How should scientists react to anthropomorphism
(defined for the purposes of this paper as the attribution of mental
states or properties to nonhuman animals)? Many thoughtful scientists
have attempted to accommodate some measure of anthropomorphism in their
approaches to animal behavior. But Wynne will have none of it. We reject
his argument against anthropomorphism and argue that he does not pay
sufficient attention to the historical facts or to the details of
Although his main target is anthropomorphism, Wynne
also displays ambivalence towards current human cognitive psychology. He
states that it has resurrected several mentalistic concepts, and while
he allows that cognitive psychology uses introspection to generate
experimentally testable hypotheses, he suggests, "there are grounds for
criticizing this practice." Nevertheless, he continues, "the issues are
not as extreme as when this method is applied across species." Like most
scientists and philosophers, we agree that inferences about animal minds
require special handling, but we believe that Wynne’s arguments commit
him to the much stronger conclusion that all mental state attributions,
even within our own species, are beyond the ken of "objective
materialistic science." We resist the temptation to speculate about why
he thinks this. Instead, we look only at what he wrote. Consider the
1. "[a] anthropomorphism is a form of mentalism,
and [b] as such is not amenable to objective study. [c] Labeling
animal behaviors with everyday terms from lay psychology does not
explain anything. Rather it is an example of the nominalist fallacy"
2. "Mentalism fails to qualify as a scientific
explanation for (at least) two reasons. [a] First, it uses
ultimately non-material causes to attempt to explain behavior....
[b] Second, mentalistic concepts are intrinsically private and thus
by definition subjective, not objective" (p. 132).
We have defined anthropomorphism in such a way that
1a is true. We will deal with 1b, amenability to objective study,
further below when we come to 2b. 1c is true but misleading. It is true
because the act of labeling never explains anything. It is misleading
for none of his main contemporary protagonist authors (Bekoff,
Burghardt, and de Waal) has ever indicated that merely labeling animal
behaviors mentalistically explains those behaviors. Nor do the direct
quotations from these authors that Wynne supplies in his section on
"Modern Anthropomorphism" establish his claim that they are guilty of
the nominalist fallacy. We distinguish three possible roles for
anthropomorphism in the science of animal behavior: (i) explaining
animal behavior, (ii) as sources of hypotheses about the causes of
animal behavior, and (iii) as targets of explanation in their own right.
In discussing the nominalist fallacy, Wynne attacks the first of these.
The three scientists mentioned by Wynne in this context consider
anthropomorphism in its role (ii), as a source of scientifically
testable hypotheses, and they endorse role (iii), treating mental states
in animals as legitimate targets of scientific investigation and
explanation. But it is only after such investigation that any of them
would endorse the explanatory role (i). Theirs is no mere labeling
strategy, but an attempt by thoughtful scientists (we assume they do
have mental states!) to understand how their thinking about animals can
be usefully informed by thinking about human mentality.
We are compelled to note that in his response to
Bekoff, Burghardt, and de Waal, all of whom have distinguished records
of scientific publication, Wynne nowhere assesses their claims in any
detail. Instead, he writes as though it is sufficient to apply the label
of anthropomorphism to their views while mounting some arguments
against anthropomorphism. Insofar as thoughtful defenders of
anthropomorphism endorse anything like an explanatory role for
mentalistic terms, it is after going through hypothesis and experimental
investigation. This is just as true of the historical figures Wynne
criticizes as their modern counterparts. We will argue below that Wynne
is no more successful in representing the history of anthropomorphism in
comparative psychology than he is in making his case against
anthropomorphism, which we turn to next.
Wynne’s objections to mentalism in statement 2 above
would, if correct, present equal difficulties for mental concepts
whether used for humans or nonhuman animals. Words are not immaterial
causes made flesh. What, then, of Wynne’s reasons for banishing mental
state attributions from science? With regard to 2a, his charge that
mentalism entails immaterial causes, we can think of only two
possibilities. Either Wynne would accuse all cognitive scientists of
being dualists, or he is committing a version of the genetic fallacy.
The scope of his reactions to cognitive science is difficult to discern
given that he did not articulate his "grounds for criticizing" the use
of mentalistic notions in human cognitive psychology. But given all the
extremely thoughtful work that has gone into providing a materialistic
underpinning for cognitive science over the past 50 years (not to
mention the long history of materialistic theories of mind, including
Descartes’ contemporary Hobbes), we should not just take Wynne’s word
for it that mentalistic terms are unavoidably committed to immaterial
causes. If it is perfectly consistent to think, as many scientists do,
that mental states can be understood in neurofunctional terms, then
Wynne’s complaint comes down to the dubious claim that we should now
throw out mentalistic terms because they were originally associated with
a dualistic worldview. This is the genetic fallacy. Just because the
terms used to be understood in that way, it does not follow that we must
regard them now as inadmissible. It would be similarly fallacious to
claim that it is unscientific to consider hypotheses about the
biological functions of the liver on the grounds that functions,
pre-Darwin, were associated with now-discredited teleological,
vitalistic, or creationist views. Wynne commits the very same fallacy
when he writes that it is a mistake to try to use the word
‘anthropomorphism’ positively because of its origins in the 13th century
as a label for a mistaken theological view. One might as well reject the
term ‘Big Bang’ in physics because it was invented by Hoyle to mock a
theory that he believed rested on a mistake.
