Allan Hazlett is Reader in Philosophy at the University of Edinburgh, having worked previously for Texas Tech and Fordham Universities. He’s the author of A Luxury of the Understanding: On the Value of True Belief (Oxford University Press, 2013) and A Critical Introduction to Skepticism (Bloomsbury, forthcoming). His research interests include the value of accurate representation, the value and nature of intellectual virtue, authenticity, testimony, knowledge attributions, and the cognitive value of fiction.
Everyone, except for Plato, agrees that you can acquire knowledge through engagement with narrative fictions, including knowledge about things other than the form and content of the relevant fiction. And there seem to be several species of knowledge from fiction. Testimonial knowledge from fiction is acquired when a fiction makes a claim about the actual world, by exploiting (contextual and genre-relative) conversational assumptions about the similarity between the fictional world and the actual world (Gendler 2000, Hazlett and Mag Uidhir 2011), and you believe that claim. Imagination-based knowledge from fiction is acquired when you draw a (potentially morally or personally significant) modal conclusion – about what would happen in counterfactual circumstances, for example – on the basis of an episode of (fiction-prompted) imagination (Currie 1995, Gaut 2006). Practical and emotional knowledge from fiction is acquired when you gain (morally significant) abilities as a result of engagement with a fiction – for example, the ability to empathize with diverse others, or the disposition to appropriate emotions (Nussbaum 2001). I think there’s another species of knowledge from fiction; I’ll sketch my idea (which I’ll be presenting at the upcoming conference on Art and the Nature of Belief at the University of York) below the fold. [Warning: Brokeback Mountain spoiler ahead]
To make my proposal plausible, I need to make three big assumptions. The first is that sometimes, if not paradigmatically, we alieve the content of fictions that we consume. My notion of “alief” is roughly the same as that articulated by Tamar Gendler (2008); the introduction of this notion is motivated by the need to explain people’s emotions and behavior in certain cases, including cases of engagement with fictions. (However, unlike Gendler, I think alief is a species of imagination (though not the only species), so I don’t deny that we imagine the content of fictions that we consume.) Charles, while watching a movie about a terrible slime, shrieks and clutches his seat. What explains his shrieking and clutching, since he knows that the slime isn’t real? The answer that Gendler and I offer: Charles alieves that the slime is real, and thus alieves that he is in danger, and this alief is manifested in his shrieking and clutching. But alief is ubiquitous: it explains our reluctance to drink a glass of water labeled “poison,” even when we know that it’s just a glass of water; it explains our fear and trembling when suspended from a great height, even when we know that we’re safe. As Gendler puts it, alief is automatic (not formed intentionally), arational (not based on reasoning or argument), and affective (essentially emotional). Moreover, alief is a representational state: a representation of things as being a certain way. I assume, then, that truth is the correctness condition for alief, and thus that alief can be accurate or inaccurate. Charles’ alief that the slime is real is a case of inaccurate alief. But accurate alief is also possible – more on which below
Second, I assume that knowledge is a species of information possession. Crucially, I assume that we should not assume that the only way to possess information is in belief. We should be open to other ways of possessing information, and thus to instances of knowledge that do not involve belief – as in the case of an information-possessing machine that is incapable of belief (Williams 1976) or the case of a person who remembers certain information without any confidence in her memory (Radford 1966).
Third, I assume that some form of reliabilism about knowledge is true. Somewhat drastically, I assume that safe true belief is sufficient for knowledge. (My argument, I think, would not be affected if we adopted a more sophisticated form of reliabilism.) When this assumption is combined with the previous assumption, I think we must conclude that safe information possession is sufficient for knowledge.
Here’s a case to illustrate the idea. You have just arrived by train in an unfamiliar city. You know that there are some dangerous neighborhoods here, but you have also received detailed directions from your trustworthy host, which are meant to describe a route that avoids the bad parts of town. You follow the directions carefully. However, you find that your surroundings are making you increasingly uncomfortable. There’s something threatening and ominous about the look of things, and about the way people are looking at you. But you realize that you have been following the directions carefully, and find good grounds to dismiss your nervousness: abandoned buildings are common during this economic downturn, even in good neighborhoods; people are always suspicious of foreigners, which explains why people seem to be staring at you. But you cannot help feeling scared, you cannot entirely dismiss the thought that you are in danger. Although you are completely and justifiably confident that you are safe, you find yourself trembling, as though you think you are in danger. And you are in danger. Your host, although trustworthy, made a mistake when preparing your directions, which have lead you straight into the city’s notoriously violent slums. You abandon your false belief that you are safe only when attacked by a gang of thugs.
