Linguistic Change, Social Network and Speaker Innovation Author(s): James Milroy and Lesley Milroy Source: Journal of Linguistics, Vol. 21, No. 2 (Sep., 1985), pp. 339-384 Published by: Cambridge University Press

Stable URL: http://www.jstor.org/stable/4175792 .

Accessed: 02/06/2011 19:03

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at . http://www.jstor.org/page/info/about/policies/terms.jsp. JSTOR's Terms and Conditions of Use provides, in part, that unless

you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at .

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Cambridge University Press is collaborating with JSTOR to digitize, preserve and extend access to Journal of Linguistics.

http://www.jstor.org

J. Linguistics 2I (I985),

_{339-384.Printedin} GreatBritain

_{Linguistic change,} social network and _speaker

I.

innovation'

MILROY

Departmentof Linguistics Universityof _Sheffield

JAMES

AND LESLEY

MILROY

Departmentof Speech, Universityof Newcastle upon Tyne

(Received9 JanuaryI985)

INTRODUCTION

Thispaperis concernedwiththe socialmechanismsof linguisticchange,and webeginby notingthedistinctiondrawnby Bynon(I977) betweentwo _quite differentapproachesto the _study of _{linguisticchange.The firstand more} idealized,associatedinitiallywith traditionalnineteenthcenturyhistorical linguistics,involvesthe studyof successive'states of the _{language',states} reconstructedby the applicationof comparativetechniquesto necessarily partialhistoricalrecords.Generalizations(in the form of laws) about the relationshipsbetweenthesestatesmaythenbe made,andmorerecentlythe specificationof 'possible'and'impossible'processesof changehasbeenseen as an importanttheoreticalgoal. Thesecondapproach,associatedwithmodernquantitativesociolinguistics, involveslessidealizationof thedatabase.Animportantobjectiveis to specify HOwlanguagespass from state A to state B in termsof both the social processesinvolvedand the effecton linguisticstructureof a givenchange. The major goal is to develop a theory which is sensitiveboth to the constrainedand regularnatureof changeand to its relationshipwithsocial structure. Some sociolinguistshave borrowedquite heavilyfrom older scholars- notablyBailey,whohastriedto applya 'wave'modelto _{contemporarydata} (Bailey,1973). Similarly,Labovhas assessed,in the lightof recentfindings, thetheoreticalapproachesof nineteenthcenturyhistoricallinguistics;in _one article,he compares'lexicaldiffusion'modelsof changewith _{those which} claimthatchangecomesaboutasa resultof theoperationof regularphonetic

_[i] Weacknowledgewiththankshelpfulcommentson a _{previousdraftof} thispaperby John Harris,DickHudson,BenRampton,PeterTrudgillandNigelVincent.Versionsofdifferent partsof it werereadat the SociolinguisticsSymposiumin Liverpooland the Societas LinguisticaEuropaeameetingin Manchester,bothin September1984.Helpfulcomments werereceivedfromparticipantsatbothmeetings.FinancialsupportfortheBelfastresearch _{on whichpartsof the paperare} basedwas receivedfromthe SocialScienceResearch _{Council(grantsHR3771,HR5777).The} secondauthoralso receivedgenerousfinancial supportfromtheSimonFund,duringhertenureof a SeniorSimonResearchFellowship

at the Universityof Manchester,I982-3.

Thishelpis gratefullyacknowledged.

339

JAMES MILROY

AND

LESLEY

MILROY

laws. Conversely, the presentcan be used to _explain the _past, i.e. to shed _light on historical linguistic problems _{(Labov, 1974a; Milroy & Harris, I980).} What seems to be well established now is that _{variability of a structured} and _regular kind is characteristic of normal _{language use and is a key to} understanding mechanisms of linguistic _change. At the _phonological level _{change appears} to _{affect contextually} defined subsets of _{phonological classes} in a (generally) regular way, spreading _through the _community in waves in

a manner controlled by extra-linguistic factors such as the _{age, sex,} social _status and _geographical location of the _{speaker. Except} where the _{ongoing change originates} with a _{high-status group} and is more or less _consciously adopted by others, spontaneous speech _appearsto be affectedearlierthan the speech characteristic of more careful _styles. These general principles have _{emerged fairly clearly} from the work of Labov carried out over the last two decades (see _{Labov, 1972)} and _are confirmed elsewhere. To _{exemplify the quantitative approach, we comment} briefly on one particular study.

account of Souletan, a dialect of _{Gascon, quite explicitly}

attempts to bring together the concerns of traditional historical _linguistics and those of sociolinguistics. She examines the _{relationships between}

long-term phonological changes which affectwhole classes of _{linguistic items,} and the 'competence' of speakers who are _involved in an _{ongoing linguistic} change. Historical, geographical and synchronic variable _data are _analysed

to illuminate the processes involved in an upwardchain shift of the back vowel

system of Souletan; the back chain shift _{is particularly advanced here and}

is still in progress.

Eckert characterises the change in _{terms of waves, which affect one word} class at a _time; as we might now predict, items _lagging behind in _{the shift} occur in the speech of older informants, as stylistic variants. Thus, the item sulament 'only' occurs with [a]in careful speech, but _{with the more innovative}

[o] in rapid connected speech. Arguing that phonological rules (which reflect speaker competence) ought to be written in such a _{way as to reveal this} pattern, Eckert gives a formal characterization of such a variable _rule. If _the language choices open to the individual are placed in _{this broader context,} they may be seen as reflections of 'earlier' and 'later' _{overlapping states of}

Eckert's (I980)

a	dynamic phonological system. The capacity of a variable rule formalism
to	handle linguistic constraints on the implementation of the rule may be seen

as _{characterizing} successive 'waves' of _{the change. Thus, it is argued,} individual language behaviour is related _{to historical changes by rules which} are seen as reflecting the competence of _{a speaker whose range of linguistic}

choices is congruent with the waves of _{change which proceed regularly} through time and space. In a sense, Eckert's work, like much of our own, straddles the two approaches distinguished by Bynon, attempting to see how they fit _together. Micro-level studies of this kind which are both 'sociolinguistic' and

340

LINGUISTIC

_CHANGE

'historical'appearto _supportthe claims _{of Weinreich,Labov & Herzog} (I968) thatlinguisticinnovationsmovesystematicallythroughspace(social, geographicaland historical)affectinglinguisticstructurealso in an _ordered manner.The task of explaininglinguisticchange _{was, they argued,best dividedinto} fivemainareas. These are, first, the _very broad _problemof UNIVERSAL CONSTRAINTSon

possible _changes. As Weinreich et al. _{note (ioi)}

this _{is part of a larger}

theoreticallinguisticissueand falls _{beyondthe scopeof quantitativesocio-} linguistics.Within historicallinguisticsit has been _{examinedby many} scholarsincludingLass(I980) _{and Vennemann(I983).} Second,theTRANSITIONproblemconcernsthe'interveningstageswhichcan be observed,orwhichmustbe _{posited,betweenanytwo} formsof a _language definedfora _{languagecommunityat} differenttimes'(Weinreichet _{al., I968:}

_IOI). As wehavealreadynoted,quantitativeanalysishascontributeda _great dealhere,showingclearlythat _{transitionis evidencedby variationbetween} conservativeandinnovatoryforms,withthe _{formergraduallygivingwayto} the latteras relativefrequencychanges. TheEMBEDDINGproblemis concernedwithdeterminingregularpatternsin both the linguisticand the extra-linguisticcontextof _{change.Includedhere} wouldbe an accountof the phoneticenvironmentsmost _{favouringchange} andthe relativerankingof theseenvironments.Muchof _{Labov'sownwork}

has addressedthisissue(andsee _{also Eckert,I980).}

the vowel analyses of _{Labov, Yaeger and Steiner (1972)}

Martinet's'chain shift' model.EXTRA-LINGUISTICaspectsof _{the embedding} questionmaybetackledindirectlybyinspectingthedistributionofinnovatory linguisticformsin speechcommunities.Labovciteshisownwork,in _addition to thatof Trudgillin NorwichandCedergrenin _{PanamaCity,as supporting} the generalizationthatwheresoundchangein _{progressis located,linguistic}

variablesdisplaya curvilinearpatternof distribution(oftenshowingup _on a graphas an unexpected'crossover'pattern).Innovatinggroupsappearto

be locatedcentrallyin the socialhierarchy,and _{arecharacterizedby Labov}

as upper-workingor lower-middleclass. _{(Labov, I980:}

_Moreover,

youngerspeakersuse _{more innovatoryforms than older speakers(both} quantitativelyand qualitatively)and _{again accordingto Labov'sanalysis,} sexualdifferentiationof _{speechoften plays a major(but as yet not clearly}

understood)role in linguisticchange. The EVALUATION problem pertains principallyto social _responsesto change'at alllevelsof awareness,fromovertdiscussionto _{reactionsthatare} quiteinaccessibleto introspection'(Labov,I982: _28). Thisembracesnotions of prestige,attitudesto languages(bothovertandcovert),aswellaslinguistic stereotypingand notionsof correctness. The principalcontributionof _{Labov himself to the EMBEDDING and} EVALUATIONproblemshas _{been,particularlyin his New YorkCitystudy,to} providea GENERALmodelof thesociallocationof a linguisticinnovationand

_{Includedalso wouldbe}

_{which follow}

_254).

341

JAMES MILROY

AND

LESLEY

MILROY

of the mannerin which it spreads from a central point upwardsand downwardsthrougha speechcommunity.Shortlywe shalllook at someof the problemsassociatedwith this modeland indeedin muchof this _paper we shallbe presentinga critiqueof partsof it. Until fairlyrecently,Labovhad not attemptedto tacklethe fifthareaof investigationoutlinedby Weinreichet _al.: this is the ACTUATIONproblem, articulatedherein its most _{challengingform:}

Whydo changesin a structuralfeaturetakeplacein a _{particularlanguage} at a giventime,butnot in otherlanguageswiththe samefeature,or in the samelanguageat othertimes?Thisactuationproblemmaybe _regardedas the veryheartof the matter(Weinreichet al., _{I968: I02).}

It is the actuationproblemwhichwe discussin this paper;appropriate strategiesaresuggestedforaddressingit andsomepartialanswersareoffered to thequestionsposedbyWeinreichet al.Mostimportantly,wetryto _explain why investigatorshave failedto makemuchheadwayin tacklingthe issue whichwas describedin I968 as 'the veryheartof the matter'. Sucha programmeas waspresentedby Weinreichet al. is not _necessarily thebestwayof organizinga systematicstudyof linguisticchange.Whilewe do not attempthere to offera comprehensivecritiqueof the paper,some difficultiesshouldbe _noted. First, although the five aspects of the problem of change have been presentedas relativelydiscrete,theydo in fact overlap.Whilethisdoes _not in itselfnecessarilyconstitutea difficulty,a readingof Labov'sI982 article whichreviewsworkon the problemsup to that timeshowsclearlythathis interpretationof the issuescoveredby eachof the fivecategoriesis _different fromWeinreich's(it was in fact Weinreichwho was mainlyresponsiblefor the early sectionsof the I968 paper from which we have quoted in _this section).So whilewehavedrawnfreelyon Labov'sreview,thecategoriesare discussedin termsof Weinreich'sformulationwherethereappearsto be a discrepancy.Thisdifficultyis compoundedby thefactthatweourselveshave sometimescategorizeda phenomenonratherdifferentlyfrom Labov. For example,whilehe regardsdiffusionof innovationas partof the TRANSITION _{question,wehavetreatedit} hereasanaspectof ACTUATION.Nowitiscertainly clearthat no _{singleaspectof linguisticchangecan} be discussedcoherently withoutreferenceto at leastsomeof the _{othersspecifiedby Weinreichet al.} Butsincethedisagreementswhichemergewhenattemptsaremadeto _specify howphenomenashouldbe categorizedaresometimesquiteradical,it seems reasonableto suggestthat the distinctionsdrawnin the _I968 paperarenot finaland that theyrequirefurthercriticalconsiderationif theyareto _serve as a comprehensiveprogrammefor the studyof linguisticchange. We look _brieflynow at Labov'sattemptto tacklethe actuationproblem by firstlocatingthe innovatorsthemselvesand then _{examiningtheirsocial}

342

LINGUISTIC

CHANGE

characteristics and

neighbourhoods2.His mainconclusionsareas _follows.

_{relationships} within

their

own

_{(Philadelphian)}

(i)	Speakerswho lead soundchangeare thosewith the higheststatusin
	theirlocalcommunitiesas measuredby a social-classindex.
(2)	Amongpersonsof equalstatus'the most advancedspeakersare the personswith the largestnumberof local contactswithinthe neigh- bourhood,yet who have at the sametime the highestproportionof
	theiracquaintancesoutsidethe neighbourhood'(I980:	26I).	Labov

thengoeson to comment'Thuswe havea portraitof individualswith the highestlocalprestigewho are _responsiveto a _{somewhatbroader} formof prestigeat the nextlargerlevelof socialcommunication.'

Bothpointsarerelevanthere.Labovpresentsin _{effectone superordinate} locus of change,viz.a centralpositionin the statushierarchy(andherehis modelis implicitlyonedependenton theexistenceof socialstratification)and onemorerefinedormicro-levellocus,withina _{groupof roughlyequalstatus.} The diffusionof changeis accomplishedby individualswho havemanyties withintheclose-knitcommunityandwhoalsohavea _{relativelylargenumber} of outsidecontacts.Ourown arguments,whicharriveat conclusionsrather differentfrom those _{of Labov, focus almost entirelyon the position of} linguisticinnovatorsin localisednetworkswhicharemadeup _{of personsof} roughlyequalstatus.Weshallalsodiscussmorebroadlythetypeof _network structureassociatedwith(oftenrapid)linguisticchangeand _{areless willing} thanLabovevidentlyis to presenta modelbasedultimatelyon _{statusorclass.} Afterall, theseareno morethanrathercontroversialconstructs(seeHalsey, 1978, for an accessiblediscussion)and the universalapplicabilityof _such constructsto theoriesof _{changeis dubious.} We pass now to a discussionof _{what is meant by the term 'linguistic} change', highlightingsome problems and ambiguities.Changes in _the realizationof twoUlstervowelsarethenreviewedinsomedetail,to _exemplify the principlethatevidenceof _{linguisticchangemay be foundin datawhich} are variableon historical,geographicaland _{social dimensions.Using the} networkconceptdevelopedpreviouslyin _{this researchprogramme(Milroy}

& Milroy, I978; L. Milroy, _I980), the informalsocial _{ties of linguistically} innovativegroupsare examined,and a model of linguisticchange,based

_partlyon ourownconclusionsandpartlyon _{workby Granovetter(I973)}

presented.Thismodel,whichsuggeststhatinnovationsflowfromonegroup

_is

_[21A 'neighbourhood'study(Labov'sterm)is _{distinctfroma surveyin thatno effortis made} to elicit comparabledata from isolatedindividualssampledin _{accordancewith some principleof} randomselection.Rather,the _{languageof speakersis investigatedwith} attentionto theirpositionin relationto othersin theirlocalneighbourhoods.Thus,both languageandsocialstructuremaybeexaminedin _{verymuchgreaterdepth,at theexpense} of somesocialandlinguisticbreadth.

343

JAMES MILROY

AND

_LESLEY

_MILROY

to anotherthrough'weak' networklinks _{is designedto offer a practical} solutionto an _{aspectof the} actuationproblem;as _{suchit is concernedwith} SPEAKERinnovation,of _{whichthe reflexin the languagesystemis a change} whichisalwaysobservedpostfactum.Finally,wesuggest(morespeculatively) that the model is capableof elucidatingparticularproblemsof _language changeandvariationwhichso _{farhave seemedquitemysterious.}

2.

LINGUISTIC

CHANGE

AND

SPEAKER

_INNOVATION

Although the ultimate aims of historicallinguisticsmay _{be to specify} universalsof change(whatis, or is not, a possiblechangeand,withinthe set of possiblechanges,thekindsof _{changethataremoreor lessPROBABLE),the} methodologyof historicallinguisticshas _{ALWAYS beencomparative.} Nineteenth-centurylinguistics('comparativephilology')aimedat _RECON- STRUCTIONof proto-languagesby comparisonof sisterlanguages,andso _the

termCOMPARATIVEacquired,inlinguistics,anassociationwithreconstruction3.

