* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Download W What`s in a a suffix? The past t tense in problem n Dutch ch ms
Scottish Gaelic grammar wikipedia , lookup
Japanese grammar wikipedia , lookup
Esperanto grammar wikipedia , lookup
Navajo grammar wikipedia , lookup
Modern Hebrew grammar wikipedia , lookup
Lithuanian grammar wikipedia , lookup
Portuguese grammar wikipedia , lookup
Latin syntax wikipedia , lookup
Ancient Greek grammar wikipedia , lookup
English clause syntax wikipedia , lookup
Georgian grammar wikipedia , lookup
Udmurt grammar wikipedia , lookup
Kannada grammar wikipedia , lookup
Spanish grammar wikipedia , lookup
Proto-Indo-European verbs wikipedia , lookup
Ukrainian grammar wikipedia , lookup
Macedonian grammar wikipedia , lookup
Old Norse morphology wikipedia , lookup
Chichewa tenses wikipedia , lookup
Old English grammar wikipedia , lookup
Lexical semantics wikipedia , lookup
Germanic weak verb wikipedia , lookup
Tense–aspect–mood wikipedia , lookup
Kagoshima verb conjugations wikipedia , lookup
Italian grammar wikipedia , lookup
Serbo-Croatian grammar wikipedia , lookup
Sotho verbs wikipedia , lookup
Spanish verbs wikipedia , lookup
Polish grammar wikipedia , lookup
Yiddish grammar wikipedia , lookup
Germanic strong verb wikipedia , lookup
Hungarian verbs wikipedia , lookup
Ancient Greek verbs wikipedia , lookup
Icelandic grammar wikipedia , lookup
Russian grammar wikipedia , lookup
Swedish grammar wikipedia , lookup
English verbs wikipedia , lookup
Grammatical tense wikipedia , lookup
Pipil grammar wikipedia , lookup
271 W What’sina asuffix?Thepastttensein nDutchch hildrenw withread ding problem ms JudithRisspensa,EliseedeBreea,&Annemarie & eKerkhoffb a UniversityofAm msterdam,bb UtrechtUn niversity Abstract: children with devvelopmental dyslexia deemonstrate profound p problems w with the P l difficulties have been hypothesized to underliie those acquisition of liteeracy skills. Phonological morphosyntacctic weaknessses have aalso been deemonstrated d. In this study we problems, but m orphophonolo ogicalprocesssofinflectingregularveerbsforthe pasttenseinwhich examinedthemo nal sound off the verb sttem) and morphology (tthe selection n of the both phonologicaal factors (fin hatchildrenw weresensitivetothe withreadinggproblemsw suffix)areinvolveed.Theresulttsshowedth ment of a past tense (i.e., morphology), but that theyy showed d different conteext requirem morphophonologicalerrorpattternsrelativvetotheirno ormallyreadiingpeers. ntroduction n 1. In blem with developingg fluent Deveelopmental dyslexia is characterized by a ssevere prob readiingdespite normal coggnitive and emotional developmeent (Snowlin ng,2000).TThere is generalconsenssusinthelitteraturethaatchildren withdevelo opmentald dyslexiashowpoor phon nologicalpro ocessing(Veellutinoetaal.,2004).In nadditionto ophonologiicaldeficits,,recent invesstigations h have indicatted subtle problems with inflecctional morrphology, ssuch as subjeectͲverb agrreement (R Rispens, Roeeleven, & Koster, K 200 04; Rispens,, Been, & Zwarts, 2006)andthep pasttense(JJoanisseet al.,2000;R Robertson,JJoanisse,Desroches,& &Terry, 2012).Research onthecom mbinationoffthesetwoelements,morphoͲphonology,islimited hildren with h dyslexia (ssee Marshaall, Ramus, & van der Lely, 2010;; Robertson n et al., in ch 