Animal sentience - An essay
What is animal sentience and how can it be investigated?
Hello again to all!
It is Sunday again and I hope you have all had a good week, and if you celebrate Easter then I hope you have been enjoying the holiday. This week I read over one of my past essays which I wrote for a university assignment. The title was “What is animal sentience and how can it be investigated?”. The topic is one which I personally feel quite strongly about and I believe it is one of the most fundamental aspects to consider when studying animal behaviour. I often like to think of the animal mind as a “black box”, wherein we cannot know the processes which occur within it, only the output in terms of behaviour. One of my key fascinations with ethology is how we can interrogate the animal mind.
Animal sentience is one of the many aspects of the “black box” - how can we infer and investigate if animals have emotions and feelings. I will present you my essay on the subject. An important side-note I feel I should make: I have written this essay based on the animal sentience laws and practices of the UK, and this is likely to be vastly different in other countries.
Animal Sentience: What is it and how can it be investigated?
“Animals are like robots: they cannot reason or feel pain” - Descartes, 1596–1650.
While this quote may seem outdated in terms of its message and its timeframe, the question must be asked, is it really that much different from the way in which some animals are still treated today? Animals are mistreated for human benefit on a daily basis, used like machines for food, research and profit. One of the major limitations to animal welfare is the lack of knowledge surrounding animal sentience, and what this means for the animal itself. One such way to exemplify this is to ask yourself: what would it take to convince you that your pet dog can feel things, such as pain or pleasure, just as you and I can? There is a great probability that at least some of the people reading this would already consider a dog to be a ‘sentient’ creature. Now consider a crustacean, such as a lobster. Does your perception of sentience differ? In other words, do you believe it can feel pleasure or pain. I’m sure for a majority of those reading, this would be the case. It is more difficult to empathise with a lobster, as we do not feel such a strong attachment or understanding of it. Many would consider it very simplistic compared to a mammal such as a dog and would dismiss the idea of such a creature being capable of feeling emotion. Despite this, recent UK government legislation (namely the Animal Welfare (Sentience) Act 2022) has declared that any vertebrate other than humans, any cephalopod mollusc, and any decapod crustacean, is to be considered sentient. But why is this the case? What kind of measures have been employed to determine if an animal should be deemed sentient under law? These are all poignant questions, but firstly it is important to address the elephant in the room: what do we mean when we talk about sentience?
A brief history of animal sentience
“The question is not, can they reason? nor, can they talk? but, can they suffer?” – Bentham, 1823.
The perception of animal sentience has greatly changed over time. The idea of animal sentience is one that was lost in common behavioural sciences such as ethology until Ruth Harrison wrote “Animal Machines” in 1964, where she sheds light upon the poor welfare of farm animals in modern industrial farming. She wrote of how physiological stress response was indicative of welfare; that a stressed animal was in poor welfare and one that was not stressed was in good welfare. However, what truly troubled her was not that the animals were stressed, but that they were sentient and could feel stressed. This book gave rise to an increasing public demand for animal welfare assessments, and a follow-up investigation by the British government was published in 1965, the so-called ‘Brambell Report”. One of the main ideas published in this report was that for animals to be in good welfare, they must be allowed the ‘five freedoms’ (now known as the ‘five domains model’). These five freedoms were the basis for modern animal welfare, and intrinsic to these freedoms was the assumption of sentience; that these animals could feel things such as pain and pleasure, and that these feelings mattered. However, sentience was not directly acknowledged at this stage, and only that the physical well-being of the animal was all that was required for good welfare – feelings didn’t matter.
In 1980, the publication of Marian Dawkins’ book “Animal Suffering” caused behavioural scientists to gradually accept the importance of feelings in their investigations into animal welfare problems. Before Dawkins’ publication, animal physical health was thought as the be-all-and-end-all of animal welfare, that if an animal grew well, was in good health, was reproduced successfully, and was relatively stress free, then it must be in good welfare. This biological functioning framework or orientation emphasised physiological measures of wellbeing, with little thought for their mental well-being. At that time, scientists did not like to imagine that animals could experience mental states such as emotions or motivations, as these were seen to be anthropomorphic speculations with no scientific basis. However, during the 1980s, there was a progression from the idea that feelings were unnecessary to welfare to one in which feelings are the only thing that matters.
“A sentient animal is one for whom feelings matter.” - John Webster, Professor Emeritus, University of Bristol, 2005.
One scientist in particular, John Webster, made a significant impact of the view of animal sentience, with the integration of biological functioning and affective states in his book ‘Limping towards Eden’. He made explicit, in his simple definition of animal sentience, what allowed an animal to be deemed sentient – “feelings that matter” (Webster, 2005). He showed that animals do not simply experience a stimulus, such as hunger, and then respond to it, by salivating, but that they interpret the information based on prior experience, which evokes an emotional response. For example, Pavlov’s dogs would salivate even in the absence of food if they heard a bell which has prior association with a food reward. Furthermore, emotional responses will be modified with prior experience, whether it be positive or negative, and this will change their motivation to perform a certain action. Therefore, sentience allows for the development of plastic responses to stimuli over time. This means that animals which are sentient will not only have distinct affective state, but will also develop a motivation or aversion to a stimulus based on what it perceives will be the likely outcome. Hence, “feelings that matter”.
