On loops

Abstract

Velar consonants are known to often show forward movement of thetongue during occlusion, resulting in elliptical trajectories in VCV sequences. To improve understanding of the influences underlying this pattern, lingual movement was analyzed by varying vowel context and manner of articulation. Therefore, two German subjects were recorded by means of Electromagnetic Articulography. The first part of the material consisted of /bV 1gV 2/ sequences with all combinations of the tense stressed vowels [i,u,a] in the second part, the intervocalic consonant was /k,g,η,x/, the initial vowels [i, u, a], and the final vowel the low schwa [∀]. In vowel contexts exclusively involving back vowels the expected elliptical patterns were found; thus the tongue may well continue moving away from V 2 even after the end of consonantal closure. Contexts involving [i] showed an asymmetry. With V 1 = [i] elliptical movement was suppressed, with V 2 = [i] it was enhanced. Regarding manner of articulation, the amount of forward movement ordered similarly to the amount of tongue raising for the consonant ([k]>[g]>[η]>[x]). In parallel with the vowel context effects, this manner of articulation effect was suppressed when V, was a high front vowel. These results indicate firstly that, for German, forward movement of the tongue is not connected with enhancing voicing in voiced stops, and secondly that it can be no more than partially due to air pressure. The present results are compared with those obtained for velar consonants in further systematically varied phonetic contexts employed both by ourselves and others. The overall conclusion is that elliptical trajectories are the robust effect of several fairly weak factors acting in combination. The required ingredients for a complete model of articulator movement are discussed.