* Dynamique du Langage
Lyon

** University of California
Santa Barbara

abstract

If we plot the increase in language capacities as a function of time over the last 4 million years, several theoretical curve shapes are possible :

- a straight line representing a slow linear evolution of language through this long period of several million years

- a step-like representation illustrating a series of punctuated equilibrium

- a curve progressing very slowly for a long period of time followed by a sharp increase after an inflexion point relatively recent in time.

In this paper, we would like to summarize the existing evidence to argue for the third scenario. We will consider three phases :

Phase 1: "Slow evolution" period (from 4 My to 100 Ky)

Phase 2 : "Inflexion point" period (around 100 Ky)

Phase 3 : "Fast evolution" period (the last 100 Ky)

We will propose that the functions of our communicative system and the mechanisms at work to modify this system through time are different in phase 1 vs phases 2 and 3. In phase 1, our "language" is used to communicate our emotions, to maintain social ties and progressively to give information about the "here and now". In phase 3, language is used to transfer knowledge : This new function will have an extremely strong snowball effect for the survival and development of our specie since social, technical, and geographical knowledge will not have to be "relearned" by each generation, it will be transmitted by articulated language. Knowledge will become cumulative. In order for phase 3 to be possible, it required a much more sophisticated linguistic system than the one used during phase 1. Drastic changes occurred in phase 2 : development of "double-articulation" (grouping of meaningless sounds to produce meaningful words- ie emergence of lexicon) and use of word sequencing (emergence of syntax). Because of the snowball effect mentioned, it is conceivable that this phase was relatively short (a few tens of thousands of years).

We will argue that if darwinian type of processes can account for early periods of language evolution (in order to create a more "efficient" communication system), recent evolution of linguistic systems are not explainable in these terms and are better accounted for by self organization processes.

Finally, we will suggest that a possible explanation for the disappearance of Neandertals could be that, unlike Homo sapiens, they did not get to a stage of "knowledge transfer".

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