ID critic and University of Waterloo computer scientist Jeffrey Shallit takes exception to my observation in a recent post that animals lack the capacity for abstract thought. Abstract thought is the ability to conceive of universals — mercy, justice, mankind, logical and mathematical relationships, and the like. It is an immaterial power of the mind, and only human beings are capable of abstraction.

I wrote:

Human beings have mental powers that include the material mental powers of animals but in addition entail a profoundly different kind of thinking… Human beings think abstractly, and nonhuman animals do not.

Shallit disagrees:

I’m really curious to know how Dr. Egnor knows with certainty that nonhuman animals cannot think abstractly. I guess he is just egnoring all the research that suggests just the opposite. It’s not like this is hidden stuff; Egnor could read, for example, the books of Dutch primatologist Frans de Waal. Maybe not all neurosurgeons are this bats— loony… Maybe you have to be immersed in jeebus-juice to believe, like Egnor does, that “Human rationality is different because it is immaterial.” I guess our thinking powers are just magic; all that neurocircuitry is just there for show.

He cites three reports of experiments that are interpreted as showing abstract thought in animals. I’ll deal with the first experiment he cites.

A 2014 Scientific American article, “Many Animals Can Think Abstractly,” describes several research studies purporting to demonstrate abstract thought in animals. The first by Jennifer Vonk:

In a study published last fall in the journal PeerJ, for example, Oakland University psychology researcher Jennifer Vonk investigated how well four orangutans and a western lowland gorilla from the Toronto Zoo could pair photographs of animals from the same biological groups.

Vonk presented the apes with a touch-screen computer and got them to tap an image of an animal — for instance, a snake — on the screen. Then she showed each ape two side-by-side animal pictures: one from the same category as the animal in the original image and one from another — for example, images of a different reptile and a bird. When they correctly matched animal pairs, they received a treat such as nuts or dried fruit. When they got it wrong, they saw a black screen before beginning the next trial. After hundreds of such trials, Vonk found that all five apes could categorize other animals better than expected by chance (although some individuals were better at it than others).

After “hundreds of trials” and free-flowing “nuts and dried fruits,” the apes were able to “categorize” animals better than expected by chance — some apes apparently responded to the training better than others.

The author concluded:

The researchers were impressed that the apes could learn to classify mammals of vastly different visual characteristics together — such as turtles and snakes — suggesting the apes had developed concepts for reptiles and other categories of animals based on something other than shared physical traits.

The author cited a study in which researchers purported to show the same capacity for abstract thinking in dogs, chimps, bears, and pigeons:

They can reliably recognize pictures of other dogs, regardless of breed, as a study in the July 2013 Animal Cognition showed. The results surprised scientists not only because dog breeds vary so widely in appearance but also because it had been unclear whether dogs could routinely identify fellow canines without the advantage of smell and other senses. Other studies have found feats of categorization by chimpanzees, bears and pigeons, adding up to a spate of recent research that suggests the ability to sort things abstractly is far more widespread than previously thought.

It’s noteworthy that Dr. Vonk is less confident of her conclusions than is Shallit.

Vonk:

There is still some question as to whether such visual categorization experiments reflect truly abstract thinking by animals, says Vonk, who noted that further work is needed to untangle the tricks various animals use in classification challenges. “I suspect the different species use different means of solving the task,” she notes.

The experiment described is characteristic of research that purports to show abstract thinking by animals, which usually involves the marginal ability of an animal to group images after intensive training. And that’s the issue — the experiments seem to represent training, rather than any innate capacity to abstract thought.

An everyday example of such “research” would be training a golden retriever to retrieve birds, not turtles. Such behavior does not suggest the dog has learned the taxonomic categories of Anas platyrhynchos and Testudines. Despite the researchers’ claims that the ability of the animals to distinguish categories determined by humans entailed genuine understanding of the categories by the animals, there is no reason to accept that conjecture.Animals show no affinity for categorization without “hundreds of trials” and buckets of treats, and the obvious explanation for the learned behavior of pointing to the image that incurred the tasty snack was that the animal was trained to do so, not that the animal actuated a capacity for taxonomy. If I have enough cheese snacks on hand, I can train my Bichon Frise to bark when I point to “4” rather than “3” or “5” after I ask “What is 2 + 2?” That doesn’t mean my dog understands mathematics.

The researchers’ claim that the apes’ ability to distinguish animals of the same kind but somewhat different morphology (snakes and turtles) demonstrates conceptual thinking is not convincing. Snakes and turtles have similar coloration, no fur, no ears, etc., and it seems enormously more likely that the apes were well trained to respond to subtle perceptual similarities than that they apprehended the Linnaean group Reptilia.

So what does it mean to have an abstract thought, a concept? A concept is an acquired disposition to recognize, understand, and discriminate kinds of particulars, and to do so when the particular is no longer perceived and (in come cases) when the particular isn’t perceptible at all.

A concept, in other words, is removable from particulars, and is contemplated independently of the particulars that instantiate it. “Justice” is removable from any perceptions or images I may have of judges, courts, or juries. Justice is an abstraction not identical to any physical thing.

In each of the experiments cited, the issue is not whether the animal (after enough trials and treats) could distinguish one image from another, but whether it could understand the difference between kinds and could conceive of these differences in the absence of the particular image (and the treats). That is what ab-stract means — to draw off, to take the concept out of the particular.

There is no evidence any animal in these experiments took the concept of species or kind or group out of the particulars and contemplated it in the absence of the particulars.

The behavior of animals is entirely sensory. It is rooted in and restricted to particulars. It is limited to the world of perceptual experience and the mental images derived from those particulars. There is no doubt that animal perception can be very acute (dogs’ smell, raptors’ vision, bats’ echolocation) and far exceed that of humans, and there is no doubt that animals via their perceptual abilities can distinguish and manipulate and respond to particulars — obstacles, prey, and predators. But there is not an iota of evidence that animals abstract universals from particulars and contemplate abstractions independently of the particulars that they encounter and remember.

Animals learn tricks, and sometimes (especially in research labs) the tricks seem analogous to abstract thinking. But tricks are not abstract thinking.

Raymond Tallis, a physician and neuroscientist (and an atheist) who has written extensively on animal perception, calls the fallacy that animals can think conceptually the “Disneyfication” of animal consciousness. Coyotes aren’t crafty in an abstract sense, and rabbits don’t contemplate the humor and irony of outwitting hunters. Animals experience pleasure and pain, but there is no evidence whatsoever that they contemplate bliss or suffering or humor or any abstract ideas.

Abstract thinking is inherently an immaterial faculty of the mind, and it is an ability exclusive to human beings. No, animals do not think conceptually.

Image: Bichon Frise, by Editor-at-Large (Own work) [CC BY-SA 2.5], via Wikimedia Commons.