Rational Choice in Love

Rational Choice in Love

Rational Choice in Love

Bender, John . "Rational Choice in Love." Occasion: Interdisciplinary Studies in the Humanities v. 1 (October 15, 2009), http://occasion.stanford.edu/node/11.


I want to focus on a single work of literature. It is not directly concerned with war among states, though its author, Choderlos de Laclos, was in fact a soldier, an artillery officer, and an expert in fortifications. It deals with other kinds of war. My goal today is to consider the place of reason—of rational choice in the commonsense version of this term—and, specifically, of rational choice theory, as they play out in the realms of unreason. In order to narrow the field, I have picked love as my example. Within the unreasonable realm of the passions, love is, if not necessarily the most irrational, at least the most often subject to attempts at regulation by theoretical and practical reason.

Among the works of literature I study, the one that has struck me as most apt in its juxtaposition of love and rational choice as strategy has been Les liaisons dangereuses, the scandalous novel of 1782 by Choderlos de Laclos. At its turning point, the novel’s two chief characters declare war—another medium in which reasoned strategy, gaming, and the passions meet.

I enjoy paradox and I’m a skeptic, so I’m asking myself what rational choice theory might have to say about unreason. One way to think about rational choice theory—and this thinking has to some extent been done—is as a way of managing unreason or even of mobilizing unreason to advance the cause of the rational management of choice. Some rational choice theorists, including Jon Elster, have gone so far as to explore the role of emotion in shaping or improving the functioning of choice—even rational choice—in the world. What I am interested in today, however, is the question of how or when rational choice can be undermined or rendered impossible or irrelevant by the play of emotion. Or can the two ever act together?


In thinking about these issues, it seemed to me that to contemplate the question of rational choice in love would be to contemplate the interaction between the most systematic among methods of gaming, planning, and choice making and the least systematic—that is, love as passion.

Laclos’ novel is paradigmatic for this discussion because we see in it not just the age-old conflict between reasonable and emotional motives, but a story in which the two principal actors consciously manage their actions following a proto-scientific version of strategic thinking devised by themselves. They take on the traits of second-order observers or analysts within the fiction of the novel. We as readers are placed as third-order observers. While watching the actors manipulate the system, we are forced into a hyper-analytic apprehension of the action in which sheer technical fascination supplants absorption in the emotional plight of the characters. The strong alliance created between the reader and the leading figures, Valmont and Merteuil, virtually forces us into the position of rational choice actors, or at least third-order rational choice observers. We become rational choice theorists avant la lettre.[1]


Schelling alone among the early analysts treats rational choice theory as intimately related to irrationality. Parfit, in turn, invokes Schelling’s example of the rational use of irrationality: this a case in which armed extortionists invade the house of a rich man and threaten to kill his children one by one until he hands over the gold kept in a safe. Since he realizes that surrendering the gold won’t save the children or himself, the only rational choice is for the rich man to self-induce an hysteric fit of despair in which he demands that the children all be killed immediately because there is no hope for any of them anyway. The invaders lose their leverage and flee.

Relevant here is Kahneman, Slovic, and Tversky’s book Judgment Under Uncertainty, which documents a variety of experiments showing that human beings typically make irrational decisions because they are guided by commonsense heuristics that they believe to be reasonable. These include: (1) faith in cause and effect sequences, (2) naive estimations of probability, (3) overprediction based on limited experience, or (4) misperception of their own capacity to control events.[2]

In this context, one may view rational choice and game theories as strategies that seem to aim at overriding the unreason of ordinary experience by marking out systematic methods of decision making. But in repressing the very unreason they work to transcend, these theories risk becoming yet another of Kahneman, Slovic, and Tversky’s heuristics. After all, the rich mad man in Schelling’s and Parfit’s example is making the seemingly rational choice to induce an episode of irrationality for the benefit of intruders. One is reminded of Carl von Clausewitz, the great nineteenth-century theorist of war as tool of diplomacy. And, one could link rational choice and game theory to the Cold War, which Schelling’s book openly theorizes. “Brinksmanship” was a version in the 1950s, and a later relative is Nixon’s so-called “madman” strategy to persuade the Russians to pressure North Vietnam for peace out of fear that he was obsessive enough to use nuclear bombs on them.

