The Dilemma of the Reliability of Geolinguistic and Dialectological Data for Sociolinguistic Research. The Case of the Andalusian Demerger of /θ/

S aipsniai / A icles 9 JUAN-ANDRÉS VILLENA-PONSODA Uni e sidad de Málaga, Spain Fields o esea ch: sociolinguis ics, social dialec ology, phonological a ia ion, Spanish, Andalusian a ie ies. THE DILEMMA OF THE RELIABILITY OF GEOLINGUISTIC AND DIALECTOLOGICAL DATA FOR SOCIOLINGUISTIC RESEARCH. THE CASE OF THE ANDALUSIAN DEMERGER OF /θ/ Geoling is ikos i dialek ologijos duomenų pa ikimumo dilema a liekan socioling is ikos y imus. /θ/ skilimo Andalūzijos ispanų kalboje a ejis ANNOTATION O e heo e ical and me hodological di e ences be ween, on he one hand, dialec- ology and geolinguis ics and, on he o he hand, sociolinguis ics make qui e di icul he essen ial ask o using da a om he o me wo disciplines o help he la e one comple e p esen esea ch on he speech communi ies, pa icula ly when i is abou econs uc ion o he pas o cu en a ie ies. The p oblem is he e o e o which ex en a e geolinguis ic and dialec monog aph da a eliable o e ospec i ely ollow bo h he his o y o a gi en a ie y and a pa icula linguis ic change. Building on ich esul s o complex analyses using mac osocial, mesocial and small- scale a iables o explain speech a ia ion, which e eal eme gence o a new a ie y in sou he n Spain, e ospec i e compa ison wi h geolinguis ic and dialec monog aph da a discloses ideological d awbacks ega ding ep esen a ion o bo h he speech communi y and he linguis ic s uc u e, as well as a qui e oubling con adic ion be ween geolinguis ic and dialec monog aph da a. KEYWORDS: speech a ia ion, sociolinguis ics, dialec ology, Spanish, Andalu- sian a ie ies. JUAN-ANDRÉS VILLENA-PONSODA 10 Ac a Linguis ica Li huanica LXXIX ANOTACIJA Aki aizdūs eo iniai i me odiniai dialek ologijos, geoling is ikos i socioling is ikos ski umai apsunkina labai s a bią užduo į – d iejų pi mųjų mokslų duomenis panaudo i aip, kad jais bū ų padedama ečiajam mokslui a liekan kalbinių bend uomenių y imą, ypač siekian a ku i esamų kalbos a mainų is o iją. Taigi, nag inėjama p oblema susijusi su klausimu, kiek geoling is iniai i monog a iniai a mių duomenys y a pa ikimi, siekian e - ospek y iai nus a y i i iamos kalbos a mainos is o iją i konk e ų kalbos poky į. Remian is gausiais mak osocialinius, mezosocialinius bei mik olygmens kin amuosius į aukiančių sudė ingų y imų, ku iais siekiama paaiškin i kalbos į ai o ę, ezul a ais, a - skleidžiančiais naujos kalbos a mainos a si adimą pie ų Ispanijoje, a lik as e ospek y inis palyginimas su geoling is iniais i monog a iniais a mių duomenimis pa odo ideologinius ūkumus, susijusius su kalbinės bend uomenės i kalbinės s uk ū os ep ezen a imu, aip pa i ne imą kelian į p ieš a a imą a p geoling is inių i monog a inių a mių duomenų. ESMINIAI ŽODŽIAI: kalbos į ai o ė, socioling is ika, dialek ologija, ispanų kalba, An- dalūzijos ispanų kalbos a mainos. INTRODUCTION Since da a ob ained in geolinguis ics and dialec ology and hose om cu en sociolinguis ics ha e been concei ed unde e y di e en heo ies on language and socie y, econs uc ion o he pas o a linguis ic a ie y in o ma ion – as well as i s ela ionship wi h o he a ie ies in he speech communi y – using ge- olinguis ic and dialec monog aph da a leads o a dilemma. Al hough he im- po ance o he indings ob ained by adi ional dialec ologis s and geolinguis s is e iden , i should be emphasised ha sociolinguis ic da a a e based on socio- logical assump ions abou sampling and a ia ion ha a e incompa ible wi h hose unde lying geolinguis ic and dialec ological da a, pa icula ly he oldes and ou da ed a lases and monog aphs (Villena-Ponsoda 2010). The p oblem is hen o which ex en geolinguis ic and dialec monog aph da a a e eliable o e ospec i ely ollow he his o y o a gi en a ie y and a pa icula linguis ic change.1 1 Da a and esul s discussed in his pape a e based on he Spanish Minis e io de Economía y Com- pe i i idad DGICYT Resea ch P ojec on he Sociolinguis ic Pa e ns o Cas ilian Spanish (ECO- PASOS, FFI2015-68171-C5-1 and FEDER unds). I am indeb ed o F ancisco Díaz-Mon esinos, who sc u inised he ALEA maps, o ganised and made comp ehensible ables om a huge mass o geolinguis ic da a, and o Pila Ga cía-Mou on, o he in aluable help and suppo . Ma ilde Vi- da-Cas o, F ancisco Díaz-Mon esinos, Godsuno Chela-Flo es and Ál a o Molina-Ga cía kindly e ised p e ious e sions o his pape . I am, howe e , esponsible o i s manne and con en . S aipsniai / A icles 11 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ This pape 2 aims a desc ibing a new u ban a ie y in o ma ion which in- ol es a ce ain numbe o changes – some o hem o a su p isingly as de- elopmen – e ealing con e gence owa ds he na ional s anda d a ie y and boos ed by adop ion o cu en main-p es ige alues and a i udes (Villena-Pon- soda 2008). Appa en ly, his a ie y could ha e o igina ed in Andalusia a ound he mid- i ies o he 20 h cen u y since only ew aces o i s main ea u es we e by hen documen ed by dialec ologis s and geolinguis s. Resul s o analy- ses cu en ly ca ied ou show ha now we a e dealing wi h a cohe en a ie y, which is pe cei ed as such by i s speake s (Villena, Vida 2017, 2018), and could ha e eme ged as he ou come o he end o mode nisa ion and u banisa ion ha ook