Richard Rorty was a philosophical provocateur. His provocations often took the form of diagnostic analyses of philosophical diseases in which he determined the causes of the illness and offered a cure. The cure was often the philosophical equivalent of “Take two aspirin and call me in the morning.” He offered, of course, his own special brand of aspirin—The Pragmatist Analgesic For Ailments You Don’t Really Have But Think You Do.
Rorty provides a pointed diagnosis of the dominant analytic theories of fiction in his essay “Is there a Problem about Fictional Discourse?”.[1] He argues that the motive for these theories of fiction all rest on two mistakes. First, they all attempt to solve an epistemological challenge (how to verify or justify a proposition) by semantic means: they attempt to guarantee that our statements are not governed simply by conventions, but by how things are in the world. Second, they all accept Russell’s proposition that “What is referred to must exist” (111). Both positions arise out of a more fundamental philosophical picture in which statements, propositions, or thoughts must “hook on to the world.” Russell’s proposition, for example, seems to provide the glue between propositions (or statements or thoughts) and the world. Rorty argues against both of these positions, as well as against the general goal of trying to get language to hook onto the world. As a consequence, he finds most philosophy of language to be both moribund and unnecessary. He also concludes that without these mistaken assumptions about reference, epistemology, and the relation between language and world, there would be no interesting problems for theories of fiction to solve (114). While this latter claim is true for most of theories of fiction in analytic philosophy, as well as those in literary studies, I do not think it is true more generally. Our need to conceptualize fiction is not simply the consequence of mistaken theories about how we know or talk about the world of objects and things. Fictions remain puzzling even without elaborate philosophical theories.
If one accepts Russell’s proposition that “What is referred to must exist,” then fiction becomes a problem. Since fictional characters seem not to exist, if we accept Russell’s premise, we cannot refer to them, and yet we do refer to them. So, how can we do this? Either fiction means in some special way not requiring such referential targeting or fictional characters exist in some special way such that we can refer to them. Or, of course, we can reject Russell’s premise. Rorty argues for this last solution.
I am sympathetic to Rorty’s solution, but my goal in this essay is neither to rehearse nor evaluate Rorty’s arguments against theories of fiction. My goal is to situate fictions within a broader category of logical distinctions, and in so doing reveal something of the oddity of fictions. What I will provide are not theories, but descriptions with supporting arguments (often arguments of demonstration, generated from examples): a philosophical net to catch fictional fish.
I begin with the essential question: What does it mean to be fictional, which is to ask, what is a minimal logical description of what counts as a fiction? We can describe the logical form of fiction in the following way: a fiction is a story or sentence or whatever from which we should not infer in any straightforward way anything about the author or the world by virtue of what is said, asserted, or claimed. It may express the beliefs of the author, but we should not conclude from the fiction that it does in the way we could if someone said I believe in leprechauns. Nor should we infer something of current or past states of affairs from a fiction in the way we could if someone said Gaius Gracchus’ son committed suicide when he learned that his father had been killed. This might be false, but it is not fictional by virtue of being false. A fiction can in this sense be neither true nor false. By utilizing other information and setting the fiction in context, we can use it to help confirm or trace aspects of a cultural or historical situation. But this is to use the fiction, not to read it as a fiction. In order to use it to make such confirmations or inferences, we must normalize aspects of the fiction relative to our understanding of evidence and relative to the logic and constraints of non-fictional statements and stories. A non-fictional statement need not be true or undistorted, but it can be evaluated relative to what is the case, and thus can be false, even if only potentially. Thus, it is necessary, when we understand a fiction, that we understand that certain inferences to the author and the world are not permissible. Consequently, what is necessary for a fiction to be a fiction—that straightforward inferences to author and world are blocked—is also sufficient for it to be a fiction.
If this last is true, then fictionality is a logical distinction, regardless of how we decide if something is a fiction and regardless of its content. Fiction is an example of the conceptual resources that allow us to talk about anything. How we evaluate and use fictions is determined by a number of psychological and cultural factors. That we can make and grasp fictions is not dependent on psychology or culture, except in that someone must have a certain minimal cognitive and probably linguistic capacity in order to grasp the distinction between fictions and non-fictions.
