Well, actually, mostly we don’t. It’s still a young field and a tricky one. We often can’t be sure even what other people are thinking, even though we can talk to them directly and share the same language. ('Yuh, sure she said that, but did she really mean it? What is she actually trying to tell me?’) And animals don’t talk, at least in the usual sense, which makes it even tougher to know what they’re thinking. Furthermore, many animal species live very different lives from us, and so, as the philosopher Wittgenstein famously cautioned, even if they could talk, we probably would not understand what they had to say.

Nevertheless, trying to understand the mental lives of other species - to get inside their heads - is an active and exciting area of research, and, notwithstanding the above caveats, a lot of it, in fact, focuses on aspects of communication, sparked by the example of research on bat echolocation by Donald Griffin and others in the 1950s. They discovered how it is that bats can navigate in the dark, and not by special powers of vision, but by sound - by producing a continuous stream of high-frequency 'click’ sounds (that we humans can’t hear) while they fly (Figure 1) and then listening for the portions of these clicks that get reflected (echoed) back off obstacles in the environment. This work helped to illuminate the dark world of the bat and it catalyzed the modern era of research on animal thinking - what is often now labeled 'cognitive ethology’. Following Griffin’s lead [1], one entire branch of this field focuses on studying the communication systems of other species, believing as he did that these offer a privileged window into the workings of their minds.

Well, as the example of bat echolocation itself showed, some other species are likely to have mental lives very different from ours. They see the world very differently and probably think very differently about it as a consequence. Bats effectively 'see’ in sound (although they can also see) and they navigate in the dark by 'talking to themselves’. Exactly what kinds of thoughts they might have about the world remains an open question, but, given their very different experience of the world, whatever thoughts they have (if any) are likely to be quite different, possibly almost unimaginable to us as a grounded, visually dominated species [2, 3].

That’s certainly a sensible intuition that is partially supported. For example, another landmark study in cognitive ethology concerns research on vervet monkeys, which are small African primates. These monkeys are vulnerable to a variety of predators, including leopards, eagles, and large snakes, and studies in the 1970s and 80s found that the monkeys produced different alarm vocalizations in response to each [4]. When other vervets heard the alarm calls of their companions, they automatically engaged escape responses that were different and appropriate for evading a leopard, an eagle or a snake even if these listeners had not themselves yet seen and identified the particular predator involved. It was as though the alarm calls on their own told others everything they needed to know, serving in effect as symbolic labels for the different predators much like our own human words for 'leopard’ , 'eagle’ and 'snake’. So, here was a case where a species more closely related to humans seemed to see the world and to communicate about it much as we do. This was exciting not just for suggesting that the capacity for symbolic communication and thought might exist in a species other than our own but also for what this fact suggested about the deep history of complex thought and language in humans. In other words, it might help us to understand another species better, and it might also help us to understand ourselves better.

Well, the evolutionary origins of complex language and cognition in the human lineage are another hot, and hotly debated, topic, susceptible to many of the same problems that plague studies of animal thinking and more [5]. It’s hard enough to ascertain what others around you are thinking, whether animal or human. It’s harder still when your subjects are fossil species long extinct. How do you reconstruct the thoughts of long dead ancestors? Here, the vervet alarm call story offered a promising lead. Contrary to some evolutionary scenarios that assume a very recent origin of symbolic thought, language and culture in humans, the evidence for symbolic communication in vervet monkeys - with whom we share only a very distant primate ancestor - suggested that the cognitive foundations for these abilities might be very deep indeed, long preceding our transition from monkey-like primates to human ones.

Not necessarily. Accumulating evidence since the original vervet work now points to some fundamental differences in the capacity for symbolic thought and communication in nonhuman primates compared to humans. For example, more recent studies of vervet alarm calls suggest that important details of the original work need to be revised and that the symbolic capacity of the monkeys’ calls may be fundamentally different from that of human words. These newer findings align with a large body of other research comparing the cognitive underpinnings of human language and primate communication [6].

In humans, language use is fundamentally intentional, in two senses. It is intentional in the formal philosophical sense that language users make implicit assumptions that the other people they interact with are mental agents, like themselves, with thoughts and beliefs about the world. It is also intentional in the more colloquial sense that speakers deliberately mean, or intend, to share what they know about the world, and influence what others think and know about it, through the use of words that codify thoughts and beliefs. In contrast, intentionality in both senses is largely absent in nonhuman primates, with the possible exception of some of the great apes [7]. In general, primates do not understand each other as mental agents with thoughts and beliefs, or knowledge about the world. Nor do they communicate with the intent of sharing or providing information about the world to others. In short, they seem not to know what others know about the world, or even that others might know things about the world at all. They therefore also fail to appreciate that what others know can be affected by one’s own communications [8]. So, in a fascinating paradox, the alarm calls that vervet monkeys produce function to alert, or inform, others about the presence of dangerous predators, but the effect is inadvertent or unwitting. The monkeys themselves do not understand the effect their calls have.

Possibly, but not necessarily. It might mean only that the signals they use are simpler and less diverse because they serve relatively broad social functions. Many primates are intensely social. Who you are and what your relationships are to others in the group are critical, and the importance of particular relationships can vary depending on the social context. In this kind of setting, the messages conveyed by vocalizations might be relatively straightforward, perhaps simply announcing your identity - in effect, simply reminding others who you are [11]. But complexity might then arise in the variety of responses that can flow from such announcements for others based on their relationship to you and to others in the group and how those relationships are best prioritized in the particular social context involved. This represents a complex social calculus for which primates appear to be especially adapted. So, it’s possible that the relatively sophisticated social-inferential abilities of primates have effectively relieved signalers from having to develop a larger, more complex repertoire of signals to convey many different messages. Perhaps for primates, the onus of communication is on listeners more than on signalers. Primates do, after all, have large brains and we have to assume all that brain tissue is there for a reason and doing something. Still, it remains an important puzzle to solve.

Arguably, the most important lesson is that the evolutionary process can yield a complex and sometimes surprising mosaic of outcomes. As the examples of communication and cognition in humans, primates and songbirds nicely illustrate, closely related species with more similar histories (human and nonhuman primates) are often similar in many respects. But they can also be very different. The psychological intentionality that undergirds complex language in humans, and distinguishes it from the communications of nonhuman primates, also supports a variety of other complex human traits, including our capacity for pedagogy, for empathy, and for broad-scale cooperation, all of which might also prove unique to humans [6]. So, it’s a mental difference that makes a difference. At the same time, very distantly related species, with divergent evolutionary histories (humans and songbirds), are usually quite different but they can be surprisingly similar in some respects. There is nothing yet in the corpus of general evolutionary principles that would lead us to expect in advance that the communication systems of songbirds would have so much in common with human language. Why there is such communicative convergence between songbirds and humans remains an important puzzle to be solved, but one thing we know is that it has also produced some interesting convergences in the way bird and human brains are structured and organized (despite their vast differences in size).

Ultimately, both the similarities among species and their differences are equally informative in our efforts to understand the organization and evolution of various kinds of minds. But the mosaic pattern of outcomes we’ve seen so far certainly reminds us that there can be a host of different solutions to life’s common problems and therefore myriad evolutionary paths species can take. In exploring these possible paths, it’s important that we, as researchers, not pre-judge the places they are likely to lead, or allow our human biases and expectations to color what we think we see in other species. We may be in for a lot of surprises yet.