Here,however,we use the termCOMPARATIVEin _{a moregeneraland literal} sense, without any necessaryimplicationthat _{reconstructionis aimed at.} Thus,the comparisonof two attestedhistoricalstatesof _{the samelanguage} is also a comparativemethod. Sociolinguisticsalso uses a comparativemethod,in _{that the languageof} differentindividualsorgroupsis _{compared.Thedifferenceis thatthechanges} are observed,or arguedfor, at a micro-levelratherthana _macro-level. In a sociolinguisticanalysis,the _{observationof changeis narroweddown} to comparisonsbasedon ageandsexof speaker,stylisticvariationandsocial grouping;observedsynchronicvariationcan _{be viewedas the counterpart} ofchangeinthediachronicdimension.Inpractice,thesemicro-levelsynchronic patternsareusuallysupplementedby'real-time'observations.Thetestimony of nineteenth-centuryand early twentieth-centuryobserversof _{a speech-} communityare used to help to establishthe long-termdirectionof _change (Labov,I972: I63-I7I). Noticethatas _{soon as themethodologyis extended} to takeaccountof paststatesof language,it becomesto thatextentthesame as comparisonof two ormoreattestedpaststates.Insteadof _comparingtwo paststates,wearecomparinga _{presentstatewitha paststate.Theimportant} differencesthatdo existbetweensociolinguisticsandhistoricallinguistics(as describedby Bynon, 1977)dependfundamentallyon _{the fact that socio-} linguisticmethodis rootedin the present,whichmeansthat _{thereis direct} accessto therichdetailofvariationinspeech-communities.Thus,itispossible to observeveryfullyboththelinguisticandthesocialembeddingof _observed changes. It becomes possible to specify the constraintsimposed _{by a}

[31'In using the comparative _{method we contrast forms of two or}

more related languages _to

most simply by recon-

determine their precise relationship. We _{indicate this relationship}

structing the forms from which they developed' _{(Lehmann, I962: 83).}

344

LINGUISTIC

_CHANGE

pre-existinglanguagesystemon _{thepossibilitiesof changewithinthatsystem}

(Eckert, I980;

suggestand evaluatethe possiblesocialmotivationsof observedchanges. Thequestionof socialmotivationis not uncontroversial.Ontheonehand, thereis a view,pioneeredby Weinreich,LabovandHerzogandassumedin thispaper,thatthe _studyof socialmotivationsconstitutesanimportantpart of any possibleexplanationof _{change.On the otherhand,thereis a strong}

traditionin language study of separatinglanguagesfrom _speakersand looking for some of the ultimateexplanationsfor _{changein languagesas}

Labov, Yaeger & Steiner, _I972;

J. _{Milroy, I976)} and _to

systems. As Lass _(I980:

_{120) has put it:}

Linguistshave,I wouldmaintain,normallytreatedlanguageas _{if it were} in factan autonomousnaturalobject(or _{an autonomousformalsystem):}

'languagechanges'- it is not (necessarily)speakersthatchangeit It is temptingto suggestthat the separationof languagesfrom _speakers

is partlya hangoverfromthe nineteenth-centuryinsistenceon the _{'life' of} thelanguageindependentof speakers- a viewverycommonlyexpressed,e.g.

by Trench(i 888: _224):

tree . '.

_{'For a languagehas a life as trulyas a man or a}

Although functionalexplanations(avoidanceof _{homophony,etc., as}

discussedby Lass,I980:

theseandmostothercurrentexplanations(e.g.physiologicalandpsychological ones)do not normallymakea PRIORdistinctionbetweenspeakerbehaviour on theone hand,andlanguageas a formalsystemon theother;theyaddress themselvesto the explanationof changesobservedin languagesratherthan explanationof speaker-behaviour.Someof _{themare,in anycase, seriously} flawed(as Lasspointsout). Whatis clearis thatfunctionalexplanationsdo not addressthe ACTUATIONPROBLEMas _{formulatedby Weinreichet al. Such} explanationsmayaccountforsomeinstancesof,e.g.avoidanceofhomophony, buttheydo notexplainwhyhomophonywasNOT _{avoidedin otherinstances.} In general,theydo not explainwhya particularchangetookplaceat _{a given} timeand in a givenlanguageor dialect,but not in similarcircumstancesin otherlanguagesanddialectsorat _{othertimesin thesamelanguageordialect.} If we areto addresstheactuationproblem(whichis _{'the veryheartof the matter'),we} mustbreakwithtraditionandmaintainthatit _{is not languages} thatinnovate;itisspeakerswhoinnovate.Thereflexesof _{speaker-innovations}

are then observedin languagestates,wherethey _{appearas systematicand} rule-governedlinguisticchange. As the best-knownfindingsof sociolinguisticshavetendedto _concentrate on phoneticand phonologicalmatters,it _{has been possible for some to} dismissthemassuperficial,non-explanatoryandpurelydescriptive(Chomsky, 1975). Evenat thephonologicallevel,however,theseapproacheshavecalled intoquestionsomeof thetheoreticalpositionsof _{thedominant(Chomskyan)} paradigm.Partsof thephonologicalmodelproposedin _{Thesoundpatternof}

64-9o) seemto takespeaker-strategiesintoaccount,

345

JAMES MILROY

AND

LESLEY

_MILROY

if applied to variation in modern _English

English (Chomsky & Halle, I968),

phonology, necessarily distort and misrepresent the 'competence' of native _speakers. More suitable _{phonologies have,} in _practice, to be constructed _and

many of the _{assumptions of SPE phonology} are not borne out in such cases (J. Milroy, 1976; I98I). Furthermore, the findings of sociolinguistics are not confined, as is often _suggested, to the _description of _{configurations} of _surface PHONETICvariants: it is in _{sociolinguistic} work and _{not, as} far as we _know, in other approaches, that we can observe cases of rulechange in _{progress (and} therefore hope to explain such _phenomena). We now _briefly discuss an example: the gradual loss of _{/a/ raising} after velars in _Belfast. David Patterson (i86o) attests that _/a/ was raised to _{[E] after the velar} consonants /k, _{g/, and} lists _{given by Gregg (I964)} for the Ulster Scots _town of Larne suggest that this rule applied regardless of FOLLOWINGconsonant. In present-day Belfast we have _attested no _{cases of raising} after _/g/, and the rule is variable after /k/. It is variable _to the _{extent that male working class} speakerscan vary between [c]and [a, a] in the same lexical items. Their choice reflects the application of different rules, rather than application or non- application of a single rule. Thus TM _(Clonard) has _[kig . _{n]: 'can', rapidly} followed by threetokens of [kia . n]: 'can', in succeedingutterances.As we shall

see in Section _3, the trend in the _/a/

since Patterson's day (conditioned by the _{following consonant),} and in _these post-velar environments the rules _are in _{conflict. The choices open to speakers}

in monosyllables may be listed as.instructions, as follows:

_{system has been towards backing of /a/}

After /k/, choose either mid or _{low, unless /r/ follows, in which case} low must be selected, If mid is chosen, select low-mid, short _[a] before _{voiceless stop;}

otherwise select _{mid, long, (3)} If low is chosen, select short front [a] before _{voiceless stop; otherwise} select _long, back _{[a ].}

Clearly the rule for raising after velars is recessive: it has disappearedafter

/g/

'car'), and is otherwise _{variable for many speakers. For many younger East} Belfast speakers, however, it has actually disappeared. In general, it _{is the} following consonant _{more than the preceding one that dictates which} realizations of _/a/ are adopted. We shall see in Section 3 that the choice _listed in 3 (above) is somewhat idealized: in fact _{there are greater and lesser} probabilities of _backing in _{an ordered series (depending on following} consonant), and some environments are more _{likely than others to allow} back-raising and _rounding to _{[3]. (For other examples of rule-change in} progress, restructuringand _{merger, see J. Milroy, i984b; J. Milroy & Harris,}

(Patterson has care to represent the pronunciation of

(i)

(2)

and before /r/

1980).

In presentingsuch a configuration of change it is _{clear that we are primarily}

346

LINGUISTIC

_CHANGE

describinga state of languageratherthan the _{'competence'of individual} speakers.The speakersthemselvesmay, or _{may not,} have accessto all _the possiblevariants,and _(as shown _{in J. Milroy, i982a), some middle-class} speakersexhibitlittlevariation;individualsmayconvergeforallitemsEITHER onfront[a]orback[a].Suchspeakersmaybesaid- _{inalinguistically-oriented}

dimension- to have _'lost' the _rules for _{raisingand backingin different}

environments.In a speaker-orienteddimension,however,these _speakers merelydisplaya differentpattern,andwe cannotassumethattheyeverhad the rulesfor _{frontingand backingin} theiractivecompetence.Nor _{do we} knowwhethertheyareawareof themin theirpassivecompetence.In _other cases, speakersmay be observedto _vary in theirrealizationsof the _same lexicalitems in the same phonologicalenvironments;such _speakershave variablerules. Thus, when we consider speaker-competence,there are difficultiesin specifyingwhata linguisticchangeactuallyis andhow it is _{implemented.At} the macro-level,claimsfor _{changehave normallybeen assumedto reston} an observeddifferencebetweenStateA _{andStateB, andhavenot depended} on speakerintuitionorcompetence(twentieth-centuryspeakers,forinstance, arenot assumedto haveintuitionsaboutfourteenth-centurystates).At _the micro-level,in which _{observedchange depends on variationin speech-} communities,speakerintuitionhas _{been assumedto be relevant,in that}

speakersmay have access to both recessiveand incomingvariantsand know when to use them. Even _{at this level, however, it seems that}

speaker-behaviourvaries,andit is possiblethatindividualspeakershave- _to

a degree- differentialcompetenceand _{intuitions. The difficultyis that}

linguisticchange must presumablyoriginatein _{speakersrather than in} languages.We thereforefindit convenientto distinguishbetweenlinguistic CHANGE, on the one hand,and speakerINNOVATION _{on the other.It is the} originand diffusionof SPEAKERINNOVATIONSwith _{whichwe are concerned}

in this paper. Speakerinnovations,likeotherinnovations,maybe _{classifiedin termsof}

theirsuccessin subsequentdiffusion,as _follows:

(i)	A speakerinnovationmay fail to diffusebeyondthe speaker.
(2)	A speakerinnovationmaydiffuseintoa communitywithwhichhe/she

has contact,and go _{no further.} (3) A speakerinnovationmaydiffuseintoa _{communitywithwhichhe/she} has contactand then subsequentlydiffusefromthat _{communityinto} othercommunitiesvia a furtherinnovatorwho _{has tieswithboth the} relevantcommunities.Whenthe _{resultsof this processare observed,} we tendto labelthe resultsas 'linguisticchange'.The _{set of possible} communitiesthroughwhichsucha _{changecan diffuseis in principle} infinite,andalthoughlinguisticandsocialconstraintson _{a changecan}

347

JAMES MILROY

AND

_LESLEY

_MILROY

in some instances be specified, the limits _of POSSIBLEdiffusion _cannot be precisely stated - either in terms of _space and time or in terms _of the possible states of language or society that may favour or disfavour the change.

It is not suggestedin the present state of _{our knowledge, that the innovators} can be preciselylocated. The linguistic innovator to whom _{we referis as much} an idealization as Chomsky's 'native speaker-listener', and it is our aim to model the sources and processes of _linguistic innovation in _{more detail than} has been possible in the past. We consider arguments about probabilistic grammarsand _{the status of variable rules (Romaine, I98 I) to be, in principle,} irrelevant here. _{For, although much of the data presented in this paper has} been collected from speakersand (necessarily)subjectedto _{quantification,our} arguments are not based on quantities, but on processes that have been observed to take place in speech communities. _{Although such processes may} have been analysed quantitatively, they are not in _{themselves quantitative} phenomena. _{By using such methods, however, we may have made some}

progress in locating the idealized speaker-innovator. We end this section by commenting on _(I)-(3) above _{(pp. 347-348). Notice} that speaker innovation is not _{identical with linguistic change. As (I) implies,} some innovations _may not be _{accepted by a community and hence may not} lead to change. On the other hand, speakerinnovation _may lead to a _change

in one segment or part of the grammar,which then _sparksoff a chain reaction

that seems to be _{internal to the language system. Thus, in the English Great}

Vowel Shift, it may be argued that ME _{a was first raised, and that as a} consequence of _{this, the ME vowels above it in phonetic space werealso raised} (or diphthongized in _{the case of the highest vowel). In such a case, it is possible} that speaker innovation is _{relevant only to one vowel and that movements}

in the other vowels are motivated by the _{language system. Yet even here it}

must be admitted that speakers have been motivated to _{keep vowels distinct} within the system. With referenceto _{(3) above, we must also note that, as the limits of possible} speech communities (like _{the limits of social networks) cannot be specified,}

we do not know that a change observed to have entered a community (through the activities of _{certain speakersor groups) is in fact originalto those} who are observed to carry the innovation. The _{apparent innovation may} already have been well established in some _{other community, and this in turn} may have adopted the innovation from _{elsewhere. In observing change in a} given community, therefore, we do _{not know beforehand at what point in}

a cycle of change we have entered the _{community. Although, from a} synchronic point of view, certain individuals _{and groups may be identified}

as innovatory (see Section _{3 below) and as responsible for introducing an} innovation _to their _{immediate communities, it is possible that the change} concerned has _{had a long history elsewhere. We shall see that this is so in}

348

LINGUISTIC

CHANGE

the case of changesin the vowels _/E/ and /a/ in Belfast- to whichwe now turn.

_3. REVIEWING

_VOWELS

_/?/

THE EVIDENCE

FOR CHANGE

AND /a/

IN

HIBERNO-ENGLISH

IN PROGRESS:

THE

_{If we comparethe range of variationin} Belfastvernacularvowels with text-bookdescriptionsof RPit is obviousthatmanyof themhavea startlingly widephoneticrangeofrealizations(J.Milroy,1976; I98I; I982). Realizations _{of /a/ rangefrom[?]} or abovebeforevelarconsonants,as in bag, bang, etc, _{to /c/, in hand, bad, etc.} This is furthercomplicatedby a variationin vowel-lengthanddiphthongization.Briefly,vowelsin monosyllablestendto _{be short before voiceless stops} and before clusters consisting of _{sonorant+ voicelessstop; they are long} before fricativesand voiced con- sonants.Closingdiphthongs([ai]) can also occurbeforevoicedvelars,and centringdiphthongs([a-a][a-e])occur when the vowel is back, long and, especially,also raised and rounded. The range from front to back is representedin TableI.

[?-]	[m]	[a]	[a:]
bag	back	bat	bad	bad
bang		snap	grass	grass
		ant	hand	hand
		back	snap

Front only		Velarenvironments
Back only		Fricative& voiced consonantenvironments(excluding velars)
Front	back	Voiceless stop environments(excludingvelars); back variantsattestedonly amongstEastBelfastyouths

Table i

Simplifiedrepresentationof phoneticrangeof /a/ in Belfastvernacular,usingkey words

_{The rangefor /e/, in e.g. step, bed} is also wide. Qualitatively,the range _{is from[a]to around[e]:similarrulesof} vowel-lengthapply,withcentring _diphthongsof the _{type [q-a]} tendingto developin long environments(see

Table2).

