Reflectorites
Here is a short article from the Science & Technology pages of The Economist
a month or so back. It reports on experiments with chimps which indicate that,
unlike human children, chimps don't view humans (and therefore other chimps)
as having minds.
This suggests that chimps can be trained to emulate human behaviour which
looks like it is based on human-like intelligence, but it is really just
trial-and-error learning, like that which some other intelligent animals can do.
Steve
=====================================================================
[...]
Animal behaviour
Chimps or chumps?
ONE of the trademarks of being human is an understanding that others also
have beliefs, intentions and desires. It has been called mindreading, or a
"theory of mind", and it is what allows people to work together, to
manipulate others, and even to deceive each other. The current belief is
that the mindreading abilities of humanity's closest relatives, the great apes,
are not fundamentally different from those of people; they are simply not as
highly developed. But work just presented to a meeting of the Association
for the Study of Animal Behaviour, at London Zoo, suggests this belief is
wrong.
According to Daniel Povinelli of the University of Southwestern Louisiana,
the idea that an ape understands that another ape has any kind of mental
life is an illusion. Their behaviour may seem to mimic that of people, but it
is based on an entirely different understanding of the world-one that does
not involve viewing others as psychological agents in their own right.
Dr Povinelli and his colleagues looked at a very simple test of mindreading,
the ability to understand seeing. When a person looks in a certain direction,
children as young as three interpret that movement of the head or eyes in
terms of the person's underlying psychological state-what the person is
attending to, and what he intends to do next-and act accordingly. They
are unlikely, for instance, to make a gesture towards someone who is not
paying attention to them.
To find out whether chimps did the same, Dr Povinelli's team confronted
the animals with two experimenters. By extending an arm through a hole in
a perspex cage towards one or other of the two, the chimp could expect a
reward of an apple or a banana. But in a series of subtle variations, the
experimenters' positions were altered so that only one of them could see
the chimp. One experimenter was blindfolded while the other wore a
blindfold over her mouth; one wore a bucket over her head while the other
did not; one covered her eyes with her hands; one presented her back to the
chimp.
The results were startling. In three out of the four sorts of test, the chimps
made their begging gestures just as frequently to the experimenter who
could not see them as to the one who could. Occasionally, after receiving
no reward from the "unseeing" one, they would gesture again, as if puzzled
by the lack of response. Only in the case when an experimenter turned her
back on the chimps did they stop gesturing to her.
That raised the possibility that the chimps might be reasoning on the basis
of some kind of postural cue-that a person's frontal aspect was a better
predictor of reward than her back-rather than assessing line-of-sight. To
test this, the two experimenters assumed identical positions, facing away
from the chimp but twisting their bodies to look back at it over their
shoulders. But only one of them actually turned her head to look at the
chimp. The second continued to look away. Once again, the chimps failed
the test, gesturing equally to both.
These findings suggest that chimps are unable to equate seeing with
knowledge in the way that human infants do. But that does not mean they
are stupid. After enough practice, when the "correct" response was
consistently rewarded, the chimps learned that front was better than back,
face was better than back of the head, eyes open was better than eyes
closed. Soon, they were reacting to the eyes just as a young child might.
Their behaviour, however, was based on trial-and-error, not on an
understanding of seeing. And when they were tested again a year later, they
had forgotten this behaviour.
Chimps seem to be able to hoodwink each other, and can follow the gaze
of another chimp or human who has been distracted by something
interesting. But according to Dr Povinelli, it is possible that in chimps,
unlike in humans, those behaviours evolved separately from an ability to
read another's thoughts. So it may be that our closest relatives do not share
our empathy, but are instead "mindblind".
84 THE ECONOMIST DECEMBER 11TH 1999
("Chimps or chumps?", The Economist, December 11th, 1999, p84)
=====================================================================
--------------------------------------------------------------------
"Older folk in the know told me that selection didn't operate to make
complicated things out of complicated things, only to make complex things
out of simple ones. I couldn't understand how anything of the sort could be
true, because, unlikely as it was, it would surely be less difficult to make a
rabbit out of a potato than to make a rabbit out of sludge, which is what
people said had happened, people with line after line of letters after their
names who should have known what they were talking about, but
obviously didn't." (Hoyle F., "Mathematics of Evolution", [1987], Acorn
Enterprises: Memphis TN, 1999, p2)
Stephen E. Jones | sejones@iinet.net.au | http://www.iinet.net.au/~sejones
--------------------------------------------------------------------
This archive was generated by hypermail 2b29 : Thu Jan 27 2000 - 08:55:45 EST