Skip to main content
EMBO Reports logoLink to EMBO Reports
. 2021 May 25;22(7):e53199. doi: 10.15252/embr.202153199

The emergence of consciousness

Research on animals yields insights into how, when and why consciousness evolved

Philip Hunter 1,
PMCID: PMC8406393  PMID: 34031966

Abstract

Whether some animal species possess consciousness is no longer the question; rather how their environment and evolution shaped species‐specific forms of self‐awareness

graphic file with name EMBR-22-e53199-g001.jpg

Subject Categories: Ecology, Neuroscience


Ever since humans acknowledged consciousness in themselves, they speculated whether animals could have a similar sentience or awareness of their internal and external existence. But although human philosophers had pondered on consciousness for centuries, it was not until 1927 when the American psychologist Harvey Carr laid the foundations for research on animal consciousness. He argued that awareness in animals could be only understood and measured when we had developed “an accurate and complete knowledge of its essential conditions in man” (Carr, 1927).

This may have provided a springboard for the field, but a definition of the essential conditions of consciousness in Homo sapiens has proved elusive to this day—hence, research on animal consciousness has struggled to achieve a sound basis for formulating and evaluating testable hypotheses. However, there has been some progress in developing correlates of human consciousness that can be applied to study animals, while brain scanning and imaging has recently allowed comparative studies of human and animal neurological activity while performing mental tasks. It has also become possible to observe animal behaviour and communication in much greater depth and identify examples of activities—such as advance planning, or recognition of individuals through their vocalization—that can be associated with human consciousness. Overall, there is a growing consensus that this research has moved beyond merely questioning whether animals can be conscious or aware of themselves to defining different dimensions along which this can be assessed.

research has moved beyond merely questioning whether animals can be conscious or aware of themselves to defining different dimensions along which this can be assessed.

Abandoning anthropogenic definitions

According to John Birch, Associate Professor of Philosophy and Principal Investigator of the Foundations of Animal Sentience project at the London School of Economics, UK, there is an emerging consensus that various mammals, birds and at least some cephalopod molluscs—notably octopuses, squid and cuttlefish—exhibit some form of consciousness (Fig 1). The debate has therefore moved on to the questions of which species are conscious and what form their sentient experiences take. As Birch and others see it, the debate was bogged down by an anthropogenic view of consciousness, derived from clinical models that place coma at one extreme and full wakefulness at the other.

Figure 1.

Figure 1

The African grey parrot (Psittacus erithacus) is one of the most intelligent animal species and has a high level of consciousness and self‐awareness. Source: Wikimedia/Papooga.

This is now being replaced by a multidimensional model that caters for the differences in mental capabilities between animals. The ERC‐funded Foundations of Animal Sentience project, Birch explained, aims to develop a “consciousness profile” for each species. But this requires more insights, not least relating to the evolutionary forces behind the emergence of consciousness. “At the moment we know very little about how evolutionary processes shape the consciousness profile of a species”, Birch said. “We can make conjectures. To give a simple example, it is plausible that a species which ecologically relies on caching, like a scrub jay, has a resultant need for integration of information across long time periods: episodic memory and future planning”. Clever and carefully designed experiments are needed to provide evidence for such conjectures. “A long‐term goal for animal consciousness research should be to formulate and test big‐picture hypotheses about the pressures that shape the consciousness profile, drawing on as much comparative data as we can”, Birch added. Such experiments actually date back almost two decades, when Nicky Clayton and colleagues at Cambridge University, UK, studied cognition in members of the crow family, including jackdaws, rooks and jays. This led her to challenge long‐held assumptions that only humans can plan for the future and understand the minds of others.

Meanwhile, there is some agreement over the multiple dimensions of animal consciousness that might be used to construct the unique species profiles Birch mentioned (Birch et al, 2020a; Birch et al, 2020b): perceptual richness (p‐richness), evaluative richness (e‐richness), integration at a time (unity), integration across time (temporality) and self‐consciousness (self‐hood). P‐richness is a measure of perception across one or multiple senses, which clearly varies between species: humans score particularly highly on visual perception but many animals including dogs and elephants are much more tuned towards the olfactory sense. E‐richness is related to experiences that might have a subjective element such as pain, thirst, hunger or fear, leading to more complex emotions such as love and greed mainly associated with humans. Determining where animals fit into this E‐richness matrix is an important goal of consciousness research, Birch commented.

