Jelena Krivokapic – jelena.krivokapic@yale.edu
Department of Linguistics
Yale University
370 Temple Street, 302
New Haven, CT 06520-8366
Popular version of the paper 5aSCa9
Presented Friday morning, November 4, 2011
162nd ASA Meeting, San Diego, Calif.
In the process of communication, speakers tend to converge to each other’s way of speaking. This happens on a number of levels: for example, speakers tend to converge at the lexical and grammatical level by choosing words and grammatical constructions that have already been used in the conversation; also the speech rate of participants in a dialogue adjusts to that of their co-speaker, as does their pitch (e.g., Pickering & Garrod 2004, Kim & Nam 2009, Babel & Bulatov 2011). A speaker’s consonants and vowels have also been shown to become more similar in their properties to the consonants and vowels of another speaker in the course of an interaction (Sancier & Fowler 1997, Babel 2010, Nielsen 2011). It has been suggested that the convergence of speakers to each other is an automatic process, although influenced by a number of factors (such as a speaker’s attitude), and convergence is believed to facilitate communication (Pickering & Garrod 2004, Babel 2010). The present study builds on this research and examines whether co-speakers become more similar to each other in the rhythmic properties of their speech.
Languages differ in their rhythmic properties. For example, American English has linguistic stress (stress refers to the prominence of a syllable, thus in a record , the stress is on re and in to record, the stress is on cord). It has been suggested that in English, but not across all other languages, the duration between stressed syllables (an interstress interval) in speech tends to be regular. Thus in the sentence below, the interstress intervals (labeled 1-4, with the stressed syllables shown in bold) tend to be of equal duration when spoken.
[Mary]1 [bought a]2 [record in the]3 [store]4.
It has been argued that this type of rhythm is maintained by adjusting the duration of the stressed syllable as a function of the overall number of syllables that occurs in the interstress interval. The stressed syllable shortens as more unstressed syllables are packaged into the interstress interval (Kim & Cole 2005). In contrast, some other languages such as French, Spanish, Italian have been argued to have a regularly timed inter-syllable, not the interstress, interval.
To examine whether convergence is possible at the rhythmic level across languages with differing rhythmic patterns, an acoustic experiment investigated how interactions between speakers of American and speakers of Indian English play out in the rhythm of their speech. The two languages differ with respect to their rhythmic properties, with Indian English reported to have regular intersyllable intervals and American English reported to have regular interstress intervals.
Speakers’ productions were examined in an experiment consisting of two parts: in the first part, speakers read a short story by themselves, and in the second part, a few days later, they read the exact same text again but this time synchronously with a co-speaker (Cummins 2002). The two productions of the same story—solo and synchronous—are compared to look for changes in rhythmic properties when speakers of different dialects are paired to read together synchronously.
Data from six adult subjects (three Indian English and three American English speakers) were collected. In the synchronous reading part of the study, an American English and an Indian English speaker were paired. Acoustic measurements were conducted to examine a) the rhythmic properties of the two languages as reflected in differences in the duration of stressed syllables and b) the changes that occurred in these durational properties in the synchronous speech task.
Results of the data analyzed to date indicate that speakers converge to each other in the synchronous reading task. Specifically, with the increase in the number of syllables in an interstress interval, the stressed syllable shortens for speakers of Indian English in the synchronous compared to the solo condition, and shortens less for speakers of American English. This is the result that would be expected if speakers are converging to each other’s productions.
This finding contributes to the body of evidence showing that there is a tight link between production and perception in speech. Adults make use of this link in adapting their production to that of their co-speaker. Such adaptations can form the basis for ongoing language learning and accent change throughout the lifespan.
References:
Babel, M. (2010). Dialect divergence and convergence in New Zealand English. Language in Society, 39, 437–456.
Babel, M. & D. Bulatov. 2011. The role of fundamental frequency in phonetic accommodation. Language and Speech, 1-17 (doi:10.1177/0023830911417695)
Cummins, F. 2002. On synchronous speech. Acoustic Research Letter Online, 3, 7-11.
Kim, H. & J. Cole. 2005. The stress foot as a unit of planned timing: Evidence from shortening in the prosodic phrase. Proceedings of Interspeech 2005, Lisbon, Portugal.
Kim, M. & H. Nam 2009. Pitch accommodation in synchronous speech, Journal of the. Acoustical Society of America, 125:4, 2575.
Nielsen, K. 2011. Specificity and abstractness of VOT imitation. Journal of Phonetics, 39, 132-142.
Pickering, M. J., & S. Garrod. 2004. Toward a mechanistic psychology of dialogue. Behavior & Brain Sciences, 27, 169–226.
Sancier, M. L. & C. A. Fowler. 1997. Gestural drift in a bilingual speaker of Brazilian Portuguese and English, Journal of Phonetics, 25, 421–436.