408 Animal Memory and Cognition
humans have an innate “theory of mind module.” According
to Baron-Cohen, autistic children, some of whom seem not to
be aware of others (as evidenced by their sitting alone, rock-
ing back and forth, and in other respects living in a private
world), lack a theory of mind.
Premack and Woodruff (1978) asked whether the chim-
panzee has a theory of mind. This question has sparked
considerable research and much controversy over the last
20 years or so. C. M. Heyes (1998) concluded that no convinc-
ing evidence has been produced to suggest that chimpanzees
have a theory of mind. Reaction to her criticisms has been var-
ied. At one extreme, Gordon (1998) suggested that the ques-
tion itself is ill conceived and thus not worth asking. Byrne
(1998) suggests, on the other hand, that Hayes misrepresents
findings and that the theory-of-mind approach is a useful one.
Heyes (1998) identified six areas of investigations empha-
sized in theory-of-mind research: imitation, self-recognition,
social relationships, rule taking, deception, and perspective
taking. In this chapter, two of these areas will be discussed in
enough detail to give, hopefully, an adequate idea of what is
intended by the termtheory of mind. These areself-recognition
andimitation.
Self-Recognition
Gordon Gallup (1970) presented chimpanzees with mirrors.
Initially the chimpanzees reacted to the mirror images as
though they were other chimpanzees. Following hours of
experience with mirrors, the chimpanzees dropped other-
directed behavior in favor of what Gallup termedself-directed
behavior. These self-directed behaviors were interpreted by
Gallup to indicate that a chimpanzee recognized the image
in the mirror as itself. To provide better evidence for self-
recognition, Gallup devised themark test: He anesthetized the
animals and marked then with an undetectable (i.e., odorless)
dye over one eye and the opposite ear—areas that could not
be seen without the aid of the mirror. The basic finding was
that chimpanzees that had mirror experience showed mark-
directed behavior, whereas control chimpanzees lacking mir-
ror experience did not.
Gallup’s (1970) initial conclusion was that chimpanzees
are capable of recognizing themselves and therefore have a
sense of self-awareness. Subsequently, Gallup (1977) ex-
tended his conclusions. The ability to self-recognize, Gallup
suggested, implied consciousness and self-consciousness, the
latter encompassing the ability to think about thinking and to
be aware of one’s own state. In 1982, Gallup went still fur-
ther. An animal that is self-aware, he suggested, has a mind,
and having a mind includes having empathy and the ability to
deceive.
Human children, of course, have passed the mark test, as
have some orangutans. After several failures, gorillas have
been shown to pass the mark test. In general, monkeys fail
the mark test, and even in instances in which behavior has
been directed at the mark, the observation is equivocal. A va-
riety of additional findings have been reported: Not all chim-
panzees pass the mark test; young chimpanzees (below age 3
years, 6 months) may prove likely to fail the mark test.
Epstein, Lanza, and Skinner (1981) claim to have trained a
pigeon to pass the mark test, a claim that has been disputed
(e.g., Gallup, 1982).
Criticism of the self-recognition claim ranges from the ob-
servation that failing the mark test may not imply a dimin-
ished mental capacity, to the observation that passing the
mark test may not indicate advanced mental capacity. As an
example of the former, it has been observed that monkeys
may not look in the mirror because eye-contact is a threaten-
ing gesture. As an example of the latter, mirror recognition
may imply no more than that the animal has a “body con-
cept,” one that is used in, say, ordinary locomotion.
A test similar to the mark test has been employed with
children; it is called the rouge test. Children are given a small
rouge mark below the right eye that can be recognized only
by using a mirror. By about 19 months of age, 52% of chil-
dren immediately direct behavior to the rouge, indicting self-
recognition (Asendorpf, Warkentin, & Baudonniere, 1996).Imitation and Social LearningBy observing another engage in some extended act that in-
volves a number of different and discrete steps, a person may
learn in a matter of minutes what might otherwise require
hours or days of individual effort without guidance. Any
number of such activities comes to mind, ranging from
changing a tire to setting a VCR. Perhaps because learning by
observing others is so important and widespread in the
human species, some have taken it to be a hallmark of intelli-
gence in other species. For example, Romanes (1882) pro-
vided a number of rich examples of animals’ engaging in
quite complex behaviors established by imitation. The prob-
lem is that Romanes’s examples were based on anecdotes and
thus his data by modern and entirely reasonable standards are
deficient.
Determining the extent to which other species learn by im-
itation is a more difficult problem than it appears to be at first
blush. For example, what sort of behaviors should be selected
for analysis? Well-fed chickens can be induced to continue to
eat by watching other chickens feed. It would seem that
species-typical behavior, such as chickens’ pecking for food,
would provide relatively unconvincing evidence for true