“What is an emotion?” That question was asked in precisely that form by William James, as the title of an essay he wrote for Mind over 100 years ago (James, 1884). But philosophers have been concerned and often worried about the nature of emotion since Socrates and the “pre-Socratics” who preceded him, and although the discipline has grown up (largely because of Socrates and his student Plato) as the pursuit of reason, the emotions have always lurked in the background–as a threat to reason, as a danger to philosophy and philosophers, as just plain unreasonable. Perhaps that is why one of the most enduring metaphors of reason and emotion has been the metaphor of master and slave, with the wisdom of reason firmly in control and the dangerous impulses of emotion safely suppressed, channeled, or (ideally) in harmony with reason. But the question “What is an emotion? has proved to be as difficult to resolve as the emotions have been to master. Just when it seems an adequate definition is in place, some new theory rears its unwelcome head and challenges our understanding.
The master-slave metaphor displays two features that still determine much of the philosophical view of emotion today. First and foremost, there is the inferior role of emotion–the idea that emotion is as such more primitive, less intelligent, more bestial, less dependable, and more dangerous than reason, and thus needs to be controlled by reason (all argument that Aristotle and other enlightened Athenians used to justify the political institution of slavery as well). Second, and more profoundly, there is the reason-emotion distinction itself–as if we were dealing with two different natural kinds, two conflicting and antagonistic aspects of the soul. Even those philosophers who sought to integrate them and reduce one to the other (typically reducing emotion to an inferior genus of reason, a “confused perception” or “distorted judgment”) maintained the distinction and continued to insist on the superiority of reason. It was thus a mark of his considerable iconoclasm that the Scottish skeptic David Hume (1739/1888), in the 18th century, famously declared that “reason is, and ought to be, the slave of the passions.” But even Hume, despite an ingenious analysis of the structure of emotions, ultimately fell back on the old models and metaphors.
Whatever else it may be, philosophy is a historical discipline, and the theories and debates of today cannot be understood or appreciated without some understanding of philosophy’s rich and convoluted past. Even when a philosopher pretends to understand the phenomenon of emotion “in itself ” or analyze the language of emotion without reference to history or to any earlier attempts to do so, both the wisdom and the folly of generations of accumulated reflection and argument are already embedded in the subject matter. And although one might impatiently demand from the outset that one “define the terms” before the current discussion commences, the truth is that a definition will emerge only at the end of a long discussion, and even then it will be merely tentative and appropriate only within a limited context and a certain model of culture and personal character.
In what follows, I have tried to sketch a somewhat selective history of philosophical attempts to understand emotion, followed by a brief summary of questions still central to philosophical debate. Given the nature of philosophy and its emphasis on reason, however, we would expect that the focus of most philosophical analysis has been and remains the more cognitive aspects of emotion, with the physiological and to a certain extent the social and behavioral dimensions of emotion diminished or in many cases even denied. That will be, I should admit from the outset, the bias of this account as well. But the dialectic of philosophy tends to go back and forth in its emphasis and rediscovery of these often neglected dimensions. Sometimes emotions are dismissed as mere feelings and physiology, utterly unintelligent, even inhuman. In reaction, emotions are ascribed the virtues of true wisdom; they are seen as the proper masters of reason and the very foundation of our being-in-the-world. Most philosophers, however, try to find some more moderate, multidimensional position.
One might object that philosophical theories of emotion tend to be “armchair” speculation, devoid of the empirical support supplied by social scientists. However, this objection ignores the fact that philosophers, contrary to their own self-styled reputations as men and women of pure reason, have emotions themselves, and in most (but not all) cases a sufficiently rich repertoire of emotions to fund and support a dozen theories of emotion. As Descartes (1649/1989) said, in his introduction to the subject, “everyone has experience of the passions within himself, and there is no necessity to borrow one’s observations from elsewhere in order to discover their nature.” Ultimately, there is no need for the perennial (in fact century-old) feud between philosophy and psychology, and the phenomenon of emotion lies equally open to both of them.
THE HISTORY OF THE PHILOSOPHY OF EMOTION
Although the history of philosophy has often been described as the history of the development of reason-for example, by the great 19th-century German philosopher G. W. F. Hegel-philosophers have never entirely neglected emotion, even if they have almost always denied it center stage. It would be a mistake, however, to put too much emphasis on the term “emotion,” for its range and meaning have altered significantly over the years, in part because of changes in theories about emotion. So, too, the word “passion” has a long and varied history, and we should beware of the misleading assumption that there is a single, orderly, natural class of phenomena that is simply designated by different labels in different languages at different times. The language of “passion” and “emotion” has a history into which various feelings, desires, sentiments, moods, attitudes, and more explosive responses enter and from which they exit, depending not on arbitrary philosophical stipulation but on an extensive network of social, moral, cultural, and psychological factors. Thus we will often find that the focus is not emotion as such, but rather some particular class of emotions or particular emotion and its role in the manners or morals of the time.
