Search

(1974). Some problems connected with the third palatalization of Velars in Slavic. Scando-Slavica: Vol. 20, No. 1, pp. 187-195.

A process of palatalization is assumed where Greek shows a development of velar or labiovelar stop to dental stop or affricate, and of dental stop to affricate or fricative, in an environment conductive to palatalization. The greater palatalizability of labiovelar as compared with velar is explained by the conversion, in a palatal environment, of the labial element to a labio-palatal, with subsequent loss of the labial component. The greater palatalizing effect of the vowel [e] as compared with the more palatal [i] is shown to be entirely in accordance with Greek phonological tendencies; palatalization before [i] is in fact exceptional, and is attributed to the effect of structural pressure. Certain other developments are explained by the hypothesis of a generally pa of the dental stops in Greek, for which there is dialectal evidence. The principles invoked are further applied to a suggested solution of certain problematic IE reconstructions.

(1969). Palatalization in cretan and cypriot Greek. Paper in Linguistics: Vol. 1, No. 3, pp. 593-598.

Acoustic analysis and synthesis of Russian palatalization and Arabic ‘emphasis’ were carried out in order to investigate the Jakobsonian tonality features called sharp and flat. The results indicate that the above features form the extremes of a pitch-like auditory dimension which is related to the separation between the two commonly accepted pitches of vowels — one of which is stimulated by F1, the other by the F2+ group of formants. This suggests a purely auditory three-dimensional vowel space which not only provides theoretical justification for the use of traditional vowel diagrams but also furnishes discrete perceptual correlates for all three of Jakobson's oral resonance dimensions; that is, sharp—ilat, acute—grave, and compact—diffuse.

The author posits that in word final position the Common Baltic group ∗∗∗- Cjai > ∗∗- Cei > ∗- Cei , i.e. in the final step palatalization was lost. Etymological depalatalization is found in those cases where East Baltic ∗- Cei passed to ∗- Cie , cf. nom. pl. masc. pesti pestie pestjai and in the dat. sg. of - e stem nound, e.g. vertei vertjai . In some cases the palatalized consonants have been restored by analogy, e.g. valdžiai (not ∗ valdei as our theory supposes) with dž on the basis of the nom. sg. valdžia . The ∗- ē (Lith. - e ) stem nouns owe their origin to analogy with the ∗- ā stems, i.e. - Cai is to ∗- Cā as - Cei is to ∗- Cē .

In this reply to Harms 1973, a critical look is taken at Harms' arguments against 'abstract phonology'. Particular attention is focused on his unwarranted dismissal of the data from borrowing. In addition, it is argued that considerations of simplicity or economy do not provide any significant insights into the problem at hand. Finally, a modified approach to Nupe palatalization and labialization is outlined.

The Strassburg Oaths of A.D. 842 have long been recognized as occupying a place of prime importance in the documentation of the history of French, since they represent the earliest attempt to set down the spoken language. Their brevity has presented some difficulties to those who have attempted to determine their exact linguistic nature, but conversely has enabled others to form hypotheses as to their dialect which have little support save that the Oaths show nothing to contradict such hypotheses. Modern scholarship, including the rigorous techniques of descriptive linguistics, has done much to correct ill-advised generalizations of this sort. Most recent and most important among the attempts to analyze the Oaths and place them in their true perspective is the study by Robert A. Hall Jr. in Lg. 29.317-21 (1953). Hall first gives the text of the Oaths as it is generally accepted, then phonemicizes it, referring to what is known of earlier and later developments to clarify points for which criteria are lacking in the text itself. Hall assumes the consonants/p bfvmnlrtdkgc s/, the vowels /i e a o u a/, and a contrast of tenseness vs. laxness applying only to the stressed mid vowels /e o/, symbolized /A/; the vowels /a/ and /a/ both occur in final unstressed position. Hall's attempt to classify the dialect of the Oaths is based primarily on the criteria that 'The Oaths do not show any of the specific characteristics of Old North French, such as vowel-raising and diphthongization in PGRom. free syllables, or reduction of unstressed /a/ to /a/, or palatalization of /k/ before /a/, etc.' On the other hand they show no distinctive features of Old South French or South East French. Hall therefore concludes that 'Between PGRom. and the ONFr. and OSFr. literary languages, we must assume an intermediate stage showing the West Romance sound shift which is common to both of these, but showing none of the features peculiar to them individually ..., labelling it simply Pre-French. The dialect of the Oaths of Strassburg is extremely close to this stage, and we might characterize them as having been written in nearly undifferentiated, conservative Pre-French.' If this conclusion is true, attempts to localize the dialect of the Oaths 'are nearly futile'. Furthermore, as a document of such a stage of Pre-French, the Oaths would assume an importance even greater than is usually attributed to them. It is the purpose of this paper to show that such an interpretation is certainly incorrect. Hall infers the existence of final unstressed /a/ from the fact that those forms which show in both ONFr. and OSFr. a final /a/ as a supporting vowel, show in the Oaths a spelling which fluctuates between -a, -o, -e, and -u (fradre, fradra; Karle, Karlo; Deo, Deus; poblo; sendra). There can scarcely be any argument with the conclusion that this 'clearly points to an attempt on the part of the

In the tt-group of Outer South-Ethiopic (Gafat, Northern and Western Gurage), certain morphemes contain or consist of distinctive features to be superposed on verbal rootmorphophonemes. A labial appendice w may be superposed on the last available labial, velar or postvelar consonant, this is the Internal Labialization. If the last root-consonant is a dental or a sibilant, it may be palatalized (End Palatalization). The Internal Labialization is a compensation for the reduction, and later loss, of a final labial element. The End Palatalization is a concomitant dissimilatory phenomenon. The Internal Labialization emerged in the history of these languages several times. This article contains descriptive and comparative statements and a historical reconstruction of these phenomena.

