Adam Smith and the Paradox of Tragedy

david hume moral philosophy sympathy love impartial spectator mutual sympathy paradox of tragedy gary becker propriety of action

Elias Khalil for AdamSmithWorks

February 10, 2020
Is the Loss of a Mistress More Tragic than the Loss of a Leg?
Solving the Paradox of Tragedy à la Adam Smith

The Paradox
Adam Smith notes a paradox. Viewers are ready to flock to the theaters to watch a tragedy about the loss of a mistress, but would hardly care about the tragedy of a lost leg, when obviously leg-loss is more tragic than lost-love:
The loss of a leg may generally be regarded as a more real calamity than the loss of a mistress. It would be a ridiculous tragedy, however, of which the catastrophe was to turn upon a loss of that kind. A misfortune of the other kind, how frivolous soever it may appear to be, has given occasion to many a fine one (Smith, 1976, p. 29).

Smith’s paradox is a manifestation of a broader phenomenon known as the “paradox of tragedy.” It is paradoxical indeed that people derive pleasure from watching tragedies per se—irrespective of whether they are about leg-loss or unrequited love.

Three Solutions
As Siraki reports, the paradox of tragedy was already recognized by ancient Roman philosophers but it only gained great attention in the 18th century at the hands of British philosophers.2 From Sirakis’s (2010) report, one may surmise three basic resolutions of the paradox.
As espoused by Thomas Hobbes, the spectating of tragedy is a pleasure only when the spectator is standing at a safe distance. The pain and suffering from a tragic event give the spectator a kind of pleasure through realizing that he or she is free from the observed pain and suffering. However, Hobbes’s view is limited to watching unfortunate events happening to others, such as traffic accidents, floods, hurricanes, and so on. It does not capture what is distinctive about tragedies on the stage where the spectators enter into the violent emotions that the protagonist undergoes, to the point sometimes of shedding tears.
As espoused by David Hume in his essay “Of Tragedy” published in 1757 (Siraki, 2010, p. 216), the production of tragedy on the stage requires artistic ability and eloquence. The audience seeks tragedy because it satiates an aesthetic need. However, Hume’s theory also explains comedies and other performances on the stage. If the audience is seeking solely aesthetic pleasure, it can attain it more cheaply by watching non-tragedy performances, i.e., without undergoing the pain and suffering associated with tragedies.
As advanced by Edmund Burke, what Siraki (2010, pp. 216-217) calls “the pleasure of sympathy,” the watching of tragedy is pleasurable—but not because of the supposed relative safety of the spectator. Rather, it allows the spectator to become benevolent toward the sufferer, and such sympathy affords some sublime feeling of humanity. Siraki argues that we have to read Adam Smith’s Theory of Moral Sentiments, and focus on his notion of sympathy, in order to appreciate Burke’s solution to the paradox. Indeed, Smith recognizes that people care about the wellbeing of others, about their suffering as well about their joys and pleasures, and such sympathy expresses benevolence and humanity. However, as shown next, contrary to Siraki, taking pleasure in sympathy cannot solve the paradox.3

Smith’s Sympathy cannot be the Solution
We need to commence with Hume’s definition of sympathy (see Khalil, 2015). Only then can we come to appreciate the origin of the paradox of tragedy and how Smith’s notion of sympathy cannot be a solution.
For Hume, sympathy is a matter of social interaction that involves a spectator who mirrors the original feeling of the person under focus. If the original feeling is sadness and pain, the spectator would feel likewise. If the original feeling is joy and pleasure, the spectator would feel likewise. Hume’s theory takes on a more sophisticated form in Gary Becker’s (1974) social interaction theory. For Becker, the spectator gains utility by watching the pleasure of others—and hence loses utility by watching their pain. Becker indeed reasons that much of altruism is the motive to make the person under focus (the beneficiary) feel good. So, the giver (the benefactor) provides goods to the receiver (the beneficiary) in order to enhance his or her own pleasure (see Khalil, 2004).
If Hume and Becker are correct, why would people flock to theatres to watch heartbroken men mourn over lost love? Tragedy is paradoxical only if one holds a Humean/Beckerian view of sympathy or social interaction.
However, Smith does not hold such a view of sympathy. Smith argues that sympathy is not merely about Humean mirroring. We actually express sympathy with the observed “disgust” of the person under focus only if we agree that the cause justifiably gives rise to disgust:
There are some passions of which the expressions excite no sort of sympathy, but before we are acquainted with what gave occasion to them, serve rather to disgust and provoke us against them (Smith, 1976, p. 11).

