University of Pennsylvania Working Papers in Linguistics

Volume 4 Issue 1 A selection of papers from NWAVE 25 Article 12

1997

Adaptive sociophonetic strategies and dialect accommodation: /ay/ monophthongization in Cherokee English

Bridget L. Anderson

Follow this and additional works at: https://repository.upenn.edu/pwpl

Recommended Citation Anderson, Bridget L. (1997) "Adaptive sociophonetic strategies and dialect accommodation: /ay/ monophthongization in Cherokee English," University of Pennsylvania Working Papers in Linguistics: Vol. 4 : Iss. 1 , Article 12. Available at: https://repository.upenn.edu/pwpl/vol4/iss1/12

This paper is posted at ScholarlyCommons. https://repository.upenn.edu/pwpl/vol4/iss1/12 For more information, please contact [email protected]. Adaptive sociophonetic strategies and dialect accommodation: /ay/ monophthongization in Cherokee English

This working paper is available in University of Pennsylvania Working Papers in Linguistics: https://repository.upenn.edu/pwpl/vol4/iss1/12 U. Penn WorkingPapers in Linguistics Volume 4.1 (1997)

Adaptive Sociophonetic Strategies communities and source language transfer features. Leechman and Hall (1955) even propose that a more expansive pan-lectal variety and Dialect Accommodation: of English developed out of the various situations in which English /ay/ Monophthongization in Cherokee English was learned and used in relation to the Native American language. Although some sociolinguistic situations involving Southwestern Bridget L. Anderson varieties have now been investigated, comparable situations in the eastern United States have received little attention from the lin guistic research community. 1. Introduction Thisstudy is a preliminary investigation of a language contact situation between two very distinct linguistic groups who Developing varieties of Naiive American English offer unique havebeen in close contact with each other for at least the past two insights into the sociolinguistic dimensions of language contact hundred years in isolated, mountainous Graham County in the situations. More specifically, an investigation of how these groups heart of the Great Smoky Mountains of Western North Carolina. utilize assimilative features, such as those adopted from local, non- Appalachian whites of thearea speak a Southern Highland variety Native American contact communities, and, at the same time, fea of English, comparable to what is described in general by Wolfram tures unique to the Native American English variety, such as those and Christian (1976) and more particularly, for the Smoky Moun which have developed as e result of source-to-target language tain region, by Joseph Hall (1942). The Snowbird Cherokee of the transfer, is particularly diagnostic in terms of how Native Ameri Eastern Band of the Cherokee Nation who reside in Graham cans situate themselves sociolinguistically with respect to sur County primarily spoke their ancestral language of Cherokee until rounding non-Native American contact communities and other early in this century when a shift toward bilingualism in Cherokee Native American groups. Such an investigation must take into and English began. An investigation of the contact situation in account the effects of source-language interference, the English Graham County, focusing on the patterning of a diagnostic vowel language learning situation, and dialect competition from sur variant, the monophthongization of /ay/ as in ride [ra:d] and type rounding non-Native American communities. [ta:p] for the two ethnic groups will yield insights into the mecha Previous studies of Native American varieties of English nisms of language contact, language assimilation, and language in the Southwest (Craig 1991; Leap 1977; Wolfram et al. 1979, shift. The monophthongized variant of /ay/ is widespread through Wolfram 1980, 1984), indicate that these varieties utilize both the out the South and is a prominent feature of Appalachian English. assimilated dialect features of surrounding non-Native American This variable is expected to be a fairly diagnostic variable of as similation. *I wish to thank my colleagues at North Carolina State University, Walt Wolfram, Erik Thomas, and Natalie Schilling-Estes, for their help with 2. The Cherokee Situation hi Western North this study, their insights into the analysis, and their comments on this paper. In addition, Erik Thomas spent many hours working through the Carolina data with me. I also wish to thank Natalie Schilling-Estes and Kirk Hazen for their assistance with the VARBRUL analysis and Kevin Wall for his Neely (1991:15) estimates the Cherokee to have been living in the invaluable assistance with the fieldwork. Finally, I wish to thank my Southern Appalachian Mountains for at least the past four thou informants, members of the Snowbird and Qualla Boundary communities, sand years. Furthermore, she notes that in the early part of the for their time, patience, and kindness. This work was funded by National nineteenth century the estimated 20,000-member Cherokee nation Science Foundation Grant Number SBR 96 16331 and by the William C. was one of the largest Indian nations north of Mexico. One-fourth Friday Endowment at North Carolina State University.

