The Complexity of Animal Emotions: What Are They, and How Can We Measure Them?

Read time: 20 minutes

What you will learn in this article:
  • How do scientists view animal emotions?
  • What are Cognitive Bias Tests, and what can the brain tell us?
  • Are animal emotions sentient experiences?

As the dominant species on our planet, we are quick to anthropomorphise. We readily assign human emotions to much of the non-human animals we share our planet with, especially our domesticated cat and dog companions – yet we deny it in others who we are less able to relate to, like reptiles and crustaceans. In the scientific community, there is concern for how the general public often view animal welfare as a ‘beauty contest’, favouring species more visually appealing to us as humans. But does an animal have to look or behave a certain way to experience emotional processes? What about the brain – does size matter? Every dog owner is familiar with that jealous paw tap should you dare to pet another canine, but measuring or even defining such complex emotions as jealousy in animals is difficult. Is that attention-grabbing paw tap an exhibition of conscious jealousy as we know it – or an unconscious evolutionary survival mechanism, a benign form of social resource guarding concerning dominance hierarchy? What about other species that don’t possess such obvious physical indicators as a wagging tail, E.G. fish, birds or rodents – do they have similar experiences of emotion? Without the luxury of linguistic report, fully understanding animal emotion is a difficult endeavour, but science has concluded that at least ‘basic’ emotions are continuous expressions across taxa.

Emotions vs Moods


Scientists define animal emotions as ‘short-lived responses to stimuli, designed to guide animals towards rewards and away from danger’. In this sense, emotions serve as adaptive survival mechanisms with an array of neural, physiological and behavioural components in order to respond effectively to stimuli and situations. This theory is supported by Darwin’s evolutionary approach.


Where emotions are short-lived responses to stimuli, like a startle response to a predator, moods are instead considered longer-term ‘affective states’ unrelated to stimuli, like depression or excitement. Affective states can be negative like chronic fear, or positive like contentment, and are not caused by a single stimulus but are the result of an accumulation of experiences.

Why measure mood? And how?

Measuring these affective states gives us a glimpse into an animal’s frame of mind, so to speak: do they view the glass as half full, or half empty? Assessing an individual’s emotional state is important for welfare as these states can impact cognitive processes like memory, attention and judgement – as well as being an indicator of psychological well-being in general.

Scientists measure these states in various ways: physiological measurements (glucocorticoid measurements, heart rate etc) combined with behavioural observations (pupil dilation, ear posture etc) can indicate states, however are difficult to interpret. For example: an increase in heart rate, adrenaline or cortisol can indicate either negative or positive valence, like escape from predation or anticipation of positive stimuli. Some behavioural measurements, like stereotypies, are widely accredited as useful indicators of poor psychological well-being – but that’s a whole other article in itself. Behavioural and physiological measurements are probably the most feasible indicators of animal moods for many settings like zoos, farms, sanctuaries, equine centres, pet owners etc because they don’t involve any extensive training periods – but, they are arguably less accurate due to difficulty in interpretation. Scientists also measure affective states using a variety of tests. Cognitive bias testing is widely accredited as a useful method of assessing animal moods and is probably the most accurate way to determine an individual’s emotional state (as well as busting myths certain animals don’t experience these psychological states like rats, sheep and birds to name a few) – but it requires some animal training, as you’ll see below.

The world’s first Professor of Animal Welfare Science was only appointed in 1986 [Donald Broom, Cambridge University]. There is still a wealth of knowledge left to discover, and much we don’t yet understand about the animal kingdom.

Examples of Cognitive Bias Tests

Cognitive bias testing measures how an individual’s emotional state has affected how they process information. Attention bias testing is an effective method, and sheep have been widely used to demonstrate this, especially for depression and anxiety. With humans, anxious people tend to spend more time worrying about threatening information than non-anxious people – the same is true for animals, who can also possess this frame of mind. Unfortunately, animals can’t tell us how they’re feeling – but we can confirm whether these animals are experiencing anxious states through attention bias testing. Accuracy of these tests have been evidenced by using environmental and pharmaceutical manipulations to induce anxiety, then measuring attention response rates towards threatening stimuli. For example: the time it takes for sheep to approach a food reward after a threatening stimulus (like a dog seen through a glass window) has been removed, and the time spent attentive to the potential threat. In this example, anxious sheep will spend more time being vigilant than the non-anxious Control sheep. This 2016 research is a good example; and this 2018 study shows how depression can also be identified.

