Mate Kapović ORCID: 0000-0001-7938-4323 University of Zagreb, Faculty of Humanities and Social Sciences, Croatia kapovic@gmail.com On archaic oxytonesis in Slovene Ter dialect DOI: https://doi.org/10.3986/15.1.04 Slovenski jezik / Slovene Linguistic Studies 15/2023. 75–87. ISSN tiskane izdaje: 1408-2616, ISSN spletne izdaje: 1581-127 https://ojs.zrc-sazu.si/sjsls Slovenski jezik – Slovene Linguistic Studies 15 (2023): 75–87 @language: sl, en, de, it, fr @trans-language: sl, en, de, it, fr @publisher.id: id @doi: 10.3986/00.0.00 @article-type: 0.00 @article-category: category @pages: 75–87 @history-received: dd. mm. yyyy @history-accepted: dd. mm. yyyy *** Žurnal meta *** @issue: xx @volume: 15 @pub-year: 2023 @pub-date: dd. mm. yyyy *** Oprema *** @avtorji: Mate Kapović @running-header: On archaic oxytonesis in Slovene Ter Mate Kapović (ORCID: 0000-0001-7938-4323) University of Zagreb, Faculty of Humanities and Social Sciences, Croatia kapovic@gmail.com DOI: https://doi.org/10.3986/15.1.04 On archaic OxytOnesis in s lOvene t er dialect The paper discusses the oxytonic mokȁ ‘flour’, bradȁ ‘beard’ type accent in the central Prosnid–Porčinj–Subid belt of the Ter dialect base on the material from Janoš Ježovnik’s 2022 monograph. The author discusses various problems concerning the oxytonesis and concludes that these marginal Slovene dialects preserve the old Slavic final accent even after the originally long root. Keywords : Ter dialect, Tersko dialect, accent, accentology, oxytonesis Prispevek obravnava oksitonski naglas mokȁ ‘moka’, bradȁ ‘brada’ v osrednjem prosnidsko–porčinjsko–subidskem pasu terskega narečja na podlagi gradiva iz monografije Janoša Ježovnika (2022). Avtor obravnava različne probleme v zvezi z oksitonezo in ugotavlja, da ta obrobna slovenska narečja ohranjajo stari splošnoslovanski končni naglas tudi po prvotno dolgem korenu. Ključne besede : tersko narečje, tersko narečje, naglas, akcentologija, oksitoneza 1 Introduction Šekli (2006: 168, 173) published the material where he shows different reflexes of Proto-Slavic *nva and *zīma̋ in the local dialect of Subid, a part of the Slovene Ter dialect (tersko narečje). The words in question are ńíu̯a ‘field’ and zímȁ/zīmȁ ‘winter’. 1 Unlike Šekli, who interpreted the zímȁ/zīmȁ type accent as an innovation (thus *zīmȁ > *zíma > zímȁ/zīmȁ), Kapović (2015: 84) on basis of this very scant evidence cautiously proposed that zīmȁ is actually an archaism, i.e. the old preserved final accent. Recently, Ježovnik (2020 and especially his 2022 monograph) provided much more material from Ter dialects which have the same type of accentuation – 1 Unlike Šekli (2006) and Ježovnik (2020, 2022), in this article I write the Slovene tonemes in the traditional manner: ̏ – short (falling), ̑ – long falling, ´ – long rising. 76 Slovenski jezik – Slovene Linguistic Studies 15 (2023) however, not from Subid but from neighboring dialects of Porčinj and Prosnid, which exhibit the trau̯ȁ ‘grass’ < *trāva̋, bradȁ ‘beard’ < *borda̋ type accent with the apparently preserved old final accent. Ježovnik (2020: 673–675, 2022: 377–382, 391) provides a thorough and very balanced overview of both possibilities – one being that the bradȁ type accent is archaic and the other that it is an innovation. However, he surmises (2020: 674) that “each is near impossible to prove conclusively” (of course, this can be said of everything in historical linguistics) and gives preference to the traditional communis opinio in Slovene dialectology (Ježovnik 2022: 377) that bradȁ type accent is an innovation (due to a progressive shift from the earlier retracted *bráda). This is clear, among other things, already from the title of his 2020 paper (“progressive acute-accent shift”) and the actual reconstructions he gives in Ježovnik 2022, e.g. Proto-Slovene *zːma (his transcription) for Porčinj zimȁ (: 286). It also has to be said that though one can indeed find certain arguments to try to argue for the *zīmȁ > *zíma > zimȁ type process (as Ježovnik does), there is in reality no real factual reason in the Ter dialect why one should even try to look for those (except for reluctance to stray away from the traditional suppositions on accentual development in Slovene historical linguistics) and not just take the attested factual material at face value. In this paper, I shall shortly discuss the very valuable material presented by Ježovnik and argue for the interpretation of words like zimȁ as archaisms and not innovations. Most of the problems have already been well discussed by Ježovnik in his two works, but I shall try to add my perspective and interpretation to it, as well as to discuss certain additional examples (like the accent of the infinitive). 