CC BY
0
UDC 811.12, 811.14
Philologia Classica. 2023. Vol. 18. Fasc. 2
Once Again about Archaisms in Modern Greek Dialects*
Maxim L. Kisilier
St. Petersburg State University,
7-9, Universitetskaya nab., St. Petersburg, 199034, Russian Federation; Institute for Linguistic Studies of the Russian Academy of Sciences, 9, Tuckov per., St. Petersburg, 199053, Russian Federation; m.kisilier@spbu.ru
For citation: Kisilier M. L. Once Again about Archaisms in Modern Greek dialects. Philologia Classica 2023, 18 (2), 339-359. https://doi.org/10.21638/spbu20.2023.214
Modern Greek identity is heavily based on the idea of the continuity of Greek culture and the Greek language. Most specialists in Modern Greek regard Ancient Greek and Modern Greek as different stages of the same language despite multiple differences and innovations at all levels. During the 19th century, a number of European classical philologists tried to find Ancient Greek features in Modern Greek dialects. As a result, they have singled out Tsako-nian as the sole dialect which descends directly from Ancient Doric Laconian but not from Hellenistic Koine as the rest of the modern dialects. Nevertheless, it is important to point out that Tsakonian is not the only Modern Greek variety with some unique peculiarities inherited from Ancient Greek. This contribution analyzes the phenomena of the Ancient Greek origin in vocabulary, phonetics, morphology and syntax in Modern Greek dialects. The research is focused on those archaisms which exist in the dialects but are absent from Standard Modern Greek. The data was mostly collected by the author of this paper and his colleagues between 2000 and 2023. The analysis demonstrates that the majority of unique peculiarities of the Ancient Greek origin are found in Pontic and Tsakonian, although most varieties of Modern Greek have some archaisms. However, the quantity of archaisms is not a consistent indicator of the antiquity of the dialect since the history of Modern Greek dialects is still terra incognita and there is no good explanation why some dialects keep their archaisms better than the others.
Keywords: Modern Greek dialects, lexical archaisms, phonetic archaisms, morphological archaisms, syntactic archaisms.
1. Intoduction
1.1. Preliminary remarks
The relations between Ancient Greek (= AG) and Modern Greek (= MG) are one of the most important subjects for MG studies, since Greek identity (at least on the official
level) fully exploits the idea of cultural and linguistic continuity. There is no unanimous
* This research was supported by the Russian Science Foundation Grant no. 19-18-00244 "Balkan bilingualism in dominant and equilibrium contact situation in diatopy, diachrony and diastraty".
I would like to use this opportunity and to thank my friends and colleagues for their help: Angela Ralli (Lesbian), Nikos Mavrelos (Northern Chiotic), Alexandra Vyatkina and Panos Marneris (Tsakonian), Archontoula Konstantinidou and Karina Avdzhan (Pontic), Andreas Kameris (Cypriot), and Aristotle Spiro (Greek dialect of Dropull in Albania). All studies were conducted according to the guidelines of the Declaration of Helsinki and I obtained consent from all subjects involved in the research.
© St. Petersburg State University, 2023
decision among the scholars whether AG and MG are one language (Browning 1983, vii),1 or two different languages (Hamp 2003, 67), but strong connections between them on all linguistic levels, especially in vocabulary, are hardly ever argued. However, one should also take into consideration that most Greek words have changed either their pronunciation or some aspects of their morphological form with the course of time, and thus they are not carry-overs, i. e. words which have remained "more or less intact over the years" (Joseph 2009, 369). Moreover, it is well known that AG texts are incomprehensible to the MG speakers without any special education. Several years ago, Alexey Kassian in the frame of the project "The Tower of Babel" created annotated Swadesh wordlists (110 items) for the so-called "Greek group" (Kassian 2018) in which he included Ancient Ionian Greek of Herodotus, Ancient Attic Greek of Plato, Standard Modern Greek (= SMG),2 and three highly divergent MG dialects (Southern Tsakonian, Pharasiot and Cappadocian).3 According to Kassian's data, only 19 AG words completely coincide with their Modern Greek counterparts if they are read with the modern pronunciation:4
(1) fÁoióc; 'bark', ai^a 'blood, spxo^ai 'come', ntvw 'drink', aKoúw 'hear', ¿y« 'I', noÁúc; 'many', Kp¿a; 'meat, OTÓ^a 'mouth', avGpwno; 'man', pi.(a 'root', Ssp^a 'skin', Kanvó; 'smoke', ^Xio; 'sun', ¿Ksívo; 'that, Súo 'two', tí 'what', Xsntó; 'thin, avs^o; 'wind'.
Many other MG words from the Swadesh wordlist also are of the AG origin, some of them have radically changed from antiquity:
(2) AG ovu^ 'nail' > SMG vúxi AG oú; 'ear' > SMG aqm
AG o^po; 'rain' > SMG Ppox^, etc.
A number of basic AG lexemes (some of them are outside of the Swadesh wordlist) disappeared (at least in vernacular) during the Byzantine period:
(3) AG apTo; or ovroc; 'bread' vs SMG ^w^t (< AG ^w^óc; 'morsel')
AG IxQú; 'fish' vs SMG ^ápi (< HellK ó^ápiov < AG o^ov 'cooked food') AG KÚwv 'dog' vs SMG OKuM (< AG OKÚXa^ 'young dog, puppy') AG oIko; 'house' vs SMG ontTi (< LAT hospitium)
AG oivo; 'wine' vs SMG Kpaot (< AG Kpaoi; [o'ivou]5 (Andriotis 1967, 172) 'mixing, blending [of wine]')
AG nostw (noiw) 'do' vs SMG Kávw (< AG Ká^vw 'work, labour')
AG üSwp 'water' vs SMG vspó (< AG vsapóv [üSwp] (Andriotis 1967, 227) 'youthful,
i. e. fresh [water]'), etc.
1 Other renown historical grammars of MG (Horrocks 1997; Adrados 1999; Tonnet 2003), actually, agree with Robert Browning as they start from AG.
2 Recorded in Moscow from three SMG speakers with PhD.
3 I intentionally make use of parenthesis here, because Kassian's Greek group, unlike Slavic group, Germanic group or Romance group consists of varieties which either belong to various chronological periods or are not generally acknowledged as separate languages.
4 It also very important to keep in mind that some words may have retained their form but they have undergone a semantic shift, like 5ouA.su« — AG 'to serve, to be a slave' and SMG 'to work'; for more examples see Kavcic et al. 2020, 130.
5 Av5piwtr|c; N. n. 'etu^oAoyiko As^iko Trjc; Koivr|c NeoeAAr|viKr|;c. 2 ek5. 0saaaAoviKr|, Ksvrpov 'EAAr|viKrjc; rAwaaac;, 'Ivotitoutov NeoeAAr|viKa>v Znou5a>v [I5pu|ia MavoAr| TpiavTacpuAAi5r|], 1967.
As I have argued elsewhere,6 MG speakers are generally familiar with these "lost" words, and their knowledge results not only from the linguistic continuum7 but also from religious8 and cultural tradition,9 school education and katharévousa, a puristic high-style variant of MG which to a large extent had AG as a paradigm and was the official language of the Greek state until 1976.10
1.2. Setting the problem
If a researcher wants to get a more "natural" image of linguistic continuity in Greek free of political and ideological implications, he should also take into account the data from the local varieties. MG dialects11 first attracted serious scholarly attention in the first half of the 19th century when Hellenists from many countries rushed to Greece in order to find AG features in the real living language. Famous professor of classics and Modern Greek at Harvard University Evangelinos Apostolides (better known as Evangelinos Sophocles) mockingly described it in his report (Sophocles 1860-1862, 299): "Now whenever a classical scholar goes to Greece to find Dorians and Ionians, it is ten to one but that he succeeds in finding Dorians and Ionians. He selects such words that are agreeable to his hypothesis, and takes no notice of those which contradict it. He lays much stress upon coincidences and disregards differences". The words of Sophocles were true both for European scholars and for local enthusiasts. One of the most vivid examples is Tsakonian (see section 2) which became a bottomless source for Dorisms. Some researchers in their desire to emphasize the connections between Tsakonian and AG even claimed that this dialect unlike all other Greek varieties (including SMG), retained dative. In fact, they described as dative the combinations of a preposition with accusative (Kisilier 2021, 227, fn. 40).
Today, it is generally believed that all MG dialects are not descendants of the AG dialects12 (cf. Andriotis 1981, 6), and only Tsakonian is a modification of Ancient Laconian (e. g. Liosis 2014, 446).13 However, nearly all dialects demonstrate peculiarities of the AG
6 Kisilier 2022 (Кисилиер M. Л. Еще раз о древнегреческом и новогреческом или помнят ли древнегреческий сами греки? // M. Л. Кисилиер (ред.) Verus convictor, verus academicus. К 70-летию Николая Николаевича Казанского. Санкт-Петербург, ИЛИ РАН, 2022, 351-370).
