripts is a debated matter in the literature (for an overview of classic proposals, cf. Coulmas 1996). Much of the confusion can be resolved, it is claimed in this talk,
if the following theoretical assumptions are made:
1. A natural language which is both spoken and written has a linguistic
system with a poken em, a iing em, and an ineface
between them (a similar view is taken by Primus 2003).
2. A writing system, in turn, consists of one or several cip and a
pelling em (also assumed by Coulmas 1996).
3. A script provides a set of graphemes; the written units formed from
them are determined by the spelling system.
4. Written and spoken units are linked by coepondence elaion
in the interface.
5. BASIC correspondence relations are presupposed in the identification of deied correspondence relations between more complex
units.
Given those assumptions, terms like phonographic, logographic, etc. –
understood as denoting properties of scripts, not of writing systems as a
whole – can be defined with respect to basic correspondence relations in
the interface of such linguistic systems. On this basis, a general classification system will be proposed for major and minor script types.
References: • Coulmas, Florian. 1996. Typology of writing systems. In Günther, Hartmut &
Otto Ludwig (eds.), Schri und Schrilichkeit/Writing and Its Use: Ein internationales Handbuch
271
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 272 — #284
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
zeitgenössischer Forschung/An International Handbook of Contemporary Research (Handbücher
zur Sprach- und Kommunikationswissenschaft 10), vol. 2, 1380–1387. Berlin: de Gruyter. • Primus,
Beatrice. 2003. Zum Silbenbegriff in der Schrift-, Laut- und Gebärdensprache – Versuch einer
mediumübergreifenden Fundierung. Zeitschri ür Sprachwissenscha 22. 3–55.
Schrift- und Lautsprache: Eine Übersetzungstheorie
Hartmut Günther
Universität zu Köln
[email protected]
AG9
Mittwoch, 24. 02. 2016, 15:00–15:30, Raum: E 403
Willy Haas hat in seinem Buch Phono-graphic translation von 1970 vorgeschlagen, dass Phoneme und Grapheme in einem Übersetzungsverhältnis
zueinander stehen. Die Beziehung zwischen Strukturen des schriftlichen
Texts und seines mündlichen Äquivalents sind kontingent.
Die von Haas im strukturalistischen Paradigma allein auf die GPK-Ebene
bezogene Theorie lässt sich verallgemeinern, sie gilt ür alle Ebenen einer
Sprache, insbesondere auch ür Syntax und Morphologie. Der Schriftkundige bezieht die beiden Systeme qua allgemeiner Sprachähigkeit aufeinander, wie dies auch der einer Fremdsprache mehr oder weniger Kundige
tut.
Diese Überlegungen werden systematisch entwickelt und auf ihre Grenzen hin überprüft. Es wird weiter überlegt, inwieweit Theorien zum Fremdspracherwerb und zum Schriftspracherwerb kompatibel sind und ob eine
Übertragung in das je andere Feld sinnvoll sein kann, insbesondere im Hinblick auf die Didaktik.
Literatur: • Willy Haas. 1970. Phono-graphic translation. Manchester. (in Teilen bei Google-Books)
272
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 273 — #285
AG 9, Raum E 403
Modalitätsübergreifende und modalitätsspezifische
Strukturaspekte sprachlicher Äußerungen: Auf welchen
Beschreibungsebenen können sich modalitätsspezifische
Unterschiede manifestieren?
Monika Budde
TU Dortmund
[email protected]
Mittwoch, 24. 02. 2016, 15:30–16:00, Raum: E 403
Zu einer natürlichen Sprache wie dem Deutschen oder der Deutschen Gebärdensprache können Äußerungen und damit Varietäten unterschiedlicher Modalität gehören: zum Deutschen gehören z.B. sowohl lautsprachliche als auch voll entwickelte schriftsprachliche Varietäten, während zur
Deutschen Gebärdensprache (die keine Varietät des Deutschen ist)
die Entwicklung schriftsprachlicher Varietäten kaum begonnen hat und
jedenfalls ür die Kommunikation zwischen Gebärdensprachsprechern keine Rolle spielt.
Sprachliche Varietäten lassen sich durch sprachliche Systeme charakterisieren („modellieren“), wobei sprachliche Äußerungen mit Bezug auf
sprachliche Systeme geäußert und verstanden werden. Sprachliche Systeme werden üblicherweise als aus Teilsystemen bestehend rekonstruiert,
die Gegenstand entsprechender Theorien sind. Eine zentrale Frage ist daher, ob sprachliche Systeme ein und derselben Sprache, die zu Varietäten
unterschiedlicher Modalität gehören, sich nur auf der „phonologischen“
Ebene (i.w.S.) unterscheiden, also insbesondere im Wesentlichen gleiche
morphologische und syntaktische Teilsysteme aufweisen (die dann unifiziert werden könnten), oder ob ür entwickelte Schriftsprachen wie das
Deutsche potentiell unterschiedliche, aber aufeinander beziehbare modalitätsspezifische Teilsysteme auf weiteren oder sogar allen Beschreibungsebenen anzunehmen sind. In dem Vortrag werden Argumente ür die 2.,
bisher weitgehend übersehene Auffassung überprüft und einige zentrale
theoretische und didaktische Konsequenzen dieser Auffassung skizziert.
Dabei soll auch gezeigt werden, dass gleichwohl zentrale phonologische,
273
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 274 — #286
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
morphologische und syntaktische Begriffe in einer modalitätsübergreifenden axiomatisch aufgebauten Theorie eingeührt werden können und sollten und modalitätsspezifische Begriffe wie „Phonem (i.e.S.)“, „Graphem“
und „Cherem“ von diesen durch einfache Definitionen abgeleitet werden
können. Mit dem Vortrag wird dabei zugleich der Schnittstellenbegriff weiter präzisiert.
Das Wort als zentrale Einheit der Graphematik
AG9
Karsten Schmidt
Universität Osnabrück
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: E 403
Obwohl phonozentrisch orientierte derivationelle Modelle der Graphematik als überholt gelten dürfen, ist im Bereich der Wortschreibung noch immer die Vorherrschaft eines phonographischen Prinzips – wenn auch unter
korrespondenztheoretischen Vorzeichen – zu beobachten. Bei allen konzeptionellen Unterschieden zwischen einzelnen Ansätzen suggeriert allein
die zentrale Stellung der Untersuchungen zur Phonographie, dass diese ür
die theoretische Modellierung des graphematischen Wortes primär sei.
Der Vortrag erörtert einige erkenntnistheoretische Gründe daür, warum (1) den phonographischen Korrespondenzen in der theoretischen Modellierung von Wortschreibungen eher ein nachrangiger Status eingeräumt
werden sollte und warum (2) im Bereich der Phonographie – im Rahmen
einer nicht-linearen Modellierung (vgl. Primus 2010; Evertz & Primus 2013)
– die Wort-Ebene als primäre Korrespondenzebene anzusetzen ist.
In einem ersten Schritt erfolgt der Nachweis, dass sich die graphematischen Einheiten Wort und Buchstabe unabhängig von Bezügen zum Gesprochenen als sprachliche Einheiten eigenen Rechts bestimmen lassen.
Dies geschieht unter Rückgriff auf die semiologische Formbestimmung von
Ferdinand de Saussure (v.a. 1997), die sich hier als besonders gewinnbringend erwiesen hat. Für Buchstaben ist zeichentheoretisch in erster Linie relevant, dass sie hinreichend differenzierbar sind und als Buchstabenkombi-
274
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 275 — #287
AG 9, Raum E 403
nationen wiederum hinreichend differenzierbare Wort- und Morphemrahmen konstituieren, die, indem sie voneinander abgrenzbar sind, bestimmte
(Wort-, Stamm- und Affix-)Bedeutungen eingrenzbar machen. Zur Identität von Buchstaben gehört dabei ihre – fuß- und silbenstrukturell gerahmte
– Position im Wort.
Das hat – in einem zweiten Schritt – Konsequenzen ür die Konzeption der Phonographie: Den Bezugnahmen zwischen geschriebener und
gesprochener Sprache stellt die Wortebene „den geringsten Widerstand“
(Enderle 2005: 244) entgegen – alle weiteren Korrespondenzen wie die zwischen Lauten und Buchstaben ergeben sich als rekursive Ausdifferenzierungen innerhalb des Rückkopplungsverhältnisses, in dem das graphematische und das phonologische Wort als jeweils (analytisch) autonome Zeichen stehen (vgl. auch Stetter 2005). Es sind die dabei auftretenden Mediatisierungseffekte, die ab einem bestimmten Grad der Literalisierung (ontowie soziogenetisch) die geschriebene und die gesprochene Sprache als Modalitäten eines Sprachsystems erscheinen lassen und die es auch nötig machen, von einer relativen Autonomie zu sprechen.
Literatur: • Enderle, Ursula (2005): Autonomie der geschriebenen Sprache? Zur eorie phonographischer Beschreibungskategorien am Beispiel des Deutschen. Berlin: Erich Schmidt. • Evertz,
Martin & Beatrice Primus (2013). The graphematic foot in English and German. Writing Systems Research 5.1: 1-23. • Primus, Beatrice (2010). Strukturelle Grundlagen des deutschen
Schriftsystems. In Bredel, Ursula, Astrid Müller & Gabriele Hinney (eds.), Schrisystem und
Schrierwerb: linguistisch – didaktisch – empirisch, 9-45. Tübingen: Niemeyer. • de Saussure, Ferdinand (1997): Linguistik und Semiologie. Notizen aus dem Nachlaß. Texte, Briefe und
Dokumente. Gesammelt, übersetzt und eingeleitet von Johannes Fehr. Frankfurt a.M: Suhrkamp. • Stetter, Christian (2005): System und Performanz. Symboltheoretische Grundlagen von
Medientheorie und Sprachwissenscha. Weilerswist: Velbrück Wissenschaft.
275
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 276 — #288
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Nur ein Reflex der Morphosyntax? Das graphematische
Wort in Norm und Gebrauch
Vilma Symanczyk Joppe
Bergische Universität Wuppertal
[email protected]
Mittwoch, 24. 02. 2016, 17:00–17:30, Raum: E 403
AG9
Das graphematische Wort – definiert als Schreibung, die durch Satzzeichen
bzw. Spatien von benachbarten Schreibungen abgegrenzt ist, selbst aber
keine solchen enthält – gilt in der deutschen Schriftsystemforschung als
bloßer Reflex des (morpho-)syntaktischen Wortes. So charakterisiert Maas
(1992) Spatien als „syntaktische Sollbruchstellen“, und die Literatur zu orthographischen Zweifelsällen der Getrennt- und Zusammenschreibung
(GZS) bedient sich gern morphosyntaktischer Tests (z.B. Günther 1997,
Fuhrhop 2007). Jacobs (2005) analysiert das Kernsystem der deutschen GZS
als Wettbewerb zweier hierarchisierter Beschränkungen: Die niedrigere
fordert die Getrenntschreibung von syntaktischen Teilausdrücken, die höhere erzwingt eine Zusammenschreibung von morphologischen Bildungen. All diesen Ansätzen ist eine Annahme gemein: Was ein graphematisches Wort ist, wird in der Syntax bzw. Morphologie entschieden und über
ein minimales Inventar an Interface-Regeln in die Schrift transferiert. Eigens ür das Schriftsystem angesetzte Regeln sind überflüssig.
Die obigen Modelle basieren allerdings sämtlich auf orthographischen
Schreibungen. Jacobs (2005: 6) weist explizit auf ein Phänomen des Gegenwartsdeutschen hin, das sich seiner Analyse entzieht, sogenannte „Komposita Getrenntschreibungen“ (KG), wie sie u.a. auch Dürscheid (2000) behandelt hat. Obwohl es sich bei diesen KG in morphosyntaktischer Hinsicht
eindeutig um Wörter handelt, werden sie mit internen Spatien realisiert.
Auf Grundlage einer Untersuchung von knapp 15.000 Komposita aus
normfernen Texten soll gezeigt werden, dass a) KG in jenen Texten in einem erklärungsbedürftig hohen Umfang vorkommen, b) ihr Auftreten sich
auf bestimmte einschlägige Faktoren zurückühren lässt und c) in diesen
Fällen sogar versierte Schreiber KG produzieren. Eine adäquate Erklärung
der KG aber erfordert eine Ergänzung des bestehenden Regelapparates.
276
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 277 — #289
AG 9, Raum E 403
Mindestens ein Teil der erforderlichen Zusatzregeln ist „substanzbasiert“
im Sinne von Primus (2003) – sie lassen sich nur über dem sprachlichen
Subsystem „Schrift“ eigene Gesetzmäßigkeiten und die Bindung an den
visuellen Wahrnehmungsbereich plausibilisieren. Der Vortrag plädiert in
diesem Sinne ür eine nicht bloß derivationelle Analyse des graphematischen Wortes.
Literatur: • Dürscheid, Christa. 2000. Verschriftungstendenzen jenseits der Rechtschreibreform. Zeitschri ür Germanistische Linguistik 28/2, 237-247. • Fuhrhop, Nanna. 2007. Zwischen Wort und Syntagma. Zur grammatischen Fundierung der Getrennt- und Zusammenschreibung. Tübingen: Niemeyer. • Günther, Hartmut. 1997. Zur grammatischen Basis der Getrennt-/Zusammenschreibung im Deutschen. In Dürscheid, Christa/Ramers, Karl Heinz/Schwarz,
Monika: Sprache im Fokus. Festschrift ür Heinz Vater. Tübingen: Niemeyer. 3—16. • Jacobs,
Joachim. 2005. Spatien. Zum System der Getrennt- und Zusammenschreibung im heutigen Deutsch.
Berlin/New York: de Gruyter. • Maas, Utz. 1992. Grundzüge der deutschen Orthographie. Tübingen: Niemeyer. • Primus, Beatrice. 2003. Zum Silbenbegriff in der Schrift-, Laut- und Gebärdensprache – Versuch einer mediumunabhängigen Fundierung. Zeitschri ür Sprachwissenscha 22. 3–55
Expressive Intensitätspartikeln in gesprochener und
geschriebener Sprache
Fabian Renz
Universität Tübingen
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:00, Raum: E 403
Phänomen Expressive Intensitätspartikeln wie voll, total oder übelst, die im
gesprochenen Deutsch insbesondere junger Sprecher hochfrequent auftreten, weisen einige bemerkenswerte Fähigkeiten auf: Sie können zum einen,
wie in (1a), bei der Modifikation von Adjektiven in DP-externer Position
auftreten (Gutzmann & Turgay 2012, 2015) und zum anderen, wie in (1b),
auf diese Weise sogar Nomen modifizieren. Gleiches gilt erstaunlicherweise auch ür zahlreiche intensivierende Präfix(oid)e wie mega-, sau-, ur-,
über-, end(s)- oder hammer-.
277
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 278 — #290
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
(1)
a.
b.
Gestern war bei uns voll/übelst/mega/sau die geile Party.
Die Party vorgestern war aber voll/übelst/mega/sau der Reinfall.
Fragestellung Angesichts dieser Daten stellt sich die Frage, welcher syntaktischen Kategorie vermeintlich gebundene Morpheme wie mega oder
sau zuzurechnen sind, gerade wenn sie DP-intern modifizieren. Bei einer
Verschriftlichung in diesen Fällen zeigt sich nämlich eine Varianz, was die
Getrennt- und Zusammenschreibung anbelangt, wie folgende Webbelege
zeigen:
(2)
AG9
Es war mal wieder eine mega geile Party.
https://de-de.facebook.com/clubelectribe/posts/10150530608620378
(3)
Es war wieder eine megageile Party, hat irre Spass gemacht.
https://de-de.facebook.com/gip.party/posts/349373761774548
Ansatz In meinem Vortrag möchte ich daür argumentieren, dass aus Präfix(oid)en wie mega und sau Intensitätspartikeln grammatikalisiert wurden
(vgl. Androutsopoulos 1998: 111). Neben den Erkenntnissen aus (1) wird
zur Beweisührung unter anderem auch die Orthographie herangezogen:
Im DECOW (14AX)-Webkorpus lassen sich z.B. ür mega_geil mehr Belege mit Getrennt- (1050) als mit Zusammenschreibung (818) finden. Ob dabei in der Mehrheit der Fälle eine Korrelation mit dem jeweils gegebenen
bzw. vom Sprecher „beabsichtigten“ kategorialen Status besteht, ist fraglich: Wo die gesprochene Sprache mittels Intonation oder Emphase Möglichkeiten zur Disambiguierung zur Verügung stellt, bleibt in der geschriebenen Sprache nur die Setzung eines Spatiums oder der Verzicht darauf.
Lediglich die Verwendung von Majuskeln (eine MEGA geile Party) kann
zusätzlich hilfreich sein, disambiguierende Betonung abzubilden, um eine Intensitätspartikel als solche auszuweisen. Unter Berücksichtigung all
dieser Aspekte möchte ich auf der Basis einer DECOWKorpusstudie untersuchen, wie sich mega, sau und vergleichbare Partikeln bei der Intensivierung von Adjektiven und Nomen (z.B. mega Trend vs. Mega-Trend) in der
geschriebenen Sprache im Verhältnis zu unzweifelhaften Intensitätspartikeln sowie zu unzweifelhaften Präfix(oid)en verhalten.
278
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 279 — #291
AG 9, Raum E 403
Literatur: • Androutsopoulos, Jannis K. 1998. Deutsche Jugendsprache. Frankfurt: Lang. Gutzmann, Daniel & Turgay, Katharina. 2012. Expressive intensifiers in German: syntax-semantics mismatches. Empirical Issues in Syntax and Semantics 9 (ed. Christopher Piňón). 149–166.
• Gutzmann, Daniel & Turgay, Katharina. 2015. Expressive intensifiers and external degree
modification. e Journal of Comparative German Linguistics 17.3. 185–228.
Verknüpfungsstrukturen listenmodaler Texte:
Schnittstellenphänomene und modalitätsspezifische
Ausprägungen
Tilo Reißig
Universität Hildesheim
AG9
[email protected]
Donnerstag, 25. 02. 2016, 09:30–10:00, Raum: E 403
Für die Verknüpfung sprachlicher Einheiten haben sich in der Syntax die
Begriffe Subordination und Koordination etabliert. Sie stellen modalitätsunabhängige Verknüpfungsstrukturen des Sprachsystems dar. Betrachtet
man schriftsprachliche Erscheinungen, bei denen die räumliche Anordnung der Schriftzeichen als sekundäres Zeichensystem fungiert, kann man
feststellen, dass die Begriffe Subordination und Koordination, so wie sie
in der Linguistik Verwendung finden, nicht ausreichen, um alle sprachlichen Verknüpfungsstrukturen zu beschreiben. So weisen Listen, Tabellen, Pools und Mindmaps einzigartige Möglichkeiten der Verknüpfung auf,
die bislang in der linguistischen Theoriebildung unberücksichtigt blieben.
Im Rahmen meines Vortrags möchte ich die Verknüpfungsstrukturen von
Listen und Tabellen näher beschreiben. Hierbei werde ich zum einen auf
Schnittstellenphänomene eingehen und zeigen, dass Listenkomplemente
additiv verknüpft sind (vgl. Reißig 2015) – das Äquivalent hierzu könnte
in der gesprochenen Sprache die spezifische „Listenprosodie“ sein (vgl. zur
Intonationsforschung z. B. Gilles 2005; Selting 2004). Zum anderen werde
ich zeigen, dass Tabellen eine modalitätsspezifische Verknüpfungsstruktur
aufweisen, da sie die ür die gesprochene Sprache unbekannte räumliche
279
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 280 — #292
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Struktur auf eine Art nutzen, wie sie unmöglich in der gesprochenen Sprache dargestellt werden könnte.
Literatur: • Gilles, Peter. 2005. Regionale Prosodie im Deutschen: Variabilität in der Intonation von Abschluss und Weiterweisung. Berlin: de Gruyter. • Reißig, Tilo. 2015. Typographie
und Grammatik. Untersuchung zum Verhältnis von Syntax und Raum. Tübingen: Stauffenburg.
• Selting, Margret. 2004. Listen: Sequenzielle und prosodische Struktur einer kommunikativen Praktik – eine Untersuchung im Rahmen der Interaktionalen Linguistik. Zeitschri ür
Sprachwissenscha 23. 1–46.
AG9
Right dislocations and afterthoughts: The effects of
punctuation and discourse structure on prosody
Janina Kalbertodt1 , Beatrice Primus1 & Petra B. Schumacher1
1
Universität zu Köln
[email protected]
Donnerstag, 25. 02. 2016, 10:00–10:30, Raum: E 403
In this study, we are interested in the effects of discourse structure and
punctuation on the prosodic marking of two distinct constructions at the
right periphery, right dislocation (RD) and afterthought (AT). Both parameters are likely to affect prosodic marking, as discourse structure ensures the
appropriateness of a certain marking, while punctuation correlates with
phrasing. The literature disagrees on the nature of the correspondence of
these parameters.
RDs differ from ATs (among others) in terms of punctuation, with RDs
being predominantly marked with a comma, while ATs are separated with
a full stop (Kalbertodt 2011), and also in terms of their discourse status,
with RD-elements being in focus, whereas AT-elements are merely activated (Fretheim 1995). In terms of prosody, Kalbertodt (2013) showed that
RDs and ATs differ in the strength of phrasal break before the last constituent ((weaker) ip-boundary in RDs, (stronger) IP-boundary in ATs) but
not in accent strength (both RDs and ATs were marked with a nuclear accent). Instead of accentual strength, RDs and ATs seemed to differ more
280
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 281 — #293
AG 9, Raum E 403
in the amount of pitch range reduction, which was significantly greater in
RDs. On the basis of this previous research, we investigate which parameter (punctuation or discourse) triggers the prosodic realization of RDs and
ATs. We test the following hypotheses: 1) if discourse structure guides the
identification of RD and AT, then these constructions will be prosodically
marked as in the previous study, irrespective of punctuation; 2) if punctuation triggers identification, then sentences with a comma should be prosodically marked with an ip-boundary and sentences with a full stop with an
IP-boundary, irrespective of a non-matching discourse structure (cf. e.g.
Bredel 2011).
In our experiment, we used a 2x2 design with the factors discourse and
punctuation, with a total of 32 critical items. 24 participants were asked to
read out these items. For the analysis, strength of phrasal break and accent
were examined and pitch range was calculated using Praat. Results show
that both boundary strength and accent strength are strongly influenced
by discourse structure rather than punctuation. The amount of pitch range
reduction, however, was not found to differ significantly between RDs and
ATs, which was unexpected. Importantly, these results strengthen an account of an indirect correspondence between punctuation and prosody
mediated by syntactic structure, as for instance proposed by Kirchhoff (in
press).
References: • Bredel, Ursula (2011): Interpunktion. Heidelberg: Universitätsverlag Winter GmbH.
• Fretheim, Thorstein (1995): “Why Norwegian right-dislocated phrases are not afterthoughts”.
Nordic Journal of Linguistics 18, 31-54. • Kalbertodt, Janina (2011): Literarische Stilistik und
syntaktische Konstruktion am Beispiel von Rechtsversetzungen. (Bachelor Thesis, University of
Cologne). • Kalbertodt, Janina (2013): Rechtsversetzung und Aerthought im Roman: Eine empirische Studie. (Master Thesis, University of Cologne). • Kirchhoff, Frank (in press): “Interpunktion und Intonation”. In: Primus, Beatrice / Domahs, Ulrike (eds.): Handbuch Sprachwissen: Laut – Gebärde – Buchstabe. Berlin: de Gruyter.
281
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 282 — #294
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Interpunktion und Intonation bei Interjektionen im
Deutschen
Ilka Huesmann1 & Frank Kirchhoff1
1
Universität zu Köln
[email protected]
Donnerstag, 25. 02. 2016, 10:30–11:00, Raum: E 403
AG9
Aufgrund ihrer formalen und funktionalen Exzentrik stellen Interjektionen in der Linguistik einen besonderen Forschungsgegenstand dar. Aus
syntaktischer Sicht handelt es sich nach Mehrheitsmeinung bei Interjektionen um eine Herausstellungsform, die lediglich lose mit dem Trägersatz verknüpft ist (vgl. u.a. Peterson 1999 und Altmann & Hofmann 2004).
Im syntaktischen Kommamodell nach Primus (2007) lizenziert diese nichtsubordinative Verknüpfungsart ein Komma in der geschriebenen Sprache.
Eine erste Stichprobe in COSMAS II von Kirchhoff (i.E.) zur Kommasetzung
von ach-Interjektionen zeigte jedoch eine inkonsistente Interpunktion dieser Interjektionen, die durch den Einfluss anderer nicht-syntaktischer Faktoren begründet werden könnte.
Da Interjektionen primär in der gesprochenen Sprache verwendet werden, wollen wir in diesem Vortrag diskutieren, welchen Einfluss intonatorische Faktoren wie Sprechpausen, Grenztöne und phrasenfinale Dehnung auf die Interpunktion von Interjektionen haben. Vor diesem Hintergrund stellen wir eine umfangreiche Korpusauswertung der deutschsprachigen E-Book- und Hörbuchfassung des ersten Harry Potter-Romans
vor. Die Auswertung des E-Books bestätigt eine inkonsistente Interpunktion der unterschiedlichen Interjektionen. Mithilfe einer Praat-Analyse des
Hörbuchs wird geklärt, inwiefern die Intonationsphrasengrenzen der Interjektionen mit der Interpunktion der geschriebenen Fassung korrelieren.
Die Ergebnisse sollen darüber Aufschluss geben, inwiefern Intonation und
Interpunktion die zugrundeliegende Syntax von Interjektionen abbilden
und darüber hinaus hinterfragen, ob ihr Status als Herausstellungsform
gerechtfertigt ist.
282
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 283 — #295
AG 9, Raum E 403
Literatur: • Altmann, Hans & Hofmann, Ute. 2004. Topologie ürs Examen. Verbstellung,
Klammerstruktur, Stellungsfelder, Satzglied- und Wortstellung. Wiesbaden: VS Verlag ür Sozialwissenschaften. Kirchhoff, Frank. i.E. Interpunktion und Intonation. In: Primus, Beatrice /
Domahs, Ulrike (Hgg.) Handbuch Sprachwissen: Laut – Gebärde – Buchstabe. Berlin: de Gruyter. • Peterson, Peter. 1999. On the boundaries of syntax: Non-syntagmatic relations. In: Collins, Peter / Lee, David (eds.) e clause in English. Amsterdam: John Benjamins, 229-250. • Primus, Beatrice. 2007. The typological and historical variation of punctuation systems: Comma
constraints. Wrien Language and Literacy, 10(2), 103-128.
Zur (ortho)grafischen Markierung von sekundären
Inhalten: Eine empirische Studie
1
AG9
2
Katharina Turgay & Daniel Gutzmann
1
Universität Koblenz-Landau, 2 Universität zu Köln/ Goethe-Universität
Frankfurt
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: E 403
Im Deutschen gibt es verschiedene Konstruktionen, die syntaktisch und
semantisch desintegriert sind und die einen sekundären Inhalt kommunizieren. Während in der gesprochenen Sprache phonologische Hinweise
verwendet werden können, um den desintegrierten Status dieser Konstruktionen anzuzeigen, muss die geschriebene Sprache auf andere Mittel zurückgreifen. Das Interpunktionssystem stellt dazu zahlreiche Indikatoren
zur Verügung: Kommata, Doppelpunkte, Gedankenstriche, verschiedene
Klammern etc. können den sekundären Status bestimmter Inhalte anzeigen. Anhand eines Korpus von (nicht edierten) geschriebenen Texten wollen wir ür diesen Bereich ein funktionales Profil ür die verschiedenen
Interpunktionszeichen in diesem Vortrag entwickeln. Es zeigt sich, dass es
keine 1:1-Beziehung zwischen den Interpunktionszeichen und den Funktionen gibt, die sie in Bezug auf sekundäre Inhalte erüllen können. Die
Interpunktionszeichen überlappen sich zwar in ihren Funktionen, unterscheiden sich aber in den syntaktischen Kontexten, in denen sie verwendet
werden können. Wir unterscheiden verschiedene Faktoren, die relevant ür
283
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 284 — #296
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Anwendbarkeit der Interpunktionszeichen sind. Erstens, die Position der
sekundären Konstruktion in Bezug auf den Rest des Satzes: Vor oder hinter
Interpunktionszeichen, oder eingebettet durch diese.
AG9
(1)
Kurzum: Man hat sich an das „überüllte“ Leben gewöhnt.
(2)
Sitzen wir im Taxi unterhalten wir uns nicht mit dem Fahrer, sondern bedienen unser iPhone oder unseren Laptop – natürlich mit
MP3-Player oder iPod im Ohr.
(3)
Der Drang nach Macht einzelner Menschen (Lenin/Hitler) brachte
die einfache Bevölkerung so oft in große Probleme.
Unsere Daten zeigen zudem eine Variation bezüglich der Beziehung des
sekundären Gehalts und dessen Anker. Der sekundäre Gehalt kann eine
Exemplifikation des Ankers darstellen und prinzipiell ür diesen eingesetzt
werden, oder er stellt eine Modifikation dar und ist folglich additiv zu interpretieren. Als weitere Funktion können wir eine diskursstrukturierende
Rolle des sekundären Gehalts annehmen. Unsere Korpusanalyse zeigt jedoch, dass nicht alle Funktionen und Positionen gleichermaßen alle Interpunktionszeichen erlauben. Die Unterschiede lassen sich dabei im Wesentlichen auf zwei Faktoren zurückühren: die direktionale Orientierung der
Interpunktionszeichen und ob die Zeichen paarig auftreten (können oder
müssen). So ist der Doppelpunkt bevorzugt vorwärtsorientiert und kann
nicht paarweise auftreten. Das heißt, er kann nicht zur zentralen Einbettung von sekundären Material genutzt werden und das sekundäre Material
steht bevorzugt rechts. Dies steht im Kontrast zu Klammern, die in Paaren
kommen müssen und meistens rückwärtsgewandt genutzt werden.
(4)
*Der Drang nach Macht (Lenin/Hitler) einzelner Menschen brachte…
Dieses Verhalten hingegen kontrastiert mit Kommata, die, wenn sie paarweise auftreten, sekundäres Material umspannen, das bevorzugt vor dem
Anker steht.
(5)
*der Moter des, neuen, Autos
Die systematische Anwendungen von Überlegungen wie den zuvor skiz-
284
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 285 — #297
AG 9, Raum E 403
zierten ührt zu einem funktionalen Profil der verschiedenen Interpunktionszeichen, das die Prinzipien hervorhebt, die den Gebrauch von Interpunktionszeichen zur Markierung sekundären Inhalts beinflussen.
Homophone und Heterographen
Kristian Berg
Universität Oldenburg
[email protected]
Donnerstag, 25. 02. 2016, 12:00–12:30, Raum: E 403
Verschiedene lexikalische Einheiten haben manchmal dieselbe Form: So
bezeichnet /baŋk/ im Deutschen sowohl das Geldinstitut als auch das Sitzmöbel; es handelt sich hier um homonyme Formen. Differenziert man weiter nach dem Medium, so handelt es sich beim Beispiel oben um homophone und homographische Formen, und damit um Homonyme im engeren
Sinne. Daneben treten aber auch a) heterophone homographische Formen
auf wie z.B. <modern> ‚zeitgemäß‘/ ‚verfaulen‘ und b) homophone heterographische Formen wie <Lid>/<Lied>.
Es kann nun beobachtet werden, dass der Typ <modern> im Deutschen
seltener ist als der Typ <Lid>/<Lied>. Daraus ist z.T. auf ein „Homonymieprinzip“ (auch „lexikalisches“ oder „semantisches“ Prinzip) geschlossen worden, vgl. z.B. die Übersicht bei Rahnenührer (1980: 248). Das Geschriebene hat aus dieser Perspektive eine stärkere Tendenz, lexikalische
Einheiten distinkt auszuzeichnen, als das Gesprochene. Eisenberg (4 2013:
317) weist allerdings darauf hin, dass dieses Prinzip im Deutschen marginal
ist und zahlreiche Ausnahmen hat: Obwohl beispielsweise die Homonyme
Weide, Kiefer und Ton graphisch differenziert werden könnten, sind sie homographisch. Doch in wie vielen Fällen werden Homophone differenziert,
in wie vielen unterbleibt eine solche Differenzierung?
Diese Frage soll mithilfe der Datenbank CELEX (Baayen et al. 1995) beantwortet werden. Da dieses Korpus (unter anderem) über graphematische
und phonologische Repräsentationen von Wörtern verügt, kann ein Korpus homophoner Wortformen extrahiert werden, das die Grundlage der
285
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 286 — #298
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
Untersuchung ist. In einem zweiten Schritt kann untersucht werden, mit
welchen Mitteln die Differenzierung geleistet wird.
Literatur: • Baayen, R. Harald, Piepenbrock, R. & Gulikers, L. (1995). The CELEX lexical database (CD-ROM). Linguistic Data Consortium, University of Pennsylvania, Philadelphia, PA.