The Oxford English Dictionary may define sentient as “able to perceive or feel things”, but I think this definition is rather ill equipped to explain the complexity of sentience, and Webster, even with his simple definition, does a much better job. However, the fullest and most encompassing definition is one provided by the RSPCA – “the capacity to experience positive and negative feelings such as pleasure, joy, pain, and distress that matter to the individual”. Both Webster’s and the RSPCA’s definition both include the sentiment that feelings ‘matter’. But what does this word mean in a biological context?
The word relates to both positive and negative experiences, implying that these emotions have a significant impact of the animal feeling them and will influence future decisions. Negative experiences ‘matter’ in the sense that some negative experiences are critical to survival, such as hunger or pain. This motivates the animal to engage in behaviours which improve its chance of survival, such as eating or escaping danger. Other negative experiences are not directly survival related, but more situational, such as frustration, anxiety or fear. These negative experiences become associated with certain stimuli, and so the animal will try to avoid these. Positive experiences matter because they not only feel good, but the anticipation of such experiences will derive positive emotions and will enhance an animal’s welfare. Therefore, feelings which matter to an animal are important to the idea of sentience because an animal will react accordingly to stimuli which increase the intensity or duration of these emotions. This in itself is very important, because without proof that animals are sentient, we would not have legal acknowledgement of such a phenomenon.
Measuring sentience: The scientific paradox
“It’s very important that we try and find ways of answering these questions because really a lot hangs on it.” – Jonathan Birch, Philosopher, 2020.
Now that we have established what sentience is, and why it matters to animals, it is now a question of how we can measure it scientifically. The acceptance that welfare was all about feelings brought with it a seemingly impossible task, one which can never be conclusively proved nor disproved. How can subjective feelings be proved in an objective manner? Well, until the day that a dog or a cow or an octopus learns to talk, these feelings are only available for the animal experiencing them. Nonetheless, I would argue that just because something is difficult to study scientifically, doesn’t mean it shouldn’t be studied at all. It is clear that others share my feelings on this topic, one example being Marian Dawkins, who in her 1993 paper “Through Our Eyes Only? The Search for Animal Consciousness” (makes a plea for) methods which can investigate how animals are feeling. Since this paper, methods have been developed to indirectly study animal sentience, some which were already in existence, but were refined to study sentience, and some novel.
There are some physiological measures which can be used as a baseline to infer sentience. Firstly, the presence of a centralised nervous system (or CNS). Scientists have shown that simply the presence of a nervous system is not sufficient for sentience unless that nervous system is centralised. This is because the structure and associated functioning between receptors and synapses which are directed to the CNS allow for conscious experience rather than simply reaction to a stimulus. However, CNS complexity varies greatly, from simply a collection of nerve ganglia all the way to fully formed brains. Further, the degree of centralisation is also extremely varied. For example, octopus have a more complex CNS than many vertebrates, and despite their difference in life history and brain structure to ours, exhibit many complex behaviours indicative of consciousness. This suggests that the CNS organisation needed for sentience may in fact be quite simple. It is on this basis that cephalopod molluscs are deemed sentient by law. Another potential more neural measure of sentience is the presence of nociceptors. Nociceptors are receptors which allow the animal to experience and respond to noxious stimuli, that have the potential to be damaging. These are present in many animals but are shown to cause the subjective feeling of pain in sentient animals. While the presence of these receptors alone don’t prove sentience, they are certainly a necessary precursor to it.
Physiological measures alone are not sufficient to prove sentience, as they are simply measures taken at a given time, and do not reflect consciousness, simply responses to a stimulus. However, they are still very useful when combined with behavioural indicators to compound the evidence for sentience. Nonetheless, it is worth noting that the philosopher Jonathan Birch promotes the idea of a ‘precautionary principle’, by which the “bar” for sentience is set quite low, so that little evidence is needed before preventative action is taken, such as protecting a species under law. Thus, physiological measures alone are not enough to conclusively prove sentience, but they are enough to apply the precautionary principle.
