ABSTRACT
Intonation has been studied at three levels, the physical, the perceptual and the functional. The physical level of study includes extraction of fundamental frequency and time-varying overall intensity level, registration of Fo(t) and possibly I(t) and description of the data in statistical terms. Classification of the data in terms of linguistic entities is highly desirable. At the physical level of description, there is little variation in the mode of representation. Fundamental frequency is recorded as a linear, logarithmic or hyperbolic function of time and the scale of the independent variable are almost invariably linear. There are basically two philosophical approaches to objective phonetic data; metaphysical and empirical. In matters intonational, as with other phonetic/phonological problems, it is possible to take either one or the other position. In the few works in which an attempt is made to relate experiments to a theory of intonation, the relation of the results of the experiments to the theory is obscure.