• Eisenberg, P. (4 2013). Grundriss der deutschen Grammatik. Band 1: Das Wort. Stuttgart: Metzler. • Rahnenührer, I. (1980). Zu den Prinzipien der Schreibung des Deutschen. In: Nerius,
D. & Scharnhorst, J. (Hgg.). eoretische Probleme der deutschen Orthographie. Berlin: Akademie-Verlag, S. 231-259.
AG9
Silbenkerne im Spannungsfeld zwischen Laut und Schrift
oder Silbenkerne als modalitätsübergreifende Einheit
Nanna Fuhrhop
Universität Oldenburg
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: E 403
Über verschiedene Schriftsysteme (hier zunächst Deutsch, Englisch, Französisch, Niederländisch) ist recht auällig, dass sich die Silbenkerne der
phonographischen Abbildbarkeit sehr deutlich entziehen. Neben der fast
schon traditionellen Diskussion über so genannte Dehnungs- und Schärfungsgraphien sind die Diphthonge augenällig – so alleine schon die Tatsache, dass die ‚Diphthongigkeit‘ in keiner Richtung (zumindest auf den
ersten Blick nicht) systematisch korrespondiert. Das Französische hat (zumindest keine eindeutigen) Sprechdiphthonge, aber vermutlich eine Reihe von Schreibdiphthongen; im Englischen können Sprechdiphthonge mit
Schreibmonophthongen korrespondieren (/ai/ <-> <i>) oder umgekehrt (<ea>
<-> /i/) usw. Dies ührt auch im Fremdwortbereich des Deutschen zu vielältigen Beziehungen, zum Beispiel bei den im Deutschen bekannten Schreibdiphthongen wie <eu> in Friseur, <au> und <ai> in au Lait und <ei> in
beige. In diesem Vortrag soll es also – neben einer Beschreibung und möglichst systematischen Erfassung der sprachlichen Daten – auch um die Frage gehen, ob dem Silbenkern an und ür sich eine besondere Autonomie
zukommt.
286
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 287 — #299
AG 9, Raum E 403
One phonotactic restriction for reading and listening: The
case of the no geminate constraint in German
Silke Hamann
University of Amsterdam
[email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: E 403
In this talk, I present a linguistic model that formalizes a cross‐modal approach to reading and listening by employing the phonotactic restrictions
that restrain the process of speech perception for the process of reading.
Speech perception and reading both involve an arbitrary, language‐specific
mapping of an input form onto a phonological surface form. In the case of
speech perception, the input is an auditory form, in the case of reading it
is a visual/orthographic form. For both mappings, the output phonological surface form is restricted by the same native phonotactic constraints.
For the speech perception process, this has been formalized in Optimality Theory with the so‐called perception grammar (Boersma 2007), involving cue constraints for the mapping of auditory forms onto phonological
output and structural constraints for the phonotactic restrictions. Hamann
& Colombo (2015) transferred this idea to the reading process, and formalized a reading grammar, with orthographic constraints for the mapping between orthographic input and phonological output, and the same
structural constraints for phonotactics as employed already in the perception grammar. With a combined reading and perception grammar, they
accounted for the interaction of orthographic and perceptual influences in
Italian loanwords borrowed from English.
In the present talk, I illustrate the cross‐modal application of a phonotactic no geminate constraint, disallowing geminate consonants within a
morpheme in German. In speech perception, non‐native forms with geminates like Italian [latːe] are perceived as /latə/ due to this constraint. In
the reading process, native (and non‐native) orthographic forms with double consonantal graphemes like <Wall> and <Teller> are read as /wal/ and
/tɛlɐ/, not */walː/ and */tɛlːɐ/. I argue that this is due to the same phonotactic
constraint rather than orthographic mappings that reduplicate phonotactic
287
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 288 — #300
AG 9: Geschriebene und gesprochene Sprache als Modalitäten …
knoweldge (as e.g. “two (or more) identical consonant graphemes should be
mapped onto a single consonant”, similar to a constraint proposed by Neef
2012: 221). A formalization in terms of Optimality theory will be provided
for both modalities.
References: • Boersma, Paul (2007). Some listener‐oriented accounts of h‐aspiré in French.
Lingua 117: 1989–2054. Hamann, Silke & Ilaria E. Colombo (2015, submitted). e role of orthography in the borrowing of English intervocalic consonants into Italian: a formal account.
Manuscript University of Amsterdam. • Neef, Martin (2012). Graphematics as part of a modular theory of phonographic writing systems. Writing Systems Research 4: 214–228.
AG9
Gesprochen, geschrieben, gesungen: Sprache in
Werbespots
Sabine Wahl
Universität Bremen/Katholische Universität Eichstätt-Ingolstadt
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: E 403
In Werbespots als multimodalen Kommunikaten begegnet den Rezipienten
die Zeichen-modalität Sprache wiederum in mehreren „Modalitäten“: gesprochen, geschrieben und gesungen. Deshalb eignen sich diese Kommunikate in besonderem Maße ür die Untersuchung von möglichen SchnittstellenPhänomen zwischen den Sprachmodalitäten. Zu den Besonderheiten bei
der Kreation von Werbespots gehört jedoch, dass die gesprochene Sprache – wie beispielsweise bei anderen Filmen und Hörspielen auch – nicht
spontan hervorgebracht wird, sondern zunächst sehr bewusst entwickelt
und in schriftlicher Form fixiert wird, bevor sie dann wieder gesprochen erklingt. Dasselbe Verfahren gilt ür Songtexte jeder Art und damit auch ür
Texte, die in Werbespots gesungen werden – seien es originale Songtexte,
Werbelieder oder kurze Jingles.
Dieser Beitrag untersucht anhand eines umfangreichen diachronen Korpus deutscher Werbespots, wie die Zeichenmodalität Sprache im Laufe der
Zeit eingesetzt wird. Dadurch werden auch Einblicke in die Arbeit mit multimodalen Korpora und Datenbanken gegeben. Zentrale Fragen der Studie
288
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 289 — #301
AG 9, Raum E 403
lauten: Wie ist das Verhältnis von gesprochener, geschriebener und gesungener Sprache? Welche Sprachen und Sprachvarietäten werden jeweils
eingesetzt? Wie stark unterscheiden sich die gesprochene, geschriebene
und gesungene Sprache voneinander? Welche Schnittstellen-Phänomene
lassen sich beobachten? Treten in der vorher schriftlich entwickelten gesprochenen Sprache der Werbespots trotzdem vermehrt Strukturen auf, die
in der sprachwissenschaftlichen Forschung als typisch ür spontan gesprochene Sprache gelten? Wie kreativ geht die Sprache in Werbespots mit dem
Sprachsystem um? Vor allem der letztgenannte Aspekt macht Werbespots
und ihre Sprache auch zu einem interessanten Thema ür die Sprachdidaktik.
289
AG9
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 290 — #302
Selected titles
Evaluating Cartesian Linguistics
From Historical Antecedents to Computational
Modeling
By Christina Behme
Frankfurt am Main, 2014. 267 pp.
Potsdam Linguistic Investigations. Vol. 12
Edited by Peter Kosta, Gerda Haßler, Teodora Radeva-Bork,
Lilia Schürcks and Nadine Thielemann
hardback • ISBN 978-3-631-64551-2
€D 59.95 / € A 61.60 / € 56.– / CHF 68.– / £ 45.– / US-$ 72.95
eBook • ISBN 978-3-653-03710-4
€D 66.65 / € A 67.20 / € 56.– / CHF 71.65 / £ 45.– / US-$ 72.95
The book evaluates Noam Chomsky’s contributions to linguistics and focuses on the historic justification for Cartesian
Linguistics, the evolution of Chomsky’s theorizing, empirical language acquisition work, and computational modeling
of language learning. It is shown that calling Chomsky’s linguistic Cartesian cannot be historically justified.
Building Bridges for Multimodal
Research
International Perspectives on Theories
and Practices of Multimodal Analysis
Edited by Janina Wildfeuer
Frankfurt am Main, 2015. 380 pp., 12 tables, 43 graphs
Sprache – Medien – Innovationen. Vol. 7
Edited by Jens Runkehl, Peter Schlobinski and Torsten Siever
hardback • ISBN 978-3-631-66266-3
€D 69.95 / € A 71.90 / € 65.40 / CHF 79.– / £ 52.– / US-$ 85.95
eBook • ISBN 978-3-653-05431-6
€D 77.85 / € A 78.50 / € 65.40 / CHF 83.25 / £ 52.– / US-$ 85.95
The book takes differences in multimodality research as
a starting point to discuss old and new theoretical, methodological as well as analytical ideas for building bridges
between various disciplines and approaches.
Please send your order to :
Peter Lang AG, P. O. Box 350, 2542 Pieterlen, Switzerland
[email protected]
www.peterlang.com
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 291 — #303
Arbeitsgruppe 10
Morphological effects on word order from a
typological and a diachronic perspective
Þórhallur Eyþórsson1 , Hans-Martin Gärtner2 &
Tonjes Veenstra3
1
University of Iceland, 2 RIL-HAS Budapest, 3 ZAS Berlin
AG10
[email protected], [email protected], [email protected]
Raum: G 308
Workshop description
The overall topic of this workshop is to re-examine the relation between inflectional morphology and word order, and how changes in the morphological component lead to changes in the syntactic component, and vice versa.
It will be considered from a typological and a diachronic perspective, with
particular emphasis on language contact and creolization. There is a longstanding tradition in historical linguistics from Meillet (1908) onwards attributing syntactic change to the loss of inflectional morphology (e.g. Jespersen 1922, Lightfoot 1979, Weerman 1989). In a specific case, this line of
research has been formalized in the last decades as the Rich Agreement
Hypothesis (RAH: Holmberg & Platzack 1991, Roberts 1993, Rohrbacher
1994, Vikner 1995, Bobaljik & Thráinsson 1998). The discussion has centered on the question whether the correlation is bi-directional (strong version: rich inflection V-to-I movement) or mono-directional (weak version:
rich inflection ⇒ V-to-I movement). The strong bi-directional version of the
RAH has recently been resurrected by Koeneman & Zeijlstra (2014), based
on the micro-variation within Germanic and Romance languages, but with
291
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 292 — #304
AG 10: Morphological effects on word order from a typological …
AG10
strong typological claims that go beyond these particular language families: on this version of the RAH, there are no languages that have rich
inflection but not V-to-I, and no languages that lack rich inflection but still
have V-to-I. Although well-known exceptions seem to abound, these have
neither been properly catalogued, nor tested, let alone accounted for.
The main goal of this workshop, therefore, is to create a platform to stimulate a structured discussion of the well-defined claims on the correlation
between morphology and word order, on the basis of diachronic data and
creole data. As to the diachronic evidence, there seems, for example, to be
a time lag in the loss of inflectional morphology and the disappearance
of V-to-I in Scandinavian languages (cf. Sundquist 2002), which begs the
question how much time is required between the trigger of the change and
its syntactic manifestation. As to the creole languages, some of them seem
to exhibit V-movement although they do not have rich agreement, whereas
others potentially count as having rich agreement but no V-movement (cf.
Baptista 2002, McWhorter 2013, Roberts 1999, Veenstra 1996, 2008).
The topics to be addressed include:
(i) When does the verbal and/or pronominal (subject clitic) paradigm
count as having RA, and are these subtypes of RA independent of,
or interdependent on, each other?
(ii) How long can the time lag between different diachronic stages be, in
order to still be able to speak of a causal relation of morphology and
syntax between such stages?
(iii) Can language contact account for apparent exceptions to the RAH?
(iv) What is the typological and historical evidence bearing on the RAH
outside of modern Germanic and Romance, in particular in creole
languages, and how is it to be interpreted?
The workshop will bring together scholars from theoretical linguistics,
historical linguistics, creole studies and language contact studies to address
the above issues.
292
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 293 — #305
AG 10, Raum G 308
The Rich Agreement Redux
Olaf Koeneman1 & Hedde Zeijlstra2
U. Nijmegen, 2 U. Göttingen
1
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 14:00–15:00, Raum: G 308
In Koeneman & Zeijlstra (2010, 2014), we argued on the basis of various
Indo-European languages that V to I movement in natural language adheres to a bi-directional generalization, thereby rehabilitating the Rich
Agreement Hypothesis in its strongest form: V moves to I if and only if
agreement morphology is rich. The agreement paradigm is rich if it expresses the same features as those that are part of the world’s most impoverished argument systems (i.e. those encoding two person and one number
distinctions). In this talk, we will shortly present our theory behind this
generalization but mainly concentrate on the predictions it makes and the
data that support it:
(i) The proposed generalization should hold universally and not just for
Indo-European (or worse: Germanic) languages.
(ii) When a language becomes poor because of deflection, V to I is either lost or the patterns associated with V to I movement (namely
V-Adv and V-Neg) lead to a reanalysis of the grammar that allows
the learner to stay within the boundaries of the bi-directional generalization.
(iii) V to I movement cannot be acquired by children before the featural
distinctions that make the paradigm rich.
In the second half of the talk, we will (i) point out which predictions
the proposal does not make, so as to flesh out our proposal more clearly,
and (ii) discuss recent literature that takes issue with our proposal (such as
Harbour 2013 and Heycock & Wallenberg 2013).
293
AG10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 294 — #306
AG 10: Morphological effects on word order from a typological …
Hand in hand or each on one’s own? On the connection
between morphological and syntactic change
Eric Fuß
IDS Mannheim
[email protected]
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: G 308
AG10
It is a long-standing idea that syntactic change may be triggered by changes
affecting properties of inflectional morphology. Standard examples include
the impact of the loss of verbal inflections on the availability of verb movement and pro-drop, or the relation between the loss of case marking and
changes affecting basic word order. However, it has not gone unnoticed
that the diachronic correlation between syntax and morphology is sometimes less tight than one might suppose. First, it is a well-known fact that
a given morphological change and its supposed syntactic effects may be
separated by a considerable temporal gap. Second, there are cases where a
proposed causal link between morphological and syntactic change does not
hold up against scrutiny (cf. e.g. Joseph 1983 vs. Anttila 1972 on the loss of
nonfinite constructions in Greek). Finally, it has been pointed out that syntactic change may take place despite conflicting morphological evidence.
Phenomena of this last type have led some researchers to claim that the
relationship between syntactic and morphological change actually holds
in the opposite direction, i.e., syntactic change may feed morphological
change (cf. Anderson 1980, Cole et al. 1980, Disterheft 1987, and Fischer
2010). In this talk, I will re-examine the relationship between morphological and syntactic change, adopting the view that language change is to be
modeled in terms of grammar change (Hale 2007), i.e., a set of discrete differences between the target grammar and the grammar eventually acquired
by the learner. It will be argued that from this perspective, the assumption
of a strong causal link between a morphological property M and a syntactic property S necessarily leads to a conflict: At the point when the learner
fails to acquire M, M will still be part of the target grammar. As a result,
syntactic patterns linked to M will continue to be part of the input the
learner receives, leading to a situation where morphological and syntactic
294
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 295 — #307
AG 10, Raum G 308
cues for a given parameter contradict each other. It will be demonstrated
that this conflict becomes even more acute when the focus is shifted from
the loss of M to the rise of M. I will then discuss a set of issues these considerations raise for a theory of language change and the interface between
syntax and morphology. In particular, I will address whether it is possible to maintain the view that there is a (strong) causal relation between
morphology and syntax. In the model presented above, this seems to boil
down to the question of how learners deal with conflicting syntactic and
morphological cues in the input; possible scenarios include:
• Some part of the evidence available to the learner is ignored (possibly
relating to the robustness/frequency with which individual cues are
attested in the input).
• Conflicting morphological and syntactic evidence in the input leads
to grammar competition and eventually the loss of one of the competing options (cf. Haeberli 2004).
• The syntactic effects of M are obscured by other syntactic processes
in the target grammar and are therefore less visible to the learner.
As time permits, we may touch on a number of additional issues which
need to be addressed if we are to gain a more complete understanding of
the interaction of syntax and morphology in processes of language change:
• Do cases in which a syntactic change occurs in the face of robust
morphological counter-evidence necessitate a view of morphology
as “formal baggage” (Anderson 1980) that is dragged along (after a
reanalysis of its function) or disposed of after syntactic change has
rendered it superfluous or non-interpretable?
• How do we account for cases where morphological changes seem to
depend on previous syntactic changes? How do we account for the
time lag between morphological and syntactic change?
295
AG10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 296 — #308
AG 10: Morphological effects on word order from a typological …
From rags to riches: The RAH from a creole perspective
Tonjes Veenstra
ZAS Berlin
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:30, Raum: G 308
AG10
In Koeneman & Zeijlstra (2014) the Rich Agreement Hypothesis has been
reformulated as follows:
A language exhibits V-to-I movement if and only if the regular paradigm
manifests featural distinctions that are at least as rich as those featural distinctions manifested in the smallest pronoun inventories universally possible.
In this paper we review different creole systems that pose serious problems for any version of this research programme.
Fren-based creoles: Most of these creoles display a long/short alternation in their verbal paradigm. In those creoles where the alternation is
present, it correlates with syntactic properties. Interestingly, the syntactic
correlate differs in (almost) each French creole. In Louisiana Creole it marks
a Tense distinction (present/past), leading to concomitant movement of the
verb into the INFL-domain, when the verb is in the short form. Arguments
come from adverb placement and negation, as well as the incompatibility
of short forms and preverbal TMA markers. Thus, such creoles exhibit Vto-I movement without the minimal necessary featural distinctions in their
paradigm.
Spanish-based creoles: The pronominal paradigm of Papiamentu is shown
to be an Agreement system, like colloquial French. Nevertheless, Papiamentu does not exhibit V-to-I movement. In expletive constructions, however, it does show V-movement to a high position in the lexical domain.
English-based creoles: Like Papiamentu, the pronominal paradigm of Saramaccan is an Agreement system. Evidence comes from perception verb
constructions. Nevertheless, there is no sign of any V-movement in this
language.
296
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 297 — #309
AG 10, Raum G 308
Thus, there is much more variation between creole languages with respect to V- movement:
(i) movement to a high position in the I-domain (presumably T, e.g.
Louisiana Creole);
(ii) movement to a low position in the I-domain (presumably Asp, e.g.
Berbice Dutch);
(iii) movement to a high position in the lexical domain (presumably Pred,
e.g. Papiamentu, Kriyol);
(iv) no verb movement at all (e.g. Saramaccan).
AG10
These movement operations are shown to be independent of any morphological properties, in terms of featural distinctions in their respective
paradigm, these languages have.
The contribution of contact linguistics to the Rich
Agreement debate
Peter Slomanson
U. Tampere
Peter.Slomanson@staff.uta.fi
Mittwoch, 24. 02. 2016, 17:30–18:30, Raum: G 308
I will argue, based on cross-linguistic data from languages with different
types of collective acquisition history, that diachronic and external evidence merit consideration in research on the association of inflectional
paradigms and syntactic movement.
According to the RAH, a rich agreement paradigm associated with finite verbs can be predictably correlated with the extent of movement of
such verbs out of VP to a position in the inflectional domain. Icelandic is
an example of a language to which this generalisation applies, the diagnostic being the surface position of negation in both main and subordinate
clauses. Icelandic contrasts with Danish, in which the finite verb follows
297
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 298 — #310
AG 10: Morphological effects on word order from a typological …
AG10
negation in complement clauses. However there are also types of language
that diverge from this pattern, featuring robust V to I in the absence of
agreement.
We can also find V to I of a weaker type in languages lacking agreement
entirely, but not lacking finiteness. In those languages, the verb raises to a
lower part of the inflectional domain, and no higher. We see this in the
fact that temporal morphology, typically marking aspect, is spelled out
post-verbally, as we would expect, but also pre-verbally, typically marking tense. We find this pattern in a small number of contact languages,
including Berbice Dutch, Cape Verde Portuguese, Sri Lanka Malay and Sri
Lanka Portuguese.
One potential way to make sense of the apparent contradiction between
an empirically promising approach such as the RAH and those languages
that appear to constitute counter-examples is to recognize that the diachrony of a grammar may have residual effects on its synchronic organization, that these effects are not necessarily short-lived, and that collective
acquisition history may be a factor influencing whether or not a particular
syntactic configuration is retained after the morphology that was once its
trigger is lost. It may be the case that the syntactic residue of an earlier
grammar is only necessarily lost when there has been a break in transmission. This would help us to understand the status of languages of intermediate type (such as Afrikaans, with residual robust V to I, but no agreement
morphology, and the contact languages referred to above, with tense morphology and weak V to I). Finally, it is important to consider the majority
of creoles, which demonstrate various types of grammatical complexity,
but no agreement and no V to I, irrespective of their parent languages.
Creoles and other types of contact language are useful for testing theories associating inflectional morphology with syntactic movement, since
the canonical claim in contact linguistics has been that many of these languages, due to a break or breaks in transmission, developed from largely
deflected grammars (pidgins or pidginesque interlanguages). When functional contrasts are later restored via grammaticalisation, they are grammatically new, rather than vestigial. When we examine a cross-section of
these languages to see which contrasts are recreated over time, we find
roughly the following implicational order: aspect, mood, tense, (non-)finite-
298
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 299 — #311
AG 10, Raum G 308
ness, and agreement, with the last two contrasts highly infrequent and exceedingly infrequent, respectively. This order seems to reflect a potential
hierarchy of communicative salience (from high to low), inversely related
to a hierarchy of feature strength (low to high), as reflected in the extent
of movement. Syntactic movement supporting rich agreement may persist
long after its loss, barring a break in transmission.
V2 and verbal morphology in Övdalian
Ásgrímur Angantýsson
U. Iceland
AG10
[email protected]
Donnerstag, 25. 02. 2016, 9:00–10:00, Raum: G 308
The purposes of this paper are twofold. First, it aims at placing Övdalian
(see Bentzen et al. 2015 and references there) among the Scandinavian languages with regard to verbal morphology and embedded V2. Secondly, it
attempts to formalize and test hypotheses predicting that languages/dialects
that have the relevant morphological differences also show certain syntactic differences.
The results presented here are from two written questionnaires administered to 52 speakers of Övdalian (12 adolescents and 33 adults) during
fieldwork in Älvdalen in central Sweden. The first questionnaire included
minimal pairs contrasting Vfin/Adv order and Adv/Vfin order in various
types of subject-initial embedded clauses with sentence adverbs like int/
it ‘not’, older/aldri ‘never’ and oltiett ‘always’. The second questionnaire
included minimal pairs/triplets of embedded topicalization and transitive
expletive constructions (TECs). A subset of the speakers (34 in total) also
performed verbal paradigm fill-in tasks. The method can be described as
‘supervised questionnaire completion’ (see discussions in Cornips & Poletto 2005).
It turns out that the older speakers of Övdalian allow the Vfin-Adv order more freely than the younger speakers, and the conditions for subjectinitial V2 depend to a certain extent on the type of embedded clause as well
299
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 300 — #312
AG 10: Morphological effects on word order from a typological …
AG10
as the type of finite verb and adverb. The results from a verbal paradigm fillin task reveal substantial variation in the use of verbal affixes and, interestingly, a tendency, especially by the younger speakers, to simplify the verbal morphology. TECs receive very low acceptance scores. The relevance of
this for different versions of the Rich Agreement Hypothesis is discussed
in the paper (Holmberg & Platzack 1995, Vikner 1995, 1997, Jonas 1996,
Rohrbacher 1999, Bobaljik 2002, Thráinsson 1996, Bobaljik & Thráinsson
1998, Thráinsson 2010, Koeneman & Zeijlstra 2012) and a possible scenario
for the historical development of embedded clause word order in Övdalian
is presented. In short, it is maintained that while the V2-order in embedded clauses is on its way out in Övdalian, presumably because of diminishing morphological support and partially ambiguous syntactic evidence,
the language has not yet reached the Mainland Scandinavian state in this
respect.
On the role of verbal mood in licensing dependent V2
clauses
Hans-Martin Gärtner
RIL-HAS Budapest
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: G 308
Holmberg & Platzack (1995) and Koeneman (2010) correlate the distinction between broad dependent V2 (bDV2) (“general embedded V2”) in Icelandic and narrow dependent V2 (nDV2) (“limited embedded V2”) in Mainland Scandinavian with the respective presence vs. absence of “rich agreement.” This view is challenged by the coexistence of bDV2 and nDV2 varieties both in Icelandic (Angantýsson 2011; Gärtner 2003; Jónsson 1996;
Thráinsson 2011) and Mainland Scandinavian (Bentzen 2014). Concentrating on the Icelandic part of the puzzle, I will explore the role of verbal
mood in licensing dependent V2 clauses. The idea here is that the force/
(non-)commitment signaling properties of ±V2 and indicative vs. subjunctive can interact in different ways, resulting in bDV2 and nDV2 varieties.
300
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 301 — #313
AG 10, Raum G 308
What defending this approach requires is a better understanding of mood
variation in contemporary Icelandic (Óladóttir 2011; Þórðardóttir 2012) as
well as of German bDV2 structures licensed by the subjunctive. The resulting account will then have to be squared with purely topic-based ones as
pursued by Hrafnbjargarson and Wiklund (2009) and Heycock et al. (2010).
Re-challenging the RAH: Problematisation of structural
and social aspects in 19th-century Icelandic
Heimir van der Feest Viðarsson
University of Iceland
AG10
[email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: G 308
For over three decades now, linguists have tried to argue in favour of the
hypothesis that the amount of verbal inflection predicts to a large extent
the possible structural position of the finite verb in the clause (e.g. Roberts
1985, Pollock 1989, Vikner 1995, Bobaljik & Thráinsson 1998, Rohrbacher
1999, Koeneman & Zeijlstra 2014). While this so-called Rich Agreement
Hypothesis has been investigated in a variety of languages, including Modern Icelandic (see esp. Angantýsson 2011), earlier stages in the history of
Icelandic have received much less attention. This is all the more unfortunate in that previous research on Icelandic has revealed historical variation which, at least on the face of it, speaks quite strongly against the RAH
(e.g. Þorbjörg Hróarsdóttir 1998, Heycock & Wallenberg 2013). A systematic study of 19th-century Icelandic is, therefore, an important source of
potential counterexamples to the RAH meriting further study.
Although the various formulations of the RAH that have been proposed
in the literature have radically different predictions concerning the empirical data, they have at least one thing in common. A richly inflected
language like Icelandic should only allow the word order in (1), where the
finite verb undergoes V-to-I movement, crossing the negation, according
to traditional assumptions about clause structure, whereas the word order
in (2), where the verb occurs below negation, ought to be ungrammatical:
301
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 302 — #314
AG 10: Morphological effects on word order from a typological …
AG10
(1)
[ Á borðinu er bréf [ sem Jón vildi ekki brenna ]]
on the-table is letter which Jón wanted not burn
(Vfin-Neg)
(2)
[ … [ sem Jón ekki vildi brenna ]]
(Neg-Vfin)
which John not wanted burn
‘On the table is a letter which John did not want to burn.’
As it turns out, the word order in (2) is not ungrammatical for all speakers of
Modern Icelandic, albeit very restricted (see Ásgrímur Angantýsson 2011).
Various pieces of evidence have been put forward to demonstrate that the
finite verb in the configuration in (2) has still undergone V-to-I movement
(e.g. Bobaljik & Thráinsson 1998, Ásgrímur Angantýsson 2007, Höskuldur
Thráinsson 2010, Koeneman & Zeijlstra 2014). This evidence rests on various structural restrictions as well as its relatively low frequency in Modern
Icelandic.
In this paper I present the results of a systematic study of the placement of the finite verb in embedded clauses in a range of different 19thcentury sources, modelled to a large extent after the study of Heycock et al.
(2012) on Faroese and Danish. Previous research on pre-Modern Icelandic
has shown that the supposedly ungrammatical word order in (2) had become quite frequent in the period 1600-1850 but that it was confined to
the very formal language of highly educated speakers imitating the word
order of Danish (cf. Heycock & Wallenberg 2013). However, the problem is
that scholars have either focused on a very limited set of sources consisting mostly of rather formal, narrative texts as opposed to more informal
language such as private letters, or the results have suggested a different
picture in which the Neg-Vfin order in (2) may not have been confined to
educated intellectuals after all (Hróarsdóttir 1998).
To address these issues, I contrast the distribution of (1) Vfin-Neg and
(2) Neg-Vfin in two types of texts on a much larger scale than attempted
before: (i) a corpus of 19th-century newspapers and periodicals, and (ii)
19th-century private letters of male and female speakers of different social
backgrounds, each approx. 1.5 million words. Subsequently, the language
302
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 303 — #315
AG 10, Raum G 308
use in these different genres is compared both against known restrictions
in Modern Icelandic (cf. Angantýsson 2011) and the distribution across different clause types in Faroese and Danish which either do or do not easily
allow V-to-C movement (cf. Heycock et al. 2012). The conclusion will be
that the distribution among the two variants patterns to some extent with
a V-to-C/V-in-situ language like Danish, but always alongside a grammar
which also generates structures that can be analysed as involving V-to-I on
basic assumptions. This is a conclusion similar to Heycock & Wallenberg
(2013), except that the use of Neg-Vfin as in (2) is shown to be much less
restricted, both structurally and in terms of the sorts of speakers capable of
producing the supposedly ungrammatical Neg-Vfin order. Importantly, the
restrictions frequently observed in Modern Icelandic do not seem to apply
either. As a result, a different analysis will be proposed which is in better
agreement with sociolinguistic aspects of variation, where certain groups
of speakers can be regarded as the leaders of language change (cf. Tagliamonte 2012), and where verb-second is a constraint that can be violated
along the lines of Zwart (2005).
Revisiting the RAH in light of diachronic data from the
history of Danish
John Sundquist1 & Caroline Heycock2
1
U. Purdue, 2 U. Edinburgh
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 12:00–13:00, Raum: G 308
In this paper, we analyze new diachronic data on changes in word order in
the history of Danish and provide evidence against a bidirectional implication between rich verbal agreement morphology and local movement of
the finite verb (here, V-to-I movement). In particular, we use Koeneman &
Zeijlstra (2014) as a point of departure, discussing their attempts to rehabilitate a “strong”—bidirectional—version of the Rich Agreement Hypothesis
(RAH) (Rohrbacher 1994, 1999, Vikner 1997, Bobabljik & Thráinsson 1998).
Using data from Middle and Early Modern Danish (1350 - 1750), we address the diachronic implication of the strong version of the RAH, namely,
303
AG10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 304 — #316
AG 10: Morphological effects on word order from a typological …
AG10
that the loss of rich verbal agreement must immediately trigger the loss
of V-to-I movement. Data from a corpus of early Danish texts corroborate
findings in Sundquist (2002, 2003) and Heycock & Sundquist (2015) that the
diachronic data support a weaker version of the RAH in which there is only
a one-way entailment between rich agreement and verb movement. First,
we present new results from an empirical analysis of an expanded corpus
of early Danish texts. Unlike Sundquist (2002, 2003) in which the corpus
was limited to texts written between 1500 and 1700, we include texts from
as early as 1350 and as late as 1750. The earlier texts allow for analysis
of the time period closer to the changes in verbal morphology, while the
inclusion of texts after 1700 provides a more robust data set closer to the
outer limit of Verb-Neg word order in embedded clauses in the history of
Danish. Next, we examine more closely the changes to the Middle Danish and Early Modern Danish verbal paradigm. As pointed out in Heycock
et al (2012), citing evidence from Weyhe (1996) on Faroese, it is necessary
to examine a wide variety of conjugations of both weak and strong verbs
to determine exactly how robust the agreement paradigm is, rather than
rely exclusively on the regular paradigm for weak verbs (cf. Rohrbacher
1999). To give a clearer picture of the status of agreement distinctions in
the spoken rather than written language of the period under investigation,
we review descriptions of the early Danish verbal system that contemporaries like Peder Syv (1685) and Henrik Thomsen Gerner (1678) provide.
Evidence from this expanded corpus and more detailed analysis of inflectional paradigms indicates that there is a significant time gap between verbal deflection in the 14th century and a loss of verb movement in the 18th
century. In the discussion of these results, we address each of the types of
reanalysis that Koeneman & Zeijlstra (2014) suggest to explain this gap.
The first type, the reanalysis of e.g. subject clitics as agreement markers
(as in French), is not a plausible explanation for Danish, where there is no
evidence for subject clitics. The second type of reanalysis that could allow
for the persistence of verb movement is that V-to-I movement was reanalyzed as V-to-C movement, that is, embedded V2. However, the examples
here and in Sundquist (2002, 2003) were carefully selected to control for
embedded V2, and when any possible ambiguous examples are excluded,
the rate of verb movement remains as high as 15% in the 18th century.
304
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 305 — #317
AG 10, Raum G 308
Finally, Koeneman & Zeijlstra (2014) suggest that V—Adverb orders may
be reanalyzed in terms of an innovative lower (vP-internal) positioning of
sentential adverbs. However, as discussed by Koeneman & Zeijlstra in this
context, sentential negation provides clear evidence for the edge of the vP.
