Short communicationNeutralization or truncation? The perception of two Russian pitch accents on utterance-final syllables☆
Section snippets
Background and hypothesis
Till the present day most descriptions and transcriptions of Russian intonation have been based on the holistic approach of Bryzgunova, 1977, Bryzgunova, 1980, Bryzgunova, 1984. Non-holistic treatments are, for instance, Fougeron, 1989, Kodzasov, 1996, Kodzasov, 1999, Nikolaeva, 2000, Odé, 1989, Svetozarova, 1982. Recent works by Yokoyama (2001) and Igarashi, 2002, Igarashi, 2004a, Igarashi, 2004b discuss issues on Russian intonation in an autosegmental phonological framework. In my view,
Pitch accents LH*L and LH* defined
According to Odé’s classification, Russian has six types of rising pitch accent, four types with large excursion, and two with normal excursion (for a full description see Odé, 1989, p. 119). Based on a thoroughly analysed corpus of 15 minutes (Odé, 1989, 115pp.), the experimentally verified accents LH*L and LH* discussed in this article have on average the following specifications: an excursion size of >15 semitones measured from the average low level reached by a given speaker in final falls,
Recordings
Recordings were made in order to arrive at a set of short utterances with accents LH*L and LH* realized in words in utterance-final position with ultimate, penultimate and antepenultimate word stress. These utterances could then be used for the verification of the hypothesis in an experimental setting. The recording of these utterances was carried out in Moscow and St. Petersburg. The recordings were made on a Marantz CDR300 digital CD-recorder with a Sony electret stereo directional microphone.
Perception experiment
The perception experiment was conducted to verify the hypothesis that types LH*L and LH* occurring on utterance-final syllables are neutralized. Listeners fulfilled a paired-comparison task; pairs consisted of two stimuli containing realizations either of the same pitch accent or of different pitch accents. Utterances for the stimuli pairs were selected from the recordings described in Section 3. The perception experiment consisted of 180 pairs and 30 native listeners carried out the task.
Discussion and conclusion
The results presented in Fig. 4 clearly show that pitch accents LH*L and LH* on utterance-final syllables are not neutralized and the hypothesis must be rejected. In this position, there is a significant difference between pairs with LH*L vs LH*L and LH* vs LH* on the one hand, and pairs with LH* vs LH*L on the other (see Section 4.2). Scores for the latter pairs are almost at chance level. If there were neutralization of the contrast between the two accents, the middle bar in Fig. 4 would have
Acknowledgements
This research is financially supported by the Netherlands Organisation for Scientific Research, NWO (dossiernr. 355-75-004).
The author expresses her gratitude to Carlos Gussenhoven, Rob van Son and especially to an anonymous reviewer for their useful comments on earlier drafts of this article.
References (21)
- Boersma, P., Weenink, D., 2005. Praat: doing phonetics by computer. Computer program, version 4.3.16. Available from:...
Zvuki i intonatsiia russkoi rechi
(1977)Emotsional’no-stilisticheskie razlichiia russkoi zvuchashchei rechi
(1984)Prosodie et organisation du message. Analyse de la phrase assertive en russe contemporain
(1989)Transcription of Dutch intonation, prosodic typology—the phonology of intonation and phrasing
- Gussenhoven, C., Rietveld, T., Kerkhoff, J., Terken, J., 2003. ToDI, Transcription of Dutch Intonation, Courseware,...
Tak nazyvaemaia neitralizatsia intonatsii—fonologicheskoe opisanie russkoi intonatsii
Bull. Jpn. Assoc. Study Russ. Lang. Lit.
(2002)- Igarashi, Y., 2004a. Segmental anchoring of F0 under changes in speech rate: evidence from Russian. In: Proceedings of...
Fonetika i fonologiia intonatsii v voprositel’nykh predlozheniiakh v russkom iazyke
Bull. Jpn. Assoc. Study Russ. Lang. Lit.
(2004)
Cited by (12)
On the neutralizing status of truncation in intonation: A perception study of boundary tones in German and Russian
2013, Journal of PhoneticsCitation Excerpt :These results support the assumption by Dinnsen (1985, p. 276) mentioned above that the acoustic cues for neutralized forms do not necessarily coincide with the main phonetic manifestation of the underlying phonological distinction. Perceptual impact of truncation in Russian and whether it leads to a complete neutralization of the tonal contrast was explicitly addressed by Odé (2005) who investigated truncated and fully realized forms of LHH- vs. LHL-tonal sequences using the classical ‘perceptual equivalence’ paradigm ('t Hart, Collier, & Cohen, 1990). The stimuli for this perception experiment contained productions of the two patterns in various segmental and syllabic environments by eight fully trained linguists who were native speakers of Russian.
On the acoustic correlates of high and low nuclear pitch accents in American English
2010, Speech CommunicationCitation Excerpt :Another possible explanation is that the height of the F0 peak for the late accented syllable may be influenced by the natural F0 declination which can occur over the course of declarative statements. While the effects of tonal crowding have been described in a number of contexts (Arvaniti et al., 2006; Arvaniti and Garding, 2007; Grabe et al., 2000; Ode, 2005), the principles that govern these effects have not been fully and systematically explored. Moreover, tonal crowding effects can be seen as part of the more general question of how F0 and other acoustic exponents of tonal targets are realized with respect to the segmental content of an utterance.
Phrase prosody of a Southern Russian dialect: Rogovatoe
2024, Voprosy JazykoznanijaSpoken corpora of spontaneous speech as a source to study polar question intonation in Russian dialects
2022, Komp'juternaja Lingvistika i Intellektual'nye Tehnologii
- ☆
The issue discussed in this article was presented at the International Conference on Tone and Intonation in Europe, 9–11 September 2004, Santorini, Greece.