2 Material Here, we shall shortly present some of the relevant material (word- forms with original length and accent on final open syllable) from Ježovnik 2022 (a short list of the relevant forms is available in Ježovnik 2020: 670 as well), without going into all the details (e.g. we will not list all types of nouns/adjectives/verbs that have the oxytonesis in some form, we will not discuss all obvious analogies at length, we will not adduce all examples from the monograph from both villages, etc.). Unfortunately, Ježovnik’s material is not comprehensive and certain Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 77 possibly interesting forms with original oxytonesis are missing, e.g. accentual paradigm c locative singular o-, ā- and i-stem forms, a. p. c imperative forms, etc. (while e.g. neuter nominative forms are rare because of the demise of the neuter). 1) ā-stem singular (a. p. B & C): Porčinj/Prosnid mokȁ ‘flour’ – acc sg mokȕ; trau̯ȁ ‘grass’ – Porčinj acc sg trau̯ȕ 2 ; bradȁ ‘beard’, zimȁ ‘winter’ 3 etc. (cf. the original initial accent in kráu̯a ‘cow’, lípa ‘linden’, níu̯a ‘field’ 4 ; Porčinj žẹ́ja ‘thirst’ 5 ; Porčinj kóža 6 ). 2) o-stem gen sg (a. p. B): Porčinj/Prosnid (ɣ)riẹxȁ ‘sin’, Porčinj kjučȁ ‘key’, križȁ ‘cross’, klabukȁ ‘hat’, etc. 7 (cf. the original initial accent in Porčinj siẹ́ra ‘cheese’, zẹ́ta ‘son-in-law’ 8 ) 3) neuter o-stem (a. p. B): Prosnid ɣniẹẕdȍ ‘nest’ – gen sg ɣniẹẕdȁ, Porčinj/ Prosnid mliẹkȍ ‘milk’ – gen sg mliẹkȁ, vinȍ ‘wine’ – gen sg vinȁ 9 (cf. the original initial accent in Porčinj liẹ́to ‘summer, year’, Prosnid s̱íto ‘sieve’ 10 ) 4) adjective (a. p. B & feminine singular a. p. C): Porčinj/Prosnid biẹlȁ ‘white’ (feminine) – Prosnid biẹlȍ (neuter) 11 , Porčinj/Prosnid liẹpȁ 2 Ježovnik 2022: 282–283 (cf. Croatian dialectal mūkȁ, trāvȁ). The word ẕviẹẕdȁ ‘star’ shifts to a. p. C in Prosnid (acc sg ẕviẹẕdo with a progressive shift of the circumflex) – in Porčinj, the circumflex is then regularly phonetically retracted in the accusative and generalized, thus yielding Porčinj innovative zviẹ́zda – acc sg zviẹ́zdu. 3 Ježovnik 2022: 286 (cf. Croatian dialectal brādȁ, zīmȁ). The word bradȁ shifts to a. p. B in both Porčinj and Prosnid (secondary acc sg bradȕ); ɣlau̯ȁ ‘head’ and ẕimȁ remain a. p. C in Prosnid (acc sg ɣlau̯o, ẕimo), while lau̯ȁ and zimȁ shift to a. p. B in Porčinj (secondary acc sg lau̯ȕ, zimȕ); petȁ ‘heel ’ and rokȁ ‘arm’ remain a. p. C in Prosnid (acc sg peto, roko), while the accusative form with the regular phonetic retraction of the circumflex is gene- ralized in Porčinj (thus pta, rka – acc sg ptu, rku) 4 Ježovnik 2022: 277 (cf. Croatian krȁva, lȉpa, ńȉva). 5 Ježovnik 2022: 281 (cf. Croatian dialectal žẽja). 6 Ježovnik 2022: 282 (cf. Croatian kȍža). 7 Ježovnik 2022: 296–298 (cf. Croatian dialectal grīxȁ, kļūčȁ, krīžȁ, klobūkȁ). 8 Ježovnik 2022: 290 (cf. Croatian sȉra, zȅta). 9 Ježovnik 2022: 311 (cf. Croatian dialectal gńīzdȍ – gńīzdȁ, mlīkȍ/mlēkȍ – mlīkȁ/ mlēkȁ, vīnȍ – vīnȁ). 10 Ježovnik 2022: 309 (cf. Croatian dialectal lȉto/ļȅto/lȅto, sȉto). 11 Ježovnik 2022: 332 (cf. Croatian dialectal bīlȁ/bēlȁ – bīlȍ/bēlȍ). 78 Slovenski jezik – Slovene Linguistic Studies 15 (2023) ‘beautiful’, mladȁ ‘young’, Porčinj suxȁ ‘dry’ 12 (feminine) (cf. the original initial accent in Prosnid dúɣa ‘long’ – dúɣo, s̱tára ‘old’ – s̱táro 13 ) 5) l-participle: Porčinj treslȁ ‘shook’ 14 As can be seen from these and other examples, the difference of the original oxytone forms (like bradȁ) and original forms with an initial old acute (like lípa) or neo-acute (like žẹ́ja) is systematic and regular, in spite of a few analogies and later secondary developments. 