2012),althoughitcanbeasssumedthattdifficultiesswillariseinthisarea.Thepresen ntstudy ologyin chilldrenwithrreading focussesontheiinteraction betweenphonology andmorpho probllemsbystudyingtheproductiono oftheregulaarpasttense. TTheDutchrregularpastttenseisa morphopho onologicalp phenomeno onastheseelection oftheesuffix‘Ͳte’’or‘Ͳde’isguidedby aninteractiionbetweenmorpholo ogyandpho onology (see Rispens&d deBree,20 014).Verbsstemsendin nginanund derlyinglyvoicelessobstruent nestus & Baaayen, 2001 1). For exam mple, the in nfinitive take Ͳte, all other verbs take Ͳde (Ern bakkeen (to bakee), with thee stem endiing in a voiiceless obsttruent (bak), becomes bakͲte 272 (baked)inthepasttense,whereasnoemen(toname)withastemͲfinalsonorantbecomes noemͲde (named). For the acquisition of the Dutch regular past tense, sensitivity to the interactionbetweenmorphologyandphonologyisthusnecessary. Previousexperimentalevidencehasshownthatlexicalfrequencyinfluenceschildren’s past tense marking of regular verbs (e.g. Oetting & Horohov, 1997; Matthews & Theakston,2006). In other words,ahigherfrequencyoftheinflectedformfacilitatesits retrieval.ThisfindinghasalsobeenattestedforDutch:ErnestusandBaayen(2001)found evidence for lexical frequency effects in the productions of past tense forms of Dutch adults.Similarly,RispensanddeBree(2014)foundthatlexicalfrequencyinfluencedpast tenseproductionin7ͲyearͲolds. Inadditiontolexicalfrequency,typefrequencyhasalsobeenshowntoplayarolein development of the production of past tense. Type frequency refers to the number of instancesthatlexicalitemsoccurinacertainpattern(Bybee,2007).Forregularpasttense production, this refers to the frequency of past tenses expressed by the allomorph –te versus –de. Type frequency is assumed to affect morphological productivity, as highly frequent suffixes tend to be applied to newly learned items rather than low frequent suffixes(Bybee,2008).Highlyfrequentpasttensemarkers(types)arethusexpectedtobe most productive. In Dutch, the combination of voiceless consonants + Ͳte is phonotactically more frequent than voiced consonants + Ͳde (Rispens & de Bree, 2014). Rispens and de Bree (2014) found that past tense production was influenced by type frequency based on the frequency of the verb stem + allomorph (Ͳte > Ͳde). Both monolingual TD Dutch 5Ͳ and 7ͲyearͲolds showed an effect of allomorph type on past tense production. The 5ͲyearͲolds showed an effect of type frequency (Ͳte > Ͳde) when theyinflectedlexicalandnovelverbsforthepasttense;the7ͲyearͲoldsonlyshowedthis effect of type frequency for the inflections of novel verbs. Furthermore, for the 7ͲyearͲ olds, but not the 5ͲyearͲolds, the past tense production of lexical regular verbs was influenced by lexical frequency, with verbs high in frequency being inflected more accurately. The absence ofalexicalfrequencyeffectintheyoungergroup suggeststhat vocabularysizeinfluencestheoccurrenceoflexicalfrequencyeffectsinverbproductions. Nexttofrequencyeffects,phonologicalpreferencescanalsoimpactoninflection(see Song, Sundara, & Demuth, 2009; Marshall & van der Lely, 2007). In Dutch, there is evidencethatearlydevelopmentofvoicedsegments/phonemesisdifferentfromthatof voicelessones.Dutchchildrenaroundtheageofthreehavebeenfoundtoshowaclear preferenceforvoicelessovervoicedsegments(Kager,vanderFeest,Fikkert,Kerkhoff& Zamuner, 2007) and