Nyelvtudományi Közlemények 96. kötet (1998-1999)

Tanulmányok - Dezső László: Typological Comparison of Root Structuring in Uralic and Early Indo-European. [Az uráli és korai indoeurópai tőstruktúrák tipológiai összevetése] 3

Typological Comparison of Root Structuring in Uralic and Early Indo-European 21 items. There are two level differences in a limited number of etymons: i-Ä (3), i-A (1). However, y can be related to A more frequently (8 items). The frequency data are relevant if the differences are significant and these data can only be considered from a typological point of view. They-A correspondence again raises the question of the quality of y, which was discussed earlier. In V2

position

y was unstressed and could be posited between и and о and between front and back vowels closer to the latter because у could correspond to reduced и and о according to my assumption. The unaccented i could also be between high and middle positions because it can result from both i and e. If one compares the position of i and у in system S and the unstressed e in R, the difference turns out to be minimal. The co-occurence of unaccented e with both accented front and back vowels shows a pronounciation of e close to front and back positions. A double pronounciation could have been preserved in late PU, reflecting an earlier situation, but the difference could not have existed in late PU (similar to i, y > i or ë, e > e in Hungarian). 3.1.2. Vowel harmony The analysis of Vi and V2 in root structures raises the question of vowel harmony. The question of front-back harmony is well established in the literature. This was a major typological characteristic of PU root structuring. I shall be interested in possible level oppositions. It is not labeled vowel harmony as far as it does not have the frequency, consistency and, mostly, systemic character of harmony. As V2 is reconstructed differently in S and R, and the number of etymons on the comparative list is limited, I shall examine the PU vocabulary of UEW from the point of view of vowel harmony and level opposition. The following analysis will be based on Décsy (1990) for Vi position and on Fancsaly (1988) for V2 (there is a small difference in the total figures). Décsy gives the following figures for Vi ä (48), e (48), i (57), the total front vowels are 153. Back vowels are much more frequent (318): a (110), о (93), и (115). The total of Vi is 471. Fancsaly's data for V2:

a (121), ä (36), e (ПО), -з (201), и and и/о (4); total 472 (to which H (4) and -a (1) is to be added). If we examine front-back harmony, the reconstructed e and non-reconstructed 3 strongly limit our possibilities. However, if we consider the correspondence between front and back vowels in S to E and з in R in V2 position of the comparative list and extrapolate it, then с 50-50% of E represents front and back vowels and с 90% of з „hides" back, 10% front vowels. If we re-calculate the frequencies, the total number of front vowels in V2 will be с 115: ä (36), H (4) and E (55), -з (20). The total back vowels are 361: a (121), u, o/u (4), E (55), 3 (181). If we compare these figures with those of Vb the extrapolation does not