Stress in English: Four Levels or Three?

Extract

Now that the CJL/RCL has completed a quarter of a century of publication, it is interesting to look back through the early numbers and to note the liveliness of the contributions to the fledgling review of the newly-founded Association. Strong opinions were sometimes offered on matters of linguistic practice and principle; one of these in fact touched off a debate that appeared in the pages of the CJL/RCL in the late fifties. It concerns the interpretation of the English suprasegmentals, especially the number of stress distinctions to be found in English. It was a controversy that went on over a period of a year and a half, and was never resolved to complete satisfaction. On the one side there were the defenders of the American behaviourist position outlined by Trager and Smith (Robinson, Theall and Wevers 1959). This position—that there are four distinctive levels of stress in English—had been given quasi-official status by its reappearance in an introductory textbook (Hill 1958), and had obviously provoked extensive discussion at the 1958 Canadian Summer School of Linguistics in Alberta, out of which discussion came the original article (Drysdale, Mackey and Scargill 1958) declaring that whereas one could get all kinds of phonetic differences of stress, there were only three distinctive levels, not four. The then-editor of the CJL/RCL, Jean-Paul Vinay, took the side of Drysdale, Mackey and Scargill, and added his name to theirs.