Emotions are typically felt experiences, which often involve strong motivations and are caused by things that are of concern to individuals, such as dangers. Emotions play a central role in people’s lives. They determine, at least in part, whether a person’s life goes well or not, positive emotions being essential to well-being, whereas negative emotions, such as despair, being the mark of ill-being. Emotions also play an important role in determining what people think and do. Whether this influence makes for more or less rationality is still a matter of debate. The question of what emotions are also remains highly controversial. Although emotion theorists mostly agree on the typical characteristics of emotions, they disagree on the nature of emotions. Three main approaches are currently prominent: feeling theories, motivational theories, and evaluative theories. Another possibility is to opt for a hybrid theory, according to which emotions involve a mix of different components.

History

In the Western tradition, interest in emotion theory goes back to Ancient Greece. Both Plato (circa 429–347 BCE) and Aristotle (384–322 BCE) held distinctive views about emotions (or passions, in their terms), which are still influential. Plato and Aristotle opposed emotions to reason and held that emotions are an important source of irrationality. Conflicts between emotions and reason could give rise to akratic actions, that is, paradigmatic irrational actions, which the agent performs in spite of their better judgment. Plato and Aristotle stated that to avoid irrationality, individuals need to control or educate their emotions. For Aristotle, the education of emotion was essential to ethics because he claimed that feeling emotions that are fitting to their objects is required for being a virtuous agent (Brown, 2009). Aristotle held a hybrid view of emotions, according to which emotions comprise several components: motivations, pain or pleasure, evaluations, and bodily changes. For instance, he defined anger as a desire for revenge, accompanied with pain, on account of a real or imaginary slight to oneself or to someone close.

Most emotion theorists have followed Plato and Aristotle in considering emotions to be a potential source of irrationality and indeed of immorality [see Foundations of Rationality]. A notable exception being David Hume (1711–1776), who challenged the idea that emotions have to be ruled by reason and instead stated that “reason is, and ought only to be the slave of the passions” (Norton & Norton, 2000, p. 3). Hume’s thought is that because reason is unable to oppose passions due to its lack of motivational power, its only function is to establish the means necessary to satisfy passions.

Interestingly, this overall negative verdict on emotions is not based on an agreement about the nature of emotions. The most prominent critics of emotions are the Stoics, such as Seneca (c. 1 BCE–CE 65), and Immanuel Kant (1724–1804). According to the Stoics, emotions are judgments, whereas according to Kant, emotions are motivations. In spite of this, they shared a negative view of emotions. The Stoics thought that most emotions had to be educated, controlled by reason, or even eradicated. Seneca, for instance, claimed that instead of merely attempting to control anger, one should destroy it, given its fundamental wickedness (Seneca, 2010). The Stoics held the view that most emotions consist of erroneous judgements based on a mistaken and irrational conception of what is important. In the same vein, Kant held that practical reason, which he considered to be the source of morality, asked people to act on principles of rationality, as opposed to inclinations, which include emotions. He considers emotions and more generally motivations to be foreign to the self of agents so that they could only result in behavior-lacking autonomy.

For most of the past, interest in emotions was motivated mainly by the thought that one has to understand the nature of the threat they may constitute. In recent times, thinkers have taken on a more positive stance towards emotions. Nevertheless, even if much of the work of psychologists such as William James (1884), Walter Cannon (1927), Silvan Tomkins (1962), Paul Ekman (1972), and Carroll Izard (1971, 1977) focused on emotions, overall interest in emotions declined throughout most of the twentieth century. Since then, emotion theory has made a spectacular comeback. What motivated this renewed interest is the conviction that emotions are crucial to both theoretical and practical reason. Following the work of philosopher Ronald de Sousa (1987) and neuroscientist Antonio Damasio (1994), who independently underscored the importance of emotions in decision-making, most contemporary emotion theorists hold that emotions have an important function.

This renewed interest in emotion theory is noticeable across a wide range of disciplines; foremost in cognitive science, emotions are now recognized as one of the main driving forces of the mind. It has indeed been suggested that affectivism should be the new research paradigm (Dukes et al., 2021). Affectivism holds that emotions and, more generally, affect, play a central role in the explanation of human and nonhuman thought and action. Accordingly, affectivism should thus replace both behaviorism—the view that behavior can be explained without appealing to thoughts and feelings—and cognitivism—the view that emerged as a response to behaviorism and according to which mental processes involved in cognition are key [see Cognitive Ontology].