place in sou he n Spain du ing he las decades o he 21 h cen u y and, abo e all, a e F anco’s ule (1939–1975). Eme gence o in e media e a ie ies be ween e nacula dialec s and he na- ional s anda d is qui e a common p ocess in he Eu opean con ex , and i seems o be ela ed o he ac ha u ban educa ed middle-class speake s need means o exp ession di e en om bo h he na ional s anda d – oo o mal o e en posh – and e nacula a ie ies – oo ude and socially ma ked (see Aue 2005: 27–28). Al hough i is ue ha an Andalusian spoken egional-s anda d ac ually exis s – based on he u ban dialec o Se ille and mean o ul il he social unc ions asc ibed o he na ional s anda d – i s in luence does no each any longe u ban a eas om eas e n (G anada, Malaga) o a wes e n (Huel a) Andalusia (see Map 2 below), so ha middle-class speake s om hese a eas ying o lee he e nacula end o shi o he na ional s anda d and adop i s main ea u es (Villena-Ponsoda 1996, 2005). The mos in e es ing ai o his end o con e gence is ha i seems o be ela ed o he indi idual speake s’ p o iles, in pa icula wi h ega d o hei mo i a ion o acqui e knowledge, hei deg ee o media exposu e and awa eness o he low o in o ma ion in socie- y (i.e. Bou dieu’s no ion o he so-called inco po a ed cul u al capi al), mo e han me ely o he cul u al capi al acqui ed a home o a school, i.e. he ins i- u ionalised and he objec i ied cul u al capi al (Bou dieu 1977, 1984, 1986). Then, he idea o a p oac i e speake who assesses s anda disa ion as a means o social mobili y and in eg a ion eme ges and p o es o be o he g ea es in e es o explain how he new a ie y has de eloped so a . 2 The p esen pape is a e ised e sion o he ex o my p esen a ion o he IX Cong ess o he In e na ional Socie y o Dialec ology and Geolinguis ics, SIDG (23–27, July 2018, Vilnius, Li hua- nia). I am g a e ul o he o ganise s o hei kind in i a ion and pa icula ly o Ju gi a Ja osla ienė om he Resea ch Cen e o he S anda d Language a he Ins i u e o he Li huanian Language o he suppo and unde s anding. JUAN-ANDRÉS VILLENA-PONSODA 12 Ac a Linguis ica Li huanica LXXIX Recons uc ion o he changes unde lying his a ie y ac oss eal ime e- eals i sel as a ma e o impo ance bu , a he same ime, b ings o ligh he ecu en p oblem o he eliabili y o he dialec ological and geolinguis ic da a o sociolinguis ic esea ch. The o ganisa ion o his pape is as ollows. In he i s sec ion (The in e media e a ie y) he new Andalusian u ban mid- dle-class a ie y in o ma ion is desc ibed, ocusing pa icula ly on i s phonology is-à- is bo h he na ional s anda d a ie y and he Andalusian e nacula dialec s. The second sec ion (The social meaning o he new a ie y) analyses sociolin- guis ic a ia ion and cohe ence o he in e media e a ie y, wi h emphasis on he e ec o mac osocial a iables on he speake s’ beha iou . In he hi d sec ion (The p oac i e speake . The impo ance o indi ec e ec s) we deal wi h he e ec o in e ening mesosocial and small-scale a iables be- ween mac osocial and dependen speech a iables – i.e. he speake ’s p obabil- i y o using he new a ie y. We show ha i is he speake ’s s anda d o ien a- ion – based on his o he capaci y o being in o med and connec ed wi h he ou e wo ld beyond he local speech communi y – wha de e mines his o he decision o using a cohe en se o ea u es whose social meaning is ela ed o mode ni y and s anda disa ion. The ou h sec ion (The dilemma) app oaches he cen al issue o his pape , namely, he limi a ions o dialec ological da a o he pas o econs uc he his- o ical de elopmen o oday’s a ie ies. To do so, wo main poin s a e consid- e ed. Fi s , ou knowledge o he pas speech communi y, which leads o un- damen al me hodological di e ences be ween, on he one hand, dialec ology and geolinguis ics and, on he o he , sociolinguis ics. Second, he p oblem o he accessible linguis ic da a, whose obse a ion e eals con adic ions be ween geolinguis ic da a and hose om dialec monog aphs. Finally, he i h sec ion (A cheoling) b ie ly accoun s o an ongoing p oj- ec which in ends o ob ain eal da a om he ime when he a ie y eme ged, based on public o p i a e spoken documen s. THE INTERMEDIATE VARIETY The ongoing o ma ion o an in e media e u ban middle-class a ie y be- ween he Spanish na ional s anda d (Mad id) and he sou he n a ie ies in Andalusia has been unde s ood as he ou come o he end o le elling and con e gence owa ds he cu en p es ige no m o speech used in Spain by middle-class speake s, whose adi ional sou he n iden i y has been changing since he mid- i ies and ge ing close o mode n iden i ies o globalisa ion, S aipsniai / A icles 13 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ mode nisa ion and u banisa ion (see Villena-Ponsoda 2008; He nández-Cam- poy, Ville na-Ponsoda 2009; Regan 2017a, b; Villena-Ponsoda, Vida-Cas o 2018). This pa icula end expe ienced a sha p ise a e he e a o F anco (1939–1975), when ansi ion om a mos ly u al, adi ional and local ideolo- gy o mode nisa ion ook o e (Die z 2001). The linguis ically mos in e es ing poin he e is ha his a ie y in ol es a syn hesis – o a ce ain ex en , con adic o y – o social meanings o (1) mo d- e nisa ion (i.e. s anda d o ien a ion) and (2) adi ion (i.e. adi ional sou he n iden i y). To do so, a combina ion is made o na ional s anda d ea u es in syllable onse posi ion wi h sou he n e nacula