U. Perot Working Papers in Linguistics, Volume 4.1, 1997 186 Volume 4.1 (1997) Adaptive Sociophonetic Strategies Anderson V. Perm Working Papers in Linguistics to one-half of the 16,000 Cherokee people forced to march west in person must demonstrate that he or she is "certifiably" of at least 1830 to what is now Oklahoma in what has become known as the one-eighth American Indian ancestry. Full-bloods, of course, are "Trail of Tears" died during their tragic relocation (Neely of total Native American ancestry. Native American activist and 1991:22). This event, of course, considerably altered the lifestyles scholar Ward Churchill (1994) notes that in 1900 about one-half of of the surviving members of the Cherokee Nation. About one federally recognized, racially defined Native Americans in the thousand Cherokees hid in the Great Smoky Mountains in order to United States qualified as "full-bloods." By 1990, this proportion elude the forced removal, and it is their descendants who now was only at about twenty percent (Churchill 1994:92). In spite of make up the Eastern Band of the Cherokee situated in Western the significant decline of the number of federally recognized full- North Carolina. bloods among American Indians, the Snowbird Community has The Eastern Band consists primarily of Cherokees living maintained a large percentage of full-bloods. In the mid-1970s, in Western North Carolina on the Qualla Boundary of Swain and 91.4 percent of Snowbird Cherokee adults were legally three- Jackson Counties and, fifty miles to the southwest, in the Snowbird fourth to full-blood range (Neely 1991:7). and Cheoah mountains of Graham County. There is also a small Perhaps the high percentage of full-bloods in the Snow number of Eastern Cherokees who live in the Tomotla area of bird Community is the reason the community has also been suc Cherokee County. The Eastern Band holds 56,572 acres of com cessful in maintaining its ancestral language. Full-bloods tend to munal lands in Swain, Jackson, Cherokee, and Graham Counties in have more traditionalist Native American values, such as native Western North Carolina (Neely 1991:24). language maintenance, than people with minimal Native American ancestry. Most adults over age forty in Snowbird are bilingual,

2.1. The Sociolinguistic Situation while the significantly higher populated Qualla Boundary is esti mated to have less than 10 percent native language speakers (King The Snowbird Cherokee are considered to be the most traditional 1975:2). The tiny Snowbird Community comprises only 6.9 per of the three Cherokee groups residing in Western North Carolina, cent of the North Carolina Cherokees, but it contains nearly one- and it is this community that is the focus of the present study. The third of the total Cherokee-speaking population in the East (Neely Snowbird group is distinctive from other groups of Cherokees in 1991:147). Western North Carolina in terms of Native American "traditional The Snowbird Community is unique in its success in ism" and "conservatism," their assimilation to encroaching white maintaining a large number of foil-bloods and native language culture, and their percentage of native Cherokee speakers. speakers. Fifty miles to the northeast of Snowbird in the Qualla The Snowbird Community comprises only a small per Boundary the Cherokee language seems to be disappearing rap centage of Eastern Cherokees. Most of the 9,000 members of the idly. Both groups have had extensive contact with white English Eastern Band live on the Qualla Boundary. The 380-member speakers. The high percentage of ancestral language speakers in Snowbird Community, however, comprises only 6.9 percent of all the Snowbird Community indicates that the Snowbird Cherokee resident North Carolina Cherokee and 5.2 percent of Graham have made a group effort to maintain their cultural identity as County's total population (Neely 1991:38), making them a small "traditional" Cherokee Indians. The Qualla Boundary group has a minority in relation to Qualla Boundary Band members and the high percentage of what one of my informants referred to as 7,217- member white population of Graham County (1980 cen "white Indians," or people with minimal Cherokee ancestry who both look and "act" like white people. Snowbird, however, has sus). Snowbird has the highest percentage of full-bloods of any few "white Indians " Consequently, Snowbird Cherokees are a Eastern Cherokee community (Neely 1991:7). The government much more homogeneous group than the more acculturated Qualla "standard" for self-identification as an American Indian is that a Boundary group. Snowbird Cherokee, therefore, do not face the

187 188 Volume4.1(1997) Anderson U.PermWorkingPapersinLinguistics AdaptiveSociophoneticStrategies

3. TheHistoryandStatusMonophthongal of sameintraethniccompetitionbetweentraditionalistNativeAmeri

canvalues,whichseemtohaveastrongconnectionwithbeinga /ay/

full-bloodCherokee,andanglicized"whiteIndian"valuesof

peoplewithonlyminimalCherokeeancestry.SeveralSnowbird Themonophthongizationof/ay/asin[ra:d]rideand[la:t]lightis

informantsindicatedtomethattheyconsiderthemselves,butnot oneofthemostsalientfeaturesofSouthernspeech(Bernsteinand

QuallaBoundaryCherokees,tobe"realIndians." Gregory1993).Inaninvestigationofhowthisvariablepatternsin

PerhapsonefactorthathasaidedSnowbirdCherokeesin CherokeeEnglishthereareseverallinguisticandsociolinguistic

maintainingtheirancestrallanguageandotherimportantcultural dimensionsuniquetoacontactsituationthatmustbeconsidered.