Core concepts

There is a huge variety of emotions, ranging from the negative emotions of fear, disgust, anger, sadness, and shame to the positive emotions of hope, trust, joy, pride, and admiration, to name only a few. Because of this, developing a general account of emotion is a challenge. In particular, one central difficulty when trying to develop a theory of emotions is that the ordinary term “emotion” seems to refer to a heterogeneous group of phenomena. On the face of it, there seems to be nothing of substance in common between emotions such as fear, religious awe, exuberant delight, pity, loving devotion, panic, regret, anxiety, nostalgia, rage, disdain, admiration, gratitude, pride, remorse, indignation, contempt, disgust, resignation, or compassion (see Rorty, 1980 and Griffiths, 1997). Some thinkers conclude that emotions do not form a natural kind, which allows for interesting generalizations. The challenge to define folk understanding of emotions is even more difficult when considering non-Western cultures. For example, consider song—“A feeling of admonition with moralistic overtone but no disposition to revenge”—used by the Ifaluk people, who live on a Micronesian atoll (Lutz, 1988), Japanese amae—“A pleasurable feeling of dependency” (Doi, 1973)—or Russian toska, a “melancholy-cum-yearning” (Wierzbicka, 1992).

The conclusion that emotions fail to form a natural or, more generally, a theoretical, kind is warranted only on the condition that a unifying theory of emotion proves impossible. The best way to see whether such a theory is feasible is to focus on what are considered by most researchers to be paradigm cases of emotion, such as fear, anger, or disgust, in the hope of developing a satisfactory theory of emotions.

A second difficulty is that even if one focuses on paradigm cases, there are different things this could mean. For example, if Alea fears dogs, this does not mean she currently experiences fear but that she is disposed to have fear experiences when there is a dog nearby. Given this, it is common to distinguish between emotional dispositions and occurrences of emotions and generally focus on the latter (Deonna & Teroni, 2012).

Even if the focus is on paradigmatic occurrent emotions, there is still an important difficulty to overcome. Typically, occurrent emotions involve a variety of components. Consider a typical case of disgust, such as disgust felt by Alea when opening her fridge and finding spoiled food. She will feel nauseous at the sight of the food, her disgust will come with physiological changes such as heartrate deceleration, she will express her disgust in facial expressions such as wrinkling her nose and down turning the corners of her mouth, her thought will be focused on ways of getting rid of the food, and she will be strongly motivated to avoid touching or ingesting it. Overall, it also appears that in some sense or another, her disgust involves a negative appraisal of its object. Thus, it is commonly agreed that a typical emotional episode comprises a phenomenological component, that is, a way it is like to experience the emotion, a physiological component, an expressive component, a motivational component, and a mental component as well as an evaluative component.

The question is whether any of these components are essential to emotions. There are at least as many answers to what is known as the problem of parts (Prinz, 2004) as there are emotion theories. Three main approaches have formed in contemporary emotion theory, each focusing on one of the typical components of emotions.

Feeling theories

The felt quality of emotions is often considered to be a necessary feature of emotions. Indeed, according to some, this feature is not only necessary but essential to emotions. On this view, emotions consist of specific kinds of feelings, which are distinguishable by their specific phenomenology. The most prominent version of this approach holds that emotions are bodily feelings (James, 1884). According to this view, emotions are feelings of bodily changes that are caused by the perception or thought of some exciting fact. As James puts it, “Our feeling of [bodily] changes as they occur IS the emotion” (James, 1884, pp. 189-190). The bodily changes in question consist of internal changes, such as the quickening of heart rate, and external changes, such as facial expression. They also include motivational changes, such as the impulse to run, and even the action of running. 

The main attraction of the feeling theory is that it focuses on what is often considered a central and possibly necessary feature of emotions, namely, its phenomenological component. Moreover, by tying emotional feelings to physiological changes, it offers a nonmysterious account of what emotional feelings consist of.

The main objection against the feeling theory is that it either loses or misplaces the so-called intentional object of emotions, that is, what the object of the emotion is about. When Alea fears an approaching bear, her fear is about the bear. Similarly, her disgust at rotten food is about the food. The bear is the intentional object of fear, whereas the food is the intentional object of disgust. So, either feelings of bodily changes are pure feelings, which lack an intentional object, or they have bodily changes as intentional objects. The problem is that none of these possibilities are satisfactory. When experiencing fear of being confronted with a bear, the fear is not a pure feeling that lacks an intentional object. Similarly, when experiencing fear when confronted with a bear, the fear is not about bodily changes. Given this, feeling theories fail to properly account for the intentional objects of emotions. 

