Linguistics for laypeople: inflection vs derivation - Quirky Case

Inflection is often defined as a type of affix that distinguishes grammatical forms of the same lexeme. When we talk of lexemes in linguistics ... Skiptoprimarycontent Postnavigation ←Previous Next→ Source:http://artsfaculty.auckland.ac.nz/courses/online/?P=6550 InthestudyofMorphology,whichisconcernedwiththestructureofwords,therehastraditionallybeenadistinctiondrawnbetweentwotypesofaffixes,inflectionalandderivational.AnaffixisbasicallywhatyourtraditionalLatinorGermangrammarswouldhavecalledan‘ending’,thoughthetermismoregeneral,asitcanrefertobitsofwordsthatcomeatthebeginning(aprefix),orinthemiddle(aninfix)orattheend(suffix)etc. Inflectionisoftendefinedasatypeofaffixthatdistinguishesgrammaticalformsofthesamelexeme.Whenwetalkoflexemesinlinguisticswe’reusuallyreferringtothefactthattherearesomewordformsthatdifferonlyintheirinflectionalproperties.So go and wentaredifferentwordforms,buttheybelongtothesamelexeme,whereasgoand walk belongtodifferentlexemes.Withthatinmind,let’sturntoanexampleofinflection.TheEnglishpluralsuffix -sin book-sisaninflectionalsuffixbecauseit distinguishesthepluralform booksfromthesingularform book.Booksand bookarethusdifferentgrammaticalformsofthesamelexeme. Derivationreferstoanaffixthatindicatesachangeofgrammaticalcategory.Takeforexampletheword person-al. Thesuffix -aldoesnotdistinguishbetweengrammaticalformsofthesamelexeme: personand personalaredifferentlexemes,and personalbelongstoadifferentwordclass(i.e.itisanadjective)from person(whichisobviouslyanoun). That’sallwellandgood,butunfortunatelythingsdon’tstopthere.Oncloserinspectionitbecomesclearthattherearesignificantproblemswiththeabovedefinitions.First,theycomewiththeoreticalassumptions,thatis,anaprioridistinctionbetweenlexemesandwordforms.Therearetheoreticalimplicationshere,aslexemesareconsideredtobethoselinguistictokenswhicharestoredindividuallyineachperson’slexiconor‘mentaldictionary’,whereasanythingtodowithgrammaristraditionallyconsiderednottobestoredthere.Moreproblematic,however,isthatmanyaffixescannotneatlybeidentifiedaseitherinflectionorderivation.Someseemmoreinflection-likethanothersbuthavederivation-likequalitiestoo,andviceversa.Thisisproblematicforpeoplewhobelieveinadichotomous dualmechanismmodel,i.e.whothink thatgrammaticalinformationandlexicalinformationarestoredinseparatecomponentsoftheoverallgrammar. Haspelmath(2002)discussesseveralmoredistinctionsbetweeninflectionandderivation,buildingonthenarrowdefinitionsgivenabove.Hegroupsthedistinctionsintotwocategories, ‘all-or-nothing’ and ‘more-or-less’criteria.Thatis,inhisview,the‘all-or-nothing’criteriaunambiguouslydistinguishinflectionfromderivation,whereasthe‘more-or-less’dosotoalesserextent.Iwon’tgothrougheverycriterionasthatwouldbetedious,butyou’llsoongetasensethatthereareproblemswithprettymuchallofthem. Hisfirst‘all-or-nothing’criterionisbasicallytheoneweusedtodefineourtermsatthebeginning:derivationindicatesachangeofcategory,whereasinflectiondoesnot.However,considertheGermanpastparticiple gesungen,(‘sung’).Onfirstglancethisseemstobeanexampleofbog-standardinflection,Thecircumfix ge--en indicatesthat gesungenisadifferentgrammaticalformofthelexeme singen (‘tosing’)from,say, singst(‘you(sg)sing’).Theyareallthesamecategory,however,astheyareallverbs.However, gesungencan