Volume 19: pp. 49-53

Differences Teach Us More Than Similarities: The Need for Evolutionary Thinking in Comparative Cognition

Stephan A. Reber

Department of Philosophy and Cognitive Science, Lund University

Reading Options


A persistent anthropocentric school of thought prevents comparative cognition from truly joining the evolutionary sciences, which often view “cognition” as an alien subject to the study of life. In this article, I argue that cognition is indeed inherent to all life and that we could study the evolution of cognitive skills like any other species-specific trait if we stop elevating convergence over differences, adopt an inclusive working definition of cognition, and choose new model organisms with a strong focus on phylogeny.

Keywordsembodied cognition, phylogeny, cognitive evolution, neuron number, animal cognition

Author Note Stephan A. Reber, Department of Philosophy and Cognitive Science, Lund University, Box 192, 221 00, Sweden.

Correspondence concerning this article should be addressed to Stephan A. Reber at stephan.reber@lucs.lu.se

The declared goal of comparative cognition is to understand the evolution of cognition. For the longest time, cognition always seemed to imply human cognition. Unlike other disciplines studying the evolution of life, the “comparison” aspect is more often than not an afterthought and still hinges on contrasting cognition in any species with its human counterpart. Similarities to humans are often treated as the most enticing results, and convergent evolution (e.g., between primates and corvids) is considered to be particularly special. Using human cognition as a template and elevating convergence over differences alienates comparative cognition from other evolutionary sciences and contributes to its perception (e.g., by behavioral ecologists) as a kind of addition to the study of life, and not as what cognition actually is: its driving force.

Comparisons are the base method of all empirical disciplines, and it is essential that they are made at the same level and between equal entities. For instance, in paleontology, the femurs of many species, alive or extinct, can be compared. From the differences in the bone, the evolution of the femur can be deduced, as well as the interrelatedness of the different species. Notably, it is differences, explicitly not similarities, that allow us to draw conclusions. Further, a bone comparison does not require a standard femur as a reference. Granted, our access to a cognitive skill comes from our own inner experience, which makes this “human template trap” so seductive. It is essential to be aware that each species has as much its own particular cognition as it has its own anatomy. Cognition is a set of attributes of a species that we can compare to answer questions about evolution.

In the rest of this article, I outline three aspects that I consider relevant for the future of comparative cognition: (a) the need for a productive definition of cognition, (b) the integration of the advancements in other fields, and (c) the need for a drastic change in attitude toward broadening the scope of study species and, well-controlled, negative results.

The Need for a Productive Definition of Cognition

How do you define cognition for your own research? I explicitly did not write “What is cognition?” because no human alive has the complete answer to this question. Researchers in most disciplines related to cognitive science appear to have an unspoken agreement to not settle on a definition of cognition before the results of a myriad of future studies make the true meaning of the term self-evident. Allen (2017) even suggested that the attempt to definitively define “cognition” is not an endeavor worth the time. However, researchers in comparative cognition should know what it is they are comparing.

A relatively modern framework in cognitive science is to see cognition not as centralized in the body (in our case, the brain) but as embodied: Interactions of the body with the environment results in (embodied) cognition, which also extends out into the external world (Osvath et al., 2014). This view is ideal for large-scale comparisons, as it is minimally restrictive and basically can be applied to all forms of life. For instances, associative learning does not require an actual brain, as seen in the box jellyfish Tripedalia cystophora (Bielecki et al., 2023). This result also suggests that forming associations is at least as old as the shared ancestor of Cnidaria and Bilateria, which would have lived in the Precambrian period. Beings entirely without neurons, such as the slime mold Physarum polycephalum, can modify their behavior depending on previous experiences (Murugan et al., 2021). The mechanisms by which this is possible are strikingly similar to those in an animal brain (Boussard et al., 2021), although achieved by different means.