ea u es in he coda. This combina ion is mean o le speake s symbolically lee ei he he u al o u ban wo king-class backg ound – i.e. in doing so, he in e media e- a ie y speake s in end o aban- don speech ea u es ela ed o social meanings o he ‘ adi ional’, he ‘oldes ’, he ‘ u al’, e c. – and, mo eo e , consciously accep s anda d ea u es, e en hough hey can en ail s opping and e en e e ing mul isecula changes and/o acqui - ing new phonological con as s. A he same ime, hey p ese e a g ea deal o he sou he n accen – he so-called accen o andalú (Andalusian accen ), which is pa icula ly iden i ied wi h unma ked e osi e a ia ion p ocesses o he leni ion and dele ion o codas. Thus, as a gued, social meanings o s anda disa ion and mode nisa ion, on he one hand, and hose ep esen ing loyal y o he sou he n communi y alues, on he o he , can be, o a ce ain ex en , compa ible. MAP 1. Mosaic o con empo a y Peninsula Spanish dialec s and languages (He nán- dez-Campoy, Villena-Ponsoda 2009: 184). JUAN-ANDRÉS VILLENA-PONSODA 14 Ac a Linguis ica Li huanica LXXIX On i s way owa ds he na ional s anda d a ian s, he Andalusian u ban middle-class a ie y speake s mee dialec use s om sou he n Cas ile, Mu cia and Ex emadu a (see Map 1). These dialec s sha e a good deal o ea u es wi h he Andalusian a ie ies so ha a con inuum has adi ionally been p oposed o ep esen he geolinguis ic a ia ion a s ake he e (Na a o-Tomás, Espinosa, Rod íguez-Cas ellano 1933). The e o e, he idea o an in e media e sup a- e- gional a ie y would apply no only o he sociolinguis ic dimension, bu also o he geolinguis ic space (see Villena-Ponsoda, Vida-Cas o 2017). Bo h he linguis ic con e gence and he unde lying new blended iden i y in- ol e he acquisi ion, among o he s, o some s anda d ea u es, no ably spli ing o co onal ica i e /θ/. This change implies econs uc ion o he Middle-Age Cas ilian con as – which me ged in he ea ly Andalusian dialec s since he 13 h cen u y – be ween wo d se s wi h apical /s/, (casa ‘house’) and laminal /s/, p e iously a ica e / s/ (caça ‘hun ing’), which has been main ained in he s anda d a ie y h ough con as be ween apical /s/ [ˈkasa] and (in e ) den al /θ/ [ˈkaθa] (see Table 1 below): TABLE 1. In en o y o obs uen consonan s in he Andalusian e nacula (le ) and in he in e media e a ie y ( igh ) labial den al pala al ela labial den al pala al ela p ʧ k ense p ʧ k ense b d ʝ g lax b d ʝ g lax θ h ica i e s x ica i e ense θlax S aipsniai / A icles 15 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ The spli ing o Andalusian /θ/ implies econs uc ion o he abo e-men- ioned con as be ween Middle-Age wo d se s wi h /s/ and wi h / s/, hen /s/, by imi a ion o he s anda d. F om wha we know now, he deme ge o /θ/ mus ha e been a ha d indi idual p ocess in ol ing he building-up o an acous ic mul i-pa ame ic dis ance be ween bo h ica i es (Regan 2017b; Mo- lina-Ga cía 2018). This phonemic con as is cu en ly used by young educa ed u ban speake s. Sc u iny o da a om linguis ic a lases, no ably he ALPI o A las de la Penín- sula Ibé ica ‘A las o he Ibe ian Peninsula’ ( i s qua e o he 20 h cen u y) and he ALEA o A las Lingüís ico y E nog á ico de Andalucía ‘Linguis ic and E h- nog aphic A las o Andalusia’ (ea ly i ies), as well as he esul s om dialec monog aphs desc ibing some u ban dialec s (no ably Al a 1974, 1993 [1973]), sugges s ha /θ/ spli ing has been a quick change om abo e. This idea is suppo ed by he ac ha e y ew – i any – aces o i s exis ence ha e been de ec ed up o he eigh ies (see Ca bone o 1982; Sawo 1980; Sal ado -Sal a- do 1980; Villena-Ponsoda 2001). The change has sp ead apidly o e he u ban cen es o Andalusia, pa icula ly in he eas e n pa o he a ea, and a ec ed he lowes s a us g oups only pa ially (Villena, Á ila-Muñoz 2014; see Map 2). MAP 2. U ban Andalusia A se o o he pa allel ea u es showing he same ajec o y would ha e like- ly soa ed by he same ime. These ea u es ha e been ac ually analysed so ha JUAN-ANDRÉS VILLENA-PONSODA 16 Ac a Linguis ica Li huanica LXXIX esea che s a e acquain ed wi h he idea o he eme gence o an in e medi- a e – ela i ely cohe en – a ie y be ween no he n s anda d-close dialec s and sou he n s anda d-dis an dialec s (see Villena-Ponsoda, Vida-Cas o 2018). In sho , i could be said ha , in addi ion o he deme ge o /θ/ – which is, o cou se, he socially mos salien change among he se – he in e media e a i- e y adop s ei he s anda d o hyb idised a ian s able o a oid e nacula uses (T udgill 1986; Chambe s, T udgill 2004 [1980]; Siegel 2010 [2006]; Be u o 2005; Almeida 2018; see Table 2). TABLE 2. Phonological changes. Me ge s and deme ge s (adap ed om Villena-Ponso- da, Vida-Cas o 2018) Va ie ies Wo d se s: <caza> <casa> <caja> <cacha> <cada> S anda d ˈkaθa ˈkasa ˈkaxa ˈkaʧa ˈkaða In e media e ˈkaθa ˈkasa ˈkaxa∼ˈkaha ˈkaʧa ˈkaða∼ˈka Ve nacula /θ/ = /s/, seseo ˈkasa ˈkasa ˈkaxa∼ˈkaha ˈkaʧa∼ˈkaʃa ˈkaða∼ˈka /θ/ = /s/ ceceo ˈkaθa ˈkaθa ˈkaha∼ˈka·a ˈkaʃa∼ˈkaʧa ˈkaða∼ˈka /θ/ = /s/ = /x/ = /ð/, heheo ˈkaha ˈkaha ˈkaha∼ˈka·a ˈkaʃa ˈka /θ/ = /s/ = /ʧ/, ex ended seseo ˈkasa ˈkasa ˈkaha∼ˈka·a ˈkasa ˈkaða∼ˈka Gloss hun ing house box bu ock each As shown in Table 2, on he one hand, he new a ie y s ops e nacula p ocesses, including ei he equen and ela i ely accep ed me ge s o co onal ica i es (seseo, ceceo) o e en hei mos