DANIEL M. WEARY

University of British Columbia, Animal Welfare Program, Faculty of Land and Food Systems, Vancouver, British Columbia, Canada

Suffering is a central issue of animal welfare, and the alleviation and prevention of suffering is a core moral responsibility to those for whom we care. One motivation in writing this chapter is to serve as a reminder of the fundamental importance of this concept and responsibility it carries. The rapid emergence of the field of animal welfare science has led to many important advances, including the development of improved assessment methods for negative affective states like pain and fear, and a more recent focus on positive affective states. Clearly a better understanding of how animals feel is important to assess and improve animal welfare, as is a sense of how positive experiences may somehow compensate for negative ones to contribute to the animal’s overall sense of well-being (Webb et al., 2019). But let not the merely important distract us from the fundamental; prevention and alleviation of suffering lies at the very heart of our duties toward animals under our care. The overall aim of this chapter is to contribute to the still embryonic body of scholarly work explicitly addressing the issue of animal suffering.

One reason for the reluctance of scholars to seriously address the issue of animal suffering may be that they are still influenced by the charming but outdated notion that the job of the scientist is to study the ‘facts’, and leave the difficult discussions around ‘values’ to others. The term suffering is infused with moral loading, explaining why it is often used in the rhetoric of animal advocates and in criminal law, and this moral loading has likely led academics to shy away from the concept. In the following section I briefly review how the term has been used in the academic literature, and compare this with what I argue is a more meaningful usage found in the human clinical literature.

12.2 Academic Usage of the Term ‘Suffering’

In the scientific literature on animal welfare the word ‘suffering’ is typically used in conjunction with the experience of some negative affective state. Most often this affective state is pain, and this is the state I focus on in most of the examples I use this chapter.

As illustrated in Table 12.1, the simplest usage of the term is adjunctive, as in ‘pain and suffering’, without attempt to distinguish the two ideas. This renders the term decorative.

A second usage of the term ‘suffer’ adds some value in that it suggests that the animal must consciously experience the negative state. It is this sense of the word that is used by authors interested in the conscious experience of pain in invertebrates (Elwood, 2011) and other animals for which there may be some questions about the animal’s level of sentience. Used in this way the word makes no claims about the quality or magnitude of pain, only that the pain is in some way felt by the animal. The search for methods that allow for strong inferences about the extent to which affective states like pain are consciously felt by animals is still ongoing (Weary et al., 2017). For academic usage I suggest that this distinction is too important to obfuscate by using the term ‘suffer’, unless specifically defined it this way. Moreover, to apply this technical meaning of ‘any felt negative affective state’ would seem to strain the boundaries of common usage of the word. One might say, for example, that ‘I felt the prick of the injection when I went for my flu shot’, but it would seem theatrical to call this ‘suffering’.

A stronger usage of suffering considers the magnitude and duration of unpleasant affective experiences. For example, Dawkins (1980, p.76) states ‘Not all fear, frustration or conflict indicates suffering. But prolonged or intense occurrences of these same states may indicate great suffering.’ Later in her book, Dawkins acknowledges the difficulty in establishing a clear line where suffering begins: ‘There is a subjective element, for example, in deciding how much fear, conflict etc. constitutes “suffering”.’ Dawkins’ academic usage comes closer to what might be considered common usage, such as that captured in the phrase ‘Last month I suffered from severe back pain.’ That said, even in this example there is some redundancy (how does the term ‘suffer’ add to the sentence given that we already know that the pain was severe?). We could instead say ‘Last month I suffered from back pain,’ thus using the term ‘suffer’ to mean ‘experience severe pain’, but the meaning would be clearer if the sentence simply specified the magnitude (and duration) directly. In addition, the use of ‘suffer’ to quantify the level of negative affect to be above some (unspecified) cut-off for severity and duration would seem to miss any additional, qualitative aspects that the moral imperative of the term connotes.

12.3 Human Descriptions of Suffering

A more nuanced understanding of suffering can be found in the academic literature describing human self-reports. As with the work on nonhuman animals described above, this literature on humans illustrates the importance of experiencing a negative subjective state of some considerable magnitude or duration. In addition, this literature points to other factors that are also important in conceptions of suffering.

In the classic paper in this field, Cassell (1982) recounts that one patient required ‘small doses of codeine’ for pain when she thought that this pain was due to sciatica, but required much higher doses when she found out that the cause was cancer. This and other examples suggest that when pain is associated with fear the likelihood of suffering increases: fear that the pain will increase to the extent that it can no longer be controlled with analgesics, fear that it will last forever, fear of becoming overwhelmed by the pain, or fear that the pain is a sign of a serious disease.