3 d iscussion In this section, we shall discuss certain issues concerning the oxytonic forms and their origin, some of which were already discussed by Ježovnik (2020, 2022). The communis opinio in Slovene dialectology/ accentology is that the retraction of the short accent in final open syllables (e.g. *trāvȁ > tráva) is old and that it encompasses all Slovene dialects – cf. e.g. Greenberg 2000: 120 (though Greenberg’s wording is not completely clear – in any case, this monograph was written before the new data from the Ter dialect was published), Šekli 2018: 152–153 (thus after and in spite of the data in Šekli 2006). From this follows that the data from the Ter dialect of the bradȁ type is considered to be secondary (cf. Ježovnik 2020: 673, 2022: 377). As we shall see, this traditional view has to be corrected. 12 Ježovnik 2022: 335 (cf. Croatian dialectal līpȁ/lēpȁ, mlādȁ, sūxȁ). 13 Ježovnik 2022: 327 (cf. Croatian dialectal dȕga – dȕgo, stȁra – stȁro). Porčinj has an innovative C-end-accent in feminine mərzlȁ ‘cold’, slabȁ ‘weak’, rusȁ ‘red’, starȁ ‘old’ – cf. Prosnid original and expected mə ̀ rẕla, rús̱a, s̱tára. Prosnid s̱ȋta ‘fed up’ is secondary just like Porčinj sȋta and slȃba, probably due to analogy with the original long/definite forms (the same analogical accent occurs also in Porčinj dȗa ‘long’ – cf. the expected accent in Prosnid dúɣa). The end-stress in Porčinj forms like starȁ was probably enhanced due to neuter forms like stáro, where the rising accent (originally stemming from the old acute) can be misanalyzed as the rising accent originating in the regular retraction of the circumflex in Porčinj. 14 Ježovnik 2022: 343 (cf. Croatian dialectal trēslȁ). In other l-participle forms, one or another kind of analogy occurred, e.g. Porčinj strslo is an innovative C-form, zrástla is analogical to the infinitive and other forms, začla ‘begun’ is analogical to the (already analogical to non-prefixed forms) neuter form (Ježovnik 2022: 343–44), etc. Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 79 While Subid, Prosnid and Porčinj 15 have a final accent in mokȁ type words, which used to have length in the root (and still do to an extent – see below), all the Ter local dialects have a final accent in the kozȁ ‘goat’ type (cf. Croatian dialectal mūkȁ but kozȁ for the original presence of root brevity/length), e.g. Porčinj sestrȁ ‘sister’, tetȁ ‘aunt’, ženȁ ‘woman’, Prosnid meɣlȁ ‘fog’, temȁ ‘darkness’, etc. 16 While words like bradȁ are usually interpreted as secondary by Slovenian dialectologists, nobody doubts that forms like kozȁ are old. Looking from a perspective of the Ter dialect itself and general historical Slavic accentology, this does not make much sense – why would kozȁ be old while bradȁ would not? Why not take the data simply at face value? Why try to imagine complex hypotheses in order to avoid a simple archaism? The Ter dialect is the westernmost Slovene dialect and marginal dialects very often preserve archaisms that are lost elsewhere – thus, it is not strange at all that bradȁ type accent could be preserved there. The problem of the Slovene dialectology is that it looks at Slovene dialects as one isolated whole and thus finds it problematic that final accent is lost everywhere in Slovene except in a part of the Ter dialect (though this is also not true – see below). However, modern national borders of Slovenia and modern political unity of Slovene dialects do not really have much meaning in historical linguistics and accentological isoglosses, especially if we keep in mind that the whole of South Slavic linguistic area is one huge dialectal continuum. The fact that there is no oxytonesis e.g. in glava and brada in Prekmurje (in the far north-east of Slovenia) should not really be more important than the fact that Croatian/Čakavian central Istra 17 does have oxytonesis in those words – in fact, Istra Čakavian is geographically closer to the Ter dialect than many eastern Slovene dialects are. Buzet Čakavian dialects, indeed, show certain traits in common with Slovene dialects (e.g. progressive shift of the circumflex) and do mostly preserve the 15 Also perhaps Malina/Forame in that area (Ježovnik 2020: 667 7 ). 