Abstract: "Place Assimilation"
by Jongho Jun
In Bruce Hayes, Robert Kirchner, and Donca Steriade, eds., Phonetically Based Phonology, Cambridge University Press (2004).
Attested patterns of place assimilation in consonant clusters display variability in targets, triggers, and domains. However, the crosslinguistic generalizations about place assimilation in (1) (Mohanan 1993, Ohala 1990) show that certain acoustic constraints govern the range of variability: more likely targets are acoustically weaker (less salient) than less likely ones, whereas more likely triggers are acoustically stronger (more salient) than less likely ones.
(1) a. Stops are more likely targets than fricatives.
b. Nasals are more likely targets than stops.
c. Coronals are more likely targets than non-coronals.
d. Non-coronals are more likely triggers than coronals.
e. Stops are more likely triggers than nasals.
f. Codas are more likely targets than onsets.
For instance, regarding (1a,b), it has been claimed that fricatives in general have stronger place cues than stops (Kohler 1991), and in turn, stops have stronger cues than nasals (Ohala 1990).
The purpose of the present study is to provide an explicit formal analysis of such variable, but constrained, patterns of place assimilation. Based on in-depth discussion of perceptual and articulatory mechanisms in speech production, we offer a formalization of phonetically based accounts of place assimilation within the framework of Optimality Theory (Prince & Smolensky 1993; McCarthy & Prince 1993, 1995). In Optimality Theory, the phonology is composed of sets of ranked violable constraints. Individual grammars are formed by ranking constraints in a particular way. If we further follow an age-long claim in phonetics that speech production is the result of reconciling two conflicting demands (ease of articulation and ease of perception), place assimilation may be viewed as the result of reconciling constraints guided by these two conflicting demands (cf. Mohanan 1993). A constraint motivated by ease of articulation (2) has the effect of reducing or eliminating any consonantal gesture, which may lead to place assimilation in consonant clusters.
(2) Weakening Constraint: Conserve articulatory effort
The weakening constraint may be outranked by perceptually-based Preservation constraints motivated by ease of perception. Preservation constraints have the effect of preserving consonantal gestures; thus gestural reduction or elimination is blocked and consequently, place assimilation does not occur. Preservation constraints can be classified into several sets on the basis of criteria such as place, manner, and prosodic domain. Each such set has its own internal ranking which is determined based on a hypothesis suggested by Kohler (1990, 1991, 1992), Steriade (1993), and Byrd (1994):
(3) Production Hypothesis
Speakers make more effort to preserve the articulation of speech sounds with powerful acoustic cues, whereas they relax in the articulation of sounds with weak cues.
Hypothesis (3) states that speakers exert more effort for those sounds which will produce dividends in terms of enhanced perceptibility. Thus, acoustically more salient segments are more likely preserved in articulation than acoustically less salient segments. This leads to a general strategy for ranking Preservation constraints: constraints preserving acoustically more salient segments must be ranked above those preserving acoustically less salient segments. The Production Hypothesis (and rankings based on it) explains the above mentioned acoustic facts which are deduced from generalizations on place assimilation shown in (1).
In summary, attested patterns of place assimilation are determined by the interaction of different sets of constraints guided by ease of articulation and ease of perception, maintaining the universal ranking within each set.
View draft version of this chapter
Return to the Phonetically-Based Phonology home page