convergence in blue collar columbus, oh aave and eae vowel systems? 1

161

7. COnVerGenCe in bLue- COLLar COLumbus, OhiO,

afriCan ameriCan and whiTe VOweL sYsTems?

daviddurian robindodsworth OhioStateUniversity NorthCarolinaStateUniversity

jenniferschumacherOhioStateUniversity

The types and extentofregionalphoneticandphonologicalvariationinafricanamericanenglish(aae)arelargelyunknownand understudied, despite sociolinguists’ detailed knowledge ofaaemorphosyntax(baileyandthomas1998).systematiccompar-ativestudiesofregionalaavevarietiesalsoremainrare(thoughseethomas2001,2007b;Labov,ash,andboberg2006).additionally,therearerelativelyfewinstrumentalcomparisonsofaaetolocalpredominantvarieties(butseebaileyandthomas1998;wolframandthomas2002;fridlandandbartlett2006).becauseoftherel-ativelackofresearchinthisarea,thesequestionsremainopen:(1)whatarethephoneticsimilaritiesanddifferencesamongregionalaaevarieties?(2)inwhatwaysdoregionalaaevarietiescomparephonetically with corresponding local and regional varieties? asfought(2006,60)notes,ifregionaldifferentiationofaaefollowsthepatternsofotherenglishdialects,thenthereshouldbegreatervariabilityinphoneticfeaturesthaninmorphosyntacticones.

columbus, ohio, a metropolis located in the heart of thenorth american midland, provides an informative context forexploringtherelationshipbetweenethnicityandvocalicvariation.asofthe2000u.s.census,columbushasapopulationof711,470residents in the columbus metropolitan statistical area. amongthepopulation,24.5%areafricanamericanand67.9%arewhite

162 pads 94: aae speakers & local sound changes

(u.s.censusbureau2000).intheurbancore,thereis frequentcontactbetweenblue-collarwhitesandblue-collarafricanameri-cansbecauseofmigrationpatternsamongbothethnicitiesduringthelatenineteenthcenturyandtwentiethcentury.asdiscussedinmurphy(1970)andbryant(1983),asignificantnumberofafricanamericansmovedtothesouthernandeasternareasoftheurbancoreofcolumbustopursueindustrial jobsinfactoriesfollowingthecivilwarandagaininthepost–worldwariandworldwariiperiods.somemigrateddirectlyfromthesouth(inparticular,virginia,northcarolina,southcarolina,Georgia,Kentucky,ten-nessee,alabama,andwestvirginia),whileothersmovedfirst toeasterncitiessuchasPhiladelphiaandPittsburghandlaterreset-tled incolumbus. inmost cases,africanamericansmigrated toareasintheurbancore,wheretheyfoundthemselvesindailycon-tactwithrecentwhitemigrantsofpredominantlyuppersouthern,Lowernorthern,andappalachianbackgrounds, aswell as long-timecolumbusresidents,whosefamilieshadbegunsettlinginthecolumbusareaintheearly1800s(Lentz2003).

until the 1970s, most african americans in columbus wereworking class as a result of columbus being a predominantly“separatebutequal”community,essentiallysincethefoundingofcolumbusin1812( jacobs1994).sincetheendofthecivilwar,the“separatebutequal”policyledtodecadesofdiscriminationinhiringpracticesbylocalbusinesses,aswellashousingsegregation,resultingfromrestrictivedeedcovenantsandthedisplacementofmembersof theafricanamerican communityduring the1960sdue to the construction of interstates 70 and 71 (oriedo 1982;burgess 1994). these factors prevented african americans fromobtaininghigherskilledlaborpositionsinthecommunity,eitherphysically,duetogeographicdistance,orsocially,duetojobacces-sibilitylimitations.

duringthelate1960s,however,thesituationbegantochangeasaresultofthepassageofthecivilrightsactof1968,whichputanendtotheenforcementofoverthousingandemploymentdis-criminationpracticesinthecommunity( jacobs1994).inconjunc-tionwiththesechanges,thecolumbuspublicschoolsunderwentchangesfroma“separatebutequal”systemtoonethatwas,atfirst,

Columbus, Ohio 163

voluntarilydesegregated, in the late1960s, toone thatwas latercourt-orderedtodesegregateviatheuseofbusing,in1979.thispolicywasineffectuntil1995,whenthecolumbuspublicschoolsendedtheformaluseofbusingasameansofdesegregation(fos-ter1997).

these changes notwithstanding, racial segregation contin-ues tohavea significant impacton thesociogeographicmakeupof columbus. based on 1990 u.s. census data, columbus hadataeuberdissimilarity index scoreof63.0 (harrisonandwein-berg1992).1asaresult,theafricanamericanpopulationremainspredominantlybluecollarandlivinginareascloselysurroundingtheurbancore,andcontactamongmanyblue-collarwhitesandafricanamericanscontinuestooccurinareasclosesttothecore,mostintenselyonthesoutheastandeastsides.theseareas,whichincludetheportionsofcolumbusfromwhichourinformantsweredrawn,areindicatedinfigure7.1.

the dialect features of blue-collar urban columbus speechthathaveemergedasaresultof thiscontactareacomplexmix-tureofnorthernandsouthernfeatures.thomas’s(2001)instru-mentalreanalysesofdatacollectedfortheDARE(1985–)dialectsurvey,alongwithhis(1989b)primarilyimpressionisticanalysesofblue-collarwhiteandafricanamericanspeech,revealmuchaboutthecontinueddevelopmentofthesedialectpatternsintheregionduringthelatenineteenthcenturyandearlytomid-twentiethcen-tury.duringthisperiod,thepredominantvernacularincolumbusincludedfeaturestypicallyassociatedwiththesouthernshift,suchas the frontwardmovementof thenucleiofbout,2boat,book,toot,andboot,andnorthmidland(Lowernorth)features,suchas r -fulness, the backing of bot, and the merger of border andboar.

Lessdiachronicinformationaboutcolumbusaaeisavailable.thomas(1989b)posits that features traditionally assumed tobemoststronglyaffiliatedwithsupraregionalaaenorms—suchasr -lessness,glide-weakeningofbuy and bide(similartothepatternfoundinsouthernspeech),andthetendencyforbout, boat,andboottoremainback—wereastrongelementofblue-collaraaeincolumbusinthelatenineteenthandearlytwentiethcenturybut


declinedduringthesecondhalfofthetwentiethcentury.basedondata collected in 1985 from blue-collar african americans bornafter1965,thomasnotonlyreportedthedecreasedpresenceofthesefeatures,butalsoprovidedevidencethatcolumbusafricanamericanspeakershavebeguntorealizeapartialmergerofbot-boughtbefore /t/andthefrontwardmovementofthenucleiofbout, boat, boot,andtoot.heargued thatafricanamericanvowel systems in columbus appear to be showing some conver-gencewithlocalwhitenorms,likeotherrecentlystudiedcommu-nitiessuchashydecounty,northcarolina(wolframandthomas

figure 7.1concentrationofafricanamericansincolumbus,ohio

(u.s.censusbureau2000)