16 Ježovnik 2022: 283–285, 287–288 (cf. Croatian dialectal kozȁ, sestrȁ, tetȁ, ženȁ, maglȁ). 17 Here I use the Croatian (and Slovene) name Istra and not the Italian version Istria, which is otherwise more frequent in English. 80 Slovenski jezik – Slovene Linguistic Studies 15 (2023) final accentuation (except certain dialects near the Slovenian border) – cf. now Klarić 2022: 129. If Istra Croatian Čakavian could preserve old accentuation in words like that, the same can be true for the Ter dialect of Slovene, the modern national identities notwithstanding. In fact, the Ter dialect preserves other archaisms, just like Čakavian does (and most of Slovene dialects do not) – for instance, the opposition of č and ć. 18 What is more, Ježovnik (2020: 675, 2022: 381) adduces some other instances of preservation of the old bradȁ type accent in other Slovene dialects, of which Breznik’s examples like rokȁ ‘arm’, zimȁ ‘winter’ from Kobarid (a Nadiško dialect just east from the Ter dialect) are the most relevant. This would point to the preservation of the old final accent as a feature of a wider territory in the westernmost Slovene dialects up until recently. The whole idea that old *brādȁ yielded first *bráda (while kozȁ remained unchanged), and then later changed again to bradȁ is possible theoretically but is completely unnecessary – Ježovnik (2022: 380) admits himself that the archaism is more probable according to the Ockham’s razor principle. Considering the contemporary differences of zimȁ and lípa, one would also have to assume that the supposedly retracted *zíma had a long rising accent that was somehow different from the long rising accent in lípa (Ježovnik 2020: 674, 2022: 378, 391) because only one of those eventually shifted to the right again. 19 Having two different rising tonemes is not impossible (cf. e.g. Old Štokavian Croatian dialects in Posavina which have both sũša ‘drought’ with the neo-acute, which can be phonetically rising, and rúka ‘arm’ with a Neo-Štokavian long rising accent), but is not very likely and there is no independent evidence for that – except for the apparent unwillingness of some Slovene dialectologists to consider the possibility that marginal dialects can indeed preserve archaisms that are not usual in non-marginal dialects. Ježovnik (2022: 381) argues that one could find a parallel for the supposed possible *brādȁ > *bráda > bradȁ in Common Slavic process of the rightward shift of Dybo’s law and then 18 Ježovnik 2022: 194. 19 The difference cannot be in the original brevity in lípa because there is no forward shift in the žẹ́ja type either (which is originally long). Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 81 subsequent leftward shift of Ivšić's (Stang’s) Law. However, it must be said that Stang’s law is highly dubious in Slavic, one of the arguments being that it is unnecessary to suppose first the rightward and then later the leftward shift in order to get the same form as in the beginning. 20 Ježovnik’s (ibid.) point that we find a forward and then backward shift of the circumflex in some Slovene dialects, including some of the Ter dialects (like Porčinj), is not a good parallel because in case of the development such as *mȇso > *mesȏ > *méso we do not posit such a right-left development simply because it is assumed that all Slovene dialects must first have a progressive shift of the circumflex – Porčinj mẹ́so indeed has a rising (!) accent and, for instance, Prosnid still has the older long falling accent in mes̱ȗọ. 21 Thus, in case of that right-left shift there are plenty of independent evidence for such a process – in case of the bradȁ type Porčinj/Prosnid accent there is no such thing. The preservation of the pre-oxytonic length, with the identical development of vowel quality in pretonic and stressed positions 22 (cf. e.g. Porčinj niẹzdȍ and koliẹ́no ‘knee’ 23 ), is