Bias testing is an effective means of assessing animal moods: however is impractical for real-time welfare assessments in zoos or farms because of the extensive training period

Judgement bias testing is one of the most effective ways to assesses an animal’s state of mind. The underlying hypothesis is that an animal with a negative state will make negative judgements about ambiguous stimuli, or view the world in a negative light to be more general. The first published study on cognitive bias illustrated this in 2004; rats in unpredictable housing conditions were hypothesised to have negative affective states, contrary to other rats in stable control conditions. The experiment involved training rats to associate sounds with events. When presented with two cues in the form of a tone of particular frequency, both sets of rats were trained to perform a response which would either A) avoid a negative event, or B) illicit a positive event respectively. What about when the rats then heard an unenforced ambiguous tone, which was not associated with either event? Will the ‘sad’ rats categorise it as negative, while the ‘happy’ rats don’t? Yes, that’s exactly what happened: confirming the researchers’ hypothesis that rats with a negative cognitive bias will categorise the ambiguous tone as negative, and perform the response to avoid a negative event. This approach to measuring states has been widely used in lab and farm settings, using monkeys, rats, dogs, sheep, birds and humans – and results support the validity of the cognitive bias testing hypothesis.

People tend to remember criticism and forget compliments. The same is true for non-human animals, who appear to be more sensitive to losses than gains, especially if they have a negative cognitive bias. A well known 2008 study evidencing this looks at the speed at which rats run towards a food reward, which will become slower after researchers decrease the reward value – whereas there is no change in speed for rats who have had the lower reward value from the beginning. “You don’t miss what you never had”, as the saying goes.

Is Bigger Better?

Historically, brain size was viewed as a proxy for cognitive abilities, but today scientists dismiss this simplistic view and the topic remains in debate. Many animals have much bigger brains than humans – our brain weighs roughly 1350g, whereas an elephant brain is about 5kg, and a blue whale weighs in around 7kg: simple comparisons of brain size across species does not equate intelligence. Brain size relative to body mass is another avenue science has probed, but again with contradicting results. The human brain occupies 2% of our body mass, whereas the blue whale’s brain only occupies about 0.01% – this would be a good theory aligning with our expectations, except for the fact that the shrew and harvest mouse come up tops with a brain occupying between 4% – 10% of their total body mass. Back to the drawing board. The importance of certain parts of the brain in intelligence is still hotly debated – for example, the cerebral cortex especially the neocortex is considered responsible for higher cognition and so dubbed the ‘seat of consciousness’. Some believe it is necessary for animals to possess this structure to experience pain or emotion, whilst others consider lack of a neocortex to be rather flimsy grounds for denying pain in non-mammals, as evidence has pointed that emotions can derive from parts of the brain outside this. This is something supported by the 2012 Cambridge Declaration on Consciousness (a prominent international group of cognitive neuroscientists, neuropharmacologists, neurophysiologists, neuroanatomists and computational neuroscientists), which stated

‘the neural substrates for emotions do not appear to be confined to cortical structures’. The latest theory compares brain size – body mass again, but this time using ‘allometry’: in simple terms, evaluating characteristics across species and identifying species whose brains are larger or smaller than expected for their body size. Results for the main vertebrate groups largely align with the behavioural complexity observed in each species, so this is the theory most scientists are running with today, but of course it is a theory and not yet a fact. However, brain complexity and brain size can change independently of one another, and a brain can become more intelligent relative to its ancestral species whilst decreasing in size. Neural complexity in insects is among the most noteworthy displays in the animal kingdom, and the underlying component of the nervous system which deals with cognitive and social behaviour has changed in complexity for several insects including butterflies and bees, and according to some the latter can be compared to the folding of the mammalian cortical surface. The role brain size plays in determining intelligence is still under much investigation and there are many contradicting studies – E.G. this research on guppies in 2018 suggests brain size influences behavioural plasticity. Biological complexity and intelligence is evidenced across species, and the exact selection pressures remain in debate today.

What Else About The Brain?