7 Some AG lexemes have become constituents of well-known compounds: ибраушуею 'water-pipe', артояо1£ю 'bakery, etc.
8 Cf. a line from the Lord's Prayer: TÓv aptov TÓv snioúmov 5óc; ^|iív ar||i£pov-give us this day our daily bread'.
9 See, for example, the frequently used expressions like épu9pó; oivo; 'red wine' (instead of kókkivo Kpaaí) and Леик0; Oíko; 'the White House' (instead of Ааяро Бят).
10 The term Katharévousa (кабереиоиаа 'purified [language/dialect]') was proposed by Adamantios Korais (1748-1833), famous intellectual of his time, as a compromise between arkhaízousa, supported by Phanariotes and aimed at revival of AG, on the one hand, and attempts to create a "new" language based on vernacular and dialects, on the other; for more details about Korais and his linguistic activities see Mack-ridge 2009, 102-125.
11 In this contribution, I use the term 'MG dialect' for all existing varieties of MG regardless of their origin. Thus, Tsakonian will be named here "a MG dialect", although most scholars believe that it is a direct successor of Ancient Doric, and according to Brian Joseph (2022, 66), Tsakonian could be rather treated as a Hellenic dialect, than a MG one.
12 Av5piwtr|t; N. П.'H yévsari rwv vsosAAriviKwv бюЛектши ФйоХоушх: nspio5iK^ ёк5оаг| auWóyou ánocpoÍTwv Ф1\оаоф1к^; Exo^fj; ^ven^'T^iiou 'I^w^^v 1981, 5, 5-22.
13 Tsakonian is currently spoken in the municipality of South Kynouria (Arcadia, Peloponnese) in the region which is called Tsakonia. The population of this region (Tsakonians or Tsakones) believes that their ancestors were Spartans and their dialect has developed from Ancient Laconian. This belief is the basis for
origin. Some of these phenomena are the same as in SMG like kappa aorist forms (from AG perfect) which both exist in the literary standard and in the dialects14 (cf. Konto-sopoulos 2001, XXXIII):15
(4) SMG (p|Ka '[I] found' < AG sup^Ka
GREC ekatevika '[I] went down' (Rohlfs 1977, 107) < AG KaTa(3s(3r|Ka TSAK oraka '[I] saw' < AG ¿opaKa PONT epika '[I] did' < AG nsnot^Ka, etc.
This paper observes the features of the AG origin which exist in MG dialects but are absent in SMG. Various aspects of this problem were discussed previously (Tsopanakis 1955; Tzitzilis 2013), but this research has at least two important differences:
a) it strives to describe the current situation (cf. section 1.3) but not to create some hypothetical or ideal image;
b) since AG "traces" in MG dialects are not restricted to some specific linguistic level, this contribution gives a general overview of all of them: vocabulary (Section 2), phonetic peculiarities (Section 3), morphological phenomena (Section 4) and syntactic archaisms (Section 5).
1.3. Data
The researchers of MG dialects have no dialectal corpora at their disposal. There are few atlases of separate dialects like Cretan16 (Kontosopoulos 1988) and Lesbian,17 but only the DiCadLand18 was helpful for the task of this paper, and some Cappado-cian examples were taken from this atlas. Most MG dialects have already been described, and some of them even more than once. However, wherever possible, I tried to use the data which were collected by my colleagues and myself in various parts of the Greek-speaking world from 2000 to 2023. The examples from all dialects, but Grecanico, which has a special orthographic tradition based on Italian, are given in phonemic transcription.
2. Vocabulary
The archaisms in vocabulary seem to be more vivid and easier to find than ancient relicts in morphology or syntax. It is not a secret that nearly all MG dialects have some words which most likely descend from AG and have not radically changed their "original" form:
local identity and is unanimously accepted outside the community. Tsakonian, indeed, seems very different from all other MG dialects and it demonstrates multiple phenomena of Ancient Greek origin at all linguistic levels: in phonetics, vocabulary, morphology and even syntax.
14 KovToaonouAo; N. r. AiaAsKToi Kai i5iw|iaTa Tr|c; vsa; sAAr|viKr|c;. A9|va, EK5oasi; Tpr|Y6pr|, 2001.
15 These aorists may not coincide in different varieties (as TSAK oraka vs SMG si5a) but more important here is that such forms exist and they do not result from the influence of SMG.
16 KovToaonouAo; N. r. rA«aaiKo; aTAac; Tr|c; Kpr|Tr|c;. HpaKAsiov, navs-maTr||iiaK£c; EK5oasi; Kp|Tn;, 1988.
17 URL: https://lesvos.lmgd.philology.upatras.gr/el (accessed: 10.02.2023) and Cappadocian (DiCadLand): URL: http://cappadocian.upatras.gr/atlas/ (accessed: 16.02.2023).
18 URL: http://cappadocian.upatras.gr/atlas/el (accessed: 16.02.2023).
(5) CYPR19 kufi 'snake' < AG opt;20 vs SMG 9161 DROP tekno 'child' < AG tskvov vs SMG nai6i
LESB isus 'even, straight' < AG iaoc; /isos/ vs SMG iaioc; /isjos/ PONT erexkome '[I] like' < AG opsyo^ai 'desire' vs SMG ^ou apsasi '[I] like' RUMftir[a] 'louse' < AG p0dp /fths:r/ vs SMG ^dpa /psira/ TSAK ujOi 'snake' < AG 091; /ofis/ vs SMG cpiSi /fiSi/
However, only Tsakonian vocabulary was systematically described several times from the point of view of glottochronology.21 Unlike a glottochronological study, the main goal of this subsection is not to demonstrate how divergent or non-divergent the dialects are. Some dialects may have lexical archaisms with different roots to express the same meaning:
(6) TSAK niu 'hear' < AG vosw 'apprehend' DROP kuo 'hear' < AG &kouw
Despite a certain semantic shift in Tsakonian and phonetic changes in the both dialects in (6), there is no doubt that niu and kuo descend from AG and still are verbs. So, it is possible to hypothesize that these lexemes belong to the most archaic part of the vocabularies of Tsakonian and of the dialect of Dropull. Another important parameter is the "uniqueness" of the word, i. e. whether it exists in SMG or not, cf.
(7) TSAK kali 'wood' < AG KaXov
DROP ksilo 'wood' < AG ^uAov, cf. SMG ^uAo
From this point of view, Tsakonian word in (7) is unique, while the word from the dialect of Dropull is not.
Since it is not possible to study the whole dialectal vocabulary, the analysis is focused on 194 words:
(8) 1. white, 2. near, 3. big, 4. all, 5. wind, 6. see, 7. water, 8. hair, 9. louse, 10. eye, 11. say, 12. year, 13. head, 14. mountain, 15. breast, 16. give, 17. far, 18. two, 19. tree, 20. long, 21. rain (noun), 22. road, 23. smoke (noun), 24. eat, 25. yellow, 26. woman, 27. burn, 28. belly, 29. fat, 30. star, 31. green, 32. earth, 33. snake, 34. know, 35. ashes, 36. tooth, 37. go, 38. name, 39. stone, 40. skin, 41. knee, 42. bark (of the tree), 43. root, 44. short, 45. bone, 46. red, 47. blood, 48. round, 49. wing, 50. who, 51. bite, 52. lie, 53. fly (verb), 54. leaf, 55. moon, 56. small, 57. many, 58. man, 59. we, 60. meat, 61. not, 62. new, 63. foot, 64. nail, 65. nose, 66. night, 67. cloud, 68. fire, 69. one, 70. sand, 71. liver, 72. drink, 73. swim, 74. full, 75. come, 76. bird, 77. horn, 78. mouth, 79. hand, 80. fish (noun), 81. seed, 82. heart, 83. sit, 84. hear, 85. dog, 86. sun, 87. salt, 88. sleep, 89. stand, 90. dry, 91. warm, 92. thin, 93. that, 94. thou, 95. heavy, 96. kill, 97. die, 98. ear, 99. tail, 100. cold, 101. good, 102. person, 103. worm, 104. black, 105. what, 106. neck, 107. this, 108. I, 109. tongue, 110. egg, 111. run, 112. beat, hit, 113. fear, be afraid of, 114. scold,
19 See the list of abbreviations at the end of the paper.
20 Some native speakers of Cypriot erroneously connect kufi with kufos 'deaf'.
21 Vaclav Blazek (2010, 18-19) believed that 78.2 % of words from the Swadesh list were the same in Tsakonian and SMG while 63.4 % were common cognates between Tsakonian and the Attic. According to Blazek, this percentage indicates 2450 years of independent development. A very detailed study by Nick Nicholas (2019, 60-65) points out multiple inconsistences of the Blazek's study and in general supports Alexei Kassian's (2018, 11) conclusion: "Because of its mixed nature, the Tsakonian wordlist can hardly be used for lexicostatistical phylogeny of Ancient Greek dialects, not to mention for calibration of glottochronological formulae".