This explains why the feeling theory has given rise to accounts that are neo-Jamesian in emphasizing the importance of bodily feelings but that take James’s account into new directions. Neo-Jamesians include motivational theorists (Deonna & Teroni, 2012) as well as evaluative theorists (Prinz, 2004). 

Motivational theories

Emotions, such as fear, anger, or disgust, are closely linked to actions, and so it seems plausible to hold that emotions consist of motivational states. Fear, for instance, is often linked to behaviors such as fight, flight, or freeze, which aim at mitigating danger. According to the most developed motivational account, emotions are “action tendencies” or “states of action readiness to execute a given kind of action” (Frijda, 1986, p. 70) The actions at stake are defined in terms of specific goals, such as protection or removing obstruction. Thus, each action tendency involves a variety of actions aimed at a specific goal.

On this view, each kind of emotion is identified with a particular action tendency. Fear, for instance, is defined as the action tendency of avoidance aimed at protection of the target of the emotion, such as the individual who feels the emotion. Disgust is defined as the action tendency of rejecting, also aimed at protection. What distinguishes the states of action readiness that constitute emotions from other kinds of action tendencies is their urge-like nature: such states tend to take precedence over other action tendencies, for instance, by interrupting current activities. A virtue of this approach is that it plausibly explains why emotional actions tend to be impulsive, in the sense that they are characterized by a preference to act earlier rather than later, as well as kind of shortsighted, the information on which they are based tending to be limited and biased. 

The thesis that emotions are, in essence, motivational states goes well with the view defended by biological determinists (e.g., Ekman, 1972; Tomkins, 1962; Tooby & Cosmides, 1990). According to the so-called basic emotion theory, many pancultural emotions are adaptations, which allowed our ancestors to meet recurrent evolutionary challenges such as dealing with rivals, escaping predators, or avoiding contaminated food (Ekman, 1972). Such basic emotions include anger, fear, disgust, happiness, sadness, and surprise and are thought to be shared with nonhuman animals, whereas nonbasic emotions are seen as depending on higher cognitive states that are unique to the human species (for criticism, see Russell, 1994; Barrett, 2006). For emotions to play an adaptative role, it appears crucial that they involve motivations. It is natural for this approach to hold that the motivational component constitutes the core of emotions. However, since biological determinism holds that emotional episodes involve a cascade of coordinated changes, which result from appraisals and include facial and vocal expressions, physiological changes, and mental changes, their account works well with hybrid theories of emotions. Indeed, it is compatible with all the main theories of emotions (Tappolet 2023a). 

The main challenge to motivational theories is that it is far from clear that all emotions are associated with action tendencies. Consider the emotion of joy. It seems that experiencing joy when winning a race, for instance, need not come with any goal-oriented action. In reply, motivational theorists have the option to adjust their conception (Frijda, 2008; Scarantino, 2014). They can suggest that an emotion like joy involves a generic goal, such as the goal to relate to the world in general. Negative emotions such as sadness or grief are equally problematic, for they tend to reduce the tendency to act. Here again, motivational theorists can reply that such emotions have the generic goal of not relating to the world (Frijda, 2008; Scarantino, 2014). It is unclear, however, whether a similar strategy can be made to work for emotions, like regret or relief, which are oriented towards the past—something that cannot be changed—or for emotions that are directed at fictions, such as compassion for Anna Karenina, the heroine of the eponymous novel by Leo Tolstoy.

A second challenge concerns differentiation between emotion kinds. People tend to avoid what they fear and what disgusts them, so merely appealing to that kind of action is not sufficient to classify it as an emotion of fear or disgust. Moreover, the goals of fear and disgust are not clearly distinct, since each aims at protection from harm. It is possible to argue that there is a difference between kinds of protection, with fear being associated with the goal of protection against danger and disgust protection against contaminants. However, this raises the question that for an individual to have a motivation with such a specific goal, it is necessary that the individual represents the relevant object as dangerous or contaminated. If so, emotions would require evaluations as well as motivations. 

A third challenge to motivational theories concerns the direction of fit of emotions (Searle, 1983). According to motivational theories, emotions appear to be like desires in having a world-to-mind direction of fit. It is the world that has to change in order to fit a motivational state. The aim of the motivational state is to be realized, something that requires a change in the world. This direction of fit contrasts with the mind-to-world direction of fit that characterizes beliefs. It is the belief that has to change in order to match how things are in the world. The problem for motivational theories is that emotions appear to have a mind-to-world direction of fit. Emotions are assessable in terms of how they fit the world, and it is the emotion that has to change in case there is a discrepancy between that emotion and the world. When a person is afraid of something harmless, for instance, the emotion should change or disappear because there is no danger involved. This again suggests that emotions involve evaluations.

