Phonotactic Illegality and Probability in Speech Perception
Evidence from second language listeners
This book presents empirical evidence on the influence of phonotactics on
speech perception, as well as a theoretical inventory and discussion. There are
many different viewpoints on phonotactic knowledge: it can be described ca -
tegorically, as the theory of legality of sound combinations, or gradiently, as
a theory that explains the probability of sound combinations. This thesis does
not start from linguistic considerations about phonotactic theory, but from
evidence on the application of phonotactic knowledge in speech perception.
Phonotactic knowledge allows listeners to resolve ambiguities, filter out incor -
rect structures and segment speech into words. Hence, the acoustic input is
transformed to be in accordance with phonotactics. By contrasting the pre -
dictions of different theories of speech recognition and of phonotactics, the
experiments reported in this thesis give insight in the psychological represen -
tations of phonotactics, as well as the process of applying this knowledge to
the perception of speech. The experiments include second language listening,
as the application of phonotactics from one language to another language
reveals the independent contribution of phonotactics. The last chapter of this
book includes a discussion of the evidence, leading to the proposal to include
a phonotactic preparser in models of word recognition. This preparser has the
function of ordering all phonotactically possible interpretations of the acoustic
input and removing the impossible percepts. The benefit of the preparser is
that it makes it unnecessary to look up improperly interpreted speech in the
lexicon. The preparser can explain the experimental evidence on phonotactic
effects in speech perception.