Download Daniel Trott East

Research Centre for Japanese Language and Linguistics
University of Oxford
オックスフォード大学 日本語研究センター
www.orinst.ox.ac.uk/research/jap-ling/
Daniel Trott
[email protected]
East Asian Linguistics Seminar
University of Oxford, 5 February 2013
1

Aim:
◦ To show that the OJ -(i)-wor- construction does not
only denote activities, but also result states
◦ To discuss explanations of this multifunctionality

Structure:
◦
◦
◦
◦
1.
2.
3.
4.
Introduction
Previous accounts
Analysis of -(i)-worExplanations
 (i) for resultative-progressive multifunctionality
 (ii) for -(i)-wor- specifically
2
3

Periodization:
◦
◦
◦
◦

Old Japanese (8th century)
Early Middle Japanese (800–1200)
Late Middle Japanese (1200–1600)
Modern Japanese (1600–)
Varieties:
◦ Central (or Western) Old Japanese
◦ Eastern Old Japanese

Main sources:
◦ Man’yōshū (MYS), Kojiki (KK), Nihon shoki (NSK),
Bussokuseki (BS)
4

The Infinitive (連用形) of a verb followed by
the verb wor- ‘be sitting, be still’
◦ e.g. iri-wori, mati-woreba, kwopwi-wora-mu

There are only 54 attestations
◦ Compare: 948 attestations of Perfective -(i)n5
1. UTTERANCE
aspectual construal
2. VERB
aspectual potential
3. CONSTRUCTION
aspectual function(s)
6

An utterance has an aspectual construal
◦ the way an event is construed aspectually by the
speaker
◦ the same event can be construed aspectually in
different ways

Examples:
◦
◦
◦
◦
◦
result state
activity
habitual
achievement (punctual directed change)
accomplishment (durative directed change)
7

In a subjective result state, the subject of an
intransitive verb is construed as having come
to be in a particular state:
8

In an objective result state, the object of a
transitive verb is construed as having come to
be in a particular state:
9

In a possessive result state, the subject is
construed as possessing the object in a
particular state:
10

An undirected activity is an activity that does not
tend towards a qualitative boundary, i.e. one that is
atelic:
11


A directed activity is one that tends towards a
qualitative boundary, i.e. one that is telic:
-yer- and -(i)te ar- cannot express directed
activities (Kinsui 1995: 18)
12

A verb does not have a fixed aspectual type
◦ The same verb may be used in utterances with
different aspectual construals, cf:
 He’s broken his ankle. [result state]
 He broke his ankle. [achievement]

Instead, a verb has aspectual potential:
◦ the particular combination of construals it has in
different aspect constructions (Croft 2012)

Verbs with similar aspectual potential can be
grouped into classes
◦ In Old Japanese, the combinations of verbs with the
major aspect constructions produce five classes:
13
14

Constructions often allow more than one
aspectual construal
◦ E.g., the English Perfect is often analysed as having
four:





current relevance of anterior event
experiential
recent past
anterior continuing
So the first step is to list the aspectual
construals of a construction
◦ Then to consider its function(s)
15

Three ways of analysing multifunctionality
(Haspelmath 2003):
◦ monosemy
◦ polysemy
◦ homonymy

The most reliable way to decide between
these is by psycholinguistic experiment
◦ I will remain uncommitted as to the nature of the
multifunctionality of -yer-, -(i)te ar-, and -(i)-wor-
16

Cross-linguistic comparison can tell us
◦ what combinations of functions are likely
◦ what changes are common

But constructions with the same name in
different languages are different
◦ It is not enough to say ‘This is a perfect’
◦ If cross-linguistic categories exist at all, they are
prototype categories (Dahl 1985; Bybee, Perkins,
and Pagliuca 1994)
17
18

-(i)-wor- is usually called a ‘progressive’, e.g.
by Vovin (2009) and Frellesvig (2010)
◦ It is seen as a construction formed on activity verbs
like the English Progressive be V-ing
◦ This is correct for the examples where it combines
with activity verbs (e.g. katarap- ‘talk’, mat- ‘wait’)

But it is also found with accomplishment,
achievement, and inceptive state verbs
◦ e.g. ir- ‘enter’, sok- ‘become parted’, komor- ‘be
shut up’
◦ In most cases these cannot be interpreted as
activities with -(i)-wor19

According to Watanabe (2008), the nonprogressive examples are lexical
◦ The verb wor- still means ‘be sitting’ or ‘be still’

