Exploring Animal Minds: Philosophical and Scientific Perspectives

Do animals think about what other animals are thinking? It sounds like a simple question, but it sits at the intersection of some of the deepest problems in philosophy and science. When a chimpanzee avoids the gaze of a dominant rival, or when a scrub jay relocates its hidden food after noticing it was being watched, is it reasoning about another mind – or just reacting to a behavioral cue? This question has occupied philosophers and cognitive scientists for nearly five decades, and the debate is far from settled.

What is theory of mind – and why does it matter for animals? 🔗

In 1978, David Premack and Guy Woodruff coined the term theory of mind (ToM) to describe a cognitive capacity they wondered whether chimpanzees possessed: the ability to attribute mental states – such as beliefs, desires, intentions, and knowledge – to oneself and to others. As the Stanford Encyclopedia of Philosophy notes, this capacity is called a “theory” precisely because mental states are not directly observable; the organism uses them to predict and explain the behavior of others. In humans, this is considered a cornerstone of social life. The central philosophical question is whether any non-human animals share it.

The stakes go beyond academic curiosity. If animals possess even a partial theory of mind, this has profound implications for how we understand animal consciousness, moral status, and the evolution of cognition. As the Internet Encyclopedia of Philosophy outlines, the debate touches on foundational questions: Can animals have beliefs? Do they reason? And what would it even mean for a non-verbal creature to understand that others have inner mental lives?

The challenge of studying animal cognition 🔗

One of the most persistent problems in this field is methodological. Researchers cannot rely on verbal reports from animals, as they do with human children. This forces scientists to design experiments where behavior alone serves as evidence of mental state attribution – a requirement that immediately raises interpretive difficulties.

The logical problem 🔗

At the heart of the debate is what researchers call the “logical problem”: any behavioral result that seems to suggest an animal is reading minds could equally be explained by the animal simply reading and responding to observable behavioral cues – body posture, gaze direction, spatial proximity – without any genuine representation of another’s mental state. Philosophers at the Stanford Encyclopedia of Philosophy have documented how this problem has generated decades of inconclusive debate, with some researchers proposing new experimental designs to disentangle the two explanations, and others arguing the problem itself is a philosophical impasse that cannot be settled empirically.

Anthropomorphism versus “killjoy” science 🔗

Researchers in animal cognition face a fundamental tension in how they frame their findings. Some adopt an anthropomorphic perspective – interpreting complex animal behavior in mentalistic terms, drawing parallels to human cognition. Others, described by Daniel Dennett as “killjoy” scientists, resist attributing human-like mental capacities to animals and demand simpler mechanistic explanations. Both approaches risk distorting the science: one by over-attributing, the other by systematically under-attributing. Achieving genuine objectivity in this domain remains elusive, because the very choice of what counts as sufficient evidence is itself a philosophical and theoretical judgment.

The false belief task: the gold standard and its limits 🔗

The false belief task has long served as the definitive test of theory of mind – both in human children and in animals. In the classic version, a subject must demonstrate that they understand another agent holds a belief about the world that is actually false. In human children, this understanding typically emerges around age four. Inspired by commentaries from philosophers Dennett, Jonathan Bennett, and Gilbert Harman on the original Premack and Woodruff study, the false belief paradigm became central to decades of animal cognition research.

What chimpanzees can and cannot do 🔗

Research on chimpanzees has produced notably mixed results. A landmark study published on PubMed found that while children succeeded on both verbal and nonverbal versions of the false belief task, neither chimpanzees nor orangutans passed the nonverbal version, despite demonstrating mastery of the task’s basic demands. A subsequent competitive false belief study published in PLOS ONE confirmed that great apes – including chimpanzees, bonobos, and orangutans – can distinguish between true and false beliefs in helping tasks designed for human infants, suggesting at least a basic level of false belief understanding may exist, though it falls short of the full propositional belief attribution seen in adult humans.

The current consensus, as reviewed in research published by ScienceDirect, is that chimpanzees grasp the distinction between knowledge and ignorance in others – they can tell if a competitor saw where food was hidden or not – but do not reliably understand false beliefs. Chimpanzees appear to track what others have seen, but not what others believe in the deeper representational sense that would constitute a full theory of mind.

The behavior-reading versus mindreading debate 🔗

Even where positive results exist, skeptics argue they can be explained without invoking genuine mental state attribution. As a review published in PMC puts it, the entire controversy hinges on whether a cognitive system produces information specific to another agent’s perspective, rather than simply responding to observable behavioral regularities. Critics have argued that chimpanzees may be highly skilled behavior-readers without being genuine mindreaders – and that distinguishing between these two possibilities has proved extraordinarily difficult in practice.

Evidence from non-primate animals 🔗

While chimpanzees have dominated the debate, research has expanded significantly to include other species – particularly corvids (the crow family), whose cognitive abilities have surprised researchers and challenged assumptions about the evolutionary origins of social cognition.

