Fictional entities

1. Distinguishing the fictional from the nonfictional

It might be supposed that what determines whether an entity is fictional or not is whether it exists. Perhaps fictional just means nonexistent. Certainly it is commonplace to contrast what is fictional with what is real. The crucial difference between Socrates and Sherlock Holmes is that the one existed and the other did not. The trouble comes when we find serious philosophers, such as Russell, describing even beings like Socrates as ‘fictions’. That invites the thought that there are different kinds of fiction and even different kinds of nonexistence. If that is right it compromises any simple identification of the fictional with the nonexistent. Also philosophers in the twentieth century have become increasingly wary of unqualified talk about what exists and what does not. Among logicians and philosophers of science there has been a pronounced shift away from asking what entities exist (or are real) towards asking what entities particular theories are committed to. Underlying this shift and acting as a constraint on theory construction is the heuristic principle called Ockham’s razor (after the medieval logician William of Ockham), namely ‘Do not multiply entities beyond necessity’.

In seeking to refine the association of the fictional with the nonexistent it is helpful to distinguish the role of fictional entities in logic, epistemology and literary narrative. There is no compulsion to suppose that some uniform account of fictional entities must span these three applications. Indeed it becomes evident that different criteria for distinguishing the fictional from the nonfictional operate in each case.

First of all, fictions in logic are most commonly associated with eliminability by logical paraphrase. The idea is roughly this: that a fictional entity is the purported referent of an eliminable singular term. It relies on a distinction between apparent reference and genuine reference; the object of a genuine reference must exist, or be assumed to exist by a theory, while apparent but non-genuine reference is that which is eliminable by paraphrase without loss of content. The conception probably originated with Bentham, who thought that rights, for example, were fictitious entities, precisely because sentences containing ‘rights’ as a noun (such as ‘The people demanded their rights’) can be paraphrased into sentences in which no such noun or singular term appears (such as ‘The people demanded fair treatment under the law’). As long as the paraphrase captures the significant content of the original it shows that the apparent reference in the original (to a specific class of entities) is not a genuine reference: thus are the entities ‘fictional’.

Bentham’s theory was a clear forerunner of the school of twentieth-century logical analysis epitomized by such philosophers as Bertrand Russell, Rudolf Carnap and W.V. Quine. Russell’s Theory of Descriptions eloquently showed that singular descriptive phrases, such as ‘the highest prime’ or ‘the golden mountain’, do not need to function logically as naming expressions (that is, having, in our earlier terminology, ‘genuine reference’) in order to be meaningful (see Russell, B. §9). Quine extended the thesis to all singular terms, including proper names, arguing that each is eliminable by logical paraphrase, in favour of quantifiers and predicates. Only bound variables, for Quine, are the true bearers of referential commitments (see Quine, W.V. §5). We will return to Russell’s and Quine’s theories in the next section, for they have been highly influential in deflationary accounts of fictional entities.

While fictions in logic connect with reference and paraphrase, fictions in epistemology rest on the idea of ‘construction’. Ordinary objects can be thought of as fictions, in epistemological theories, just to the extent that they are thought of as ‘constructed’ (by the human mind) out of more basic, perhaps more real, elements. A characteristic kind of empiricism equates what is real with what is knowable and what is knowable with what is given in experience. For those empiricists who hold that the only immediate objects of experience are subjective entities like ‘impressions’ (Hume) or ‘sense-data’ (Russell), it is a short and natural step to the idea that enduring objects in space and time are mere ‘constructs’ or posits of the mind that round out the flux of experience. Quine, albeit rejecting any privileged foundation of knowledge, has famously compared physical objects with the gods of Homer and spoken of both as ‘myths’.

No doubt for those who think of fictional characters in legend or literature as paradigmatic fictional entities, it will seem tendentious, even paradoxical, to describe physical objects as ‘fictions’. However, etymology associates fiction more with ‘making’ or ‘feigning’ than with nonexistence and empiricist epistemologists can be taken to be emphasizing the different kinds of ‘making’, or imaginative invention, involved in human knowledge. Nelson Goodman has described knowledge itself as a species of ‘worldmaking’. None of these philosophers wants to collapse the distinction between a fictional character, in the familiar literary-critical sense, and a real person.

So what sets apart the characters of literary narratives from the ‘fictions’ in logic and epistemology? One common idea is that fictional entities such as the Time Machine and Sherlock Holmes are logically rooted in the narratives in which they are first introduced; their very conception derives from the descriptive content of those narratives. To find out about Holmes one must read the Holmes stories. This already hints at a fundamental difference from nonfictional objects, for the latter do not derive their existence and nature from any narrative; to find out about them we must investigate the world, not merely a source text. Fictional entities of the literary type are ‘made’ or ‘made up’ precisely by being described in an act of storytelling, under the appropriate conventions. The fact that they do not exist in the real world can be seen to be as much a consequence of their origin (real objects are not made up by storytellers) as a defining feature of their fictionality.