Introduction

In the preceding chapter, I identified two basic approaches to contrastive specification in phonology, and discussed their status from a logical point of view. I argued that pairwise comparison based on full specifications has severe logical problems, whereas contrastive specification by feature ordering is logically sound. I showed that Martinet's (1960) analysis of Standard French bilabial stops is an example of the former approach, and that Jakobson and Lotz's (1949) analysis of the same phonemes exemplifies the latter approach.

In this chapter I will review some work in structuralist phonology that bears on this issue. This chapter has three main aims. First, I wish to show the extent to which contrastive specification was central to the project of phonological theory in its formative years, roughly the period 1925–50. Though the authors I will survey are central figures in the field whose work has been widely read, I believe that this work has been misconstrued in various ways. There are several reasons for this. First, the authors sometimes use terminology that is unfamiliar to contemporary readers. Second, they are not always very explicit about how their theory is supposed to work, and in some cases, their writings do not add up to a consistent theory. Third, we tend to read these works through the prism of our own preoccupations.

The second aim of this chapter is to investigate in detail the narrower question of how these phonologists determined which features of a phoneme are contrastive in a given language; that is, how they arrived at contrastive specifications.