The Epistemic Rationality of Fictional Emotions — Epoche C2
The paradox, stated so that it bites A viewer who tenses as the Kim family scrambles out from under the coffee table in Bong Joon-ho's Parasite (2019) is in no doubt at all that there is no Kim family, no house in Seongbuk-dong, and nobody under any table. The tension is nevertheless real: measurable, unpleasant, and hard to switch off. The question this essay addresses is whether that response — and the graver responses that films about historical atrocity elicit — can be epistemically rational: whether it can stand in evidential relations, be well or badly calibrated, and contribute to the justification of a judgement about the work. The answer defended here is that it can, but not for the reason usually given, and the reason usually given is wrong in a way that has kept the debate stuck. Colin Radford put the difficulty into its modern form in 1975 by asking how we can be moved by the fate of Anna Karenina. The difficulty is standardly regimented as three claims, each independently plausible, which cannot all be true: We have genuine emotions directed at fictional characters and their situations. Having a genuine emotion directed at something requires believing that it exists and has roughly the properties the emotion attributes to it. We do not believe that fictional characters exist. Any pair of these is comfortable; the three together are not. Radford's own response was to accept all three and conclude that we are, in this corner of our mental lives, incoherent — that the response is real, common, and not defensible. Positions in the subsequent literature are best sorted by which claim they sacrifice. Kendall Walton, in "Fearing Fictions" (1978) and at length in Mimesis as Make-Believe (1990), sacrifices (1). Noël Carroll, in The Philosophy of Horror (1990), sacrifices (2). The argument below takes a third route: it grants the triad for the narrow class of cases in which the object of the emotion really is a non-existent individual, and argues that the responses which do epistemic work have objects that do exist — kinds of act, historical events, forms of loss — so that premise (3) does not apply to them. Before any of this can be assessed, Walton has to be stated correctly, because the objection most often pressed against him — including in an earlier version of this essay — is one he anticipated and conceded. What Walton actually claims, and why "the tears are real" does not touch it Walton's example is Charles, who watches a film in which a green slime oozes towards the camera. Charles's heart rate rises, his muscles tighten, he grips the arms of his chair; he may cry out. Walton grants every bit of this. He calls the resulting state quasi-fear , and quasi-fear is not a pretend feeling, a simulated feeling, or a feeling Charles is play-acting. It is an actual physiological and phenomenal state, as real as any fear Charles has had. The word "quasi" marks not the unreality of the sensation but Walton's refusal to classify the whole state as fear. The refusal rests on a functional argument, not on a bare stipulation about belief. Fear, on any account, is individuated partly by what it disposes one to do: to escape, to warn others, to remove the threat. Charles does none of these. He does not leave the cinema; offered the chance to leave, he declines; he does not telephone anyone; he pays for a ticket in order to be in the room with the slime, and afterwards reports having enjoyed it. A state stripped of the motivational profile that distinguishes fear from other unpleasant arousal is not, Walton argues, fear — it is the arousal without the emotion. His positive account then explains the phenomenon without the emotion: Charles is a participant in a game of make-believe in which the film is a prop, and his quasi-fear is what makes it fictional — true within the game — that Charles fears the slime. This is why the standard rebuttal fails. Pointing at a weeping audience and insisting that the tears are not make-believe tears, that they are genuine physiological and psychological events, concedes what Walton asserts and denies nothing he claims. Walton's thesis is not that the audience is dry-eyed; it is that weeping plus the absence of any belief about, and any disposition to act towards, the person wept for does not add up to grief. An argument against him must engage the functional claim, not the phenomenology. That is the correction on which the rest of this essay depends, and it disposes of the objection an earlier version of this essay rested its case on. Where the pressure on Walton actually lies Two lines of attack survive the correction. The first is Carroll's. He denies (2) by relocating the object of the emotion: what horrifies the viewer is not the monster, which does not exist, but the thought of the monster — a content entertained without being asserted. Entertaining a content is a perfectly ordinary mental act that requires no belief in its truth, and on Carroll's account the emotion of art-horror consists of that entertained content, an evaluation of it as threatening and impure, and the physical agitation that follows. If a mere unasserted thought can be the object of an emotion, premise (2) is false as stated and the triad dissolves without denying (1). The second line is Tamar Gendler's, and it matters here because it breaks a dichotomy Walton's argument quietly assumes. Gendler's "Alief and Belief" (2008) introduces the alief : an automatic, associatively activated state with representational, affective and behavioural components, which can persist in flat contradiction to what one believes. Her cases are non-fictional — the visitor to a glass-floored walkway over a canyon who believes the floor is safe and steps back anyway; the subject who will not drink from a brand-new, sterile bedpan. The alief explains the behaviour that belief cannot. Applied to the cinema, it yields a third category that Walton's argument does not consider: a state that is neither a belief-in