J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ...  Joseph Schallert and Marc L. Greenberg University of Toronto and University of Kansas The Prehistory and Areal Distribution of Slavic *gъlčěti ‘Speak’ Slovenski jezik – Slovene Linguistic Studies 6 (2007): –76 Razprava podrobno obravnava zemljepisno razvrstitev in pomenski razvoj leksema *gъlčěti v južnoslovanskih narečjih, tj. na področjih z migracijsko poselitvijo. Zastavlja tudi vprašanje o prvotni povezavi med jezikovnimi skupnostmi, v katerih se je leksem razvil v glavni izraz za pojem ‘govoriti’, in sicer v južnoslovanskem pa tudi v širšem slovanskem kontekstu. Posebna pozornost se posveča bolgarščini in vzhodni slovenščini, ki izpričujeta ta razvoj. The paper examines in some detail the diatopic distribution and semantic development of *gъlčěti in South Slavic, i.e., the Slavic dialect areas settled by migration, and raises the question of the nature of the relationship among those dialects that have developed *gъlčěti as the primary neutral verb meaning ‘speak’ both in its South Slavic and broader Slavic contexts. Special attention is given to Bulgarian and Eastern Slovene dialects, which share this development. 0. Introduction The student of the Slavic languages whose purview extends only to the standard languages and canonical Old Church Slavic might suspect that a small number of lex- emes expresses the notion ‘speak’. S/he might presume that the putatively original Proto- Slavic word, attested by OCS glagolati, was, at an early stage following the beginning of literacy, replaced by a relatively small number of lexemes, e.g., BCS govoriti, Be гаварыць, Bg говоря, Ru говорить, Sk hovorit’, Sn govoriti; Cz mluvit, Po mówić, Uk мовити; Ma зборува; LS rjac, US rěčeć. This ostensible simplicity disappears when one confronts the dialect variation, which reveals a highly variegated and dynamic picture. To give just a thumbnail sketch, in his survey of the Indo-European lexemes expressing this meaning, Buck (5a, 5b) noted 75 developments, of which 22 are found in Slavic languages, and the list of which could be extended further. Among those lexemes for ‘speak’ that occur only in the dialects is PS *gъlčěti, a verb whose meaning was originally ‘make sound/noise’, as is evidenced by older attesta- tions, e.g., OCS gъlkъ, ‑a ‘hluk, шум, Lärm, tumultus’ (SJS /8: 405); OCz hlučěti, ‑u, ‑íš ‘hlučeti, křičeti, schallen, lärmen, schreien, rufen’, hluk , -a ‘hřmot, křik, prudkost, nepokoj, Lärm, Geschrei, Unruhe, Ungestüm’ (Gebauer 1970: 431, 432); ORu гълчати,  It has not, however, completely disappeared. It is attested in Čakavian dialects, e.g., on Dugi otok: glagolȁti ‘razgovarati’ (Sali, Piasevoli 1993: s.v.); and on Vis: glagoljât ‘govoriti narod- nim jezikom (spram talijanskom u svakodnevnom i spram latinskom u crkvenom govoru)’ (Roki 1997: 122–123). These usages may exemplify reintroductions from Church language into everyday speech. 0 Slovenski jezik – Slovene Linguistic Studies 6 (2007) гълчу ‘кричать’, гълчание ‘шумъ, крикъ’, гълка ‘шумъ, мятежь’ (Срезневский 1893: 611). The verb shifted to the meaning ‘speak’ in a subset of Slavic dialects, currently attested in three disparate regions of Slavic—northern and central Russian dialects and, in the South Slavic area, Bulgarian and Slovene dialects, a semantic shift that is paralleled in a number of cases both at the Indo-European and Slavic lev- els. The lexeme is lacking altogether in the BCS standard languages, notwithstand- ing similar-sounding verbs gúčati, ‑čīm ‘say the sound “gu”, babble’, gúkati, gȗče2 ‘coo’ (< *gou̯k-, cf. Ru dial. гучать, Sk hučat’; Ru гукать, Sk húkat’), and all but marginally in the dialects associated with them. With regard to Russian dialects, the verb голчaть means ‘говорить’ in northern Russian dialects north of Moscow from Novgorod to Kostroma to Vladimir; голчить has the same meaning in a wider band running from north to south in a crescent-shaped area around Moscow from Vologda to Voronezh; in virtually all of these areas both forms also mean ‘кричать’ (ЭССЯ 7: 190; Даль I: 366; see Map 1). Material for the words that are of interest to us here have not been collected systematically for the Slovenski lingvistični atlas (Vera Smole, p.c.), currently in preparation. This is not surprising, given that govoriti is generalized for most of the Slovene territory and the areal of *gъlčěti has long been known to be associated (ex- clusively?) with the Pannonian Slovene dialect. Perhaps for the reason just mentioned there is no discussion of the lexeme in isolexical studies that include the Slovene territory prominently (e.g., Neweklowsky 1987, Куркина 1992). The word is mentioned in passing in Tolstoj’s lexical analysis of the South Slavic area in support of a view that South Slavic may be divided into peripheral and central zones, with Slovene and Bulgarian lying towards the periph- eries, contrasting with the Western vs. Eastern division going back to Kopitar and Miklosich. A partial listing of Tolstoj’s lexical items is given in Table , where PS forms are given in place of the author’s modern forms (1974 [1997]: 234–235). The central vs. peripheral division of South Slavic goes back to the work of ethnogra- phers (Gavazzi 1936, Bratanić 1951) studying types of plows and their nomenclature among the South Slavs (a thorough discussion of the topic, with references, is found in Šivic-Dular 87). We agree with Bezlaj’s view that “[r]aba pluga in rala je bila odvisna od mnogih faktorjev in zelo dvomljivo je, da bi bilo današnje stanje v celoti podedovano” (Bezlaj 55: ). Bezlaj’s take on the matter is sensible in light of the fact that the innovations in technology and their reflection in the lexicon constitute a problem to be considered at the Indo-European and not the Proto-Slavic level (see Puhvel 64). Nor do we endorse a view of prehistory that views material culture, language and ethnicity as isomorphic. Nevertheless, it is worth noting that the lexeme under study in this paper fits into a pattern of correspondences linking all or part of Slovene with all or part of Bulgarian/Macedonian.3 At the same time, even a casual 2 These have become (nearly) homophonous in most BCS dialects because of the merger of PS *u and *l̥. 3 Note also Bezlaj’s pertinent remarks on Slovene-Bulgarian isolexical correspondences and the better reliability of lexical than phonological evidence for discerning the heterogeneity of South Slavic settlement (Bezlaj 55: 2; 67, in particular p. 0). J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ...  Map 1. Голчать/голчить ‘говорить’ in Russian (by oblast’) glance at the equivalences in Table  demonstrate that there are no clear-cut lexical boundaries that would separate a central from peripheral zones. Rather, the corre- spondences are of a statistical and relative nature. The present paper examines the areal distribution of the innovation in South Slavic in some detail and attempts to give a plausible theory of the development of the lexeme within Slavic with the hope that the discussion will contribute to a more detailed understanding of the dialect differentiation of the Slavic speech territory. 2 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Table 1 Central Peripheral meaning *kys’a *dъzd’ь ‘rain’ *pravъ *desnъ ‘right, dexter’ *znojь *potъ ‘sweat’ *čadь ~ *čad’ь *sad’a ‘soot’ *vatra *ogn’ь ‘fire’ *lubenica *dyn’a ‘melon’ *ručati *obědati ‘eat the main meal’ *gurati *pьxati ‘push’ *govoriti *gъlčěti ‘speak’ The root *gъlk- evidently may be grouped with other Slavic verbal bases which are, according to Vaillant ‘d’origine expressive’ (1966: 395–396), and end in k ~ č, a subset of which have meanings relating to sound phenomena and are imitative of them, e.g., PS *bręčěti (IP), *bręknǫ/yti (P) ‘make sound, strum’ (OCS bręčati, Ru брячать/бренчать, брякнуть/бренкнуть); PS *kričěti (IP);*kriknǫ/yti (P) ‘scream’ (Ru кричать, крикнуть; Sn kričáti, kríkniti); PS *kvičěti (IP), *kviknǫ/yti (P) ‘make shrill sounds’ (Po kwiczeć, kwiknąć; Sn cvičáti, cvíkniti); PS*vъrčěti (IP), *vъrknǫ/yti (P) ‘growl’ (Po warczeć, warknąć; Sn vrčáti). Some of these words were formed at least as early as the Balto-Slavic stage, e.g., Li brìnketerėti ‘clink, jangle’ (with “mo- mentive” suffix, see Stang 1942: 168), krỹk(š)ti ‘scream’, kvỹkti ‘squeal’, ver̃kti ‘cry’ (Vaillant: loc. cit.), formed with the extension -k-, e.g., *bręk- is related to a simplex form attested in Sn brnẹ́ti ‘buzz’, Po brzmieć, Bg бръмча, related to Ve bhramará- ‘bee’, La fremere ‘make noise, buzz’ (< PIE *bh(e)rem-) (see Snoj 2003 s.v. brnẹ́ti). Likewise, PS *gъlk- may be analyzed as *gъl‑k‑, whereby the first element goes back to a PIE onomatopoeic root, either *gal- (Pokorny 1959: 350) or *ghel- (428). The latter appears to be the better fit insofar as apophonic variants are attested, cf. OHG gellan ‘scream’, nahti‑gala ‘nightingale’, En yell, nightingale, and the Baltic and Slavic correspondences both point to reduced grade, BS *GL ̥K- > La gulkstêt ‘yell’, Bg глъча. Whatever the PIE provenience, of particular interest to the internal history of Slavic is the semantic shift of *gъlk‑ from ‘yell, make noise’ to ‘speak’. The se- mantic development from ‘make (a) noise’ to ‘speak’ is a common one, for which Buck lists some twenty-two cases attested in Indo-European languages alone, e.g., PIE *(s)preg- ‘crackle’ > Li spragė ́ti ‘crackle’, Sn prážiti ‘fry’ vs. OHG sprahhan, OEn sp(r)ecan, Al shpreh ‘I express’ (Buck 5a: ); PIE *gu̯eu̯- ‘make noise’ > Gk βοή ‘a shout’, Ve jóguve ‘shout’, Li gaũsti ‘make noise’, OCS говоръ ‘tumult’ vs. Ru говор ‘speech’ (ibid.: 0; ЭССЯ 7: 75–76). The evidently universal nature of the de- velopment makes it difficult to decide whether the developments in the three distinct areas of Slavic go back to a single innovation or evidence parallel innovations. The matter must be decided on circumstantial evidence and the solution can only be of a probabilistic nature. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 13 The semantic shift can be more easily understood when the formally parallel, but semantically antonymical roots BS *ML ̥K- (< PIE *m(e)lh2k-) and *GL ̥K- (< *g(he)lk-) are contrasted, cf. StSn molčáti, molčí ‘ne govoriti’ (SSKJ); Ru молчать, молчит; Bg мълча < *mъlčěti, *mъlči-. Table 2 context meaning progress of semantic shifts *ML̥K- *GL ̥K- . sound/perception silence noise 2. human activity being quiet making noise 3. modal keeping one’s mouth shut gossiping 4. speech remaining silent speaking 1. Western South Slavic The Western South Slavic standard languages (Slovene, Bosnian, Croatian, Serbian) operate with a highly similar repertoire of verba dicendi, the neutral imperfective verb expressing the notion ‘speak, be able to speak’ usually being Sn govoríti, BCS govòriti, while zbòriti is favored in Montenegrin. Transactional imperfectives are de- rived from the same base, Sn pogovárjati se, pogovȃrjam se ‘converse’; BCS razgov‑ árati, razgòvārām ‘idem’. Other verba dicendi, mostly perfective, come from a num- ber of sources, Sn povẹ́dati, povẹ́m ‘say, tell’; Sn réči, réčem; BCS rȅći, rȅčēm ‘say’; Sn práviti, prȃvim ‘say, tell’ (perfective and imperfective); BCS kázati, kȃžēm ‘say’; Sn velẹ́ti, velím ‘order, say’; BCS vèlīm (imperfective, no infinitive). Some combina- tion of these verbs occurs throughout the area and they are widely represented in the dialects. Of these, praviti and (veleti), velim are the only ones not to be inherited as verba dicendi from Proto-Slavic. Apparently isolated is the verb slovı̏t, slõvin ‘speak’, marked as “archaic” for the locality of Bejska Tramuntana on Cres (Velčić 2003: 395).5 Not without historical significance is the attestation of *glagolati ‘dicere’ in the Freising Folia: imper. 2nd, ‘malediction’, both in the first of the three texts (BS: s.vv.). 4 The western South Slavic material was assembled by MLG, the eastern South Slavic material by JS. The general parts were written collaboratively. MLG wishes to thank Grant Lundberg (Brigham Young University), Bernard Rajh (University of Maribor), Vera Smole (ZRC SAZU), and Marko Snoj (ZRC SAZU) for their assistance in preparing the data, the first three for providing data for particular dialect points, the fourth for arranging for speedy delivery of newly published materials pertaining to the Prekmurje dialect as well as, as always, providing a helpful discussion of etymological problems. 5 This verbum dicendi is noteworthy in that it gives credence to explanations of the ethn- onym *slov’an- as ‘speaker, Slav’ (in contrast to *něm- ‘mute, German’). If this is so, it indi- cates a highly archaic usage for Bejska Tramuntana. ↓ 4 Slovenski jezik – Slovene Linguistic Studies 6 (2007) 1.2 Slovene As in BCS, govoriti is the most widespread neutral term for ‘speak’. Among the earliest attestations, confirming central Slovene usage, we find in the Kranj Manu- script (st half of the 6th c.) a passage that fortuitously attests a range of verba di- cendi: <...Khar khulj Vſwety wode naprey Vſetu. olli gouoryenu. khar edi Vſackhaterj retſche, olli gouorj. tiga Nicomer pouedati...> (Mikhailov 1998: 183). De Sommaripa (60 [7]) lists s.v. parlare (the second term from Ge schreien ‘scream’).6 The Slovene reflex of *gъlčěti is attested first in the 18th century in texts originating in Prekmurje (75) ‘they did not speak of him’, <ſtero bom miszlo, gúcsao> (754) ‘what(ever) I think I shall say’, ‘speak the truth’ (1754), (1833) ‘all spoke as with a single tongue’ (Novak 2006, all s.v. gúčati). The dialect word also appears normalized (based on etymological considerations such as *l̥ > ol) in the modern standard language as a reflection of dialect usage, e.g., in the prose of the Prekmurje writer Miško Kranjec: “V mlin pa nikdar?” ji je ušlo, čeprav se ni kanila pogolčavati z mlinarjem, z ‘očom’, ‘očimom’, ali kar ji je že bil ‘“And never into the mill?” she blurted out, although she hadn’t intended to converse with the miller, with “Dad”, “Stepdad”, or whatever he was supposed to have been to her’ (Kranjec 77). The core area of the continuation of *gъlčěti in the meaning ‘speak’ is well known in Slovene dialectology to be within the Pannonian group, which includes Prekmurje, Slovenske gorice, Prlekija and Haloze. In particular, the usage has been known from the Prekmurje literary tradition, dating to the early 8th century, which can now be mined efficiently with the publication of Novak 2006. In this dialect not only is the verb gučati the protypical term for the meaning ‘speak’, but it serves also as the basis for numerous derivatives, e.g., nagúčati ‘pregovoriti’, pogúčati se ‘pogovoriti se, dogovoriti se’, vö̀ zgúčati se ‘izgovoriti se’; the verb govorìti is known in literary Prekmurje usage (Novak 2006: s.vv.). The Prekmurje dialect, including the Porabje region in Hungary, is fairly unified internally, with some minor phono- logical isoglosses dividing the area into three subdialects (Greenberg 1993) and a fairly distinct boundary running along the Mura between Prekmurje to the north and Prlekija and Međimurje to the south (Greenberg 2005). In the 1980s, when Greenberg was conducting fieldwork in Prekmurje and Porabje, a dialect speaker from Cankova explained to him that a salient characteristic of the region is that Prekmurci gučimo, oni tam prek pa govorijo ali se spominjajo, a proposition that cannot be adequately translated into English. The utterance employs three different regional synonyms meaning ‘to speak’—gučati (standardized Sn golčati), govoriti and spominjati se, the latter two being used as neutral verbs for this activity (or ‘converse’) in locali- ties in the Croatian Kajkavian and Slovene speech areas neighboring Prekmurje. The first is used as the neutral term for ‘speak’ in Prekmurje, e.g., Doma smo slovenski gučali ‘at home we spoke Slovene’ (Felsőszölnök/Gornji Senik) (Kozar-Mukič 1988: 6 If the term šrajati meant ‘govoriti’ for de Sommaripa, his younger neighbor Janez Svetokriški (1647–1714) understood the word closer to its German meaning: , (Snoj 2006: 426). J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 5 0). When used in a locally native context, the verb gučati serves also as a marker of in-group identity in addition to denoting the activity of speaking itself, insofar as the dialect is used alongside Standard Slovene (except in Hungary) by virtually all speakers. When used by outsiders as a borrowed term, the verb carries a different and richer meaning, ‘to speak as they do in Prekmurje’. Outside of the Pannonian dialects we find scant and sporadic evidence for this verb, in fact, only one piece of it for the NW Sn dialect of Bovec: gučáti, gučím ‘brneti, neprekinjeno bučati’ and ‘govoriti, govoričiti, dajati nauke’ (Ivančič Kutin: s.v.),7 meanings which closely parallel those found in Čakavian (see below). A spu- rious connection to the verb is found in Šašel’s lexicon for the Rož dialect (Carin- thia), which has an inexact formal correspondence: hóvšat, hovším ‘šumeti’, pud Húmprščim mûstam pa Drava hovší ‘and the Drava rustles under the Humperk (Höl- lenburg) Bridge’ (Hafner and Prunč IV: 59).8 This verb is probably not from *gъlčěti, but *glušěti ‘be/go deaf’ after vowel reduction (hoṷš‑ < *həṷš‑ < *həl̥š‑ < *hl̥š‑). Below are further details on the data in the Slovene Pannonian dialects. 1.2.1 *gъlčěti Pleteršnik notes gółčati, ‑ím vb. impf. with the meanings ‘sprechen, reden’, which are both taken from earlier dictionaries dating to the 1860s (Cigale, Janežič) and Caf’s material, the latter of which is marked for eastern Styria and Prekmur- je (vzhŠt. and ogr.‑C., respectively) (84: 227). The material adduced from Josip Jurčič’s texts, representing Lower Carniolan usage, includes the closely related mean- ing ‘schwätzen’ with the example tiho bodi! kaj pa golčiš! ‘shut up! what are you prattling about!’0 Other meanings are associated with making sound or noise, either produced by animals or in the atmosphere, slavec golči ‘die Nachtigall schlägt’ (Caf), grlica golči ‘die Taube girrt’ (marked for Bela krajina), ‘dröhnen, tönen’ (Kojsko, Štrekelj), and ‘tosen (o gromu)’ (Caf for Resia). Prekmurje. The dialect is divided into three subdialects, Goričko (including Porabje) in the N, Ravensko in the SW, and Dolinsko in the SE, characterized by phonological isoglosses (for details see Greenberg 1993). The following forms may be adduced: Goričko: gu'čəi̭mo (Martinje, Greenberg 0: 62); Ravensko: gúčatı̣, zgučávatı̣ sı̣, vö̀ sä zgúčatı̣; pregolči ‘elszólja magát, megszakítja a csendet’ (Cankova, Pável 0: 2), gu'čiːjo, za'guːčali, nei̭ 'guːčti! (Cankova, Greenberg field notes); 7 MLG is grateful to Vera Smole for this datum from Ivančič Kutin’s dictionary, which is still in preparation. 8 With gratitude to Vera Smole for this datum and to Tom Priestly for the identification of the bridge. Priestly (p.c.) notes that č > š would be a sporadic, if not isolated change in Rož dialect.  MLG is grateful to Marko Snoj for this suggestion (p.c.). 0 There are no specifically Lower Carniolan dialect glossaries to check. A query was put to dialectologist Vera Smole, a native speaker of eastern Lower Carniolan, who affirms that the word is unknown in her dialect.  Štreklelj adduces the forms γòwčat, γowčím ‘dröhnen, tönen’ alongside morphologically parallel mòwčat ‘schweigen’ (887: 478). This dialect does not distinguish word-level contras- tive pitch. 6 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Dolinsko: gúčati, ‑ín ‘govoriti’; nazáj gúčati ‘ugovarjati’; gúčau je (Beltinci, Novak 6: 4); pogučávati si, ‑vlen si ‘pogovarjati se, meniti se’, pogúčati se, ‑ín se ‘do- govoriti se, zmeniti se’ (Novak 6: 0); pregúčati ‘pregovoriti (koga)’, kòmaj smo ga pregúčali (Novak 6: ); zgučávati si, ‑an si ‘pogovarjati se’; zagúčati se, ‑ín se ‘zagovoriti se’; zgúčati si, ‑ín si ‘dogovoriti se, reči (sam zase)’ (Novak 1996: 78); 'gou̯čati, po'gou̯čo, 'gou̯čalo sɛ (Bistrica, Greenberg field notes); 'gučo, 'guːčala (Bükovje, Greenberg field notes). Pavel’s proposed Prekmurje standard: ‘beszélni’ (Pável 42: 00).2 Slovenske gorice. gu'čiː (Črešnjevci, Koletnik 2001: 64, 83),13 zgu'či:mo (Sveta Trojica, Koletnik 200: 78), gu'čiːmo (Benedikt, Koletnik 200: 78; Negova, Ko- letnik 200: 8), gu'čiːn (Radenci, Koletnik 200: 85), 'guːča Msg l-pcp (Spodnji Ivanjci, Koletnik 200: 0), 'guča Msg l-pcp (Radenci, Koletnik 200: 0), 'uni̥ si z no'benin 'nẹčejo 'gučati̥ (Koletnik 200: 205), 'tixo 'guːči̥ 2sg-imper (Koletnik 200: 207). Prlekija. 'guːčati, gu'čiːn, 'guːča je ‘he spoke’ (Cerkvenjak, Rajh 2002: 7 and p.c.), 'gučati ‘govoriti’ (Gomile pri Kogu, Zorko 1998: 72), 'gučati, gu'čiːn (Radomerščak, Zorko: 1992: 473). Rajh (p.c.) confirms that this is the only neutral term for ‘speak’ in Slovenske gorice and most of Prlekija, for which he notes in particular Sv. Ana (Slovenske gorice), Trnovska vas, the environs of Ljutomer, and Markovci near Ptuj. In addition, speakers in the area around Ormož (Litmerk, Ivan- jkovci) employ the verb spominati se in the sense of a transactional verbum dicendi ‘pogovarjati se (o čem)’. Derivatives: z'guːčati si ‘agree’, doj z'guːčati ‘dissuade’, coj 'guːčati ‘persuade’; gučliv ‘talkative’; z'guːčano ‘agreed’ (Cerkvenjak, Rajh, p.c.). Središče in the SE of Prlekija has govoriti as of the late th-early 20th centuries ac- cording to Ozvald’s materials and lacks gučati even today (Greenberg 2: 42).4 Haloze. Central and eastern Haloze attest 'gučati (Belavšek, Lundberg, p.c.), 'gučati, gu'čiːn, gu'chiːmo ‘to talk, I talk, we talk’ (Gorenjski vrh, Lundberg : 100 and p.c.), but the western zone of Haloze has govo'riː (Žetale, Lundberg, p.c.). To the east, immediately across the Croatian border the meaning ‘speak’ is rendered by gevueː'riti, ge'vuːeril ‘speak, he spoke’ and a reflex of *gъlčěti is evidently unknown (Brežani, Lundberg, p.c. and 2005). This is in accord with what can be surmised from 2 Pável’s unpublished standard, prepared in 1942, reflects something of an average of Prekmurje varieties without distinctly local characteristics (e.g., it lacks the change of jV ́ > d’V ́ (jȅzik > d’ȅzik ‘tongue’). Following Hungarian orthographic practice, long monophthongs are marked by the acute accent sign (gúcsati), which in the case of Prekmurje Slovene always co- incides with the place of stress. Unstressed and stressed short monophthongs are left unmarked. In his dialect descriptions, Pável marks short stressed vowels with the grave accent mark, a practice followed also by Novak. The practices used in these original publications are carried over in the present paper. 