LINGUIST List 7.551

Sat Apr 13 1996

Disc: Outline of evolution/emmergence (Was re:Formalism)

Editor for this issue: Ljuba Veselinova <lveselinemunix.emich.edu>

Directory

1. Stephen P Spackman, Re: 7.533, Disc: Economy, Minimalism, Formalism
2. Stirling Newberry, Outline of evolution/emmergence Was re:Formalism
3. Peter Daniels, Re: 7.533, Disc: Economy, Minimalism, Formalism
4. Allan Wechsler, Natural selection and Linguistics

Message 1: Re: 7.533, Disc: Economy, Minimalism, Formalism

aeulenbeindiana.edu (Alex Eulenberg) wrote in 7.533 (responding to
Robin Allott, 7.512):

> Robin, you seem to be saying that any theory which finds linguistic
> categories in more general laws of human physiology or psychology must be
> compatible Darwin's theory of evolution in order for it to be
> scientifically valid.
> [...]
> I think what you say is untrue, in particular when you say "Darwinian
> processes" (by this I assume you mean survival of the fittest) "account
> for all other animal and human forms and behaviours." In fact, Darwinism
> can only account for trivial differences, such as variations in shape,
> size and color -- but Darwinism does not answer questions about the origin
> of truly novel structures.

On the contrary, I believe it can be (and has been) established that
_Darwinian_ evolution is not merely necessary but sufficient to
explaining the biological diversity that we see around us; and
understanding how and why this can be is vital to linguistics. Note
that I underlined "Darwinian", since the theory that is supported by
the evidence is the one involving the perpetuation of the gene-lines
of highly reproducing individuals, the "survival of the fittest [=most
surviving]", rather than the positively mediaeval notion of "survival
of the fittest [=most like me]" which is more usually ascribed to
Darwin in school and on television.

One conceptual breakthrough we have seen in the last couple of decades
is "selfish gene" theory, which explains "altruistic" behaviour in
genetic terms by sharpening the identification of the "individuals" to
which evolution applies to genetic structures rather than their
macroscopic manifestations per se. This is important because it allows
us to see how things that are not encoded purely genetically
(including linguistic or cultural structures) can also get in on the
act, once genetics has "licensed" them by providing suitable hardware
(that is, there is automatically a meme -> gene transfer once the gene
-> meme system is established).

That still leaves the puzzle discussed under names such as "missing
links" or (in other circles) the "punctuated equlibrium" model. What
is going on here is that genes exhibit a quasi-heirarchic control
structure in which one gene can control another gene's expression and
even the _error rate_ permitted when it is reproduced.

The consequence is that the evolutionary mechanism _itself_ is subject
to evolution - all well within the framework of Darwin's very elegant
theory - and genetic organisations that favour favourable mutations
and reduce the likelihood of fatal (or more importantly, inefficiently
fatal!) ones are, in the extreme long term, themselves
favoured. "Rapid-deployment evolution" is an evolved strategy that
relies on segregating the mechanics of biochemistry from those of, for
instance, body plan, or on a more superficial level still, colouration
(I'm a decade behind the times on this stuff, but if you want some
hard evidence of this phenomenon I'd guess you could do worse than to
look into work on the "homeobox" genes).

The relevance that this has for linguistics is twofold. First, it
re-establishes the plausibility of evolution as the principle
underlying the biological stratum - and in the absence of any
mechanically satisfactory competing proposal that I know of, Occam's
razor suggests that we maintain it as our working hypothesis (while,
of course, turning any further evidence that may turn up to its
validation, as appropriate). Second, it meshes with and is reinforced
by what is hopefully actually clearer in diachronic linguistics than
in biology: that certain structures (whether biological, syntactic,
lexical or morphological) are common _because_ they are enduring once
they occur, and they are enduring because they are either
"self-protecting" in their context or readily protected by related
structures in their environment (which might in turn be enduring or -
and this is important - cyclically recurring).