e ol ed ou comes (heheo, ex ended seseo); on he o he hand, i adop s phonologically-ma ked mains eam-p es- igious a ian s (a ica e [ʧ], ica i e [x] and app oximan [ð], ins ead o hei e nacula unma ked coun e pa s ( ica i e [ʃ], aspi a ed [h] o dele ed /x/ and dele ed /d/). THE SOCIAL MEANING OF THE NEW VARIETY How his new a ie y has sp ead ou and o wha ex en we can alk abou a cohe en a ie y is a ele an ma e o discussion. Howe e , a c ucial issue o S aipsniai / A icles 17 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ be poin ed o is he social p o ile o i s speake s. Resul s o p e ious esea ch end o indica e ha i is u ban young – mos ly emale – uni e si y-g adua e speake s who lead he use o he se o ea u es shaping he in e media e a i- e y, which seems o sp ead bo h socially and geog aphically (Villena-Ponsoda 2001; Moya-Co al, Sosinski 2015; Regan 2017b). Unexpec edly, he con em- po a y image o he Andalusian speech communi ies shows g owing di e gence be ween wo se s o a ie ies; namely, on he one hand, adi ional u ban wo k- ing-class and u al e nacula dialec s, which a e ela i ely ou o he in luence o he cu en s anda d- end o con e gence and, on he o he hand, he new in e media e middle-class a ie y (Villena-Ponsoda, Á ila-Muñoz 2014). Fea u es being pa o his a ie y co ela e in such a way ha hei use seems o be a ec ed by cohe ence es ic ions. Va iables ela ed o he s anda d alues and a i udes end o occu in con ex s and be used by speake s who a e unlike- ly o p omo e a iables ep esen ing e nacula social meanings, and he oppo- si e is also ue (see Table 2 abo e). Then, co-occu ence o he single a ia- bles o m clus e s whe e we can iden i y mo e complex unde lying a iables o componen s, able o gi e us an idea o hei common dis ibu ion, which is in ac compa ible wi h he no ion o a cohe en a ie y (see Guy, Hinskens 2016). Gi en ou pu pose o explo ing he exis ence o a cen al-sou he n (s an- da d- e nacula ) in e media e a ie y and since ou objec i e is o p o e i s co- he ence, a good way o p oceeding could be o iden i y clus e s o a iables on he basis o hei simila a iance. Doing so, we would be able o seize which a iables a e likely o co-occu and he kind o speake s who would use hem – i.e. which ea u es shape a pa icula a ie y and hei social dis ibu ion. To ind ou which ea u es ga he acco ding o hei espec i e a iance, a ac o analysis was ca ied ou .3 Based on mul iple co ela ion, wo p incipal compo- nen s o unde lying clus e s o a iables we e ound, which a e ac ually able o explain c oss-dependence be ween he indi idual a iables. In o he wo ds, hese complex unde lying a iables, componen s o ac o s would ep esen he common basis o he single a iables. The o e all a ia ion o he da a was educed hen o wo dimensions o ac- o s ha explain 69.5% o he a iance (39.1 and 30.4, espec i ely). A P incipal Componen s Sa u a ion Ma ix is shown in Table 3, whe e he p opo ion o a iance o each a iable explained, espec i ely, by ac o 1 and 2 is displayed. Figu e 1 below shows he ela i e posi ion o each single a iable in he ideal space delimi ed by bo h ac o s gi ing us an idea o how hey co ela e. 3 In he ame o a b oade esea ch p ojec , we wo ked wi h a sample o 72 speake s o bo h gende s (m = 36, = 36), s a i ied by educa ion and age. Fieldwo k o his s udy was ca ied ou du ing he i s decade o he 21s cen u y. See Villena-Ponsoda 2008. JUAN-ANDRÉS VILLENA-PONSODA 24 Ac a Linguis ica Li huanica LXXIX The unde lying pa h analysis e eals hen an image o complex in e ac ions be ween a iables (see Figu e 5 below). On he one hand, he o e all nega i e e ec o educa ion (x1) (Be a = −.562) is explained by bo h i s di ec in lu- ence (Be a = −.293) and indi ec ly h ough he in e ening a iable x5, Media/ S anda d o ien a ion, Be a = −.210): he highe he speake s’ educa ion and he highe hei sco es on Media/S anda d o ien a ion (Be a = .765), he lowe hei posi ion on he F1 dimension (Be a = −.276). On he o he hand, educa ion ac s also as in e media o o he indi ec in luence o he a iables ep e sen ing he con ex o linguis ic acquisi ion (x2, Be a= −.266 and x3, Be a = −.233) and he speake ’s age (x4, Be a = −.139). Di ec e ec s o hese a iables on he speake ’s educa ion a e posi i e bu indi ec ly, h ough educa ion as an in e - ening a iable, hey nega i ely a ec F1 o jus minimise hei o iginal posi i e in luence. Finally, he speake ’s ne wo k has no only a di ec posi i e e ec on F1 (Be a = .272), bu also almos cancels he e ec o educa ion (−.059). All e ec s a e signi ican (p < .05). Gende (x6) is a dummy a iable, whe e 0 = male, 1 = emale x1 Speake ’s educa ion x2 Mo he ’s educa ion x3 Fa he ’s educa ion x4 Age x5 Media/S anda d o ien a ion x6 Gende x7 Ne wo k s eng h x8 Speech use FIGURE 5. Diag am o he mul i-causal p ocess o he e ec s o he speake ’s a iables on speech use S aipsniai / A icles 25 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ Resul s o he pa h analysis a e hen compa ible wi h he idea o he eme - gence o an in e media e a ie y be ween he na ional s anda d and he Andalu- sian e nacula dialec s. Speake s who cu en ly use speech ea u es cha ac e - ising his a ie y ha e low sco es on he unde lying a iable F1 ha ep esen s e nacula indexicali ies. As easily deduced om esul s displayed in Table 4 and Figu e 5, his new u ban a ie y is