traditionsisthegeographicalisolationwhichhasservedtoprotect First,itisimportanttodeterminewhattherelationshipof/ay/isto

themfromtourism,whichaffectsmanyNativeAmericanreserva thephonologicalsystemofCherokee,thefirstlanguageofmost

tions,includingtheQuallaBoundary.Snowbirdisuniqueinthatit middle-agedandolderspeakersintheSnowbirdCommunity.A

hasvirtuallynotourism,dueinlargepartthe todepressedecon secondconsiderationisthehistoryandstatusof/ay/inthewhite

omyandruggedterrainofthecountyinwhichitissituated. contactcommunity.Inaddition,sinceSnowbirdhasahighper

Eighty-fivepercentofGrahamCountyisundevelopedforests, centageofbilingualspeakers,itisimportanttoconsiderthestatus

someofwhichareamongtheonlyvirginforestseastoftheMis of/ay/intheEnglishlanguagelearningmodelforolderandmid

sissippi.OnlyoneotherNorthCarolinacountyhaslessland dle-agedspeakerswholearnedEnglishinschool.Andfinally,

clearedfor industryandfanning(Neely1991:37).Cherokeesin whatisthesynchronicsociolinguisticdistributionofthisvariable

thiscountycontinuetoresideontheirancestralhomeland,once amongcurrentspeakers?

theCheoahtownshipoftheCherokeenation,whichtheCheoah Inacommunitywheremostadultsoveragefortyarebi

Cherokeeactuallypurchasedfromthestatetheweeks in immedi lingualinCherokeeandEnglish,linguisticinterferencefromthe

atelyfollowingtheremoval(22).TheSnowbirdCherokeeare sourcelanguagetothetargetlanguageistobeexpected.Thoma-

descendantsoftheCheoahandcontinuetoresideonthisland, sonandKaufman(1988:37)indicatethatinthecaseoflanguage

2,249acresofscatteredtractsconcentratedinwhatiscommonly shift,interferencewillmostlikelybestructural—thatis,pho

referredtothe asSnowbirdareaofGrahamCounty.Thusfar,I nological,phonetic,orsyntactic—interferenceratherthanlexical

haveconductedsociolinguisticinterviewswithtwenty-fiveChero interference.AlthoughCherokeehasnoclear-cutcasesoftauto-

keeEnglishspeakersofdifferentagesinthearea.Forpreliminary syllabicnucleuscombinationssuchas[al],vowelcombinations

comparativepurposes,Ihavealsoconductedafewinterviewswith withepenthetic[y],suchas[aye]and[ayo],dooccur(Huff

CherokeesfromtheQuallaBoundaryandmembersofthewhite 1977:23).Thus,thereisaphonologicalmodelforupglidinginthe

contactpopulationofGrahamCounty.Evidencegleanedfrom sourcelanguage,althoughitisnottautosyllabic.

theseinterviewsdemonstratesthatbothCherokeegroupsexhibitat Thereisalso,however,aphonologicalmodelformono

leastsomeassimilationtothelanguagenormsofthesurrounding phthongal[a:]inthesourcelanguage.Huff(1977)observesthe

mountainwhitecommunities.Inthisstudy,Iwillattemptto followingvowel-glidesequencepatternsforCherokee:/a/plus

quantifythedegreeofassimilationthroughaquantitativeanalysis anyvowelexcept/a/and,mostsignificantly,IM,intheunderlying

ofmonophthongal/ay/,asin [ra:d]rideand[fa:t]fight,apromi formyieldsasurfaceform[a] of+ epenthetic[y]+vowel.A

nentfeatureofAppalachianEnglishandoneofthemostsalient vowelcombinationof/a/+/i/or/a/occurringintheunderlying

featuresofSouthernspeechingeneral. formwill,therefore,berealizedinthesurfacefromas[a].Inother

words,when/a/precedes/a/orIVintheunderlyingformthe of

sourcelanguagetheresultingsurfaceformis[a],butwhen[a]is

189 190 Volume4.1(1997) Anderson U.PennWorkingPapersinLinguistics AdaptiveSociophonelicStrategies

(1983:73)indicatesthat/ay/forseventyEastTennesseaninfor combinedwithvowelsotherthanIMor/a/intheunderlyingform

mantsisrealizedmostoftenasamonophthongand,lessfre thesurfaceformwillberealizedasavowel-glidesequence.So,

quently,ashortdiphthong. as Hefurthernotesthat/ay/istypi thesourcelanguageofCherokeeprovidesmodelsforboth

callymonophthongalbeforevoicelessconsonants,asinwriteor monophthongal[a:]andvowel-glidesequencesinvolving[a]+

light,forallageandsocialgroupsoftheregion(75). epenthetic[y]+vowel. My dataforthewhitecontactpopulationofGraham Thenextconsideration,then,isthehistoryandstatusof