Wynne’s mention of the subjective/objective contrast
(1b and 2b) provides the more interesting challenge: namely, how to
bring mental states into the scientific fold. We will argue for two
points about this. First, Wynne’s history of anthropomorphism distorts
the attempts of earlier thinkers to meet that challenge. Second, with
respect to current science, this is a challenge that cognitive
scientists and neuroscientists are vigorously pursuing. Wynne ignores
the details of this scientific work. The result in both cases is that he
commits the straw man fallacy.
Wynne presents the history of comparative psychology
as one of decreasing appeals to anthropomorphism. But the story is more
complicated than he lets on. For instance, although Darwin used a lot of
anthropomorphic anecdotes in his effort to establish continuity, he also
took mental states to be target of empirical investigation (role iii
above). In his final book he described experiments aimed at
understanding the intelligence of earthworms with respect to the
petioles and other objects they use to plug their burrows (Darwin, 1881;
see also Crist, 2002).
Romanes collected many anecdotes, but he also tried
to justify this practice. Wynne describes nothing of Romanes’s attempt
to struggle with these issues, choosing instead to simply label him as
credulous and gullible because he included reports from "persons bearing
names more or less unknown to fame" (Romanes, 1883 as cited by Wynne).
In fact, however, many of the more unbelievable anecdotes in Animal
Intelligence come from known observers. It was Darwin who provided
Romanes with an anecdote that snails are capable of communicating
complex information to one another. Similarly, when a bishop reported a
kind of trial of a jackdaw by rooks, Romanes felt obliged to publish
(see Boakes, 1984, p.26). Wynne acknowledges Boakes as a careful
commentator, but instead of stopping to analyze the history carefully
himself, he jumps ahead ten years to Lloyd Morgan’s 1894 publication of
his Introduction to Animal Psychology.
The problem with this leap is that it causes Wynne to
neglect the fact that during the 1880s Lloyd Morgan was one of the most
important critics of Romanes and Darwin’s uses of anthropomorphism. In
1886 Lloyd Morgan published a short article in the journal Mind
in which he described the state of the science at that time as "a
chaotic mass of anecdotal fact and fiction" (Lloyd Morgan, 1886, p.
174). He began by quoting an anecdote from Romanes in which an orangutan
unties several knots in order to get a key. Lloyd Morgan criticizes
Romanes’s inference that the animal understood the nature of the
problem. He rejects Romanes’s anthropomorphic method of considering
one’s own mental states when performing similar activities. Lloyd Morgan
argued that such a method carried a risk of error even when applied
cross-culturally within humans, and he thought that the situation was
worse for animals, partly because of the absence of language. He argued
that the scientific study of animal intelligence should focus on the
habits and activities of animals and concluded that knowledge of the
subjective states of animals is unattainable.
By 1894, however, Lloyd Morgan was more willing to
employ anthropomorphism. Missing from Wynne’s story is any indication of
why Lloyd Morgan would change his mind. Lloyd Morgan’s reasons are still
relevant to current research, and they can be summed up in one word:
evolution. Lloyd Morgan may have been Romanes’s biggest critic, but he
was also Romanes’s great friend. Romanes’s use of anthropomorphism and
his views about the subjective states of animals were challenges to
Thomas Huxley’s automaton view of animals. Huxley (1874) believed that
the subjective states of animals were causally irrelevant for
understanding animal behavior: Consciousness was essentially an
epiphenomenon. Huxley believed that human subjective states were more
complex than those of animals. Romanes argued that if subjective states
are becoming more complex and are different among various species, then
they must be selected for. On this view, subjective states were causally
efficacious and could therefore explain behavior.
Lloyd Morgan would come to accept Romanes’s argument
that the subjective states of animals were in fact important to the
understanding of animal behavior, but not until he had experimental
evidence. After publishing Animal Life and Intelligence in 1890,
Lloyd Morgan was informed by T. Mann Jones that he had wrongly described
many of the behaviors of young chicks (Boakes, 1984). Lloyd Morgan had
relied on experiments that Douglas Spalding published in 1872 describing
the pecking behavior of chicks as completely unlearned and totally
instinctive. Upon repeating these experiments T. Mann Jones found that
there was a learned component. Lloyd Morgan repeated the experiments
himself and saw that he could not ignore the learned aspect of the
chick’s behavior—and he thought that he needed to appeal to the
experience of the chick in order to explain this.