115. breathe in, 116. rope, 117. entrails, 118. inside, 119. turn, twist, 120. you (polite or plural), 121. where, 122. swallow, 123. rotten (adjective), 124. dirty, 125. share, divide, 126. day, 127. hold, 128. others, 129. think, 130. blow, 131. if, 132. press, squeeze, 133. animal, 134. live, 135. slaughter, 136. here, 137. and, 138. play, 139. measure (verb), 140. vomit, 141. as, 142. throw, 143. when, 144. ice, 145. forest, 146. mother, 147. sweep, 148. wet, 149. sea, 150. inflate, 151. on/to the left, 152. on/to the right, 153. sky, 154. some (plural), 155. few, 156. smell (verb), 157. lake, 158. sharp, 159. father, 160. hunt, 161. fall, 162. stick, 163. sing, 164. spit, 165. bad, 166. why, because, 167. right, correct, 168. dust, 169. five, 170. baby, kid, child, 171. cut, 172. river, 173. straight, 174. dig, 175. with, 176. bind, tie, 177. laugh, 178. snow, 179. wash, 180. there (not here), 181. rub, 182. push (verb), 183. father, 184. grass, 185. three, 186. mist, fog, 187. pull, 188. narrow, 189. fruit, 190. flower, 191. scratch, 192. four, 193. wide, 194. sew
All these words belong to the basic vocabulary, that is why one can expect to find them in all dialects and to suppose that they could remain more "intact" and "original" than other words. For example, in Tsakonian which is considered the most archaic of MG dialects, AG lexemes are concentrated in the basic vocabulary while most terms of agriculture, cattle-breading and seamanship (all three were very important domains of the Tsakonian everyday life) are borrowed either from other Balkan languages or from Venetian (more details and examples in Kisilier 2021, 247-250). However, it does not mean that the core Tsakonian vocabulary is free of loanwords (9a) and innovations (9b).
(9a) TSAK tsea/dzea 'home, house' < LAT cella 'room'22
(9b) TSAK kabzi 'child, kid' < Kapniov 'little fruit'23 (Deffner 1923, 162)24
The wordlist in (8) was collected from the speakers of seven MG dialects which may be regarded as enclave dialects and thus are most likely to preserve multiple archaisms25 (cf. Zhirmunsky 1976, 492):
(a) Cypriot, Lesbian and Northern Chiotic are spoken within the island communities;
(b) Azov Greek, Greek dialect of Dropull and Pontic are outside of Greece;
(c) although Tsakonian always was inside Greece,26 it is so different from all adjacent local variants of Greek that it can be also treated as an enclave dialect.
While comparing the wordlists, I took into account two parameters: a number of words of the AG origin (= SimAG) regardless of any semantical shifts (cf. ex. 6) and a number of words of the AG origin which are absent from SMG (= NSimSMG) and meet
22 SMG аят 'home, house' is also a borrowing from Latin — hospitium 'shelter'. Another very similar example is TSAK koleya 'friend' < LAT collega 'partner, associate, fellow'. In Tsakonian the loanwords of the Latin origin could have come through Venetian and not directly from Latin.
23 Аёфуер M. Ле^коу rfjc; raaKwviKrjc; SiaXsKTou. 'Ev A9r|vaic;, Tunoypacprfov "Eatia"; K. Maiavsp Kal N. Kapya5oupr|, 1923.
24 A different etymology (< LAC яартшс;) is suggested by Peter Jernstedt (1922, 85-89), but it seems more doubtful.
25 Жирмунский В. М. Избранные труды. Общее и германское языкознание. Ленинград, Наука,
1976.
26 Propontis Tsakonian from Asia Minor is not examined here.
the parameter "uniqueness" as TSAK kali in (7). The results of the comparison may be seen in the following table:
Table 1. AG words in MG dialects
MG dialect Total27 SimAG SimAG % NSimSMG NSimSMG %
CYPR 194 140 72.2 9 4.6
DROP 194 135 69.6 3 1.5
LESB 190 140 73.7 6 3.2
NCHIOT 194 141 72.7 1 0.5
PONT 193 143 74.1 30 15.5
RUM 189 138 73 12 6.3
TSAK 194 156 80.4 29 14.9
According to Table 1, SimAG is the highest in Tsakonian, while in Cypriot, Lesbian, Northern Chiotic and Pontic, it is almost the same with a slight fall in Azov Greek and in the Greek dialect of Dropull. It means that the quantity of the AG words in a dialect itself cannot be seriously regarded as an AG feature. These words could easily come from the school education which was rather widespread in Byzantium (Markopoulos 2008, 787), or be affected by the continuous and constantly increasing influence of SMG. The parameter NSimSMG seems to be more reliable. It demonstrates that Pontic and Tsakonian have much more unique archaic words than any other dialect, and Northern Chiotic is almost free of them. Unexpectedly, Azov Greek which has less words of the AG origin than all other varieties, except Greek dialect of Dropull, demonstrates a relatively high NSimSMG.
Almost all lexical archaisms follow the morphological patterns that are relevant for the current state of the dialect, but not for AG. Thus, Tsakonian feminine nouns generally get the flexion -a even if in AG they belonged to the 3rd declension, and evidently it is not the influence of SMG:
(10) TSAK a xera 'hand' <f28)>vs AG ^ x£ip <f> and SMG to xspi <n> 3. Phonetics
Phonetic peculiarities of the AG origin can be divided into two groups:29 general
(3.1) which are not connected with any specific AG dialect or dialectal group, and Doric
(3.2). Some MG dialects (e. g., Tsakonian) may have features of the both types.
3.1. General
Phonetic phenomena which descend from AG in general are well-known to specialists in MG dialectology and are frequently used as an important parameter for dialectal classifications (cf. Trudgill 2003, 54, 57).
27 Total number of the examined words.
28 All morphological abbreviations are given according to the Leipzig glossing rules (Comrie et al. 2015) and CorpAfroAs Team. 2017.
29 Christos Tzitzilis (2013) also finds some Arcado-Cypriot and Aeolic features, but they are so scarce that it is difficult to regard them as a separate group.
3.1.1. /y/ > /u/
It is known that in Attic-Ionian, there was a shift from /u/ (u) to /y/ (u) which (due to the expansion of Hellenistic Koine) affected all local Greek varieties. In the 9th-11th centuries, in the most cases /y/ became /i/ (Holton et al. 2019, 12-13):
(11) AG ovu^ /onyks/ 'nail' > SMG vuxi /nixi/ 'nail', CYPR nifi, DROP nixi, LESB nix, NCHIOT nixi, PONT nif, RUM nif, TSAK nixi30
Some examples in certain local varieties demonstrate the regression /y/ > /u/ instead of the expected /y/ > /i/:31
(12) AG |uta /myja/ 'fly' > CYPR muja vs SMG |uya /miya/
AG Kupio; /kyrios/ 'having power, authority' > GREK ciuri 'father' (Violi 2007, 152; see also32 Liosis, Papadamou 2011, 205-214) vs SMG /kirios/ 'sir'
AG axupov /axyron/ 'straw' > MAN axuro (Mirambel 1929, 75) vs SMG ¿xupo /axiro/ AG yuv| /Yyne:/ 'woman' > MEGAR yuneka (Newton 1972, 21), TSAK yuneka vs SMG yuvaixa /jineka/, etc.
Peter Trudgill (2003, 54) supposed that /y/ > /u/ was limited to Attica and adjacent regions, and some parts of Peloponnese, and this area has fallen apart after the penetration of Arvanitika. However, the examples from Cyprus, South Italy (12), as well as from many other regions like Rhodos, Karpathos, Crete, etc. (published in Tsopanakis 1955, 57-58) demonstrate that it is not just a local phenomenon.