Limiting our dataset here to examples with negation, we have clear empirical evidence for the persistence of V-to-I movement, several centuries
after the impoverishment of the verb agreement paradigm. The implications that such a significant time gap pose for the weak and strong version
of the RAH are discussed in the conclusion of the paper.
Rich agreement in the Shetland dialect of Scots
AG10
Elyse Jamieson
U. Edinburgh
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: G 308
The Shetland dialect of Scots is unusual in that it is a variety of English
that, until very recently, retained main verb movement from V-to-I. This
phenomenon was explored by Jonas (2002), who concluded that Shetland
dialect was a counterexample to the rich agreement hypothesis (RAH). She
argued that Shetland dialect did not have rich morphological agreement,
and that it did not exhibit any of the other characteristics we might expect
from a variety with rich agreement. Instead, she posited that residual V-toI was available in the dialect because of the continued presence of V-to-C
movement. Appealing to the Head Movement Constraint, she argued that
the cyclic movement required for the verb to reach the higher CP location
provided evidence of an available “stopover” projection in I. However, in
this paper, I will argue that this is not the correct analysis for Shetland dialect’s one-time presence of V-to-I movement. Instead, I will contend that
Shetland dialect was, at least until recently, a variety with rich agreement,
with an agreement system requiring the [speaker], [participant] and [plural] phi- features specified by Koeneman & Zeijlstra (2014).
I will argue that the presence of a T/V distinction between singular second person pronoun du, which inflects –s, and the null inflected plural
305
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 306 — #318
AG 10: Morphological effects on word order from a typological …
AG10
second person pronoun you, invokes both [speaker] and [plural] features.
While this alone is not enough to constitute rich agreement, the Northern
Subject Rule (where third person plural NP subjects take –s verbal inflection e.g. the birds sings) also invokes the [participant] feature. The combination of all three of these features gave Shetland dialect the richness of
morphological agreement it needed to have evidence for V-to-I movement.
This analysis is further supported by evidence of transitive expletive constructions in the dialect, another type of construction frequently cited as a
reflex of the RAH (e.g. Bobalijk & Thráinsson 1998).
Despite the above, however, V-to-I has gone from being “residual” (Jonas
2002) to near “unacceptable” (Jamieson 2014) in the dialect in recent years.
Although I argue that there is some levelling of the morphological agreement paradigm taking place, it nevertheless remains productive. This raises
wider questions about the relationship between the morphological “trigger” and the phenomena it supposedly permits. I will thus argue that in
Shetland dialect, grammar competition (e.g. Kroch 1989) due to language
contact has also played an important role in the presence and subsequent
reduction in acceptability of V- to-I movement in the dialect.
The Rich Agreement Hypothesis: Diachronic (lack of)
evidence from English
Eric Haeberli1 & Tabea Ihsane1
U. Geneva
1
[email protected], [email protected]
Freitag, 26. 02. 2016, 12:00–13:00, Raum: G 308
This paper examines the validity of the Rich Agreement Hypothesis (RAH)
on the basis of evidence from the loss of verb movement in the history
of English. The RAH postulates that there is a relation between the Vmovement properties of a given language and its verbal agreement morphology. The RAH comes in two flavours, a strong one and a weak one.
Under the strong version of the RAH (e.g. Koeneman & Zeijlstra 2014),
V-movement is possible only if the verbal agreement morphology is sufficiently rich. If that is not the case, it is not possible for the verb to move out
306
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 307 — #319
AG 10, Raum G 308
of the VP. As for the weak version of the RAH (e.g. Bobaljik & Thráinsson
1998), it only claims that V-movement must occur if verbal agreement is
sufficiently rich but it does not have anything to say about cases where
agreement does not meet the requirements for being rich. Although the
two approaches do not make exactly the same diachronic predictions, they
have two important consequences in common:
(1)
V-movement is considered in these approaches as an all-or-nothing
option – either a language has it, or it does not have it. Hence, if a
language loses V-movement, it would be expected to lose it at once.
No intermediate steps, i.e. movement out of the VP but to a lower
landing site than previously, would be expected.
(2)
A language should not start losing V-movement before rich agreement morphology is lost.
Both of these consequences can be evaluated on the basis of the history
of English. With respect to (1), a close analysis of the development of the
two main diagnostic elements for V-movement, i.e. adverbs and negation,
reveals that they do not pattern together. Whereas the adverb data suggest that there is a rapid decline in V-movement past adverbs starting in
the middle of the 15th century and being completed in the middle of the
16th century, the negation data show a very slow decline starting from
the beginning of the 16th century, with V-movement past negation remaining strong up to 1650. The final decline in V-movement then occurs
from 1650 onwards. We conclude from this evidence, that V-movement is
lost step-wise in the sense that initially V-movement to a high inflectional
head is lost and replaced by movement to a lower inflectional head before
V-movement is lost completely (cf. also Han 2000, Han & Kroch 2000).
The question that arises then is what the role of agreement could have
been in this development, if any. We will show that, in terms of two prominent approaches postulating the RAH, i.e. Bobaljik & Thráinsson (1998)
and Koeneman & Zeijlstra (2014), the crucial morphological property that
turned English from a rich agreement language into one with poor agreement is the loss of the 2nd person singular -(e)st morpheme. As for the timing of the loss of 2nd person singular morphology, we conclude on the ba-
307
AG10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 308 — #320
AG 10: Morphological effects on word order from a typological …
AG10
sis of corpus evidence and observations made in the literature (Hope 1993,
Lass 1999, Walker 2007) that it must be situated somewhere in the 17th century. Hence, both the first development (loss of V-movement past adverbs
starting in the middle of the 15th century) and the second one (replacement
of V-movement past negation by do-support starting in the early 16th century) begin at a point when, according to the two versions of the RAH
considered here, verbal morphology should still require V-movement. The
decline of V-movement in English is therefore not easily compatible with
the RAH. We will argue that a plausible account of the developments in
English can be formulated that is entirely independent of agreement morphology and instead accounts for the overall development of V-movement
in the history of English in terms of a combination of factors: changes in the
verbal morphosyntax (loss of subjunctive, rise of periphrastic forms), an acquisitional bias towards simpler structures, the decline of the subject-verb
inversion grammar found in early English, and effects of dialect contact.
We will also show how the continued occurrence of movement with have
and be (another potential problem for the RAH) can be integrated into our
framework.
‘If It’s Tuesday, This Must Be Belgium’: Some alleged
syntax-morphology correlations re-examined
Þórhallur Eyþórsson
University of Iceland
[email protected]
Freitag, 26. 02. 2016, 13:00–14:00, Raum: G 308
It is commonly assumed that languages such as Old Norse, Modern Icelandic and (a variety o) Faroese have “independent V-to-T” movement
(see Angantýsson 2011 for a comprehensive discussion). There is an ongoing debate about whether this movement to T is related to rich agreement
or not (the Rich Agreement Hypothesis, existing both in a strong and a
weak form; e.g., Koeneman & Zeijlstra 2014). On an alternative approach,
the verb moves to a position within the CP system, not within TP (e.g.,
308
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 309 — #321
AG 10, Raum G 308
Wiklund et al. 2009). Regardless of the actual position to which the verb
is said to move, what is usually lost sight of in these analyses is that, just
as in main clauses, the verb also occurs in second position in embedded
clauses in (standard) Icelandic, exhibiting strict adjacency with the preceding phrase. In other words, there is a generalized V2 in Icelandic (whereby
V2 is not necessarily to be equated with V-to-C). A disturbing factor which
obscures the picture relates to the fact that a subset of V3/verb-late orders
seem to involve lower TP-internal movement, occurring for example in
Old Germanic (Eythórsson 1995), Northern Norwegian varieties (Bentzen
2010) and varieties of Icelandic (Angantýsson 2010). Whatever the motivation for this lower TP-internal movement, it is clearly unconnected to
either rich agreement or V2.
309
AG10
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 310 — #322
Sprachwissenschaft
bei Winter
Universitätsverlag
winter
Heidelberg
luth, janine
poulain, elfie
Semantische
Kämpfe im Recht
Einführung in
die Literaturpragmatik
Eine rechtslinguistische
Analyse zu Konflikten zwischen
dem EGMR und nationalen
Gerichten
mit einer Beispielanalyse von
Kafkas Roman Der Prozess
2015. 304 Seiten. (Schriften des
Europäischen Zentrums für Sprachwissenschaften (EZS), Band 1)
Geb. € 40,–
isbn 978-3-8253-6325-3
2015. 110 Seiten. (Sprachwissenschaftliche Studienbücher)
Kart. € 19,–
isbn 978-3-8253-6484-7
kloudová, v Ě ra
2015. 72 Seiten. (Literaturhinweise
zur Linguistik, Band 2)
E-Book, € 14,–
isbn 978-3-8253-7534-8
butterworth, judith
Redewiedergabeverfahren in der Interaktion
Individuelle Variation bei der
Verwendung einer kommunikativen Ressource
innerwinkler, sandra
Neologismen
2015. 120 Seiten. (Literaturhinweise
zur Linguistik, Band 1)
E-Book, € 14,–
isbn 978-3-8253-7511-9
2015. 363 Seiten, 30 Abbildungen,
10 Tabellen. (Sprache – Literatur
und Geschichte, Band 47)
Geb. € 68,–
isbn 978-3-8253-6525-7
brenning, jana
Sprache und Migration
Syntaktische
Ko-Konstruktionen im
gesprochenen Deutsch
2015. x, 101 Seiten. (Kurze Einführungen in die germanistische
Linguistik, KEGLI, Band 18)
Kart. € 13,–
isbn 978-3-8253-6454-0
2015. x, 256 Seiten,
2 Abbildungen. (OraLingua,
Band 11)
Geb. € 40,–
isbn 978-3-8253-6502-8
peterson, john
D-69051 Heidelberg · Postfach 10 61 40 · Tel. (49) 62 21/77 02 60 · Fax (49) 62 21/77 02 69
Internet http://www.winter-verlag.de · E-mail: [email protected]
Synonymie und
Antonymie
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 311 — #323
Arbeitsgruppe 11
Indefinites between theory and language change
Chiara Gianollo1 , Klaus von Heusinger1 & Svetlana Petrova2
1
Universität zu Köln, 2 Bergische Universität Wuppertal
[email protected], [email protected],
[email protected]
AG11
Raum: G 201
Workshop description
Language users employ indefinites, pronouns (someone, anything, whatever) and different types of noun phrases (a book, a certain student, some
time, any teacher) to encode (non-)referentiality, but also other crucial
properties, such as degree of identifiability, speaker-hearer knowledge status, discourse saliency. Recent typological and theoretical studies have uncovered a wealth of variation in this domain, on various grammatical levels (morpho-syntax, semantics, pragmatics). Also research on the history
of indefinite articles and some classes of indefinite pronouns in individual
languages has advanced substantially. The emerging picture needs now to
be complemented by a comparative evaluation of the observed diachronic
patterns. We face scenarios that challenge well-known models of development and therefore need a broader cross-linguistic perspective on evolutionary tendencies, also encompassing non-Indo-European languages.
A more fine-grained study of the diachronic clines involving indefinites
may shed light on some of their intriguing synchronic properties (morphosyntactic complexity, multifunctionality, context dependence), and on the
way systems of indefinites are structured (complementarity, blocking). The
investigation further promises to disclose more general conclusions on the
311
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 312 — #324
AG 11: Indefinites between theory and language change
systematic nature of change affecting functional elements of the lexicon.
We therefore invite contributions from linguists of various persuasions,
reconciling in-depth theoretical analysis with comparative and diachronic
evidence, and seeking answers to the following questions:
(a) Which properties account for the systematicity in diachronic processes involving indefinites? To what extent are different sub-classes
(specific, epistemic, free-choice, polarity-sensitive indefinites) affected by cyclical developments? How can theories account for the fact
that such developments may be multidirectional?
AG11
(b) In the system of indefinite markers, the indefinite article stands out
in many respects (source of grammaticalization, interaction with the
definite article and with number marking). How can we reconcile
the cyclical model traditionally proposed for its evolution with the
evidence provided by many historical varieties, in which various
functions (referential, generic, predicative, etc.) already co-exist at
an early grammaticalization stage?
(c) Which ingredients are needed to provide satisfactory theoretical models of the crosslinguistic micro-variability attested by indefinites, and
of the system synchronically and diachronically organizing the observed functions?
Indefinites as fossils
Maria Aloni
Universiteit van Amsterdam, Institute for Logic, Language and
Computation
[email protected]
Mittwoch, 24. 02. 2016, 14:00–15:00, Raum: G 201
In the first part of the talk I will summarise the results of three diachronic
corpus studies establishing the patters of development of three marked indefinites in three different languages: German irgend-series; Spanish cualquiera and Dutch wie dan ook (see http://maloni.humanities.uva.nl/Inde-
312
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 313 — #325
AG 11, Raum G 201
finites/corpus.html). In the second part I will discuss a number of theoretical repercussions of these studies on various issues at the semanticpragmatic interface, including the debate about the status of the modal
inferences (ignorance and free choice) triggered by these indefinites.
Anti-specificity and the role of number: The case of
Spanish algún/algunos
Urtzi Etxeberria1 & Anastasia Giannakidou2
CNRS-IKER, 2 University of Chicago
1
[email protected], [email protected]
AG11
Mittwoch, 24. 02. 2016, 15:00–16:00, Raum: G 201
The Spanish singular determiner algún is claimed to be an anti-specific indefinite, as in (1) (Giannakidou and Quer 2013, Alonso Ovalle and MenendezBenito 2010), while plural algunos is claimed to show ‘context dependence’
(cf. (3)). This results in an ambiguity analysis, which is undesirable because
it does not capture the role of the plural. In this paper we argue that algún and algunos are the singular and the plural version of each other and
propose a unifying analysis by showing that the difference between singular algún and plural algunos is illusory and that in both usages referential
vagueness is satisfied. We ague that the context dependency of algunos
arises only in anaphoric contexts like (3) where a discourse referent has
previously been explicitly introduced, and this discourse referent sets up
an antecedent (note that in (5) with no antecedent algunos is indistinguishable from unos). It is conceivable then that the plural in this case functions
as an anaphoric pronoun just like in Mary brought the yellow T-shirts, and
Ariadne the red ones (Kester 1995). We will propose that a plural anaphor
is triggered and is reflected in the plural number, thus, what appears to
be a plural is really an anaphoric pronoun (6). So, the plural introduces
the pronoun proPL which is also an anaphor, but this happens only in the
context of an overt antecedent. Our idea is close in spirit to Marti’s C variable, but we do not assume that alg- introduces it (cf. (4)); rather, if an antecedent is available the pronoun will be triggered, as expected generally
313
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 314 — #326
AG 11: Indefinites between theory and language change
in ellipsis contexts. As a consequence, partitivity appears to be epiphenomenal, a consequence of the presence of the anaphoric plural. In sum, we are
proposing a fully compositional analysis that retains the anti-specificity
of the alg- indefinite (cf. (2)) and attributes the illusory specificity to the
plural ellipsis).
Examples:
AG11
(1)
Ha llamado algún estudiante. #Era Pedro.
have called some student
was Pedro
‘Some student called. #It was Pedro.’
(2)
Referential vagueness as anti-specificity
a. A sentence containing a referentially vague indefinite α will
have a truth value iff:
∃ w1 , w2 ∈ W: 〚α〛w1 ̸= 〚α〛w2 ; where α is the referentially vague
indefinite.
b.
The worlds w1 , w2 are epistemic alternatives of the speaker.
(3)
{Teachers A and B are on an excursion with [a group of children, of
whom they are in charge]K . Teacher A comes to teacher B running:}
a. Teacher A: ¿Te has enterado? [Algunos niños]K, #J se han perdido en el bosque.
b. Teacher A: ¿Te has enterado? [Unos niños]K, J se han perdido
en el bosque.
‘Have you heard? Unos/algunos children got lost in the forest.’
(4)
〚alg-〛= λR<et,<ett» .λP<et> .λQ<et> .R(P∩C)(Q); Implicature: R(P∩C)({x:
Q(x) = 0})
(5)
Llegaron algunos/unos chicos a la oficina.
Arrived boys
to the office
‘Some boys arrived to the office’
(6)
〚algunos (niños)〛= algun + proPL [+anaphoric]
314
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 315 — #327
AG 11, Raum G 201
On (negative) indefinites in Old Italian
Irene Franco1 , Olga Kellert2 , Guido Mensching2 & Cecilia Poletto1
Goethe Universität Frankfurt am Main, 2 Universität Göttingen
1
[email protected], [email protected],
[email protected], [email protected]
Mittwoch, 24. 02. 2016, 16:30–17:30, Raum: G 201
In this talk, we concentrate on the morphosyntax of Old Italian indefinites
in negative contexts. The main questions we address are the following:
(i) What is the difference between indefinites as n-words and as NPIs?
(ii) How do these indefinites behave with respect to Negative Concord
(NC)?
(iii) How can we account for their distribution within the clause?
We propose that NC depends on an Agree mechanism (see Zeijlstra 2004)
that is sensitive to the internal structure of n-words (Martins 2000, Déprez
2011, a.o.), while NPI licensing is a different type of dependency (Giannakidou 2002). Old Italian (OI) is the perfect laboratory to tackle questions
(i)-(iii), since it has apparently optional NC, and there is a change around
the turn of the XIV century towards non-strict NC, like in Modern Italian.
The apparent NC optionality is shown in (1) (see Garzonio & Poletto 2012).
(1)
a.
b.
E neuno di voi si
spaventi…
and no.one of you efl= fear-bj
‘And may none of you get scared…’ (OVI, VeV 69)
E neuno non andasse
poscia
in paradiso…
and no.one not went-3SG-SBJV afterwirds in heaven
‘(So) no one would go to heaven afterwards.’ (ibid. 78)
NC is apparently optional with preverbal and postverbal n-words. However, we will show that the optionality is systematically restricted by the
internal and external syntax of n-words. NC is always attested with adverbial n-words, whereas argumental n-words have a mixed behavior. Specifically, argumental n-words generally obey NC, unless they are merged a)
315
AG11
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 316 — #328
AG 11: Indefinites between theory and language change
in a copular/existential construction; b) inside a PP, where P=from/to; or
c) inside a manner/reason adverbial PP. We argue that n-words can lexicalize two different structures, which in turn depend on their semantics.
N-words obeying NC are QPs, on a par with NPIs like alcuno (=‘any’), see
(2a). Negative expressions that do not obey NC are instead of an adjectival
type (Giusti & Leko 2001, Cardinaletti & Giusti 2005), see (2b).
(2)
AG11
a.
b.
[QP ne-/alc- [nP uno]]
[PP da[DP ∅ /-l [AdjP ni- [nP ente]]]]
In our talk, we will present a further test, which reveals an asymmetry in
the distribution of NPIs like alcuno, and n-words like neuno, in (2a). NPIs
cannot precede n-words (whether or not there is a non-veridical licensor
in the clause). We thus propose that n-words can locally license an NPI. We
conclude that NPI licensing is established first at a local level (cf. Haegeman & Lohndal 2010) and, if no licensor is met, it creates a non-local dependency.
Another route towards epistemic indefinites: A case for
VERUM?
Remus Gergel
Karl-Franzens-Universität Graz
[email protected]
Mittwoch, 24. 02. 2016, 17:30–18:30, Raum: G 201
We analyze comparative and historical aspects of epistemic indefinites in
Romance as currently exemplified by Romanian vreun, ‘any’ (Farkas 2002,
Fălăuş 2014) and relate them to focus. Fălăuş discovered that vreun is licensed by obligatorily non-factive epistemic operators. The item seems to
be an NPI w.r.t. most contexts. But it is barred in (most) directly negative
contexts, where genetically unrelated, overtly negative indefinites must be
used, a fact that has been attributed to some version of blocking.
We address two sets of questions. (A) What is the cycle of the vreun family: can the development be seen in line with other indefinites (Haspelmath
316
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 317 — #329
AG 11, Raum G 201
1997, Eckardt 2012)? Are epistemic effects more recent or are they an older
vestige of some Latin indefinite (cf. Gianollo 2013 on Lat. aliquis). (B) What
is the genesis of vreun in its use as an NPI and as a designated epistemic
indefinite? What are the structural building blocks?
While items related to vreun are not in the common stock of indefinites
known from Classical Latin, such an item must have been in use in early
Romance. A similar item is attested in Aromanian (e.g. texts in Saramandu
2008) and Old Italian had veruno (Ramat 1997). The trajectory of veruno
must have been that of an NPI. Old Italian allowed the item properly under negation, including covert uses, and early Romanian varieties showed
likely vestiges of such a behavior too. How did the potential for NPI and
epistemic flavor arise? Looking into the morphological composition of the
ancestor, i.e. < vere+unus (=‘truly’+‘one’), negation must be added to the
initially developing negative item (as in Old Italian). A plausible scopal relationship was truly > not > (even) one. We explore the idea that the high
adverb was the key towards creating the right types of alternatives (despite
possible appearances to the contrary). It introduced em as a function
acting on the epistemic alternatives available to an individual in a particular world (Han & Romero 2004, Romero 2006). Consequently, not only the
more general sensitivity to alternatives is expected (available for NPIs), but
also the fact that sensitivity to epistemic states of affairs in the minimizing
context of unus could develop.
From indefinite NP to bare NP: Why does the indefinite
article disappear?
Ljudmila Geist
Universität Stuttgart
[email protected]
Donnerstag, 25. 02. 2016, 9:00–10:00, Raum: G 201
In this talk, I will discuss conditions for the omission of the indefinite article
in German. In the predicate position, most sortal nouns in modern German
require an indefinite article.
317
AG11
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 318 — #330
AG 11: Indefinites between theory and language change
(1)
Udo ist ein Held / ein Idiot. ‘Udo is a hero/ an idiot.’
Other sortal nouns and relational nouns may occur without an indefinite
article. Two cases can be distinguished: sortal nouns referring to wellestablished groups, such as names of professions and nationalities (2), and
relational nouns (3).
AG11
(2)
Udo ist Lehrer (profession) / Deutscher (nationality).
‘Udo is a teacher / German.’
(3)
Udo ist Teil des Teams / Kunde bei BASE.
‘Udo is part of the team / a customer of BASE.’
Diachronically, the bare use of nouns corresponds to the early phase of Old
High German (OHG, c. 800-1050), where all nouns were bare in the beginning. Ein appeared as an indefinite article with referential nouns in the late
9th century (Oubuzar 2000). As Petrova (2015) shows, ein also began to be
used with predicate nouns, which are non-referential. She observes that
NPs with ein in the predicate position co-varied with bare nouns until the
end of the OHG period. Given this, two competitive hypotheses about the
evolution of the indefinite article, from OHG to modern German, can be
assumed:
Hypothesis 1: Ein spread to predicate nouns but had stopped before it
could reach nouns denoting socially established groups and some relational
nouns.
Hypothesis 2: Ein spread to all predicate nouns but later was omitted in
those combinations with nouns denoting socially established groups and
with some relational nouns.
My pilot study of predicate nouns in the Bonner Frühneuhochdeutschkorpus corpus of Early New High German (1350-1650) revealed that during
this time period indefinite predicate nouns of the type in (2) and (3) were
used with ein, thus Hypothesis 2 is more likely to be correct. But, how
can the omission of the indefinite article be accounted for on the syntaxsemantics interface? I will argue that the omission of the indefinite article
in (2) is triggered by a process similar to pseudo-incorporation (Dayal’s
2011), occurring with a reduction of nominal structure. In (3), however,
318
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 319 — #331
AG 11, Raum G 201
the omission of the indefinite article can be accounted for by head movement of the predicate noun to the position normally occupied by the indefinite article. Although pseudo-incorporation and head-movement have
similar effects on predicate NPs – i.e., the omission of ein and restricted
modifiability (not shown here) – these processes can be distinguished, as it
is only pseudo-incorporation that presupposes social-establishedness and
is restricted to human referents. Head-moved predicate nouns, however,
do not require social-establishedness and need not be restricted to human
individuals (4).
(4)
Der Baikalsee ist Teil der Baikal-Riftzone.
‘Lake Baikal is part of the Baikal Rift Zone.’
AG11
Scalar epistemic indefinites: A case study of weiß Gott win Present Day German
Patrick G. Grosz
Universität Tübingen
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: G 201
Baground: This talk investigates epistemic indefinites of the form weiß
Go w- (wGw) ‘God knows w-’, as in (1), attested in a wide range of European languages (Haspelmath 1997:131).
(1)
Vor einem Mülleimer findet der Knirps einen Schuhkarton, den weiß
Gott wer dort in die Landschaft geschmissen hat. [DeReKo: Braunschweiger Zeitung, 08.10.2008]
‘[…] the toddler finds a shoe box, which God knows who has thrown
into the landscape.’
Core Proposal: While wGw phrases originate as separate clauses (CPs),
(2a), parenthetically inserted into a host clause, (2b), I argue that weiß Go
(wG) ‘God knows’ in Present Day German has been diachronically reanalyzed as an indefinite particle (like German irgend): it combines with a
319
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 320 — #332
AG 11: Indefinites between theory and language change
wh-element to form a complex word, (2c). As an indefinite particle, wG
can no longer occur outside a preposition (cf. auf irgendwen ‘for someone
or other’ vs. *irgend auf wen), i.e. (2b) cannot yet be Sage III. (Note that
(2a)-(2c) coexist in Present Day German.)
(2)
a.
b.
c.
AG11
OK
Er hat gewartet. [CP Weiß Gott, auf wen.] Sage I (wGw as
separate clause)
‘He was waiting. [CP God knows, for whom.]’
OK
Er hat [CP weiß Gott auf wen] gewartet. Sage II (wGw as
parenthetical clause)
‘He was waiting [CP God knows for whom].’
OK
Er hat auf [DP [D weiß Gott wen]] gewartet. Sage III (wGw
as complex word)
‘He was waiting for [DP [D God knows whom]].’
Syntactic Evidence: I provide syntactic evidence to show that German
wGw has reached Sage III. Standard analyses of parentheticals such as
(2b) (e.g. Kluck 2011) assume that the wGw phrase is a complete CP with
deletion (“sluicing”). This entails that the wGw part can always be expanded
into a complete sentence at Sage II. While this is possible for (2b), as in
(3), it is impossible for (2c), as in (4). Example (2c) is thus unambiguously
Sage III.
(3)
Er hat noch weiß Gott auf wen gewartet.⇒ OK Weiß Gott, auf wen
⟨er gewartet hat⟩.
(4)
Er hat noch auf weiß Gott wen gewartet.⇒ *Weiß Gott, wen ⟨er
(au) gewartet hat⟩.
Synronic Semantics: I argue that wGw indefinites combine indefiniteness with a scalar component. They [i.] existentially quantify over alternatives that the wh-element introduces [ii.] which are high on a salient scale.
This often gives rise to a scalar effect (e.g. weiß Go was ‘God knows what’
≈ ‘something remarkable’). I show that this scalarity is part of the truthfunctional content of a sentence (cf. Potts 2015); e.g. it can be targeted by
clausal negation, (5).
320
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 321 — #333
AG 11, Raum G 201
(5)
Man kann nicht weiß Gott was erwarten […]
[DeReKo: Süddeutsche Zeitung, 2000]
‘One can not expect God knows what.’(≈ not [sth. remarkable] →
only [sth. average])
From correlative protases to existential pronouns in
Basque
Ricardo Etxepare
CNRS, IKER UMR5478
[email protected]
AG11
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: G 201
Basque has a rich system of quantificational expresssions based on socalled “indeterminate pronouns” (Kuroda, 1968): indefinite bases formally
identical to wh-words. Among the wh-word based expressions are existential pronouns equivalent to the English “someone, something” series. Those
pronouns are formed by combining the wh-word and a suffix, formally
identical to the prefixal complementizer bait-, used in various relations of
subordination, particularly causals and relatives (Oyharçabal, 1987):
(2)
a.
b.
zer-bait
what-comp
“Something”
Jin bait-a gizona
come comp-is man.de
“The man who came”
Unlike the other complex expressions in the wh-pronoun series, existential pronouns present a high degree of morphosyntactic variation in their
lexical realization. The range of morphological variation existing in the category of existential pronouns in Basque can be properly understood under
the hypothesis that existential pronouns come from correlative protases.
Free relatives, closely related to correlative protases, are a common source
in the emergence of existential quantifiers (Haspelmath, 1997). Correlative
321
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 322 — #334
AG 11: Indefinites between theory and language change
constructions exist as a productive form in a few areas of the Basque country, and they were general before. The forms to be compared to (2a) is (3):
(3)
AG11
Zer (ere) baita
(correlative protasis/free relative)
what even COMP-is
“What(ever) it is”
I show that the dialectal variation in the expression of existential quantifiers corresponds to the selective lexicalization of different portions of (3),
with the complementizer itself a constant element. The aim of this paper
is to show that dialectal variation, together with the historical record, can
shed light on some of the finer details of the process leading from correlative protases to existentials. The paper also discusses the progressive
phenomenon of semantic weakening to the extent allowed by the existing
sources.
Strong polarity contexts and evolution of n-words
Amel Kallel1 & Pierre Larrivée2
1
University of Tunis, 2 Normandie Université, Unicaen, CRISCO EA4255
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 12:00–12:30, Raum: G 201
The Grammaticalisation framework has suggested that there is a tendency
for items to evolve historically into increasingly more abstract elements on
a pathway of change (Traugott & Dasher 2002 i.a.). An example of pathways of change is provided by n-words (such as French aucun ‘none’) that
generally evolve out of negative polarity items (Haspelmath 1997). Such
pathways of evolutions have recently been proposed to relate to a model
of feature acquisition: learners would attribute items the most specified feature compatible with the input (Willis 2011). Items can thus gain a stronger
feature but cannot lose it, leading to an assumption of irreversible change.
Once negative polarity items have become n-words, they are expected to
neither retain nor recreate polarity uses. This strong claim is invalidated by
the behaviour of declining n-words in early French (Larrivée 2014), some
322
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 323 — #335
AG 11, Raum G 201
oh which recreate negative polarity functions. Such recreations are however constrained: only movement along adjacent spaces on a semantic map
is attested. This is speculated to be because adjacent functions share specific bridging contexts where an item can be analysed as expressing either
function. What this predicts is that during or before the period where a
polarity item is becoming a n-word, there is a preponderant proposition
of bridging strong polarity context such as those commanded by without.
That would be because sans is a bridging context where a NPI can be readily reanalysed as a n-word. The role of strong polarity contexts for this
pathway of change is what is tested on the basis of French n-word aucun
‘no (N)’. Its evolution into a n-word from a NPI is shown to occur in the
16th century as the competing nul is declining (Kallel & Ingham 2014). The
relative weight of the weak, strong and n-word uses of aucun is quantified
in a corpus of remission letters (pleas written by condemned criminals to
the Chancery) from the late 14th to the end of the 16th . Preliminary results
support the view that strong polarity contexts are a bridging context in
the evolution of items into n-words. It substantiates the view that change
progresses along a pathway of functions because these share bridging contexts.
References: • Haspelmath, Martin. 1997. Indefinite pronouns. OUP. • Kallel, Amel & Richard
Ingham. 2014. Evidence from a correspondence corpus for diachronic change in French indefinites 1450–1715. Hansen and Visconti (Eds). e Diachrony of Negation. Benjamins. 213-234.
• Larrivée, Pierre. 2014. L’unidirectionalité irréversible du changement linguistique comme
conséquence de l’acquisition ? Le cas d’expressions négatives déliquescentes en français ancien. CMLF 2014. • Traugott, Elizabeth C. & Richard B. Dasher. 2002. Regularity in semantic
change. CUP. • Willis, David. 2011. Negative polarity and the quantifier cycle: Comparative
perspectives from European languages. Larrivée and Ingham (Eds). e Evolution of Negation:
Beyond the Jespersen Cycle. de Gruyter. 285-323.
323
AG11
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 324 — #336
AG 11: Indefinites between theory and language change
Indefinite polarisation and its scalar origin: Evidence from
Japonic
Moreno Mitrović
Karl-Franzens-Universität Graz
[email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: G 201
AG11
It is has been well investigated by Szabolcsi (2013), and Kratzer & Shimoyama (2002), among many others, that Modern Japanese (MdJ), among
other languages, constructs universal and polar terms by combining a whword and the particle mo (henceforth μ). Compositionally, the semantic
role of the μ particle obtains a universal reading as μ0 obligatorily activates
the alternatives of its complement (i.e., the wh-abstract with an existential presupposition), and asserts that all alternatives be true. What remains
formally unexplored, however, is the historical dimension of this compositional behaviour in light of the absence of polar pattern in the earliest
stage of the language, since wh+μ terms were terms not licensable under
negation. This paper shows not only (i) that polarity system in Japonic is diachronically derived from scalar universals but also (ii) when and how this
process took place by adopting Chierchia’s (2013) theory of grammaticised
scalar implicatures (SIs).
