I read all of Kendall Walton’s Mimesis as Make-Believe betwixt and between sunny sessions on Thai sand. Overall verdict: quite interesting but ultimately unconvincing.
Let’s take up where I left off in previous posts. The nature and proper definition of ‘fiction’. Analytic philosophy stuff, not very flashy. Anyone who wants to help me out is welcome, but probably you will not be entertained.
And really this is just me recording impressions, trying to express them before I forget them; so everything should be qualified with 'I'm really not sure, but it seems to me that ...' But that's dead boring to keep repeating every other sentence. So I sound more confident than I really am just to punch it up a little.
§1 Walton’s Thesis: Fiction As Props in Games of Make-Believe
Walton’s central notion is that of ‘representation’ but he uses the term almost interchangeably with ‘fiction’. His thesis is that works are fictions if and only if their function is to serve as 'props in games of make-believe'. Something is a prop in a game of make-believe if and only if its function is to 'prescribe imaginings'. More precision and elucidation needed – not to mention defense - but we readily intuit more or less where Walton is coming from.
The intuition behind the proposed analysis and definition is genealogical. “In order to understand paintings, plays, films, and novels, we must look first at dolls, hobbyhorses, toy trucks and teddy bears. The activities in which representational works of art are embedded and which give them their point are best seen as continuous with children’s games of make-believe" (p. 11). So sitting down to read War And Peace is like lying on the rug, playing with toy soldiers and dollies. Toy soldiers and dollies prompt imaginings of real soldiers and real families. This is their function. Likewise, sentences in the novel prompt imaginings of real soldiers and real families. This is their function. Not that there aren’t differences. But there is, Walton maintains, essential continuity. Reading novels is an activity crucially identical in kind with playing make-believe. Somehow this is going to be the key that unlocks many puzzles.
§2 Conceptual Analysis or Conceptual Revision?
At points Walton suggests that his account – his definition of ‘fiction’ – is mildly but not terribly revisionary of our ordinary notions of what is, and is not, fiction. I think this is mistaken, and the mistake leads at the very least to obscurity in presentation. Walton adduces examples of fiction and non-fiction in support of his account – relying on our ordinary, classificatory intuitions – without considering whether these very examples may not stand in need of reclassification if his account is accepted. It seems to me our ordinary concept of fiction is likely to stand in need of substantial revision as a precondition of its being amenable to interesting theoretical treatment. So I don't object to revision, per se. But it should be clearly marked as such to avoid confusion.
§3 Our Ordinary Notion of Fiction
Let me lay out the elements of what I take to be our ordinary, everyday notion of fiction. I make no claims for the timelessness or universality of this concept. (I'm sure it is neither.) I merely catalogue what I take to be standard English usage at the present time. The point of noting how we in fact do use the term ‘fiction’ is not to lock us into this usage but just to clarify the degree of Walton's departure from ordinary usage.
I think ordinarily we look for four features. Anything clearly exhibiting all four will be easily, confidently and securely classed as ‘a work of fiction’. Anything lacking even one will be, at best, a borderline case. Probably most three-out-of-four cases fall into the 'yep, it's a work of fiction but it seems weird to call it that' category. Anything lacking two or more of the following features probably does not qualify as a work of fiction (though I would not want to insist on that as any sort of iron law.)
1. Fictions are works of intentional, undeceptive untruth.
We don’t call it ‘fiction’ if the author believes it all really happened; nor if the author hopes to trick us into believing it all really happened; nor if it all really happened. Works of fictions are not records of delusions, nor lies, nor essentially reliable reports of data about the actual world around us.
All three conditions are ticklish, needing qualification; the third very much so. Few works of fiction contain no true sentences; few works of non-fiction contain no false ones. So it seems we are going to be looking at a true/false ratio, when we were vaguely hoping for a cleaner, binary relation. Furthermore, propositions are the standard candidates for truth and falsehood; whereas works – containing many propositions, or perhaps none (if they are pictorial) – are the standard candidates for being fiction or non-fiction. So what does ‘work of … untruth’ mean? A few lies and delusions (what artist can resist?) and some incidental factual reportage around the edges, would seem to be quite consistent with the commission of an act of fiction – even of pure fiction. But how exactly is that possible?
