Paul Postal, 1968
During the 25 or so years from the early 1930s to the late 1950s, linguists developed a model of phonology in which the phonemic level was independent of syntactic structure. This led to the entire system being designated as Autonomous Phonemics. Postal makes the claim that during the late 1950s and early 1960s, a new phonological structure came into being in which the phonemic level depended on information from syntactic Surface Structure. Derived from the newly developing ideas of Generative Grammar , this new Systematic Phonemic level was somewhat more abstract the the older Autonomous Phonemic level and specifically depended on syntactic information.
Linguists were pretty well agreed that morphemes must be composed of some sort of segments (the Phonemes). In Autonomoous Phonemics, these phoneme segments were indivisible units with no internal structure. Systematic Phonemeics replaced these units with bundles of binary phonetic features. The features would be based on articulatory facts, but used for various higher level constructions, including stress assignment and inflectional effects.
One peculiar facet of all this is that the informational dependency was not in terms of any sort of mental or real-time computational issue, but was, rather, in terms of what information a linguist needed in order to form a complete and logical description of a language's phonological system. This was expressed in terms of whether a level of phonemic representation existed from which the linguist could tell whether two utterances in the language were tokens of the same or different words.
The real question was whether a Narrow Phonetic Transcription (NPT) could be taken as an absolute representation of the phonetic detail of a language, without reference to syntactic or any other details of the language. Referring to ideas from Saussure (1959), Postal labels such a depencency on other language factors as "mentalism".
Top of Page | APT Summary | Sort by Topic | Sort by Title | Sort by Author