acquire voiceless phonemes before voiced ones (e.g. Beers, 1995). Dutch shows final devoicing, and has the tendency of voiceless wordͲinternal clusters 273 (Zonn neveld,1983).TheadvantageandpreferenceeforͲteallo omorphsin Dutchinthestudy byRispensandd deBree(201 14)canalso obeaccounttedforbythisphonolo ogicalpreferrence. hypothesizeed that the phonologiccal deficits that occur in develop pmental It can be h opment of f past tensse morphology, due to the strructural dysleexia affect the develo relatiionship bettween the final f phoneeme of the verb stem and the allomorph itt takes. Robeertson et al. (2012) examined thee productio on of the paast tense in n English sp peaking childrrenwithdeevelopmentaldyslexiaw milartothe Dutchpastttense, whospeak English.Sim theEEnglishpasttenseismo orphophono ologicalinnature,asth heselectionofallomorp phs/t/; /d/; and a /Id/ deepends on tthe phonolo ogical propeerties of th he final pho oneme of th he verb stem (compare baked; kissed, and waited). w Thee children with develo opmental dyslexia d inflecctedregularrverbscorrrectlysignifficantlyless oftenfortthepastten nsecomparedtoa chron nologically ageͲmatche a edcontrolggroup.Productionofth hepastten nse ofnovellwords wasn notsignificaantlydiffereent.Furthermore,perfo ormanceon nbothregularandnoveelverbs wasssimilarforthechildrenwithdevelo opmentaldyslexiaandreadingͲageͲmatchedcontrol childrren. These findings su uggest a d developmen ntal lag compared to the ageͲm matched childrren(Roberttsonetal.,2 2012). TThe literatu ure on morphophonolo ogy and liteeracy in Du utch does n not pinpoin nt clear morp phophonolo ogical diffcu ulties for p poor readerrs. One stu udy has been reporteed with respeect to literaacy and past tense inflection in Dutch. Kerrkhoff, de B Bree, Hoeben and Vreuggdenhil(in press)asseessedallomo orphselection(ͲteorͲͲde)fortheeDutchpassttense specifically for vverbs that can c differ in voicing value in thee infinitive aand stem, ssuch as leven nͲleefde ‘lived’, compaared with blaffenͲblaft b te ‘barked’. There weere no diffeerences betw ween normaal and poor readers o on past tense producction in a spoken con ndition. Howeever,whereeascorrectp pasttenseinflectionoffwordsand dnovelword dsincreased dwhen ortho ographic support wasp provided(ee.g.presentingbothau uditoryand writtenlevven)for theggroupofnorrmallyreadiingchildren n,thisbeneffitwasnotaattestedforragroupoffpoorly readiing children n. These ffindings do o not pointt towards severe mo orphophono ological difficulties in poor readerrs. Note, ho owever, that no read dingͲageͲmaatched grou up was ded. Additionally, a sttudy by dee Bree and Kerkhoff ((2010) did not find m morphoͲ includ phon nological diffficulties in children w with a famillial risk of d dyslexia (i.ee., with a d dyslexic paren nt).Themo orphophono ologicalareaaunderinveestigationw wasvoicing alternation nofthe pluraal.Criticalitemsweren nounsthatd differedinu underlyingvvoicingofthelastconssonant: comp pare handͲhanden ´haands´ and krantͲkranten ´newsp papers´. Th he compariison of voicin ngalternationintheplluraloffiveͲͲyearͲoldch hildrenwith handwithoutafamilialriskof dysleexia showed d that both h groups sh howed a prreference for f nonͲalteernation (i.ee., Ͳten pluraals rather th han Ͳden plurals). Inteerestingly, the t children n at risk for dyslexia sshowed 