Volume 63, Issue 4

Vowel normalization in speech perception was investigated in three experiments.The range of the second formant in a carrier phrase was manipulated and thisaffected the perception of a target vowel in a compensatory fashion: A low F2 rangein the carrier phrase made it more likely that the target vowel was perceived as afront vowel, that is, with a high F2. Recent experiments indicated that this effectmight be moderated by the lexical status of the constituents of the carrier phrase.Manipulation of the lexical status in the present experiments, however, did notaffect vowel normalization. In contrast, the range of vowels in the carrier phrase didinfluence vowel normalization. If the carrier phrase consisted of mid-to-high frontvowels only, vowel categories shifted only for mid-to-high front vowels. It is arguedthat these results are a challenge for episodic models of word recognition.

The two word accents in Stockholm Swedish (Accent I and Accent II) are distinguishedby a consistent falling pitch contour on the stressed syllable of AccentII words. The current study presents new types of evidence that this feature ofword accent can be systematically found in words produced by 16- to 18-montholdSwedish-speaking children. Compared to other disyllabic productions, thestressed syllable in Accent II words have a larger F0 decline, more sequences ofhigh-low turning points identified by a stylized contour algorithm and an earlier F0peak. In addition, a negative correlation is found between the value of F0 peak andF0 change in the stressed syllable. Taken together, these findings indicate that childrenlearning Swedish have internalized a subtle but lexically relevant pitch contourby a very early stage of word production.

The general goal of this study was to expand our understanding of what ismeant by ‘foreign accent’. More specifically, it deals with the role of prosody (timingand melody), which has rarely been examined. New technologies, includingdiphone speech synthesis (experiment 1) and speech manipulation (experiment2), are used to study the relative importance of prosody in what is perceived as aforeign accent. The methodology we propose, based on the prosody transplantationparadigm, can be applied to different languages or language varieties. Here,it is applied to Spanish and Italian. We built up a dozen sentences which are spokenin almost the same way in both languages (e.g. ha visto la casa del presidenteamericano ‘you/(s)he saw the American president’s house’). Spanish/Italian monolingualsand bilinguals were recorded. We then studied what is perceived whenthe segmental specification of an utterance is combined with suprasegmental featuresbelonging to a different language. Under these conditions, results obtainedwith Spanish and Italian listeners suggest that prosody is important in identifyingSpanish-accented Italian and Italian-accented Spanish.