I am situating fiction as a logical category amidst other logical distinctions, all of which do not depend on psychology or culture (except contingently). I will describe a few of these logical distinctions in order to sketch the landscape in which fiction exists and thrives. My first example requires some sleight of hand. It is a kind of magic trick in which I turn a pencil into a concept. The trick is simple. I raise the pencil in my hand and tell my class that they must use this on their test. I do not expect my students to all rush desperately towards me to grab the pencil out of my hand, so that they can use that particular pencil. They understand that I mean the kind of thing called a pencil. I turned the pencil into a type, into the concept pencil: I translated the physical pencil into a logical pencil. It is as if I am putting single quotes around the pencil, turning it into a kind of symbol of the pencil I want my students to use. By turning the pencil into an example of the pencil I want my students to use, I am doing something akin to mentioning a set of words, and thus I am framing the pencil-as-kind as logically distinct from the actual pencil I am holding in my hand. This is like saying “I want you to use the words ‘a person is a person and not a chair’ in your essay.” I am mentioning this sentence and not using it. Mentioning words alters their logical form or status, and thus alters our relation to them in a way that is akin to our logical relationship to fictions and to fictional characters. In my example, I am mentioning the pencil (by demonstration), abstracting the physical object, which is still there, into the kind of thing I want my students to use.
The relationship between the pencil I hold and the logical pencil I refer to is like the relationship between a stop sign and the laws about stop signs. The sign tells us that we should abide by the law, governing our behavior relative to such signs. But the traffic law has no physical relationship with the sign. Even if we put a statute book next to the sign, the law itself, which is not the words by which it is expressed, is not at that moment close to that one sign and farther away from others. If it is a law, its domain is the domain of law—and “domain” in this case characterizes the scope of its legitimate application. And while that scope includes the physical space that includes all stop signs in the jurisdiction where the law is in force, the law remains neither close nor far from any stop sign in that domain.
Asking how far or close a law is to the stop sign is a nonsensical statement that is akin to (but not the same as) asking how statements hook onto the things in the physical world. The dualism between laws and signs is benign since it need not imply two kinds of substances. In the case of traffic laws, we can note that laws in general do not need to interact with or hook onto signs since they govern (or provide the normative rules for) our behavior with these signs.
Again, to imagine that there is a distance between laws and stop signs is not to say something false, it is to say something nonsensical. This is a characteristic of a logical confusion. The joke on which my pencil magic trick turns is that it would be nonsense to imagine that I am saying the test can only be taken using the one pencil that I happen to have.
We are the link between the stop sign and traffic laws, just as we are between the pencil and the concept of pencil. To understand how these objects relate to our concepts, laws, ideas and so on is just to account for our conceptual abilities tout court. The distinctions I am highlighting here are mundane ways of pointing to what Sellars calls the logical space of reasons. I want to stick as close to the mundane as I can. In these cases, the distinctions, it needs to be repeated, are logical. We have ways of talking about stop signs that are different from ways of talking about traffic laws, and these differences are partly founded on the fact that these are different kinds of things. What it means for them to be different kinds of things is simply that we know (or should) that we cannot measure the number of atoms in any specific traffic law, although we could of a stop sign. Similarly, we know that we cannot reduce the law to the marks on the page by which it is articulated, even if to understand the law we interpret the words made from those marks. Certain things are made of atoms, but other things are not. When we speak of kinds we are articulating differences from within and by means of our ability to speak, to represent and to conceptualize.
Language remains the most powerful matrix through which we conceptualize things. And it provides, of course, the primary tools for making and understanding fictions, as well as for articulating things like traffic laws. The sentence you are reading now both is something and means something. The previous sentence is a sentence, but it is also a string of words, a string of letters, marks on a page, something to read or vocalize; it is meaningful, self-referential, part of English, something I wrote, and so on. Some of what the sentence is is tied to how we understand it relative to particular aspects and descriptions of it—descriptions highlighting that it is in English or that I wrote it, for example. Frege argues that a sentence means just the thought that it expresses, which suggests that a sentence just is the thought that it expresses. That there is some relation between the thought a sentence expresses and the logical form of the sentence suggests, even if one rejects Frege’s theory of meaning, that the attempt to separate what a sentence is from what it means will not succeed. To individuate a sentence is in many ways to say what it means.
