ASL 'Syllables' and Language Evolution: A Response to Uriagereka

I am grateful to Juan Uriagereka (2000) for his thorough and thoughtful review of my book The Origins of Complex Language (henceforth Origins; Carstairs-McCarthy 1999).¹ The book tackles fundamental questions about the relationship among syntax, semantics, and cognition, and Uriagereka is not persuaded by all my suggestions about the prehistory of this relationship. I will not pursue these large issues here; rather, I want to address a more circumscribed issue that is nevertheless crucial to the argument of the book, so that my failure to discuss it is an important omission, as Uriagereka points out. This issue is whether the syllable, as a unit of phonological description, is modality-neutral (so as to be equally at home, with fundamentally the same sense, in descriptions of signed and spoken languages), or whether the syllables of signed and spoken languages are really different phenomena, so that the use of the term syllable for both draws attention to resemblances that are more accidental than fundamental. I will argue that the evidence supports the latter view more strongly than the former; therefore, when discussing language evolution, it is legitimate to appeal (as I do) to aspects of spoken syllables that are undoubtedly modality-dependent, such as their physiological underpinnings in the vocal apparatus.

Before addressing this issue directly, I would like to summarize briefly why it is important in the context of my book. Second, by way of reassurance, I will explain why the conclusion that I reach does not belittle sign languages, nor imply any old-fashioned skepticism about their entitlement to be recognized as real manifestations of the human language capacity.

1. The importance of speech for the Origins scenario

Among primates, humans are unique in two respects: they have language and they are habitually bipedal. A peculiar aspect of the adult human vocal tract, by comparison with that of other primates, is its L-shaped configuration, with the pharynx at right angles to the oral cavity. This facilitates speech, and hence spoken language, by equipping human vocalization with independently variable first and second formants (the first associated mainly with the volume of the pharynx, the second mainly with the volume of the oral cavity). These formants make it easy to produce a wide range of vowel distinctions, and contribute importantly (through formant transition patterns) to the acoustic and auditory discrimination of consonants. But the existence of the pharynx is due to the low position of the larynx in adult humans, as a result of which no self-contained tube can be formed to conduct air from the nose to the lungs. Adult humans, therefore, unlike nearly all other mammals, cannot breathe while swallowing, and are subject to a relatively high risk of choking.

One widely held view of this peculiarly human anatomical configuration is that it is a byproduct of selective pressure favoring the development of spoken language: the risk of choking was outweighed by countervailing linguistic advantages. But, as I point out in Origins (citing evidence from biological anthropology), the original impetus for larynx lowering may have been independent of language. Bipedalism itself, and the consequent ninety-degree reorientation of the skull in relation to the spinal column, made it harder to accommodate the larynx in the standard mammalian position, close under the skull base and contiguous to the soft palate. The kind of more varied vocalization that the L-shaped tract made possible, whereby alternations in sonority due to successive widenings and narrowings of the tract could be combined with a larger vowel repertoire (creating syllable nuclei and margins), could therefore have arisen for reasons quite independent of, for example, brain expansion or social intelligence, and, indeed, quite independent of language in anything like its modern sense.

This scenario implies two things about syllables, however. First, they were historically a very early ingredient in language evolution, long predating syntax, so that their neural underpinning was available later as raw material out of which a neural underpinning for syntax could be tinkered into existence. This accounts (I suggest in Origins) for some of syntax’s central oddities, such as the sentence/NP...