With this background sketched in, let’s consider how two different historical love systems play out in Laclos’ novel—one system in which love is stringently and explicitly rule-bound, the other in which love is detached from rational management. These are, broadly speaking, the classic absolutist system in which marriage is an economically, socially, and politically motivated contract independent of love. And, on the other hand, the more modern system in which marriage follows upon romantic love—that is upon what Niklas Luhmann in his book Love as Passion has called a real or phantasmatic “ecstatic interpenetration” of the passions of two subjects.[3] This is the kind of love that provokes war in Laclos’ novel.

In the first system—at least in the aristocratic absolutist era toward the end of which Laclos’ novel plays out—love no less than marriage was managed through rule-bound affairs that were themselves tightly coupled to the power hierarchy of community politics, of which the court was the most extreme and stylized version. Even loves that played out at the margins of these rules, and may seem on the surface to be modern in character, like that of the Princesse de Clèves in Madame de Lafayette’s novel, in fact were rigidly contained by protocol and subordinated within a single systemic realm. Lafayette’s novel enacts the ultimate force of the absolutist system to contain passion within the conventions of courtship and marriage while stretching the membrane of its power to its extreme limit. Technical signs of this stress in the system, as portrayed in the novel, are those remarkable passages narrating consciousness, and using the modern technique of free indirect discourse, that break out as the Princesse struggles to contain her love for Nemours within the chastity of thought and action that defines spousal duty to her husband.

The romantic love system, by contrast with the contractual, runs in tandem with the modern functional differentiation of socio-political domains. Subject-centered romantic love goes hand in hand with the individual mobility required by societies in which hierarchy is based on organizational function. Paradoxically, the untethered irrationality of modern romantic love enables functionally differentiated societies precisely because the love system becomes relatively autonomous from other social systems. This love operates to cut individuals loose from formal social networks and to make them free operators in a modern marriage market and in a functionally differentiated economic system. Individualism increases and large-scaled family ties decrease as skills and function supplant long-term social networks.


The subjective manifestation of this romantic love is of an absolute, profound, and irrefutable incapacity to make rational choices. At least in its idealized (or ideological version), this self-surrender works in modernity as a secular version of empowerment in the world that Luther attributed to his surrender to God’s mercy through faith. It enables bold, improvisatory actions in a mobile and unpredictable world and, as such, goes in tandem with Weber’s sense of the professional “calling” as an enabling force. Love in the private realm, and the professional calling in the economic domain, are the engines of productivity.

Conflict between the old and new systems of love and marriage is the force that actuates the plot of a famous novel that shares the formal perfection of Laclos’ work without its overt conceptual complexity. I have in mind Austen’s Pride and Prejudice. Austen puts all of the machinery of the managed marriage system in play. The Bennets will lose their wealth and social position without profitable marriages for their daughters. The highly placed and rich gentleman, Bingley, is the object of Mrs. Bennet’s strategy to profit from the marriage of Jane, her most beautiful daughter. The aristocratic Darcy—already destined to marry his socially well-placed cousin even before the novel begins—is the plot’s candidate to marry the heroine, Elizabeth Bennet. The planned marriage to a cousin is an unmistakable marker that the absolutist system is operative. Elizabeth is at the largest plausible social remove from Darcy within the plot’s requirement that the two be able to meet socially. At that, their crucial encounter at his noble house in Derbyshire and in its environs is depicted as an accident. In the immediate aftermath of her angry refusal of his first proposal after that chance encounter, Darcy serves up a litany of reasons against their union straight out of the lexicon of managed marriage. All of these were overcome, he says, by his love. The only force to countervail managed marriage is an ecstatic love than can be assuaged solely in marriage. When Darcy’s aunt, Lady Catherine, descends unannounced on the Bennet house, her purpose is to declare the impossibility of her nephew’s marriage to Elizabeth on the ground that every kind of social decency forbids it, no less than does a pre-existing lineal alliance with his cousin. Ecstatic love of course prevails over managed marriage and countless misunderstandings.


In the romance of the novel, this love brings with it wealth far beyond the hopes of the doomed manager—Mrs. Bennet—and also dashes the hopes of her spiritual sister Lady Catherine. Lest we miss the point, however, Elizabeth’s friend Charlotte Lucas, having accepted marriage to the obsequious and officious Mr. Collins, who ultimately will inherit the Bennet property by entail, delivers a long statement to Elizabeth on the merits of rational marriage and the limitations that force her to accept it as an alternative to romance. Nor does Austen fail to show the dangers of love as passion. For the brilliantly intertwined plot of her sister’s elopement with Darcy’s prime enemy, Wickham, both reveals the evil of passion run amok precisely because it is not governed by the rigid frameworks of the managed system and also occasions philanthropic actions by Darcy that prove his worth to Elizabeth.