A sentence is not individuated simply by its being a certain set of marks or vocalizations. We might think that some nonsensical string of words is a sentence syntactically yet still not count it as a meaningful sentence. We can reasonably claim that a sentence that means nothing is not a sentence, or that nonsense sentences just look like sentences but that they are in fact factitious. Such nonsense phrases, which we mistakenly call sentences, can provide the resources and the constraints for us to rewrite them into some kind of sense. This is often what we do when we interpret poems, for example.
A number of conclusions follow from these descriptions. First, sentences are not physical objects, individuated causally; they are individuated relative to their meanings. Thus, sentences are strange objects—and those linguistic forms, like poems and fictions, which are constituted by them, are likewise strange. One way in which they are strange is that while we must grasp them as they are manifest to us (by sounds or in marks), we do not stand toward them as we do toward other objects. We are, it is tempting to say, inside the meaning of sentences, because their meaning is something we grasp from the inside out—from within the possibilities of meaningfulness that constitute our involvement with language (including our ability to speak and understand one). This brings me close to Rorty’s holism about language. It suggests, as Rorty would agree, attitudes towards fictions are reflections of our understanding of what it means to ask about how sentences relate to that which they speak about.
Another example will help me re-focus our involvement with language towards the oddity of fictions. I can do to myself what I did to the pen in my earlier example. Imagine that I am playing myself playing myself in a fictional play. What I am doing in the play is what I am doing in playing myself: I am pretending to pretend. My playing myself, however, is still different than my playing myself in the play. The Brett I am is different than the Brett I am playing. There is a logical difference, just like that between the pencil I am holding and the concept pencil I indicate by saying “Use this on your test.” An actor who plays herself acting as herself within a fictional play fictionalizes herself, but only in the way that I conceptualized the pen. I cannot become logically completely fictional: there remains a distinction between outside and inside the fiction.
Fictions are fictions because they are framed as fictions. It is neither the content nor the fictional status of the characters or anything else that makes a fiction a fiction. A fictional character, however, is ontologically no more odd than a stop sign described in the non-fictional statement, “I have learned to stop at stop signs.” Of course, someone might deny this equivalence if they accepted Russell’s premise that such a sentence must include a reference to an existent stop sign, here understood as a general concept (does that commit us to believing that we are referring when we say this sentence to a thing called both a concept and a concept of stop signs?). I will not reenter the debate about this here. I will, however, reiterate that the inclusion of a non-existent thing in a story is not what makes that story fictional. We can turn a non-fictional statement into a fictional statement simply by treating it as such. What this means is that fictions are logically distinct not from the actual world but from non-fictional statements. If that is the case, then the distinction that matters, that tells us about the logical kind of thing fictions are is not between fictions and the world but between fictions and non-fictions. What we count as a non-fiction or a fiction might be determined culturally, or sociologically as Rorty would say, but what it means to be the one or the other is a logical distinction.
The logical distinctions manifest as the differences between signs and laws, pencils and ‘pencils,’ and fictions and non-fictions define some of the conceptual limits we live amidst. To collapse these limits produces nonsense. To fail to understand a fiction as a fiction, for example, as happens to hilarious effect in the film Galaxy Quest, is to act nonsensically. We live in a world of logical possibilities. Fictions are one form of this, akin to but not the same as hypotheses, plans, pictures, and even fears (our fears of abstract possibilities—of disease, for example). Fictions rely on and thus provoke our own attention and confusions about the realm of logical possibilities. Those possibilities are not actual in the same way my keyboard is as I write, but both the world and these possibilities are the world we human beings live in. That we live and think amidst and through these logical possibilities is what not only allows us to write fictions, and it is the reason we want to read them. In saying this, I am resisting Rorty’s tendency to understand fictions as grasped by epistemological means and not logical means.