Scrub jays and the cache protection problem 🔗

Western scrub jays provide some of the most compelling behavioral evidence of social cognition outside primates. In natural settings, these birds compete for food by caching it and then relocating their caches when they suspect competitors know where the food is hidden. Research on corvid cognition has shown that scrub jays will re-cache food after being observed storing it – but crucially, only those birds with prior experience of pilfering other birds’ caches exhibit this protective behavior. This suggests the birds are not simply following a fixed rule, but are drawing on their own experience of being a thief to infer what a competitor might do.

However, critics have offered an alternative account. A computational model published in PLOS ONE proposed that re-caching behavior in scrub jays could be explained entirely by stress responses and learned behavioral rules, without invoking any reasoning about mental states. This counter-hypothesis illustrates a recurring pattern in animal cognition research: for nearly every positive finding, an alternative, simpler explanation is proposed.

Ravens and attributed visual access 🔗

Ravens have attracted particular attention for what appears to be a sophisticated understanding of what others can and cannot see. A landmark study published in Nature Communications found that ravens guard their caches in response to sounds of competitors when a peephole is open, but not when it is closed – even without seeing the competitor directly. This suggests ravens can generalize from their own perceptual experience to infer the possibility of being observed, going beyond a simple response to gaze cues. The study’s authors concluded this provides strong evidence that ravens are more than mere behavior-readers. Even so, debate continues over whether this constitutes genuine mental state attribution or an especially sophisticated form of associative learning.

Controversies and interpretive challenges 🔗

Despite decades of research, there remains little consensus on whether any non-human animal possesses anything resembling a full theory of mind. The controversies are not just empirical – they are deeply philosophical. Researchers often agree on the raw behavioral data but disagree sharply about what it means.

The question of minimal theory of mind 🔗

One response to the persistent impasse has been to shift the framing. Rather than asking the binary question “do animals have a theory of mind?”, researchers increasingly ask which components of ToM are present in which species, and to what degree. As discussed in a review article in Learning & Behavior, the research agenda has moved toward identifying which aspects of mental state attribution – perception, knowledge, desire, belief – are present in different animals, rather than treating theory of mind as a single, all-or-nothing capacity. This has led to the concept of a minimal theory of mind: a reduced set of abilities that allows an organism to track others’ perceptions and knowledge states without necessarily representing full propositional beliefs.

The value of non-human research 🔗

One of the most significant contributions of animal cognition research to philosophy is precisely what it reveals about the limits of our methods and concepts. Studying non-verbal animals forces researchers to design experiments that rely entirely on behavioral indicators – anticipatory looking, gaze following, competitive foraging choices – as proxies for mental state understanding. These methods have proven invaluable not just for animal cognition research, but for understanding infant cognition as well, since human infants share the constraint of being non-verbal. As the Stanford Encyclopedia of Philosophy observes, the scientific study of animal social cognition has long been of deep interest to philosophers, who have in turn contributed to shaping the experimental paradigms used by cognitive scientists.

Evolutionary considerations 🔗

The discovery of sophisticated social cognition in corvids – birds that diverged from the primate lineage approximately 300 million years ago – raises important questions about the evolution of mind. Rather than treating theory of mind as a uniquely human achievement built on a primate cognitive foundation, these findings suggest that elements of social cognition may have evolved independently in multiple lineages in response to the cognitive demands of complex social life. This shifts the philosophical discussion: the question is no longer just whether animals have minds, but how different kinds of minds evolved to track the mental states of others.

Where philosophy and science converge 🔗

The debate over animal theory of mind is an exemplary case where philosophy and empirical science are genuinely interdependent. Philosophical analysis is needed to define what theory of mind actually requires, to assess the adequacy of experimental designs, and to evaluate competing interpretations of the data. Science, in turn, generates the findings that challenge philosophical assumptions and force conceptual refinements. As reviewed in Notre Dame Philosophical Reviews, the fundamental epistemic question – “How could we ever know if an animal has a theory of mind?” – remains as pressing as the empirical question itself, and answering one without the other is impossible.

What makes this debate so productive is precisely its difficulty. It forces scientists to be more rigorous about what counts as evidence, and it forces philosophers to be more precise about what they mean by belief, knowledge, and mental representation. Both disciplines are better for the encounter.

What do you think? If an animal consistently acts as though it knows what another animal is thinking – but we cannot rule out a simpler behavioral explanation – does that uncertainty matter morally? And should the standard of evidence we require from animals be any different from the standard we apply when inferring mental states in other humans?

References

1. https://plato.stanford.edu/entries/animal-social-cognition/
2. https://iep.utm.edu/animal-mind/
3. https://plato.stanford.edu/entries/cognition-animal/
4. https://pubmed.ncbi.nlm.nih.gov/10218261/
5. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0173793
6. https://www.sciencedirect.com/science/article/abs/pii/S0010027708001881
7. https://pmc.ncbi.nlm.nih.gov/articles/PMC2346530/
8. https://www.questjournals.org/jrhss/papers/vol10-issue10/1010173177.pdf
9. https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0032904
10. https://pmc.ncbi.nlm.nih.gov/articles/PMC4740864/
11. https://link.springer.com/article/10.3758/s13420-017-0268-z
12. https://ndpr.nd.edu/reviews/mindreading-animals-the-debate-over-what-animals-know-about-other-minds/