DIVERGENCE OF PROTO-INDO-EUROPEAN VOWELS [ɐ] AND [ǝ] IN PROTO-SLAVIC

ABSTRACT

It is believed that short Proto-Indo-European (PIE) phonemes *[ŏ] and *[ă] became a single Proto-Slavic phoneme *[o], and long *[ā] and *[ō] became a single Proto-Slavic phoneme *[a]. Proto-Slavic short vowel ъ, according to the theory, was also formed from two PIE vowels: *[ŭ] and *[ŏ]. The convergence of PIE vowels in Proto-Slavic looks strange against the background of the divergent development of vowels in Latin and Ancient Greek. This strangeness is a flaw in the existing theory.

In the article, the vowel *[ɐ] with intermediate articulation is considered as a PIE prototype of vowels [o] and [a], in lines with Fortunatov's view on PIE vocalism. The choice is prompted by uncertain articulation of stress-free o/a, which persists in some Slavic languages. It is also shown that the short vowel *[ǝ] was the prototype of the vowel ъ.

Studies have shown that the divergence *[ɐ] > {[o]; [a]} is observed in IE languages regardless of the length of the prototype. Several exceptions have been found: short [a] resulting from long *[ɐ:] and long [a:] resulting from short *[ɐ]. The elimination of quantitative distinction of [o] and [a] in Slavic languages could result from this irregularity, which is more natural than convergence of *[a] and *[o] in the existing theory. Examples of divergent development *[ǝ] are given.

Replacement of PIE prototypes *[ŏ], *[ă] with short *[ɐ], prototypes *[ā], *[ō] with long *[ɐ:] and prototypes *[ŭ] and *[ŏ] with short *[ǝ] allows to uniformly explain the correspondences of a number of vowels in IE languages, regardless of their length, using the divergent evolution of PIE *[ɐ] and *[ǝ] and to explain the observed facts of a-accent and o-accent in Slavic dialects, which are not conditioned either positionally or combinatorically.

АННОТАЦИЯ

Считается, что краткие протоиндоевропейские (ПИЕ) фонемы *[ŏ] и *[ă] перешли в единую праславянскую фонему *[o], а долгие *[ā] и *[ō] перешли в единую праславянскую фонему *[a]. Праславянский краткий гласный ъ, согласно теории, также образовался из двух ПИЕ гласных: ПИЕ *[ŭ] и ПИЕ *[ŏ]. Конвергенция гласных ПИЕ в праславянском языке выглядит особенно странно на фоне дивергентного развития гласных в латинском и древнегреческом языках. Это выглядит недостатком существующей теории.

В статье в качестве PIE-прототипа гласных [o] и [a] рассматривается гласный *[ɐ] с промежуточной артикуляцией, что близко к взглядам Фортунатова на вокализм ПИЕ. Выбор обусловлен неопределённостью артикуляции безударных o/a, которая сохраняется в некоторых славянских языках. Показано также, что прототипом гласного ъ был краткий гласный *[ǝ].

Исследования показали, что дивергенция *[ɐ] > {[o]; [a]} наблюдается в ИЕ языках независимо от долготы прототипа. Было найдено несколько исключений: краткий [a], происходящий из долгого *[ɐ:], и долгий [a:], происходящий из краткого *[ɐ]. Исчезновение количественного различия [o] и [a] в славянских языках может быть результатом аналогичной нерегулярности, что более естественно, чем конвергенция *[a] и *[o] в существующей теории. Приведены примеры дивергентного развития *[ǝ].

Замена ПИЕ прототипов *[ŏ], *[ă] кратким *[ɐ], прототипов *[ā], *[ō] – долгим *[ɐ:] и прототипов *[ŭ] и *[ŏ] – кратким *[ǝ] позволяет единообразно объяснить соответствия ряда гласных в ИЕ языках, независимо от их долготы, с помощью дивергентной эволюции ПИЕ *[ɐ] и *[ǝ] и объяснить наблюдаемые факты a-кания и o-кания в славянских диалектах, не обусловленные ни позиционно, ни комбинаторно.

Keywords: long vowels, short vowels, divergence, convergence, quantitative, qualitative, Proto-Indo-European (PIE) language, Proto-Slavic language, articulation.

Ключевые слова: долгие гласные, краткие гласные, дивергенция, конвергенция, количественные, качественные, протоиндоевропейский (ПИЕ) язык, праславянский язык, артикуляция.