274 more difficulty in inflecting novel words for the plural, indicative of phonological processingdifficulties. Returning to the current research, it is thus an open question whether or to what extent past tense difficulties will be attested in children with reading difficulties. We therefore examined past tense productions of Dutch speaking children with reading problems1)togaininsightintomorphophonologicalprocessinginchildrenwithreading problems,2)toinvestigatetheeffectsoflexicalfrequencyandtypefrequencyinchildren withdevelopmentaldyslexiarelativetotypicallydevelopingchildrenand3)toaddcrossͲ linguisticdataonmorphophonologicalabilitiesinchildrenwithreadingproblems. 2. Methods Participants Two groups of children participated in this study. ThirtyͲeight children with reading problems (RP) (n=38; 16 girls) between 7;1–10;8 years of age participated. They all attendedregularelementaryschools.Thechildrenwerereferredbyspeechandlanguage therapistsorbyspecialeducationspecialistsworkingattheprimaryschools.Toascertain thereadingproblems,tworeadingtestswereadministered:therealwordtask(RWT;Brus &Voeten,1973)andapseudoͲwordtask(PWT;vandenBos,Spelberg,Scheepstra,&de Vries,1994).IntheRWTthechildisrequiredtoreadoutalistofexistingwordsasquickly and accurately as possible within one minute. The PWT follows the same principle, but uses nonͲwords and takes two minutes. Timed reading tasks were chosen as in the Netherlands speed is judged to be a better indicator of reading development than accuracy alone due to the relative transparency of the orthography (De Jong & van der Leij, 2003). Both tasks are commonly used as part of the diagnosis procedure of developmentaldyslexia.ThechildrenwereclassifiedashavingRPwhentheyscoredmore than1SDbelowthemeanofthecombinedscoresoftheRWTandthePWT(amaximum score of 6). All children who participated in this study within the group of RP met this criteria. Children who had a history of speech output problems, such as dyspraxia, or showedevidenceofsuchproblemsatthemomentoftestingwereexcluded.Allchildren hadnormalhearingandnormalorcorrectedͲtoͲnormal,vision.Notethattheparticipants inthisgroupwerenotallofficiallydiagnosedwithdevelopmentaldyslexia;somechildren werestillintheprocessofbeingofficiallydiagnosed. The second group consist of control children without RP who were matched on chronological age with the group of children withRP (n=41, 26 girls) between 7;8Ͳ11;6 participated.Thesameexclusionarycriteriawereappliedtothecontrolgroup.Allchildren 275 hadaameanscoreontheco ombinedsccoreofthe PWTandRW WTofatleaast7,indicaatingat least averagereeadingabilitty.Anindep pendentttestdemonsstratedthe twoexperiimental psdidnotd differwithreespecttoth heirage,t(77)=.99,p==.32. group Taskss Past tenseprodu uctionwas investigatedthrougheelicitationo ofpasttenseeformsofeexisting Rispensand ddeBree(2014). verbssandnovelverbs,seeR Task 1:Pasttensseproductio onofexistin ngverbs.Tw welveregularverbsweereselected dbased wtokenfreq quencycountsoftheC CELEX(Baayyen,Piepen nbrock,&vaanRijn, onhiighandlow 1993) database. The verbss were furtthermore cclassified acccording to the type o of past morph with which thee verbs aree inflected. Half of th he verb steems ended d in an allom undeerlying voiceeless obstru uent (four times t /k/ an nd twice /s/), the otheer half in a voiced conso