Countyfor/ay/indicatesthatcurrent-daySmokyMountainEng /ay/inthesurroundingwhitecontactcommunity.Hall(1942:43)

lishislargelymonophthongalfor/ay/inallfollowingphonetic describesapatternofglideweakeningforhisdatafromtheSmoky

environments.Tabulationsofthe/ay/ variableforninelifelong Mountains,indicatingthat/ay/ismostoftenrealizedas[a:]inall

whiteresidentsofGrahamCountyindicatemattheseinformants phoneticenvironments.Henotes,infact,thatalthoughtheten

are categoricalmonophthongizersof/ay/inallphoneticenviron dencyingeneralSouthernspeechatthattimewastomonoph

ments.Sothecurrentcontactmodelisoneofexpansiveand gen thongize/ay/invoicedenvironmentsbuttoretainthediphthongin

eralizedmonophthongization. voicelessenvironments,thepatterndidnotholdtrueforSmoky

Anotherimportantconsiderationis,ofcourse,thecontact MountainEnglish,wheremonophthongal[a:]waspreferredinall

modeloftheinitiallanguagelearningsituationofmanyofthe phoneticenvironments(Hall1942:43).KurathandMcDavid

middle-agedandolderspeakers.Beginningin1880,whiteQuak (1961)foundtokensof[a9]and[a*]inWesternNorthCarolinafor ersbeganusingformaleducationinanattempttoacculturatethe

thewordtwiceandtokensof[a']inMaconCounty,whichborders CherokeeintoAnglo-Americansociety.Theseschoolsemphasized GrahamCounty,forthewordsnineandmight.Thedataforthe Anglo-Americancultureandvaluesandgavelittleattentionto

wordmightprovidedbytheLAMSASofficeattheUniversityof Cherokeeculture(Neely1991:29).Theteachersoftheseschools

Georgia*indicates/ay/wasmonophthongalinWestern North werenotlocaltothearea.Theyareexpected, not therefore,to Carolinainbothprevoicedandprevoicelessenvironmentsinthe haveservedastheagentsoftransmissionformonophthongal/ay/.

1930s,andthatprevoicelessdiphthongal/ay/wasalreadyarelic TheQuakerschoolsclosedwhentheBureauofIndian

forminthisarea.WolframandChristian(1976:64)foundthatAp Affairs(BIA)gainedcontroloftheCherokeeeducationalsystem

palachianEnglishspeakersintheirstudyparticipatedinthe intheearly1900s.Neely(1991:29)characterizestheBIA-run

monophthongizationof/ay/,andtheydeterminedthelinguistic boardingschoolsas"dictatorial,"as didseveralofmyolder

constraintorderforfollowingphoneticenvironmentsforthisfea Cherokeeinformantswhoattendedtheboardingschoolonthe turebe topause> voicedobstruent> voicelessobstruent.This QuallaBoundary.Studentsweretaughttoadoptwhitecultural

orderingfallsinlinewiththetraditionalconstraintpatternfor attitudesandwereseverelybeatenforspeakingCherokeeatany

SouthernspeechandisincontrasttoHall's(1942)observation time.A fewmiddle-agedandolderinformantsinmystudywho

that/ay/wasmonophthongaiinallfollowingphoneticenviron didnotspeakCherokeeindicatedmethattheir to parents,who mentsintheSmokyMountainregionofWesternNorthCarolina. werefluentinEnglishandCherokee,chosenottoteachtheirchil Williams(1992:14)alsocontendsthatin /ay/AppalachianEnglish drenCherokeebecauseoftheirexperiencesintheboarding

ismostoftenmonophthongal,and,althoughhedoesutilizethe schools.Again,alltheteacherswerewhiteandfewofthemwere

classicexampleofthegeneralSouthernpronunciationof[a:s]for fromtheSouth,sotheyalsoarenotexpectedtohavebeenagents

ice,hedoesnotgointoadiscussionoftheeffectoffollowing oftransmissionformonophthongal/ay/. phoneticenvironmentonthepatterningofthevariable.Pederson SnowbirdCherokeeattendedanall-IndianBIAday

schoolfortheelementarygradesuntil1965(Neely1991:31).

Snowbirdstudentswhowishedtoattendhighschoolwereforced *ThankstoWilliamA.Kretzschmar,Jr.forprovidingthelistmanuscript.

