The aim of this study is to examine the development of OE pre-modals (sculan, willan, magan, cunnan, motan) and to find whether there is a theory of change beyond a theory of grammar. When we view the history of the English modals we can find five early changes, affecting different aspects of the grammar at different times, and making the pre-modals exceptional verbs with the special properties. This was followed by a re-analysis best described as the introduction of a new Modal category, a reformulation of two transformations and the appearence of new lexical items have to, be able to, and be going to. Reasons for these five early changes cannot be provided, but they must be regarded provisionally as a function of chance, caused by extra-linguistic factors of some kind. However, the net result of these changes was a situation where categorial membership of must, may, will, etc., was opaque and the degree of opacity approached the limits described by the theory of grammar. Therefore, a therapeutic re-analysis was required, imposed by the theory of grammar. With the introduction of a category modal, further changes were required-as a matter of necessty, imposed this time by the theory of change. If there were no further changes, the existing rules of Neg-Placement and Subject-Verb Inversion would yield quite different surface patterns, of a kind to strain communicability across generations to breaking-point. We have shown that grammars are created anew by each successive generation and that changes may be radical and are to be interpreted in terms of the system which gives rise te them, and that we should not expect to find a ``diachronic universal`` or the theory of change independent of a theory of grammar.