[UPDATE: Somehow I forgot to mention all the well-known, well-worn puzzles about truth in fiction. It's true in the story Sherlock is a detective, etc. So it's no good just saying that sentences/propositions that are true in the story are false, simpliciter. Something more complicated has to be devised. Lots of literature on this.]
Obviously the tricky cases are going to be things like historical novels, which may be very hard to distinguish from histories written with literary flair. Biography and autobiography will also skate close to the line, causing us to scratch our poor heads.
Since I cannot sort this briefly, I will not sort it. I take it to be obvious that we do in fact operate with some such standard, even if there are problems working out what it means to say 'works of fiction should be intentional, undeceptive untruths.' We do in fact sort delusion, from lies, from factual reportage, from fiction. Or at least we think this is what we do. Any account of our ordinary notion of fiction needs to acknowledge that we at least try to do this.
2. Fictions tell stories.
Works of fiction are, paradigmatically, narrations of stories. (In a somewhat more strained sense, stories themselves – e.g. the story of Little Red Riding Hood, which may be conveyed by all manner of narrative vehicles – are fictions but not works of fiction.) By contrast, a painting, joke, poem, philosophical dialogue, diagram, forged document, map, tableau, diorama, sculpture, theatrical set, prop, dance, song, individual sentence, may intentionally and undeceptively represent – or imply - that which never was. But we begrudge the likes of these the title ‘work of fiction’ unless we detect not just characters but actions and events in a temporal order. Not to put too fine a point on it, there should be a beginning, a middle, and an end (not necessarily in that order.) So a picture of a face, even of a non-existent subject, is a poor candidate ‘work of fiction’. A slightly less bad candidate would be an illustration accompanying the narration of a story. But I think we are not much inclined to call illustrations included with works of fiction, or gracing their covers, ‘works of fiction’ in their own right. A yet less bad candidate would be an image representing a scene from which the viewer could infer the occurrence of a series of actions – man on ground, knife in chest, horrified woman and children standing by, blood-spattered fellow frantically attempting to get out the window, police pounding on door. Some pictures tell a story. One-frame “New Yorker”-style comics often do this, but often not. Two people, holding drinks, saying something funny, is not much story. Three-frame comics, such as appear in the newspaper, are typically more story-like. Serial story strips – e.g. “Prince Valiant” – are clear cases (although they lack endings.) Strips whose characters age – e.g. “Doonesbury” – are more intuitively story-like than strips whose characters stay the same age. We perceive a sort of over-the-years story-arc in “Doonesbury” that is absent from “Calvin & Hobbes”. On the other hand, “Calvin & Hobbes” more often tells a little one-strip story, as opposed to just telling a joke. Nor is it just sequential visual art that often teeters on the cusp of storyhood like this. Tove Jansson has a Moomintroll novel (anyone remember which one?) in which one character tells a ‘story’ that runs, in its entirety: “There was a water-rat named Poot.” That's not a story. Philosophical dialogues tend to be most infirm in the narrative department. ‘Socrates met someone and they talked philosophy.’ Poems are often not very rich, story-wise, however rich otherwise.
The general point is: jokes, poems, philosophical dialogues, narrative essays, short stretches of sequential visual art, so forth, tend to be only uncomfortably classifiable as works of fiction. We just don’t call them that (even if we are prepared to admit, somewhat head-scratchingly, that they are that.) And our willingness tracks, precisely, the degree of story-ness we detect. The more paradigmatically story-like the representational content of a work, the more comfortable we are classing it as a work of fiction.
One exception: experimental literary works that strive to undermine narrative and story conventions – i.e. that willfully lack beginnings, middles and ends, characters, so forth – are usually quite easily classified as fiction (unless they seem to be turning into poems.) I take this ease of classification to be the result of a sort of grandfather-clause. Or maybe an Oedipus-like father-clause: if you are trying so hard to overthrow story, you must be story, ergo fiction. (A bit of a puzzle.)