191 192 Volume4.1(1997) Anderson U.PermWorkingPapersinLinguistics AdaptiveSociophoneticStrategies

thisgroupheldjobs,suchasweldingandboiler-making,thattook toleavetheareatoattendboardingschoolseitherontheQualla

themoutoftheregionforextensiveperiodsoftime.Itisimpor Boundaryor outofstate.Understandably,manyolderSnowbird

tant,also,tokeepinmindthatmiddle-agedSnowbirdspeakersin residentschosenottoattendhighschool.TheSnowbirdday

bothgroupsattendedhighschoolwiththeirwhiteneighborsafter school,whichwasinoperationuntil1965,wasrunby twonon

theSnowbirdSchoolclosedinthemidsixties,andyoungerspeak localwhiteteachers.StudentswereallowedtospeakCherokeeto

ersattendedthepublicschoolinRobinsvillc eachother.Again,wedonotexpecttheseteacherstohave been

/ay/monophthongizers,andoneinformantreferredtoonethese of

4. Monophthongal/ay/inAppalachianand teachers"theYankee." as In1954theboardingschoolonthe

QuallaBoundaryclosedandSnowbirdstudentsbeganattending CherokeeEnglish

GrahamCounty'sRobinsvillcHighSchool(Neely1991:31)where

monophthongizationfor/ay/wouldhave beenthelanguagelearn Usingtheprecedingsociolinguisticbackgroundasa framework,

ingmodel. nowconsidertheincidenceof/ay/monophthongizationinthree

Finally,itisimportanttoconsiderthedifferentgroupsof speakergroups:low-interactionCherokee,high-interactionChero

speakerswithinSnowbird.Therearestrikingdifferencesinterms kee,andtheexternalreferencegroupofAppalachianwhites.The

offrequencyofcontactwithwhiteGrahamCountyresidents.I whiteexternalreferencegroupconsistsoffivemalesandfourfe

dividedtheCherokeesinthisstudyintotwogroupsbasedprimar malesranginginagefrom24to90.Table1givestherawfigures

ilyoninteractionfrequencywiththesurroundingwhitecommu andmonophthongizationpercentagesforthe threegroupsbysev

nity.Cherokeesthatfallunderthecategory"low-interaction"are eralfollowingphoneticenvironments:liquid,nasal,voicedobstru thoseCherokeeswhohavehadminimalcontactwithwhites.They ent,voicelessobstruent,wordboundary+ vowel(asineyeap

typicallyhavenotworkedoutsidethecommunityorintermarried pointment),wordboundary+consonant(asinliedown)andutter

withwhites.Speakersfromthisgroupincludesevenwomen ancefinalposition(asinOh.my).

ranginginagefrom37to83andsixmenranginginagefrom31 Table1indicatesthathigh-interactionspeakershavea

to94.Allofthespeakersinthisgroup,exceptfortheoneQualla significantlyhigherpercentagerateformonophthongizationthan

Boundarywomanwhoisincludedinthisanalysisonlyforpre dolow-interactionspeakers.In thedataunderinvestigation,high-

liminarycomparativepurposes,havemaintainedregularsocial interactionspeakersweremonophthongalfor/ay/mostofteninthe networksprimarilythe withinSnowbirdCommunityandhave followingenvironmentofliquid,followedbyvoicelessandvoiced

marriedotherCherokees.Allspeakersinthisgroup spokeChero obstruents,wordboundary+ consonant,andnasal.Rawpercent

keeastheirfirstlanguageanddid notlearnEnglishuntiltheyat agesareclearlymuchlowerformonophthongizationinthefollow

tendedelementaryschool. ingenvironmentsofwordboundary+pauseandwordboundary+

Cherokeesclassifiedas"high-interaction"tendtohave vowel.

moreextensivecontactwiththesurroundingwhitecommunityin Low-interactionCherokeeEnglishspeakersalsopartici

theirjobsand,insomecases,throughmarryingmonolingual pateinthemonophthongizationof/ay/,butnotnearlytotheextent

whites.Speakerscomprisingthisgroupconsistsofthreefemales, oftheirhigh-interactioncounterpartsorwhitecohorts.Low-

ranginginagefrom16toearly50's,andelevenmenrangingin interactionCherokeesshowthehighestincidenceofmonoph

agesfrom22to83.Threeofthemeninthisgroupmarriedmono thongizationwiththefollowingenvironmentsofvoicelessand

lingualwhitewomen,andallspeakersinthisgroup,withtheex voicedobstruents,followedbypre-nasalandpre-wordboundary+

ceptionofthesixteen-year-oldstudent,haveprimarilyheldjobs consonantenvironments.

whichbroughtthemintocontactwithlocalwhites,suchasforest

serviceandwagelaborjobs.Additionally,severalofdiemenin

193 194 Volume 4.1(1997) Adaptive Sociophonetic Strategies Anderson V. Perm Working Papers in Linguistics

