Volume 34, Issue 3

This paper attempts a new departure both in German dialectology and in phonemic analysis: (i) It is based on an open corpus of spontaneous, colloquial speech. (ii) The segmentals (phonemes and clusters) are integrated into the hierarchy of supra-segmentals (syllable types, shape types, intonation contours). (iii) Phonemic variability is recognized within the dialect. Its major isoglosses are restated in terms of phoneme correlations. (iv) Certain unstressed words do occur in isolation. (v) Vowel length is recognized as redundant, but the difference between the dental /t/ and the alveolar /t/ is distinctive. (vi) The inventories of stressed and unstressed vowels differ radically, the unstressed suffixes containing the vowels /i a/ only. (vii) The peripheral status is analyzed of the aspirates (other than /kh/), of the affricates /tš kχ/, and of the lenis /t š n/, but none of these phonemes is limited to loanwords.(viii) The fusion is analyzed of adjacent stops in initial k- and {-clusters.

The phenomenon of emphasis is studied with regard to the phonetic variation of the vowel and consonant units of the word structure in Telugu, a Dravidian language spoken mainly in South India. The electro-aerometer and the mingograph are used to provide phonetic evidence. The study reveals that different means are used to emphasise words of different structures so that the phonological structure is not disturbed.

Four German children aged 4–9 learned English as their second language (L2) the natural way, i.e. without formal teaching. They acquired the local L2 /r/ – Which varied from central frictionless continuant to retroflex – in the same developmental sequence. The first non-zero substitute for the L2 /r/ was [w], then a frictionless continuant [ɹ] and, finally, a retroflex [r]. Additional evidence from other first-language (LI) and L2 studies suggests that, at least under non-school-room conditions, the L2 acquisition of the various types of r’s is not accomplished by relying on the prior LI /r/. This contrasts sharply with the acquisition of other types of phonetic-phonological elements like the L2 target vowels, plosives, and others, which are first replaced by LI elements.

Subjects listened to a speech perception test. It consisted of a series of items obtained by tape splicing from one recorded English utterance it’s still . The subjects assigned one of four possible English utterance choices to each item. Combined durations of the /s/ friction and closure of the following stop account for their choice between it’s dill , it still or it’s still . A delay in voice onset of the stop is primarily responsible for items consistently heard as it’s till . There is no difference in closure duration between the traditionally forlis and lenis stops following the /s/.