Can animals think about thinking?

An overview of metacognitive studies in non-human animals

By: Maria Lagakos

This may make you feel guilty about killing that insect crawling on your wall at 1A.M.…

Within the first few days of taking the biology course titled “Evolution of Brain and Behaviour” at McGill University, we were assigned a reading entitled “The frontiers of insect cognition” by Perry et al (8). This article offers an account of the, often disregarded, cognitive capacities of various insect species. With humans and non-human primates often being the center of sophisticated cognitive studies, Perry et al. showcase the seemingly complex forms of cognition found in insects, ranging from spatial navigation abilities to metacognition. Metacognition has long been studied in the field of philosophy of mind and cognitive neuroscience, as a primarily human attribute. With Perry et al.’s paper in mind, I wondered how and if metacognition was present in non-human species.

Metacognition

Metacognition is generally defined as the ability to think about one’s own thoughts – thinking about thinking (2). Concisely put by Nancy Chick, “metacognition includes a critical awareness of a) one’s thinking and learning and b) oneself as a thinker and learner” (5).  Metacognition has, thus, come to be associated with consciousness in humans, notably because thinking about one’s own thoughts seems to be a private and internal reflective experience (2).  

During certain tasks, we may utilize metacognitive abilities by sorting through our memory and picking the most suitable option, or refraining from doing something due to lower levels of confidence. In this sense, and in the cognitive science domain, metacognition is often referred to as the ability to monitor and regulate one’s own cognitive processes. This regulation could be denoted as recognizing uncertainty and opting for alternative strategies to reach a certain goal (4). 

In the last few decades many researchers, such as evolutionary biologists and behavioral neuroscientists, have sought out an understanding of the non-human animal mind, and have debated its internal mechanisms and processes.

So, what about animals?

Asides from metacognition, a plethora of executive functions have been denoted in non-human animals. But, what are executive functions?

In cognitive science, “executive function” is an umbrella term that refers to skills  that enable us to regulate our behavior: that allow us to control ourselves, focus on one task at a time, plan for the future, etc (6). These are functions that are generally viewed as “higher order,” more complex and effortful, and that sometimes take into account our values, morals and goals. Some of the most agreed upon executive functions are inhibition (i.e. refraining from impulsive action), selective attention (i.e. selectively attending to a specific stimulus) and working memory (i.e. holding information in memory to complete a given task) (7). From these building blocks, we are able to engage in more complex behaviors and cognitive processes, such as learning, social interaction and reasoning, among others (7).  

While generally neglected as a set of species with complex behaviors, some insects have been suggested to display higher-order cognition. Take bees and wasps as an example. Interestingly, they seem to engage in “social learning and cultural transmission” (8). Perry et al. note that, generally, insects have been understood as separate from primates, the latter being the group who exhibit social learning and individual recognition within their community. That being said, their article blurs the lines of this stereotyped dichotomy between “higher” and “lower” species. For example, they tell us that some groups of wasps actually learn the facial features of their colony’s members. This enables individuals to understand where others fall in the colony’s social hierarchy, based on their fighting skills (8). 

Furthermore, honeybees are capable of transmitting knowledge. They do this via “waggle dancing” as seen here (9). These insects display a certain sequence of movements which indicate to their observers the “direction, distance, and quality” of a location of food resources and suitable new residences (8, 9). Not only does this behavior demonstrate working memory, but it also shows that this communication enables learning. Bees who are deprived from observing this dance in others display more errors in their own waggle dances, by not properly encoding distances, for example (9).

A famous study done on monkeys claims that they have a sense of “fairness.” In this experiment, two capuchin monkeys are put in chambers side by side where they can see each other. Both monkeys learn a task where once they are given a rock, they must return it to the researcher in order to receive a reward. The reward for one of the monkeys is a grape, a highly attractive reward, and the reward for the other is a piece of cucumber, a less attractive reward in comparison to the grape. When the monkey receiving the cucumber realizes that his neighbor is receiving the more delicious reward, he launches the cucumber piece back to the researcher (i.e. the giver) as if to say “Hey! Where’s my grape?”. In this task, we see that the cucumber-receiving monkey senses that this situation is unequal and unjust.