25.4–46.6% 49.3–69.4% 72.8–92.8%

Columbus

270

70

16

40

62

161

71

70

71

270

670

33

23

62

23

33

161 161

315

104

franklin co pickaway co

delaware co

union co

mad

iso

n c

o

lick

ing

co

fair

fiel

d c

o

Columbus

40

Columbus, Ohio 165

2002),texana,northcarolina(childs,mallinson,andcarpenter2010[thisvolume]),Pittsburgh(eberhardt2010[thisvolume]),and memphis, tennessee (fridland 2003; fridland and bartlett2006).thismarks thedifferencesbetweengroups incontrast tothenonconvergencenotedearlierinothercommunities,suchasspringville and silsbee, texas (bailey and thomas 1998), Phila-delphia(Graff,Labov,andharris1986),andwilmington,northcarolina(thomas1989a).althoughthomas’sdatawerecollectedatonlyone timepoint, theassumption thatearlieraaespeakersmayhavedisplayedmoresouthernorolder supra-regionalaaefeaturesisconsistentwiththehistoricalmigrationpatternsofafri-canamericanresidentsincolumbus.however,asmoreextensivedatawereunavailabletohim,manyofhisconclusionsremainopentofurtherexploration.

morerecentstudiesofsociolinguisticvariationincolumbus,bothinthesuburbsandintheurbancenter,havefocusedonthevariable fronting and lowering of the nucleus of the boat class,a feature central to thomas’s convergence argument. in thesestudies,boat-frontinghasbeenfoundamongwhitespeakersandappears to be led by young white females. for example, durian(2008a)foundthatyoungfemalesledinthefrontingofbothboatand bout nuclei among white-collar white columbusites; dods-worth (2005) found similar results for boat in the white-collarcolumbus suburbofworthington,asdidthomas(1996) in theblue-collarexurbofjohnstown.

thomas (1989b) found that, among blue-collar speakers,femalesledinthefrontingofboat,withthemostextremefront-ingamongwhitefemales.nosinglelinguisticenvironmentsignifi-cantly conditioned boat nucleus-fronting except that followingliquidsdisfavoredit,especiallyamongwhites;however,centraliza-tionbefore / l /occurredata44%rateamongafricanamericans.fromthisevidence,thomasargued that theafricanamericans’centralizationof thenucleusofboat resulted fromcontactwithblue-collar whites and that the centralization was generalized toincludepre-/ l /position.

inthiscontext,weaddressthefollowingquestionsconcerningtherelationshipbetweenaaeandthepredominantvernacularas


spokenbyblue-collarresidentslivingintheurbancoreofcolum-bus.first,whatdocolumbusaaevowelspaceslookliketoday,andhowdotheycomparewiththoseofblue-collarwhitesincolumbus?second,consideringthecentralityofboattothomas’sargumentforconvergence, is thereapparent-timeevidence thatcolumbusaaeisshiftingtowardparticipationinboat-fronting?third,howdoessystematicvocalicvariationincolumbusaaecomparetothatinotherregionalaaevarieties?

meThOdOLOGY

data were collected from four distinct populations. the first setwascollected in1992 foraproject focusingonmorphosyntacticandphonologicalvariationamongblue-collarafricanamericanslivingonthesoutheastsideofcolumbus.fifty-fourafricanameri-canswereoriginallyrecorded,thoughthecurrentstudyfocusesonasubsetof14bornbetween1942and1977.samplesconsistedofconversationalspeech,andallfield-workersandparticipantswereafricanamericanandwithin-groupmembersofafamilyandtheirclosestneighbors(seeweldon1994;mcGuire2003;anddurian2008b formoredetailson this study). individual vowelplots forfourspeakersarepresentedinthenextsection,whileboatdatafromall14arepresentedinthesectionafterthat.

thesecondsetofrecordingswasobtainedfromfourblue-col-larwhitespeakersbornbetween1950and1980fromthebuckeyecorpus(Pittetal.2007),acollectionof40one-on-onesociolin-guisticinterviewsconductedbywhiteresearchersattheohiostateuniversity in2000.thedatafromthesefourspeakersappear intheindividualvowelplotssectionaswellastheboatcomparison.athirdsetofrecordings, largelyofreadspeech,wasmadeavail-ablebyerikr.thomasfromhiscorpusofspeakersbornbetween1946and1964andnow living in johnstown,ohio,anexurbofcolumbus(seethomas1996formoredetailsonthisstudy).theserecordings were made in 1994 with nine blue-collar whites whogrewupincolumbusandareusedintheboatanalysispresented.in addition, awhitemale speaker,borncirca1984, andawhite

Columbus, Ohio 167

femalespeaker,borncirca1977,recordedin2007and2008viahour-long sociolinguistic interviews by the first author, are alsoincludedintheanalysisofboat.

all blue-collar african american informants were recordedinteractingwithotherblue-collarafricanamericaninformantsbyan african american field-worker, and all tokens were extractedfromconversationalspeech.forthewhiteinformants,thespeak-ersrecordedforthebuckeyecorpusandbydaviddurianspecifi-callyforthisprojectwereinterviewedbywhiteinterviewers,andalltokens fromthese interviewswereextracted fromconversationalspeech.forthethomas(1996)studyspeakers,datawereelicitedfromwhite speakersbyawhitefield-worker, andall tokenswereextractedfromamixofwordlist,readingpassage,andconversa-tionalspeechenvironments.3

all white informants were selected because they were raisedorcurrentlyliveinsectionsoftheurbancoreinwhichthesocio-economicstatus,aswellas thecontactsituationbetweenafricanamericansandwhites,was similar to that found in southeasterncolumbus(wherearelativelyhighpercentageofblue-collarwhitesandafricanamericansliveinclosequarterstoeachanother).theoccupationlevelofadultinformantswasalsousedtoensurespeak-erswererepresentativeofblue-collarspeech.

sex,birthdate,race,localeinwhichspeakerswereraised(ifknown),4andoccupationofall29speakers(14africanamericansand15europeanamericans)areprovidedintable7.1.interviewercharacteristics(includingsexandrace),thedateoftherecordingsessionwiththespeaker,andthetypeofspeechelicitedfromthespeaker during each type of recording session are also listed intable7.1.

for the individual speaker vowel plots, classes appearing insmallcapitallettersrepresentthemeanvalueoftentokens,whilethoseappearingin<>notationrepresentthemeanvalueofthreeto eight tokens, except for theboat class.5token selections forthese classes were limited to no more than 3 repetitions of thesamelexicalitem.forboat,15tokensperspeakerwereextracted,withnomorethan4repetitionsofthesamelexicalitem.themeanvalueforthe15tokensofboatisplottedintheindividualaswell


table 7.1demographiccharacteristicsofthespeakersamplePopulation

Speaker Interviewer Date of Speech# Sex Birthdate Race Location Occupation Sex Race Interview Style1 F c1950 AA SE Side warehouse F AA 1992 Casual office mgr2 F c1950 W Short N nurse M W 2000 Casual Int3 F c1969 AA SE Side student F AA 1992 Casual4 F c1976 W NE Side homemaker F W 2000 Casual Int5 M c1957 AA SE Side car rental F AA 1992 Casual service mgr6 M c1959 W E Side HVAC worker M W 2000 Casual Int7 M c1975 AA SE Side student Casual8 M c1980 W NE Side landscaper F W 2000 Casual Int9 F c1945 AA SE Side cashier F AA 1992 Casual10 F c1947 AA SE Side warehouse F AA 1992 Casual worker11 F c1952 AA SE Side warehouse F AA 1992 Casual worker12 F c1952 AA SE Side warehouse F AA 1992 Casual worker13 F c1954 AA SE Side receptionist F AA 1992 Casual14 F c1953 W N Side house cleaner M W 1994 Read15 F c1960 W S Side factory worker M W 1994 Read16 F c1960 W NE Side homemaker M W 1994 Read17 F c1976 AA SE Side student F AA 1992 Casual18 F c1977 W W Side teacher’s aide M W 2007 Casual Int19 M c1942 AA SE Side exterminator F W 1992 Read20 M c1951 W NE Side fire truck M W 1994 Read builder21 M c1955 W Short N trucker co. M W 1994 Read employee22 M c1956 W N Side sales rep M W 1994 Read23 M c1957 W N Side baker M W 1994 Read24 M c1958 W NE Side car painter M W 1994 Read25 M c1958 W N Side mason M W 1994 Read26 M c1976 AA SE Side high school F AA 1992 Casual student27 M c1977 AA SE Side student F AA 1992 Casual28 M c1977 AA SE Side student F AA 1992 Casual29 M c1984 W E Side waiter M W 2008 Casual Int

