How lexical semantics constrains inflectional allomorphy

Grammatically conditioned allomorphy, paradigmatic structure, and the Ancestry Constraint

“…What I say here is based on evidence from only three languages (Latin, Itelmen and Hungarian), so is eminently vulnerable to disconfirmation. On the other hand, it gains some plausibility from the fact that it fits in with already-published suggestions about connections between the learning of inflectional morphology and of ordinary lexical items (Carstairs-McCarthy 1994, 1998.…”

“…What we do not observe is a third conceivable category: affixes that are neither screeveshape-identifiers nor screeveshape-defaults, and which are therefore`blurred' in the terminology of Carstairs (1994). I have argued elsewhere (Carstairs-McCarthy 1998) that this striking gap supports extending the notion of`blur avoidance' outside the 236 transactions of the philological society 99, 2001…”

“…The alternative that I will propose in section 5 exploits the kind of binary branching tree structure over which morphosyntactic information is distributed in Bobaljik's approach, and does so in a way that explains Carstairs's hypothesis that outward sensitivity involves whole categories, in the Matthewsian sense, rather than individual properties. This explanation exploits the notions`screeve' and`screeveshape' that have proved useful elsewhere in relating the organisation of inflection class systems to the general issue of inflectional content (Aronson 1982, Carstairs-McCarthy 1998. The resulting proposal certainly needs much more empirical testing; but, from a theoretical point of view, it seems to offer a way of combining insights from rather different morphological perspectives.…”

The No Blur Principle meets Norwegian dialects*

“…If we should find cases of more than one class‐default affix within the same tense (for example), they would constitute an empirical embarrassment for this idea of analogy between grammatical and lexical items. The reason is that these suffixes would then both (alternatively all) have the meaning ‘ present tense ’ and nothing more, and in order to distinguish them semantically, we would have to resort to disjunction, which is outlawed in ordinary lexical items (Carstairs‐McCarthy 1998). Carstairs‐McCarthy terms this blurring , and claims that it would be analogous to having a word with genuinely disjunctive meaning.…”

“…This article deals with morphological theory and Norwegian diachrony and dialects. The central theoretical claim is that the No Blur Principle (henceforth NBP) (Carstairs‐McCarthy 1994, 1998, 2001, 2005, Cameron‐Faulkner & Carstairs‐McCarthy 2000) is a Good Thing for diachronic morphology in general and for Natural Morphology (henceforth NM) in particular. It has been suggested elsewhere that the NBP and NM are useful in the study of Norwegian genders and declensions (Enger 2004, 2005); this article focuses on conjugations.…”