3. Fiction is complex. [maybe complex isn't the right word? artful? crafted?]
‘Fiction’ comes from the Latin, ficti: ‘the act of shaping, a feigning, that which is feigned.’ It might be thought that the bit about feigning is what we have chiefly retained, domesticating it (for entertainment purposes) by breeding out that anti-social strain of intentional deception. But, on reflection, I think we have also retained our sense that fiction should, by rights, be decently complex, should bear the hallmarks of having been wrought, crafted. A fiction should be elaborate and artful, or at least not too pathetically brief and simple. Hence our intuition that a very complex, non-narrative deception or lie – e.g. a spy’s scrupulously maintained cover, a con man’s elaborate set-up – may be counted as ‘a fiction’, even though we ordinarily insist on the story-form and exclude fraud, as per above. (Possibly ‘legal fictions’ can be shoehorned in here, too.) By contrast, a thoroughly artless lie – ‘I didn’t do it’, when really I did – will not be awarded the palm of ‘fiction’. (And would you consider the following to be a work of fiction: ‘Smith was born, grew up, worked for a living, died of old age.’ It’s got a main character, a beginning, a middle and an end. It’s a story. But it needs shaping, I should think.)
The requirement that fiction be complex is vague. But it may shed some incidental light on the first requirement. As per 1, we do not ordinarily deem stand-alone propositions (or sentences) to be suitable candidates for being fiction or non-fiction. Works are assessed as a whole to determine whether they are fiction or non-fiction. It may be that we are hereby insisting on a certain overall complexity - holistic artfulness in shaping. (Not that sentences cannot be complex and artful, so perhaps I am just speculating.)
4. Works of fiction function as props in games of fiction appreciation.
Obviously this is circular if we are trying to define what makes for fiction. But let me try to salvage something. (Really I am just working my way back to Walton, whose ‘props in games of make-believe’ is in the vicinity. More on that momentarily.)
It is tempting for a number of reasons to define ‘fiction’ at least in part in terms of what audiences do (or should do) with the stuff. The trouble is that there is disagreement about what audiences do (and should do) with the stuff. You wade into these disputes and never make it out. Aesthetes swear up and down that fiction should pleasure us. A more moralizing breed of critic denounces this is callow and shallow. Fiction serves to express eternal verities about life and the human condition – in short, it serves to educate us morally. There are far more than just these two views, and I am stating them about as baldly and lamely as possible. But they will do, for illustrative purposes. We could try to finesse the need to settle disputes between the likes of aesthetes and moralists by saying something is a work of fiction if aesthetes will regard it as a prospective source of pleasure; and if moralists will regard it as a prospective source of moral education. And so forth for other views about what the stuff is for. So this requirement ends up saying, roughly: fiction is that which is exquisitely diplomatical enough to promise to give everyone what they think they ought to get from fiction.
There are difficulties with this oblique strategy, not the least of which being that it will be turn baroque quick. And it is, as initially suspected, at least semi-circular. It almost comes down to saying: fiction is what everyone thinks it is. True, but unilluminating. Nevertheless, there may be some content to the notion that fiction is whatever manages to be acknowledged by everyone (or nearly everyone) as fiction. The content comes in when we actually look at everyone’s wish-lists and sets of demands – the habitual uses to which everyone collectively put the stuff, the notions with which they burden it, the work they insist it do, so forth. These wishes and demands may be articulable without circular reference to fiction, and so we may learn something. To be an artifact that can be both flexible and stiff enough to serve all ends – pleasure, moral improvement, etc. – may be a non-trivial achievement.
And certainly it is true that, to some extent, we really do decide whether something is a work of fiction by looking and seeing whether everyone else says so, and then saying: ‘what they said’.