Sucha widerangein twoneighbouringvowelsresultsin overlapping.Some

_{realizationsof /E/} are like realizationsof /a/,

backwardsin time,it is possibleto arguethat restructuringhas takenplace

349

and vice versa. Projecting

JAMES MILROY

AND

LESLEY

_MILROY

Mid

Low

_[e:,,e -a, ?'-, ? * _]

bed, bend,best (Fricativeand voiced consonantenvironments)

_{[a, w] wet, went} (Voicelessstop, _{sonorant+ voicelessstop environments)}

Table 2

Simplifiedrepresentationof phoneticrangeof _/?/ usingkey words

in Belfastvernacular,

at some _time in the _past and _that, for _example, the _/a/-/c/

distinction _may

havebeenneutralisedbeforevelars,with _{bagand beghavingbeenpossibly} identical.However,restructuring(with _{transferfrom /a/} to _/?/ and vice versa)cannot be adequatelydemonstratedfrom the _{present-dayevidence.} Althoughsomespeakershavedifficultyin _{disentanglingpairslikepack/peck,} speakersareawarein mostcasesthat[E]realizationsbeforevelarsaretokens of /a/, whereas[a]realizationsbeforevoicelessstops _{aretokensof /?/. As} raisingappliesto /a/ beforevoicelessandvoicedvelars,itemslikeback,bag areoftenrealizedwith [c].Howeversincelowrealizationsof /E/ applybefore ALL voicelessstops(includingthe velar),itemslikeneck,wreck(withvelars) tend to be realizedwith [a]. This resultsin an apparentflip-flop,and the followingexamplesaretypical:

'The back[bEk]of my _neck[nak]; 'Will _{you pay by} Access[EksEs]cardor _{by cheque [tfak]';} 'Jet _[tat] - lag [E 9]'.

Therearetwo overlappingsystems,informallystatedas _follows:

/a/-[E]/-

_{/?/+ [a]/-}

^Velar

VoicelessStop

The complexityof such _{systems, together with the range of socially} motivatedvariationthat _{occursin the realizationsof the vowels,presenta} considerablechallengeto our _{abilitiesto identifythe directionof changein} progress,but the sheeramountof variationprovidesmanyclues.First,we considerthe regionaland socialrangeof _{realizationsof /?/.}

(i) Raisingof /?/ FigureI showstheresultof a quantitativeanalysisof _/?/

Belfastouter-citycommunities(Andersonstownand _{Braniel)and a smaller} town(Lurgan)situatedI _{7milessouth-westof} Belfast.ThesymbolTindicates

a followingvoicelessstop or sonorant+voicelessstop _cluster;C$ indicates that the vowel is in the stressed syllable of a polysyllabicword (this

_{realizationsin two}

350

			LINGUISTIC		CHANGE
70	-	ANDERSTOWN (n = 1104)
60	-
50
40
30
20-
10
	_
0
		TC$D	TC$D	TC$D	TC$D	TC$D	TC$D
		a	e:		?	E:	e
70	-	BRANIEL	(n=	800)
60
50
	_
40
30
20
10
0
		TC$D	TC$D	TC$D	TC$D	TC$D	TC$D
		a			6		e
70-		LURGAN	(n=1484)
60
50-
40
30	-
20
10
0
		TC$D	TC$D	TC$D	TC$D	TC$D	TC$ D
		a	6eW			:	e

Figure i Percentage distribution of _/E/ (bed, bet) variants _{by following} environment in outer- city Belfast (Andersonstown, the _Braniel) and _{Lurgan. (After} Harris, _{1983: 157.)}

35'

JAMES MILROY

_AND

_LESLEY

_MILROY

environment tends to favour short _{realizations); D indicates following} fricative or voiced consonant _{(excluding /r/). Notice that the lowest short} realization, [a],is _{not favoured,} but that in _Lurgan short and _{low realizations} in short environments _{(T, C$)} are more favoured than elsewhere _{(see also} below), and that _long realizations _[ae:,?'] in these short environments _{are rarer} in Lurgan. The inner-city figures _{(Ballymacarrett, Clonard, Hammer) in} Table 3 clearly show some contrasts with the _{outer-city figures. Before}

Men 40-55

Women40-55

T	B	100	68
	C	97	8I
	H	97	75
C$	B	73	56
	C	8i	67
	H	76	68

Men 18-25 WomenI8-25

IOO	56
84	73
98	67
78	50
75	6o
76	52

Table _{3 in typically 'short' phonetic contexts in}

three inner-cityBelfastcommunities, _{Ballymacarrett(B), the Clonard (C) and}

Percentage low realizations of _/?/

the Hammer _(H)

voiceless stops, a low short realization ([a], [2]) is categorical for _{many male} speakers, while _{the women more often prefer higher and often lengthened} realizations. Thus, for typically low vowel environments, as in _{wet, went} females often have [w?:t,w*-:nt]for _{'vernacular' [wat, want]. In this respect} the inner-city female pattern is similar to that found generallyin these _higher status outer-city communities. These variabledata give us a basis for _{examining processes of change, since} they suggest initially that either the _{higher or lower variants are innovatory,} or - more properly - that the direction _{of change is either raising or lowering}

of /?/.

In fact, an examination of historical documentation _{(real-time evidence)} suggests that the direction of _{change is towards raising. Moreover, it appears} that mid realizations are gradually appearing in environments (such _as pre-voiceless _{stop) where low realizations were once the norm. It also appears} that as the low variants are _{replaced by higher ones, the relevant vowels are} lengthened and sometimes diphthongized: thus, _{as the rules are applied,} conservative variants such as _{[rant, rent]: 'rent', are replaced by [rE} . nt] (raising and lengthening) and [re.ant] (diphthongization). (For a _discussion see J. Milroy, 1976). The options open to speakers for the realization of _/c/

352

LINGUISTIC

CHANGE

beforeVoicelessStoporbeforeConsonant+ VoicelessStopmaybedescribed as follows:

(i)	Chooseeithermid or low;
(2)	If low, realizeas short;

(3) If mid,realizeas _{long; (4)} If mid-long,realizeas _{monophthongor diphthong.}

Thisis of coursean idealizedand simplifiedaccount,and the aim of _listing suchoptionsis descriptiveonly.Wedo not claimthatwe knowthe _ordering of rules,and if we do wishto orderthem,it is _possiblethat LENGTHshould precedeHEIGHTorthatlengtheningandraisingaresimultaneous.Nor _{is there} any implied claim that all individualspeakershave the same rules _or

rule-order- far from it. For the great complexitiesthat do exist when

speaker-variation is _studied, see now _Milroy et al. _(I983),

Harris _(I983).

Acceptingthis as a broaddescriptionof the currentstate,we now _examine somereal-timedatain orderto confirmthe directionof _change. Pattersongivesa listof fivewordsof the /s/ class,whichwerethen(i860)

pronounced in Belfast with low realizations: wren,wrestle,wretch,grenadier,

desk.Thesefew examplesare enoughto show that the low realizationwas thenmorewidespreadthantoday:wrenanddeskdo _{not satisfytheVoiceless} Stopor Sonorant+Voiceless _{Stopconditionin monosyllables,and arenow} categorical[e:]or [j ^- a]environments.Eventhe _{disyllablewrestleis unlikely} to appearwith [a], as the rule for raisingand lengtheningbefore[-s] _now almostalwaysoverridesthetendencyto _{lowerandshortenin disyllablesand} polysyllables.4Items _{like wretchand grenadierare now variable.Staples}

(I898) and Williams(I903),

the vowelin the city, whichallowus _{to inferthat low variantshad a much} widerdistributionthen _{than they do today. The completelist, takenfrom} thoseearlywritersallowsusto seethatthelowvowelappearedinenvironments whereit wouldnot appearnow- forexample,beforevoicelessfricativesand voicedstops(HarrisI983:I60). ThedistributioninpresentdayBelfastisquite different,as is shownby Table3 andFigureI. Inconservativeworkingclass speech,low variantsaremaintainedin 'short' environments,verymuchas in the nineteenthcentury:but _{low realizationshave been almost entirely} replacedin long environmentsby mid realizationsof _/c/. Moreprestigious and less conservativespeakersareless _{likelyto use 'low' realizations,even} in shortenvironments.

additionallygive quitedetaileddescriptionsof

It is evidentthat overthe last hundredyearsor so mid realizationshave

beenspreadingat the expenseof low realizations.Mid _/?/

totally replacedlow /E/ in 'long' contexts (pre-voicelessstop, _pre-sono- rant+voiceless stop, and in polysyllables).Low statusinner-cityspeakers

_{has now almost}

[41Itemslike wren,wretch,wrestlewerehistoricallyaffectedby loweringafter _/w/, andstill appearin manyIrishandAmericanvarietieswitha low vowel.

353

JAMES MILROY

AND

LESLEY

MILROY

(males)sometimesstillhavecategoricallylow realizationsin shortenviron- ments,butinthemoreprogressiveouter-cityhousingestates,thevowelisnow categoricallymidforsomespeakers.Interestingly,thedistributionof _variants in Lurganis moresimilarto thatof theinner-cityareasthanthatof _theouter

areas(a patternthat appliesalso to othervowel and consonantvariables). Thisrelativelyrapidlinguisticchangein _{Belfasthas accompaniedits risein}

populationfromabout _120,000

yearsof thiscentury,and _{Belfastmay be takenas an exemplarof linguistic} changein fast-growingcommunities(whileruraltownsandvillagesadhereto older patterns).The characteristicNETWORK _{structuresof these different} types of communityare also relevantto the mannerin whichchangemay come about, in so far as urban growth tends at first to _{weaken strong} pre-existingruralnetworks. Wemaysupplementourobservationson _{/?/ byconsideringevidencefrom} present-dayUlsterdialects.Thesearedividedinto _{two distincttypes.Ulster} Scots dialectsare found in East Ulster in a belt extendingfrom _around ColeraineintheNorth,throughmostof _{CountyAntrimandmuchof County} Down (whichis south of Belfast- see map).Most _{of Ulsterto the west of} thisbeltis English-basedor mixedScots-English.Present-dayBelfastdialect is often describedas an intrusionof this Mid-Ulstertype _{into the Scottish} easternbelt.Now, thelongmidvariantsof /?/ areoverwhelminglyassociated withpresent-dayUlsterScotsdialects(Gregg,I972) _{andarecharacteristicof} moderncentralScots dialectsgenerally(an _{exceptionis very conservative}

Galloway Scots, on which _see J. _{Milroy, I982b). Traditional Mid-Ulster} English, on the other hand, is characterized by lower realizations in all environments (Harris, I983: _I8I). The pattern of distributionin _{these dialects} is remarkably similar to that of nineteenth-century Belfast _{vernacular as} described in Patterson, Staples and Williams. We _may infer _{that this pattern} is a residue of some earlier English vowel pattern that has not been well identified or described _by historical _{linguists. There is sixteenth-century} orthographic evidence _{(discussed by} J. _{Milroy, I984b)} that _{suggests some}

distribution of low vowel realizationsfor _/?/ in _{London English of the period:}

it seems possible that this pattern of lowering of historic _{short vowels has been}

overtakenin recent StandardEnglish and _{CentralScots by a patternof raising} and (in the latter case) _lengthening. The Mid-Ulster dialects _{may therefore} have preservedto a great extent an older _{general English vowel pattern, and}

they may help us to project knowledge of the _present on _{to the past.} The historical and geographical evidence _{then both suggest that the low}

realizations of /E/ (conservative English in _{background) are giving way in}

a linguistically ordered way to the long mid realizations characteristic _of present-dayScots. It is clearthat this _change carriesprestigein _{Belfastin terms} of social class hierarchy and _status, as it is the _{more prestigious groups that} tend to adopt it and the more 'advanced' (generally female and _younger) group who introduce it to the conservative inner-citycommunities _{(which are}