It is believed that in Proto-Slavic dialect of PIE language there was a loss of quantitative differences characteristic of PIE vowels. Wherein the theory says (more details in [1, p. 143]) particularly about the following transformations:

Proto-Slavic *[о] (probably, short *[ŏ] firstly) corresponds to PIE *[ŏ], *[ă], while there are [ŏ], [ă] in other languages (Proto-Slavic *[o] < Proto-Slavic *[ŏ] < *[ŏ], *[ă]):

Proto-Slavic *domъ (> Old Russian домъ) – PIE *dom- (> Latin dŏmus, Ancient Greek δόμος),

Proto-Slavic *оsь (> Old Russian ось) – PIE *ak's- (> Latin ăхіs, Ancient Greek ἄξων);

Proto-Slavic *[a] (probably, long *[ā] firstly) corresponds to PIE *[ā], *[ō] (while there are [ā], [ō] in other IE languages (Proto-Slavic *[а] < Proto-Slavic *[ā] < PIE *[ā], *[ō]):

Proto-Slavic *mаtі (> Old-Russian мати) – PIE *mā-t-er- (Latin mātеr, Ancient Greek Doric μάτηρ);

Proto-Slavic *dаrъ (> Old Russian даръ) – PIE *dō(w)- (Latin dōnum 'gift', dō 'to give', Ancient Greek δῶρον 'gift', but also Latin dare 'to give', datum 'gift', Ancient Greek δάνος 'loan').

In other words, the existing theory postulates that different short PIE phonemes have turned into one Proto-Slavic *[о], and different long PIE phonemes have turned into one Proto-Slavic *[a], i.e. quantitative contrast had been allegedly replaced by the qualitative one. The convergence of different vowels in Proto-Slavic looks particularly strange against the background of the divergent evolution of long PIE *[ō] in Latin and Ancient Greek. Besides, long *[ā] and *[ō] are distinctly articulated vowels, even if they are in unstressed positions, and the confusion of them in several PIE dialects also needs an explanation.

Proto-Slavic vowel ъ, in theory, also comes from two different PIE vowels: from PIE *[ŭ] and from PIE *[ŏ] (in some consonant endings) [2]. I.e., the unmotivated convergence of reflexes of different initial sounds is also assumed with regard to ъ.

We will base the reconstruction of PIE vowels corresponding to Proto-Slavic *[o] and *[a] on the observable phonetic reality, accepting the evolution from uncertainly articulated sounds to distinctly articulated sounds as one of the trends of sounds evolution.

F.F. Fortunatov [3, p. 24 – 46] already assumed the presence of uncertainly articulated vowels in PIE language [see also table 1]:

- short syllabic vowels of full formation:

ă^e – ă tending to e;

ă^o₁ – ă tending to o and close to o, approximately equal to o open ([ɔ]) and of the same origin as ă^e;

ă^o₂ – ă tending to o and close to a, but not of the same origin as ă^e;

- other syllabic vowels of full formation:

ā^e – ā tending to e;

ā^o – ā tending to o (the indexes 1 and 2 have coincided here);

- short syllabic reduced vowel ə (“schwa”) evolving from reduction of long ā;

- irrational super-short vowel α₀ (syllabic and non-syllabic), probably evolving from reduction of short ă ([ʌ]?), which could produce IE long close e in stressed position.

Table 1.

Reflexes of some PIE vowels in IE languages according to F.F. Fortunatov

The introduction of *ă^o₁, *ă^o₂ and *a^a variants did not solve the problem of irregularity of correspondence of the PIE [a] and descendant sounds. There is divergence of *ă^o₁ > {ă, ā} in Ancient Indian, and in Old English we discover a transition into [e] of both *ă^o₂ prototype (in ēowu 'sheep'), and *a^a prototype (in eax 'axis'); in Old Norwegian ɔxull 'axis' we see a transition of *a^a into [ɔ], also. Probably, the solvation of this problem is not in referring to prototype vowels of more distant "before PIE" times, but in accepting the divergence of pronunciation of a PIE vowel with an uncertain articulation, just like the divergence of unstable palatal consonants was taking place [4], [5], [6].

For example, the uncertain vowel *[ɐ] fits this condition (this is how the first vowel in Russian вода [vɐˈda] sounds). As a result, in some IE dialects [ɐ] > [o], in others – [ɐ] > [a], cf. Slovenian vóda and Lithuanian vanduõ.

The sound [ɐ] is similar to *ă^o offered by Fortunatov (a tending to o) [3, p. 24], but the difference with the Fortunatov’s hypothesis is that here this sound becomes the source of different more stable vowels in IE languages.

