The Value of Surprise in Science

Abstract

Scientific results are often presented as ‘surprising’ as if that is a good thing. Is it? And if so, why? What is the value of surprise in science? Discussions of surprise in science have been limited, but surprise has been used as a way of defending the epistemic privilege of experiments over simulations. The argument is that while experiments can ‘confound’, simulations can merely surprise (Morgan, 2005). Our aim in this paper is to show that the discussion of surprise can be usefully extended to thought experiments and theoretical derivations. We argue that in focusing on these features of scientific practice, we can see that the surprise-confoundment distinction does not fully capture surprise in science. We set out how thought experiments and theoretical derivations can bring about surprises that can be disruptive in a productive way, and we end by exploring how this links with their future fertility.

Putting the ‘Experiment’ back into the ‘Thought Experiment’

“…24 It is not always just a matter of lack of logical omniscience: cf. (French & Murphy, 2021) and the element of genuine surprise of TEs. 25 Nor often are artworks: paintings and novels are often telling about the real world.…”

“…On TEs and surprise, see also(French & Murphy, 2021).6 For an exhaustive analysis of Norton's view on TEs, see(Brendel, 2018).…”

Emotions in scientific practice

“…In this sense, one might conclude that surprise about the output of theoretical models is epistemically different from the genuinely epistemic surprise about experiments (Morgan 2005). 2 This view has been nuanced by Currie (2018), who suggests that models and simulations can genuinely be surprising in the epistemic sense (see also French and Murphy 2021). This is the case if the model's output challenges knowledge structures relevant to the target of the model and not just to the model itself.…”

“…In line with Velleman's idea, aesthetically and literarily successful thought experiments give a sharper focus and a better grasp of their puzzle, while clumsy ones might fail to do so (Stuart 2016; Murphy 2020a, 2023). A successful thought experiment lays the shortest path toward the issue at its heart, provides a unique vantage point at a paradox and essentially captures something truly surprising and curious (French and Murphy 2021).…”

Grounding interventionism: Conceptual and epistemological challenges

“…25 Moreover, science has a track record of repeatedly overturning commonsense metaphysical intuitions (Shtulman and Harrington 2016), including even the strongest modal intuitions, such as the intuitions "that non-Euclidean geometry is impossible as a model of physical space, that it is impossible that there not be deterministic causation, [and] that non-absolute time is impossible" (Ladyman and Ross 2007, 16). Since surprise is a measure of scientific success (French and Murphy 2021), we might even think that counterintuitiveness is characteristic of scientific discovery. These points cast serious doubt on the evidential weight of intuitions.…”