There are two possible reactions a motivational theorist can make. The first is to accept that emotions have both a mind-to-world and a world-to-mind direction of fit (Scarantino, 2014). Emotions then represent evaluative features at the same time as involving goals. The main problem with this kind of hybrid account is basically the same as for the standard version: it is unclear whether it can account for emotions that fail to be tightly connected with motivation. The other reaction is to embrace the so-called attitudinal theory, which holds that the evaluative component can be fully explained in terms of the motivational profile of emotions.

The attitudinal theory

The main idea of the attitudinal theory is that emotions are states of felt action readiness directed at intentional objects (Deonna & Teroni, 2012). Such feelings of action readiness involve the experience of the intentional object as having certain evaluative features. In anger, for instance, individuals not only feel their body ready to attack but they also evaluate the object of their anger as offensive. The crucial point is that, according to the attitudinal theory, these evaluations do not consist in evaluative representations but in evaluative attitudes. Succinctly, the claim is that emotions are attitudes towards nonevaluative contents (e.g., persons, animals, etc.), which have evaluative features (e.g., offensiveness, danger, etc.) as their formal objects. Anger at one’s boss, then, is a state of felt action readiness that consists in an attitude of “taking-to-be offensive” directed at the boss. Because of this, the anger would be correct if and only if the boss was really offensive. 

One important issue is whether one can make sense of the idea of an emotion being correct on condition that it corresponds to some evaluative feature of the world without appealing to the notion of representation (Rossi & Tappolet, 2019). The attitudinal theory also shares difficulties with standard motivational accounts. First, the attitudinal theory has to meet the challenge of emotions that do not appear closely linked to motivation. Second, it is not clear that the states of felt action readiness associated with each kind of emotion are sufficiently distinct to allow the differentiation between kinds of emotions. The attitudinal theorists cannot invoke the different evaluations that are involved, for, according to their theory, these evaluations are explained in terms of motivations. According to the attitudinal theory, fear of a dog is “an experience of that dog as dangerous, precisely because it consists in feeling the body’s readiness to act so as to diminish the dog’s likely impact on it (fight, preemptive attack, etc.)” (Deonna & Teroni, 2012, p. 81). It is not clear that this kind of action readiness differentiates fear from disgust, or even pity. The reason is that it is possible to act so as to diminish a dog’s impact on a person’s body out of disgust (think of a slimy dog) or out of pity (think of a fragile dog). In order to distinguish between the action readiness that corresponds to each emotion kind, there seems to be no choice but to accept that they involve representations of their object as dangerous, disgusting, or pitiful (Ballard, 2021). 

Overall, it is not clear that motivational states are necessary for emotions, and when motivations are present, they seem to presuppose an independent evaluation. Whether these challenges can be overcome remains an open question.

Evaluative theories: Judgmental and quasi-judgmental 

Evaluative theories divide into several families: judgmental and quasi-judgmental theories, on the one hand, and perceptual and quasi-perceptual theories, on the other.

According to judgmental theories (Nussbaum, 2001; Solomon, 1993), emotions are evaluative judgments. Being angry at someone, for instance, consists in judging that this person has wronged or offended oneself, whereas being afraid of something amounts to judging that this thing is dangerous or threatening to oneself. 

A similar approach has been defended by psychologists under the name of appraisal theory, an approach according to which emotions involve appraisals of situations (Arnold, 1960; Lazarus, 1991). However, appraisal theorists need not consider emotions to consist in, rather than being caused by, appraisals. On one account, an emotion is a “felt tendency toward anything intuitively appraised as good (beneficial), or away from anything intuitively appraised as bad (harmful)” (Arnold, 1960, p. 171). On another account, emotions are complex processes, which involve the appraisal as cause and a combination of action tendency, physiological changes, and subjective affect as effect (Lazarus, 1991).

The attractiveness of judgmental theories is that it gives a straightforward account of the intentionality of emotions. Judgments clearly have intentional objects. In contrast to motivational theories, judgmental theories have no difficulty differentiating between kinds of emotions, since different kinds of judgments correspond to different emotions. Because evaluative judgments need not be emotions, however, judgmental theorists have to hold that the judgments that constitute emotions involve some other component, such as a feeling or motivation.