13 Both guč and gučati, evidently from Peter Dajnko’s (1787–1873) native dialect of Črešnjevci, entered the regional standard that he had invented, viz. (using Dajnko’s alphabet) ‘govor’, ‘govoriti’, ‘govori!’ (Rajh 8: 65, 5). 4 Moreover, Lundberg has recently confirmed in the field that gučati is not used in Središče today (p.c). J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 7 Jedvaj’s description for the Kajkavian dialect of Bednja, which similarly lacks a re- flex of *gъlčěti and instead has *govoriti (Jedvaj 56). 1.2.1.1 *gъlkъ The Slovene continuation of PS *gъlkъ is attested in Pleteršnik gȏłk ‘das Tosen des Donners’ (84: 228) and is marked as originating only in the Resian dialect (NW), but it is evidently unattested in Steenwijk’s modern material for Resia (Steenwijk 200: 2). The meaning ‘speaking, talking’ is given in SSKJ for gọ̑lk and marked as archaic in contemporary Slovene, though the term is limited to phrases contrast- ing with mo ̣̑lk ‘silence’ (e.g., golk je srebro, molk je zlato ‘talk is silver, silence is golden’). No Slovene dialect data indicating the existence of golk outside of these sources has been found. 1.2.1.2 *gъlčь Pleteršnik lists gòłč, gółča ‘das Reden, die Rede’ and gółčati, ím (impf.) ‘spre- chen, reden’ (Cig., Jan, vzhŠt., ogr.‑C.) (84: 227). Prekmurje: gùč, (Cankova, Pável 0: 2); gùč, ‑a ‘govor, govorjenje’, nèga gùča ‘no way’ (Beltinci, Novak 6: 4). Prlekija: 'guč, -a, loc. sg. pr ̥ 'guːči ‘speech’. Collocations indicate the range of uses of the term, from ability to speak to concrete instantiations of speech: po guči spoznati ‘recognize someone by his/her voice’, guč je bija o ... ‘it was a matter of ...’, meti lepi guč ‘give a fine public speech’, dosta je guča ‘that’s enough of that talk’, kaki guč je to ‘what sort of drivel (is that?)’, iti na guče ‘go for some conversation, gossip’ (Cerkvenjak, Rajh, p.c.). The short-stressed form in Prekmurje gùč, from which the standard form listed in Pleteršnik (gòłč, gółča) was constructed, is unlikely to have been inherited from Proto-Slavic, as there is no comparative evidence pointing to a laryngeal in the root nor, for that matter, any potential comparanda outside of a few parallels in Bulgarian (see below). Moreover, the related verb is mobile stressed. We can conclude, then, that gùč is a back formation. The short stress was assigned by default as the unmarked stress for monosyllabic nouns in the Prekmurje dialect, cf. còl, ‑a ‘an archaic mea- surement = 2.5 cm’ (< Ge Zoll, StSn cọ̑la), cùg , -a ‘train’ (< Ge Zug), vr̀č, ‑a ‘jug’ (StSn vr ̑č). Given the proximity of the lexeme’s areal to Hungarian and the high number of words borrowed from Slavic dialects into it, we also consulted Kniezsa 74. The word is not listed there.5 1.3 Bosnian/Croatian/Serbian The reflexes of *gъlkъ, (*gъlčь), *gъlčěti are conspicuously absent from BCS sources, e.g., no mention of them is found in the historically oriented Akademijin rječnik, Skok’s and Gluhak’s etymological dictionaries, no evidence of them in a several dozen dialect grammar dictionaries and descriptions consulted, nor are BCS 5 However, famously, the borrowings beszéd ‘speech’ and beszélni ‘speak’ (< PS *besěda) do occur in Hungarian. 8 Slovenski jezik – Slovene Linguistic Studies 6 (2007) reflexes adduced in broader Slavic dictionaries such as ЭССЯ. Only a few peripheral indications have been found, which are listed below. 1.3.1 Kajkavian The RHKKJ s.v. gučati (Sv. 3, 1986: 688; there is no entry for **guk‑ nor **guč‑) points to Belostenec’s Gazophylacium of 740, representing Kajkavian usage of the early 8th century, where is found in the Latin-“Illyrian” volume s.v. Ajo, which is glossed as , , and . Intriguingly, is marked as , indicating that the word was deprecated in the proto-standardization process owing to its rural provenience. It is conceivable that the word was known to Belostenec from usage in the areas that are today known as Slovene Pannonian dialects, though there is no way to know for certain. In any event, this establishes that gučati was known, at least to some extent, in Kajkavian speech territory in the past. In addition to this sociolinguistic hint, we note that the item is absent from Houtzagers’ description of dislocated Kajkavian dialects, which were transplanted from the 5th century onwards, having become enclaves within Map 2. Distribution of *gъlk‑ reflexes in Slovene NE: Prekmurje: Bi = Bistrica, Bl = Beltinci, Bü = Bükovje pri Polani, Ca = Cankova, GS = Felsőszölnök/Gorni Senik, Ma = Martinje; Slovenske gorice: Bn = Benedik, Čr = Črešnjevci, Ra = Radenci, SIv = Spodnji Ivanci; Prlekija: Cr = Cerkvenjak, Go = Gomile pri Kogu, Iv = Ivanjkovci, Li = Litmerk, Mr = Markovci, Rd = Radomerščak, Tr = Trnovska vas; Haloze: Be = Belavšek, Gv = Gorenjski vrh, Že = Žetale; SW: Ko = Kojsko J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ...  the Hungarian speech territory; rather, Hidegség and Fertőhomok Kajkavian (both spoken near Sopron) have govȍrit, govorı̑m, as expected (Houtzagers 1999: 251). It is possible that none of these forms was inherited in Kajkavian or, for that matter, elsewhere in the dialects corresponding to BCS. In other Kajkavian lexicographi- cal works from the 7–8th centuries a similar picture emerges with regard to verba dicendi, e.g., ‘loquor’, ‘narro’, ‘dico’ (no **) (Habdelić 1670); s.v. Fabulo , s.v. Loquor , Narro (Jambrešić 1740). 1.3.2 Čakavian The north-western Čakavian dialect of Kastav near Opatija attests the verb gūčȅt, gūčı̑n ‘govoriti o čemu po selu’; Sȍ selo guči ‘govori se po cijelom selu’ (Jardas 57: 392). The same usage is confirmed nearly a half century later for nearby Rukavac: gūčȅt, gučin (no accent listed on present tense) ‘brujati, na veliko govoriti o čemu’, So selo guči (Mohorovičić-Maričin 2001: s.v.) with the additional information that the word has both the older and newer meanings simultaneously, viz. ‘brujati’ ‘buzz, hum’ and ‘gossip’. There is no reason to think that the word has become the neutral word for ‘speak’ in these localities. Of particular interest are the attestations of the lexeme in outlying Čakavian dialects in Austrian Burgenland, where the speakers have been cut off from contact with the bulk of Čakavian for several centuries. Here the meaning ‘make a noise’ as well as the conflation of the two roots *guk‑, *gl ̥k- in- dicate an earlier stage of the semantic development, e.g., gučàti, gûčem6 (Hrvatski Cikljin/Spitzzicken), gučîm (Stari Hodas/Althodis) ‘gurren’, Gòlub gûče na kròvu (Hrvatski Cikljin/Spitzzicken) (Tornow 1989: 146). 1.3.3 Štokavian No examples of the verb *gъlčěti have been found for the majority of Štokavian dialects with the exception of intrusions from neighboring Eastern South Slavic dia- lects. See below 2.3.9 Southeast Serbian (“Torlak”) dialects. 1. Analysis of the Western South Slavic material In contrast the the more complex picture that will be seen below with regard Eastern South Slavic, where the verb *gъlčěti in the meaning ‘speak’ competed with several other forms for the same or similar meaning, the Western South Slavic picture is relatively simple. The verb *gъlčěti has two distinct denotations that correlate with distinctly different areal distributions. In those areas that have the meaning ‘speak’, that is, the Slovene Pannonian dialect (Prekmurje and Porabje, Slovenske gorice, Prlekija and Haloze), the word has long been established as the primary verb for ‘speak’ and has given rise to derivatives with ‘speech/speaking’ as their common denominator. The isogloss between the Pannonian dialects is fairly sharp against both other Slovene as well as Croatian Kajkavian dialects. In the SE periphery of the Pan- nonian dialects, the important marketplace dialect of Središče ob Dravi in Prlekija and bits of W Haloze indicate a switch to the prevailing reflex of *govoriti, a situation 6 No tones. The grave marks short, the caret long stress. 20 Slovenski jezik – Slovene Linguistic Studies 6 (2007) which can be explained by influence from surrounding dialects both on the Slovene and Croatian sides of the border. In remaining areas of Slovene and Čakavian, the attestations of *gъlčěti point to the more archaic meaning ‘to make noise’, ‘raise a ruckus’. Their attestations are sporadic and are found unsurprisingly in the periphery, indicating not so much that the form has all but disappeared in these areas, so much as that, in contrast to the Pannonian situation, the verb is highly marked and not central to everyday usage. The one outlier is Bovec in NW Slovene, which attests as one of its meanings ‘speak’ along with marked meanings ‘make a buzzing sound, constant muffled roaring noise’ (‘brneti, neprekinjeno bučati’) as well as a secondary mean- ing ‘lecture, dress someone down’ (‘dajati nauke’). With regard to the plain meaning ‘speak’ there are three logical possibilities: () the Bovec word could be organically connected to Pannonia through a later migration; (2) it could be connected to Pan- nonia as a rare retention of an earlier and much more widespread situation prior to the expansion of *govoriti; or (3) it could be an independent local innovation following the same path of semantic development as elsewhere, i.e., ‘yell’/‘make a noise’ > (‘gossip’/‘prattle’ >) ‘speak’. There is no evidence to point to (). Possibility (2) is weak in light of the fact that related dialects in the surrounding areas have the more archaic meaning, to the extent that the lexeme is attested. Because the entire range of meanings is attested in Bovec, including the rather specific reference to the kind of noise noted for the lexeme in both Russian dialects and the archaic Slovene dialect of Rezija (i.e., a buzzing/roaring noise in the distance), scenario (3) presents itself as the most plausible, i.e., that it is an independent local innovation. The Pannonian usage is well attested in the earliest documents found there, that is, at the beginning of the 8th century. In view of the centrality of the notion ‘speak’ to everyday language, the compactness of the Pannonian dialect area in which *gъlčěti is found, the high productivity to which the verb is subject to there, as well as indica- tions of its use as a marker of local identity, there is every reason to believe that the meaning ‘speak’ for is an old, inherited one, conceivably as old as the Slavic settle- ment of the “Pannonian” territory. The other ‘noise’ meanings are older still. We shall now turn to the evidence in Eastern South Slavic before considering whether and, if so, how, to connect the Western South Slavic picture with that of other parts of the Slavic-speaking world. 2. Eastern South Slavic 2.1 Bulgarian Modern Literary Bulgarian operates with two basic imperfective verbs of speech which do not implicitly require a direct object, viz. говóря and прикáзвам ‘speak, talk, converse’ (in contradisintction to казвам ‘tell, say’). Whereas dictionaries of the standard language make mention of equivalent dialect forms such as думам and хортувам, the verb гълчъ is mentioned either not at all or only as a dialect equivalent of ‘shout; scold’. A similar brief notice is usually accorded a small series of nouns and J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 2 adjectives which are derived from the base *gъlk-, e.g., глъч, -ъ́́́т, глъ́́́чка ‘clamor, uproar’ (see below).7 Nevertheless, dialect-colored texts from Bulgarian literature occasionally evi- dence the usage of *gъlčěti in various senses of speech, as indicated, for example, in lexicographical excerpts from the works of Ivan Vazov, cf. гълчá 1) ‘приказвам, говоря; converse, talk’, Стига си ми по седенки с Камена безняла, заран‑вечер на извора със него гълчали; 2) ‘карам се на някого, мъмра; scold, nag’ ...и му гълча дълго с твърде решителен вид, със силни ръкомахания и с чести поглеждания към килията (Антонова-Василева, Керемедчиева 2001: s.v.). In one source which is partially devoted to dialectisms in Bulgarian literature of the –20th cc., there is no entry for *гълчъ itself, but the verb is used (along with бърборя) as a gloss for the ‘dialect’ form гълчуля, which occurs in the work of M. Javorski, cf. Гълчуля аз, но не обичам да докачам чиляка... (Илчев, Иванова, Димова, Павлова 1974: 92).8 In addition, Gerov’s monumental and widely inclusive dictionary of the Bulgar- ian language (Геров 1895–1904) provides the following entry for *gъlčěti with the meanings ‘shout; converse with; upbraid’, cf. глъчъ ́́́, -и́́́ш, -чáл 1) ‘говоря яката, викам (Ru кричать, орать, горланить)’, Глъчат като Чифути в хавра; 2) ‘глъча с някого, говоря, разговарям се, сборвам, ромоня (Ru говорить, разговаривать)’, Глъчим си с Драгана; 3) ‘хокам, карам се, викам, бъбра, мъмра, накръпелясвам (Ru бранить, журить, тазать, нарекать, порицать, распекать)’, Аз го (му) глъчах, ала не слушна; глъчи се ‘чуе се глъч, вика се (Ru слышны крики)’, Из пътя се глъчи, някои минуват. We also find the prefixed verb разгълчá (да се), -иш, -чал се, разглъчáвам се with the meaning ‘(start to) raise a ruckus’, cf. ‘начинам да глъчъ твърде много; развикам се (Ru раскричаться)’, Разглъчали се като Цигани на пазар (като в Циганска механа, като на Циганска сватба) (5: 8). It is interest- ing to note that the examples with the meaning ‘shout; raise a ruckus’ involve pejo- rative comparisons with the (unintelligible) speech of marginalized and/or foreign- speaking minorities (e.g., Jews and Gypsies) in a context where a group of people are shouting. 7 The multi-volume Речник на българския език (8) presents quite a different situa- tion with respect to *dumati and *duma. Here, the verb is acknowledged (and well documented with citations from Bulgarian literature) as a “dialect” form in the sense ‘говоря, казвам, при- казвам’, whereas the noun is clearly the basic Literary Bulgarian term of expression for ‘word, speech’ and is used in an extremely wide range of meanings and idioms. 8 Various linguistic features of this example point to NE provenance, cf. the form of the first person pronoun (аз) (which excludes southern Bulgaria), apocope of the middle syllable of *čelověkъ (widely encountered in E Bg dialects, see БДА), and vowel reduction in the first syllable of this same word  Although the phonology and accentuation of Gerov’s dictionary reflect the west-central Stara Planina dialect base of his native Koprivštica (see with specific respect to accentuation, p. 50 of Gerov’s “Introduction”), the lexicon itself includes many words from both western and eastern dialects, all without indication of provenance (a telling example is the inclusion of both W Bg разбой and E Bg стан as words for ‘loom’). This of course is not surprising, since Gerov in the introduction to his monumental work tells us that he drew all of his data from the spoken speech and folklore of his compatriots in his quest to fashion a dictionary of the ‘liv- ing language of the Bulgarian people’ (‘речник на жив български народен език’ (see Геров 22 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Map 3. Bulgarian dialect areas 2.1.1 Sources for verba dicendi in Bulgarian dialects Our sources for verba dicendi in Bulgarian dialects fall into two main categories, viz. the Bulgarian Dialect Atlas (hereafter, БДА) and all the others. In geographical terms, the four primary volumes of the БДА (i.e., those devoted to dialects situated within Bulgaria proper) are divided geometrically into four quad- rants (SE, NE, SW, NW), approximately equal in area and defined by two straight lines, perpendicular to each other and intersecting at roughly the midpoint of Bulgar- ia. These four quadrants correspond only approximately to linguistically determined entities, such as “Northwest Bulgarian”, “Southwest Bulgarian”, “East Bulgarian”, “South Bulgarian”, etc., as defined in standard works on Bulgarian dialectology (see notably Стойков 1968). 1894, p. IV). These considerations notwithstanding, in the case of *gъlčěti it is quite possible that this particular verb occurred in the Koprivštica dialect of Gerov’s day, since БДА-SW indicates a continuous block of “гълчъ-dialects” immediately to the east of the town. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 23 In addition to these four cardinal volumes, all of which are the product of col- lective teamwork directed by Stojkov, there are two additional continuations of the БДА, each devoted to dialects spoken in regions adjacent to the republic of Bulgaria, viz. eastern Aegean Macedonia (БДА-Aeg; Иванов 1972) and extreme southeastern Serbia (БДА-Trans; Божков 1986). Although these last two volumes are produced in the same format as the first four, in contradistinction to the latter they were compiled by single authors who did not conduct fieldwork in situ, but rather worked with in- Map. . Eastern South Slavic dialectal area. Key to place names Serbia: 1. Timok, 2. Lužnica, 3. Pirot, 4. Caribrod, 5. Svrljig, 6. Bosilegrad; Macedonia: 7. Kumanovo, 8. Skopje, 9. Polog, 10. Debъr, 11. Kičevo, 12. Galičnik, 13. Radožda-Vevčani, 14. Dihovo, 15. Ohrid, 16. Bitola, 17. Prilep, 18. Mariovo, 19. Veles, 20. Radoviš, 21. Maleševo, 22. Gevgeli, 23. Dojran; Albania: 24. Korča; Greece: 25. Lower Prespa, 26. Kostur, 27. Le- rin, 28. Kajlar, 29. Kukuš, 30. Kulakia, 31. Solun, 32. Struma, 33. Belica, 34. Valovište, 35. Ziljaxovo, 36. Čeč, 37. Gjumurdžina, 38. Dedeagač; Turkey: 39. Lozengrad, 40. Čanakča, 41. Mandъr; Bulgaria: 42. Vidim, 43. Kula, 44. Belogradčik, 45. Montana, 46. Berkovica, 47. Lom, 49. Kozloduj, 50. Orjaxovo, 51. Pleven, 52. Bjala Slatina, 53. Vraca, 54. Botevgrad, 55. Svoge, 56. Sofia, 57. Godeč, 58. Graovo, 59. Trъn, 60. Burel, 61. Pernik, 62. Pernik, 62. Ra- domir, 63. Kjustendil, 64. Dupnica, 65. Gorno Pole, 66. Elin Pelin, 67. Makocevo, 68. Pirdop, 69. Ixtiman, 70. Samokov, 71. Blagoevgrad, 72. Petrić, 73. Sandanski, 74. Goce Delćev, 75. Razlog, 76. Pazardžik, 77. Panagjurište, 78. Etropole, 79. Koprivštica, 80. Teteven, 81. Trojan, 82. Loveć, 83. Sevlievo, 84. Nikopol, 85. Veliko Tъrnovo, 86. Gabrovo, 87. Trjavna, 88. Elena, 98. Kotel, 90. Erkeč, 91. Karlovo, 92. Kazanlъk, 93. Nova Zagora, 94. Ajtos, 95. Svištov, 96. Bjala, 97. Silistra, 98. Razgrad, 99. Popovo, 100. G. Čamurla, 101. Šumen, 102. Sъrt, 103. Pre- slav, 104. Zagorci, 105. Fakija, 106. Strandža, 107. Xaskovo, 108. Krumovgrad, 109. Plovdiv, 110. Asenovgrad, 111. Xvojna, 112. Smoljan, 113. Široka Lъka, 114. Devin, 115. Peštera, 116. S. Pavlikjan, 7. N. Pavlikjan; Romania: 8. Bucharest, . Banat 24 Slovenski jezik – Slovene Linguistic Studies 6 (2007) formants who were either refugees from displaced populations (cf. БДА-Aeg) or who had emigrated to or were visiting in northwest Bulgaria (cf. БДА-Trans). Sources other than the БДА are diverse in nature. They include dialect dictionar- ies of varying dimensions, lexicographical notes, dialect descriptions, unpublished archival materials, and Schallert’s own field notes (for Ardino and Široka Lъka). Of these sources, the most useful have proved to be the dialect dictionaries and lexical observations. Since the authors of these latter two sources are themselves frequently native speakers of the given regional dialect or are thoroughly conversant with it, their testimony regarding lexical peculiarities and semantic nuances is of particular value. 2.1.2 Inventory of verba dicendi in Bulgarian dialects Because the lexical possibilites for expressing the notion ‘to speak, talk’ are so numerous in Bulgarian, and have therefore been collected in a relatively systematic fashion in Bulgarian dialectology, something must first be said here about the variety of forms encountered. According to the five volumes of the БДА which provide rele- vant maps and data, the sense ‘приказвам, (говоря)’ (‘speak, talk’) is rendered by the following thirteen verbs (and their morphological variants) in Bulgarian dialects:20 (a) бъбра; (b) думам; c ) говор’а; (d) гъгра; (e) гълча; (f) кликам; (g) кунуштисвам се; (h) лафим, лафовам, лафувам, лафуем; (i) приказвам, прикажувам, сказвъм; (j) оратим; хортувам, хортовам; (k) вел’а; (l) вревим; (m) зборувам, зборовам, зборуем. On the basis of semantic and etymological criteria these verbs may be bro- ken down into seven groups: () ‘indicate’ приказвъм, сказвъм; (2) ‘think’ думам; (3)‘will’ велим; (4) ‘make noise’ бъбра, говор’а, гъгра, гълча, вревим;2 (5) ‘call’ кликам; (6) (loanwords) ‘talk’ лаф-, хорт-/орати-, куношти́́́свам са;22 (7) ‘gather’ 20 The only volume of the БДА which does not provide a map and commentary for dialect equivalents of standard Bg ‘говоря, приказвам’ is БДА‑SE, the first volume published. The only published reference to the БДА materials for this questionnaire item that we have found is in Bunina (1963) (see discussion below of Bulgarian émigré dialects in Ukraine). 2 Cf. *bъbre- (from *bъrbor- with incomplete reduplication; БЕР :6), *gъgre- (from *gъrgor- with incomplete reduplication), *vrěvěti (from IE *w[e]rē-, perhaps by incomplete reduplication of *vrě‑vra-; БЕР : 28). 22 From Turkish laf ‘word’ (cf. Persian laf), horata ‘jest; loud noise’ (archaic, see Red- house) (cf. Gk χωρατά), and konştu pret. of konuşmak ‘converse with’ (with Gk suffix -is-). For the etymology of конуштисвам, see БЕР 2:602. For the etymology of Bg лаф masc., лафувам, see БЕР 3:325–326. For the etymology of Bg хоратá fem., хортувам, хорати, etc. see Buck 1915b: 145 and Филипова-Байрова 1969: 17 (with literature). In the case of хората, etc., scholars disagree as to the extent to which Turkish served as a vehicle for trans- mission to Slavic. Tk horata itself is derived directly from Greek χωρατά, the pl. of χωρατό neut. ‘jest’, from χώρα ‘countryside, region, place; town’, which also served as the base for old loanword Bg хора ‘country, region (obs.); people (post-13th c.)’