Columbus, Ohio 169

as group plots. on average, speakers each contributed roughly20–30minutesof speech,generallywithmore from theone-on-one interviewsand less from thegroup interviews.forall vowelclassesanalyzed,tokensoccurringinenvironmentswithapreced-ingorfollowingliquidorafollowingnasalwereexcluded(withtheexceptionofbarandboar).alldatawereanalyzedacousticallyinPraat(boersmaandweenick2006),usingavariablewindowof10–14LPccoefficientsdependingonthequalityofthetoken.initialmeasurementsweretakenbyall threeauthorsandanadditionalresearcher, aidedbya custom-made formantextraction script inPraat,withadjustmentsmadebyhandwhennecessary.thedatawerecheckedforinter-raterreliabilityacrossmeasurementsfollow-ingtheinitialcoding.

traditional monophthongs, bit, bet, bat, bot, bought,book,bar,andboarareplottedwithameasurementofthesteadystatetakenatthe50%pointofthevowel’sduration.vowelsthatarecommonlytreatedasdiphthongs,beet,bait,bide,bite,bout,boat,boot,andtoot,areplottedusingmeasurementstakenat20%and80%torepresentthenucleusandoffglide,witharrow-headsmarkingtheoffglide.Tootistreatedasaseparateclassfromboot because preceding alveo-palatals and alveolars ordinarilyinducefrontingof /uw/(Labov,ash,andboberg2006),whilebarandboararetreatedseparatelyfrombotandboughttoprovideamoredetailedpictureofthebackportionofthespeakervowelspaces.bideisalsotreatedasaseparateclassfrombite,giventhetendencyforvoicedsegmentsfollowing /ai/toinduceshorteningoftheoffglideincentralohio.inourdata,theboarandborderclassesaretreatedasasingularmergedcategory.

PresenT-daY bLue-COLLar COLumbus VOweL sY sTem VariaTiOn

forthecomparativeplotspresentedinthefollowingsection,datawerenormalizedusingtheLobanov(1971)z-scoreformula,afterthe raw hertz values were first transformed into bark using theformulaprovidedintraunmüller(1990).thedatawerefirsttrans-


formedtobarkinanefforttoincreasecompatibilitybetweenthegeneralvowelplotsandtheboatspecificanalysis,whichwerenor-malizedusingthemeanZ3implementationofthebarkdifferencemetric(thomasandKendall2007).followingnormalization,thedatawerethenscaledtohertz-likevaluesacrossalleightspeakersfordisplaypurposesusingtheformulasprovidedbythomasandKendall(2007).

speakervowelplotsaregroupedsothatoneafricanamericanspeaker and one white speaker of the same gender and relativeageappearinthesameplot.figure7.2illustratesthevowelsystemforspeaker1,anolderafricanamericanfemale,andspeaker2,anoldereuropeanamericanfemale.thecenterofthewordrep-resentsthevowelnucleus.offglidesaredenotedbylinesmovingawayfromthenucleusandendinginanarrowheadtoshowtheirterminus.although it isnotnotedexplicitly in thefigures,mostspeakersshowanapparentnear-mergerof thebotandboughtclasses,atleastbefore /t/.however,forseveralspeakers,ourdatahaveahighconcentrationoflexicalitemswithword-final/t/,soour

figure 7.2speakers1and2,olderblue-collarfemalesfromcolumbus

Nor

mal

ized

F1

(Hz)

Normalized F2 (Hz)

300

400

700 2200

500

600

1800 1400 1600 1200 1000 2000

bat <height> <side>

bought

beet

beet

bootbit

bit book book

toot

bait

toot

bet

bet

butbait

bat

<white><five>

bat

but

<height>

bout bot

bout

bot <side>

bought bought

bar<far>

boatboat

boot

boarboar

S1 (AA, b. 1950)S2 (W, b. 1950)

Columbus, Ohio 171

findingsmaybesomewhatskewedbythisenvironment.itshouldalsobenotedthat intheplots inthissection,moreconservativeblue-collarwhitespeakervowelsystemsaredisplayedforourspeak-ers in each age group, as these are the speakerswith whom theblue-collarafricanamericanspeakersinourdatashowthestron-gestsignsofconvergence.thisismostclearlydemonstratedinthecontextoftheboatspecificanalysisthatfollows.

ourplotsshowthatallspeakershaveanessentiallymonoph-thongalbeet,withslightloweringamongonlysomeafricanamer-icanwomen(asinfigure7.3).however,ourdatasuggestthatthisloweringislargelydiminishingacrossethnicityaswellasage.thebaitclassalsoreflectssomenuclearlowering,butonlyamongafri-canamericanspeakers,andappearstobereversing,asouryoungspeakers(speakers3and7)shownosignsofloweringcomparedwiththeolderfemaleafricanamerican(speaker1).thiscontrastswiththomas’searlierfindings,inwhichbait-loweringwasfoundto some extent among both european americans and africanamericans.

figure 7.3speakers3and4,Youngerblue-collarfemalesfromcolumbus

Nor

mal

ized

F1

(Hz)

Normalized F2 (Hz)

300

400

2200

500

600

1800 1400 1600 1200 1000 2000

<north>

beet

<court>

<hard>

beet

bit

tootbook

book

toot boot

bit

bait bait

betbut

but boat

boatbar

bought

boughtboutbot

bet

batbat

bout

<height>

<side>

boot

bot bidebite

S3 (AA, b. 1969)S4 (W, b. 1976)


bitandbetvaryconsiderably.6raisingofbetisfoundamongtheyoungerafricanamericanmale(speaker7)andalsotheolderafricanamericanwoman(speaker1),butitisnotfoundamongtheolderafricanamericanmale(speaker5),whoactuallyshowssomemildlowering,ortheyoungerafricanamericanfemale(speaker3).incontrast,noraisingofbetwasfoundamongthewhitespeak-ers,althoughsomeretractionofthenucleuswasfoundamongtheolderwhitewoman(speaker2).bitalsoappearstoberaisedonlyamongafricanamerican speakers, and raisingappears forbothyounger african american speakers (speakers 3 and 7) and theolder african american woman (speaker 1). however, raising ofbitisnotfoundamongthewhitespeakers.takentogether,theseresultsindicatemovementawayfromearlierdocumentedcasesofsouthern-styletensingandraising(thomas2001)amongwhites,whileraisingisstillcommonamongafricanamericans.

anothernoteworthytendencyisthemildraisingofbatamongsomeafricanamericansandwhites.Previously,thomas(1989b)found /á/ toberelatively stableamongbothethnicgroups,with

figure 7.4speakers5and6,olderblue-collarcolumbusmales

Nor

mal

ized

F1

(Hz)

Normalized F2 (Hz)

300

400

2200

500

600

1800 1400 1600 1200 1000 2000

<tight> <bite>

<court>

beet

<five><side>

toottoot

beet

bit

bit

baitbait

book book

butbut

boat boat

boot

boot

betbetbat

bat

bout

bar bar

bought

boughtbot

botbout

<tight>

boar

<bite>

S5 (AA, b. 1957)S6 (W, b. 1959)