Appalachian whites are nearly categorical monophthong- VI Obst. Word Word Word izers of /ay/. Only one speaker, a forty-seven year-old male, has Bound + Bound* Bound + even slight evidence of diphthongal /ay/, which occurred twice Vowel Con. Pause with a following environment of voiceless obstruent. The age ay range of the speakers in this group (the oldest being ninety) indi Speaker a: ay a: ay a: ay a: cates that white speakers in this region have been ungliding in all Groups environments at least since the early part of this century. Low- 98 76 0 27 7 \b 0 7 The results of a VARBRUL analysis, including bom in Interaction ternal and external factor groups, is provided in Table 2. Cherokee Internal constraints consist of the following phonetic en English vironments: nasal, voiced and voiceless obstruents, word boundary n=13 56.3 % 0% 31.8% 0% + consonant, word boundary + vowel, and word boundary + pause. Pre-liquid following environment is not included as a constraint High- 170 38 6 18 14 t> 6 11 Interaction Cherokee Liquid Nasal Vd Obst English n=14 81.7 % 25.0% 73.7% 35.3 %

Speaker a: ay a: ay a: ay 0 Groups Appalachian 122 2 11 0 25 0 21 Low- 0 0 35 56 51 50 White Interaction n=9 98.4 % 100% 100% 100% Cherokee English Table 1-continued. Incidence of/ay/ Monophthongization for n=13 NA 38.5 % 50.5% Three Speaker Groups Monoph. High- 14 0 37 17 69 20 because it was thrown out as a knockout constraint in the initial Interaction run of VARBRUL. External constraints consist of low-interaction Cherokee and high-interaction Cherokee English speaker groups. English The data indicates that high-interaction speakers favor n=14 100% 68.5 % 77.5 % monophthongal /ay/ over low-interaction speakers. Results of ANOVA tests, given in Table 3, indicate that the correlation be Appalachian 9 0 89 0 90 0 tween group affiliation (high-interaction, low-interaction, and White white) and monophthongization of /ay/ is statistically significant n=9 100 % 100 % 100 % atthep<00l level. The VARBRUL weightings indicate that the following environments of voiceless and voiced obstruents most strongly Table 1. Incidence of/ay/ Monophthongization for Three favor monophthongization, followed by nasals and word boundary Speaker Groups (continued on the next page). + consonant. Clearly, the following environments of word

195 196 U. Penn Working Papers in Linguistics Volume 4.1 (1997) Adaptive Sociophonetic Strategies Anderson

English by Wolfram and Christian (1976:64). The constraint order Table 2. VARBRUL Probabilities for Cherokee English suggests a disyllabic interpretation at the end Input Probability=.6O (Chi-Square/ Ce!K334) of a word boundary when the next word starts with a vowel. In other words, since fyl is being interpreted as the onset of the next

Social Factors: syllable in the source language, it follows that upgliding in the target language is expected to occur most frequently in the envi Low-Interaction Cherokee=.34 High-Interaction Cherokee=.66 ronment of word boundary + vowel. Cherokee is a CV language,

Linguistic Factors: and this is the expected pattern of interference. Both Cherokee English speaker groups show evidence of VD Obstruent=.56 VL Obstruent=.62 monophthongal [a:], although high-interaction speakers clearly Nasal=.44 Word Bound. + Con-.42 favor monophthongization over low-interaction speakers. One Word Bound. + Pause= . 13 Word Bound. + Vowel=.07 potential explanation for monophthongal [a:] in Cherokee English Table 3. ANOVA tests of significance of monophthongization is source language interference. In Cherokee, [a] is mono of /ay/ and speaker group affiliation phthongal except when /a/ is followed by vowels other than /a/ or l\S in the underlying form (Huff 1977:23). Weinreich (1968) maintains that phonological interference is the result of bilinguals Source Sum of degrees of Mean F Squares freedom Square identifying a phoneme in the target language with a phoneme from the source language and then subjecting this phoneme to the pho between 1.588 2 .794 26.47* nological rules of the first language when reproducing it in its sec within 1.006 35 .030 ond language production. More specifically, Romaine (1995:53) total 2.594 notes that this type of interference may result in a process of over- ♦p<.001 differentiation, which occurs when speakers transfer phonological boundary + pause, with VARBRUL weighting of .13, and word distinctions from the source language to sounds in the target lan boundary + vowel, with VARBRUL weighting of .07, disfavor guage. Source language interference may play an important role monophthongization. What, then, are possible explanations for the patterns in both monophthongal [a:] and diphthongal [al]. In this case, suggested by the analysis? The fact that Cherokee English speak speakers show transfer in their English by ungliding, or deleting /i/ ers, particularly low-interaction speakers, are not typically when it follows /a/, unless /a/ is followed by a vowel other than /a/ monophthongizers of/ay/ in the environment of a following word or /i/, in which case it is interpreted as the Cherokee /a/ plus a or syllable boundary followed by either another vowel or a pause vowel-glide sequence involving epenthetic [y] and thus is upglided is most reasonably attributed to source language interference. As to match the corresponding pattern in the source language. This noted earlier, although Cherokee has no clear-cut cases of tautosyl- explanation accounts for both the Cherokee English monoph labic vowel-glide sequences such as [ay], combinations of vowel- thongization of /ay/ and the upgliding of /ay#/ with a following glide sequences such as [aye] do occur. The constraint order for environment of word boundary + vowel. monophthongization in Cherokee English (voiceless obstruent > Although source language interference can account for voiced obstruent > nasal > word boundary + consonant > word both realizations of the variant, monophthongal [a:] and diphthon boundary + pause > word boundary + vowel) is a reversal of the gal [al], the process of dialect assimilation also surely must play an traditional Southern white pattern and the pattern of pause > important role in the monophthongization of /ay/ in Cherokee voiced obstruent > voiceless obstruent described for Appalachian English. Monophthongal [a:] is a pervasive phenomenon of the