. The Greek formations χωρατό and denominative χωρατέυω ‘to jest’ both apparently postdate the 11th century, since neither word is attested in Sophocles’ monumental dictionary of the Roman and Byzantine periods (Sophocles 1887/1951). It is also unclear to what extent Gk χωρατέυω [xoratevo] served as the prototype of the Slavic -(u)va-, -ova- suffixed formations. It would seem that the direct adapta- tion хоратéвам (cited by Filipova-Bajrova) was quite rare in Bulgarian (it is unattested in the J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 25 (if from *sъbor-) зборува-. Our survey of other dialect sources has revealed no ad- ditional verbs of this kind, excluding onomatopoeic and metaphorical expressions which have not acquired neutral meaning. In terms of chronological layering, these verbs of speech can be assigned to three distinct strata: ) the older Slavic stock (regardless of etymological type); 2) Ottoman Balkan loanwords (based primarily on Turkish) or lexical calques (thus, perhaps, zbor ‘word’, according to the most widely accepted etymology); 3) Literary Bulgarian (cf. the intrusion of приказвам as a neutral verb of speech into regions other than its apparent “natural habitat” in the Stara Planina). Of course, in order to distinguish between приказвам as the product of the influence of Literary Bulgarian, on the one hand, and as the result of autochthonous semantic innovation, on the other, one must take into account factors such as the observations of the investigating dialectologist, the shape of the suffix, and the location of the dialect. It is important to note at the outset that our verificatory survey of sources other than BDA has uncovered instances where *gъlčěti (or some other non-standard verb of speech) is not indicated as a neutral verbum dicendi for a given region in BDA, but is in fact attested with this meaning in a dialect dictionary or description. Where such discrepancies arise, we are inclined to assume that the given usage of *gъlčěti has probably eluded the notice of the BDA investigator(s). The fact that such an oversight is possible should come as no great surprise if we bear in mind the multiplicity of shades of meaning involved in the expression of verbs of speech and the compara- tively rich etymological inventory from which Bulgarian dialects can derive words to express these different nuances. 2.1.3 Summary of the areal distribution of other verba dicendi in Bulgarian dialects The geographical extent of the verba dicendi listed in the preceding section var- ies considerably. Detailed knowledge of this distribution can be determined for all Bulgarian dialects with relevant data in the БДА (i.e., all regions except those covered in БДА-SE). This picture can be supplemented (and at times revised) by the testimony of other sources. The resulting composite areal distribution can be summarized as fol- lows (beginning with the two literary variants then proceeding alphabetically through the other types). () According to БДА-NE, NW, SW, and Aeg, the verb “приказвъм” (as in Lit- erary Bulgarian) occurs in a wide range of dialects, particularly in N Bg, cf. (a) the Danubian plain to the north of the Stara Planina (except for Vidin, Lom, and most of the stretch from Nikopol to Bjala); (b) throughout much of the Stara Planina itself, from Berkovica in the west to Kotel in the east (except for most of Trojan); (c) Erkeč and its colonies in the eastern end of the Stara Planina and to the north of Varna; (d) Mizija o-dialects, in Grebenci, Tъrgovište, Veliko Tъrnovo-Pavlikene, and other scat- tered points. To the south, we find “приказвам” in e) W. Pazardžik, Panagjurište, E. БДА). On the other hand, direct adaptation from the nominal base *xorat- is clearly manifested in the denominative i-verb, which is still well attested in the dialects (particularly in the west) and is cited as early as the 7th c. (see our discussion of Bg Damascenes, below). 26 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Goce Delčev; the variant “приказувам” in f) Aegean Macedonia (Čeč, N. Ziljaxovo and C. Valovišta); and the variant “прикажувам” in g) N. Goce Delčev, Sandanski, N. Petrič, S. Blagoevgrad. Lastly, an extremely rare prefixed form “сказвам” is at- tested in Botevgrad and Etropole. (2) According to БДА-NW, NE, SW, the verb “говоря” (also as in Literary Bul- garian) serves as the basic verb of speech in a very small number of scattered dialects, almost all of which are located in N Bg, cf. Lom (pt. 175), the mouth of the Iskъr (pt. 204), NE of Pleven (pt. 705), N. Skopje (pt. 2183), an o-dialect to the west of Razgrad (pt. 903), in variation with приказвъм to the south of Pleven (7 pts.) and to the north of Pavlikene (pts. 605, 62), in variation with “бъбра” near Pirdop (5 pts.), and in variation with “приказвам” to the north of Koprivištica (pt. 2322). (3) According to БДА-NW, SW, the verb “бъбра” is attested as a neutral verb of speech more frequently than “говоря”, but is found only in two disconnected clusters in W Bg, cf. (a) Vidin, Kula (in extreme NW Bg; approx. 30 villages); (b) Makocevo, Pirdop, Ixtiman, Panagjurište, Pazardžik (approx. 30 villages); (4) According to БДА-Aeg, the verb “вел’а” ocurs only in Aegean Macedonia, cf. (a) a roughly continuous band of five pts. in Belica and 12 pts. in Ziljaxovo. Other sources further indicate that *velěti occurs in many Macedonian dialects, as well as in Literary Macedonian, but only in the sense ‘say, tell’, thus functioning as the perfec- tive of “рече” and the equivalent of Bg казвам. One might suppose that the use of *velěti in this function is reduced in dialects where *dumati is used for ‘say, tell’, but this aspect of the question requires further investigation. (5) According to БДА-NW, SW, Trans, Aeg, the verb “вревим” is limited to ex- treme W Bg and a small number of Transitional dialects, and parts of Aegean Mace- donia, cf. (a) Belogradčik, N. Temska (Serbia), and a band of dialects stretching from Trъn, Radomir, Dupnica, Kjustendil Pijanec, and some Kjustendil Polčane, N. Blago- evgrad, some S. Blagoevgrad, S. Petrič; (b) Aegean Macedonian dialects of Struma, Valovište, S. Ziljaxovo. Other sources indicate that *vrěvěti occurs as far west as W Rup dialect of Goce Delčev, as far east as the W Rup dialect of Goce Delčev and in Macedonian dialects throughout much of SE Mac, including Solun, Gevge- li, Radoviš. БЕР 1: 183 lists more particular or pejorative meanings for “вревя” in Western dialects, cf. ‘разговраям’ (Bansko in Pirin Macedonia), ‘бърборя, шумя’ (Debъr in peripheral Western Macedonia), ‘говоря много’ (Trъn in the Transitional dialect zone). (6) According to БДА-SW, the verb “думам” is attested as the basic verb of speech only in the SW and South, cf. (a) the Transitional dialect of Bosilegrad and some adjacent points in N. Kjustendil (Kamenica, some Kraište); (b) N. Kjustendil Polčane and Kamenica; (c) 3 pts. in Radomir, 1 pt. in Dupnica; (d) Razlog; (e) South- Central Rhodope. Other sources indicate that “думам” is used as a basic verb of speech in parts of EMac (e.g., Maleševo), as well as in Bg émigré dialects in Ro- mania and Ukraine, and in the 8th c. Svištov Damascene. The cognate дума (‘word, expressed thought’) is more widespread and, in contradistinction to the verb, survives well in Literary Bulgarian. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 27 (7) According to БДА-NW, NE, the verb “гъгра” occurs as the basic verb of speech in just two small clusters of dialects, cf. (a) 5 pts. in E. Vraca in NW Bg (1361, 1378, 1379, 1380, 1433; in variation with ‘приказвам’ except in 1433); (b) 3 o-dia- lect pts. to the south of Kotel (2531, 2576, 2577). (8) According to БДА-NW, SW, derivatives of *sъbor-, occur as the basic verb of speech in several morphophonological variants in the West, cf. (a) “зборувам” in NW Bg (*ǫ > -ъ-) dialects to the north of Sofia, 4 pts. to the south of Razlog; (b) “зборуем” in W. Sofia, Transitional dialects of Slivnica, Pernik, Godeč, N. Caribrod (Serbia); (c) “зборовам” in 4 pts. in S. Goce Delčev (in variation with лафовам). Other sources indicate the occurrence of (d) “зборва‑” (cf. type in -ъva-) in SW Mac (Korča); (e) “зборим” in Skopje (cf. the influence of Serbian sboriti); (f) “зборува-” in Central Macedonia. () According to БДА-SW, NW, Aeg, derivatives of Turkish laf‑ have supplanted other verbs as the basic verb of speech almost exclusively in various Rup dialects, cf. (a) S. Goce Delčev, Dospat, Central Rhodope (sometimes in variation with “гълчъ”), SE Drama (Aegean); but only rarely elsewhere, cf. “лафуем” in 2 Transitional dialect pts. in Kjustendil Kamenica, “лафувам” 1 village in Vraca in NW Bg (1362). In view of the widespread influence of Turkish on colloquial Balkan Slavic, the areal domain of this verb, at least as a variant (or marked variant), may actually be greater than БДА suggests. This hypothesis is supported by attestations from other dialect sources, as well as by the following listings for лаф in БЕР 3: 325–326: SE Mac: Negotino, Tikveš, Mariovo; Lower Vardar; Peripheral W Mac: Debъr; SW Bg: Pirdop (Smol- sko); NCe Bg: Elena, Sevlievo (Krъvenik); N. Thrace: Xaskovo; S. Thrace: Ksantija (Gabrovo); Gjumjurdžina (Sъčanli). (0) According to БДА-NE, NW, SW, derivatives of the Greek-Turkish Bal- kanism xorata occur as the basic verb of speech in a wide variety of dialects, cf. “хортувам” in (a) NW Bg dialects primarily in the vicinity of Vidin, Lom, but in isolated pts. in variation with “приказвам” near Bjala Slatina, Montana, Lukovit; (b) a large mass of NE Bg dialects in the Danubian plain from Nikopol to Bjala and Pavlikene, as well as scattered pts. in the Stara Planina (Botevgrad, Teteven, Trjavna, Ajtos dialects near the Erkeč group); (c) a plurality of Mizija o-dialects, including N. Zagorci dialects between Kotel and Erkeč; (d) in 5 N. Thracian dialects to the north of Asenovgrad; in the form “оратим” in (e) a mass of dialects to the south and east of Sofia (cf. S. Sofia, E. Sofia, Samokov, Ixtiman); (f) a few dialects to east of this mass (in variation with “бъбра”); (g) Transitional dialects of Trъn and S.Caribrod; (h) a small number of Rup dialects, cf. 4 pts. in S. Goce Delčev; in the form “хорто‑ вам” in 3 N.Rhodope dialects near Asenovgrad. Other sources attest to the presence of “хорати-, хоротовам” in dialects as widely separated as Strandža (in extreme SE Bg) and the Bg émigré Pavlikjan dialect in Banat (a trait held in common with N.Pavlikjan). The verb does not seem to occur in Macedonian (including Ser-Drama and other Aegean dialects). () According to БДА-Aeg, the verb “клика” serves as the basic verb of speech in a small number of Čeč dialects near the Mesta river to the south of the Bulgarian- Greek border and in variation with “прикажувам” in 2 villages to the south thereof (cf. pts. 4141, 4143). 28 Slovenski jezik – Slovene Linguistic Studies 6 (2007) (2) According to БДА-SW, the Balkan Turkish loanword derivative куношти́́́свам са occurs in variation with гô ́л’чем in 2 villages in N. Devin in the Central Rhodopes. Map 5. Areal distribution of verba dicendi (as per БДА‑NE, NW, SW, Aeg, Trans)23 2.2 *gъlčěti Having thus outlined the dialect geography of other verba dicendi in Bulgarian, we will now summarize the areal distribution of the verb “гълчъ” (primarily as the basic verb of speech, but also as a marked or pejorative term), then briefly discuss 23 Note that the map only reflects attestations in БДА, therefore a significant *gъlči cluster in Ixtiman (SW) is not included. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 2 patterns of variation involving “гълчъ” and other verbs of speech. These introductory remarks are followed by a detailed presentation of the data provided by БДА and other sources. 2.2.1 Summary of areal distribution of *gъlčěti in Bulgarian dialects In terms of geographical configuration, the distribution of *gъlčěti (primarily as a neutral verb of speech) in Bulgarian can be summarized as follows. () According to БДА-NW, “гълчъ” predominates in a continuous belt of ca. 60 points situated in Northern Transitional (*ǫ > -u-) dialects (Belogradčik, W. Montana), as well as immediately adjacent NW Bg dialects to the east thereof (E. Belogradčik, W. Lom, W. Montana, Berkovica). This extensive zone of “гълчъ” dia- lects is bounded on the north (i.e., towards Vidin), by “бъбрем” dialects, on the east by “приказвам” dialects, on the south by the uninhabitated massif of the Stara Plani- na (on the southern side of which stretches a large zone of Transitional “зборуем” and W. Sofia-Svoge “зборувам” dialects), and in part on the west by Transitional “вре‑ вим” dialects. Roughly half of the “гълчъ” dialects exhibit variants with “бъбрем”, “приказвам”, or “вревим”. Our investigation of other sources reveals that in Serbian territory, to the immediate west of the zone of Transitional “гълчъ” dialects, lexicog- raphers attest traces of “гълчъ” (but only in pejorative usage) in the Timok region (Northeastern Torlak).24 We also find isolated occurrences of “гълчъ” to the north (cf. village of Gъrci near Vidin) and east (cf. Studeno Buče and Vъrbovčec in Montana). (2) According to БДА-NW, to the SE of Berkovica the zone of “гълчъ” dialects is interrupted by a solid wedge of “приказвам” dialects (NW Svoge, W. Vraca), then resumes in the form of a pocket of about 0 villages, all but one of which are situated in remote, mountainous, and scattered locales to the south and east of the Iskъr river. To the east of this “Svoge pocket,” the zone of attestation of “гълчъ” is once again interrupted by a wide, continuous stretch of “приказвам” and “сказвам” dialects in S. Vraca, Botevgrad, most of Teteven, and E. Sofia. To the southeast of dialects in the E. Sofia plain, our investigation of other sources indicates (pace БДА-SW) that “гъл‑ чъ” is also found in the hilly region surrounding Ixtiman (in variation with “оратим” and “бъбра”) and (in the marked sense ‘говоря силно’) to the southwest thereof in the Samokov plateau (where “оратим” serves as the unmarked verb of speech). In connection with this marked usage in Samokov, it is interesting to note the neutral use of “гълчъ” in the émigré dialect of Čanakča. Although this dialect was once spoken in the vicinity of Istanbul and is now preserved in a small quarter in Šumen in NE Bg, it still exhibits clear traces of SW Bg origin (arguably from the vincity of Samokov), a circumstance which links it with the “Fakija” émigré dialect type still found in SE Bg to the west of Strandža (see below). On the other hand, there is no evidence either in БДА-SW or in our other sources which would indicate that “гълчъ“ has breached the massif of the Pirin mountain range, which rises up to the south of Samokov (cf. “думам” in Razlog, which is situated beyond this barrier). 24 Note that more detailed information is required for verbs of speech in Torlak before we can draw any conclusions regarding the areal distribution of marked (and potentially, un- marked) occurrences of *gъlčěti in this region. 30 Slovenski jezik – Slovene Linguistic Studies 6 (2007) (3) According to БДА-NW and БДА-SW, to the east of the “приказвам, сказ‑ вам” dialects of Botevgrad and W.Teteven, a massive and virtually continous zone of “гълчъ” dialects (ca. 80 villages in number) begins in the Stara Planina in E. Teteven, then extends through Trojan and southward down into the Tundža basin around Kar- lovo, whence it occupies the eastern Sredna Gora and Marica valley plain to the north of Plovdiv (including S. Pavlikjan dialects).25 In the Marica valley to the southeast of Plovdiv and the northeast of Asenovgrad, БДА-SW indicates the presence of a small cluster of dialects of the N. Thracian (rather than Rhodope) type, in which the com- mon verb of speech is “хортувам”, rather than “гълчъ”. With this exception, to the south of Plovdiv, the continuous zone of “гълчъ” attestation resumes in the Marica valley to the north of Asenovgrad, then continues into the north-central Rhodopes, extending in a broad band to the south and southwest as far as Smoljan and Devin (cf. the dialect regions of Xvojna, N. Smoljan, Široka Lъka, and N.Devin and Peštera dialects to the west thereof).26 As to the delimitation of the territory with “гълчъ”, БДА-SW indicates that to the south of Trojan, this second and geographically more imposing massive of “гълчъ” dialects is bounded to the west by the nearly uninhab- ited eastern Sredna Gora centered around Koprivštica (beyond which there lies a belt of “бъбра” dialects in Panagjurište). Somewhat to the west of Plovidv one encounters “приказвам” dialects. In the SCe Rhodope region,”гълчъ” gives way to “лафовам” and “думам”. As noted above, in the north, the Trojan pocket of “гълчъ” dialects is succeeded to the east by a large body of “приказвам” dialects, which extends east- ward through the Stara Planina, encompassing Sevlievo, Gabrovo, Trjavna, Velikovo Tъrnovo, and Elena, thus reaching as far as Kotel. Farther to the south, the territory to the east of the large zone of “гълчъ” dialects centered in the Sredna Gora, Marica valley and Rhodopes, falls within the purview of БДА-SE. Since this volume of БДА provides no map for dialect equivalents of Bg приказвам, говоря, we must have recourse to other sources, which with a small number of notable exceptions provide little lexical information pertinent to our topic.27 Before proceeding to a summary of this remaining material, we note that considerations of general dialect continuity render it fairly likely that “гълчъ” occurs at least to the immediate east of Karlovo, Asenovgrad, and the upper course of the Arda river (in the southcentral Rhodopes). On the other hand, it is equally probable that “думам” continues to prevail in the extreme south (along the upper course of the Vъrbnica) and fairly likely that reflexes of “хортувам” predominate in some of N. Thrace. Beyond the geographical scope of these projections into the uncharted domain of БДА-SE, we have the direct evidence of other sources which indicate the presence of “гълчъ” in Rup dialects, albeit those 25 For a more detailed discussion of the Pavlikjan dialects, which occupy a special position in Bulgarian dialectology, see below. 26 In the course of fieldwork, Schallert noted the occurrence of “гълчъ” as a neutral verb of speech (‘speak, talk’) in the Široka Lъka dialect of Stojkite (гô ́lча) and in the Ardino dialect of Drenka (not listed in БДА-SW or SE) (галчи́́́ш or perhaps гал’чи́́́ш). 27 With the exception of Strandža (in extreme SE Bg) (Горов 1961) and the S. Thrace dialects of Dervent (Бояджиев 1970, 1979) and Sъčanli (1971, 1972) , there are no available published monographs devoted to dialects which fall within the quadrant of БДА-SE. In ad- dition, the published dialect decriptions which occur in article format are generally not very informative regarding lexical details. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 31 situated at some remove from the central Rhodopes, cf. as a neutral verb in the S. Thracian dialect of Gjumjurdžina (Sъčanli) and in the sense ‘scold’ in the extreme SE dialect of Strandža. We also find “гълчъ” with the meaning ‘scold’ in a remote Rup colony (cf. N. Pavlikjan) near the Danube in the vicinity of Nikopol. This scattered SE Bg material can be further supplemented with the evidence of certain Bulgarian émigré dialects in Ukraine, although none of these are of the classic Rup type (see below for discussion, as well as the list in the immediately following paragraph). Otherwise, there is also evidence for “ороти”, “урто(в)а” as the basic verb of speech in at least one S. Thracian dialect (Dervent). (4) According to БДА-NE, “гълчъ” serves as the neutral verb of sppech in a tiny number of N. Mizija o-dialects (cf. 2 villages of the Razgrad “Kapanci” dialect and 3 villages of the Sъrt dialect). This sparse evidence of relic status can be supported with important testimony from a variety of other sources. First, the occurrence of “гълчъ” as ‘scold’ or ‘talk loudly’ is documented in autochthonous o-dialects in Silistra, Pre- slav, and Šumen. Second, Mladenov has described émigré dialects in Romania where “гълчъ” occurs as a basic verb of speech among Mizija dialect speakers in villages near Bucharest (see 2.2.3.5). Third, among Bulgarian émigré dialects in Ukraine (see 2.2.3.5), we find further support for the erstwhile existence of “гълчъ” among o- dialects speakers in E Bg, cf. an unglossed occurrence in the Kriničnoe Sъrt dialect, neutral usage in the Suvorovo East Thracian dialect and the Loščinovki “Thracian” dialect, and marked (‘scold’) usage in the Ol’šane dialect. Finally, an unglossed oc- currence of “гълчъ” is also attested in émigré dialects of the Čijšija type, which origi- nated in NE Bulgaria, but which are differentiated from both the Mizija and Stara Planina types (see 2.2.3.5). 