Knowledge through the scientific imagination

 Fiora Salis is Associate Lecturer at the University of York. She works on imagination, fiction and scientific models. Currently, she is developing a new project on epistemic uses of imagination in science and art.

Fiora Salis is Associate Lecturer at the University of York. She works on imagination, fiction and scientific models. Currently, she is developing a new project on epistemic uses of imagination in science and art.

A post by Fiora Salis.

The problem of how scientists gain knowledge of reality through imagination in scientific models is still largely unresolved. Consider the Lotka-Volterra model of predator-prey dynamics. The model is usually identified with two differential equations interpreted as describing the growth rates of two populations, one prey and one predator, dynamically interacting with each other. To facilitate mathematical treatment, the model makes a number of assumptions, including that predators have infinite appetite, prey have limitless supplies of food, and the environment never changes. These assumptions enable the isolation of certain features of predator-prey interaction (including for example the density of the two populations) not by abstracting away from some particular predators and prey interacting with each other, but by stipulating that some populations having certain features interact in such and such a way. These assumptions, of course, are false (the idealized populations do not exist in reality), but they are not lies either. They are the product of creative uses of imagination that divert from reality to generate a model system as the object of study. They describe two imaginary populations interacting with each other under imaginary conditions. And this enables the generation of certain hypotheses and the assessment of their truth-likeness. The model predicts that the dynamic interaction between predators and prey will show a cyclical relationship in their numbers. Imagination is therefore vital both to the construction and development of the model and to the generation of plausible hypotheses.

What sort of knowledge is produced through scientific imagination? An answer to this question is inspired by the similar problem of how we learn through imaginative engagement with works of fiction. When creating and developing a work of art, the artist considers alternative possible ways the work could be developed, including possible outcomes (similar to the scientist when creating and developing a scientific model). Sometimes readers and literary critics do the same in forming an evaluative judgment of how the work could have been better or worse than it actually is (similar to scientists assessing the truth-likeness of hypotheses generated through imagination in the model). For example, by imagining a different ending to a story, we may come to a reliable judgment about the strengths and weaknesses of the work. So reliable critical evaluation sometimes requires imagining alternative possibilities. Knowing what those other possible versions are requires imagination. So, justified artistic evaluation in some important cases requires imagination (just like justified scientific evaluation of models requires imagination).

Imagination also plays an integral role in the ways in which people can gain knowledge of reality. Flaubert created and developed Madame Bovary through his own imagination and wrote about the vicissitudes of Emma Bovary, a doctor’s wife in provincial Northern France who lived beyond her means and had adulterous affairs. We, as readers, imaginatively engage with the story and learn about Emma’s psychology, including the reasons for and consequences of her actions. But we also learn about contemporary rural French people, their social aspirations and the ways they aped an urban bourgeoisie.

Philosophers of art usually distinguish between knowledge-claims about the fiction and the imaginary scenario associated with it, and knowledge-claims about reality that are generated through the fiction (Novitz 1987). A similar distinction can be made also in the context of modelling. Scientists can make knowledge-claims about the imaginary system described in the model, and they can make knowledge-claims about reality that are produced through imagination in the model. For example, they can claim to know how imaginary predator and prey populations interact in the Lotka-Volterra model and, on the basis of this claim, they can also claim to know how real predator and prey populations dynamically interact with each-other. This distinction provides a useful basis for the systematic investigation of the sort of knowledge involved in these contexts.

This investigation, however, presupposes a previous understanding of the ways in which imagination is constrained. Imagination is often thought of as being completely free and subject to the will. But the key to a genuine understanding of knowledge through imagination is in those uses of imagination that are constrained in certain relevant ways (Kind and Kung 2016). I hypothesise that there are three main types of constraints on imagination, architectural, context-specific and epistemic.