spoken mos ly by young educa ed and well-in- o med s anda d-o ien ed emale speake s wi h loose-kni pe sonal ne wo ks. THE DILEMMA Bea ing in mind such a complex p ocess o sociolinguis ic a ia ion, he e- cons uc ion o he pas o he in e media e a ie y seems o be an exci ing chal- lenge. Howe e , he cen al issue now is whe he he a ailable da a o he pas we ha e access o a e comple e and sa is ac o y enough o answe he main ques- ions aised in he cou se o he esea ch on he p esen . This is because he da a o he pas we e ob ained using he me hods o sampling, ieldwo k and analysis which a e no always compa ible wi h hose used o deal wi h he complexi y o con empo a y speech communi y da a. In spi e o he eno mous igou and gene osi y dis illed om hei wo k, he e a e p o ound unde lying ideological di e ences be ween he esul s ob ained by dialec ologis s and geolinguis s and hose achie ed in he ame o he p esen -day sociolinguis ic pa adigm. On he one hand, basically, in he ield o bo h dialec geog aphy (di ec ep- esen a ion o local a ia ion using ocabula ies and maps, om Wenke o Gil- lié on h ough o he con empo a y p ojec s) and dialec monog aph ( econs uc - ing he p o o ypical local dialec , om Ascoli and Wegene h ough o mode n sociolinguis ics), dialec ologis s ha e ca ied ou an essen ial and indispensable wo k. Howe e , hei aim was o show ha laws o linguis ic change a e excep- ionless and ha le pa ois au hen ique o a local speech communi y is he pu es o m o a language, wo h p ese ing om ex e nal con amina ion and s udying as a ese oi o o iginal pu i y (see Villena-Ponsoda 2010: 614–615 e passim). On he o he hand, da a o he p esen ha e o do wi h a e y di e en ou - look, which e lec s a d ama ic and eal change in socie y. In ac , adop ing Du kheim’s (1902 [1893]) concep ual con as be ween mechanical and o ganic solida i y, he ansi ion om adi ional o mode n indus ial communi ies can be connec ed wi h he e olu ion o me hods. Du kheimian mechanical solida - i y suppo ing adi ional u al closely kni communi ies, and e lec ing uni- o mi y o he speake ’s é a s de conscience, was eplaced by he con empo a y u ban sha ply s a i ied loose-kni communi ies based on o ganic solida i y and di ision o labou (Be hele 2004: 732–734). The mode n socie y p esen s an JUAN-ANDRÉS VILLENA-PONSODA 26 Ac a Linguis ica Li huanica LXXIX image o he e ogenei y which only ex alinguis ic co ela es (occupa ion, edu- ca ion, age, gende , e c.) could help o o ganise. The e o e, ideological di e - ences be ween bo h pa adigms became i s ly heo e ical and, e en ually, me h- odological. Dialec ology and sociolinguis ics inished up as sepa a ed om each o he as he communi y ypes hey we e mean o desc ibe. Wi hin his amewo k, wo issues should be conside ed when compa ing – as i is he case he e – con empo a y and his o ical da a o such a di e en ype: (1) The speech communi y. Wha do we know abou he speech communi y o he pas ? (2) The ac ual linguis ic da a. How was he linguis ic s uc u e we know o- day o ganised in he pas ? The speech communi y dilemma. Da a om he Malaga u ban speech communi y ega ding he key ime-pe iod we a e conce ned wi h he e a e a ailable only om he A las Lingüís ico y E nog á ico de Andalucía, ALEA (1961–1973). Fieldwo k o his a las was ca ied ou du ing he 50’s (see Al a 1955; Ga cía-Mou on 1992; Mo illo-Vela de 2001) using “mul iple and epea - ed ques ionnai es” and “op imal subjec s” (Al a 1959: 10, 27–30). Since he idea was – ollowing Gillié on, Pop, Pusca iu and o he s – o ob ain […] “ he snapsho o he speech in a pa icula momen ”, […] we sough he pe son able o be he subjec o his “linguis ic pho og aph”, a oiding s idency – no he eldes , he mos pe sonal o he newes da a – jus he a e age, al hough some imes i may be necessa y o ha e ecou se o he old- es people o pick up a declining phenomenon o a disappea ing piece o c a - wo k. Tha is he eason why a man be ween i y and six y yea s old will p o- jec e y likely he p ecise image o wha we a e looking o (Al a 1959: 29, my ansla ion). To design samples o ci ies and some o he bigges owns, local a eas and he in o man ’s age, gende and social class we e conside ed as a sou ce o a - iabili y, ollowing Jabe g and Jud (1928–1960) p ocedu es o he AIS (Jabe g 1936: 20; see also Jabe g 1959). Howe e , a he end o he day, and in spi e o he bes in en ions, only ou subjec s we e in es iga ed in Malaga ci y (mo e han 270 000 inhabi an s by hen).4 In ac , al hough his pa icula a las con- ains a good deal o in o ma ion abou he sociolinguis ic a ia ion o he ime – including he so-called “polimo phism” o unconscious a ia ion – he ask 4 Th ee men, one o hem de ined as a uni e si y g adua e and he o he wo as illi e a e, and one uni e si y g adua e woman; no pa icula de ails o hei age a e p o ided. S aipsniai / A icles 27 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ o loca ing and ga he ing his in o ma ion om each map ou he six olumes o he a las is exhaus ing, and di ec da a on sociolinguis ic di e ences – ak- en om pa allel ques ionnai es wi h di e en subjec s – a e no always clea ly sepa a ed om obse a ions by he esea che based on he in o man ’s linguis- ic awa eness (see Mo illo-Vela de 2001: 15–21 e passim). As a ma e o ac , a sys ema ic sc u iny o pa icula sociolinguis ic co ela ions be ween speech