not a typological problem. As Ježovnik (2022: 380) notes, a parallel process (the identical development of long pretonic and long stressed vowels), can be seen in Bednja Kajkavian, cf. e.g. Bednja 24 zvīēzdȍ ‘star’ = črȋešńo ‘cherry’ (with the same diphthong in long pretonic position and under long falling accent). The same process can be seen e.g. in Orbanići Čakavian 25 zvi̯ezdȁ ‘star’ = bi̯ẽli ‘white’ (with the same diphthong in long pretonic position and under the long neo-acute). Here it also has to be mentioned that the preservation of pretonic length in bradȁ type words in the archaic part of the Ter dialect is not just a theoretical supposition. The oldest of Ježovnik’s informants in Prosnid (and to a lesser extent some other, younger, informants) seem to preserve inconsistent pretonic length in the bradȁ type words (cf. Ježovnik 20 Cf. e.g. Hendriks 2003, Kapović 2015: 103–134, 2017b: 391 22 , 2020a. 21 Ježovnik 2022: 315. 22 Ježovnik 2020: 675, 2022: 379. 23 Ježovnik 2022: 310–311. 24 Jedvaj 1956: 301–302. 25 Kalsbeek 1998. 82 Slovenski jezik – Slovene Linguistic Studies 15 (2023) 2020: 667, 2022: 380). 26 It may just be that pretonic length has started to disappear in that position only in the last couple of generations (Ježovnik 2020: 675). One of the possible pro-innovation arguments Ježovnik (2020: 674, 2022: 378–379) uses is the circumflex in the nominative plural of neuter accentual paradigm B (oxytonic) nouns. For instance, in Porčinj we find the expected singular oxytonic niẹzdȍ in a. p. B (as opposed to liẹ́to in a. p. A) but plural nȉẹzda (Bː) = lȉẹta (A). Now, Ježovnik thinks that this analogy could point to the original *gniẹ́zdo (with retraction) = liẹ́to, which made the analogy in the plural possible. However, as he himself admits (Ježovnik 2020: 674, 2022: 379), the supposed retraction has to entail a different (!) rising tone, which makes the whole supposition improbable. In any case, presupposing a retraction and then a progressive shift is hardly necessary to account for the secondary nȉẹzda type. First of all, this is basically just one example (there is also a morphologically secondary nom pl mlȉẹke from mliẹkȍ) and could have easily originate in a simple analogical spread of the circumflex in neuter nom pl regardless of the accent in singular. One may even speculate that this might have also been helped via contact influence of other dialects with the retraction (cf. Breginj ɣnię́zdo – nom pl ɣnȉę̄zda). Analogies are often inconsistent so it is not strange that there is no secondary circumflex in pluralia tantum u̯ráta ‘door’ and pléća (Porčinj). The form rébra (Porčinj) also shifts (partially) to the feminine gender 27 so this may have had an influence as well. In any case, it is just one or two examples and a simple enough analogy – this is certainly not something that would prove the *brādȁ > *bráda > bradȁ process. As for the progressive shift of the acute in propenultimate syllables that occurs in the area of Prosnid, Porčinj and Subid 28 (e.g. *jùžina ‘lunch’ > Porčinj južȅ̥na 29 ), this is clearly a different process 30 (in which the old 26 As a parallel for an inconsistent retention of prefinal length cf. e.g. the case of the Čakavian dialect of Susak in Croatia (Kapović 2020b: 509–510). 27 Ježovnik 2022: 378 267 . 28 Cf. the map in Ježovnik 2022: 384. 29 For examples cf. Ježovnik 2020: 667–668, 2022: 278–279, 291, 309, etc. Additional examples for Subid are given in Zuljan Kumar 2018: 42. 30 Acknowledged also by Šekli 2006, Ježovnik 2020: 666, 673, 2022: 377. Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 83 acute from the propenultima is shifted to a medial syllable, while the old acute does not shift from the penultima to a final open syllable in the lípa type), and one that is clearly young as apparent from it being blocked by the disappearance of internal -i- (e.g. Porčinj *pàlica > pálca). 