Introduction

This chapter aims to describe those aspects of Latin inflectional morphology which remain substantially intact from Latin into Romance, focusing especially on cases where the relation between grammatical or lexical meaning, on the one hand, and morphological form, on the other, is arbitrary and idiosyncratic. Much of Latin inflectional morphology is of the ‘fusional’ type, characterized by allomorphy (more than one form corresponds to one meaning), cumulativeness (one form simultaneously expresses more than one morphosyntactic property) and, sometimes, ‘emptiness’ (there are formatives to which no grammatical meaning can be independently ascribed). Studies of Romance historical morphology usually highlight what has changed, assuming tacitly or explicitly that the change is motivated, at least in part, by preference for formally simpler, more ‘transparent’, form– meaning relationships. It should be obvious, however, that the ancient Romans were no better endowed to cope with morphological complexity than any subsequent generation of native speakers, and a priori there is no reason why morphology should get simpler. Overall, it does not. Some of the most eye-catching changes in inflectional morphology, such as the complete loss of the future imperfective inflections from Latin (see below), or the disappearance of the passive inflectional endings in favour of auxiliary + past participle constructions, probably have more to do with the existence of alternative structures, than with a move towards ‘simplicity’ (see Herman 2000a:71–74; also 59, 68f.). Indeed, the Latin imperfective passive inflections were (with the exception of second person endings -ris and -mini) characterized by an extremely transparent ending -(u)r, yet such structural transparency did not impede their complete disappearance.