The dual-process framework (e.g., Evans & Stanovich Reference Evans and Stanovich2013; Kahneman Reference Kahneman2011a) has long conceived human thinking as an interplay of fast and intuitive processing (“System 1” thinking) and slower, more demanding deliberate processing (“System 2” thinking). The characterization of rationalization in this popular framework is often problematic.
On one hand, rationalization is typically conceived as epiphenomenal. It is considered as a mere “making-up-excuses-after-the-facts” in which reasoners post hoc look for justifications for an (often inappropriate) intuitively cued choice. For example, one might think here of classic reasoning and heuristics-and-biases tasks in which people fail to correct biasing intuitions but are afterwards all too eager to find reasons to support their (erroneous) intuition (e.g., Evans & Wason Reference Evans and Wason1976; Kahneman Reference Kahneman2011a). As such, rationalization would have no functional “rational” role to play in sound reasoning.
At the same time, dual-process theorists also tend to characterize rationalization as a deliberate (System 2) process. Indeed, people often spend considerable time and effort to come up with justifications and rationalize their answers (Pennycook et al. Reference Pennycook, Fugelsang and Koehler2015; Wason & Evans Reference Wason and Evans1975). This poses a puzzle. Why would we waste scarce resources on a pointless epiphenomenon? The fact that the human cognitive miser – who typically prefers to refrain from demanding deliberation – nevertheless engages in it to rationalize its behavior suggests that rationalization must serve an important function (Evans Reference Evans2019). Unfortunately, this functional role of rationalization has received little attention in traditional dual-process models.
However, recent dual-process work has started to hint at a possible role in social communication and argumentation (Bago & De Neys Reference Bago and De Neys2019; De Neys Reference De Neys2017; Evans Reference Evans2019). A key observation is that reasoners rationalize not only incorrect intuitions but also correct ones. One intriguing finding comes from two-response studies in which reasoners first have to answer as fast as possible with the first response that comes to mind and afterwards can take the time to deliberate and give a final answer (Bago & De Neys Reference Bago and De Neys2019; Newman et al. Reference Newman, Gibb and Thompson2017). Results indicate that sound reasoners do not necessarily need to deliberate to correct an initial erroneous intuition (e.g., “10 cents” in the infamous bat-and-ball problem); their initial intuitive response is often already correct. However, without subsequent deliberation, they struggle to give an explicit justification of their (correct) intuitive answer (Bago & De Neys Reference Bago and De Neys2019). In other words, good reasoners seem to intuitively know the correct response, but don't seem to know why it is correct in the absence of further deliberation. This indicates that sound reasoners do not necessarily deliberate to correct their intuition but to rationalize it and look for an explicit justification.
As Mercier and Sperber (Reference Mercier and Sperber2017) have stressed, such a justification process in which we look for explicit reasons in support of our intuitions can be critical to efficiently sway others. Clearly, if I want to convince my peers that my solution to a problem is right, I will be more successful when giving them an explicit, verifiable argument than by simply telling them that I “felt” it was right (Bago & De Neys Reference Bago and De Neys2019).
Whereas the recent dual-process findings (and the work of Mercier & Sperber Reference Mercier and Sperber2017) point to a possible functional role of rationalization in social persuasion, Cushman's account points to an additional contribution to our own “internal” information processing. In my opinion, such “internal” and “external” functions do not need to be mutually exclusive. However, my goal here is not to comment on the specifics of Cushman's proposal. The key point I want to highlight is that by pinpointing a rational role of rationalization, Cushman's work offers dual-process theorists a possible way out of the rationalization paradox.
As I tried to clarify, the lack of a functional account of rationalization is problematic for dual-process theories. If rationalization is not rational, it would be hard to explain why we spend our dearest resources on it and still survive as a species. Cushman builds a convincing case for the rationality of rationalization. Therefore, any dual-process proponent (or critic) should take note of it. My hope is that this will instigate renewed empirical research on rationalization in the dual-process field.