onant(twicee/l/,twice/r/andtwice/w/).Theesestemsw werechosen nastheyrendered unam mbiguous Ͳtte or Ͳde eexpectations. All verbs in the in nfinitival fo orm had a CVCVC structure. Task2:Pasttensseproductio onofnovelverbs.Fifteeennovelveerbswereprresented.Seevenof ndeight thevverbsendedinanunderlyingvoiceelessobstrueent(4timess‘p’and3ttimes‘k’)an in a voiced conssonant (4 times t ‘m’, twice t ‘n’ an nd twice ‘r’). Only verb b stems en nding in onantsthattdemandunambiguouslyaͲteor Ͳdesuffixw werechosen.Theverb bswere conso contrrolledforph honotacticffrequencyb byusingthedatabaseo ofAdriaans((2006). Proceedure Thecchildrenweereshownaapicturein aPowerPointpresentationona screenofalaptop comp puter that d depicted an n action (seee also Rispens & de B Bree, 2014). Children h heard a prereecordedfem malevoicep pronouncetthetarget. Picturesweerepresenteedtothech hildren, one at a time, and they were w accom mpanied by a little sto ory during which the picture n in the remaained on the screen. TThis three liine story first describeed what waas going on picture and then stated th hat characteers perform med this acttion often or every daay. The henaskedto ofinishthefinalsenten nceinwhich htheywereepromptedtogive childrrenwereth apassttenseform m. TThe responses of the children w were record ded and traanscribed o onͲline and scored afterw wards.Anitemwould bescored ascorrectifthechoiceeofthepasstͲtensesufffixwas correect. Violatio ons of agreeement betw ween the veerb and thee subject weere not taken into accou unt.Theord derofthep presentationofthetw wotaskswaspseudoͲraandomized: halfof 276 the children were first presented with task 2 (novel verbs) followed by task 1 (lexical verbs),theotherhalftheotherwayaround. 3. Results Regularlexicalverbs Table1showsthecorrectpasttenseproductionsofthetwogroups.Theresultsindicate that the children with RP have an average accuracy performance of 90% of past tense productionscorrect.Thecontrolchildrenproducedonaverage85%correctpasttenses.A repeatedmeasuresANOVAwithlexicalfrequency(high/low)andtypefrequency(Ͳte/Ͳde) as the withinͲsubjects variables and group (RPͲcontrol) as the betweenͲsubject variable wascarriedout.Nosignificantmaineffectsorinteractionswerefound(lexicalfrequency: p=.91;typefrequency:p=.1;group:p=.38;allinteractionsp>.33). Table1:Meanproportionscorescorrect(standarddeviationbetweenparentheses)ofthe pasttenseproductionsofthelexicalandnovelverbs Lexicalverbs Group HF–te LF–te HF–de LF–de RP .87(.30) .89(.27) .92(.20) .93(.21) Control .84(.28) .84(.27) .89(.24) .86(.29) Novelverbs Ͳte Ͳde RP .75(.29) .78(.27) Control .84(.29) .69(.41) Novelverbs TheaccuracyscoresonthepasttenseproductionofthenovelverbsaredisplayedinTable 1. A repeated measures ANOVA with type frequency (Ͳte /Ͳde) as the withinͲsubjects variable and group (RPͲcontrol) as the betweenͲsubject variable was carried out. The resultsshowedthatthattherewasnomaineffectoftypefrequency(p=.14),noramain effect of group (p = .93), but that there was a significant interaction between type frequency×group,F(1,73)=4.36,p=.04.Toinvestigatetheinteractionoftypefrequency × group, post hoc analyses were carried out. A paired samples tͲtest showed that the childrenwithRPdidnotdiffersignificantlyinproducingthepasttenseofnovelverbswith –te and –de, t(37) = .44, p =.66. The control children, in contrast, showed a significant 