a ian s and speake a iables o a ce ain u ban locali y wi hin he a las e eals sca ci y o occu ences and, hence, lack o i mness o conclusions. None heless, he main ouble is he exis ence o wo di e en ways o p e- sen ing dialec da a. On he one hand, we ha e he geolinguis ic da a ha a e accessible o each locali y and each phonological, mo phological, lexical o syn ac ic a iable. This da a we e p o ided by one pa icula NORM in o man o each poin in he map – wi h supplemen a y da a by a woman o a ce ain numbe o local- i ies and a iables – and, as men ioned abo e, wi h a ia ion a leas o gen- de and educa ion, in he case o ci ies and he bigges owns (see Al a 1959: 29–30). O cou se, his so o da a a e di ec and undoub edly a ec ed by he skewness linked o he NORM speake sampling me hod. On he o he hand, he dialec ological da a, which consis o a se o addi- ional obse a ions on he social meanings ela ed o he speake ’s s a us, age o gende . These obse a ions a e included bo h as supplemen a y no es wi hin he maps hemsel es o by he use o symbols wi h glosses on he maps’ ma - gins. Based appa en ly bo h on hese wo sou ces o da a and on addi ional e- co dings and quali a i e obse a ions, dialec monog aphs we e accomplished and hen dialec monog aph da a a e also a ailable (Al a 1993 [1973], 1974). This ype o da a is in luenced hen no only by he same ideological and me h- odological bias as he geolinguis ic da a, bu also by he di ec e ec o quali a- i e judgmen s by he esea che s hemsel es. The dilemma o he linguis ic da a. Based bo h on he o e all ALEA in- o man s’ p o ile (men be ween 50 and 60 yea s old) and on he in o ma ion ega ding he speake s s udied in Malaga ci y (4 speake s, 1 uni e si y-educa - ed woman and 3 men, one o hem uni e si y-educa ed and he o he wo illi - e a e), i is likely o be asce ained ha , in acco dance wi h he appa en - ime dis ibu ion commen ed on p eceden pa ag aphs, among hese speake s – who would be 85–90 yea s old by he ime when he ieldwo k o he Malaga in- e media e a ie y s udy was ca ied ou – he illi e a e male in o man s would no use he s anda d-like a ian s, while he uni e si y g adua e speake s (man and woman) pa ially would. (De ails on hese speake s a e shown in Table 5). JUAN-ANDRÉS VILLENA-PONSODA 28 Ac a Linguis ica Li huanica LXXIX TABLE 5. ALEA in o man s in Malaga ci y (sou ce: adap ed om Al a (1993 [1973]) In o man 45 50 583 506 Local a ea Ci y cen e Ci y cen e Semi-u ban Semi-u ban Locali y Malaga Malaga Chu iana Izna e Gende Male Female Male Male Educa ion Uni e si y Uni e si y Illi e a e Illi e a e The Uni e si y-educa ed in o man s (45 and 50) p o ided esponses only o he phone ic ques ionnai e, while he illi e a e in o man s (583 and 506) esponded o all he i ems o e e y pa o he whole ques ionnai e. A “linguis ic ou line” consis ing o ou o i e pages – which was included in he phone ic pa o he ques ionnai e – as well as a comple e page o phone ic an- sc ip ion om each in o man ’s speech, we e p epa ed including also eco dings and pala og ams. This ma e ial is no a ailable. The image o he speech communi y based on he da a p o ided by hese speake s is, o cou se, e y limi ed bu one would expec ha he geolinguis ic and dialec ological in o ma ion gi en was close enough – and no con adic- o y – o allow he econs uc ion o he his o ical ajec o y ollowed by he changes leading o he in e media e a ie y. Howe e , bo h sou ces o da a do no ma ch. O cou se, as poin ed ou abo e, i would be expec ed – gi en he esul s ob ained in he con empo a y Malaga s udy, based on da a collec ed by he ea ly yea s o he 21s cen u y – ha we had o ace e idence, on he one hand, o a s ongly-ma ked e nacula a ie y used by he illi e a e speake s and, on he o he hand, o a less ma ked and mo e s anda d-like bu clea ly sou he n-accen ed a ie y used by he uni e si y-educa ed speake s. Howe e , al hough his is p ecisely he image ha eme ges when we conside he dialec o- logical da a, his is no so clea when he geolinguis ic da a a e aken in o accoun . When we ocus on he mos ele an ea u es o he in e media e a ie y (see Table 2 abo e), he dialec ological in o ma ion p o ided by Al a (1993 [1973]) – which is he unique dialec monog aph dealing wi h he u ban dialec o Mala- ga o he i s hal o he 20 h cen u y – con i ms he cu en hypo heses abou he se o quick changes enhanced by con e gence owa ds he na ional s and- a d and suppo ing he eme gence o he new u ban in e media e middle-class a ie y. Al hough all he e nacula ea u es men ioned by Al a a e ela ed o he illi e a e speake s, wi h he excep ion o me ge o /ʎ/ = /ʝ/ ( he so-called yeísmo), which is accep ed as no mal by uni e si y g adua es, his desc ip ion does no sugges ha he changes leading o he p esen -day si ua ion we e in p og ess a he ime (Table 6). In ac , he mos accep ed opinion among he schola s by he mid- i ies was ha he Andalusian dialec seemed o be socially homogeneous: S aipsniai / A icles 29 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ Rega ding i s social di usion, he Andalusian a ie y does no ha e any limi s: i eaches he highes g oups in socie y. And when I call hem ‘ he highes ’, I mean he cul u ed g oups, which a e he mos likely ei he o le el o o esis di usion. P onuncia ion by a li e a e Andalusian speake is clea ly simila o ha o his o he illi e