3.1.2. "Ionic" /e/
One of the important differences between AG and SMG vocalism is the transformation of /e:/ into /i/ which happened during the period of the so-called Late Koine:33
(13) AG nAio; /Ye:lios/ 'sun' > SMG ^Xio; /iXos/, CYPR iMos, DROP iMos, GREK iglio (Violi 2007, 195), LESB iMus, NCHIOT ijos, PONT ilen, RUM iMus, TSAK iMe
This transformation consisted at least of two stages: (a) /e:/ > /e/ and (b) /e/ > /i/. Some examples demonstrate only the stage (a). The studies in historical linguistics (cf. Holton et al. 2019, 26-28) indicate that the stage (a) was typical for Asia Minor Greek. For this reason, /e/ derived from AG /e:/ is frequently called "Ionic" /e/, and nowadays it is regarded as one of the peculiarities of Pontic and Cappadocian which actually are Asia Minor dialects:
(14) AG OKwXn^ /sko:le:ks/ 'worm' > PONT skolek vs SMG okouA^ki /skuliki/, DROP skuli-ki, LESB sklikus, NCHIOT skuliki
30 One may argue that in Tsakonian, it is just a borrowing from SMG, however, in TSAK ío 'water' (< AG v5wp /Yydor/) /у/ has also transformed into /i/ although the word u5wp does not exist in SMG.
31 The examples of this regression exist even in SMG: фоиака /fúska/ 'bubble, balloon' < AG срт>акг| / fyske:/.
32 Люаг|с; N., Паяабйцои E. H s^éXi^n тои и атчс; N£osAAr|viKéc; б1а\ёктои^: n KarwiraXiK^ ae аиукрг|аг| це T|v Taaкшvlкr. NeoeXXrviK^ 51а\екто\оу[а 2011, 6, 201-223.
33 In the Early and Middle Byzantine texts, one can easily find many orthographic mistakes where <|> is replaced by <i> or <ei> and vice versa, cf. Tohtasiev 2018, 134-137 [= Тохтасьев С. Р. Язык трактата Константина Багрянородного DE ADMINISTRANDO IMPERIO и его иноязычная лексика. Санкт-Петербург, «Наука», 2018] with multiple examples from De administrando imperio.
AG n^Y^ /pe:ge:/ 'well, source' > CAPP peyadi 'well' cf. SMG n^YaSi /piyaSi/, TSAK kiyajdi
Azov Greek which is sometimes regarded as an Asia Minor dialect (Karatsareas 2014, 79) demonstrates the same phenomenon — skulek or skuletf. However, the so-called "Ionic" /e/ also exists in Tsakonian (TSAK nema 'thread' < AG v^a /ne:ma/ vs SMG v%a / nima/) and other non-Asia Minor Greek dialects and even in SMG:
(15) AG &aT^p /aste:r/ 'star' > SMG aaTepi /asteri/, CYPR asteri, LESB aster
AG ^d; /Ye:ms:s/ 'we' > SMG E^d; /emis/, CYPR emis, DROP emis, NCHIOT emis, TSAK eni
AG naT^p /pate:r/ 'father' > SMG naTepa; /pateras/, CYPR pateras, DROP pateras, LESB pateras, NCHIOT pareras
Although in (15) the influence from SMG cannot be excluded, these examples prove that the "Ionic" /e/ cannot be treated as a purely Ionic (or post-Ionic) phenomenon.
3.1.3. Retention of /-n/
The loss or retention of final /-n/ (especially for nouns) highlights the differences in morphonology between AG and SMG. The author of the first consistent classification of MG dialects, Brian Newton (1972, 99-100; cf. Trudgill 2003, 57) suggests that the final /-n/ should be regarded as a characteristic of South Eastern Greek (Cypriot and Chiotic):
(16) AG to KaXov naiSiov 'beautiful child/young slave'
CYPR to kalom bedin 'good guy' (with assimilation /n/ > /m/ and /p/ > /b/), cf. SMG to KaXo naiSi
A thorough examination of the basic vocabulary of seven Modern Greek dialects (see section 2) shows that final /-n/ is much more regular in Pontic than in Cypriot and Northern Chiotic:34
(17) AG SsvSpov /dendron/ 'fruit- or mast-bearing tree' > PONT dendron 'tree' vs SMG SsvSpo /Sendro/, CYPR dendro, NCHIOT dendro
AG cpuWov /fyllon/ 'leaf' > PONTfilon vs SMG cpuAAo /filo/, CYPRfillo, NCHIOTfilo AG a^oivtov /sxojnion/ 'small rope, thread' > PONT fkinin 'rope' vs SMG a^oivt / sxini/, CYPR fini, NCHIOT skini
AG naiSiov /paidion/ 'child, young slave' > PONT pedin 'child, guy' cf. CYPR pedin vs SMG naiSi /peSi/, NCHIOT pedi
Final /-n/ in many cases could also manifest not some kind of relations with AG but just analogical developments which were frequent in Medieval Greek (cf. Holton et al. 2019, 164-165).
3.2. Doric
About 70 years ago Agapitos Tsopanakis (1955) noticed that several dialects demonstrate phonetic features which could be connected with Ancient Doric. Although some
34 Tsakonian, in spite of being regarded as the "most ancient" dialect, has no examples with final /-n/.
points of Tsopanakis could be argued or reassessed,35 his idea gave a new insight into the relations of AG and modern dialects.
3.2.1. "Doric" /a/
One of the most renown peculiarities of Doric is /a:/ instead of /e:/. The influence of Hellenistic Koine was so strong that deeply affected all regions of the Hellenic world regardless of whether they initially were Ionic or Doric. Nevertheless, some words still have /a/ in the positions where SMG should have expected /i/ (< /e:/, see 3.1.2) and are found at least in two MG dialects which are spoken in the former Doric area:
(18) TSAK amera 'day' < DOR a^spa vs SMG n^spa /imera/ < ION ^¿pa /Ye:mera/ KARP sinomalikos 'coeval'36 (Rylik 2012, 736) < DOR auvo^aAi^ 'fellow, comrade' vs SMG auvo^AiKo; < ION auvo^Ai^
Dikaios Vayakakos37 (1994, 57) also gives an example from Simi alakati 'spinning-wheel' and from Mani alakata (cf. ION qAaxaxn). However, according to LS/,38 the Doric form is qAaxaxa, while aAaraxa is encountered in Aeolic. Multiple examples of "Doric" /a/ from various regions (Rhodos, Creta, Santorini, Kythera, Laconia, Epirus, Corfu, etc.) may be also found in (Tsopanakis 1955, 55-57). Not all of them seem really Doric like RHOD alekati 'spinning-wheel' or trofatizo 'predict' (cf. DOR, BEOT npo^axa; 'prophet' vs ION npo^^x^;39) which may have nothing to do with Doric.
3.2.2. AG /6:/ > /u/
One of the most famous Greek linguists and specialists in dialectology Georgios Hatzidakis40 (1901, 558-561) noticed that in Tsakonian AG /6:/ is regularly replaced by /u/ while /o/ may remain intact:
(19) TSAK yrusa 'tongue, language' vs SMG yAwaaa /ylosa/41 TSAK topo 'place' vs SMG Tono; /topos/
According to Hatzidakis, the examples like (19) prove that Tsakonian, unlike SMG and its dialects, has not entirely lost the opposition of long and short vowels.42 I am not going to discuss here whether Hatzidakis was right or not, but what is more important, the shift /6:/ > /u/ can be traced in some other SMG dialects which are spoken at the former Doric area:
35 On the one hand, Tsopanakis incorrectly believed that /y/ > /u/ was a "Doric" phenomenon (see section 3.1.1), and on the other hand, he omitted several important "Doric" features.
36 Рылик П. А. К вопросу о связи новогреческих говоров острова Карпатос с дорийским диалектом. Индоевропейское языкознание и классическая филология 2012, XVI, 730-741.
37 BayiaKaKo; А. В. Koiva SiaXsKTiKa aroixsia Aw5sKavr|aou Kai Mavri;. NsosXXr|viKr| 5iaXsKToXoyia 1994, 1, 43-63.
38 https://lsj.gr/wiki/r|\aKaTr| (accessed: 24.02.2023).
39 However, the alternation cpr|-/9a- is possible: фщ^ say' <prs.isg> vs фаа[ <PRS.3PL>.
40 XarZiSaKi; Г. N. rXwaaoXoyiKai |isXsTai. A9r|va, 'Ek rou тип. П. A. ZaKsXXapiou, 1901.
41 The examples in 3.2.2 are taken from the report by M. Kisilier and N. Liosis "Ha Tr|v пpoфopa Tou и as vsosXXnviKs; SiaXsKTou;: TaaKwviKa, KarwiraXiKa, Aw5sKavr|aiaKa, MaviariKa" at the 9th International conference on Modern Greek dialects and linguistic theory (MGDLT9), Leonidio, Greece, 4-5 June 2021.