31 While it is interesting that this innovative progressive shift occurs in the same area that preserves old oxytonic accent, this shift cannot be bundled together with the one that supposedly produced the bradȁ type. The accentual paradigm b is rare in i-stems (cf. Kapović 2009). In Porčinj, *lũčь ‘light’ (Czech louč ‘torch’ and Standar Slovene lúč – gen sg lúči/lȗči) shifts to o-stems (lúč – gen sg lučȁ), while preserving the old a. p. B. However, in Porčinj pt ‘path’, we find the retracted accent in gen sg pte̥ – this must be due to analogy to the nom sg , cf. in Čakavian (Vrgada 32 ) pũt – gen sg pũta (instead of the original pūtȁ, attested in other Čakavian dialects). This kind of analogical accent is also found in Prosnid gen sg lúč̱, pọót (with the loss of the final vowel). 33 Finally, the accent of the infinitive has to be mentioned. In the archaic Ter area with the bradȁ oxytonic type, the short-stem infinitives preserve the old final accentuation, e.g. Porčinj ponestȅ̥ ‘to carry with’, pejćȅ̥ ‘to bake’ 34 (cf. Neo-Štokavian pèći and younger pȅći). However, in long-stem infinitives, only the retracted accent is attested: Porčinj u̯zẹ́te̥ ‘to take’ (cf. Neo-Štokavian dialectal uzéti but standard ùzēti, dialectal uzẽti), ráste̥ ‘to grow’ (cf. Neo-Štokavian rásti and younger rȃsti, dialectal rãsti), trẹ́ste̥ ‘to shake’ (cf. Neo-Štokavian trésti and younger trȇsti, dialectal trẽsti), začẹ́te̥ ‘to start’ (cf. Neo-Štokavian dialectal začéti but standard zàčēti, dialectal začẽti), otpẹ́te̥ ‘to untie’ (cf. Neo-Štokavian dialectal otpéti but standard òtpēti, dialectal otpẽti), 31 Ježovnik 2020: 668. 32 Jurišić 1973. 33 The Ter forms are from Ježovnik 2022: 319. 34 Ježovnik 2022: 347. In some Ter dialects, curious secondary forms like ptȉ ‘to be’ are found (Ježovnik 2022: 465). This must be due to a number of factors: the secondary differentiation from bíti ‘to hit ’, the influence of short-root infinitives in -tȉ and perhaps due to influence to forms parallel to Standard Slovene b(ə)lȁ – b(ə)lọ̑ (with the allegro reduction of the first syllable) or the like. 84 Slovenski jezik – Slovene Linguistic Studies 15 (2023) otpriẹ́te̥ ‘to open’ 35 (cf. Neo-Štokavian dialectal uprijéti but standard ùprijēti, dialectal uprijẽti). Thus, Porčinj points to an original *pećȉ but *rásti (from older *rāstȉ). The second type could be taken as an example of the original retraction in the bradȁ type in general, however infinitives are hardly conclusive in this regard. In infinitives (and in some other forms), there is a tendency in Kajkavian, Čakavian and Western Štokavian – and now, as we can see, obviously in (at least part of) Slovene – to morphologically/non-phonetically unstress the ending (probably due to the fact that the infinitive *-ti is originally unstressed in most verbal classes). For this cf. Kapović 2015: 633–638, 2017a: 612–613, 615–616. Thus, the Porčinj opposition of pejćȅ̥ and ráste̥ would be like the opposition of pèći and rȃsti in some Neo-Štokavian dialects (the details of this morphological process are very diverse dialectally. This old *rãsti type (with the long neo-acute) may be quite relevant for the wider Western South Slavic dialectology and historical accentology because it points to this process being not only Kajkavian/ Čakavian/Western Štokavian, as previously thought, but also Slovene as well (in most Slovene dialects, the original *rāstȉ and *rãsti cannot be distinguished – cf. the Standard Slovene rásti). One must also note that in some Ter dialects, including Prosnid, the accent cannot even possibly synchronically be at the ending in the infinitive because it is lost – cf. Prosnid (Ježovnik 2022: 341, 343) u̯ẕẹét, rás̱t, etc. (the same is in Štokavian and Čakavian, where -ti is not always preserved). As for the l-participles, as we have already said, the old accent is preserved in Porčinj treslȁ, while in other forms it is innovative – Porčinj feminine u̯zẹ̑la is analogical to masculine u̯zẹ̑u̯, zrástlo and strẹ́slo are secondary C-forms (with the retracted circumflex), etc. 4 conclusion From what we have seen, the easiest explanation for the attested material in the central dialects of Tersko (Prosnid-Subid-Porčinj + probably Malina as well) is that they preserve the old oxytonic accentual type bradȁ, zimȁ. This is the westernmost Slovene dialect and archaisms are to be expected in the margins and, in any case, if such an accentuation 35 Ježovnik 2022: 341–346. For Štokavian, cf. Kapović 2018: 185, 188, 198, 201. Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 85 can be preserved in e.g. Buzet area of North Čakavian, some 150 kilometers away, there is no reason why it could not be preserved in the central Ter dialects. Ježovnik’s (2020: 674) pessimistic conclusion that “[b]oth hypotheses have considerable implications and each is near impossible to prove conclusively” is not really useful (though he presents both sides of the arguments fairly). The fact of the matter is that the central belt of the Ter dialect shows the oxytonic accent in bradȁ type words and that we have no real reasons to presume that this is not an archaism. 36 The only reason why there is reluctance to accept forms like mokȁ, bradȁ, zimȁ as archaisms seems to be the reluctance of Slovene dialectology to give up on the long held notion that the retraction of *zīmȁ > *zíma is pan-Slovene and “Proto-Slovene” (even though the need to accept two different rising accents in “old central- Ter” would also change the traditional reconstruction and conceptions) and thus preserve the notion of the historical unity of Slovene. However, even in that regard this is not necessary since Slovene dialects share other important accentological isoglosses such as the neo-circumflex in a wide variety of positions. In any case, the argument that this kind of oxytonesis does not exist elsewhere in Slovene (though this is also probably not true, considering the Kobarid zimȁ type accent) is not a sustainable argument – it speaks more on tradition in Slovene historical linguistics than on the Ter dialect material itself. The final conclusion can be: if it looks like a duck, swims like a duck, and quacks like a duck, then it probably is a duck. r eferences Greenberg, Marc L. 2000. A Historical Phonology of the Slovene Language. Heidelberg: Universitätsverlag C. Winter. Hendriks, Pepijn. 2003. A Note on Stang‘s Law in Moscow Accentology. In: Dutch Contributions to the Thirteenth International Congress of Slavists, Ljubljana, August 15-21, 2003: Linguistics. 107–123. DOI: https://doi.org/10.1163/9789004488366_005 Jedvaj, Josip. 1956. Bednjanski govor. Hrvatski dijalektološki zbornik 1: 279–330. 36 Porčinj gen sg pte̥ can be a simple analogy and the infinitive ráste̥ is akin to Čakavian non-phonetic rãsti, as we have shown – the rest of the forms which do not have the expected oxytonesis are also clear analogies. 86 Slovenski jezik – Slovene Linguistic Studies 15 (2023) Ježovnik, Janoš. 2020. A further look at the progressive acute-accent shift in the Tersko dialect of Slovene. In: L. Repanšek, H. Bichlmeier, V. Sadovski. Vácām ̐ si miśrkr̥ṇavāmahai. Proceedings of the international conference of the Society for Indo-European Studies and IWoBA XII, Ljubljana 4-7 June 2019, celebrating one hundred years of Indo-European comparative linguistics at the University of Ljubljana. Hamburg: Baar-Verlag. 665–677. Ježovnik, Janoš. 2022. Glasovne in naglasne značilnosti terskega narečja. Ljubljana: Založba ZRC. DOI: https://doi.org/10.3986/9789610506430 Jurišić, Blaž. 1973. Rječnik govora otoka Vrgade, uspoređen s nekim čakavskim i zapadnoštokavskim govorima. II dio, Rječnik. Zagreb: JAZU. Kalsbeek, Janneke. 1998. The Čakavian Dialect of Orbanići near Žminj in Istria. Amsterdam–Atlanta: Rodopi. Kapović, Mate. 2009. Rising mobility in Slavic i-stems. In: R. Lühr, S. Ziegler (eds.). Protolanguage and Prehistory. Akten der XII. Fachtagung der Indogermanischen Gesellschaft, vom 11. bis 15. Oktober 2004 in Krakau. 236–243. Kapović, Mate. 2015. Povijest hrvatske akcentuacije. Fonetika. Zagreb: Matica hrvatska. Kapović, Mate. 2017a. The Position of Kajkavian in the South Slavic Dialect Continuum in Light of Old Accentual Isoglosses. Zeitschrift für Slawistik 62/4: 606–620. DOI: https://doi.org/10.1515/slaw-2017-0038 Kapović, Mate. 2017b. On shortening, lengthening, and accent shifts in Slavic. Rasprave Instituta za hrvatski jezik i jezikoslovlje, 43/2. 