Columbus, Ohio 173

raising foundonly to a limitedextent amongwhites andmainlybeforenasals.ourdatadifferfromthomas’sinthatwefindmoreextensive raising in nonnasal environments among both africanamericans and whites, albeit more among african americans(speakers1,5,and7)thanwhites(speaker6).7bothofourwhitefemalespeakersdifferfromthomas’s(1989b)whitesbyalsoshow-ing some retraction. however, similar retraction has also beenfoundinnearbyworthingtonandjohnstownamongbothyoungmaleandyoungfemalewhitespeakers(thomas1996;dodsworth2005),suggestingtheretractedvariantmaybeacquiringnewsocialmeaning.

our evidence for the bite and bide classes is less definitivethanformostothervowelsbecausetheshortnessofsomeoftheinterviewsmadeitimpossibletoextractenoughtokens.afewgen-eralobservationsarepossible,though.withregardtothenuclei,allthespeakersshowquitelownucleiforbide.theformantplotssuggestthatsomeofthespeakers,especiallythemales,mayshow“canadian raising”—i.e., higher nuclei for the bite class than

figure 7.5speakers7and8,Youngerblue-collarcolumbusmales

Nor

mal

ized

F1

(Hz)

Normalized F2 (Hz)

300

400

2200

500

600

1800 1400 1600 1200 1000 2000

bout

<type> <hard>

toot toot

bitbeet

beet

bit

boot

book

book

but boat

boot

boar

bought

baitbet

bet

bait

bat

bat bout

boutbot

bite

boar

boughtbarbut

bot

<court>

<side>

<type>

<five>

S7 (AA, b. 1975)S8 (W, b. 1980)


thebideclass—butagain,welackconclusiveevidence.asfortheglides,thebiteclassusuallyshowsstrongglides,butseveralspeak-ersofbothethnicitiesshowratherweakglidesforbideclasswords,asothershavefoundforthemidlandregion(e.g.,Labov,ash,andboberg2006).

thebackvowelsbotandboughtareoftenrealizedasnearlymerged across class and racial groups in columbus. as thomas(1989b)hasdiscussed,botandboughtareoftenmergedbefore/t/ incolumbusona lexeme-by-lexemebasis,butafricanameri-cansappeartoberesistingthemergerinotherenvironments.ourspeakers show a different trend, however. the younger africanamericanwoman(speaker3)showsonlyapartialmergeroftheseclasses before /t/ in our data, while the older african americanwoman(speaker1)andyoungerafricanamericanmale(speaker7)showlittleevidenceofpartialmergerevenbefore /t/,andtheolderafricanamericanmalehasamoreorlessfullymergedpairofclasses.asthomashasalsoreported,near-mergerappearstobemore robust amongwhite speakers,who tend toexhibitmergerbeforevoicelessstopsmorefrequentlythanafricanamericans.ourwhitespeakersalsodifferinthisrespectandshowthispatternonlybefore/t/.inotherenvironments,ourplotsshowfronterboughtforallfourafricanamericanspeakersrelativetothewhitespeak-ers,aswellasfrontingofbotforspeakers3and7andloweringofbotforspeakers1and7.barandboarpresentanexceptiontothistrend,withthreewhitespeakersshowingfronterbarandboartokensrelativetotheafricanamericanspeakers,theolderwhitespeakerconstitutinganexception.however,speakersofbotheth-nicgroupsshowasimilardegreeofmergerwiththeborderclass(asdiscussedabove),regardlessofageorsex.

thefrontingand/orloweringofbot,takentogetherwiththeraisingofbat,bit, andbet, are trends thatmight indicate thatspeakers1and7,aswellaspossiblyspeaker3,areparticipatinginwhatthomas(2007b)hascalledtheafricanamericanshift.how-ever,itshouldbenotedthatspeaker3hassubstantialbit-raisingwithasomewhatloweredbeet,whilespeaker1hassomedegreeofbet-raisingwithatleastsomedegreeofbaitlowering.hence,thesouthernshiftmaydescribethesespeakers’vowelconfigura-

Columbus, Ohio 175

tions better than the african american shift, although as notedabove,theloweringofbaitappearstoberecedingovertime.moreresearchneedstobeconductedincolumbusbeforesuchadeter-minationcanbemadeconclusively.incontrast,amongwhitespeak-ers,theretractionofbatandbet,alongwiththepartialmergerofbotandboughtbefore/t/inthevowelsystemoftheolderwhitewoman(speaker2),suggeststhatsomeblue-collarspeakersmaybeshowing signsofparticipating inwhatclarke,elms,andYoussef(1995)andLabov,ash,andboberg(2006)havedefinedas thecanadianshift,atrendthatwasalsofoundrecentlyincolumbusbydurian(2008a)inthevowelsystemsofwhite-collarwhitemalesbornafter1962.

for book, most speakers in columbus evidence some front-ingasdiscussedindetailinthomas(1989b)andLabov,ash,andboberg(2006),whileblue-collarafricanamericanspeakerstendstronglytowardbut-raising(speakers1,3,and5).however,notethatspeaker6alsoappearstoshowamildtendencytowardraisingaswell,atrendpreviouslyunreportedamongwhitespeakers.forbook,whitespeakersrealizeonaveragefronterandhigherarticu-lations,whileafricanamericans’arelowerandbacker.

aspreviouslydocumented(thomas1989b,2001;Labov,ash,andboberg2006;durian2008a),thenon–lowbackvowels(boat,toot, andboot) and thediphthongbout appear tobeunder-goingfrontingofthenucleus,and,tosomeextent,theoffglides.thesetrendsappeartobeinterrelated,suggestingapossiblechainshift,thoughpreviousstudiesdisagreeastowhetherthistermcanbeappliedtothecovariantbehaviorexhibitedincentralohio(cf.Labov,ash,andboberg2006;thomas2001;durian2008a).allspeakersinfigures7.2–7.9shownuclearfrontingofthetootclassalongwithvaryingdegreesofnuclearfrontingforboot,boat,andbout.wheregroupsincolumbusdifferisintheextentofnuclearfrontingthatoccursandtheimpactofnuclearfrontingontheoff-glides.

thefrontingofthenucleusoftootcoincideswithglidereduc-tion among african americans (speakers 1, 3, and 7), whereasglides appear relatively unaffected among white speakers (seespeakers 2, 4, 6, and 8). boot fronting appears to be a general


trendamongspeakersofbothethnicities,withtheleastamountoffrontingfoundamongolderafricanamericans.themildfrontingamongyoungerafricanamericanssuggeststhebeginningofcon-vergenceofafricanamericannormswithwhitenorms.alleightspeakers inthesedatarealizebootwithabackglideandfiveofeightrealizethenucleusofbootlowerthantoot.whitespeakersclearlyleadinnuclearfrontingforbootand,toasomewhatlesserdegree,fortoot.

thomas’s(1989b)impressionisticanalysisofafricanamericanandwhitespeakersforboutfrontingsuggeststhatglidereductiontypifiesafricanamericanusageasabyproductofnucleus-fronting.ourdataareconsistentwiththomas’sfindings,andwefindthatamong whites bout realizations generally show less glide reduc-tionbutmorenuclearfrontingthanamongafricanamericans.inourdata,africanamerican speakersgenerally realizeboutwithbothalowernucleusandmorereducedglidethanwhitespeakers.however,thereissignificantoverlapbetweenthegroupsforfront-nessofthenucleusandglidesamongtheyoungermalesandolderfemales.thetwoyoungfemalesshowmoredifferentiation,how-ever,becausetheyoungafricanamericanwomanhasaretractednucleus.

inregardtoboat, thedata infigures7.2–7.5andthomas’s(1989b,2001)earlierfindingssuggestthatblue-collarwhiteandafrican american speakers both show fronting of the nucleus.the pattern is both sex-differentiated and age-graded: youngerwomenaretypicallythemostadvancedfronters,regardlessofrace,althoughastheevidenceamongyoungerwhitemenandwomeninbothfigures7.3and7.5,aswellasamongthespeakersplottedinthefollowingsection,suggests,theyoungfemaleleadmaybeshow-ingsignsofdiminishingamongyoungblue-collarurbanspeakers.forheight,whitespeakers,particularlyyoungermen,havehighernucleiandglides thanafricanamerican speakers,whileafricanamericans tend toward slightly shorterglides.for frontness, thetrendsgenerallysupportthomas’sconclusionthatboat-frontingexhibitsconvergencebetweenafricanamericanandwhitevariet-ies.however,asthenextsectionreveals,thesituationisalsomorecomplexthanwhatfigures7.2–7.5depict.