The emotions as such, accordingly, do not form one of the three aspects of Plato’s (c. 428-347 B.C.) tripartite soul as defined in The Republic (1974). These aspects are reason, spirit, and appetite; not only does what we call “emotion” seem divided between spirit and appetite, but, considering Plato’s discussion of eros as the love of the Good in his dialogue The Symposium (1989), there are emotions involved in reason as well. Aristotle (384-322 B.C.), by contrast, did seem to have a view of emotion as such; but although he had a mania for taxonomies, he spent relatively little time listing or analyzing the emotions, as he did, for example, the virtues and the various kinds of birds. In his Rhetoric (1941), however, he defined emotion “as that which leads one’s condition to become so transformed that his judgment is affected, and which is accompanied by pleasure and pain. Examples of emotion include anger, fear, pity, and the like, as well as the opposites of these.”‘ (He did not specify what these “opposites” might be.)
Aristotle discussed certain emotions at length, notably anger, which he described in remarkably modern terms. In the Rhetoric he defined anger as “a distressed desire for conspicuous vengeance in return for a conspicuous and unjustifiable contempt of one’s person or friends.” He added that “anger is always directed towards someone in particular, e.g. Cleon, and not towards all of humanity,” and mentioned (if only in passing) the physical distress that virtually always accompanies such emotion. The key to his analysis, however, is the notion of a “slight” as the cause of anger, and may be an instance of “scorn, spite, or insolence.” Aristotle made allowances for only imagined slights (in other words, unwarranted anger is nevertheless anger), and he gave a central place to the desire for revenge, thus introducing a behavioral component at the heart of the emotion.
We might note that Aristotle, who was so precocious in so many disciplines, seems to have anticipated most of the main contemporary theories. His analysis of anger includes a distinctive cognitive component, a specified social context, a behavioral tendency, and a recognition of physical arousal. He even noted that physical or psychological discomfort-sickness, poverty, love, war, breached expectations, or ingratitude yields a predisposition for anger. It is worth noting that Aristotle had little to say of “feeling,” presumably not because the Greeks were anesthetic, but rather because what we (inconsistently) call “affect” and inner sensation generally held little interest for them and played no significant role in their language or their psychology.
Perhaps the most important single point to make about Aristotle’s view of emotion is the fact that his analyses make sense only in the context of a broader ethical concern. Anger was of interest to him because it is a natural reaction to offense and a moral force, which can be cultivated and provoked by reason and rhetoric. (Thus its inclusion in a book on that topic.)
Anger (and several other emotions, notably pride) are also prominent in Aristotle’s classical list of virtues in his Nicomachean Ethics (1941), where he discussed in some detail those circumstances in which it is appropriate to get angry, those in which it is not, and what amount or intensity of anger is justified. He suggested that forgiveness may be a virtue, but only sometimes. He also insisted that only fools don’t get angry, and that although overly angry people may be “unbearable,” the absence of anger (aimed at the right offenses) is a vice rather than a virtue. In this, as in all else, Aristotle defended moderation, the “mean between the extremes.” So too, he discussed fear at length in the Ethics with regard to courage, which is not fearlessness or “overcoming” fear so much as it is having just the right amount of fear-not to be foolhardy or a coward either. The emotions, in other words, are central and essential to the good life, and the analysis of their nature is part and parcel of an ethical analysis.
So, too, in Roman times, we find the conjunction of ethics and emotion in the philosophy of the Stoics (see Rist, 1969). But whereas Aristotle took emotion to be essential to the good life, the Stoics analyzed emotions as conceptual errors, conducive to misery. In modem terms, the Stoics Seneca and Chryssipus developed a full-blooded cognitive theory of the emotions, two millenia ago. Emotions, in a word, are judgments-judgments about the world and one’s place in it. But the world of Roman society was not a happy or a particularly rational place. (Seneca served under the Emperor Nero, and ultimately committed suicide at his behest.) And as the Stoics saw the world they lived in as out of control and beyond any reasonable expectations, they saw the emotions, which impose such expectations on the world, as misguided judgments about life and our place in the world.
The emotions, consequently, make us miserable and frustrated. Accordingly, the Stoics made a careful study of the component judgments that compose the emotions-the presumptuousness of moral judgment in anger, the vulnerability of love, the self-absorption of security in fear. The alternative was seen as “psychic indifference,” or apatheia (apathy). The Stoics did believe, we might add, in a “higher” reason, one transcending the vanities of the social world. But they felt that the best life in that world could be achieved only by getting straight about the pointlessness of emotional attachments and involvement.
Throughout the Middle Ages, the study of emotion was again typically attached to ethics, and it was central to Christian psychology and the theories of human nature in terms of which the medievals understood themselves (see Hyman & Walsh, 1973). There were elaborate, quasi-medical studies of the effects of the various “humours” (gall, spleen, choler, and blood itself) on emotional temperament, but there were (as there were among the Stoics) especially rich studies of the cognitive and “conative” aspects of the emotions. Emotions were essentially linked with desires, particularly self-interested, self-absorbed desires. And so the Christian preoccupation with sin led to elaborate analyses of those emotions, passions, and desires designated as sins (notably greed, lust, anger, envy, and pride).
