o, o. e. uo, ie – kuscholarworks.ku.edu

Loading...

Taking too long?

Reload document

| Open in new tab

ON THE VOCALIZATlON OF JERS IN SLOVAK’

The rise of the modern renexes of the historically “reduccd” or centrali z
cd short vowels Uers) in the Central (C) dialeets of Slovak (Slk) has rc
mained a longstanding unresoved problem of the reconstruction of the
Common Slavie dialeets. As is well known, in some C Slk dialeets a num
ber of phonetic realizations occur as reflexes of strong jers, namely thc
phonemes o, tJ. o. e. II, i, uo, ie and d. several of which may cooccur in
a given loca! dialect. Habovštiak’s (1954, 1965) material from the C Slk
dialect of Orava illustrates variation indicative of the complexity of this
problem:
~ > a, c.g. max « *tmXlJ) ‘mass’ (Upper Orava);

*r&dtca) ‘caraway’ (see Habov~liak 1958);

> tJ, e.g. rtJsca «
> o, e.g. voš « *vM’6) ‘flea’;
> u, e.g. ku mhe (.n mbll€)
> ‘i, e.g. ľi§ka «

6 > ‘IiI’a, e.g. l’lin/ľan « ·16ll’b) ‘flax’;

> e, e.g. pes «
lakeľ «

• PIUb) ‘dog’;

·OI1nJ6) ‘elbow.

‘towards me’ [C and Upper Orava);

·h.VJca, Standard Slk lyžica) ‘spoon’;

No cohercnt set of phonological conditions has yet been set forth Ihal
would account for this heterogeneous set of correspondcnces. As a resu!t,
schol ars have of ten attributed this variation to such diverse principles as
morphologically motivated rcshuffling or the influence of a substraluml .

• I wouJd like to express my gratitude to Profs. Henrik Birnbaum and Alall Timber

lake for their comments on earlieT versions of this paper.

Citation of forms in this paper follows that traditionally uscd in the scholarly lillguj·
stk literature of the individuallanguages, e.g., length is marked in Czech and Slovak by
Ihe acute <á>. bUl in Soulh Slavie by <1>, <á>, <ä> where lenglh is concomitant
with toneme and ietus and by where it is not. The grapheme is used for the
vokele» velar frieative in Czech and Slovak forms to avoid the traditional digraph .
, ThoUJh il will not serve our purpose to argue systematically against previous ex
pIanalions of the Slk jer development, v..’C shalllist here some of the thoughts on the sub
jecl thal have been presented in the literature. Sumce il IO say that no scholar has ye:t arriv
ed at an airtight solution. Hypotheses on Ihe devdopmelll of the jen in Slk number
virlually as many as the scholan who have wriuen about them. Diels (1914) connects the

44

Marc Greenberg

There is, however, a fairly coherent isogloss pattern which suggests that
therc was some phonological regularity in Slk jer vocalization. In connee·
tion with this pattern we will suggest a possible explanation for the devel
opment as a reHc of Common Slavie prosodic distinctions. White the dis
cussion cannat aim to be definitive, we hope at (east la have shed light
on this complex problem of Slavie historical dialectology.

As with most aspects of the historical grammar of Slk, thc jer ref1exes
are considered in the literature to be divided between the Western (w) and
Eastern (E) dialeets versus the C dialects (VáŽflÝ 1934: 219 rf.)1, This is ge-

C Slk area with the South Slavie macrodialect, explaining the occurrence of the aberrant
reflex a as an anaptyctic vowel generalized from forms where the jer should have been lost
to the form where the jer would have been normally vocalized as a Thus, e.g., ·m”XU (Gsg)
> ·mxu oa > muxu, ·mbXb (Nsg) > ·mox = > max. Conev lists the Slk jer reflexes
as one of the phenomena showing the linguistie similarity of Slk and Bulgarian, a fact
which to him indicates their contiguity at the time of a presumed common Slavie dialect
continuum (1919: 4Off.). Melich argues against the transitionalist theory that C Slk links
the E and S Slavie jer developments (192S: 324-325). He claims that lo and l> retained their
original identity in Proto-Slk and subsequently devcloped into two phonetie (1) variants
of Icl in Old Slk, Le., lo > e “dunkel” (corresponding to Old Cz ue §irok4!, temné” in
Oebauer’s terminology IS94: 58) and l> > e “hell” (“uzk4!, jasn4!”). Then this e could devel
op along with etymologieal e and l in two directions, either I. > o, io or 2. > a > d,
‘a, ia (334). A similar explanation was given by Smilauer, who c1aimed that i and b became
one vowel which was reinterpreted as either e or tJ (and subsequently a or o) depending
on the consonantal environment (1930). Novák (1931, 1934, and his recent synthesis of
earlier work 19S0) sees the changes” > e, lo > o as the resular development in C Slk, while
instances of these reflexes in W and E dialects are due to misration of individual lexical
items (19S0: 155 ff.). According to Novák, the a reflex atose from the White Croatian sub
stratum that presumably had the reflex a (in his opinion a IO c. phenomenon) for the jers
and remained alongside indigenous forms (174-177; see also note 3).