Lloyd Morgan’s personal realization that seemingly
instinctive behavior was fine tuned by learning would lead him to write
his Introduction to Comparative Psychology (1903), where he first
stated his now famous canon. The historical details of how Lloyd Morgan
came to the canon help us to understand the acceptance of
anthropomorphism by perhaps the most famous psychologist in Britain of
the era. Wynne is right that Lloyd Morgan’s canon is not a blanket
prohibition against anthropomorphism. But he is wrong to insinuate that
Lloyd Morgan commits the nominalist fallacy. The canon is better seen as
a call for experimentation that can pick out where subjective states are
causally important. Lloyd Morgan famously gives the example of his own
dog opening a garden gate. The question that Lloyd Morgan poses is
whether or not the dog understands how the gate works. Lloyd Morgan does
not pretend to be able to know all of the qualities that might go into
the subjective state of understanding something, but he can experiment
and see whether the dog has general knowledge of how the gate works. He
tries to get his dog to open the latch in different ways, but the dog is
unable to open the latch except by using his nose in the way that it is
accustomed to. Lloyd Morgan concludes that the dog does not understand
how the latch works, but has associated lifting his nose against the
latch with the opening of the gate. This is an application of the canon.
However, had the dog been more flexible in its use of the latch, the
canon would have supported the conclusion that the dog understands
something about the way that the latch works. In this case, we would be
speaking of the mental states of the dog.
Of course, critics of such inferences are free to
propose (and test) alternative explanations. Detailed attention to
actual experiments is where the discussion of anthropomorphism is most
likely to be fruitfully located—not where Wynne takes it, with broad
generalizations about the demerits of anthropomorphism. Had Wynne
considered the works of Lloyd Morgan more carefully, he might have seen
this. Instead, his comments on the absence of a science of remorse do
nothing to show that anthropomorphism is generally unjustified. It fails
to engage with any actual science. Instead, we are supplied with Wynne’s
own anecdote-driven speculations ("I have noticed that…", p. 133) about
the possibility ("perhaps… perhaps… perhaps…", p. 133) of Pavlovian
explanations for the submissive behavior of puppies in the presence of
We do not rule out the power of such approaches to
explain some aspects of animal behavior. Neither do we deny that some
pro-anthropomorphism scientists sometimes provide more rhetoric than
substance. But, as we mentioned above, many scientists are vigorously
pursuing the challenge of applying cognitive approaches to animal
behavior, entirely within a non-dualistic framework. Specific scientific
work which makes use of the attribution of mental states to animals
would be worthy of analysis; for instance, de Waal’s experiments on
fairness in monkeys (Brosnan & de Waal 2003) or the experiments by Hunt,
Rutledge, and Gray (2006) and Weir and Kacelnik (2006) to test the
understanding of tools by New Caledonian crows. Metacognition, social
play, affective neuroscience, and the study of mirror neurons are all
areas where the attribution of mental states to animals has led to
interesting experiments. Instead of blanket complaints against
anthropomorphism, we enter a plea for more attention to the actual
arguments of philosophers of science who have shown that
anthropomorphism is not necessarily or always a logical mistake (e.g.,
Fisher, 1991; Sober, 1998; Keeley, 2004) and for more thoughtful
engagement with those scientists who have attempted to articulate their
own views about the various roles that anthropomorphism plays in their
empirical investigations of animal mind and cognition.
Boakes, R. (1984). From Darwinism to
behaviorism. Cambridge, UK: Cambridge University Press.
Brosnan, S. F., & de Waal, F. B. M. (2003).
Monkeys reject unequal pay. Nature, 425, 297-299.
Crist, E. (2002). The inner life of earthworms:
Darwin’s argument and its implications. In M. Bekoff, C. Allen, & G.
M. Burghardt (Eds.), The cognitive animal: Empirical and
theoretical perspectives on animal cognition (pp. 3-8).
Cambridge, MA: MIT Press.
Darwin, C. (1881). The formation of vegetable
mould through the action of worms, with observations on their habits.
London: John Murray.
Fisher, J. A. (1991). Disambiguating anthropomorphism: An interdisciplinary review. Perspectives in
Ethology, 9, 49-85.
Hunt, G. R., Rutledge, R. B., & Gray, R. D.
(2006). The right tool for the job: What strategies do wild New
Caledonian crows use? Animal Cognition, 9, 307-316.
Huxley, T. (1874). On the hypothesis that animals
are automata, and its history. Nature, 10, 362-366.
Keeley, B. L. (2004). Anthropomorphism,
primatomorphism, mammalomorphism: Understanding cross-species
comparisons. Biology & Philosophy, 19, 521–540.
Morgan, C. L. (1886). On the study of animal
intelligence. Mind, 11, 174-185.
Morgan, C. L. (1890). Animal life and
intelligence. London: Edward Arnold.
Morgan, C. L. (1903). An introduction to
comparative psychology (Rev. ed.). London: Walter Scott
Romanes, G. (1883). Animal intelligence.
New York: D. Appleton & Co.
Sober, E. (1998). Morgan’s canon. In D. D.
Cummins & C. Allen (Eds.), The evolution of mind (pp.
224-242). New York: Oxford University Press.
Spalding, D. A. (1872). On instinct. Nature, 6,
Weir, A. A. S., & Kacelnik, A. (2006). A New Caledonian crow (Corvus
moneduloides) creatively re-designs tools by bending or unbending
aluminium strips. Animal Cognition, 9, 317-334.