The tight linkage between the study of emotion and ethics is particularly evident in the curious fact that the highest virtues, such as love, hope, and faith, were not classified as emotions as such, but were rather elevated to a higher status and often (e.g., by Thomas Aquinas) equated with reason. The old master-slave metaphor remained alive and well, and as some emotions were seen as sins, the highest virtues could hardly be counted among the mere emotions.
Reviewing the ancient and medieval literature on emotion, René Descartes (1596-1650) was provoked to write that what they taught was “so slight, and for the most part so far from credible, that I am unable to entertain any hope of approximating the truth excepting by shunning the paths they followed” (1649/1989). Descartes is typically recognized as the “father” of modern philosophy, and, in a more scholarly vein, as the bridge between the scholastic world of the Middle Ages and our own. But Descartes was fundamentally a scientist and a mathematician, awed by “the natural light of reason”; accordingly, he disdained the bodily and the bestial, insisting that the mind is a separate “substance” from the body (and that beasts do not have minds).
The separation of mind and body proved to be a famously difficult problem for Descartes and his successors, however, and nowhere was that problem more evident than in his attempt to deal with the emotions. Thoughts about mathematics may be clearly “in” the mind, as stomach contractions are in the body, but an emotion seems to require the interaction of mind and body in an undeniable way. Accordingly, Descartes defended a theory in his treatise On the Passions of the Soul (1649/1989), in which the mind and body “meet” in a small gland at the base of the brain (now known as the pineal gland), and the latter effects the former by means of the agitation of “anirnal spirits (minute particles of blood), which bring about the emotions and their physical effects in various parts of the body.
But the emotions involve not only sensations caused by this physical agitation, but perceptions, desires, and beliefs as well. Thus over and above the physical agitation and familiar sensations, the emotion hatred, Descartes declared, ultimately arises from the perception of an object’s potential harmfulness and involves a desire to avoid it. Accordingly, it is not as if an emotion is merely a perception of the body; it is rather, as Descartes put it, a perception of the soul, and some perceptions (as in dreams) may in fact be of things that do not exist at all.
An emotion is one type of -passion,” and Descartes defined the passions in general as “the perceptions, feelings or emotions of the soul which we relate specifically to it, and which are caused, maintained, and fortified by some movement of the [animal] spirits.” The passions in general are distinguished from “clear cognition,” and render judgment “confused and obscure.” Emotions are particularly disturbing passions. And yet emotions can be influenced by reason. For example, writing of courage, Descartes stated:
To excite courage in oneself and remove fear, it is not sufficient to have the will to do so, but we must also apply ourselves to consider the reasons, the objects or examples which persuade us that the peril is not great; that there is always more security in defense than in flight, that we should have the glory and joy of having vanquished, while we should expect nothing but regret and shame for having fled, and so on.
And so the physiological account gives way to a cognitive account, and the emotions move from the merely bodily to an essential ingredient in wisdom: “The utility of the passions consist alone in their fortifying and perpetuating in the soul thoughts which it is good that it should preserve, and which without that might easily be effaced from it.” How then can there be “bad” emotions? “The harm is that they fortify these thoughts more than necessary, or they conserve others on which it is not good to dwell.” So, bewildered by the physiology (though he was at the head of his class in the latest scientific knowledge), Descartes too tended to a value-oriented, wisdom-minded analysis of emotion. His six “primitive passions-wonder, love, hatred, desire, joy, and sadness-are not meaningless agitations of the animal spirits, but ingredients in the good life.
Baruch (Benedict) Spinoza (1632-1677) might well be considered to be a latter-day Stoic, like Chrysippus and Seneca in ancient Rome. Just as the Stoics saw the emotions as misguided judgments about life and our place in the world, Spinoza too saw the emotions as a form of “thought” that, for the most part, misunderstand the world, and consequently make us miserable and frustrated. But unlike the Stoics, Spinoza did not aspire to that “psychic indifference” known as apatheia (apathy); rather, in his Ethics (1677/1982), he urged the attainment of a certain sort of “bliss,” which can be achieved only once we get straight our thinking about the world. In particular, we have to give up the idea that we are or can be in control of our own lives, and adopt instead the all-embracing idea of ourselves and our minds as part of God. Most of the emotions, which are passive reactions to our unwarranted expectations of the world, will leave us hurt, frustrated, and enervated.
The active emotions, by contrast, emanate from our own true natures and heighten our sense of activity and awareness. Spinoza, like the Stoics, developed an early version of the cognitive theory of emotion. But Spinoza also defended a grand and complex metaphysics, in which all substance is one and mind and body are but dual “aspects” of one and the same being. Accordingly, he did not face Descartes’s formidable “mind-body’ problem”; although he himself could not have attempted to work this out, he anticipated the subtle emotion-brain research that is being carried out today by some philosophers as well as by neuropsychologists.
David Hume (1711-1776) was one of the most outspoken defenders of the Enlightenment, that very vocal and often rebellious intellectual movement that challenged old orthodoxies, elevated science and put religion on the defensive, attacked superstition and irrationality in all quarters, practiced and encouraged vigorous debate and discussion, and put a premium on the virtues of reason. But Hume, in carrying out the directives of reason to challenge, debate, and question, came to question the role and capacities of reason itself, and in particular the power of reason to motivate even the most basic minimum of moral behavior. “It is not against reason,” he declared in one of his most outrageous proclamations, “to prefer the destruction of half the world to the scratching of my finger” (1739/1888).