Animal sentience can also be investigated by indirect methods. One such method, and the first of these to be developed, is the preference test. Preference tests established by two scientists independently, Barry Hughes, and Marian Dawkins, in the 1970s. Both scientists were interested in laying hens, at a time when it was commonplace to keep these birds in battery cages. They were interested in the question “do hens suffer in battery cages?”. To assess this, they gave the birds a choice between a spacious or confined cage and allowed the hens the freedom to choose which one they preferred. It can be assumed that the hens will choose the option which is in better interest of their welfare. Unsurprisingly, they chose the spacious cages significantly more. However, there are flaws with preference tests. Whilst they are a good first step to determining how an animal feels in a certain environment, preference tests alone cannot address the question of ‘how important is a particular choice?’. Even when a choice in a given direction is consistent, it does not necessarily mean that this choice is important, as it is possible to argue that neither choice causes suffering to the animal. Despite this seeming obvious in the example mentioned above, this question is something that needs addressed in other cases, and in particular, for animals that are not necessarily deemed as sentient. This links back to our definition of sentience, where animals that are sentient will experience positive and negative emotions that matter; feelings that are important to them.
There are several methods for measuring the strength of a preference. One such method is by making the animal work for something, and to determine how motivated it is to perform a certain behaviour or obtain a certain commodity – this is known as a motivation test. There are many examples of motivation tests, especially in farm animals. They are often used to determine how much natural behaviours matter to these animals, as in confined farming systems they are usually not able to perform these. Gareth Arnott and colleagues carried out a study in dairy cows in 2017, to see how far they were willing to walk to access open, outdoor pasture. They varied the distances that the cows were required to walk to access the pasture, to assess how important this behaviour was to them. It seemed that at nighttime, when the cows wanted to lie down and sleep, they were willing to travel any distance to access pasture, as it was assumed that the pasture was more comfortable to lie down on, showing that the pleasure of comfort was important to these animals. Another study in sows, carried out by Marie Haskell and colleagues in 1997, used a motivational test to determine how important nest-building behaviour was to sows who were close to giving birth. This behaviour is a natural behaviour which sows perform in the wild, and it is restricted in commercial farming systems. The research group gave the sows a pen which was barren, but she could press a lever to access a larger pen in which she had space to walk around and collect materials for nest building. Pressing the lever required effort, and so it shows how motivated the sow is to work to perform this behaviour. They showed that as the sow came close to giving birth, she was willing to work considerably more to access this larger pen, which tells us that this behaviour matters to sows.
Preference is not the only way to understand how an animal is feeling. As previously mentioned, our inability to communicate with animals limits us in terms of how we can infer their emotional, or affective, state. Preferences only give an indication of what matters to an animal, but fail to interpret discrete emotions such as excitement, calmness, depression or fear. These emotions are the result of subjective experiences, and make up the mental state, or affective state, of an animal. This can be difficult to measure when we cannot physically see the result of these, but we can make inferences from an animal’s behaviour. Mike Mendl and colleagues devised a very useful framework for understanding an animal’s affective state (Mendl et al., 2010). They characterise subjective experiences into valence (positive or negative emotions) and arousal (the level of activity the animal shows in a behavioural sense).
By using this framework, it is possible to make assumptions of the effect that environmental conditions and different stimuli will have on subjective emotions based on the transition between quadrants which are caused by functional responses. It helps us understand and assess the affective states of animals, which is something that is relatively understudied but is key to understanding animal sentience. I think it is fair to say that there is no question of animal sentience. It is a fact, and as we begin to move forward, we must acknowledge and take into consideration, as almost every aspect of human life has an impact on animals – direct or indirect. So, the next time that you see a dog or a lobster, or any animal for that matter, take a moment and appreciate that it is a sentient being with feelings that matter.
In summary, this essays details the thinking behind testing sentience, and I believe that we have sufficient scientific evidence to say that animals can feel emotion. It answers some important points, however, I could have added a final section where I discuss what this all means for the future of ethology:
The future of animal science is exciting and brilliant. Never has there been such enthusiasm and drive to learn about what animals think and feel, evident in the addition of decapods to the UK Animal Sentience Act 2022. These changes will hopefully promote this area of science, and showcase the growing body of research that demonstrates the diverse nature of the emotional lives of animals, especially those of farm animals. As animal welfare science continues to grow, understanding of the needs of farm animals and how we can make their lives worth living. Despite the difficulties of studying consciousness in non-human species, this does not mean that we should abandon the search for consciousness and understanding the baffling problem of the relationship between brains and experience in other species. On the contrary, facing up to the difficulties could set us free to be more objective about what we do and do not understand. We can investigate the mechanisms of behaviour and whether or not they involve conscious pathways without the pressure of feeling that it is “good” (for animal welfare) to prove consciousness in a given species or “bad” (for animal welfare) or dangerous to question it (Bekoff, 2012). The study of animal consciousness is far too important and far too difficult for us to pretend we have demonstrated it when have not or to underestimate the very real difficulties of making it 532 demonstrable in the future.
If you made it this far, thank you for taking the time to read this longer-than-normal post. I hope you found todays post interesting and insightful, and I encourage you to comment below any thoughts you may have. Thank you again for visiting my blog page and for reading what I have to say; it really means a lot!
Until next week,
Have a great day!
B.