1 The diacritic features of the C Slk dialects that have received the most attention
in Sl;holarly research are the so-called “Yugoslavisms;’ i.e., S Slavie features found in these
and no other W Slavie dialects. The following S Slavie features in C Slk dialccts are gener
ally cited: l. AP C ·ort·, “oIt· > ra/-, lat· (cf., C Slk rasriem ‘l grow! laket’ ‘elbow’; W
Slk ros/ern, loket; SCr rasti, ráste:m, Sn láket); 2. ·dl, -ti> I (cf., C Slk Jj’lo ‘awl: krilo
‘wing’; WIE Slk Jidlo. kridio; SCr!llo. krno); 3.:i > s as a result of the second velar pala
talization (ef. C Slk mň(x, mflfsi ‘monk, mon ks’; Cz mflix, mfliJi; SCr mOflah, mOflóSI);
4. thc Ipl ending -mo (ef., S-C Slk kos/mo ‘we mow’; Cz kos(me; Sn kos/mo). While it
will not be our place here to argue for or against the concept of Yugoslavisms in C Slk,
we shall agrce that there is at least some evidence that S Slavie exerted an influence on the
early development of the pre·Slk dialects. It should be left to a more detailed study to re
construct the processes (substratum, contact etc.) involved in this influence, as well as those
W Slavie features in S Slavie (e.S., lenition of g [Sn Primorsko dial. hriem < ·gred- 'I go'l, retention of dental + liquid clusters (Snprivedlaje 'she brought'J). For further ďiSl;ussion, see also Conev 1919; Palivu 1922/3; Malecki 1931; Ramovf 1933; Stieber 1933; Váfný 1934; Kniezsa 1948; Stanislav 1958; Pauliny 196]; Krajčovič 1965, 1974, 1975; and Novák 1980. On Ihe VocaHuuion Or Jcrs in Slovak 45 nerally true, though the W border of the C reflexes has moved west ward for some etyma (e.g., the rorms max/mox are round aJongside the expect ed ma in the W dialects). It appears that part or the process of vocaliza· tion characteristic or the C dialects opera ted in the E a[ca as well. Bc cause the rall or the jers is a cruciaJ early phenomenon or the Late Com mOD Slavie (LCS) period or dialectaJ dirrerentiation, an unde.-standing of the conditions which created the modern day picture will shed light on the preh.istory or Slk. Ln LUm, the Slk picture is importam rOJ the overall reconstruction or Slavic jer loss and vocalization, a process generally re cognized as the terminus ad quem in the disintegration of LeS dialcet continuity (Birnbaum 1975). There c.xist no one-Ia-one correspondences between reconstructed jers and their C Slk reflexes. No clear set of conditions (e.g., prosody, conso nantal environment) thoroughJy states the process or jer vocalization in C Slk. This is not to suggest, however, thal such conditions did not play a role in its development, but ralher that they are not immcdiately evident rrom the surraee data. Since the application or the comparalive method has failed to turn up an adequate description or Slk jer locaJization, our study will auempt to arrive at a hypothesis about the conditions indirect ly, tak ing as a starting point an analysis of the IinguJ stic geography of the modern ref1c.xes. We analyze the isoglass patlerns of the jer reflexes presented in the maps in the Slovak dialect atlas, Atlas slovenského joz.yka (1968). This source gives a maximum amount of information for the geography of certain words. Because the number of words is smaJl and only isolated forms are given, it is impossible to investigate all of the conditions (Lc., consonanlal environmem, aecent, word position, and paradigmatic alter nation, among others) that might be relevant to the problem. For this reason, the conditions giving risc lo the diverse reflexes cannot yet be stat ed exhaustively. Rather, our tentative inferences must rely upon an analy sis of the data on Iinguistic geography that is available. This analysis will yield some clues about the reconstruction, though nOI the entire picture of the process of jer vocalization. As mentioned above, Slk has more than two vowel phonemes as mo~ dern reflc.xes for the two jers traditional1y reconstructed for Common Slavie and atlestcd in early Slavie texts. For both .", and ·6 Ihe reflexes e, a (both their long and short variants), o. and occasionally i and u are represented in the diaJects (Le., M;ica, ku). We shall be concerncd only 46 Marc Oreenberg with the a, o and e reflexes, sinec the i and u reflcxcs are limited to a small number of lexical iterns. A difference between front and back jer existed in Early Slk, as palatalization (and laler assibilation in some dialeets) look place before -6 in all three major dialcet zones. "[ypically in C Slk the ref]exes of vocalized" is e with preceding palatalization or assibila tion, e.g., ťel)kí « ·tlllTllnJ6) 'thin (m. sg.)' (Orava). But also the reflex a is sometimes found as a reflex for ll, e.g., ľan « 16fl6) 'nax' (Orava). Back II may be vocalized as either o or Q and somelimes e (without preceding palatalization or assibilation). It is the a/o bifurcation that will receive most of our attention. AII vowel reflexes are found lO a greater or lesser extent over the eRlire Slk territory. GEOGRAPHY As a preliminary to OUt proposal of conditions rclevan! to the reconstruc tion of jer vocalization in Slk we shall examine the geography of modern day jer ref1exes in the Slk territory in order to determine whether pauerns of isoglosses occur. We shall assume that the patterning of isoglosses rc· presents a dyna mie piclure of the development of the various ref1exes. Moreover, we shall operate with the widely held view that linguistic inna vatian tends to movc in waves from a center to a periphery. Moreover, as a means of dealing with further complexity in the patterning of linguistic geography, we shall reref to Andersen's typology of isoglosses (1978: 3): 1. simplex. where the isogloss line delineates an innovation in one dialect area from the absence of the innovation in the neighboring dialect area; 2. duplex, where the isogloss represents the division between two logically alternative innovations; and 3. comp/ex. where the same innovation has different consequences due to underlying structural differences between the two areas. l. o II e ISOGWSSES Certain etyma show only two reflexes. o and e, without the third reflex o. Such words include von II ven 'away, 'outside! voJ II veJ 'Jouse! which separate the W (e) from the C and E (o) areas (see schematie map A). This isogJoss cuts clean1y aJong the W-C isog1oss bundle. However, the On the Vocalization of Jers in Slovak 47 piclure is somewhat obscured in the south C area by non-vocalic variants that appear to have Icveled with the oblique cases and reidentiried with the a·stem noun declension:fla, w~a> pJa. Nevertheless, the northern de
velopment is shaped identically with that of ven II VO”, such that we can
rcconstruct a nearly idemical distribution for these two etyma.

A. von X ven

B. xfaptok X xfaptek

The clearest picture of the situation is presented by the patterning of
words derived with the suffixal formants •. -“h/.-IJk1, and ··6ttkb. o
against e as a reflex of both front and back strong jers in these suffixes
has the shape of an innovation moving in (wo parallel waves from south
to north in the south C and the easternmost extent of the Edialeets. The
most rcstrictcd instanee of this isogloss is defined by the etymon xfapčok
II xlapček ‘boY, which dcfines o areas in lhc southern part of thc C and
the eastern half of the E dialects (see map B).

A funher development is represented by the etyma piesok II piesek
‘sand (N. sg.)’ and kuoročka II kuorečka ‘tree bark (dim.): where the o
innovation covers thc”entire C Sik dialect area and the eastern half of the
E dialccts. At its fulJest realization, e.g., sviečok II svietek ‘candie (Opi):
the O area covers almost the entire C and Edialeets with the exception
of the northern periphery of both areas, while the W dialccts have uni
form e reflexes. Tripartite reflexes with this suffix occur in the north
western part of the C dialects where the o innovation has reaehed this
area: slovak 11 slovok II slovek ‘a hundred’ (see also below).