One important difficulty for judgmental theories is to account for cases of recalcitrant emotions, that is, cases in which the emotion experienced conflict with judgment (Rorty, 1978). It seems people can be afraid of something judged not to be dangerous or threatening, for instance. If emotions are judgments, recalcitrant emotions would involve a serious kind of irrationality. In contrast, it seems if there is any irrationality involved in recalcitrant emotions, it is not as serious as when making contradictory judgements. Put differently, the rationality requirement that is infringed by recalcitrant emotions appears less stringent than the one that forbids believing that p while believing that not p. 

Quasi-judgmental theories (Greenspan, 1988; Roberts, 2003) have a ready answer to this worry. According to such theories, when experiencing fear, for instance, people contemplate an evaluative proposition without committing to the truth of that proposition. In this respect, emotions are like imaginings. On one account, fear would amount to discomfort at the thought that danger looms, something that can be contemplated without it being true. On another account, feeling fear consists in a construal of something as dangerous (Roberts, 2003). Such states are fully compatible with contrary judgments, so quasi-judgmental theories appear to have no problem accounting for recalcitrance. One problem with quasi-judgmentalism is that it appears to make no room for even a weak kind of irrationality (Helm, 2001). Even if it does not amount to the stark kind of irrationality involved in contradiction, recalcitrance clearly involves some irrationality.

The main problem with both judgmental and quasi-judgmental theories is that they appear too intellectual (Deigh, 1994). In principle, to judge that something is dangerous involves possessing the concept of danger. However, it is far from clear if infants and nonhuman animals possess such a concept, but most researchers would agree that infants and nonhuman animals have emotions such as fear. Mice and sparrows, for instance, often experience fear. In response, it might be denied that the judgments or quasi-judgments at stake require the possession of concepts. However, the strategy to stretch the notion of judgment might well be questionable (Scarantino, 2010). 

Perceptual and quasi-perceptual theories

Perceptual theories (Döring, 2003; Prinz, 2004; Tappolet, 2012; Tappolet 2016) are motivated by the many analogies between emotions and sensory perceptions: both have phenomenological properties, both are elicited automatically and fail to be directly subject to the will, both have correctness conditions in terms of how they correspond to the world, and both allow for cases of recalcitrance. The claim is that emotions are kinds of perceptual experiences; they are perceptual experiences of evaluative features such as dangerousness in the case of fear. By contrast, quasi-perceptual theories (de Sousa, 1987; Goldie, 2000; Mitchell, 2021) hold that the analogies only warrant the conclusion that emotions share important similarities with sensory perceptions. 

Apart from the argument based on analogies, the attraction of perceptual and quasi-perceptual theories is that they can easily account for the fact that nonhuman animals can have emotions, for even if perceptual experiences have contents, these contents need not be conceptually articulated. Thus, concept possession is not required to experience emotions. Perceptual and quasi-perceptual theories can also account for the fact that emotions are typically felt, since that is a characteristic of sensory perception. To explain the motivational impact of emotions, such accounts can appeal to the fact that by making salient evaluative features, emotions have an important influence on decisions and motivations. 

The main objection to perceptual theories is that there are important differences between emotions and sensory perceptions (see Brady, 2013; Deonna & Teroni, 2012). One important difference, for instance, is that emotions require prior informational states. A person needs to perceive the bear or at least to believe that there is a bear nearby to feel fear. The question is whether a perceptual theory can allow for such a disparity and, if not, whether a substantial and plausible quasi-perceptual account can be developed. According to a new suggestion, such an account is indeed possible on condition that emotions are modeled not on sensory perceptions but on analogical representations of magnitudes, which are at one remove from the informational periphery (Tappolet, 2023b).

Questions, controversies, and new developments

The current debate between the main approaches in emotion theory is still very active. The focus of this discussion has recently moved to the question of whether there can be reasons for emotions and, if so, what they are. 

An important objection to perceptual theories is that sensory perceptions do not allow for reasons, whereas it appears that there are reasons for emotions (Brady, 2013). One can ask why it is that someone experiences fear on a particular occasion, and when this question is raised, it seems that what is asked for is a normative reason, which would justify the experience of fear. That there seems to be a thunderstorm in the offing, for instance, is a normative reason for Alea to feel fear at the idea that her sailboat might get struck by lightning. In such a case, feeling fear would be justified. The problem is that sensory perceptions do not seem to depend on normative reasons. Perceptual theorists have three main ways to respond to this challenge. They can deny that emotions allow for normative reasons (Gubka, 2022); they can claim that contrary to appearances, sensory perceptions allow for normative reasons (Milona, 2016); or they can hold that emotions can be perceptions even if they depend on normative reason.

