- This correspondence—historically a convergent development—produces examples where, in one variety, four different forms: mæntEk ‘I save’, mæntæk ‘free (pl.) from’, mEntEk ‘you (pl.) go’, mEntæk ‘they went’, correspond indiscriminately to the single mεntεk (orthographically mentek) in the other variety.—The symbols æ, ε, E, e indicate the four qualities of the “e” phoneme family; the most open quality is symbolized by æ, the less open by ε, the more close by E (these three occur only as short), and the most close by e (this occurs only as long). These phonemes are related to each other by internal correspondences within the varieties of the Standard norm and by pattern congruity.
- Lazicius, in an important theoretical article, “Probleme der Phonologie” (Ungarische Jahrbücher, 15.495–510 [1936]), sets up, in the course of his structural analysis of Hungarian, a bilabial fricative ø as an emphaticum. This sound occurs only in the interjection øɯ, e.g., øɯ dæ mælæg vɔn ‘how warm it is’. Lazicius identifies the second part of this interjection with the long high open phoneme y. (as occurring in f y. ‘grass’). This interpretation is certainly incorrect since this interjectional w is illabial, whereas y is rounded. Therefore, one would have to set up a special vocalic emphaticum. Such an analysis is unwarranted, however, since functionally we have to do with a single global gesture, the segments of which cannot be combined with other elements to contruct signals. The relationship is the same as that between the cat's meowing, which is a single global gesture and cannot be decomposed structurally, and the similarly articulated human imitation, which can. This interjectional øw does belong, of course, to the corpus of Hungarian as a marginal phenomenon.
- I do not think the absence of these vowels has to do with age-group, as Sebeok says in the Quarterly Journal of Speech (27.449–52 [1942]) and again in Language (14.163 [1943]), stating that these phonemes are disappearing in the pronunciation of the younger generation. It would seem that the usage depends on geographical conditions and training in schools which insist upon the distinction. Nor do I agree that the variation is free; at the most, it is partially overlapping. Also, I must note that I have never heard words like ki.nɔ ‘China’, my ‘masterpiece’, or bro∫y.rɔ ‘pamphlet’ with short vowels. Even my statement that some speakers have “in der höchsten Reihe nur kurze Vokale” (Das Ungarische Sprachsystem [Stockholm, 1939, 26]) is oversimplified; the preceding: “In vielen Dialekten kommen die langen Vokale der höchsten Zungenstellung sehr selten vor…” is more cautious, but unspecified.—One might add that there are no keys provided for these letters on Hungarian typewriters. Also it should be noted that even the sequence long vowel—short vowel occurs; e.g. fy·yl ‘like grass’.
- The orthographical representation of this marginal phoneme, a is the only deviation in Hungarian spelling from a consistent usage of one letter for one phoneme. It is usually written with é, which is the normal sign for the long back open a. The existence of the phoneme a is established from such contrasts as ∫vajts ‘Switzerland’/fa·js ‘you are a pain’; ∫pajz ‘larder’/pɔjz ‘shield’. It is interesting to note that no minimal contrasts can be constructed with any other vowel and thus, if the environment is specific enough, a could be regarded as in complementary distribution with a. or ɔ, or probably with any other vowel. This shows how insufficiently the principle of complimentary distribution is formulated.
- Sebeok, in Language (Vol. 19, 1943), lists length as a special feature of the vowel pattern but later, in his Finnish and Hungarian Case Systems: their Form and Function (Stockholm, 1946), uses a geminate transcription. Hall, in Hungarian Grammar (Language Monograph No. 21, 1944), without any discussion whatsoever, regards a phonetically long vowel as phonemically geminate. Garvin also uses a geminate transcription in his article in Language (Vol. 21, 1943). The European structuralists, Trubetzkoy and Hjelmslev were more cautious. Trubetzkoy, in his Grundzüge der Phonologie (TCLP, Vol. 7, Prague, 1939, pp. 174–75), mentions that in Hungarian, the long syllables should be regarded as special phonemes and not as the sum of two short vowels. Hjelmslev, in a cryptic note on the Hungarian connective and the thematic vowel (Bulletin du Cercle Linguistique de Copenhague, 6.12–13 [1939–40]), uses the Hungarian orthographie notation for the long vowels with a basic letter and accent, as opposed to the type without an accent which “cenematically” might correspond to a short vowel or a phonemic zero. None of these writers however, refers to the complexities of the problems involved in Hungarian vocalism.
- This remark has reference to the recent discussions on whether or not the levels of analysis have to be kept apart and whether or not they constitute a hierarchy. I think that operationally, i.e. from the point of view of the subjectivism of the scientist, no valid decision can be made. On the other hand, psychological experiments carried out by proper methods might show socially valid interdependences. In the analysis above, I take the more cautious view of conformity with the physical reality. The problem of length in Hungarian is not identical with that in Finnish—as Trubetzkoy clearly saw, although, using the word, an undefined entity, as his frame of reference, he was not aware of some of the difficulties.
- See the article of Lazicius in footnote 2 and also “A New Category in Phonology,” Proceedings of the Second International Congress of Phonetic Sciences (London, 1935).
- Of the digraphs Tolnai mentions, cz was very soon reduced to a single letter, c. In the evaluation of Tolnai's results, however, there is still some question about his distinction between sound and letter: he lists ly and lly separately from j and jj, although they are only graphically distinct.
- Tarnóczy discusses the sound-symbolic effect of dark vowels and remarks that, contrary to expectations, the vowels are light and not dark in the famous first two lines of “V. László,” a ballad by Arany (1817–82). This statement is not correct, however, since y and ø, which dominate in these two lines, are labial front vowels and they show acoustically lowered formants. For relevant discussion, see Jakobson, Kindersprache, Aphasie und allgemeine Lautgesetze (Uppsala, 1941).
- The 9 vowel qualities with which Tarnóczy operates correspond to the 9 acoustically distinct timbres he distinguishes in A magyar magánhangzók akusztikai szerkezete (The Acoustic Structure of the Hungarian Vowels) (Budapest, 1941) and “Resonanzdaten der Vokalresonatoren,” Akustische Zeitschrift (8.22–31 [1943]).
- The vowel count of the first text was carried out in 1947 in Stockholm, together with a more detailed statistical investigation of Ostyak. The material is deposited at the Hungarian Institute of the University of Stockholm, from which additional information can be obtained. The count of the shorter prose texts was carried out at Columbia University in the same year. The transcription of the texts—mainly the indication of the distinction between æ and E—was made by myself and it reflects my own pronunciation.
- For comparison with Finnish, which has a similar vowel pattern, also uses suffixes as the main morphological process, and has vowel harmony—a restriction of the distribution of vowels in certain morphemic sequences consisting of a root and the following string of suffixes (simple word)—, see J. Lotz, “Speech and Language,” The Journal of the Acoustical Society of America (22.713 [1950]).
- It was not the aim of the above paper to evaluate the numerical data statistically. I should like to note, however, that the Chi-squared test (a standard device used by mathematical statisticians as a test of significance of data obtained from various samples) does indicate that the differences in our two samples, P and M, are too great to be attributed to chance alone. It seems, however, that in order to evaluate the importance for linguistics of such tests, one would have to conduct more specific statistical investigations of linguistic material.
- Choose new content alerts to be informed about new research of interest to you
- Easy remote access to your institution's subscriptions on any device, from any location
- Save your searches and schedule alerts to send you new results
- Export your search results into a .csv file to support your research