How does this relate to metacognition?

It is important to understand executive functions, as they seem to be fundamental to metacognition. An example of this is reading comprehension. Reading comprehension requires an awareness of understanding what one is reading (i.e. metacognition). However, to do this, the reader must keep in their available memory what is happening in the passage they are reading to comprehend the story (11). Metacognition depends on executive functions and these executive functions in turn offer ways for complex behaviors and cognitive processes to occur.

Can metacognition be found across non-humans? How do we study this?

While many executive functions and complex skills have been noted in non-human animals, metacognition, strongly related to what many conceive of as consciousness, is one of those traits that scholars hesitate to attribute to non-primate organisms (12). On the other hand, many also argue that by having these executive functions, it may lead to the emergence of metacognition.

Importantly, animal studies of metacognition have been modified from the traditional metacognition measurement tools in humans, which have heavily relied on verbal reports (12). Thus, many of the experiments done with non-human animals have been based on public cues, rather than private metacognitive cues. These public cues are the observable traits that may indicate the presence of metacognition in the individual (2).

Studies aimed at studying metacognition followed a similar methodological pattern: they are typically perceptual-behavioral tasks which mix easy and hard tasks in one given trial, and allow the animal to opt out of any tasks that they may choose (4, 12). The idea is that if an individual comes across a trial that is too hard or that they are uncertain of, they might choose to opt out of the task. In this case, researchers suggest that the individual could be endowed with metacognitive abilities, since they are demonstrating that they are aware that they lack the knowledge to provide a correct response (4). Lacking confidence in one’s ability to provide a correct answer might inherently require an awareness of what is right or wrong in a given context and an awareness of personal knowledge.

Non-human primates have famously been studied for this cognitive process, considering the general presumption that they may be cognitively closest to humans. In one experiment, presented by Hampton (13), rhesus monkeys were presented with a certain sample of stimuli, and had to remember which sample of stimuli they were shown after a certain delay (referred to as the “forgetting interval”). After this delay, monkeys faced two conditions: in two thirds of the trials they were given the choice to either take the memory test or decline the memory test. In the rest of the trials, they were forced to take the test. Notably, they had to make this choice prior to seeing what they would be tested on. The results indicate that when the individual can choose whether or not to partake in the test, its responses are more accurate than when forced to take the test. This is because, in the first case, the monkey can opt out of the trial when it is uncertain about its memory, whereas in the second case, even if it doesn’t remember, it is forced to respond anyways. Thus, Hampton suggests that this demonstrates that “rhesus monkeys know when they remember” (13). 

This behavior has also been demonstrated in our previously mentioned friends: bumble bees and honey bees. In the experiment by Perry et al, bees were trained to do a visual discrimination task, for which they were either rewarded or punished based on the correctness of their response. The bees were trained to enter a special two-chamber setup and to choose the stimuli that was associated with reward. These stimuli could either be placed above or below a reference bar and bees were trained to associate either a reward or a punishment to the stimuli based on its position (i.e. above or below the reference bar). In easy trials, the stimulus was evidently above or below the reference bar, whereas hard or impossible trials presented the stimulus aligned with the reference bar. During testing, bees could fly to the first chamber and choose to opt out of the trial if it was too hard. However, upon flying into the second chamber, they were forced to make a decision (14). 

To determine whether bees were intelligently opting out of the task in the first chamber or had simply associated avoidance to difficult trials, the researchers analyzed the bees’ correct performance on hard tasks in both forced and unforced trials. Bees had more correct responses in hard tasks that they chose to do versus those that they were forced to do. This seems to be because they opted out of the hard trials that were deemed most difficult to solve. Once again, this experiment demonstrated that bees “selectively avoid difficult tasks they lack the information to solve” (14). 

What does this tell us?

Many scholars understand this metacognitive process, of opting out when uncertain, to be associated with human consciousness (13). In doing metacognitive studies, humans themselves report that their uncertainty responses are based on a conscious awareness of their uncertainty (12). Whether this can be extended to animals – the capacity to be aware of their uncertainty –  is still up for debate amongst many (12, 14). 