Columbus, Ohio 177

a deTaiLed eXPLOraTiOn Of COLumbus bOaT-frOnTinG

figures7.6–7.9presentaside-by-sidecomparisonofafricanamer-icanandwhitevariationpatternsforboat,subdividedbyageandsex, for the larger sampleof29blue-collar speakers(14africanamericanand15white)analyzedinthisstudy.itshouldbenotedthatonlymeasurementsobtainedfromtheboattokensthemselveswereusedtofeedthenormalizationcalculationsforthedatapre-sentedinthissection,asadditionaldatafromtheafricanameri-canspeakerswasnotavailabletoallowtheuseofadditionalvowelclasses (suchascornervowels) in thosecalculations.asa result,avowel-intrinsicmethod,ratherthanvowel-extrinsic(suchastheLobanov 1971 technique used in the full vowel systems plottedabove),wasused—thebarkdifferencemetric(thomasandKen-dall2007).

becauselittleinformationontheshapeofthevowelspaceforthesespeakersisavailableasaresultofusingavowel-intrinsicnor-malization,theresultsappeartoshowsomesignsofdistortion.insome recordings, f3 is indistinct or poorly tracked; because thebarkdifferencemethodreliesonf3,someskewingoftheresultsmayhaveresulted. inparticular, several speakersappear tohaveinappropriately skewed normalized mean values relative to theirmeanvaluesinrawhertz.theseincludehighernucleusf2valuesin relationship toother speakers (such that theyappear tohavefronternucleithantheylikelyhave)forspeakers1and2amongthe older women, as well as speaker 26 among the young men.also,speaker12amongtheolderwomenappearstohavealowernucleusf2value(abackernucleus),andspeaker6amongtheoldermenhasalowernucleusf1meanvalue(ahighernucleus)thanhelikelyhas.8 thus,theresultsbelowarebestviewedasanestimateoftheconvergencetrendsfoundamongourspeakersclusteredasgroups,presentedforthepurposesofcomparisonwithotherstud-iescollectedinthisvolume.

toperformthenormalization,thedatawerefirsttransformedintobarkusingtheformulaprovidedintranmüller(1990).fol-lowingconversiontobark,ameanZ3valuewasobtainedfrommea-


table 7.2meanvaluesbyspeakerGroup,inmeanZ3normalizedbark

µZ3 – Z1 µZ3 – Z1 µZ3 – Z1 µZ3 – Z1 (nucleus) (nucleus) (offglide) (offglide)Older AA Men 8.64 4.75 8.85 4.91Older W Men 9.06 3.71 9.71 4.18Younger AA Men 8.61 4.32 9.11 4.73Younger W Men 9.86 3.84 10.52 4.89Older AA Women 8.84 4.28 9.20 4.74Older W Women 9.38 3.07 10.11 3.85Younger AA Women 9.15 4.22 9.70 4.83Younger W Women 9.16 3.96 10.00 4.81

surementsextractedacross15tokensofboatat50%ofthetoken’sdurationforeachspeakertoallowaspeaker-specificcentroidvaluetobeestablished.theheightdimensionofalltokens’nucleiandglideswerethencalculatedusingtheformulameanZ3 – Z1,whilethefrontnessdimensionofalltokens’nucleiandglideswerecalcu-latedusingtheformulameanZ3 – Z2.asinthefullvowelsystemsplottedabove,valuesforthenucleiwereextractedat20%ofthevowel’sduration,whilevaluesforglideswereextractedat80%ofthevowel’sduration.eachsymboldenotesthemeanvalueforthenucleusacross15tokensforagivenspeaker’sboatclass,andoff-glides are indicated by an arrowhead. table 7.2 provides groupmeanvaluesforeachgroupinthesefigures.

asshowninfigure7.6,blue-collarafricanamericanwomenbornbefore1969showamorediversemixtureofrealizationsforboat in comparison to african american female speakers bornafter 1969 (figure 7.7). speakers 10 and 13 tend toward nearlymonophthongal realizations, while the other four older africanamericanwomen(speakers1,9,11,and12)showmorediphthon-galrealizations,althoughtheirglidesarestillshorterthanthoseoftheolderblue-collarwhitewomen(speakers2,14,15,and16).

the older african american women are roughly 1.2 barkbackeras agroup than theirwhite counterparts for thenucleusandroughly0.9barkbackerfortheglide.thus,theirrealizations

Columbus, Ohio 179

are,asagroup,morecentral thanthoseofwhitewomen.olderafricanamericanwomenalsoshowapreviouslyunreportedtrendtowardhigherglidef1frequenciesthantheirwhitecounterparts,andtwoofthesixalsoshowhighernucleusf1frequenciesthantheolderwhitewomen.

incontrast,youngwomen,plottedinfigure7.7,showanethnicdifferenceofonly0.26barkinthenuclearf2meanandofonly.02barkintheglidef2mean.Lowspeakernumberslimitthestrengthofourconclusions,butbasedontheavailabledata,itwouldappearthat thesubstantialdifference inboat frontingobservedamongtheolderspeakers(figure7.6)hasbeenleveledamongtheyoungergeneration.thisconvergenceresultsmostrobustlyfromhigherf2meansamongtheafricanamericans(speakers3and17)forbothnucleusandglide,butalsofromlowerglidef1meansforthewhitespeakers(speakers4and18).thatis,youngerafricanamericanwomenaredifferentiatingthenucleusandglidemorethanolder

figure 7.6meanboatrealizations:olderafricanamericanandwhitefemales

Z3–Z

1 (B

ark)

Z3–Z2 (Bark)2.0

10.0

9.0

2.5 4.5 6.0

11.0

8.0

10.5

9.5

8.5

3.5 3.0 5.5 5.0 4.0

S15(b. 1960) S14

(b. 1953)

S16(b. 1960) S12

(b. 1952)S11

(b. 1952)

S2(b. 1950)

S9(b. 1945)

S1(b. 1950)

S10(b. 1947)

S13(b. 1954)

African American White


africanamericanwomen,whileyoungerwhitewomenareshort-ening theglidemore thanolderwhitewomen.Youngerafricanamericanwomenalsoappeartoproducehigherboatnucleiandglides(i.e.,withlowerf1)thanolderwomen.speaker17’smeannucleusf1andf2valuesarenearlythesameasspeaker4’s,whilespeaker3’smeannucleusf2value isnearly thesameasspeaker17’s,andhermeannucleusf1valueisnearlythesameasspeaker18’s.therangeofvaluesamongspeakersofbothgroups forf1alsoshowlessoveralldifferentiation,suggestingthatafricanamer-icanandwhitewomen’sproductionsareconvergingovertimeforf2andlikelyforf1aswell.

