Re: Milk Bottles & Animal IQ

Ton Maas (
Wed, 17 Dec 1997 09:17:23 +0100

Message-Id: <v03102805b0bd370ae9f4@[]>
In-Reply-To: <>
Date: Wed, 17 Dec 1997 09:17:23 +0100
From: Ton Maas <>
Subject: Re: Milk Bottles & Animal IQ

>> This may seem an overly simplistic question considering the erudite
>> nature of this debate, but how come birds can count at all when humans
>> have to be taught this skill?
>I don't know if birds can count but rather have a measure of size and
>distribution and so may have trouble with exact numbers. Comparable to
>having a glass of water and if you pour some of it out you can be
>possitive that there is less water in the glass and if you pour a lot of
>it out you can guess that it could be empty. But then again, I don't
>know any of the actuall studies, I'm just guessing from what I have read
>in this newsletter (I think it might be called Heuristic Strategy).

In "MInd & Nature" Gregory Bateson describes Koehler's experiments:

"Some birds can somehow distinguish number up to seven. But whether this is
done by counting or by pattern recognition is not known. The experiment
that came closest to testing this difference between the two methods was
performed by Otto Koehler with a jackdaw. The bird was trained to the
following routine: A number of small cups with lids are set out. In these
cups, small pieces of meat are placed. Some cups have one piece of meat,
some have two or three and some have none. Separate from the cups, there is
a plate on which there is a number of pieces of meat greater than the total
number of pieces in the cups. The jackdaw learns to open each cup, taking
off the lid, and then eats any pieces of meat that are in the cup. Finally,
when he has eaten all the meat in the cups, he may go to the plate and eat
the *same number* of pieces of meat that he got from the cups. The bird is
punished if he eats more meat from the plate than was in the cups. This
routine he is able to learn.
Now the question is: Is the jackdaw counting the pieces of meast, or is he
using some alternative method of identifying the number of pieces? The
experiment has been carefully designed to push the bird toward counting.
His actions are interrupted by his having to lift the lids, and the
sequence has been further confused by having some cups contain more than
one piece of meat and some contain none. By these devices, the experimenter
has tried to make it impossible for the jackdaw to create some sort of
pattern or rhythm by which to recognize the number of pieces of meat. [...]
It is still conceivable, of course, that the taking of the meat from the
cups becomes some sort of rhythmic dance and that this rhythm is in some
way repeated when the bird takes the meat from the platter. The matter is
still conceivably in doubt, but on the whole, the experiment is rather
convincing in favor of the hypothesis that the jackdaw is counting the
pieces of meat rather than recognizing a pattern either of pieces or of his
own actions."


This was distributed via the memetics list associated with the
Journal of Memetics - Evolutionary Models of Information Transmission
For information about the journal and the list (e.g. unsubscribing)