Perhaps the most important observations of all are (a) that modularity
is often an evolutionarily favoured phenomenon because of its
chaos-resisting properties, and (b) that evolution provides, and
continues to provide, _cross_-modular links, while synchronic
mechanisms are restricted to intra-modular interactions and limited
interface phenomena.

If evolution appears to be implausible, it may be because you are
forgetting that the mechanisms of evolution evolve. This is where our
in traditional terms not especially "fit" species cashes in - we adapt
furthest and fastest.

stephen p spackman
<stephenacm.org>

 2)
 Date: Tue, 09 Apr 1996 11:54:31 CDT
 From: aeulenbeindiana.edu (Alex Eulenberg)
 Subject: Economy, Minimalism, and Formalism, re: 7.517
	
This post makes a very common error: it is a refutation of
"gradualism" rather than a refutation of "selection", and it falls
into many of the common errors about evolution.

First: evolution occurs when a piece of data is "codified" into the
next generation. There are two basic areas of codification: location
and genetics. Most evolution is currently concerned with genetic
evolution rather than locational, or more accurately, ecological
evolution.

The steps of evolution must be small, however a small change in DNA
can have a large change in the final organism, displace one base pair
in one protien and you can have a functional genetic code become
unfunctional. The difference, litterally, between life and death. The
traits, that we see, are not the things which change gradually, to
think of gradual trait change as evolution is known in evolutionary
circles as "Lamarkian" evolution. It is a basic error.

Secondly genetic changes can pile up between large trait changes. This
is, in the language of xaos theory, known as "emmergance", that is
order seemingly suddenly arising, because what has really happened is
that a large number of changes have suddenly become dependant or
focused upon a much smaller area, changes in that area have a much
greater impact on the system as a whole. To see emmergant behavior in
action, take a pile of iron shavings and drop them on a pice of
paper. Benath the table rapidly mobe up a powerful magnet.

"Suddenly" the fillings will fall into a pattern. What has really
happened is:

1. the fillings are organized by the magnet.
2. the magnet is organized by your hand.
3. your hand is controlled by the motor sections of your brain.

Thus the movement of a few electrons in your brain can organize a much
larger numbe of electrons in the iron fillings, and so on. When seeing
macro evolutionary changes one should look for the single protien
change which triggers them and then follow the protien changes which
took place to get to that point.

This sort of non-trivial change has been seen, and does not require
any macro external force because the system is internally capable of
emmergant behavior. (There is a proof you can do out for this one if
you care to).

In order to follow evolution then, one needs to trace the micro steps
of *protien* and then see how those micor steps can produce, under
certain conditions, macro-steps of trait.

 - -

I submit that language is such an emmergant macro step, that it is
compose of individual sections of the brain, and therefore of the
protien formation mechanisms and receptor patterns that create those
sections, each one of which can evolve to the point where emmergence
is possible without language at a high level being the selecting
factor.

Once linguistic mechanisms reach this point, all that is required is a
micro (protien) mutation which causes certain very small regions of
the brain to organize the linguistically attuned areas into a larger
structure. Once this happens language is so useful that it can easily
be seen to be a large advantage to those croups of humans that
uniformly possess it. I submit that the fossil record bears this out,
we see a long period with skulls of early sapiens (which I label
"eos") with an admixture of traits found in later, branches, and now
extinct ones, along with clearly erectine features. In other words a
hodge-podge.

Then very rapidly a few skull types crystalize out: one lineage
clearly leads to neaderthal, the other to moderns. What has happened?
There were a welter of forms, each representing an admixture of
different proportions of brain section growth. At some point a few of
these forms became organized around a few small areas of the brain,
and thus codified, the small areas, being small, had little ability to
adapt to a welter of different organizations, and thus could only
organize effectively within certain patterns (two main ones and a host
of subsidiary ones), anything outside of this "fractal cone" of
possibilities yileds no lingusitic ability, or severely hampered
lingustic ability.

