Department of Linguistics - Simon Fraser University

Colloquium Series

 

 

Gunnar Ólafur Hansson
 Linguistics, The University of British Columbia

 

 

 

Non-local assimilation and its diachronic-evolutionary sources 

 

March 3, 2005

11:30am, EDB 9511

 

 

Abstract

Long-distance assimilatory phenomena--most notably consonant harmony (consonant "agreement"), but also vowel harmony systems with one or more transparent vowels--are puzzling because they appear to violate what is otherwise a very general property of assimilation: that it only operates in a strictly local fashion, where the trigger and target are adjacent. Although attempts have been made at explaining away some apparent cases of non-local assimilation as a mere illusion (in that the intervening so-called transparent segments are in fact phonetically affected after all), a significant residue still remains which cannot be reinterpreted in this way. Recent work has argued that these genuine cases of "action ą distance", i.e. non-local assimilation, have their roots in the domain of speech planning (phonological encoding for speech encoding.
Nevertheless, much is still unclear about the precise nature of the link between psycholinguistic planning factors and the phonological sound patterns which appear to be shaped by those factors.


In this talk I will focus on two particularly rare types of consonant harmony: voicing agreement and nasal agreement. Not only are such cases rare, but some of them have properties that are otherwise anomalous within the general typology of consonant harmony systems. A more detailed look at the individual attested cases reveals that their diachronic origins are in fact quite diverse, and that analogical changes of various kinds (reanalysis, "rule inversion", extension, etc.) are often involved in their historical development. As a result, many such phenomena turn out to have origins that have absolutely nothing to do with the speech-planning factors that are argued to underly consonant harmony in general. An advantage of this result is that some of the anomalous properties of the systems in question receive straightforward explanations, and the general typology of "true" consonant harmony phenomena becomes somewhat tidier.