The effect of focus prominence on phrasing

Abstract

Prosody simultaneously encodes different kinds of information about an utterance, including the type of speech act (which, in English, often affects the choice of intonational tune), the syntactic constituent structure (which mainly affects prosodic phrasing), and the location of semantic focus (which mainly affects the relative prosodic prominence between words). The syntactic and semantic functional dimensions (speech act, constituency, focus) are orthogonal to each other, but to which extent their prosodic correlates are remains controversial. This paper reports on a production experiment that crosses these three dimensions to look for interactions, concentrating on interactions between focus prominence and phrasing. The results provide evidence that interactions are more limited than many current theories of sentence prosody would predict, and support a theory that keeps different prosodic dimensions representationally separate.

Two-dimensional parsing of the acoustic stream explains the Iambic–Trochaic Law.

“…This means that listeners can interpret whether an increase in one cue encodes prominence or grouping by interpreting it relative to changes in the other cue. The relationship between the cues and its potential role in perception has not been previously noticed when looking at word segmentation, although a similar relation between the cues was observed for the encoding of prosodic phrasing in a production study reported in Wagner and McAuliffe (2017, 2019). A rational listener can make use of the cue distribution observed in production when deciding about the grouping and prominence structure of an utterance, just as the cue distribution for a phonemic contrast is directly predictive of listener’s response in a categorization task (Clayards et al, 2008).…”

“…In two perception studies, Streeter (1978) found that intensity was a less important cue for grouping compared to other acoustic cues. Production evidence in Wagner and McAuliffe (2017, 2019), however, found that words at the beginning of a phrase have a significantly higher intensity than words later in a phrase, suggesting that intensity could in principle be an important cue for phrasing. While intensity generally drops throughout an utterance (see Pierrehumbert, 1979 for the observation, and perceptual evidence that listeners compensate for an expected downdrift in intensity), this utterance-level downdrift alone cannot explain this effect: The results in Wagner and McAuliffe (2017, 2019) showed that loudness partially resets when a new prosodic phrase begins (see also Poschmann & Wagner, 2016).…”

Linguistic syncopation: Meter-syntax alignment affects sentence comprehension and sensorimotor synchronization

“…A limitation of our design is that we assumed that words aligning with amplitude accented tones would be perceived as metrically strong. Metrical strength does not necessarily correspond to acoustic intensity in either music (Lerdahl & Jackendoff, 1983) or language (Wagner & McAuliffe, 2019). We think it unlikely, but it is possible that some participants were perceiving metrical accent in different positions than assumed in our experiment, driven by their own intuitions for implicit prosody (Fodor, 2002), thus contributing additional variability to our results.…”

“…1.3 Evidence for a link A few key points of evidence help to motivate this hypothesis. First, syntactic structure cues the perception of phrasal stress over and above acoustic cues (Bishop et al, 2020;Cole et al, 2017Cole et al, , 2019Kentner & Vasishth, 2016;Wagner & McAuliffe, 2019). This underscores not only the cognitively constructed nature of metrical perception-it is not acoustic-but crucially suggests that people actively seek out meter-syntax alignment.…”

“…Exceptions (Bolinger, 1972;Ladd, 2008) drive ongoing theoretical debate (Selkirk, 2011;Wagner, 2010;Zubizarreta, 2014). For example, there are some open questions about whether prominence serving an information focus function (Breen et al, 2010) is represented metrically and whether the cues for this interact with those serving a more structural function (Kentner & Vasishth, 2016;Wagner & McAuliffe, 2019). Despite these nuances, the different proposals all acknowledge a systematic relation of meter to syntax in some form.…”