Authorial intentions may be quite determinative in this regard. A work that satisfied conditions 1-3 but whose author did not intend the thing to function as a prop in a game of fiction appreciation would be, at best, a borderline case of a work of fiction. Are there any actual cases in which works can be seen being pushed in or out of the fiction box, intuitively, through an application of this requirement? Perhaps the difference between two well-known journalistic frauds, Stephen Glass and Jason Blair, will serve. Blair was a liar. He become a liar, apparently, through a mix of laziness, instability, and temperamental unsuitability for the profession. Glass was a liar, too. He became one, apparently, because he was an obscurely compulsive fabulist – as per the title of his first novel. Glass was driven to produce verbal artifacts that function admirably as props in games of fiction appreciation. He had the knack for giving pleasure, the knack for making an audience feel morally instructed (and so on, all down the list of things you might dream of demanding from fiction.) Blair just wrote shoddy stuff for the paper.
Clearly neither Blair’s fraudulent work, nor Glass’s, will ever be offered up as a perfect paradigm of fictionality. Both writers are intentionally deceptive untruth-tellers, hence in violation of requirement 1. But we may be inclined to class Glass’s works, not Blair’s, as borderline works of fiction, in virtue of their status as exemplary props for the game of appreciation, even if they were not originally advertised as such. (This requirement dovetails with the requirement for ‘shaped-ness’, as per 3, but I will not attempt to fill this in. Glass shapes his frauds in a way that Blair does not.) Glass’s works seem so admirably engineered to trick the audience into reading them, unawares, as works of short fiction. So maybe Glass really was a (borderline) fiction author even before he turned to novel-writing.
Please note that this is not purely a point about authorial intention (though it is not purely not a point about authorial intention either.) Glass’s intentions are quite unclear, probably even to himself. What is clear is that he produced a body of work with exemplary functionality along a certain axis – the fiction appreciation axis. These are props you can lean on with confidence, unless you are looking for solid journalism. (And obviously if this thumbnail sketch of the differences between the cases and characters of Glass and Blair is disputed, the example should be regarded as hypothetical.)
Quite inadequate, I admit. Which carries us back to Walton.
§4 Where Is Walton With Respect To This?
Walton’s definition of ‘fiction’ amounts to a radical simplification of the foregoing four-fold order of our ordinary notion of fictionality. Walton simply drops 1-3 and drastically refines 4.
Running down the list:
1. Fictions are works of intentional, undeceptive untruth.
Walton demands nothing of the sort. He is happy to class delusions, lies and accurate factual reportage as fiction, at least potentially. “Could an author be claiming truth for every sentence he writes and still be writing fiction? I see no reason why not, why there couldn’t be a genre of historical novel in which authors are allowed no liberties with the facts and in which they are understood to be asserting as fact whatever they write” (p. 79). What is Walton’s motivation for departing from ordinary usage in this way? He says surprisingly little on behalf of what seems to me a substantial act of conceptual revisionism. But it is pretty clear he regards requirement 1 as analytically unworkable as it stands, and unrefinable into anything much better. Basically, the notions of ‘true’ and ‘false’ that need to be made to do work won’t, in Walton’s considered opinion. So he is placing his chip on a different number – number 4, to be exact.
2. Fictions tell stories.
Walton does not demand, or even mildly request, that works of fiction tell stories. He has no problem saying one-frame comics, poems, paintings, sculptures and a great many other things we would not ordinarily call ‘works of fiction’ – because they don’t tell stories – are perfect paradigms of fictionality. Walton says even less about dropping the story requirement than he does about dropping the intentional, undeceptive untruth requirement. But it is possible to guess what leads him not to build storytelling into his account: our ordinary insistence on representations of characters and events through time – beginning, middle and end – can easily seem an arbitrary feature of our concept of fiction, hence analytically unpromising. Any analysis of fiction that accommodates insistence on storytelling can seem ad hoc, then.
