Study Suggests Monkeys Have Ability to Think
By NICHOLAS WADE
Animals do not have language and therefore, savants from Descartes
onward have held, cannot think, because words are coin of thought. A
strong hint that animals can think, but in ways that do not involve
language has emerged from a study of Rosencrantz and MacDuff, two
rhesus monkeys whose cognitive skills are being explored at Columbia
University.
The study shows the monkeys can arrange things in serial order, an
ability that is presumably part of the human heritage too and that
could be the origin of the ability to handle syntax, the rules for
arranging words in meaningful order in a sentence.
Pressing at panels on a computer screen, the two primates have shown
in repeated trials that they can touch, in correct numerical order,
displays containing one to nine objects, according to a report in
Friday's issue of Science.
This is not the same as being able to count from one to nine, because
people when they count use the abstract concept of number. Strictly
speaking, Rosencrantz and MacDuff have so far demonstrated only that
they can recognize arranged numerical sets in correct order, like
someone saying "First, second, third."
The monkeys may also have the ability to think "One, two, three," but
that remains to be proved, said Elizabeth M. Brannon and Herbert S.
Terrace, the two psychologists who are studying them. They plan to
train the monkeys to associate sets of a given number of objects with
a specific texture. If the animals can symbolize numbers in this way,
it would suggest they conceive of cardinal numbers as people do.
Exploring how animals think is important because it promises to shed
light on the abilities from which the human mind evolved. Dr. Terrace
is well known for his experiment 20 years ago with Nim Chimpsky, a
chimpanzee named in respectful reference to the linguist Noam Chomsky.
At the time, ape-language specialists believed that chimpanzees
trained with pictures or in sign language could string symbols
together in sentences, indicating the ability to use syntax. Dr.
Terrace argued there was no evidence for true syntax in the strings of
signs Nim put together, a view that became widely applied to the other
apes, too.
Chimpanzees could be trained to press buttons for the four symbols
"Please machine give apple," but whatever fanciful interpretation
human observers might put on the event, in Dr. Terrace's view, the
chimp was simply placing four symbols in an order it knew would bring
some reward.
"I felt I could teach a pigeon to do that, and I did," Dr. Terrace
said. "Then I realized it was not a negative but a positive
experiment. The question of can animals learn language was replaced by
the question, 'Can animals think without language?' "
Dr. Terrace started to explore monkeys' adeptness at putting things in
sequence, believing the ability might be an important evolutionary
precursor in the primate line of the ability that in humans led to
language.
One day this week, when Rosencrantz and MacDuff were not working, Dr.
Terrace showed a videotape of them touching a sequence of panels
displayed on a computer monitor. The monkeys seemed absorbed in their
task, which they performed as deftly as if it were something a rhesus
did every day.
The ability to place things in order might indeed be useful to monkeys
in the wild, Dr. Terrace said, for example, in figuring out their
position in their troop's social hierarchy.
Although many other psychologists have studied animals' ability to
handle numbers, Dr. Brannon's and Dr. Terrace's new experiment is
important because it excludes the many possible confounding factors
"better than any other paper I know," said Dr. Susan Carey, an expert
on number learning at New York University.
Dr. Carey believes that animals possess several different
number-representing systems. Rats, for example, can evidently
represent numbers as large as 45 in their minds because they can be
trained to press a bar about 45 times and then switch to another bar.
Human infants know that one is less than two and two less than three
but cannot represent larger numbers. Presumably, infants are using a
different system, or mental module, from the one rats use to decide
when they have reached 45 and that people use to list the cardinal
numbers.
Although the list of whole numbers might seem an integral part of
everyone's vocabulary, some languages have a very truncated counting
system that goes "One, two, many." That system may go back to the
earliest days of language, Dr. Carey believes, because it is woven
into the syntax of all languages in the form of the single and plural
forms of verbs. Many languages also had a dual form for expressing the
actions of two people.
It remains to be discovered how the large number mental module, the
one-two-many module and the list of integers module are related to
each other and to the syntax module.
"That is a question in evolutionary history," Dr. Carey said, "and the
Brannon and Terrace experiment tells a part of that story in a very
elegant way."
Copyright 1998 The New York Times Company
*****************
Wade T. Smith
morbius@channel1.com | "There ain't nothin' you
wade_smith@harvard.edu | shouldn't do to a god."
******* http://www.channel1.com/users/morbius/ *******