Evidence for attractors in English intonation

Abstract

Although the pitch of the human voice is continuously variable, some linguists contend that intonation in speech is restricted to a small, limited set of patterns. This claim is tested by asking subjects to mimic a block of 100 randomly generated intonation contours and then to imitate themselves in several successive sessions. The produced f 0 contours gradually converge towards a limited set of distinct, previously recognized basic English intonation patterns. These patterns are "attractors" in the space of possible intonation English contours. The convergence does not occur immediately. Seven of the ten participants show continued convergence toward their attractors after the first iteration. Subjects retain and use information beyond phonological contrasts, suggesting that intonational phonology is not a complete description of their mental representation of intonation.

Exploring the relationship between intonation and the lexicon: Evidence for lexicalised storage of intonation

“…We are looking, however, into possible methodologies that could test this directly. For example, we could extend the iterative mimicry study by Braun et al (2006) to see whether the type of intonation attractors changes depending on the frequency of the words in the phrase.…”

“…However, there is evidence that the acquisition of the prosodic word is frequency driven (Vigário et al, 2006), that word stress assignment can be inferred by instance-based learning (Daelemans et al, 1994), and that distributional properties of syllables influence the predictability of their durations (Schweitzer and Mö bius, 2004;Walsh et al, 2007). For tonal parameters, there is very little research investigating frequency effects, with the exception of Braun et al (2006) who showed that random tonal contours gravitate towards frequent contours in an iterative mimicry study. There is also very little research that explicitly tackles the question whether intonation can be stored lexically.…”

“…These studies are firstly situated in the domain of psycholinguistics, where several experiments demonstrate that the familiarity or frequency of prosodic parameters influence speech processing, perception and production (Braun et al, 2006;Braun and Johnson, 2011;Mandel et al, 1994;. A second research area which provides evidence for lexicalised storage of intonation is the area of machine learning, where various studies showed that word identity helps in predicting pitch accent location (Brenier et al, 2006;Nenkova et al, 2007;Pan and Hirschberg, 2000;Pan and McKeown, 1999), and where instance-based learning of prosody outperforms other types of learning (Marsi et al, 2003).…”

Rhythm measures and dimensions of durational variation in speech

“…It is described more fully in ; it has been used in prior work, notably Alvey et al (2008) and Braun et al (2006), and is available for download (Kochanski, 2010b). It is first run to convergence (via stepper:run to bottom in mcmc helper:py) and then run to generate (in this instance) 20 samples ofC from the distribution of classifiers for each test/training split.…”

Pitch Range Variation in English Tonal Contrasts: Continuous or Categorical?

“…In production tasks, phonetic evidence of phonological contrast is taken to correspond to discreteness or discontinuity in imitation of continuous phonetic variation, e.g. f 0 maximum timing or F1-F2 values [Pierrehumbert and Steele, 1989;Viechnicki, 2002;Niebuhr, 2003;Redi, 2003;Niebuhr and Kohler, 2004;Dilley, 2005;Braun et al, 2006;Dilley, 2007]. In perception tasks, phonetic evidence of phonological contrast is typically taken to correspond to categorical perception, which corresponds to a maximum in discrimination and an s-shaped labeling function, given continuous phonetic variation [Liberman et al, 1957;Repp, 1984], but other kinds of perceptual tasks have also been taken to provide evidence relevant to evaluating phonologic representations [e.g., Gussenhoven and Rietveld, 2000].…”