Wednesday, June 15, 2011

Paradox of Fiction

Michael Gilleland recently had a post quoting Macauley on reading Homer's Iliad while out walking. This section in particular caught my eye:

I never admired the old fellow so much, or was so strongly moved by him. What a privilege genius like his enjoys! I could not tear myself away. I read the last five books at a stretch during my walk to-day, and was at last forced to turn into a by-path, lest the parties of walkers should see me blubbering for imaginary beings, the creations of a ballad-maker who has been dead two thousand seven hundred years.

I've been intending for several months now to write a post on the paradox of fiction, and this reminded me of that. So here are a few thoughts thrown together; I may expand or improve on them in a later post.

The paradox of fiction is essentially this. We human beings read, watch, and listen to a lot of fiction. We know that it is fiction. But we have emotional responses and attachments to the characters. So, according to Colin Radford, who first put it forward, this shows that there's something incoherent in our emotional responses: we feel for things we know don't exist.

I find in general that there are people who think Radford was on to something and people who think he was just confused, and this can affect discussions quite a bit in places you wouldn't expect. I remember once arguing with an atheist who claimed that, since atheists don't believe God existed, it is just nonsense to claim (as some theists do) that atheists are in fact acting out on anger at God. If I recall correctly, I responded that, while atheists might not all be angry at God, his argument for this was a bad one: people have emotional responses, including anger, to fictional characters all the time. He would have none of it.

Nonetheless, it certainly is true that people fall in love with Mr. Darcy, or with Marian Halcombe, or get outraged at the treachery of a character in a Tyler Perry movie. So if you think our emotional responses, in order to be rational, have to be sensitive to our beliefs about whether their objects exist, then this causes Radford's problem. The simplest and most natural solution to this whole thing is to say that our emotional responses are (so to speak) prior to any beliefs we might have about the existence or nonexistence of their objects -- in effect, that we can have emotional responses to things we just think about, regardless of what we judge to be true about them. Thus we rule out the view that emotional response to things we think about but believe not to exist is irrational. Not only does this fit the phenomenology of fiction pretty well, it has the additional advantage of not requiring us to say that human interaction with fiction is thoroughly irrational, without committing us to all that much. This is known as thought theory; it's tricky to give a positive defense of it, and there are several varieties, but it's also the position with the most obvious general advantages. Some people, however, will have none of it.

One of the more outrageous positions among those who will not take the easy way out is pretend theory, which has a number of forms but is usually associated with Kendall Walton. Walton handles the paradox by denying that we actually do have any emotional responses to fiction. Women reading Emma don't pity Harriet Smith; they merely pretend to do so. Actually, fear is probably the easiest point at which to make the case for pretend theory, and the one that is usually used as an example: if you go to a horror movie, you know that you are not in danger. But surely, Walton will say, at least the risk of danger is a precondition for fear? So you only make-believe that you are afraid when the bogeyman jumps out.

Yes, but surely, you might say, we experience something when the suspense builds? Certainly, says Walton; we experience make-believe fear, which is not an emotion but a quasi-emotion.

The whole problem with pretend theory is that it only actually works, and only covers all of the evidence of human experience of fiction, if quasi-emotions are pretty much exactly the same as emotions, excepting the one distinctive feature that they occur in contexts in which we believe the object doesn't exist. Much of pretend theory ends up being an error theory, because people certainly think that they have emotional responses to movies and novels and the like, and in order for it to be such, quasi-emotions have to be easily confused with the corresponding emotions. So it does in fact seem that pretend theorists are committed to saying that we have make-believe emotions and real emotions, that the former are not really emotions at all but pretenses at emotions, but that in our experience we often (and perhaps always) can't distinguish the two except by checking to see whether we believe the object to exist or not. If this is all that it amounts to, the theory not only doesn't resolve the paradox of fiction, since it just verbally portions out our emotions into two groups, one of which it arbitrarily refuses to call emotions; it also introduces needless and useless complication in our discussion of human emotional life, forcing us to explain on different principles what can actually be explained on the same principles.