For example, while some might argue that such capabilities demonstrate “meaningful self-reflection,” others note that there is little to no research on what is being directly “reflected on,” as in what is happening in the mind of the animal that enables them to make these certainty judgments (10). 

That being said, animals seem to demonstrate the use of these uncertainty responses even “under circumstances of no training, deferred reinforcement and so forth,” eliciting some scholars to believe that it is plausible that these animals display similar conscious representations as humans (12). 

Thus, while the research is not yet conclusive, the possibility of metacognition in animals should be enough to relay ethical treatment to non-human species. With the rise of studies related to animal sentience, as well as the negative effects of the anthropocene on all organisms, such ethical considerations are being further taken into account in socio-political contexts. While we still have a long way to go, disciplines such as evolutionary biology, the anthropological study of human-animal interactions and ecological conservationism, continue to explore the needs of our wild counterparts, as well as the consequences of our unbalanced relationship with them.  

References

1. Fleming S. Can Animals or Computers Have Metacognition? (2014, September 1). Scientific American. https://www.scientificamerican.com/article/can-animals-or-computers-have-metacognition/#:~:text=As%20later%20experiments%20demonstrated%2C%20macaques,answers%20it%20thinks%20is%20correct. 
2. Hampton, R. R. Metacognition and metamemory in non-human animals. (2019) Encyclopedia of animal behavior. https://psycnet.apa.org/record/2018-36382-051 
3. Barron A. & The Conversation US. Bees Are Astonishingly Good at Making Decisions. (2023, June 23). Scientific American. https://www.scientificamerican.com/article/bees-are-astonishingly-good-at-making-decisions/ 
4. Beran M. et al. Primate Cognition: Attention, Episodic Memory, Prospective Memory, Self-Control, and Metacognition as Examples of Cognitive Control in Nonhuman Primates. (2016, June 10). Wiley Interdisciplinary Review – Cognitive Science.   https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5173379/ 
5. Chick, N. Metacognition. (2013). Vanderbilt University Center for Teaching. https://cft.vanderbilt.edu/guides-sub-pages/metacognition/ 
6. Harvard University. What Is Executive Function? And How Does It Relate to Child Development? Center on the Developing Child. https://developingchild.harvard.edu/resources/what-is-executive-function-and-how-does-it-relate-to-child-development/#:~:text=The%20phrase%20%E2%80%9Cexecutive%20function%E2%80%9D%20refers,focused%20despite%20distractions%2C%20among%20others. 
7. Diamond, A. Executive Functions. (2012, September 27). Annual Review Psychology. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4084861/ 
8. Perry C. et al. The frontiers of insect cognition. (2017, May 25). Current Opinion in Behavioral Sciences. https://www.sciencedirect.com/science/article/pii/S2352154616302017#:~:text=Insects%20display%20a%20variety%20of,of%20much%20larger%2Dbrained%20animals. 
9. Dong, S. et al. Social signal learning of the waggle dance in honey bees. (2023, March 9). Science. https://www.science.org/doi/10.1126/science.ade1702 
10. Kornell, N. Where is the “meta” in animal metacognition? (2014, May). Journal of Comparative Psychology. https://psycnet.apa.org/record/2013-25592-001 
11. Metacognition and Executive Function: A Dynamic Relationship of Cognitive Functioning. (2025) Reflection Sciences. https://reflectionsciences.com/blog-metacognition-executive-function/#:~:text=Metacognitive%20strategies%20and%20executive%20function,are%20crucial%20in%20reading%20comprehension
12. Smith. D. J. The study of animal metacognition. (2009, September 1). Trends in Cognitive Sciences. https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(09)00148-X 
13. Hampton, R. R. Rhesus monkeys know when they remember. (2001, March 27). Biological Sciences. https://www.pnas.org/doi/full/10.1073/pnas.071600998#abstract 
14. Perry, C. J. & Barron, A. B. Honey bees selectively avoid difficult choices. (2013, November 4). Biological Sciences. https://www.pnas.org/doi/full/10.1073/pnas.1314571110 
15. Smith, D. J. et al. Animal Metacognition: Problems & Prospects. (2009). Comparative Cognition & Behavior Reviews. https://scholarworks.gsu.edu/cgi/viewcontent.cgi?article=1006&context=lrc_facpub