2.2.2 Dialect-internal variation involving *gъlčěti In general, the evidence of the БДА indicates that dialect-internal variation be- tween “гълчъ” and other verbs of speech exhibits one of several distinct patterns. This picture of variation tends to be confirmed by the testimony of our other dialect sources. For the purposes of a provisional typology, these patterns of variation can be classified as either “binary” (involving two members) or “ternary” (involving three members), as follows. First, we find binary variation between “гълчъ” and “приказвам”. In this sort of situation, “гълчъ” is sometimes described in the БДА as “older” (traditional) and less “cultured” than “приказвам”. These characterizations suggest that the literary language is the primary vehicle for the spread of “приказвам”. An insight into the so- ciolinguistic dynamics of this variation can be obtained by considering the commen- tary for the Montana village of Meljane (NW pt. 639), where the basic entry is глчи́́́м, прикáзвам (по-ново) and the normative value judgment is apparent in Прикáзвам, прикáзвам óкаме, глчн’áта не е култу́́́рно (“Prikazvam, we call it (what you’re ask- ing about) prikazvam, (using) glčn’ata isn’t cultured”), even though we find straight- forward descriptions such as Они́́́ глчъ́́́т а йа слу́́́шам. As might be expected, the frequency of “гълчъ” in such dialects varies, depending on the degree to which it has receded before the standard form. Thus, in some villages, the traditional lexeme 32 Slovenski jezik – Slovene Linguistic Studies 6 (2007) occurs “occasionally” and in the speech of “older” informants, cf. Vraca (NW pt. 1319) прикáзвам; понякога у старите и глчи́́́ме си; Montana (NW pt. 588) глчи́́́м (по-стapo).28 Most commonly however, there is no indication of the relative status of the two variants and one must infer their relationship on the basis of systemic consid- erations, cf. Sъrt (NE pts.1889, 1895) прикáзвам ~ гълчá. Second, we find binary variation between “гълчъ” and another “traditional” ono- matopoeic Slavic verb. This pattern is found in the NW and seems to have developed due to dialect contact.2 Most frequently it involves competition with “бъбра” (cf. es- pecially W. Belogradčik), less commonly with “вревим” (cf. W. Belogradčik pts. 127, 55, 552, 555) The encroachment of one such verb upon the more neutral domain of the other may in fact be reinforced by the prior existence of the intrusive, “more expressive” verb in the lexicon. Third, one observes binary variation between “гълчъ” and a non-onomatopoeic Slavic verb of speech, such as *думам. This pattern occurs occasionally in the Rho- dopes (cf. Asenovgrad, Madan). It is quite probable that this kind of variation reflects the natural evolution of competition between an originally marked verb of speech on the one hand and a metaphorically derived (but unmarked) verb of speech on the other. Differences in valence patterns may also have contributed to the coexistence of these two verbs, since думам can easily be accompanied by an internal object дума (and in this respect is analogous to verbs derived from *kazati), whereas “гълчъ” does not display such governance. In some dialects, “гълчъ” is limited to its expressive status, whereas “думам” is neutral (e.g., the Ol’šane émigré dialect in Ukraine). Fourth, we encounter binary variation between “гълчъ” and a borrowed non- Slavic verb of speech which has been introduced within the context of the polyglot Ottoman Balkans, cf. “хорати-, хортувам”, “лафи-, лафувам”, or “куноштисвам са”. Since these verbs have been in circulation for much longer than “приказвам” (at least in its Literary Bulgarian form), it is not surprising that they have in some cases been more successful in gradually supplanting “гълчъ”, cf. Montana (NW pt.28) орáтим, глчи́́́м “много рядко”. Fifth, some dialects exhibit a more complex, ternary dynamic, whereby com- petition between “гълчъ” and another traditional onomatopoeic verb is overlaid by the recently introduced literary form, cf. Kula (NW pt. 2) глчи́́́м, бъ́́́брем; 5, 6, 7 глчи́́́ме, бъ́́́бреме (Най-новата дума е прикáзваме). Belogradčik (NW pt.122) глчи́́́ме, прикáзвам, бъ́́́брем; И си глчáле двáмата (“And the two of them were talking together”); Бъ́́́лгарски гълчъ́́́т (“They speak Bulgarian”);30 Седи́́́м си ту́́́ка и си гълчи́́́ме (“We’re sitting here together and talking to one another”); Стой мáнко 28 Note that when more than one speaker is involved, a plural form of “гълчъ” is frequent- ly accompanied by a reflexive dative pronominal particle (cf. седи́́́м си ту́́́ка и си гълчи́́́ме Belogradčik, NW pt. 122). This does not seem to be the case with “приказваме”. 2 For the partial settlement by Transitional dialect speakers of lower-lying river valleys between the western Stara Planina and towns such as Vidin and Lom, see the historical intro- duction to БДА-NW and Стойков 1968: 101. 30 This represents an interesting example of “гълчъ” used in the sense ‘to speak a lan- guage’. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 33 да поглъчи́́́ме (“Stay a bit so we can have a chat”); Такá ви прикáзвам (“That’s what I’m telling you”);31 Бъбрá с н’éга (“I/you/he/she talked with him”). In some instances, it is clear that one of the two traditional variants is now the dialectal norm, cf. Teteven (NW pt. 2307) глъчи́́́м, прикáзва, пубъ́́́брим (по-рядко). Sixth, analogous to the preceding type, but far rarer, is the ternary opposition of two onomatopoeic verbs and a loanword, cf. “гълчъ ~ бъбрем ~ ортувам” in Vidin (NW pts. 136, 137, 140). This typology of variation can easily be applied to other combinations involving dialectal verbs of speech. 2.2.2.1 Evidence for *gъlčěti in the Bulgarian Dialect Atlas (БДА) The attestation of “гълчъ“ in the meaning ‘приказвам, говоря; talk, speak’ («Как се казва в говора: говоря с някого или вревим с някого; гъгра, гълча, зборувам, приказвам, хортувам и др.» БДА-SW, p. 22)32 in the pertinent volumes of БДА is as follows.33 (a) БДА-SE (no map for dialect equivalents of ‘приказвам’); (b) БДА-NE (Map 282 ‘приказвам’, Commentary p. 134) – *гълчъ is attested in only 5 pts. (all of them Mizija o-dialects),34 cf. Sъrt 1892 *гълчъ ́́́, 88, 85 прикáзвам ~ гълчá; Razgrad , 2 орту́́́вам ~ гъл’чъ ́́́; 31 Note that this example is not strongly diagnostic, since the sense still seems to reflect the object-oriented valence pattern of “приказвам”. 32 Note that this is just the lead question, which serves as a heading in the questionnaire. In point of fact, it appears from some of the material provided in the commentaries to the БДА that the range of appropriate contexts in which any given basic verb of speech was recorded is considerably wider (cf. ‘speak a language’, ‘tell somebody something’), although it is unclear how extensively the volume of tokens actually recorded for different points might vary. 33 As a rule, data are cited in the commentaries to the maps in the БДА only when the forms recorded in the course of fieldwork differ in some phonological or morphological respect from the headform listed in the legend of the corresponding map, or when variation is observed either in the production of a single form or due to the occurrence of more than one lexeme. Thus, the amount of data provided in the commentaries varies considerably. Fortunately for our purposes, in the case of *gъlčēti the documentation in the commentaries is extensive for a va- riety of reasons, viz. a) the root lends itself to phonological variation; b) the verb is frequently attested in NW Transitional dialects with the non-typical -m sg. desinence; (c) the verb is frequently attested in the pl. form; d) the verb is occasionally cited in a preterite form. In the present study, we also generate asterisked (*) forms for dialect points not represented in the commentary, i.e., in cases where the data are evidently identical to the form cited in the legend of the corresponding map (cf. *гълчъ ́́́). 34 The Mizija “o-dialect” type is quite distinct from other NE Bg dialects. One of its most salient features is the occurrence of -o as the reflex of the strong jer in the masc. def. sg. desi- nence (e.g., градо < *gordъ‑tъ), in contradistinction to -ъ or -ъt in the vast majority of other NE Bg dialects. This feature was first identified by Miletič, who used it effectively as a primary diagnostic in his ground-breaking study of E Bg dialects and introduced the term “o-dialect” into Bulgarian dialectology (see Miletič 1903). Amongst other distinctive traits, certain Mizija dialects (chiefly those situated near Šumen) exhibit desinential stress in the def. sg. of nouns which elsewhere in NE Bg (and in literary Bg) show uniform root stress (see Schallert 1993). 34 Slovenski jezik – Slovene Linguistic Studies 6 (2007) (c) БДА-SW (Map 301, Commentary pp. 221–222) – *гълчъ is found in an east- ern band of dialects running from south to north from the central Rhodopes (Madan, Smoljan, Ardino, Široka Lъka, Devin, Xvojna), through Asenovgrad, Plovdiv, S. Pav- likjan, 2 pts. to the west of Pazardžik (cf. pts. 3614, 3618), and Karlovo, thus forming a continuum into Trojan in БДА-NW. With the exception of Karlovo, most of Plovdiv, most S.Pavlikjan points, a few points to the south of Asenovgrad, and pt. 3614 in Pazardžik, all of these dialects exhibit the classic S Bg (“Rup”) feature of initial stress in the sg. pres. form, where proclisis has been generalized in most verb classes, re- gardless of Common Slavic accentual paradigm. Some of these same dialects add -m to the sg. vocalic desinence (cf. -em, -am), while retaining the first-syllable stress of the older form. Forms with the desinence -im are to be interpreted as pl. (e.g., гълчи́́́м, бъ́́́брим), except in the case of Transitional dialects in Montana, Berkovica, Lom, and Kula, where they represent the sg. The data are presented here by dis- trict, proceeding roughly from south to north, cf. Madan 4664 гáлч’е̥м; 4673 лáфе̥м, ду́́́ме ̥м, гô ́л’че̥м; 4676 гốл’че̥м, лáфе̥м; 4685 гốл’чем, ду́́́мам; Smoljan 4635(a?) гал’чи́́́м (pl.?),35 4643 гô ́лчем; 4647 гал’чи́́́м си (pl.?), 4660 гô ́л’чем; Ardino 4648 гал’чи́́́м си (pl.?), 4650 гô ́лчам; 465 гъ́́́лчам; Široka Lъka 4195 гô ́lча; 48 гô ́лча; 4630, 4631, 4632 гô ́лча; Devin 4141, 4143 гô ́л’чем, куношти́́́свам са; 444 гáл’че; 44 гъ́́́л’ча; 45 гъ́́́л’чe̥м; 452 гóлчам; 42 гóлчам; 4620 гáлчем; Xvojna 4169, 470, 47 лáф’ам, гъ́́́л’ч’ем; 484 *гълчъ́́́, *лафóвам; 485 *гълчъ ́́́; Asenovgrad 3703 гъ́́́л’ча, прикáзwам; 3708 гълчъ ́́́, прикáзвъм, урту́́́въм 4173, 4174 гъ́́́л’чам; 475 гълчъ ́́́, хуртóвъм; 477 гъ́́́л’ча; 478 хортóвам, гáлчам; 47, 480, 48 гълчъ ́́́; 482 гъ́́́л’чам, лафóвам; 4183 *гълчъ́́́; 48 гô ́лчам; 44 ду́́́мам, гал’чи́́́ме; 420 гъ́́́л’чам, лафóвам; 4202 гъл’чи́́́м, лафóвам’е; 4203 гъ́́́л’чам, лафóвам; 4204 гъл’чи́́́м, лафóвам’е; Peštera 3633 прикáзувам, гъ́́́л’чам; 4138 гъ́́́л’чам; Pazardžik 3614 глъчи́́́м, прикáзуваме си; 3618 гъ́́́л’ча; Plovdiv 3007, 3008 гълчи́́́м; 3028 *гълчъ ́́́, *прикáзвам; 3041, 3043, 3044 *гълчъ ́́́; 3029, 3030 гълчá; 3032 гълчи́́́м; 3038 гълчи́́́м; 3048 гъл’чи ́́́, прикáзувъм (по-рядко); 3053 *гълчъ ́́́; 3648 гълчи́́́м, прикáзуваме; 3651 *гълчъ ́́́, *прикáзвам; 3653 галчъ ́́́; 3654, 3655, 3656, 3658 *гъл‑ чъ ́́́; 3657 гълчи́́́м; 3660 гълчъ ́́́, прикáзуам; 3661 гълчи́́́м, прикáзвъм (ново); 3671, 3675 гъ́́́лча; 3677 гълчá; 3692 гълчáх; 3702 гъл’чи́́́м; 465, 466 гô ́л’чам; S. Pav- likjan 3028 *гълчъ ́́́, 3035 гълчъ ́́́, прикáзуъм; 3037, 3040 3040а *гълчъ ́́́; 3665 гъ́́́лча; 3717 гъ́́́лча; Karlovo 2368 гъл’чи́́́м; 3000, 3004 гъл’чи́́́м; 3012 гъчи́́́м; 3014 бъ́́́брим, гълчи́́́м (по-старо); 3017 *гълчъ ́́́; 3023, 3024 (not underlined on Map) гълчи́́́м си; 3025 гълчи́́́м; 3026 (not underlined on Map) гълчи́́́м си; 3042 гълчъ ́́́, прикáзвъм (по- ново); 3046, 3048 гъл’чи ́́́, прикáзузвъм (по-рядко). (d) БДА-NW (Map 377 ‘приказвам’ p. 242) – *гълчъ occurs in the SE in Trojan and in a band of western dialects in Berkovica, Belogradčik, and Kula (often of the Transitional Bg-Serbian type), as well as in isolated points in Svoge to the N of So- fia), cf. (from SE to NW) Trojan 2308 гълчи́́́ме; 2336 прикáзвъм, гълчъ ́́́; Loveč 1526 прикáзвъм, гълчъ ́́́; Teteven 47 (Galata) бъ́́́бра, гълчá, прикáзвам; 2307 глъчи́́́м, прикáзва, пубъ́́́брим (по-рядко); Botevgrad 1405 гълчи́́́м; Svoge 1352 прикáзвам, погл̥чáаме (си); 2173 гл̥чáт; 280 глъчи́́́ме си; 2247 прикáзваме, гълчи́́́ме; Vraca 35 Although forms in -im (si) are not indicated as pl. in the БДА commentary, they are interpreted here as such on the basis of the suffixal vowel and end stress. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 35 64 прикáзвам, гълчи́́́м; 1319 прикáзвам; Понякога у старите и гл̥чи́́́ме си; 1374 прикáзвам, гълчъ ́́́; 1397 прикáзвам, гл̥чи́́́м; 40, 402 прикáзвам, гл̥чá; 1347 гл̥чá; Montana36 583 гълчи́́́ми, прикáзвами; 584 гл̥чи́́́м; 585 гл̥чи́́́м, прикáзваме (ново); 588 гл̥чи́́́м (по-старо), орáтим; 5 гл̥чи́́́м, прикáзвам; 56 622 гл̥чи́́́м (по-често), прикáзвам; 627, 628 гл̥чи́́́м, прикáзвам; 631, 633 гл̥чи́́́м; 639 гл̥чи́́́м, прикáзвам (по-ново); прикáзвам, прикáзвам óкаме, гл̥чн’áта не е култу́́́рно; они́́́ глчъ́́́т а йа слу́́́шам; 28 орáтим, гл̥чи́́́м (много рядко); 1285 гълчи́́́м; Berkovica 640 глъчи́́́ме; 28 гл̥чи́́́м; 2 гл̥чи́́́ме; 1301 прикáзваме ‘разговаряме’, гл̥чи́́́м ‘говоря’; 1303 гл̥чи́́́м; 1304 прикáзвам, гл̥чи ́́́; 1306 гл̥чи ́́́; 1311 гл̥чи́́́м; 1312 гл̥чи́́́ме; 1314 гл̥чи́́́ме; 1315 гл̥чи́́́м; Belogradčik 77 бъ́́́брем, гълчи́́́м; 6 гълчи́́́м;  гл̥чи́́́мо; 22 гл̥чи́́́ме, прикáзвам, бъ́́́брем; И си гл ̥чáле двáмата; бъ́́́лгарски гълчъ́́́т; седи́́́м си ту́́́ка и си гълчи́́́ме; стой мáнко да поглъчи́́́ме; такá ви прикáзвам; бъбрá с н’éга; 123 гл̥чи́́́мо си; 24 гл̥чи́́́мо; 25 гл̥чи́́́мо си; 26 гл̥чи́́́мо; 27 врéвим, гл̥чи́́́м; 28 гл̥чи́́́мо; 2 глъчи́́́мо, бъ́́́бремо, прикáзваме; 42 бь́́́бреме, гълчи́́́ме, прикáзваме; 143 (as per pt. 22) гл̥чи́́́ме, прикáзвам, бъ́́́брем; И си гл̥чáле двáмата; бъ́́́лгарски гълчъ́́́т; седи́́́м си ту́́́ка и си гълчи́́́ме; стой мáнко да поглъчи́́́ме; такá ви прикáзвам; бъбрá с н’éга; 47 гълчи́́́ме, прикáзвам, бъ́́́брем; 4 гл̥чи́́́м; 55, 552 врéвим, гл̥чи́́́м; 554 гл̥чи́́́м; 555 врéвим, гл̥чи́́́м; 560 гл̥чи́́́м; 566 гл̥чи́́́мо; 570, 574 гл̥чи́́́м; 578 врéвимо, гл̥чи́́́мо, бъ́́́брам; 57 гълчи́́́ме; Lom 46 бъ́́́брем, гълчи́́́м, прикáзвам; Kula 2 гл̥чи́́́м, бъ́́́брем; 5, 6, 7 гл̥чи́́́ме, бъ́́́бреме (Най-новата дума е прикáзваме); 2 гл̥чи́́́ме, бъ́́́бреме (Най-новата дума е прикáзваме); 5 гл̥чи́́́м; Vidin 114, 117 гл̥чи́́́м; 135 бъ́́́брем, гл̥чи́́́ме; 136, 137 бъ́́́брем, орту́́́вам, гл̥чи́́́м; е там бъбрá с една; орту́́́вам кво ли не; гълчи́́́мо си; 139 бъ́́́бреме, гълчи́́́ме, прикáзваме; 40 бъ́́́брем, орту́́́вам, гълчи́́́м ~ гл ̥чи́́́м. In addition, there are several points in БДА-NW where the verb гълчъ occurs with the meaning ‘shout loudly’ (‘викам високо’, see Map 378, cf. Montana 655, 656; Vraca 1411, 1414; Elin Pelin 2263). All of these villages are located within the immediate or proximate vicinity of points where “гълчъ” is attested as a neutral verb of speech.37 (e) БДА-Aeg (Map 215 ‘приказвам’) – *gъlčěti is not attested. (f) БДА-Trans (Map 312 ‘приказвам’) – *gъlčěti unattested. 2.2.2.2 Evidence for *gъlčěti and other neutral verba dicendi in other sources for Bulgarian dialects The survey given here covers only a representative sample of sources for Bul- garian dialects. The focus is primarily upon geographically defined idioms for which lexical material has been collated in dictionary form, although reflexes of “vocalic” liquids and the relevant sections on verbal morphology were also consulted in a vari- ety of other sources. Wherever possible, the material is correlated with the evidence of БДА. 36 At the time of publication of БДА-NW, “Montana” was still named “Mixajlovgrad”. 37 For unglossed examples of *gъlčěti in isolated points in Montana and Vraca, see our discussion on p. 42 of data from Toдopoв, 1936. 36 Slovenski jezik – Slovene Linguistic Studies 6 (2007) The macro-dialect affiliations here are as per Stojkov (Стойков 1968), cf. NE “Balkan”, NE “Mizija” (o-dialect type), SE (“Rup”), SW (where the critical feature is *ǫ > -a-, cf. даб, together with ekavian reflex of jat’), NW where the critical features are *ǫ > -ъ-, together with ekavian reflex of jat’), Transitional (where the critical feature is *ǫ > -u-, as in Serbian). Rup dialects can be further subdivided into Western Rup (Razlog, Goce Delčev), various Rhodope dialects, Strandža, and North Thracian, as well as dialects situated outside of Bulgaria (cf. eastern Macedonian Aegean/Ser- Drama, South Thracian). 2.2.2.3 Northeast dialects (Mizija type) Silistra “Grebenci” (primarily the village of Požarevo) (Кочев 1969). When de- scribing phonological differences in the st sg. pres. form, Kočev (112) cites гъл’чé ‘укорявам, мъмря’ as characteristic of the Silistra region as a whole. Since БДА-NE indicates приказвам as the neutral verb of speech for most dialects of the Silistra re- gion, including Požarevo (NW pt. 42), we may infer that *gъlčěti has been restricted here to its marked usage. Map 6. Distribution of *gъlčěti‑ in the neutral meaning ‘приказвам, говоря’ in Bulgarian 1. Kapanci (o-dial.). 2. Sъrt (o-dial.). 3. Belogradčik, Berkovica (some Transitional dial.). 4. Svoge. 5. Ixtiman (Младенов 1967). 6. Trojan, Sevlievo. 7. Karlovo. 8. S. Pavlikjan, N. Ce. Rhodope. 9. Sъčanli (Бояджиев 1971). All as attested in БДА, except 5, 9. Émigré dialects in S. Ukraine: Loščinovki (Fakija-Thracian) and Suvorovo (Mizija Zagorci). Mizija émigré dia- lect in Romania (Младенов 1993). Istanbul-Šumen Fakija émigré dialect of Čanakča (Денчев 7). J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 37 Mizija (émigré dialects in Romania) (Младенов 1993). Mladenov documents the presence of *gъlčěti, as well two other neutral verbs of neutral speech, cf. гълчи́́́м ‘приказваме, разговаряме’ (Vъrъšt) 315; ду́́́мъм ‘говоря, казвам; мисля, пре- полагам’ (provenance not indicated) 315; урту́́́въм (Vъrъšt, Kiseletu, Cherъšt) ~ хурту́́́въм (Vъrъšt, Knjaža., Milošešt) ~ фурту́́́въм (Epurešt Sъrb) ‘говоря, разгова- рам’ 315. Note that the shades of meaning for “хортува” overlap partially with each of the other two verbs. The coexistence of continuations of *gъlčěti and *dumati re- calls the situation in S-Ce rhodope dialects and seems to be maintained at least in part by subtle differences in meaning. Of particular note is the fact that “думам” continues to retain its original meaning (i.e., ‘think, suppose’) alongside the metaphorical one. This is unusual, in view of the predominance of *mysliti in South Slavic. Popovo (Palamarca) (Miletič 1903). Miletič cites only the noun głъ́́́čkъ 93. Preslav (Imrenčevo) (Попиванов 1940b). In the dialect lexicon which accom- panies his detailed “supplement” to Miletič’s original description of the Mizija o- dialect, Popivanov adduces гъ́́́лчъ, -иш (actually probably *гълчи́́́ш) ‘говоря високо, мъмря, укорявам’ 380. Compare БДА-NE (pt. 1803), where the basic verb of speech is represented by wурту́́́wъми (*xortuva-). Razgrad (Kapanci) (Младенов 1985). In a brief, but highly informative listing of distinctive Kapanci dialect vocabulary, Mladenov (336) cites гъл’чъ́́́ ‘карам се, давам наставления, напътствувам’, with the following examples Ас кът съм и мáйкъ йъ гълчъ ́́́ (‘Since I’m her mother, I’m scolding/instructing [?] her’), Гъл’чъ́́́т мъ синувéте (‘My sons are telling me what to do’). According to БДА-NE, орту́́́вам is the basic verb of speech in Razgrad dialects, although two “Kapanci” villages (Senovo , Krivnja 2) exhibit гълчъ́́́ as a variant. Razgrad (Xъrcoi) (Ангелова 1931). Angelova makes no mention of *gъlčěti in her comparatively brief list of reflexes of vocalic */l ̥/ (p. 139), nor in the examples for palatalized -л’- (ibid.). Šumen, Sъrt (Miletič 1903). Miletič makes no reference to *gъlčěti in his de- scription of the reflexes of vocalic */l ̥/, nor in the verbal morphology of the Šumen Sъrt dialect (for głъ́́́čkъ, cf. Popovo, above). Kočev, however, adduces the 1st sg. pres. form гълчъ́́́ for Šumen in the contrast to the Silistra type гъл’чé (see above). It is not quite clear whether the pejorative gloss noted above also applies to the example from Šumen. Sъrt émigré (Northern Dobrudža) (Gorna Čamurla) (Добрева 1986). In the ex- cerpts from Dobreva’s dissertation which we consulted, the following example was cited to illustrate the simple “statement of fact” narrative aorist — То́́́й дwа́́́ ча́́́съ гъл’ча́́́ (3sg. aor.), ас купа́́́ў (1 sg. aor.) 239. The meaning appears to be neutral, i.e., ‘He talked for two hours, while I hoed’. 2.2.2.4 Northeast dialects (Stara Planina type) Elena (Петков 1974). Petkov’s extensive dictionary (ca. 170 pp.) contains no entry for *gъlčěti, *bъbri-, *velěti, *vrěvěti, but provides evidence for two neutral verbs of speech (*dumati, *xortuva-) and one (*gъgre-), which also can convey a 38 Slovenski jezik – Slovene Linguistic Studies 6 (2007) more intimate, deprecatory tone,38 cf. ду́́́мъм ‘говоря, казвам’, Ду́́́мъм му дéйди зъ дървá (‘I’m telling him to go for wood’) 37 (note that this sentence actually exempli- fies the meaning of Bg казвам); ‘наричам, именувам’ Мéнъ мъ ду́́́мът Стуйéне 37; cf. noun ду́́́ма, Пу́́́штъм ду́́́ма (‘slander’) 37; hурту́́́wъм ‘приказвам, говоря, съветвам’, Слу́́́шъй мъ къкó ти hурту́́́wъм нéма дъ збъ́́́ркъш (‘Listen to what I’m saying and/so you won’t get confused’) 163; гъ́́́гр’ъ ‘говоря, разговарям; бъбря’, Сигá нидéй ми гъ́́́гр’ъ, шъ спе 30; гъ́́́грим се ‘говорим си, бъбрим си’, Чи и́́́двъй, кугáту си хъйлáк дъ си гъ́́́грим (‘[S]he’ll come by when [s]he gets’ an urge to have a chat’) 30. Also of particular interest is the verb збуру́́́wъм, which has two mean- ings, cf. 1) ‘на сбор съм, на панаир’ (farmer’s market, fair); 2) ‘говоря много’, e.g., Сти́́́га збуру́́́wъ, дай рéт и нъ дру́́́ги да кáжът нéшту 44. The second meaning here is remarkable inasmuch as the collective testimony of БДА and our other sources indicates that “зборувам” does not occur as a verb of speech (marked or otherwise) outside of Western dialects. One fundamental feature which distinguishes the Erkeč dialect from neighboring Stara Planina and Mizija dialects alike is the reflex -ǽ- for both *ъ and *ą. If in fact not related to the generally accepted etymological origin of “зборувам” as a “Balkan” verb of speech, perhaps this use of the verb here is a secondary development involving metonymic extension from the primary meaning of “зборувам“ in the local dialect (i.e., ‘attending the market; marketing’ > ‘bargaining’ > ‘talking a lot’). Erkeč (Delčeva, unpublished ms.).39 Delčeva’s recent study of the Erkeč lexicon lists gal’čǽ́́ ‘to nag’, Má́́jka gal’čé́́še; prí́́kazka ‘speech, dialect’, cf. Prískata sǽ́́štata, nosíeta – sǽ́́štata ‘(They have) the same dialect (way of speaking) and the same clothes (as us)’; “Uló́́u” dú́́ma sigá́́ (‘Now he says he means tin’ or perhaps ‘Now he’s calling it ulou’); “Urki” mu dú́́mame (‘We call the evil eye urki’, cf. *u‑roki); Šamšír mu ká́́zwam’e; brus mu wí́́kam’e (‘We call a whetstone brus’). There is also one ex- ample of *govori-, cf. zad garbǽ́́ ni govó́́r’at. БДА-NE indicates forms of *prikazvam in all 27 Erkeč villages (with variants in *xortuva- for only two points). Erkeč (Георгиев 1907; Стойков 1956). Neither Georgiev, nor Stojkov make any reference to *gъlčěti in their brief remarks on the Erkeč lexicon. The verb is also absent in Georgiev’s list of reflexes of vocalic */l̥/ (p. 84–85), whereas говор’ǎ is cited elsewhere (p. 7). Karlovo (Vojnjagovo) (Ралев 1977). Ralev’s dialect description includes an ex- tensive lexicon (pp. 02–8), which cites the following forms of *gъlčěti: гълчъ ́́́, гълчи́́́ш, гълчáл (sans gloss) 40; гълчъ ́́́ ‘говоря, приказвам’ 115, Кóлчим (колкото пъти) съ срéшнът, виздéн гълчъ́́́т и ни съ нъгълчáвът – ц’áлу силó жъ уду́́́мът; ‘съвятвам някого’ 115, Гълчи́́́ гу дъ съ връ́́́штъ нъврéми, чи се пу сукáцити скитóри тва въџи́́́шку40 йтé ‘Advise (tell) him to get back on time, ‘cause that damn kid likes to wander around the streets’ 5. Of special interest here is the use 38 Note that somewhat to the east of the Elena dialect, *gъgre- is used as a neutral verb of speech in three o-dialects in the vicinity of Kotel. 