277 advantageforprroducingpaasttensesofnovelverb bsdemandingthe–tessuffix,t(36) =2.45, 019. p=.0 B Bothgroupssproduced around25% %errorson nthenovel verbtask. Toinvestigatethe errorr patternso of thechildrren,weuseeda qualitativeerroraanalysis.Fivve categorieeswere creatted based o on the actual occurren nce of errorrs (see Tablle 2, and Rispens & de Bree, 2014). For our research qu uestion, wee were partticularly interested in the ‘wrongg suffix’ morphophon nology. Responses that could errorr category aas this relaates to senssitivity to m notb beclassified dasoneof ourdefined derrortypeeswereassiignedtotheecategory ‘other’. This means thatt all respon nses were aassigned to a categoryy as it neveer occurred that a ondedwithsilence.RessultsarepreesentedinTTable2. particcipantrespo more suffixx errors TThe error analysis show ws that thee poor read ders made somewhat s than thenormalreaders.C Closerinspeectionindicaatesadifferrencebetw weenthete andde verbssforthecon ntrolchildreenbutnotfforthechild drenwithRP.ChildrenwithRPpro oducea wron ngsuffix(e.gg.vupdeinssteadofvup pte(11.65% %)anddaantteinsteado ofdaande(1 12.5%)) in thee novel verrbs, whereaas the contrrol children mostly pro oduce a wro ong suffix in n verbs demaanding Ͳde (daante insstead of da aande 17,2% % of errorss versus 2,5 5% errors ssuch as vupdeinsteadoffvupte). TThiserrorpatternthus matchesth hesignifican nteffectof typefrequeencyinthe control childrren:novelvverbsdemandingaͲte suffixhaveanadvantaageoverveerbsdemand dingaͲ desu uffix,andthispreferencceisalsoshowninthetypeoferro orsproduceed(Ͳde>Ͳteeoccurs moreefrequentlyythanͲte>Ͳde).TheRP Pchildrend didnotdemonstratethesamefaciilitation forveerbstakingaͲtesuffix,andtheirerrorsdidno otindicateaapreferenceeforeithersuffix. a their occurrences o in percenttages (all eerrors + acccurate Tablee 2. Error patterns and respo onses=100% %)intherea alizationsoffthetwogrroups.Stand darddeviatiionsinbracckets. Errorrcategories Exampleofnovelverb Norm malreaders Poorreaders Ͳen((infinitive) vuppen 5%(16%) 4%(10%) Zeromarking vup 0,2%(0,1%) 0%(0%) –tsu uffix(presentttense) vupt 2,4%(7,5%) 1,2%(3,4%) Incorrrectsuffix vupde(insteadoffvupte) 9,8%(15,5%) 12,1%(14,8% %) Otheer: Vario oustypesoferrors Examples: 6,9%(18.25%) a).vverb‘go’+verrb:(gingvuppen) b). change within the verb stem (e.gg.vumde) c).infinitive+‘Ͳdee’(vuppende) 6,7%(14,3% %) 278 4. Discussionandconclusions ThepresentstudyaimedtoinvestigatethepasttenseinchildrenwithRPtogaininsight into morphophonological processing, to investigate the effects of lexical frequency and type frequency in children with RP, and to add crossͲlinguistic data to studies of morphophonologicalabilitiesinchildrenwithRP. Tostartwiththeoverallresearchaim–examiningthepasttenseinchildrenwithRPͲ, theresultsofthepresentstudyindicatedthattheproductionofthepasttenseinchildren withRPiscomparablewiththatofchildrenwithoutRPmatchedonthesameage.Inline withpreviousfindings(e.g.deBree&Kerkhoff,2010;Joanisseetal.,2000;Kerkhoffetal., in press), this study does not show clear evidence of morphophonological difficulties in poor readers. For real verbs, the group of normal and poor readers do not differ in percentagescorrectinflectionandinfluenceoflexicalfrequencyandtypefrequency.For novel verbs, an interaction between type frequency and group was attested. This