It may be helpful at this point, for those of you who have seen the Stephen Frears’s 1988 film Dangerous Liaisons to recall that the two main characters are Glenn Close and John Malkovich. Glenn is a secretly libertine young widow, the Marquise de Merteuil; and John is a younger, infamously libertine playboy, the Vicomte de Valmont. They are long-term lovers and co-conspirators. Their chief victim is Michele Pfeiffer, a ravishingly chaste wife known as Madame, la Presidente de Tourvel. The film simplifies the novel, and I am reducing it even more. Literature, rational choice theory, and the history of western societies all must fall prey here to time limits!

Glenn Close—sorry, Madame de Merteuil—writes a long letter in this epistolary novel about her realization as a young widow that she could rationally master both her own passions and the classic love system in order to gain independence, pleasure, and power for herself through choice. It is her manifesto. Presenting a facade of unassailable virtue based on her reading of conventional moralists, she ranges through love affairs that allow her to shape the destiny of others by knowing their secrets, while guarding hers. On her own account, she learns how to do so from novelists and philosophers—the then current equivalent of rational choice theorists.

It is hard to have fun alone, however, and when the novel begins, she already has formed an alliance with John Malkovich—sorry, the Vicomte de Valmont. They have been physically passionate in the past but have turned for a time to a more delicious intercourse as co-conspirators in a plot Merteuil has designed for revenge against a range of enemies. The seduction of innocent women will be their means. One of these is Michele Pfeiffer—sorry, Madame de Tourvel. Valmont’s reward for executing these more than cynical conquests will be a return to Merteuil’s bed. When he has succeeded, and demands his reward, war breaks out between our conspirators. Our anti-hero is killed in a duel. Our anti-heroine is disgraced by letters he hands over on his deathbed (including her infamous manifesto) and, in an allegorical move, her face is ravaged by smallpox.

What destroys Merteuil’s and Valmont’s rational management of the classic system of love separated from marriage? What leads them to war? They have been supremely successful players at a secret second-order game based on their rational analysis of the implicit rules and strategies of the first-order love system accepted by their society. They have executed their choices to perfection. They have gained the pleasures—admittedly vicious ones—of sex with and superiority over their dupes, along with the power to destroy others at will. All the while, enjoying places of rank and prestige in a ritualized aristocratic society that continually threatened tedium for members without actual posts at court. Add the frisson that each knows enough about the other’s crimes to bring prison or exile down on their heads, and you have a transgressive scene that brought scandal upon Laclos and his book in 1782, and left an air of danger around the novel until quite recent times.

The two players are destroyed because their game depends on absolute rational detachment. No doubt, Valmont may, along the way, enjoy writing letters on the rump of a young virgin he recently deflowered—a famous scene. Merteuil may savor a complex bed trick arranged to humiliate an old fop. Likewise, now-a-days, a skilled poker player may savor reducing another person to nothing through a brilliantly executed bluff without for an instant losing the count and continuing to play the hand according to strict analysis.

Merteuil’s infamous letter describing her analysis of the world around her after marriage and the detached quest for power and pleasure that she pursued thereafter (Letter 81), no less than similar if less sustained and analytic manifestos from Valmont, assumes a degree of control and predictability of causation that may itself be understood a delusion—or a heuristic in light of Tversky et al.’s findings. Merteuil’s system for putting herself in charge as a woman is itself rule-bound and relies for its effectiveness on being lodged like a Chinese box within another system of predicable rules, reactions, and conventions. In short, she depends on everyone else in her social order to perform the rules of their game perfectly or at least to be manipulated into doing so in her skilled hands. The “classical” French style of the novel—so disciplined, simple and rigorous that it recalls Racine—works to signal that the managed social system is in play. In reality, of course, the actors and the audience are in the midst of a revolution in love systems.

Merteuil lives the fiction that love could be managed within rational choice: the most irrational of experiences could be mimicked within a clearly understood set of rules and thus brought into the domain of choice. Thus when we learn that Merteuil and Valmont were lovers, and when they say they love one another, we must understand that they were and are working within this paradoxical old system. Within this system, the more extreme the passion, the more strict the regulation it will evoke in terms of discretion, delimitation and the understanding that it will have a self-contained and secret history or story. It is no accident that the genre called the “secret history” was a leading type of fiction in the seventeenth-century heyday of absolutist marriage.