42 In Propontis Tsakonian of Asia Minor, there is /о/ instead of /u/: yosa or ywosa 'tongue, language'.
(20) DOD skulukas 'worm', CRET skulikas, cf. AG aKwAr|^ /ska:le:ks/ MAN xuma 'soil, cf. AG x&^a /xa:ma/ and PONT xoman
3.2.3. Digamma
At least in three MG dialects from "Doric" area, one can find sporadic traces of F /v/ which was present in Western AG dialects (including Doric) and was absent in Attic and Ionic:
(21) TSAK vane 'lamb', cf. DOR Fap|v43 vs SMG apvi
PIER vrozos 'knob' < PIE *wrpdjos 'branch, root', vs SMG po(oc (Tzitzilis 2013) AGR kataxovja 'in gusts' < KaTa + \oFa (= xo|) 'pouring out' (Tzitzilis 2013)
3.2.4. Traces of Laconian in Tsakonian
In several Tsakonian examples, one can find the shift from /th/ to /s/ which was relevant for Laconian (Bourguet 1927, 75-78; Kristoffersen 2019, 172), instead of /th/ > /0/ as it happened in other Greek varieties:
(22) TSAK serindu 'reap, cf. LAC osptSSw vs SMG GspiZw /Gerizo/
Another famous Laconian peculiarity is rhotacism (Kristoffersen 2019, 172-173). In Tsakonian, rhotacism is regular only with articles if a noun has an initial vowel:
(23) TSAK tar ayaki 'love' <sg.gen> vs SMG Tr|i ayannc TSAK tur aQripi 'man' <pl.acc> vs SMG Tou^ av0p«nou;
These peculiarities may exist in other MG dialects as well44 (Prombonas 1963), but at the moment, the most reliable examples are from Tsakonian.
4. Morphology
Inscriptions, literary texts, correspondence and documents make it possible to trace the evolution of Greek morphology from the earliest stages of AG till our days. Certain flexions and forms have remained almost intact while some aspects of the morphological system have either disappeared (dative, dual, optative, etc.), or radically changed (e .g., subjunctive, perfect,45 future). The language evolution had different speed in various parts of the Greek-speaking world, and sometimes MG dialects may provide invaluable data for historical linguistics. All dialects, including Tsakonian, follow in general the morphological patterns of SMG or borrow some patterns from Turkish or Italian as a result of close and prolonged contacts. However, there still exist archaic features.
In this section, I shall not take into consideration such peculiarities as negations in Tsakonian, Pontic and Azov Greek (24a) or Tsakonian forms of the verb 'to be' (see the discussion in Pernot 1910) and personal pronouns (24b), etc.:
43 URL: https://lsj.gr/wiki/&pr|v (accessed: 24.02.2023).
44 npo^nova; 'I. K. 'O poTaKia^o; Kai ^ iSioTuno; npopopa tou p06yYou \ si; to yAwaaiKov i5i«|ia OiXwtou Trj; Na^ou. 'Ev A0rvai;: 'ETaipsia KukXciSikwv MsXsTwv, 1963.
45 Only certain forms of AG perfect exist both in SMG and its dialects, but now they belong to the aorist paradigm (see ex. 4).
(24a) TSAK o < AG oú[k], PONT ki < ION oúkí, RUM tfi < ION oúkí
(24b) TSAK eníu <1.sg.obl> < DOR ¿^¿w (cf. Scutt 1912-1913, 163)
In my opinion, the examples like (24) do not represent the remnants of the AG morphological system, and should be regarded as lexical archaisms.
Some morphological archaisms are variants competing with new (more standard) forms like verb flexions in the dialect of Karpathos (25a) while other are a unique way of expression as Tsakonian medio-passive aorists (25b):
(25a) KARP pat-úsi <step.ipfv-prs.3pl>, cf. AG nat-ouoi vs KARP pat-ún <prs.3pl>, cf. SMG naT-oúv (Dawkins 1903-1904, 100)
(25b) TSAK orá-ma <see-pass.aor.1sg>, cf. AG é-Só-^nv <pst-give-mid.pst.1sg> (Liosis 2014, 449)
4.1. Nominal system
The nominal system of MG dialects is rather innovative. In 4.1 I would like to discuss two archaisms which concern possessive (4.1.1) and personal (4.1.2) pronouns.
4.1.1. Pontic possessive pronouns
AG possessive adjectives (é^ó; 'my' <m>, etc.) used as possessive pronouns seem to have disappeared from the vernacular everywhere except Asia Minor Greek during the first millennium AD (Holton et al. 2019, 914). They were replaced with a personal pronoun in genitive or oblique (if the dialect lacks the distinction between genitive and accusative):
(26) SMG [iou <1SG.GEN>, CYPR mu, DROP mu, NCHIOT mu, RUM m (< mu), TSAK mi <1sg.obl>, etc.
Unlike other dialects, Pontic still has possessive pronouns of the AG type (27a) along with more innovative forms (27b):
(27a) t e món o ándr-as
DEF.N.SG 1SG.POSS DEF.SG.NOM husband-SG.NOM
(27b) ándr-as=im
husband-SG.NOM=POSS.1SG46 'my husband'
Pontic possessive adjective (27a) has no more gender agreement: it is always neuter regardless of the gender of the noun it refers to.
4.1.2. Tsakonian personal pronouns of 1 and 2sg in oblique
AG had no clitic forms for personal pronouns of 1 and 2pl. Clitic pronouns ^a; and aa; can be encountered only after the 10th century AD (Holton et al. 2019, 884-886), and
46 The clitic possessive pronoun is m; -i- is added for euphonia between two consonants, cf. pedi-0=m <child-sg.ngen=poss.1sg> 'my child' where it is not needed as the noun has final vowel.
now all MG dialects have the opposition of strong and weak forms pronouns. However, in Tsakonian, there are some strange examples of clitic doubling:
(28a) m=orak-ate namu
1SG.OBL=see.AOR-AOR.2PL 1PL.OBL
'[you] saw us'
(28b) ethe Qe-nde na=ndi=d-u pumu
be.PRS.2PL want-PTCP.NN.PL SBJV=2SG.OBL=give-1 SG.SBJV 2PL.OBL '[you] want me to give you'
In (28a) strong pronoun namu <1pl> is doubled with a clitic pronoun of 1sg, but not of 1pl. The same thing happens with the strong pronoun pumu <2pl> in (28b). Tsakonian grammars generally distinguish strong and week pronouns for 1 and 2pl (cf., for example, Pernot 1934, 186-187, and see Table 2).
Table 2. Tsakonian pronouns 1 and 2PL in oblique
Strong Weak
1pl namu nam
2pl pumu pum
However, the examples like (28) demonstrate that the real speakers (those who have never read grammars) do not know clitic pronouns for 1 and 2 pl just like speakers of AG.47
4.2. Verb system
4.2.1. Tsakonian archaisms
Expectedly, most archaisms in the verb system can be found in Tsakonian. According to Christos Tzitzilis (2013), Tsakonian analytic forms of present and imperfect (participle + the auxiliary 'to be', see 28b) could be a Laconian heritage, cf. two Laconian glosses in Hesichius: ¿^n^P^p '[s/he] saw' (perfect participle expresses imperfect) and dnsaouxip '[s/he] was saved' with the participle of passive aorist (for another explanation see Kisilier 2021, 239).
Tzitzilis (2013) also demonstrates that some Tsakonian aorists originate from Doric perfects:
(29) TSAK ekreva '[I] stole', cf. MESS kskAsP«;
TSAK eyrava '[I] wrote', cf. ARG YSYpa^avTai
4.2.2. Infinitive
An extremely unique AG feature is encountered in verb morphology of Italian Greek (30a) and Ophitic Pontic (30b). It is the aorist infinitive which has remained in SMG only in periphrastic forms of perfect (cf. Holton et al. 2019, 1682):
47 A different interpretation may be found in (Liosis 2017, 63-64).
(30a) GREC (Rohlfs 1977, 191)
m'=e-kam-e
1SG.OBL=PST-make-PST.3SG '[s/he] made me laugh'
(30b) PONT (Sitaridou 2014, 36): prin maireps-ini
before cook.PFV.INF
'before cooking'
4.2.3. Athematic aorist
As a result of the evolution of the Greek verb system, SMG and most MG dialects have two patterns of the aorist formation (the both of the AG origin): more frequent sig-matic aorist and more rare kappa aorist (4). Some varieties of MG like those of Constantinople and Dodecanese have also examples of athematic aorists which definitely come from AG athematic aorist:
(31) CONST tvya '[I] walked', DOD tvye '[she] walked'48 (Kisilier, Fedchenko 2015, 370), cf. AG sPnv '[I] walked' vs SMG Py^Ka
5. Syntax
Greek syntax has been continuously changing throughout its long history. Although AG word order was regulated by the Wackernagel's law it was not the same at the various stages of AG and evolved from few prosodic restrictions in the Homeric Greek to almost obligatory juxtaposition of the syntactically bound words (Dunn 1989). Thus, pronominal clitics were forced to occupy the position close to the governing verd but not just the second position within the clause like other clitics (especially the particles yap, Se, S^, etc.). As the prosodic evolution went on, the particles disappeared, the Wackernagel's law lost its relevance for Greek, unlike some Balkan Slavic languages (Alexander 1994). As a result, the position of clitic pronouns became grammaticalized and dependent on the form of the verb it referred to (cf. Janse 2000). Most MG dialects have undergone this evolution and regulate the placement of pronominal clitics by means of the same principles as SMG (Condoravdi, Kiparsky 2002, 15-16).