What motivates us to right (and wrong) behavior, Hume insisted, are our passions, and rather than being relegated to the margins of ethics and philosophy, the passions deserve central respect and consideration. Accordingly, he gave the passions large middle portion of his great first book, A Treatise of Human Nature (1739/ 1888). Not surprisingly, however, most philosophers then and since have preferred to read the first and third parts, on knowledge and ethics, and to ignore the central position of the passions.
Hume’s theory is especially important not only because he challenged the inferior place of passion in philosophy and questioned the role of reason. He also advanced a theory of the passions that, although limited and encumbered by his general theory of mind, displayed dazzling insight and a precocious attempt to grapple with problems that would only be formulated generations later. Hume, like many of his contemporaries and predecessors, defined an emotion as a certain kind of sensation, or what he called an “impression,” which (as in Descartes) is physically stimulated by the movement of the “animal spirits” in the blood. Such impressions are either pleasant or unpleasant, but the differentiation of the many emotions is not to be found in the nature of these impressions as such. Rather, the impressions that constitute our emotions are always to be located within a causal network of other impressions and, importantly, ideas.
Ideas cause our emotional impressions, and ideas are caused in turn by them. The pleasant impression of pride, for example, is caused by the idea that one has achieved or accomplished something significant, and the impression in turn causes another idea, which Hume described as an idea of the self, simpliciter. The emotion, in other words, cannot he identified with the impression or sensation alone, but can only be identified by the whole complex of impressions and ideas. What Hume again acknowledged with his emphasis on the essential place of ideas in emotion is what we now call the cognitive dimension of emotion, in addition to the physiological (“animal spirits”) and merely sensational (“impression”) aspects of emotion. Moreover, his inclusion of the second idea of the self in this example indicates his grappling with the notion of intentionality (the “aboutness” of emotions)-an effort that is further reinforced by his somewhat obscure insistence that the connection between an emotion (the impression) and this consequent idea is “original” or “natural,” or something more than the merely causal associations that form the usual bonds between ideas and impressions.
The emotions, for Hume, form an essential part of ethics. There are good emotions and bad emotions. Pride, he declared, is a good emotion; humility, its opposite (an unpleasant feeling brought about by the idea that we have accomplished something), is a bad emotion, a “monkish” emotion. Here we can see again the extent to which, as so often, a theory of emotion serves to grind some larger philosophical ax-in this case, Hume’s Enlightenment attack on religion. In this regard too, we might mention another aspect of Hume’s moral philosophy, followed in kind by his illustrious Edinburgh friend and colleague Adam Smith (1723-1790, also the author of The Wealth of Nations [1776/1976], the bible of modem capitalism).
Hume and Smith both defended the importance of what they called “the moral sentiments” (see Smith, 1759/1976), the foremost of which is sympathy, our ability to “feel with other people and appreciate (if not suffer with) their misfortunes. Sympathy, they argued, is a universal feature of human nature (countering and mitigating the self-interest that Smith in particular famously championed in The Wealth of Nations), and it is the bedrock foundation of society and morality. Emotion, in other words, is not an embarrassment or part of the refuse of the human psyche, but rather the very essence of human social existence and morality. It is not to be unfavorably contrasted and opposed to reason, but, on the contrary, is to be celebrated and defended along with it.
Immanuel Kant (1724-1804) was also a champion of the Enlightenment, but he too questioned the capacities and limits of reason. He was uncompromising in its defense, however, against Hume’s skepticism, against any attempt to replace reason by irrational faith, and against any attempt to ground ethics on fleeting human feeling instead of the universal and necessary dictates of reason. Thus Kant reinforced the crucial distinction between reason and what he called “the inclinations” (emotions, moods, and desires), and dismissed the latter (including the moral sentiments) as inessential to morals at best and intrusive and disruptive or worse. And yet, although Kant felt no need to develop a theory of emotion to accompany his elaborate and brilliant “critiques” of reason, his position on the “inclinations” is more ambiguous than is usually supposed, and his “respect for feeling” more significant.
It was Kant, a quarter-century before Hegel (who is credited with it), who insisted that “nothing great is ever done without passion,” and it was Kant, in his Critique of Judgment (1793/1951; concerned in part with art and aesthetics), who celebrated the importance of shared (“intersubjective) feelng in the appreciation of beauty and the awe with which we try to comprehend the wonder of God’s creation. Indeed, even Kant’s central notions of respect and human dignity, the very heart of his rationalist ethics, are sometimes suggested to be matters of feeling as well as reason, thus calling into question the harshness of his ruthlessly divided self. When his successor Hegel (1770-1831) took over the reins of German philosophy in the early 19th century, the overstated distinction between reason and passion was again called into question, and Hegel’s own odyssey of reason (in an epochal book called The Phenomenology of Mind, 1807/1977) has rightly been called a “logic of passion” as well.