Smith investigates sympathy as a mode of judgment: is the observed emotion proportional to the cause? Smith dedicates Part I of TMS to the study of the “propriety of action” in the sense of whether the extent of the reaction to an incentive is proportional to the incentive. Sympathy-as-judgment makes sure that the decision maker commands the rational action (Khalil, 2010).
This does not undervalue the role of sympathy-as-preferences as motives from which altruistic actions and self-interested actions arise (Khalil, 1990). Indeed, Smith dedicates Part II of TMS to the investigation of the “merit of action” in the sense of whether preferences are meritorious. The preferences are meritorious, first, if they are free of aggrandizement and, second, if they are free of malevolence. The concern of merit ensures benevolence by editing out the toxic preferences of aggrandizement and malevolence.
Sympathy-as-preferences is meritorious insofar as the person under focus cares about the interest of the other (or the self) in a genuine way rather than out of aggrandizement. But still sympathy in the role of sympathy-as-preferences does not entail that people would rush to see tragedies because, following Hume and Becker, it would make them feel terrible. Even if one cares about the wellbeing of others in a genuine way—i.e., not because of mirroring à la Hume and Becker—it need not mean that one derives pleasure from watching tragedies. If one cares about the victims of floods in Bangladesh, it does not mean that one would seek to watch their misery in order to enhance one’s pleasure. To the contrary, one usually seeks to know about the affairs of the victims out of concern and out of desire to donate labor and resources to alleviate the misery of the flood victims.
So, sympathy-as-preferences cannot solve the paradox of tragedy. Neither can sympathy-as-judgment. If the judge (impartial spectator) agrees with the pain of the observed, then the person under focus is proper. That concordance, i.e., agreement, gives rise to a second-order pleasure. The second-order pleasure of the judge matches the feeling of integrity of the person under focus. While the person under focus congratulates him- or herself for carrying out the proper emotion or act, the judge likewise feels a similar emotion, a second-order pleasure that the observed person is acting properly.(Khalil, 2010).
If there is concordance, then the feeling of integrity and the second-order emotion would be positive, i.e., pleasurable. This is the case even when the original emotion/act is pain, resulting from an unfortunate event. That is, the concordance always occasions the positive feeling of integrity and the second-order emotion.
But what about the second-order emotion and the feeling of integrity if there is no concordance? That is, if the theatregoer in the role of the judge finds the emotion/act of the protagonist in the play to be a bit excessive, and the second-order emotion is negative as a result, could the theatregoer still enjoy tragic pleasure? Insofar as the theatregoer still derives tragic pleasure even in the case of disagreement (discordance), this indicates that tragic pleasure differs from the second-order emotion..
So, the agreement arising from sympathy-as-judgment cannot be the pleasure that one derives from tragedy. The reason is simple. While the former arises in the case of judgment, the latter does not arise from judgment. In fact, to enjoy the violence of the protagonist’s emotions, it is better for the theatregoer to avoid sympathy-as-judgment altogether. That is, for the tragic pleasure to take its course, the spectator would be better off retiring, at the doors of the theater, his or her potential role as a judge of propriety of action.
In short, the mere possibility of tragic pleasure even when there is discordance with respect to sympathy-as-judgment goes to show that tragic pleasure does not depend on concordance of emotions. Similar to sympathy-as-preferences, sympathy-as-judgment cannot be the solution.

Smith’s “Mutual Sympathy” contra Smith’s “Sympathy”
It is rather Smith’s notion of “mutual sympathy,” exposed at the outset of TMS (Smith, 1976, pp. 13-16), in Chapter 2, that provides the solution to the paradox of tragedy. And there is a radical difference between Smith’s “sympathy” and Smith’s “mutual sympathy” (see Khalil, 2020). This difference escapes Siraki (2010, p. 220) who mentions “mutual sympathy” while highlighting Smith’s sympathy. The difference between sympathy and mutual sympathy also escapes many other scholars (see Khalil, 2020).
Many of these scholars could have paid greater attention to Chapter 2 of TMS. But what matters is whether they realized that Smith is advancing a genus of fellow-feeling, viz., mutual sympathy, that is unrelated to the other more famous genus, sympathy. Even Smith seems to have forgotten about the uniqueness of his concept of mutual sympathy. First, Smith did not integrate mutual sympathy in his exposition of sympathy, particularly in his distinction of the two functions of sympathy. As discussed above, there is a difference between sympathy-as-judgment, what Smith calls “propriety of action,” and sympathy-as-preferences, what Smith calls “merit of action” (see Smith, 1976, p. 92).
Second, Smith failed to resurrect his mutual sympathy concept when he responded to Hume’s critique of the 1st edition of TMS. Hume expressed his critique in a letter on 28th July 1759, which the editors of the Glasgow Edition of TMS reproduced (see Smith, p. 46, n.2). Smith responded to Hume’s critique in a footnote added to the 2nd edition of TMS (Smith, p. 46 n.*). Hume’s critique consisted in pointing out that Smith argued in Chapter 2 that mutual sympathy is always pleasurable, even when the original emotion is painful, while in the rest of TMS Smith discusses “sympathy” as partaking of the original emotion. That is, for Hume and Smith sympathy should be painful if the original is painful, and joyful if the original is joyful. Hume correctly asks: how could fellow-feeling be always pleasurable, when it partakes of the original feeling?
Instead of noting that fellow-feeling involves two different concepts, viz., mutual sympathy and sympathy, Smith discusses how the second-order emotion of judgment is always pleasurable. It is true that second-order emotion of judgment is pleasurable—but only if the judgment is positive approbation. If it is negative approbation, there is discordance and, hence, the judgment is not pleasurable. This issue, viz., the possibility that the judgment could be non-pleasurable in the case of discordance, escapes Smith (see Khalil, 2020).
Thus, in his own defense, Smith’s response is non-viable. To note, many scholars (e.g., Fleischacker, 2019, p. 27), including Siraki (2010), failed to note the non-viability of Smith’s response. They basically endorsed Smith’s response.
This essay cannot detail the problem with Smith’s response. Basically, Smith should have responded by stating why he argued, in Chapter 2, that mutual sympathy is always pleasurable (Smith, 1976, pp. 14-15).
But, first, what is “mutual sympathy” for Smith and how does it differ for him from sympathy? Mutual sympathy is the feeling of bonding that arises among friends and people who love each other. In contrast, Smith’s sympathy, in either function, does not entail friendship-and-love. When a bystander sympathizes with an old man who lost his cane, such sympathy does not entail that the bystander and the old man be friends. When the old man goes into a temper tantrum, and the bystander judges that the reaction is of excessive pitch and hence cannot enter into sympathy with it (i.e., sympathize in the sense of sympathy-as-judgment), such judgment does not entail that the bystander and the old man be friends.
In contrast, mutual sympathy entails the participants to be friends, and what they are mutually sharing need not necessarily involve caring or judgment. What friends share is the re-telling, musing over, and nostalgia-like reflection on an event that has befallen one or both of them. When the friends recall and reflect upon a horrific automobile accident or a flood, they are producing a common feeling, mutual sympathy. Such a feeling, Smith (1976, p. 15) argues, is puzzling since it is always pleasurable, even when the accident or the flood was painful, a tragic event.
Smith should have responded to Hume in the following manner: “my book is not contradictory. I was not discussing in Chapter 2 sympathy, which partakes the original emotion. I was rather discussing mutual sympathy that characterizes the feeling of friendship-and-love. And it is mutual sympathy, not sympathy, that is always pleasurable.”
Given that mutual sympathy is always pleasurable, Smith (1976, p. 14) concludes that it taps “another source of satisfaction.” This source is in contradistinction to sympathy that is about pecuniary satisfaction. Friendship-and-love, the basis of mutual sympathy, affords a satisfaction that transcends pecuniary satisfaction. This explains why the transcendent feeling of friendship-and-love is always pleasurable, even when the pecuniary emotion regarding an automobile accident or flood is painful.
It is easy to conflate the transcendental satisfaction of mutual sympathy, which amounts to “another source of satisfaction,” with the second-order emotion that arises from the concordance or approbation of observed emotion/act. While transcendental satisfaction arises from the nostalgia-like reflection on an event, the second-order emotion expresses approbation, i.e., sympathy-as-judgment. Sympathy-as-judgment is not what defines friendship. While friendship may involve caring (sympathy-as-preferences) and judgment (sympathy-as-judgment), friendship is based on mutual sharing of emotions that transcends the pecuniary feelings, irrespective of whether they are pleasurable or not.