In Old Japanese (OJ; c. 8th ce), the [wh+μ] quantificational expressions
were confined to inherently scalar (σ) complements, i.e. either numeral
nominals or inherently scalar wh-terms (e.g. how-many/when), as Whitman
(2009) first noticed. Focussing on the latter μ-hosts. The only two kinds
of wh-terms which can serve as μ-hosts we find in OJ are temporal- (1)
and quantity-wh-terms (2), i.e. those wh-abstracts with only a σ-domain of
alternatives, as shown in the table
324
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 325 — #337
AG 11, Raum G 201
wh-hosts to μ
#
+cala
itu mo ‘when μ’
11
iku mo ‘how much/many μ’
8
−cala
ado/na/nado mo ‘what/why μ’
0
ika mo ‘how μ’
0
ta mo ‘who μ’
0
AG11
One of the ideas central to the proposal made in the paper is that the
original μ0 associated with scalar hosts, i.e. those elements endowed with
[σ] feature, and that activated scalar alternatives were originally existential, i.e. truncated at the low end so as to exclude no/∅. The synchronic and
diachronic analysis of [wh+μ] quantification in Japanese rests on Chierchia’s (2013) system of grammaticised scalar implicatures, where scalar (σ)
and non-scalar (D) alternatives are lexically grounded and represented as
features (σ, D), which check the exhuastifier operator. The paper will demontrate how polarity-sensitivity arose in Early Middle/Classical Japanese
via syntactic featural change from [σ] to [D] ashe restriction on scalar whcomplements declines in the beginning of the EMJ period and non-scalar
wh-complements enter the structure. The analysis also makes reference to
early Indo-European conjecturing a potentially universal scalar core of polar terms.
325
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 326 — #338
AG 11: Indefinites between theory and language change
Konstruktionen mit Indefinita in altindogermanischen
Sprachen
Rosemarie Lühr
Humboldt-Universität zu Berlin
[email protected]
AG11
Freitag, 26. 02. 2016, 11:30–12:00, Raum: G 201
Indefinita in Nominalphrasen altindogermanischer Sprachen sind entweder Substantive im Genitiv oder mit dem Bezugswort kongruierende Adjektive. Im Hethitischen zeigt sich folgendes Stellungsverhalten: Besteht
in Possessivkonstruktionen der Possessor aus einem genitivischen Indefinitum, steht dieses hinter dem Possessum. Das Possessium ist entweder ein
sortales Nomen wie Haus, Gut, oder es bezieht sich auf unikale Teile von
Entitäten wie Blut, Kopf, Augen, Fleisch.; vgl. die Stellung des Genitivs kuelqua ‚jemandes‘ hinter dem Bezugswort:
(1)
HG § 44b A (KBo 6.2) ii
[35]
pár-na-ma
ku-e-el-ka
ins Haus-DIR-aber jemand-GEN
‚aber in jemandes Haus‘
Ist der Possessor jedoch kein Indefinitum, erscheint er vor dem Possessum.
(2)
TelErl I 66 (KUB XI 5 Vs. 8’)
ad-da-aš e-eš-ḫar-še-it
Vater-GEN Blut-sein
‚das Blut des Vaters‘
Der Referent des Possessums ist durch den vorangestellten Genitiv eindeutig identifizierbar. D.h., der Definitätswert der unter D0 eingebetteten
Konstruktionen verändert sich durch die Verschiebung einer Konstituente.
Während ku-el-ka „jemand-GEN“ sich unter der am tiefsten eingebetteten
DP in situ befindet, wird ad-da-aš „Vater-GEN“ nach links (i.e. „oben“) unter die DP von FP (= AgrPPOSS ) verschoben (Lühr 2004). Die erste Frage, um
326
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 327 — #339
AG 11, Raum G 201
die es hier geht, ist, welche Elemente innerhalb der DP im Hethitischen wie
das substantivische Definitum in situ verbleiben. Die zweite Frage ist, ob
es einen Zusammenhang zwischen der Stellung des substantivischen und
des adjektivischen Indefinitum im Hethitischen gibt. Denn als attributives
Adjektiv ist das Indefinitum heth. kuiški in der Regel hinter dem Bezugswort plaziert, während Demonstrativa, modifizierende Adjektive wie auch
Genitive davor auftreten. Die dritte Frage ist, ob das Hethitische bei der
Stellung der Indefinita einen Sonderweg gegenüber den anderen altindogermanischen Sprachen Altindisch und Griechisch beschreitet.
Literatur: • R. Lühr (2004): Der Ausdruck der Possessivität innerhalb der Determinansphrase
der ältesten indogermanischen Sprache, in: Šarnikzel. ed. by D. Groddek/S. Rößle, Dresden,
415–446.
Relative and indefinite pronouns: Synchrony and
diachrony: The case of Hittite
Andrei Sideltsev
Russian Academy of Sciences
[email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: G 201
The Hittite system of indefinite pronouns belongs to one of the most standard varieties:
Wh-words existential NPI universal Free Choice
kuiš
kuiš-ki
kuiš-ki kuišš=a kuiš, kuiš kuiš
The systems of this type are standardly believed to be built around whwords (Haspelmath 1997; Kratzer, Shimoyama 2002; Kratzer 2005; Yanovich
2005). I argue, however, that in Hittite it is rather the relative/subordinate
set (kuiš), phonologically identical to wh-words, that behaves in syntactical
terms identically to existential quantifiers (EQ further on).
EQs attest two linear positions in the Hittite clause – preverbal and second. The same two positions are attested by relative pronouns/subordinators,
327
AG11
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 328 — #340
AG 11: Indefinites between theory and language change
AG11
but not by wh-words which can either be first or preverbal. The preverbal
position of wh-words is different from EQs and relative pronouns. Pace
(Huggard 2015) and following (Becker 2014), I argue that the two positions
of relative pronouns and EQs are unmarked as for specificity and scope:
the relative phrases and QPs can be specific/wide scope and non-specific/
narrow scope in either position. But the preverbal positions of relative pronouns and EQs are not fully identical: preverbal relative pronouns are to
the left of the negation markers whereas EQs are to the right. Besides, the
linear word order in relative clauses is relative pronoun–NP whereas QPs
are linearized as NP–EQ. Thus at most the diachronical connection, but not
the synchronic identity can be maintained between relative pronouns and
EQs.
Still, Hittite provides evidence for the connection between relative pronouns and EQs within the history of Hittite, namely analogical impact of
relative pronouns influence upon EQs. I will discuss two pieces of data.
The first concerns how the two EQ positions evolved in the history of Hittite. In the earliest written texts of the Old Hittite (OH) period only the
preverbal position was available. I argue that 2P EQs appeared within the
written history of Hittite, in Middle Hittite period, and that they are due to
the analogy after relative pronouns/subordinators which were 2P already
in the OH time. The analogy explains the specificity/topicality properties
of 2P EQs. The analogy is demonstrably not after wh-words: wh-words in
Hittite never attest 2P constraint whereas relative pronouns/subordinators
do.
The second piece of evidence concerning relative pronoun/subordinator
impact on EQs comes from bare interrogatives, namely the use of relative
pronouns instead of indefinite ones in conditional clauses and after negations. Again, the use is not attested in Old Script originals and appears first
in Middle Script (MS) copies of Old Hittite texts. The use is ambiguous between relative pronouns/subordinators and wh-words. However, in half of
the earliest cases (MS texts) the ‘bare interrogative’ is in the second position, which is totally unexpected if wh-words are involved, but which is
easily explained if relative pronouns are involved.
References: • Becker, K. 2014, Zur Semantik der hethitischen Relativsätze, Hamburg: Baar, 2014.
328
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 329 — #341
AG 11, Raum G 201
• Haspelmath, M. 1997, Indefinite Pronouns. Clarendon Press: Oxford. • Huggard, M. 2015,
Wh-words in Hiite. PhD Dissertation, University of California. • Kratzer, A. 2005, Indefinites
and the Operators they Depend on: From Japanese to Salish. In: G. N. Carlson and F.J. Pelletier
(eds.), Reference and antification: e Partee Effect. Stanford, (CSLI Publications), 113-142.
• Kratzer A. and J. Shimoyama, 2002, Indeterminate Pronouns: The View from Japanese. In:
Yukio Otsu (ed.), e Proceedings of the ird Tokyo Conference on Psycholinguistics. Tokyo:
Hituzi Syobo, 1-25. • Yanovich, I. 2005, Choice-functional Series of Indefinite Pronouns and
Hamblin Semantics. In: E. Georgala and J. Howell (eds.), SALT XV 309-326, Ithaca, NY: Cornell
University.
Partitive case markers and indefiniteness: A diachronic
survey
Silvia Luraghi
Università di Pavia
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: G 201
In this paper, I survey the origin and the development of partitive case
markers, including adpositions. Items discussed include partitive cases, as
in Balto-Finnic, partitive genitives/ablatives, as in various Indo-European
languages, and partitive determiners, as in some Romance languages and in
Basque (see Luraghi & Huumo 2014). These items have in common the fact
that they originated from a case marker which, when functioning as partitive, does not diplay the typical function of cases, i.e. “marking dependent
nouns for the type of relationship they bear to their heads” (Blake 1994:
1). Indeed, NPs marked by such items can typically encode both direct objects and subjects, and have a quantifying function: they indicate unbound
quantity, and tend to develop in the direction of indefinite determiners.
Diachronic developments attest to partly different origins of partitive case
markers (Luraghi & Kittilä 2014). Partitive cases in Finnic languages originated from the Proto-Uralic separative case. The Basque partitive determiner is an allomorph of the present ablative: the two case forms became
differentiated at a pre-literary stage, when the features of number and definiteness in spatial cases had not yet emerged. In most ancient and some
329
AG11
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 330 — #342
AG 11: Indefinites between theory and language change
AG11
modern Indo-European languages, the genitive also has a partitive meaning (cf. languages such as Sanskrit or Latin, in which the ablative and the
genitive are distinct). In general, partitive case markers originate within
partitive construction. However, partitive case markers are no longer used
within partitive constructions possibly after losing their separative meaning, as shown in Finnic, in which partitive constructions feature the elative
case. Setting partitive items in a cross-linguistic perspective, I show that
one can trace a diachronic cline, which moves away from partitive nominal
construction, as in English “A piece of that cake” (cf. Koptjevskaja-Tamm
2006), and leads to a more generic quantifying function.
References: • Blake, Barry. 1994. Case. Cambridge: CUP. • Koptjevskaja-Tamm, M. 2006. Partitives. In K. Brown (ed.), Encyclopedia of Languages and Linguistics, vol. 9, Amsterdam: Elsevier,
218-21. • Luraghi, Silvia & Seppo Kittilä 2014. The typology and diachrony of partitives. In S.
Luraghi & T. Huumo (eds), 17-62. • Luraghi, Silvia & Tuomas Huumo (eds.) 2014. Partitive
cases and related categories. Berlin/New York: Mouton De Gruyter. • Luraghi, Silvia & Tuomas
Huumo 2014. Introduction. In S. Luraghi & T. Huumo (eds), 1-13.
Discussion
Chiara Gianollo1 , Klaus von Heusinger1 & Svetlana Petrova2
1
Universität zu Köln, 2 Bergische Universität Wuppertal
[email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:00–14:00, Raum: G 201
330
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 331 — #343
Arbeitsgruppe 12
Presuppositions in language acquisition
Anja Müller1 & Viola Schmitt2
1
Goethe-Universität Frankfurt, 2 Georg-August-Universität Göttingen
[email protected], [email protected]
AG12
Raum: E 402
Workshop description
Evidence from adult speakers suggests that apart from standard truth-conditional content, content can also be contributed on the levels of presupposition and of implicature. The organization of these levels as well as the particular analysis of individual expressions w.r.t. these levels are core questions of current linguistic research. The investigation of when children acquire which content types helps to determine the general path of language
acquisition itself and is central for the theoretical questions mentioned
above. Work on the acquisition of implicatures suggests implicated content becomes accessible later than truth-conditional-content (cf. Noveck
2001 a.o.). The present workshop focusses on the acquisition of presuppositional content, which so far has received less attention in language acquisition research. The question we raise and that papers should address
are, a.o.:
• Data from adult speakers suggests a distinction between “semantic”
presuppositions and “pragmatic” presuppositions. Is this distinction
reflected in acquisition? Do children acquire ”pragmatic” presuppositions later or earlier than ”semantic” presuppositions?
• At what age do children interpret sentences with presupposition
331
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 332 — #344
AG 12: Presuppositions in language acquisition
triggers adult-like? Do we find differences between the levels of truthconditions and presupposition in children?
• At what age do children show an adult-like behavior w.r.t. presupposition projection and accommodation?
Given that online methods seem to be especially sensitive for children’s
competence methods studies using online techniques are of particular interest for the workshop.
AG12
References: • Noveck, Ira. 2001. When children are more logical than adults: Experimental
investigations of scalar implicature. Cognition 78. 165–188.
Cumulative universal quantification
Kenneth Drozd
Center for Language and Cognition Groningen (CLCG), University of
Groningen
[email protected]
Mittwoch, 24. 02. 2016, 14:00–15:00, Raum: E 402
The topic of the talk is children’s understanding of distributive, collective, and cumulative interpretations of quantified sentences. The first goal
would be to summarize the enormous amount of literature in this area and
compare the main families of theoretical developmental proposals made to
explain the published data. The second goal would be to show how recent
theories of universal quantification advanced by Champollion, Brasoveanu,
Szabolsci, Cable, and others bring new insights into the developmental results and new explanations of them.
332
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 333 — #345
AG 12, Raum E 402
The presuppositions of also and only: The view from
acquisiton
Francesca Panzeri1 & Fancesca Foppolo1
University of Milan-Biocca
1
[email protected]
Mittwoch, 24. 02. 2016, 15:00–15:30, Raum: E 402
Sentences like “Also/Only Lyn came” convey the information: (i) that Lyn
came (prejacent); (ii) that someone/no-one else came (alternatives). The additive particle also asserts the prejacent and presupposes that at least one
alternative is true (Karttunen & Peters 1979); only asserts that no alternative is true, and presupposes the prejacent (Horn 1969). In the acquisition
literature, previous works showed that children have problems with focus
operator like only (Crain et al. 1999; Paterson et al. 2003) and also (Hüttner et al. 2004), and that additive operators are harder than exclusive ones
(Bergsma 2002; 2006, Matsuoka 2004; Matsuoka et al. 2006), even if Berger
& Höhle (2012) showed that German children from age 3 are competent
with auch (also) and nur (only) if tested with a Reward-paradigm. We ran
a study on children s comprehension of the Italian versions of also (anche)
and only (solo), with a Reward-paradigm. We tested pre-school aged children: 16 4 year-olds and 17 5 year-olds. Children had to decide whether
some animals deserved a reward on the basis of an indirect statement in
which anche/solo were used. The two items appeared in two conditions: before the subject (Solo/Anche x V NP) or before the direct object (x V anche/
solo NP).
Results: We found that 4 year old children made significantly more errors
than 5 year olds (p=.017); in general, children made significantly more errors with also (40%) than with only (28%) (p<.001); also, a significant effect
of position (p<001) and a significant interaction of position and trigger type
is found (p<001): independently of age, children made significantly more
errors when also appeared in initial position (C1) than when it appeared
after the predicate (C2).
Conclusion: The additive particle anche is harder than the exclusive one:
this might depend on the fact that the alternatives are presupposed by also
333
AG12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 334 — #346
AG 12: Presuppositions in language acquisition
and asserted by only. As for the fact that sentence initial additive particles
are harder than middle ones, we tentatively suggest that this might due
to the processing cost to retrieve salient alternatives: this operation might
be harder when also appears sentence initially because the relevant alternatives are less constrained and the parser starts the process of retrieval
before relevant information is provided; analogously, it might be easier in
case of only because no contextual alternative has to be retrieved but in
fact negated.
Only for children
AG12
Yi-ching Su
National Tsing Hua University
[email protected]
Mittwoch, 24. 02. 2016, 15:30–16:00, Raum: E 402
Previous studies found preschool children show a VP-associated pattern for
sentences with pre-subject ‘only’ or zhiyou (e.g., Crain et al. 1994; Philip &
Lynch 2000; Zhou & Crain 2010), i.e., most children interpret (1)as ‘The cat
is only holding a balloon.’ In this study, we show that Mandarin preschool
children have the semantic knowledge for the use of pre-subject zhiyou
‘only’, and their non-adult interpretation is due to pragmatic infelicity in
the contexts. We used Truth Value Judgment Task in two conditions - Zhou
and Crain’s (2010) contexts (Condition A) and our revised contexts (Condition B). The major difference between the two conditions is that in Condition B, the other character in the story explicitly expresses the intention
to accomplish the action, whereas the presupposition required for the use
of zhiyou ‘only’ is not made clear in Condition A.
We tested 48 preschool children (age 4;07-6;10), who were randomly
assigned to the two conditions. The children in each condition were further divided into two groups - the older group (N=12, age 5;07-6;10) and
the younger group (N=12, age 4;07-5;06). For Adult-True trials, the acceptance rates were 23% and 96% for Conditions A and B, respectively, which
differs significantly (p<0.05). The difference between older and younger
334
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 335 — #347
AG 12, Raum E 402
groups was not significant. For Adult-False trials, the rejection rates were
94% and 96% for Conditions A and B, respectively, which did not differ
significantly. However, the justification reasons showed that the percentages of VP-association were 80% and 17% in Conditions A and B, respectively, which differs significantly (p<0.05). The difference between older
and younger groups was not significant.
To examine when children can accommodate the pragmatic infelicity,
we further tested 14 school-age children (age 8;0-9;08) with Condition A.
The results showed that for Adult-True trials, the acceptance rate was 54%,
which was higher than that of the preschool children in Condition A, but
still much lower than that of preschool children in Condition B. As for
Adult-False trials, the rejection rate was 100%, and 29% of which were VPassociation responses, which were significantly lower than that of preschool
children in Condition A, but still a little higher than that of preschool children in Condition B. This shows that children’s pragmatic accommodation
for the use of pre-subject zhiyou ‘only’ gradually develops with age, but is
still not adultlike before age 10.
Children fail to repair presuppositions
Tom Roeper1 & Jennifer Rau1
University of Massachusetts
1
[email protected]
Mittwoch, 24. 02. 2016, 16:30–17:00, Raum: E 402
What preconditions must be met in order for children to be able to reject
a false presupposition (Schulz 2003). Conversation demands that we repair
presuppositions if there is an error. It is harder than answering a direct
question, which is much easier for children. Conversation has overt and
covert “repairs” frequently and it is a part of being a competent speaker.
This means a presupposition or implicature failures are everywhere (and
possibly harder for autistic children). Our evidence shows that children are
often in grade school before they can repair presupposition failure appropriately.
335
AG12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 336 — #348
AG 12: Presuppositions in language acquisition
AG12
Participants and Materials: 36 participants between the age of 4;6 and 8;7.
The 28 trial sentences were each accompanied by a picture and were designed to include different levels of (in)compatibilty of a proposition with
the common ground (CG) to test for how willing children are to accommodate far-fetched propositions. The presupposition triggers used were factive predicates as in, definite articles triggering existential presuppositions
(in topic position; see von Fintel 2003) and clefts. In addition, we distinguished between visually supported conflicts and CG conflicts. Sentences
presented either in the assertive or presupposed form.
Results: 77% of pre-kindergarten accommodate, only by 8yrs to 45% reject presupposition. Children primarily attempt to Accommodate Question, then make various adjustments before outright rejection at age 8yrs.
Although the inquiry into presupposition failure is still preliminary, we
can project hypothetical stages on an acquisition path:
Accommodate question
Keep Presupposition, Adjust Common Ground
Change Question under Discussion
a. change factivity
b. adjust question
We suggest that the capacity to challenge presuppositions is linked to
the capacity to carry out normal but-implicatures. These presupposition
repairs are not as common as implicature reversals, but we argue that they
are related conversational properties and a significant dimension of pragmatics far more prevalent than previously assumed and essential to normal
conversation.
336
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 337 — #349
AG 12, Raum E 402
How children deal with a contextually canceled
presupposition
Magda Oiry
University of Massachusetts
[email protected]
Mittwoch, 24. 02. 2016, 17:00–17:30, Raum: E 402
This study reports findings on question formation in a context where the
presupposition is either satisfied or canceled. We tested three-to-six years
old with the help of small scenarios, in which the targeted embedded proposition has a true answer to it or not. The main research question was to find
out whether children would be able to produce long distance questions in
a context where the presupposition under the subordinate clause, as in
‘What do you think Mommy bought?’, was shared knowledge by all the
participants or not.
In the presuppositional context, it is completely acceptable to ask a matrix question, because there is no disparity in the shared knowledge of the
situation. Both speaker and hearer know Mommy bought something. Only
the non-presuppositional context urges the use of a long distance question,
because it is clearly established that Mommy didn’t buy anything. So it is
infelicitous to ask a root question such as ‘what did she buy?’, when she
hasn’t bought anything. There is a difference between what the utterer of
the targeted question knows and what the addressee knows. It is relevant
to ask a question to the person, who is not aware of the absence of a true
answer.
Looking at the results from 33 French-speaking children (2;11 to 6;03),
a strong contrast is displayed between the youngest group and the three
other groups. Only three year olds can’t utter any target question, they resolve to matrix questions and Scope marking questions in the first context
or don’t respond. But they do not deal well with a canceled presupposition and could only ask LD yes-no questions in the non-presuppositional
context, utter an answer or refuse to answer at all.
Overall, children have trouble dealing with the second context, but we
can see the older they get, they are getting better at producing the target
337
AG12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 338 — #350
AG 12: Presuppositions in language acquisition
question. Interestingly, one of the strategies that emerges from context 2
is the use of questions to try to accommodate the presupposition.
On the acquisition of presupposition projection
Cory Bill1 , Jérémy Zehr2 , Lyn Tieu3 , Jacopo Romoli4 , Stephen Crain1 &
Florian Schwarz2
1
Maquarie University, 2 University of Pennsylvania, 3 Ecole Normale
Supérieure, 4 Ulster University
[email protected]
AG12
Mittwoch, 24. 02. 2016, 17:30–18:00, Raum: E 402
Bill et al. (2014) report that 4-year-olds have knowledge of the global presupposition of win (1a), but struggle with local accommodation (1b).
(1)
The bear didn’t win the race.
a. Presupposition: The bear participated in the race.
b. Local accommodation: The bear didn’t both participate and win
the race.
In this study, we embed the presuppositional trigger under the negative
quantifier none (2). The nature of such presupposition projection has been
controversial, with the readings in (2) being variously endorsed by different authors (Heim 1983, Beaver 1994, Schlenker 2008). We investigate the
availability of these readings, focusing on whether children are able to access the same set of interpretations as adults.
(2)
338
None of the bears won the race.
a. Existential presupposition: At least one bear participated in the
race.
b. Universal presupposition: All of the bears participated in the
race.
c. Local accommodation: None of the bears both participated and
won the race.
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 339 — #351
AG 12, Raum E 402
Experiment: We tested 24 adults using the Covered Box paradigm (Huang
et al., 2013). Participants heard descriptions of events and saw two images,
one visible and one obscured (the ‘covered box’); they were expected to select the visible image if it was consistent with the sentence, and the covered
image otherwise. Each trial consisted of a context, followed by the critical
description and images: condition a) is inconsistent with both presuppositions; condition b) is only inconsistent with the universal presupposition.
In addition to the 4 Existential and 4 Universal trials, participants saw control conditions where the universal presupposition was satisfied.
Results: Adults made significantly more covered picture selections in the
Existential condition than in the Universal and ControlTrue conditions;
Universal and ControlTrue conditions did not differ. Adults also accessed
local accommodation readings, selecting the visible image in the Existential condition about 60% of the time. We found no evidence, however, for
a universal presupposition. Preliminary response time analyses also suggest that local accommodation is only involved in the Existential condition.
Given Bill et al.’s proposal that children rarely access local interpretations,
we expect the child participants to select the visible image in the universal
but not existential condition; such a _nding would further corroborate the
adult result that sentences like (2) are associated only with an existential
presupposition.
Exhaustivity of structural focus in Hungarian:
Presupposition or implicature?
Lilla Pintér
Hungarian Academy of Science
[email protected]
Mittwoch, 24. 02. 2016, 18:00–18:30, Raum: E 402
Resear question: The study compares the interpretation of (1) sentences
with structural focus, and (2) sentences with neutral intonation and word
order, which are claimed to express different - presupposed and accidentalpragmatic - kinds of exhaustivity, respectively.
339
AG12
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 340 — #352
AG 12: Presuppositions in language acquisition
AG12
Baground: According to the standard view (Kenesei 1986, Szabolcsi 1994,
Bende-Farkas 2009), in Hungarian the exhaustivity of syntactically and
prosodically marked structural focus is a presupposition. Sentences with
neutral intonation and word order can also be interpreted exhaustively;
however, this is an implicature arising only in certain contexts.
This assumption was questioned by Wedgwood (2005), in whose analysis exhaustive interpretation is a pragmatic implicature in the case of
sentences with and without structural focus alike. Experimental findings
of Onea & Beaver (2011), Kas & Lukács (2013), and Gerőcs, Babarczy &
Surányi (2014) seemed to support Wedgwood’s view; however, the Truth
Value Judgment Tasks they used were fundamentally unsuitable for pointing out presupposition-violations.
Experiments: I conducted sentence-picture verification tasks in which
each test sentence had the same structure: in Experiment 1, they all contained structural focus; whereas in Experiment 2, there were only neutral
SVO sentences. There were four conditions differing in the type of pictures: in addition to three baseline conditions there was a critical condition in which I tested whether or not the test sentences are accepted
in non-exhaustive situations. The task was to judge each item by using a
three-point scale in which the scores were represented by sad, straight and
happy smiley faces.
Results and conclusion: I tested four age groups in each experiment: preschoolers, 7-year-olds, 9-year-olds, and adults. The results showed that there
is a remarkable difference between the interpretation of structural focus
constructions and neutral SVO sentences. In the non-exhaustive contexts
of Experiment 1, older children and adults mostly chose the middle option,
indicating that structural focus constructions are only partially true there.
As opposed to this, the majority of the participants accepted sentences of
Experiment 2 in the same condition, which proves that speakers do not
interpret these constructions exhaustively when there are no contextual
cues to trigger the generation of an implicature.
In case of adult participants, reaction times were also measured, and the
results of the non-exhaustive condition of Experiment 1 did not differ significantly from those of the control conditions, which also provide evidence
against the assumption that exhaustivity is an implicature in the case of
structural focus.
340
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 341 — #353
Arbeitsgruppe 13
Adjective order: Theory and experiment
Eva Wittenberg1 & Andreas Trotzke2
1
University of California, San Diego, 2 Universität Konstanz
[email protected] , [email protected]
AG13
Raum: E 402
Workshop description
In linguistics, the issue of adjective order has a long history. Bloomfield
(1933) already made some remarks on the robust restriction that size adjectives usually precede color adjectives (a small black dog vs. a black small
dog). Following these early notes, many researchers in linguistic typology
investigated adjective order in the form of semantic hierarchies (Dixon
1982, Bache & Davidsen-Nielsen 1997). Our workshop aims at bringing together recent work from both theoretical and experimental linguistics to
reframe this classical topic.
In particular, we are interested in the tension between proposals of finegrained syntactic hierarchies (Scott 2002; Laenzlinger 2005) and large-scale
semantic distinctions as being relevant for ordering (Stavrou 2001; Truswell
2009). Furthermore, do non-canonical orders involve a specialized focus
position (cf. the BLACK small dog (and not the BROWN small dog); e.g.
Alexiadou et al. 2007; Svenonius 2008), or do other semantic factors explain
such patterns (Cinque 2010, 2014)? Experimental evidence has shown that
the abstract principles governing adjective order seem to constitute a separate domain of representation that can be selectively impaired (Kemmerer
et al. 2009). Experimental work can have important impact on theory in
341
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 342 — #354
AG 13: Adjective order: eory and experiment
many respects. For example, to what extent is the phenomenon syntactic at all, given that (some) non-canonical orders do not result in syntactic processing difficulties (Huang & Federmeier 2012)? Can the role of abstract semantic categories in linear precedence be reduced to mere online
abstraction (Vandekerckhove et al. 2015)? What is the linking hypothesis between adjective order and perceptual features that adjectives denote
(Belke 2006)? We invite submissions that present language specific, crosslinguistic/comparative, and experimental work on adjective order. We especially encourage submissions of relevant pragmatics and corpus work.
AG13
Property subjectivity predicts adjective ordering
preferences
Gregory Scontras1 , Judith Degen1 & Noah D. Goodman1
Stanford University
1
[email protected], [email protected], [email protected]
Donnerstag, 25. 02. 2016, 9:00–9:30, Raum: E 402
Cross-linguistically stable preferences for adjective ordering have been widely documented, yet the factors that determine these preferences are still
poorly understood. Our approach to the investigation of adjective ordering
preferences synthesizes strategies from the psychological approach, probing
the principles that underlie these preferences, and from the grammatical
approach, using descriptive semantic classes of adjectives to structure and
our investigation and smooth our data. Distilling the psychological proposals that precede us into a single feature, we advance the hypothesis that it
is the subjectivity of the property named that determines ordering preferences: less subjective adjectives occur closer to the substantive head of the
nominal projection, that is, to the modified noun. While subjectivity can
be assessed directly (by asking participants), we show it can more reliably
be measured as the extent to which two people can disagree about a description without one necessarily being wrong (i.e., the degree to which an
adjective description admits faultless disagreement). In “the big blue box,”
judgments about bigness are likely less consistent than judgments about
342
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 343 — #355
AG 13, Raum E 402
blueness; “blue” is less subjective than “big,” and so it occurs closer to the
noun “box.”
To test the hypothesis that adjective subjectivity predicts ordering preferences, we established two empirical constructs: the preferences themselves, which we measured using naturalness ratings and validated with
corpus statistics; and adjective subjectivity, which we operationalized as
potential for faultless disagreement and corroborated with a direct “subjectivity” measure. We find that an adjective’s semantics does predict its
distance from the modified noun: faultless disagreement scores account for
88% of the variance in our ordering preference data (r2 = 0.88, 95% CI [0.77,
0.95]). Our results suggest that ordering preferences likely emerge, at least
partially, from a desire to place less subjective content closer to the substantive head of a nominal construction (i.e., closer to the modified noun).
For now we can only speculate about the ultimate source of this desire:
Subjective content allows for miscommunication to arise if speakers and
listeners arrive at different judgments about a property description. Hence,
less subjective content is more useful at communicating about the world.
Whatever its source, the success of subjectivity in predicting adjective ordering preferences provides a compelling case where linguistic universals,
the regularities we observe in adjective ordering, emerge from cognitive
universals, the subjectivity of the properties that the adjectives name.
Adjective order restrictions: The influence of
temporariness on prenominal word order
Sven Kotowski1 & Holden Härtl1
1
University of Kassel
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 9:30–10:00, Raum: E 402
A central factor claimed to contribute to Adjective Order Restrictions (AORs)
is temporariness, usually stated in terms of a constraint such that permanent properties encoded by individual-level (IL) adjectives (cf. Kratzer 1995)
occur closer to the head noun than transitory (stage-level (SL)) ones (cf.
Cinque 2010; Eichinger 1992; Haliday 2014; Larson 1998) as in (1):
343
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 344 — #356
AG 13: Adjective order: eory and experiment
(1)
AG13
a.
b.
einen betrunkenen jungen Mann
?
einen jungen betrunkenen Mann
‘a drunk young man’
‘a young drunk man’
Following Kotowski (2015), this paper’s overall claim relegates the status
of temporariness to an epiphenomenon of more general criteria governing AORs: (i) AORs can most adequately be captured by assuming layered
modification classes, one of which is classificatory (generic/IL) and directly
adjacent to the head noun. (ii) In a qualificatory layer outside the classificatory one IL- as well as SL-adjectives are found, with AORs sensitive
to constraints along the lines of the scalar properties of relative and absolute adjectives (cf. e.g. Kennedy & McNally 2005). Relative modifiers (all
ILs) fairly robustly precede absolute ones (among them all SLs) in German/
English.
Against this background, we report on three studies on AORs in German. A corpus study drawing on AAN-phrases evaluated preferred positions of selected SL-adjectives (e.g. nackt ‘naked’), showing invariable
‘quality≫relational’ and robust ‘relative≫absolute’ constraints, yet no general “SL≫IL” preference. Two German questionnaire studies tested word
order preferences for IL/SL-ambiguous adjectives. First, a two-versions rating study made use of evaluative adjectives disambiguated in introductory
paragraphs, with 100-split-ratings applied to different adjective orders in
follow-up sentences. The second rating study employed deverbal adjectives ending in –bar (∼‘–ible/–able’). The results of neither study support
assumptions along the lines of ‘SL≫IL’ constraints, not corroborating hypotheses of (quality) AORs as core grammatical phenomena. The presumed
IL-SL-divide and its bearing on adjective order are epiphenomenal to more
general adjective classes and modification layers.