This means that strings like iri wori must be
interpreted as sequential events, e.g. ‘go in
and be sitting’ or ‘go in and be still’
◦ But a resultative interpretation always makes better
sense than a ‘lexical’ one
20

According to Kinsui (2006)
◦ -(i)-wor- is always progressive
◦ non-activity verbs that appear with -(i)-wor- are
behaving as activity verbs in this construction

But
◦ These verbs (ir- ‘enter’, komor- ‘be shut up’) do
not behave like activity verbs in any other
constructions
 They are directed change or inceptive state verbs
◦ In this construction they do not denote activities
21

Like Stative -yer- and Periphrastic Stative
-(i)te ar-, the -(i)-wor- construction denotes
◦ undirected activities with activity verbs
◦ result states with accomplishment, achievement,
and inceptive state verbs

This is not unheard of, especially in Japanese,
and explains the data
22
23

The following verbs that -(i)-wor- combines
with appear to be activity verbs:
24

When -(i)-wor- combines with activity verbs
it denotes an activity:
25
26

The following verbs that -(i)-wor- combines
with appear to be accomplishment or
achievement verbs:
27

When -(i)-wor- combines with directed
change verbs it denotes a result state:
28

ki-iri-wori, iri-wori, and soki-wori here
denote result states
◦ A lexical interpretation of wori is less natural to the
sense of the poems
29

The following verbs that -(i)-wor- combines
with appear to be inceptive state verbs:
30

When -(i)-wor- combines with inceptive state
verbs it denotes a result state:
31

These verbs can denote both an achievement
and a state
◦ Here they denote states
◦ -(i)-wor- can be interpreted as denoting the result
state of the achievement sense
32

-(i)-wor- should be interpreted as denoting:
◦ activities with activity verbs
◦ result states with achievement, accomplishment,
and inceptive state verbs
33
34

Constructions that denote result states
(resultative constructions) and constructions
that denote activities (progressive
constructions) are both common
◦ But constructions that denote both are rare

The two functions can be united in some
formalizations of aspect
◦ E.g., they denote the rightmost extended temporal
phase of an event
◦ But the question of how they develop diachronically
is harder to answer
35

Ebert (1995) proposes that some of the activity
verbs found in resultative-progressive
constructions began as achievement verbs
◦ From being achievement verbs in resultative
constructions they were reanalysed as activity verbs
in progressive constructions
◦ This caused reanalysis of the construction as
allowing both result state and activity construals
36


She uses the example of nite iru ‘resembles’
◦ which presumably began as an achievement verb in
a resultative construction
But for reanalysis of the construction, tokens of
such verbs must have been frequent
◦ It seems unlikely that many of the OJ activity verbs
began as achievement verbs
37

Ebert (1995) proposes that alternatively
reanalysis could result from the ambiguity of
inceptive state or inceptive activity verbs
◦ So a verb like sak- ‘bloom’ could be reanalysed
from an achievement verb in a resultative
construction to an activity verb in a progressive
construction
◦ Then the construction is reanalysed as denoting
result states and activities, and activity verbs can
begin to be used

But are there enough inceptive state verbs to
induce this reanalysis?
38

Similar to proposal (b)
◦ but with a different verb class

Activity verbs were first used in these
constructions to denote the result state of the
inception of an activity, i.e. the activity’s
occurring
◦ e.g. mati-woru ‘has begun to wait’

Then the construction was reanalysed to
express activities without the implication of
their having begun
◦ e.g. mati-woru ‘is waiting’
39

The ‘inception of activity’ construal of activity verbs
seems to have existed in OJ, especially with
Perfective -(i)n- ~ -(i)te-:
40
41

Sometimes Perfective -(i)n- seems to denote an
undirected activity:
42

These can be seen as the application of the result
state function of Perfective -(i)n- to the ‘inception
of activity’ sense of activity verbs
43

If activity verbs have a conventionalized
‘inception of activity’ sense, this could be
used as input to a resultative construction
◦ Kudō (1983) explains Uwajima dialect -(I)-tor- this
way
◦ He glosses arui-toru as ‘has begun to walk’

This need not be the correct synchronic
analysis of any resultative-progressive
construction
◦ But it is a possible diachronic explanation
44

-(i)-wor- is different from -yer- and -(i)te ar◦ -yer-: 63% result state, 30% activity
◦ -(i)te ar-: 80% result state, 14% activity
◦ -(i)-wor-: 35% result state, 65% activity

Why do -yer- and -(i)te ar- express more result
states than activities, while -(i)-wor- is the other
way round?
45

What are converbs?
◦ ‘a nonfinite verb form whose main function is to mark
adverbial subordination’ (Haspelmath 1995)
◦ e.g. the Gerund and the Infinitive