39 The Erkeč dialect exhibits a series of accentual, phonetic, and lexical traits which tend to associate it more closely with Rup or Mizija dialects than with other dialects of the Stara Planina type, to which it is traditionally assigned. 40 From важди́́́я, a regionalism of obscure origin (see БЕР 1: 111). J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 39 of the Slavic Aktionsart formation in nagъlčavati sę, for which the perfective is also cited, cf. нъгълчъ́́́ съ (as an example of “извършаване на действието до насита”). Cognate nominal forms are also adduced, cf. глъч, глъ́́́чкъ 2. Not attested are *du‑ mati, *duma; *vrěvěti; *velěti, *govoriti, *lafuva-, *sъboruva-. There is an entry for *prikazova-, *prikaza-, cf. прикáзоŏм, прикáă ‘приказвам; одумвам (i.e., ‘slan- der’)’, but it is interesting that the only example provided illustrates the specifically pejorative meaning, with direct object governance, cf. Ни нуси́́́ тоо ̆ гуч сукмáн, ътъ ни прикáзоо̆ ц’áлту силó 56. The secure position of *gъlčěti as a neutral verb of speech in the Vojnjagovo dialect is confirmed by the testimony of БДА-SW, which indicates гълчъ́́́ without variant for pt. 3017. Kazanlъk (Enina) (Китипов 1970). Kitipov’s relatively detailed lexicon (pp. 07–52) contains no entries for *gъlčěti, *dumati, *bъbre-, *lafi-, but documents урту́́́въм ‘говоря, убеждавам’, Урту́́́въм, ъмъ кóй мъ чу́́́въ 143, as well as the cog- nate fem. noun урътъ́́́ ‘събаране не жените на улицата на приказки’ 143. On the other hand, the Turkish loanword laf and the onomatopoeic base *bъbr- are repre- sented only by their nominal forms, cf. respectively лаф ‘дума, говорене’, Къжи́́́ иди́́́н лаф 27, and Óрътъ ве́́́че ста́́́нъле, пу пъ́́́т’ът съ чу́́́въ бъ́́́бър (‘the murmur of voices’) 0. Since the village of Enina is situated in the quadrant of БДА-SE (cf. pt. 246), Kitipov’s data thus shed light on a dialect which is not represented in БДА for this question. According to БДА-NE, dialects immediately to the north in Gabrovo and Trjavna show a preponderance of “приказва”, but a small number of points to the immediate north of Trjavna exhibit урту́́́въм, just as in Enina (cf. pts. 248, 2427). Kotel (Петров 1911). Petrov’s exceedingly brief lexical notes include no refer- ence to *gъlčěti, cf. rather the noun хурăтъ́́́ ‘събрани няколко души, за да работят и праказват’ 222. Nikopol (émigré dialect in Romania) (Младенов 1993). Mladenov cites the two variants уръти́́́м, урту́́́въм ‘говорим, приказвам’ 238. In БДА-NW, we also find both of these forms in the vicinity of Nikopol, sometimes in variation with “приказ‑ вам”. Pleven (Trъstenik) (Евстатиева 1971). Evstatieva’s detailed dialect lexicon (pp. 151–243) attests *gъlčěti only in its pejorative sense (‘scold’), cf. гъăчъ́́́ ‘карам се’, Не гъăчи́́́ детéто ма 65. At least one other semantically relevant verb of speech (*xoroti-, *xortuva-) is represented in a more neutral meaning, cf. оротъ́́́, орту́́́вам ‘приказвам, говоря’, Двéте орати́́́а (‘talking, conversing’) край врáтникъ 204. In addition, one of the glosses for “думам” suggests that it can also be used in a similar meaning (cf. ду́́́мам ‘казвам, говоря’ 169), although the examples cited take an ex- plicit or implict dicendum, cf. Аз му ду́́́мам (‘am telling’), ма кой мъ слу́́́ша 6, Кви́́́ грóзни ду́́́ми му думá (‘spoke, said’) 6. One of the derived nouns also indicates a more general meaning, cf. ду́́́ман’е ‘говорене’ (as well as the customary ду́́́мъ ‘дума, слово’) 169. БДА-NW Trъstenik (pt. 725) shows only прикáзвам (without variation or commentary). Popovo (Braknica) (Денчев). Denčev’s limited lexical selection of “distinctive” vocabulary reveals no examples of “гълчъ”. This village is not incoroporated in БДА- NE. 40 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Sevlievo (Krъvenik) (Ковачев 1970). The entry for *gъlčěti in Kovačev’s lexi- cal material (pp. 8–52) indicates that this verb is used in both neutral and pejorative senses, cf. гълчъ́́́ ‘говоря; мъмря някого (scold)’, Мáмъ мъ гълчи́́́ 6. There are no listings for *dumati, *bъbre-, *gъgre-, *xortuva-, or *lafuva-, although we do find лаф ‘дума, разговор’, ут лáф нъ лáф и мръ́́́кнъ 2. БДА-NE indicates only при‑ казвъм for Krъvenik (pt. 2396), perhaps since only the pejorative sense of “гълчъ” was detected by the investigators. Teteven (Стойчев 1915). Stojčev’s extensive dialect dictionary contains no entry for *gъlčěti, but the verb is cited in the sections on verbal morphology (cf. гл́́́ча [un- glossed], глч’áх, глч’áл, глчáн, глчи́́́ imv. глчá‑шта 93) and the reflexes of vocalic */l̥/ (cf. глчи́́́ш Tet. ~ галчи́́́ш G. Željazna, глчáва Tet. ~ галчáва G. Žel. 17). It is possible that the presence of *gъlčěti in the Teteven dialect has contributed to the cre- ation of what appears to be a hybrid formation with *glъtati ‘swallow’, cf. гл́́́кна, -еш ‘да престане да плача от крайно истощаване и омал’аване на гласът’ 252. Since БДА-NW indicates “гълчъ” as a neutral verb of speech in one village with a Teteven ế-dialect (pt. 2306) and one immediately adjacent village (pt. 2307), it is possible that Stojčev’s verb also could convey such a meaning. Trojan (Ковачев 1968). Kovačev’s dialect dictionary (pp. 161–243) includes the entry гълчъ ́́́ ‘приказвам, говоря’ 197, as well as гъ́́́гр’ъ ‘приказвам’ 197, Сти́́́га сти гъ́́́гръли, ъми спéти 197. In addition we find the noun лаф ‘дума, разговор’, Дъ си ръзбирéм уд лáфът сигá 22, Събрáли съ нъ лáф муъбéт (cf. Tk laf muhab‑ beti) 22. There are no entries for *bъbre- (but cf. бъбри́́́цъ ‘човек, който приказ- вам много’) 192), nor for *duma-, *xortuva-/*xorati-, *zboruva-. The presence of “гълчъ” in neutral usage is consistent with the testimony of БДА-NW for a cluster of villages surrounding the town of Trojan, although this same source cites no vari- ants with “гъгре” for Trojan (note that such forms are not cited only in two villages to the south, near Karlovo, see БДА-SW 3011 бъ́́́брим си, 3014 бъ́́́брим си, гълчи́́́м ‘по-старо’)’. According to Stojkov (Стойков 1968: 73–74) Trojan is one of the few regions where the “old Bulgarian” population has not been replaced by settlers from other parts of the Stara Planina. Sevlievo, Trojan, Teteven (Попиванов 1962). In his brief summary of the dis- tinguishing characteristics of these central Stara Planina dialects, Popivanov (pp. 165–183) notes several distinctive lexical items, including гълчъ ́́́ ‘говоря’ 179, but does not indicate its specific provenance. Once again, the inclusion of Sevlievo in this domain is not consistent with the testimony of БДА-NW, but is supported by the testimony from the village of Krъvenik just cited above. Veliko Tъrnovo (town) (Гъбюв 1896). Gъbjuv’s list of 13 roots with vocalic */l̥/ does not include *gъlčěti (p. 463). 2.2.2.5 Southeast dialects (Rup type) 2.2.2.5.1 West Rup Goce Delčev/Nevrokop (Мирчев 1936). Mirčev’s discussion of reflexes of vocal- ic */l̥/ is of little diagnostic worth for our present purposes, since it contains only three J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 4 items, cf. укъ́́́лцăм, ж’ъ́́́лч’кă, сăлзи́́́т’е 36. Nor is there any reference to *gъlčěti in the comparatively brief, yet highly informative glossary. The latter, however, does include several notable verbs of speech with commentary, cf. вели́́́ 7 (used most frequently in the 3sg. pres. as a substitute for каже, кае), врé̥в’е ‘казвам, приказвам, говоря’ 123 (“Глаголът е много обикновен по говорите на цяла Източна Македо- ния”, an observation which echoes Šapkarev’s previous claim), збōруóвам ‘говоря’ 117. Mirčev’s citation of *vrěvěti in such a wide employ constitutes the easternmost attestation of this verb and is thus of historical significance for determining the SW delimitation of *gъlčěti at a stage prior to the collection of data for БДА, inasmuch as *vrěvěti seems to have served as an in-group, expressive verbum dicendi among speakers of dialects which lacked *gъlčěti for this purpose. This historical value of Mirčev’s testimony is even more apparent when we compare it to that of БДА-SW, which indicates primarily “приказвам” (SE Goce Delčev, i.e., mainly to the east of the Mesta river), “прикажувам” (most villages to the N of the town, including Mirčev’s Ognenovo, pt. 4563), “оратим” (4 pts. in the north, including Mirčev’s Baldevo, pt. 4560), “лафовам” (SW pts., including Mirčev’s Dělčevo, pt. 4556). Several villages in БДА-SW also attest “зборовам” (including Mirčev’s Tešovo, pt. 4588), but “вре‑ ви-” is cited only in 2 pts. (422, 425) of the southern Razlog type, situated farther up the Mesta valley beyond the northern limit of the Goce Delčev dialect region (in variation with “зборовам”) and in 1 pt. (4555) on the western fringe of Goce Delčev (in variation with “лафовам”). It would thus appear that in Mirčev’s time, the Nevro- kop (i.e., Goce Delčev) dialect operated with two colloquial verbs of speech (врé̥в’е and збŏрўóвам) and that this state of affairs also characterized outlying villages of the S. Razlog type (the latter were explicitly excluded by Mirčev from consideration for the Nevrokop dialect) and continues today in БДА‑SW pts. 422, 425 (whereas the core Razlog dialect pts. of the Razlog-Bansko basin exhibit only *дума‑). The poten- tial status of suffixed derivatives of *prikazati in Mirčev’s day can not be determined, since he makes no specific mention of them. Razlog (Bansko) (Молерови, Д. и К. 1954; Велянова 1997). There is no attes- tation of *gъlčěti in the list of 7 roots with vocalic */l̥/ which is found in the gram- matical introduction to the collection of folktales assembled by the Molerov brothers. The verb is also not listed in Veljanova’s monographic lexicon of the Bansko dialect, which tends to focus more upon distinctive nominal vocabulary. Razlog (Молеров 1905, Алексиев 1931). Neither Aleksiev’s description of the Razlog dialect, nor Molerov’s study of its “double” accentuation contain any refer- ence to *gъlčěti. Molerov cites sg. pres. бъ́́́рборе, за́́́бърбо́́́ре (sans gloss) 77. 2.2.2.5.2 Rhodope Central Rhodope (Стойчев 1965). The first and most extensive installment of Stojčev’s three-part “Rhodope Dictionary” (Родопски речник, see Стойчев 1965, 1970, 1983) is devoted primarly to the Central Rhodope region (see БД 2: 119–315), where we find “гълчъ” attested in a variety of local dialects with different reflexes of vocalic */l̥/, cf. гáл’чêм (Davidkovo/Ardino),4 (“Pas”, probably a misprint for 4 The corresponding pt. 420 in БДА-SE has -ъ- as the reflex of vocalic /l̥/. 42 Slovenski jezik – Slovene Linguistic Studies 6 (2007) “Pis”, = Pisanica/ Smoljan),42 Padina/Ardino 42; гô ́л’чêм, Зби́́́раме са вéчеро, та хми гốл’чем (Smoljan, Ardino, Asenovgrad, Manastir/Smoljan, Devin) 46; гъ́́́л’чêм (Orexovec/Ardino = SW pt. 4651, Bani Lъdža/Ardino = lacking in БДА, Zlatograd = SE pt. 4758, Pavelsko/Xvojna = lacking in БДА, Xvojna = SW pt. 4171) 149. Col- lation of two separate entries reveals a rare instance of variation between *gъlčěti and *dumati, cf. гáл’чêм 42, ду́́́мêм ‘говоря’ 155 (Xambardere, Ardino). Stojčev’s materials also provide rare evidence of the pejorative use of *xorati-/*xoratova-, cf. хóратĕм ‘говоря, приказвам’, but also pejoratively ‘приказвам лоши, неприятни думи, злословя’, Нимóй хŏрати́́́ саки́́́ва ду́́́ми (Smoljan, etc.) 26; хŏрат’óвам (= хоратĕм) (Smoljan) 26; fem. noun хŏратô ́ ‘лоша дума, клюка’, Нимóй кáзва гŏлê ́ма хŏратố (Smoljan, Ardino, Asenovgrad) 26. БДА-SW provides no evidence for *xorati-, etc, in CeRhod, except on the northern periphery, cf. Asenovgrad (pts. 475, 478). On the other hand, the wide attestation of *lafi-/*lafova- is consistent with the witness of БДА-SW, cf. лáфêм ‘говоря, приказвам’ (mainly among Bulgari- an Muslims), Нимóй ми лáфи, йунáче 8; лафóвам (Smoljan, Ardino, Asenovgrad, Madan, Devin) 8; cf. also the base noun лаф ‘дума, мълва, приказка’, Нимóйте прáви лаф! (Smoljan, Ardino, Asenovgrad, Madan, Devin, Mg), ‘обещание’, Дáли са лаф на мŏми́́́ченŏ 8. The verb *gъgre- is attested only in its expressive meaning, cf. гô ́гр’ам ‘дърдоря, бръщолевя’ (Slavejno.) 145. There are no entries for *vrěvi-, *zboruva-. Devin (Mugla) (Стойков 1970). Stojkov’s discussion of verbal morphology includes the following present tense paradigm of *vrěvěti *gъlčěti (sans gloss): йê га́́́лчем, ти галчи́́́ш, то галчи́́́, не галчи́́́ме, ве галчи́́́те, тê галчếт 66. Krumovgrad (Tixomir) (Кабасанов 1963). There is no listing for *gъlčěti in Kabasanov’s brief dialect lexicon. Smoljan (Momčilovci) (Кабасанов 1956). Kabasanov cites гô ́лча̣ ‘говоря’ 72, and also the noun гô ́лче̣ва̣ ‘шум, говор’ (Канева-Николова 2001: ). Kaneva-Niko- lova’s study shows that Kabasanov’s forms are still in use nearly fifty years later, cf. гô ́л’ча ‘говоря’162, Гô ́л’чa ̥м хи дḁ си утварa ̥ очи́́́нe̥ нa ̥ чéтри (here used in the sense ‘tell’); cf. fem. noun гô ́л’чa ̥вa ̥ ‘високо и шумно говорене’, Чу йе гул’ê ́ма гô ́л’чава 62. These examples are important since they show that a pejorative mean- ing in a noun such as “голчава” does not necessarily imply the existence of this meaning in the verb. Southeast Rhodope (Стойчев 1983), West Rhodope (Стойчев 1970). These in- stallments of Stojčev’s Rhodope dictionary contain no entries for *gъlčěti or *du‑ mati. 2.2.2.5.3 Pavlikjan Pavlikjan dialects (from the “Paulician” Catholic heresy) constitute a particular type within Bulgarian. They are commonly believed to have originated in the SE Rhodope region, but their present geographical distribution is as follows: (a) S. Pav- likjan (several villages in the general vicinity of Plovdiv and to its north); (b) N. Pav- 42 The corresponding pt. 4652 in БДА-SW has -ô- as the reflex of -ъ. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 43 likjan (near the Danube in the vicinity of Nikopol); (c) Banat (in SW Romania); (d) near Bucharest. The migration from northern Bulgaria to Romania was undertaken following the Čiprovec rebellion of 1688. At a subsequent date, some of the Banat Pavlikjans returned to Bulgaria, settling near Nikopol. South Pavlikjan (БДА-SW). Since no published material of any significance on S. Pavlikjan has appeared since Miletič 1912,43 for the sake of comparison we repeat the data from БДА-SW to demonstrate the use of “гълчъ” as a neutral verb of speech, cf. 3028 *гълчъ ́́́, 3035 гълчъ ́́́, прикáзуъм; 3037, 3040 3040а *гълчъ ́́́; 3665 гъ́́́лча; 3717 гъ́́́лча. North Pavlikjan (Неделчев 1994). In his lexical notes, Nedelčev cites the marked usage of “гълчъ” for ‘scolding’, cf. гълчи́́́ ‘кара се, мъмри’, whereas “хортува” is used in a neutral sense, cf. урту́́́въм ‘говоря, разговарям’ 362. Banat Pavlikjan (Стойков 1968). In his exhaustive dictionary of the Banat di- alect (more than 400 pp.), Stojkov makes no mention of *gъlčěti, but cites rather хурту́́́вам ‘изричам, произнасим думи’ Дитê ́ту хурту́́́ва на тъ́́́нку; ‘разговарам, приказвам’ Знай да хурту́́́ва маџê ́рск’и ‘уговарям, споразумявам се’ Дóйăт у мумъ́́́та съз гъдлáре и тугáз хурту́́́вăт гудéжă 265; dialect glosses кáзвам, ричъ́́́ 265; ду́́́мам ‘говоря; мисля, смятам’ pp. 74. The data for related nouns of speech are as follows: Banat хуртъ́́́ f. ‘дума’ 265, Влáшка хуртъ ́́́, Кък’é са хурты́́́те на тáзы песмъ́́́? Ни размê ́ва бъ́́́лгарск’ите хурты́́́ 265; ду́́́ма («нов.»), Туй и палк’ếнска ду́́́ма 74; рêч ‘говор, говорене; дума’ 205, dialect glosses ду́́́ма, хуртъ́́́ 205. Bucharest Pavlikjan (cf. Popešt-Leorden, Čoplja) (Младенов 1993). On the ba- sis of Mladenov’s lexical notes, we may conclude that the Bucharest Pavlikjan system of verba dicendi is similar to that of North Pavlikjan, cf. гълчи́́́ ‘кара се, мъмри’, урту́́́въм ‘говоря, разговарям’ 362. Comparative analysis of the differences in the four branches of the Pavlikjan dialect reveals several interesting facts. First, S. Pavlikjan (i.e., the branch which is situated closer to the putative homeland of the Pavlikjan dialect) has retained *gъlčěti, the Ce Rhodope ‘speak’ lexeme of Slavic origin. Second, the other branches of Pavlikjan reflect a later stage, in which *xortuva- has replaced *gъlčěti as the basic verb of speech, although N. Pavlikjan and Bucharest Pavlikjan have retained *gъlčěti in the sense ‘scold, mutter’, whereas this verb is now lacking in the Banat dialect. Third, since the Banat Pavlikjans emigrated to Romania after the Čiprovec rebellion of 688, thereby severing direct linguistic ties with their fellow Bulgarians, it would appear that the late 7th c. constitutes the terminus ad quem for the introduction and entrenchment of *xortuva- (this is consistent with the evidence of 7th-century Bul- garian vernacular Damascene texts from the West Central Stara Planina region, see 2.3.6). Fourth, at least the Banat dialect has at some point acquired or developed *dumati as a verb of speech, albeit one which warrants less attention than *xortuva-. Fifth, whereas *xor(o)ta serves as the basic word for ‘word, speech’ in Banat, *duma is deemed to be more recent. 43 Note that only pt. 3665 (General Nikolaevo) and pt. 3717 (Sekirovo) retain the charac- teristic “Rup” feature of initial stress in the sg. present form. 44 Slovenski jezik – Slovene Linguistic Studies 6 (2007) 2.2.2.5. North Thracian Plovdiv (Češnegivoro) (Попгеоргиев 1907) Popgeorgiev provides a detailed list of words containing reflexes of vocalic */l ̥/, but cites no forms of *gъlčěti. Elsewhere, we find the 3pl. aorist or imperfect form урту́́́ваă 425. This is consistent with the evidence of БДА-SW (pt. 3701), for which “уртува” is indicated as a neutral verb of speech. Xaskovo (town) (Димчев, Кювлиева 1970). The authors provide no entry for *gъlčěti, *dumati, or other potentially neutral verbs of speech, but do cite pejorative *gъgre-, cf. гъ́́́гр’ъ ‘постоянно гълча някого, постоянно досаждам’ 66, Сти́́́гъ си мъ гъ́́́грил! 66, as well as the noun лаф ‘разговор, приказни’ 78. Xaskovo (Nova Nadežda) (Христов 1956). Xristov provides no evidence of *gъlčěti in the discussion of reflexes of vocalic */l̥/, nor in the small lexicon (ca. 8 pages). 2.2.2.5.5 South Thracian (NE Aegean Greece and European Turkey) Dedeagač (Dervent) (Бояджиев 1970). According to the oral tradition of its orig- inal inhabitants (who now reside in Elxovo SE Bg), the village of Dervent was origi- nally founded by settlers from Kruševo, which is situated to the NE of Solun. This circumstance would serve to explain the presence of SE Mac features in the dialect, such as ръзбóй 237 and the masc. def. sg. desinence -ут.44 The E Mac origin of the dialect is also manifested in the form при́́́къзнъ ‘приказване, думи’. Thus it is not surprising that Bojadžiev’s brief contrastive dialect dictionary (pp. 223–245) does not list continuations of *dumati or *gъlčěti, although the absence of *vrěvěti and *velěti is somewhat unexpected. Instead we find урóт’ъ ‘говоря’, Т’а млóгу урóти 240, уртóъм ‘говоря’, Ти уртóвъй се̣гá 24. This verb appears to have been acquired due to contact with other Thracian dialects, a circumstance which would also explain the presence of the Rup -ova- suffix (as opposed to -uva- or -va-) in the variant уртóъм. The particular example cited for *prikaz(ъ)va- suggests that the dialect speaker may be adapting his speech to that of the investigator, cf. И ни́́́е въз вас (според вас) пáк тъкá прикáзвъме (‘According to you we also speak that way’) 225. Gjumurdžina (Sъčanli) (Бояджиев 1971). In his richly documented lexicon, Bojadžiev cites гъ́́́л’чъ ‘говоря, приказвам’ 20, Гъл’чъ́́́т пу д’ук’áнету (‘People are talking in the stores’), ‘ ́́́карам се’ От’ гъл’чи́́́ш децáтъ (‘Why are you scolding the children?’). Also of particular interest is the entry ду́́́мъм ‘говоря, казвам’ 25, Ми́́́рчу пумáцим ду́́́мъше (spoke to) (folksong; note dat. pl. of ‘Pomak’), especially when this verb is used in the sense ‘to speak a language’, cf. Ту́́́рцкуту с д’áте у́́́чът ту́́́рцку дъ ду́́́мъ 25, as well as its original sense ‘мисля’, cf. Йед’áл л’ап и си ду́́́мъл 25. Further meanings of “думам” in this dialect include ‘наричам, именувам’ Тъкá му ду́́́мъхъ; ‘клеветя, разправям’ Нъ пупрéл’лки, нъ межи́́́, нъ óру и нъ с’áкъде ги ду́́́мъли, чи съ сръвли́́́ви 25. In addition we find прикáзвъм (literary form), прикáзувъм (the 44 The EMac desinence *-ot occurs only in unstressed position, elsewhere we generally find -ът, which is the more common Rup ending. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 45 dialect form -ova- with vowel reduction?) ‘говоря’,45 Йа с’а прикáзвъм със тéбе, ъмъ н’áмъм тай ми́́́слъ дъ ти прикáзвъм 73; Н’áмъ къко дъ ви прикáзувъм 73; при́́́къскъ ‘говор, език; приказка’; прикáзувъне ‘приказване’ 73. While there is no entry for *govoriti, we do find the unusual impfv. formation рéкуъм, рéкувъм (from рекъ). Boajdžiev’s glossary thus provides important evidence for the usage of *gъlčěti as a basic verb of speech in SouthThracian dialects, notably those situated in close proximity to the southeastern Rhodopes. The semantic overlap of “гълчъ” and “думам” is only partial, since both verbs can mean ‘говоря’, but there seems to be a distinction between *gъlčěti ‘приказвам; talk’ (i.e., ‘occupy oneself with talking’) and *dumati ‘казвам; tell’ (i.e., ‘convey one’s intention, express one’s thoughts’). Lozengrad (Павлова, 1988). Pavlova’s discussion of accentual alternations in- cludes a reference to *gъlčěti, cf. гъ́́́л’чъ ~ гъл’чáх (sec. 4.2). Although the Lozen- grad dialect exhibits a mixture of SW (Fakija) and SE Bg traits (contrast respectively sg. йас with initial stress in both forms of the imperative зáпъли, зáпълите), the presence of palatalized -л’- in the forms of *gъlčěti is probably attributable to the presence of the Rup Strandža dialect element, an assertion which is consistent with the presence of *gъlčěti in the latter region (v.s.). Asia Minor (Mandъr) (Цанова 1986). Consultation of the chapter on accentua- tion in Canov’s dissertation yielded the following unglossed examples of *gъlčěti: sg. pres. гъ́́́lч’ъ 68, гъ́́́lч’ъ 85, 2sg. pres. *гъlч’и́́́ш’, sg. aor., sg. impf. *гъlч’а́́́h 85. In addition to these sources, we have consulted the descriptions of reflexes of vo- calic */l̥/ and verbal morphology for all the relevant chapters in Bojadžiev’s detailed, monographic survey and atlas of S. Thracian dialects (Бояджиев 1991), but have found no examples of *gъlčěti. 