Friedrich Nietzsche (1844-1900) was a philosopher for whom passion was the watchword and reason a source of suspicion. He was the culmination of a long line of “Romantics,” beginning with the Sturm und Drang poets of the 18th century and continuing through the philosophy of Nietzsche’s own favorite influence, the neo-Kantian pessimist Arthur Schopenhauer. Nietzsche anticipated the global skepticism and conceptual chaos of the 20th century; like Freud, who admired him, he described (and celebrated) the darker, more instinctual, and less rational motives of the human mind. Accordingly, in his On the Genealogy of Morals (1887/1967), he praised the passions and, in an ironic twist, described the passions as themselves having more reason than Reason.
But this was not to say that all passions are wise; some, he declared, “drag us down with their stupidity,” and others, notably the “slave morality’ emotion of resentment, are devious and clever but to a disastrous end, the “leveling” of the virtuous passions and the defense of mediocrity. Nietzsche never developed a “theory” of emotions, but his distinctions were remarkable in their insight and subtlety. His celebration of passion scared the wits out of a great many philosophers in Europe, however, who saw more than enough passion and irrationality in the Great War an then the rise of National Socialism in Germany. Accordingly, the ancient celebration of reason would once more rule philosophy, and emotion was again relegated to the sidelines.
In the 20th century, one can trace the fate of emotion in Western philosophy through two very different tracks. In North America and in England, the emotions were given particularly short schrift, in large part because of the newly exaggerated emphasis on logic and science. The great British philosopher Bertrand Russell gave elaborate praise to love and passion in the opening pages of his autobiography (1967), but in his philosophy he said virtually nothing about them. Of course, the nature of emotion was a major concern of William James and the young John Dewey in the early years of the century, but with James’s emphasis on the physiological nature of emotion (he argued  that an emotion is a sensation or set of sensations caused by a physiological disturbance, which in turn is prompted by some “perception” or other), coupled with the subsequent and quite unfortunate split between philosophy and psychology as academic discipline, questions about emotion were relegated to the realm of psychology (where they were also treated with less than the full respect due them).
Indeed, the first major attention to emotion in Anglo-American philosophy came in mid-century, when an ethical theory named “emotivism” came to dominate both the English and the North American scene. But emotivism, which was part and parcel of an across-the-board philosophical purgative known as “logical positivism,” was essentially a dismissal of ethical (and many other) questions in philosophy as “meaningless” (i.e., unscientific and without verifiable solutions). Emotion came back onto the stage of philosophy, but only as the butt of the argument: Ethical statements were meaningless because they were nothing but expressions of emotion.
During the same period in Europe, however, the emotions enjoyed more attention. Franz Brentano (1874/1971) succeeded the British “moral sentiment” theorists in attempting to found an ethics on a foundation of emotions. (Sigmund Freud was one of his Students.) Following the “phenomenology” of Edmund Husserl (1938/1960) (another Brentano student and a mathematician who showed little or no interest in emotion), Max Scheler (1916/1970), Martin Heidegger (1927/1962), and more recently Paul Ricouer (1950/1966) developed ambitious philosophies in which emotions were given central place in human existence and accorded considerable respect.
In the shadow of World War 11, Jean-Paul Sartre offered the slim but important The Emotions: Sketch of a Theory (1939/1948), followed by his monstrous tome Being andNothingness (1943/1956), which includes embedded within its many pages a number of detailed “phenomenological” analyses of emotion. Sartre’s conception of emotions as “magical transformations of the world”–willful stratagems for coping with a difficult world-added a new “existential” dimension to the investigation of emotion. But, predictably, philosophy in both France and Germany turned again to other interests, though the study of emotion continued despite the perennial shift in fashions.
In Anglo-American philosophy, however, the fortunes of emotion were also to change. In an article simply entitled “Emotion” (indicating how rarely the topic had even been broached), Errol Bedford (1956/1964) addressed the Aristotelean Society in London on the nature of emotion and the errors of thinking of emotions as “feelings.” The essay might have sat on the shelves gathering dust except for the fact that the then dean of Oxford philosophers, J. L. Austin (1956-1957), took it upon himself to remark on one of Bedford’s claims. (Austin’s own essay was not about emotions at all.) Austin’s attention kept the article alive and occasionally anthologized until the 1960s, when the subject seemed to come to life again.
Today, one finds a rich variety of arguments about emotions on both sides of the Atlantic Ocean and the English Channel. Given the nature of philosophy and its current concern with epistemological matters, it is again not surprising that the focus is on the conceptual structures of emotion, rather than the sensory, social, or physiological aspects of emotion. But there has been a reaction even within philosophy to the “hypercognizing” of emotion; consequently, there has been a serious effort to join forces with psychologists, neurologists, anthropologists, and moral philosophers to obtain a more holistic theory of emotion.
SOME PHILOSOPHICAL QUESTIONS ABOUT EMOTION
What is an emotion? Because philosophy is a discipline concerned with the essential nature and the “definition” of things, the basic question facing theories of emotion in philosophy is still the question posed by James and answered in a fashion by Aristotle. It is, on the face of it, a quest for a definition, a conceptual analysis. But it is also a much larger quest for an orientation: How should we think about emotion-as intrusive, as essential to our rationality, as constitutive of meaning, as dangerous, as dispensable, as an excuse for irresponsibility, or as a mode of responsibility? Which of the evident aspects of emotion-that is, the various sensory, physiological, behavioral, cognitive, and social phenomena that typically correspond with an emotion-should we take to be essential?
