Abstract
AbstractValency‐changing morphemes, in those languages that allow more than one in the same word, show ordering restrictions and effects of syntactic and semantic scope. In some languages, a given pair of valency‐changing morphemes (e.g. causative and passive) may be restricted to appearing in a fixed order. When both orders of a given pair of morphemes are attested, often the syntax and semantics associated with one of the orders are different from those associated with the other order. Three different approaches to explaining the constraints on the ordering of valency‐changing morphemes are considered. One line of research assumes that scope is not the primary force at play, but rather a template is, and that scope is stipulated as a constraint that interacts with the template. Another line of research assumes that words with valency‐changing morphemes are derived through syntactic movement and that scope effects (or the mirror principle) follow from the syntactic principles involved in the derivation. The third line of research discussed here assumes that all words, including those with valency‐changing affixes, are formed in the lexicon and that scope effects are derived from the idea that, within a layered morphology, stems are formed through the pairwise combination of a stem with an affix and every such operation implies simultaneously a concatenation of morphs and a composition of syntactic and semantic information. Problems with and advantages of each approach are presented.
Reference42 articles.
1. On the Argument Structure of Causatives;Alsina, Alex;Linguistic Inquiry,1992
2. 1993 Stanford University Stanford California Alex Alsina Predicate Composition: A Theory of Syntactic Function Alternations
3. Where’s the Mirror Principle?
4. On the nonsemantic nature of argument structure