Another factor associated with suffering is that the patients are no longer able to do those things that are most important to them. Snyder (2004) states ‘Pain is a physical sensation of discomfort, whereas suffering taps the degree to which a person has let the pain prevent him or her from doing the important things in life.’ Anhedonia, a sign of depression, can be considered both a reduction in the motivation to access a reward and a reduction in the pleasure that is experienced from that reward (Treadway and Zald, 2011). Thus, reduced rates of activities can be both a cause of suffering (as in the first example of a person so incapacitated by pain that they can no longer perform an activity they previously enjoyed) and a sign of suffering (if the person is sufficiently depressed to show signs of anhedonia).

Loss of control is a recurrent theme in studies of human suffering. As described by Cassell (1999), ‘Suffering can start with anguish over the possibility that if the symptom continues, the patient will be overwhelmed or lose control — “I won’t be able to take it”.’ Loss of control is sometimes characterized as loss of the essence of who you consider yourself to be as a person, perhaps related to being able to do those things that you enjoy or believe to be important. The importance of loss of ‘personhood’ to the concept of suffering can be seen even in the definition of torture as an attempt to ‘obliterate the personality of the victim or to diminish his physical or mental capacities, even if they do not cause physical pain’ (Organization of American States, 1985).

The lack of access, or inability to cognitively process, environmental cues important to effective decision making may contribute to the feeling of being overwhelmed or losing control. One example of this importance of cognitive control comes from how patients understand the meaning of their pain. This is illustrated in relation to fear of catastrophic consequences in the cancer example discussed earlier. In addition, people sometimes report positive meanings to their pain, such as that associated with childbirth, and even pain that is self-inflicted as in attempts at ‘spiritual cleansing’ (Cassell, 1982). Thus, pain that is not perceived to have value, or is associated with other negative outcomes, is more likely to be perceived as suffering.

I conclude from these examples that control, in all its forms and effects, is key to understanding suffering. Control includes being able to take action, for example to avoid pain, as well as the ability to access and understand information to predict the occurrence, duration, and severity of the pain (for more on the benefits of control to well-being, see Chapter 6, this volume). These factors fall under the concept of agency. The literature on agency as a component of animal welfare has advanced greatly over the past few years. In the section below I will briefly review some of the key ideas from this literature, and discuss how these can be applied to better understand suffering in animals.

12.4 Agency

Research on quality of life in humans illustrates that assessments of our own well-being are only partially related to what we have. Our sense of agency (how we gain access to the things we value, including our ability to learn about different outcomes and to make informed choices) is also critical (Higgins, 2012). Recent thinking in the animal welfare literature has begun to argue that such processes are central to the welfare of nonhuman animals as well (Špinka and Wemelsfelder, 2011; Franks and Higgins, 2012; Špinka, 2019).

Špinka and Wemelsfelder (2011) defined agency as ‘the propensity of an animal to engage actively with the environment with the main purpose of gathering knowledge and enhancing its skills for future use’, and added that this will include ‘goal-oriented behavioural sequences such as foraging, mate seeking and predator avoidance’ as well as ‘agency-based patterns such as exploration and play’. As further explained by Špinka and Wemelsfelder, the ability to express these aspects of agency may be beneficial to welfare by improving feelings of competence and allowing animals to improve their success in dealing with environmental and social challenges. These authors also discussed how the lack of opportunity to express agency can be negative for animal welfare, for example by inducing feelings of boredom (increasingly recognized as a welfare harm; see Meagher, 2019) and even affecting measures of health. Most importantly from the perspective of the current chapter, they linked diminished agency with increased negative affect. For example, they argued that ‘declining agency may affect welfare … through the consequences of underdeveloped competence, such as heightened fear and anxiety and compromised social coping’, and that ‘animals from impoverished backgrounds may be overwhelmed by events when they arise, fail to cope and experience intense fear or anxiety’. The sense of being overwhelmed, failing to cope, and experiencing intense fear and anxiety correspond with how human patients characterize suffering as discussed in the previous section.

Špinka and Wemelsfelder (2011) also suggested that keeping animals in low agency conditions can result in them becoming ‘less well able to classify and evaluate perceived environmental stimuli, and will be less ready to deal with challenges once they arise’. I will return to ideas about the perceptual and cognitive capacities in the next section.