A third position is illusion theory. Illusion theorists say that we do have real emotional responses to fictional characters; but we are able to do so because we don't fully believe that they don't exist. That is, we have a partial belief (one is tempted to call it a quasi-belief, but illusion theorists think that it really is a sort of belief, just an incomplete or weak one -- which makes me think instead that the theory is unfortunately named). They often claim for themselves Coleridge's famous phrase, "willing suspension of disbelief". This is not a popular position, but I think philosophers generally don't do justice to it. It actually coheres fairly well with a common (albeit wrong) philosophical view that beliefs can be adequately analyzed in terms of 'credences'. Credences can get so great that we can say that, for practical purposes, you have no doubt about something, without necessarily implying that there is no credence on the other side; and if there is, we have no particular reason to think that these residual credences can't generate emotional responses. In this sense emotional response to fiction would be the result of the difficulty of obtaining perfect certainty. Sure, you don't believe there are vampires; but it's still the case that your eyes see them on the movie screen, and your ears hear people talking about them as if they existed, and it's not too hard to imagine that this might add a tiny little force to the vampires exist side of the balance -- too small ever to change your considered position about vampires, but just enough to allow for a tiny bit of trepidation, maybe a little anxiety, a little jumpiness. In any case, I don't think illusion theory is right, but I think, contrary to the usual view, that it is massively superior to pretend theory. It fits the evidence better than is usually admitted (I mostly reject it because of my views on the nature of belief). Think of how people sometimes have to remind themselves that it's all in their heads. And it coheres with other features of our emotional lives, as well. In Toronto there is a very noticeable landmark called the CN Tower. It is very, very tall (until 2007 it was both the world's tallest tower and the world's tallest freestanding structure). If you go up to the top, you can try to walk on a glass floor. The floor is marvellously engineered: the glass is very thick and it's supported by steel. The only way anything's falling through that glass floor is if the whole tower comes tumbling down. But it's remarkable how difficult it is to walk on a glass floor 342 meters (1122 feet) above the ground, no matter how much you are sure that the engineering behind it is good. Now, setting aside the paradox of fiction, consider the paradox of the CN Tower and ask yourself the following question: which is the more likely explanation, that you are pretending to be afraid of falling or that part of you, however small, really still believes that you might fall to the ground?

A fourth position is counterpart theory. Counterpart theories say that we do have emotional responses when faced with fictional characters, but that these responses aren't to the fictional characters themselves. Rather, our mind associates the fictional characters and events with real-life characters and events (counterparts), and we have emotional reponses to these. I actually think this is superior to pretend theory, too, since obviously we do sometimes associate fictional characters with real-life characters and events and have emotional responses to the latter. It's just extremely difficult to see how this could cover the whole panoply of emotional responses to fiction, and the account seems like it would inevitably be strained.

Of course, none of these views would help out my atheist interlocutor: on thought theory it's entirely possible to be angry at things you think don't exist; on pretend theory, you might not be angry, but you could still be quasi-angry, which (it turns out) is a lot like being angry; on counterpart theory, you may not be angry at the thing you think doesn't exist, but you could still be angry at something it reminds you of (one way really to annoy some atheists is to tell them that they are motivated by the fact that they have serious issues with father figures!); and on illusion theory, you can be an atheist and angry at God because you believe in him still a little bit (suggesting which is another way really to annoy an atheist, which I guarantee you is why some theists so gleefully suggest it). There's really no theory of emotions, or any that does justice to the paradox of fiction, anyway, on which you can be guaranteed not to get angry on thinking about something that you believe doesn't exist. Which, of course, is not to say that you necessarily would, either.

So you know that I prefer thought theory, followed by illusion theory, as the best account of emotional responses to fiction, like Macauley's "blubbering for imaginary beings". Which theory do you think best?


  1. Brendan Hodge11:49 AM

    This has actually been a topic somewhat on my mind lately, due both to recent reading and thinking about getting back to writing fiction.

    I think both of these fit in with your explanation of thought theory, but I'm curious as two how these two issues lay in all this:

    1) We (or at least I and some other people I've talked to about the topic) seem to naturally want to think and feel about what happens to fictional characters outside of the work of fiction itself.  For instance, I just finished re-reading Robertson Davies Mixture of Frailties which ends by stating that the main character has decided how she will respond to a marriage proposal she's received, but fails to tell the reader what she's decided.  I have my own idea as to what that response is, but I find myself caring about what that response really is.  Despite knowing that this is a fictional character who does not exist beyond the events described in Davies' narrative, I can't help caring about what the character does afterwards and wanting to know if I'm right.  Similarly, one of the things which really annoyed me reading Atonement was the way the author played with the notion that the novel was in part a narrative written by a character in the novel which intentionally falsified what had actually happened in the "real life" of the narrative -- but left it unclear where exactly this falsification picked up and thus what was or was not true.  This coyness on the author's part left me feeling cheated by the author, because he didn't reveal what "really" happened (or perhaps more so because he intentionally toyed with you on the topic, rather than giving you a fair set of evidence and letting you draw conclusions.)

