Re: Dunbar Grooming

From: Stevan Harnad (harnad@coglit.ecs.soton.ac.uk)
Date: Fri Dec 03 1999 - 08:58:02 GMT


On Fri, 3 Dec 1999, Thuma Orsolya wrote:

> It also means that we, human beings, are also
> constrained in our capacity to form integrated
> groups. What are the specially cognitive limitations?

To a first approximation, it seems to be remembering
who did and who owes what to whom, who dominates whom,
etc. Dunbar feels (probably rightly) that this cannot
all be remembered through passive vision and hearing
alone. There must be an interaction that rehearses
and practices these relationships -- and that that
interaction is nonhuman primates in grooming, and in us
is chatting.

> What about the complexity of the relationship
> between all the individuals in the group and
> the different interrelated levels of groupings,
> e.g. family, dominance hierarchy, friendships?
> How do they load the cognitive capacity?

See the above. And note that even as an explanation
of grooming, the group-size/cognitive-capacity
hypothesis is merely a hypothesis -- though a rather
more probable one than its generalization to language.

> > from an individual animal's point of view, current
> > neocortex size sets a limit on the number of
> > relationships that it can maintain through time,
> > and hence limits the maximum size of its group.

Tell that to Don Giovanni!

> Earlier Dunbar claimed a correlation between
> brain size and group size. Now he says that
> it is a causal relationship.

You are right; this is an inference: How good do
you think his evidence and arguments are in support
of this causal interpretation? Can you (or anyone)
think of alternative interpretations?

> Even if the reader has some reservations about the
> causal link between brain size and group size, the
> correlation permits us to predict any of the two factors
> knowing the value of the other factor.

Correct. And these are the hasards of correlational thinking,
especially pairwise correlational thinking among a small
set of (perhaps arbitrary) variables one picks out to consider.
Once interpreted causally, these variables always yield self-
fulfilling prophecies.

To see whether this is just a just-so story, each of you should
think of alternative possibilities for the basis of the correlation
(unless you think it too is too weak to be taken seriously) as
well as its relationship to the origin of (1) grooming, (2) language.

> Grooming in the case of primates means fiddling in the
> fur of the other animal, looking for insects, so it
> has both a cleaning and a socal function.

But do you believe the cleaning function is primary?

> So the bigger a group, the more time is spent on
> grooming. There is a correlation between group size
> and grooming time, but we still do not know if it is
> a causal relationship. If time is a limiting factor,
> why do we need to refer to cognitive factors?

First, causal theories try to make sense of mere correlations.
If you doubt Dunbar's causal theory, the best way to
support your doubts is to provide equally plausible
alternative causal interpretations (possibly bringing
in more neglected variables).

Second, time is definitely a cognitive limitation. There
are many things one cannot do in real-time ("on-line" as
it is now called in computer terms) that one can do "off-line,"
given enought time to rehearse and reflect on it. The
off-line storage, rehearsal, recoding and reflective capacities
involved are indeed cognitive.

(The online/offline dichotomy will arise again, in our discussion of
language origins: Reflect, for now, as an example, on the number of
items for which one can judge their "numerosity" [how many of
them there are] by "subitizing" [by just looking and immediately
perceiving, hilistically, how many there are] and by COUNTING.
Counting takes time, it requires symbols and a code, abstraction
and re-coding; but once available, it allows rehearsal, and the
remembering of quantities in ways that subitizing made impossible --
except in the case of certain autistic exceptions: See the other
Student Skywriting discussion threads in the other archives about
Luria's "The Mind of a Mnemonist" and Borges's "Funes the Memorious"
for the cognitive advantages and disadvantages of online/offline
processing, abstraction, remembering and forgetting.)

> > However, the relationship between group size and
> > time devoted to grooming appears to be a
> > consequence of the intensity with which a small
> > number of key "friendships" (the primary network)
> > is serviced rather than to the total number of
> > individuals in the group.
>
> Is this remark not in contrast with the earlier
> claim that is in only the total number of relationships
> that matter?