interactionreflectsthatfactthatonlythenormalreadersshowedapreferencefortheͲte suffix.Incontrasttootherstudies,noeffectsoflexicalfrequencywereobservedineither group.Apotentialexplanationcouldbethatthesechildrenwereolderthanchildrenthat participatedinpreviousexperiments,andthatthelexicaleffectsarenolongervisibleat thislaterage. TheliteratureonDutchpasttenseinflectionreportsdifferentpatternsrelatedto–te and–depreferences.WhereasRispensanddeBree (2014)founda–tepreferencefor5 and 7ͲyearͲold typically developing children, Kerkhoff et al. (in press) did not find a preference in either direction for 9ͲyearͲold normal and poor readers. In the present study, the group of poor reading children did not show a preference on novel verbs for eithersuffix,andthenormalreadersa–tepreference,thusaddingtothemixedfindings. Whereas the normal readers seem to adhere to phonological patterns of voicelessness withinclusters(Kageretal.,2007;Zonneveld,1983)andahigherfrequencyofconsonant +Ͳtecomparedtoconsonant+Ͳde(Rispens&deBree,2014),thegroupofpoorreaders does not, at least for novel verbs. Also, when the children with RP produced the wrong suffix,theydidnotshowapreferenceforthe–tesuffix,whereasthiswasthecaseforthe controlchildren(presenceofdaanteerrors,butvirtuallyabsenceofvupdeerrors) A highly tentative suggestion is that the –de realizations in contexts where a /t/ is required,arecausedbyspellinginstruction.GiventhatdevoicingoccursinDutch,children heara/t/butareofteninstructedtowritea‘d’,asinnounshand´hand´,mond´mouth´, bed´bed´andpastparticiples,suchasgezegd´said´andgehoord´heard´.Fornovelverbs, no representation is available. Through a strategy, based on orthography, the children mightthusoptfora–deallomorph(seeKerkhoff,deBree&Buesseler,inpress,onvoicing 279 inspokenandw writtenpluralnounpro oduction).C Childrenwitthreadingp problemsm mayuse orethanchiildrenwitho outreading problems, astheyofteenhavehadmore thissstrategymo andm moreexplicitspellingsstrategiesto ohelpthem movercome theiroftenoccurringsspelling probllems.ChildrrenwithouttRPwillthereforenotu usethisstraategy(changgea/t/into oa/d/), butrratherwillu useaphono ologicalprin nciplewhen ntheyhave toinflectn novelverbs forthe pastttense:optfforthephon nologicallylessmarkedͲteinDutch. Ackno owledgemeents: Wew wouldliketo dgeFransZw wartsforbeeingthepro omotoroftthefirst oofcourseeacknowled autho or (Judith R Rispens), an nd for bein ng the direcctor of the ‘Dutch Dysslexia Progrram’ in which hthesecon ndauthor(EElisedeBreee)wasalso oinvolvedaasaPh.D.sttudent.Thaankyou Franss! 5.Reeferences Adriaaans,F.(200 06).PhonottacTools(Teestversion)).[Computeerprogram]].UtrechtIn nstitute o ofLinguisticssOTS,UtrecchtUniversity,theNetherlands. Baayeen,R.H.,Piiepenbrock,,R.&VanR Rijn,H.(1993).TheCELEXlexicald database.CD DͲROM. P Philadelphiaa,PA. Brus,B.Th.,&Voeten,M.J.M.(1973)..EenͲMinuu utTest.Vorm mAenB.Nijmegen:Beerkhout TTestmateriaaal. deBrree, E. &Keerkhoff,A. (2010).Bem mpenor bemben:Diffeerences bettweenchild drenatͲ riskofdyslexxiaandchildrenwithSSLIonamorphoͲphono ologicaltaskk.Scientific Studies o ofReading,1 14,85Ͳ109. Bybee, J. (2007). Frequenccy of use a and the org ganization o of language. Oxford: Oxford U UniversityPress. Bybee,J. (2008)).UsageͲbassedgrammar andseco ondlanguaggeacquisitio on.InP.Ro obinson aand N. Ellis (eds.), Handbook of Cognittive Linguisstics and SSecond Lan nguage A Acquisition.NewYork:Routledge.216Ͳ236. Ernesstus,M.,&Baayen,H.