Do Non-Human Animals Have a “Theory of Mind”?

Theory of mind (ToM), or mentalising, is a fundamental aspect of human cognition. It is defined as the ability to attribute mental states—such as desires, beliefs, intentions, and emotions—to oneself and others, enabling the explanation and prediction of behaviour based on these mental states (Wimmer & Perner 1983).

David Premack and Guy Woodruff (1978) first introduced the concept and term "theory of mind," defining it as the capacity to attribute mental states to oneself and others. They argued that this ability constitutes a "theory" because mental states are not directly observable but can be inferred to predict behaviour. In their seminal paper ‘Does the Chimpanzee Have a "Theory of Mind"?’, they explored whether chimpanzees could understand human goals and intentions. Their findings suggested that chimpanzees can comprehend goals and intentions, providing initial evidence of a non-human ToM. However, they found no evidence that chimpanzees understand false beliefs, concluding that chimpanzees operate within a perception-goal psychology rather than a human-like belief-desire psychology (Call & Tomasello 2008).

Premack and Woodruff’s study inspired extensive research into whether animals genuinely understand other minds or merely respond to learned associations and environmental stimuli. This has led to the development of alternative methodologies aimed at addressing the question.

Some researchers, however, have expressed skepticism about whether non-human animals can possess ToM. For instance, Penn and Povinelli (2007) questioned whether non-verbal animals could understand the concept of a mental state.

It’s strange that some researchers assume that the absence of verbal thinking and verbal language equates to a lack of conceptual understanding? And why do we impose human categories onto animals' non-verbal languages?

In fact, the concepts found in verbal language, such as nouns, verbs, and adjectives, are arbitrary. The absence of certain linguistic elements, like adverbs, in animals' non-verbal languages does not imply a lack of understanding of concepts like speed, direction, or spatial relationships.

Similarly, the absence of verbal concepts like "love" or "morality" does not mean that non-verbal animals are incapable of experiencing or expressing love, or embodying moral traits such as cooperation, empathy, fairness, justice, and trust.

Heyes (1998) observed that little progress had been made on Premack and Woodruff's question, while Povinelli and Vonk (2004) criticised many experimental protocols for being fundamentally incapable of testing ToM hypotheses.

Call and Tomasello (2008) reviewed studies of ToM in chimpanzees and concluded that the evidence supports the idea that chimpanzees understand certain mental states, such as perception and knowledge. However, laboratory studies testing false belief understanding suggest that chimpanzees lack this specific cognitive ability. Critics argue that such artificial experimental contexts may not reflect animals' natural cognitive abilities.

To address this limitation, researchers like Nissani et al. (2004) developed tasks that simulate naturalistic environments. Some of these studies revealed evidence of abilities that earlier, more artificial protocols failed to detect. Similarly, Hare et al. (2006) adapted study designs to better align with natural behaviours and found evidence suggesting that chimpanzees can manipulate others’ mental states, challenging the notion that intentional deception is uniquely human (e.g., Cheney & Seyfarth, 1990; Tomasello & Call, 1997).

Research has since investigated ToM in other social animals, including dolphins, elephants, and dogs. Elephants, for instance, exhibit empathy-related behaviours, such as comforting distressed individuals and cooperating in tasks. Similarly, dogs are highly attuned to human social cues, such as gaze and pointing gestures, which suggests they can infer human intentions and beliefs.

Cats’ understanding of their humans’ ToM is even better – they can detect when their owners are trying to deceive them, e.g., cats somehow know when they're being taken to the vet, even when there are no obvious signs inside the house and the owners may try to act casual and pretend like everything is normal.

While the extent to which animals possess a full-fledged ToM remains debated, mounting evidence indicates that certain species can attribute mental states to others. This ability likely varies across species and individuals, with some animals demonstrating a more sophisticated understanding of mental states than others.

PS: Let's be honest, do humans understand ToM of animals? In majority of cases, the answer is ‘No’. Even when it comes to animals like cats, who have been living with humans for centuries. Can you honestly claim to know what that cat in the empty room is thinking? What are her intentions, desires, and so on? And there's a fun little challenge for you to try out: the Theory of Feline Mind Test.