Englisches Seminar der Universität Kiel
Leibnizstr. 10
24098 Kiel, Germany
OfficeLing@anglistik.uni-kiel.de

abstract

Attempts to reconstruct the evolution of the capacity for phonological systems in homo sapiens require consideration of the anatomy of the vocal tract (e.g. Liebermann 1975, 1984, 1990, Kent 1992), perception as well as the interaction between the two. In this paper I will review the impact of the auditory system on these issues along lines originally suggested by Kuhl (e.g. Kuhl/Miller 1978, Kuhl/Padden 1982, 1983, Kuhl 1988, 1993). I will argue that the structure and the functioning of the auditory system as required for natural human languages to function the way they do, antidates the evolution of the vocal tract as required for speech by quite a margin. The data to be considered include speech perception in human neonates, pre-speaking infants, children and adults, as well as non-human animals. The argument is focussed primarily on distinctive features, their typology, and the ability of human beings to categorize the sound wave into segments. It is argued that the distinctive features are based on the heightened sensitivities of the auditory system. They restrict the universal set of sound contrasts that are used for phonemic purposes, i.e. these auditory sensitivities constitute the biological basis for the distinctive features; and categorical perception constitutes the biological basis for the phoneme.

The first step in the argument is to point out that the distinctive features and phonemes are based on the functional potential of the auditory system and not on the anatomy of the vocal tract. That is, perception is primary and articulation secondary. The crucial evidence for this point is the typology of distinctive features and the fact that there is nothing in the anatomy of the vocal tract to explain why, for example, only certain places of articulation are used for phonemic purposes and not others.

The second step is to show that human neonates and infants can do two things: (a) They have categorical perception; and (b) during the first half of their first year of life they can detect those sound contrasts that are used for distinctive purposes in the languages of the world, including those that do not occur in their native language (e.g. Streeter 1976, Trehub 1976, Lasky et al. 1975, Werker/Tees 1984, Best et al. 1988, Best 1994). Apparently, these ablities have an innate basis, and can, therefore, serve as the biological basis for the functional potential of phonological systems.

Two objections are possible here. One is that categorical perception is a phenomenon derived from the research on synthetic speech and it needs to be shown to also apply to non-synthetic, i.e. natural stimuli. The second objection is that so far we are far from having checked all or most of the sound contrasts that are utilized phonemically in the languages of the world. The first objection is valid. However, there are some experiments that have relied on natural stimuli, such as the dental vs. retroflex stop contrast of Hindi (Werker/Tees 1984). Moreover these findings have been replicated for synthetic stimuli (Werker/Lalonde 1988). Although these studies were not on categorical perception, they do show that the perception does not neccessarily differ for natural and synthetic stimuli.

The third step is concerned with the animal evidence. The earlier research on chinchillas, monkeys, Japanese quail, and European starlings (e.g. Kuhl/Miller 1978, Kuhl/Padden 1982,
1983, Kluender et al. 1987, Kluender et al. 1997 etc.) had shown that these animal species had categorical perception and that their points of heightened auditory sensitivity along a given acoustic dimension parallelled those of human beings. This kind of evidence allowed for several conclusions. First, categorical perception and the points of heightened auditory sensitivity are not unique to homo sapiens, hence they are not species-specific. It followed, second, that they are not language-specific either. Third, since the same properties occur in such diverse species including homo sapiens, it is reasonable to assume that these species share a common ancestor. Fourth, given the differences in the anatomy among the present-day species, the development of their anatomical differentiation needs to be taken as postdating the shared auditory system by a considerable margin. Fifth, it is not quite clear at the present time whether it can also be claimed from the above evidence that perception has had an impact on shaping the vocal tract in such a way that the latter was brought into line to closely match perception in the sense of Stevens' notion of the quantal nature of speech (Stevens 1972, 1989).

Whereas the earlier animal evidence referred to above suggested that the points of heightened auditory sensitivity were the same for humans and non-humans, there is a recent study by Sinnott/Brown 1997 that indicates that that may not neccessarily hold for all sound contrasts used for phonemic purposes in the languages of the world. Sinnott/Brown studied the perception of the /r/-/l/ contrast in American English by adult native and non-native speakers and Japanese macaques using a synthectic continuum ranging from /ra/ to /la/. The contrast was discriminated by both groups of subjects. But the category boundary was located in different areas of the cotinuum. With the monkeys it was shifted more towards the /r/ end of the continuum, with the humans it was located more towards the /l/ end.

Obviously, it is no longer correct to assume that animals and humans share all points of heightened sensitivity. Apparently, this is correct only for some contrasts, notably, VOT in stops, place contrasts like labial, dental, velar, the (some?) vowels, and probably others. Nonetheless, it still seems justified to maintain the view that natural human languages and the capacity for them have evolved in a natural way from common, i.e. non-language-specific properties of the organisms according to Neo-Darwinian principles.

 Conference site: http://www.infres.enst.fr/confs/evolang/