Maartje Schreuder, University of Groningen
For some decades phonologists strive for grounding their findings with empirical bases. The rise of Laboratory Phonology is the product of this ambition. Laboratory phonologists combine phonological hypotheses with phonetically testing. The rationale is that all things we can hear must be triggered by features in the acoustic signal. Several phonological phenomena are even only to be explained using acoustic information.
I will argue, however, that not everything we hear can be found in the sound signal. In psychology, psycholinguistics and even musicology many examples are given to show that the mind of the listener also plays an important role in perception. This psychological influence is widely overlooked in phonology.
I will illustrate the position that phonological perception is not only influenced by acoustic cues of the speech signal, but also by communicative strategies of the listener, with examples of my own experiments on rhythmic variability. In these experiments I first tried to explain the phonological processes of secondary stress shift by investigating the acoustic characteristics of the signal. It turned out, however, that none of the phonetic correlates of stress – neither duration, nor pitch, intensity, nor spectral balance - could identify the perceived stress shifts.
What did play a role was timing. I found that listeners focus on equally-spaced distances in time, intervals of about 300 msec. They seem to expect important information ‘on the beat’, and a secondary stress is perceived on the syllable which is nearest to that beat. I will therefore argue that a communicative strategy is involved, an ‘Internal Metronome’, helping the listener to anticipate on the most relevant parts of the spoken message. This implicates that the listener may hear what he expects to hear, even if it is not present in the acoustic speech signal.
