Research

Some questions prompt intuitions. When a cashier asks, “Would you like to donate to [a non-profit organization]?”, our first impulse might be “not today.” Of course, we can step back and reflect on the intuition. “Why not today?” And reflection can change either change or confirm our initial impulse. “Upon reflection, if I can afford to help people in need, then I should.” Reflection is rife in the history of ideas, science, and everyday life.[i]

I construct arguments and deploy scientific methods to clarify the nature and normativity of reflective reasoning (Byrd, 2022; Cullen & Byrd, in prep.) in a wide range of areas: disagreement (Cullen, Byrd, et al., in prep.), philosophical decision-making (e.g., Byrd & Conway, 2019; Byrd & Białek, 2021Byrd, 2022b; Biebel & Byrd, under review), implicit bias (Byrd, 2019; Byrd & Thompson, 2022), technology (Byrd, 2020), and their intersections (Byrd, under review). The results have implications for both method and theory, both facts and norms, both individuals and groups.

Here is a selection of my research questions:

  • How are our evaluations of arguments and evidence biased?
  • How does our sense of identity bias our reasoning?
  • How do our biases relate to our moral, political, religious, and other philosophical tendencies?
  • How do our biases impact ours and others’ well-being?
  • How can we improve each others’ reasoning, decision-making, and well-being?
  • How can technology help (or hinder) these improvements?

Bounded Reflectivism: The Nature & Normativity Of Reflective Reasoning

One innovation of my research is a testable, algorithmic model of intuitive and reflective reasoning (Figure 1). The model captures how reflection can either help or hinder our reasoning. For example, reflection can lead to equilibrium, but it can also lead to polarization.[ii] The key is that reflection can be bound by factors like epistemic identity—identities that involve particular beliefs such as religious and political identities. When we sense a threat to one of our epistemic identities, we may reflectively defend its beliefs rather than dispassionately follow logic and evidence. For the bounded reflectivist, the solution is not to suppress or veil epistemic identity but embrace it—e.g., with shared, superordinate epistemic identities (Byrd, 2022a).

The visual algorithm of bounded reflection.
Figure 1. The Bounded Reflectivist Model of Reflective Reasoning.

Some data already align with bounded reflectivism: reflection can result in converge (Byrd & Conway, 2019) and even rational decisions (Cullen, Byrd, et al., in prep.), but sometimes reflective reasoners diverge from society’s beliefs (Byrd, 2022b). Funding from private and public institutions allows me to investigate the descriptive and normative aspects of reflection in our theories (Byrd & Sytsma, in prep.) as well as high-stakes, real-world thinking—e.g., rumination (Byrd, 2020), bias (Byrd, 2019), democracy (Byrd, 2022a), and intelligence analysis (Gilliam, 2022).

Modeling Intuition, Reflection, Norms, and Identity. When we step back from an impulse to reflect on it, we represent the impulse as well as some reasons for or reasons against the impulse—e.g., the consequences of acting on the impulse.[iii] Bounded Reflectivism incorporates reflection’s antecedents (e.g., perception), its mediators (e.g., use of digital technology), or its consequences in an algorithmic model that contains testable descriptions and predictions as well as normative implications.

Reflection & Debiasing . In additions to models, experiments can test the effect of epistemic identity on reasoning and disagreement. Consider disagreement and polarization. Support from the US Intelligence Community has allowed my colleagues and me to develop software to facilitate discussion-based experiments that improve decisions (Cullen, Byrd, Chapkovski, et al., in prep.) and mitigate partisan biases in people’s evaluation of arguments for controversial policies (Cullen, Byrd, and Oppenheimer, in prep.).

Mediators of Reflection’s Impact On Other Factors. The link between reflection and certain beliefs may vary across cultures or levels of education (Byrd, 2022b; Byrd & Sytsma, in prep.).[iv] This raises the question of how cultural and other factors might feature in the bounded reflectivist model of reflection. For instance, the effect of reflection on decisions might be mediated by the kinds of technology involved (Byrd, 2020). Another possibility is that reflection intervenes on the predominant views and norms around us that we unreflectively adopt. If that is right, reflective reasoning involves, among other things, countenancing more counter-cultural ideas or norms. The John Templeton Foundation has supported our research into this possibility—e.g., why correlations between reflection and religiosity vary by country and religion.

Progress Through Collaboration, Diversity & Innovation

With a diverse group of students and colleagues around the world, I am developing novel tools and revealing new insights about the nature, benefits, and limits of reflective reasoning (Byrd, Joseph, Gongora, & Sirota, in prep. Cullen, Byrd, et al., in prep.; Schwenkler, Byrd, Lambert, & Taylor, 2021). Our results continue to prompt conversations with more people that introduce additional perspectives, skillsets, and resources—something that, historically, has led to scientific progress.[v]


[i] Intuition in philosophy: Chalmers, 2014; Kornblith, 1998; Mallon, 2016. Reflection in philosophy: Goodman, 1983; Hursthouse, 1999; Korsgaard, 1996; Rawls, 1971; Sosa, 1991; Velleman, 1989. Intuition in science: Tallant, 2013. Reflection in psychology: Pennycook, 2018; Sperber, 1997.

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

[iii] E.g., Gervais et al, 2018.

[iv] Korsgaard, 1996; Frith, 2019

[v] E.g., Bickle, 2016; Braganza & Beck, 2018; Sullivan, 2022