changecategorywhenitfunctionsasanattributiveadjective,asin(1): 1.Eingesungen-esLied Asing.PP-NOMsong ‘Asongthatissung’ Inthiscase,then,anexampleofwhatappearstobeinflectioncanalsochangecategory. Haspelmath’s(2002)thirdcriterionisthatof obligatoriness.Thesayinggoesthatinflectionis‘obligatory’,butderivationisnot.Forexample,in(2),therightkindofinflectionmustbepresentforthesentencetobegrammatical: (2)Theyhave*sing/*sings/*sang/sung. Bycontrast,derivationisneverobligatoryinthissense,andisdeterminedbysyntacticcontext.However,someexamplesofinflectionarenotobligatoryinthesensedescribedaboveeither.Forexample,theconceptofnumberisultimatelythespeaker’schoice:shecandecidewhethershewishestouttertheform bookor booksbasedonthediscoursecontext.Becauseofthis,Booij(1996)distinguishesbetweentwotypesofinflection, inherentand contextual.Inherentinflectionisthekindofinflectionwhichisdeterminedbytheinformationaspeakerwishestoconvey,liketheconceptofnumber.Contextualinflectionisdeterminedbythesyntacticcontext,asin(2).Keepthisdistinctioninmind,we’llcomebacktoit! Inaddition,thereareproblemswithallofHaspelmath’s(2002)further‘more-or-less’criteria.I’lltakethreeofthemhere,butI’llcoverthemquickly. i.Inflectionisfoundfurtherfromthebasethanderivation Example:in personalitieswehavethebase person,thenthederivationalsuffixes -aland -itybeforewegettheinflectionalsuffix -s.Youdon’tget,e.g.*person-s-al-ity Problem: Affect-ed-nesshastheoppositeordering(i.e.inflectionalsuffix -edisclosertothebasethanthederivationalsuffix -ness). ii.Inflectionalformssharethesameconceptasthebase,derivationalformsdonot. Example: person-shassameconceptas person,but person-aldoesnot. Problem: It’sveryvague!Whatisa‘concept’?WhataboutexampleslikeGerman Kerl-chen(‘littletyke’)?-chenisusuallyconsideredtobeaninflectionalsuffix,but Kerldoesn’tmean‘tyke’,itmeans‘bloke’.Thereissurelyachangeinconcepthere? iii.Inflectionis semanticallymoreregular(i.e.lessidiomatic)thanderivation. Example: inflectionalsuffixeslike -s and-ed indicateobvioussemanticcontentlike‘plural’and ‘pasttense’,butit’snotalwaysclearwhatderivationalsuffixeslike -alactuallyrepresentsemantically.Derivation,suchasintheRussian dnev-nik(‘diary’,lit.‘day-book’)ismoreidiomaticinmeaning(i.e.youcan’tworkoutitsmeaningfromthesumofitsparts). Problem: Whataboutinflectionalformslike sand-s,whichisidiomaticinmeaning?(i.e. sandsdoesnotequatewiththepluralof sandinthesamewaythat booksdoeswith book.) So,whydoesthismatter?Ialludedtotheproblemabove.Basically,manylinguists(e.g.Perlmutter(1988))arekeentoholdtoadichotomousapproachtogrammaticalandlexicalcomponentsintermsofhowlinguisticinformationisstoredinthebrain.Theywantinflectionandderivationtobedistinctinaspeaker’slinguisticcompetenceinaccordancewiththedualmechanismmodel,withderivationoccurringinthelexiconandinflectionoccurringsubsequenttosyntacticoperations.Butthenaturallanguagedataseemtoindicatethatthedistinctionbetweeninflectionandderivationissomewhatfuzzier. Sohowdopeoplegetaroundit?Thereareseveralways,butI’lloutlinetwoofthemhere.Thefirstisknownasthe Continuumapproach,advancedbyscholarssuchasBybee(1985).Asthenamesuggests,thisapproachentailsthatthereisacontinuumbetweeninflectionandderivation.Takealookatthefollowingtable,adaptedfromHaspelmath(2002:79)(sorryit’ssosmall): Inthedescendingrows,thedifferenttypesofinflectional/derivationalaffixescanbeplacedinanorderaccordingtohowprototypicallyinflectionalorderivationaltheyare.Forexample,the-spluralsuffixisprototypicallymoreinflectionalthantheGermandiminutivesuffix -chen. Butthisapproachcan’taccountfortheorderpreferenceofbase-derivation-inflection,whichisoneofthepropertieswediscussedabove.Inaddition,itcarrieswithitgreattheoreticalimplications,namelythatthegrammarandthelexiconformacontinuum.Thisisnottheplacetogetintothisdebate,butIthinktherearegoodreasonsforkeepingthetwodistinct. Booij(1996;2007)comesupwitha tri-partiteapproach togetaroundthisproblem,anditgoesbacktothedistinctionmadeabovebetweeninherentandcontextualinflection.Hisapproachisneat,becauseitattemptstoaccountforthefuzzinessoftheinflection/derivationboundarywhilemaintainingadistinctionbetweenthegrammarandthelexicon.Bydividinginflection/derivationphenomenaintothreeratherthantwo(soderivationplusthetwodifferenttypesofinflection),wecanaccountforsomeoftheproblematicphenomenawediscussedabove.Forexample,‘inherent’inflectioncanaccountforlackofobligatorinessininflectionwhenthisoccurs,aswellasaccountingfortheoccasionalbase-inflection-derivationorder,whenthatoccurs.‘Contextual’inflectiontakescareofobligatoryinflectionandtheusualorderingofbase-derivation-inflection. There’smoretobesaidonthis:canBooij’stripartiteapproachreallyexplainwhy,forexample,theorderingbase-derivation-inflectionissomuchmorecommonthantheotherordering?WhatabouttheproblemswithinflectionthatcanchangecategorysuchasineingesungenesLied?Nevertheless,we’veseenthat asharpdistinctionbetweeninflectionandderivationcannotbedrawn,whichhasconsequencesforadichotomyapproachtothegrammar.ThisdichotomycanbemaintainedifwefollowBooij’sdistinctionofcontextualversusinherentinflection. References Booij,G.1996.Inherentversuscontextualinflectionandthesplitmorphologyhypothesis, YearbookofMorphology1995,1-16. Booij,G.2007.TheGrammarofWords.AnIntroductiontoMorphology.Oxford:OUP. Bybee,J.1985.Morphology.TheRelationbetweenFormandMeaning.Benjamins: Amsterdam. Haspelmath,M.2002.UnderstandingMorphology.London:Arnold. Perlmutter,D.M.1988.Thesplitmorphologyhypothesis:evidencefromYiddish,inM. Hammond&M.Noonen(eds),TheoreticalMorphology.SanDiego,CA:AcademicPress, 79-100. Sharethis:ShareFacebookLikethis:LikeLoading... Related 1thoughton“Linguisticsforlaypeople:inflectionvs derivation” Pingback:Linguisticsforlaypeople:languagecontact|QuirkyCase LeaveaReplyCancelreply Enteryourcommenthere... Fillinyourdetailsbeloworclickanicontologin: Email(required)(Addressnevermadepublic) Name(required) Website YouarecommentingusingyourWordPress.comaccount. ( Log Out /  Change ) YouarecommentingusingyourTwitteraccount. ( Log Out /  Change ) YouarecommentingusingyourFacebookaccount. ( Log Out /  Change ) Cancel Connectingto%s Notifymeofnewcommentsviaemail.Notifymeofnewpostsviaemail. Δ Privacy&Cookies:Thissiteusescookies.Bycontinuingtousethiswebsite,youagreetotheiruse. Tofindoutmore,includinghowtocontrolcookies,seehere: CookiePolicy Follow Following QuirkyCase Join870otherfollowers Signmeup AlreadyhaveaWordPress.comaccount?Loginnow. QuirkyCase Customize Follow Following Signup Login Copyshortlink Reportthiscontent ViewpostinReader Managesubscriptions Collapsethisbar   LoadingComments...   WriteaComment... Email(Required) Name(Required) Website %dbloggerslikethis: