RE: [xmca] Reference for ontological and phylogenetic languagecomparison

From: Jay Lemke (
Date: Thu Jan 11 2007 - 15:42:02 PST

I wonder just how people are interpreting 'recapitulation' here? You
can't usefully think about development and evolution separately. What
evolves is the typical developmental trajectory of the members of
some species. At any given phylogenetic time, there is some
similarity in the development of all members of the species, with
individual variations around a common and typical normal ontogenetic
trajectory. Across phylogenetic timescales, that typical trajectory
itself changes.

But the later stages of the trajectory are more likely to change than
the earlier ones, because the very possibility of the organism making
it to some stage where it might do things differently depends
critically on its having followed a path up to that point that was
"pioneered" by its ancestors, who found successful ways to stay alive
up to that stage. You can think of this as our dragging our
phylogenetic baggage around with us, or as our building on the
achievements of our ancestors. But their biggest achievement was to
figure out how to get from an egg through an embryo to a newborn, and
any changes that mess with their traditional formula for success is
likely to lead to a very dead end. That leaves the successful changes
to operate mainly on the later end of the trajectory, with the
earliest bits being, by default, recapitulative.

In the case of language, we don't really know much about what bits
are early vs. late phylogenetically. We just have what we see
ontogeneticallly in language development. So we can be pretty sure
that the capacity to learn and use language at all is at the early
end. Maybe some parts of the Chomskyan argument are correct (and I
don't buy a lot of those arguments) that certain core features of
grammar, shared by all languages (or at least shared by them from our
point of view ... remember that a lot of the world's original
languages must have been totally lost a very long time ago) are also
early enough that all future languages have to build on them. My
guess is that there may be some such features, but that current
theories on these matters posit way too many of them.

If you look at child language development, you can make a case like
Halliday does in _Learning how to mean_ for how grammar itself
emerges from two-sign utterances (can be two words, can also be a
word and a gesture, or a word and an intonation). Some such principle
as he identifies for a language to have a grammar (i.e. something
more complex than a one-to-one sign-to-meaning correspondence) may
also be phylogenetically early enough to be necessary for all later
language development. Does the capacity for it rest on some genetic
mutation? some increase in brain complexity? very hard to say when
all those brains are gone to dust.

All this however is still very individualistic. And language for the
most part is not, it is a social phenomenon that we may internalize
for some private purposes. So it may be the evolution of social
complexity, rather than genetic mutation per se, that made more
advanced language (i.e. with grammar rather than protolanguage with
it) have enough value-added to get added onto the growing end of the
developmental trajectory. In which case, there are mutual advantages
in a community to helping other people (and growing infants) learn
your way of talking, and to using those ways of talking, etc.

And here we can make the football analogy. What is important is that
the game get played. To get played it needs players with learned
skills, or evolved-and-learned skills. The players have developmental
trajectories on the biographical timescale. And the game itself, as a
dynamic unfolding process, has a developmental trajectory (start to
finish) and a phylogenetic one (typical game-play styles evolve on
the historical timescale, in football, or chess). The analogy would
say that opening gambits tend to be conserved over time, and games
individuate more, and playstyles change more, with respect to later
moves in the game. Is that true?

At the individual player level, we could look at player training.
There would be a typical newbie-to-pro trajectory, which might change
over historical time (new ideas in coaching players), but the basics
that come early in training would stay the same. Do they? And all
players would tend to play more or less the same way when they start
playing the game, and only individualize their play style later in
their years of playing. Do they? Note that we have to talk here about
usual and typical cases, not exceptions.

Exceptions, of course, like mutations or innovations, may get
conserved and reproduced, or not. One factor in whether they do
continue may be how they fit into the constraints of successful
overall trajectories (both for player careers and for gameplay).


At 10:36 AM 1/11/2007, you wrote:
>What a nice, useful analogy Michael.
>I'm thinking about how to make it more precisely parallel. The (ontogenic)
>development of language ability in the child could be compared with the
>(phylogenic) development of a player's football skills (I'm thinking
>basketball might work better, since -- at least in US "gridiron" football --
>most players on the field have specialized roles not requiring as great a
>range of versatile skills as in basketball [IMHO: a defensive left guard
>might think otherwise]). So, the development of a [basketball] player's
>skills would not recapitulate the development of the game itself. Skills
>that might have had value in the game as it was played in the early history
>of the game might have no value for players today, and would not be part of
>a developmental stage that today's players go through on their way to
>development of skills they use today.
>-----Original Message-----
>From: [] On
>Behalf Of Wolff-Michael Roth
>Sent: Thursday, January 11, 2007 10:22 AM
>To: eXtended Mind, Culture, Activity
>Subject: Re: [xmca] Reference for ontological and phylogenetic
>A CHAT perspective built on the dialectic of individual and
>collective, the person realizes cultural possibilities available to
>any one else. From this perspective, children grow up in a different
>material context, hearing different utterances in the context of
>different situation. This would lead to the contention that ontogeny
>does not recapitulate phylogeny, much in the same way that a present
>day football game would not recapitulate the first football game ever
>played or its precursor. (The referent of "football" can be taken the
>British or American way).
>On 11-Jan-07, at 6:46 AM, wrote:
>Dan I. Slobin has an article, "From Ontogenesis to phylogenesis: what
>child language tell us about language evolution?" that appears in IN
>j. Langer, S.T. Parker edited volume, "BIology and Knowledge.
>The questions he poses in the article are: Does linguistic ontogeny
>recapitulate phylogeny?, Does linguistic diachrony recapitulate
>OD children create grammatical forms?
>good read but not a CHAT perspective but rather biologicaly based.
>xmca mailing list
>xmca mailing list
>xmca mailing list

Jay Lemke
University of Michigan
School of Education
610 East University
Ann Arbor, MI 48109

Tel. 734-763-9276
Website. <>
xmca mailing list

This archive was generated by hypermail 2b29 : Thu Feb 01 2007 - 10:11:31 PST