The sympathy we feel for physical suffering, and which constitutes part of what we call humanity, would be a sentiment too transient to be often useful, my dear C***, were we not capable of reflection as well as sensation. But because reflection prolongs the ideas brought to us by our senses, it extends and preserves in us the effect brought on by the sight of suffering, and it is that, one might say, which makes us truly human. Indeed, reflection will fasten in our souls an instance of suffering that was only present to our eyes for one brief moment, and make us want to be relieved from it and from the unwelcome and painful idea of it. It is reflection which, making up for our natural changeability, forces our compassion into action by presenting it anew with objects which had only made a momentary impression. It is reflection which, when we see someone oppressed by pain, reminds us that we too are subject to that same tyrant, destroyer of life, and through emotion and self-pity moves us closer to her, making her sufferings interesting to us, even when they are more repulsive than attractive.--Sophie de Grouchy, Letters on Sympathy, translated by Sandrine Berges, Letter II.
Aesthetics -- and in particular, the effects of beauty -- is of central importance to the argument of the Letters on Sympathy. Grouchy brackets the metaphysical question about the nature and origin of beauty. Rather, she settles for a pragmatic definition: something is beautiful if it gives us pleasure to watch (or hear, etc.) it. This means it is grounded in sensibility. And this means that for her, paying attention to beauty and beautiful people can be a significant source of "happiness" (Letter III), especially if we sympathize with them. Elsewhere I have discussed the significance of beauty to her account (recall) of demagoguery. So, here I want to focus on her 'solution' to the problem of tragedy.
The problem of tragedy—a topic widely discussed in eighteenth century thought--can be stated (with thanks to Wasserman)* in simplified form as follows: why do viewers find pleasure, or something like pleasure, in the dramatic representation of an action that is painful and repellent in real life. This problem generated a huge amount of reflection in the middle of the eighteenth century. As Grouchy notes in Letter IV, Adam Smith had also taken it on in The Theory of Moral Sentiments. Modern students may encounter a version of the problem when they enjoy a horror movie.
The problem of tragedy is an especially acute one for Grouchy for two related reasons: first, given her account of sympathy, we do not find the suffering of others pleasant, and we try to remove the causes of such suffering. But while audiences sometimes leave a theater in disgust (or avoid going to a horror movie), we generally do not try to help those suffering on stage. So, how come we willingly go watch tragedies and enjoy them? Second, she claims (and here she adapts a view standard since Scaliger) that the "point of tragedy" is in part ethical: it is in "great part to make our sympathy for the misfortunes of others seem pleasant." So, tragedy serves a didactic purpose in virtue of being pleasing. But given her account of sympathy, why would it do so?
It is central to Grouchy's argument that we sympathize more with pains that we are likely to experience (Letter IV). This is as true for physical as psychological discomfort, even if according to her, we tend to sympathize more with physical suffering than the psychological ones.
So, her strategy in dealing with the problem of tragedy is to emphasize two features of a tragic performance: first, the events portrayed are unlikely to happen to the viewers. Most of us are not kings or heroes, and so according to Grouchy we really do not identify directly with the protagonists and so do not really sympathize with them. Second, tragedies prepare our sensibility slowly and avoid "suddenly confronting" us "with the heartbreaking sight of physical pain." So, de facto, these two features entail that we actually sympathize very little or very lightly with the agents portrayed. The exceptions -- pity for the fallen mighty -- prove the rule: "we have more compassion for their misfortunes than for those of others, it is only because kings who seem to be preserved from misfortune by their elevation, strike us as more sensitive to them." (Letter IV; here she explicitly disagrees with Adam Smith.)
As a solution to the problem of tragedy this is unsatisfying. And because she rejects Smith's idea that the outcome of the sympathetic process is always pleasing, she seems in a bad position to solve the problem. But she is clearly assuming some background commitments that mitigate the problems with her position. For example, when she first introduces the problem of tragedy, she discusses the nature of boredom, which on her account we do anything to avoid. It is so "unbearable" that "we do not fear, in order to avoid it, to give ourselves over to painful sensations, and that it is the desire to avoid it which leads us to seek the proximity of the idea of suffering." (Letters II.) This means that we subject ourselves to tragedy in order to avoid a worse fate (boredom), especially if the ideas encountered in it are fairly novel. (This is why it matters to Grouchy that in tragedy we encounter events we are unlikely to experience.) These ideas were already used (go read Wasserman) by two influential and widely read French theorists, Du Bos and Fontenelle, in addressing the problem of tragedy. (It would be a surprise if she had not encountered either in her education and readings.)
Grouchy adds an important point. We can identify with the victims of tyranny in a tragedy because we can be subject to the same tyrant. She seems to evoke Antigone at one point in the passage quoted above in Letter II. But we do not find this pleasant or attractive at all. Rather, this identification activates "self-pity" and so we find the actions of the tragedy of interest (even if unpleasant). This serves the didactic purpose of generating an interest in humanity. So, the other part of Grouchy's strategy in addressing the problem of tragedy is to deny simply one of its premises altogether: we do not really enjoy a tragedy even when we find it of interest.
*I thank James Harris for calling my attention to Wasserman.