Okay, so neural measurement is another way to measure emotional processes. Research has mostly focused on mammals, so the basic common design of the limbic system, brainstem, cerebellum and cerebral cortex is now well understood – albeit knowledge on evolution and function has come along in leaps and bounds in the last two decades, with revisions of longstanding major hypotheses such as the Triune brain supposition. Similarly, the role of key brain areas (hippocampus, amygdala and ventral pallidium) in emotional processing is a hot topic in animal emotion research. Old experiments included electrical stimulation and lesioning – no way these methods would pass today’s rigid ethical standards! Neuroimaging study has focused a lot on correlations between animals and humans to better understand which areas of the brain are involved with which processes – aka, what parts of the brain light up under different conditions. Studies have used awake non-human primates for more detailed comparisons of the brain responding to stimuli and performing cognitive functions (so cool!), and has shown there is a striking similarity between humans and dogs in the caudate nucleus, which is associated with emotion and pleasure – investigations have shown dogs experience positive emotions, empathetic responses and human bonding as well as understanding emotional vocal cues and processing faces! So yes, neuroscience now backs up what dog owners could already tell you.

Though we have learnt a lot, much remains to be discovered. For example, for decades the amygdala has been recognised as playing an important role in fear – but general consensus now holds that it is also important in various other contexts, and our understanding of its role in reward learning and social behaviour falls short. To summarise, there is still a ton we simply don’t understand about different areas of the mammalian brain – nevermind the how, why, where and when of emotion in animal brains of different classes. Measuring neural function is technically complex, and isn’t a practical method of assessing animal emotions and welfare outside of an academic context – so although it provides a worthwhile basis for neural processes of emotion, it’ll have to remain in the lab, as it’s not a feasible tool for determining emotional responses in zoo or farm settings.

Are Animal Emotions Sentient Experiences?

Ah, this is the question. We know that basic or ‘primitive’ emotions are continuous expressions across taxa. For example, fear is a basic emotion and an essential survival tool, guiding animals away from harm. In its most basic sense, animal emotion can then be considered a product of evolution – unconscious survival mechanisms, void of subjective experience. However, as we’ve discussed, animals can experience different frames of mind as results of their environment – and what is ‘sentience’ if not the ability to experience psychological states of pleasure or pain? Determining just how aware animals are of these psychological states they are experiencing is an enigma science has only just begun to unravel, and as empathetic humans considering animals to be void of any kind of conscious experience just feels downright wrong. Thankfully, this is a perspective becoming increasingly outdated within the scientific community.

This reminds us that we often underestimate that which we do not understand

Behavioural plasticity (adjusting behaviour in response to complex environmental conditions) has long been thought to indicate superior cognitive ability, mainly in terms of learning and adaptability – thus an inference for consciousness. Complex behaviours certainly require a greater degree of awareness, but in the words of Balcombe, ‘intelligence’ should be recognised as a suite of abilities specific to each species’ ecological challenges, not one single property. Fish are a good example. Because fish do not scream when impaled with a hook, and have no discernible facial expressions, we assume they do not feel pain – that they are simple stupid creatures, only capable of basic cognitive processes. Research has shown us we are wrong – that fish have the cognitive capacity to plan, learn, use tools, hold long-term memories; there is evidence of perspective taking, social cohesion, complex relationships, and individual recognition. Goldfish, viewed by many as the pinnacle of fish stupidity, are actually viewed very differently by scientists; contrary to popular belief, the urban myth that goldfish have 3 second memories has been disproved time and time again by research which has instead shown them to possess not only an impressive memory, but also problem-solving skills: E.G. in this 2003 research where goldfish learned to operate a lever for food and,

even more impressively, independently discriminate which 1hr time slot out of 24hrs was the correct time to do so! In 2012 a now famous experiment showed adult cleaner wrasses outperform chimpanzees, orangutans and capuchin monkeys! Two plates were presented to the animals – one was blue and one red. When the animals took the food from the blue plate, the red was taken away (so halving their meal), but when animals took food from the red plate first, the blue plate remained. It took the fish about 45 trials to figure this out – some of the primates never did, even after 100 trials. The experiment was based on decision making, which is important for the wrasses who are ‘cleaner fish’ and must choose to service visiting reef fish before resident reef fish in order to get both meals. This shows again intelligent behaviour depends on what challenges a species faces in the wild. If complex behaviours indicate sentience, then we should regard fish as we do our furry companions – a striking claim, hard to believe for most of us, yet a logical conclusion nonetheless. This reminds us that we often underestimate that which we do not understand and we, the human race, are but one of the millions of species on this planet – a mere pin-drop in an ocean of life.