However, allowing that emotions can be justified by normative reasons leaves many questions open. Depending on what normative reasons for emotions turn out to be, different accounts of emotions become plausible. Thus, one can take normative reasons for emotions to be epistemic, that is, reasons that speak to the truth of beliefs or, more generally, the accuracy of representational states. If normative reasons for emotion are of this kind, this would speak in favor of representational theories, for it follows that emotions would consist in states that can have truth or accuracy conditions. Alternatively, normative reasons for emotions could be practical normative reasons in the sense that they are reasons that speak to the question of why an action, an intention, or some mental state is good. This would strengthen the case of motivational theorists as well as for a different approach, according to which emotions have to be modeled on actions (Naar, 2022). One way to spell out this analogy with action is to hold that emotions are responses that are deemed fitting because of the evaluative stance the individual takes (Müller, 2019; Mulligan, 2009). Finally, reasons for emotions could sui generis, in the sense of being of their one kind (Silva, 2022). If that is the case, emotions could not be modeled on states such as beliefs, perceptions, or motivations.

In general, accounts of what reasons for emotions amount to is inferred from specific theories of emotions (but see Maguire, 2018). Given that the debates in emotion theory are far from settled, it would be fruitful to proceed instead from the nature of reasons (see Alvarez, 2010). Hence, the search for independent arguments concerning the nature of reasons for emotions should be high on the agenda of emotion theorists. A better understanding of the nature of such reasons would also allow cognitive scientists to reassess the fraught question of the rationality of emotions.

Broader connections

Emotions are thought to play a central role in many distinct phenomena. Theories of psychological happiness and well-being often emphasize the importance of emotions (Haybron, 2008; Rossi & Tappolet, 2023). The thought is that a happy life is a life in which the individual experiences more positive emotions (joy, hope, or pride, for instance) than negative emotions. Similarly, a life of well-being, that is, a life that is good for the person whose life it is, is often taken to be one marked by a positive balance of emotions. Emotions are also plausibly taken to be important in accounts of personal identity, the idea being that emotions make us the person we are (Kristjánsson, 2010) [see Social Identity]. Another domain in which emotions are important is value theory. According to a prominent account, values have to be analyzed in terms of emotions: something being good, for instance, such as to be the fitting object of positive emotions (D'Arms & Jacobson, 2000). Finally, responsibility theorists have stressed that people respond with resentment or gratitude to agents who harm or who benefit them (Strawson, 1962). Accordingly, some have argued that responsibility is tied to emotions. Holding someone responsible would consist in feeling emotions such as resentment or gratitude, whereas being responsible amounts to being the fitting target of such emotions (Shoemaker, 2015). More generally, emotions such as respect or compassion have often been to be central to the psychology of virtuous agents (Hursthouse, 1999; Sreenivasan, 2020). 

A further set of questions concerns the implications of emotion theory on neighboring empirical fields. A prominent question in animal cognition research is whether nonhuman animals really experience emotions, and if so, what explains that they cannot experience the whole range of human emotions [see Animal Cognition]. In developmental psychology, scientists want to know at what age children can experience full-blown emotions and recognize emotions in others. Finally, the field of artificial intelligence has given rise to a new set of questions, including whether artificial systems (i.e., beings that are neither flesh nor blood) have emotions and whether they can have emotions, even if they are not able to experience anything. This leads to broader questions, such as what is it, in fact, to feel one’s emotions? There are also questions about the reliability of artificial systems in detecting the emotions humans experience. More deeply, there are concerns about whether it is desirable for an artificial system, however smart, to be able to have emotions and what potential impact this would have on people should artificial systems be trained to detect emotions in human beings.

Given the prominence of emotions in the debates about all these distinct phenomena, the importance of becoming clear about the nature of emotions cannot be underestimated.

Acknowledgments

I am grateful to the Social Science and Humanities Research Council of Canada for supporting my research.

Further reading 

• Deonna, J. A., & Teroni, F. (2012). The emotions: A philosophical introduction. Routledge. 

• Scarantino, A., & de Sousa, R. (2021). Emotion. In E. N. Zalta (Ed.), The Stanford encyclopedia of philosophy (Summer 2021 ed.). Stanford University Press. https://plato.stanford.edu/archives/sum2021/entries/emotion

• Tappolet, C. (2023). Philosophy of emotion: A contemporary introduction. Routledge.