    2) Sometimes we end up having very strong emotions indeed as a result of hearing something about real people we know, only later to discover that the happening we had been told of didn't actually happen.  However, unlike our reactions to fictional characters and events, we generally stop feeling these reactions when we find out that our understanding was incorrect, and indeed, sometimes swing in the opposite direction and feel the opposite of the emotions that we felt when deceived.  Yet, the reminder that fictional characters are fictional never seems to have this result.

  2. branemrys12:15 PM

    Both of these are very good issues; the sort of experience you note in (1) is the foundation of perhaps the strongest argument in favor of illusion theory, and  the sort you note in (2) is the foundation of perhaps the strongest argument against thought theory, so they are both very relevant to the problem. I'm not really sure what to do with (1), but to some degree I think it's tied to the role we naturally attribute to the author.

    With regard to (2), I suspect it has a great deal to do with the kind of emotional response one has: some emotional responses affect all those that come afterward. But if that's so, then the matter necessarily gets very complicated. I think it also is related to another (much older) paradox, the paradox of tragedy, which I hope to do a post on at some point this week.

  3. skholiast3:37 PM

    Why not a kind of counterfactual theory: our responses are to the scenarios as depicted, as if they were real-- not because we are pretending, but because we respond to them as hypotheticals? This is more or less Aristotle's account, yes? Not "what happend" (history) but t"he kind of thing that would happen" (poetry). I don't actually hold to this account, but I am not sure it is isomorphic with any of the others you present.

  4. branemrys4:27 PM

    Any of the theories are capable of accommodating a counterfactual version, so it would depend on details how it would fit into the scheme. What does it mean to response emotionally to a counterfactual hypothetical: Is it the mere thought that makes it possible to have the emotional response? Is it treating the counterfactual situation as if it were real that makes it possible? And so forth.

  5. I wonder if you've read Experiencing Narrative Worlds by Richard Gerrig? He takes a position that the fictionality of a narrative is not particularly important with regard to how we experience or understand what we are reading. Rather, we tend to engage in the same comprehension processes whether the information is from fiction or non-fiction sources.

    With regard to emotional (and other) responses, Gerrig follows Spinoza in suggesting that we first understand what we read and only afterwards sometimes appraise that information for accuracy. Thus, rather than a willing suspension of disbelief, we actually will tend to believe the information unless we willingly construct disbelief.

    If I understand correctly, this would be consistent with a thought theory.

    Psychologists, such as Gerrig, have found some evidence for this position in laboratory experiments. For instance, even when readers know a fact to be untrue (i.e. that mental illness is spread through viruses and you can catch schizophrenia if somebody breathes on you) readers have much difficulty rejecting these misconceptions after reading about them in narratives.

  6. branemrys2:01 PM

    I haven't, although it sounds interesting. From what you've said here, Gerrig's position sounds like it could end up being either  thought theory or an illusion theory, depending on precisely how one understood the 'willingly construct disbelief' part.

  7. pdorrell11:04 PM

    A more basic question, that perhaps needs to be answered first, is why do we read, listen to and watch so much fiction?

    After all, if fiction is false, then we are wasting our time perceiving and learning about things that are false.

    One possible answer is that we consume fiction, in order to practice our perception and understanding of reality.

    If this is the case, then we have to have emotional responses to our perception of fiction, because emotional responses are an important part of our perception of reality, and if emotional responses to fiction were excluded, then our responses to fiction would be incorrect, and therefore we would be practicing an incorrect response.

    It remains true that there is a danger that emotions about fiction may become confused with emotions about reality, given that emotions are a somewhat global aspect of our perception of current reality. This suggests that there is no ideal level of emotional response to fiction: too much response will lead to confusion with reality, too little emotional response will result in an invalid response to the fictional content. Whatever level of emotional response we do have to fiction is a compromise, and there may indeed be factors that influence to what degree this compromise leans in one direction or the other (i.e more or less emotional response to fiction).

  8. branemrys7:03 AM

    That's a very interesting argument that I haven't come across before.


Please understand that this weblog runs on a third-party comment system, not on Blogger's comment system. If you have come by way of a mobile device and can see this message, you may have landed on the Blogger comment page, or the third party commenting system has not yet completely loaded; your comments will only be shown on this page and not on the page most people will see, and it is much more likely that your comment will be missed.