ABSTRACT
Before1980,studiesofBa~rainiArabictreateditwithinan arealdialectalcontext.Johnstone's(1963)and(1965)articles, forinstance,focussedrespectivelyontwophonologicalchanges whichhaveaffectedalargegroupofEasternArabiandialectsthebackingofthealveolaraffricate/j/toapalatalfrictionless continuant/y/andthefrontingandaffrictionofthevelar stops/g/and/k/toalveolaraffricates/j/and/~/.His lengthier(1967)studyprovidedthefirstthoroughdescription ofthephonologyandmorphologyoftheGulfdialectsasawhole. GiventhepracticaldifficultieswhichstillhinderthecollectionofdatainsociallyconservativeMuslimcountries(and Johnstone'sdatawascollectedinthelate50s)itisnotsurprisingtofindthatJohnstone'streatmentofindividualcountry dialectswithinthisgroupisincomplete.IntheBaorainicase, manyofthefeaturesattributedtoallEasternArabiandialects (1967:1-17)donotinfactapplytotheBaharnadialects. Morerecentstudies(Prochazka1981:Al-Tajir.l982;Holes1981, 1983a)havefilledinmanyofthegapsandprovidedamuchfuller contrastivetreatmentofBAdialects.Inbroadtypological terms,the'ArabdialectsofBaorainfollowtheclassic'nomadic' patterninkeypointsofphonologyandmorphology,whereasthe Ba!Jarnadialects·sharemanyfeatureswithlong-established non-nomadiccoastalpopulations(Holes1983a:35-7).Thefollowing are,perhaps,themostobvious'nomadic'('Arab)versus'semisedentary•(BaQarna)contrastingfeatures:
'Arabdialect
Ba9arnadialects
l}(voiceless) fd~ GaG (e.g.\iahwa) GaGaGaG/GiGGaG (e.g.~arabator
AnumberofotherfeatureslinktheBaoarnadialectswiththe sedentarycoastaldialectsofSouthernArabiaandOman,e.g. the2ndpersonfemininesuffix-~(versus'Arab--15);the obligatory-in-infixinstructuresinvolvingasuffixedactive participle,insuchphrasesas/ka:tbinnah/'he'swrittenit' (versus'Arab/ka:tbah/);theinterrogativesuffix~h/,e.g. /~a:fkum9h/?'Didheseeyou?'Thepictureisconfused,however, bythepreservationinsomevillageBa~rnadialectsofanumber offeatureswhicharenormallythoughtofinotherdialectal contexts(c.f.Ingham1982:33ff)asmoretypicalofconservative, nomadicspeechnorms.Inthespeechofmanyelderlyvillage Baharna,forexample,internalpassives(insteadoftheanalogicallyreduced-/in/and/ta/- formations)arecommon,as istheuseoftheindefinitemarkerinnounphrases.Thelack ofreliabledemographicdatafortheareaasawhole,aswell asofdataonspeechformsinuseatearlierperiodsofBaQrain's recenthistorymakeitextremelydifficulttopiecetogether acoherentpictureofhowthepresentdistributionofdialectal formscameabout.Whatisimportantfromthepointofview ofthisstudy,however,isthatinthefaceofthereductional changeswhichareaffectingallGulfdialects,andtheincreasing linguisticeffects,bothsyntacticandlexical,ofMSA,some dialectalfeatureshaveforsomereasonremainedmoreimpervious thanotherstothetideoflinguisticchange.Sometwenty-one linguisticfeatures,eightphonological,thirteenmorphophonologicalformthefocusofinvestigationofthiswork.Thequestions wehavetoaskare:
1.Whyaresomeofthesefeaturesrelativelymoreresistant toreplacementthanothers?
