School of International Liberal StudiesWaseda University

The speech of learners of a non-native language generally shows features that result from an imposition of the rules and structures of their native language or languages.

In the phonetic domain, this imposition can affect the pronunciation of suprasegmental features, such as rhythm, intonation, or speech rate; or segmental features, that is, the way individual segments are pronounced.

While traditionally suprasegmental deviations have been widely studied, the segmental aspect of foreign accent has received more attention in the last years, partly due to the advances on signal manipulation. In this part, I will focus on the effects of segmental foreign accent in the interaction of English, Japanese and Spanish. Particularly, in my research I employ computational techniques that are aimed to gain control over the degree of foreign accent with which a word is pronounced in a foreign language.

Some of the conclusions that will be presented are in accordance with part of the literature, that demonstrates that a foreign accent does not necessarily implies a loss in intelligibility that could distort the message intended by the speaker.

The novelty of this type of research is the finding that there are segments that require more deviation from what native listeners understand as native pronunciation to be perceived as non-native, implying that the phonological representations of segments play a crucial role in the decoding of speech uttered by non-native speakers.