Are slime molds as complex as chimpanzees? Of course not, but that does not mean that what they are doing is not cognition. Many researchers still insist on a dichotomy between “mechanical” and “cognitive” to describe the same behavior in a slime mold and an ape (e.g., following signals to find food). This is as absurd as saying that a single-celled organism cannot have locomotion because it does not have a skeleton. Cognition varies along a continuum from relatively basic to highly complex (Levin et al., 2021). Did the advent of nervous systems lead to a massive qualitative improvement of decision making? Without a doubt, but it also has to be clear to us that those nervous systems evolved specifically for improving cognitive skills that already existed. Nervous systems did not come about and then decisions could be made. Like everything else in evolution, life used what it already had to adapt to the environment.

We do not yet have a universally agreed-upon definition of cognition, which can account for its entire continuum of complexity. However, to draw conclusions about evolution, it is essential to know what authors throughout the field consider “cognitive” or “part of cognition.” Hence, each should have a definition of cognition they can refer to. For instance, there clearly are those who consider perception part of cognition and others who consider it a step preceding cognition. Even though I belong to the former, it does not matter as long as I am aware that authors adhering to the second group differ in their definition; I could still use their work to propose hypotheses taking all findings into account. I would like to end this section by suggesting one possible and, in my opinion, highly useful description of cognition first created by Pamela Lyon (2020):

“Cognition comprises the sensory and other information-processing mechanisms an organism has for becoming familiar with, valuing, and interacting productively with features of its environment [exploring, exploiting, evading] in order to meet existential needs, the most basic of which are survival/persistence, growth/thriving, and reproduction.” (Lyon et al., 2021, p. 4)

What Biological Features Should We Compare Cognitive Performance With?

Matching the differences in cognitive performance across species to measurements of the brain is an established practice in comparative cognition (e.g., brain mass, encephalization coefficient). As discussed in the previous section, no brains or even neurons are required for cognition. But the recent advancements in neuroscience may very well allow us to understand cognition at a level that applies to all life, the cellular one. A still underused, yet fairly available, metric is the number of neurons. Excellent efforts of Pavel Němec and colleagues led to an ever-expanding data set with the neuron numbers of hundreds of species (Kverková et al., 2022; Olkowicz et al., 2016). Importantly, we can identify the number of neurons in each section of the brain, which allows us to trace the evolution of the brain better and better. Matching neuron numbers to findings in animal cognition leads overall to a fairly convincing scaling of cognitive complexity. Famously, the pallium of a common raven has slightly more neurons than that of a capuchin monkey, although it is only one quarter of its weight (Olkowicz et al., 2016). We can expect the method of comparing cognition across species and using neuron numbers per brain region as predictors of performance to become a standard in comparative cognition.

It is also possible to reconstruct the connections in an entire brain, a so-called connectome, using electron microscopy. The latest breakthrough was the completion of the connectome of a larval Drosophila melanogaster, which had 548,000 synapses in a brain with 3,016 neurons (Winding et al., 2023). This was the biggest brain fully mapped so far, and the challenges for larger brains are still substantial. However, the architecture of brain circuits is deeply conserved across Animalia (Winding et al., 2023) and lessons learned from Drosophila might yet again start a revolution of a biological field. We are still years, probably decades, away from having a database of connectomes of key species in comparative cognition. But the day will likely come when we can use the actual neuronal connections of different animals to understand the evolution of cognition.

Which Species Should We Study and Compare?