as shown in figure 7.8, african american men born before1975(speakers5and19)tendtowardshorterandmoremonoph-thongal realizations of boat than those of either african amer-ican men born after 1975 (figure 7.9) or any of the white menanalyzed in this study. they also produce variably front-glidingor back-gliding realizations with shorter and lower glides, whichthemeansrevealonlypartially.olderwhitemen(speakers6and

figure 7.7meanboatrealizations:Youngerafricanamericanandwhitewomen

Z3–Z

1 (B

ark)

Z3–Z2 (Bark)2.5

10.0

9.0

4.5 6.0

11.0

8.0

10.5

9.5

8.5

3.5 3.0 5.5 5.0 4.0

S18(b. 1977)

S3(b. 1969)

S4(b. 1976)

S17(b. 1976)


Columbus, Ohio 181

20–25),incomparison,showrealizationsrangingfrombackvari-antstostronglyfrontedones.Likeolderafricanamericanwomen,olderafricanamericanmenshow lowerglides thanolderwhitemen,althoughincontrasttothedifferenceshownamongsomeoftheolderafricanamericanandwhitewomen, theolderafricanamericanandwhitemen’snucleiareofcomparableheight.

asrevealedbyfigure7.9,youngafricanamericanmen(speak-ers7,26,27,and28)exhibitasimilarmovementtowardfronterproductions forbothnucleus andglide, aswell asmoreoverlapinthemeanoff2forthenucleusofboatwiththeirwhitecoun-terparts (speakers 8 and 29). there is also a tendency towardmorediphthongalformsofboatlikethosedisplayedbyyoungerwomen.theyoungerwhitemalesalsotendtoshowlongerglidesthantheolderwhitemales,thoughsomeoldermales(speakers23and24)showcomparableglides.allfouryoungafricanamericanmales show much higher nuclear and glide f1 frequency valuesthantheyoungwhitemales.thus,youngafricanamericanmalestendtoresemblethemoreconservativeoftheyoungerwhitemales

figure 7.8meanboatrealizations:olderafricanamericanandwhitemen

Z3–Z

1 (B

ark)

Z3–Z2 (Bark)2.5

10.0

9.0

4.5 6.0

11.0

8.0

10.5

9.5

8.5

3.5 3.0 5.5 5.0 4.0

S25 (b. 1959)

S20(b. 1951)

S22(b. 1956)

S23(b. 1957)

S24(b. 1958)

S21(b. 1955)

S6(b. 1959)

S19(b. 1942)

S5(b. 1957)

S25(b. 1959)



(speaker8)forf2ofthenucleusandthelessconservativeofthewhite males (speaker 29) for f2of the glide, but they resembleneitherforf1ofeitherthenucleusortheglide.

in sum, our findings are consistent with those of thomas(1989b)inshowingconvergencebetweenafricanamericanandwhitespeakersforf2atboththenucleusandoffglideofboat.ourdata also reveal a previously unreported trend toward loweringoftheglideamongolderafricanamericanwomenandyoungerafricanamericanmen,suggestingpossibledivergencefromwhitenormsforheight.

sOCiaL and reGiOnaL imPLiC aTiOns Of COLumbus VOweL VariaTiOn

as the above findings reveal, blue-collar african american andwhite speakers showconvergencealong the frontnessdimensionforthevowelstoot,boot,book,boat,andbout,aswellascon-

figure 7.9meanboatrealizations:Youngerafricanamericanandwhitemen

Z3–Z

1 (B

ark)

Z3–Z2 (Bark)2.5

10.0

9.0

4.5 6.0

11.0

8.0

10.5

9.5

8.5

3.5 3.0 5.5 5.0 4.0

S29(b. 1984)

S8(b. 1980)

S26(b. 1976)

S28(b. 1977)

S7(b. 1975)

S27(b. 1977)


Columbus, Ohio 183

vergencealongtheheightdimensionfortootandboot.ontheotherhand,theyshowdivergenceforbought,bot,book,boat,bout,bet,bit,andpossiblybutalongtheheightdimension,andboughtandbotalongthefrontnessdimension.forboatinpar-ticular,ourlargersamplerevealsthatthetrendssuggestedbytheindividual plots generally reflect variation within columbus forspeakersbornafter1940.thestatusofbatisunclearbecauseoftherangeofrealizationsshownbyourwhitespeakers,whilethatofbeetandbaitisunclearbecauseofconflictingpatternsamongafrican americans. for blue-collar african americans, takentogether,theraisingofbat,bet,andbitandthefrontingofbotmay indicate that somespeakersareparticipating in theafricanamerican shift, although the data are inconclusive because ourolderafricanamericanfemalespeakershowsbait loweringandouryoungerafricanamericanfemaleshowsbeetlowering,whichmayinsteadrepresentthesouthernshift.

within thecontextof thecolumbus speechcommunity, theconvergent trends may be motivated by some of the social pat-ternsdiscussedabove.first,therehasbeenlongstandingcontactbetweenthegroupswithinmostareas fromwhichspeakersweresampledinthisstudybecauseofthehistoricalmigrationpatternsdiscussed above. second, continued contact has been fosteredbetweengroupsbecausethesocioeconomicprofileoftheseareashas remained low over time. third, most of the speakers bornafter1969ofbothracesattendedhighschoolwithcourt-ordereddesegregation and busing leading to increased contact betweenthe races throughoutcolumbuspublic schools.considering theincreasein“facetime”fosteredbyschooldesegregation,itseemsplausiblethatthisextraexposurewouldpromoteconvergencefortoot,boot,book,boat,andboutamongouryoungestspeakers.thisisdemonstratedinparticularbythegrowingsimilarityinthebackvowelsubsystemovertime,notonlyinourstudy,butalsointhomas’s(1989b)studyofhighschool–agedteens,whichwascon-ductedduringtheheightofdesegregationincolumbusschoolsinthe1980s.

nevertheless, developments in white speech may be leadingtonewethnicdifferences.forinstance,betandbatappeartobe


undergoingthewesternandcanadiantendencytowardretractionamongsomeofourwhitefemaleinformants.thismovementisnotoccurringamongtheafricanamericans,who,ifanything,appeartobeintensifyingtheraisingofbetandpossiblybatandstillshowlatentevidenceofbaitandbeet lowering.thus,divergencefortheseclassesappearstobearesultofchangesfromearliercommu-nityspeechnormsnotonlyamongtheafricanamericanspeakers,butalsoamongthewhitespeakers.clearly,theinterrelationshipofthesesocialandlinguisticfactorsisacomplexmatter,andonethatrequiresadditionalresearchtargetingeachfactorspecifically.

with respect to alignment with regional dialect norms, ourdataappeartoconfirmthatcolumbusvowelsystemsamongspeak-ersofbothethnicgroupsaretypifiedbythenorth/southtransi-tional flavor that thomas’s earlier studies (1989b, 2001) wouldleadustoexpect.

however,thepatternsofbetandbatretractionshowastron-gerwesterndialectalignmentelementamongourwhitespeakersthan has been noted in previous research (e.g., thomas 1989b,2001;Labov,ash,andboberg2006),withtheexceptionofdurian(2008a).inaddition,thetendencyofourspeakerstorealizeonlyapartialmergerofbotandboughtbefore/t/differsfromthomas’s(1989b)earlierresultsaswell.

withregardtotherelationshipofcolumbusblue-collarafri-can american speech to african american speech elsewhere, itwouldappearthat,overtime,thestrongertrendtowardincreasedfrontingofthenucleiofbackupglidingvowelsincolumbusaaeissimilartowhathasbeenfoundrecentlyincertainothercommuni-ties,namelyhydecounty,northcarolina(wolframandthomas2002),texana,northcarolina(childs,mallinson,andcarpenter2010[thisvolume]),Pittsburgh(eberhardt2010[thisvolume]),and memphis, tennessee (fridland 2003; fridland and bartlett2006).thesecommunitiesalsoshowconvergencewithlocalwhitespeechnorms fortoot,boot,book, andboat.the rangeandmeanvaluesforboatamongour14speakersshowstrongparallelswithvowelplotdataorresultsdiscussedinthosestudies.ourplotslikewiseshowclearparallelswithpatternsnotedfortoot,boot,andbookinthosecommunities,althoughthetrendsaregenerally