in _{I 860 to nearlyhalf a millionin the early}

354

LINGUISTIC	CHANGE
	Rathlin
cO	Island

<

~~~~~~~~~~~~~~~North

|	YB~~~~~~allymena	Lrn
	Lurgan*	<	N	;

_{Map representing'core' Ulster} Scots _areasof north-east Ulster _{(shaded areas): adapted} from Gregg (1972).

_{characterizedby dense} and _multiplex network ties that tend to resist innovationandmaintainconservativeforms).Thetensionbetweeninnovative and conservativesocial _{mechanismsgives} rise to a identifiablepatternof gradualdiffusion,whichmayberepresentedasa historicalshiftfromanolder English-typepatterntowardsa _{patterncharacteristicof} modernScots.As we haveimplied,themannerinwhichthechangeproceedsisconditionedbyboth socialand phonologicalfactors.We now turnto a _{descriptionof change} in _/a/, withwhichthe _{/e/ systemcan be compared.}

(ii) Backing of _/a/

As we have indicatedabove _{(p. 349), the range of realizationsof /a/}

_in

present-dayBelfastvernacularis considerable - from[e]through[a]and _[a]

355

JAMES MILROY

AND

LESLEY

MILROY

to back raised and rounded _{[3]. Again,} as for _{/?/, patterns} of _{lengthening and} diphthongization are _present, with _long vowels _being associated _{mainly with} back realizations _and with _{the higher front realizations before voiced velars}

(see Table I).

In what _follows, we are concerned _{only with backing and}

retraction, and _{we therefore largely exclude the pre-velar environments (in} which backing is not found). Table I also shows that back realizations are _{favoured by following} fricatives, non-velar voiced _{stops and non-velar nasals (on the rules for} preceding velars, as in cab, _carrot, see p. _{346 above).} Nasals favour _backing particularlystrongly. Middle-class urban _{speakers (J. Milroy, i982 a) tend to} narrow the extreme _range described above _and in _{some cases converge on a} point somewhere in _{the middle} of _{the range, around [a] (but see below). The} widest range is found mainly in the speech of inner-city male _speakers. Furthermore, it is the MALES of _{Ballymacarrett (East Belfast) who use the} backed variants most and who show evidence of _{spreading the backed} realizations into VOICELESS stop environments _(as in _{that, wrap),where short,} front variants are _{expected. If there is evidence of change in progress towards} backed variants of _/a/, it will therefore be _{male speakers who are leading it,} rather than the females who lead _{the change towards raised /?/.} Historical documentation _{suggests that /a/ backing is a recent trend. The} elocutionist _{Patterson (i86o) does not comment on /a/ backing at all. On} the contrary, his _{remarks suggest that the Ulster tendency was towards}

fronting and raising and that the most salient Belfast feature was fronting and raising in velar environments.

In some places [presumably in _{the north of Ireland: JM, LM] the short}

sound of e is improperly substituted for _{a, in almost every word in which}

it occurs; in Belfast, however, this error is almost exclusively _{confined to} those words in which a is preceded by c or g, or followed by _{the sound}

of k, hard g or _ng.

(Patterson, _{i86o: I5)}

A very few of Patterson's _{spellings may indicate that /a/ backing and}

rounding had been observed sporadically in -r and -l environments: he _has form for 'farm' and canaul for 'canal'. However, examples of _{this kind are} so few that they indicate only a slight tendency (possibly _{confined to some} pre-sonorant environments), _{which is not enough for /a/ backing to be} discussed as a stereotype. The item carappearsin _{Patterson as 'care', in which} the now highly recessiverulefor fronting and raisingaftervelarsis _clear.Items like hand,band,in which [o]is now stereotypically _{expected, are given simply} as han, ban, etc. Frequently, _{however, items that now have low and/or back} vowels, are given with [E]: these include rether for 'rather' (a _{rural Scots} residue), e for a in single nasal environments in polysyllables such _{as exemine,} Jenuary and in nasal cluster environments _{such as demsel, exemple, Entrim} ('Antrim'), _{slent, bendy '(bandy'), brench.}

356

LINGUISTIC

CHANGE

WhereasPatterson'saccountindicatesa _{systemgenerallyinclinedtowards}

front-vowelrealizations,Staples(I898),writingnearly40 _{yearslater,reports}

a 'low backwide'vowelbeforenon-velarnasals,ine.g.man,hand,land.Since

Patterson's time -/r/ environmentshave become _{categorically} back realizations.Otherwise,the figureson present-dayvariationconfirmthat sincethen it is nasal environmentsthat have _{subsequentlyled} the _change,

_{closelyfollowedby} fricativeand voicedstop environments.In East _Belfast,

_{as we have} noted

environments,and this is most clearlyattestedin youngmen _(thosein our samplewereaged _I8-20).

THUS,ALTHOUGHRAISING AND LENGTHENINGOF _f/? AND BACKING OF/a/ ARE BOTH CHANGESASSOCIATEDWITH MODERN CENTRAL _SCOTS, THE FORMER _IS AT PRESENTLED IN BELFAST BY FEMALESAND THE LATTER BY _MALES. It _is

clear from patternsof stylisticvariationthat (as we might _alreadyhave inferred)thetwo changeshavedifferentprestigevaluesattachedto them.As Table4 indicates,thebackingof /a/ tendsto beresistedbyspeakersincareful

_{above, backing} is _{spreadingeven} into voiceless _stop

EastBelfast(Ballymacarrett)

	Men (40-55)	Women (40-55)	Men (I8-25)	Women(I8-25)
IS	3.03	I.75	2.89	I.89
SS	3.58	2.58	3.43	2.IO
		WestBelfast(Clonard)
	Men (40-55)	Women (4-55)	Men (I8-25)	Women(I8-25)
IS	2.79	1.77	2.36	2.36
SS	2.79	I.85	2.33	2.6I

Table 4

Incidenceof retractionand _{backingof /a/ by age, sex and conversational stylein} two Belfastcommunities,calculatedby an _{indexscorerangingfrom}

o (minimum)to 4 (maximum).IS, interviewstyle;SS, _{spontaneousstyle}

'interview'style(whereasraisingof _{/F/ is MORElikelyincarefulstyles).Thus,} men seem to be principallyassociatedwith a changethat _{speakersdo not} consciouslyviewas beingof highprestige,whilewomenareassociatedwith one adopted by speakersin their more carefullymonitoredstyles (for furtherdiscussionof _{thesefigures,see Section6 below).} Ourrealtimeevidenceconfirmsthatthe _{movementin /a/ is phonetically} fromfrontto back.Thismeansthatsporadicfront-raising(foundmainlyin

West Belfast) in words likeflat, trap ([flet, _trep]) must be seen as _{residues and}

not as innovations.The beliefof manycasualobserversthat _{raisingbefore} velars(andveryoccasionallybeforevoicelessnon-velarstops)are _attempts

357

JAMES MILROY

AND

LESLEY

MILROY

to imitateconservativeRP ('ReceivedPronunciation',as _{described,e.g. by}

Gimson, I980)

is also shown to be wrongby quantitativeand _diachronic

evidence.Thequantitativeevidenceshowsthatthe _{frontingandraisingrule} in Belfastvernacularis _{virtuallyconfinedto} velarenvironmentsandcannot applyto wordslike bad,hand,stab(whicharefrontin RP). The diachronic evidenceshows _that, for _{a centuryor more, the} trendhas _{been towards} retractionand _backing. Theevidencealsoindicatesthattheruleforbackingdiffusesgeographically fromEast to WestBelfast(see Table4). Scoresfor _{/a/ backingarehigher} for East Belfastmalesthan for _{any othergroupsstudied,and the rangeof} environmentsin which _{backing operatesis extended to voiceless stops} amongstyoungerEast Belfastmales. It _{appearsto be inner East Belfast} (Ballymacarrett)thatprovidesthemodelforworking-classspeechin _thecity (L. Milroy, I980); this is discussedby Harris(I983) in _{termsof a 'labour} aristocracy' representedby the (relatively)fully employed protestant populationof EastBelfast. Both /a/ backingand /?/ raisingare relativelyrecent _phenomenain Belfast(butseebelow),andbothareassociatedwitha _{backgroundin Scots.} Patterson'saccount _{of Belfastshows characteristicsof conservativerural} Scotslexicaldistribution,muchof _{whichappearsto havebeenresidualand} is now obliteratedby restructuring.Howeverlengtheningandraisingof _/e/ andbackingof /a/ aremodernScots.Gregg's(I972) _{accountof UlsterScots} givesoverwhelminglybackrealizationsof _{/a/ anddescribes/c/ as oftenlong} in realization(contrasttheveryshortlow _{realizationsin conservativeBelfast} vernacular,suchas _{[stap, t3at] for step,jet). Similarly,/a/ backingseemsto} be a verygeneralmodernScotsfeature(Lass,1976).EastBelfastadjoinsthe Ulster-Scotsregionof _{NorthDown (wherebackingis strong),whereasWest} Belfastpoints south-westdown the _{LaganValley, the speechof whichis} Mid-Ulsterwith less Scots influence;furthermore,immigrationto _West Belfastis recent and is largelyfrom a Mid- and West-Ulsternon-Scots hinterland.Presentdayquantitativestudiesin _{Lurgan,a smallcountrytown} south-westof Belfastin the LaganValley,confirmthe _{existenceof an /a/} systemwith little backing(frontvowelshave _{been noted in that areaeven} before[r]andfinally),whichis _{quitesimilarto Patterson'si86o accountof} Belfastin this respect(Pitts, _I982). Finally,wemustnotethatif we takea generaloverview,thesetwo _vowels appearto be movingawayfromone anotherin phoneticspace,ratherthan in the samedirection(as we wouldexpect,e.g. _{in a chain-shift).We arenot} in thispaperprimarilyconcernedwiththeembeddingof _{changesin language} systems(and argumentsbasedon _{this could suggestthat one change- /-/} raising - is slightly more recent in origin than the other), but we _may commentthat if data for individualspeakersand _{homogeneousgroupsare} examined,theoverallpictureof vowelsmovingawayfromoneanotherdoes not appearso prominentlyas it does whenwe focuson the languagerather

358

LINGUISTIC

CHANGE

than the speaker.Speakerswho have [6] raisingtend to havemorefronted realizationsof /a/ (thesespeakersareoftenfemale),andthosewhohavelower

reali"ationsof /?/

male). Thus, an account based on what speakersactuallydo looks very differentfroma generalizedaccountof changein the languagesystem.

aremorelikelyto have[a]backing(thesespeakersareoften

In the next section,we move froman accountof the languagesystemto

a discussionof speakersand theirsocialrolein phonologicalinnovations.

_4. SOCIAL NETWORK STRUCTURE AND SPEAKER INNOVATION: _AN ANALYSIS OF /a/ AND /6/

_{In additionto the} variablesof _age, sex and statusdiscussedin Section3, _a furthersocialvariableassociatedwitha _{speaker'sDEGREEOFINTEGRATIONinto} his closeknitcommunityappearedto affect the probabilityof his being linguisticallyinnovativewithrespectto choiceof vowelvariants. _{Generallyspeaking,it} seemsto be truethattheclosertheindividual'sties to a local communitynetwork,the more likely he is to approximateto vernacularnorms(see L. Milroy, I980, for details).Followingsome well- developed anthropologicalfindings,we have suggestedthat a closeknit _network has an intrinsiccapacity to function as a norm-enforcement mechanism,to the extent that it operatesin opposition to larger scale institutionalstandardisingpressures.One corollaryof this _capacityof a closeknitnetworkto maintainlinguisticnormsof a non-standardkindis that the LOOSENINGof sucha networkstructurewill be associatedwithlinguistic

change (L. Milroy, I980:

I85; Gal, 1979).

_corollaryon whichwe concentratehere.

It is the implications of that

A majorpoint emergingfrom our earlieranalysisof language/network

relationshipswas that the variableNETWORK neededto be consideredin

relationto

(I982: 71), the networkvariableis in generalcloselyassociatedwithmany others,includinggenerationcohort,geographicallocation,andsocialstatus. Thus,ournexttaskhereis to pickoutbrieflytherelevantpartsof ouranalysis

of the social distributionof innovatoryrealizationsof /a/ and _/?/, as identifiedin _Section3.

arestronglyaffectedby thevariable

SEX OF SPEAKER.Thus,althoughincomingvariantsof _{bothvowelsappearto}

have originatedin the same hinterlandScots dialect,each has assumeda diametricallyopposedSOCIALvaluein its new urbansetting.

are, in the low status inner city, associated

particularlywith women and with careful speech styles. They are also associatedgenerallywith slightlymore prestigiousOUTER city speech,and datacollectedby surveymethodsconfirmsthat the higherthe statusof the speaker,themorelikelyhe is to use raisedvariants(seeMilroyet al., I983). Differentlevelsof use accordingto SEX OF SPEAKERareparticularlyevident

the variableSEX OF _{SPEAKER.Indeed,as Gumperzhas remarked}

Firstof all,realizationsof /a/ and/?/

Raised variants of /6/

359

JAMES MILROY

AND

LESLEY

MILROY

in Ballymacarrett,where it appears to be YOUNGER FEMALE speakers who are most strongly associated with the incoming raised variants. The incoming variants of /a/ show an almost perfectly converse pattern of social distribution. _High levels of _backing are associated with males (particularly Ballymacarrettmales, although levels in other inner city areas are still _{quite high)} and with casual _{styles appropriate} to interaction between _peers. The most _extremely backed variants do not _appear at all in outer _city

speech. Interestingly,the sex differentiationpattern across the threeinner city

areas is not as consistent for /a/

that the young Clonard WOMENareincreasingtheiruse of backed realizations when compared with other female groups (see Table 4). They also use these variants MORE than theirmale counterpartsalthough they follow the expected

sex differentiation patterns with respect to other phonological variables (see Section 6 below for a discussion of the Clonard pattern).

as it is for _/?/;

there is some _indication

In summary then, it appears that incoming variants of /a/	are associated
with core Belfast vernacular, while incoming variants of /?/	are associated

with careful higher status speech. If we look at the relationship between speaker choice of variant and

individual network structure, the picture becomes even more complicated. _{With respect} to _{both vowels,} choice of variant shows a correlation with _personal network _structurein some subsections of the innercity _communities;

but the details of

The vowel /a/ is particularly sensitive to variation _according to the network structureof the speaker; but WOMEN appear to correlate theirchoice of variant more _closely with their _personal network structure than do men. This means that among women a relatively large amount of /a/ backing is more _likely to be associated with a _high level of _integration into the network than is the case among men - a relationship analysed by Spearman's Rank Order Correlation (L. Milroy, I980: _I55). Although, as we _{have noted,} women are much less _likely than men to select back variants of _/a/, this

generallylower level of use does not prevent individual women from varying their realization of /a/, within the female norms, according to their social network structure.Thus, the DEGREEOF FIT between phonological choice and network structuremay be seen as an issue quite separate from the _ABSOLUTE LEVELOF USE of a particular range of variants. We may thus argue that /a/

functionsfor womenas a NETWORKMARKERto a greaterextentthanit does

for _{men; by} this we mean that there is for them a _higher correlation between choice of variant and network structure,a tendency to select relativelybacked variants _being associated with _{higher levels} of _{integration into the} community. When we look at the relationship between choice of /?/ realization and individual social network structure, we find a pattern emerging converse to the one described for _/a/; recall also that the _incoming variants _{of the two}

this correlation are quite different for each vowel.

36o

LINGUISTIC

CHANGE

vowelsshowedan almostconversesocialdistributionwithregardto _status, sex of speaker,and speechstyles. Mostimportantly,thereappearsto _{be no tendencyat allforwomento use}

/s/

significantcorrelationbetweennetworkscoresof _{MALEspeakers(particularly}

young male speakers)and choice of _/f/

relativelyLOW (conservative)variantsis associatedwitha relativelyhighlevel

of integrationinto the community(see L Milroy, I980:

Thiscomplexrelationshipbetweennetworkstructure,sex _{of speakerand} languageuse is summarisedin Table _5. However,our interesthereis in a

as a NETWORK MARKER in the sense describedabove; but there is a

realization.A tendencyto select

155 for details).

/a/

_/E/

Changeled by

Males

Females

Highcorrelationwith

networkstrength

Table ₅

Females

_Males

Contrastingpatternsof distributionof two vowels involved in _change, accordingto sex of speaker,relativefrequencyof _{innovatoryvariantsand} levelof correlationwithnetworkstrength

_{generalizationwhichwe} arenow ableto makeconcerningon the one hand the relationshipbetweenlanguageand networkstructure,and on the other the socialidentityof _{theinnovatinggroup.IN} THE CASEOF BOTH _/E/ AND _/a/

IT _IS THEPERSONSFORWHOMTHEVOWELHAS _{LESSSIGNIFICANCEAS A NETWORK} MARKERWHO SEEMTO BE LEADING THE LINGUISTICCHANGE. It is as if absence

of this language/networkrelationship(a _{relationshipthat} fulfilsa cohesive