a e ellow ci izens. Lexis and p ocedu es o exp ession may change bu p onuncia ion is he same. In ac , o s udy he phone ics om G anada, uni- e si y-g adua e speake s we e in es iga ed, e en uni e si y p o esso s (Sal ado 1967 [1963]: 64, my ansla ion). On he o he hand, an exhaus i e sc u iny o he geolinguis ic da a om he ALEA e eals a a he di e en image. The in o ma ion de i ed om he A las and he one p o ided as a esul o quali a i e analyses by dialec ologis s end o con as . The geolinguis ic image om he ALEA is compa ible wi h a scena io whe e mos o he changes ha ac ually shape he in e media e a ie y we e in p og ess al eady by he 50’s, pa icula ly among he u ban middle-class speake s. Table 6 displays a compa ison be ween bo h sou ces o in o ma ion. Fea u es as ele an o he cha ac e isa ion o he in e media e a ie y as he con as be ween /s/ and /θ/ help us unde s and bo h posi ions. Al a (1993 [1973]) poin s ou ha he mos equen pa e n o ealisa ion o me ged /θ/ among li e a e speake s is sibilan ica i e [s] (seseo), while illi e a e speake s end o use non-sibilan ica i e [θ] (ceceo). Howe e , esul s om sc u iny o he ALEA maps show ha a ia ion seems o ha e s a ed by he ime o ieldwo k o he a las, including some aces o he spli ing o /θ/ (see Table 7). TABLE 6. Educa ional di e ences in he use o he Malaga e nacula ea u es (adap ed om Al a (1993 [1973]) and ALEA) Dialec Monog aph da a Geolinguis ic da a S anda d Illi e a e Uni e si y Illi e a e Uni e si y /s/ /θ/ poso/pozo Ceceo[ˈpoθo] Seseo[ˈposo] Ceceo/Seseo/Spli Seseo/Spli /ʧ/ pocho [ˈpoʧo] F ica i e [ˈpoʃo] A ica e [ˈpoʧo] A ica e A ica e -0d- dedo Dele ion [ˈdeo] Main . [ˈdeðo] Va ia ion Main enance -#d- dado Dele ion [ˈdao] Dele ion [ˈdao] Dele ion Va ia ion The same can be applied o he es o ea u es in Table 6. The compa ed dialec ological and geolinguis ic esul s speak o hemsel es (see also Tables 7–9). The appa en homogenei y o he speech beha iou p ojec ed by he di- alec monog aph analysis is o a ce ain ex en con adic ed by he esul s ob- ained a e a ho ough s udy o he maps, including quan i a i e accoun o he esponses. JUAN-ANDRÉS VILLENA-PONSODA 30 Ac a Linguis ica Li huanica LXXIX TABLE 7. Educa ional di e ences in he ealisa ion o wo d se s wi h <s> and <z, ce, ci> in Malaga (sou ce: adap ed om he ALEA) Uni e si y Illi e a e Male Female Male Male E im. [s] [θ] To al [s] [θ] To al [s] [θ] To al [s] [θ] To al s13 114 14 014 9258 267 204 38 242 92.9 7.1 100.0 0.0 3.4 96.6 84.3 15.7 z, ce/ci 526 31 35 136 0316 316 149 115 264 16.1 83.9 97.2 2.7 0.0 100.0 56.4 43.6 To al 18 27 45 49 1 50 9574 583 353 153 506 TABLE 8. Educa ional di e ences in he ealisa ion o in e ocalic /d/ in Malaga (sou ce: adap ed om he ALEA) Uni e si y Illi e a e Male Female Male Male /0d/ [d] ∅To al [d] ∅To al [d] ∅ N [d] ∅To al 8 2 10 8 2 10 32 69 101 43 54 97 80.0 20.0 80.0 20.0 31.7 68.3 44.3 55.7 /#d/ 257 4 3 7 3 106 109 192 93 28.6 71.4 57.1 42.9 2.8 97.2 1.1 98.9 10 717 12 517 35 175 210 44 146 190 58.8 41.2 70.6 29.4 16.7 83.3 23.2 76.8 TABLE 9. Educa ional di e ences in he ealisa ion o /ʧ/ in Malaga (sou ce: adap ed om he ALEA) Li e a e Illi e a e Male Female Male Male [ʧ] 16 76.2 23 100.0 136 97.8 116 87.2 [ʃ] 52.8 00.0 32.2 17 12.8 To al 21 23 139 133 S aipsniai / A icles 31 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ Be o e his con as o esul s be ween dialec monog aphs and geolinguis- ic wo k, key ques ions abou he eme gence and de elopmen o he Andalu- sian in e media e a ie y emain unanswe ed. We eally do no know whe he mos o he phonological a iables which cu en ly shape his a ie y we e ab- sen om he speake s’ epe oi e by he mid wen ie h cen u y and suddenly and unexpec edly a ose boos ed by ce ain social and cul u al changes o whe - he hey had been de eloping – maybe since he old imes – and, hence, hey we e al eady obse able in ha same pe iod as p ocesses o socially condi ioned a ia ion. In he i s case, we would be acing an example o a quick change om abo e, pe ec ly analysable using da a om spoken co pus since he las decade o he 20 h cen u y. C oss-sec ional appa en - ime analyses would hen be app op i- a e and hei esul s would p obably ma ch wi h da a om dialec monog aphs. In he second case, we would need mo e de ailed e-analyses o he geo- linguis ic da a and sys ema ic compa ison wi h cu en sociolinguis ic esul s. Longi udinal end s udies based on, le ’s say, a chaeological da a would hen be desi able. Al hough a discussion abou he exac na u e o he la e so o da a would be he subjec o a whole pape , a b ie ou line is summa ised in he nex pa ag aph. ARCHEOLING In addi ion o he well-known dispa i y o me hods and heo e ical assump- ions be ween dialec ology/geolinguis ics and sociolinguis ics – no o men ion again he unde lying ideological pe spec i es in ol ed – ou e ospec i e a - emp o iden i y clea aces o li ing sound changes in dialec monog aphs and linguis ic maps has come ac oss a supplemen a y sou ce o ouble. Subjec i e assessmen s o dialec use mee he NORM speake bias and make compa ison o p esen -day da a wi h hose de i ed om dialec ological and geolinguis ic wo k a di icul ask. Con empo a y wo k on hese adi ional disciplines has comple ely changed he si ua ion. Ou s anding esul s in bo h ields will make hings easie in he u u e. Ne e heless, now we ha e