Many philosophers hold onto the old “Cartesian” view that an emotion cannot lack its “subjective” or “introspective” aspect, although what this means (and how accessible or articulate an emotion must be on inspection) is itself a subject of considerable dispute (Lyons, 1980; Sartre, 1943/1956; Freud, 1915/1935). Many philosophers have become skeptical about such subjective essentialism, however; like their associates in the social sciences, they have pushed the analysis of emotion toward more public, observable criteria (formulating their own versions of behaviorism, physiologism, and social construction theory, for example). But the seemingly self-evident Cartesian demand that first-person experience is seemingly ineliminable is evident in even the most radical of them, although its place and significance are greatly diminished. Can one have an emotion without feeling?
What is a “feeling”? The virtue of the Jamesian theory is that it ties down the nature of emotional sensation to quite particular and therefore verifiable visceral responses. Unfortunately, the Jamesian theory is wrong, at least in its details. How specifically are emotional feelings tied to physiological processes? To be sure, whatever goes on in the mind must now be supposed to have some correlate and cause in the brain, but can we not and should we not describe the “phenomenology” of those feelings quite apart from their brain correlations and causes?
Some theorists have tried to save feeling theory by employing the vague, general (and technical) notion of “affect” and its cognates (“affective tone”). But do such terms do anything more than cover up the problem with another word, whose meaning can only be explained by “the kind of feeling you get when you have emotion XT It is a mistake, moreover, to suppose that such feelings are indescribable or “ineffable,” whether out of excessive romanticism (as if understanding always undermines passion) or dismissive scientism (why talk about feelings if we can’t experimentally test them?). Most feelings have at least an “as if” recognizability (“It feels as if I’d known him for years” or “it felt as if he had shot me through the heart, it was so sudden and so traumatizing”). Many feelings have a distinctive structure, which (not surprisingly) emerges in the thoughts (and then in the verbal expression) of the emotion.
In general, one must ask how much cognition and learning is presupposed in the feelings that we identify as emotions. One does not need an elaborate Schachter and Singer (1962) scenario to do the Gedanken experiment, which shows that certain feelings typical of, say, fear and anger do not actually constitute fear and anger if there are no appropriate beliefs accompanying them. A person may well feel flushed, uncomfortable, and “as if” he or she wanted to flee or start a fight with someone, but if there is no fearful object (more precisely, if the person has no sense of a fearful object) or if there is nothing objectionable, frustrating, or offensive (to the person), then those feelings do not count as fear and anger (or even as “feeling afraid” or “feeling angry”). Whatever else we may say about the place of feeling in emotion, feeling alone is not sufficient.
Of course, this same term “feeling” can be expanded to include all sorts of thoughts, cognitions, and attitudes about the immediate situation, and not only tendencies to behave but even the behavior itself (as subjective experience rather than observable action). But this only shows that the seemingly innocuous notion of “feeling” also needs careful analysis, and the commonsense notion that an emotion is basically a feeling (perhaps a feeling in a certain context or brought about by a certain kind of cause) is accordingly still a prominent focus in philosophy. (For example, see Kraut, 1987; Stocker, 1990.)
Recent advances in neurology disclose structural and functional patterns in the central nervous system that are correlated with, and that under experimental conditions bring about, certain emotional reactions. Do these patterns dictate the structure of an adequate theory of emotion, or are those findings simply one more set of (contingent) considerations for inclusion in an all-embracing theory? Whatever the case, it is now clear that philosophers cannot ignore or neglect the rich neurophysiological literature on emotions. Indeed, there is now a new (inter)discipline in philosophy called “neurophilosophy,” which makes the new neurology central to any adequate analysis of emotion and “the mind” (Churchland, 1986). Philosophers may continue to argue that Aristotle knew all about emotions even though he did not know beans about the brain, but they do so at their peril – and in the face of the obvious fact that among the factors that have altered the history of philosophy and its concepts most radically have been new advances in previously unknown or undeveloped sciences.
Virtually all emotions get expressed (however minimally) in behavior. Should behavioral tendencies or sequences of actions or certain basic gestures be taken as essential? A great deal of detailed work in psychology has shown the enormous subtlety and the seemingly “hard-wired” nature of basic patterns of facial expression. Philosophers remain skeptical (Neu, 1989). The data are not in question, but the implied shift in conception from the emotion to the symptom of emotion is. What is it that causes the twitch or a gesture? The emotion would seem to be the perception, the awareness, the realization that is expressed, not the expression itself.
On the other hand, many philosophers of a somewhat behaviorist bent (following Wittgenstein’s later Philosophical Investigations  and Gilbert Ryle’s The Concept of Mind ) have suggested that an emotion is nothing but its behavioral expression, though certainly not a single gesture but an openended sequence of actions. An emotion is not a “ghostly inner event,” Ryle tells us, but a “multitrack disposition” to behave in any number of recognizable ways. So, too, philosophers have tried to understand emotion not as an inner feeling but as a value-laden description of a social situation. Thus Errol Bedford (1956/1964), in his pioneering article in the 1950s, suggested that the difference between shame and embarrassment, for example, is not some shade of difference between internal qualia, but the differences between descriptions of an awkward situation in terms of responsibility or innocence.