I want to offer a final example of how we live amidst the possibilities of form and sense that allow us to think about the world. The example is also mundane in its way, but slippery to grasp in its implications and force. In The Peregrine, a masterpiece of descriptive writing, J.A. Baker describes his efforts to get close to a set of peregrines in the fen lands of eastern England. At one point in his pursuit of these birds, he sees a peregrine hunt. The raptor circles, swoops, passes nearby. All the animals around him quiet into silence. Baker hears the quiet; the world shifts from one tone to another. The peregrine makes its kill. And then it shrieks and all is silent again. But this silence is different from the first:
The curved hook of its beak protruded from the heart-shaped disc of the mask, like a single claw. The dark eyes were rimmed with wine. It flew overhead, and in the first coldness of the spring night, suddenly called. A hoarse bellowing shriek drew out to a sharp edge, and bristled away to silence. But not the silence that was there before (158).[2]
First, we have a moment of sound given shape—“a hoarse bellowing shriek drew out to a sharp edge, and bristled away to silence.” Then, we have the aftermath in which all is changed—“but not the silence that was there before.”
This second silence is a memory of the first, with a difference: something has been killed. The silence is an after-something silence, not a before-something silence. That is its difference. It is a silence arising from motives and reactions slightly altered from the first disturbance into silence. That first silence expressed wariness and fear. The second happens amidst a kind of relief for those that react as prey to the peregrine’s hunting. The hunt is temporarily in abeyance. Baker feels paused by the event of that killing, halted in his own pursuit of the raptor.
This silence is meaningful. That meaningfulness is not mere projection. I will make three interrelated arguments in support of this last claim. First, silence is not anything in itself. How it is heard is what it is; it is constitutive of silence that it be heard. Anything that is heard must be heard in some way. Consequently, how silence is heard is directly implicated in what it is. Thus it is not mere projection.
We can strengthen this argument by attending to how silence is heard. Silence is a kind of auditory form, and a second silence takes its form relative to the form and fact of the first. We hear silence as a form as surely as we hear a melody, and for the same reason. We do not live in the moment. We are dispersed backwards and forwards in time. We hear patterns of tone, not single pitches, and what we take as a single pitch is itself an oscillation of frequencies. After the peregrine has killed, the second silence emerges as Baker’s concentrated awareness of the negative space, or rather, the negative sound that outlines the raptor’s shriek of rage and power. Again, it is not mere projection.
We can combine the previous two arguments into a third. Silence, like sound, has duration. It is understood as part of a series. Consequently, it is grasped comparatively. The second silence fits with and recalls the silence that preceded the hunt. Both silences frame death and express fear. This second silence, however, mimics the first. The first silence becomes the concept by which the second is grasped as similar but different. This should recall my earlier example of the pencil that I turned into a concept. The first silence is like the physical pencil that becomes the concept by which the second silence can be distinguished as both alike and different. In other words, we recognize this silence like we might a spoken word or the concept pencil when offered as an example. But with the second silence, it is not an action of a person that is recognized, but the event that follows from the action of a raptor. This is not projection; it is the way the world emerges as meaningful for us within our own involvement with it. In some sense, therefore, all silences are second silences of previous silences. This is not the kind of deflationary argument Rorty makes about how our words relate to the things, but it is the direction in which he points us.
These arguments also suggests that our philosophical interest in fiction, as well as other conceptual questions about similar limits of sense, are not exhausted by resisting the question “How do our statements hook onto the world?” What is mysterious, strange and wondrous is how we inhabit the world in the ways we do. Words can hold something dear and silent. “Silence is a shape that has passed,” Stevens writes towards the end of his life, listening for death, the sun less and less a hope for him (“The Green Plant”). The shape of silence is second silence. We hear what is passed because that past is over (the sound has ceased) and because this silence recalls an earlier quiet. We do not need theories about semantics, reference, or language to explain silences’ shapes. But that silence has a shape is wondrous, and our philosophical descriptions help reveal that wonder.