Why would this be so? Think about a painting of a knight on horseback, a sculpture of a knight on horseback, a little plastic knight on horseback for the kids to play with, a description of a knight on horseback in a novel, a poem about a knight on horseback. For Walton the thread running through all this is not any story about a knight on horseback but just prescriptions to imagine a knight on horseback. In other words, Walton thinks he sees an analytically fruitful grouping of objects – of ‘representations’, as he calls them – which only very imperfectly overlaps the set of things that tell stories (hence the set of things we are ordinarily inclined to call ‘fictions’.) For the sake of the analytic fruit he thinks he sees, Walton is prepared to equate ‘representation’ with ‘fiction’, which means dropping the story requirement.
I think a serious slip is hereby made. It has to do not with the fact that ordinary usage of ‘fiction’ is hereby flouted (that’s all right if you have your reasons) but with the fact that make-believe – Walton’s official focus – is not just a matter of imagination but also of role-play; and role-play is storytelling. So Walton ends up not talking about something big and important. We will come back to this.
3. Fiction is complex.
Again, Walton says nothing about why he does not insist on this. Here the silence is understandable, as this requirement is less obviously an essential component of our ordinary notion (I think it really is an essential component.) It is also analytically unpromising. It does not promise to shed light on what sort of complexity makes for fictions. Obviously not just any sort will do. And it is inherently vague. On we go.
4. Works of fiction function as props in games of fiction appreciation.
Here Walton places his bet. According to him fictions function as props in games of make-believe. This is his definition of ‘fiction’. Now clearly this is an improvement on – or, at any rate, a less circular specification of – my more generic formulation. With ‘make-believe’, Walton is explicitly committing himself to a specific answer to the question: wherein consists the activity of appreciating fiction? He is saying it consists in audiences making-believe, imagining things as prompted by props. This clearly avoids the triviality with which my formulation is at least somewhat threatened. On the other hand, in philosophy when your steer clear of triviality you usually say something wrong. Is Walton right that playing the game of fiction appreciation in some deep sense consists in making-believe?
I think not, but before making the negative case let me say something positive. There is something shrewdly diplomatic about Walton’s position, and that is a good thing on its face. Take my two sample quarreling critics: the aesthete and the moralist. They cannot, it seems, be brought to agree about what one is supposed to do with fiction – what the stuff is for. Is it for pleasure, or for moral education? But once upon a time these two were children, down on the rug with toy soldiers and dolls and so forth. As adults, they cannot see eye to eye about War and Peace, but as children they played together happily enough. The strategic promise of Walton’s account, then, is to finesse this childish unity into something that bears on adult activities, without having to settle all the tedious arguments adults have. (Not obviously a good plan, but at least it's a plan.)
§5 Brief Overview of My Problems With Walton’s Thesis
This will probably be pretty telegraphic. Let me just try to get it down
First, I think Walton overemphasizes the degree to which make-believe can be construed as a matter of being obedient to prescriptions to imagine. At the very least he hereby mistakes a part – an inessential part – for the whole.
To take one of Walton’s own examples: if you are making-believe that every stump is a bear, the way to play is to imagine a bear whenever you see a stump. The trouble is that this conflates two different games. First, there could be a game in which the rule is that, when you see a stump, you are supposed to form a mental image of a bear standing where the stump is (probably mentally erasing the stump in the process.) Second, there could be a game in which the rule is that when you see a stump, you are supposed to suppose there is bear there. And behave accordingly. (Exit, pursued by bear. Much shrieking had by all.) In the second case, there is no special obligation to form an actual mental image of any bear, although you might find one handy as an internal prop, for purposes of inhabiting your role more thrillingly and dramatically.
The second game is the one children play; but Walton consistently talks as though make-believe were the first game. But it isn’t. For that matter, no one actually plays the first game, which isn’t really a game, more a recipe for private phenomenological activity of a highly particular sort. (Suppose I made a private resolution to conjure as vivid a mental image of Christ as I could, whenever I witnessed someone sinning. To do so would not be to play a game of make-believe, I take it. Certainly it would be an extremely unusual resolution to make.)
To imagine is to spectate, after a fashion. To make-believe is to play– to role-play. These stances are not mutually exclusive but neither are they mutually entailing. Walton does not deny this but neither does he affirm it. He consistently hints at an account of the nature of games of make-believe while speaking purely in terms of prescriptions to imagine, which is like hinting that it is possible to explain what it is to do in terms of what it is to watch, which does not really make sense.
