Retour vers liste

Détail de la contribution

Auteur: Gregory STUMP

Titre:
Polyfunctionality and inflectional economy


Abstract/Résumé: In many languages, the same morphology is used for two or more functions. I argue that such polyfunctionality is a manifestation of the complex way in which two distinct kinds of paradigms interface in the definition of a language’s grammar. A lexeme L has a content paradigm, whose purpose is syntactic: its cells specify the morphosyntactic property sets (MPSs) with which L may be associated in syntax. A stem X has a form paradigm, whose purpose is morphological: its cells specify the MPSs whose association with X is available for realization by rules of inflectional morphology. Often, the two sorts of paradigm are isomorphic: where lexeme L has stem X and s is a MPS appropriate to L, the s-cell in L’s content paradigm (a content cell) is linked to the s-cell in X’s form paradigm (its form correspondent) in such a way that the two cells share their realization. Although the two sorts of paradigm are often isomorphic, many morphological phenomena make it necessary to distinguish them: in instances of defectiveness, a content cell lacks a form correspondent; in instances of deponency, a content cell and its form correspondent are associated with distinct MPSs; in instances of suppletion, cells in the same content paradigm have their form correspondents in the form paradigms of different stems; and in instances of syncretism, two or more content cells have the same form correspondent. Like these phenomena, polyfunctionality involves a complex correspondence between content paradigms and form paradigms: if two content cells have the contrasting properties p1 and p2 and the distinction between these properties is neutralized as p in the corresponding form cells, then the morphology realizing p is polyfunctional, expressing p1 in some cases and p2 in others. Thus, in Khanty, properties of (a) subject agreement and (b) object number in the content paradigms of transitive verbs are distinguished from properties of (c) possessor agreement and (d) possessum number in the content paradigms of nouns; but in the corresponding form paradigms, the (a)/(c) distinction and the (b)/(d) distinction are both neutralized. As a consequence, the form correspondents of verbs specified for properties of types (a) and (b) and those of nouns specified for properties of types (c) and (d) are realized by many of the same morphological patterns. I contend that the synchronic architecture of grammar actually facilitates the incidence of polyfunctional inflection. The use of a single kind of form paradigm in realizing more than one kind of content paradigm simplifies the definition of a language’s inflectional morphology and enhances both the predictiveness of known word forms and the predictability of unencountered forms.