Segment duration as a cue to word boundaries in spoken-word recognition

Abstract

In two eye-tracking experiments, we examined the degree to which listeners use acoustic cues to word boundaries. Dutch participants listened to ambiguous sentences in which stop-initial words (e.g., pot , j jar) were preceded by eens (once); the sentences could thus also refer to cluster-initial words (e.g., een spot , a spotlight). The participants made fewer fixations to target pictures (e.g., a jar) when the target and the preceding [s] were replaced by a recording of the cluster-initial word than when they were spliced from another token of the target-bearing sentence (Experiment 1). Although acoustic analyses revealed several differences between the two recordings, only [s] duration correlated with the particip pants' fixations (more target fixations for shorter [s]s). Thus, we found that listeners apparently do not use all available acoustic differences equally. In Experiment 2, the participants made more fixations to target pictures when the [s ] was shortened than when it was lengthened. Utterance interpretation can therefore be influenced by individual segment duration alone.

Possible words and fixed stress in the segmentation of Slovak speech

“…In support of the first approach, research has shown that the speech signal contains regularly occurring properties such as the aspiration of word-initial plosives or durational cues, which correlate with word boundaries (Lehiste, 1960;Nakatani & Dukes, 1977) and are exploited by listeners (e.g., Dumay, Content, & Frauenfelder, 1999;Gow & Gordon, 1995;Mattys & Melhorn, 2007;Quené, 1987Quené, , 1992Shatzman & McQueen, 2006;Spinelli, McQueen, & Cutler, 2003). But languages differ in how word boundaries are marked by physical cues, and hence the nature of these cues depends on language-specific phonology (Lehiste, 1964).…”

Speaking rate from proximal and distal contexts is used during word segmentation.

“…Recently, it has also been demonstrated that rate modulates perception of duration as a suprasegmental lexical-stress cue (Reinisch, Jesse, & McQueen, in press). The present series of experiments investigated speaking rate in relation to another type of durational cue: duration as fine phonetic detail used in word segmentation (e.g., Salverda, Dahan, & McQueen, 2003;Shatzman & McQueen, 2006). We ask whether linearly manipulated speaking rate influences the interpretation of the duration of juncture phonemes, and hence, where word boundaries are perceived in the segmentation of continuous speech.…”

“…For example in Dutch, "eens speer" ("once spear") and "eens peer" ("once pear") contain the same sequence of sounds but differ in whether the [s] belongs to both words or only the first one. Word boundaries are marked by a variety of acoustic cues that listeners can use (e.g., Cho, McQueen, & Cox, 2007;Mattys, White, & Melhorn, 2005;Nakatani & Dukes, 1977) and duration is an important one of them (e.g., Gow & Gordon, 1995;Klatt, 1976;Quené, 1992;Repp, Liberman, Eccardt, & Pesetsky, 1978;Salverda et al, 2003;Shatzman & McQueen, 2006;Spinelli, McQueen, & Cutler, 2003;Tabossi, Burani, & Scott, 1995). The longer a boundary sound is (e.g., the [s] in the example above), the more it supports a word-initial interpretation (i.e., "eens speer").…”

“…Moreover, because the use of durational cues to word segmentation has been shown to occur during the early phases of word recognition (Shatzman & McQueen, 2006), we also asked whether speaking rate information is also taken into account early in processing or whether it is used later to reinterpret durational cues. Studies on the perception of speaking rate so far mostly used offline tasks such as phoneme categorization or category goodness judgment.…”

Models of spoken‐word recognition

“…Moreover, the activation flow of candidate words over time in Fine-Tracker has been successfully linked to word activation in eye-tracking studies 33 that examined the use of durational cues in word recognition. 67,68 A shortcoming of real-speech models is that due to limitations of the speech conversion module, i.e., the imperfect conversion of the speech signal to prelexical representations, such models are currently only able to use a small subset of a language's vocabulary. 63,69 Obviously, if the speech conversion module fails, everything downstream will as well.…”