Is it accurate to claim that autosegmental phonology is a theory of suprasegmental prosodic phenomena?

Question

I'm a little unclear on what autosegmental phonology is vis-a-vis suprasegmentals. Although I understand the nuts-and-bolts of autosegmental phonology, it is not clear to me if it negates a division between segmentals and suprasegmentals. So far as I read, autosegmental phonological descriptions do not make reference to suprasegmentals...

Answer 1

Actually, negating the difference between segmental and suprasegmental is what autosegmental phonology is all about (see Goldsmith 1976 and similar publications of the era). The premise of "suprasegmental phonology" qua approach is that we have segments, and things above the segment. Autosegmental phonology does not take the segment to be the representational primary, it takes the feature to be primary (each feature is "it's own segment"), thus each feature has full autonomy in the way that tone classically is autonomous from the segment. In suprasegmental phonology there is limited separation of properties and the notion of "above" the segment makes sense, but in autosegmental phonology there is no clear up and down. On paper you have to represent things in an "up/down" fashion, but the logic is simply that every property is on a tier of its own and the ith element of the ath tier can associate with the jth element of the bth tier, and there is no special ordering of the tiers. "Property" is a more appropriate term, since some of the classical suprasegmental "things" are not considered to be features with valus, for example "syllable" is not a feature, nor is "mora" / "timing slot", likewise "stress"; "property" generalizes over "feature" and "structural node".

The "goal" of complete dissolution of representations into a gestural score is laid out in the last chapter: however, it wasn't long before people were uncomfortable with this no-segment approach, and McCarthy's (dissertation) introduction of "C" and "V" elements as morphological stabilizing elements with a special affinity for the major class features led to CV phonology, which effectively reintroduced the "segment" as a coherent unit of representation.