2.2.2.5.6 Strandža Our principal source for this dialect region (Горов 1962) provides important evidence for the areal domain of *gъlčěti (at least in its marked usage) to the east of the Rhodopes. In the lexicon (pp. 64–159) we find гъ́́́лче̣ ‘карам се’, Фáта̣й си рáбо̣та̣та̣ о̣д вр’ếме, че тáтко̣ ти ше те гълчи́́́ ‘do a bit of work once in a while or your father will give you a talking to (scold you)’ 78. In the section on verbal mor- phology, we have pres. гъ́́́лче, гълчи́́́ш, гълчи ́́́, imperf. гълчáх, гълч’ếше, aor. гълчáх, гълчá 35. Otherwise, the basic verb of speech appears to be хурăтóвăм ‘приказвам, говоря’, Друк пък’ дă знáĕш кăко хурăтóвăш ‘Next time you’ll know better what to say/how to speak’ 153.46 The generalized meaning of this verb is reflected in the cognate fem. noun хурăтă ́ ‘говор, диалект’, Йа не мó дă се с’ê ́те нă хурăтă ́тă нă 45 The dialect of Sъčanli is not a mixed one, therefore the presence of the literary vari- ant here can likely be attributed to the interaction between the investigator and the dialect speaker. 46 Note that in addition to the characteristic Rup -ova suffix (rather than -uva), the Strandža form also displays the rare (and etymologically correct) -a- vowel of the second syllable of the root (in contradistinction to the more widely attested forms with apocope or -o-, cf. “хортува” and “хоротува”). 46 Slovenski jezik – Slovene Linguistic Studies 6 (2007) зäгóрците ‘I can’t seem to understand the dialect those Zagorci speak’ 153. There are no entries for *vrěvěti, *dumati. 2.2.3.1 Southwest dialects Blagoevgrad (Padež, Leško) Стоилов 1904, 1905). Stoilov cites no examples of *gъlčěti in his general description of the dialect (05), nor in his study of its accen- tuation (04). Whereas Stoilov’s list of 20 ě/i-verbs does not include *gъlčěti, the list of i-verbs cites гово́́́ри (1905, p. 213). Botevgrad (Trudovec, formerly Lъžane) (Тодоров 1936). Todorov cites the in- triguing example поглчи́́́ 20 (sans gloss), but does not identify the form. The latter can nonetheless be determined by accentual criteria, which indicate that this is the imperative sg. rather than a form of the aorist singular.47 This appears to be a hypo- choristic usage, which has sheltered “гълчъ”. According to БДА-NW the correspond- ing pt. 2254 is a “сказвам” dialect. Botevgrad (Попиванов 1940a); (Илчев 1962). Ilčev’s small dictionary (pp. 185– 205) contains no entries for *gъlčěti, *dumati, or other remarkable verbs of speech. Botevgrad (Skravena) (Тодорова ). Todorova’s concise, informative lexicon (pp. 104–133) includes an entry for *gъlčěti in its marked usage, cf. глъ́́́ча ‘говоря високо’ 108. In БДА-NW (p. 242), Skravena (pt. 2252) shows прикáзвам, óрута. Fakija (SW Bg émigré dialects in SE Bg) > Čanakča (Istanbul/Šumen) (Ден- чев 1979).48 Denčev’s brief lexical material (pp. 343–349) includes гълчá ‘говоря’, Гълчá с момчéто си 344.4 Since the ancestors of Fakija dialect speakers most prob- ably originated from south of the Stara Planina (cf. Samokov and regions to the west thereof), this attestation is very intriguing, although we can not completely exclude the possibility of influence from the dialect of the town of Šumen, where Čanakča speakers occupy a small quarter. Ixtiman (Младенов 1967). Mladenov’s richly documented lexicon (pp. 3–197) includes глъ́́́ча ‘говоря, приказвам’, 51, Йа му глъ́́́ча, cf. also глъчóли се ‘говори се, шуми се’, Глъчóли се вáнка 5; глъчли́́́ф’ призказлив’ 51; глъч ‘глъчка, шум, врява’ 51; глъ́́́чка ‘викавица, шум’ 51. Although the Ixtiman dialect region is not 47 Lъžane/Trudovec is situated in the zone of “semi-retracted” stress (where all prefixed aorists are stressed on the root, see БДА-NW, map 86, where no variation is indicated for pt. 2254), but also in the zone of oxytonic stress in the imperative sg. (see БДА-NW, map 2, where no variation is indicated for pt. 2254). 48 Fakija dialects are situated in SE Bg to the west of the Strandža region. There is con- siderable linguistic evidence to support the ethnographic and historical indications that these dialects originated in SW Bg (ca. 6th c.), perhaps in the general vicinity of Samokov (see Кочев 1964; Schallert 2001). The Fakija region itself served as the base for colonization into southern Thrace (Lozengrad, Čanakča), where speakers of the Fakija dialect often came into close contact with speakers other dialects, then later emigrated to Ukraine (cf. probably Ter- novka), Dobrudža, and Šumen. 4 Oxytonic, rather than initial, stress in the sg. pres. form is not a characteristic SW Bg dialect feature and may reflect influence of the Mizija o-dialect of the town of Šumen, to which Čanakča speakers emigrated in the 20th century. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 47 marked for *gъlčěti in БДА-SW, one is loathe to challenge the authority of Mladenov, an eminent and highly observant dialectologist, who worked as a school teacher for many years in the Ixtiman region (see introduction to Младенов 1967). According to БДА-SW, for ‘приказвам’ N. Ixtiman dialects usually have forms of *оратим (rarely in variation with *приказвам), whereas S. Ixtiman dialects exhibit either forms of *бъбра or *бъбра ~ *оратим. Mladenov also lists forms of the latter two verbs (and cognates of *орати), cf. орáта ‘приказвам, говоря’ 126, Т’а и не орати́́́. Орати́́́ си каквó си ни́́́е орати́́́м, оратли́́́в ‘приказлив, общителен’ 126; оратá ‘разговор, приказване’, Оди́́́хме на оратá у ни́́́х ‘говор, език’, Такáва си ни е оратáта. Той се познáва по самáта оратá; бáбра ‘говоря’ 37, Бáбре му каквó и́́́ска; бáбре се ‘говори се’, На нóго местá се дру́́́гойче бáбре; бáбре си ‘разговарам с някого’, Сéдни и слу́́́шай каквó си бáбрем. Ixtiman (Dolna Banja) (Тодоров 1936). Not listed in БДА-SW, Dolna Banja is presently a small town located several kms. south of Ixtiman and to the east of Raduil (Samokov). Todorov cites глчи́́́ме 20 (without gloss). Kjustendil (Умленски 1965). There is no reference to *gъlčěti in Umlenski’s detailed monograph of the Kjustendil dialect in its regional varieties, cf. Kjustendil Pijanec, Polčane, Kamenica, Kraište (the last-named being a Transitional dialect). Kjustendil Polcane (Бояджиев 1932). Neither Bojadžiev’s lexicon (pp. 320– 329), nor his summary of reflexes of vocalic */l̥/ list *gъlčĕti. Makocevo (Стоянов 1972).50 Although Stojanov’s monograph-length article on the Makocevo dialect contains no reference to *gъlčĕti in the modest-sized lexicon, we do find unglossed forms of the verb in the detailed discussion of vocalic */l̥/ and verbal morphology, cf. respectively гл̥́́́ча, гл̥чи́́́ш 0 and гл̥́́́ча, гл̥чи́́́ш, гл̥чáх, гл̥чáне 24. БДА-SW (pt. 2288) indicates бáбра and óрота as the neutral verbs of speech. It is therefore likely that we are dealing here with marked usage of *gъlčěti. Pirdop (Smolsko) (Кънчев 1968). There is no entry in the detailed dialect dictionary (pp. 86–55) for *gъlčěti, although we do find глчáва, гл́́́чка ‘глъчка, кавга, шумни расправия’ 95. There are also no listings for *govori-, *veli-, *vrěvi-, *duma-, or *zborva-. Instead, the two most basic verbs of speech appear to be “бъ‑ бръ” and “орати”, cf. óрата, -иш ‘говоря’, Орати́́́ли си за млáдото врéме 28; оратá ‘говор, реч’, Оратá му е сé на кавгá 28; бáбра, -еш ‘говоря, приказвам, преговарям, обсъждам’, Ни́́́а вéче бабрáhме за телéто, ама тóй мáлко дáва 88. Other verbs of speech have more limited meanings, cf. прикáзувам, прикáжа, -еш (the formal equivalent of Literary Bulgarian приказвам, прикажа), which is used specifically to refer to a brief conversation held upon meeting someone (‘заговарям при среща, не отминавам мълком или само с поздрав’), e.g., Като ма срéшне, се ште са зáпре и ште са прикáже 136, whereas the noun при́́́каска has a wide range of meanings (‘реч, говор, начин на говорене’), Нему при́́́каската е сé на кавгá, ‘пословица, поговорка’ 136. As in Botevgrad, situated somewhat to the northwest, 50 Although not listed as a separate regional dialect in standard works on Bulgarian dia- lectology, the dialect of Makocevo as described by Stojanov in effect constitutes a transition between those of E. Sofia and Pirdop. 48 Slovenski jezik – Slovene Linguistic Studies 6 (2007) we also find скáзвам ‘разказвам’, Нáшата бáба чéсто ни скáзваше за чу́́́мата 42. Samokov (Вакарелска-Чобанова 2005). Vakarelska-Čobanova’s authoritative dictionary (pp. 17–392) cites *gъlčěti, but only in its expressive sense, cf. глъчи́́́м ‘говоря силно’ 67 (cf. дерем се, джаголим, джакам); Глъчи́́́, та гóра вéе; Днéска дóма глъчáхте нéшто 67. The neutral verb of speech is evidently орáтим (си) ‘говоря (си)’ (cf. also говóрим си, моабéтим си) 246, Нóгу оби́́́чам с вáзе да си орáтим, Орáтиме си сос нéа ци́́́гански 246, Васи́́́лко, ми ду́́́ма, не óрати такá 9б, although the dialect glosses would suggest that говóрим си and the adapted Turkism моабéтим си are equivalent in meaning to оратим. All told, this description is consistent with that of БДА-SW, where оратим is indicated as the basic verb for ‘приказвам’ throughout the Samokov region. Not cited by Vakarelski-Čobanova are *орати (predominant to the west according to БДА-SW), and *бъбре, which occurs to the east in Ixtiman. The verb *dumati functions as a verb of speech, but only in the sense of ‘tell, say’, cf. ду́́́мам ‘казвам’ 96, Васи́́́лко, ми ду́́́ма, не óрати такá, thus serving as a counterpart to *kazuvati/*kazati, cf. казу́́́вам (impfv.), кажем (pfv.) ‘казвам’ 146, Йа не казу́́́вам на мажó ми, да не óка. In addition, *dumati retains its primary meaning, ‘think’, cf. Ни се ду́́́мам, ни се разду́́́мувам ‘не искам да зная’ 6. Samokov (Raduil) (Ангелова 1948). In her listing of reflexes of vocalic liquids (pp. 310–312), Angelova cites (but does not translate) both гл̥чáт, глъчи́́́ли (sic!) and the nouns гл̥́́́чка, глъ́́́чка. БДА-SW (pt. 3555) indicates only “орáтим” as a basic verb of speech. Samokov (Govedarci) (Стойков et al. 1956). There is no attestation of *gъlčěti in the list of reflexes of vocalic /l ̥/ (p. 264), nor in the lexicon. Instead we once again find орáтим ’говоря’ 317 (thus also pt. 3519, БДА-SW). West Sofia (Dobroslavci) (Гълъбов 2000). Gъlъbov’s monumental dictionary of the dialect of Dobroslavci (pp. 59–835) contains no entry for *gъlčēti, nor for *vrěvěti or *velěti, in any usage, but does document the presence of several related nouns, cf. гл̥ч ‘шум, викане, врява’, Какóв бéше тоа гл ̥ч нощéска у Ки́́́рови? 139; глчáва, гл́́́чка (cf. глч) 139. If such forms presuppose the erstwhile existence of the verb *gъlčěti, then it most likely would have been used only in the original (non-neu- tral) sense. The attested unmarked verbs of speech are as follows: ) збру́́́вам, збру́́́ем ‘говоря, разговарям, пирказвам’, Цáла sáран збру́́́ва саз жени́́́те на чешмáта 267; збор, -тá ‘говор, говорене’, У збортá би́́́е на бащá си, ‘приказка, разговор’ Отишлá на збóр при жени́́́те 267; 2) Говóри (‘збрува’) налéво и надéсно 267; 3) прикáзвам ‘говоря, разправям’, Не мóже да прикáзва, óти го боли́́́ гр́́́лото 631; ‘разговарям, беседвам’, Приказвáа си нéшто, ама не разбрáх заштó бéше 631. The verb *dumati is used for speech, but (in contradistinction to Kjustendil dialects to the west) only with an implicit or stated complement, cf. ду́́́мам ‘говоря’, Сéки ден му ду́́́мам да не ргá текá по у́́́лиците , Нáна му ду́́́ма, а óн па си тáа ; Кáзвам (‘думам’) прáво/напрáво у́́́очи (у очи́́́те) . Note that neither думам nor дума (8–) means ‘think’, ‘thought’, cf. rather ми́́́слим 42. The particular form збру́́́вам (with apocope) is also the one cited in БДА-NW for pt. 222. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 4 West Sofia5 (Gorna Banja, Bojana, Gurmazovo, Knjaževo, Mirovjane, Suxodol, Filipovci) (Божков 1962). Božkov’s lexical material indicates the presence of the pejorative nominal formation гълчáва ‘глъчка’ 246. Otherwise, we find *zboruva- as a basic verb of speech, *duma‑ in the sense ‘say, tell’, and *vrěvi‑ only in its onomato- poeic and pejorative usage, cf. збору́́́вам ‘говоря’ 249, ду́́́мам ‘казвам’ 247, врéвим ‘карам се, говоря сърдито’ 244 (despite the absence of an entry for this last verb in Gъlъbov’s dictionary of the Dobroslavci dialect). The entries are not attributed to specific villages. 2.2.3.2 Northwest dialects Aside from БДА-NW, an important source providing evidence for “гълчъ” in NW Bg is Todorov’s monumental survey of NW Bg dialects (see Тодоров 1936), which contains at least 0 citations of this verb (two of which actually pertain to Transitional or SW Bg dialects, v.s.). Only one of these examples is glossed, albeit significantly as ‘говорили’ (cf. Gъrci, Vidin). Other sources consulted provide no attestations of “гълчъ” in NW Bg (see below for references from Mladenov’s study of Bulgarian émigré dialects in Romania and Xitov’s lexicon of Radovene, a Vraca dialect). Bjala Slatina (Enica) (Тодоров 1936). Cf. the noun глчáва 20, 244. Since the village of Enica is the product of settlement from Botevgrad and Etropole to the south, it was not listed in БДА-NW. Given that neighboring villages between Bjala Slatina and the Iskъr river attest приказвам (sometimes in variation with хортувам or гъгра) as the neutral verb of speech, one may assume that the noun глчава which Todorov cites most likely designates ‘uproar, tumult’. Bjala Slatina (émigré dialect in Romania) (Младенов 1993). Whereas Mladenov adduces only ду́́́мам ‘казвам, приказвам’ 169, БДА-NW indicates “приказвам” near Bjala Slatina, once in variation with “хортувам”. This suggests either that думам is an archaism (and thus no longer used in Bjala Slatina) or that the investigators of БДА-NW observed its use only in the sense of ‘казвам’. Kula (Gramada) (Тодоров 1936). Todorov cites the unglossed verbal forms гълчáла, гълчáле 2. БДА-NW (pt. 6) indicates variation between pres. глчи́́́ме, бъ́́́бреме (with the note “Най-новата дума е прикáзваме”). Thus, Todorov’s verb may also represent a neutral term of speech. Kozloduj (Čibr-Ogosta émigré dialect in Romania) (Младенов 1993). Mladenov cites only ороту́́́ва ‘говори, приказва’ 117. БДА-NW provides no data for dialects in the immediate vicinity of Kozloduj, but the adjacent villages of E. Lom and W. Or- jaxovo indicate only “приказвам”, a circumstance which suggests that “оротувам” was more common in the NW between Lom and Nikopol in the past and that it has subsequently been replaced by the standard form. Lom (Dolno Linjavo) (Тодоров 1936). Todorov provides гълчъ́́́т (sans gloss) 2. Although not included in БДА-NW, Dolno Linjavo (local pronunciation “Li- 5 For Kovačica the data are as follows: згъ́́́сти, зъб, къ́́́шта, мъ́́́чат, съ́́́бота, съ́́́штото ~ гу́́́ските, му́́́ти, му́́́шка, прут, пру́́́чка, ру́́́ка БДА-NW, p. 28. 50 Slovenski jezik – Slovene Linguistic Studies 6 (2007) nevo”) is situated on the Danube several kms. to the east of Lom. According to БДА- NW, the nearest village Kovačica (pt. 175) attests говóрим, whereas other proxi- mate points indicate either прикáзвам or орту́́́вам. In fact, according to БДА-NW the nearest point which exhibits “гълчъ” even as a variant neutral verb of speech is situated considerably to the west (cf. глчи́́́м, pt. 46). Thus, the testimony of БДА- NW militates against interpreting the form cited by Todorov in an unmarked sense. On the other hand, БДА-NW (Map 14) indicates that Kovačica and pt. 178 both ex- hibit variation between /ъ/ and /u/ as the reflex of *ǫ,52 although once again the near- est villages showing similar variation are situated considerably to the west. Thus, there is evidence that villages immediately to the west of Lom could represent earlier settlements from Transitional dialects in the northwestern Stara Planina, where “гъл‑ чъ” commonly occurs as a basic verb of speech.53 If such is the case, then the form гълчъ́́́т which Todorov cites may actually represent another example of this usage. Montana (Gabrovnica) (Тодоров 1936). Cf. гълчáле 121 (Studeno Buče) (Тодо- ров 1936). Cf. гълчи́́́м 2. Although neither of these villages is listed in БДА-NW, Gabrovnica is located 13 kms. to the N of Montana and Studeno Buče 4 miles to the NW thereof. Villages in this vicinity attest “приказвам” as neutral, but villages in the foothills of the Stara Planina to the west show “гълчъ“. In addition, one village farther to the east (pt. 64) displays “гълчъ ~ приказвам” (thus representing the east- ernmost outcropping of “гълчъ” in dialects to the north of Vraca). Since the village of Studeno Buče was excluded from the network of БДА-NW due to its “extremely heterogeneous” population (see p. ), it is quite possible that the presence of “гъл‑ чъ” in this case should be attributed to the influence of Transitional dialect speakers. There is no discussion of the grounds for excluding Gabrovnica. Pleven, Maraš (émigré dialect in Romania) (Младенов 1993). Mladenov lists орати ́́́ ‘говори’ 192. БДА-NW indicates that the form орати survives in two Pleven villages (pts. 1439, 751) as well as in two adjacent Nikopol points (pts. 746, 751). In other villages, the most common forms are “приказвам” and “говор’а” followed by “(х)ортувам”. Vidin (Gъrci, formerly Gradec) (Тодоров 1936). Todorov’s citation of гълчáле with the gloss ‘говорили’ (p. 387) is quite unexpected, since the corresponding БДА- NW pt.  is situated in the midst of a cluster of uniform “бъбрем” dialects. This would require the loss of *gъlčěti in unmarked usage within less than one full genera- tion. Vidin-Lom (émigré dialect in Romania) (Младенов 1993). In his remarks on the lexicon Mladenov cites no data on special verba dicendi (p. 64). Vraca (Тодоров 1936). Since the form гъ́́́лча 2 is not glossed, we cannot deter- mine if this verb represents the unmarked usage which in БДА-NW is attested in pt. 1347 to the south and pt. 1374 (in variation with “приказвам”) to the east. 52 See Стойков 1968: 101 and the historical introduction to БДА-NW, p. , for discussion of such migrational patterns. 53 The dialect of Radovene (situated on the Iskъr, to the west of Lukovit), exhibits many E Bg lexemes (cf. крак, риза, котка, късам, недей, той vs. NW нога, кошул’а, мачка, немой, он), but cf. also йа/йазе vs. NE аз. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 5 Vraca (Radovene) (Хитов 1979).54 Xitov’s extensive and well-illustrated dic- tionary (pp. 223–342) cites no entries for continuations of *gъlčěti. Instead, we find two neutral verbs of speech, cf. прикáзвам, прикáзам (i), при́́́кажна, при́́́кажа (p) 308 (for приказвам cf. also БДА-NW pt. 408) and óрата ‘говоря, приказвам’ (not cited in БДА-NW), Каквó ми орáти тóлко! 20. The latter verb also can be used pejoratively (as in some Rhodope dialects), cf. ‘пускам слухове’, Ис сéло орáтат штó ли нé 20. The cognate base noun is also used in both senses, cf. оратъ́́́, оротъ ́́́ ‘говор, реч; дума; слух, мълва, клюка’, Извáдиа на момичéто оротъ ́́́, без да и́́́ма заштó 20. There are also no listings for *bъbre-, *dumati, or *vrěvěti, but *gъgre- is found in its pejorative onomatopoeic sense, cf. гъ́́́гра ‘бъбря, говоря неразбрано’, Моми́́́те гъ́́́грат на пъ́́́к’а 238. 2.2.3.3 Transitional dialects The term “Transitional” is employed in contemporary Bulgarian dialectology (as in Стойков 1968: 115–117) to designate linguistically cognate dialects which are situated in extreme NW Bulgaria and immediately adjacent Serbian territory, cf. (in Bulgaria) Belogradčik, western Berkovica, Trъn, Breznica, and (in Serbia) Cari- brod, Bosilegrad. In general these dialects exhibit classic typological traits of the Balkansprachbund type (such as the postposed definite article, drastically reduced case system, near absence of an infinitive) together with genetically determined pho- nological and morphological features of a Serbian type (cf. *ǫ > u, *ъ and *ь > ъ, retention of voiced stops in auslaut, medial /l/, absence of the phoneme /f/, pl. -mo, fem. pl. -e, 3pl. aor. -še), whereas still others admit of more complex interpretations (cf. *tj, *dj > ч, џ and t’, d’). In geopolitical terms, the regions concerned which are now situated on the Serbian side of the border were part of Bulgaria prior to 8. Although published sources for Transitional dialects other than БДА provide little information in general regarding continuations of *gъlčěti, our survey yielded examples of *gъlčěti only from the Belogradčik region. No attestations of this verb are cited in Transitional dialects to the south of the Stara Planina massif (cf. Godeč, Burel, Graovo, Trъn). This distribution is consistent with the detailed picture one ob- tains from БДА-NW, SW, where “гълчъ” predominates to the north of the Berkovska Planina, but does not breach the Stara Planina chain to the south thereof. Belogradčik (Ol’šane) (Берберска 1931). In her discussion of reflexes of vocalic */l̥/, Berberska cites both the verb and a cognate noun (both sans gloss), cf. гл̥чиш 2, гл̥чáва 2. This does not not contradict the testimony of БДА-NW (pt. 25), where we find гл̥чи́́́мо си in the sense ‘приказвам’. Belogradčik (Vъrbovčec) (Тодоров 1936). Cf. гълчъ́́́т 398, кълчи́́́мо (sic!) (‘гълчимо’) 121. The foothill village of Vъrbovčec was excluded from the network of БДА-NW (p. ) due to the strong presence there of western Stara Planina (i.e., Transitional dialect) settlers. Note further that Vъrbovčec is located several kilome- 54 The dialect of Radovene (situated on the Iskъr, to the west of Lukovit), exhibits many E Bg lexemes (cf. крак, риза, котка, късам, недей, той vs. NW нога, кошул’а, мачка, немой, он), but cf. also йа/йазе vs. NE аз. 52 Slovenski jezik – Slovene Linguistic Studies 6 (2007) ters to the east of Belogradčik, i.e., where according to БДА-NW neighboring villages attest “гълчъ” with the meaning ‘приказвам’. Belogradčik-Timok (Vratarnica)55 (Sobolev 1994). In his discussion of reflexes of vocalic */l̥/, Sobolev cites several forms of *gъlčěti (including an Aktionsart form with *za- + sę), cf. гл̥чи́́́мо, гл̥чáмо, загл ̥чáла се 0, as well as the cognate noun гл̥́́́чка 0. Burel (Любенов 1993). Ljubenov’s extensive and well illustrated dialect lexicon (pp. 5–55) contains no entires for *gъlčěti, *bъbre-, *veli-, nor *dumati, but does cite *vrěvěti and *zboruje- as ‘speak, talk’ in neutral usage, cf. вре́́́вим ‘говоря, при- казвам’, Му́́́жу, тре́́́бе да вре́́́вимо на синото́́́га ве́́́чим да се же́́́ни 26, збору́́́ем ‘говоря’, Мло́́́го е срамл’и́́́ф – у народ не збору́́́е, само си тра́́́е и гле́́́да 48, че збору́́́емо 0, збор ‘говор’. In contrast, the verb *prikazati is restricted in its usage, cf. прика́́́зуем се, прика́́́жем се ‘общувам и уважавам стари роднински връзки’ 2. Godeč (Виденов 1978).56 Videnov’s otherwise extremely thorough description contains neither a dictionary nor a systematic discussion of the lexicon. There is no evidence of “гълчъ” in the brief summaries of the reflexes of the jers (33–34) and vocalic liquids (50–52). On the other hand, a discussion of verbal polysemy notes that думам can mean both ‘разказвам’ and ‘заповядвам’ 93. BDA-NW indicates only “зборуем” for Godeč. Graovo (Мартинов 1958). Martinov’s condensed lexicon of the Graovo dialect contains no entries for *gъlčěti, *bъbre-, and *velěti, but cites instead *vrěvěti and *zboruje- as basic verbs of speech, cf. вре́́́вим ‘говоря’, Шо ми вре́́́виш ода́́́мо 775, збору́́́ем, зазборува́́́ли се (from збор ‘говор’, Збор се чу́́́е некуде 777). This is con- sistent with БДА-SW, NW, where the Graovo region to the north of Breznik straddles the somewhat diffuse isogloss between “вревим” dialects to the west and “зборуем” dialects to the east. Trъn (Господинкин 1921; Петричев 1931). Gospodinkin’s brief glossary makes no reference to verbs of speech, nor does his fairly extensive list of reflexes of vo- calic */l̥/ contain any example of *gъlčĕti (pp. 44–45). Petričev makes no reference to *gъlčĕti in his exahustive list of words with vocalic */l̥/ (p. 173), while citing only *vrĕvĕti as a distinctive, but neutral verb of speech, cf. врéвим ‘говоря’, Дóста ви 55 The village of Vratarnica was founded near the middle of the 19th c. by settlers from the neighboring Transitional dialect region of Belogradčik in Bulgaria (see Sobolev 1994: 14–15). Situated at the same latitude as Rakovica (Belogradčik) Vratarnica represents the northernmost point in Serbian territory for which an attestation of *gъlčěti has been found in the sources consulted for the present study. 