Do Animals Think?

In the animal behavior popular press, battle lines are being drawn. On one side are the anthropomorphists, authors like Jeffrey Masson, Elizabeth Marshall Thomas, Peter Singer, and Jane Goodall, who believe animals experience life just like we do. Animals solve problems the same way we do, they learn the same way we do, they have the same emotions, the same capacity for language, the same -- well, humanity. On the other side are authors like Stephen Budiansky, Marc Hauser, and Clive Wynne. Folks on this side of the fence argue that intelligence is not a single thing, but a many-splendored thing; animals are all intelligent in very different ways. Do Animals Think? lobs quite a few shells into the anthropomorphic camp, and Wynne isn't shy about naming names.

I'll come right out and say it: I'm with the latter group. Different animals have different sensory organs, and take in different information from their worlds. Dogs see different visual patterns than humans do because they don't see in color and yet they do see infrared frequencies. Bats can "see" in the dark with their ears, via echolocation. In addition to ascertaining what sensory data is available to an organism, one must determine what types of patterns the animal is looking for in the data. Is it a herbivore or a predator: what does food look/smell/feel like? Does it have predators to look out for? Is it social or solitary? Is the animal a picky eater looking for a specific pattern in the data, or a generalist looking for broad classes of patterns? Once the information is processed, how can the animal interact with its world; what kind of body does it have? Does it run, swim, crawl, or fly? Does the animal have to chase and kill something in order to eat? Is manipulation of food required, such as avoiding spines or thorns, breaking off shells, or neutralizing stingers? What structures does it have for eating or defense? All of these things -- senses, ecology, and morphology -- help define an animal's intelligence. Different sensory inputs combined with different physical outputs yield different brains, processing different types of information for different reasons. Any non-human animal will by definition experience a different world than we do.

Looked at another way, each animal has to have the intelligence to solve the problems that are ecologically placed before it. A rattlesnake has to solve rattlesnake problems; how do you find, capture, and swallow prey if you don't have any arms or legs? Rattlesnakes do not have to memorize sequences of flashing lights in order to get fed, and so they tend to do poorly on human-based intelligence tests. However, a human would do just as poorly on a rattlesnake intelligence test; bind a human's arms and legs and toss them in a pitch-black room with a live rat, and they will probably not successfully eat the rat. When people assume intelligence is a single thing that is the same in all animals, they miss out on how amazingly complex life on this planet really is.

In Do Animals Think? Wynne combines species accounts with general topics. There are chapters devoted to what remarkable critters bees, bats, pigeons, and dolphins are. Each of these chapters discusses what we know about the animals' behavior, and contrasts this with popular myths and misinterpretations. In all instances, the animals' actual behavior is shown to be much more intricate than anthropomorphism allows for. Pigeons are able to find their way home using a variety of tools unavailable to us limited humans; an acute sense of smell, ability to see polarized light, and even perception of magnetic fields may contribute to pigeon orientation. Rupert Sheldrake's half-baked notion that pigeons find their roosts via quantum entanglement is unceremoniously shot down. Also debunked is the "100th monkey" myth -- invented by Lyall Watson and perpetuated by Sheldrake -- which states when a certain number of animals learn how to do something, then suddenly other animals elsewhere will magically know how to do it.

Between chapters on individual animals are general discussions on topics such as language and learning. Wynne examines the body of ape sign-language literature and finds it lacking. A fundamental question scientists must always ask themselves is, "Can this data be explained with a simpler mechanism?" Can an ape seeming to use language to ask for food be explained by simple associative learning -- make this gesture with your hand, get a banana? Almost all "statements" made by apes, whether through sign language or computer keyboards, are single words, double words, or repetitious concatenations. The chimpanzee Nim's longest "statement" is, "Give orange me give eat orange me eat orange give me eat orange give me you." Communication is clearly happening -- Nim wants an orange -- but true grammatical language is just as clearly not happening. No grammar or comprehension of grammar has been shown in any ape language studies. Word combinations are often nonsensical. The vast majority of communications, like Nim's above, are simple demands for food. Associative learning explains the data, and frees the researcher to see apes as not like us, but wonderfully different. Wynne shows again and again that clinging to the belief animals use language skews researchers' interpretation of their own data. Unfortunately, some of these researchers will not allow others to independently examine their data, never a good sign in the science world (sort of like when movie studios don't preview a movie for the press). When one independent researcher analyzed the chimpanzee Washoe's utterances from videotape, Washoe's handlers threatened to sue for breach of copyright.

A further danger of approaching a data set with assumptions is you often have to invent unseen mechanisms to account for your assumptions. Sheldrake assumed pigeons are just like us, so in order to explain homing he had to invent "morphic resonance," which in turn he had to explain with a pseudo-physical quantum entanglement model. Take away the assumption that pigeons are just stupid people with wings, and you can see that they have completely different sensory structures that are sampling different parts of the informational world. Thus, you can explain homing without invoking mystical energy fields. Assumptions mask the world, and anthropomorphism is a particularly blinding assumption. Wynne invites us to see the world for what it is. Along with Stephen Budiansky's If A Lion Could Talk, Do Animals Think? is a good primer in seeing past ourselves to view the rich tapestry of animal behavior with unclouded eyes.

• Read more about this book