344
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 345 — #357
AG 13, Raum E 402
Remarks on the order of adjectives cross-linguistically
Guglielmo Cinque
Ca’ Foscari University, Venice
[email protected]
Donnerstag, 25. 02. 2016, 10:00–11:00, Raum: E 402
Some aspects of the attested order of adjectives across languages will be
considered, both with respect to the relative order among them and with
respect to the position they take in relation to the noun. Languages will be
seen to either have a rigid order of adjectives if they only have attributive
direct modification adjectives (Yoruba), no rigid order if they only have
indirect modification adjectives (Malayalam), or just an unmarked order if
they have both sources (English). Since the position of the adjectives with
respect to the noun is intimately related to the head-initial or head-final
character of the language a derivational account of word order typology
will also be sketched.
A functional-cognitive analysis of the order of adjectival
modifiers in the English NP
Tine Breban1 & Kristin Davidse2
University of Manchester, 2 University of Leuven
1
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 11:30–12:00, Raum: E 402
Our analysis of the order of adjectival modifiers in the English noun phrase
(NP) has two central tenets. We adhere to the cognitive premise that grammatical categories are not primitives; rather, the whole construction defines its elements of structure (Langacker 1987, Croft 2000). Our approach
hence starts from the assembly of modifier relations making up the English NP, with a focus on those that can be realized by adjectives. We adhere to the cognitive-functional assumption that a construction’s meaning
is coded by its lexico-grammatical form (e.g. Langacker 1987, McGregor
345
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 346 — #358
AG 13: Adjective order: eory and experiment
AG13
1997). As such, the relative ordering of elements in the English NP is considered a reflection of their meaning, and, by consequence, if we want to
explain the order of modifier relations realized by adjectives, we have to
relate them to the different functions they code.
The descriptive approach we take to the meaning of the English NP and
its elements is a functional analysis building on the tradition of Bolinger
(1968), Halliday & Hasan (1976), Dixon (1982), Halliday (1994) and Bache
& Davidsen-Nielsen (1997), which we have refined and expanded in our
corpus-based work (e.g. Breban 2010, Breban & Davidse forthcoming, Vandewinkel & Davidse 2008). In addition to the four well-known functions of
classifier, epithet (or qualitative attribute), noun-intensifier and secondary
determiner, we include two further functions, categorization evaluators
and focus markers. We will illustrate all functions with examples from our
own corpus studies of Present-day and historical English. After defining
and illustrating the functions, we focus on how they are encoded by the
syntagmatic combinatorics of the modifier zone of the English NP.
We propose that these combinatorics and hence the order of adjectives
in the English NP is primarily motivated by the distinct head-modifier or
modifier-submodifier relations the different functions are involved in (in
addition to these general ordering principles, however, pragmatic and contextual factors create secondary variation). We make a main distinction between 2 types of relations: descriptive modifiers, which add descriptive material to their head and are processed compositionally, and interpersonal
modifiers, which do not ‘add’ their speaker-related meanings to the lexical
material in their scope, but change and mould it (McGregor 1997: 236). Both
relations are syntagmatically encoded in such a way that the dependent element is typically positioned to the left of the element it is dependent on
(although in some cases, as we will explain, it occurs to the right). Our
proposed ordering is based on attested orders in Present-day English and
historical corpus data.
346
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 347 — #359
AG 13, Raum E 402
Linear vs hierarchical, two accounts of premodification in
the of-binominal noun phrase
Elnora ten Wolde
University of Vienna
[email protected]
Donnerstag, 25. 02. 2016, 12:00–12:30, Raum: E 402
This talk will present, based on a corpus study of the premodification patterns of of -binominals, a comparison between linear, predominantly semantic zone-based approaches to English premodification ordering (e.g.
Dixon 1982; Feist 2012) to the hierarchical approach in Functional Discourse Grammar (FDG; Hengeveld & Mackenzie 2008).
Premodification patterns play a central role in the categorization of of binominals in general, and particularly, in the grammaticalization of the
evaluative binominal noun phrase (EBNP; a beast of a man) into an evaluative modifier (EM; a beast of a Hollywood year), as the first noun integrates
itself into the pre-existing premodification patterns (ten Wolde & Keizer
forthc.). This paper examines to what extent the two different approaches
to premodification can capture and account for this development.
For the zone-based approach, the project has adopted Ghesquière’s (2014)
construction based, functional-cognitive model of the NP, which incorporates both subjectivity and inter-subjectivity dimensions as well as captures scopal and modificational relations. This model is compared to FDG’s
function to form approach to premodification, where the linear ordering of
modifiers (as well as grammatical elements) at the Morphosyntactic Level
is determined by the function of these elements at the pragmatic and semantic levels in a top-down, outside-in manner (Keizer 2015: 218-231).
The paper comprises two parts: First, based on an empirical study of
data from the Corpus of Contemporary American English and the Corpus of
Historical American English, the development of the premodification distributional patterns of the EBNP and the EM will be presented. Second,
the paper will discuss the theoretical implications of the two approaches
mentioned when explaining these patterns, focusing on (i) the subjective
347
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 348 — #360
AG 13: Adjective order: eory and experiment
and objective distinctions and (ii) the transition from EBNP to EM. It is expected that the best results will most likely be obtained from an integration
of the two approaches.
References: • Dixon, R.M.W. 1982. Where have all the adjectives gone? Berlin: Mouton de
Gruyter. • Feist, Jim. 2012. Premodifiers in English: eir Structure and Significance. Cambridge:
CUP. • Ghesquière, Lobke. 2014. e Directionality of (Inter)subjectification in the English Noun
Phrase. Berlin: Mouton de Gruyter. • Hengeveld, Kees; Mackenzie, J. Lachlan. 2008. Functional
Discourse Grammar: A typologically based theory of language structure. Oxford: OUP. • Keizer,
Evelien. 2015. A Functional Discrouse Grammar for English. Oxford: OUP. • ten Wolde, Elnora;
AG13
Keizer, Evelien. forthc. Structure and substance in Functional Discourse Grammar: The case
of the Binominal Noun Phrase.
Free not-so-free adjectival word order in Latin
Giuliana Giusti1 & Rossella Iovino1,2
1
Ca’ Foscari Univ. of Venice, 2 Roma Tre
[email protected], [email protected]
Donnerstag, 25. 02. 2016, 12:30–13:00, Raum: E 402
1. e issue: Latin is well known for having free adjectival word order
inside the nominal expression (henceforth NE) as well as for allowing adjectives to appear in the clause in discontinuous order with respect to the
NE. These are issues for any configurational approach and real problems
for generative approaches that propose universal hierarchies (Cinque 2005,
2010), and/or strict motivation for internal and external merge, and/or the
elimination of optionality (Chomsky 1995).
2. Aims and goals: The paper presents an overview of the different patterns
in which adjectives are found in Latin and how they relate to discourse
features. Although the Latin DP is clearly configurational (cf. Devine &
Stephens (2006), Gianollo (2005), Giusti & Oniga (2007), Ledgeway (2012),
Iovino (2012), Langslow (2012), pace Spevak (2010) and Bošković (2005),
there is a large degree of optionality of word order in NEs.
3. e data: Based on quantitative and qualitative data, provided by previous literature, we show that the position of N when combined with a single
348
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 349 — #361
AG 13, Raum E 402
modifier (determiner, numeral or adjective of any class of direct or indirect
modification) is rather free, not always related to discourse features:
(1)
a.
b.
(2)
a.
b.
hunc populum (Cic. orat. 2,225) / populum hunc (Plaut. Pseud.
204) ‘this people’
Dem N (dominant) / N Dem (solidly attested)
nostra amicitia (Sall. Iug. 102,7) / amicitia nostra (Cic. fam. 3,8,6)
‘our friendship’
Poss N / N Poss (neither dominant)
Punico bello (Liv. 23,13,3) / bello Punico (Cic. agr. 1,20) ‘Punic
war’
Rel.A N (solidly attested) / N Rel.A (dominant)
armati homines (Cic. Sest. 34) / homines armatos (Cic. Sest. 127)
‘armed men’
indir.mod. N (attested) / N indir.mod (dominant)
The possibilities become more restricted in complex NE, where the noun
is modified by two elements.
a) Postnominal adjectives may appear in any order with no apparent
discourse motivation: In any case, the hierarchy is respected and it seems
that N(P) movement (5a) can freely vary with roll-up (5b):
(3)
a.
b.
[DP haec libro
[FP2 vetere
libro [FP1 linteo
libro]]]
book.ABL.M.SG. old.ABL.M.SG
linen.ABL.M.SG
(Liv. 10,38,6)
“old linen book”
[AgrP2 [[AgrP1 Ova [FP1 anserina
ova]]
egg.acc.n.pl of-goose.acc.n.pl
[FP2 pilleata
[AgrP1 ova anserina]]
with-pilleum.acc.n.pl
(Petr. Sat. 65,2)
“goose eggs with pilleum”
b) Indirect modification can be prenominal: as in (4). As in (3) above, it is
difficult to distinguish direct from indirect modification. The relative order
349
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 350 — #362
AG 13: Adjective order: eory and experiment
of adjectives appears to be as expected by Cinque (2010):
(4)
a.
b.
AG13
[indMod splendens [dirMod stella candida]]
shining.nom.f.g star.nom.f.g white.nom.f.g
(Plaut. Rud. 3)
“the shining white stars”
[indMod calefacta
[dirMod bubula
urina]]
warmed-up.nom.f.g cow.adj.nom.f.g urine.nom.f.g
(Col. 6,15)
“warmed-up cow urine”
The orders in (3)-(4) show that optionality is limited to orders which respect Cinque’s adjectival hierarchies. This shows that the freedom regards
whether to move N at all, or move it alone, or applying roll-up. This is restricted to the non-phasal (NP-)layers of the NE and never involves DP as
shown by the fact that a demonstrative is mandatorily in the leftmost position of Complex NEs, preceding possessive adjectives and numerals and
only following universal quantifiers.
c) ere is no N(P) or roll-up movement across the demonstrative in
complex NEs: (unlike (1a) above):
(5)
a.
[DP Illam
[PossP meam [FP2 pristinam [AgrP lenitatem]]]]
that.acc.f.g my.acc.f.g former.acc.f.g mildness.acc.f.g
(Cic. Catil. 2,6)
b.
[QP omnes [DP haec
[NumP tres
[AgrP partes
all.nom.f.pl these.nom.f.pl three.nom.f.pl part.nom.f.pl
[FP1 purgationis partes]]]]] (Cic. Rhet. Her. 2,16,24)
of-purging
The remaining few apparent violations of Cinque’s hierarchy are statistically rare and pragmatically marked. They are the only orders that can
be analyzed as internal merge of the AP carrying discourse features to be
checked in the nominal Left Periphery, as proposed by Giusti and Iovino
(to appear):
350
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 351 — #363
AG 13, Raum E 402
(6)
a.
b.
[DP haec
[FP equestrian
[LeftPer equestria
equestrian.acc.n.pl this.acc.n.pl
spolia]]]
(Liv. 8,7,13)
spoil.acc.n.pl
[LeftPer alexandrina
[DP/indMod belvata
Alexandrian.nom.n.pl decorated.nom.n.pl
[FP2 tonsilia
[FP1 alexandrina tappetia]]]]
trimmed.nom.n.pl
carpet.nom.n.pl
4. e proposal: The apparent optionality of N in simple and complex NEs
is accounted for in the projection-proposal Giusti (2009, 2015), according
to which, N is bundled with all its functional features (case, number and
gender in Latin) and is internally merged with each modifier. The iterated
internal merge creates a segmented head scattered along the extended projection and crucially lower than DP. The rich nominal morphology and lack
of articles which characterize Latin result in the projection of a number of
identical N-segments. At spell-out, only one of them is realized. The choice
of which one is quite open.
In simple NEs, the projection is limited to the phasal and the lexical layer.
In these cases the non-phasal layer is missing. This allows for the realization of N in D, leaving the demonstrative in SpecNP, a position in which it
is first merged (Giusti 2015 a. o.).
Note that the Left Peripheral layer cannot host NP or a constituent containing it, opposite to Cinque’s (2005) roll-up movement. This makes it crucially different from the optional positioning discussed in (1)-(4).
Finally, the left periphery will be shown to be the escape hatch for fronted
or scrambled adjectives, which give Latin the apparent resemblance of a
non-configurational language.
351
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 352 — #364
AG 13: Adjective order: eory and experiment
An electro-encephalography study on Dutch-Papiamento
code-switching production
Myrthe Wildeboer
Leiden University
[email protected]
Freitag, 26. 02. 2016, 11:30–12:00, Raum: E 402
AG13
Intra-sentential code-switching in Dutch-Papiamento bilingualism may create a conflict within the determiner phrase, because in Dutch the adjective
precedes the noun, whereas in Papiamento the adjective follows the noun.
The Matrix Language Framework (MLF – Myers-Scotton, 1993) suggests
that the matrix language will provide the grammatical frame and that the
embedded language will supply some content elements. The matrix language will thus determine the word order in a code-switched determiner
phrase. In the case of Dutch-Papiamento intra-sentential code-switching,
the MLF will predict an [adjective-noun] order when the matrix language is
Dutch, and the MLF will predict a [noun-adjective] order when the matrix
language is Papiamento.
The MLF predictions were tested by using an advanced psycholinguistic
method, namely electro-encephalography (EEG). The integration of a psycholinguistic method in a code-switching experiment is an innovative way
of testing the predictions of a theoretical model. In this study, an EEG signal was recorded while Dutch-Papiamento bilingual speakers conducted a
modified picture naming task. The conditions were analysed by looking at
naming latencies and by looking at the part of the EEG signal following
target presentation.
Based on results of previous picture naming tasks (Christoffels, Firk &
Schiller, 2007; Rodriguez-Fornells, Van Der Lugt, Rotte, Britti, Heinze &
Münte, 2005; Misra, Guo, Bobb & Kroll, 2012), I expected slower naming
latencies and a more negative waveform for the conditions that violate the
predictions of the MLF. The expected slower naming latencies were observed in two MLF- conditions: Papiamento adjective followed by a Dutch
noun (Papiamento matrix language) and Papiamento noun followed by a
Dutch adjective (Dutch matrix language). The expected negative waveform
352
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 353 — #365
AG 13, Raum E 402
was observed in only one MLF- condition: Papiamento adjective followed
by a Dutch noun (Papiamento matrix language). Furthermore, a P300 and a
late positive component seem to be elicited in code-switching production.
The amplitude of the P300 peak was higher in the conditions that contain
a violation of the MLF, which could be explained by the higher complexity of the MLF- conditions. The occurrence of the P300 could be explained
in terms of the context-updating theory (Donchin, 1981; Donchin & Coles,
1988) or the neural inhibition theory (Polich, 2007). On the whole, the results do not provide conclusive support for the predictions of the MLF.
Adjective order in the Cimbrian DP
AG13
Claudia Turolla1 , Andrea Padovan2 & Ermenegildo Bidese1
1
University of Trento, 2 University of Verona
[email protected], [email protected], [email protected]
Freitag, 26. 02. 2016, 12:00–12:30, Raum: E 402
Cimbrian is a German-based minority language currently spoken only by
the inhabitants of Luserna, in the province of Trento (Italy).
In our talk we present the results of investigation carried out with a
group of Cimbrian native speakers. The main topic of our research is the
syntax of adjectives (see also Alber, Rabanus & Tomaselli 2012 for the Cimbrian of Giazza), focusing on:
(i) the position of one adjective (A) with respect to the noun (N);
(ii) the position of two or more adjectives with respect to the noun.
As regards (i), all informants put the A in prenominal position (1):
(1)
Di Maria lebet in a roat-n
haus ka dar Tetsch.
The Mary lives in a red-DAT;N house at the Tezze.
‘Mary lives in a red house at the Tezze’.
If two adjectives occur, the reciprocal position of A and N shows a great
deal of variation. As a matter of fact, both modifiers can appear together
353
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 354 — #366
AG 13: Adjective order: eory and experiment
prenominally (2) or postnominally (3), or separately, the first one showing
up before and the second one after the noun (4):
AG13
(2)
Disar taütsch-ar
plabe-∅
auto iz naüge.
This German-NOM;M blu-NOM;M car is new.
‘This blu German car is new’
(3)
Di Maria lebet in a haus khlumma un roat ka dar Tetsch.
The Mary lives in a house small
and red at the Tezze.
(4)
Disar taütsch-ar
auto plabe iz naüge.
This German-NOM;M car blu is new.
While the order with one adjective is clearly connected to the prototypical Germanic position (Cinque 2010), the variation provided by two modifiers raises a theoretical question: are these phenomena to be traced back
to independent internal evolution that might have been speeded up by the
contact with the surrounding Romance varieties (see Guardiano’s 2014 resetting of parameters in language contact) or are they directly induced by
the contact which has caused a “structural reorganization” (Savoia 2008) of
the Cimbrian syntax?
In our talk we investigate both hypotheses, although we capitalize on the
idea that the syntax of adjectives in Cimbrian is likely to follow an internal
path of development that makes more positions available for adjectives
w.r.t. Standard German.
Adjective ordering is not just semantics: A language
contact perspective
Fryni Panayidou
Queen Mary University of London
[email protected]
Freitag, 26. 02. 2016, 12:30–13:00, Raum: E 402
Sproat and Shih (1991), among others, argue that adjectives which refer
to absolute properties (e.g. colour, nationality) are closer to the noun than
354
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 355 — #367
AG 13, Raum E 402
adjectives which refer to relative properties (e.g. quality, size). This talk
argues that adjective ordering patterns cannot be explained by simply appealing to semantic constraints such as absoluteness. Supporting evidence
comes from Cypriot Maronite Arabic (CMA), an endangered language in
heavy contact with Greek, in which colour adjectives vary as to whether
they will precede or follow nationality adjectives:
(1)
a.
th avli
l-italiko l-aχmar
table.def the-Italian the-red
‘the red Italian table’
(N < Nationality < Colour)
b.
th avli
li-prasino l-italiko
table.def the-green the-Italian
‘the green Italian table’
(N < Colour ? Nationality)
AG13
Another puzzling property of CMA colour adjectives is that some are restricted to a postnominal position, while others are free to surface either
pre- or post-nominally:
(2)
a.
(*l-aχmar) th avli
l-aχmar
the-red table.def the-red
‘the red table’
b.
(li-prasino) th avli
li-prasino
the-green table.def the-green
‘the green table’
The above examples appear, at first blush, to be problematic for the assumption that adjectives of the same semantic class behave uniformly. I
argue that the variation observed in CMA is systematic and is only captured under a strict adjectival hierarchy:
(3)
Quality > Size > Shape > Colour > Nationality > N
(Cinque 1994)
355
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 356 — #368
AG 13: Adjective order: eory and experiment
AG13
The analysis will reveal that, in CMA, there is a correlation between the distribution of colour adjectives and their morphology. Colour adjectives with
a native Arabic nonconcatenative root (‘red’, ‘black’ and ‘white’) are always
post-nominal and surface in the mirror image order of (3), while colour adjectives borrowed from Greek have retained their concatenative morphology and always follow the order in (3), both pre- and post-nominally.
The discussion will point to three theoretical conclusions: a) variations of
the order can only be captured in a restricted grammar in which a universal
adjective hierarchy exists, b) the underlying hierarchy could be controlled
by semantic constraints, but adjective ordering is far more complicated
than that, and c) roots must have some (morpho-)phonological content,
as different types of roots in CMA are associated with different types of
movement.
Greek noun-adjective ordering revisited
Melita Stavrou
Aristotle University of Thessaloniki
[email protected]
Freitag, 26. 02. 2016, 13:00–13:30, Raum: E 402
First, a survey of the basic accounts of noun-adjective ordering variations
in modern Greek (G) will be given and, second, a novel account will be
proposed whereby A<N and N<A are both possible in both definite and
indefinite noun phrases (with the meaning differences also found inside
the Romance group); the obligatory presence of the definite article will be
attributed to formal reasons, not to do with meaning.
Noun-adjective orderings in Greek have been a lively issue of discussion
ever since the eighties. The distribution of articulated postnominal adjectives in a definite DP was reduced to some kind of ‘appositional’ structure,
whereby the ‘definite’ adjective was held to parallel a definite nominal in an
equational clausal structure, as exemplified in (1)-(2) (Horrocks & Stavrou
1986,1987):
356
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 357 — #369
AG 13, Raum E 402
(1)
a.
b.
o Andreas o politikos [the Andrew the politician]
o Andreas ine (o) politikos [the Andrew is (the politician)]
(2)
a.
to vilio to megalo [the book the big]
(Also: to megalo to vivlio (the big the book )
to vivlio ine (to) megalo [the book is (the) big] (the book is the
big one/the book is big)
b.
This line of thought survived in more recent accounts claiming (with minor
differences amongst them) that definiteness spread (or polydefiniteness)(2a)-encodes a close appositional structure involving noun ellipsis-(2b) (Marinis & Panagiotidis 2011, Lekakou & Szendroi 2012). But based on a number
of significant differences between noun phrases like those in (2a) and similar ones like that illustrated in (3) below, the claim was made by Alexiadou
et al 2007 that (2a) should not be identified as an instance of attributive
modification but as a basically different nominal construction.
(3)
to megalo vivlio the big book simple modification (A<N) cf. (2a) with
a polydefinite.
Alexiadou et al, hypothesize that the interpretation of adjectives in a polydefinite DP parallels one of the two sets of interpretation of postnominal
adjectives in Romance (also Cinque 2010): they are interpreted as indirect
modifiers, viz. intersectively, restrictively and as stage-level modifiers. This
idea is further elaborated in Stavrou 2012. Not only adjectives in polydefinite DPs but also postnominal adjectives in indefinite DPs are interpreted
that way. Here it is proposed that G is like Romance as far as postnominal adjectives are concerned; they are equally possible in all types of noun
phrases and they obey the restrictions that postnominal adjectives in Romance obey. The definite article in front of the adjective does not contribute to the interpretation of the nominal phrase but is the spell out of a
Pred head; it values the phi- and, crucially, case features of the postnominal
adjective inside the PredP.
357
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 358 — #370
AG 13: Adjective order: eory and experiment
Preferences in adjective order: Hierarchical and semantic
approaches reconciled
Eva Wittenberg1 , Andreas Trotzke2 , Emily Morgan1 & Roger Levy1
1
UC San Diego, 2 Universität Konstanz
[email protected], [email protected], [email protected],
[email protected]
Freitag, 26. 02. 2016, 13:30–14:00, Raum: E 402
AG13
We investigate word order in combinations of English ie/hape adjectives with colo adjectives, such as big blue and circular blue. In the literature, on the one hand, finegrained syntactic hierarchies postulate the ordering ie > hape > colo and thereby predict that both ie and hape adjectives preferably precede colo adjectives (e.g., Scott 2002; Laenzlinger
2005). A rivaling approach states that fine-grained ordering restrictions are
less relevant and claims that only a broader semantic difference in terms of
sectivity matters for adjective order (e.g., Truswell 2009). Other accounts
view adjective order as the product of broader cognition, such as accessibility (operationalized as frequency, cf. Wulff 2003).
To arbitrate between hierarchical, semantic, and cognitive accounts of
adjective order, we collected a corpus of adjective combinations by choosing 53 ie/hape adjectives and 43 colo adjectives, and obtained bigram
counts for every pair of adjectives in both possible orders from the Google
n-grams corpus. Our choice of colo adjectives takes into account that
some colo adjectives denote subsective properties (e.g., bright in bright
sunset vs. bright jeans). We predict ordering preferences with a mixedeffects regression model. Predictors were the difference in logarithmic frequency between adjectives, sectivity (hand-coded), and semantic categories
(ie/hape vs colo). We find significant effects of all predictors: There
is a preference for intersective adjectives to be closer to the noun, a preference for color adjectives to be closer to the noun, and more frequent
adjectives tend to be produced first. The sectivity effect is larger than the
semantic category effect, which is larger than the frequency effect, as indicated by magnitude of change in loglikelihood when one factor at a time
358
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 359 — #371
AG 13, Raum E 402
is removed from the model. No two- or three-way interactions reach significance.
Our results show that both hierarchically determined structural constraints and broader semantic distinctions, as well as cognitive factors such
as frequency, have independent roles to play in determining adjective ordering preferences. We discuss how such a multitude of factors needs to be
accounted for in a theoretical model of the faculty of language.
359
AG13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 360 — #372
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 361 — #373
Teil V.
Sektionenprogramm
361
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 362 — #374
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 363 — #375
Postersession der Sektion Computerlinguistik
Die Postersession findet im Foyer auf der Ebene A5 statt.
Poster 1
Methoden zur Erstellung einer lexikalisch-semantischen
Ressource für die Meinungsinferenz
Jasper Brandes1 & Josef Ruppenhofer1
1
Universität Hildesheim
[email protected], [email protected]
Nachdem sich die Forschung im Bereich der Sentimentanalyse in den letzten 15 Jahren vor allem mit expliziten Meinungen befasst hat, ist in letzter
Zeit das Interesse an der sogenannten Meinungsinferenz, also dem automatischen Schließen von impliziten Meinungen, gewachsen. Die Arbeiten
von Anand und Reschke (2010) zur Ereignisevaluation sowie von Choi und
Wiebe (2014) zur Erstellung von +/-EffectWordNet – einer Ressource basierend auf WordNet, welche Effekt-Informationen auf Synset-Ebene enthält
– sind hierbei besonders hervorzuheben.
In (1) ist zu sehen, wie implizite Meinungen in der Interaktion von expliziten Meinungen (fett) und Effekten (teletype) aufgedeckt werden können: Tim hat eine implizit negative Haltung gegen Bayern.
(1)
Tim freut sich, dass Bayern verloren hat.
Auf unserem Poster diskutieren wir unterschiedliche Methoden zur Erstellung einer deutschsprachigen lexikalisch-semantischen Ressource ür
die Meinungsinferenz. Hierbei gibt es zwei grundlegende Ansätze: i) Abbildung von +/-EffectWordNet auf ein deutsches Wortnetz wie GermaNet
oder dem deutschen Teil von BabelNet (Navigli und Ponzetto, 2010) oder ii)
363
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 364 — #376
Postersession der Sektion Computerlinguistik
der automatischen Annotation von GermaNet-Synsets mittels eines Graphen-basierten Ansatzes nach Choi und Wiebe (2014). Wir evaluieren diese
Methoden mittels eines von zwei Annotatoren nach erweiterten Guidelines (Ruppenhofer und Brandes, 2015) annotierten Goldstandards von über
1600 GermaNet-Synsets.
Literatur: • Anand, Pranav und Reschke, Kevin (2010): Verb classes as evaluativity functor
classes. In Proceedings of Verb 2010, Seiten 98–103. • Choi, Yoonjung und Wiebe, Janyce (2014):
+/-EffectWordNet: Sense-level Lexicon Acquisition for Opinion Inference. In Proceedings of
the 2014 Conference on Empirical Methoods in Natural Language Processing (EMNLP), Seiten
CL
Poster
1181–1191, Doha, Qatar. • Navigli, Roberto und Ponzetto, Simone Paolo (2010): BabelNet:
Building a very large multilingual semantic network. In Proceedings of the 48th annual meeting of the association for computational linguistics, Seiten 216–225. • Ruppenhofer, Josef und
Brandes, Jasper (2015): Extending effect annotation with lexical decomposition. In Proceedings of 6th workshop on computational approaches to subjectivity, sentiment and social media
analysis (WASSA 2015), Seiten 67–76.
Poster 2
Automatic detection of German senitive structures with
GI-Tutor
Alessia Battisti1 , Rodolfo Delmonte1 & Peter Paschke1
Department of Linguistic Studies & Department of Computer Science,
Università Ca’ Foscari, Venice, Italy
1
[email protected], [email protected], [email protected]
The prediction of German genitive endings in NLP is currently a controversial topic [1, 2]. Recently, we developed GI-Tutor [3], a computer system
designed for Italian native speakers learning German that operates checking an input text and giving learners suitable exercises and feedbacks. The
rule-based system works with a context free grammar, therefore it is customized for the analysis of a large number of sentences with different sentences and phrase structures.
364
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 365 — #377
Postersession der Sektion Computerlinguistik
This contribution focuses on the identification and analysis of German
genitive structures both in well-formed and ill-formed sentences by GITutor. As many learners mistakes are due to the interference with Italian,
we included the prepositional phrase ‘von+dative’ denoting relationship
and possession. In order to analyze these structures, we properly modified
three modules: first, the tagger for the correct word labeling and disambiguation; second, the transition networks in the parser to implement a
specific phrase; third, the checker for the analysis of this new constituent.
Using 1262 sentences of training corpus, we manually extracted 18 genitive phrases of masculine and neutral nouns, 31 genitive phrases of feminine nouns and 13 prepositional phrases (von+dative) and we analyzed
them with GI-Tutor. Preliminary results of this study show that GI-Tutor
can perform an automatic detection of the majority of the grammatical
and agrammatical structures of German. Specifically, the system identifies
properly the genitive, despite the ambiguities this structure can present.
The overall performance of the system was evaluated by comparing it to
ParZu [4] and to Stanford Parser for German [5].
To the best of our knowledge, this is the first study in Italy which explores the relationship between parser performance and the automatic detection of genitive structures in the context of foreign language acquisition.
References: • [1] Schneider, R. (2014). GenitivDB – a Corpus-Generated Database for German Genitive Classification. In: Proceedings of the Ninth International Conference on Language
Resources and Evaluation. • [2] Genitive DB www.hypermedia.ids-mannheim.de/call/public/
korpus.genitivdb • [3] Delmonte, R., Battisti, A. (2014). GI-TUTOR: Grammar-Checking for
(Italian) Students of German. In: 1st Workshop on Language Teaching, Learning and Technology,
by ISCA, Leipzig. www.docs.google.com/viewer?a=v/&pid=sites/&srcid=ZGVmYXVsdGRvbW
FpbnxsMXRlYW NoaW5nYW5kdGVjaG5vbG9neXxneDo2ODRjMjgwYTUzZjA5NTM
• [4] www.kitt.ifi.uzh.ch/kitt/parzu/ • [5] www.nlp.stanford.edu/software/lex-parser.shtml
365
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 366 — #378
Postersession der Sektion Computerlinguistik
Poster 3
Focus annotation by non-experts: Exploring what the
crowd can do
Kordula De Kuthy1 , Ramon Ziai1 , Detmar Meurers1
Universität Tübingen
1
[email protected], [email protected], [email protected]
CL
Poster
We present a focus annotation study in which we set out to explore how
broader samples of authentic data can be successfully annotated using untrained annotators available through crowd-sourcing.
While previous of approaches trying to annotate focus in authentic data
had difficulties obtaining realiable annotations (e.g. Ritz et al., 2008), Ziai &
Meurers (2014) showed that focus annotation is feasible if provides explicit
questions and explicitly takes them into account in an incremental annotation process. However, manual annotation of focus by experts is still timeconsuming, both for annotator training and the annotation itself. We thus
conducted two crowd-sourcing experiments in order to explore whether
the more time-efficient focus annotation by a large but untrained crowd of
annotators provides comparable results to an expert annotation. Two data
sets were annotated by non-experts recruited via the crowd-sourcing platform Crowdflower, one consisting of 40 Q/A pairs from the Questionnaire
for Information Structure (QUIS, Skopeteas et al., 2006), the other consisting of 1087 Q/A pairs from the Corpus of Reading Comprehension Exercises in German (CREG-1032, Ziai & Meurers, 2014). As the gold standard
to which the focus annotations produced by the crowd workers were compared we used manual annotations by two expert annotators. In order to
calculate the percentage agreement between the gold standard annotation
and the crowd-sourced annotation, we calculated all possibilities of combining up to 10 crowd workers into one „virtual“ annotator using majority
voting on individual word judgments. The overall percentage agreement
(PA) between a gold annotator and the crowd annotators reached of 79% for
the CREG-1032 data (compared to 87% between the two expert annotators).
Further results of our study show: a) the PA between crowd workers and an
expert reaches expert level agreement for specific cases (who-, when-, and
366
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 367 — #379
Postersession der Sektion Computerlinguistik
where-questions), and b) the agreement can reach expert level when the
annotations of a larger number of crowd workers is taken into account. Interesting patterns emerged with respect to learner-data-specific properties
of the CREG-1032 data set: correct student answers were annotated with
a higher PA than the incorrect ones by the expert annotators, whereas for
the incorrect student answers four or more crowd workers compared to an
expert reached a higher PA than the two expert annotators.
Summing up, our study on crowd-sourcing focus annotation shows that
this type of non-expert annotation is promising a) for exploring the impact
of different subtypes of data and instructions on the annotation of authentic data and b) for the large-scale annotation of some types of data. Further
research needs to explore how to improve the focus annotation for those
types of data where the non-expert annotations do not reach the level of
the expert annotations.