We see these assumptions played out in La Princesse de Clèves where the tragedy is actuated by the husband’s knowledge and the heroine’s profound reaction to this knowledge. In this paradoxical system, every lover is always already a former lover just as every person is always a potential lover. The system also had attributes that lead to its own internal acceleration and intensification. In its normal playing out, the affair between sophisticated lovers who know and follow the rules, the system works or worked within the frame of conventionalized marriages. Merteuil and Valmont introduce a mutation by gaming a system that is itself a game: they join in an alliance to manipulate others in part to gain specific goals like humiliating an enemy but largely for the displaced erotic pleasure of displaying their manipulations of the system to one another.

The old system had a certain stability but their gaming of its rules is intrinsically unstable because, like all games, it may be imperfectly played by the dim witted or inexperienced. Or imperfectly played by the impassioned! Our gamesters Merteuil and Valmont believe on the basis of insufficient data, to mention another of Tversky’s heuristics, that they can predict the playing out of their maneuvers even though their private game is parasitical and therefore intrinsically unstable. This instability can be understood historically and sociologically as a function of changes in the love system as Luhmann accounts for it. They are gaming the game at the edge of the historical period in which the old set of rules pertained. The condition of possibility for their gaming might even be seen as the breakdown of that old system of managing marriage and love. That system had itself included strict protocols for the management of passion.

Valmont’s ruthless management of the genesis and efflorescence of the Presidente’s horribly destructive passion positions him as a kind of third-order actor playing out a script that has been given him by Merteuil. The moment he moves from equality with Merteuil as a player of the second order, he is doomed. At the fatal climax of his affair with the Presidente, Merteuil literally scripts the words of his breakup with the Presidente—“It isn’t my fault.” Both conspirators are fatally weakened at this point as they move even toward fourth-order positions in which she must script him in order to trust him and he must in effect become a reader to his own actions. In a sense, they join us in attempting the role of rational choice theorists. As I have noted, the emotional impact of the horrible destruction of the Presidente is defused or short-circuited by the layers of observation. Even in Michele Pfeiffer’s brilliant film performance, it is hard as a viewer to enter into full empathy with the Presidente. At the end of the novel this house of cards—this system of second-, third-, fourth-order observations—collapses, and Valmont and Merteuil become first-order actors deprived of their tools of manipulation. They move out of their roles as classical rational choice theorists into the position of ordinary actors.

The fatal flaw is that Valmont falls hopelessly into romantic love with the Presidente, Madame de Tourvel, just upon the completion of his most impossibly difficult, most extended, most ingenious, and most vicious seduction. This truly chaste woman submits after months of agony—for she has been relentlessly manipulated and does love him passionately. She submits sexually when, in a clichéd strategy straight out of heroic drama plots or opera, he convinces her that he will kill himself if she does not. This is a variant on the madman strategy, which clearly pre-dates rational choice theory.


When Valmont goes to collect his prize from Merteuil and describes his triumph in novelistic—even confessional—detail, she quickly sees that he is in love. Not in lust, not in the libertine quest to expand his repertory of pleasure, but in modern romantic love. Valmont is so enmeshed in the old game with Merteuil, and so unable to recognize the fate that has overtaken him, that he accepts her challenge to break off the affair with Madame Tourvel before gaining his wages (for that is what Merteuil’s submission would be). He acts in this way, contrary to his own interests, to prove that he is bound by their game, not by love of the Presidente.

The break is to be accomplished brutally by Valmont’s catechistic invocation, fully scripted by Merteuil and rehearsed with her, of the many reasons ranging from boredom to continuing lust for another woman that he must end his affair with his newly found, deeply felt love. Each reason is followed by the phrase, “It isn’t my fault.” The strategy is cruel, not least because it echoes religious formulas dear to the devout Madame de Tourvel. Counter to Valmont’s own deep feelings, it is so successful that she believes him and actually dies of a broken heart.