What remains at the core of all such theories, however, is an awareness that all emotions presuppose or have as their preconditions certain sorts of cognitions–an awareness of danger in fear, recognition of an offense in anger, appreciation of someone or something lovable in love. Even the most hard-headed neurological or behavioral theory must take account of the fact that no matter what the neurology or the behavior, if a person is demonstrably ignorant of a certain state of affairs or facts, he or she cannot have certain emotions. If neurologically induced rage does not include some object of anger, that rage (whatever else it may be) cannot be anger. So, too, Freud’s “free-floating anxiety” would count as an emotion only insofar as it does indeed (as Freud [1915/ 19351 argued) have an object, albeit “unconscious.”
Philosophers (following Aristotle and the scholastics of the Middle Ages) have come to call this the “formal object” of emotion, and one might well think of this as the minimum essential set of “beliefs” defining an emotion and an emotional experience. The formal object of fear, to take an obvious case, is a fearful object, and the beliefs that constitute the awareness of the presence or threat of such an object. Other emotions are more complicated and, accordingly, more the topics of debate and disagreement. Anger would seem to require a formal object involving an offense, but some authors would allow frustration alone to count as anger (Gordon, 1987). Jealousy is more difficult still, for its object seems to involve not only a threatened loss but a perpetrator as well (perhaps the threatened object as a perpetrator too), and possibly the larger social situation in which jealousy involves not only loss but humiliation as well (Neu, 1980).
But though the exact natures of the formal objects and requisite beliefs of various emotions is a matter of lively debate (and there is some doubt and debate over the very possibility of a generalized formal object for all emotions or emotions sui generis), the presumption is that every emotion must have a cognitive basis and an object. (There is some corollary debate concerning the status of moods and mood-like emotions [e.g., joy], which do not have a determinate object).
There is also considerable debate over the nature of cognition itself. I have used the word “belief ” above, but that seems to me to be unsatisfactory for a variety of reasons Beliefs are too much like established states rather than the spontaneous acts or events that characterize many emotions. Beliefs seem to be too fully articulate and already verbalized for the unreflective reactions that characterize most emotions. For that reason, I prefer to use the concept of “judgment” in this regard (like the ancient Stoics), whereas others (Neu, 1977) prefer the term “thought” (like Spinoza, for example). Some have simply stuck with the notion of evaluation (Pitcher, 1965), while others have preferred the less cognitively commital notion of a way of seeing (“seeing as”) sometimes as a rejection of the cognitive view, but more appropriately, I believe, as a refinement of it (Calhoun, 1984). The nature of an emotional cognition, and whether it must be fully conscious or capable of articulation, remain matters of considerable debate. Indeed, if certain holistic suggestions can be worked out, it may be that the very distinctions that philosophers have so long presupposed among cognition, behavior, physiology, and feeling are themselves inadequate and ought to be integrated.
One way of putting the point that emotions must have a cognitive component-that they cannot be simply feelings or physiological processes, or even “mindless” bits of behavioris to insist that they have intentionality. “Intentionality” is a technical notion, but its common-sense meaning can be captured by the idea that emotions are always “about” something or other. One is always angry about something; one is always in love with someone or something (even if one is also “in love with love”); one is always afraid of something (even if one doesn’t know what it is) Thus we can understand the “formal object” of an emotion as its essential intentionality-the kind of object (event, person, slate of affairs) to which it must be directed if it is to be that emotion. But intentionality has also been the object of philosophical consternation for over a century now, because despite its appeal as a way of understandin the nature of perception and other mental acts” (which gets us away from the image of images or representations “in” the mind), intentionality has its own peculiar cornplications (Kenny, 1963; Searle, 1983).
Most troubling for philosophers is the obvious fact that an emotion may be “about” some non-existant, merely imagined object. The object of fear may be nowhere around. The imagined threat in jealousy may not exist. The person one still loves may be dead. (Indeed, the problem seems to remain whether the lover knows of the death or not. In either case, the emotion is directed at a person who is in no position to receive it.) Moreover, the object of an emotion would seem to be one and the same object, whether or not it exists or not. (It is one and the same devil that is the object of a child’s fear, whether the devil exists or not.) Thus the ontological status of the intentional object of emotion causes considerable commotion. In recent decades, many Anglo-American language-oriented or “analytic” philosophers have reduced the seemingly mysterious notion of intentionality to the supposedly more manageable notion of “intensionality,” a precisely defined feature of certain sorts of sentences (Dennett, 1978, 1987). But whether intensionality does in fact capture the necessary features of intentionality is itself a topic of considerable debate, and at least seems to confuse the language with which we ascribe emotions with the nature of the emotions themselves (Searle, 1983).