However, this pattern is not unique. In Pontic, pronominal clitics are so-called syntactic X0 clitics (Condoravdi, Kiparsky 2002, 18-22) and always follow the verb regardless of its form and possible syntactic or prosodic constraints:
(32a) eyo elep-o=sen
1sg.nom see=2SG.ACC 'I see you'
(32b) kj=e-Qeln-en na=d(n-e=me esena
NEG=PST-want.IPFV-PST.3SG SBJV=give.IPFV-PST.3SG=1SG.ACC 2SG.ACC '[my mother] did not want to give me to you'
48 Кисилиер М. Л., Федченко В. В. Некоторые замечания об архаизмах в системе современных греческих диалектов. Индоевропейское языкознание и классическая филология 2015, XIX, 368-373.
jel-is-i
Iaugh-PFV-INF
The position of pronominal [en]clitics in Pontic may be explained in terms of Wack-ernagel's law: in (32a) a clitic pronoun occupies the second position, and in (32b) it is in the third position because it was moved by the particle na. This particle could be treated as a syntactic barrier in terminology of Andrey A. Zaliznyak49 (2008, 47-51) which makes VP "retreat" one step backwards and thus moves the clitic pronoun to the third position.
In Byzantine Greek, the Wackernagel's law remained one of the word order regulators but the prosodic system was constantly changing and the appearance of any modal particle in front of VP would "attract" the pronominal clitic to the position in front of the verb (cf. Mackridge 1993) in order to create a prosodic cluster with a particle. Azov Greek (33a), Cap-padocian (33b), Cypriot (33c) and some other MG dialects reflect this medieval pattern: the pronominal clitic is behind a finite verb but precedes it if the VP follows a modal particle.50
(33a) RUM:
e-kleps-iz=mi
pst-steal.pfv-pst.2sg=1sg.obl '[you] have stollen me' Qa=ta=fay-u avir
fut=3pl=eat.pfv-prs.1sg tomorrow '[I] shall eat it tomorrow'
(33b) CAPP (Janse 1998, 260-261): dixn-i=se=to
show.IPFV-PRS.3SG=2SG.GEN=3SG.NGEN '[s/he] shows it to you' as=to=fay-o=m
let=3.NF.SG.NGEN==eat.PFV-PRS.1SG=Q 'will [I] eat him?'
(33c) CYPR: lal-o=su
say-PRS.1SG=2SG.GEN '[I] say to you' na=mu=pi-s
SBJV=1SG.GEN=say.PFV-PRS.2SG 'tell me'
Tsakonian, at first glance, seems more innovative: in all examples, collected from contemporary speakers, a clitic pronoun precedes the finite verb as in SMG:
(34) mj=epetsere
1 sg.obl=say.aor.2sg '[you] told me'
However, when the VP has the archaic negation o, the verb is always behind the clitic pronoun:
(35) o=pets-ere=mi neg=pst-say.aor-aor.2sg=1sg.obl '[you] did not tell me'
49 Зализняк А. А. Древнерусские энклитики. Москва, Языки славянских культур, 2008.
50 In (Condoravdi, Kiparsky 2002, 5-15), this type of clitic pronouns is called Xmax clitics.
It looks as if o "switches on" the same pattern with the Wackernagel's law as in Pontic (32b). This pattern does not work with the modern negation de (< SMG Se[v]):
(36) de=mj=e-pets-ere NEG=1SG.OBL=PST-say.AOR-AOR.2SG '[you] did not tell me'
According to Nick Nicholas (2022), about a century ago, Tsakonian clitic pronouns could follow the verb in an affirmative clause:
(37) e-pets-e=ni tu vafiKi-a PST-say.AOR-AOR.3SG=3SG.OBL DEF.NF.GEN king-SG '[she] told the king' (Scutt 1913-1914, 19)
There are even examples when the clitic pronoun follows the verb after the particle na:
(38) na=afi-the=mi s endeni to fin-a SBjV=leave-IMP.2PL=1 sg.obl in this.M def.m.sg.acc mountain-SG.NGEN 'leave me in these mountains' (Scutt 1913-1914, 27)
However, (37) and (38) are somewhat problematic. In (37), the postposition may be caused by the fact that the clitic pronoun is a doubling pronoun and its position may be affected by the position of the object it refers to; cf. (39) where the object precedes the verb and, as a consequence, the doubling pronoun is also in front of the verb:
(39) ta Majiyula p=e-ksexats-e
DEF.SG.OBL Marigula 3SG.OBL=PST-forget.AOR-AOR.3SG
'[the king] forgot Marigula' (Scutt 1913-1914, 19)
In the same collection of Tsakonian texts one shall easily find numerous examples where the clitic pronoun is placed as in SMG:
(40) tse p=aq-ai
and 3sG.OBL=take.AOR-AOR.3PL '„and [they] took her' (Scutt 1913-1914, 19)
In (38), na=afithe=mi semantically is an imperative and it may be the reason for the postposition. In the previous phrase of the same text, there is a proper subjunctive and the clitic pronoun is placed immediately after the particle:
(41) jatsi na=mi=skotu-the
why SBJV=1SG.OBL=kill-SBJV.2PL
'why should [you] kill me?' (Scutt 1913-1914, 27)
Since in all other cases with na the clitic pronoun precedes the verb, I am inclined to believe that (38) it is just a lapsus linguae.
6. Conclusions
SMG and its dialects are full AG features, and this fact, in my opinion, manifests the continuity of the Greek language. The majority of these features is an important part of the vocabulary, phonetics and grammar of SMG and MG dialects. They cannot be called "archaisms" because they are productive and are in common use today. This contribution
is about "real" archaisms, i. e., unique particularities of the AG origin which were natural for AG but no longer exist in SMG. Amazingly, they can be found almost in all SMG dialects and may refer to any linguistic level. The archaisms are neither productive nor interdialectal but some of them like final /-n/, "Ionic" /e/ or "Doric" /a/ are encountered in more than one dialect. The use of some Doric archaisms is restricted to the "original" Doric-speaking area, so it is very tempting to speak about the Post-Doric dialects. This group could include Tsakonian, Cretan, the dialects of Rhodos, Santorini, Kythera, Laco-nia, Epirus, Corfu, etc. However, we lack enough data to make something more than just hypothetical conclusions.
There also are multiple archaisms of the Ionian origin but the fact of their existence in a dialect may just indicate that the region where the dialect is spoken was highly influenced by the Hellenistic Koiné or had good schools in the Byzantine and the Post-Byzantine periods. It is well-known that many regions had multiple highly educated intellectuals who could read original AG texts and produce their own texts in AG like the verses in "Homeric" style written in the 18th century by a Tsakonian schoolmaster from Prastos51 (Koulouri 1992).
It is evident that various dialects have different archaisms. Most archaisms in vocabulary were in Pontic and in Tsakonian. I dare suppose that if I had some relevant Cappado-cian data at my disposal, Cappadocian would have provided the results comparable with Pontic and Tsakonian. In phonetics, morphology and syntax, Pontic and Tsakonian also demonstrate most specific peculiarities. However, except vocabulary, Tsakonian analytic forms of present and imperfect (if it really is an archaism and not some kind of innovation), or Wackernagel's law in Pontic, all archaisms are extremely sporadic. They are quickly disappearing, and in couple of generations will be no longer in use.
Probably, the most important question is: can the quantity or the type of archaisms tell us anything not about the history of Greek but about the dialect itself? Do we have any real ground to say that, for example, Pontic or Tsakonian are more ancient than the dialects which have less archaic features? It is so tempting to answer "yes" that most scholars and amateurs who write about Tsakonian are eager to describe it as almost an Ancient Doric dialect. But being more archaic does not mean being more ancient. In some aspects modern Lithuanian has no less archaisms than Classical Greek, still they are not equally ancient. Unfortunately, we know almost nothing about the history of MG dialects (including Tsakonian), we cannot explain why some dialects have archaisms while others do not. Why Azov Greek has much less archaisms than Pontic although both of them are enclave dialects, had similar environment, and, probably both belong to Asia Minor Greek? The only thing we know for sure, is that all MG dialects, despite their differences, contacts with other languages, etc., have common Ancient Greek basis which unites them into one language and provides one general identity to speakers of all varieties of Greek, regardless of their location.