socialfunction)enablesa particularsocialgroupto _{adopttheroleof linguistic} innovators.Thisappearsto be thecaseregardlessof _{whethertheinnovation} is evaluatedby thewiderurbancommunityas _{beingof highor of low status.}

For althoughit is clearthat/e/ raisingis diffusingon a muchbroadersocial frontthan /a/ backing,the generalizationstill seemsto hold truethat _it is thosepersonsin the innercity for whomthe vowelfunctionslessclearlyas

a network marker who are the principal innovators into their own communities. Itis importantto notethateventhoughbackedvariantsof _{/a/ arestrongly} emblematicof vernacularspeech,they are neverthelessspreadingto _higher statusgroupsin thewidercommunity.Butthisdiffusionisbeingimplemented

in a manner very different from that affecting _/?/.

_{We have noted that [?]}

raisingis characteristicboth of low-statusfemalespeechandmoregenerally

13

_36I

LIN

21

JAMES MILROY

AND

LESLEY

_MILROY

_{of higherstatusspeech.The} diffusionof _{[E] raisingon} thiswidesocialfront is confirmedboth _{by linguisticsurveydata} and _by moredetailedouter-city communitystudies. Whenwe look at the socialdistributionof variantsof _{/a/ (on} whichsee J.Milroy,i982 a)wefindbywayofcontrastthathigherstatusBelfastspeakers avoid both extremefront _{AND extremeback realizations,as they converge} aroundCardinalVowel4 in the middleof _{the phoneticrange.However,a} veryinterestinggroupof young,male,middle-classspeakerscanbeidentified in the sample of speakersstudied in the _{survey. They} also show _the characteristicmiddle-classtendencyto _{convergearounda} limitedphonetic area,withrelativelylittleconditionedvariation.However,phonetically,the pointat whichtheyconvergeis furtherbackthanthatcharacteristicof _older middle-classspeakers. It appearsthereforethatthe mechanismof diffusionassociatedwitheach of t;hevowelsis different.Raisedvariantsof _/?/ areapparentlyspreadingin alinguisticallyorderedway,with'long'environmentsaffectedfirst.Formany outer-cityandmiddle-classspeakers,a _{raisedvowelis alreadycategoricalin} all environments.Althoughbackedvariantsof _{/a/ appearto be diffusing} historicallyandlaterally(throughthelow _{statusinner-citycommunities)in a} linguisticallyorderedmannerparallelto _{the processesaffecting/E/, the} mechanismof diffusionupwards(socially)throughthe _{communityis quite} different.Whatseemsto be involvedhereis a 'drift'phoneticallyto _theback of the characteristicmiddle-classrealization. Thedatapresentedheresuggestthatsocialnetworkstructureis _implicated

in processesof linguisticchangein at leasttwoways.First,a _{strongcloseknit} networkmaybe seento functionas a conservativeforce,resistingpressures to changefromoutsidethe network.Thosespeakerswhoseties _areweakest arethosewho approximateleastcloselyto _{vernacularnorms,and aremost} exposedto pressuresfor changeoriginatingfromoutsidethe _network. Second,a detailedsociolinguisticanalysisof _{[?] raisingand [a]backing} - processeswhichhavea commondialectalpoint _{of originbut havetakenon} very differentsocial _{values in theirnew urbancontext} - _{suggeststhat the} VERNACULAR speakersassociatedmost _{stronglywith the innovationare in} eachcasethoseforwhomthevowelfunctionsleastprominentlyas _{a network} marker.It is as if a strongrelationshipbetweenthe _{networkstructureof a} given group and choice of phonetic realizationof a particularvowel disqualifiesthat _{group from fulfillingthe role of innovatorswith respect}

to that vowel. Conversely,it may

language/networkrelationshipwith _{respectto a group of speakersis a} necessaryconditionfor thatgroupto fulfilthe role of linguisticinnovators. Both of these observationssuggest _{that since the variable NETWORK} STRUCTURE is implicatedin a negativeway in linguisticchange, _{a closer} examinationof WEAK networktieswouldbe _{profitable.For it mightwellbe}

_{be the case}

_{that dissolution of the}

362

LINGUISTIC

_CHANGE

that it is speakerswho lack _{strongnetworkties} or are _loosely attachedto closeknitgroupswho arecharacteristicallylinguisticinnovators. The problemis that a generalweaknessof socialnetworkanalysisis _its superiorabilityto handleCLOSEKNITtiesas _{opposedto weak,diffusetypesof} networkstructure.Thisdifficultyarisesfromthefactthatpersonalnetworks arein principleunbounded;the numberand _strengthof ties whichbindan individualto others are not, in the last analysis,definable.Howeverin closeknitterritoriallydefinedgroupsit _{is possibleto treatpersonalnetworks} AS IF theywereboundedgroups(seeMilroy,I980: _{Ch. 3)whereasin socially} andgeographicallymobilesectorsof societythisisnotfeasible.Ourownwork hasreflectedthisin thatit hasconcentratedon thefunctionof _{closeknitties,} observedwithina definedterritory,as _{an importantmechanismof language} MAINTENANCE. Yet, it is evident that a _{very large numberof speakers,} particularlyin cities,do not havepersonalsocialnetworksof _{this type.We} havesuggestedthat,in Britishsocietyat least,closeknitnetworksarelocated primarilyat thehighestandtheloweststrata,witha _{majorityof sociallyand} geographicallymobile speakersfallingbetweenthese two _{points. (But see} Kroch,M.S., for an interestingstudyof a closeknitupper-classAMERICAN network).Significantly,Labov _{and Kroch have noted that in the United} Stateslinguisticchangeseemsalwaysto _{originateanddiffusefromsomepoint} in this centralareaof the social hierarchy- neverfrom _{the highestor the}

lowest social groups (Labov, _{I980; Kroch, I978).}

Thus,despitethedifficultiesof studyinglooseknitnetworktiesin _theouter cityusingthemethodsadoptedintheinner-cityareas,a _{searchforsomeother} meansof followingthroughtheirevidentassociationwithlinguisticchange seemedwell worthwhile.5Thiscannotbe _{accomplishedby analysingstatis-} ticallyrelationshipsbetweenlanguageand _{network,as was possiblein the} innercity communities,simplybecausethereis _{no obviousway of charac-} terisingquantitativelylooseknituniplextieswhichextendovervastdistances and are often contractedwith _{large numbersof others. Indeed,such an} undertakingmightbe neitherpossiblenordesirable,giventheverydifferent role fulfilledby the closeknitgroupsat eitherend of the socialhierarchyin maintainingpolarisedsets _{of linguisticnorms.It is certainlynot clearthat} quantitativeexaminationof thelooseknitnetworkscontractedby _{a majority} of speakersin thecentreof thathierarchywouldbeparticularlyilluminating.

_[5] The initialdifficultieswerethe _{practicalones whichmightbe predicted.We foundthat} networktiesof outer-cityindividualsinthekeynetworksectorsof kin,friendshipandwork oftenstraggledoverextensiveareas.Conversely,tiesof _{neighbourhood,whichwerecrucial} in the innercity, oftenseemednot to be significant,sincepeoplefrequentlyhardlyknew theirneighbours.Thus,in theinnercity,wheretiesweredenseandterritoriallybounded, it seemedreasonableto studycommunitylinguisticnormsusinga _{networkmodelwhich} wasitselfpartof a theoryof languagemaintenance.Butit _{wasnot at all clearwhatkind} of hypothesiswe mightderivefroma comparablestudyin _{the outer-cityareas} - _{or even} whatmightconstitutea _{comparablestudy.}

363

I3-2

JAMES MILROY

AND

LESLEY

MILROY

We thereforeproceed to examine the relationshipbetween looseknit networktiesandlinguisticchangein _quitea differentway.First,a theoretical modelof the socialfunctionof 'weak' networktiesis presented;second,we look at the socialcharacteristicsof innovatorsin _general,and on this basis suggesta newmodelof linguisticinnovationand diffusion.

5.

The discussionin this section dependsheavilyon a _{suggestivepaper by} Granovetter(1973), who sees 'weak' ties betweenindividualsas _important linksbetweenmicro-groups(small,closeknitnetworks)andthewidersociety. Perhapsit is best at this _point to _{graspthe nettle,and attempta definition} of whatis meantby 'weak' and _{'strong'ties,for} thiscontrastcannoteasily be characterizedquantitatively.Granovettersuggeststhe _following: 'the strengthof a tie is a (probablylinear)combinationof the amountof _time, the emotionalintensity,the intimacy(mutualconfiding)and _{the reciprocal}

services which characterisea tie' _{(I 36I). Note that by this measure multiplex}

ties- i.e.thosewithmultiplecontent- wouldbecountedasrelativelystrong; the notion of multiplexitywas an importantbasis of the networkstrength measuresusedin the Belfastinner-citystudies. Granovetter'sdefinitionis _{probablysufficientto satisfy most readers'} intuitivesenseof whatmightbe meantby a 'strong'or 'weak'interpersonal tie, correspondingas it (approximately)does to an everydaydistinction betweenan 'acquaintance'anda 'friend'.It is certainlysatisfactoryfor our purposehere. Granovetterremarksthatmostnetworkmodelsdealimplicitlywithsmall, well-definedgroupsWITHIN whichmanystrongtiesarecontracted(cf.p. ₃₆₃ above).Hisfundamentalargumentis thatweaktiesBETWEENgroupsprovide bridgesthroughwhichinformationandinfluencearediffused,andthatweak ties aremorelikelyto linkmembersof _{DIFFERENTsmallgroupsthan strong} ones, whichtendto be concentratedWITHIN particulargroups.Thus,while strongties give rise to local cohesion,they lead, paradoxically,to overall fragmentation. _Only weak ties can form _{a bridge between cohesive groups, for the} followingstructuralreason(whichGranovetterexpressesasa _{hypothesisand} initiallysupportsby aprioristicargumentratherthanby _{adducingempirical} evidence):

WEAK

TIES AND

INNOVATIONS

If we considertwo arbitrarilyselectedindividuals,A _{and B andthe set S,}

consisting of C, D, E

of all persons who have ties with _{either or both of}

them,the strongertherelationshipbetweenA _{andB, themorethenetworks} of each arelikelyto overlap.Extensiveoverlap,whichwill _{inhibitthe flow} of NEW informationbetweenA _{andB, is predictedto be leastwhentheA-B} tie is absentand to increasein proportionto its strength.Thisrelationship betweennetworkoverlapandstrengthof _{tieresultslargelyfromthetendency}

364

LINGUISTIC

CHANGE

for strongties to involvemore timecommitment;for exampleif A spends

a largeproportionof his timewith B, it is likelythat this time investment will bringhimultimatelyinto contactwiththe individualsC, D and E who _{initiallyformedpartof B's} network.Conversely,the networksof A and B _{arelesslikelyto overlapif} thetiebetweenthemis _weak,andso wemayderive the moregeneral(andfor our purposemoreimportant)principlethatlinks BETWEENcloseknitgroupsare normallyWEAKties betweenthe individuals whohavecontractedthem.Theseweaktiesbetweennon-overlappinggroups provideimportantbridgesfor the diffusionof innovations. Examiningvariouslikelyand unlikelyconfigurationsof strongandweak ties,Granovetternotesthatindividualsvaryin theproportionof eachwhich they contract.Whilenot all weak ties functionas bridgesbetweengroups, all bridgesmust, Granovettersuggests,be weak ties. For the sake of the argument,a bridgeis definedas the ONLYroutethroughwhichinformation flowsfromA to B,orfromanycontactof A to anycontactof B(seeFigure2).

E

F

A--------BI

Figure 2 A bridge between two networks. ----,

G

Weak ties;

H

K

,strong ties.

Granovetter'sinterestis in exploringtheinterpersonalmechanismswhich connectsmallgroupsto each otherand to a largersociety,and his model predictsthatinnovationandinfluencewillflowthroughweaktiesratherthan strongones.Itis therelationshipbetweenstrengthof tieandnetworkoverlap whichleadshim to suggestthatNO STRONG TIE CAN BE A BRIDGE. And while it mustbe acknowledgedthatin practicethereis likelyto be morethanone link betweengroupsof any size, the principlethat theselinksare likelyto be weakis of greatimportancehere.Weakintergroupties,by Granovetter's argument,arelikelyto becriticalin transmittinginnovationsfromonegroup to another,despitethecommonsenseassumptionthatSTRONGtiesfulfillthis role (see for exampleDownes (I984: 155) who suggeststhat networksmay be importantin developinga theoryof linguisticdiffusion,butassumesthat it is strongties whichwill be critical). AlthoughGranovetter'sprinciplemightat firstseemcounter-intuitiveand paradoxical,a littlethoughtconfirmsthatit worksout wellempirically.First of all, it is likely(in the networksof mobileindividualsat least)that weak ties aremorenumerousthanstrongties. _{Second,it is clearthatmanymore} individualscan be reachedthroughweaktiesthanthroughstrong;consider

365

JAMES MILROY

AND

LESLEY

MILROY

forexamplethenumberof contactsmadeby a salesmanin _{thecourseof his} business,duringwhichhe _{buildsup an elaboratestructureof bridges.Con-}

sider also the elaborate bridges set up by participantsat academic conferences,whichlinkthecohesivegroupsassociatedwitheachinstitution. It is via these bridgesthat new ideas pass from one institutionalgroupto another.Conversely,informationrelayedthroughstrongtiestendsnot _{to be} innovatory;as Granovetterremarks,'If _{one tells a rumourto all his close} friendsand theydo likewise,manywill _{hearthe rumoura secondandthird} time, since those linkedby strong ties tend to sharefriends'(1366). _But it is evidentthatgenuinediffusionof _{therumourwilltakeplaceif eachperson} tellsit to acquaintanceswithwhomheis _{onlyweaklylinked;theyin turnwill} transmitit to a largenumberof non-overlappinggroups,so _{thatthe'retelling} effect'will _{not occur.} It has often beennoted (see, for example,Turner,I967) _{that a closeknit} networkstructurewillusuallynot _{survivea changeof location,andit is clear} in generalthatsocialor geographicalmobilityis _{conduciveto theformation} of weakties.Moreover,a mobileindividual'sweaktiesarelikelyto _bemuch morenumerousthanhis _{strongties.If a manchangeshisjob, he is not only} movingfromone networkof tiesto another,butestablishinga _linkbetween eachrelativelycohesivegroup.Thus,mobileindividualswhoarerichinweak ties,but (asa _{consequenceof} theirmobility)relativelymarginalto _anygiven cohesivegroupare,it is argued,in a particularlystrongpositionto diffuse innovation.Note thatthiscontentionisinlinewiththetraditionalassumption by historiansof languagethat the emergent,mobile merchantclass _were largelyresponsiblefor the appearanceof Northern(and _{other) dialectal} innovationsin EarlyModern(Standard)English(see, _{for example,Strang,}

1970:

2I4 f.; Ekwall, I956;

Baugh & Cable, I978:

194);

if it is correct,

Granovetter'sprinciplethatthe _{overlapof two individuals'socialnetworks} variesdirectlywiththe strengthof theirtie to one anotherhasconsiderable implicationsfor any _theory of diffusion.(Strengthof _{tie is of course a} continuousvariablealthough'weak'and'strong'tieshavebeentreatedhere as if theywerediscrete.)It mightappearthatthisrelativelyclearhypothesis could _easily be _supportedor disconfirmedempirically;but _{unfortunately} networkor sociometricstudiescannoteasilybe useddirectlyas a sourceof corroboratory(or disconfirmatory)evidencesimplybecausetheirresearch designusuallyentailsrelativeneglectof _{weakties.Thus,for example,when} personsareaskedto nameothersfromwhomtheyhavereceivedinformation (orfriendship,as in Labov's'lames'study)thenumberof _{permittedchoices} is usuallyrestrictedso that the namingof weakties is effectivelyinhibited. Evenif the researchdesignpermitsidentificationof _{personswith weakties} to specifiedothers, as did our own _{(see Milroy, I980, for details),it is} extremelydifficultto _study those ties _{just because they ARE weak and} perceivedas relativelyunimportantto EGO.

366

LINGUISTIC

CHANGE

Fortunately,empiricalevidence to _support Granovetter'smodel _has

emergedfromelsewhere-

thediffusionof aroundfifteenhundredinnovations.Somegeneralprinciples can be extractedfrom this largebody of empiricalevidencewhichtend to supportGranovetter'scontentionthat innovationsfirstreacha _{group via} weakties. A distinctiondiscussedinsomedetailbytheauthorsisbetweenINNOVATORS and EARLY ADOPTERS of an innovation.This distinctionturns _{out to be} importantwhethertheinnovationis _{agricultural(theintroductionof hybrid} seedcornto an Iowancommunity);technological(machineryto _engineering firms);educational(new methodsof mathematicsteaching);or _concerned with publichealth(introducingthe _{habitof boilingcontaminatedwaterto} Peruvianvillagers).Thereis evenanearlylinguisticstudyof _{theintroduction} of lexicalinnovationsto an oilfield(Boone, _I949). All _{of thesestudies,and} verymanymore,confirmthe principlethat INNOVATORSaremarginalto _the groupadoptingthe innovation,often beingperceivedas _{underconforming} to the point of deviance. The EARLY ADOPTERS of the innovationare, on _{the other hand, central} membersof thegroup,havingstrongtieswithinit, _{andarehighlyconforming} to groupnorms;they frequentlyprovidea modelfor othernon-innovative membersof thegroup.Afterits adoptionby thesecentralfigures(frommore marginalpersons),an innovationis typicallydisseminatedfrom _{the inside} outwardswithincreasingspeed,showingan _{S-curveof adopterdistribution} throughtime.Whileit is clearthatlinguisticinnovationsdifferin _{a number} of respectsfrom,for example,technicalinnovations(seeTrudgill,I983: _63, fora discussion),theydo notappearto be DIFFUSEDbymechanismsmarkedly differentfromthosewhichcontrolthediffusionof _{innovationsgenerally.For} linguisticinnovationsalso _{show this characteristicS-curveof distribution} throughtime(seeChambers& _{Trudgill,I980: I76-I8i; Bailey,I973).} Bearingin mindthe norm-enforcingcharacterof a _{groupbuiltup mainly} of strongties, and its consequentlack of susceptibilityto outsideinfluence, we can see why innovatorsare likelyto be personswho are _weaklylinked to the group. Susceptibilityto outsideinfluenceis _{likely to be greaterin} inverseproportionto _{strengthof tie withthe groupandby implicationalso} in inverseproportionto susceptibilityto norm-enforcingpressurefromthe group. Thus, where groups are linked by many weak ties they will be susceptibleto innovationpartlyfor this _{(social)reason,and partlybecause} innovationis for structuralreasonsunlikelyto _{be transmittedvia a strong} tie (seepp. 364-365above).

