A Postulate for Linguistics in the Sixties

Abstract

The approach to language study represented by postulates and corollaries was once the basis of Bloomfieldian linguistics, yet it is now somewhat out of fashion. Sets of postulates have a tendency to grow in number and unwieldiness, and transformational investigators favor a more operational analysis and statement. Yet transformational study itself rests upon a postulate which is at present implicit. Stated without setting it into the Euclidian language and style favored by Bloomfield, it is that the relations between utterances are capable of being stated in the form of manipulations.l Such an approach to utterances forces us to assume that all utterances contain items in combination, and that all items can appear in other utterances, which are then new combinations. Thus language is described as having three constituents: items, combinations, and relationships operationally statable between combinations. I believe that this postulate of possibly fictitious manipulability is fundamental both to transformational and to structural analysis. It is quite obvious that the formulae for 'deriving' a passive sentence from an active one could not be stated as a manipulation unless the active sentence contained isolable items capable of being manipulated at least in the form of rules on paper. It is equally obvious that when a structuralist sets up such units as morphemes and phonemes, and describes their sequences in morphotactics and phonotactics, he is setting up items, combinations of items, and relationships capable of being described as manipulations. Further, it seems to me that the statement I have just given is quite properly formulated as a postulate or axiom, since I do not see how it is capable of proof. At best, it is an empirical statement about observed languages. Since not all languages have been, or will be, observed, there is always the possibility that some language does not contain items in manipulable arrangements. Yet since the postulate is necessary for more than one kind of analysis, these analyses themselves constitute good reasons for acceptance. The fact that it underlies more than one kind of analysis is merely further strengthening. Since the postulate does not seem to have been explicitly stated, it is not surprising that its implications have not been fully explored. A beginning at such an exploration is the purpose of this paper. The first implication is that any utterance is separable into discrete units, since only chunks and pieces can be arranged and manipulated. Continua can be blended and stirred, perhaps, but