197 198 Volume4.1(1997) Anderson U.PernWorkingPapersinLinguistics AdaptiveSociophoneticStrategies

outtheSouth(Baileyetal.1996);itiscertainlyprominentinthe mountainwhitecontactcommunity.Sincereservationtractsare

whitecontactcommunitywherespeakersshownear-categorical interspersedwithprivatetractsoflandownedbywhites,Snowbird

monophthongizationregardlessoffollowingenvironment.Al Cherokeeshavehadwhiteneighborssincetheypurchasedtheir

thoughmonophthongal[a:]inSmokyMountainEnglishisnowa landsaftertheremoval.Middle-agedtoyoungerCherokeesat

generalphoneticprocess,prevoicelessmonophthongizationissali tendedGrahamCountypublicschools,andCherokeesinvolvedin

entsocially,particularlytonon-Southerners.Perhapsthe current wage-laborindustryworkwithwhites.Monophthongization,es

contactmodelofmonophthongizationinallphoneticenviron peciallyforhigh-interactionspeakerswhohavea highfrequency mentsandthesaliencyofprevoicelessmonophthong-izationhave ofcontactwithwhites,couldovert beassimilationofthesurround affectedthevariablelevelsofCherokeeEnglishspeakerswhoas ingwhitedialectnorm.Thedifferencesbetweenspeakergroupsin

similatetothedialectnormofthecontactcommunity. theanalysissupportthisexplanation.High-interactionCherokees

haveaVARBRULratingof.66formonophthongization;low-

interactionCherokeesreceivedaVARBRULweightingatalmost 5. Conclusion

halfthefigureoftheirhigh-interactioncounterparts.TheANOVA

analysisalsodemonstratesthesignificanceofmonophthongization Inhersocio-culturalstudyoftheSnowbirdCommunity,anthro

andgroupaffiliation. pologistSharlotteNeely(1991)describestheSnowbirdCherokee

TabulationsforthetwospeakersfromQuallaBoundary,a as"persistent",andthistermalso canbe usedinadescriptionof

marriedcouplebothaged83.alsosupporttheexplanationthatthe thecommunity'slinguisticsituation.TheSnowbirdpeoplehave

participationinmonophthongal[a:]mayrepresentovertassimila alwaysbeenpeopleofpersistence.Thisisevidentintheirrefusal

tiontothedialectnormsofthewhitecontactcommunity.Both toberemovedontheTrailofTearsin1830andintheircontinued

speakerslearnedCherokeeastheirfirstlanguageandattendedthe occupationoftheirancestralhomeland.SinceNativeAmericans

BIA-runboardingschoolonQuallaBoundary.Thewoman,cate couldnotlegallypurchaselandatthattime,theyenlistedthehelp

gorizedasalow-interactionspeaker,wasahomemakerandthus ofthreelocalwhitemenwhopurchasedthelandfortheChero

hadlittleneedtointeractwithwhites.However,theman,catego keesintheirownnames.Thissituationissignificantbecauseit

rizedasahigh-interactionspeaker,foughtinWorldWarIand illustrateswhatseemstobetheprimarystrategythiscommunity

workedfortheparkserviceformanyyears.Hehadamuchhigher usestomaintainitsancestrallanguageandothercharacteristics

incidenceofmonophthongizationthandidhiswife.Takinginto associatedwithculturalautonomy.Low-frequencyCherokee

considerationallfollowingphoneticenvironments,themanreal Englishspeakers'limitedinteractionwithwhitesisreflectedin

izedthevariantasmonophthongal[a:]in66percentofhistokens. theirlimitedparticipationinthemonophthongizationof/ay/.

