<i>r</i>, Hypercorrection, and the Elsewhere Condition

<jats:title>Abstract</jats:title> <jats:p>Our purpose in this article is to contribute to the current debate regarding the differing theoretical assumptions between Optimality Theory (OT) and the more traditional theory of derivations employing ordered rules. Such a discussion of fundamental issues is greatly facilitated if there is a body of data at hand that satisfies two desiderata: (1) the data are of a degree of complexity that makes it necessary to invoke more than just the most elementary resources of the competing theories, and (2) the phenomena are well-enough understood so that obviously inadequate solutions can be rejected without extensive investigation of the most basic facts. One body of data that satisfies these desiderata is the well-known shibboleth of the Eastern Massachusetts r. We are fortunate, moreover, in that these data have been reviewed in two recent papers by John McCarthy (1991; 1993), a native speaker of the dialect, who happens also to be one of the leading phonologists of our generation. In addition, much of the same data have been discussed in chapter V of Harris (1994), where parallel developments in other dialects of modern English are also considered.</jats:p>