References: • Ritz, Julia, Stefanie Dipper & Michael Götze. 2008. Annotation of information
structure: An evaluation across different types of texts. In Proceedings of the 6th international
conference on language resources and evaluation, 2137–2142. Marrakech, Morocco. • Skopeteas,
Stavros, Ines Fiedler, Samantha Hellmuth, Anne Schwarz, Ruben Stoel, Gisbert Fanselow,
Caroline Féry & Manfred Krifka. 2006. estionnaire on information structure (QUIS): reference manual, vol. 4 Interdisciplinary studies on information structure (ISIS). Universitätsverlag
Potsdam. • Ziai, Ramon & Detmar Meurers. 2014. Focus annotation in reading comprehension data. In Proceedings of the 8th Linguistic Annotation Workshop (LAW VIII, 2014), 159–168.
COLING Dublin, Ireland: Association for Computational Linguistics.
Poster 4
A workflow for creating, analysing, and storing multi-layer
corpora: Pepper, Atomic, ANNIS and LAUDATIO
Stephan Druskat1 , Thomas Krause2 , Carolin Odebrecht2 & Florian Zipser2
Friedrich Schiller University Jena, 2 Humboldt-Universität zu Berlin
1
[email protected]
The creation and analysis of corpus linguistic resources can be a costly and
error-prone process. Apart from the complexity of the annotation process
367
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 368 — #380
Postersession der Sektion Computerlinguistik
CL
Poster
itself, there are larger technical obstacles to be overcome. Single tools have
to be combined in a common workflow, and different formats taken into
account. This poster presents a family of well-aligned open source tools
which support the conversion, annotation, and analysis of linguistic corpora, as well as securing their long-term accessibility, in a complete workflow.
The interoperability of these tools is guaranteed by the use of a common data model – Salt (Zipser & Romary, 2010) – which, among other
things, is used as an intermediate model for the conversion framework
Pepper (Zipser et al., 2011). With Pepper, many linguistic formats can be
converted into each other, thereby allowing existing data to be included in
the workflow. The support for a multitude of linguistic formats allows for
the replacement of single components as well as the integration of further
tools into the workflow presented here.
The annotation of corpora is carried out in Atomic (Druskat et al., 2014),
an extensible annotation platform. Atomic also utilizes Salt – in this case
as its concrete data model – and thus allows for theory-neutral annotation
which is independent of tagsets and annotation types. By embedding Pepper, it supports a wide variety of source formats for further annotation, as
well as target formats for export. Additionally, its plugin-based architecture makes it possible to easily extend the software, e.g., with additional
editors, data views, or processing components. For a new annotation type,
for instance, a dedicated editor can thus be created and integrated.
At any point in the annotation process, the annotated data can be transferred to the search and visualization tool ANNIS (Krause & Zeldes, 2014)
for visualisation and analysis. Conclusions from the analysis can then, for
example, also flow back into the annotation process.
When a corpus is ready for publication, it can be released in different
formats to a public repository – in the case of historical text corpora, for
example, the LAUDATIO-Repository (Odebrecht et al., 2015). Third parties
can then download, reference and re-use the data.
References: • Druskat, Stephan, Lennart Bierkandt, Volker Gast, Christoph Rzymski & Florian
Zipser. 2014. Atomic: an open-source software platform for multi-level corpus annotation. In
Josef Ruppert & Gertrud Faaß (eds.), Proceedings of the 12th Konferenz zur Verarbeitung natür-
368
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 369 — #381
Postersession der Sektion Computerlinguistik
licher Sprache (KONVENS 2014), 228–234. • Krause, Thomas & Amir Zeldes. 2014. ANNIS3:
A new architecture for generic corpus query and visualization. Digital Scholarship in the Humanities. http://dx.doi.org/10.1093/llc/fqu057. • Odebrecht, Carolin, Thomas Krause & Anke
Lüdeling. 2015. Austausch von historischen Texten verschiedener Sprachen uber das LAUDATIO-Repository. Poster presented at 37. Jahrestagung der Deutschen Gesellscha ür Sprachwissenscha, 5 March, Leipzig University, Leipzig, Germany. • Zipser, Florian & Laurent Romary. 2010. A model oriented approach to the mapping of annotation formats using standards. In Proceedings of the Workshop on Language Resource and Language Technology Standards, Seventh International Conference on Language Resources and Evaluation (LREC 2010),
Valletta, Malta. • Zipser, Florian, Amir Zeldes, Julia Ritz, Laurent Romary & Ulf Leser. 2011.
Pepper: Handling a multiverse of formats. Poster presented at 33. Jahrestagung der Deutschen
Gesellscha ür Sprachwissenscha, 24 February, Göttingen University, Göttingen, Germany.
Poster 5
JuReko – Das juristische Referenzkorpus
Isabell Gauer1 , Prof. Dr. Friedemann Vogel1 , Dr. Hanjo Hamann2 &
Yinchun Bai2
1
Albert-Ludwigs-Universität Freiburg, 2 Max-Planck-Institut Bonn
[email protected]
Linguisten und Juristen haben eine Gemeinsamkeit: Beide arbeiten an und
mit Sprache. Im Recht werden sprachlichen Gebrauchsmustern zusätzlich
institutionelle Bedeutungen beigemessen. Trotzdem sind Rechtstexte „nicht
Behälter der Rechtsnorm, sondern Durchzugsgebiet konkurrierender Interpretationen“ (Müller et al. 1997: 19). Die Rechtssprache als Fachsprache und das Verhältnis von Sprache und Recht beschäftigen also Linguisten und Juristen gleichermaßen. In der Rechtslinguistik sowie in der empirischen Rechtsforschung steigt das Bewusstsein, dass viele Forschungsfragen nur mittels großer Textsammlungen beantwortet werden können.
Im interdisziplinären Akademieprojekt „Vom corpus iuris zu den corpora iurum Konzeption und Erschließung eines juristischen Referenzkorpus
(JuReko)“ (vgl. Vogel & Hamann 2015) soll dieser Bedarf nun adressiert
369
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 370 — #382
Postersession der Sektion Computerlinguistik
CL
Poster
werden. Dazu wird zunächst ein Korpus mit deutschen juristischen Texten
zusammengestellt und aufbereitet. In einem zweiten Schritt sollen erste
Studien durchgeührt und so Analysemethoden erprobt werden. Weiterhin soll das deutsche Kernkorpus mit britischen Case-Law-Texten erweitert werden.
JuReko wurde konzipiert, um quantitative aber auch qualitative Datenanalysen durchzuühren. Ein besonderes Augenmerk bei der Textaufbereitung und -auszeichnung liegt deshalb auf Genauigkeit und Fehlerfreiheit.
Bei JuReko handelt es sich um ein Spezialkorpus mit einer Zielgröße von
einer Milliarde Token. Es werden nur Textsorten mit juristischer Ausrichtung aufgenommen, das heißt Aufsätze aus juristischen Fachzeitschriften,
Entscheidungstexte und Normtexte. Die Textdaten werden zunächst im
html-Format gewonnen und anschließend in mehreren Konvertierungschritten TEI P5 konform kodiert. Daür kommen xsl-Transformationen zum
Einsatz, die auf die unterschiedlichen Webseitenstrukturen angepasst werden. Im Anschluss werden die Texte mit Part-of-Speech und weiteren Annotationen und Metadaten angereichert, wobei die speziellen Anforderungen einer rechtslinguistischen Textanalyse und –verarbeitung im Vordergrund stehen.
Literatur: • Müller, Friedrich & Christensen, Ralph & Sokolowski, Michael. 1997. Rechtstext
und Textarbeit. Berlin: Duncker & Humblot. • Vogel, Friedemann & Hamann, Hanjo. 2015.
Vom corpus iuris zu den corpora iurum – Konzeption und Erschließung eines juristischen
Referenzkorpus (JuReko). Jahrbuch der Heidelberger Akademie der Wissenschaen ür 2014.
Heidelberg: Winter.
370
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 371 — #383
Postersession der Sektion Computerlinguistik
Poster 6
Features of compositionality in English and German
noun-noun compounds
Anna Hätty1 , Sabine Schulte im Walde1 , Stefan Bott1 &
Nana Khvtisavrishvili1
1
Institut ür Maschinelle Sprachverarbeitung, Universität Stuttgart
[email protected]
Noun-noun compounds are complex words with two simplex nouns as constituents. A compound may have different degrees of compositionality regardings its constituents. For example, the German compound Lederhose
(‘leather trousers’) is highly compositional, because its meaning can be
obtained by combining the meanings of its constituent words. The compound Sündenbock (‘sin buck’, meaning scapegoat), in contrast, is rather
non-compositional, because there is no obvious synchronic relation between its meaning and the meanings of its constituents.
A common NLP approach to compute the degree of compositionality
are vector space models (Reddy et al., 2011; Schulte im Walde et al., 2013;
among others): Context words for the compounds and constituents are extracted from large corpora, and the similarity between the compound and
the constituent context vectors predicts the degree of compositionality of
the compound. Our work considers German and English noun-noun compounds, and as the basis for our models, the German and English COWCorpora (cf. www.corporafromtheweb.org) are used.
As compound datasets, we rely on three existing resources: 90 English
noun-noun compounds introduced by Reddy et al. (2011), a subset of the
1,443 English compounds introduced by Ó Séaghdha (2007), and 244 German noun-noun compounds introduced by Schulte im Walde et al. (2013).
In addition, we created a new dataset of German compounds comprising
1,208 noun-noun compounds. All compound datasets have been extended
to include human compositionality ratings, and semantic relations between
modifiers and heads (the first and second constituents of a compound).
Our specific interest is to determine how different properties of the compounds influence their degree of compositionality. One factor is the pro-
371
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 372 — #384
Postersession der Sektion Computerlinguistik
ductivity of a constituent (i.e., the number of words a constituent can be
combined with in a compound, either in the modifier or head position),
e.g. Brunnenwasser ‘well water’ has a highly productive modifier. Other
factors we explore are corpus frequency, ambiguity (e.g. Porzellanrohling
‘porcelain blank’, where the head can stand for brute or blank) and semantic relations (e.g. Rechtsgeschichte ‘legal history’ with the relation ABOUT).
References: • Diarmuid Ó Séaghdha. 2007. Designing and evaluating a semantic annotation
scheme for compound nouns. In Proceedings of Corpus Linguistics. • Silva Reddy, Diana McCarthy & Suresh Manandhar. 2011. An empirical study on compositionality in compound
CL
Poster
nouns. In Proceedings of IJCNLP. • Sabine Schulte im Walde, Stefan Müller & Stephen Roller.
2013. Exploring vector space models to predict the compositionality of German noun-noun
compounds. In Proceedings of *SEM.
Poster 7
Von Audiodaten und ihren Transkripten zur linguistischen
Analyse – die Aufbereitung von Radio-Interviews
Fritz Kliche1,2 , Kerstin Eckart1 , Katrin Schweitzer1 , Markus Gärtner1 &
Ulrich Heid2
1
Universität Stuttgart, 2 Universität Hildesheim
[email protected]
Wir stellen eine Verarbeitungskette vor, mit der Audiodaten ür die linguistische Analyse verügbar gemacht werden können. Die Schritte umfassen
die Überührung in ein linguistisch verarbeitbares Format, die Aufbereitung der Transkripte sowie die Alignierung von Text und Ton.
Als Primärdaten dienen Audiodateien von Radio-Interviews und die zugehörigen Transkripte. Interviews liegen in Bezug auf ihren Sprachstil zwischen vorgelesenen Texten wie z.B. Radio-Nachrichten (Eckart et al., 2012)
und spontanen Dialogen (Schweitzer et al., 2015): Die Redebeiträge sind
klar voneinander getrennt, und die Sequenzen der befragten Personen weisen aufgrund verschiedener Hintergründe (Politik, Wirtschaft, Vereinswesen) verschiedene Stufen konzeptueller Mündlichkeit auf. Annotationen,
372
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 373 — #385
Postersession der Sektion Computerlinguistik
die sich auf die Transkripte beziehen, werden auf die Zeichenebene der textuellen Primärdaten abgebildet. Annotationen, die sich auf die Audiodaten
beziehen, werden analog dazu auf Zeitstempel abgebildet. Dies ermöglicht
eine gemeinsame Abfrage von verschiedenen linguistischen Annotationen.
Über Werkzeuge des Instituts ür Maschinelle Sprachverarbeitung (IMS,
Uni Stuttgart) können wir die Interviewdaten aufbereiten und ür die phonetische Analyse zugänglich machen. Am Beginn der Pipeline verwenden wir die Explorationswerkbank aus dem Digital-Humanities-Projekt eIdentity. Die Werkbank ist eine Web-Anwendung ür die Aufbereitung von
semistrukturierten Textdaten (Kliche et al., 2014). Wir überühren mit ihr
die Transkripte im PDF-Format in Dateien im TSV-Format, in denen die
Fragen der Interviewer, die Antworten der Gesprächspartner sowie Metadaten (Sendezeit, Titel der Sendung, Redaktion) abgebildet werden. Am Ende der Pipeline können die Daten von ICARUS verwendet werden. ICARUS
ist ein Visualisierungswerkzeug ür linguistische Suchanfragen, das auch
multimodale Suchanfragen anbietet, z.B. können phonologische Realisierungen in bestimmten syntaktischen Strukturen gesucht werden (Gärtner
et al., 2015). In ICARUS können u.a. auch Label-Files aus Festival importiert
werden.
Wir zeigen die Verarbeitungskette und die integrierten Werkzeuge in
ihrer Anwendung. Wir geben Beispiele ür Prosodie-Analysen und zeigen
die Schritte, mit denen die Daten ür existierende linguistische Analysewerkzeuge zugänglich gemacht werden können.
Literatur: • Eckart, Kerstin; Riester, Arndt und Schweitzer, Katrin (2012). A discourse information radio news database for linguistic analysis. In: Christian Chiarcos, Sebastian Nordhoff und Sebastian Hellmann (Hrsg.), Linked Data in Linguistics. Representing and Connecting
Language Data and Language Metadata. Heidelberg, Springer. S. 65-75. • Gärtner, Markus;
Schweitzer, Katrin; Eckart, Kerstin und Kuhn, Jonas (2015). Multi-modal visualization and
search for text and prosody annotations. In: Proceedings of the 53rd Annual Meeting of the
Association for Computational Linguistics (ACL 2015): System Demonstrations. Peking, China. Juli 2015. • Kliche, Fritz; Blessing, André; Heid, Ulrich und Sonntag, Jonathan (2014). The
eIdentity Text ExplorationWorkbench. In: Proceedings of the 9th International Conference on
Language Resources and Evaluation (LREC’14), Reykjavik, Island. European Language Resources Association (ELRA). Mai 2015. • Schweitzer, Antje; Lewandowski, Natalie; Duran, Daniel
373
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 374 — #386
Postersession der Sektion Computerlinguistik
und Dogil, Grzegorz (2015). Attention, please! – Expanding the GECO database. In: Proceedings of the 18th International Congress of Phonetic Sciences (ICPHS 2015), Glasgow, GB. August
2015.
Poster 8
TraMOOC: Translation for Massive Open Online Courses
Valia Kordoni1
1
Humboldt Universität zu Berlin, Germany
[email protected]
CL
Poster
Funding agency: The European Commission
Funding call identification: H2020-ICT-2014-1 -ICT-17-2014
Type of project: Innovation Action
Project ID number: 644333
http://www.tramooc.eu
List of partners:
Humboldt Universität zu Berlin (UBER), Germany (coordinator)
Dublin City University (DCU), Ireland
The University of Edinburgh (UEDIN), UK
Ionian University (IURC), Greece
Stichting Katholieke Universiteit (Radboud University & Radboud UMC),
The Netherlands
EASN Technology Innovation Services BVBA (EASN TIS), Belgium
Deluxe Media Europe Ltd (Deluxe Media Europe Ltd), UK
Stichting Katholieke Universiteit Brabant Universiteit van Tilburg,
The Netherlands
IVERSITY GMBH (iversity.org), Germany
KNOWLEDGE 4 ALL FOUNDATION (K4A), UK
Project duration: February 2015 — January 2018
374
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 375 — #387
Postersession der Sektion Computerlinguistik
Summary:
Massive open online courses have been growing rapidly in size and impact.
TraMOOC aims at developing high-quality translation of all types of text
genre included in MOOCs from English into eleven European and BRIC
languages (DE, IT, PT, EL, DU, CS, BG, CR, PL, RU, ZH) that are hard to
translate into and have weak MT support. Phrase-based and syntax-based
SMT models will be developed for addressing language diversity and supporting the language-independent nature of the methodology. For a high
quality, automatic translation approach and for adding value to existing infrastructure, extensive and advanced bootstrapping of new resources will
be performed. An innovative multimodal automatic and human evaluation schema will further ensure translation quality. For human evaluation,
an innovative, strict-access control, time-and cost-efficient crowdsourcing
setup will be used. Translation experts, domain experts and end users will
also be involved. Separate task mining applications will be employed for
implicit translation evaluation: (i) topic detection will be applied to source
and translated texts and the resulting entity lists will be compared, leading to further qualitative and quantitative translation evaluation results;
(ii) sentiment analysis performed on MOOC users’ blog posts will reveal
end user opinion/evaluation regarding translation quality. Results will be
combined into a feedback vector and used to refine parallel data and retrain
translation models towards a more accurate second phase translation output. The project results will be showcased and tested on the Iversity MOOC
platform and on the VideoLectures.net digital video lecture library.
375
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 376 — #388
Postersession der Sektion Computerlinguistik
Poster 9
Alpenwort. Korpus der Zeitschrift des Deutschen und
Österreichischen Alpenvereins - Vorstellung des Projekts
und erster Ergebnisse
Bettina Lar1 & Irina Windhaber1
1
Institut ür Sprachen und Literaturen, Bereich Sprachwissenschaft,
Universität Innsbruck
[email protected], [email protected]
CL
Poster
Das Projekt Alpenwort ist am Institut ür Sprachen und Literaturen, Bereich
Sprachwissenschaft der Universität Innsbruck angesiedelt und wird von
der Initiative go!digital der Österreichischen Akademie der Wissenschaften
geördert. Projektleitung: Claudia Posch, Gerhard Rampl.
Die Zeitschrift des Deutschen und Österreichischen Alpenvereins (ZAV,
Jahrgänge 1872–1998) bildet die Grundlage des Korpusprojekts. Sie ist aufgrund ihrer inhaltlichen Homogenität und des langen Zeitraums über den
sie durchgehend erscheint eine einzigartige Textquelle. Schließlich nimmt
Österreich im Alpenraum sowohl geographisch als auch bei seiner wissenschaftlichen und touristischen Erschließung eine zentrale Stelle ein. Eine
wertvolle Kooperation besteht mit dem Projekt Text+Berg digital an der
Universität Zürich, das die Jahrbücher des Schweizer Alpenclubs (SAC) bereits als vollständig annotiertes Korpus zur Verügung stellt.
Im ersten Projektschritt von „Alpenwort“ wird die ZAV digitalisiert und
anschließend linguistisch annotiert, d.h. tokenisiert, POS-getagged und mit
NER-Annotation angereichert. Das Ergebnis soll als elektronisches Korpus,
den CLARIN-DARIAH-Standards folgend, der Forschungsgemeinschaft zur
Verügung gestellt werden.
Das so entstehende Korpus bildet eine facettenreiche Basis ür Forschungsfragen auf vielen Gebieten. Bisher geplant sind Untersuchungen aus den
Bereichen der Onomastik, der Sentiment-Analyse und der Feministischen
Linguistik.
Im Rahmen der Posterpräsentation werden bisher absolvierte Schritte
auf dem Weg vom Buch zum Korpus erläutert und anhand konkreter Beispiele illustriert.
376
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 377 — #389
Postersession der Sektion Computerlinguistik
Poster 10
Linguistic methodology to help German and French
non-translator users writing bilingual specifications
Claire Lemaire
ILCEA4 (Institut des Langues et Cultures d’Europe, Amérique, Afrique,
Asie et Australie), Université Grenoble Alpes
[email protected]
In the SAP implementation industry, more and more German and French
IT-Engineers are required to write bilingual functional design specifications, because most of the programming is now done by Indians. As the
domain is too specific, and no budget is available for linguistic purposes,
translators are not integrated in the process. Our poster, which describes
a part of our doctoral study, focuses on how to help those writers and programmers with linguistic methods and tools.
A survey is conducted to find out which linguistic tools are used among
professional translators (English, German, French). These tools are mainly
based on translation memories, but also on machine translation, spell checker, electronic dictionaries and comparable corpora.
We propose some of those tools to the IT-Engineers but, above all we
propose to alter the linguistic “raw material”, by focusing on pre-editing
according to 2 criteria: linguistic quality and clarity of the source text, and
fidelity of the translated text, by activating the passive knowledge of the
writer thanks to NLP tools. For the “clarity criteria”, we propose also to
substitute words by graphics and give a graphical template which could
help the communication. We then propose an experiment to compare the
quality of the output texts, before and after pre-editing the input texts with
our linguistic methodology and terminological resources.
References: • Delpech, E., Daille, B., Morin, E., Lemaire, C. (2012). “Extraction of domain-specific bilingual lexicon from comparable corpora: compositional translation and ranking.” COLING 2012, Dec 2012, Mumbai, India. pp.745-762. • Bouillon, P., Gaspar, L., Gerlach, J., Porro,
V., Roturier, J. (2014), ‘Pre-editing by forum users: a Case Study’, in: Proceedings of the 9th Edi-
377
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 378 — #390
Postersession der Sektion Computerlinguistik
tion of the Language Resources and Evaluation Conference (LREC), CNL Workshop, Reykjavik,
Islande.
Poster 11
Die Architektur von CoreGram:
Grammatikimplementierung mit Demo
Stefan Müller1 , Antonio Machicao y Priemer2 & Elodie Winckel1
1
Freie Universität Berlin, 2 Humboldt-Universität zu Berlin
[email protected], [email protected], [email protected]
CL
Poster
Mithilfe von Grammatikimplementierungen kann die Konsistenz von Theorien verifiziert und die Interaktion von sprachlichen Phänomenen besser
beobachtet werden. Diese Ziele werden auch im CoreGram-Projekt verfolgt (Müller, 2015). In diesem Projekt werden Grammatikfragmente verschiedener Sprachen implementiert. Grammatiken (bzw. Grammatikfragmente) ür das Deutsche, Dänische, Persische, Maltesische, Chinesische,
Jiddische, Französische und Spanische sind bereits implementiert worden
und werden kontinuierlich weiter ausgebaut. Alle Grammatiken werden
im Rahmen der Head-driven Phrase Structure Grammar (Pollard und Sag,
1994; Müller, 2013) und deren Semantik in Minimal Recursion Semantics
(Copestake et al., 2005) entwickelt.
Die Beschränkungen und Generalisierungen in der Behandlung von sprachlichen Phänomenen sind in erster Linie sprachspezifisch begründet. Nichtsdestotrotz lassen sich Generalisierungen allgemeinerer Art finden, welche
von mehreren oder auch von allen (bisher implementierten) Sprachen geteilt werden.
In unserer Präsentation sollen zum einen die theoretischen Grundlagen
und das technische Grundgerüst des TRALE-Systems (Meurers et al., 2002)
kurz erläutert und diskutiert werden. Außerdem werden bei der Demo verschiedene Grammatiken vorgeührt und es wird gezeigt, wie sprachübergreifende Generalisierungen erfasst werden.
Literatur: • Copestake, A., D. P. Flickinger, C. Pollard und I. A. Sag (2005). Minimal Recursion
Semantics: An Introduction. Research on Language and Computation 3(4), 281–332. • Meurers,
378
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 379 — #391
Postersession der Sektion Computerlinguistik
W. D., G. Penn und F. Richter (2002). A Web-based Instructional Platform for Constraint-Based Grammar Formalisms and Parsing. In D. Radev und C. Brew (Hg.), Effective Tools and
Methodologies for Teaching NLP and CL, New Brunswick, NJ, S. 18–25. • Müller, S. (2013). Head-Driven Phrase Structure Grammar: Eine Einührung. Tübingen: Stauffenburg. • Müller, S.
(2015). The CoreGram Project: Theoretical Linguistics, Theory Development and Verification.
Journal of Language Modelling 3(1), 1–66. Pollard, C. und I. Sag (1994). Head-Driven Phrase
Structure Grammar. Chicago: University of Chicago Press.
Poster 12
KOLIMO: Aufbau und Annotation eines Korpus der
literarischen Moderne
CL
Markus Paluch1 , Christina Schmidt1 , Benedict Spermoser1 &
J. Berenike Herrmann1
1
Seminar ür Deutsche Philologie, Universität Göttingen
[email protected]
Das vorgeschlagene Poster dokumentiert das Ineinandergreifen literaturwissenschaftlicher und computerlinguistischer Fragestellungen und Arbeitsschritte bei Aufbau und (POS-)Annotation eines Korpus der deutschsprachigen literarischen Moderne (KOLIMO), das in der Folge computerstilistische
Analysen im Rahmen eines Projektes zur Quantitativen Analyse der Literarischen Moderne (Q-LIMO) ermöglichen soll. Obwohl inzwischen viele digitale Ressourcen (wie TextGrid, Deutsches Textarchiv [DTA], GutenbergDE) zur Verügung stehen, ist ein repräsentatives und computerlinguistisch
solide aufbereitetes Korpus von narrativen fiktionalen Erzähltexten der literarischen Epoche der Moderne (ca. 1880 – 1930) bisher nicht vorhanden.
Beim Aufbau des Korpus stehen zwei Kriterien im Vordergrund: (a) eine
näherungsweise Repräsentativität der abgebildeten Epoche, (b) eine hohe
Akkuratheit bzw. Konsistenz bei POS-Annotation, Metadaten und Textmarkup, eingebettet in einen praktikablen Workflow samt XML-Datenbank
(EXist). Durch Metadaten (Autor, Titel, Publikationsdatum, Publikationsort, Gattung, Geschlecht) und linguistische Parameter (zunächst v.a. POS)
379
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 380 — #392
Postersession der Sektion Computerlinguistik
CL
Poster
werden philologische Fragestellungen in präzise und praktikable Kategorien umgewandelt.
KOLIMO (ca. 1.800 Texte) wurde aus TextGrid, DTA und Gutenberg-DE
zusammengestellt, und wird derzeit um einige weitere Werke ergänzt. Besondere Anforderungen sind dabei (a) die Berücksichtigung von kanonischen und populären Werken, und, zu Vergleichszwecken, nichtliterarischer Texte der Epoche. Eine Subsektion von KOLIMO hat den als „prototypisch modern“ geltenden Autoren Franz Kafka als Referenzpunkt; hier
werden Texte kompiliert, die mit hoher Wahrscheinlichkeit durch Kafka
rezipiert wurden, um später vergleichende Stilanalysen zu ermöglichen.
Auf computerlinguistischer Seite (b) ist neben der Vereinheitlichung von
Format (xml-Stripper), Metadaten (händisch/automatisch) und Textmarkup (TEI) die Überprüfung der Sprachmodelle der verügbaren Tagger ausschlaggebend. Es handelt sich überwiegend fiktional-narrative Texte handelt (vom modernen Standarddeutschen abweichender Sprachgebrauch).
Eine manuelle Fehleranalyse des Outputs gängiger POS-Tagger soll ür das
Training eines Taggers verwendet werden. Ziele sind hier eine reliable automatische Annotation des Korpus und die Erzeugung eines Taggers, der
auf Besonderheiten der deutschen Prosa der literarischen Moderne gerade
auch in ihrer internen Variation trainiert ist.
Insgesamt soll KOLIMO nicht nur eine wertvolle neue Ressource einer
interdisziplinären Forschungsgemeinschaft sein, sondern seine Entwicklung stellt in den Einzeldisziplinen jeweils neue Fragen.
Poster 13
Die DBÖ als Ressource - Vergangenheit, Gegenwart und
Zukunft
Melanie Siemund1 , Jack Bowers1 & Barbara Piringer1
lexlab (lexicography laboratory), Austrian Centre for Digital Humanities
(ACDH), Österreichische Akademie der Wissenschaften, Wien
1
[email protected], [email protected], [email protected]
Das Poster „Die DBÖ als Ressource – Vergangenheit, Gegenwart und Zukunft“ informiert über den Inhalt, die Entstehung sowie die Zukunft der
380
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 381 — #393
Postersession der Sektion Computerlinguistik
Datenbank der bairischen Mundarten in Österreich (DBÖ). Diese besteht
hauptsächlich aus einer Belegdatenbank ür Dialektwörter, aber auch aus
einer Quellendatenbank, welche die gesamte Bibliographie beinhaltet, sowie zusätzlichem Material, welches beispielsweise Sammler*innen, Orte
und kulturelle Informationen umfasst. Eingeschlossen ist eine Datenbank
von Pflanzennamen, welche mundartliche Synonyme ür lateinische Pflanzennamen enthält. Des Weiteren existiert eine der Belegdatenbank angegliederte Bilddatenbank, die Abbildungen und Skizzen, die die Sammler*innen zur Bedeutungserklärung angefertigt haben, einbezieht. Die DBÖ im
hiesigen Sinne besteht derzeit genau genommen aus zwei Datenbanken:
der DBÖ, einer TUSTEP-Datenbank, die Daten anhand von rund 200.000
Lemmata auflistet,sowie der dbo@ema, einer MySQL-Datenbank, welche
Daten anhand von rund 50.000 Lemmata sortiert. Die Belege wurden im
Zeitraum 1913 bis 1989 von zahlreichen Sammler*innen in den Gebieten
der (ehemaligen) Habsburgermonarchie, in denen bairische Dialekte gesprochen wurden, erhoben. Dazu zählen Österreich (außer die alemannischsprachige Region Vorarlberg), Südtirol, im Osten und Norden an Österreich angrenzende Gebiete sowie Sprachinseln. Die hierür benutzten
Fragenbögen sind ebenso in der DBÖ enthalten. Zukünftig sollen Schnittstellen ür TEI/XML geschaffen werden, um die Organisation der Metadaten, Konzepte und linguistischen Daten zu verbessern. Ebenso sollen
Schnittstellen ür die Nutzung von Linked Open Data (LOD) geschaffen
werden, um ontologische Ressourcen verwenden zu können und somit die
Darstellung der konzeptionellen und semantischen Informationen zu gewährleisten. Schließlich soll das Poster überblicksartig Aspekte des neuen
Projektes exploreAT! vorstellen.
381
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 382 — #394
Postersession der Sektion Computerlinguistik
Poster 14
Kontextbasiertes morphologisches Parsing
Petra Steiner1
1
Universität Hildesheim
[email protected]
CL
Poster
Für viele computerlinguistische Anwendungen über deutsche Sprachdaten ist die Wortbildungsproduktivität eine Herausforderung: Zum Beispiel
bei der computergestützten Übersetzung, der Terminologie-Extraktion und
dem Information Retrieval ist die morphologische Analyse komplexer Lexeme eine Voraussetzung ür die weitere Verarbeitung und Bedeutungsanalyse sprachlicher Daten.
Bisher beschränken sich Wortanalysetools ür das Deutsche auf die Segmentierung zu flachen morphologischen Strukturen. Um die Bedeutung
komplexer Wortformen erschließen zu können, ist jedoch meist die Kenntnis ihrer hierarchischen Strukturen erforderlich. Da flache Analysen ebenfalls oft ambig sind, ist es sinnvoll, bereits auf der Ebene der Morphe Auswahlverfahren anzuwenden. Für die Zerlegung deutscher Komposita im
Hinblick auf die maschinelle Übersetzung hat Cap (2014) ein solches Auswahlverfahren entwickelt, das auf den Ergebnissen der flachen Analysen
des Morphologie-Tools SMOR (Schmid, Fitschen u. Heid 2004) aufbaut. In
diesem Poster zeigen wir einen quantitativen Ansatz, der ebenfalls die von
SMOR produzierten Analysen verwendet und Teile des Verfahrens von Cap
(2014) aufgreift. Analysiert werden jedoch nicht nur Komposita, sondern
u.a. auch Derivate. Das Verfahren besteht aus drei Schritten, wobei die
Schritte 2 und 3 ür jede Analyseebene der morphologischen Struktur bis
zu einer vollständigen Derivation wiederholt werden. 1. Die Menge der
flachen Strukturen wird mittels verschiedener Gewichtungsmaße eingeschränkt. 2. Die Menge der rein kombinatorischen Analysemöglichkeiten
ür die jeweils höhere Analyseebene wird mittels linguistischer Kriterien
eingeschränkt. 3. Die verbleibenden kombinatorisch möglichen Strukturen
werden mit verschiedenen Gewichtungsmaßen gefiltert.
Als Häufigkeiten ür die Gewichtungsmaße werden verwendet: (a) Die
Frequenz von Morphen als Bestandteil von Lexemen eines Lexikons, (b)
382
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 383 — #395
Postersession der Sektion Computerlinguistik
die Frequenz von Konstituenten als Bestandteil von Lexemen eines Lexikons und (c) die Frequenz von Morphen und Konstituenten in Korpora,
Subkorpora und im Kontext.