The dual-process framework (e.g., Evans & Stanovich Reference Evans and Stanovich2013; Kahneman Reference Kahneman2011a) has long conceived human thinking as an interplay of fast and intuitive processing (“System 1” thinking) and slower, more demanding deliberate processing (“System 2” thinking). The characterization of rationalization in this popular framework is often problematic.
On one hand, rationalization is typically conceived as epiphenomenal. It is considered as a mere “making-up-excuses-after-the-facts” in which reasoners post hoc look for justifications for an (often inappropriate) intuitively cued choice. For example, one might think here of classic reasoning and heuristics-and-biases tasks in which people fail to correct biasing intuitions but are afterwards all too eager to find reasons to support their (erroneous) intuition (e.g., Evans & Wason Reference Evans and Wason1976; Kahneman Reference Kahneman2011a). As such, rationalization would have no functional “rational” role to play in sound reasoning.
At the same time, dual-process theorists also tend to characterize rationalization as a deliberate (System 2) process. Indeed, people often spend considerable time and effort to come up with justifications and rationalize their answers (Pennycook et al. Reference Pennycook, Fugelsang and Koehler2015; Wason & Evans Reference Wason and Evans1975). This poses a puzzle. Why would we waste scarce resources on a pointless epiphenomenon? The fact that the human cognitive miser – who typically prefers to refrain from demanding deliberation – nevertheless engages in it to rationalize its behavior suggests that rationalization must serve an important function (Evans Reference Evans2019). Unfortunately, this functional role of rationalization has received little attention in traditional dual-process models.
However, recent dual-process work has started to hint at a possible role in social communication and argumentation (Bago & De Neys Reference Bago and De Neys2019; De Neys Reference De Neys2017; Evans Reference Evans2019). A key observation is that reasoners rationalize not only incorrect intuitions but also correct ones. One intriguing finding comes from two-response studies in which reasoners first have to answer as fast as possible with the first response that comes to mind and afterwards can take the time to deliberate and give a final answer (Bago & De Neys Reference Bago and De Neys2019; Newman et al. Reference Newman, Gibb and Thompson2017). Results indicate that sound reasoners do not necessarily need to deliberate to correct an initial erroneous intuition (e.g., “10 cents” in the infamous bat-and-ball problem); their initial intuitive response is often already correct. However, without subsequent deliberation, they struggle to give an explicit justification of their (correct) intuitive answer (Bago & De Neys Reference Bago and De Neys2019). In other words, good reasoners seem to intuitively know the correct response, but don't seem to know why it is correct in the absence of further deliberation. This indicates that sound reasoners do not necessarily deliberate to correct their intuition but to rationalize it and look for an explicit justification.
As Mercier and Sperber (Reference Mercier and Sperber2017) have stressed, such a justification process in which we look for explicit reasons in support of our intuitions can be critical to efficiently sway others. Clearly, if I want to convince my peers that my solution to a problem is right, I will be more successful when giving them an explicit, verifiable argument than by simply telling them that I “felt” it was right (Bago & De Neys Reference Bago and De Neys2019).
Whereas the recent dual-process findings (and the work of Mercier & Sperber Reference Mercier and Sperber2017) point to a possible functional role of rationalization in social persuasion, Cushman's account points to an additional contribution to our own “internal” information processing. In my opinion, such “internal” and “external” functions do not need to be mutually exclusive. However, my goal here is not to comment on the specifics of Cushman's proposal. The key point I want to highlight is that by pinpointing a rational role of rationalization, Cushman's work offers dual-process theorists a possible way out of the rationalization paradox.
As I tried to clarify, the lack of a functional account of rationalization is problematic for dual-process theories. If rationalization is not rational, it would be hard to explain why we spend our dearest resources on it and still survive as a species. Cushman builds a convincing case for the rationality of rationalization. Therefore, any dual-process proponent (or critic) should take note of it. My hope is that this will instigate renewed empirical research on rationalization in the dual-process field.