Abbreviations
AG — Ancient Greek, AGR — dialect of Agrafa, ARG — Argolic, BEOT — Beo-tian, CAPP — Cappadocian, CONST — dialect of Constantinople, CRET — Cretan,
51 KouXoúpn X. «Ztíxoi ^pwiKOÍ» roü 'E|i|iavou^\ Tpoxávr| SaaráAou ánó TÓv npáato (1775), as: npaKtiKa ApKaSiKoü nv£U|ariKoü au|nóaiou 1992 (AswviSíwv 21-23 Nos|ppíou 1992). A9r|vai, 'Eraipsía nsXoHovvriaiaKwv EnouSwv, 1994, 215-241.
CYPR — Cypriot, DOD — dialect(s) of Dodecanese, DOR — Doric, DROP — Greek dialect of Dropull (Albania), GREC — Grecánico, HellK — Hellenistic Koiné, ION — Ionic, KARP — Greek dialect of Karpathos, LAC — Laconian, LAT — Latin, LESB — Lesbian, MAN — Maniot, MEGAR — Megarian, MESS — Messenian, MG — Modern Greek, NCHIOT — Northern Chiotic, NSimSMG — number of words of Ancient Greek origin absent from Standard Modern Greek, PIE — Pre-Indo-European, PIER — dialect of the Pierian mountains (Macedonia), PONT — Pontic, RHOD — dialect of Rhodos, RUM — Azov Greek, SimAG — number of words similar with Ancient Greek, SMG — Standard Modern Greek, TSAK — Tsakonian.
References
Adrados F. R. Historia de la lengua griega: de los orígenes a nuestros días. Madrid, Gredos, 1999.
Alexander R. The Balkanization of Wackernagel's Law. Indiana Slavic Studies 1994, 7, 1-8.
Andriotis N. P. Etymological Dictionary of Modern Greek. 2nd ed. Thessaloniki: Institute of Modern Greek
Studies [Manolis Triandafyllidis Foundation], 1967 (in Modern Greek). Andriotis N. P. The Origin of Modern Greek Dialects. Philologica: The Periodical of the Association of the Graduates of the Faculty of Philosophy of the University of Ioannina 1981, 5, 5-22 (in Modern Greek).
Blazek V. Glottochronological Analysis of the Greek lexicon: Modern, Tsakonian, Old and Mycenaean
Greek. Graeco-Latina Brunensia 2010, 15 (1), 17-35. Bourguet É. Le dialecte laconien. Paris: Librairie ancienne honoré champion [Libraire de la Société de linguistique de Paris], 1927. (Collection linguistique publiée par la Société de linguistique de Paris XXIII). Browning R. Medieval and Modern Greek. 2nd ed. Cambridge, Camridge University Press, 1983. Comrie B., Haspelmath M., Bickel B. The Leipzig Glossing Rules: Conventions for Interlinear Morpheme-by-Morpheme Glosses. 2015. URL: http://www.eva.mpg.de/lingua/resources/glossing-rules.php (accessed: 15.02.2023).
Condoravdi C., Kiparsky P. Clitics and Clause Structure. Journal of Greek Linguistics 2002, 2, 1-39. CorpAfroAs Team. The CorpAfroAs and CorTypo List of Glosses. 2017. URL: https://bit.ly/Glosses-
CorpAfroAs (accessed: 16.02.2023). Dawkins R. M. Notes from Karpathos. The Annual of the British School at Athens 1903-1904, 10, 83-102. Deffner M. Dictionary of the Tsakonian Dialect. Athens, Estia, K. Meisner & N. Kargadouri, 1923 (in Modern Greek).
Dunn G. Enlitic Pronoun Movement and the Ancient Greek Sentence Accent. Glotta 1989, 67, 1-19. Hamp E. P. Prehistoric Greek Phonological Chronology, with some Reflections on Sibilance, Aspiration, and
Spirancy. Journal of Greek Linguistics 2003, 4, 65-68. Hatzidakis G. N. Linguistic Studies. Athens, P. D. Sakellariou, 1901 (in Modern Greek). Holton D., Horrocks G., Janssen M., Lendari T., Manolessou I., Toufexis N. The Cambridge grammar of Medieval and Early Modern Greek (1: General Introduction and Phonology; 2: Nominal Morphology; 3: Verb Morphology). Cambridge, Cambridge University Press, 2019. Horrocks G. C. Greek. A History of the Language and its Speakers. London; New York, Longman Publishing group, 1997.
Janse M. Cappadocian Clitics and the Syntax-Morphology Interface, in: B. D. J oseph, G. C. Horrocks, I. Philippaki-Warburton (eds). Themes in Greek Linguistics II. Amsterdam — Philadelphia, John Benjamins Publishing Company, 1998, 257-282 (Current Issues in Linguistic Theory 159). Janse M. Convergence and Divergence in the Development of the Greek and Latin Clitic Pronouns, in: R. Sornicola, E. Poppe, A. Shisha-Halevi (eds) Stability and Change of Word-Order Patterns over Time. Amsterdam, John Benjamins Publishing Company, 2000, 231-258. Jernstedt P. Mittelgriechisches und Zakonisches, in: N. A. Bees (Hrsg.) Byzantinisch-Neugrichische Jahrbücher. Internationales wissenschaftliches Organ. Unter Mitwirkung zahlreicher Fachgenossen III. Berlin — Wilmersdorf, Verlag der "Byzantinisch-Neugriechische Jahrbücher", 1922, 81-89. Joseph B. D. Greek, in: B. Comrie (ed.) The World's Major Languages. New York, Routledge, 2009, 347-372.
Joseph B. D. What is Meant by "Modern Greek Dialect"? Some Thoughts on Terminology and Glossonymy, with a Glance at Tsakonian, in: M. Kisilier (ed.) MGDLT9. Proceedings of the 9th International Conference on Modern Greek Dialects And Linguistic Theory. Leonidio, 4-6 June 2021. Leonidio, Tsakonian Archives, 2022, 55-71 (A. Ralli, B. D. Joseph, M. Janse (eds). Modern Greek Dialects and Linguistic Theory (MGDLT)).
Karatsareas P. On the Diachrony of Gender in Asia Minor Greek: The Development of Semantic Agreement in Pontic. Language Sciences 2014, 43, 77-101.
Kassian A. Annotated Swadesh Wordlists for the Greek Group (Indo-European Family). 2018. URL: http:// starling.rinet.ru/new100/grk.pdf (accessed: 26.03.2018).
Kavcic J., Joseph B. D., Brown C. Teaching Modern Greek to Classicists: Taking Advantage of Continuity. Keria: Studia Latina et Graeca 2020, 22 (2), 119-139.
Kisilier M. L. Once again about Ancient Greek and Modern Greek, or do Greeks Remember Ancient Greek themselves?, in: M. L. Kisilier (ed.) Verus Convictor, Verus Academicus. In Honour of the 70th Anniversary of Nikolay N. Kazansky. St Petersburg, Institute for Linguistic Studies, 2022, 351-370 (in Russian).
Kisilier M. L. Reconstructing Past Coexistence: Problems and Mysteries in the Multilingual History of Tsakonia, Greece, in: A. N. Sobolev (ed.) Between Separation and Symbiosis. South Eastern European Languages and Cultures in Contact. Boston — Berlin, De Gruyter Mouton, 2021, 215-263 (Y. Matras (ed.) Language Contact and Bilingualism [LCB] 20).
Kisilier M. L., Fedchenko V. V. Some Remarks on Archaic Phenomena in Modern Greek Dialects. Indoevropeiskoe iazykoznanie i klassicheskaia filologiia 2015, XIX, 368-373 (in Russian).
Kontosopoulos N. G. Dialects and Idioms of Modern Greek. Athens: Grigoris Publishers, 2001 (in Modern Greek).
Kontosopoulos N. G. Linguistic Atlas of the Crete. Heraklion, Crete University Press, 1988 (in Modern Greek).
Koulouri Ch. "Heroic Verses" by Emmanouil Trochanis, the schholmaster from Prastos (1775), in: Proceedings of Arcadian Spiritual Conference 1992 (Leonidio, 21-23 November 1992). Athens, Foundation of Pelopponesean Studies, 1994, 215-241 (in Greek).
Kristoffersen T. R. The Artemis Orthia Inscriptions and Spoken Laconian in the Imperial Period: In Defence of Dialect Survival at Sparta. Glotta 2019, 95, 169-189.
Liosis N. Auxiliary Verb Constructions and Clitic Placement. Evidence from Tsakonian. Journal of Greek Linguistics 2017, 17, 37-72.