The third dimension of integration at a time, or unity, is interesting in that animals exhibit big differences between themselves and from humans in terms of connectivity between the hemispheres of the brain and especially the halves of the visual cortex fed by each eye. While humans enjoy a highly unified conscious experience thanks to the corpus callosum that connects the two hemispheres of the brain, birds, for instance, lack this structure, which raises the question of whether they have two parallel consciousnesses cooperating with each other. Temporality is related to episodic memory—the “what, where and when”— and its processing for determining present actions and anticipation of future scenarios. As Clayton found, California scrub‐jays can retrieve food items discriminately by remembering what they had cached, where and when. They also know that some food needs time to ripen to become edible and so caching decisions clearly involve some future planning based on episodic memory.

Selfhood is then awareness of oneself as distinct from other individuals. It has long been proven at a basic level mainly by the mirror‐mark test, which has been passed by dolphins, primates and birds of the corvid family.

There is growing support now for the multidimensional theory of animal consciousness, even if there may be room for wriggle over the exact components and their definitions. “I think this is inevitable if you think in evolutionary terms”, commented Eva Jablonka, an evolutionary theorist at Tel‐Aviv University, Israel. “It is clear that there will be different ecologically relevant styles of consciousness, different ways of perceiving the world and the body, different notions of self and so on. This does not mean that one cannot discern a progressive trend in some lineages, along some dimensions. But it is clear that the conscious styles (or ‘Bauplans’) of octopus, corvid, dog and human are different and have evolved to be different”.

It is clear that there will be different ecologically relevant styles of consciousness, different ways of perceiving the world and the body, different notions of self and so on.

Experimental evidence

There are though some properties or benefits of consciousness or awareness that should be more universal, and one of those is the ability to judge how other individuals of a species might behave under given circumstances. Humans clearly have this capability, but it seems that corvids are also able to draw on their own experiences to infer how others may behave, especially among those who adopt anti‐social behaviour such as stealing food. (Emery & Clayton, 2001). At a similar level, male Eurasian jays can feed their female partners the food she would appear to like and adapt to her changing desires (Ostojić et al, 2013). Research is now exploring whether these abilities amount to self‐consciousness, a question that may never be answered beyond all doubt.

What can be done experimentally though is use brain imaging to identify which forms of sensory output may be associated with a degree of sentience. This was applied recently in a study comparing facial recognition in dogs and humans for identifying both members of the same species and individuals within it (Bunford et al, 2020). The authors in Hungary and Mexico used brain scanning on 20 pet dogs and 30 human volunteers to measure their responses to faces. The dogs were trained to lie still inside an MRI tube while watching four types of two‐second video clips: the front or back of a human head, and the front or back of a dog head. The same videos were played to the human volunteers.

This firstly confirmed results of earlier studies that, when shown a face, a large part of the visual cortex became active in both humans and dogs, but less so when only the backs of heads were shown. Human brains were sensitive both to faces of their own species and to differences between individuals. The dogs however only exhibited the former, being more oblivious to differences between individuals. It suggests that dogs may not have a sense of self derived from visual processing, which may seem surprising given that dogs appear to be capable of distinguishing between people in the flesh. But this may also involve processing of animation and other clues, as well dogs' much more evolved olfactory sense.

Naked mole rats' societies

Vocal communication between individuals of a species is another aspect of intelligence closely associated with consciousness since it involves distinguishing between self and others. Although not considered an immediate correlate, vocalization overlaps with all five of the dimensions identified by Birch and others and has been studied in that context among dolphins and primates in particular, given that individuals can recognize each other and their social groups from sounds they make. The more advanced features of vocalization, such as dialect and recognition of individuals, had been thought to be confined to higher mammals, some birds and cetaceans. However, two species of rodent, including the African naked mole rat, are also capable of dialect and recognizing individuals by their sound.

Vocal communication between individuals of a species is another aspect of intelligence closely associated with consciousness since it involves distinguishing between self and others.

