This paper offers the initial design for a functional theory of morphology and addresses the question of how much unification or separation of domains and units is required in such a theory. Given that morphology is largely driven by principles that also rule syntax and semantics, the main thrust of this proposal is that the interaction between morphology and other areas of the theory can be adequately accounted for by the combination of three descriptive-explanatory resources: layered structures, templates and constructions. Other descriptiveexplanatory principles such as feature percolation and the characteristics of bases and adjuncts remain specifically morphological. Unification and separation in morphology must be carried out on functional grounds. Moreover, Old English evidence shows that no strict separation can be postulated, either between morphology and the rest of the theory or among the different morphological processes.

No abstract

This article carries out an analysis of the Old English verbal prefix ge-that is based on a paradigmatic organization of the lexicon, which allows for the inclusion of synchronic and diachronic data into a panchronic approach. On the synchronic side, the focus is on the spread of ge-in the derivational morphology of the language and the assessment of productivity, while on the diachronic side the emphasis is on the lexicalization and grammaticalization of the affix at stake. A path of grammaticalization is described as starting from the Proto-Germanic derivational prefix *ga-and leading to inflectional gein Old English. Quantitative evidence is gathered in favour of a progressive loss of the derivational productivity of the prefix that is paralleled by a gradual increase in inflectional distribution. The conclusion is reached that the neuter noun which is zero derived from the strong verb contributes through its derivatives to the decategorialization of the prefix ge-, which undergoes grammaticalization for reasons of over-markedness, becomes exclusively inflectional and, ultimately, disappears.

This article is a contribution to the development of the theory of morphology of Role and Reference Grammar in line with the more central role that constructions play in the latest version of this linguistic theory. In order to propose a principled definition of morphological processes that is not dependent on fuzzy notions like class membership, this article applies the distinction between constructions and constructional schemas to morphology in the following way. As regards morphological constructions, which are typologically relevant, the defining criterion is the distribution of markedness, that is, whether the morphologically relevant features are in the Nucleus or not. Two possibilities arise in this respect: the nuclear element or one or more non-nuclear elements are marked. If the nuclear element is marked, there is projection of morphological features. If the non-nuclear elements are marked, there is percolation as well as projection of features. Regarding constructional schemas, which are language-specific instantiations of constructions and can combine with one another, they fall into the following types: recursive/non-recursive, analytic/synthetic and continuous/discontinuous. After an application to the language of analysis, Pitjantjatjara/Yankunytjatjara, the conclusion is drawn that derivation (including compounding and affixation) can be endocentric or exocentric, whereas inflection is endocentric. Inflection and derivation can be analytic and synthetic as well as continuous and discontinuous. Unlike derivation, inflection is typically non-recursive, but the evidence provided by double case in Pitjantjatjara/Yankunytjatjara is relevant for this question.

No abstract