56 The Godeč region in many respects represents a zone of overlap between classic “Tran- sitional” and NW Bg (East Sofia) features, cf. considerable local variation in the form of an east-west gradation of features, with productive spread of W. Sofia dialects features, not those of Lit-Bg or the “pure” Transitional dialects to the west; thus, Videnov, the author of the pri- mary source, cites much of his data from centrally located villages, where a prominent W. Sofia feature such as the masc. def. sg. -o co-exists with “Serbian” phonetic reflexes of *ǫ > u, and morphological traits such as 3pl. aor. In -še and pl. -mo) J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 53 вреви́́́л – главá ме заболи́́́ 200; Си́́́де де, не вреви́́́ ми млóго 200. БДА-NW shows mainly “вревим” for the Trъn region. For Transitional dialects on the Serbian side of the border, we have the evidence of БДА-Trans, which confirms the absence of *gъlčěti (at least as a neutral verb of speech) to the south of the Timok and Pirot dialect regions, cf. (from north to south) “збору́́́ем” (N. Caribrod), “орáтим” (S. Caribrod, Trъn), and “ду́́́мам” (Bosilegrad). All of these forms constitute direct continuations of the isoglosses which begin on the Bulgarian side of the border, south of the Stara Planina. For evidence of *gъlčěti in Serbian Torlak dialects to the immediate north of these regions, see the following sec- tion. Where attested, the past tense forms of *gъlči- exhibit -i- rather than the reflex of -ě- as the stem vowel, as in some Russian dialects (see Introduction). 2.2.3.4 Southeast Serbian (“Torlak”) dialects Our survey of East South Slavic (“Balkan Slavic”) may be extended to include those dialects which in many respects constitute a continuation of the “Transitional” dialects on the Serbian side of the border with Bulgaria, but are located to the west and north thereof, (cf. Timok-Lužnica, Svrljig-Planina). Sources for these dialects provide evidence for the occurrence of *gъlčěti, primarily in districts which are situ- ated in farily close proximity to the Serbian-Bulgarian border. When glossed, the verb is onomatopoeic, as are the cognate nouns (with one notable exception in Dinić’s material from Timok). Timok (Динић 1988a) (specific provenance not indicated) гл̥чи́́́ ‘галами; приго- вара, гунђа’, Мóре, чу му однесéм тоj едан‑пу́́́т да ми ви́́́ше не гл̥чи́́́ 4. Cf. also гл̥ч ‘тишина, мир; жагор’, гл̥чáње ‘галама, врева, мумлање’ 48; (Динић 1988b) (no indication of specific provenance) гл̥́́́чиjе ‘jасниjе, гласниjе’, Де мáлко гл̥́́́чиjе, слáбо те чу́́́ем 397 (note that Dinić derives this comparative form from the “impera- tive” of гл̥чи), cf. also гл̥́́́чка ‘врева, разговор, ђаскање’ 397. Of particular interest here is the semantic extension of гл̥чка to include a more neutral meaning (cf. ‘разго- вор’). Timok-Lužnica (Белић 1905). In his monumental survey of SE Serbian dialects, Belić cites several examples of *gъlčěti and its cognate nouns in his discussion of reflexes of vocalic /l̥/. Where glossed, the verb is still strongly onomatopoeic, cf. Kraljevo Selo/Novi An. гл̥́́́че 94; Radoševci гл̥чи́́́ 4; Tijelovac57 гл̥чи́́́ли су ‘ровеоше’ 4. The same can be said of its cognates, cf. Novo Korito58 гл̥чи́́́jе ‘више, jасниjе, пре’ 94, G. Roman гл̥ъач ‘урлање’106; although there is one apparent exception, cf. Novo Korito гл̥́́́чка ‘разговор’ 94. Timok (Sobolev 8). Sobolev’s linguistic atlas of SE Sb and NW Bg (for our purposes equivalent to E. Torlak and Transitional Bg) is based on extensive fieldwork as well as an exhaustive collation of material from published sources. In the detailed 57 Tijelovac is located near the southern bank of the Nišava, ca. 20 kms. NW of Pirot and 7 kms. SE of Bela Palanka. 58 Novo Korito is located within a few kms. of the Sb-Bg border at the same latitude as Rabiša (Belogradčik). 54 Slovenski jezik – Slovene Linguistic Studies 6 (2007) discussion of reflexes of vocalic */l ̥/ (pp. 137–164), we find Crni Vrh5 gl̥čú́́ (sans gloss) (Exp. 5)60 142. For neighboring regions, we find the following attestations of *gъlčĕti and cognate nouns, all unglossed. Pirot (Sobolev 8). The Pirot dialect region is situated to the south of the Timok region and immediately NW of the Caribrod transitional dialect. Sobolev adduces So- pot6 iskĺ́́čim 143 (citing Златковић 1989: 247); Gradašnica62 isklčíl 143 (citing Злат- ковић 1989: 39); Rosomač63 glčí́́lo 144 (citing Златковић 1988: 38); Vojnjegovac64 iskĺ́́čim 145 (citing Златковић 1989: 247).65 Lužnica (Sobolev 1988). Sobolev lists Donji Striževac66 klčé́́mo 144 (citing Curić 1983: 25–28). Svrljig (Sobolev 88). This region represents the westernmost limit of attesta- tions of *gъlčěti in our study. Sobolev quotes Beli Potok67 gləčí 142 (citing Богдано- вић 1979: 15–18). Thus, when compared to Belič’s data, the most recently gathered material would indicate the relative stability of *gъlčěti, etc. over the course of the 20th century. A review of Broch’s general survey of Torlak dialects (Broch 1903) and of Alex- ander’s description of their accentual systems (Alexander 75) yielded no further examples of continuations of *gъlčěti or its cognates. 2.2.3.5 Bulgarian dialects in Ukraine Bulgarian dialects in Ukraine are spoken by the descendants of immigrants who settled primarily in Bessarabia and Tavrija over the course of nearly a century from the middle of the 8th to the middle of the th century. Almost all of the coloniza- tion came from NE and SE Bg, although some of the dialects exhibit W Bg features, which probably reflect prior migration and settlement in S. Thrace. Some villages exhibit evidence of dialect mixture due to different waves of settlement. The two major sources on all of these dialects are Deržavin’s monumental survey (Державин 1915) and a series of publications by Russian scholars who conducted fieldwork after WWII (see primarly the series СМБД). Here we cite only attestations of *gъlčěti and its cognate nouns. Čijšija dialect type68 (Чешко 1952). Cf. гαлчá (not glossed) 7. 5 Crni Vrh is located adjacent to the Bg border just south of the latitude of Gorni Lom. 60 “Exp. 5”, etc. refers to the year of dialect expedition during which the material for the Atlas was collected. 6 Sopot is located  kms. N of Pirot. 62 Gradašnica is situated 3 kms. N of Pirot. 63 Rosomač is situated due W of Pirot near the Bg obrder and Berkovska Planina. 64 Vojnjegovac is located ca. 10 kms. S of Pirot, near the Caribrod dialect region. 65 We assume provisionally that forms with root-initial voiceless velar (*iz‑kъlči-) are cognate with the similar form cited for Belogradčik by Todorov (see above, Bg Transitional dialects) and glossed by him as equivalent to *gъlči-. 66 Donji Striževac is located ca. 30 kms. west of Pirot. 67 Beli Potok is located ca. 20 kms. W of Knjaževac, to the north of the Svrljiški Timok basin. 68 According to Deržavin, the Čijšija dialect is of NE Bg origin, but does not embody the J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 55 Kriničnoe6 (Журвалев 1962). Cf. гал’чи́́́ (unglossed) 32. Loščinovki (Чешко 1952). Cf. гά́́л’чи,70 Гά́́л’чи с нéзи, т’а у тилифóна 27, гαл’чáти pl. imv. 65. The occurrence of *gъlčěti in a neutral sense here is of particu- lar significance, due to the complex nature of the dialect’s composition. Češko identi- fies the Loščinovki dialect as primarily of SE Bg “Kajraklija” Thracian origin, with admixtures of influence from Eastern Rhodope and NE Bg “Čišija” types. Our own analysis of certain features in the dialect indicates that the “Thracian” element here is most probably of SW Bg origin, cf. mobile stress in the imperative, lexical features such as разбóй ‘loom’ (op. cit.) 27 (vs. E Bg стан), штéрка (vs. E Bg дъщеря) 27 (albeit in variation with дαщтерá 65), and perhaps even the sporadic occurrence of masc. def. sg. -o in the speech of some older informants (p. 30), e.g., пу пъ́́́т’у 31. Further, the regular occurrence of -’a- (and less commonly -’ä‑, -čä-) as the reflex of jat’ in Loščinovki is not inconsistent with our hypothesis regarding an older SW Bg substratum in the dialect, since the presence of such reflexes (particularly the archaic -ä-) is clearly documented in SW Bg émigré dialects of the Fakija type (see maps for reflexes of jat’ in БДА-SE). It is impossible to determine if the neutral which is at- tested in Loščinovki represents the former state of affairs in SW Bg dialects such as Samokov, where today we find only the pejorative sense of ‘loud talking’. Ol’šane (Бунина 1953, 1954). In her glossary of the Ol’šane dialect, Bunina (54) cites гαл’чä ́ ‘ругать’ 16, гал’чä ́ сα ‘ругаться, браниться, ссориться’, but also ду́́́мам ‘говорить, называть’ 20. Whereas the apparently neutral sense of *du‑ mati as a verb of speech may represent an archaism which has vanished in modern NE Bg, the occurrence of *gъlčěti is consistent with other evidence for the Ol’šane diaect, which appears to combine features of the Mizija type (cf. masc. def. sg. in -o, notably in conunction with unusual instances of mobile stress, e.g. крäс, крαстó, боп, бубó Бунина 1953:57) with striking characteristics of the Erkeč dialect type (cf. /ä/ < *ǫ, *ъ, *ъr/*ьl). Although the Sъrt and Erkeč dialects are situated in relatively close proximity to each other, it is conceivable that the assortment of traits exhib- ited by the Ol’šane dialect reflects dialect mixture which occurred after settlement in Ukrkaine rather than in Bulgaria. We do note, however, that the pejorative (rather than neutral) sense is the one attested for *gъlčěti in the modern Erkeč dialect (v.s.). Suvorovo (Полтораднева-Зеленина 1955). Poltoradneva-Zelenina notes the occur- rence of *gъlčěti in a neutral sense, cf. гαлча́́́ ‘говорить, беседовать’ 75. Generally classified as an East Thracian o-dialect (see Бернштейн, Журавлев, Сенкевич Толс- той 1953) by Deržavin, Suvorovo exhibits accentual features which tend to preclude Rup origin (cf. oxytonesis in sg. present and the sg. imperative). On the other hand, standard Stara Planina type, despite the presence of -ъ (rather than Mizija -o) in the masc. def. sg. form. 6 The Kriničnoe dialect is of the classic Sъrt o-dialect type, see Журавлев 1955: 18–63. As such it retains stress on the first syllable in nouns such as *motyka, cf. мóт’ка скъ́́́рца 2 ‘the hoe got broken’. 70 This symbol was first used by Deržavin to indicate the “non-tense” lower schwa-like vowel found in most E Bg dialects, in contradistinction to the so-called “tense” (higher, more closed) articulation which occurs in Sъrt and Sъrt émigré dialects. Later Russian scholars often followed this tradition. 56 Slovenski jezik – Slovene Linguistic Studies 6 (2007) the masc. def. sg. o-desinence and the archaic jat’ reflex are incompatible with the Stara Planina type. All of this points towards the so-called “Zagorci” o-dialect pres- ence as significant, although the -a sg pres. desinence is admittedly problematic for this interpretation. The usage of *gъlčěti as a neutral verb of speech enhances the probability that the Suvorovo dialect has Zagorci origin, in view of the genetic con- nection between Zagorci dialects and Mizija o-dialects to the north, where this same usage has been retained in a small number of Sъrt and Razgrad dialects. Map 7. Distribution of *gъlčěti (onomatopoeic-pejorative usage; unglossed examples) Onomatopoeic, pejorative usage: () NE Torlak, (2) Samokov, (4) Botevgrad, (6) N. Pavlik- jan, (7) Pleven, (8) Silistra, (9) Šumen, (10) Preslav, (11) Erkeč, (12) Strandža; Unglossed: (1) NE Torlak, (3) Makocevo, (5) Teteven, (13) Lozengrad, (14) Mandъr, (15) Montana, (16) Vraca. Romania: Pavlikjan (pejor.); Ukraine: Ol’šane (pejor.), Čišija (unglossed), Kriničnoe (unglossed). 2.2.3.6 Bulgarian Damascenes of the 17–18th centuries The so-called “Damascene” texts provide unique and invaluable evidence on the state of the Bulgarian and to a lesser degree Macedonian vernacular language of the 7–8th centuries. The texts derive their name from the Greek author Damascene Studite, whose religious anthology Θησαυροσ (Slavic Сокровище) was translated into Church Slavonic of the Bulgaro-Serbian recension in the late 6th century, then into the vernacular. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 57 The majority of the best Damascenes originate in the West Central Stara Plani- na region of Bulgaria. Of particular note are the so-called Tixonravov, Trojan, and Koprivštica Damascenes. In her magisterial study of the Tixonravov Damascene, the Russian scholar Demina (Демина 1968) documents the presence in this manuscript of two characteristic imperfective verbs of speech, viz. хорати- (cf. “статьи груп- пы тогива”)7 and дума- (cf. “статьи группы тогива и группы тогази”). Demina further notes that modern dialects of the Loveč region appear to lack дума utterly (thus Демина 1985: 222–223, and, tacitly, БДА), while usually exhibiting хортува or прикажува (actually БДА shows приказвам). Having previously established that the Tixonravov version was composed in the mid-7th century in the general vicinity of the Loveč eparchy, Demina proposes the possibility that хороти- was more widely used at that time than in the mid-20th c. and was subsequently partially replaced by приказвам. Thus, it appears that *dumati was the older Slavic verb of speech, which was in the process of being replaced by *xorati- in the 7c. The latter verb was then superseded by the morphologically productive type хортува- , which in turn has been subject to replacement by literary (and prestige dialect?) прикажува-. A similar development seems to have occurred in the Mizija o-dialect region farther to the east, except that here the old Slavic verb was probably *gъlčěti (which, unlike *dumati, still survives in some isolated localities as the basic verb of speech). The Svištov Damascene of 1753 is of particular interest from the point of view of Bulgarian dialectology, inasmuch as it documents the possible erstwhile presence of the Mizija o-dialect to the north and west of the modern borders of this type (see Милетич 1923: 3–4). Our preliminary survey of selected passages from this text re- veals of selected passages reveals no occurrence of *xorati-, but suffices to document the usage of *dumati as a neutral verb and *duma as its nominal counterpart, whereas *kazati, *kaže- is used to introduce quotations, cf. чюдеши са и думаше, що ще да бъде туй 369;72 начеха да думать в’ бъди‑кой род’ и племе 203; И като чю царъ Валакъ т<зи думы, каже Валааму: чи аз та зовах’ да ги проклънешъ, а ты ги благослов<ваш. Каже му Валаамъ: азъ ты и по‑напред’ рекох’ туй, оти щото ми рече богъ мой, туй щъ да думамъ, що са гн<вашъ на мен = pre-modern textual attestation; BCS = Bosnian/Croatian/Serbian (standards), Be = Belarusan, Bg = Bulgarian, BS = Balto-Slavic, Cr = Croatian, Cz = Czech, En = English, Gk = Greek, Hi = Hittite, IP = imperfective, IE = Indo-European, La = Latvian, Li = Lithuanian, Lit = Literary, LS = Lower Sorbian, Lv = Latvian, OCS = Old Church Slavic, Ma = Macedonian; MHG = Middle High German, O- = Old, OPru = Old Prussian, P = perfective, Po = Polish, PS = Proto-Slavic, Ru = Russian, Sb = Serbian, Sk = Slovak; Sn = Slovene; St = Standard, Uk = Ukrainian, US = Upper Sorbian, Ve = Vedic References Akademijin rječnik = Rječnik hrvatskoga ili srpskoga jezika. Zagreb: JAZU, 1880–. Alexander, Ronelle. 75. Torlak Accentuation. (= SlBei 4). Munich. Andersen, Henning. 1999. The Western South Slavic Contrast Sn. sah‑ni‑ti || sah‑nu‑ti. Slovenski jezik/Slovene Linguistic Studies: 47–62. BÉB = Bibliotheque d’études balkaniques. Paris. Belostenec, Ivan. 1740 (1973). Gazophylacium, vols. 1–2. Zagreb: Liber, Mladost. [Reprint edition.] Bezlaj, France. 1955. Krčevine. Slavistična revija 8/1–2: 1–23. Reprinted in F. Bez- laj, Zbrani jezikoslovni spisi I: 20–40. Ljubljana: ZRC SAZU. – – 1967. Slovenščina v krogu slovanskih jezikov. III. seminar slovenskega jezika, lit‑ erature in kulture. Predavanja: 1–13. Ljubljana: Filozofska fakulteta. Reprinted in F. Bezlaj, Zbrani jezikoslovni spisi I: 305–313. Ljubljana: ZRC SAZU. – – 77, 82, 5, 2005. Etimološki slovar slovenskega jezika. vol. I (A–J), II (K–O), III (P–S), IV (Š–Ž). Ljubljana: ZRC SAZU, Inštitut za slovenski jezik/ Mladinska knjiga, 77, 82. BLS = Bibljoteka Ludu Słowiańskiego. Cracow. Bratanić, Branimir. 1951. Uz problem doseljenja Južnih Slavena: nekoliko etnograf- sko-leksičkih činjenica. Zbornik radova: 221–250. Zagreb : Sveučilište u Za- grebu, Filozofski fakultet. Broch, Olaf. 1903. Die Dialekte des südlichsten Serbiens. SBKLA‑SD, v. 3. BS = Bernik, France, et al. ,eds. 1993. Brižinski spomeniki. Znanstvenokritična izda‑ ja. Ljubljana: ZRC SAZU. Buck, Carl D. 5a. Words of Speaking and Saying in the Indo-European Languag- es: First Paper. The American Journal of Philology 36/1: 1–18. – – 5b. Words of Speaking and Saying in the Indo-European Languages: Second Paper. The American Journal of Philology 36/2: 125–154. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 67 Delčeva, ms. Towards a Lexicon of the Erkeč Dialect. Gavazzi, Milovan. 1936. Problem karakterističnoga razmještanja nekih etnografskih elemenata na Balkanu. Comptes Rendus du IVe Congrès des géographes et des Ethnographes Slaves. Sofia. Gebauer, Jan. 70. Slovník staročeský, Díl I (A–J). Prague: Academia. Gluhak, Alemko. 1993. Hrvatski etimološki rječnik. Zagreb: August Cesarec. Greenberg, Marc L. 0. A Historical Analysis of the Phonology and Accentuation of the Prekmurje Dialect of Slovene. Unpublished Ph.D. Dissertation, UCLA. – – 1993. Glasoslovni opis treh prekmurskih govorov in komentar k zgodovinske- mu glasoslovju in oblikoglasju prekmurskega narečja. Slavistična revija 4/4: 465–487. – – 1999. Slovarček središkega govora (na osnovi zapisov Karla Ozvalda). Slovenski jezik/Slovene Linguistic Studies 2: 28–75. [Also available at http://hdl.handle. net/808/80] – – 2005. Dialect Variation along the Mura.” Croatica et Slavica Iadertina : 07– 24. Groen, B. M. 77. A Structural Description of the Macedonian Dialect of Dihovo. Lisse: de Ridder. Habdelić, Juraj. 1670 (1989). Dictionar ili réchi szlovenszke. Zagreb: Kršćanska sadašnjost. [Reprint edition.] Hafner, Stanislaus and Erich Prunč. 1982–. Thesaurus der slowenischen Volkssprache in Kärnten. Vienna: Verlag der Österreichischen Akademie der Wissenschaften. Hendricks, Peter. 1976. The Radožda‑Vevčani Dialect of Macedonian. Lisse: de Rid- der. Hill, Peter. 1991. The Dialect of Gorno Kalenik. Columbus: Slavica. Houtzagers, Peter. 1999. The Kajkavian Dialect of Hidegség and Fertőhomok (= Studies in Slavic and General Linguistics 27). Amsterdam, Atlanta: Rodopi. Ivančič Kutin, Barbara. To appear. Slovar bovškega govora. Jambrešić, Andrija. 1742 (1992). Lexicon Latinum interpretatione Illyrica, Germanica et Hungarica locuples. Zagreb: Zavod za hrvatski jezik Hrvatskoga filološkoga instituta. [Reprint edition.] Jardas, Ivo. 57. Kastavština. Građa o narodnom životu i običajima u kastavskom govoru (= ZNŽOJS 39). Zagreb: JAZU. Jedvaj, Josip. 56. Bednjanski govor. Hrvatski dijalektološki zbornik 1: 279–330. Zagreb: JAZU. Kalsbeek, Janneke. 8. The Čakavian Dialect of Orbanići near Žminj in Istria (= Studies in Slavic and General Linguistics, 25). Amsterdam, Atlanta: Rodopi. Kniezsa, István. 74. A magyar nyelv szláv jövevánszavai, vols. –2. Budapest: Akadémiai kiadó. Koletnik, Mihaela. 200. Slovenskogoriško narečje (= Zora 12). Maribor: Slavistično društvo. Kozar-Mukič, Maria. 1988. Felsőszölnök–Gornji Senik. Szombathely, Ljubljana: Sa- varia Múzeum, University of Ljubljana. Kranjec, Miško. 19773. Strici so mi povedali. Murska Sobota: Pomurska založba. Kuryłowicz, Jerzy. 1945–49 (1966). La nature des procès dits “analogiques.” Acta Linguistica 5. Reprinted in Eric P. Hamp, Fred W. Householder, and Robert Aus- 68 Slovenski jezik – Slovene Linguistic Studies 6 (2007) terlitz, eds., Readings in Linguistics II: 58–74. Chicago: University of Chicago Press. Lundberg, Grant H. 1999. Preliminary Report on Dialectological Fieldwork in Haloze, Slovenia. Slovenski jezik/Slovene Linguistic Studies 2: –08. [Also available at http://hdl.handle.net/808/805] – – 2005. Dialect Divergence on the Slovene-Croatian National Border. Balkanistica 8: 7–84. Małecki, M. 1936. Dwie gwary macedońskie. Solun i Wysoka w Soluńskiem. II. Słownik. BLS 3, A. Mazon, A. 1936. Documents, contes et chansons slaves de l’Albanie du Sud. BÉB 5. Mazon, A., Vaillant, A. 1938. L’Évangéliaire de Kulakia/ Un parler slave du Bas‑ Vardar. BÉB 6. Mikhailov, Nikolai. 