The Cambridge Handbook of Communication Disorders examines the full range of developmental and acquired communication disorders and provides the most up-to-date and comprehensive guide to the epidemiology, aetiology and clinical features of these disorders.
Date: Wed, 8 Oct 2003 05:46:18 -0400 (EDT) From: Nancy Hall <firstname.lastname@example.org> Subject: Phonological Representation and Phonetic Phasing
Kehrein, Wolfgang (2002) Phonological Representation and Phonetic Phasing, Niemeyer.
Nancy Hall, University of Haifa
This study addresses two questions about phonological representations: the nature of affricates and the level at which the laryngeal node is licensed. Based on an extensive typological survey of 281 languages, Kehrein argues that the conventional view of affricates and laryngeality predicts many kinds of contrasts that are not attested.
The first part of the book concerns the nature of affricates (which comprise, in this taxonomy, pre- and post-nasalized stops and laterally released stops as well as the familiar strident affricates). Kehrein argues that "affricate" is not a phonological category: affricates themselves do not form a natural class. Furthermore, affricates and fricatives do not together form a class [continuant], contrary to theories which treat affricates as involving the features [stop, continuant]. However, affricates do form a natural class with stops, and hence are analyzed as kinds of stops.
According to this theory, the sounds described as affricates fall into two classes. In the first group, affrication is the realization of one of the manner features [strident], [nasal] or [lateral] on a stop. This type of affricate can contrast with a plain stop at the same place of articulation, since the two are featurally different. In the second group, affrication is a phonetic strategy for maximizing the perceptibility of small differences in place of articulation. The segments thus distinguished are simply [stop]s, without extra manner features. This type of affricate cannot contrast with a plain stop at the same place of articulation: for example, it is claimed that no language contrasts bilabial stops with bilabial affricates, although some contrast biliabial stops with labio-dental affricates. Only the place feature is phonologically represented; the affrication which serves to enhance the place contrast is a purely phonetic detail. Thus, neither type of affricate involves a phonological stricture contour.
The second part of the book deals with the nature of the laryngeal node. The basic claim is that laryngeal features are not licensed by segments. The laryngeal node (which consists of the three privative features [spread glottis], [constricted glottis], and [voice]) is licensed directly by the subsyllabic constituents of onset, nucleus, and coda. Within each of these constituents, laryngeal features are temporally unordered with respect to supralaryngeal features.
The evidence for this claim comes primarily from a typological study of contrasts. It is argued that no language has more than one laryngeal node in a single onset, nucleus or coda, nor does any language contrastively order laryngeal and supralaryngeal articulations within these constituents. For example, no language contrasts laryngealization and laryngeal segments within a constituent: glottalized [p] cannot contrast with a sequence of [p] and glottal stop. Conflicting laryngeal features do not occur within a constituent, and pre- and post-laryngealized constituents cannot contrast. Phonological processes that affect laryngeals, such as assimilation, OCP, and neutralization effects, apply over whole subsyllabic constituents rather than over individual root nodes.
Several apparent counter-examples to these claims are brought up and argued against. For example, Huautla Mazatec has been claimed to contrast pre- and post- aspiration; Kehrein argues that the consonants transcribed as post-aspirated are actually plain consonants preceding breathy vowels. Other counterexamples include languages, such as Georgian, Bella Coola, and Imdlawn Tashlhiyt Berber, that apparently allow long strings of consonants in onsets or codas. These long strings do allow contrastive ordering of laryngeal specifications. Kehrein argues, following other researchers such as Dell & Elmedlaoui 1988 and Bagemihl 1991, that these strings are actually polysyllabic, and hence do not contradict the generalization that ordering is non-contrastive within a single subsyllabic constituent.
Other evidence for the independence of the laryngeal node from segments comes from long-distance laryngeal movement, such as Grassmann’s Law in Sanskrit, in which laryngeal features move about words, leaving supralaryngeal features behind.
There are a large number of phonetic strategies for realizing laryngeal articulations, varying along the articulatory dimensions of degree of glottal constriction, vertical larynx movement, and gestural phasing of laryngeal constrictions with other laryngeal constrictions or supralaryngeal constrictions. However, Kehrein argues that these phonetic details are not relevant to systems of contrast, or natural classes.
This attempt to distill a small number of phonological contrasts from a wealth of phonetic realizations is essentially the unifying theme of the book. Both affricates and laryngeals are argued to participate in fewer contrasts than usually assumed. Just as non- strident affricates are argued to be only a phonetic variation on plain [stop]s, many phonetic realizations of laryngeality are argued to be featurally identical and hence incapable of contrasting with one another.
The study does not assume any particular theoretical framework, such as Optimality Theory or a rule-based framework. This neutrality allows the question of representations to be kept distinct from the system of mapping underlying forms to surface forms. In the final section, Kehrein discusses implications of the theory for a variety of frameworks, such as Aperture theory and Articulatory Phonology.
I found the typological claims made in this work fascinating. They are a challenge for many linguistic theories, and if accepted, require significant changes in thinking, not only about the place of laryngeal and stricture features, but possibly about syllable structure. The theory about laryngeals depends on a concrete, structural definition of subsyllabic constituents (onset, nucleus, coda), which some theories do not currently assume.
Kehrein makes a convincing argument that the wealth of phonetic realizations of certain features has sometimes obscured the fact that the possibilities for contrast are relatively limited. This reduction of the number of possible phonological contrasts, and disentangling of featural contrast from phonetic detail, is a desirable result. As with any typological work, the claims will of course need to be evaluated by linguists who are experts in the particular languages discussed. The number of languages mentioned is enormous, and the book could well be used as a reference for locating languages with particular types of contrasts. The presentation of data is clear and organized, with good use of charts.
The lack of formal demonstrations of input - output mapping systems (whether rules or constraints) does cause unclarity at a few points. For example, Kehrein occasionally suggests possible Optimality Theoretic constraints, but it was not clear to me in some cases how the constraints proposed would actually capture the pattern. Inclusion of tableaux might help. However, since the book’s main goal is to describe possible systems of contrast rather than their implementation in a particular framework, this was not a major problem.
Bagemihl, Bruce (1991) Syllable structure in Bella Coola. Linguistic Inquiry 22. 589-646.
Dell, François & Mohamed Elmedlaoui (1988). Syllabic consonants in Berber. Linguistics 34 357-395.
ABOUT THE REVIEWER:
ABOUT THE REVIEWER
Nancy Hall is currently a visiting instructor at the
University of Haifa, Israel. Her research interests
include Articulatory Phonology and the phonetics /
phonology interface. Her recent Ph.D. dissertation
(University of Massachusetts- Amherst) concerns the
phenomenon of intrusive vowels: non-segmental, vowel-like
percepts that are heard in consonant clusters.