Špinka (2019) proposed four categories of agency:

The passive/reactive category includes the simplest and arguably most fundamental aspects of agency, and the aspirational category includes the autobiographical component that juxtaposes well with concerns about sense of self described in the human suffering literature. Also convenient is that the intermediate categories (action and competence building) link with the conceptions of Franks and Higgins (2012) (‘control effectiveness’ and ‘truth effectiveness’, respectively), as described below.

Franks and Higgins (2012) center their ideas around the concept of ‘effectiveness’, suggesting that ‘animals want to be successful in having desired results (value effectiveness), establishing what is real (truth effectiveness), and managing what happens (control effectiveness)’. They conclude that high welfare results ‘when these three domains work together to create organizational effectiveness’.

The first of these domains (value effectiveness) is the easiest to grasp as it fits neatly into traditional conceptions of well-being as in having what you want. This will include access to suitable housing, food and water, social companionship, etc., as well as having pleasurable experiences and avoiding painful or otherwise unpleasant experiences.

The second domain (truth effectiveness) is the motivation for information and understanding about the physical and social environment; this helps the animal avoid confusion and develop a sense of confidence in their knowledge of the world around them. Truth effectiveness is established, in humans at least (Higgins, 2012), by asking questions like ‘What is that?’, ‘Why did this happen?’, ‘Is this what I expected?’, and ‘Are my beliefs consistent with those of my group mates?’, questions that correspond well with the Bayesian framework discussed in the next section of this chapter. Franks and Higgins (2012) provide many examples showing how animals appear motivated to learn and acquire information about their world, even about features that may be fear-inducing or dangerous.

The third domain (control effectiveness) relates to the animal’s motivation to exert control over themselves (e.g., perform the behaviors they are motivated to perform) and over the physical and social environment in which they live (e.g., where they sleep, what they eat, and whom they interact with). As with the other domains, exerting control effectiveness may be a way of getting what you want (i.e., achieving value effectiveness, in this case a comfortable bed, a good meal, and some time with a friend), but it is also considered inherently important. Thus, animals may be motived to exert some choice, regardless of whether this provides them access to better options. Animals may also be motivated to work for access to options, even if they can have access to the same options for free (called ‘contrafreeloading’).

In Section 12.3 on human descriptions of suffering I reviewed how the inability to perceive or process information from the world around us can be an important contributor to suffering in our species, and how the experience of severe negative affective states like pain and fear can reduce our ability to sense and comprehend environmental cues. In the current section on agency I have reviewed the related conceptions of ‘competence-building agency’ (Špinka, 2019) and ‘truth effectiveness’ (Franks and Higgins, 2012), which argue that animals are inherently motivated to perceive and learn from environmental cues and to use this information to inform their actions (i.e., ‘control effectiveness’ and ‘action-driven agency’), and to ultimately obtain the things they desire (i.e., ‘value effectiveness’). Together, these ideas indicate that a reduced ability to perceive and make sense of new information, and use this to inform our actions, contributes to the experience of suffering.

12.5 The Bayesian mind

12.5.1 Bayes’ theorem and its application to pain

Over the past decade an important idea has begun to transform the way we think about cognitive processes in humans and other animals: the way we perceive and understand cues from our environment is affected by our expectations, and vice versa (Clark, 2015). In this section I will briefly introduce Bayes’ theorem and its influence on the way we think about animal cognition in general and pain perception in particular.

According to Bayes’ theorem, our belief about the probability of an event is based on prior expectations, updated by any new data that are available. To illustrate the power of this simple idea, consider the example of a coin toss. If you were to take a coin and toss it five times and each time the coin came up ‘heads’, how certain would you be that the coin had ‘heads’ on both sides? To be sure, five out of five heads is a rare result with a fair coin, and would only be expected to occur by chance about 3% of the time. But in addition to considering the results of this specific test, the Bayesian would ask about ‘priors’; that is to say, their belief that the coin was fair before they received the results of the coin toss. Say that for this type of coin it was known that only one in a million were two-headed. According to Bayes’ theorem, this prior should be considered in combination with the test result; the combined probability is still small, showing that despite the unusual sequence of heads it is still unlikely that the coin is two-headed.

Humans or other animals do not routinely calculate exact probabilities, but this is not required of Bayesian conceptions of cognition (Sanborn and Chater, 2016). The important point is that we do not simply respond to our bottom-up perceptions. Our top-down expectations affect what we perceive and how we make sense of these perceptions, and our experiences contribute to future expectations. This process is described by Ongaro and Kaptchuk (2019) as follows:

Stay updated, free articles. Join our Telegram channel

Join