Any discipline in biology would profit from studying as many species as possible because, as elaborated on earlier, the more differences we find, the more steps in evolutionary history could be understood. Animal cognition is, probably more than other fields, relying on a disproportionally small set of study organisms, and there is a straightforward reason: High-quality cognition studies require a massive investment. Detailed studies on isolated cognitive skills can usually be performed only in captivity or in semiwild conditions, which requires costly keeping facilities or extensive man-hours. Particularly, when we work with species that cannot live in traditional lab settings, it may take the better part of a dedicated researcher’s entire career to perfect the methodology to study the cognition of a given genus or even a single species. The apparent lack of resources makes simply broadening the scope of species to study unfeasible. Comparative cognition should choose novel study species with a particular focus on phylogenetic relatedness. As an example, let’s look at avian cognition: Extant birds belong to two major clades, the Palaeognaths (ostriches, rheas, tinamous, kiwis, emus, cassowaries) and the Neognaths (every other bird). A snapshot review of avian cognition literature between 2015 and 2020 found not a single study on Palaeognaths (567 studies/141 bird species; Lambert et al., 2022). Passeriformes (perching birds) accounted for close to half the species studied (47%). If one would like to understand the evolution of avian cognition, it seems obvious that the longest strides could be made by comparing the two major clades of living birds. In a recent study, we investigated visual perspective taking in three species of Palaeognaths (Zeiträg et al., 2023). We found that all of them can follow the gaze of a conspecific around a barrier, that is, repositioning themselves to take another’s perspective. Previously, only three species of Passeriformes were known to exhibit this behavior (Bugnyar et al., 2004; Butler & Fernández-Juricic, 2014; Schloegl et al., 2008), which invited the assumption that this is a special ability absent or at least rare among other bird species. Our study has now shifted these expectations. Visual perspective taking might very well be present in all bird lineages alive and probably predates the Palaeognath–Neognath split. Choosing study species with a focus on phylogeny advances our understanding of cognitive evolution more rapidly. This approach by no means implies that replicating studies in more closely related species should be considered redundant. But by targeting major clades that are, as of yet, understudied, we can learn more about how widespread certain cognitive abilities are, which in turn allows for more accurate predictions for studies within any clade.

Most of the established model organisms in comparative cognition are evolutionarily relatively young (e.g., apes and corvids). They can also be tested rather rapidly once they are accustomed to experiments. Other species (e.g., nonavian reptiles) are more time-consuming to test, and their cognition is (probably) less similar to humans. Yet, convincingly documented limitations in or even absence of an aspect of cognition in less-well-studied taxa have the potential to teach us more about evolution than yet another similarity born out of homology or convergence. Such studies, of course, require controls. Let’s take birds as an example again: In an observational study on allopreening (i.e., grooming the plumage of an affiliated conspecific), four species of Palaeognaths and common ravens were filmed for 429 hr. Ravens showed extensive allopreening, but none of the Palaeognaths displayed this behavior even once. By using ravens as the control species, Jensen and colleagues demonstrated that allopreening most likely arose in Neognaths (Jensen et al., 2023).


In sum, we should embrace the differences and stop elevating similarities to human cognition. The inclusion of species that are harder to work with should be encouraged, which will happen automatically if we start to value evidence for limitations. By adopting this more objective view, by defining our concepts clearly, and by integrating our work with that of other disciplines, evolutionary theory will become inherent to comparative cognition (as it should be).


Allen, C. (2017). On (not) defining cognition. Synthese, 194(11), 4233–4249. https://doi.org/10.1007/s11229-017-1454-4

Bielecki, J., Dam Nielsen, S. K., Nachman, G., & Garm, A. (2023). Associative learning in the box jellyfish Tripedalia cystophora. Current Biology, 33(19), P4150–4159. https://doi.org/10.1016/j.cub.2023.08.056

Boussard, A., Fessel, A., Oettmeier, C., Briard, L., Döbereiner, H.-G., & Dussutour, A. (2021). Adaptive behaviour and learning in slime moulds: The role of oscillations. Philosophical Transactions of the Royal Society B: Biological Sciences, 376(1820), Article 20190757. https://doi.org/10.1098/rstb.2019.0757

Bugnyar, T., Stöwe, M., & Heinrich, B. (2004). Ravens, Corvus corax, follow gaze direction of humans around obstacles. Proceedings of the Royal Society of London. Series B: Biological Sciences, 271(1546), 1331–1336. https://doi.org/10.1098/rspb.2004.2738

