Contents

Number 2 2005

Taive Särg, The Early History of Estonian Speech Prosody Studies
Abstract. The paper provides an overview of research on Estonian speech prosody from the 17th century to the beginning of the 19th century, focusing on the discovery of three levels of word stress and quantity. The priests of mainly German origin founded Estonian linguistic tradition. The lexical stress on the first syllable was probably known before it was first mentioned in the grammar (Stahlen 1637); the existence of secondary stress was clearly stated by J. J. A. Hirschhausen (1827). Till the 19th century it was difficult to realize that the duration was not a correlate of stress in Estonian. The Estonians' perception was of utmost importance in discovering the three quantity degrees, as the (supposed) first recording of three quantities in vowels ("Agenda Parva" 1622 by unknown author, presumably of South-Estonian origin) as well as their theoretical description (Masing 1824) were given by native speakers.

Meelis Mihkla, Jüri Kuusik, Analysis and Modelling of Temporal Characteristics of Speech for Estonian Text-to-Speech Synthesis
Abstract. A text-to-speech system must be capable of generating sounds and pauses with such durations that do not noticeably differ from natural speech. Currently, the prosodic modelling of Estonian text-to-speech synthesis is largely based on generalized measurements of speech units in isolated words and sentences, and as a result the synthesized speech is often monotonous and has poor fluency. In this work the first attempts are made to improve the naturalness of the output speech of the speech synthesiser with the help of statistical duration models of fluent speech. The source material consisted of (a) prose read out by a professional actor, and (b) news broadcasts read by announcers. On the basis of this material variability of the duration of pauses and boundary lengthenings was investigated. It turns out that in the case of a read text at normal speech rate the classification of speech pauses is perfectly possible and can be applied in speech synthesis. An attempt was also made to establish whether and to what extent the syntactic parsing of a text is related to the prosodic parsing of speech. A generalized regression analysis revealed what features are essential in predicting sound durations in speech and a statistically optimal model was developed. Curiously the quantity degree of a foot, despite being the cornerstone of Estonian word prosody, was not a significant feature for prediciting the duration of a sound on the basis of this material. The results of the modelling were then compared with the expert opinions of some Estonian phoneticians.

Karl Pajusalu, Toomas Help, Pärtel Lippus, Ellen Niit, Pire Teras, Tiit-Rein Viitso, On the Temporal Structure of Estonian Secondary-Stressed Feet
Abstract. The article focuses on Estonian words of two and more feet, comparing the temporal structures of the feet in words of four to six syllables that consist of short primary- and secondary-stressed syllables. A phonological introduction to the problem of secondary stress in Estonian is followed by the treatment of lengthening of vowels in secondary-stressed feet. The article then analyses duration ratios on the basis of phonetic data. The pronunciations of the western and eastern peripheries of Estonia — western Saaremaa and eastern Võrumaa — are compared to follow possible regional differences in Standard Estonian. It appears that only in Saaremaa the unstressed vowel in a secondary-stressed foot has generally lengthened, and the duration ratios of the secondary-stressed foot are similar to that of the primary-stressed foot. In both areas the temporal structure of feet depends on the general structure of the prosodic word.

Eva Liina Asu, Intonational Contour Alignment in Saaremaa and Standard Estonian
Abstract. The dialect of Estonian normally singled out as differing from the standard in its "sing-songy" melody is that of the island of Saaremaa. The characteristic intonation of this variety is attributed to historic contacts with Swedish and has been shown to be phonetically manifested in differences in peak alignment. The present paper extends the analysis of two previous studies (Asu 2004a; 2005) dealing with tonal alignment in the Saaremaa variety by concentrating on the study of the nucleus in declarative read sentences.The nucleus in Standard Estonian is often low and level, the pattern which in autosegmental-metrical analysis of intonation is labelled as H+L*. The delay of intonation peaks which implies a rightward shift in Saaremaa Estonian represents a trend towards lower pitch on the accented syllable. It is tested here whether this shift to the right precludes a distinct low accentuation. It appears that this is not the case and that both intonation patterns are distinct also in the Saaremaa variety. The finding suggests that the peak delay is a local phenomenon involving specific pitch events rather than a global displacement of the whole intonation contour.