Hiswife,however,realizedthevariantasmonophthongal[a:]in High-frequencyCherokeeEnglishspeakersshowmoreassimila

only16.7percentofhertokens. tiontothecontactnormofmonophthongal[a:],buteventheydo

Theconstrainthierarchiesformonophthongizationin nottypicallydisplaymonophthongal[a:]inthelinguisticenviron

CherokeeEnglishmustalsobetakenintoaccountinanexplana mentwhereupglidingwouldbeexpectedinthesourcelanguage.

tionoftheanalysis.TheconstrainthierarchiesofCherokeeEnglish Thesepatternssuggestamixedalignment,acombinationofsource

donotfallinlinewiththetypicalSouthernconstraintpatternin languageinterferenceanddialectassimilationworkingtogetherto

whichprevoicedandprenasalenvironmentsfavormonophthongi affectthevariablenormsofthecommunity.Thelinguisticsitua

zationoverprevoicelessenvironments.Theyare,infact,reversed tionofthisgroupisalsoindicativeofSnowbird'sabilitybothto

inCherokeeEnglishwheremonophthongizationisslightlyfavored persistinculturaltraditionandtobeadaptiveintheirdealingswith

inprevoicelessenvironments.Researchhasshownthatmono thesignificantlylargermajorityofGrahamCounty's Appalachian

phthongal[a:]inprevoicelessenvironmentsisspreadingthrough-

199 200 Volume 4.1(1997) Adaptive Sociophonetic Strategies Anderson U. Perm Working Papers in Linguistics

Leap, William L. ed. (1977). Studies in Southwestern Indian English. San white population. What appears at first glance to be an overt as similative phenomenon, and may even be utilized as such— Antonio: Trinity University. Leechman, Douglas and Robert A. Hall, Jr. (1955). "American Indian especially in the case of high-interaction Cherokees—does not Pidgin English: Attestations and Grammatical Peculiarities." preclude substratal effects of source language transfer. Nor does American Speech 30: 163-171. contact-induced language change necessarily reflect language Neely, Sharlotte (1991). Snowbird Cherokees: People of Persistence. change as it occurred in the contact community. This mixed Athens: University of Georgia. alignment is one way a group can be both adaptive in regard to Pederson, Lee (1983). East Tennessee Folk Speech. New York: Verlag pervasive external dialect norms and, at the same time, maintain Peter Lang. important cultural and social distinctions. Romaine, Suzanne (1995). Bilingualism. Oxford: Blackwell Publishers, Inc. Thomason, Sarah and Terrence Kaufman ( 1988). Language Contact, Creolization, and Genetic Linguistics. Berkeley: University of References California Press. Weinreich, Uriel (1968). Languages in Contact. Mouton Press. [First Anderson, Bridget, Jessica Schridcr, and Walt Wolfram (1996). Cherokee edition 1953. New York: Linguistic Circle of New York Publi English in the Great Smoky Mountains: A Continuum of So- cation No. 2]. ciotinguistic Assimilation. Paper presented at SECOL 52. Col Williams, Cratis D. (1992). Southern Mountain Speech. Berea: Berea lege Station, TX. College Press. Bailey, Guy, Tom Wikle, Jan Tillery, and Lori Sand (1996). "The Conse Wolfram, Walt, Donna Christian, William L. Leap, and Lance Potter quences of Catastrophic Events: An Example from the Ameri (1979). Variability in the English of Two Indian Communities can Southwest." Sociolinguistic Variation: Data, Theory, and and its Effect on Reading and Writing. Washington, DC: Center Analysis. Selected Papers from NWAV 23 at Stanford. Ed. Jen for Applied Linguistics. nifer Arnold, Rence Blake, Brad Davidson, Scott Schwenter, Wolfram, Walt (1980). "Dynamic Dimensions of Language Influence: and Julie Soloman. Stanford: CSLI Publications. The Case of American Indian English." Howard Giles, W. Peter Bernstein, Cynthia and Elizabeth Gregory (1994). The Social Distribution Robinson, and Philip M. Smith (eds.) Language: Social Psycho of Glide Shortened/ai in LAGS. Paper Presented at SECOL 50. logical Perspectives. Oxford: Pergamon Press. 377-388. Memphis, TN. Wolfram, Walt (1984). "Unmarked Tense in American Indian English." Churchill, Ward (1994). Indians Are Us? Monroe, Maine: Common American Speech 59: 31-50. Courage Press. Wolfram, Walt and Donna Christian (1976). Appalachian Speech. Arling Craig, Beth (1991). "American Indian English." English World Wide ton: the Center for Applied Linguistics. 12:25-61. Hall, Joseph S. (1942). "The Phonetics of Great Smoky Mountain Speech." American Speech Reprints and Monographs, No. 4. Department of English New York: Columbia University Press. Norm Carolina State University Huff, Charles (1977). The Phonology of Qualla Cherokee, unpublished Raleigh, NC 27607 Master's Thesis. University of North Carolina at Chapel Hill bianders@unity. ncsu.edu Department of Linguistics. King, Duane (1975). A Grammar and Dictionary of the Cherokee Lan guage. Unpublished Ph.D.dissertation. University of Georgia. Kurath, Hans, and Raven I. McDavid, Jr. (1961). The Pronunciation of English in the Atlantic States. Ann Arbor: University of Michi gan Press.

201 202