It is often said that humans are the only animals to use symbols. So many other claims of human uniqueness have fallen away—thoughts, emotions, intelligence, tool use, sense of fairness—what's so special about symbols, you ask? I share your skepticism, dear reader, and in the next few paragraphs I'll tell you why.

Let's begin by clarifying what "symbol" means here. One way to do this is to contrast symbols with signs. A sign, such as a red light, a grimace, a growl, or a thunderstorm, signifies something direct and tangible, making us think or act in response to the thing signified. Issuing and responding to signs is commonplace in Animalia. A symbol, on the other hand, is "something that represents something else by association, resemblance, or convention". A symbol allows us to think about the thing or idea symbolized outside its immediate context, such as the word "water" for the liquid, "7" for a certain quantity, and "flag" for a community. What is symbolized doesn't even have to be real, such as God, and herein lies the power of symbols—they are the building blocks of abstract and reflective thought. Evidence of material symbols used by humans dates back at least 60-100K years, when burial objects and decorated beads start to appear in archaeological finds. Linguistic symbols were almost certainly in use long before then.

According to Susanne Langer, symbols serve "to liberate thought from the immediate stimuli of a physically present world; and that liberation marks the essential difference between human and nonhuman mentality ... Words, pictures, and memory images are symbols that may be combined and varied in a thousand ways." It is only through symbolic thought that we imagine the past or the future—mental time-travel, including episodic memory, requires the use of symbols. Indeed, language is really a system of symbolic communication, combining words (which are symbols) and syntax. If non-human animals lack symbols, what and how do they really think?

Do Animals Live Only in the Moment?

If it's true that we are the only species that uses symbols, then nonhuman animals (henceforth animals), including intelligent problem-solvers, live purely in the moment and perceive only their immediate environments. They follow habit and reflex, have at best associative feelings and emotional states, and cannot think about anything in the past or the future, nor wilfully imagine fond objects they cannot see. For instance, per this view, a hungry dog cannot imagine in her mind a food bowl, nor imagine her pup that died yesterday (because this needs a mental image, i.e., a symbol for the bowl and the pup), nor plan with intent for a future event. Sure, past events may impact animal moods in the present—as from a beating a dog may have received—but the dog would not be able to recall the face or the location of the beater in her mind, until he reappears in person and awakens associative feelings. A lack of symbols would also imply that animals have no consciousness, if we assume that consciousness needs at least a minimal "awareness of self", which is an abstract, symbolic idea. Animals like birds and mammals then, implies this view, live entirely in a state of "momentary sentience". They do little more than feel pleasure and pain, act out of instinct, and respond to signs using the physical abilities that evolution has granted them. 

Could this portrait be true? As support for this "no symbols" model of the animal mind, defenders claim that animals display no actual evidence of symbol use. But how solid is this claim? Can we reliably say that symbols are not being used in some cases? For instance, take the mating dance ritual of Blue-footed Boobies, during which the male gives the female "a small stone or stick. He then tips his beak, tail, and wing tips to the sky and whistles." If there were any symbolic import in the stone vs. the stick, or its size and weight, or in the type of whistle, how would we know it? When a chipmunk stashes away nuts for the winter, is that purely instinctive or is a certain symbolic conception of a cold, nutless future mixed in? Do elephants invest a certain amount of symbolism in mourning their dead? There may well be no symbolic import in these examples, but if there were some, how would we know?

Then there is the vast range of non-visual expression in the animal world where symbols may exist but to which we are usually oblivious. For instance, vervet monkeys, writes James R Hurford, "use a 'bark', a 'cough' or a 'chutter' to communicate the presence, respectively, of a leopard, an eagle, or a snake. There is nothing (as far as we know) inherently leopardlike in a bark, or inherently barklike in a leopard. It seems more reasonable to grant that the vervets are using genuinely arbitrary symbols", especially since the calls do not follow a stimulus-response mode and depend on context; the calls are likelier if one's own offspring need protection rather than unrelated juveniles, or if a female needs to be impressed" (read more). Prairie dogs have dozens of unique alarm calls for different predators, including "different ones for humans with or without guns." Perhaps all this is one hundred percent biological programming untainted by symbol use, but how do we know for sure? Herein lies a genuine epistemological problem. How do we know the inner experience of animals like chipmunks, monkeys, and elephants, each with its unique and frequently prodigious capacities of sight, smell, memory, taste, sound, locomotion, spatial cognition, echolocation, geomagnetic navigation, and more?

