Palatalisatsioonist ühesilbilistes i-tüvelistes pika vokaaliga sõnades. Roos närtsis, sest vaas oli tühi
Palatalisation of Consonants after a Long Vowel at the End of Monosyllabic i-Stemmed Words

Summary/Abstract: The study focuses on the palatalisation of the consonants l, n, s, t and d at the end of i-stemmed monosyllabic Estonian words, e.g. saal ‘hall’, veen ‘vein’, roos ‘rose’, pruut ‘bride’, kood ‘code’. The purpose of the study is to ascertain the extent of palatalisation. According to the Dictionary of Standard Estonian, which provides the basis for Estonian text-to-speech synthesis, the final consonant of an i-stemmed word with a (C)VVC structure should be palatalised. There is, however, large variation in actual pronunciation. The investigation addresses 2274 pronunciations of 57 words with the above structure. There are 36 i-stemmed words and 21 words with some other stem vowel, chosen to enable a comparison of the acoustic features of the palatalisation. The target words have been embedded in sentences read out by 40 informants representing 10 different historical dialect areas of Estonia. The target words were analysed both by audio assessment and by measuring the transition phase inside the long vowel. Previous speech studies have shown that the characteristic feature of Estonian palatalisation is an i-like transition phase, which constitutes more than a quarter of the total duration of the vowel. The results of this study show that according to the audio assessment, 29% of the consonants l, n, s, t, d at the end of the i-stemmed monosyllabic words analysed were perceived as palatalised; according to the duration of the transition phase the percentage of palatalisation is 33%. Palatalisation occurs in more than half of the pronunciation instances only in 5 words out of the 36. The results confirm that the palatalisation of the consonants l, n, s, t, d at the end of i-stemmed Estonian words with a (C)VVC structure is generally untypical. In addition, some linguistic and sociolinguistic factors were investigated. It occurs that the palatalisation of the final consonant is influenced by the preceding vowel. According to the audio assessment the consonant is not palatalised when following the vowel ää; it is often palatalised when following the vowel ee. The acoustic analysis, however, does not support those results. The informant’s dialectal background was studied as a sociolinguistic factor. The results show a rather slight correlation between the present-day pronunciation and the palatalisation used in the respective historical dialects, but the general tendencies are still findable.

• Details
• Contents