We can see the boundaries of this cone in action today based on
lesion, stroke and malformation of brain parts. Certain other forms
which are radically different in brain topology can also fall under
this fractal cone: most specifically ones where highly excited regions
of the brain map all of the functions normally spread over the entire
brain. I know of only low level examples of this in stroke recovery,
but there is no reason, in principle, that even people with large
amounts of brain lesion/disfunction, could not also fall under this
fractal cone.

The probabilty as the amount of brain material disfunction occurs
drops rapidly, also people whose divisions of cognitive function are
more specialized (ie men in general, and anyone exposed to larger
amounts of androgens in the critical fetal period where brain
specialization occurs) will have a harder time recovering from lesser
lesions, but greater probability of recovering from larger lessions.

- -

I am sorry to bring in so much that is normally outside of the range
of linguistics proper: however it seems that there are some
misconsceptions both as to the functioning of evolutionary systems in
general, and emmergant linguistic centers in particular that need to
be pointed out. The outline presented above is meant as merely a back
drop: it does not favore either side of the question of lingusitically
specific mechanisms, we often see organs with multiple functions
become devoted to one later on, and conversely we often see sinlge
purpose organs/systems widden their base later.

This means that there could be sections of the brain, which whatever
their previous function now do language and only language. Similarly
there could be sections of the brain that developed almost exclusively
for language, which now do other things as well. Only PET scans and
neurobiology will tell for sure.

But this, in itself, should be reassuring to lingusitics in general:
we can begin to attempt to map the sorts of brain pattern functioning
that we would expect to see given certain lingusitic to general
cognition pairings, and then test to see whether that pattern does
indeed come up, find our errors and cycle through again. It also means
that we can come to general definitons of root function, since we can
make definitions based on the universal mappings of brain function,
rather than structural functions in some particular language or
language group.

Further we can also extend lingustic analysis into other modes of
behavior where we can see that the organizing sections of the brain
are linguistically oriented and we would then expect them to organize
their section of the activity along lines similar to language, and
thus create patterns of behavior outside of an individuals language,
which never the less relate clearly to their native language.

Message 3: Re: 7.533, Disc: Economy, Minimalism, Formalism

While I do not believe LINGUIST List is an appropriate forum for the
discus- sion of evolution vs. "creation science", I'd like to
recommend to the sceptics like Alex Eulenberg that they consider *The
Beak of the Finch* by Wiener. This is mostly the story of a team of
ornithologists who have studied the Galapagos finches for many years
and, literally, observed evolution in action; it also describes a
number of other studies showing evolution at work. I do not know the
Denton book; but typically, attacks on evolution claim that because
biologists argue about the details, the entire observed edifice is
shaky. In essence, "evolution" is a fact; numerous theories are
proposed that describe it more or less accurately, and, just as in any
science, progress is made by accumulating disproofs, stimulating
refinements, etc. Evolution hap- pened, and contunues to
happen. Antibiotic-resistant bacteria; killer bees; the lemurs of
Madagascar.

At the risk of getting off the subject, I feel compelled to respond to
Alex Eulenberg's comments on Darwinism.

If the main point is that synchronic linguistic theory need not be
"compatible" with Darwinian theory, I agree, trivially, because I feel
the purview of the theories are orthogonal.

But I must disagree with all of Eulenberg's subsequent points. In
particular, I firmly believe:

1. Darwinism (evolution through inherited random variation and natural
selection) is quite capable of accounting for all observed differences
between species.

2. Darwinism is very well supported by direct and indirect evidence
from paleontology and biology.

3. Darwinian theory is the result of standard scientific method of
inference, involving (as do all empirical sciences) theories that are
in principle falsifiable by observation, and experiments whose results
agree with the predictions of the theory.

I hope this list is not coopted by debate on these points (I believe
they are basically irrelevant to most linguistic research), but I
could not leave Eulenberg's statements unchallenged, as this could
lead to the incorrect impression that his views were generally
accepted.

-A