The Double Meaning of “Tragedy”
Smith’s concept of mutual sympathy is the candidate for solving the paradox of tragedy. Tragedy certainly makes the spectator feel the protagonist’s pain along with the usual Smith’s sympathy (or along Hume’s). But a successful tragedy allows the spectator to immerse him- or herself in the protagonist’s misfortune, to feel his or her pain as a friend, and consequently enjoy the pleasure of such bonding. The bonding between the spectator and the protagonist gives rise to mutual sympathy. The emergent mutual sympathy affords a transcendental emotion of friendship that is always pleasurable.
But given the pecuniary pain that defines all tragedies, the emergent bonding should afford enough transcendental pleasure to offset such pain. Otherwise, the tragedy would be a failure; the theatre house could not attract the needed audience. So, it is not sufficient that the play is tragic. The friendship must be potent enough to make the play successful.
As much as lost love is frivolous in comparison to the loss of a leg, the former evokes greater mutual sympathy than the latter. The demand for mutual sympathy explains why people go to see plays about lost love, not the loss of a leg. If they rather demand sympathy, they would prefer going to hospitals rather than to playhouses. (Even when they go to hospitals, they do not go to enjoy the suffering of the patients, but rather to offer volunteer labor to alleviate the pain of the patients.)
The solution of the leg-loss/lost-love puzzle is now clear. The puzzle arises from the double meaning of “tragedy.” In the pecuniary (sympathy) sense, leg-loss is more tragic than lost-love. In the mutual sympathy sense, lost-love is more tragic than leg-loss.
Lost-love is more tragic in the mutual sympathy sense than leg-loss for different reasons. First, lost-love is an emotion between two people and, hence, is prima facie stronger than the emotion between the sufferer of leg-loss and him- or herself. Second, it can be presumed that a finely-directed tragedy about leg-loss may evoke more mutual sympathy than a badly-directed tragedy about lost-love. But supposing the directing skill is the same, it is easier for the audience to bond with a protagonist who is in pain for unrequited love than with a protagonist who is in pain for leg-loss because the pain of unrequited love lasts (almost) a life-time. In contrast, the pain of leg-loss usually lasts a few months or a year. The person usually adjusts to leg-loss, although not as easily as supposed by set point theory (e.g., Brickman & Campbell 1971; Brickman et al., 1978)). Researchers in happiness studies are finding out that the adjustment to great pecuniary loss or great pecuniary gain is never complete (see Easterlin, 2004). But still, protagonists adjust easily to bodily pain than to emotions related to friendship-and-love.
Third, the emotion of love between two people is not pecuniary; it is already a mutual sympathy. As such, it is easier for the spectator to feel the mutual sympathy, via mirroring, than to feel whatever supposed mutual sympathy that a sufferer of leg-loss shares with him- or herself. That is, there is more emotional effort and cost on the part of the spectator to bond with the protagonist who suffers from leg-loss than with a protagonist who suffers from lost-love.
Fourth, there is little pecuniary pain in watching lost-love, while there is pecuniary pain in watching leg-loss. So, everything else being equal, a play about leg-loss must generate much more powerful mutual sympathy than a play about lost-love in order to attract the same audience.
Smith made a passing remark regarding the leg-loss/lost-love puzzle. He did not solve it or even attempt to solve it. But his concept of mutual sympathy offers the key to a solution.

