Research

Some questions prompt only intuitive responses. When we ask, “How much should I donate?”, our first, intuitive response might be something like, “whatever feels right.” However, we might step back and reflect on this intuition. “Must I donate that much?” “Must I donate anything?” This reflection can change or confirm our initial intuition. “Upon reflection, there are more effective charities.” Interestingly, intuition also features in philosophical and scientific discourse.i So does reflection.ii

My research clarifies the nature and normativity of intuition and reflection in philosophy, science, and everyday reasoning. By employing both empirical and philosophical tools I advance theory (e.g., social psychology, moral psychology, epistemology, etc.) and practice (e.g., education, religion, ritual, etc.).

Great Minds Do Not Think Alike—E.g., not all who ponder count costs

In two large studies (N = 1299), one pre-registered, I find that much of what we observe among non-philosophers replicates among philosophers (Byrd, in prep.). Specifically, less reflective philosophers preferred many of the philosophical views that less reflective non-philosophers have preferred.iii However, philosophical views were sometimes better predicted by culture, education, gender, numeracy, and personality than by reflection per se. This aligns with my earlier studies of morality: Is it appropriate to cause harm for the sake of the greater good? In two studies (N > 450), accepting such utilitarian harm tradeoffs correlated with numeracy rather than reflection per se (Byrd & Conway, 2019).

Together these findings suggest that there is more to the link between reflection and philosophy than expected. Moreover, these findings undermine arguments for the superiority of certain ethical judgments according to their correlation with reflection.iv

Explicating ‘Reflection’: Construct validity and epistemic identity

A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the ball. How much does the ball cost? “10 cents” comes to most peoples’ minds without any reflection. However, upon reflection, one can realize that the correct answer is 5 cents. There are many other reflection tests that lure people toward a demonstrably incorrect answer. As such, they are a reasonable measure of the kind of reflection that philosophers talk about—e.g., Korsgaard’s metaphor about backing up reconsidering our initial impulse.

However, some reflection tests are mathematical—like the bat-and-ball problem—whereas others are logical, or else not obviously mathematical or logical.v Further, some measures of reflection assume, a priori, that the correct answers were reflective and the lured answers were unreflective. This results in ambiguity: reflection tests may track reflection per se, logical competence, mathematical competence, or an a priori assumption about reflection. I argue for two solutions to these ambiguities (Byrd, in prep.): think aloud protocols and process dissociation.vi I am applying both methods to measures of reflection (Byrd & Sirota, in prep.). Participants in the US and UK will complete reflection tests as usual or while thinking aloud. This will allow us to test the a priori assumptions of reflection tests as well as the effect of thinking aloud (reflecting) on reflection test performance.

Theories of reflection must explain why reflection can either help or hinder reasoning—e.g., productive equilibrium vs. undesirable polarization.vii Bounded reflectivism explicates ‘reflection’ (Figure 1) in terms of—among other things—epistemic identity: treating some beliefs as part of our identity. Suppose that I identify with a religion. If you challenge my religious identity, I might reflectively defend my religious beliefs rather than dispassionately submit to the best arguments and evidence. Rather than suppress identity, I should embrace it—e.g., by focusing on a shared, superordinate epistemic identity. With funding I can test the view (Byrd, in prep.).

A flowchart depicting the algorithm of intuition and reflection.
Figure 1. Bounded Reflectivist Model of Reflection

I also collaborate to develop resources for researchers in my area. For example, David Yaden, Derek Anderson, and I are establishing the Psychology of Philosophy Network to build connections between researchers investigating philosophical reasoning and pool resources wherever possible. For example, we plan to replicate and clarify findings that philosophers’ beliefs can be predicted, in part, by psychological factors such as cognitive ability, personality, wellbeing, psychedelic drug use, and others (Yaden & Byrd, in prep.).


i In philosophy: Chalmers, 2014; Kornblith, 1998; Mallon, 2016. In science: Tallant, 2013

ii In philosophy: Goodman, 1983; Hursthouse, 1999; Kennett & Fine, 2009; Korsgaard, 1996; Rawls, 1971; Sosa, 1991; Velleman, 1989; 2000; Wallace, 2006. In psychology: Pennycook, 2018; Sperber, 1997.

iii E.g., Hannikainen & Cova, in prep.; Pennycook et al., 2016.

iv Baron, 1994; Bazerman & Greene, 2010; Greene, 2013

v Mathematical: Frederick, 2005. Logical: Janis & Frick, 1943. Other: Sirota et al., forthcoming

vi Think Aloud Protocol Analysis: Ericsson & Simon, 1998. Process Dissociation: Jacoby, 1991.

vii E.g., Kahan, 2013; Kahan et al., 2017; Pennycook & Rand, 2018a; cf. Pennycook & Rand, 2018b.