The question of animal emotions

SCIENCE • 24 Mar 2022 • Vol 375Issue 6587 • pp. 13511352 • DOI: 10.1126/science.abo2378
There is increasing evidence that some invertebrates, such as lobster, are sentient and experience emotions. This means they should be killed ethically, rather than being boiled.
If the UK joins a handful of other nations to recognize the sentience of invertebrates, such as cephalopod mollusks and decapod crustaceans, by, for example, prohibiting the boiling of live lobsters, this will be based on evidence that emotions and felt experiences (i.e., sentience) are not limited to animals close to humans, such as the mammals. This topic has been heavily debated in both affective neuroscience (how to define an emotion?) and philosophy (what is the moral relevance of animal experiences?), but a consensus on the criteria for and implications of recognizing animal sentience seems to be emerging (1).
Over a decade ago, the same debate revolved around fish. Do fish feel pain? It was thought that these animals only had nociception—i.e., unconscious reactions to noxious stimuli mediated by the peripheral nervous system, such as when humans reflexively withdraw a hand from a hot stove before they are aware of the burn. Fish may react to pain, it was thought, but without any associated feelings. Because nociception does not necessarily reach the central nervous system and consciousness, it does not amount to sentience. Sentience in this case requires valenced experiences—i.e., experiences judged by the organism as either attractive/positive or aversive/negative. This debate was settled when fish were found to learn from encounters with negative stimuli by avoiding dangerous locations. The best explanation is that fish remember these locations because they felt and neuronally processed aversive experiences (2).
The same logic has been followed for arthropods, such as crabs, which in experiments learn to avoid locations where they have been shocked. Nociception cannot explain associative learning because there is no motivation to avoid the stimulus if it does not feel bad at the time. The negative experiences underlying their behavior may even be integrated with other evaluations, because hermit crabs need stronger shocks to evacuate a high-quality shell that protects their abdomen than a poor-quality shell (3). If valenced experiences guide behavior in complex ways, the central nervous system must be involved. In vertebrates, the typical pathway to do so is through the emotions. Emotions are temporary physical and mental states that prepare the organism for adaptive action. The suggestion from studies on crabs, then, is that arthropods have felt emotions, hence they are sentient.
After a long period of Cartesian denials of animal emotions owing to doubts about what they feel, or even skepticism that they feel at all, most affective scientists have settled on a distinction between feelings and emotions. Feelings are private conscious states that are not publicly observable and hence are inaccessible to science, whereas the emotions are measurable physiological and/or neural states that are often reflected in behavior. Affective science concentrates on the emotions of animals while generally staying neutral on their associated feelings (4).
The distinction between emotions and feelings is less relevant in our species because humans communicate feelings by means of language. Humans accept verbal statements about internal states as evidence of those states, and conversely sometimes equate the absence of language with the absence of those states. Until well into the 1980s, this argument was even extended to members of our own species. Skeptical that preverbal human babies felt anything, the medical profession freely subjected them to surgery without anesthesia (5). Verbal reports of private states are not the same as the states themselves, however, but rather they are a source of evidence. In relation to human babies, the linguistic focus has since been abandoned and their pain is taken into account. We think it is time for the same step in relation to nonhuman animals. To deny them felt emotions does not seem a reasonable position given the fundamental similarity between the nervous systems of humans and other animal species and the shared evolutionary history that has promoted similar emotionally mediated reactions to the environment and social partners.
Affective neuroscience relies on animals to study which brain areas and neural circuits are activated during specific emotional reactions, such as fear, anger, disgust, and attraction. These findings are often extended to humans. This work began with observations of animal behavior suggestive of emotional states. For example, the face—the proverbial window on human emotions—expresses emotions through similar muscular contractions to those in other primates owing to a homologous facial musculature, which is nearly indistinguishable between humans and chimpanzees (6). Obviously, increasingly distant species have increasingly different expressions of the emotions, but research has found that, for example, physiological changes, lowered temperature of the extremities, and activation of the amygdala during fear are notably similar in fearful rats and fearful humans. Even outside of the mammals, continuities in the expression and underlying physiology of the emotions are observed. Research in comparative endocrinology, for example, finds a role for glucocorticoids (stress) and an oxytocin-/vasopressin-type peptide system (reproduction, bonding) across a wide range of animals that includes invertebrates (7).
Behavioral tools to evaluate emotional responses have been sharpened, too, such as the cognitive bias test. In this paradigm, investigators first manipulate the animals’ environment, making it rich versus impoverished, or calm versus disturbing. Then they train subjects on two different stimuli, such as a high versus a low tone, which triggers either positive or negative reinforcement. By subsequently presenting the animals with an ambiguous stimulus, such as an intermediate tone, they can gauge their optimism (approach) or pessimism (avoidance) about the stimulus. The typical outcome in mammals and birds has been that a poor environment negatively affects their expectations about the world (8).
