1. For the Ehrenfest model see Barberousse and Ludwig (2009). For the plum pudding, see Suarez (2009, 162-63).
2. Achinstein (1968) characterised models largely in terms of the known falsity of assumption, useful because they combine simplification and approximation. Vaihinger, more or less the founder of fictionalism, says little about literature or the imagination. There is brief discussion of “aesthetic” and “poetic” fictions, and of mythology; centaurs, griffins, and other nonexistents are said to be important for the theory of existential propositions but of “minor importance for our present theme” (1924, 82).
3. Theories of fiction often appealed to in this context have been Currie (1990), and Walton (1990). For negative assessments of this idea see Giere (2009) and Teller (2009).
4. Later on I will question this way of putting the matter; see below, Section 2.
5. Versions of this view are developed in Walton (1990) and Currie (1990). One dissenter is Matravers (2014), though he does not claim that belief is the appropriate response to fictions. For more dissent, confined to the cinematic case, see Quilty-Dunn (2015).
6. Williams (1973) offered the first of these as the reason why the second is true. There is now a substantial literature on both, with no agreement as to the truth of either or even their best formulations.
7. I ignore some complications here. For more detail see my (2014a).
8. See Walton (1990; 2015).
9. See Lewis (1983), Currie (1990, chapter 3), Byrne (1993).
10. There is currently a debate over whether we can define fiction in terms of its functioning or being intended to secure a distinctive attitude of imagining on the part of an audience (see e.g., Friend 2012; Currie 2014b). But even those who deny this might agree with the sentence above. If they do not, that's one less difference between fiction and nonfiction for me to worry about, at least for the purposes of this paper.