Perceiving Focus

Many linguists approach intonational matters from a purely speaker-oriented perspective¹. For instance, in different studies, in as far as these are empirical in nature, evidence for particular tonal distinctions is often solely based on acoustic analyses of fundamental frequency (F0) traces. However, if one wants to gain full insight into how intonation ‘functions’, such an approach is arguably incomplete. That is, a prosodic feature, as any other linguistic feature, can only be said to be communicatively relevant if it is not only encoded in the speech signal by a speaker, but if it also has an impact on how an utterance is processed by a listener. In other words, claims about important intonational categories and their respective meanings are somewhat premature if they are not backed up with results that show that these are also relevant at the receiving end of the communication chain. Ideally, such an analysis should be more than an individual linguist’s interpretation of a prosodic phenomenon.

In this chapter, we argue that controlled perceptual studies allow us to investigate the communicative importance of intonational features. Rather than concentrating on subtle differences between intonational categories, we will illustrate this viewpoint with a series of studies on the cue value of pitch accents. In languages such as Dutch and English, the distribution of accents has been claimed to be exploited as a means to distinguish important bits of information in an utterance from unimportant ones. That is, in such languages, pitch accents serve as a linguistic strategy to put ‘new’ or ‘contrastive’ information in focus, whereas speakers take care not to highlight ‘given’ information, i.e., information which is present (explicitly or implicitly) in the preceding context. However, it is unclear whether such observations generalize to other languages as well; in addition, most studies on pitch accent distribution have not looked at the relation of such accents, which are encoded in the speech signal itself, to visual cues speakers may send to a communication partner. Therefore, in order to gain insight into the relative cue value of pitch accents for signaling focus, we will tackle this question in different perception tests, and approach it from (1) a multilingual and (2) a multimodal perspective. In the studies presented below we shall ignore newness accents, for reasons that will become clear later, and hence in the present study being contrastive is equivalent to being in focus.