The examples of reflexes *[ɐ] and *[ɐ:] in IE languages are given in table 2, the lexemes containing the studied reflexes are from [7].

Table 2.

PIE *[ɐ] reflexes in IE languages [8].

In the offered scheme of the phonetic evolution long PIE *[ɐ] usually have reflexes in IE languages as long [o] and [a], the short [a] may be observed in Ancient Greek δανείζω and Latin dare 'to give'. Short PIE *[ɐ] usually have reflexes as short [o] and [a], the long [a] is found in Tocharian ā(u)w and in Old Icelandic āsauδr 'sheep' (tab. 2), i.e., we assume the *[ɐ] > {[o]; [a]} divergence regardless of the prototype length. This is quite consistent with the observed facts as a-accent and o-accent in Slavic dialects, which is not due to either positional or combinatorial conditions. The quantitative differences of [o] and [a] in Slavic languages have eventually disappeared, and the uncertain articulation of unstressed о/а is still present in some Slavic languages.

The divergent evolution of PIE *[ɐ] is not something exceptional in the evolution of PIE vocalism.

With respect to the above Proto-Slavic ъ we can also accept that it had an uncertain short vowel prototype in PIE language, which had transformed into more certain variants in different daughter languages. The best candidate is surely “schwa” (*[ǝ]). It is well-seen in numerous examples from Indo-Iranian languages (cf. Avestan mǝrǝɣa- with Ossetian marg and Persian murg 'bird', or cf. Urdu ləmbə with Hindi lembe, Gujarati lambu and Bengali lomba 'long', examples are from [9]) where we observe the split of PIE *[ǝ] > {[ǝ]; [ŭ]; [ŏ]; [a]; [e]}. In Bulgarian the sound [ə] is still depicted with the letter ъ.

References:

1. Bernshtein S.B. (2005), Sravnitel'naja grammatika slavjanskih jazykov [Comparative Grammar of Slavic Languages]: uchebnik / M.: Izd-vo Mosk. un-ta. – 352 pp. [in Russian].
2. Remnjova M.L., Dedova O.V., Kuz'minova E.A. et al. (2009), Paleoslavistika [Paleoslavistics]. Uchebnyj kurs. © OOO "OSZ", / URL: http://www.philol.msu.ru/~tezaurus/library.php?view=d&course=1&raz=1&pod=2&par=5 (accessed 29.01.2019) [in Russian].
3. Fortunatov F.F. (1957), Izbrannyje trudy [Selected Writings], t. 2. / M.: Gos. uch.-ped. izd. – 472 pp. [in Russian].
4. Telezhko G.M. (2016), O divergentsiji palatal'nogo bokovogo approksimanta v slavjanskih jazykah [On the Divergence of Palatal Lateral Approximant in Slavic Languages] // Universum: Filologija i iskusstvovedenije : elektron. nauchn. zhurn. 2016. № 12(34). / URL: http://7universum.com/pdf/philology/12(34)/Telezhko.pdf (accessed 29.11.2022) [in Russian].
5. Telezhko G.M. (2017), O divergentsiji palatal'nogo frikativa v prajindojevropejskom jazyke [On the Divergence of a Palatal Fricative in Proto-Indo-European Language]// Universum: Filologija i iskusstvovedenije : elektron. nauchn. zhurn. 2017. № 2(36). / URL: http://7universum.com/pdf/philology/2(36)/Telezhko.pdf (accessed 29.11.2022) [in Russian].
6. Telezhko G.M. (2019), On the Divergence of a Proto-Indo-European Velar Syllabic Nasal in Indo-European Languages. DISCOURSE, vol. 5, no. 5, pp. 114-122. DOI: 10.32603/2412-8562-2019-5-5-114-122.
7. Starostin S.A., Indo-European etymology. © 1998–2003. / URL: http://starling.rinet.ru/cgi-bin/query.cgi?basename=\data\ie\piet&root=config&morpho=0 (accessed 29.11.2022).
8. Telezhko G.M. (2019), O vozmozhnosti divergentnoj prirody glasnyh [o] i [a] v praslavjanskom [On the Possibility of Divergent Nature of Vowels [o] and [a] in Proto-Slavic] // Modern Humanities Success, no. 2. С. 103-107 [in Russian].
9. The Indo-European Lexical Cognacy Database (IELex) / URL: http://ielex.mpi.nl/wordlist/all/ (accessed 14.01.2019).