This finding is perhaps less surprising to those familiar with naked mole rats and how they evolved to cope with the extreme conditions in their underground habitats in arid regions of Africa. They are almost unique among mammals, and indeed vertebrates, in being Eusocial like ants, bees and termites, defined by cooperative brood care and division of labour into reproductive and non‐reproductive functions. It often involves a single queen sired by a group of reproductive males, along with non‐reproductive members of both genders split into castes to perform well‐defined roles. This is the case for naked mole rats, which also exhibit human‐like behaviours of promotion, relegation, coup d’etats and regicide.

The latter may happen when a younger female aspiring to the throne recruits some males to help assassinate the queen. It leads to a period of societal upheaval, which demonstrates the role of vocalization. “It may take time for a new queen to become established when the old queen goes. In order for a colony to work cooperatively, a queen has a large controlling effect”, explained Gary Lewin, Group Leader at the Max Delbrück Centre for Molecular Medicine in Berlin, Germany. “When the queen is there, the colony speaks with one voice, but if you take the queen away, the dialect changes and becomes less coherent, much more variable. The queen makes sure everyone does their job”.

The mole rats build complex networks of tunnels to interconnect chambers allocated for specific purposes, such as storing food, excretion and as quarters for the queen. They are exclusively vegetarian and their sole food source are various plant tubers. Such tubers are often quite scarce so evolution has taught the rats that long‐term survival requires moderation and leaving some carefully intact so that they can regenerate. In that sense, mole rats developed agriculture some 30 million years before humans.

None of these factors equate directly to consciousness, but, according to Lewin, the extreme degree of social organization is at least indicative of an awareness of self as animals learn their place in the hierarchy and identity through vocalization. He also noted evidence of empathy in the animal’s ability to respond to pain in others, although that appears to be shared by a wider number of animals. “Studies in various animals suggest empathy is quite ancient”, Lewin suggested.

One former postdoc in Lewin’s team, Alison Barker, who is starting her own research group at the Max Planck Society in Germany, studies the neural mechanisms underpinning the unusually high degree of vocalization in mole rats and how that relates to self‐awareness and social behaviour. “There are certainly elements of naked mole‐rat behaviour that are relevant to the question of how consciousness evolved”, Barker explained. “One of the things that I find most intriguing about the naked mole‐rat use of vocalizations is the high level of sensory discrimination that is possible. We showed that naked mole‐rat vocalizations can encode information about individual identity and colony membership and that information about colony membership is actively used to drive behaviour”.

Consciousness shaped by the environment

The 40 species of aquatic mammals commonly known as dolphins are another example of how environmental constraints may have driven evolution of higher levels of consciousness or self‐awareness. While the lack of manipulative limbs and especially opposing thumbs has prevented dolphins from developing equivalents of human tools or settlements, they have nonetheless mastered their environment, commented Diana Reiss at Hunter College, City University of New York (CUNY). “Despite not having arms, hands or fingers, dolphins are highly manipulative and have found highly effective ways to live in their environment”, she explained. “They are opportunistic foragers and use many cooperative foraging strategies that require precise social and self‐awareness in timing, for example mug‐ring feeding and stranding feeding, that create their own objects of play, such as bubble rings”.

Reiss also pointed out that dolphins develop self‐awareness and mirror recognition at the age of 7 months, at least 6 months before human babies do. “It is an observable measure or index of consciousness”, she said. “As I often say with all puns intended—it is a clear reflection of a level of conscious awareness of an external representation of self and the understanding that the mirror can be used as a tool to gain access to viewing oneself”.

The origins of consciousness

If we accept that at least some animals do have higher levels of consciousness or self‐awareness, the question arises of when and why it evolved. Making a reasonable estimate depends on breaking down what the foundation of consciousness is in terms of cognitive abilities. “One conjecture in this area is that conscious experience is intimately related to real‐time coordination of complex actions in active, mobile animals”, explained Birch. “The idea is that conscious experience helps an animal disentangle the changes in sensory input caused by its own actions from changes caused by the outside world. Bjorn Merker and Peter Godfrey‐Smith have powerfully articulated that view. A different view ties the evolution of conscious experience to certain forms of learning—in particular, to learning about subtle relationships that cut across different sense modalities and that reach across time. This is the ‘unlimited associative learning’ view, powerfully advocated by Simona Ginsburg and Eva Jablonka”.

