ideology,power,and linguistic theory

Ideology,Power,andLinguisticTheory∗GeoffreyK.PullumUniversityofCalifornia,SantaCruzMyaiminthispaperistodiscussanintenselycomplexclusterofinterlinkedconceptsinvolvingdis-tinctionsbetween(i)descriptiveandprescriptivegrammar,(ii)constitutiveandregulativerules,(iii)conservativeandliberalattitudes,and(iv)standardandnon-standarddialects.Icannothopetobecom-prehensive,butIwilltrytobeclear.Sometimesitseemstomethattheissueswearedealingwithherehavebeenmiredinthesamecontroversiesforagoodfortyyears.Othertimesitseemsmorelikeahundredandforty.Therearefewsignsofanyknowledgeaboutgrammardatingfromafter1900havingbecomeknowntoabroadcross-sectionofthegeneralpublicorhavinghadanimpactoneducation.Perhapsitistimetoattempttounderstandthesituationbetter,ratherthansimplytodeploreit.CorrectnessconditionsIbeginbytakingitforgrantedthatthereareconditionswemightcallcorrectnessconditionsfornaturallanguages.(Whethertheyarestandardlanguages,non-standarddialects,orundescribedtriballanguagesofpreliteratepeoplesdoesnotmatter:allhavecorrectnessconditions.)AndIwillalsoassumethatitispossibleinprincipletobeperfectlyexplicitaboutsuchconditions.Intermsofthedistinctiondrawnfamiliarthirty-fiveyearsagobyJohnSearle,1Theyareconstitutive,notregulative.Theydonotregulatetheuseofthelanguage,inthesensethatonecoulduseiteitherinwaysthatcomplyorinwaysthatdon’t;theyconstitutethelanguage,inthesensethatnotrespectingthemamountstonotusingitatallbutdoingsomethingelseinstead.Moderndescriptivelinguiststrytofigureoutfromtheavailableevidencetheprinciplesthatconsti-tutethelanguagebeingdescribed,andtogiveexplicit,andpotentiallyfalsifiable,formulationsofthem.Linguistsoftenmakelittleefforttodistinguishtheconditionsthemselves(thesubjectmatterunderstudy,theconditionsastheytrulyare)fromproposedstatementsoftheconditions(hypothesesaboutthegram-maticalstructureofexpressionsinthelanguage).Theyoftenrelyona‘systematicambiguity’thatletssuchtermsas‘thegrammarofEnglish’refereithertotheconditionsthatactuallydoconstituteEnglishortothelinguist’scurrenteffortatmakingastatementofthem,whicheverthecontextmaycallfor.Butthedifferenceisimportant.Foronething,arbitrarilymanydifferentstronglyequivalentstatementsofthecorrectnessconditionsforEnglishcouldbegiven,butthatdoesn’timplythattherearearbitrarilymanydifferentobjectsofstudy.∗ThisisarevisedversionofthetextonwhichIbasedmypresentationattheannualmeetingoftheModernLanguageAssociation,onDecember30,2004,inPhiladelphia,PA(tobeprecise,theversionofDecember13,2006).Itoriginatedasatalk,andisconsequentlyratherlightonbibliographicalreferences.IamgratefultoHectorTorresforsettingupthesession,toDennisBaron,EricHyman,JohnMorse,andothersatthesessionfortheirinterestingcomments,andtoStuartShieberforpointingoutseveralerrors.Aboveall,IthankmyphilosophyoflinguisticscollaboratorBarbaraC.Scholz,whosuppliedsubstantivecommentaryandorganizationalcriticismsthatledtoacompleteredesignofthepaper,andcontributedseveralkeypoints.1JohnSearle,SpeechActs,CambridgeUniversityPress,1969.1SupposealinguiststatesitasaconditionthatinStandardEnglishanindependentdeclarativeclausebeginningwithapreposednegativeadjunctmusthaveatensedauxiliarybeforethesubject:(1)a.NeverbeforehadIseensuchathing.b.*NeverbeforeIhadseensuchathing.(2)a.Atnotimedidheleavetheroom.b.*Atnotimehelefttheroom.TheclaimbeingmadeisnotthatspeakersofStandardEnglishOUGHTtopositionsubjectsofindependentclausesbeforethetensedauxiliarieswhenthereisnopreposednegativeadjunct,asinthe(a)examples;theclaimisthattheyactuallyDOpositionthemthus(settingasideunintentionalfailuresliketypingoreditingerrorsthatsometimespreventpeoplefromdoingwhattheyintended).ThisofcourseimpliesthatyouwouldbewelladvisedtopositionthemthusifyouwanttoberegardedasusingStandardEnglish;butnooneistellingyouthatyouSHOULDspeakStandardEnglish.Thecorrectnessconditionsofalanguage,bydefinition,providefulljustificationforanytrueclaimtotheeffectthatsomeexpressionisorisnotwellformedinthelanguage.Butit’simportantthatjustificationofaquitedifferentsortisneededforthehigher-levelclaimbyalinguistthatacertainsetofstatementscapturestheRIGHTcorrectnessconditionsforagivenlanguage.Thelinguistcanbewrongaboutwhethersomeproposedstatementofconditionsisaccurateforsomelanguage,evenwhenitisthelanguagethatthelinguistspeaksnatively.Hencethephrase‘potentiallyfalsifiable’above.Itshouldbeobviousthatdifferentdialectshave(atleastslightly)differentcorrectnessconditions.IfanelderlyBritishspeakersaysHaveyouapen?andayoungAmericanspeakercouldneversaythat(butwouldsayDoyouhaveapen?instead),itisnotsensibletoassumethereisjustoneanswertothequestionofwhatisgrammatical,andoneofthetwospeakersiswrong.TheAmericanandBritishvarietiesofEnglishexhibitaveryclosesimilarity,buttheyarenotquitethesame.Theydifferveryslightlyinwhatcorrectnessconditionshold.Thisclaimonitsownisquiteenoughtosparkcontroversyinsomecircles.WilliamLabovpointedouttolinguistsfortyyearsagothatitcanbeapplied,andshouldbeapplied,toAfricanAmericanVer-nacularEnglish(henceforth,AAVE).WhenalinguistsaysthatinAAVEatensedauxiliarymaybeginadeclarativeclauseifthatauxiliaryismarked