Literatur: • Cap, Fabienne. 2014. Morphological processing of compounds for statistical machine translation. Universität Stuttgart [hp://elib.uni-stugart.de/opus/volltexte/2014/9768].
• Schmid, Helmut, Arne Fitschen u. Ulrich Heid. 2004. SMOR: A German computational morphology covering derivation, composition and inflection. Proceedings of the International Conference on Language Resources and Evaluation (LREC). Lisbon, Portugal. 1263-1266. • Steiner,
Petra u. Josef Ruppenhofer. im Druck. Growing trees from morphs: Towards data-driven morphological parsing. Proceedings of the International Conference of the German Society for Computational Linguistics and Language Technology.
Poster 15
Developing a hybrid supertagger: Domain adaptation
methods for low-resource texts
Kyoko Sugisaki
Institute of Computational Linguistics, University of Zurich, Switzerland
[email protected]
We present a supertagger for an automatically annotated linguistic corpus and NLP applications. The addressed problem is that state-of-the-art
parsing methods are biased by the text type, and as a consequence, do not
perform well on domain texts different from the training data (cf. Versley,
2005; Gildea, 2001). The legislative domain, which we are interested in, is
largely out of reach for state-of-the-art parsers.
In this poster, we report on domain adaptation methods that do not
require a large amount of annotated in-domain texts. The presented supertagger automatically annotates topological dependency grammar relations and is hybrid: It combines a rule-based tagger and a statistical tagger
with state-of-the-art parsers. The evaluation showed that our hybrid supertagger achieved a F1 score of 94.22% in label accuracy, outperforming
state-of-the parsers.
383
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 384 — #396
Postersession der Sektion Computerlinguistik
References: • Gildea, D. (2001). Corpus variation and parser performance. In Proceedings of
2001 Conference on Empirical Methods in Natural Language Processing (EMNLP), pages 167–202.
• Versley, Y. (2005). Parser evaluation across text types. In Proceedings of the Fourth Workshop
on Treebanks and Linguistic eories (TLT ).
Poster 16
SaltInfoModule - Automatische Metadatenextraktion und
Dokumentation für Korpora
CL
Poster
Vivian Voigt1 , Florian Zipser1 & Carolin Odebrecht1
Humboldt-Universität zu Berlin
1
[email protected], [email protected], [email protected]
Linguistische Korpora sind vielältig aufbereitet und liegen daher in unterschiedlichen Formaten abhängig von der Art der Analyse und Veröffentlichung vor (z.B. in EXMARaLDA wie Tatian, Petrova et al. 2009 oder in
PAULA wie PCC2, Stede & Neumann 2014 uvm.). Um diese Korpora ür
Dritte zugänglich und nachvollziehbar zu machen, werden sie u.a. mit einer Auflistung aller im Korpus enthaltenen Texte und deren Annotationen
dokumentiert (Odebrecht 2015). Dieser Prozess ist oft ein Manueller und
daher sehr aufwendig und fehleranällig. Um diese Dokumentationsarbeit
zu automatisieren haben wir ein Plug-In (SaltInfoModule) ür die Konvertierungsplattform Pepper entwickelt (Zipser 2014).
Das SaltInfoModule extrahiert aus einem beliebigen Korpus dessen Größe, vorhandene Metadaten, Namen der Annotationsebenen sowie das verwendete Tagset. Neben einer XML Darstellung zur weiteren maschinellen Verarbeitungen wird auch eine menschenlesbare HTML-Darstellung
erzeugt. Letztere kann sowohl zur Fehlerüberprüfung als auch als Dokumentationshomepage ür das Korpus verwendet werden. Weiter bietet es
die Möglichkeit durch die Korpusstruktur (Korpora, Subkorpora, Dokumente) zu browsen und die jeweiligen Annotationen zu betrachten. Diese generierte Homepage kann durch eine Konfigurationsdatei und Stylesheets individuell angepasst werden. So können bspw. Beschreibungen zu
den Annotationen, zum Korpus selber oder Bilder zur Illustration eingeügt
384
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 385 — #397
Postersession der Sektion Computerlinguistik
werden. Zusätzlich können Links zu Suchanfragen ür einzelne Annotationen in ANNIS (Krause & Zeldes, erscheint) automatisch generiert werden.
Durch die Unterstützung einer Vielzahl linguistischer Formate durch Pepper (wie EXMARaLDA, PAULA, TEI, MMAX2, TIGER-XML, TCF uvm.)
können viele unterschiedliche Arten von Korpora durch das SaltInfoModule automatisch dokumentiert werden.
Literatur: • Krause, T. & Zeldes, A. (erscheint) ”ANNIS3: A New Architecture for Generic
Corpus Query and Visualization”. Digital Scholarship in the Humanities. • Odebrecht, C. (2015)
Interdisziplinäre Nutzung von Forschungsdaten mithilfe einer technisch-abstrakten Modellierung. Vortrag. Von Daten zu Erkenntnissen. 2. Jahrestagung des Verbandes der Digital Humanities im deutschsprachigen Raum. 25.02.-27.02.2015, Graz. • Petrova, S., Solf, M., Ritz, J.,
Chiarcos, C. & Zeldes, A. (2009) Building and using a richly annotated interlinear diachronic
corpus: The case of old high German tatian. Traitement automatique des langues 50(2), 47–71.
• Stede, M., & Neumann, A. (2014). Potsdam Commentary Corpus 2.0: Annotation for Discourse Research. In Proceedings of the Language Resources and Evaluation Conference, LREC 2014,
Reykjavik. • Zipser, F. (2014) SaltNPepper und das Formatpluriversum. LAUDATIO Workshop
2014. Berlin, 07.-08.10.2014.
Poster 17
Semantische Suche mit Ontologien
Magdalena Wolski1 , Michael Dembach1 & Ulrich Schade1
1
Fraunhofer-Institut ür Kommunikation, Informationsverarbeitung und
Ergonomie (FKIE)
[email protected], [email protected],
[email protected]
Ontologien stellen eine wertvolle Technologie dar, Wissen eines bestimmten Gegenstandsbereichs formal zu repräsentieren und nutzbar zu machen.
Dazu bedient man sich an Konzepten der Semantik und Logik, um Objekte
und ihre Relationen zu beschreiben. Es werden zwei Forschungsprojekte
des Fraunhofer–Instituts FKIE vorgestellt, in denen Ontologien auf unterschiedliche Weise genutzt werden. Im Projekt „Cyber Radar“ ür die Deutsche Telekom wird eine Ontologie ür den Bereich „Cyber Security“ ge-
385
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 386 — #398
Postersession der Sektion Computerlinguistik
CL
Poster
nutzt, um themenrelevante Nachrichten von Experten innerhalb der von
ihnen genutzten sozialen Medien zu finden und zu priorisieren. Die Ontologie ermöglicht es zum einen, die Suche durch verwandte Begriffe zu
erweitern. So werden Meldungen zum Thema „Schadsoftware“ auch erkannt, wenn sie nicht dieses Schlüsselwort, sondern stattdessen bspw. „Virus“ oder „Elkcloner“ beinhalten. Zum anderen lässt sich die enorme Nachrichtenflut durch Gewichtung bestimmter Teilbäume eingrenzen, um die
wichtigsten Meldungen herauszufiltern. Im Projekt EnArgus1,2 wird durch
die Projektpartner unter der Leitung des Fraunhofer–Instituts FIT ein zentrales Informationssystem entwickelt, um die Energieforschungsörderung
in Deutschland transparenter zu machen. Mit dem System werden staatliche Datenbanken genutzt, um zu ermitteln, zu welchen Fragestellungen
bereits Forschungsprojekte existieren. Damit werden nicht nur die geörderten Projekte gefunden, die bspw. „Windenergie“ explizit in ihrer Projektbeschreibung erwähnen, sondern auch jene, die sich mit „Rotoren“ und
„Blattverstellmechanismen“ beschäftigen. Die Projekte Cyber Radar und
EnArgus unterscheiden sich nicht nur durch ihre Gegenstandsbereiche,
sondern auch durch die zu untersuchenden Quellen, die verschiedenartige
Anforderungen stellen.
1 https://enargus.fit.fraunhofer.de/index.html
2 geördert
durch das Bundesministerium ür Wirtschaft und Energie (BMWi) auf Grundlage
eines Bundestagsbeschlusses. Förderkennzeichen: 03ET4010B
Poster 18
Annotation as a didactic means
Heike Zinsmeister1 , Melanie Andresen1 , Fabian Barteld1 , Jiayin Feng1 ,
Johanna Flick1 & Renata Szczepaniak1
1
Universität Hamburg, Institut ür Germanistik
[email protected]
In this poster/demo, we explore annotation, i.e. both the systematic process
of adding interpretative information to linguistic data and the resulting
data (Leech 1997), from a didactic perspective.
386
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 387 — #399
Postersession der Sektion Computerlinguistik
First, we make use of annotation as a method for students to explore linguistic concepts in a systematic, data-driven way: (a) inductive, based on
a given annotation scheme; (b) deductive by inducing generalizations and
classifications for a given phenomenon from data. For this purpose, we detail the Annotation Cycle (Zinsmeister 2011), an iterative process, which
is derived from the Hermeneutic Circle in the humanities (e.g. Bögel et
al. 2015) and the MAMA Annotation Development Process from computational linguistics (Pustejovsky & Stubbs 2012). We evaluate the didactic
use of annotation in linguistic seminars by means of a questionnaire that
combines usability and linguistic issues.
Second, we present a didactic methodology helping students to become
familiar with ways of creating sustainable annotations. This is especially
important since many linguistics students lack the computational expertise
needed for corpus linguistics. By conducting exercises for part of speech
and syntactic analysis with easy to use annotation programs students are
introduced to annotation guidelines and tools, which they can rely on in
their subsequent studies e.g. when writing a term paper. In the demo, we
will give examples for such exercises.
References: • Bögel, Thomas, Michael Gertz, Evelyn Gius, Janina Jacke, Jan Christoph Meister, Marco Petris, and Jannik Strötgen. 2015. “Gleiche Textdaten, Unterschiedliche Erkenntnisziele? Zum Potential Vermeintlich widersprüchlicher Zugänge zu Textanalyse.” In: Von
Daten Zu Erkenntnissen. Book of Abstracts - Vorträge, 119–26. Graz. www.gams.uni-graz.at/
o:dhd2015.abstracts-vortraege. • Leech, Geoffrey. 1997. “Introducing Corpus Annotation.” In:
Corpus Annotation. Linguistic Information from Computer Text Corpora, edited by Roger Garside, Geoffrey Leech, and Tony McEnery, 1–18. London/New York: Longman. • Pustejovsky,
James and Amber Stubbs. 2012. Natural Language Annotation for Machine Learning. O’Reilly.
• Zinsmeister, Heike. 2011. “Exploiting the ‘Annotation Cycle’ for Teaching Linguistics.” Slides
presented at the Workshop Corpora in Teaching Languages and Linguistics, Humboldt-Universität zu Berlin. www.linguistik.hu-berlin.de/en/institut-en/professuren-en/korpuslinguistik/
events-en/CTLL/ctll_abstracts/ctll_zinsmeister.
387
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 388 — #400
Postersession der Sektion Computerlinguistik
Poster 19
Nutzen von Parsing in der Termextraktion: eine qualitative
und quantitative korpusbasierte Untersuchung
Marie Zollmann1 , Ina Rösiger2 , Ulrich Heid1 & Michael Dorna3
1
Universität Hildesheim,2 Universität Stuttgart, 3 Robert Bosch GmbH
[email protected]
CL
Poster
Die meisten Verfahren zur Extraktion von mehrwortigen Fachtermini aus
Texten beruhen auf Mustersuche in wortartannotierten Texten, kombiniert
mit Verfahren zur Sortierung der musterbasiert extrahierten Kandidaten
nach Frequenz, Text-spezifizität, Assoziation zwischen den Bausteinen (z.B.
durch Assoziationsmaße) oder dem C-value-Maß (Frantzi/Ananiadou 2000).
Wir ühren derzeit Experimente mit einem solchen Termextraktor (Gojun
et al. 2012, Rösiger et al. 2015) durch, anhand von Expertentext und usergenerated content aus dem Bereich des Heimwerkens.
Bei der Extraktion von Mehrwort-Termkandidaten des Musters N+Prp+(Art)
+N werden neben grammatisch sinnvollen Kandidaten (wie z.B. Bohrer
mit Diamantspitze, Fräse mit Führungssiene) auch Sequenzen extrahiert, die strukturell unplausibel sind, weil sie nicht innerhalb einer Nominalphrase liegen, sondern die Grenze zwischen zwei von einander unabhängigen Phrasen (NP und PP) überspannen; solche Fälle treten z.B. oft
im Zusammenhang mit Verben auf, die PPen subkategorisieren oder bestimmte PP-Adjunkte präferieren (überschüssiges *Öl mit Leinenläppen
aufnehmen, *Regal an die Wand schieben), oder im Zusammenhang mit
idiomatischen Wendungen (den *Nagel auf den Kopf treffen).
Wir berichten über Experimente zur Extraktion aus dependenzgeparsten Texten (mate, cf. Bohnet 2010, Björkelund et al. 2010), in denen außer den grammatischen Funktionen auch Konstituenten und ihre Köpfe
annotiert sind. Wir präsentieren eine Evaluierung der Auswirkungen auf
die Termextraktionsqualität, die ein constraint hat, der Sequenzen von NP
und PP als Schwesterknoten verbietet. Aus der qualitativen Evaluierung
von ca. 2000 Satzinstanzen wird deutlich, welches zusätzliche Sprachwissen ür eine weitere Verbesserung der Extraktionsqualität nötig wäre; die
388
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 389 — #401
Postersession der Sektion Computerlinguistik
quantitative Evaluierung zeigt die Verteilung der Phänomene (welche Präpositionen sind besonders schwierig, welche unkritisch?) und erlaubt es,
den Qualitätszuwachs zu quantifizieren.
Literatur: • Bernd Bohnet. 2010. Top Accuracy and Fast Dependency Parsing is not a Contradiction. e 23rd International Conference on Computational Linguistics (COLING 2010), Beijing, China. • Anders Björkelund, Bernd Bohnet, Love Hafdell, and Pierre Nugues. 2010. A
high•performance syntactic and semantic dependency parser. In Coling 2010: Demonstration
Volume, pages 33-36, Beijing, August 23-27 2010. • Gojun, A., Heid, U., Weissbach, B., Loth, C.,
& Mingers, I. (2012). Adapting and evaluating a generic term extraction tool. In Proceedings
of LREC (pp. 651-656). • Ina Rösiger, Johannes Schäfer, Tanja George, Simon Tannert, Ulrich
Heid and Michael Dorna. Extracting terms and their relations from German texts: NLP tools
for the preparation of raw material for e-dictionaries. Proceedings of the eLex 2015 conference,
11-13 August 2015, Herstmonceux Castle, UK.
Poster 20
An XLE-based pragmatic parser
Mark-Matthias Zymla1 & Maike Müller1
1
Universität Konstanz
[email protected], [email protected]
We are presenting ongoing work on a linguistically motivated pragmatic
parser for discourse structures that combines theoretic research on discourse models with existing NLP applications. Various research projects
such as LingVisAnn, VisArgue and Tense & Aspect in Multilingual Semantic
Construction focus on semantic and pragmatic phenomena and interest in
NLP applications.1
Our pragmatic parser is built upon the existing German ParGram grammar (Dipper 2003) and the German AKR semantics (Zarrieß 2009). The core
mechanisms for this system are introduced in Zymla et al. (2015). There, we
show how the interaction between grammar and semantics/pragmatics can
be utilized to interpret discourse structures using the example of German
discourse particles.
389
CL
Poster
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 390 — #402
Postersession der Sektion Computerlinguistik
In this poster, the details of implementation with respect to the translation of Gunlogson’s discourse model (Gunlogson 2002) including a contextsensitive algorithm for discourse move determination are described. The
system provides an interesting perspective for both computational linguists
and theoretic researchers in the field of (computational) pragmatics.
1 See
http://ling.uni-konstanz.de/pages/compling/projects.html for more information.
References: • Dipper, Stefanie. 2003. Implementing and Documenting Large-Scale Grammars —
German LFG. Ph.D. thesis, IMS, University of Stuttgart. • Gunlogson, Christine. 2002. Declara-
CL
Poster
tive Questions. In B. Jackson, editor, Proceedings of Semantics and Linguistic eory XII, Pages
124–143. Ithaca, NY: CLC Publications. • Zarrieß, Sina. 2009. Developing German semantics
on the basis of parallel LFG grammars. In Proceedings of the 2009 Workshop on Grammar Engineering Across frameworks, Singapore, Pages 10–18. Association for Computational Linguistics. • Zymla, Mark-Matthias, Maike Müller, and Miriam Butt. 2015. Modeling the Common
Ground for Discourse Particles. In M. Butt and T. H. King, editors., Proceedings of LFG15, CSLI
On-line publications.
390
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 391 — #403
Tutorium der Sektion Computerlinguistik
Visual Analytics for Linguistics
Lectures: Miriam Butt1 & Dominik Sacha2
Universität Konstanz 1 FB Sprachwissenschaft, 2 FB Informatik und
Informationswissenschaft
[email protected], [email protected]
Dienstag, 23. 02. 2016, 10:00–17:00, Raum: G 227
The aim of this course is to provide an introduction to the emerging field
of the visualization of linguistic information (LingVis). LingVis combines
techniques developed in the fields of Information Visualization (InfoVis)
and Visual Analytics with methodology and analyses from theoretical and
computational linguistics. Besides standard visualization techniques such
as bar charts, scatterplots or line charts, a large number of advanced novel
methods have been developed. Prominent examples are treemaps, pixel
displays, sunburst visualizations and glyphs of varying complexity. We
present concrete use cases of LingVis for synchronic and diachronic linguistic questions. A part of the course will include a hands-on session in
which students can experiment with pre-prepared data sets and freely accessible LingVis software in order to investigate how complex linguistic
questions can profit from visual analysis.
The course will build on the ESSLLI 2014 course held by Miriam Butt and
Chris Culy, see http://ling.uni-konstanz.de/pages/home/butt/main/material/
esslli14-vis
Expected level and prerequisites
Introductory. Basic knowledge of linguistics is required. Elementary programming experience will be helpful but is not a priori required.
391
CL
Tutorium
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 392 — #404
Branchenvertreter und Datenschützer die Veröffentlichu
Privaten in Social Media und populären Medienformaten
empirischer Erkenntnisse werden Aspekte einer Ethik de
sowie Konzepte der Privatheit und deren Ökonomisierun
mögliche Regulierung diskutiert.
Im Mittelpunkt steht dabei das Verständnis von Privathei
Petra Grimm / Oliver Zöllner (Hg.)
Schöne neue Kommunikationswelt
oder Ende der Privatheit?
Die Veröffentlichung des Privaten
in Social Media und populären
Medienformaten
2012.
360 Seiten mit 33 Abbildungen.
Kart.
¤ 49,–
ISBN 978-3-515-10296-4
Michael Elmentaler / Markus Hundt /
Jürgen Erich Schmidt (Hg.)
Deutsche Dialekte. Konzepte,
Probleme, Handlungsfelder
2015.
516 Seiten mit 57 Abbildungen,
40 Farbabbildungen auf 20 Tafeln
und 55 Tabellen.
Kartoniert.
€ 79,–
& 978-3-515-10984-0
@ 978-3-515-10986-4
Jetzt auf unserer Homepage bestellen:
www.steiner-verlag.de
Michael Elmentaler / Markus Hundt /
Nutzung von Sozialen Netzwerken zu Grunde liegt: Gibt
Jürgen Erich Schmidt (Hg.)
schiedliche Vorstellungen von Privatheit in der Online- u
Deutsche Dialekte. Konzepte,
Offlinewelt? Wie sollen der Einzelne und die Gesellschaf
Probleme, Handlungsfelder
Herausforderungen der Social Media umgehen? Wie kön
Akten des 4. Kongresses der Internationalen Gesellschaft
Nutzer
zu(IGDD)
einem selbstbestimmten Privacy Management
für Dialektologie des
Deutschen
Zeitschrift für Dialektologie
und Linguistik
– Beiheft
werden?
Das Buch
bietet158einen Überblick über den Forsc
Die Dialektologie des
Deutschen
hat ihren neue
Gegenstandsund
zeigt zudem
Perspektiven und Lösungsansätze
bereich in jüngerer Zeit auf das gesamte regionalsprachliche
Spektrum erweitert. .und
. . . . theoretisch
. . . . . . . . . . . . .wie
. . . .methodologisch
.........................................................
neue Perspektiven entwickelt. Neben den Basismundarten
werden auch regionale Umgangssprachen und landschaftliche Ausprägungen des gesprochenen Standards systematisch erforscht. Die Rezeption kontaktlinguistischer Ansätze
ermöglicht ein besseres Verständnis der Genese und des
u. hasebrink: Das Social Web im Alltag von Jugendlichen
Wandels regionaler Varietäten.
Der vorliegende Band,
der ausgewählte
des Kieler
k. neef:
WandelBeiträge
des Privatheitsverständnisses
und die H
IGDD-Kongresses von 2012 versammelt, trägt der Vielfalt
rungen
für
Gesellschaft
und
Individuen
|
b.
debatin:
Soz
der modernen dialektologischen Forschung Rechnung und
bietet einen aktuellen
Einblick
in
Fragestellungen,
Methoden
Netzwerke aus medienethischer Perspektive | r. funiok:
und Ergebnisse. Außer den beiden programmatischen
tierte
der Privatheit in Sozialen N
Hauptvorträgen von
IngridStrategien
Schröder und zum
Alfred Schutz
Lameli sind
Beiträge zu den Bereichen
Phonologie
und Morphologie,
r. capurro:
Verwende
niemals deine wahren Daten! | h
Syntax, Sprachatlanten, Wahrnehmungsdialektologie,
Konzept
„Privatheit“sowie
in den Medien | k. dörr / m. herz /
Minderheitensprachen
und Mehrsprachigkeit
Regionalsprachenforschung enthalten.
Aus dem Inhalt
Realitätsentwürfe in Scripted Reality-Dokumentationen
. . . . . . . . . . . . . . . . . . . . . . .Strukturen
. . . . . . . . . . . . . . . . . .und
. . . . . . .Probleme
. . . . . . . . . . . . . . . .des
. . . . . .Zusammenlebens in media
| e. wagner: Virtuelle Kommunikation be
Mit Beiträgen Gesellschaften
von
und Co. | e. clausen-muradian: Das rechtliche Instrume
Hauptvorträge: Ingrid Schröder, Alfred Lameli
Daten- und
Persönlichkeitsschutzes
| u.a.
Phonologie und Morphologie:
Raffaela
Baechler, Hanna
Fischer, Peter Gilles, Cristian Kollmann, Jens Philipp Lanwer
Syntax: Jürg Fleischer, Simon Kasper, Thilo Weber
Pröll
/ Simon
Pickl / Aaron Spettl
Franz
Steiner
Verlag
Hier geht es zu Bestellung!Sprachatlanten: Simon
Wahrnehmungsdialektologie:
Joachim44
Gessinger
/ Judith
Birkenwaldstr.
· D – 70191
Stuttgart
Butterworth, Markus Hundt / Nicole Palliwoda / Saskia
Telefon:
0711
/
25
82
–
0
·
Fax:
Schröder, Andrea Kleene, Susanne Oberholzer, Anja 0711 / 25 82 – 390
E-Mail: [email protected]
Schaufuß
Minderheitensprachen
und Mehrsprachigkeit:
Steffen Krogh,
Internet:
www.steiner-verlag.de
Stefan Rabanus
Regionalsprachen: Michael Elmentaler / Peter Rosenberg,
Roland Kehrein, Marie Josephine Rocholl
www.steiner-verlag.de
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 393 — #405
Doktorandenforum
Good Scientific Practice (GSP)
Organisators: Alexandra Rehn, Janina Reinhardt & Yvonne Viesel
Universität Konstanz, FB Sprachwissenschaft
[email protected]
Dienstag, 23. 02. 2016, 9:30–16:30, Raum: F425
The 6th DGfS Doktorandenforum will, like in the past years, precede the
annual DGfS-conference and it will take place at the University of Konstanz
on the 23rd of February. It is called Doktorandenforum not only because it is
organised by and for doctoral students, but also because it is meant to be a
networking event by providing an opportunity to establish ties, to initiate
contacts, to simply get in touch with peers.
Consequently, we chose a topic which is of great importance to anyone
at the beginning of his/her academic career: Good Scientific Practice (GSP).
Intuitively, everyone knows what the rules of GSP are: Be honest, make
proper documentations of your research and respect the work of others.
No half-truth, no concealment of data and no plagiarism. But is it always
that easy? What about the grey areas? And most importantly, what can we
do, when the rules of GSP are not respected?
Problems may arise at any point in an academic career; but especially
PhD students, who mostly work with people more advanced in their career,
may experience conflicts caused by hierarchical relations. It is the duty
of every supervisor to coach their PhD-students. But who do you talk to
when things go wrong? What do you do when you have the feeling your
supervisor is hardly ever available or does not want to help you? This is
tricky because on the one hand you need their help, on the other hand you
do not want to argue with them. Sometimes difficulties also come up when
working closely with your supervisor. He or she might ask you to work in
393
DF
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 394 — #406
Doktorandenforum
DF
a way you do not approve of or they might not be happy with the outcome
of your work – and what if the results of your experiment simply do not
tell the story your supervisor wants to hear? Problems can also come up
when publishing a paper: who has done the major part of the work, is your
supervisor entitled to be named as co-author just because he provided you
with the facilities to run your experiment? Despite the question of who has
done most of the work it is also of importance who came up with the initial
idea that made it happen in the first place.
Issues like the ones listed above will be addressed in the workshop and
the aim is to provide you with possible strategies to overcome them. Every German university has a so-called ombudsman/woman who is a neutral person and an expert concerning everything that is subsumed under
the term ‘good scientific practice’. We are happy that our local ombudsman Prof. Dr. Maret will share his expertise with us on that matter! There
will be enough room to discuss the individual topics (and in addition, your
questions as well as possible examples you might be able to provide from
your own experience)!
The language of instruction will be English. Participation is free, but you
should register in advance as the number of participants is limited! If you
want to participate or if you have any questions please send an e-mail to
[email protected].
394
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 395 — #407
Doktorandenforum
Schedule
09:30–10:15
Welcome & Introduction to GSP
10:15–11:15
Problems of Power Asymmetry
11:15–11:30
Coffee Break
11:30–12:30
Authorship and Competition
12:30–13:30
Lunch Break
13:30–14:15
Proper Documentation
14:15–15:00
Whistleblowing and Procedures
15:00–15:15
Coffee Break
15:15–16:30
Guideline, Conclusions & Farewell
Honorary Speaker
Prof. Dr. Georg Maret, Ombudsman of the University of Konstanz
DF
395
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 396 — #408
Handbücher zur Sprach- und
Kommunikationswissenschaft /
Handbooks of Linguistics and
Communication
XXXX
Science (HSK) 40
Peter O. Müller, Ingeborg Ohnheiser,
Susan Olsen, Franz Rainer (Hrsg.)
WORD-FORMATION
An International Handbook of the Languages of Europe
ff
ff
ff
State of the art in word-formation
207 articles in XVI chapters from a wide variety of perspectives,
written by leading international scholars
74 portraits on word-formation in the individual languages of Europe
Preise pro Band:
Gebunden UVP € 299.00 [D]
eBook UVP € 299.00 [D]
Print + eBook UVP € 499.00 [D]
VOLUME 1
VOLUME 3
VOLUME 5
03/2015, XXII, 802 Seiten
ISBN 978-3-11-024624-7
PDF ISBN 978-3-11-024625-4
EPUB ISBN 978-3-11-039320-0
ISBN 978-3-11-214951-5
09/2015. XII, 826 Seiten
ISBN 978-3-11-037566-4
PDF ISBN 978-3-11-037573-2
EPUB ISBN 978-3-11-042361-7
ISBN 978-3-11-037574-9
03/2016. Ca. XII, 740 Seiten
ISBN 978-3-11-043094-3
PDF ISBN 978-3-11-042494-2
EPUB ISBN 978-3-11-042751-6
ISBN 978-3-11-042495-9
VOLUME 2
VOLUME 4
06/2015, XII, 758 Seiten
ISBN 978-3-11-024626-1
PDF ISBN 978-3-11-024627-8
EPUB ISBN 978-3-11-039468-9
ISBN 978-3-11-214733-7
01/2016. Ca. 1000 Seiten
ISBN 978-3-11-037897-9
PDF ISBN 978-3-11-037908-2
EPUB ISBN 978-3-11-039354-5
ISBN 978-3-11-037909-9
degruyter.com
n
en Sie unsere
Bitte besuch
tellung
ss
au
ch
Bu
r
Stand in de
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 397 — #409
Infotag der Lehramtsinitiative der DGfS
Mehrsprachigkeit fördern – von der Kita bis zum
Gymnasium
Organisatoren: Björn Rothstein1 & Svenja Kornher2
Ruhr-Universität Bochum, 2 Universität Konstanz
1
[email protected], [email protected]
Dienstag, 23. 02. 2016, 15:00–19:00, Raum: C 230, C 421, C 422, C 423,
C424, C427
Um mit Kindern das Deutsche zu reflektieren, es mit anderen Sprachen zu
vergleichen und die Mehrsprachigkeit zu ördern, bedarf es linguistischer
Kenntnisse. Sprachreflexion, Sprachbewusstheit und Sprachvergleich sind
Ziele aller sprachlichen schulischen Fächer und der vorschulischen Bildungsinstitutionen. Der Infotag der Lehramtsinitiative im Rahmen der Jahrestagung der Deutschen Gesellschaft ür Sprachwissenschaft am 23. Februar 2016 an der Universität Konstanz thematisiert daher mit Fortbildungsangeboten ür aktive und in Ausbildung befindliche ErzieherInnen, Lehramtsstudierende, ReferendarInnen und aktive LehrerInnen Mehrsprachigkeit, auch bezogen auf dialektale und umgangssprachliche Variationen des
Deutschen.
Sprachwissenschaftliche und sprachdidaktische ExpertInnen bieten in
Workshops wissenschaftlchen Input einschließlich Best Practice Unterrichtsbeispielen und Anwendungsreflexionen ür Bildungseinrichtungen von der
Kita bis in die Sekundarstufe.
397
LAI
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 398 — #410
Infotag der Lehramtsinitiative der DGfS
Programm
15:00 - 15:15
Begrüßung
15:15 - 16:15
Eröffnungsvortrag von Prof. Dr. Janet Grijzenhout, Dr.
Tanja Rinker (beide Zentrum ür Mehrsprachigkeit,
Universität Konstanz): – Raum C 230 –
Platz ür viele Sprachen – individuelles Können und
Klassenzimmeralltag
16:15 - 16:30
Pause
16:30 - 17:30
Workshopblock
Workshop 1: Leseörderung auf Satzebene im
mehrsprachigen Klassenzimmer
Workshop 2: Heute Grammatik, morgen wieder was
Schönes
Workshop 4: Vorlesesituationen in mehrsprachigen
Kontexten gestalten
Workshop 5, Teil I: Kinderreime, Gedichte und
Geschichten ür Vor- und Grundschule im
mehrsprachigen Kontext
LAI
Workshop 6, Teil I: Schulische Mehrsprachigkeitsprojekte
in Klasse 5 und 6: Wer lernt da eigentlich was?
17:30 - 18:00
Pause
18:00 - 19:00
Workshopblock
Workshop 1 (Wiederholung): Leseörderung auf
Satzebene im mehrsprachigen Klassenzimmer
398
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 399 — #411
Infotag der Lehramtsinitiative der DGfS
Workshop 3: Über die Wortarten und Satzglieder hinaus
– Sprachliche Strukturen entdecken und vergleichen.
Grammatikunterricht ab Klasse 8
Workshop 4 (Wiederholung): Vorlesesituationen in
mehrsprachigen Kontexten gestalten
Workshop 5, Teil II: Kinderreime, Gedichte und
Geschichten ür Vor- und Grundschule im
mehrsprachigen Kontext
Workshop 6, Teil II: Schulische
Mehrsprachigkeitsprojekte in Klasse 5 und 6: Wer lernt
da eigentlich was?
Workshop 1
Leseförderung auf Satzebene im mehrsprachigen
Klassenzimmer
Alexandra Zepter1 & Sabine Zepnik1
Universität zu Köln
1
[email protected], [email protected]
LAI
Dienstag, 23. 02. 2016, 16:30–17:30 und Wiederholung um 18:00–19:00,
Raum: C 421
Sicheres Lesen auf basaler Ebene ermöglicht es, im weiteren Verlauf der
Schulzeit auch längere und komplexere Texte zu verstehen. Davon sind
Lernende mit mehrsprachigem Hintergrund genauso betroffen wie monolinguale Lernende mit Deutsch als Erstsprache. Damit der Schriftsprachund Orthographieerwerb ohne Umwege verläuft, ist es von zentraler Bedeutung, bereits in der Grundschule auf die Systematik unserer Schrift aufmerksam zu machen. Im Workshop möchten wir Lehrkräfte mit einem systembasierten, gleichsam spielerisch konzipierten Leseunterricht bekannt
399
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 400 — #412
Infotag der Lehramtsinitiative der DGfS
machen und dabei insbesondere syntaktische Strukturen (vorwiegend Topologie des Satzes und der Nominalphrase im Deutschen) und deren orthographische Markierungen in den Mittelpunkt stellen. Die spielerischen
Konzeptionen bieten v.a. die Möglichkeit zum impliziten Lernen; daneben
lädt das Material auch zu aktiver Auseinandersetzung mit syntaktischen
Strukturen und damit zur Sprachreflexion im Kernbereich ein.
Workshop 2
Heute Grammatik, morgen wieder was Schönes
Sandra Döring
Universität Leipzig
[email protected]
Dienstag, 23. 02. 2016, 16:30–17:30, Raum: C 422
LAI
Warum wird Grammatikunterricht als langweilig, staubtrocken, aber schwierig wahrgenommen? Kann Grammatik Spaß machen oder ist das ein Widerspruch in sich? Was gibt es im Grammatikunterricht zu entdecken? Wie
können SuS zu einer Entdeckungsreise in die Welt der Sprache(n) begeistert und angeleitet werden? Der Workshop lädt ausdrücklich zum Erfahrungsaustausch und zur Diskussion ein. Er richtet sich gleichermaßen an
LehrerInnen, die froh sind, wenn die Lerneinheit ‚Grammatik‘ vorbei ist,
und an solche, die gern mehr Grammatik unterrichten würden (Deutschunterricht Klassen 5-8).
400
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 401 — #413
Infotag der Lehramtsinitiative der DGfS
Workshop 3
Über die Wortarten und Satzglieder hinaus – Sprachliche
Strukturen entdecken und vergleichen.
Grammatikunterricht ab Klasse 8
Sandra Döring
Universität Leipzig
[email protected]
Dienstag, 23. 02. 2016, 18:00-19:00, Raum: C 422
Der Workshop bietet die Gelegenheit des Austauschs und der Diskussion
ür DeutschlehrerInnen (Deutschunterricht Kl. 8-12/13) und FremdsprachenlehrerInnen, ür LehrerInnen, die gern Grammatik unterrichten und
ür LehrerInnen, die sich selbst unsicher im Grammatikunterricht ühlen,
ür LehrerInnen, die auch gern in höheren Klassen im Deutschunterricht
Grammatik unterrichten würden, und ür LehrerInnen, die froh sind, dass
man ‚in der 8. Klasse damit weitestgehend durch ist‘, ür solche, die auf die
Erstsprachen der SuS im Sprachunterricht zurückgreifen, und ür solche,
die dies vermeiden, weil sie diese Sprachen ja selbst nicht sprechen.
Workshop 4
Vorlesesituationen in mehrsprachigen Kontexten gestalten
LAI
Linda Stark
Universität Würzburg
[email protected]
Dienstag, 23. 02. 2016, 16:30–17:30 und Wiederholung um 18:00–19:00,
Raum: C 423
Vorlesesituationen können sich erwiesenermaßen auf unterschiedliche Spracherwerbsbereiche positiv auswirken. Diese positive Wirkung macht sich
insbesondere die Sprachörderpraxis mit Kindern zunutze, die das Deutsche als Zweitsprache – nicht nur aber vor allem in der Kita – erlernen.
401
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 402 — #414
Infotag der Lehramtsinitiative der DGfS
Gerade in Vorlesesituationen mit mehreren Kindern, deren Deutschkenntnisse unterschiedlich weit vorangeschritten sind, stellen sich im Rahmen
der Sprachörderung mit Bilderbüchern jedoch vielältige Herausforderungen an die Vorlesenden. Im Workshop sollen Möglichkeit aufgezeigt und
ausprobiert werden, um diesen Herausforderungen zu begegnen. Darüber
hinaus soll gemeinsam erarbeitet werden, inwiefern sich diese Handlungsmöglichkeiten auch auf Rezeptionssituationen anderer Medien, so z.B. textloser Bilderbücher oder Kamishibais, übertragen lassen.
Workshop 5
Kinderreime, Gedichte und Geschichten für Vor- und
Grundschule im mehrsprachigen Kontext
Eva Belke1 , Friederike von Lehmden1 & Claudia Müller1
1
Ruhr-Universität Bochum
[email protected], [email protected], [email protected]
Dienstag, 23. 02. 2016, Teil I 16:30–17:30 + Teil II 18:00–19:00, Raum: C 424
LAI
Kinderlieder, -reime und -bücher sowie Sprachspiele sind eine ideale Grundlage ür implizites sprachliches Lernen. Sie enthalten sprachliche Einheiten
und ihre strukturellen Beziehungen in hoch konzentrierter und redundanter Form und laden zum Spielen mit diesen Einheiten und Strukturen ein.
Die generative Textproduktion (GT) nutzt dies, indem sie Lerner anregt,
eigene Texte auf der Basis vorgegebener Textmuster zu generieren. Wir
wollen in diesem Workshop aufzeigen, inwiefern sich vor allem Texte der
Kinderliteratur und Lieder daür eignen, diejenigen kognitiven Ressourcen und impliziten Lernmechanismen, die insbesondere jüngere Kinder
zum Spracherwerb nutzen, zu aktivieren. Hierzu stellen wir ein Liederbuch (Kauffeldt et. al 2014) sowie sprachörderliche Bilderbücher vor und
analysieren diese in Bezug auf ihre sprachlichen Lernmöglichkeiten.
402
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 403 — #415
Infotag der Lehramtsinitiative der DGfS
Workshop 6
Schulische Mehrsprachigkeitsprojekte in Klasse 5 und 6:
Wer lernt da eigentlich was?
Carolina Luisio
Pädagogische Hochschule Bern, Pädagogische Hochschule Zürich
[email protected]
Dienstag, 23. 02. 2016, Teil I 16:30–17:30 + Teil II 18:00–19:00, Raum: C 427
Die Mehrsprachigkeit hat Einzug gehalten in den Klassenzimmern, auch
in den Gymnasien. Nebst der Förderung der Schulsprache ist der Einbezug der Erstsprache der Schülerinnen und Schüler ein wichtiges Anliegen
der modernen Sprachdidaktik. In verschiedenen Projekten und Ansätzen
wie language awareness, ELBE, HSK etc werden solche Zugänge beschrieben. Im Workshop sollen Projekte vorgestellt und Beispiele zur konkreten
Umsetzung im Unterricht aufgezeigt und ausprobiert werden. Wichtig ist
es hierbei den Blick auf kleine, im Alltag verankerte Elemente zu lenken,
welche unspektakulär und mit wenig Aufwand umgesetzt werden können.
Dabei sollen auch kritische Fragen zur Wirksamkeit und Praktikabilität
solcher Unterrichtssequenzen gestellt werden.
LAI
403
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 404 — #416
Angelika Redder, Johannes Naumann,
Rosemarie Tracy (Hrsg.)
FORSCHUNGSINITIATIVE
SPRACHDIAGNOSTIK UND
SPRACHFÖRDERUNG
– ERGEBNISSE
2015, 240 Seiten, br., 29,90 €,
ISBN 978-3-8309-3288-8
E-Book: 26,99 €, ISBN 978-3-8309-8288-3
D
ie „Forschungsinitiative Sprachdiagnostik und Sprachförderung
(FiSS)“ schlägt in ihrer zweiten
Laufzeit die Brücke von der anwendungsbezogenen Grundlagenforschung hin
zur Intervention. Wie sind zielgenaue
diagnostische Verfahren und wirksame
Programme zur Sprachförderung auszugestalten? Welche Faktoren müssen bei
ihrer Konzeption und Durchführung beachtet werden? Was können Erzieherinnen und Erzieher, was können Lehrkräfte
und Eltern tun, um die sprachliche Qualifizierung ihrer Kinder mit positiven Impulsen anzustoßen und zu begleiten?
Der zweite Band der Forschungsinitiative versammelt wichtige Ergebnisse und
resümiert darüber hinaus das Forschungs­
ensemble als Ganzes. Er entwickelt Perspektiven für die Implementierung von
Sprachförderung in Schulen und Kitas
ebenso wie für die Weiterentwicklung der
Forschung.
www.waxmann.com
[email protected]
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 405 — #417
Tagung der Arbeitsgemeinschaft Linguistische
Pragmatik (ALP e.V.)
Sprachliche Verfestigungen und sprachlich Verfestigtes
Dienstag, 23. 02. 2016, 8:45–18:15, Raum: G 201
Sprachgebrauch kommt ohne Routinen und ohne Rückgriff auf vorgeprägte Muster nicht aus. Neuere sprachtheoretische Ansätze wie etwa konstruktionsgrammatische Zugänge oder die Perspektive sprachlicher Prägungen stellen denn auch solche verfestigten Muster, angefangen bei idiomatischen Wortverbindungen über Kollokationen bis hin zu syntaktischen
Schemata, geradezu ins Zentrum ihrer Betrachtungen. Grundlegend pragmatisch ist der Blick auf verfestigte Muster in zweifacher Hinsicht: Erstens
entstehen sie im Gebrauch und sind als Ergebnisse von Routinisierungsprozessen zu beschreiben; zweitens sind sie als Einheiten mit übersummativer Qualität häufig nur unter Verweis auf ihre typischen Gebrauchskontexte hinreichend zu erfassen.
Während in neueren sprachtheoretisch-pragmatischen Ansätzen insgesamt ein zunehmendes Interesse an Verfestigungen auf verschiedenen sprachlichen Ebenen spürbar ist, haben sich umgekehrt auch in der Phraseologie
pragmatische Beschreibungsansätze etabliert. Die Forschungen zu pragmatischen Phraseologismen wie Routineformeln oder Phrasemkonstruktionen sind hier ebenso zu nennen wie die Arbeiten zur Funktionalität von
Phraseologismen in verschiedenen Textsorten und Kommunikationsbereichen.
Auf der ALP-Tagung sollen theoretische und empirische Zugriffe auf
Verfestigungen und Verfestigtes auf allen Ebenen des Sprachgebrauchs diskutiert werden. Das Hauptaugenmerk liegt dabei auf Beiträgen, die theoretische Aspekte des Themas anhand von konkreten empirischen Fallstudien
behandeln.
405
ALP
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 406 — #418
Tagung der Arbeitsgemeinscha Linguistische Pragmatik
Website
http://www.alp-verein.de
Programm
8:45–9:00
Einührung
9:00–9:45
Dmitrij Dobrovol’skij:
Verfestigungen im Lexikon: diskurspragmatische
Faktoren
9:45–10:15
Natalia Filatkina:
Trying to chart the directions. Sprachhistorische
Verfestigungsprozesse am Beispiel der Routineformeln
10:15–10:45
Tilo Weber:
Sprichwörter und Anti-Sprichwörter, substanzielle und
schematische Konstruktionen – unterschiedliche
Perspektiven auf Produkte und Prozesse sprachlicher
Verfestigung
10:45–11:15
Kaffeepause
11:15–11:45
Nadine Proske:
Sprachliche Verfestigungen in der Interaktionalen
Linguistik
11:45–12:15
Rita Finkbeiner:
Konstruktionen der Narration. Serialisiertes Erzählen in
Bilderbüchern ür Vorschulkinder
12:15–12:45
Katrin Hee:
Schultypische sprachliche Muster. Wortverbindungen in
Erwerbsperspektive
12:45–14:00
Mittagspause
ALP
406
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 407 — #419
Tagung der Arbeitsgemeinscha Linguistische Pragmatik
14:00–14:45
Kathrin Steyer:
Wortverbindungsmuster. Zur funktionalen Verfestigung
sprachlicher Einheiten durch rekurrenten Gebrauch
14:45–15:15
Sören Stumpf:
Phraseografie und Korpusanalyse
15:15–15:45
Sarah Brommer:
Sprachlich Verfestigtes in wissenschalichen Texten. Ein
neuer Blick auf einen bekannten Gegenstand:
Möglichkeiten, Grenzen und erste Ergebnisse einer
induktiv korpuslinguistischen Untersuchung
15:45–16:15
Kaffeepause
16:15–16:45
Juliane Schröter:
‚Geühlte Objektivität‘. Kulturanalytisch-pragmatische
Untersuchung eines neuen Phraseologismus
16:45–17:15
Marcel Dräger:
Mitverfestigtes. Handlungs- und Erfahrungswissen in
Kollokationen
17:15–17:45
Philipp Dreesen:
„wilde“ Südseeinsulaner. Ausdrücke in
Anührungszeichen als metasprachlich-referierende
Verfestigung und ihre paradoxe Distanzierungsfunktion
17:45–18:15
Abschlussdiskussion
18:30
Mitgliederversammlung der Arbeitsgemeinschaft
Linguistische Pragmatik e.V.
ALP
407
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 408 — #420
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 409 — #421
Teil VI.
Anhang
409
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 410 — #422
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 411 — #423
Notizen
.
411
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 412 — #424
Notizen
.
412
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 413 — #425
Notizen
.
413
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 414 — #426
Notizen
.
414
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 415 — #427
Notizen
.
415
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 416 — #428
Notizen
416
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 417 — #429
Gesamtübersicht der Arbeitsgruppensitzungen
417
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 418 — #430
Gesamtübersicht der Arbeitsgruppensitzungen
2
F 426
3
G 530
4
F 425
5
Mittwoch, 24. 02. 2016
1
G 300
AG
G 309
Raum
Adani et
al.
Gyuris
Lohnstein
Eldeen
BacskaiAtkari
Öhl
Zebib et
al.
Roberts &
SzczepekReed
Schouwenaars et
al.
Pagliarini
& Arosio
Valian
Rizzi
Philipp &
Primus
Bott et al.
Graf et al.
Martin
Apresjan
Kretzschmar
et al.
Kasper
Polinsky
14:00–
14:30
Vikner
Wiese et
al.
Ganenkov
Irwin
Grillo et
al.
Brandt &
Schumacher
Dimroth
& Sudhoff
14:30–
15:00
Los & van
Kemenade
Casalicchio
&
Cognola
15:00–
15:30
15:30–
16:00
16:00–
16:30
16:30–
17:00
17:00–
17:30
17:30–
18:00
18:00–
18:30
6
7
E 404
8
E 403
9
G 308
10
G 201
11
E 402
12
E
402
13
Veenstra
Fuß
Gergel
Franco et
al.
Etxeberria
&
Giannakidou
Su
Panzeri &
Foppolo
–
–
–
–
–
Bill et al.
–
Oiry
Pintér
Roeper &
Rau
–
–
D 406
Slomanson
Drozd
F 420
Renz
Aloni
Evertz &
Kirchhoff
Nolda
Häussler
& Juzek
Schippers
—
Bunt
Benedicto
Danon
Symanczyk
Joppe
Schmidt
Günther
Schembri
Budde
Dachkovsky
K
Hou
Cormier
Meier
—
Culbertson
Külpmann
& Symanczyk
Joppe
Koeneman
&
Zeijlstra
Janzen &
Maaß
Villalba
& Koller
van
Deemter
Franck et
al.
418
Bader
Raum
9:00–
9:30
13:00–
14:30
Kiss
Sekerina
Arnhold
et al.
Seeliger
& Repp
Blanchette
Sorace
Wierzba
Ellsiepen
E 404
8
P S CL  M
—
Quer
Repp
Price &
Witzel
12:30–
13:00
van Valin
Jr.
Sanfelici
et al.
Goodman
12:00–
12:30
Andorno
& Crocco
Lourenço
Santoro
et al.
Schembri
et al.
Meir
D 406
7
Balvet &
Garcia
Meyer
Franke &
Bergen
Stevens
F 420
6
Gould
Donazzan
& Tovena
Claus et
al.
Turco
F 425
5
11:30–
12:00
Kiparsky
Hinterwimmer Haendler
Vasishth
G 530
4
K
Braun &
Smolka
10:30–
11:00
Stockall
et al.
Iordachioaia
OltraMassuet
et al.
Koukoulioti
& Stavrakaki
F 426
3
G 300
2
11:00–
11:30
de Swart
&
van
Bergen
10:00–
10:30
9:30–
10:00
1
G 309
AG
Donnerstag, 25. 02. 2016
—
Berg
Turgay
& Gutzmann
Huesmann
& Kirchhoff
Kalbertodt
et al.
Reißig
—
E 403
9
Kallel &
Larrivée
Sundquist
&
Heycock
Mitrović
Etxepare
Grosz
Geist
G 201
11
Viðarsson
Gärtner
Angantýsson
G 308
10
–
–
–
–
–
–
–
E 402
12
Giusti &
Iovino
ten Wolde
Breban &
Davidse
Cinque
Kotowski
& Härtl
Scontras
et al.
E 402
13
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 419 — #431
Gesamtübersicht der Arbeitsgruppensitzungen
419
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 420 — #432
Gesamtübersicht der Arbeitsgruppensitzungen
1
2
F 426
3
G 530
4
F 425
5
F 420
6
D 406
7
Freitag, 26. 02. 2016
AG
G 300
E 404
8
Fuhrhop
E 403
9
Jamieson
G 308
10
Lühr
G 201
11
—
E 402
12
Wildeboer
E 402
13
Stavrou
Besnard
Krebs et
al.
G 309
Zymla
Raum
Matić &
Nikolaeva
de Hoop
de
Villiers
&
Roeper
Djärv et
al.
Lensch
11:30–
12:00
–
Turolla et
al.
Gianollo
et al.
–
Pavlič
Eyþórsson
Wittenberg
et al.
Sideltsev
Kuhn
Costello
Final
comments
–
Hamann
Horch
Nishida
—
Woods
Gutzmann
et al.
Kuhn
Schrinner
Panayidou
Trabandt
et al.
Musgrave
et al.
Pfau &
Salzmann
–
Gabrovska
& Geuder
Garassino
& Jacob
Concluding
remarks
Luraghi
Demirdache
et al.
Panizza &
Lohiniva
Sano
Haeberli
& Ihsane
Czypionka
& Eulitz
Ahdout
de Ruiter
et al.
Wahl
Catasso
Pappert
et al.
Final discussion
Nykiel
Roeper &
Woods
Reuters
et al.
12:00–
12:30
13:00–
13:30
Bott et al.
12:30–
13:00
13:30–
14:00
420
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 421 — #433
Personenverzeichnis
Adani, Flavia, 149, 150
Ahdout, Odelia, 145
Alday, Phillip, 128
Alexiadou, Artemis, 39, 103
Aloni, Maria, 312
Andorno, Cecilia, 189
Andresen, Melanie, 386
Angantýsson, Ásgrímur, 299
Apresjan, Valentina, 130
Aristodemo, Valentina, 232
Arnhold, Anja, 192
Arosio, Fabrizio, 154
Augusto, Marina R. A., 176
Bacskai-Atkari, Julia, 184
Bader, Markus, 79
Bai, Yinchun, 369
Baker, James, 123
Balvet, Antonio, 235
Barteld, Fabian, 386
Battisti, Alessia, 364
Baumann, Michael, 120
Belke, Eva, 402
Benedicto, Elena, 219
Benz, Anton, 201, 216
Berenike Herrmann, J., 379
Berg, Kristian, 285
van Bergen, Geertje, 80
Bergen, Leon, 208
Besnard, Anne-Laure, 261
Bettelou, Los, 77
Bidese, Ermenegildo, 353
Bill, Cory, 338
Blanchette, Frances, 260
Bornkessel-Schlesewsky, Ina, 128
Bott, Oliver, 69, 71, 101
Bott, Stefan, 371
Bowers, Jack, 380
Brandes, Jasper, 363
Brandt, Patrick, 105
Brandt, Silke, 174
Braun, Bettina, 82, 192
Breban, Tine, 345
Brommer, Sarah, 407
Brynjólfsdóttir, Elísa Guðrún, 101
Budde, Monika, 273
Bunk, Oliver, 71
Bunt, Harry, 202
Butt, Miriam, iii, 391
Bögel, Tina, iii
Casalicchio, Jan, 75
Catasso, Nicholas, 96
Cheng, Lisa Lai-Shen, 16, 63
421
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 422 — #434
Personenverzeichnis
Cinque, Guglielmo, 345
Claus, Berry, 188
Cognola, Federica, 75
Cormier, Kearsy, 225, 230
Corrêa, Letícia M. S., 176
Costello, Brendan, 241
Crain, Stephen, 338
Crocco, Claudia, 189
Culbertson, Jennifer, 252
Czypionka, Anna, 119
Dachkovsky, Svetlana, 222
Danon, Gabi, 252
Darby, Jeannique, 123
Davidse, Kristin, 345
van Deemter, Kees, 206
Degen, Judith, 342
Dehé, Nicole, ii, iii, 3
Delmonte, Rodolfo, 364
Dembach, Michael, 385
Demirdache, Hamida, 144
Dimroth, Christine, 177, 179
Djärv, Kajsa, 91
Dobrovol’skij, Dmitrij, 406
Dold, Simon, iii
Domaneschi, Filippo, 192
Donazzan, Marta, 138
Dorna, Michael, 388
Dreesen, Philipp, 407
Drozd, Kenneth, 332
Druskat, Stephan, 367
Dräger, Marcel, 407
Döring, Sandra, 400, 401
Eckart, Kerstin, 372
422
Eckhard, Regine, iii
Eldeen, Unaisa Khir, 185
Ellsiepen, Emilia, 256
Etxeberria, Urtzi, 313
Etxepare, Ricardo, 321
Eulitz, Carsten, 119
Evertz, Martin, 269, 270
Eyþórsson, Þórhallur, 291, 308
Feng, Jiayin, 386
Fenlon, Jordan, 230
Filatkina, Natalia, 406
Finkbeiner, Rita, 406
Flick, Johanna, 386
Foppolo, Fancesca, 333
Franck, Julie, 254
Franco, Irene, 315
Franke, Michael, 208
Freitag, Constantin, iii, v, 69, 71,
101
Fritzsche, Tom, 149, 150
Fuhrhop, Nanna, 286
Fuß, Eric, 294
Gabrovska, Ekaterina, 142
Ganenkov, Dmitry, 110
Garassino, Davide, 196
Garcia, Brigitte, 235
Gauer, Isabelle, 369
Gehrke, Berit, 107
Geist, Ljudmila, 317
Geraci, Carlo, 232
Gergel, Remus, 316
Geuder, Wilhelm, 142
Giannakidou, Anastasia, 313
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 423 — #435
Personenverzeichnis
Gianollo, Chiara, 311, 330
Giusti, Giuliana, 348
Goodman, Noah, 210
Goodman, Noah D., 342
Gould, Isaac, 84
Graf, Tim, 132
Grijzenhout, Janet, ii, iii, 3, 398
Grillo, Nino, 107
Grosz, Patrick G., 319
Gutzmann, Daniel, 195, 283
Gwilliams, Laura, 113
Gyuris, Beáta, 179
Gärtner, Hans-Martin, 291, 300
Gärtner, Markus, 372
Günther, Hartmut, 272
Haeberli, Eric, 306
Haendler, Yair, 160
Hamann, Cornelia, 156
Hamann, Hanjo, 369
Hamann, Silke, 287
Hartmann, Katharina, 195
Haugh, Michael, 215
Hautli-Janiz, Anette, iii
Härtl, Holden, 343
Heck, Fabian, v
Hee, Katrin, 406
Heid, Ulrich, 372, 388
Hendriks, Petra, 158
Herrmann, Annika, 217
von Heusinger, Klaus, 311, 330
Heycock, Caroline, 91, 303
Hinterwimmer, Stefan, 136
Hirsch, Nils, 107, 123
de Hoop, Helen, 118
Horch, Eva, 213
Hou, Lynn Y-S, 224
van Hout, Angeliek, 144
Huesmann, Ilka, 282
Hänel-Faulhaber, Barbara, 217
Hätty, Anna, 371
Häussler, Jana, 247, 248
Ibrahim, Lina Abed, 156
Ihsane, Tabea, 306
Iordachioaia, Gianina, 134
Iovino, Rossella, 348
Irwin, Patricia, 108
Jacob, Daniel, 196
Jamieson, Elyse, 305
Janzen, Sabine, 204
Juzek, Tom, 247, 248
Jónsson, Jóhannes Gísli, 101
Kaiser, Georg, iii
Kaiser, Katharina, iii
Kalbertodt, Janina, 280
Kallel, Amel, 322
Kasper, Simon, 127
Kellert, Olga, 315
van Kemenade, Ans, 77
Khvtisavrishvili, Nana, 371
Kiparsky, Paul, 115
Kirchhoff, Frank, 269, 270, 282
Kiss, Tibor, 116
Klabunde, Ralf, 201, 216
Kleinmann, Achim, iii
Kliche, Fritz, 372
Koeneman, Olaf, 293
423
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 424 — #436
Personenverzeichnis
Koller, Alexander , 205
Kordoni, Valia, 374
Kornher, Svenja, iii, 397
Kotowski, Sven, 343
Koukoulioti, Vasiliki, 135
Krause, Thomas, 367
Krebs, Julia, 238
Kretzschmar, Franziska, 128
Krifka, Manfred, 188
Kuhn, Florian, 212
Kuhn, Jeremy, 243
Kupisch, Tanja, iii
De Kuthy, Kordula, 366
Külpmann, Robert, 250
Lar, Bettina, 376
Larrivée, Pierre, 322
vonLehmden, Frederike, 402
Lemaire, Claire, 377
Lensch, Anke, 141
Levy, Roger, 358
Lieven, Elena V. M., 174
de Lima Júnior, João C., 176
Liu, Jinhong, 144
Lohiniva, Karoliina, 172
Lohnstein, Horst, 180
Lourenço, Guilherme, 233
Luisio, Carolina, 403
Luraghi, Silvia, 329
Lühr, Rosemarie, 326
Lüll, Svenja, 128
Maaß, Wolfgang, 204
Machicao y Priemer, Antonio, 378
Manouilidou, Christina, 113
424
Mantovan, Lara, 232
Marantz, A., 111
Marantz, Alec, 113
Maret, Georg, 395
Marinis, Theodoros, 149, 150
Martin, Fabienne, 131, 144
Matić, Dejan, 193
Matthewson, Lisa, 195
McNally, Louise, 16, 63
Meier, Richard P., 227
Meijer, A. Marlijn, 188
Meir, Irit, 228
Mensching, Guido, 315
Meurers, Detmar, 366
Meyer, Lars, 162
Mitrović, Moreno, 324
Morgan, Emily, 358
Musgrave, Simon, 215
Müller, Anja, 331
Müller, Claudia, 402
Müller, Maike, 389
Müller, Stefan, 378
Neophytou, K., 111
Ney, Anna, iii, 3, 7
Nikolaeva, Irina, 193
Nishida, Chiyo, 264
Nolda, Andreas, 271
Nykiel, Joanna, 262
Odebrecht, Carolin, 367, 384
Öhl, Peter, 182
Oiry, Magda, 337
Oltra-Massuet, I., 111
Padovan, Andrea, 353
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 425 — #437
Personenverzeichnis
Pagliarini, Elena, 154
Paluch, Markus, 379
Panayidou, Fryni, 354
Panizza, Daniele, 172
Panzeri, Francesca, 333
Paolazzi, Caterina, 107
Pappert, Sandra, 120
Paschke, Peter, 364
Pavlič, Matic, 239
Pechmann, Thomas, 120
Penke, Martina, 122, 176
Petrova, Svetlana, 311, 330
Pfau, Roland, 245
Philipp, Markus, 125, 126, 132, 147
Pintér, Lilla, 339
Piringer, Barbara, 380
Poeppel, David, 16, 64
Poletto, Cecilia, 315
Polinsky, Maria, 104
Price, Iya Khelm, 164
Primus, Beatrice, 125, 126, 132, 147,
280
Proske, Nadine, 406
Prévost, Philippe, 156
Quer, Josep, 236
Rathmann, Christian, 217
Rau, Jennifer, 335
Rehn, Alexandra, iii, 393
Reinhardt, Janina, iii, 393
Reißig, Tilo, 279
Renz, Fabian, 277
Repp, Sophie, 89, 188, 190
Reuters, Sabine, 122
Reuße, Sebastian, 201, 216
Rinker, Tanja, 398
Rizzi, Luigi, 150
Robert-Tissot, Aurélia, 124
Roberts, Leah, 186
Roehm, Dietmar, 238
Roeper, Thomas, 98
Roeper, Tom, 168, 335
Rohde, Hannah, 91
Romero, Maribel, iii, 192
Romoli, Jacopo, 338
Rothstein, Björn, 397
Rothweiler, Monika, 176
Ruigendijk, Esther, 158
de Ruiter, Laura E., 174
Ruppenhofer, Josef, 363
Rösiger, Ina, 388
Sacha, Dominik, 391
Salzmann, Martin, 245
Sanfelici, Emanuela, 87, 170
Sano, Kyoko, 197
Santi, Andrea, 107
Santoro, Mirko, 232
Schade, Ulrich, 385
Schalley, Andrea C., 215
Schembri, Adam, 221, 230
Schippers, Ankelien, 248
Schlegel, Jana, iii
Schlesewsky, Matthias, 128
Schlotterbeck, Fabian, 69, 71, 101
Schmidt, Christina, 379
Schmidt, Karsten, 274
Schmitt, Viola, 331
Schouwenaars, Atty, 158
425
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 426 — #438
Personenverzeichnis
Schrinner, Fabian, v
Schrinner, Matthias, 265
Schröter, Juliane, 407
Schulte im Walde, Sabine, 371
Schulz, Petra, 87, 170
Schumacher, Petra, 105
Schumacher, Petra B., 280
Schwarz, Florian, 338
Schweitzer, Katrin, 372
Scontras, Gregory, 342
Seeliger, Heiko, 190
Sekerina, Irina A., 166
Sharpe, V., 111
Sideltsev, Andrei, 327
Siemund, Melanie, 380
Sigwarth, Gloria, iii
Slomanson, Peter, 297
Smith, Garrett, 254
Smolka, Eva, 82
Sorace, Antonella, 259
Spermoser, Benedict, 379
Stark, Linda, 401
Stavrakaki, Stavroula, 135
Stavrou, Melita, 356
Steinbach, Markus, 217
Steiner, Petra, 382
Stevens, Jon, 201, 207, 216
Steyer, Kathrin, 407
Stockall, Linnaea, 113
Strangmann, Iris M., 144
Stumpf, Sören, 407
Su, Yi-ching, 334
Sudhoff, Stefan, 177, 179
Sugisaki, Kyoko, 383
Sulger, Sebastian, iii
426
Sundquist, John, 303
de Swart, Peter, 80
Symanczyk Joppe, Vilma, 250, 276
Szczepaniak, Renata, 386
Szczepek-Reed, Beatrice, 186
Tabor, Whitney, 254
Theakston, Anna L., 174
Þráinsson, Höskuldur, 16, 65
Tieu, Lyn, 338
Tovena, Lucia M., 138
Trabandt, Corinna, 170
Trabandt,Corinna, 87
Trotzke, Andreas, iii, 341, 358
Tuller, Laurice, 156
Turco, Giuseppina, 187
Turgay, Katharina, 283
Turolla, Claudia, 353
Valian, Virginia, 152
van Valin, Robert D., 139
Vasishth, Shravan, 159
Veenstra, Tonjes, 291, 296
Verhoeven, Elisabeth, 39, 103
Verlage, Sarah, 122
Viesel, Yvonne, iii, 393
Vikner, Sten, 73
Villalba , Martín , 205
de Villiers, Jill, 168
Viðarsson, Heimir van der Feest,
301
Vogel, Friedemann, 369
Voigt, Vivian, 384
Wahl, Sabine, 288
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 427 — #439
Personenverzeichnis
Weber, Tilo, 406
Widera, Carmen, iii
Wierzba, Marta, 257
Wiese, Heike, 71
Wilbur, Ronnie, 238
Wildeboer, Myrthe, 352
Wimmer, Eva, 176
Winckel, Elodie, 378
Windhaber, Irina, 376
Wittenberg, Eva, 71, 341, 358
Witzel, Jeffrey, 164
Wochner, Daniela, iii
ten Wolde, Elnora, 347
Wolke, Daniela, iii
Wolski, Magdalena, 385
Woods, Rebecca, 93, 98
Zebib, Rasha, 156
Zehr, Jérémy, 338
Zeijlstra, Hedde, 293
Zepnik,Sabine, 399
Zepter, Alexandra, 399
Ziai, Ramon, 366
Zinsmeister, Heike, 386
Zipser, Florian, 367, 384
Zollmann, Marie, 388
Zymla, Mark-Matthias, 210, 389
427
“DGfS_2016_booklet” — 2016/1/19 — 13:53 — page 428 — #440