On Wed, 17 May 2000, Margareta Lutzhoft wrote:
> Dear Professor Harnad,
>
> I'm a MSc student (Computer/Cognitive science)
> and I've been reading your discussion lists (student skywriting)
> with great interest.
>
> I wonder if you could tell me where to find what you
> call 'the original critique' against the Cangelosi/Parisi paper,
> mentioned in the "1999-2000 Student Skywriting CogSci-Szeged99"
> since it might be useful for my master's project about the
> evolution of communication.
>
> best regards,
>
> Margareta Lutzhöft
>
> University of Skovde, Sweden.
Dear Margareta:
That critique was given orally in the seminar, but I'll repeat it
here:
In the Cangelosi/Parisi paper, the (predecessors of) (1) "toilers" and
(2) "thieves" were foragers who (1) learned which mushrooms were edible by
trial and error-correcting feedback or (2) learned which mushrooms
were edible by "hearsay" (overhearing the toilers say "edible": there
were plenty of mushrooms, so the toilers lost nothing by vocalizing).
In the competition, the "thief" strategy beat the "toiler" strategy, and
the proportion of thieves in the population grew.
But, as would have become evident if the simulation had continued for
several more generations, "theft" under these conditions is not an
"evolutionarily stable strategy." For once there were few or no toilers
left, no one would know what was edible. For toilers, the knowledge is
grounded in having learned the features of the edible mushrooms; so the
toilers can forage alone; for the thieves, the only "feature" of an
edible mushroom is the vocalization of the toilers: "edible". So when
there are no more toilers, the thieves cannot eat.
(Hence, across generations there would be a constant oscillation, with
thieves' proportions rising and falling; this is not a viable model of
language-origins because we are ALL thieves now, stably!)
The solution is always to have the bottom-level categories learned by
toil, by everyone. The categories acquired by "theft" need to be
higher-order categories, describable by (e.g., boolean) combinations of
the names of the lower-order categories, as in the Cangelosi/Harnad
paper.
Best wishes,
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 - 16:23:07 GMT