Mietta Lennes

Speech-sound discrimination in neonates as measured with MEG

Magnetic brain responses to speech sounds were measured in 10 healthy neonates. The stimulation consisted of a frequent vowel sound [a:] with a steady pitch contour, which was occasionally replaced by the vowel [i:] with a steady pitch, or the vowel [a:] with a rising pitch, manifesting a change of intonation. The magnetic mismatch-negativity response (MMNm) was obtained and successfully modelled to the speech sound quality change in all infants and to the intonation change in 6 infants. The present results indicate that auditory-cortex speech-sound discrimination may well be studied with magnetic recordings as early as in newborn infants.

Is the phonetic quality of unaccented words unpredictable? An example from spontaneous Finnish

Mietta LennesNina AlarotuMartti VainioDepartment of Phonetics, University of Helsinkimietta.lennes@helsinki.finina.alarotu@helsinki.fimartti.vainio@helsinki.fiIn spontaneous speech, occurrences of the same word form vary phonetically.Words that do not receive sentence accent tend to be less clearly articulated andshorter in duration in comparison to prominent words. If such an unaccented wordis isolated from continuous speech, it may not even be identifiable by listeners, butas part of the speech stream, it is usually easily identified and helps the listener toperceive and organize the spoken message as a whole. Thus, the variability ofunaccented words is not random but must follow certain patterns determined bycontext. In this study, we take a closer look at the Finnish unaccented word-likeconstruction niinku in spontaneous conversational speech samples collected fromone female Finnish speaker. We provide an analysis of the phonetic variability ofthe niinku structure found in the data, including such properties as segmentalduration, segmental make-up, and phonetic quality. We then show what kind ofproperties are common to all occurrences of this unaccented construction anddiscuss the conditions that probably govern its acoustic-phonetic quality.