One reason to be skeptical of the "no symbols" model of animal minds is based on a consideration of the astonishing problems many animals solve to survive and the complex social behaviors they actually display. Is all that possible without any symbols? Some of it may be but some may not be. Can a baboon order, classify, and track over a lifetime its complex social-hierarchical relations with hundreds of individuals without symbolic concepts about them? Mary Midgley adds, "Many animals move continually from one food source to another, often with their young to provision, and sometimes with responsibility for a whole pack or herd. They have to be able to think how long this or that will last, or when it will recur. If they had not enough memory and anticipation of order to fit their plans into the probable train of events, with alterations for altered circumstances, they often could not survive." 

In recent decades, experiments and systematic observations have seriously questioned the "no symbols" model. We now know that many animals, including chimpanzees, whales, dolphins, and even African Grey parrots, at least have the capacity for symbol use. The behavior of the bonobo in Susan Savage-Rumbaugh's video is a great argument for why symbol use is likely continuous with other animals. Three years ago a chimp "jolted the research community by providing some of the strongest evidence yet that nonhumans could plan ahead". At least one study describes how Capuchin monkeys were taught the use of certain symbols. While evidence for a capacity does not constitute evidence for its active use, these findings are notable. Does evolution commonly provision unused capacities?

In fact, a range of studies have revealed animal behaviors that, in humans, involve symbol use. Some birds "can recall past events and use the information to plan for the future". Many species "show behavioral manifestations of different features of episodic memory". Tool-using beavers have "been observed gathering material they need before starting to build, which shows forethought." The ability to correctly order numerical quantities exists in many species. Flexible and contextual deceptive behaviors, such as playing-dead to escape certain types of predators, are practiced by some birds and mammals. Behaviors suggesting metacognition—the knowledge of what one knows—have been detected in rats. Pigs are apparently capable of "visual perspective taking, which is the ability to assume what the other [weaker pig] sees and to adjust one's own behavior accordingly", suggesting "a degree of theory of mind". Several species have passed the mirror self-recognition test—including magpies, which belong to a distant evolutionary lineage from us—even as many other animals may get more of their "sense of self" from better developed aural, olfactory, tactile, and other sensory information. While not conclusive, these findings are significant enough to have thrown wide open the big question: What can we justifiably say about the inner lives of animals? 

§

Since the 1970s, many cognitive ethologists and other specialists who study animals have challenged old orthodoxies about animal minds, including many implications of the "no symbols" model. This has invited accusations of naïve anthropomorphism from some who appeal to "reason" and "parsimony" (as if these weren't eternally corruptible human constructs. Their accusations in fact remind me of the "relativism" bogey). To be fair, it is also true that many animal lovers tend to inflate or misrepresent the capacities of animals; our folk stories, animated cartoons, and casual talk routinely assign human-like thought to them. Though we share a lot in common, humans are clearly not the same as other animals, who almost certainly don't sit around and argue about their models of the human mind. Having said that, how deeply does one's head have to be buried in the sand to defend the "no-symbols" model today, versus one that is more elastic and non-binary on the question of symbol use and is better able to account for observed animal behaviors?

A model, for instance, that embraces a gradualist and nonlinear evolutionary approach to animal minds, and permits many species some symbols both inside and out, giving them a certain sense of time, episodic memory, and intent. (Arguably, we too use lots of prelinguistic symbols that remain inside, and of which we are only partly aware. Human social and sexual relations, for instance, abound in subterranean and primordial symbols of power and desire—all part of the substrate in which our newly learned and linguistic symbols likely take hold.) Migratory cranes and whales have navigation skills far superior to anything humans possess, but this does not mean humans have zero navigation sense. Likewise for symbol use. Where humans have the crane-like advantage is in our ability to "learn massive numbers of arbitrary symbols" and to give expression to them (in art, ritual, language, etc.)—adaptations fueled by our more complex social lives.

Frans de Waal writes, "There will always be tension between those who view animals as only slightly more flexible than machines and those who see them as only slightly less rational than human beings." The problem of other minds is even worse across the species barrier. Science may never be able to settle whether animals use symbols in their inner lives, or whether they live entirely in the moment. We're stuck with reasoned interpretation of carefully observed behavior. And at the end of it, long after running the animals through our speciocentric hoops, if certain behaviors leave room for doubt about their symbolic content, why not give animals the benefit of the doubt? As JM Coetzee puts it, "Why should it be the doubters who always get the benefit of the doubt?"

This essay was motivated by an exchange on another 3QD thread with my friend, Chris Schoen, to whom I'm grateful for that discussion. More writing by Namit Arora?

Read a companion piece: On Eating Animals.