The aim of this poster is to motivate a fiction-based account of counternomics in science in terms of make-believe, the type of imagination employed in Walton’s (1990) pretense theory of fiction and recently carried over to fiction views of models, most notably by Frigg (2010), Nguyen & Frigg (2020), and Frigg & Salis (2020).

Why counternomics? Counternomics are counterfactual conditionals with nomologically impossible antecedents. Despite their appeal to nomological impossibilities, they feature in scientific explanations and model-based reasoning with non-trivial truth values (Tan 2019, Stuart et al. 2021), hence the importance of an account which can provide both the right semantics for counternomics and the right description of counternomic-reasoning processes.

Regarding semantics, one natural place to look at is the analyses of counterfactuals proposed by Stalnaker (1968) and Lewis (1973), which have traditionally been used to assign truth-values to counterfactuals. I argue that these accounts have several shortcomings when applied to counterfactual reasoning in science. In a nutshell, 1) they offer a bad reconstruction of counternomic reasoning in practice, as they only appeal to philosophical intuitions, 2) possible worlds contain too much information for successful counterfactual evaluation (Salis & Frigg 2020), 3) possible worlds don't allow logical inconsistencies, and 4) the implicit epistemology of modality that results from these semantic accounts is underdeveloped. Furthermore, and this is another reason for the focus on counternomics, 5) when conjoined with the view that laws of nature are metaphysically necessary possible-worlds semantics output vacuous counternomic truths, as there are no possible worlds where the laws are violated. In short, the prospects for factualist possible-worlds-based views of CFs are not good.

Regarding counternomic reasoning, research in the cognitive processes underlying counterfactual reasoning (Byrne 2005, 2016) highlights the central role of the imagination, constrained in a reality-oriented way, as well as the need for a uniform treatment of all types of counterfactuals across the board (that is, counterfactuals' evaluation is independent of the modal status of their antecedents.)

Putting all these results together, I propose to examine counternomics through the lens of fiction as imagination, an underexplored strategy which evades most of the problems of the Lewis-Stalnaker semantics while doing justice to mental modelling psychological theories of counterfactual reasoning. Imagination is here understood as pretense in terms of games of make-believe à la Walton (1990). Briefly, games of make-believe are initiated by props - objects or texts which trigger some direct imaginings - which together with principles of generation (PGs) prescribe additional imaginings. Applied to CNs, make-believe results in the following framework: their antecedents become prescriptions to engage in legitimate pretenses, where legitimacy is no other than the demand that these imaginings serve scientific aims such as driving empirical discovery or enhancing theoretical understanding. Given that make-believe is a type of propositional imagination (Currie 1990), counterfactuals get fictional truth values: they are fictionally true (f-true) iff their consequents are prescribed for imagination by their antecedents and pertinent PGs (See Kimpton-Nye 2020), which in the case of science are laws of nature, mathematical knowledge, and rules of inference. Whether these imaginings represent features of the world is a question that falls outside of make-believe but can be dealt with the theory of scientific representation of choice.