ProPro 2017: Abstracts

You find the invited speakers' abstracts on this page. The participants' abstracts will be published in our workshop booklet.

What is special about prosody in processing?

rosodic processing differs across languages in some very important ways. For instance, the way listeners segment continuous speech signals varies as a function of the prosodic feature rhythm. Language learners are sensitive to this feature very early – even before birth, in fact. Moreover, the function of segmenting speech is a vital part of language development, and evidence that the ability to segment speech has been achieved is a significant predictor of language skills later in childhood. Thus prosodic processing (and language-specific prosodic processing at that) is one of the most crucial steps in our early life. Though prosodic processing continues to be important in language use, and is language-specific in interesting ways (at the word level, the sentence level and potentially more), there is surprisingly little targeted cross-language research in this area (though research on prosody itself, and on within-language prosodic processing, both seem to be growing). The most likely underlying reason is the relative lack of awareness of prosody in general, and it is time for us to consider whether there is something that can be done to rectify this, in particular by making prosodic knowledge more accessible both to language scientists of any kind, and to the general public.

Listeners interpret rising and falling intonation prior to the final boundary

It is still not well understood how full intonational contours are processed and integrated during online sentence comprehension. Though there is a body of work investigating pitch accents specifically, comparatively less work has been dedicated to the processing of other parts of the contour, like boundary tones. Boundary tones are hypothesized however to be crucial in signaling various kinds of higher-level meanings, such as distinguishing questions from statements. We investigated how listeners process such contours in the context of a “targeted language game.” Participants played a card game where on critical trials the speaker produced an utterance that was structurally ambiguous (e.g. “Got an armadillo”) but was distinguished as a question or a statement by virtue of the intonational contour. Using eyetracking, we asked when in the contour do listeners interpret the utterance as a question or statement: it is only after they have heard the full contour (i.e. post boundary tone) or do they rely on earlier cues to distinguish the contours? Though an initial experiment suggested listeners need to wait until hearing the full contour, later experiments with more tightly controlled stimuli suggested that listeners will use acoustic information prior to the end of the boundary tone.

Investigating development of linguistic and non-linguistic prosody

In this seminar, I will talk about how to deal with the interaction between linguistic prosody and affect prosody, which has been traditionally considered as a non-linguistic component of communication. I will discuss how investigations of linguistic prosody and those of non-linguistic prosody have been separated theoretically and methodologically, and will propose that we should consider them simultaneously for understanding face-to-face daily oral communication. While the myth that emotion recognition is fundamental and easy and should precede the acquisition of language-specific prosody seems to have led the separation of research fields, affect processing and linguistic processing are equally complex and impact each other during communication. Potential experimental questions, tasks, measures and the associated challenges for future studies will be discussed.

Testing for processing advantages of linguistic effects of focus in L1 and L2

Listeners exploit prior sematic context to determine focussed information but not prosodic context for the same purpose. A phoneme-detection task (Experiment 1) shows this finding in Sepedi, a language with no grammmaticalized prosodic expression of focus. Sepedi listeners detected phoneme targets more quickly when the phoneme-bearing words were focussed (than unfocussed) but not when these words were emphatic (vs non-emphatic). The lack of interaction between the two effects suggests that focus and prosodic emphasis do not share the same function (searching for focus) in Sepedi. Experiment 2 tested cross-linguistic differences on the processing of Sepedi learners of English (a language with grammaticalized focus-to-accent mapping). Non-natives detected phoneme-bearing words more quickly in focussed condition and in accented condition. Like in Sepedi, no interaction was found. The study suggests that non-natives acquire the L2 prosodic structure (even if it is very different from their L1) while they remain unaware of its underlying discourse representation.