Butler, S. R., & Fernández-Juricic, E. (2014). European starlings recognize the location of robotic conspecific attention. Biology Letters, 10(10), Article 20140665. https://doi.org/10.1098/rsbl.2014.0665

Jensen, T. R., Zeiträg, C., & Osvath, M. (2023). The selfish preen: absence of allopreening in Palaeognathae and its socio-cognitive implications. Animal Cognition, 26(5), 1467–1476. https://doi.org/10.1007/s10071-023-01794-x

Kverková, K., Marhounová, L., Polonyiová, A., Kocourek, M., Zhang, Y., Olkowicz, S., Straková, B., Pavelková, Z., Vodička, R., Frynta, D., & Němec, P. (2022). The evolution of brain neuron numbers in amniotes. Proceedings of the National Academy of Sciences, 119(11), Article e2121624119. https://doi.org/10.1073/pnas.2121624119

Lambert, M. L., Farrar, B. G., Garcia-Pelegrin, E., Reber, S. A., & Miller, R. (2022). ManyBirds: A multi-site collaborative open science approach to avian cognition and behavior research. Animal Behavior and Cognition, 9(1), 133–152. https://doi.org/10.26451/abc.

Levin, M., Keijzer, F., Lyon, P., & Arendt, D. (2021). Uncovering cognitive similarities and differences, conservation and innovation. Philosophical Transactions of the Royal Society B: Biological Sciences, 376(1821), Article 20200458. https://doi.org/10.1098/rstb.2020.0458

Lyon, P. (2019). Of what is “minimal cognition” the half-baked version? Adaptive Behavior, 28(6), 407–424. https://doi.org/10.1177/1059712319871360

Lyon, P., Keijzer, F., Arendt, D., & Levin, M. (2021). Reframing cognition: getting down to biological basics. Philosophical Transactions of the Royal Society B: Biological Sciences, 376(1820), Article 20190750. https://doi.org/10.1098/rstb.2019.0750

Murugan, N. J., Kaltman, D. H., Jin, P. H., Chien, M., Martinez, R., Nguyen, C. Q., Kane, A., Novak, R., Ingber, D. E., & Levin, M. (2021). Mechanosensation mediates long-range spatial decision-making in an aneural organism. Advanced Materials, 33(34), Article 2008161. https://doi.org/10.1002/adma.202008161

Olkowicz, S., Kocourek, M., Lučan, R. K., Porteš, M., Fitch, W. T., Herculano-Houzel, S., & Němec, P. (2016). Birds have primate-like numbers of neurons in the forebrain. Proceedings of the National Academy of Sciences, 113(26), 7255 LP–7260. https://doi.org/10.1073/pnas.1517131113

Osvath, M., Kabadayi, C., & Jacobs, I. (2014). Independent evolution of similar complex cognitive skills: The importance of embodied degrees of freedom. Animal Behavior and Cognition, 1(3), 249–264. https://doi.org/10.12966/abc.08.03.2014

Schloegl, C., Schmidt, J., Scheid, C., Kotrschal, K., & Bugnyar, T. (2008). Gaze following in non-human animals: The corvid example. In F. Columbus (Ed.), Animal behaviour: New research (pp. 73–92). Nova Science Publishers.

Winding, M., Pedigo, B. D., Barnes, C. L., Patsolic, H. G., Park, Y., Kazimiers, T., Fushiki, A., Andrade, I. V, Khandelwal, A., Valdes-Aleman, J., Li, F., Randel, N., Barsotti, E., Correia, A., Fetter, R. D., Hartenstein, V., Priebe, C. E., Vogelstein, J. T., Cardona, A., & Zlatic, M. (2023). The connectome of an insect brain. Science, 379(6636), Article eadd9330. https://doi.org/10.1126/science.add9330

Zeiträg, C., Reber, S. A., & Osvath, M. (2023). Gaze following in Archosauria—Alligators and palaeognath birds suggest dinosaur origin of visual perspective taking. Science Advances, 9(20), Article eadf0405. https://doi.org/10.1126/sciadv.adf0405