(Note: This post is the fifth in a series of posts exploring the contents of Jean-Louis Dessalles' book, Why We Talk. For a summary of what came before see Why We Talk: Summary. I had been posting these summaries every Wednesday, but that way promises to take forever. So I'm going to do a bunch in a row and see how that goes.)
Homo erectus survived between one and two million years and during all that time the species seems unlikely to have had the grammatical skills of a three year old. At best they may have spoken what we would call a pidgin:
- mamasan go.
- bring horse.
But as the man in the Geico commercial notes, they were smart enough to spread across Asia and use fire.
Dessalles agrees with Derek Bickerton that pidgin/protolanguage "is not the result of a rough simplification of language; it is a tool for communicating meanings that has its own organization." (p. 171). It also had its own function.
One of Dessalles; strengths is his insistence that we must take seriously the notion that evolution is not goal+oriented. Speech without syntax was not speech striving to become grammatical, it was functional in its own terms:
it is unwise to use syntax as some sort of miracle that hominids eventually stumbled upon after much trial and error. ... it is wiser to see protolanguage as a system which, well adapted as it was to its function, had no need for syntax as we know it. (p. 173)
From our perspective, coherent speech is always useful, always better than no speech or pidgin, but just as chimpanzees do not need language, erectus did not need syntax. It was sufficient for protolanguage to group words into semantic components.
Dessalles is in complete agreement with the general proposition urged on this blog that speech is a system for directing attention. At the protolanguage level it was (for Homo erectus) and is (for modern-day users of pidgins) a way of drawing attention to "concrete scenes, either experienced or imagined."
Protolanguage can be used in a few basic ways. One is to point out something present, 'zebra there.' Another is to negotiate, 'two axe.' 'three shell.' These kinds of exchanges are all possible in pidgin and, assuming Homo erectus had any speech at all, appears to have been sufficient (when accompanied with energetic pointing) for perhaps two million years.
What was the value of protolanguage? In an off-list exchange that I had with Massimo Piattelli-Palmarini a few weeks ago, he told me:
The protolanguage issue is quite moot, in my opinion. Animals can surely do categorization, and chimps (as David Premack has shown many years ago) can analyze objects into features (color, shape, texture etc.). Something like conjunction of features and a symbol-object correspondence can be granted. But in order to have a language of thought, the creature needs predication, and that comes with Merge, and with edge features, so possibly the protolanguage already was language. In other words, no protolanguage.
I think he is saying that without predication you don't get a "language of thought" and therefore whatever early "protolanguage" existed it wasn't just something for chimpanzee-level understanding. Thus the "protolanguage" was really language. But I have to admit that the remarks are so pithy that I could be way wide of understanding. Dessalles (I think) would reply that we are not talking about a "language of thought" but one that draws attention to salient things and did so without having predicates. Indeed, Dessalles is clear that he sees the introduction of predicates as coming after protolanguage.
We can see something like protolanguage in the speech of a child of 30 months, but a society of perpetual two-year-olds would have no use for speech because children of that age do not converse. They listen to their parents, but not their age-mates. It is probably more accurate to imagine Homo erectus engaged in a kind of pidgin interaction than mere childish pointing. Dessalles describes protolanguage as having no syntax, but having a protosemantics. The excited cry of lion was probably less typical than of what they had to say than the more ordinary back and forth of practical action:
A: fig yonder. (points beyond a hill)
B: (shifts attention to a third figure) koko. fig.
C: fig good.
A: no axe.
B hands A a stone axe. C and A go off to find the figs.
Obviously this dialog is pure fiction, but it illustrates the kinds of thing that protolanguage makes possible, the most important being practical cooperation and mutual dependence. This scene imagines that figures A and C can be trusted to go off and bring back figs, rather than simply eating them all themselves. If Homo erectus enjoyed that kind of trust, they lived lives that were very different from apes today. If not, it is hard to see what they had to talk about.
For anyone familiar with full speech, one of the striking features of pidgin conversations is their abruptness. They promote no intimacy. Dessalles ignores this fact and never discusses it as a limitation, but syntactical speech enables the inclusion of respect-assuring and trust-building phrases. Plainly, in a world of abrupt speakers, people are not going to develop grudges over such rude speech, but they are not going to develop trust either. If protolanguage speakers were going to trust one another to the point of sharing useful information, they had to be developing that trust outside of speech itself. Dessalles does raise the matter of 'verification' and says that with protolanguage verification must be deferred. If A says that figs are over yonder, the truth or falsity can only be learned later. But there is more to trust than truthfulness. In the example, two individuals are being trusted to share food.
What protosemantics could do is point and evoke perceptions. By itself, perception recognizes objects. Somebody sees things on trees and recognizes them as a sweet food. Speech adds recall to the mix. Somebody sees objects on trees and recalls the word fig. Somebody hears the word fig and recalls an image of the fruit, or maybe recalls the scent or taste of the fruit. If described in terms of perceptions, the conversation above would be:
A recalls perception and says fig yonder, inspiring B to recall a similar perception.
B directs attention toward another person and says, koko fig, inspiring koko to recall theimage of a fig. She is left to determine from the context why B has said such a thing. A might add a word like bring because protolanguage allows for combining recalled perceptions into a string in which one perception follows another.
C recalls the taste of figs and says fig good. If statements such as this were possible for Homo erectus, they could acknowledge what they have been told to do.
A recognizes that it lacks a tool for harvesting figs and says no axe. B recalls the image of an axe and supplies a real one for the task at hand.
For a hundred thousand generations Homo erectus was a prosperous species, spreading out of Africa and across Asia and into at least the eastern European region, probably becoming the most successful ape species ever. It was not an especially creative species, making and using the same stone "axe" throughout the whole of its existence, but it appears to have been highly adaptive, able to occupy a variety of niches and climate zones. Imagining exchanges like the simple one presented here helps make sense of this success. They were not philosophers, but they could interact purposefully. But of course that success raises another question, one Dessalles presses. If this kind of perception-based communication was sufficient for well over a million years, what could have happened to force a change?
Comments