§1. Evans claimed that, in the case of a speaker S who makes a use of a singular term that has no referent in a normal information-invoking way, nothing is said by S (343). However, he readily recognized that he owns us an explanation of the contrary intuition according to which there seem to be perfectly intelligible uses of such terms. For example, negative existential sentences containing purely fictional names. In order to explain such uses, Evans employed the notion of ‘games of make-believe’, which is borrowed from Kendall Walton.
§2. Kendall Walton has notoriously defended the thesis according to which representational works of art can be seen as props in games of make-believe. Here I am not interested in giving a precise account of his up to date views on the topic, but rather to summarise the content of two papers which have influenced Evans, that is, (1973) Pictures and make-believe. The Philosophical review, 82 (3): 282-319; and Fearing Fictions. The Journal of Philosophy, 75(1): 5-27. First a merely exegetic note: Evans took from Walton the notion of make-believe and part of the symbolic notation (see Evans (1982: 354)). Evans further elaborated the various rules of these games, and applied these ideas to negative existential sentences. Walton, in his later book Mimesis as make-believe, subsequently borrowed from Evans this treatment.
§3. Statements like ‘the monster is coming’ or ‘I have a gun’ (while holding a stick), uttered while playing a game of make-believe (MB1), are literally false. However, they can be regarded as true in the context of the
make-believe itself, given the acceptance (or presupposition) of certain rules. In other terms, these statements can be fictionally true. Or so Walton says. He further adopts the following notation (see Walton (1973: 287)): a proposition p is fictionally true = *p*(F). According to Walton, fictional truths cannot be reduced to make-believe truths: for example, sentences true in the context of a dream are not make-believe truths. If there are features of the real world somehow related to certain rules proper of a game of make-believe, then certain corresponding fictional truths can be generated. A particular fictional truth p that is also a make-believe truth can be represented as *p*(MB).
§4. An important point in Walton’s papers is “What one says in asserting “This is a raisin pie” in a game of mud pies thus involves essentially the notion of something’s being a raisin pie (literally).” (Walton 1973: 296) This can be understood as suggesting that in order to get what is said in a game of make-believe, we should have some previous understanding of the embedded proposition. Evans defended a similar view.
§5. Walton claimed that we do not need to posit non-existent objects in order to provide an account of those propositional attitudes which allegedly involve fictional entities. This point is vigorously defended by Evans as well. It seems to me that Walton wants to suggest that when we appreciate works of fiction (for example, novels), we are literally caught up into the story. In fact, he claims that we actually engage in games of make-believe in which the works function as props. We do not literally root for the villain of the novel (as it always happens to me…), we *root for the villain*(MB), we create, so to speak, a fictional world in which certain proposition (literally false) can be true. We are engaged in games of make-believe in which we have quasi-emotions, that is, physical reactions not accompanied by the full range of dispositions to act typical of proper emotions. Certainly, games of make-believe have rules, if only implicit, which somehow regulate the appropriateness and derivability of certain fictional truths. They are not merely idle fantasies. In particular, it makes perfectly sense to say that we can find out fictional truths belonging to a specific fictional world, even though we might have not been aware of them since the very beginning.