As I have stated several times on other occasions, the more one attempts to compare the two Tocharian 'dialects' with a view to concluding something about the nature of Proto-Tocharian and its position among the Indo-European languages, the more one becomes impressed with the remarkable divergence of development which took place from the period of Proto-Tocharian unity to the time between 500 and 800 A.D. when the dialects appear. About all that really connects the language of Turfan in the east to that of Kucha in the west is a close similarity of grammatical system and a fairly high coincidence of vocabulary. Everywhere we find either the same grammatical function served by morphemes of entirely different phonetic shapes and of different formal origins, or else formal elements the same in origin but serving different grammatical functions. Nowhere is this situation more apparent than in the verbal system. Both Turfanian (Toch. A) and Kuchean (Toch. B) possess four modes-indicative, subjunctive-future, optative, and imperative--and, in the indicative, a present, an imperfect, and a preterit tense. But there the coincidence all but ceases. There are of course some close parallelisms in stem formations in most verbal categories, but there are also marked discrepancies; and of all the verbal categories, it is the formation of the imperfect in Turfanian and its relation to the present system on the one hand and to the preterit system on the other that is most confounding. It is also in the formation of the imperfect that the two languages go farthest asunder. The present paper treats the identification of so-called imperfects and preterits and their ultimate origins. Some of the facts are obvious and already pretty well agreed upon, others are in dispute, and some of the problems have received no attention at all. In our treatment we shall in general proceed from what is most certain to what is more obscure. The formation of the imperfect in Kuchean seems clear enough. It is formed by the addition of a characteristic sign -i(-i-, -y-) or -ai(-ey-) to the present stem. Before this sign, an immediately preceding consonant or consonant group is subject to palatalization: k > 9, ik > c, tk > cc, sk > gs, t > c, tt > cc, nt > Ac, n > K, I > ly, s > f.1 Exceptions to this rule are the imperfects of Present Classes III and IV (cf. below),2 and the present stems in original d, which form the im

[Hittite shows no trace of assibilation or palatalization of the gutturals, and so the development of the Indo-European palatals seems to have been subsequent to the Indo-Hittite period. Hittite shows an independent phoneme (u or M) corresponding to the labial element of the IE labiovelars, and this state of affairs must be assumed for IH also. The loss of this labial sound had begun under certain circumstances in IH, and it continued in primitive IE and in the historic IE languages.]

Few detailed studies of the so-called Tocharian palatalization have appeared, though off-hand mention of it has been current. Perhaps the first serious treatment was that by W. Schulze in an article dealing with the reduplicated preterit in Tocharian and Germanic (BSB 166-74 [1924], phil.-hist. K1. = Kl1. Schr. 239-48),1 but there the discussion was only incidental to the main topic. Recently two more discussions have appeared, one by A. J. van Windekens in his monograph De indo-europeesche Bestanddeelen in de tocharische Declinatie 56 ff., esp. 66 (Louvain, 1940), the other by H. Pedersen in his recent volume Tocharisclf vom Gesichtspunkt der indoeuropaiischen Sprachvergleichung 235 ff. (Copenhagen, 1941). That the subject is a confused and difficult one is generally recognized-but so are most of the problems of Tocharian phonology. And perhaps in this case, as already in others, we have been too prone to jump at conclusions by drawing analogies from our experience in other languages. Indeed the very term 'palatalization' shows our preconception of the nature of these consonant changes before we have established their conditions. In fact, even the consonant changes themselves are established only in the most general outlines. There is general agreement however that the affricates c [tf'] and ts, the sibilants A and ?, the nasal K, and the lateral ly have, in most instances at least, arisen secondarily. And there is also general agreement that both c and ts arise from dentals which are normally reflected by Tocharian t; that ? arises from s, and A from velars, labio-velars and palatals which appear usually as k; and that i and ly reflect normal n and 1. There is less agreement that my and py bear a

[Latin -ksis represented in standard Italian by two 'popular' developments: -ssand -s(saxum > sasso; coxa > coscia). Examination of the geographical distribution of words whose Latin etyma contained -ksshows that the -ssdevelopment was originally characteristic of central and southern Italy (hence also of Tuscan and the standard language), as well as of northeastern Italy with simplification to -s-; and that palatalization of -ksto -isand 4-was originally characteristic of northwestern Italy, whence words with -swere borrowed into standard Italian and thence spread elsewhere in Italy.]