Becker, Gary S. “A Theory of Social Interactions.” Journal of Political Economy, 1974, 82, 1063-93.
Brickman, P. & D.T. Campbell. “Hedonic Relativism and Planning the Good Society.” In: Adaptation-level theory: A symposium, ed. M. H. Appley, pp. 287-302. New York: Academic Press, 1971.
Brickman, P., D. Coates, and R. Janoff-Bulman. “Lottery Winners and Accident Victims: Is Happiness Relative?” Journal of Personality and Social Psychology, 1978, 36, pp. 917-27.
Easterlin, Richard A. “The Economics of Happiness.” Daedalus, 2004, 133, pp. 26-33.
Feagin, Susan. “The Pleasures of Tragedy.” American Philosophical Quarterly, 1983, 20, pp. 95-104.
Fleischacker, Samuel. Being Me Being You: Adam Smith and Empathy. Chicago: University of Chicago Press, 2019.
Griswold, Charles L. Adam Smith and the Virtues of Enlightenment. Cambridge:
Cambridge University Press, 1999.
Keen, Suzanne. Empathy and the Novel. Oxford: Oxford University Press, 2007.
Khalil, Elias L. “Beyond Self-Interest and Altruism: A Reconstruction of Adam Smith's Theory of Human Conduct.” Economics and Philosophy, October 1990, 6:2, pp. 255-273.
________. “What is Altruism?” Journal of Economic Psychology, Feb 2004, 25:1, pp. 97-123.
________. “Adam Smith's Concept of Self-Command as a Solution to Dynamic Inconsistency and the Commitment Problem.” Economic Inquiry, January 2010, 48:1, pp. 177-191.
________. “The Fellow-Feeling Paradox: Hume, Smith and the Moral Order.” Philosophy, October 2015, 90:4, pp. 653-678.
________. “Why is the Commodification of Friendship-and-Love Repugnant? The Implication of the Hume-Smith Dispute over Fellow-Feelings.” Journal of Business Ethics, 2020, under revise-and-resubmit.
Siraki, A.T. “Adam Smith’s Solution to the Paradox of Tragedy.” Adam Smith Review, 2010, 5, pp. 213-230. (eds., V. Brown and S. Fleischacker).
Smith, Adam. The Theory of Moral Sentiments, eds. D.D. Raphael and A.L. Macfie. Oxford: Oxford University Press, 1976.
Smuts, A. "The Paradox of Painful Art." Journal of Aesthetic Education, 2007, 41, pp. 59-77.
_______. "Art and Negative Affect." Philosophy Compass, 2009, 4, pp. 39-55.

  1. This essay benefited from the comments of Janett Khalil and the editorial help of Maks Sipowicz. The usual caveat applies.
  2. The paradox of tragedy still attracts the occasional attention of contemporary theorists of aesthetics (e.g., Feagin, 1983; Smuts, 2009). Smuts (2007) in particular argues that there is a toleration of horror when it appears in art. However, the paradox of tragedy shows that the issue is the opposite: it is not a matter of toleration but rather of active seeking of tragedy.
  3. In addition, the fact that Smith (e.g., 1976, p. 10; see Griswold, 1999, p. 65; Keen, 2007) did not draw a sharp distinction between fictions/novels and real life is not, contrary to Siraki’s argument, analytically relevant to the solution of the paradox of tragedy.