Psychology Department, Aberdeen University
Aberdeen, Scotland.
e.salzen@abdn.ac.uk

Speculations on the evolution of speech have rejected an origin from animal calls and vocalizations because of an apparently unbridgeable divide between Vocal signals which are innate, conative, emotional, involuntary, & high intensity and Speech symbols which are learned, cognitive, rational, voluntary, & low intensity. Furthermore the Neural systems of emotional calls are sub-cortical, limbic and right hemisphere dominant whereas those of speech are cortical and left hemisphere dominant. This divide can be bridged in Phylogeny, Ontogeny and in Neurology using existing studies that seem to have been overlooked in the controversy on the animal origins of speech.

In the case of phylogeny, an ethological analysis of social signals in animals (Morris, 1956) shows them to be patterns of intention movements due to frustration and conflict (thwarting). These movements and the vocalizations produced by these movements have referential significance because of their nature, orientation and releasing stimuli. Similar emotional conflict and thwarting signalling in humans induces co-operative social responses that may end the thwarting (Salzen, 1991). The evolutionary process of "ritualisation" (Blest, 1961) involving "typical intensity " and "typical form " (Morris, 1957) accounts for animal signals and calls with repeated elements and constant form. Such calls are comparable with the consonant/vowel or c/v/c or c/v/c/v of the first utterances and words of babies (cf. Lewis, 1936). According to some studies of comparative philology (Diamond, 1959; Swadesh, 1971) the first words of early man may have been of the same form and were verbs used as imperatives to get other people to do things (Diamond, 1959; Revesz, 1956). Words signifying the use of an implement or tool would be verb-nouns and lead to a naming lexicon. As in the case of animal signals where repetition can indicate intensity, in some modern languages repetition of the cv elements is indeed used to indicate plurality, intensity, size, and adverbs (Farb, 1974). In this way further variation of the elements can extend the lexicon and allow the development of a syntax.

In the case of ontogeny, a rarely cited study (Lewis, 1936) preempted the ethological use of intention movements in signalling by showing how the first speech sounds of infants arise from anticipatory and residual feeding/sucking actions in hunger and satiation. It showed how these sounds are single or duplicated labials or labial-dentals with "a" vowels i.e. cvcv and form the majority of infant first words in European languages. Similar words occur in many other languages (Wundt, 1924) and have similar meanings (cf. Clark & Clark, 1977). These first words have whole sentence meanings (holophrases) such as "do x with y " (cf. Greenfield & Smith, 1976) and they also show the syllabic repetition essential for the development of a lexicon. The same infant study (Lewis, 1936) also suggested that these first speech sounds give rise to babbling because they reinstate their associated motivational and motor states which then reinforce the emitted behaviour. In this way babbling provides the bridge between involuntary and voluntary vocalizations. Babbling becomes involved in first words when adults pick up and use the sounds made in babbling by speaking in "motherese " with simplified consonants and "cvcv " repetitions (mama, papa). Subsequently when the infant makes these particular babbles again it is reinstating the parental reinforcement sounds and comfort states. This provides a sensori-motor rehearsal that facilitates the developing voluntary motor control of the speech apparatus and verbal mimicry, which is undoubtedly a major means of speech and language learning.

In the case of neurology, another less well-known work (Jurgens, 1992) describes a brainstem articulatory coordination system, a forebrain motor control system, and a limbic motivational initiation system for phonation in primates. Humans, unlike other primates, appear to have a direct connection between the primary motor cortex and the brainstem nucleus ambiguus which contains the laryngeal motoneurons (Kuypers, 1958), and this may be how humans obtain exquisite voluntary control of articulation. Control of sequential articulations and hence of syntactic language develops along with voluntary control of sequential arm-hand movements during reaching and grasping in the neonate (cf. Young, 1977). In both cases this occurs primarily in the left hemisphere, which has long been implicated in sequential single stimulus-response processing (cf. Levy, 1969) and in the lateralisation of both manipulation and speech. If the present view that speech arose from the effects of thwarted action states on the autonomic nervous system is correct, then this system may still be involved in the instigation of speech. One might expect, therefore, that lesions of higher autonomic centres would affect the impulse to vocalize or speak rather than give specific speech defects. Human patients with parasagittal tumours near the anterior cingulate gyrus, a structure through which emotive motivational information enters the cortical system from the brainstem, may show poverty of speech. In one significant case (Damasio & Van Hoesen, 1983) the patient could repeat speech and with subsequent partial recovery reported that she had simply had no will or desire to speak. This is consistent with a thwarting theory of social signalling and its role in the origin and evolution of speech.

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