This innovative testing paradigm, which offers an index of the animals’ felt emotions, was recently extended to honey bees. Bees subjected to vigorous physical agitation (shaking) to simulate a predatory attack proved less willing to explore new tastes, and hence were negatively biased by their experience. They also showed reduced amounts of hemolymph dopamine, octopamine, and serotonin. Changes in these neurotransmitters mark anxiety or depression in humans. In any mammal, a pessimistic behavioral response associated with physiological indications of stress would be taken as a sign of negative emotions. The same logic should be applied to insects (9).
Animals are exquisitely sensitive to signs of emotion in conspecifics, resulting in empathic reactions, such as consolation of distressed individuals by means of body contact, and synchronized behavior, such as when the fear and alarm of one individual arouses fear and alarm in nearby conspecifics (10). When humans show empathy, be it through a hug or verbally with “I feel your pain,” this is perceived as them also experiencing the other’s emotion—and if they do not show the appropriate emotional reactions, the empathy is not judged to be genuine. In a recent study of prairie voles, the stress of another individual increased its mate’s circulating corticosterone concentrations and led to a comforting response that could be blocked by an oxytocin receptor antagonist injected into the anterior cingulate cortex. This suggests an empathy response regulated by neural mechanisms that are conserved between rodents and humans (11).
Even though feelings can (as yet) not be directly demonstrated in other animals, these findings strongly hint at their existence. Recognizing that there is only indirect evidence of human feelings as well—humans do not directly experience what other people feel—it seems appropriate to adopt a stance of consistency and evolutionary parsimony such that similar emotional expressions and processes in related species call for similar psychological descriptions when it comes to sentience. The frequent exhortation to avoid anthropomorphic language is often misguided, especially for species that are evolutionarily close to humans, because it forces us into linguistic distinctions between their behavior and ours that remain unsupported by evidence for an actual distinction. Warnings of anthropomorphism assume that there are stark differences between humans and animals, an assumption that is questionable given that humans are animals (1213). Although every animal species is distinct, science recognizes fundamental similarities when it comes to neurophysiology, cognition, emotions, and sentience.
It is not hard to see that the denial of animal emotions, and by extension animal feelings, has been morally convenient during human’s history of animal exploitation. Conversely, their recognition is bound to shake up our moral decision-making. This is because sentience is widely taken to be sufficient for what ethicists call moral standing, or the status of having one’s interests morally matter. What this means in practice is that to act morally, a person must take into account those individuals with moral standing. Although we are used to thinking about how our actions affect other humans, recognizing widespread animal sentience requires us to also notice—and consider—our impact on other species. This way, animal sentience is bound to complicate an already complex moral world.
If crabs experience emotional states, then they have an interest in these states being positively valenced. Current research indicates that a wide range of animals have interests in avoiding felt pain, and that they would not consent to painful procedures if given the opportunity. This makes pain a harm to them, and thus places a moral obligation on humans to recognize that harm and to avoid causing it if possible.
With an increasing number of species whose emotional states have moral standing, what follows about how exactly we should treat them? Here there remain more questions than answers. To make progress requires knowing the range and variety of animal emotions and feelings. It is not much of a stretch to presume that for many species, avoidance of pain is just one interest among many. For example, animal empathy research may find that cows are distressed by seeing their calf in pain, an additional interest that goes beyond the desire to avoid their own pain. These added interests also need to be considered when making decisions about good husbandry and agricultural practices (14). With more interests among more species come more pronounced and numerous conflicts of interest between humans and other animals.
Philosophical ethics offers theories and tools for managing conflicts of interests, although without understanding the wide range of features that may be morally relevant, the tools are only of limited help. For this reason, we think that greater integration between ethics and affective science by ethologists, comparative psychologists, and neuroscientists is needed to reach moral conclusions. This integration cannot come from ethicists alone but requires students of animal behavior to engage with the ethical implications of their work, which until now most have been reluctant to do (15).
When the medical community recognized infant pain in the 1980s, it was because the evidence was so overwhelming that physicians could no longer act as if infants are immune to pain. A similar point is being reached where invertebrates can no longer be treated as if they only have a nociceptive response to harmful stimuli. If they can no longer be considered immune to felt pain, invertebrate experiences will need to become part of our species’ moral landscape.

