Morphological autonomy and diachrony

The primary aim of this study has been simply to show that autonomously morphological structure need not be an inert, defunct, residue of an earlier etat de langue, nor a kind of diachronic ‘dead end’.35 It can be a dynamic, pervasive, self-reinforcing factor in morphological change. If morphology, and in particular autonomous morphology, is a ‘disease’ of language, it must be an extremely benign one. Indeed, so innocuous is it that speakers can actually pass up golden opportunities to align allomorphs with morphosyntactic properties (cf. the generalization of the preterite 1sg. PYTA alternant, described in 3.3), in favour of the ‘morphomic’ distribution. I have also sought — albeit speculatively — to suggest that the autonomously morphological may permeate phenomena which, prima facie, seem to be motivated by universal principles of iconic alignment between form and meaning. I proposed that complete levelling out of allomorphy — a common cross-linguistic phenomenon — could just as easily be formulated in ‘morphomic’ as in extramorphological terms, and that there was some evidence from Romance to suggest that such a perspective could not be excluded a priori. I have further argued that an autonomously morphological signatum, namely the very fact of being a formative, may be present even in simple, linear, concatenations of formatives, and therefore potentially present not only in any language, but indeed even in formatives which might have a lexical meaning. But the least claim I want to make is that morphologists, and especially historical morphologists, should not regard the autonomously morphological as a stagnant backwater of linguistic structure.