Converbs play an important role in forming
periphrastic constructions
◦ Anterior converb + stative > result state
◦ Simultaneous converb + stative > activity

Converbs can change their meaning over time
◦ The Japanese Gerund broadened in LMJ from being only
an anterior converb to also expressing temporal
simultaneity (Ohori 1994)
46


The Infinitive shifted from denoting temporal
sequentiality to denoting temporal simultaneity at
some point pre-OJ (Frellesvig 2010)
◦ This made -(i)-wor- more iconic (and popular) as a
progressive than as a resultative construction
◦ -yer- was not affected, since its Infinitive origin was
obscured
◦ -(i)te ar- was not affected, since the Gerund still
denoted temporal sequentiality
-(i)-wor- could originally have been an alternative
to -yer◦ like -(i)-imas-
47


-(i)-wor- could denote both result states and
activities, like -yer- and -(i)te arIt probably began by denoting result states
and expanded to denote activities
◦ Possibly via denoting the ‘result state’ of the
inception of an activity

The changing function of the Infinitive caused
its progressive function to dominate
48







BYBEE, Joan, Revere PERKINS, and William PAGLIUCA. 1994. The Evolution of Grammar:
Tense, Aspect, and Modality in the Languages of the World. Chicago; London: University
of Chicago Press.
CROFT, William. 2012. Verbs: Aspect and Causal Structure. Oxford: Oxford University
Press.
DAHL, Östen. 1985. Tense and Aspect Systems. Oxford: Basil Blackwell.
EBERT, Karen H. 1995. ‘Ambiguous perfect-progressive forms across languages’. In Pier
Marco Bertinetto, Valentina Bianchi, Östen Dahl, and Mario Squartini (eds), Temporal
Reference, Aspect and Actionality, Vol. 2: Typological Perspectives, Turin: Rosenberg &
Sellier, pp. 185–203.
FRELLESVIG, Bjarke. 2010. A History of the Japanese Language. Cambridge: Cambridge
University Press.
HASPELMATH, Martin. 1995. ‘The converb as a cross-linguistically valid category’. In
Martin Haspelmath and Ekkehard König (eds), Converbs in Cross-Linguistic Perspective:
Structure and Meaning of Adverbial Verb Forms – Adverbial Participles, Gerunds –, Berlin;
New York: Mouton de Gruyter, pp. 1–55.
——. 2003. ‘The Geometry of Grammatical Meaning: Semantic Maps and CrossLinguistic Comparison’. In Michael Tomasello (ed.), The New Psychology of Language:
Cognitive and Functional Approaches to Language Structure, Vol. II, Mahwah, NJ:
Lawrence Erlbaum Associates, pp. 211–242.
49






KINSUI Satoshi [金水敏]. 1995. ‘ “Shinkootai” to wa nani ka’ [「進行態」とはなにか, ‘What is
“progressive aspect”?’]. Kokubungaku: kaishaku to kanshoo [国文学 解釈と鑑賞, ‘National
Literature: Interpretation and Appreciation’] 60–7: 14–20.
——. 2006. Nihongo sonzai hyōgen no rekishi (日本語存在表現の歴史, ‘The History of
Existential Expressions in Japanese’). Tokyo: Hituzi Syobo.
KUDŌ Mayumi [工藤真由美]. 1983. ‘Uwajima hoogen no asupekuto’ [宇和島方言のアスペクト
, ‘Aspect in the Uwajima dialect’]. Kokubungaku: kaishaku to kanshoo [国文学 解釈と鑑賞
, ‘National Literature: Interpretation and Appreciation’] 48–6.
OHORI, Toshio. 1994. ‘Diachrony of Clause Linkage: TE and BA in Old through Middle
Japanese’. In William Pagliuca (ed.), Perspectives on Grammaticalization, Amsterdam:
John Benjamins, pp. 135–149.
VOVIN, Alexander. 2009. A Descriptive and Comparative Grammar of Western Old
Japanese, Part 2: Adjectives, Verbs, Adverbs, Conjunctions, Particles, Postpositions.
Folkestone, Kent: Global Oriental.
WATANABE, Kazuha. 2008. ‘Tense and aspect in Old Japanese: Synchronic, diachronic,
and typological perspectives’. Cornell University PhD dissertation.
50
Research Centre for Japanese Language and Linguistics
University of Oxford
オックスフォード大学 日本語研究センター
www.orinst.ox.ac.uk/research/jap-ling/
Daniel Trott
[email protected]
51