Ellen Niit, Vowels in Initial Syllables in Saaremaa
Abstract. The article provides an overview of vowel quality in the initial syllable of words having various degrees of quantity in the insular dialect of Estonian. The material consists of eight sub-dialects of Saaremaa. The article presents measurement findings of the formant analysis and their discussion. The acoustic-phonetic data served as a basis for establishing the systems of vowels in the sub-dialects of Saaremaa. The results allowed distinguishing 8 vowels (a, o, u, ä, e, ö, i, ü). Length affects vowel quality — overlong mid-high monophthongs are raised. The area under discussion has no unrounded central vowel õ of Standard Estonian; it has developed into a mid-high rounded front vowel, the quality of which is close to the short and long ö but not the overlong ö. The vowels ö and open ö have merged. Only in some words in western Saaremaa ä has developed into a mid-high short or overlong monophthong, the quality of which is close to the short and long e but not the overlong e. The sound ɛ is an allophone of e.

Sulev Iva, Glottal Stop in Võro South Estonian
Abstract. One of the most striking characteristics of Võro South Estonian is the glottal stop ʔ that occurs as a peculiar consonant phoneme. The present article investigates the occurrence, quality, and duration of the glottal stop, and the F0 contour of the words with the glottal stop, in controlled and spontaneous speech from the East-Võro area. The acoustic analyses have been carried out with the program PRAAT. The results show that the Võro glottal stop occurs always as a word-final stop whose duration is similar to that of word-final semi-long consonants. The glottal stop is pronounced more frequently in certain word and sentence structures and often gets assimilated with the initial consonant of the following word. The Võro glottal stop has characteristics triggering both phonemic and morpho-phonological processes as well as intonational characteristics. The Võro glottal stop is a phoneme that has certain prosodic features that are above all used for highlighting grammatical and communicational meanings.

Niina Aasmäe, Jaan Ross, How Free is Alternating Stress in Erzya?
Abstract. The present paper addresses the subject of the variability of stress assignment in Erzya. The aim of the analysis reported here was establishing speaker-, utterance- and word-related data that might account for the alternation of the initial and non-initial stress patterns. The material used for the analysis contained spontaneously produced utterances recorded by 33 speakers of Erzya originating from different areas of the Mordvin Republic and diaspora. The overall distribution of the stress patterns in the data of spontaneous speech proved to be more variable than it has been observed in script reading. The occurrences of initial stress exceeded those of non-initial stress; however, non-initial stress was used more often than in reading. The analysis revealed differentiation in the distribution of the patterns of stress in the idiolect-related data. The tendency towards the assignment of initial stress was more salient in the data of the idiolects characterized by the reduction of vowels in non-initial syllables. There was less difference between initial and non-initial stress occurrences in the data of the idiolects which use only full-formation vowels. In the dialects that use reduced vowels, reduction might have been the consequence of the diminishing mobility of stress. The assignment of stress, as the test material showed, could not be conditioned by the type of word structure. It is rather the functional role of the word in an utterance that might be associated with the mobility of stress. The recurrent responses of the speakers recorded in a dialogue were found to contribute to the increase of non-initial stress occurrences.

REVIEWS

N. S. Aduškina, M. D. Imajkina, K stoletiju so dnja roždenija N. F. Cyganova (1904—2004)

Sirkka P a i k k a l a, Se tavallinen Virtanen. Suomalaisen sukunimikäytännön modernisoituminen 1850-luvulta vuoteen 1921. Suomalaisen Kirjallisuuden Seura. Helsinki [—Tampere 2004]. Rez. von Marja Kallasmaa

Tat'jana T r o j a n o v a, Antropotsentritšeskaja metafora v russkom i estonskom jazykach (na materiale imen suštšestvitel'nych), Tartu 2003 (Dissertationes Philologicae Slavicae Universitatis Tartuensis 11). Rets. Pille Eslon

M. V. Mosin, D. V. Cygankin, Aleksandr Feoktistov (27 avgusta 1928 — 6 sentjabrja 2004)

E. N. Mustajev, Pamjati Fedora Ivanovitša Gordejeva (1929—2005)

Auf der Titelseite: Phonetikbücher von Paul Ariste