Columbus, Ohio 185

lessrobustintermsofthemaximaldegreeoffrontnessexhibitedbyspeakersthaninthenorthcarolinacommunities.hence,itisperhapsunsurprisingtofindsimilarpatternsincolumbus,sincehistorically, thesevowelshavealsoshownsouthernshift tenden-cies incentralohio(thoughperhapsfromwesternPennsylvaniainfluence insteadof fromthesouthproper,hence thesimilarityaswellwithPittsburgh).additionally,asidefromPittsburgh,thesecommunitiesshowthereversalofthefrontvowelsubsystemclassesbetandbait,althoughbait-loweringappearstobeonthedeclineinmemphisamongyoungerspeakers,adevelopmentthatrendersmemphismoresimilaroveralltothepresent-daycolumbusspeechcommunitythantothenorthcarolinacommunities.

thetrendsfoundamongourafricanamericanspeakers forthefrontvowels,particularlytheraisingofbat,bet,andbit,alsoresemblethosefoundinmemphisbyfridlandandbartlett(2006),as well as in a variety of locales (including brooklyn, new York;austin, texas; and cleveland heights, ohio) by thomas (2001,2007a).theraisingbehavioroftheseclassesinthoseareas,alongwithcovariantfrontingand/orloweringofbot,constitutethecoreevidence thomas (2007b) used to argue for the african ameri-canshiftasasupraregionalfeatureofaae.however,aspreviouslynoted,ourspeakers’alignmentwiththisputativeaaenormcan-notbeconfirmeddefinitivelyinourdatasetbecauseitisatpres-entunclearfromthemixedresults inourstudywhetherafricanamericansinsteaddisplaythesouthernshift.

beyondthesepatterns,asnotedearlier,but-raisingamongafri-canamericans incolumbusappears toalignwith supraregionalaaenorms,whilethedivergenceintheheightofbook,boat,andboutmaybeuniquetocolumbusaae,asthistrendhasnotbeenreportedinothercommunitiesinwhichafricanamericanspeak-ersevidencefronting(e.g.,wolframandthomas2002;fridland2003;childs,mallinson,andcarpenter2010[thisvolume]).thereasonafricanamericanspeakersshowsuchtrendsrequiresmoreextensiveresearch,thoughitseemspossibletheymayindexracialidentity,sincefrontnessnolongerdifferentiatesafricanamericanandwhitespeakersincolumbusrobustly.


COnCLusiOns

intermsofthesocialmotivationsunderlyinginterrelationshipsinethnicvowelvariationincolumbus,ourstudyhasraisedasmanyquestions as it answers. rather than attempting to disentangletheseissues,givenconstraintsofspace,weprovidedinsteadonlyadescriptiveanalysisofthevowelsystemsinordertocontributetothebroaderinvestigationoflocalaaeconvergenceto/divergencefrompredominantregionalvarietiesaswellassupraregionalaaenorms.

four additions to this analysis would allow more confidentcomparisonsbetweenlocalpredominantandaaevernacularsandacrossregionalaaevarieties.first,agreaternumberofspeakerscouldconfirmthepatternsofage-gradingsuggestedbyourdata.second, as wolfram and thomas (2002) observe, an adequateinvestigation of vocalic variation in any community would con-sidermore thoroughly the identity-basedmotivationsunderlyinglanguagemaintenanceandchange.third,morestudiesinvestigat-ingtheoccurrenceoftheafricanamericanshiftincommunitiesacrosstheunitedstatesarerequiredsothattheramificationsofthe shiftas itaffectscolumbusaaevowel systemscanbebetterunderstood.

finally, there is the matter of the normalization technique(barkdifferencemetric)utilized inouranalysisofboat.asdis-cussedinthatsection,thetechniquewasutilizedbecauseofalackofavailabledatafromsomeoftheafricanamericanspeakerstopermittheuseofavowel-extrinsicmethod.thiswouldlikelyhaveimprovedthequalityoftheresults,asvowel-extrinsicmethodshavebeenshowninarecentcomparativeanalysisofnormalizationtech-niques(adank,smits,andvanhout2004)toperformbetterthanvowel-intrinsictechniques.inparticular,thebarkdifferencemet-ricappears tobequite sensitive tofluctuations inf3,and if thedatahavepoorlyformedorindistinctf3,asinthecaseofatleastsomeofourafricanamericanrecordings,theeffectivenessofthetechniqueappearstobestronglyimpacted.

Columbus, Ohio 187

nOTes

we wish to thank bridget smith for her significant contribution in theanalysisofdataaswellascontentsuggestionsappearinginthischapter.wealsothankerikr.thomasforcommentsandthecontributionofspeakersfromhis1990sstudyofjohnstown,ohio.inaddition,wethankdonaldwinford,cynthiaclopper,marybeckman,andmalcahYaeger-dror forcommentsthatstrengthenedouranalysis.

1. ataeuberindexscoreof100indicatescompletesegregation(acom-pletelyunevendistributionofminorities),whereasascoreof0indi-catescompleteintegration(acompletelyevenanduniformdistribu-tionofminorities)ofresidentsinacommunity.inotherwords,thehigherthescore,thegreaterthelevelofresidentialsegregation.

2. notethatthroughoutthisdiscussion,vowelclassesaremarkedusingamodifiedversionofthenotationforwordclassesprovidedbywells(1982).

3. forthesereasons,thomas’s(1996)resultscanbecomparedonlyten-tativelywiththeotherstudies’inthisvolume.

4. with regard to speakers for whom we were unable to sufficientlydeterminethisinformation,thelocationinwhichtheinformantcur-rentlylivesislistedinstead.thesespeakersinclude1,7,10,15–18,and23.

5. asecondexceptionisthat,forspeaker7,only1tokenofthebootclass( food)isplotted,while10tokensofthelexememoveareplottedfor speaker8.hence, in theirplots,weuse<food>and<move> torefertotheirmeanvaluesforboot.

6. althoughnotplottedhereduetolowfrequencycountsinourafri-canamericandata,ourimpressionisticanalysisofbinandbentokensamongourwhiteandafricanamericanspeakersrevealsthatspeak-ersofbothethnicitiesfrequentlysubstitutethebitforthebetvowelbeforenasals.thistrendismostpronouncedamongafricanameri-canspeakers,afindingwhichagreeswiththomas’s(1989b)earlierobservations.

7. althoughnotinstrumentallyanalyzedduetoaudioqualityissuesintheafricanamericandata,impressionisticanalysisofbantokensamongoureuropeanamericanandafricanamericanspeakersrevealssimi-larraisingtrendstothosefoundinthomas’s(1989b)study.