Question: at what point did you know that you were in danger? On my view, you knew that you were in danger at some point prior to your being attacked. You were aware of the danger, but this awareness had not yet risen to the level of belief. The notion of alief has intuitive application here: prior to being attacked, you alieved that you were in danger – which explains your trembling. On my view, this alief amounted to knowledge – because this alief was an instance of safe information possession. (The idea that your alief was safe is based on an assumption about the story – that your nervousness was a response to evidence that you were in danger, as opposed to (say) a nervous disposition that would have made you tremble even if you had been in an obviously safe neighborhood.)
Safe information possession is sufficient for knowledge. Putting this together with my first assumption (above), we can appreciate a novel way in which we can acquire knowledge from fiction. Here’s a case of the kind of knowledge-acquisition I have in mind. In the movie Brokeback Mountain (2005) there is a scene in which Ennis Del Mar (Heath Ledger) is told that Jack Twist (Jake Gyllenhall) was killed by an exploding tire. Ennis immediately imagines that Jack was in fact murdered – we are briefly shown images of a gang attacking Jack. Ennis, and now the viewer, cannot help thinking that Jack may have been murdered on account of his sexual orientation, but we are left without any evidence of this, apart from Ennis’ instinctive imaginative response. On my view, this sequence effectively expresses the idea that, at least in 20th century America, (actual) gay people were terrorized by the threat of anti-gay violence. Call this proposition T. Furthermore, on my view, T is something that you (if you didn’t know it already) can learn through engaging with the aforementioned scene.
How is this knowledge acquired, on my model? First, you alieve the explicit fictional content of the scene: that Ennis immediately imagines, upon hearing about Jack’s relatively violent death, that he was murdered for being gay. Second, you realize – in alief – that (fictional) Ennis is terrorized by the threat of anti-gay violence. Third, in virtue of this, you alieve the fictional analogue of T; call it T*: that (fictional) Ennis is a member of a (fictional) community of people are terrorized by the threat of anti-gay violence. Fourth, by alieving T*, you thereby alieve T. And that state of alief amounts to knowledge: it amounts to your knowing T.
The move from the third to the fourth step is crucial. When we think about about fictional characters and events we often, if not always, also think about actual people and events. On my view, we think about actual people and events by thinking about fictional characters and events. This explains why we can morally evaluate someone’s emotional responses to fiction (Gaut 1997, 2007, Hazlett 2009). Our emotional responses are responses to actual people and events, because when we think (e.g. form aliefs) about fictional people and events, we thereby think (e.g. form aliefs) about actual people and events.
If this is right, we can see how you comes to alieve T, as a result of engaging with the aforementioned scene. But why should we think of this alief as amounting to knowledge? Why, given my assumption (above), think that this alief is a case of safe information possession?
I assume that T is true. But I also assume that your alief, in the present case, is safe. Why think that? Two assumptions combine to make this thought plausible. First, I assume that this particular movie is a reliable source of information, such that the filmmakers responsible for it would not easily have included a scene expressing a false idea about actual people. That this premise is plausible depends both on facts about the movie’s generally realistic genre, on facts about the sincerity of the filmmakers, and on facts about their reliability of the relevant topics and questions. Second, I assume that you manifests certain competences, such that you would not easily have gone along with the movie’s claim if it had been false. Although we can think about real people and events by thinking about fictional people and events, we will resist doing so when we think that the relevant fiction, or the relevant part of the relevant fiction, is unrealistic (Hazlett and Mag Uidhir 2011). Suppose the filmmakers behind Brokeback Mountain had depicted Ennis imagining that Jack was killed by marauding aliens, in such a way that it was clear that, in the world of the fiction, gay people are terrorized by the threat of alien attack. The sensible viewer would not alieve that actual gay people are terrorized by the threat of alien attack, as a result of engaging with this scene. The scene’s unrealistic content would undermined any putative connection between the fictional world and the actual world, at least with respect to this scene. So in saying that, when you come to alieve T, you come to know T, I assume that you are a sensible viewer, in this sense.
I have sketched an argument that there are instances of alief that amount to knowledge from fiction. If this argument is sound, then I think there are two exciting payoffs. First, we have a way of seeing not only how the imagination can assist the intellect in acquiring knowledge, but how the imagination itself is capable of knowledge. (This is based on the controversial idea mentioned above: that aliefs are states of the imagination.) Second, we have a way of explaining the essential role of emotion in knowledge from fiction. Testimonial knowledge from fiction and imagination-based knowledge from fiction do not essentially involve emotion, but because alief is an essentially emotional state, alief that amounts to knowledge from fiction will be an essentially emotional state. We shall be able to explain the sense in which (even propositional) knowledge from fiction is essentially different from the kind of knowledge we acquire in philosophy or in the empirical sciences.