Food For Thought

Research results are consistent, and although it may never be possible to prove beyond a doubt that animals experience sentient emotions, many scientists believe we should err on the side of caution. Translating this science into real life changes however is a failing endeavour – E.G. the everyday act of boiling crustaceans like crabs and lobsters for dinner. From my own personal experience, every summer a trip to our holiday home in the West coast of Ireland would be accompanied by throwing in the lobster pots, hopefully catching one or two for tea. As a child I was distressed to see the lobsters boiled alive, and the high pitched whines and whistles that bubbled up relentlessly from under the lid ensured however yummy the tea was, it came sprinkled with a side of guilt. I was reassured calmly by my father that they didn’t feel anything, because placing them in the freezer for a few hours beforehand induces a state of hibernation – then when placed in the boiling water, it’s over before they know it. Reluctantly I accepted – as do millions of people across the globe today casting in their own lobster pots or ordering in restaurants. Now I know there is no clear evidence that freezing crustaceans to torpor induces anaesthesia rather than just paralysis, and freezing could be harmful in itself. The technique originates not from welfare scientists but chefs, in order to immobilise the lobsters who otherwise thrash wildly and scrape the sides of the pot in desperate attempts to escape. I’ve reflected greatly on why lobsters and crabs earn no sympathy; perhaps the misconception lies with size, and people deem these animals too small to feel pain or fear? This seems unlikely though as many other furrier animals are of similar size,

and closing the lid on either a budgie or a rat is likely to be met with outcries of animal cruelty. Perhaps it’s the way they look, then; does the scaly exterior remove it in our minds as an animal, replacing it instead as some strange, hard creature from an alien world? If they had fur, would that change anything? What if they communicated vocally; if lobsters sang like birds, or huffed like bears would we still boil them alive? Maybe our understanding is too limited, and as a rule only extends so far into the underwater world so opposite to our own, despite being two sides of the same coin. If growing evidence suggests lobsters can feel pain, should they not be slaughtered in a more humane way? Recent research has also shown crustaceans display the hallmarks of stress and anxiety like those we accept in sheep. Boiling lobsters alive has been illegal in New Zealand since 1999, and as of 2018 Switzerland too – but remains the norm in many other countries. The science is there, but change it seems is dependent upon social attitudes and norms. Perhaps lobsters are just too removed from us as humans, and too different from the other forms of life whose validity we are comfortable accepting. This need for a wiser compassion extends well beyond how we put food on our plates, but also to what we do for fun; the pets we keep, the places we visit, the clothes we wear and even the beauty treatments we buy: if you could see where those mink eyelashes came from, would you still want them on your face? Nobody’s perfect, but being kinder to animals is easy, it just requires asking the right questions: is this coat real fur? Are these lashes synthetic or mink? Where did my pet shop source my ________ from, and where did the breeder get the parents / grandparents?

How do non-human animals perceive and experience the world? This remains to be one of the greatest mysteries only partially unravelled by science, and we still have a long way to go before we fully understand it

Can intelligence be a true indicator of sentience? Can emotional intelligence infer an animal’s level of consciousness? Moreover – should animal welfare standards be dependent upon the degree to which suffering is experienced in a subjective sense? We are but one of the millions of living species on the planet. We know that animals possess emotions. The interesting question now lies with what degree of conscious awareness accompanies such emotive states: are such responses survival mechanisms, devoid of self / external awareness, or are they indications of sentience? This is a mystery science may never unravel, however the multitude of evidence supporting noteworthy cognition is enough to warrant changing global perspectives on the morality of welfare standards for animals across many industries.

“A mind does not have to be human to suffer”

Jonathan Balcombe

Read More:

Frustrated some papers aren’t free?This blog definitely doesn’t advise copying & pasting the paper’s DOI link (usually found at the top under the title) into SciHub. SciHub is technically an illegal ‘pirate’ site (equivalent to movie streaming sites except without the pop-up ads or viruses) which believes access to science should be available for everyone. University lecturers and students definitely never use this site to access free scientific papers… The SciHub domain changes frequently but the current one is Don’t do it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s