Previous part – Back to the first page
The author presents his own opinion on many of the touched issues here.
In the most early stage of the development of the Indo-European language only one vowel, the A, may be reconstructed.
In unstressed syllables the Indo-European vowel A completely disappeared or was reduced to an indistinctly articulated sound: A. The reduction took place especially in the neighbourhood of the sonorants R, L, M, N, J, W or laryngeals H1, H2, H3.
Things also came to splitting of the vowel A into 3 variants, depending on the kind of the laryngeal in the neighbourhood. By the H1 the front e appeared, by the H2 – a, and by the H3 – the rear o. In other positions one time the front variant e, another time the back one o was introduced (the variant a did not occured here) – the choice depended surely on morphological factors. And then also the qualitative alternation e : o developed. Usually it is considered that similar process did not affect the reduced variant A; in the light of Greek, Slavic and Baltic data it is not so certain.
Early, new vowels i, u, ə (“schwa”) developed as a consequence of total reduction of unstressed vowels in the neighbourhood of the consonants J, W, H (AJA > i; AWA > u; AHA > ə). The development of similar groups after a stressed vowel as well as when some unstressed vowels gathered, caused the origin of diphthongs and triphthongs (AJA > ei, ai, oi; AWA > eu, au, ou; AHA > eə, aə, oə; AHAJA > əi; AHAWA > əu; AJAHA > iə; AWAHA > uə; AJAHA > eiə, aiə, oiə; AWAHA > euə, auə, ouə).
Similar development spread also to groups containing R, L, M, n. From groups of the type ARA, the syllable-forming varieties r̥, l̥, m̥, n̥ are said to have been originated. Perhaps that process occured however only in the next stage and it did not envelop all the dialects.
We can reconstruct for that stage:
The subsequent processes did not envelop all originating dialects to the same degree. An important tendency was the deterioration of laryngeal phonemes, including the vowel ə. And so, in dialects of which the Slavic languages developed, the vowel “schwa” either disappeared completely (e.g. *dhughəter- > *dukter- > *dutjerukā > *dъtjerъka > *dcerka > córka ‘daughter’), or (when being the first vowel within the morphem) it identified with a. When it was an element of diphthongs or triphthongs, it disappeared, but it had caused lengthening and specific rising (acute) intonation of vowels (the rising of the voice tone had remained as a trace after the laryngeal). Perhaps the origin of long acuted vowels let the development of length under other conditions too. It was probably a process determinated mainly by morphological factors, and the new long (lengthened) vowel did not get the rising intonation (but falling – circumflex – or it was articulated with the equal tone). In various words the same root contained either a short or a long vowel – so, the quantitive alternation of vowels originated.
The result of such development was a very complicated vocalic system in which beside short vowels (i, e, a, o, u) also long acuted (í, é, á, ó, ú) and non-acuted (ī, ē, ā, ō, ū) as well as reduced vowels existed. The system of diphthongs was similarly differentiated (the triphthongs which had existed before, developed into acuted diphthongs). Groups of a vowel and a sonorant of the type er, ol, an etc, had close to diphthongs character. Those groups could also be short or long acuted (from groups of the type ARAHA), and perhaps also long non-acuted (because of a secondary, morphologically conditioned, lengthening). And at last, there existed sonants of the type r̥, even if in some dialects (among others Proto-Slavic), instead of them, groups of the type er, or could remain. And here also acuted varieties as well as (what is testified e.g. in Sanskrit) lengthened non-acuted ones existed.
During the development leading to separating of the Balto-Slavic protolanguage, that extremely complicated system underwent considerable simplifications.
Similarly like in many other Indo-European dialects, short o, a mixed together. Also the reduced vowels often mingled, however (contrary to other languages) the opposition of the (more frequent) front e and the back o remained. Their further development caused mixing the front reduced vowel with i, while the back one with u. That way a system of 4 short vowels originated: i, e, a, u.
The long vowels, as well acuted as non-acuted, had all identical development, even if some differences in their intonations remained (a system of intonations has remained as a big archaism till today in some Slavic and Baltic languages, and its traces can be observed also in the Polish language). Late things came to mix ā with ō: till today the Lithuanian language has preserved some traces of separate character of their continuants. Finally, 4 long vowels remained however in the Proto-Slavic dialects: ī, ē, ā, ū (acuted and non-acuted).
Soon after these changes also the system of alternations underwent transformation. Because of the development e > i, o > u also i, u began to be felt as reduced forms of the e, a. The alternations i : e, u : a were diffused perhaps onto many roots where they had not appeared before. Things used to come to development of new i, u (alternating with e, o), and also to their secondary lengthening into ī, ū. Perhaps, as a result of such morphologically determinated reductions, then the new diphthong ui developed of the ai under the same conditions as when in other roots the u existed instead of the a. It is possible to assume (hipothetically) also analogic development of the iu in the place of the eu.
After execution of those changes and simplifications, there remained 6 short diphthongs: ei, ai, ui, iu, eu, au. Baltic facts can testify also a very early change of the eu (and the hypothetical iu) into the group jau, what reduced the number of the short diphthongs to 4: ei, ai, ui, au. On the other hand the old long diphthongs either shortened and mixed with the short ones, or lost their second element and mixed with the long vowels (cf. the variancy of the Polish forms chmara ‘countless number, clouds (of gnats)’ : chmura ‘cloud’ which is a reflexion of the Proto-Slavic alternation āu : au, later ā : au). Similarly the groups of the type ēr, ér as well as īr, ír (from the old lengthened er as well as from the acuted erə) lost their length (or their second element). Like everywhere, the acute lasted as a trace of the once existed vowel “schwa”, and a laryngeal consonant already before. Finally there remained the 16 clusters: ir, er, ar, ur, il, el, al, ul, im, em, am, um, in, en, an, un. Regarding their diversified intonation we can treat them as diphthongs.
Before a spirant the diphthongs containing the nasal element m, n underwent the simplification to the nasal (long) vowels į, ę, ą, ų. Much later the nasal element disappeared in other positions.
Main page – Polish grammar