In Kahn (1976), one finds a syllable-based reanalysis of certain well-known phonological rules of English. Prominent among these are aspiration and flapping. Given that the belated recognition of the syllable by generative phonologists has generally yielded very interesting and positive results, this reinterpretation of the facts must be seen as a welcome development in that respect.
On the other hand, however, Kahn’s decision to base his syllable-oriented generalizations on the concept of ambisyllabicity must be seriously questioned since the phonological evidence for such a phenomenon is so tenuous, and also because this concept is virtually unfalsifiable as it is not known to have any physical-acoustic or articulatory-correlates whatsoever.
Also, over and above Kahn’s recourse to such an unsubstantiated mechanism, there is the additional problem that the two rules he proposes simply do not correctly account for all the data. A later attempt by Kiparsky (1979) to replace Kahn’s ambisyllabic treatment of these two rules by a metrical analysis involving the foot runs into exactly the same kind of trouble.