Personsat the centreof a norm-enforcinggroup(i.e. _{personswho share} strong ties within it) will, as a corollary,not be susceptibleto outside pressures.Becauseof _{the investmentin time and commitmentneededto} maintainthese strongties, they will almostcertainlylack _{opportunitiesto}

notablyRogers'andShoemaker's(I97I)

_studyof

367

JAMES MILROY

AND

LESLEY

MILROY

_{formmanybridges(weakties)with} othergroups.Thus,typically,for _these

variousreasons,innovators(as _{opposedto earlyadopters)will be persons marginalto} theircommunity,withmanyweakties to othergroups. It appearsat firstto be difficultto _{explainhow thesemarginalinnovators} coulddiffuseinnovationssuccessfullyto centralmembersof the _group;but two relatedpointscanhelpus here.First,in viewof theverygeneralfinding

of sociolinguisticresearchthat the _{prestigevaluesattachedto languageare}

oftenquitecovertanddifficultto _{tapdirectly,wemaysuggestthata successful}

innovationneedsto beevaluatedpositively,eitherovertlyorcovertly.Thisis

of coursea necessarybutnot a sufficientconditionforits _{ultimateadoption,}

and is bindingon non-linguisticinnovationsalso.

Second,wemaysurmisewithGranovetterthatsinceresistancetoinnovation

is likelyto be greatin a norm-conforminggroup,a _{largenumberof persons}

willhaveto be exposedto it and adoptit in the earlystagesfor it to _spread successfully. Now weak ties are, in a mobile _{society, likely} to be _{very much} more numerousthanstrongties,andsomeof themarelikelyto _{functionasbridges}

to the groupfromwhichthe innovationis _{flowing;thusan innovationlike}

the Cockney mergerbetween /v/:/6/ and /f/:/0/ reportedin _teenage

Norwichspeakersby Trudgill(I983: ₇₃₎ is _likelyto be transmittedthrough

a great many weak links contractedbetween _{Londonersand Norwich}

speakers.Quite simply,beforeit standsany chanceof _acceptanceby the centralmembersof _{a group,the} linksthroughwhichit _{is transmittedNEED}

to be numerous(cf. Granovetter,I973:

Returningto ourfirstpoint,we assumethatsomekindof _{prestige,either} over or covert,is associatedwith the _{innovation.In} otherwords,Norwich

speakers,whethertheyaremarginalor centralto theirlocalgroups,in some senseviewvernacularLondonspeechas _{desirable- moredesirablethanthe} speechof othercities.6Again,followingthroughthe _{argumentspresentedin} thissection,we suggestthatpersonscentralto _{thenetworkwouldfinddirect} innovationa riskybusiness;but adoptingan innovationwhichis _already widespreadon the edgesof the groupis muchless risky.Thus,insteadof askinghowcentralmembersof _{a groupareinducedto acceptan innovation} frommarginalmembers,wecanviewthisas _{a sensiblestrategyon theirpart.}

In orderto adoptan innovationwhichis seenas desirable,theydiminishthe

risk of a potentiallydeviantactivityby _{adoptingit frompersonswho are}

alreadylinkedto the group,ratherthanby directimportation.

_1367).

[6] The merger between dental and labio-dental fricatives has been _{noted in the speech of} Sheffieldadolescents _{also. By the reasoning which we are using here, we must assume first} that weak ties exist between Sheffield adolescents and London speakers and second (crucially) that London speech has some kind of prestige for Sheffieldspeakers. _Although we cannot at this stage enumeratethe factors which give rise to _{covert attitudes of this kind,} it seems reasonable to suggest that for reasons of (for example) local loyalty, _combined with perceptions of relative autonomy, not all cities will _{share them.}

368

LINGUISTIC

CHANGE

_{We arenow} in _{a positionto} relatethe substantivepointsemergingfrom this _discussionto earlierargumentsconcerningthe _{relationshipbetween} socialnetworkstructureandlinguisticchange.It is clearthatthe linknoted earlierbetweenthe _{dissolutionof} closeknitnetworksand the _{susceptibility}

of a groupto linguisticchangefits in with the observationsof Granovetter

and

stronglyassociatedwiththediffusionof theinnovativeraised/?/ andbacked _{/a/ variantsappearedto be} thoseverygroupswho tendedleastto use _these phonologicalelementsas 'networkmarkers'.It is _{likelythatthe} sociallocus

of theinnovationsis, at leastin _{part,a consequenceof} theuse a _givengroup is making(or failingto _{make)of themas} networkmarkers. Ifwereturnto Labov'sdiscussionoftheactuationproblem(seepp.342-343 above),it is clearthatthemodelelaboratedheredoesnot _{entirelyagreewith} his accountof the individualswho actuatelinguisticchange,i.e. _introduce an innovationto a definablegroup.Recallthattheyaredescribedas _persons

who havehighprestigeanda _{largenumberof}

the small local group. They do not sound at all similarto the _typical innovator,describedby Rogersand _{Shoemakeras underconformingto} the point of deviance. One seriousdifficultyappearsto be _that thereis _{apparentlyno easy way} forempiricalstudiesoflinguisticchangeinprogress(particularlyphonological change)to makethecrucialdistinctionbetweenINNOVATORS(whoaresocially marginal)and EARLY ADOPTERS (who occupy a central _position in the network).Wecan onlytrackan innovationthroughhistorical,geographical and socialspace,finallylinkingit _{witha specificgroup.It is not clearhow,} without being able to pinpoint the time of _{the first introductionof an} innovationto a community,we could identifythis group confidentlyas innovatorson theonehandorearlyadopterson _{theother.Butit is important} in principleto distinguishbetweenthe two _{groupsand it seemslikelythat} phonologicalinnovationwill alreadyhave begunto diffusethroughoutthe groupif it is sufficientlywell establishedto be _{observable.We shallshortly} discussthisquestionin relationto thegroupwhichappearsto _{be leadingthe} changeto /a/ backingin the Clonard,Belfast. Most probably,thepersonsdescribedby _{LabovareEARLYADOPTERS.But} thereis still a problemin thatit is not at all clearhow theirgroup-internal tiescouldbe strongwhentheyhavesimultaneouslya _{largenumberof such} ties (relative-toothers)and a high _{proportionof ALL theirties outsidethe} group.Onedifficultyin assessingLabov'sworkfromtheperspectivewehave

adoptedhereis thatheseemsto relyfundamentallyon theexplanatorypower of the notionof the PRESTIGEof the innovators,payingless _{attentionto the} contentor structureof INTERPERSONALLINKS. We haveargued,on _{the other} hand, that although a successfulinnovation _{needs in some sense to be} positivelyevaluated,generalizationscanbemadeaboutthesocialmechanisms controllinginnovationanddiffusionquiteindependentlyof _{theprestigevalue}

369

Rogers and Shoemaker.Further,we showed that the _{groups most}

ties _BOTH INSIDE _{AND OUTSIDE}

JAMES MILROY

AND

LESLEY

MILROY

attached to _{any given} innovation _{(see the discussion of /a/ and /?/ in Section} 4 above). Despite these difficultiesarisingpartlyfrom _{differencesin theoretical orientation,} the _persons describedby Labov _do in fact _{correspond reasonably} closely to _Rogers' and Shoemaker'saccount of _{highly conforming individuals} with strong ties inside the _group who serve as models to others. _What is _clear is that the marginals who are identifiedas _typical innovators are _{precisely the} kind of individuals to whom _Labov, in the best tradition of _small-group studies, is _likely to _pay little attention. In _{fact, they closely resemble the} famous 'lames' of the Harlem _study, who _{belong centrally neither to the} community youth networks nor to other _{networks outside} the _community. They are marginal to both, providing a tenuous link between them. We thus _emerge with a model of _linguistic innovation and diffusion _which at first sight seems counter-intuitive, although we have tried to _suggest at various points that it agreesreasonably well with historicaland _{sociolinguistic} observations. Specifically, it is suggested that at the small _{group level} linguistic innovations are transmitted across _{tenuous and marginal links.} Thus, for the very reason that persons who actuate linguistic _{change may} do so in the course of fleeting, insignificant encounters with others _occupying a similarly marginal position in their social _groups, direct _{observation of the} actuation process may be difficult,if not impossible. What we most probably CANobserve is the take-up of the innovation by the more _socially salientEARLY

ADOPTERS.

At the macro-level, societies undergoing social processes which entail social and geographicalmobility and the dissolution of _{closeknit networks (processes} associated with industrialization) provide the conditions _{under which} innovations can be rapidly transmitted along considerable social _and geographical distances _{(see Trudgill (1983, Chapter 3) for a relevant study} of geographical _diffusion). Bearing in mind the difficulty of studying _{directly the early stages of an} innovation, we proceed _{now to assess the usefulness of the model developed} here. Specificproblemsassociated _{with innovation and diffusionarediscussed,} first at the level of small groups and then at _{the level of larger national} communities.

6.

The possible explanatory value of a _{theory of weak ties can be considered}

in relation to

patterns are difficult _{to explain in terms of the usual assumptions about} linguistic diffusion, viz. that it is encouraged by frequency of contact _and relatively open channels of communication, and _{discouraged by boundaries}

observed patterns of language variation. In certain _{cases, these}

WEAK

TIES AND

LINGUISTIC

CHANGE:

A MICRO-LEVEL

EXAMPLE

of one sort or another, or weaknesses in lines of communication (see, for example, Labov (I974b) for an empirical study which links the _{location of} dialect boundaries with a trough in north-south links).

370

LINGUISTIC

CHANGE

In Belfast,two instancesstand out whichare difficultto explainin this apparentlycommon-senseway. Theyare (i) the socialconfigurationof the

spreadof /a/ backingfromthe protestanteast of thecityinto the Clonard,

a WestBelfastcatholiccommunityand(ii) thecity-wideyoungergeneration

_{consensuson evaluationof} variantsof the (pull)variable(as againstgreater _{variabilityin} the oldergeneration).Detailsof thesevariables,referredto as

(a) and(A),aremosteasilyaccessiblein Milroy& Milroy(1978),anddetails for /a/ arealso givenin Table4 (above,p. 357)and Figure3.

_(a) index score

350
300
250	-		- -	-H H
200		-		B
150	1	l
	Men	Women	Men	Women
	40-55	40-55	18-25	18-25
Backing of /a/ Figure 3 in Ballymacarrett,the Clonard and the Hammer.

_{Thebackingof /a/, aswehaveseen,is} ledbyEastBelfastmales:thismuch _{is indicatedby} Table4. However,as the significanceof the detailsin Table ₄ is difficultto interpret,wediscussthembrieflyhereinthelightof thegeneral _argumentof this _{paper.Figure3} is a diagrammaticrepresentationof the _{spontaneousstylepatternfor all} threeinner-citycommunities;it showsthe 'cross-over'patternthat tends to characterizechangein progress(Labov, _{1972a). The changeappearsto be carried,not by West} Belfastprotestant males (as might be expected),but by the younger FEMALEgroup in the _{CATHOLICClonardcommunity.Thisis} thegroupthatexhibitsthecross-over pattern. It maybe objected,however,that thereis a moderatelyhighincidenceof backingamongstolder Clonardmales, even though this group shows no stylisticdifferentiation(on which see below). But it is the young Clonard

femaleswhoREVERSETHEGENERALLYEXPECTEDPATTERNS.Amongstthem,the

city-widefemalepattern(awayfrom/a/ backing)is reversed:the incidence of /a/ backinginthisgroupishigherthaninolderandyoungerfemalegroups, _higherthanin the olderClonardfemalegroup,and - surprisingly- higher than amongsttheiryoungermalecounterpartsin the Clonardarea.When measuredagainstothergroups,theseyoungwomenappearto be reversing

a trend. WhenSTYLISTICpatterningis additionallytakenintoaccount,it isclearthat

37I

JAMES MILROY

AND

_LESLEY

_MILROY

this young female group is the _only Clonard _group with _{significant stylistic} differentiation on _the East _{Belfast model (see Table 4). Their usage is} innovatory in West Belfast in that the social value attached by them to the variants is the same as the social evaluation _evident in _{the East Belfast data,} but not well established in the _west of _{the city. Thus, while superficial} consideration of the _{figurescited might suggest that the young Clonard female} pattern is modelled on older _{Clonard males, such an explanation would not} account for reduction of /a/ _backing in _{other groups, nor would it account} for the use of (a) as a stylistic marker _{by the Clonard girls.} The social barriersthat inhibitcontacts betweenworking-classcommunities have been well described (see for _{example a discussion of this work in} L. Milroy, _ig80) and it is clear, as Boal (1978) has _{shown, that the inter-} community conflict in Belfast has strengthened these barriers. In _{fact, the} major traditional _{sectarian boundary in West Belfast is now marked} physically by a brick and barbed wire structure, which is described by the military authorities, _{apparently without intentional irony, as 'The Peace} Line.' The puzzle is, that an East Belfast pattern can be carried across _these boundaries, evidently by a _group of _{young women whose physical movements} and face-to-face contacts _{have been constrained from a very early age. It is} clear that the diffusion of patterns of /a/ backing from _{east to west, pro-} gressing in a linguistically and stylistically orderedmanner, _{is a continuation} of the long term shift in the Belfast vowel system (together with the _social values attached to it) described in Section _{3. That this shift is continuing} apparently unhindered across _{the iron barriers, both physical and psycho-} logical, which _{separate protestant East and catholic West Belfast, is a fact} for which up until this point we have not felt _{able to propose any principled} explanation. The continuation of the change may now be consideredin termsof _{the claim} that INNOVATORS who aremarginalto a group introduceinnovations, _{to EARLY} ADOPTERS who are _{central figures within that group. The innovation is likely} to _be transmitted _{by means} of _{weak, rather than strong, ties.} In addition to scoring high on _{/a/ backing} - a score _{reflecting both} quantitatively and _{qualitatively} the _speaker's choice of _realization - _the Clonard girls scored _{extremely high} on the _{Network Strength Scale, which} was designed to measure relativecentrality of _position in _{the closeknit group.}

Hence, they resemble Rogers' and Shoemaker'sEARLY

_strictly, INNOVATORS.

As describedin L. Milroy (I980), the girlswere all in _employment and _were all associated with the same rather poor city-centre store. This store was located in North _Street, a _{shopping area on the sectarian interface which} served both protestants and _{catholics, mainly} those _living west _of the _river (the girls' male _counterparts contrasted _sharply with them in _being un- employed and scoring low on network strength). We need to _emphasize at this _{point that when we argue, with Boal, that}

ADOPTERS _ratherthan,

372

LINGUISTIC

CHANGE

