Evolang Post #5 – d’Errico and Bickerton

There were just two more plenary speakers at the end of the conference and both delivered fascinating and controversial presentations.

The archaeological positions on language evolution are complicated and frequently bitter. Archaeologists are reluctant to speculate on anything that doesn’t fossilise and the emergence of language is embedded into several, broader debates about how humans emerged. Firstly there is a debate about how late or early modern human behaviour emerged. Secondly a debate about when anatomically modern humans emerged, and finally a debate over the geographical pattern of emergence of modern humans (both behaviourally and biologically).

D’Errico is a world famous archaeologist and one of the leading proponents of a nuanced multi-regional model of modern human evolution. He argues that emergence was gradual, patchy and complex, rather than rapid and sweeping like some have envisioned. He also promotes the idea that humans weren’t exclusive in their development of some, if not all of the advanced skills we associate with ‘being human’ and in this talk he really went to town on the idea that neanderthals had advanced symbolism that is likely to indicate a capacity for language.

D’Errico convincingly argued that although there is no categorical proof of neanderthal language this is no reason to determine that this rules them incapable of language. He held up the increasing evidence for both symbolic body painting and ceremonial decorations as evidence for advanced symbolic ability in neanderthals. It is more parsimonious, he argued, that symbolic ability indicates language, than to argue for symbolic ability without language. Beads and craftwork are good signs of not only interactive symbolic value, but of oral traditions necessary for the continuation of skills. Behavioural complexities like these indicate transmission of complex symbolic information at a higher level than mere primitive body painting or self decoration. These were in keeping with the evidence for FOXP2 in neanderthals, and other complex non-symbolic behaviours.

I’m sure many will already see the traditional objections to these arguments. Firstly that there is some evidence for speech impairment in neanderthals, and the old classic, that you can’t infer lingusitic or even communicative value from material evidence. These are all valid points but D’Errico reminded me of an old quote I heard about archaelology, describing it as “the necessary inference from limited means”. Language doesn’t fossilise and as the weight of evidence reveals increasing behavioural complexity, so the arm of parsimony swings firmly towards the claim that neanderthals had some sort of complex communicative ability. This is far more plausible than a bunch of mute, uncommunicative neanderthals engaging in complex cultural behaviour.

Whether it could ever be called full human language is a far more difficult question to answer, with lots of difficult related questions about what exactly do we mean by human language, and what influence would interbreeding have had, had if it it occurred? It also acutely poses the question of why humans persisted and neanderthals did not. Overall, a very robust presentation that had people talking about it afterwards.

Finally, the old warhorse Bickerton took to the podium. He ditched his advertised presentation and got stuck into linguistics, surfing on the general anti-Chomskian sentiment withing the room. He attacked the famous paper by Chomsky, Hauser and Fitch in which it was argued that the unique defining feature of human language faculty is the capacity for recursion (the ability to embed identical structures within structures of the same kind). This skill, it has been argued, gives human language its intense creativity and flexibility. The evolutionary value of recursion if often confused, and recursion is frequently conflated with ‘discrete infinity’, the linguistic ability make infinite combinations from finite means.

Bickerton attacked the very foundations of the C,H & F claims and argued that recursion is itself a fiction. In reality it was a dressed-up way of describing the iteration of lexical processes. Here’s a few choice quotes from the presentation:

The core process of language consists of an iterative process applied so as to satisfy lexical dependencies

He argued that the evolution of syntax is not the core issue of language evolution. In keeping with Deacon (1997) he argued that the semantic value of language is the most important, as manifested in the mental lexicon:

It thereby becomes possible to produce a more parsimonious theory in which words, with all their lexical and semantic properties evolved first, and then a simple iterative process” .. allowed these lexical values to combine together.

He argued that merge, one of the fundamental components of recursion, which allows the combining of smaller units into larger units, was actually part of the lexicon and had an underlying neural basis. He didn’t extrapolate as to what or where this neural basis might be.

As you probably know already, Bickerton is best known as a proponent of proto-language, the idea that there was a stage of primitive language that preceded full-blown language. Bickerton argued that the difference between proto-language and language would be in how the words were assembled by the lexicon. In proto-language they would project singularly to the motor-articulatory system, whereas in language they would be assembled beforehand.

A very robust and pleasing talk. Deacon was very persuasive in his arguments although I think a lot of detail remains to be filled in. I’ve never been convinced by the primacy of recursive syntax in either our linguistic ability or its evolution. This was a very plausible alternative argument, but I feel bashing Chomssy, Hauser and Fitch is increasingly becoming a straw man argument. As Chris Knight mentioned in the question session afterwards, they are all now beginning to disagree about the claims of the paper, with only Chomsky sticking rigidly to his guns. Hauser and Fitch are far less interested in the value of the broad language faculty, than just the narrow faculty and recursion.

Advertisements

4 Responses to “Evolang Post #5 – d’Errico and Bickerton”


  1. 1 watercat May 2, 2008 at 4:15 pm

    A proto-language with only single words and no syntax contradicts the evidence. Pre-linguistic gesture, both in acquisition and in language creation, has semantic frames with clearly marked theta roles, null arguments, and other features. Children transfer part of this semantic frame to the spoken word and retain the rest in their co-speech gesture; they don’t forget it all and then reinvent it later after producing their first word. It is more parsimonious to say pre-linguistic gesture evolved first, since we see it in other species. Then a simple iterative process for combining units, as in both linguistically trained apes and in humans without language, would have enabled Neanderthals to produce beads, craftwork and cultural traditions.

  2. 2 languageevolution May 3, 2008 at 7:37 am

    I agree with you completely. To be honest, I suspect that to a certain extent Bickerton might do too. Although he would frame it by saying it doesn’t matter whether protolanguage is gestural or spoken (see Bickerton’s “A Brief Guide to Language Evolution for Linguists”).

  3. 3 watercat May 4, 2008 at 1:03 am

    Well, the issue isn’t the medium but whether words or syntax came first, and DB has always defined proto-language as lacking any syntactic structure. His attempt to refute Wray’s holophrastic stage, which ignores the gestural components of language, leaves him baffled about how the hearer could assume a pronoun ‘referred to a female beneficiary’—that the speaker would be pointing at! Using their example, you can’t gesture just “give”, you can only gesture the holophrastic phrase [give| agent, beneficiary]. Speech, sign, and gesture all indicate arguments by pointing, because the symbol grounding problem is nearly insurmountable otherwise. The deWaal study shows that even wild Bonobos do this, making their “help” gesture as a directional proto-verb with movement from subject to object. This is clear hierarchical structure, before the advent of words, yet Bickerton insists that this stage of evolution was followed by one with words but no structure, and then a further stage where the structures re-appear.

  4. 4 Nonsupporting June 19, 2008 at 4:59 am

    Somehow i missed the point. Probably lost in translation 🙂 Anyway … nice blog to visit.

    cheers, Nonsupporting.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s





%d bloggers like this: