Not just a function of function words: Distal speech rate influences perception of prosodically weak syllables

Abstract

Listeners resolve ambiguities in speech perception using multiple sources, including non-local or distal speech rate (i.e., the speech rate of material surrounding a particular region). The ability to resolve ambiguities is particularly important for the perception of casual, everyday productions, which are often produced using phonetically reduced forms. Here, we examine whether the distal speech rate effect is specific to a lexical class of words and/or to particular lexical or phonological contexts. In Experiment 1, we examined whether distal speech rate influenced perception of phonologically similar content words differing in number of syllables (e.g., form/forum). In Experiment 2, we used both transcription and word-monitoring tasks to examine whether distal speech rate influenced perception of a reduced vowel, causing lexical reorganization (e.g., cease, see us). Distal speech rate influenced perception of lexical content in both experiments. This demonstrates that distal rate influences perception of a lexical class other than function words and affects perception in a variety of phonological and lexical contexts. These results support a view that distal speech rate is a pervasive source of information with far-reaching consequences for perception of lexical content and word segmentation.

Syllable as a unit of information transfer in linguistic communication: The entropy syllable parsing model

“…The rate of degraded speech (low-pass filtered or sinewave), or tone sequences surrounding syllables in function words did not alter perception. Similar effects have been found for perception of weak syllables in both function words and lexical items (cease versus see us) (Baese-Berk, Dilley, Henry, Vinke, & Banzina, 2019), in native and nonnative speakers, as well as languages with distinctly different morphosyntactic properties, such as Russian and Mandarin (Dilley, Morrill, & Banzina, 2013;Lai & Dilley, 2016).…”

Contextual speech rate influences morphosyntactic prediction and integration

“…If that were the case, the early perceptual cue should remain active in the system for as long as it is relevant for linguistic processing, and we may observe rate effects even after the onset of the disambiguating target information. This is especially interesting given that phoneme-level contextual rate effects have been claimed to be "fragile" (Baese-Berk et al, 2019). As such, our experiment offers novel insights into how the brain infers linguistic cues from the acoustic signal, and how these inferential cues might be combined with information from higher levels of linguistic hierarchy during online sentence comprehension.…”

“…In fact, reduced highly coarticulated linguistic units can even be missed entirely by listeners when presented in slow contexts. For instance, a reduced "terror" can be perceived as "tear", omitting the second unstressed syllable "-or", when embedded in a slow sentence (Baese-Berk, Dilley, Henry, Vinke, & Banzina, 2019). Similarly, the function word "or" in a phrase such as "leisure (or) time" can be perceived as present or absent depending on contextual speech rate (Dilley & Pitt, 2010), and the determiner "a" in a sentence such as "The Petersons are looking to buy (a) brown hen(s) soon" can perceptually "appear" or "disappear" when embedded in fast or slow contexts (Brown, Dilley, & Tanenhaus, 2012).…”

Spectral contrast effects are modulated by selective attention in “cocktail party” settings

“…This is in stark contrast with recent findings concerning a related contextual phenomenon involving temporal contrast effectsalso known as rate normalization (Bosker & Ghitza, 2018;Kaufeld, Ravenschlag, Meyer, Martin, & Bosker, in press;Maslowski, Meyer, and Bosker, 2019a, b). That is, duration cues on target speech sounds are perceived relative to the surrounding speech rate: a word with a reduced unstressed syllable (e.g., "-um" in "forum") is perceived as "shorter" ("form") in a slow context, but "longer" ("forum") in a fast context (Baese-Berk, Dilley, Henry, Vinke, & Banzina, 2018;Dilley & Pitt, 2010). However, when such ambiguous target words are perceived in the context of two carrier sentences, one slow and the other fast (like in Experiment 2), target categorization is not influenced by selective attention to one or the other sentence (Bosker, Sjerps, & Reinisch, 2018).…”