8. thisassessmentisbasedonacomparisonoftheplacementofboatamongspeakersforwhomwedidhavefullvowelsystemdatarather


thanonlyboatdata,usingthemeanZ3methodversustheplacementofboatforthespeakers,usingboththeLobanov(1971)andnearey(1978)“logmean”methods.

referenCes

adank,Patti,roelsmits,androelandvanhout.2004.“acomparisonofvowelnormalizationProceduresforLanguagevariationresearch.”JournaloftheAcousticalSocietyofAmerica116:3099–107.

bailey,Guy,anderikthomas.1998.“someaspectsofafrican-americanvernacularenglishPhonology.”inAfrican-AmericanEnglish:Structure,History,andUse,ed.salikokos.mufwene,johnr.rickford,Guybai-ley,andjohnbaugh,85–109.London:routledge.

boersma,Paul,anddavidweenink.2006.Praat:doingPhoneticsbycom-puter.version4.4.30.http://www.praat.org/.

bryant, vinnie vanessa. 1983. “columbus, ohio, and the Great migra-tion.”m.a.thesis,ohiostateuniv.

burgess,Patricia.1994.Planning for thePrivate Interest:LandUseControlsandResidentialPatternsinColumbus,OH,1900–1970.columbus:ohiostateuniv.Press.

childs,becky,christinemallinson,andjeanninecarpenter.2010.“vowelPhonology and ethnicity in north carolina.” in Yaeger-dror andthomas,23–47.

clarke,sandra,fordelms,andamaniYoussef.1995.“thethirddialectofenglish:somecanadianevidence.”LanguageVariationandChange7:209–28.

DARE.DictionaryofAmericanRegionalEnglish.1985–.ed.fredericG.cas-sidyandjoanhoustonhall.4vols.todate.cambridge,mass.:belknapofharvarduniv.Press.

dodsworth,robinm.2005. “Linguisticvariationandsociologicalcon-sciousness.”Ph.d.diss.,ohiostateuniv.

durian,david.2008a.“anewPerspectiveonvowelvariationthroughoutthe 20th century in columbus, oh.” Paper presented at the 37thannualconferenceonnewwaysofanalyzingvariation(nwav37),houston,texas,nov.6–9.

———.2008b.“thevocalizationof / l /inurbanbluecollarcolumbus,ohafricanamericanvernacularenglish:aQuantitativesociopho-neticanalysis.”OhioStateWorkingPapersinLinguistics58:30–51.

eberhardt,maeve.2010.“africanamericanandwhitevowelsystemsinPittsburgh.”inYaeger-drorandthomas,129–57.

http://www.praat.org/

Columbus, Ohio 189

foster,Pauln.1997.“‘whichseptember’:segregation,busing,andreseg-regationinthecolumbusPublicschools,1944–1996.”b.a.honorsthesis,harvarduniv.

fought, carmen. 2006. Language and Ethnicity. cambridge: cambridgeuniv.Press.

fridland, valerie. 2003. “network strength and the realization of thesouthernvowelshiftamongafricanamericansinmemphis,tennes-see.”AmericanSpeech78:3–30.

fridland, valerie, and Kathy bartlett. 2006. “the social and LinguisticconditioningofbackvowelfrontingacrossethnicGroupsinmem-phis,tennessee.”EnglishLanguageandLinguistics10:1–22.

Graff,david,williamLabov,andwendella.harris.1986. “testingLis-teners’reactionstoPhonologicalmarkersofethnicidentity:anewmethodforsociolinguisticresearch.”inDiachronyandDiversity,ed.davidsankoff,45–58.amsterdam:benjamins.

harrison,roderickj.,anddanielh.weinberg.1992.RacialandEthnicRes-identialSegregationin1990.washington,d.c.:u.s.censusbureau.

jacobs,Gregorys.1994.“GettingaroundBrown:desegregation,develop-ment, and the columbus Public schools, 1954–1994.” m.a. thesis,ohiostateuniv.availableathttp://etd.ohiolink.edu/send-pdf.cgi/jacobs%20Gregory%20scott.pdf?acc_num=osu1220536463.

Labov,william,sharonash,andcharlesboberg.2006.TheAtlasofNorthAmericanEnglish:Phonetics,Phonology,andSoundChange.berlin:mou-tondeGruyter.

Lobanov,b.m.1971.“classificationofrussianvowelsspokenbydiffer-entspeakers.”JournaloftheAcousticalSocietyofAmerica49:606–8.

Lentz,ed.2003.Columbus:TheStoryofaCity.charleston,s.c.:arcadia.mcGuire,Grant.2003.“therealizationofinterdentalfricativesincolum-

bus, oh, aave.” Paper presented at the montreal-ottawa-torontoPhonologyworkshop,toronto,feb.7–9.

murphy,melvinL.1970.“thecolumbusurbanLeague:ahistory,1917–1967.”Ph.d.diss.,ohiostateuniv.

nearey,terrancemichael.1978.PhoneticFeatureSystemsforVowels.bloom-ington:indianauniv.Linguisticsclub.

oriedo, evelyn. 1982. “african american business development: itsimpactontheeconomicstatusoftheafricanamericancommunityincolumbus,oh.”m.a.thesis,ohiostateuniv.

Pitt, mark a., Laura dilley, Keith johnson, scott Kiesling, william ray-mond,elizabethhume,andericfosler-Lussier.2007.BuckeyeCorpusofConversationalSpeech.2ndrelease.columbus:dept.ofPsychology,ohiostateuniv.availableathttp://www.buckeyecorpus.osu.edu/.

http://etd.ohiolink.edu/send-pdf.cgi/Jacobs%20Gregory%20Scott.pdf?acc_num=osu1220536463

http://etd.ohiolink.edu/send-pdf.cgi/Jacobs%20Gregory%20Scott.pdf?acc_num=osu1220536463

http://www.buckeyecorpus.osu.edu/


thomas,erikr.1989a.“theimplicationsof/o/frontinginwilmington,northcarolina.”AmericanSpeech64:327–33.

———.1989b.“vowelchangesincolumbus,ohio.”JournalofEnglishLin-guistics22:205–15.

———. 1996. “a comparison of variation Patterns of variables amongsixthGradersinanohiocommunity.”inFocusontheUSA,ed.edgarw.schneider,149–68.amsterdam:benjamins.

———.2001.AnAcousticAnalysisofVowelVariation inNewWorldEnglish.Publicationoftheamericandialectsociety85.durham,n.c.:dukeuniv.Press.

———.2007a. “conclusions:symposiumonvowelPhonologyandeth-nicity.”PaperpresentedattheannualmeetingoftheLinguisticsoci-etyofamerica,anaheim,calif.,jan.4–7.

———. 2007b. “Phonological and Phonetic characteristics of africanamerican vernacular english.” Language and Linguistics Compass 1:450–75.

thomas,erikr.,andtylers.Kendall.2007.norm:thevowelnormal-izationandPlottingsuite.available athttp://ncslaap.lib.ncsu.edu/tools/norm/.

traunmüller,hartmut.1990. “analyticalexpressions for thetonotopicsensoryscale.”JournaloftheAcousticalSocietyofAmerica88:97–100.

weldon,tracey.1994.“variability innegationinafricanamericanver-nacularenglish.”LanguageVariationandChange6:359–97.

wells, j.c.1982.Accents ofEnglish.3 vols.cambridge:cambridgeuniv.Press.

wolfram,walt,anderikr.thomas.2002.TheDevelopmentofAfricanAmeri-canEnglish.oxford:blackwell.

u.s.censusbureau.2000.“stateandcountyQuickfacts:columbus(city),ohio.”http://quickfacts.census.gov/qfd/states/39/3918000.html.

Yaeger-dror,malcah,anderikr.thomas,eds.2010.AfricanAmericanEng-lishSpeakersandTheirParticipationinLocalSoundChanges:ACompara-tiveStudy.Publicationof theamericandialectsociety94.durham,n.c.:dukeuniv.Press.

http://ncslaap.lib.ncsu.edu/tools/norm/

http://ncslaap.lib.ncsu.edu/tools/norm/

http://quickfacts.census.gov/qfd/states/39/3918000.html

convergence in blue collar columbus, oh aave and eae vowel systems? 1

Documents