Birch suggests these views are not incompatible though but could combine into a unified theory of consciousness. “Interestingly, both views converge on the Cambrian period (about 500 million years ago) as being the most likely time of origin for conscious experience, a view also defended, via a third different route, by Todd Feinberg and Jon Mallatt in The Ancient Origins of Consciousness”. That book, published in 2016, proposed that consciousness appeared much earlier in evolutionary history than had been commonly assumed. The authors posit that the great “Cambrian explosion” of animal diversity produced the first complex brains, which were accompanied by the ability to distinguish between self and the environment. They argue that simple reflexive behaviours evolved into a unified inner world of subjective experiences, deducing that all vertebrates are and have always been conscious. Feinberg and Mallatt even suggest that arthropods, including insects and probably crustaceans, meet many of the criteria for consciousness.

The authors posit that the great “Cambrian explosion” of animal diversity produced the first complex brains, which were accompanied by the ability to distinguish between self and the environment.

At around the same time, Jablonka and colleagues argued that the emergence of unlimited associative learning (UAL) around the Cambrian explosion marked the transition to minimal consciousness or sentience (Bronfman et al, 2016). UAL is the ability to associate an event with other stimuli across senses, such as hearing, sight and smell, along with combinations of motor actions. This can only be executed by animals that possess some form of central coordinating system and therefore can be seen as a marker of consciousness or sentience.

However, building on the concept of UAL to establish how higher orders of consciousness around the self emerged, or even how to define them, remains a challenge, as Jablonka acknowledged. “A sense of self is very ancient, but it is not self‐consciousness in the human sense”, she said. “Human self‐consciousness leads to a very peculiar and special kind of sense of self and feeds very strongly into the evolution of social emotions and social behaviours”.

Jablonka has collaborated with Birch and Simona Ginsburg to explore further how UAL can be extended to establish a finer‐grained foundation for consciousness and to test for both its presence and absence (Birch et al, 2020a; Birch et al, 2020b). As their article suggests, despite all the advances to identify correlates of consciousness, the underlying mystery embodied in the “mind‐body” problem formulated by philosopher René Descartes in the 17th century remains unsolved. Research on consciousness in animals to comprehend lesser forms of self‐awareness and how and why it evolved may therefore provide the clues to better understand human consciousness aptly described by Descartes’ Cogito ergo sum—“I think, therefore I am”.

… despite all the advances to identify correlates of consciousness, the underlying mystery embodied in the “mind‐body” […] remains unsolved.

EMBO reports (2021) 22: e53199.

References

  1. Birch J, Ginsburg S, Jablonka E (2020a) Unlimited associative learning and the origins of consciousness: a primer and some predictions. Biol Philos 35: 56 [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Birch J, Schnell AK, Clayton NS (2020b) Dimensions of animal consciousness. Trends Cogn Sci 24: 789–801 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bronfman ZZ, Ginsburg S, Jablonka E (2016) The transition to minimal consciousness through the evolution of associative learning. Front Psychol 10.3389/fpsyg.2016.01954 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bunford N, Hernández‐Pérez R, Borbála Farkas E, Cuaya LV, Szabó D, Szabó AG, Gácsi M, Miklósi A, Andics A (2020) Comparative brain imaging reveals analogous and divergent patterns of species‐ and face‐sensitivity in humans and dogs. J Neurosci 40: 8396–8408 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Carr H (1927) The interpretation of the animal mind. Psychol Rev 34: 87–106 [Google Scholar]
  6. Emery NJ, Clayton NS (2001) Effects of experience and social context on prospective caching strategies by scrub jays. Nature 414: 443–446 [DOI] [PubMed] [Google Scholar]
  7. Ostojić L, Shaw RC, Cheke LG, Clayton NS (2013) Evidence suggesting that desire‐state attribution may govern food sharing in Eurasian jays. Proc Natl Acad Sci USA 110: 4123–4128 [DOI] [PMC free article] [PubMed] [Google Scholar]

Articles from EMBO Reports are provided here courtesy of Nature Publishing Group

RESOURCES