My second objection flows from the first. Walton wants to explain fiction in terms of make-believe. This is a promising enough gambit. But since Walton has an unduly imagination-based conception of make-believe, he ends up with an unduly imagination-based account of fiction, among other problems.
Let me offer one example of how Walton goes wrong - not a terribly important example but symptomatic, so it seems to me. Both my objections score here. "Appreciating paintings and novels is largely a matter of playing games of make-believe with them of the sort it is their function to be props in. But sometimes we are interested in the props themselves, apart from any particular game" (p. 53). This is wrong with regard to painting and probably wrong with regard to novels. Paintings first: if it were the case that paintings were usually just props to stimulate imaginings of their subject matter, then it would be not too difficult for me to produce a painting that does the essential artistic work of (to pick one of Walton's own frequent examples) George Seurat's "A Sunday Afternoon on the Island of La Grande Jatte". I am just good enough at drawing and painting to make it decently clear to the viewer that this is supposed to be a woman, that a man, an umbrella, a dog, boats in the background, so forth. And really that is all that is needed to get the viewer imagining the proper thing, if that is what you want. The reason no one is going to accept my poor substitute for Seurat is that, frankly, the prop's the thing - not sometimes, always (except in vanishingly few cases.) But really this is just the thin edge of the wedge. It is wrong to imply that the main artistic function of paintings is to get us to imagine their represented subjects. (Not that apprehending the subjects may not be necessary. I doubt you could properly appreciate Seurat's painting if you couldn't tell that it was a representation of human beings in a park. But apprehending that it is a representation of human beings in a park is a means to a further artistic end - one that I will not at the present time venture to specify.) But it is even wronger to imply that it is at all common, let alone standard practice, to play games of make-believe with the likes of Seurat's painting. Seurat is actually the exception that proves the rule, because Stephen Sondheim and James Lapine have gone and written a whole musical - that is, a story - based on his painting. And, plausibly, we identify with the character of George when we watch "Sunday in the Park with George". And, plausibly, that leads to a sort of role-play, i.e. make-believe. But this whole development is so wildly non-standard and out of the ordinary that it only serve to underscore - emphatically - the extent to which we do not play games of make-believe with every picture in the gallery. We just don't make such a beginning, middle and an end production out of it, mostly. We don't make up stories. We don't wonder what happened before, or happened next. In vanishingly few cases do we make up a story and then, as children will imagine that they themselves are driving the little truck as they vroom it around on the rug, insert ourselves into it, i.e. strongly personally identify with the characters: make-believe.
Now with regard to novels there is certainly at least a little something to the idea that, in at least some cases, there is at least something inessential about the prop, i.e. all the words - because the story is the thing, and any given story can be narrated any number of different ways. But indifference to the specifics of the narrative, i.e. all the words, is really not a standard attitude. You don't have to be much of a critic to care how a given story is told, i.e. to care about the prop itself, as well as what the prop makes us imagine (if it does make us imagine.) More generally, what Walton's slip with regard to painting has done is run together two questions regarding novels, to which the answers are not at all self-evident, and running them together does not help. First, to what extent do novels function by making us imagine what happens in them? Second, to what extent are novels occasions for make-believe, i.e. to what degree does fiction function to draw us into feeling that we ourselves are part of the action?
Lurking underneath all this is the rather stubborn fact that Walton and I are not at all agreeable about the usage of terms. Specifically, we are using 'imagination' and 'make-believe' and a couple other terms differently. Walton would probably say that I appear to be perversely determined to construe these terms very narrowly, which makes his claims come out sounding like nonsense. But he did not intend the terms so narrowly, so all my objections are nonsense. For my part I would have to reply that the narrow senses are the ones that make sense to me. But I'll leave it at that for tonight. Maybe tomorrow I'll take up at this point. Or not. Walton's book really is quite interesting, I find. It's turning out to be one of those 'disagree with everything but thereby figure out what you really think' type books.