a super(b) paper on human evolution

[Tom Givon]

Mike's question is to the point. When we discuss language evolution, we are
honor-bound to concede that  making testable predictions is--at the
moment--hard to come by, much harder than in physics or biology. Being an
incipient (if well-disguised)  optimist, I think that some day we will be able
to make testable predictions. But they are not going to be exactly of the same
kind (thus, perhaps, alas,validity?) as predictions one can make about
'hard-wired' bio-evolution. In biology, one can already manipulate  both
mutation & selection. In linguistics, one cannot, and maybe never will be able
to. But predictions can also involve complex, indirect if-then inferences, and
I think some day, with a bit more ingenuity (and a richer theory) we might be
able to make test such prediction. At the moment we must concede we are at the
hypothesis-formation stage, groping for a theory that WILL make predictions.
If you want to do science in the  complex area of culture, behavior and
cognition, you have to be a bit smarter than the physicists. The most
interesting potential science--of mind--is surely the most difficult. So the
odds of scoring big are lower, but the prize is surely worth it.  And since
when is difficulty a reason to quit?  Best, TG

================

Mike_Cahill at sil.org wrote:

> A question on Tom's hypotheses, repeated below:
>
> "Here are some quantifiable hypotheses:
>    (i) All other things being equal, typological features that are more
> widely attested
>         cross-languages may appear earlier in evolution.
>    (ii) Typological features that are more frequent in live communication
> may
>          appear earlier in evolution.
>    (iii) In Grammaticalization  Chains, earlier stages, those that tend to
> be 'source
>          constructions', may have also evolved earlier.
>     (iv) Likewise in morphemic development, concrete lexical senses have
> most likely
>           evolved before more abstract, metaphoric and/or grammatical,
> uses.
>     (v) Communicative behavior (function) most likely has preceded
> grammaticalization
>           (structure) in evolution."
>
> These may be quantifiable (in the sense that one might develop a
> hypothetical numerical model), but how can they be testable, given that
> we're talking about the emergence of language features that arose before
> writing, and are presumably undatable? And if not testable, then how
> valuable are they?
>
> Mike Cahill
>
> Tom Givon <tgivon at uoregon.edu>
> Sent by: funknet-bounces at mailman.rice.edu
>
> 08/08/2005 01:42 PM
>
> Dear FUNK people,
>
> As some of you may know, I am a slow reader & live out in the boonies.
> So it took me a while to get to a superb (tho long & complicated) paper
> on human evolution that I would like to alert y'all to. It is of great
> interest first in term of general evolution (human or otherwise), but
> also in terms of its profound implications for an eventual
> undedrstanding of  language evolution. The full  reference is: S.
> McBrearty & A. Brooks (2000)  "The revoluition that wasn't: A new
> interpretation of the origins of modern human behavior", J. of Human
> Evolution, vol. 39, pp. 453-563. It is a careful, massively-documented
> re-evaluation of the henceforth prevalent model of  a "Modern Human
> Revolution", reputed to have occurred ca. 40k BC. According to that
> 'revolutionary' model (see summary in C. Li, 2002), all 'modern human
> behavioral traits' (complex tool btechnology, complex social structure,
> organized domestic space, expanded ecological range, sophistacated
> hunting, trade networks, art, symbolic behavior & by inference "language
> as we know it now") emerged suddenly, in Europe (Cromagnon Man) without
> apparent gradual development. The paper reviews the archaoeological
> evidence--skeletal, ecological, nutritional, artefactual--in  Africa and
> Europe, as well as the behavioral implications that can be drawn from
> the physical evidence.
>
> Some of the main conclusions drawn there are: (1) the development of
> 'modern human traits' already occured in  Africa, gradually, and is
> attested there as early as 300,000 BC. (2) The apparent discontinuity in
> the European archaeological records is due to repeated out-of-Africa
> migrations & subsequent extinctions (glaciation). (3) The graduality was
> not only a matter of cultural evolution, but also of the existence of
> many sub-variants of 'early'  homo sapiens in Africa. That is--as
> elsewhere in biology & diachrony--graduality & variation go
> hand-in-hand. (4) As elsewhere in biological & cultural evolution  (cf.
> many works by E. Mayr & the recent book by Boyd & Richerson), there was
> no firm boundary between biological (genetic) and behavioral (cultural)
> evolution. Rather (to paraphrase Mayr), "behavior is the pacemaker of
> evolution".
>
> The implications of the McBrearty/Brooks paper to the evolution of
> language are many & fairly obvious. (1) Out of the windows goes the
> Chomsky/Gould revolutionay model of sudden emergence. (2) Both lexical
> and grammatical development lend themselves quite naturally to a
> gradualistic model. (3) With a certain amount of caution (see critical
> article by Dan Slobin in our recent TSL volume "The Evolution of
> Language out of Pre-Language" [2002] and my  response to it in the same
> volume), one could infer from the non-evolutionary developmental data
> accessible to us now (child learning, pidginization, diachrony) some
> possible gradual courses of both vocabulary & grammar evolution. These
> inference may follow, in the main, suggestions made in Givon (1979) and
> Bickerton (1981) about the "fossils of language". But there is more
> possible room for treating gradual evolution seriously & responsibly.
> Here are some quantifiable hypotheses:
>    (i) All other things being equal, typological features that are more
> widely attested
>         cross-languages may appear earlier in evolution.
>    (ii) Typological features that are more frequent in live
> communication may
>          appear earlier in evolution.
>    (iii) In Grammaticalization  Chains, earlier stages, those that tend
> to be 'source
>          constructions', may have also evolved earlier.
>     (iv) Likewise in morphemic development, concrete lexical senses have
> most likely
>           evolved before more abstract, metaphoric and/or grammatical,
> uses.
>     (v) Communicative behavior (function) most likely has preceded
> grammaticalization
>           (structure) in evolution.
>
> Since abstraction and complexity in both vocabulary and grammar are a
> matter of degree, and since the semantic space (of vocabulary) and
> discourse-pragmatic space (of grammar) are complex & multi-dimentional,
> graduality in evolution, acquisition and diachrony is probably the norm.
> This brings language and culture back into line with the rest of the
> biologically-based universe, where variation, behavioral exploration and
> gradual emergence are the norm.
>
> I hope this is not too far out for y'all. But if we are ever to come up
> with a viable theoretical perspective on language, a
> developmental-diachronic-evolutionary framework is, it seems to me, the
> only way to go.
>
> Best,  TG