It is not clear why: This may be a matter of coarser vs.
finer gained analysis and abstraction.

> Within the larger group smaller coalitions
> or friendships can be identified where the individuals
> groom relatively more than in other realtionships.
> What explains then the observation that the correlation
> between group size and grooming is linear?

I don't quite understand your objection. The linear
correlation is empirical, but frequency-independent.
The close groomings are a subset of the general groomings.

But maybe alternative interpretations on the correlations
will cast light on this: Does anyone have any alternative
interpretations?

> But does he have evidence that these groups are really
> based on personal knowledge of all the members? And what
> is really special about this level of grouping among
> smaller and larger units beside the fact that it smashes
> with the number he predicted?

I agree that it does seem to be picked arbitrarily out of a
hat. But there is a (weak) methology to this kind of inference
from correlations: It is open to counter-interpretations,
just as theories are open to refutation by testing.

Your doubts are hence better formulated RELATIVELY, by giving
one or more at least equally plausible rival interpretations,
rather than absolutely, by simply asking "why should I accept
this interpretation." (Even a wrong causal interpretation is
better than none, because it ponts to where to look next, and
at least it SEEMS to make sense of otherwise meaningless
correlations.)

> Thus, Dunbar's conclusion is that it is language
> that makes it possible for us to overcome the
> limitations of physical grooming and as a result
> permitting larger groups to maintain.
> Language is a symbolic means of communication,
> we can talk about things that are not present,
> talk about other people to follow what is happening
> to them even when we cannot see them. And speech
> is transmitted auditorily, it does not require
> physical contact between the interacting individuals.
> This permits the speakers to persue other activities,
> thus to save time.

This is the critical core of the paper. What do you think
of this? Is it the right account of the functional utility
of language, why it is so important and useful, why it
evolved and how it involved (the transition scenario)?

> > My suggestion, then, is that language evolved as a
> > "cheap" form of social grooming, so enabling the
> > ancestral humans to maintain the cohesion of the
> > unusually large groups demanded by the particular
> > conditions they faced at the time.
>
> Finally, Dunbar concludes that the time saving nature
> of linguistic grooming was the main selective pressure
> on the evolution of language.
> What does he mean by "cheap" form?

That's easier to answer than the other questions. He means
it accomplished everything grooming had accomplished, and
a lot more, with minimal energy and an easy transition form
what was before.

> Does language really serve a grooming function?

It is almost trivially true that it does. (You only need to go to a
hairdresser or manicure studio to see that.) But the real question is,
was that it's ORIGINAL and PRIMARY function, and was that the key to its
adaptive value and success?

> > The acquisition and exchange of information about
> > social relationships is clearly a fundamental part
> > of human conversation. I suggest that it implies
> > that this was the function for which it evolved.
>
> This statement is in contrast with traditional ideas
> of language evolution, which originate language from
> the need to communicate factual information and to control
> common actions, such as hunting.

Indeed it is: and what do you make of that?

> highly intellectual communities behave in the same
> way as the groups we know from our ordinary life
> as far as the content of talking is concerned.

Yes, but since it is obvious that language today,
AMONG OTHER THINGS, is indeed our form of social
"grooming," this universality in what people chat
about and when is less surprising. The real question
is: Was that the original and essential function of
language, the one that drove its evolution?

> However, to put all the pieces of Dunbar's arguments
> we should know if there was any evolutionary
> advantage of larger groups in our ancestors' life?

That is accepting the critical part of his causal story as given.
Groups could be an advantage yet the hypothesis about language origins
could be completely wrong.

--------------------------------------------------------------------
Stevan Harnad harnad@cogsci.soton.ac.uk
Professor of Cognitive Science harnad@princeton.edu
Department of Electronics and phone: +44 23-80 592-582
Computer Science fax: +44 23-80 592-865
University of Southampton http://www.cogsci.soton.ac.uk/~harnad/
Highfield, Southampton http://www.princeton.edu/~harnad/
SO17 1BJ UNITED KINGDOM



This archive was generated by hypermail 2b30 : Tue Feb 13 2001 - 16:23:06 GMT