thereare few ties betweenworking-classgroupsin Belfastwe mean more properlythat _thereare few STRONG TIESsuchas those of _{kin, friendshipor work,particularlyacrossthe} sectariandivide.But thereareplentyof weak ties(to whichwehavehithertopaidlittleattention)between,on theonehand, WestBelfastcatholicsand _{protestantsand} on _{the other,betweenEastand} WestBelfastprotestants.Someof the _settingsin whichtheyregularlymeet are(asreportedbyinformantsin theinnercitystudy)shops,hospitalwaiting rooms,socialsecurityofficesandjob _{centres.The Clonardgirlsworkingin} the shop wouldbe extremelywell _{placedto adopt} innovationstransmitted _bypersonson theedgeof _{theirnetworkwhoin} turnprovidedweaklinkswith othercommunities.Wehavealreadyarguedthatif _{an innovationis to stand} anychanceofadoption,theseweaktieencounterswouldneedto befrequent- thatis, theywouldneedto be witha _{largenumberof} back[a]users.It _may be surmised,giventhe numberof serviceencountersin the shopin _any one day,thatweak-tieencounterswithback[a]userswhotransmittheinnovation will greatlyexceedin numberstrong-tieencounterswithnon-back[a]users. Hence the capacity of innovation-bearingweak ties to _compete with innovation-resistingstrongties. If we havea theoreticalperspectivesuchas _{the one developedhere,which} explicitlypredictsthatan innovationwillbe transmittedthroughweakties, perhapsin casualserviceencountersperceivedby _{participantsto be of no} affectivevalue,the back[a]diffusionproblemdissolves.Theproblemarises in the first place only if we assumethat strong ties must be involvedin diffusionof innovations;forin thatcase,a searchforanexplanationincasual encountersin waitingrooms,shopsanddolequeueslooksliketheworstkind

of ad hoc-ery.

The secondpuzzleconcernsthe _{(pull)variable,whichis associatedwitha} small numberof lexical items alternatingbetweenthe two _phonological classes/u/ and /A/ - examplesarepull,push, _{took, shook,foot. Thecomplex} historyof this subset(seeJ. Milroy I980 for details)is apparentlyreflected in greatinstabilityamongall _{but the youngerinnercity speakersboth with} regard to the specific lexical items assigned _{to one or another of the} phonologicalsets, and with _{regardto the social value assignedto the [A]} variant.Thus,forexample,somespeakersexplicitlystigmatized[A] _realizations of itemslikepullandpush,whileothers,inso _{farastheyusedan [A] realization}

for readingthemon a wordlist, apparentlyconsideredformslike _{[pAl] and [pAJ]} to be 'correct'.Overall,[A] realizationswereparticularlyfrequentin the Clonard,especiallyamongthe olderwomenand evenin carefulstyles. Whenwe turnto theyoungergeneration,thepicturechangesradically,as can be seenfromFigure4. Whatthisdiagramreflectsis a _processof lexical diffusion,wherebyitemswhichalternatebetween[A] and[u]realizationsare

_{graduallystabilisingin the /u/}

alternatehaveassumedveryconsiderablesociolinguisticsignificance,the _[A] realizationsbeing perceivedas _{stronglysymbolicof} Belfastworkingclass

_set. But the _few items whichcontinueto

373

JAMES MILROY			AND	LESLEY	MILROY
(A)	%
75
25					c
	Men	Women		Men	Women
	40-55	40-55		18--25	18-25

Figure 4 Distribution of the (pull) _{variable (% of [A] variants are shown) by age, sex and area} in inner city Belfast.

languageandculture.As such,theyareconsistentlyusedby _youngmenvery muchmorethanby _{youngwomen,as Figure4 shows.} Thepuzzleis of coursehowyoungpeoplelivingin _{theclosedcommunities}

of Ballymacarrett,ClonardandHammer,whosecontactwithothersoutside their areas has been _only of a _very tenuous _kind, have come to reach cross-communityconsensuson the social value to be _assignedto the two variantsof the (pull)variable.Paradoxicallytheir _parents,who formed friendshipsmuchmore freelyacrosssectarianand _{communityboundaries} (untilthe _beginningof the civildisordersin _I969), stillshowedconsiderable variationbothin the use andthe evaluationof the variable.Thisvariability was apparentlyin the first place partly a consequenceof the different phonologies of various hinterlanddialects. _Yet, the _process of _lexical diffusionand the absorptionof the (pull)variableinto the _{regularsocio-}

Belfast's urban dialect continued _unhindered,

linguistic structure of

apparentlyunaffectedby theinabilityof the _{youngergenerationto contract} any strong interpersonalties across the sectariandivide. For it is these youngsters,and not their parents,who show dramaticagreementon the formwhichthesesociolinguisticpatternsshouldtake.

Althoughthereisstillagreatdealto explainaboutthechangingdistribution of a complexphono-lexicalset like _{(pull),the questionof how city-wide} consensusonitsuseandevaluationwasreachedbytheyoungerspeakersdoes not now seempuzzling.Likethe _{diffusionpatternof /a/ backing,the (pull)} problemdissolvesif we _{acceptthat weak ties are the normalchannelfor} diffusionof innovations.

374

LINGUISTIC

_CHANGE

Having discussedthese details of _change and diffusionin _present-day Belfast,we turnin the next sectionto the _place of weak ties in _long-term languagechange.

7.

WEAK

TIES AND

LANGUAGE

CHANGE:

A MACRO-LEVEL

_EXAMPLE

It is wellknownthatin the courseof historysomelanguageshavechanged moreradicallythanothers.In theIndo-Europeanfamily,certainlanguages, such as Lithuanian,are acknowledgedto be highlyconservative,whereas others(e.g.English,Dutch,French,Portuguese)havedivergedverymarkedly fromtheirancestralforms.Furthermore,in _{thehistoryof certainlanguages} there have been periodsof rapidchange and _periodsof slow _change.A comparisonof thesocialandculturalconditionsobtainingin _{periodsof slow} and rapidchangeshouldcast _lighton the socialmotivationof _changes. Manyargumentshavebeenadducedto _{accountfor largescalelinguistic} changes;forexample,substratumtheoriesandaccountsof _{lexical,syntactic} andphonologicalborrowing.Culturalfactorshavealsobeendiscussed,such as languagecontactfollowingconquestand _{settlementof alienspeakers.In} recentdecadesmuchattentionhasbeenpaidto _{pidginizationandcreolization} (Todd, I974), and pidginlanguagesare of _{coursethe paradigmaticcase of} linguisticinstability;theycan _{changeveryrapidly.As argumentsbasedon} substratum,conquest,etc., _{arenot uniformlyapplicableto all situations,it} maybethata moregeneralcondition(inlinewiththeargumentof _thispaper) can be proposed,that will encompassthesevariedsituations.This _{can be} statedas follows:

LINGUISTICCHANGEISSLOWTOTHEEXTENTTHATTHERELEVANTPOPULATIONS ARE WELL ESTABLISHEDAND _{BOUND BY STRONG TIES, WHEREASIT IS RAPID} TO THE EXTENTTHAT WEAK TIESEXISTIN POPULATIONS.

Wecan seeksupportfor thishypothesisby _{comparingtwo languagesthat} havechangedat verydifferentrates. Amongstthe Germaniclanguages,Icelandicand _{Englishprovidea sharp} contrastin rate of change and degreeof variation.WhereasEnglishhas changedradicallysincethetwelfthcenturyandhasat _{allrecoverableperiods}

exhibited gross dialectal variation, Icelandichas _{altered little since the} thirteenthcentury _{and reportedlyshows very little dialectal variation.} Icelandicmaintainsa fullinflexionalsystemforcase,number,gender,person, tenseandmood;phonologicalchangehasbeenslight,involvingtwomergers of low functionalyieldandveryminorconsonantchanges;phoneticchanges includediphthongizationof _{long vowels and some allophonicchangesin}

_{clear how far these had already}

consonants _{and vowels, but it is not}

progressedin the Middle Ages (some fifteenth-centuryspellingsalready

indicatediphthongizationof certainlongvowelspresumablysometimeafter

375

JAMES MILROY

AND

LESLEY

_MILROY

the changes had occurred). _{Change in English,} on the other _hand, is _quite radical - amounting to a typological change from a _highly inflected to a weakly inflected _language. There are also _{many phonological changes,} word-order changes and partial relexification from Romance and Classical sources (for a brief _history see J. _{Milroy, I984a).} Notice _that the _geographical isolation of Icelandic _{(although relevant) cannot be a} sufficientexplanation for its _long-term conservatism. If _geogra- phical _isolation were the most _{important factor,} we should _expect the _dialects of Icelandic to have divergedconsiderably. Iceland is _comparable in size with Britain, but the centre of the _country is _glaciated, and settlements are scattered _around the _coastal areas. _The climate and terrain are _such that in the Middle Ages little communication was _possible in the winter months (conventionally October to _{April). According} to the Icelandic _sagas, the _{journey to the} main _assembly at _Thingvellir could take weeks. In _Hrafnkels Saga, the hero's _journey from eastern _Iceland to _{the west} is _described, and the writer comments:

Sui6r6r Fljotsdal eru sjautjaindaglei6ir a'Dingvdll (South from Fljotsdal it is a seventeen day journey to Thingvellir).

Hrafnkel's _{rival, Saimr,had an even longer journey:}

Ok f6rsk honum jvi _seinna, at hann aittilengra lei6 (And his journey was so much slower in that he had a _{longer route).}

A theory of change based mainly on the separation of communities would surely predict that varieties would diverge _rapidly in these conditions. Our hypothesis on the other hand predicts that if widely separated communities maintain the same linguistic forms, ties between them must in _some sense _be strong, and evidence _from the _{Icelandic family sagas (c. 1200-1300) seems} to bear this out. Iceland was colonized in the late ninth _{and tenth centuriesby independently} minded Norwegians, some of _{whom had settled in the Orkneys, Shetlands} and Hebrides prior to their emigration to Iceland. There was little _social

stratification in the _{Icelandic Commonwealth:}

the feudal system had no effect until after the _{annexation of Iceland by the}

Norwegian crown in the late thirteenth century. Although Christianity was accepted officiallyin IOOO,the _{temporal power of the Church appearsto have}

been less than elsewhere. In Icelandic writings, the early missionaries are

represented as thugs, and _{the status of}

been hardly better than that of farm-servants. In short, _{institutional power} seems in general to have been weak enough to allow informal _{kinds of social} organization to _flourish. The thin population was widely distributed, but an early _{form of quasi-} democratic government evolved. The country was divided into districts, and, in these, assemblies were held at which attempts were made to settle _disputes

376

_{there was no aristocracy, and}

_{priests seems for some time to have}

LINGUISTIC

_CHANGE

andpronouncejudgmenton wrong-doers.Everyyear,thenationalassembly (theAlthing)washeldatThingvellir,nearReykjavik,andpeoplewouldtravel verylongdistancesto this.Accountsinthesagassuggestthatthisinstitutional superstructurewas not verysuccessfulin settlinglegaldisputes,and it was certainlyunableto carryoutpunishments.Inorderto _{getredressforoffences,} peoplewerein practicewhollydependenton thesupportof theirfamiliesand friendsandthosewhohadobligationsto _{them.It wasveryimportant,in the} absenceof stronginstitutionalpower,thatstrongties shouldbe _maintained withthose who mighthelpin a timeof need. Theassemblieswere,in practice,a meansof maintainingstrongtiesacross long distances,and the sagas furthershow the greatimportancethat was attachedto personalidentity,kin and _{friendship.Whena new characteris} introducedby the saga-writer,a paragraphor moreis typicallydevotedto naminghisparentsandgrandparents(andsometimesdistantancestors),his brothersand sisters,his wife and _{family (and sometimesother relatives).} Whena strangerappearsin the story,he is oftenquestionedabouthisname, his home,his relativesandhis status.WhenSaimr,in HrafnkelsSaga,meets

a stranger(whomaybeableto assistin a law-suit),heaskshisname,whether or not he is a localleader(godordsmadr)or _{farmer(bondi),who his brothers} are,andso _on. Thestranger'srepliesgivemoreinformationthanthemodern readermightthinknecessary.In HrafnkelsSaga,thestrangertellsSaimrthat hisbrother'snameis _Pormo6, thatPormo'6rlivesatGar6aronAlptanesand

thatheismarriedtoDordis,whoisthedaughterofD0rolfr,sonofSkalla-Grim,

fromBorg.Thiskindof _{exchangeof informationis typicalof the saga;it is} alsotypicalof communitiesthatdependonmaintainingstrongnetworklinks. Similarexchanges,the purpose of which is to declareidentity,political affiliationand personalrelationships,were _{reportedby informantsin the} Belfastproject(L.Milroy,I980: 55); theseinformantswereattachedto _strong territoriallybasedsocialnetworks. The conservatismof _{Icelandicand the relativelack of variationin that} languagemaythereforebeattributedlargelyto _{thegreatpracticalimportance} attachedtomaintainingstronglyestablishedkinandfriendshipnetworksover long distances and through many generations.As _{in the low-status}

communities described by Lomnitz _(I977) and discussed by L. _{Milroy (I980:}

70 ff), the patternsof exchangeand obligationimposedby _{such network} structuresensurespracticalsupportin _{timeof need.Sucha socialstructure} (basedon informallinks)could _{flourishin medievalIcelandbecauseof the} inabilityof pan-Europeaninstitutions(theChurchandthefeudalsystem)to

establishtheirpowerfully.Oneof _{theresultsof thisinformalsocialstructure}

is the impositionof linguisticnormson _{its members(in commonwithother}

norms). Hence the failure of the language to exhibit _{much change or} variation,despitethe _{difficultiesof distanceand terrain.}

The history of English, which is dramaticallydifferentfrom _that

_of

Icelandic,canhardlybeunaffectedbypopulationhistory.Inearlytimes,there

377

JAMES MILROY

AND

LESLEY

MILROY

is a _history of _repeated incursions. Danish settlers in Eastern _{England in the}

ninth and tenth centuries found Old _{English (Anglo-Saxon)} well _established, whereasthecontemporaryNorwegian settlersin Icelandfound an uninhabited

country. The numerous Scandinavianplace-names of _Yorkshire, Lincolnshire _and elsewhere _{strongly suggest} that _{Danish-speaking} communities _survived

in these areas for some time. If _so, the ties contracted between the Danes and

the _{English could not,} for social and _{linguistic reasons,} have been _strong in _the first _place. The communication that must have taken _place in the course of _{trade and farming} seems to have been carried on in an _{Anglo-Danish contact language.} This is indicated _by the nature of the _language that _emerges

in the Middle English texts of these eastern areas _(e.g. The _Peterborough

Chronicle, I 137) which is an Anglo-Saxon-based language with _gross in- flexional loss, absence of grammaticalgender, and _partial reflexificationfrom Danish and Norman French. On the other _hand, the _English of the West Midlands around 1200 - an area _largely unaffected _by the Danes - _provides

a _startling contrast. The Ancrene _Wisse, for _example, is _{morphologically}

conservative (in that gender and case inflexions are largely _retained), and Danish _loanwords are _very rare. _Thus, we _appear to have _{relatively rapid} change in areas where pre-existing strong networks are disrupted and where influence through weak ties is made possible: on the other hand _{we have} a conservative _language in areas of the West Midlands where _Anglo-Saxon institutions remained more stable, and where neither Danish nor _Norman influence was initially strong. The success of the Norman Conquest imposed a tight and _organised administration on much of the _country; rule was more _{centralized, and class} divisions more fully institutionalized _by the feudal _{aristocracy.While Iceland} remained a yeoman _democracy, England acquired an institutional system of social stratification. One of the effects of stratificationis the creation of social _{distance between} sectors of the population. Two developments in English may be _{a general} consequence of social distance and weak ties. _{The first is the character of} relexificationfrom Norman _{French. Thereis a rapiddevelopment of English/} French synonyms of the type child/infant, love/charity, _board/table, stool/chair: the French synonyms tend at first to be limited to _{more formal} social contexts. The second development is the use of the polite _{pronoun of} address, which was marked for status and social _{distance: it was used} asymmetrically and non-reciprocally by inferiors to superiors. _{Brown and} Gilman (1972) estimate that the non-reciprocalpolite plural pronoun _entered

most Europeanvernacularsbetween I _I00

_{and I300,}

_{with French very}

advanced in this respect. In Icelandic, this development is relatively late. It

for use between

is unknown in the sagas of the Icelanders (c. I200-I300)

Icelanders, although Icelandic adventurers (around _{I300) are occasionally}

representedas addressing some European monarch _with the _{polite pronoun.}

378

LINGUISTIC

_CHANGE

Itisnotuntilchivalricsagason non-Icelandicthemesappearinthefourteenth centurythatthenon-reciprocalpolitepronounbecomesreasonablycommon in literature(oftentranslated).As _{faras we areaware,thereis no indication} that, at this date,nativeIcelandersusedit _{amongstthemselves.} A finaldevelopmentin Englandthattendedto _{encouragethe break-upof} strongties and the _{developmentof} weaktieswas _{the risein the importance} andpopulationof London.Londonbecamethe _{seatof the Court,themain} commercialcity and the centre of the wealthiestpart of the _country. Immigrationto London(Ekwall,1956;Strang,I970: _{2I4 f.) wasfrommany} areas,but largelyfromthe EastMidlands(resultingin _{a gradualchangein} the dialectfromsouthernto EastMidlands).Therapidinflexionalloss _that diffusedthroughoutthe ME _{periodcan be seen,not only as a resultof the} influenceof weaklyinflected(E. Midland)dialectson stronglyinflectedones, butas a productin Londonof thecontactsituationitself,in _{which'mergers}