(2001).Cho oosingbetweentheDuttchpastten nsesuffixesͲteand Ͳde.LinguistticsintheNeetherlands, 77Ͳ88. R.,Keating, P.,&Seiden nberg,M.SS.(2000).Laanguagedeficitsin Joaniisse,M.F., Manis,F.R d dyslexic children: Speeech perceeption, pho onology, and morpho ology. Jourrnal of eexperimenta alchildpsycchology,77((1),30Ͳ60. 280 Jong,P.F.,de.,&vanderLeij,A.(2003).Developmentalchangesinthemanifestationofa phonologicaldeficitindyslexicchildrenlearningtoreadaregularorthography.Journal ofEducationalPsychology,95,22Ͳ40. Kager,R.,vanderFeest,S.,Fikkert,P.,Kerkhoff,A.&Zamuner,T.(2007).Representations of [voice]. Evidence from acquisition. In: J. van de Weijer, E. J., van der Torre (eds). VoicinginDutch:(de)voicingphonology,phonetics,andpsycholinguistics.Amsterdam: JohnBenjamins. Kerkhoff, A., de Bree, E., Hoeben, I., and Vreugdenhil, A. (in press). De invloed van orthografie op verleden tijdsvorming door zwakke lezers. Stem,ͲSpraak en Taalpathologie. Kerkhoff, A., de Bree, E., Buessler, H.A. (in press). Saying ‘t’ and writing ’d’: The case of alternatingplurals.FestschriftforWimZonneveld. Marshall, C., & van der Lely, H. (2007). The impact of phonological complexity on past tense inflection in children with GrammaticalͲSLI. Advances in SpeechͲLanguage Pathology,9,191Ͳ203. Marshall,C.R.,Ramus,F.&vanderLely,H.K.J.(2011)'Dochildrenwithdyslexiaand/or SLI compensate for place assimilation? Insight into phonological grammar and representations.',CognitiveNeuropsychology27,563Ͳ586. Matthews,D.E. & Theakston, A. (2006). Errors of omission in EnglishͲspeaking children’s production of plurals and the past tense: The effects of frequency, phonology, and competition.CognitiveScience,30,1027Ͳ1052. Oetting,J.&Horohov,J.(1997).PastͲtensemarkingbychildrenwithandwithoutspecific languageimpairment.JournalofSpeech,LanguageandHearingResearch,40,62Ͳ74. Rispens, J., & Been, P. (2007). Subject–verb agreement and phonological processing in developmental dyslexia and specific language impairment (SLI): a closer look. InternationalJournalofLanguage&CommunicationDisorders,42,293Ͳ305. Rispens, J.E., Roeleven, S., & Koster, C. (2004). Sensitivity to subjectͲverb agreement in spokenlanguageinchildrenwithdevelopmentaldyslexia.JournalofNeurolinguistics, 17,333Ͳ347. Rispens,J.E.,&deBree,E.(2014).PasttenseproductivityinDutchchildrenwithSLI:The roleofphonologyandfrequency.JournalofChildLanguage,41,200Ͳ225. Robertson, E. K., Joanisse, M. F., Desroches, A. S., & Terry, A. (2013). PastͲtense morphology and phonological deficits in children with dyslexia and children with languageimpairment.Journaloflearningdisabilities,46(3),230Ͳ240. Snowling,M.J.(2000).Dyslexia.2ndedition.Oxford:Blackwell 281 Song,, J. Y., Sun ndara, M., & Demuth h, K. (2009 9). Effects of phonolo ogy on children’s rd p production of o English 3 person singular s –s. Journal of Speech, Lan nguage & Hearing H R Research,52 2,623Ͳ642. van d den Bos, K.P., Spelbergg, H.C., Sch heepstra, A.J.M., & De Vries, J.R. (1994). De Klepel. N Nijmegen:B BerkhoutTestmateriaall. Vellutino,F.R., Fletcher,J. M.,Snowliing,M.J.,& &Scanlon,D D.M.(2004 4).Specificrreading d disability(dyyslexia):Wh hathaveweelearnedin nthepastfo ourdecadess?.Journal ofchild p psychologya andpsychia atry,45(1),2 2Ͳ40.