Native ‘um’s elicit prediction of low-frequency referents, but non-native ‘um’s do not

Abstract

a b s t r a c tSpeech comprehension involves extensive use of prediction. Linguistic prediction may be guided by the semantics or syntax, but also by the performance characteristics of the speech signal, such as disfluency. Previous studies have shown that listeners, when presented with the filler uh, exhibit a disfluency bias for discourse-new or unknown referents, drawing inferences about the source of the disfluency. The goal of the present study is to study the contrast between native and non-native disfluencies in speech comprehension. Experiment 1 presented listeners with pictures of high-frequency (e.g., a hand) and lowfrequency objects (e.g., a sewing machine) and with fluent and disfluent instructions. Listeners were found to anticipate reference to low-frequency objects when encountering disfluency, thus attributing disfluency to speaker trouble in lexical retrieval. Experiment 2 showed that, when participants listened to disfluent non-native speech, no anticipation of low-frequency referents was observed. We conclude that listeners can adapt their predictive strategies to the (non-native) speaker at hand, extending our understanding of the role of speaker identity in speech comprehension.

Phonetic variation in bilingual speech: A lens for studying the production–comprehension link

“…The finding that listeners responded to the slightly slowed speech rate preceding cued codeswitches in some respects resembles findings that monolingual listeners are sensitive to naturalistic hesitation phenomena (Bosker, Quené, Sanders, & de Jong, 2014; Corley, MacGregor, & Donaldson, 2007) 8 . We think it quite reasonable that habitual codeswitchers could develop associations between hesitations and codeswitching much in the same way that monolinguals learn to associate hesitations with new information, for example (Arnold, Tanenhaus, Altmann, & Fagnano, 2004).…”

Variation and Change in the Use of Hesitation Markers in Germanic Languages

“…For example, Arnold et al (2013) found that when encountering disfluent speech, listeners were more likely to expect a discourse-new referent. In line with this, Bosker et al (2014) showed that listeners were more likely to expect a low frequency as opposed to a high frequency word after a disfluency marker, though listeners adapted this expectation on the basis of the speaker the UM/UH ratio in American English based on the relatively recent Speed Dating Corpus (SDC; Jurafsky et al, 2009) and the older Switchboard corpus (Godfrey and Holliman, 1993) and obtained similar results, with women showing a greater UM/UH ratio than men in both corpora. Based on the Switchboard corpus, Acton (2011) also showed that this pattern persisted at the dialect-region level and when the gender of the hearer was taken into account (i.e.…”

Counting ‘uhm’s: How tracking the distribution of native and non-native disfluencies influences online language comprehension

“…Other studies have since shown the same effect in children as young as 2 years of age (Kidd, White, & Aslin, 2011;Orena & White, 2015;Owens & Graham, 2016;Owens, Thacker, & Graham, 2018;Thacker, Chambers, & Graham, 2018a, 2018b. Adult listeners have also been shown to be able to predict other types of complex referents, such as discourse-new (Arnold, Fagnano, & Tanenhaus, 2003;Arnold, Tanenhaus, Altmann, & Fagnano, 2004;Barr & Seyfeddinipur, 2010), compound (Watanabe, Hirose, Den, & Minematsu, 2008), and low-frequency referents (Bosker, Quené, Sanders, & De Jong, 2014a) upon hearing a disfluent filler uh.…”

“…However, none of the eye-tracking studies cited in this paper report any order effect (neither positive nor negative; cf. Corley & Hartsuiker, 2011 for an overall order effect in reaction time data, but no interaction between order and delay), except for Bosker et al (2014a) and Thacker, Chambers, and Graham (2018b).…”

“…Listeners' so-called disfluency bias for relatively more complex referents has been attributed (at least in part) to prior exposure to the non-arbitrary distribution of disfluencies in native speech (Bosker et al, 2014a). Given that disfluencies tend to occur more often before complex lexical items than before relatively simple words (Arnold & Tanenhaus, 2011;De Jong, 2016;Hartsuiker & Notebaert, 2010), listeners can cleverly use disfluencies as a predictive cue in comprehension.…”