Fictional Emotions Through Chinese Aesthetics — Epoche C2
A reader who weeps at Anna Karenina's death has not forgotten that there is no Anna Karenina. That combination — a full-blooded emotional response together with the settled knowledge that its object is invented — is what philosophers since the mid-1970s have called the paradox of fiction, and the standard Anglophone resolutions all work by showing that the response is not quite the emotion it appears to be. This essay argues that the classical Chinese pairing of qing 情, the affective and dispositional side of a person, with li 理, the pattern or coherence that runs through things, supplies a better-motivated alternative: one on which feeling correctly is itself a form of discernment, so that nothing needs quarantining. Before that case can be made, however, the target has to be described accurately, because the version of it in the earlier text of this essay conflated three distinct positions. The paradox, stated so that it can be solved Colin Radford (1975) set the problem in the form still used. Take a reader $r$ and the proposition $\varphi$ that Anna throws herself under a train. Three claims each look compelling on its own: $$\text{(P1)}\ \ \mathrm{Emo}(r,\varphi)\qquad\text{(P2)}\ \ \mathrm{Emo}(r,\varphi)\rightarrow\mathrm{Bel}(r,\varphi)\qquad\text{(P3)}\ \ \neg\mathrm{Bel}(r,\varphi)$$ P1 says the reader is genuinely moved by $\varphi$; P2 says genuine emotion towards a state of affairs requires believing that it obtains; P3 says the reader believes no such thing. P1 and P2 entail $\mathrm{Bel}(r,\varphi)$, which contradicts P3, so at most two can stand. Every position in the literature is identified by which limb it drops, and this is where the earlier text went wrong. It attributed to Gregory Currie something it called his "thought theory", and said that on this theory our responses are "quasi-emotions" that are not genuine. Those are three different views belonging to three different people, and one of them holds the reverse of what was claimed. Position Limb dropped What the response is said to be Radford (1975) none genuine, and therefore incoherent — we are simply irrational here Walton (1978) P1 quasi-fear, a real felt state inside a game of make-believe; it is fictional of us that we fear Lamarque and Carroll (1990) P2 a genuine emotion whose object is the entertained content, not a believed-in particular Currie (1990) P1 real affect driven by simulated, off-line counterparts of belief and desire The "quasi-emotion" is Kendall Walton's coinage, from "Fearing Fictions" (1978): watching the slime advance, the viewer undergoes quasi-fear — genuine arousal, genuinely unpleasant — while it is only fictionally true that he fears the slime. The "thought theory" is Peter Lamarque's and Noël Carroll's, and its whole point is the opposite of what was ascribed to it. Carroll's The Philosophy of Horror (1990) argues that art-horror is a real emotion, on the ground that emotions in general are directed at contents that need only be entertained; what the horror film supplies is a thought held before the mind unasserted, and being frightened by a thought is not a defective way of being frightened. Dropping P2 rather than P1 costs nothing, on this view, because P2 was never true of emotion generally. What Currie actually argues, and what it leaves out Currie's The Nature of Fiction (1990) takes the other route. His account is simulative: in reading, one does not form the belief that Anna despairs, but runs a belief-like state with that content off-line, together with desire-like states, and feeds them into the same affective machinery that processes real beliefs and desires. The output is real feeling; what is make-believe is the intentional structure that produced it. Currie is therefore dropping P1, but with a mechanism rather than a stipulation, and the mechanism earns two things. The first is behavioural inertness. Nobody climbs onto the stage to stop Othello, and any theory must explain why. On the simulation account the explanation is structural: the states feeding the affective system are quarantined from the practical system precisely because they are not beliefs, so they produce affect without producing action. The second is a set of correct asymmetries. Reading about a murder and witnessing one differ in more than intensity — they differ in what one is subsequently disposed to do, what one takes oneself to know, and whom one holds responsible — and a theory that made them the same state would be wrong. Here is the turn on which everything after this depends. The complaint against Currie is not that his psychology is mistaken; it is that his account is deliberately a psychology and stops there. It says how the state is produced and why it does not issue in rescue attempts. It says nothing about why the state is worth having, why a tragedy that produces it well is better than one that produces it badly, or how a response could ever be the right or wrong response. Those are the questions the Chinese material was developed to answer, which is why it is worth importing — not as ornament, and not because a tradition is old, but because it is a worked-out rival on the point where the simulation account is silent. Qing and li: what the terms actually are Two cautions before the terms can be used, both of which the earlier text needed and did not have. The first concerns qing 情. A. C. Graham (1989) argued that in pre-Han texts qing does not mean "the passions" at all: it names what a thing genuinely is, the facts of a matter, so that the qing of something is its actual condition as against its name or appearance. The affective reading is a later development, dominant by the medieval period and central to the literary criticism of the sixteenth and seventeenth centuries. An argument that reads "emotion" back into Warring States occurrences is therefore anachronistic, and the pairing this essay needs is the late one. The second concerns li . Two different characters are romanised identica