Exploring the role of exposure frequency in recognizing pronunciation variants

Abstract

Words can be pronounced in multiple ways in casual speech. Corpus analyses of the frequency with which these pronunciation variants occur (e.g., Patterson & Connine, 2001) show that typically, one pronunciation variant tends to predominate; this raises the question of whether variant recognition is aligned with exposure frequency. We explored this issue in words containing one of four phonological contexts, each of which favors one of four surface realizations of word-medial /t/: [t], [ʔ], [ɾ], or a deleted variant. The frequencies of the four realizations in all four contexts were estimated for a set of words in a production experiment. Recognition of all pronunciation variants was then measured in a lexical decision experiment. Overall, the data suggest that listeners are sensitive to variant frequency: Word classification rates closely paralleled production frequency. The exceptions to this were [t] realizations (i.e., canonical pronunciations of the words), a finding which confirms other results in the literature and indicates that factors other than exposure frequency affect word recognition.In casual speech, talkers pronounce words in ways that deviate from their canonical pronunciations. For example, talkers of American English often flap intervocalic /t/s. For some words (e.g., pretty), flapping occurs with such frequency that the flapped variant (e.g., [prɪɾi]) is much more common than its citation form (e.g., [prɪti]). For communication to succeed, listeners must learn to recognize these alternative pronunciations of words. How does this occur?A partial answer to this question is that learning occurs through exposure. Listeners encode the variation they experience to the degree (frequency) they experience it, thereby tuning their perceptual system to the pronunciation variability found in the environment (e.g., Connine, 2004;Connine, Ranbom, & Patterson, 2008; Ernestus & Baayen, 2007;Mitterer & Ernestus, 2006). Although there may be other means of recognizing pronunciation variants (generalization of rules), the ever-growing literature on statistical language learning (Gomez, 2007;Saffran, 2003) and exemplar theoretic models of language perception and production (Bybee, 2001;Johnson, 2006;Pierrehumbert, 2003) suggest that an experiencebased account is both plausible and likely.Correspondence information: Mark A. Pitt, Department of Psychology, 1835 Neil Avenue, Ohio State University, Columbus, OH 43210-1222, office (614) 292-4193, fax (614) 688-3984, pitt.2@osu.edu, http://lpl.psy.ohio-state.edu. Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain. NIH P...

Online perception of glottalized coda stops in American English

“…In our study, we similarly did not find any facilitation of word recognition when listeners heard glottalized [ʔ͡ t] compared to non-glottalized [t]. These results are consistent with the robust finding that canonical forms are privileged, despite their actual production frequency across contexts (McLennan et al, 2003(McLennan et al, , 2005Ranbom & Connine, 2007;Pitt, 2009;Pitt et al, 2011;Ranbom et al, 2009;Tucker, 2011). It is an open question, however, how much of a perceptual mismatch glottalized voiced stops are.…”

“…Several researchers have examined listeners' ability to recognize familiar words produced with a glottal variant of /t/ both in word-medial (Pitt et al, 2011) as well as in word-final coda position (Sumner & Samuel, 2005). In particular, word-final glottal variants of /t/ (both [ʔ͡ t or ʔ]) were as effective as canonical non-glottalized [t] at priming a semantically-related target.…”

Pronunciation Variation in Infant-Directed Speech: Phonetic Reduction of Two Highly Frequent Words

“…First of all, there is evidence that adults store representations of reduced pronunciation variants in their mental lexicons (Bürki et al, 2010;Pitt et al, 2011;Ranbom and Connine, 2007). Given the high frequency of reduced pronunciation variants in our data, it could well be the case that these representations are acquired in infancy.…”

“…They showed that the relative frequencies of the pronunciation variants are correlated with response latencies: participants produced a variant faster the more frequent it is relative to the other variant. Finally, Pitt, Dilley and Tat (2011) investigated the correlation between the frequencies of different pronunciation variants of word-medial /t/ in American English ([t], [ʔ], [ɾ] and a deleted variant) and recognition in a lexical decision task. They found that the more commonly produced variants are also more easily recognized.…”

Perceptual adaptation to segmental and syllabic reductions in continuous spoken Dutch

“…Such episodic storage, however, does not make prelexical reconstruction superfluous. It has been repeatedly found that, compared to reduced forms, canonical forms have an advantage in speech comprehension (e.g., Ernestus, 2009;Pitt, Dilley, & Tat, 2011). Adapting to a given speaker's reduction style via prelexical reconstruction would allow listeners to access canonical forms, which would in turn facilitate recognition.…”