8. Frühslowenische Sprachdenkmäler. Di handschriftliche Periode der slowenischen Sprache (XIV. Jh. bis 1550) (= Studies in Slavic and General Linguistics 26). Amsterdam, Atlanta: Rodopi. Miletič, L. 1903. Das Ostbulgarische. SBKLA‑SD 2. Mohorovičić-Maričin, Franjo. 2001. Rječnik čakavskog govora Rukavca i bliže oko‑ lice. Rijeka: Adamić. Mukič, Francek. 2005. Porabsko‑knjižnoslovensko‑madžarski slovar. Szombathely. Neweklowsky, Gerhard. 87. Lexikalische Übereinstimmungen im nordwestlichen Südslawischen. Slavistična revija 35/1–2: 3–16; 187–209. Novak, Franc. 6. Slovar beltinskega prekmurskega govora (ed. and amended by Vilko Novak). Murska Sobota: Pomurska založba. Novak, Vilko. 2006. Slovar stare knjižne prekmurščine. Ljubljana: ZRC SAZU. Pável, Ágost. 0. A vashidegkuti szlovén nyelvjárás hangtana (= Magyarországi szláv nyelvjárások, 1. füzet). Budapest: A Magyar Tudományos Akadémia. – – 42. Vend nyelvtan. Szombathely/Vashidegkút (unpublished ms.). Piasevoli, Ankica. 1993. Rječnik govora mjest Sali. Zadar: Ogranak Matice hrvatske Zadar. Pokorny, Julius. 5. Indogermanisches etymologisches Wörterbuch (Bern: Francke, 5). On-line version at http://www.indoeuropean.nl Puhvel, Jaan. 64. The Indo-European and Indo-Aryan Plough: A Linguistic Study of Technological Diffusion. Technology and Culture 5/2: 76–0. Pulevski = G. Pulevski’s quatralingual dictionary (Macedonian, Albanian, Aroma- nian, Turkish), as cited by Belić in his study of the dialect of Galičnik. Rajh, Bernard. 8. Dajnkovo berilo (= Zora 4). Maribor: Slavistično društvo. – – 2002. Od narečja do vzhodnoštajerskega knjižnega jezika (= Zora ). Maribor: Slavistično društvo. Reiter, Norbert. 64. Der Dialekt von Titov Veles. Berlin. Rix, Helmut, et al. 2001. LIV – Lexikon der indogermanischen Verben. Die Wurzeln und ihre Primärstammbildungen. Wiesbaden: Dr. Ludwig Reichert Verlag. Pleteršnik, Maks. 1894. Slovensko‑nemški slovar, Prvi del A–O. Ljubljana: Knezoškofijstvo. RHKKJ = Rječnik hrvatskoga kajkavskoga književnog jezika, ed. B. Finka et al. Za- greb: JAZU/HAZU, 1984–. Roki, Andro. 7. Libar viškiga jazika. Toronto: Libar Publishing. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 6 SBKLA‑SD = Schriften der Balkankommission. Linguistische Abteilung. Süds‑ lawischen Dialektstudien. Wien. Schallert, Joseph. 200. Southwest Bulgarian Dialect Features in the Fakija (Gru- dovo) Dialect of Southeastern Bulgaria: (с)кина ‘to Pluck’. In Honor of Howard I. Aronson, On the Occasion of His 60th Birthday, V. Friedman, D. Dyer, eds. (= Indiana Slavic Studies, v. 12): 359-367. Schmidt, Johannes. 87. Zur Geschichte des indogermanischen Voсalismus. Wei- mar. Šivic-Dular, Alenka. 1987. O dosedanjem preučevanju slovenske lingvogeneze. XXIII. Seminar slovenskega jezika, literature in kulture. Zbornik predavanj: 53– 67. Ljubljana: Filozofska fakulteta. SJS = Kurz, J., et al., eds. 58–7. Slovník jazyka staroslověnského. Prague: Československá akademie věd. Skok, Petar. 7–74. Etimologijski rječnik hrvatskoga ili srpskoga jezika. Zagreb: JAZU. SlBei = Slavistische Beiträge. Munich. Snoj, Marko. 2003. Slovenski etimološki slovar. Ljubljana: Modrijan. – – 2006. Slovar jezika Janeza Svetokriškega (=Slovenska akademija znanosti in umetnosti, Razred za filološke in literarne vede, Dela 4/7), vol. –2. Ljubljana: Ljubljana: ZRC. Sobolev, Andrej. 8. Sprachatlas Ostserbiens und Westbulgariens, 3 vols. Mar- burg. Sophocles, Evangelinus Apostolides. 887. Greek lexicon of the Roman and Byzan‑ tine periods (from B.C. 146 to A.D. 1100). New York: F. Ungar. Sommaripa, Gregorio Alasia da. 607 (7). Vocabolario Italiano‑sloveno. Ljublja- na, Devin-Nabrežina, Trieste/Trst: Mladinska knjiga. [Reprint edition.] Sophocles, Evangelinus Apostolides. 887. Greek lexicon of the Roman and Byzan‑ tine periods (from B.C. 146 to A.D. 1100). New York: F. Ungar. SSKJ = Ahlin, Martin, Ljudmila Bokal, et al. Slovar slovenskega knjižnega jezika (In- ternet version). 2001. Ljubljana: Inštitut za slovenski jezik Frana Ramovša ZRC SAZU. URL: http://bos.zrc-sazu.si/sskj.html Stang, Christian S. 42. Das slavische und baltische Verbum. Oslo. Steenwijk, Han. 2001. The Contribution of Oroslav Caf to Resian Lexicography. Slovenski jezik/Slovene Linguistic Studies 3: 102–134. Šašel, Josip. Ca. 1910–1957. Rožanski narečni besednjak. Ms. in possession of Inštitut za slovenski jezik Frana Ramovša. Štrekelj, Karl. 1887. Morphologie des Görzer Mittelkarstdialektes. Sitzungsberich‑ te der Wiener Akademie der Wissenschaften. Philologisch‑historische Klasse CXIII, 1: 377–496. Tornow, Siegfried. 8. Burgenlandkroatisches Dialektwörterbuch. Die vlahischen Ortschaften (= Osteuropea‑Institut an der freien Universität Berlin. Balkanolo‑ gische Veröffentlichungen, Bd. 15). Wiesbaden: Otto Harrassowitz. Vaillant, André. 1966. Grammaire comparée des langues slaves, t. III. Paris: Éditions Klincksieck. Velčić, Nikola. 2003. Besedar Bejske Tramuntane (= Čakavska biblioteka, knj. ). Mali Lošinj: Katedra Čakavskog sabora Cres-Lošinj. 70 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Vermeer, Willem. 1982. On the principal sources for the study of čakavian dialects with neocircumflex in adjectives and e‑presents”. Studies in Slavic and General Linguistics 2: 279–340. Amsterdam: Rodopi. ZNŽOJS = Zbornik za narodni život i običaje južnih Slavena. Zagreb: JAZU. Zorko, Zinka. 1992. Prleško narečje v Radomerščaku, rojstem kraju F. Miklošiča. Miklošičev zbornik (= Obdobja 13): 463–474. Ljubljana: SAZU, Univerza v Lju- bljani. – – 1998. Govor Gomile pri Kogu, rojstnega kraja Božidarja Borka. Zinka Zorko, Haloško narečje in druge dialektološke študije (= Zora 6): 64–73. Maribor: Slavistično društvo. Алексеев, Н. 1931. Разложкият говор. МПр 3: 91–125. Ангелова, Р. 1948. Село Радуил, Самоковско. Народопис и говор. ИССФ 8–, 131–411. Антонова-Василева, Л., Керемедчиева, С. 2001. Кратък речник на диалектните думи. Sofia. *БД = Проучвания и материали по българска диалектология. Sofia. БДА‑Aeg = Български диалектен атлас. Български говори от Егейска Македо- ния. Sofia, 1966. Иванов, Й. БДА‑NE = Български диалектен атлас, т. 2. Североизточна България. Sofia, 66. БДА‑SW = Български диалектен атлас, т. 3. Югозападна България. Sofia, 1975. БДА‑NW = Български диалектен атлас, т. 4. Северозападна България. Sofia, 8. БДА‑Tran = Български диалектен атлас, т. 2. Североизточна България. Sofia, 66. БЕР = Български етимологичен речник. Т. 1–5. Sofia. 1971–1993. БгСт = Български старини. Sofia. Белић, А. 1905 Диjалекти источне и jужне Србиjе. СДЗб . – – 1935. Галички диjалекат. СДЗб 7. Берберска, А. 1931. Говорът на село Ошане, Белоградчишко. ИССФ 7: 7–20. Бернштейн, C. Б., Журавлев, В. К., Сенкевич, М. Г., Толстой, Н. И. 1953. Отчет о диалектологической экспедиции в болгарское село Суворово летом 5. СМБД 3: 110–149. Бинев, Кр., Кацаров, Г. Народописни материали от Морихово. МП 1, 5–6:160. Богданович, Н. 1972. Говор Бучума и Белог Потока. СДЗб 25. Божкова, З. 1962. Принос към речника на софийская говор. БД 1: 242–273. Боjковска, Ст. 2003. Радовишкиот говор. ПИз 40. Бояджиев, Т. 1970. Из лексиката на с. Дервент, Дедеагачко. БД 5: 223–245. – – 1971. Речник на говора на с. Съчанли, Гюмюрджинско. БД 6: 5–137. – – 1979. Говорът на тракийските преселници в с. Орешник, Елховско. БД , 3–79. – – Българските говори в западна (беломорска) и источна (одринска) Тракия. Бояджиев, Т. 1996. Характепистика на говорите. В кн.Странжда. Духовна и ма- териална култура. ЕИМ. Sofia. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 7 Бояджиева, Й. 1932. Кюстендилските полчане и техният говор. ИССФ 7: 8– 330. Бунина, И. К. 1953. Звуковой состав и грамматический строй говора ольшанс- ких болгар. СМБД 4: 5-24. – – 1963. К истории говора с. Дермендере. ВСЯ 7: –47. Вакарелска-Чобановска, В. 2005. Речник на самоковския говор. Sofia. Вачева-Хотева, М., Керемедчиева, С. 2000. Говорът на село Зарово Солунско. С оглед към лексикалната му система. Sofia. Верковикћ, Ст. 1985. Македонски народни песни, кн. 1. Женски песни. Подго- товил и редактирал Кирил Пенушлиски (primarily on basis of Верковић’s publication of 860). Skopje. Виденов, М. 1978. Годечкият говор. ТБД 0. Видоески, Б. 1957. Кичевскиот говор. МJ 8, 1: 31–90. – – 1958. Говорот и топонимиjата на кичевските села од таjмишата група. МJ , –2: 7–65. – – 62. Кумановскиот говор. ПИ 3. – – 85. Битолксиот говор. При 0, 2. – – 8. Диjалектите на македонскиот jазик, т. 1. Skopje. – – . Диjалектите на македонскиот jазик, т. 2. Skopje. ГЗ‑ФФУС = Годишен Зборник на Филозофскиот факултет на Универзитетот во Скопjе. Skopje. ГСУ‑ИФФ = Годишник на софийския университет. Историко‑филологически факултет. Sofia. ГСУ‑ФСФ = Годишник на софийския университет. Факултет на славянска фи‑ лология. Sofia. Георгиев 1907. Еркечаните и техният говор. ИССФ 2: 133–200. Геров, Н. 1895–1904. Речник на блъгарский язик. Sofia. Горов 1962 = Горов, Г. Странджанският говор. БД 1: 13-165. Господинкин, Д. 1921. Трънчаните и трънският говор. ИССФ 4: 48–2. Григоров, М. 1907. Говорът на Малореканците (мияците) в Дебърско. ИССФ 2: 201–304. Гъбюв, П. 1896. По говора на град Велико Търново. СНУ 13: 442–473. Гълъбов, Л. 2000. Речник на говора на с. Доброславци. Sofia. Даль, В. 1880 (1978). Толковый словарь великорусского языка, т. первый: А–З. St. Petersburg, Moscow. [Reprint edition] Демина, Е. И. 1968. Тихонравовский дамаскин. Болгарский памятник XVII в. Часть I. Исследование и текст. Sofia. – – 1985. Тихонравовский дамаскин. Болгарский памятник XVII в. Часть III. Ти- хонравский дамаскин как памятник болгарского языка XVII в. на народной основе. София. Денчев, В. 1970. По-особени думи и изрази в говора на с. Бракница, Поповско. БД 5: 245–256. – – 1979. Речников материал от говора на преселниците от с. Чаначка, Чаталд- жанско, в Шумен. БД 9: 343–349. Державин, Н.С. 1914–1915. Болгарские колонии в России, vols. –2. St. Peters- burg, Prague. 72 Slovenski jezik – Slovene Linguistic Studies 6 (2007) Димчев, К., Кювлиева, В. 1970. Речник на хасковския градски говор. БД 5: 53- 05. Дмитров, Л. 1905. Кратки етнографски бележки за Дебърско с оглед към говора в тоя край. ИССФ 1: 263–275. Динић, J. 1988. Речник тимочког говора. СДЗб 34: 7–337. Динић, J. 1990. Речник тимочког говора (други додатак). СДЗб 38: 379–586. Добрева, Д. 1986. Северодобруджанският говор на с. Горна Чамурла, Тулчанс‑ ко. Unpublished dissertation: Sofia. ДР = Дипломни работи. Skopje. Дрвошанов, В. 1993. Каjларскиот говор. Skopje. Евстатиева, Д. 1971. Лексиката на говора на с. Тръстеник, Плевенско. БД 6: 5- 243. Журвалев 1962 = Журавлев, В. К. Фонетика говора село Криничное. СМБД 0: 3-45. Златковић, Драгољуб. 1989. Фразеологија страха и нађе у пиротском говору. СДЗб 35: 175–457 Иванов, Й. 1932. Гевгелийският говор. Sofia. Илчев, С. 1962. Към ботевградската лексика. БД 1: 183–207. ИССФ = Известия на Семинара за славянска филология. Sofia. Кабасанов, Ст. 1956. Говорът на с. Момчиловци, Смолянско. ИИБЕ 4: 5–103. – – 1963. Един старинен говор, Тихомирският говор. Sofia. Каневска-Гиколова, Е. 200. Говорът на село Момчиловци, Смолянско – половин век по‑късно. Китипов, П. 1970. Речник на говор на с. Ениниа, Казанлъшко. БД 5: 07–4. Ковачев, С. 1968. Троянският говор. БД 4: 161-243. Ковачев, Н. 1970. Речник на говор на с. Кръвеник, Севлиеско. БД 5: 8–52. Конеска, М. 1951. Мариовскиот говор. ДР 3. Конески, Б. 1949. Прилепскиот говор. ГЗ‑ФФУС 2: 247–301. ---. 7. Речник на македонски jазик. Т. 1–3. Skopje. 2nd ed. Кочев 1969. Гребенският говор в Силистренско. ТБД 5. Куркина, Л. В. 1992. Диалектная структура праславянского языка по данным южнославянской лексики (= Slovenska akademija znanosti in umetnosti, Razred za filološke in literarne vede, Dela 38). Ljubljana: SAZU. Кънчев, И. 1968. Говорът на с. Смолско, Пирдопско. БД 4: 5–6. Кювлеива, К. Димчев, К. 1970. Речник га хасковския градски говор. БД 5: 60– 05. Любенов, Р. 1993. Бурел и с. Неделище, Софийско. Sofia. Мартинов, А. 1958. Брезнишки речник. СНУ 49: 773–789. Милетич, Л. 1923. Свищовски дамаскин. Новобългарски паметник от XVIII век. БгСт 7. Мирчев 1932. Неврокопският говор. ГСУ‑ИФФ 32. Младенов, М. 1967. Лексиката на ихтиманския говор. БД 3: 3-197. – – 1985. Характерисника на капанския говор. В Капанци. Бит и култура на старото българско население в севеноизотчна България. Етнографски и езикови проучвания: 311-338. Sofia. – – 1993. Българските говори в Румъния. Sofia. J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 73 Молеров, Д. 1905. Вторично ударение в разложкия говор. ИИСФ 1: 173–182. Молерови, Д. и К. 1954. Бележки за разложкия говор. СНУ 48: –6, 42–574. МПр = Македонски преглед. Sofia. Неделчев, Е. 1994. Диалект на Българите‑католици. Северен павликянски го‑ вор. Велико Търново. ОСА 1990 = Дыбо, В. А., Замятина, Г. И., Николаев, С. Л. 1990. Основы славян‑ ской акцентологии. Москва. Павлова, Н. 1988. Лозенградският говор. Канд. дисс. Sofia. Пеев. К. 1988. Кукушкиот говор, кн. II. Skopje. – – . Речник на македонските говори во jугоисточниот егеjски дел, т. 1. Skopje. Петков, П. 1974. Еленски речник. БД 7: 8–75. Петричев, Д. 1931. Принос към изучаване на трънския говор. ИССФ 7: 35–75. Петров, К. 1911. Принос към говора на град Котел. ИССФ 3: 195–222. ПИ = Посебни изданиjа. Skopje. Полтораднева-Зеленина, Э. И. 1955. Словарь говора болгарского села Суворо- во. СМБД 5: 64-134. Попгеоргиев, Ср. 1907. По говора в с. Чешнегировр Нова-Махала, Станимашко. ИССФ 2: 4–445. Попиванов, Г. 1940а. Допълнителен речник на Орханиския говор. СНУ 34: 309– 324. – – 1940b. Особености на шуменския говор. Допълнения към описанието му в “Das Ostbulgarische” на проф. Л. Милетича. СБАН 34: 329–464 – – 1962. Бележки за говорите в Севлиевско. Троянско, Тетевенско. БД : 65– 82. Попов, К. 1956. Говорът на с. Габаре, Белослатинско. ИИБЕ 4: 103–177. Попоски, А. 1959. Реканскиот говор. МJ 0: 07–54. Ралев, Л. 1977. Говорът на с. Войнягово, Карловско. БД 8: 3-201. РБЕ = Речник на българския език, т. 3. Г – Деятел. Sofia, БАН-ИБЕ. 1981. Реметић, С. 1989. Српски призренски говор I (гласови и облици). СДЗб 35: 319- 64. Речник на македонската народна поезиjа. Skopje. Савов, В. 1931. Ловчанските помаци и техният говор. ИССФ 7: 1–35. СБАН = Списания на Българска академия на науките. СДЗб = Српски дијалектолошки зборник. Селищев, А.М. 1929. Полог и его болгарское население. Sofia. СМБД = Статьи и материалы по болгарской диалектологии. Moscow. Соболев, Андрей Н. 1994. Говор села Вратарница в восточной Сербии в исто‑ рическом и ареальном освещении. (= SlBei 320). Munich. Срезневский, И. И. 1893 (1989). Словарь древнерусского языка, Том первый: А–К (reprint edition). Moscow. Стоилов, Х. П. 1904. Горно-джумайски говор. СНУ 20: 1–33. – – 1905. Ударението на горноджумайския говор. ИССФ 1: 183–217. Стойков, С. 1962. Българска диалектология. Sofia. Стойков, Ст. 1956. Днешното състояние на еркчекия говор. ИИБЕ 4: 356–365. – – 1968. Лексиката на банатския говор. ТБД 4. 326 с. 74 Slovenski jezik – Slovene Linguistic Studies 6 (2007) – – 1970. Рупските особености в съртския говор. ИИБЕ : 70–72. – – 1971. Говорът на с. Мугла. ИИБЕ 20: 3–90. Стойков, Ст. и др. 1956. Говорът на с. Говедарци, Самоковско. ИИБЕ 4: 25– 331. Стойчев, К. С. 1915. Тетевенският говор. СНУ 31. Стойчев, Т. 1965. Родопски речник. БД 2: 119-315. – – 1970. Родопски речник. БД 5: 153–220. – – 1983. Родопски речник. РСб 5: 287–353. Стоянов, Ст. 1972. Говорът на с. Макоцево, Софийско. ГСУ‑ФСФ 65: 151–312. ТБД =Трудове по българска диалектология. Sofia. Тодоров, Ц. 1936. Северозападни български говори. СНУ 4. Тодорова, Д., Коцев, Д. 1999. Езикът на моите деди. Sofia. Толстой, Н. И. 1974 (1997). Некоторые вопросы соотношения лингво- и этног‑ рафических исследований. Проблемы картографирования в языкознании и этнографии: 16–33. Лениград. [Reprinted in: Н. И. Толстой. Избранные труды, том. I: 223–242.] Ћирић, Љубисав. 1983. Лужнички говор. СДЗб 2. Угринова, Р. 1951. Говорите во Скопско. Skopje. Умленски, Ив. 1965. Кюстендилският говор. ТБД . Филипова-Байрова, М. 1969. Гръцки заемки в съвременния български език. Sofia. Хитов, Хр. 1979. Речник на говора на с. Радовене, Врачанско. БД 9: 223–342. Христов, Г. 1956. Говорът на с. Нова Надежда, Хасково. ИИБЕ 4: 24–24. Цанова, С. 1986. Говорът на с. Мандър, Мала Азия. Канд. Дисс., Sofia. Чешко, Е. В. 1952a. К истории кайраклийского говора (Болгарский говор села Лощиновки Болградского района Измайловской области УССР). СМБД 2: 2-76. – – 1952b. Из итогов второй диалектологической экспедиции в болгарские села измайловской области УССР, проведеной летом 1949 года. СМБД 2: 76-84. Шапкарев, К. А. 1891–1904/2001. Градиво за български речник. I. Общи българ‑ ски думи. Швецова, Т. В. 1952. Материалы для словаря болгарских говоров СССР. Сло- варь говора села Твардицы. СМБД 2: 84-. Шклифов, Б. 1973. Речник на костурския говор. БД 8: 201–328. – – 1977. Костурският говор. ТБД 8. – – 1979. Долнопреспанският говор. ТБД . Шклифов, Б., Шклифова, Е. 2003. Български диалектни текстове от егейска Македония. Sofia. Цанова, С. 1986. Говорът на Мандър, Мала Азия. Unpublished dissertation: Sofia. ЭССЯ = Этимологический словарь славянских языков. Moscow. Якимова, М. 1911. Охридският говор. ИССФ 3: 223–256. Prispelo januarja 2007, sprejeto februarja 2007 Received January 2007, accepted February 2007 J. Schallert and M. L. Greenberg, The Prehistory and Areal Distribution ... 75 Prazgodovina in zemljepisna razvrstitev psl. *gъlčěti ‘govoriti’ V slovanskih knjižnih jezikih je izpričano sorazmerno majhno število glagolov za pojem ‘govoriti’. Poleg stcsl. glagolati najdemo še bolg. говоря, bhs. govoriti, blr. гаварыць, rus. говорить, slš. hovorit’, sln. govoriti; češ. mluvit, pol. mówić, ukr. мовити; mak. зборува; dls. rjac, gls. rěčeć. Slika v narečjih je precej bolj zapletena. Med leksemi za pojem ‘govoriti’, ki so izpričani izključno v narečjih, je tudi psl. *gъlčěti, ki je prvotno pomenil ‘povzročati hrup, šumenje; kričati’, kot kažejo starejši pisani primeri, npr. stcsl. gъlkъ, ‑a ‘hluk, шум, Lärm, tumultus’; stčeš. hlučěti, ‑u, ‑íš ‘hlučeti, křičeti, schallen, lärmen, schreien, rufen’, hluk, -a ‘hřmot, křik, prudkost, nepokoj, Lärm, Geschrei, Unruhe, Ungestüm’; strus. гълчати, гълчу ‘кричать’, гълчание ‘шумъ, крикъ’, гълка ‘шумъ, мятежь’. V nekaterih slovanskih narečjih se je pomen razvil v ‘govoriti’, ki je danes izpričan na treh različnih področjih: v osrednjih in severnih ruskih, v bolgarskih in v slovenskih narečjih. Enak pomenski razvoj se opaža tudi pri drugih leksemih tako v indoevropščini kot v slovanščini. Lek- sem *gъlčěti je povsem odsoten v bhs. knjižnih jezikih; izpričan je samo v obrobnih narečjih, in sicer v kajkavščini pod vplivom slovenščine in v srbščini pod vplivom bolgarščine. Razprava podrobno obravnava zemljepisno razvrstitev in pomenski razvoj lekse- ma *gъlčěti v južnoslovanskih narečjih, tj. na področjih z migracijsko poselitvijo, za- stavlja pa tudi vprašanje o prvotni povezavi med jezikovnimi skupnostmi, v katerih se je leksem razvil v glavni izraz za pojem ‘govoriti’. V obravnavi se precizira zemlje- pisna razvrstitev leksema, ki je tu prvič podana sintetično za celotno južnoslovansko področje. Avtorja prihajata do zaključka, da je verjetna povezava med slovenskim panonskim narečjem in arhaičnimi bolgarskimi narečji in da ta povezava sega v čas naselitve. Leksem so prinesli migranti po Donavi na Balkan s severovzhoda (kot je razvidno iz enakega razvoja v osrednjih in severnih ruskih narečjih), nato pa v sloven- sko panonsko narečje. Analiza podpira domnevo, da so slovanske selitve v podalpski in balkanski prostor vključevale raznovrstne praslovanske narečne skupine. The Prehistory and Areal Distribution of Slavic *gъlčěti ‘Speak’ The Slavic standard languages attest a relatively small number of expressions for the notion ‘speak’. In addition to OCS glagolati one finds today BCS govoriti, Be га‑ варыць, Bg говоря, Ru говорить, Sk hovorit’, Sn govoriti; Cz mluvit, Po mówić, Uk мовити; Ma зборува; LS rjac, US rěčeć. However, the picture in Slavic dialects is much more complex. Among those lexemes for ‘speak’ that occur only in the dialects is PS *gъlčěti, a verb whose meaning was originally ‘make sound/noise’, as is evi- denced by older attestations, e.g., OCS gъlkъ, ‑a ‘hluk, шум, Lärm, tumultus’; OCz hlučěti, ‑u, ‑íš ‘hlučeti, křičeti, schallen, lärmen, schreien, rufen’, hluk, -a ‘hřmot, křik, prudkost, nepokoj, Lärm, Geschrei, Unruhe, Ungestüm’; ORu гълчати, гълчу ‘кричать’, гълчание ‘шумъ, крикъ’, гълка ‘шумъ, мятежь’. The verb shifted to the meaning ‘speak’ in a subset of Slavic dialects, currently attested in three disparate regions of Slavic–central and northern Russian dialects and, in the South Slavic area, Bulgarian and the Slovene dialects, a semantic shift that is paralleled in a number of 76 Slovenski jezik – Slovene Linguistic Studies 6 (2007) cases both at the Indo-European and Slavic levels. The lexeme is lacking altogether in the BCS standard languages and is attested all but marginally in the dialects associ- ated with them, this being due to secondary influence from the direction of Slovene as regards Kajkavian Croatian and from Bulgarian as regards Serbian. The paper examines in some detail the diatopic distribution and semantic devel- opment of *gъlčěti in South Slavic, i.e., the Slavic dialect areas settled by migration, and raises the question of the nature of the relationship among those dialects that have developed *gъlčěti as the primary neutral verb meaning ‘speak’. In the process, the precise geographical distribution of the lexeme is made for the first time in a synthetic manner, covering the whole South Slavic area. The authors reach the conclusion that there is a probable connection between Pannonian Slovene and archaic Bulgarian dialects with regard to this lexeme, dating to the time of settlement, that was carried via the Danube to the Balkans from the northeast (as reflected in central and north- ern Russian dialects today) and then northwards to Slovene Pannonian dialects. The analysis supports the view that the Slavic migrations to the sub-Alpine and Balkan regions were of heterogeneous dialectal origins.