Backwards signing and ASL syllable structure

A recurrent question raised by the study of signed languages concerns the linguistic effect of the modality in which the language is produced. Is the modality difference between speech and sign reflected merely in the nature of the phonetic features that map into production and perception, or is it the case that there might be higher level organizational differences between the two linguistic modalities? The present study addresses the nature of the modality effect inside the syllable, namely whether syllables in ASL display evidence of segmental composition. Data from backwards signing are presented to demonstrate that the phonological representations that must be available to signers when they perform backwards signing tasks cannot be adequately represented with the current models that posit segmental composition of ASL syllables. Instead, it is argued that it is sufficient to make reference to distinctive features, in syllable initial and syllable final positions, and that there is no support for any further internal segmental divisions.