Valmont is trapped by the binary logic of rational choice games in which there is always a “dominant” position that wins over other possible but strategically inferior positions or choices. Jean-Pierre Dupuy has shown that this binary, which underlies so much of modern rational choice theory, is intrinsically flawed because it forecloses the multi-variable realities produced by human bonds, especially the collective social order of affect. One can be in the dominant position, as Merteuil is and as Valmont seeks to be, but one can fail to realize options outside the game or, one may say in historical terms, the emergence of new games with more variables. Laclos already is exploring the limitations of game-oriented choice.[4]


In the final pages of Les liaisons dangeureuses, the conditions of possibility for the game have vanished. Valmont is still in love with Tourvel and Merteuil knows it. Further, in a twist that recalls René Girard’s plot of mimetic desire, Merteuil’s old-style erotic relationship with Valmont morphs into jealous love. The more he presses for his wages, the more jealous she becomes at his love for her rival. The more he insists, the more, as Merteuil says, he resembles a husband, and she had long ago resolved never again to have another of those. Jean-Pierre Dupuy notes that this letter shows Valmont to have lost the possibility of making the dominant move in his game with Merteuil. She writes, “Oh well, I shall do your reasoning for you. Either you have a rival or you don’t. If you have one, you must set out to please, so as to be preferred to him; if you don’t have one, you must still please so as to obviate the possibility of having one. In either case the same principle must be followed” (Letter 152). Merteuil is describing the “sure thing principle” that rides in tandem with the dominant position in games. But Merteuil, who was the perfect player when allied with Valmont, now makes the mistake implicit in the binary of dominant and inferior. She fails to see that there are options outside the stylization of the game.

This bitter standoff ends with Valmont’s explicit challenge to Merteuil that she submit or go to war. She chooses war. It appears at first that the kind of war Merteuil has in mind is the kind of limited war that paralleled the managed love system and marriage arrangements in absolutist society. By the rules of her game, she might have submitted cynically. It surely would not be the first time for her to do so as part of a long-term strategy. Later, she could move against Valmont. In fact, Merteuil has actuated the entire plot of the novel by setting up a scheme of managed revenge against a former lover. But no. The game is over and the players are destroyed because they love one another in a way surprising and incalculable for them. Their private version of modern, absolute war breaks out. They make the irrational choice.

Luhmann describes the emotions as products of internal crises in psychic systems or, as we might say, consciousness. Psychic systems work to maintain their own coherence and continuity. They are, for Luhmann, fundamentally like Locke’s “understanding.” That is, they are self-governed by their own intrinsic capacities to absorb and process information and to sustain and reproduce the chains of cause and effect that their own internal coherence requires. Luhmann accounts for emotions as the immune function of psychic systems. That is, emotion emerges when systems of consciousness are overloaded or confused. Emotions, thus, are not “representations that refer to the environment but internal adaptations to the internal problem situations in the psychic system that concern the ongoing production of the system’s elements by the system’s elements. Emotions are not necessarily formed in an occasional and spontaneous manner; but one can be more or less disposed to an emotion-laden reaction.”[5] Emotions are, instead, effects of process in the psychic system. For this reason, they are transient and unreliable. They also can be socially conditioned and, as one may infer here and from Love as Passion, historically conditioned. It is now possible to understand more fully what Luhmann means when he calls romantic love an “ecstatic interpenetration” of the passions of two subjects.

Luhmann allows us to theorize the conditions under which rational choice, and even rational choice theory itself, as an exercise may be highly mobile historically. How the passion of love is managed or becomes evident in different historical and culture situations can vary phenomenologically. The example of La Princesse de Clèves illustrates a very powerful system for managing the passions—a system that normally contained them within social conventions and yet allowed considerable play of passion within the rigid structures of marriage. The breakdown occurs in Lafayette’s novel because the Princesse’s husband truly does love her and cannot recover from the blow of imagining that she loves someone else. She is doomed because neither her psychic system nor the social system she has internalized can process the overload. The breakdown that occurs in Laclos is at a historical node between systems that renders impossible the management of passion either by accepting it in its modern version or confining it in its classical version. Even Merteuil is infected by the demonic double of the modern version of passion, which erupts in her jealousy of Valmont and the Presidente. Toward the end of the novel she displays jealousy as a version of mimetic desire that contrasts with the revenge motivated plot with which she actuates the whole drama at the beginning of the work.

I leave you with a few questions. Does rational choice have a history analogous to that of love systems? This history would take commonsense gaming that we might call first-order or implicit rational choice and place them into relationship with meta-theoretical, second-order phenomena like rational choice theory itself as a body of knowledge. This history would perhaps define eras in the past when rational systems were instrumental—analogously to their importance in recent decades: here I have in mind Scholasticism and its practical interactions with religious belief in the late middle ages. In the same vein, are we nearing the end of a historical period in which concepts like utility and the rational management of actors and outcomes can guide actions in a world that delimits choice—whether choice by corporate systems, orderly protocols, or individuals—and delimits choice through the sheer force of unreason? Valmont and Merteuil, on my analysis, are on the cusp between two systems for managing the sexual impulse. They are hyper-rationalists and brilliant players at the end of an epoch. When environmental conditions change around them—including their own responses to these conditions—their rational system assumes qualities of a doomsday machine.