o cope wi h he Anda- lusian pano ama as i is. Re-analysis o geolinguis ic and dialec esea ch da a, as well as gaining ac- cess o b oadcas adio and ele ision eco dings and documen a y ilms a - chi es, would allow us o (1) sc u inise in eal ( hough se e ely biased) ime he expansion ac oss socie y o he cu en ly ongoing changes and (2) explo e also he ideological lag a ec ing his kind o e ospec i e ajec o ies. Assessmen o selec ion c i e ia unde lying NORM speake s in geolinguis ics and adi ional JUAN-ANDRÉS VILLENA-PONSODA 32 Ac a Linguis ica Li huanica LXXIX dialec ology (i.e. he idea o a ‘genuine’ o ‘pu e’ dialec ) is an in e es ing way o acing his pa icula issue. A cheoling is an ongoing esea ch p ojec which in ends o ha e access o supplemen a y sou ces o spoken and w i en da a allowing he possibili y o a compa ison.5 Longi udinal end s udies on adio a chi e sou ces (see Van de Velde, an Hou , Ge i sen 1996, 1997; He nández-Campoy 2018), as well as complemen a y w i en public and p i a e documen s (He nández-Campoy, Conde-Sil es e 2012), will p o ide compa a i e e idence o language change and he his o y o he o ma ion o he Andalusian in e media e a ie y. REFERENCES ALEA – A las Lingüís ico y E nog á ico de Andalucía, 1–6 ols, 1961–1973, ed. by Al a Manuel, A. Llo en e, G. Sal ado . Mad id: CSIC. Almeida Manuel 2018: Language hyb idism: On he o igin o in e dialec al o ms. – Language Va ia ion. Eu opean Pe spec i es VII, ed. by Juan-And és Villena-Ponsoda, F ancisco Díaz Mon esinos, An onio Á ila-Muñoz, Ma ilde Vida-Cas o. Ams e dam: SILV, John-Benjamins, o hcoming. ALPI – A las Lingüís ico de la Península Ibé ica, 2016, ed. by Ga cía Mou on Pila (co- o d.), Inés Fe nández-O dóñez, Da id Heap, Ma ia Pila Pe ea, João Sa amago, Xulio Sousa, ALPI-CSIC: h p://www.alpi.csic.es (22-10-2018). Al a Manuel 1955: Las encues as del “A las Lingüís ico de Andalucía (diciemb e 1953 – ma zo 1955)”. G anada: Uni e sidad: h ps://g upo.us.es/ehandalucia/pd /lec- u as/a las_linguis ico_ andalucia.pd (26-10-2018). Al a Manuel 1959: El A las Lingüís ico-E nog á ico de Andalucía. G anada: Uni e sidad. Al a Manuel 1974: Se illa, mac ocosmos lingüís ico. – Es udios Filológicos y Lingüís icos. Homenaje a Ángel Rosenbla , ed. by Qui oga-To ealba Luis, Ma io To e- alba Lossi, Ped o Díaz-Seijas. Ca acas: Ins i u o Pedagógico, 13–42. Al a Manuel 1993 [1973]: No as de asedio al habla de Málaga. – Pue a Nue a. Re- is a de Educación, 6–13: h ps://g upo.us.es/ehandalucia/pd /lec u as/a las_linguis- ico_andalucia.pd (26-10-2018). 5 In he ame o he abo e men ioned (on oo no e 1) P ojec on he Sociolinguis ic Pa e ns o Cas ilian Spanish (ECOPASOS, FFI2015-68171-C5-1 and FEDER unds). S aipsniai / A icles 33 The Dilemma o he Reliabili y o Geolinguis ic and Dialec ological Da a o Sociolinguis ic Resea ch. The Case o he Andalusian Deme ge o /θ/ Aue Pe e 2005: Eu ope’s Sociolinguis ic Uni y, O , a Typology o Eu opean Dia- lec /S anda d Cons ella ions. – Pe spec i es on Va ia ion, ed. by Nicole Delbecque, Jo- han an de Auwe aand, Di k Gee ae s. Be lin: Mou on de G uy e , 7–42. Be u o Gae ano 2005: Dialec /s anda d con e gence, mixing, and models o lan- guage con ac . The case o I aly. – Dialec Change. Con e gence and Di e gence in Eu- opean Languages, ed. by Pe e Aue , F ans Hinskens, Paul Ke swill, Camb idge: Cam- b idge Uni e si y P ess, 81–95. Be hele Raphael 2004: Dialek soziologie – Soziolinguis ische Aspek e de Dial- ek ologie. – Sociolinguis ics: An In e na ional handbook o he Science o Language and Socie y 1. Second edi ion, ed. by Ul ich Ammon, No be Di ma , Klaus J. Ma heie , Pe e T udgill. Be lin/New Yo k: Mou on de G uy e , 721–738. Bou dieu Pie e 1977: L´Économie des échangeslinguis iques. – Langue F ançaise 34, 17–34. Bou dieu Pie e 1984: Capi al e ma ché linguis iques. – Linguis ische Be ich e 90, 3–24. Bou dieu Pie e 1986: The o ms o capi al. – Handbook o heo y and esea ch o he sociology o educa ion, ed. by J. Richa dson. New Yo k: G eenwood, 241–258. Ca bone o Ped o 1982: El habla de Se illa. Se illa: Ayun amien o. Chambe s J. K., Pe e T udgill 2004 [1980]: Dialec ology. Camb idge: Cam- b idge Uni e si y P ess. Die z Gun he 2001: F on ie hyb idiza ion o cul u e clash? T ans-na ional mi- g an communi ies and sub-na ional iden i y poli ics in Andalusia, Spain. – Jou nal o E hnic and Mig a ion S udies 30(6), 1087–1112. Du kheim Émile 1902 [1893]: De la di ision du a ail social. É ude su l’o ganisa- ion des socié éssupé ieu es. Second edi ion. Pa is: Alcan. Ga cía-Mou on Pila 1992: El A las Lingüís ico y E nog á ico de Andalucía. Hom- b es y muje es. Campo y ciudad. – Ac as del Cong eso In e nacional de Dialec ología, IK- ER 7, 667–685: h p://digi al.csic.es/handle/10261/13687?locale=es (22-10-2018). Guy G ego y, F ans Hinskens 2016: Linguis ic cohe ence; sys ems. epe oi es and speech communi ies. In oduc ion. – Linguis ic cohe ence: Sys ems, epe oi es and speech communi ies. Lingua 172–173, 1–9. He nández-Campoy, Juan-Manuel 2018: Vi ual sociolinguis ics: F om e- al- ime su eying o i ual- ime a chi al sou ces o acing change longi udinally. – Language Va ia ion. Eu opean Pe spec i es VII, ed. by Juan-And és Villena-Ponsoda, F ancisco Díaz Mon esinos, An onio Á ila-Muñoz, Ma ilde Vida-Cas o. Ams e dam: SILV, John-Benjamins, o hcoming.