References and Notes

J. Birch et al., Review of the Evidence of Sentience in Cephalopod Molluscs and Decapod Crustaceans (London School of Economics and Political Science, 2021).
V. A. Braithwaite, F. A. Huntingford, Anim. Welf. 13, S87 (2004).
M. Appel, R. W. Elwood, Appl. Anim. Behav. Sci. 119, 120 (2009).
D. J. Anderson, R. Adolphs, Cell 157, 187 (2014).
E. N. Rodkey, R. Pillai Riddell, J. Pain 14, 338 (2013).
A. M. Burrows, B. M. Waller, L. A. Parr, C. J. Bonar, J. Anat. 208, 153 (2006).
H. Ando, K. Ukena, S. Nagata, Eds., Handbook of Hormones Comparative Endocrinology for Basic and Clinical Research (Elsevier, ed. 2, 2021).
M. Mendl, O. H. P. Burman, R. M. A. Parker, E. S. Paul, Appl. Anim. Behav. Sci. 118, 161 (2009).
M. Bateson, S. Desire, S. E. Gartside, G. A. Wright, Curr. Biol. 21, 1070 (2011).
F. B. M. de Waal, S. D. Preston, Nat. Rev. Neurosci. 18, 498 (2017).
J. P. Burkett et al., Science 351, 375 (2016).
F. B. M. de Waal, Philos. Top. 27, 255 (1999).
K. Andrews, How to Study Animal Minds (Cambridge Univ. Press, 2020).
S. Monsó, J. Benz-Schwarzburg, A. Bremhorst, J. Ethics 22, 283 (2018).
C. E. Webb, P. Woodford, E. Huchard, BioEssays 42, 2000030 (2020).


Science shows that dogs feel things like us. Legislation must catch up

Research supports what Darwin said in 1872 – dogs express emotions in a way recognisable to humans. Governments must do more to protect them, says Jules Howard in New Scientist: SOCIETY COMMENT 25 May 2022: https://www.newscientist.com/article/mg25433881-100-science-shows-that-dogs-feel-things-like-us-legislation-must-catch-up/#ixzz7VAQsJaco 

“Research has shown: how the emotional centres of dog brains, observable through fMRI scans, fire up like human ones; that the concentrations of hormones and neurotransmitters associated with emotional responses rise and fall predictably, just as they do in our own brains; that the sociability of dogs – specifically, their interest in us – is found in their genes; and that, in the words of psychology, dogs are capable of strong attachments with us that mirror those seen between human caregivers and children.” —Read more: https://www.newscientist.com/article/mg25433881-100-science-shows-that-dogs-feel-things-like-us-legislation-must-catch-up/#ixzz7VAVvG6Yd