John Bender is Jean G. and Morris M. Doyle Professor in Interdisciplinary Studies at Stanford University, in the Departments of English and Comparative Literature. Publications include “Hume’s Learned and Conversible Worlds” (with Robin Valenza) in Just Being Difficult?, ed. Jonathan Culler & Kevin Lamb (Stanford, 2003); “From Theater to Laboratory,” Journal of the American Medical Association, March 6, 2002; "Matters of Fact, Virtual Witnessing, and the Public in Hogarth's Narratives" (in Hogarth: Representing Nature's Machines,” 2001); "Enlightenment Fiction and the Scientific Hypothesis" (Representations, 1998).

He is author of Spenser and Literary Pictorialism, Imagining the Penitentiary: Fiction and the Architecture of Mind in Eighteenth Century England (which won the 1987 Gottschalk Prize for the best book on an eighteenth-century topic) and co-editor (with Simon Stern) of Tom Jones (Oxford, 1996), (with David Wellbery) of The Ends of Rhetoric: History, Theory, Practice and Chronotypes: The Construction of Time (Stanford, 1990 and 1991), and (with Michael Marrinan) of Regimes of Description: In the Archive of the Eighteenth Century (Stanford, 2005). His book The Culture of Diagram, written with Michael Marrinan, was published with Stanford University Press in 2010.


[1] During this adventure, I have consulted several basic works on the classical theory of games and rational choice. Among them were John von Neumann and Oskar Morgenstern, Theory of Games and Economic Behavior, 60th ann. ed. (Princeton, NJ: Princeton University Press, 2004); Morton D. Davis, Game Theory, A Non-technical Introduction (New York: Basic Books, 1983); J. C. C. McKinsey, Introduction to the Theory of Games (New York: McGraw-Hill, 1952); Thomas C. Schelling, The Strategy of Conflict (Cambridge, MA: Harvard University Press, 1960); and Derek Parfit, Reasons and Persons (Oxford: Oxford University Press, 1984). On the interaction of reason and emotion, the more recent works of Elster and John C. Harsanyi have been crucial. See Jon Elster, ed., Rational Choice (Oxford: Basil Blackwell, 1986); Solomonic Judgements: Studies in the Limitations of Rationality (Cambridge: Cambridge University Press, 1989); Alchemies of the Mind: Rationality and the Emotions (Cambridge: Cambridge University Press, 1999); and John C. Harsanyi, Rational Behavior and Bargaining Equilibrium in Games and Social Situations (Cambridge: Cambridge University Press, 1977).

[2] Daniel Kahneman, Paul Slovic, and Amos Tversky, eds., Judgment Under Uncertainty: Heuristics and Biases (Cambridge: Cambridge University Press, 1982), esp. chap. 4.

[3] Niklas Luhmann, Love as Passion: The Codification of Intimacy, trans. Jeremy Gaines and Doris L. Jones (Cambridge: Harvard University Press, 1986), esp. chap. 16.

[4] Jean-Pierre Dupuy, “Quand la stratégie dominante se révèle irrationnelle,” in Décision, organisation, prospective. Mélanges en l'honneur de Jacques Lesourne (Paris: Dunod, 2000). Dupuy and I both spoke at the conference at which this paper first was delivered. During the question period, he declared that a paper of his own began with a quotation from letter number 152 of the novel (see below).

[5] Niklas Luhmann, Social Systems, trans. John Bednarz, Jr., with Dirk Baecker (Stanford: Stanford University Press, 1995), 274. H. U. Gumbrecht pointed me to this discussion by Luhmann in his “Mozart Presences: Can We Describe the Pleasures of Listening?,” published as “Mozarts‚ Präsenzen’: Ist es möglich, Hörvergnügen zu beschreiben,” in Mozarts Klavier- und Kammermusik, ed. Matthias Schmidt (Laaber: Laaber 2006), 483-98. My understanding of Luhmann is indebted as well to David E. Wellbery, “Aesthetic Media: The Structure of Aesthetic Theory Before Kant,” in Regimes of Description: In the Archive of the Eighteenth Century, ed. John Bender and Michael Marrinan (Stanford: Stanford University Press, 2005), 199-211.