381–402. Kapović, Mate. 2018. Povijest glagolske akcentuacije u štokavskom (i šire). Rasprave Instituta za hrvatski jezik i jezikoslovlje 44/1: 159–285. Kapović, Mate. 2020a. Stang’s Law. In: M. L. Greenberg (ed.). Encyclopedia of Slavic Languages and Linguistics. Leiden–Boston: Brill. DOI: http://dx.doi. org/10.1163/2589-6229_ESLO_COM_036054 Kapović, Mate. 2020b. Notes on the Phonetics, Phonology and Prosody of the Čakavian Dialect of Susak. Zeitschrift für Slawistik 65/4: 498–534. DOI: https://doi.org/10.1515/ slaw-2020-0024 Klarić, Alvijana. 2022. Fonološke i morfološke značajke govora buzetskoga dijalekta. PhD dissertation. Zadar: Sveučilište u Zadru. Šekli, Matej. 2006. Tersko narečje v kraju Subid. In: M. Kožuh (ed.). Terska dolina/ Alta Val Torre/Val de Tor. Terska dolina v besedi, sliki in pesmi Viljema Černa. Celje–Gorica: Celjska Mohorjeva družba, Goriška Mohorjeva družba. 161–173. Šekli, Matej. 2018. Tipologija lingvogenez slovanskih jezikov. Ljubljana: Založba ZRC. DOI: https://doi.org/10.3986/9789610501374 Zuljan Kumar, Danila. 2018. Tersko narečje/Il dialetto del Torre. In: B. Balloch, R. Dapit, L. Trusgnach, D. Zuljan Kumar (eds). Lučice na oknah. Naš sviet pouan naposebnosti/ Il nostro mondo pieno di meraviglia. Čedad/Cividale del Friuli: Kulturno društvo Ivan Trinko/ Circolo culturale Ivan Trinko. 38–49. Mate Kapović, On archaic oxytonesis in Slovene Ter dialect 87 Prispelo oktobra 2022, sprejeto februarja 2023. Received October 2022, accepted February 2023. Prva različica prispevka je bila predstavljena 22. 9. 2022 na 4. Slovenskem dialektološkem posvetu v Ljubljani. s uMMary o n archaic oxytonesis in s lovene ter dialect The paper discusses the oxytonic mokȁ ‘flour’, bradȁ ‘beard’ type accent in the central Prosnid–Porčinj–Subid belt of the Ter dialect based on the material from Janoš Ježovnik‘s 2022 monograph. This type of accent is often interpreted in Slovenian accentology and dialectology as being innovative because the retraction of the short accent from a final open syllable to a preceding length is usually regarded as a Common Slovene innovation. This type of accent is found in the following types of forms: words like mokȁ, words like gen sg kjučȁ, words like vinȍ, adjectives like mladȁ and participles like treslȁ (the oldest speakers still seem to preserved inconsistent length in these forms). The progressive shift of the *jùžina ‘lunch’ > Porčinj južȅ̥na type is clearly a distinct process and the infinitival opposition of *pećȉ but *rásti (with the innovative accent in the second long- root type) is to be connected with a typical Kajkavian-Čakavian-Western Štokavian non-phonetic ‘retraction’ of accent in certain forms (such as the infinitive). The author concludes that central (Prosnid–Porčinj–Subid) Ter dialects (and some other Slovene dialects like Kobarid) preserve the old Slavic final accent in open syllables even after the originally long root. o arhaični oksitonezi v slovenskem terskem narečju Članek obravnava oksitonični naglas tipa mokȁ ‘moka’, bradȁ ‘brada’ v centralnem pasu terskega narečja, ki temelji na gradivu iz monografije Janoša Ježovnika (2022). Ta vrsta naglasa se v slovenski akcentologiji in dialektologiji pogosto tolmači kot inovacija, saj se umik kratkega naglasa iz končnega odprtega zloga na prejšnjo dolžino običajno šteje za splošnoslovensko inovacijo. Ta vrsta naglasa se pojavlja v naslednjih oblikih: besede kot mokȁ, besede kot gen sg kjučȁ, besede kot vinȍ, pridevniki kot mladȁ, in participi kot treslȁ (se zdi da najstarejši govorci še ohranjajo neskladno dolžino v teh oblikah). Progresivni premik tipa *jùžina ‘kosilo’ > Porčinj južȅ̥na je jasno ločen proces, in nasprotje infinitiva *pećȉ, vendar *rásti (s inovativnim naglasom v drugem obliku z dolgim korenom) je povezano s tipičnim kajkavsko-čakavsko-zahodnoštokavskim ne-fonetičnim »umikom« naglasa v določenih oblikah (kot je infinitiv). Avtor sklene, da centralni del (Prosnid–Porčinj–Subid) terskega narečja (in nekatera druga slovenska narečja, kot kobariško) še vedno ohranjajo stari slovanski končni naglas v odprtih zlogih tudi po prvotno dolgem korenu.