_{expandat theexpenseof} distinctions'(Herzog,quotedin _Labov,I972:

In thirteenth- and fourteenth-centuryIceland, there _{were no such} developments.Icelandersin searchof theirfortunehadno _{largetownto settle} in; theytendedto go abroadfor a timeandthenreturnto _{theirruralhomes} in Iceland.In suchconditions,strongnetworksremainedto _{a largeextent} intact. Thus,thecontrastbetweenEnglishandIcelandicseemsto _{be an exemplar} of thecontrastbetweensocialconditionsthatencourageweaktiesandthose thatencouragestrongties. _{Rapidchangesin Englishseemto havedepended} ontheexistenceofindividualsandgroupswhoweresociallyandgeographically mobileand whosestrongnetworktieswereweakenedor _{brokenup by this} mobility.A highdegreeof socialdistanceseemsto haveresulted.Icelandic society, on the other hand, dependedin earliercenturieson _{the strong} networkstypicalof _{rurallife. Hence,despitethe difficultiesof climateand} terrain,socialnetworksprovedto bea cohesiveforce,notonlyinmaintaining socialnorms,but _{also in maintainingthe normsof language.} We havediscussedthe _{case of Icelandicand Englishin orderto support} the generalizationstatedon p. 375 above that 'linguisticchangeis slow _to the extentthat the relevantpopulationsare _{well establishedand boundby} strongties,whereasitisrapidto theextentthatweaktiesexistinpopulations'. Casesof conquestand colonizationare takenas relevanttypesof _weak-tie situationsforthereasonthatrapidchangeis _{oftenassociatedwithsuchcases.} Nigel Vincent(personalcommunication)drawsour _{attentionto a possible} Romanceanalogue.Sardinianis _{generallyregardedas themostconservative} of the Romancelanguageson _{a numberof counts,and this stateof affairs} canplausiblybecorrelatedwiththefactthataftertheperiodof _Romanization (3rdcenturyB.C.), suchincursionsand _{occupationsas therewerehad only} a marginaleffecton the socialorganizationof _{the inhabitantsof the island,} and even then only in peripheralareas(see BlascoFerrer,I984). _Sicily,by

_300).

379

JAMES MILROY

AND

LESLEY

MILROY

contrast, has a _{long history} of _conquest and colonization _{by Greeks,} Normans, Lombards and Arabs, and linguistically Sicilian _displays a _good deal of innovation and stratification _{(see Varvaro, I98I).} However, our _{generalization} is intended to _{encompass any} situation _where

contacts between _people lead to the establishment _{of many weak ties. It} therefore _includes, in _principle, situations where warlike incursions _{are not} of major _importance. Two further broad and _{overlapping types} of _contact spring to mind. One is the _{peaceful in-migration} of _populations who _speak other _languages or dialects. This _{might help to account for change} in _the maritime colonial _languages of Western _{Europe (as against} those of _the interior), as metropolitan centres are characterizedby some ethnic and _social diversity; this is also _clearly relevant to _{rapid change} in _{large post-Industrial} Revolution cities, in which much of the _in-migration from _{the hinterland is} by speakers of different dialects of the _{same language.} The _{second type is} sustained commercial and cultural contact _(which leads in extreme _{cases not} only to simplificationbut to _{'language death').} These _types of _{contact would} seem to be relevant to cases like Danish (an old established _maritime and

language closely related to _Icelandic, but which has _{undergone rapid}

colonial

change); the history of that country has been characterizedby _quarrelswith neighbouring states, but also by sustained _commercial and _{cultural contact} with these states. Therefore, we do not consider _{incursion and conquest to} be a necessary condition in _{itself for rapid linguistic change. The correct} generalization must _{account more broadly for the spread of speaker in-} novations through _{weak ties.}

8.

CONCLUSION

We have here presented a model designed to explain why linguistic change seems commonly to take place in some social conditions but not in _others. Specifically, we have tried _{to show as explicitly as possible that innovations} are normally transmitted from _{one group to another by persons who have} weak ties with both groups. Further, at the macro-level, it is _{suggested that} in situations of mobility or social instability, where _{the proportion of weak} links in a community is consequently high, _{linguistic change is likely to be} rapid. Social groups who characteristicallycontract _{many weak ties} - _{and in} Western society these could consist of persons who _{belong neither to the} highest nor to the lowest social _groups - _{are likely to be closely implicated} in the large scale diffusion of linguistic innovations. These claims are supported by empirical observations. For _{example, it has} been noted that innovations seem to _{hop from one centre of population to} another, along main lines of communication such as _{roads and railways} (Trudgill, I983: Chapter 3). This is to be expected if we assume firstthat _they are carried by persons from community A who have _{weak ties with those in} community B, and _{second that ties contracted in these contexts are likely to}

380

LINGUISTIC

_CHANGE

be numerous.Ourargumentsherealso fit _{in with Labov'sfindingthat the} locusof changeis _alwaysat somecentralpointin thesocialhierarchywhere, we havesuggested,tiesareweak.Forthisreason,an accountbasedon _weak ties seemsto be at a higherlevelof _{generalitythan} one based on class _or status. Nor doestheevaluativenotionof PRESTIGE(overtorcovert)havea _central partto playin themodelpresentedhere.Labovis _{correctin his observation} thatlinguisticinnovationsmaydiffusebothupwardsanddownwardsthrough the socialhierarchy;someappearto _{originatewithhighand somewithlow} statusgroups.Conversely,a _{comparisonof the diffusionmechanismsand} distributionalpatternsof /a/ and _/E/ inUlstershowsthatelementsoriginating from the same (rural)dialectcan take on, apparentlyarbitrarily,entirely differentsocialvaluesin _{theirnewurbancontext.} Weassume(althoughanydiscussionis _{beyondthescopeof thispaper)that} perceptualandacousticfactors,as _{wellas a rangeof moregenerallinguistic} constraints,willsharplylimittheclassof _{possibleinnovations(cf.Weinreich}

et al., I968:

IOO; Labov,I982:

27;

Comrie,I98I:

195). Butwithinthelimits

set by thislargerclassof _{constraints,the notionof prestigedoes seemto be}

importantin explainingwhy one particularlinguisticelementis _{a realistic} candidatefor innovationwhileothersare _{not. We have suggestedthat the} workingclassof EastBelfast,whoformeda kindof labouraristocracy,were particularlystronglyassociatedwiththe _{UlsterScotshinterlandfromwhich} thecontemporaryurban/a/ and _/?/ changeshaveoriginated.Thisassociation helpsexplainwhythesephonologicalelementsandnot, _{for example,others}

associatedwith mid-Ulsterdialects,have been successfullyintroducedand diffused.Butsincethistopicalsoliesbeyondourscopehere,we _simplynote, withTarde,thatweneedto learnwhy,if a hundredinnovationsareconceived simultaneously,ten will _{spreadwhileninetywill be forgotten(I903: 140).} It has also been necessaryto distinguishsharplybetweenINNOVATION (whichis the act of a speakeror speakers)and CHANGE, whichis the _reflex of a successfulinnovationin thelanguagesystem.Presentdaysociolinguistics (althoughsensitiveto _{social phenomena)is in fact stronglyorientedto a} 'system' approachand has often not made a sufficientlysharpdistinction betweenthelinguisticbehaviourof _{speakersandtheeffectof thatbehaviour} on the languagesystem. Finally, by makinga furtherdistinctionbetweenthe _{INNOVATORS of a} linguisticchangeand _{the EARLY ADOPTERS, we have suggesteda principled} reason for the difficultyexperiencedin _{observingthe introductionof an} innovationinto a community.This may be seen as the earlieststage of _a

linguisticchange - _{at least fromthe point of view of}

_{the communitywhich}

is adoptingit. Weinreichet al. havedescribedthis actuationof a _changeas

'the _very heart of the matter'. _However, since innovators tend to _be marginalindividualsattheedgesof networkswhodiffuseinnovationviaweak ties with others,the personswhom investigatorsactuallyidentifyas _being

38I

JAMES MILROY

AND

LESLEY

MILROY

stronglyassociatedwitha changearemostprobablythemoresociallycentral earlyadopters. It is importantto notethelimitationsof theclaimspresentedhere.We _are

attemptingto shedlighton the actuationof a linguisticchange,notingthat thereis not necessarilya one-to-onerelationshipbetweenevena successful speakerinnovationand the changein the languagesystemwhichreflectsit.

A singleinnovationmaytriggeroff a seriesof _{changesin a chainshiftwhich}

canthenbestbeexplainedbyexaminingtheinternalorganisationalprinciples underlyingthe languagesystem.It is _assumedthat an _{appropriatelyex-} planatoryaccountof languagechangemustsupplementthemodelpresented herein at leasttwo importantways. First,it mustspecifythepsycholinguisticandlinguisticconstraintswhich limitthe classof candidatesfor innovation.Second,it mustaccountfor _the regularand orderlymannerin which _{successfulinnovationsare diffused} throughoutthe system,so thateventuallytheyareperceivedas _instancesof linguisticchange.

REFERENCES

Bailey, C.-J. (1973). Variation and linguistic theory. Washington D.C.: Center for _Applied

Linguistics.

Baugh, A. C. & Cable, T. (1978). History of the English language, 3rd ed. London: _Routledge

& KeganPaul. _{BlascoFerrer,E. (I984).} StorialinguisticadellaSardegna.Tiibingen:MaxNiemeyerVerlag. Boal,F. W.(I978). TerritorialityontheShankill-Fallsdivide,Belfast:theperspectivefromI976.

In Lanegran,D. A. &Palm,R. _(eds),An invitationto _{geography. 58-77.}

	Boone,L. (1949). Patternsof innovationin the languageof the oil-field.AmericanSpeech 24.
	31-37.
	Brown,R. &Gilman,A. (1972). Pronounsof powerandsolidarity.Reprintedin Giglioli,P. P.
	Languageandsocialcontext.Harmondsworth:Penguin.252-82.
	Bynon,T. (I977).	Historical linguistics. Cambridge:CambridgeUniversityPress.
	Chambers,J. K. &Trudgill,P. J. (I980). Dialectology.Cambridge:CambridgeUniversityPress.
	Chomsky,N. (1975). Language and responsibility.Sussex:HarvesterPress.
	Chomsky,N. & Halle,M. (I968). Thesoundpatternof English.New York:HarperandRow.
	Comrie, B. (I98I).	Language universalsand linguistic typology. Oxford: Blackwell.
	Downes,W. (I984). Language and society. London:Fontana. Eckert,P. (I980). The structureof a long-termphonologicalprocess:the backchainshiftin SouletanGascon.In Labov,W. (ed.).Locatinglanguagein timeandspace.179-2 I9.
	Ekwall,E. (I956).	Studieson thepopulationofmediaeval London.Lund:LundStudiesinEnglish.
	Gal, S. (1979). Language shift. New York:AcademicPress.
	Granovetter,M. (I973). Thestrengthof weakties.AmericanJournalof Sociology 78. I360-I380.

Gregg,R. J. (I964).

Scotch-Irishurbanspeechin _Ulster. In _{Ulster dialects -a}

_symposium.

Holywood,Co. Down:UlsterFolk Museum.I63-I9I. Gregg,R. J.(1972). TheScotch-Irishdialectboundariesof _{Ulster.InWakelin,M.(ed.),Patterns}

in thefolk speech of the British Isles. _I09-139.

Gumperz,J. J. (I982).

Halsey,A. H. (1978). Change in British society. Oxford:OxfordUniversityPress. Harris,J.(I983). Linguisticchangeina nonstandarddialect.UnpublishedPhDthesis,University of Edinburgh.

Heilmann, L. (ed.) (1974).

Bologna:IIMulino. Kroch,A. S. (I978). Towardsa theoryof socialdialectvariation.Language _{in Society 7. I7-36.}

Discourse strategies. Cambridge:CambridgeUniversityPress.

Proceedings of the Eleventh International Congress of Linguists.

382

LINGUISTIC

_CHANGE

Kroch, A. S. (I979). Dialect and style in the speech of upper class Philadelphia. _Mimeo:

University of Pennsylvania. Labov, W. (1972). Sociolinguisticpatterns. Philadelphia: Philadelphia University Press. Labov, W. (1974a). On the use of the present to explain the past. In Heilmann, L. (ed.), Proceedings of the EleventhInternationalCongressof Linguists. Ii ii-I 135. Labov, W. (1974b). Linguistic change as a form of communication. In Silverstein, A. (ed.), Human communication: theoretical explorations, Hillsdale, New Jersey: Lawrence Erlbaum Associates. Labov, W. (I980). Locating language in time and space. New York: Academic Press.

Labov, W. (I98I). Resolving the neogrammarian controversy. Lg 57. 267-308.

Labov, W. (I982).

Perspectiveson _{historical linguistics. Amsterdam and Philadelphia: John Benjamins. 17-92.}

Labov, W., Yaeger, M. & Steiner, R. _(1972). Report on NSF Project No. _65-3287.

Lanegran, D. A. & Palm, R. (eds) (1978). An invitation to geography (2nd _ed.). New _York:

Building on empirical foundations. In _Lehmann, W. P _{& Malkiel, Y. (eds),}

A quantitativestudy _of sound _change in _progress.

McGraw Hill.

Lass, R. (I976).

Press. Lass, R. (I980).

Lehmann, W. P. (I962).

Englishphonology _{andphonological theory. Cambridge: Cambridge University}

On explaining language change. Cambridge: Cambridge University Press.

Historical _{linguistics: an introduction.New York: Holt, Rinehart &}

Winston. Lomnitz, L. (1977). Networks and marginality.New York: _{Academic Press.} Milroy, J. (1976). Length and height variations in the _{vowels of Belfast vernacular. Belfast} WorkingPapers in Language and Linguistics I. 3. Milroy, J. (I980). Lexical alternations and the _{history of English. In Traugott, E. (ed.), Papers} from the InternationalConferenceon _{Historical Linguistics. Amsterdam: Benjamins. 355-62.}

Milroy, J. (I98I). Regional accents of English: Belfast. Belfast: Blackstaff. Milroy, J. (I982 a). Probing under the tip of the iceberg: phonological 'normalization' and the shape of speech communities. In Romaine, S. (ed.), Sociolinguistic _{variation in speech} communities.London: _{Edward Arnold, I982. 35-47.} Milroy, J. (I982b). Some connections between _{Galloway and Ulster speech. Scottish Language} I (Winter issue). Milroy, J. (I984a). The history of English in the British _{Isles. In} Trudgill, P. (ed.), _Language in the British Isles. Cambridge: _{Cambridge University Press. 5-3I.}

Milroy, J. (I984b). Presentday evidence for historical _changes. In _{Blake, N. F. & Jones, C. (eds),} Progress in English historical linguistics. Sheffield: University of Sheffield. Milroy, J. & Harris, J. (I980). When is a merger not a merger?: the MEAT/MATEproblem in a present-day English vernacular. _{English World-WideI. 199-210.} Milroy, J. & Milroy, L. (1978). Belfast: _change and variation in an urban _{vernacular.In Trudgill,}

P. (ed.), Sociolinguisticpatterns in _{British English.}

Milroy, J., Milroy, L., Harris, J., Gunn, B., Pitts, A. & Policansky, L. (I983). Sociolinguistic variationandlinguisticchangeinBelfast. _{Report to the Social Sciences ResearchCouncil, Grant} No. HR5777. Milroy, L. (I980). Language and social _{networks.Oxford: Blackwell.} Patterson, D. (i 86o). Provincialismsof _{Belfast. Belfast.}

Pitts, A. (I982).

Urban influence in Northern Irish English: a comparison of variation in _two

communities. Unpublished PhD dissertation, _University of _Michigan.

Rogers, E. M. & Shoemaker, F. F. (1971).

Communicationof innovations(2nd _{ed.). New York:}

Free Press. Romaine, S. (I98 I). The status of variable rules in sociolinguistic _theory. JL _{I7. 93- I19.} Staples, J. H. (I898). Notes on Ulster English Dialect. Transactionsof the _{Philological Society.}

357-387.

Strang, B. M. H. (1970). A History of Engish. London: _Methuen.

Tarde, G. (1903). The laws of imitation (translated by Elsie Clews Parsons). New York: _Holt. Todd, L. _(I974). Pidgins and creoles. London: Routledge & Kegan Paul. Trench, R. C. _{(I888). On the study of words, 20th ed. London: Kegan Paul, Trench & Co.}

Trudgill, P. (ed.) (1978).

Trudgill, P. (I983). On dialect. Oxford: _Blackwell.

Sociolinguisticpatterns _{in British English. London: Arnold.}

383

_{JAMES MILROY}

AND

LESLEY

MILROY

_{Varvaro, A. (I98I). Lingua} e storia in Sicilia. Palermo: Sellerio. _{Vennemann, T. (I983). Causality} in _{language change.} Folia Linguistica Historica IV/I. 5-26.

_{Wakelin, M. (ed.) (I972). Williams, R. A. (1903).}

Quarterly6. 129-I35. _{Weinreich, U., Labov, W. & Herzog,} M. _(I968). Empiricalfoundations for a theory of language

_Remarks on Northern Irish _{pronunciation} of English. ModernLanguage

_Patterns in thefolk speech of the British Isles. London: Athlone.

_{change. In Lehmann, W. & Malkiel,} Y. _(eds), Directions for historical linguistics. Austin:

University of Texas Press. 95-195.

384