Philosophers have also become concerned with the “why?” of emotions-their function and their explanation. Most of the work here has been done on the explanation of particular instances of emotion, although a few investigators have recently tackled the much larger question of the evolution and function of emotions as such (de Sousa, 1987; Gibbard, 1990). Particular instances of emotion seem to be subject to two different sorts of explanations. On the one hand, because they are intentional and essentially involve beliefs (also desires, needs, attitudes, and values), emotions seem to require an explanation that invokes a person’s beliefs and attitudes toward the world. A person is angry because he believes that so-and-so wronged him, or someone is saddened because she has found out that she has just lost a loved one, and so on.
But this cannot be a complete account of emotional explanation. We also explain emotions by citing the fact that a person has been sleepless all week, or is ill, or has been given some medication. In other words, explanation of emotion may cite an underlying cause that may or may not make mention of the object of emotion. The cause may be physiological-for example, an underlying state of irritability, an ingested drug, or a direct surgical stimulation of the brain. The cause may be some state of affairs or incident that “triggered” the person’s emotion, but this may not be the object of the person’s emotion, nor need he or she have any memory or awareness of it. (“Subliminal” messages presumably work this way.)
But how this causal explanation can be reconciled with an explanation in terms of beliefs and attitudes is not always obvious, and many philosophers tend to emphasize the importance of one over the other or to reduce all explanations to causal explanations or belief-and-desire explanations. On the one hand, one provides a fuller account of the intentionality of an emotion by describing not only its formal object (“He’s angry because he’s been offended”), but the specific details of the situation, as well as the person’s beliefs and various attitudes.
On the other hand, one provides an explanation in terms of an underlying cause that may or may not make mention of the object of emotion. Very often, however, the citation of a cause of emotion (its initiating stimulus or “trigger”) and the account of the object of the emotion will be nominally the same (“He got mad because she stepped on his toe”). The problem that has been addressed by many philosophers (and has been the subject of several weighty studies) is the relation between these two and the various problems in understanding thern together (Rorty, 1980).
The cognitive basis of emotions also raises another question, one that was often a matter of deep concern for earlier philosophers: the question of the rationality of emotions. Many thinkers have written as if the emotions were not only irrational but also nonrational-not even candidates for intelligence. Accounts of emotions as mere feelings or physiological processes would make them no more than nonrational (one cannot have a “stupid” headache, except by way of a roundabout complaint about its inconvenience). Aristotle, on the other hand, simply assumed that an emotion can be appropriate or inappropriate, foolish or prudent, not just on the basis of whether or not it is acceptable in the circumstance in question (though that social dimension is certainly essential), but on the basis of the perceptions, beliefs, and desires of the individual.
The fact that emotions consist at least in part of cognitions means that they can be evaluated in terms of the same epistemic and ethical criteria that we use to evaluate beliefs and intentions: Are they appropriate to the context? Do they consider the facts of the matter? Are their perceptions fair and their evaluations reasonable? Indeed, the argument is now prevalent and persuasive that we cannot understand emotions without grasping their reasons, and these reasons in turn give us a basis for evaluation (de Sousa, 1987; Greenspan, 1988). The current debate, however, concerns how these reasons are to be understood, and whether the rationality of emotions can indeed be fairly compared to the evaluation of more fully deliberative, articulate activities.
The rationality of emotions also moves to center stage the question of emotions and ethics that we have been following through the history of philosophy. How does emotion enter into ethical understanding, and how do our ethics affect our emotions? One thing is clear: The commingling of emotions and ethics is not grounds for dismissing either ethics or emotion, as the old emotivists suggested. It is worth noting that a new conception of the emotional foundations of ethics has taken root in the Anglo-American tradition and, an appropriate irony, has taken the name “emotivism” (Gibbard, 1990). Of course, one of the questions that remains, left over from the old rationalist charges that emotions are “merely subjective,” is that emotions vary too much from culture to culture to provide firm basis for ethics; in other words, they are “relative.” But though philosophers cannot (and should not try to) answer the empirical question of the universality or relativity of emotions, they can and should clear away the dogmatic assumptions and mistaken conceptions that have often occupied philosophy in the past.
There is nothing in the nature of emotion (including the human brain, which changes significantly with experience and varies considerably from person to person) that assures universality, out neither is it so obvious that emotions differ so much from place to place either. (This is indicated not only by studies of facial expression, but by the logic of the “human condition” and its more general features.) This also raises the question of emotions and choice-the supposed passivity of emotions. Sartre (1939/1948, 1943/1956) suggested that the emotions are willful, but many philosophers who do not share Sartre’s extreme voluntarism would agree that emotions are indeed ways of coping, whether inherited through natural selection or cultivated in the less articulate practices of a society.
But are we at the mercy of our emotions? Do we simply “have” them, or do we perhaps, to some extent, cultivate and “do” them ourselves? Obviously, a good deal of ethics and our attitudes toward ourselves depend on this. The study of emotion in philosophy is, accordingly, not a detached and marginal discipline, but the very core of our inquiry into ourselves and our own natures. it was Socrates, the great charnpion of reason, who took as his mottos the slogan at Delphi (“Know thyself ) and the rather extreme injunction that “The unexamined life is not worth living.” But part of that knowledge is surely our understanding and appreciation of our emotions, which are, after all, much of what makes life worth living.