Liosis N. Tsakonian, in: G. K. Giannakis, V. Bubenik, E. Crespo, C. Golston, A. Lianeri, S. Luraghi, S. Matthaios (eds). Encyclopedia of Ancient Greek Language and Linguistics 3 (P-Z, Index). Leiden — Boston, Brill, 2014, 446-450.
Liosis N., Papadamou E. The Development of v in Modern Greek dialects: Italiot Greek in Comparison with Tsakonia. Neoelliniki dialektologia 2011, 6, 201-223 (in Modern Greek).
Mackridge P. H. An Editorial Problem in Medieval Greek Texts: The Position of the Object Clitic Pronoun in the Escorial "Digenes Acrites", in: N. Panayotakis (ed.) Arches tis Neoellinikis logotechnias. Proceedings of the 2nd International Conference "Neograeca medii aevi". Offprint from the 1st vol. Venice, Crete University Press, 1993, 325-342.
Mackridge P. H. Language and National Identity in Greece, 1766-1976. Oxford, Oxford University Press, 2009.
Markopoulos A. Education. Ch. 17. Language, Education, and Literacy, in: E. Jeffreys, J. Haldon, R. Cormack (eds). The Oxford Handbook of Byzantine Studies. New York, Oxford University Press, 2008, 785-795.
Mirambel A. Étude descriptive du parler maniote méridional. Paris, E. de Boccard, 1929. (École Française d'Athènes, Travaux et mémoires 1).
Newton B. The Generative Interpretation of Dialect. A Study of Modern Greek Phonology. Cambridge, Cambridge University Press, 1972.
Nicholas N. A Critical Lexicostatistical Examination of Ancient and Modern Greek and Tsakonian. Journal of Applied Linguistics and Lexicography 2019, 1(1), 18-68. URL: https://journall.org/index.php/main/ article/view/9/13 (accessed: 25.02.2023).
Nicholas N. The Shift from Enclisis to Proclisis in Older Tsakonian, in: M. Kisilier (ed.) MGDLT9. Proceedings of the 9th linternational Conference on Modern Greek Dialects and Linguistic Theory. Leonidio, 4-6 June 2021. Leonidio, Tsakonian Archives, 2022, 85-109 (A. Ralli, B. D. Joseph, M. Janse (eds) Modern Greek Dialects and Linguistic Theory (MGDLT)).
Pernot H. O. Introduction à l'étude du dialecte tsakonien. Paris: Société d'édition "Les Belles Lettres", 1934. (Collection de l'Institut Néo-Hellénique de l'Université de Paris II).
Pernot H. O. Le verbe être dans le dialecte tsakonien. Extrait de la Revue des Études Grecques. Tome XXIII, no. 101, Janvier — Février 1910. Paris, Ernest Leroux, 1910.
Prombonas I. K. Rhotacism and Special Pronunciation of the Sound l in the Dialect of Filoti of Naxos. Athens, Foundation of Cycladic Studies, 1963 (in Modern Greek).
Rohlfs G. Grammatica storica dei dialetti italogreci (Calabria, Salento). Nuova edizione interamente rielabo-rata ed aggiornata. München, C. H. Bech'sche Verlagsbuchhandlung, 1977.
Rylik P. A. On a Connection between Modern Greek Idioms Used on the Island of Karpathos and the Dorian Dialect. Indoevropeiskoe iazykoznanie i klassicheskaia filologiia 2012, XVI, 730-741 (in Russian).
Scutt C. A. The Tsakonian Dialect. — I. The Annual of the British School at Athens 1912-1913, 19, 133-173.
Scutt C. A. The Tsakonian Dialect. — II. The Annual of the British School at Athens 1913-1914, 20, 18-31.
Sitaridou I. The Romeyka Infinitive: Continuity, Contact and Change in the Hellenic Varieties of Pontus. Diachronica 2014, 31 (1), 23-73.
Sophocles E. A. Four Hundred and Ninety-Ninth Meeting. October 8, 1861. Monthly Meeting. Remarks on the Dialect of Tzakonia. Proceedings of the American Academy of Arts and Sciences 1860-1862, 5, 296-313.
Tohtasiev S. R. The Language of the Treaty DE ADMINISTRANDOIMPERIO by Constantine Porphyrogeni-tus and its Foreign Vocabulary. St Petersburg, Nauka Publ., 2018 (in Russian).
Tonnet H. Histoire du grec modern. Formation d'une langue. 2 éd. Paris: Langues & mondes; LAsiathèque, 2003. (Collection "Langues-INALCO").
Trudgill P. Modern Greek Dialects. A Preliminary Classification. Journal of Greek Linguistics 2003, 4, 45-64.
Tsopanakis A. G. Eine Dorische Dialektzone im Neugriechischen. Byzantinische Zeitschrift 1955, 48 (1), 49-72.
Tzitzilis C. Archaisms in Modern Dialects. In: G. K. Giannakis, V. Bubenik, E. Crespo, C. Golston, A. Lian-eri, S. Luraghi, S. Matthaios (eds). Encyclopedia of Ancient Greek Language and Linguistics. 2013. URL: http://dx.doi.org/10.1163/2214-448X_eagll_COM_00000034 (accessed: 13.07.2016).
Vayakakos D. Common Dialectal Features of Dodecanese and Mani. Neoelliniki dialektologia 1994, 1, 4363 (in Modern Greek).
Violi F. Vocabolario paradigmatico efraseologico Grecanico — Italiano, Italiano — Grecanico. 2nd ed. (rivedu-ta e corretta con note grammaticali e differenze diatopiche di Bova, Galliciano, Roghudi e Roccaforte esemplificato sul Lessico del Karanastasis e del Rohlfs). Bova, Apodiafàzzi, 2007.
Zaliznyak A. A. Enclitics in Old Russian. Moscow, Iazyki slavianskikh kul'tur Publ., 2008 (in Russian).
Zhirmunsky V. M. Selected Papers. General and Germanic Linguistics. Leningrad, Nauka Publ., 1976, 491516 (in Russian).
Еще раз об архаизмах в новогреческих диалектах*
Максим Львович Кисилиер
Санкт-Петербургский государственный университет,
Российская Федерация, 199034, Санкт-Петербург, Университетская наб., 7-9;
Институт лингвистических исследований РАН,
Российская Федерация, 199053, Санкт-Петербург, Тучков пер., 9; m.kisilier@spbu.ru
Для цитирования: Kisilier M. L. Once Again about Archaisms in Modern Greek Dialects. Philologia
Classica 2023, 18 (2), 339-359. https://doi.org/10.21638/spbu20.2023.214
Одним из важнейших аспектов идентичности современных греков является связь с Древней Грецией, причем как на культурном, так и на языковом уровне. Большин-
* Работа выполнена при поддержке гранта РНФ № 19-18-00244.
ство неоэллиннистов рассматривает древнегреческий и новогреческий не как разные языки, а как разные этапы единого языкового континуума, тем более что лингвистические данные дают для этого множество оснований, несмотря на значительные изменения, затронувшие практически все языковые уровни. В XIX в. ряд европейских специалистов по классической филологии обратился к изучению диалектов новогреческого языка в поисках сохранившихся в них черт древнегреческого. Именно тогда впервые было высказано предположении об уникальности цаконского диалекта, как единственного, возникшего не из эллинистического койне, а восходящего напрямую к древнему дорийскому. Не обсуждая здесь справедливость или ошибочность этой гипотезы, важно отметить, что цаконский далеко не единственный из современных новогреческих диалектов, где встречаются разнообразные особенности, унаследованные из древнегреческого языка. В настоящей статье предпринимается попытка рассмотреть, какие древнегреческие черты представлены по диалектам новогреческого языка в лексике, фонетике, морфологии и синтаксисе. Прежде всего, для описания подбирались особенности древнегреческого происхождения, встречающиеся в диалектах, но исчезнувшие в стандартном новогреческом языке. Материалом для исследования в основном послужили полевые материалы автора и его коллег, собранные в период между 2000 и 2023 гг. Предпринятый анализ показывает, что наибольшее количество уникальных древнегреческих черт обнаруживается в понтийском и цаконском, хотя и в большинстве других диалектов в разной степени присутствуют те или иные архаизмы. Представляется, что число архаизмов не может быть достаточным свидетельством о древности того или иного диалекта, поскольку история развития новогреческих диалектов до сих пор крайне плохо изучена, и непонятно, почему в одних диалектах архаизмы сохраняются лучше, чем в других.
Ключевые слова: новогреческие диалекты, лексические, фонетические, морфологические, синтаксические архаизмы.
Received: April 9, 2023 Accepted: August 12, 2023