Architectural constraints are determined by the cognitive structure of the imagination and operate on all uses of imagination. Two main architectural constraints emerge from contemporary studies in cognitive psychology, mirroring and quarantining (e.g. Leslie 1987; Perner 1991; Nichols and Stich 2000). Mirroring is displayed when imagination carries inferential commitments that are similar to those carried by belief. Quarantining is that feature of imagination according to which imaginings have effects only within an imaginative context and do not guide action in the real world. Of course, this does not mean that nothing of real-world importance can be learned through imagination. But learning through imagination requires moving one step away from it, by exporting what one has learned within an imaginary context outside of it and into reality.

Context-specific constraints are determined by disciplinary conventions and interpretative practices. Walton (1990) appeals to reality orientation and mutual belief orientation as contributing to the generation of truths in fiction. Reality orientation operates when features of an imagined episode are derived from features of reality. Mutual belief orientation operates when features of an imagined episode are derived from the mutual beliefs of the community where the episode originated. Both constraints are highly problematic, but I do not have the space to engage in a critical discussion here. Others have already emphasised that imagination is constrained differently in different artistic genres, e.g. epic, tragic, comic (Shen-yi Liao 2016). And Roman Frigg and I (forthcoming) argue that different (context-specific) constraints operate on imagination in different types of models, e.g. mathematic, mechanistic, computational. However, context-specific constraints have not been studied in any rigorous or systematic way and they are still poorly understood.

Finally, epistemic constraints are determined by the particular epistemic purposes (the sort of knowledge we want to acquire) of an episode of imagination. Not all constrained uses of imagination are aimed at gaining knowledge of reality. A writer of fiction and her readers might respectively create a work of fiction and imaginatively engage with it to escape reality or to enjoy a particular imaginative experience. Scientists typically have some particular epistemic purpose in mind when building and developing a model. I can see at least two epistemic constraints on imagination: justification and rationality. What makes a particular imagining justified is the possession of evidence that is provided by the original set of assumptions together with the relevant context-specific constraints. We are justified in imagining that Emma Bovary was the wife of a country doctor if this conforms to the story’s stipulations that this was the case. Similarly, we are justified in imagining that predator and prey populations dynamically interact with each other in cyclical ways if this conforms to the imaginings generated in the Lotka-Volterra model according to the relevant context-specific constraints. A particular imagining is minimally rational if it is a member of a set of imaginings that is internally consistent. A set of imaginings is internally consistent if for any proposition p in the set it is the case that either p or not-p holds. Imagining that a predator population has infinite appetite is rational if this is true in the model and it is not the case that the same population does not have infinite appetite in the same model. Imagining that Emma was the wife of a country doctor is rational if this is true in the story and it is not the case that she was not a country doctor’s wife in the same story. This minimal rationality condition can be strengthened by adding further conditions relating to the particular evidence provided in the imaginary context of the fiction or the model. 

There is no philosophical or scientific understanding of how scientific and aesthetic imagination generate knowledge of reality. However, philosophers and cognitive scientists have a common interest in this issue and bringing together both perspectives may lead to a better understanding in the future. 


Kind, Amy and Kung, Peter 2016. Introduction: The puzzle of imaginative use. In Amy Kind and Peter Kung (eds), Knowledge Through Imagination, Oxford University Press.

Leslie, Alan, 1987. Pretense and representation: The origins of ‘Theory of Mind’. Psychological Review 94/4: 412-26.

Liao, Shen-yi, 2016. Imaginative Resistance, Narrative Engagement, Genre, Res Philosophica 93/2: 461-482.

Nichols, Shaun and Stich, Steven 2000. A cognitive theory of pretense. Cognition 74: 115-47.

Novitz, David 1987. Knowledge, Fiction & Imagination. Temple University Press.

Perner, Josef 1991. Understanding the Representational Mind. Cambridge: MIT Press.

Salis, Fiora and Frigg, Roman forthcoming. Capturing the scientific imagination. In Peter Godfrey-Smith and Arnon Levy (eds), The Scientific Imagination, Oxford University Press.

Walton, Kendall 1990. Mimesis as Make-Believe. Harvard University Press, Cambridge MA.