We question Knobe's thesis that in acknowledging the asymmetries observed in reasoning associated with moral judgment we must perforce abandon a “theory-theory” characterization of naïve psychological reasoning. Certainly there are pervasive asymmetries, at least when the situation affords opportunity for moral evaluation, but such asymmetries may not, in themselves, invalidate a theory-theory perspective.
We suspect that underlying the asymmetries is a tendency to view prosocial dispositions as the norm, a tendency that makes us especially vigilant about antisocial dispositions. In other words, we firmly expect others to have prosocial intentions and behavior. When asked about actions that involve harm-infliction or rule breaking, we assume these anomalies are the events that require explanation and attention. This vigilance, which recognizes moral culpability as a conspicuous and interesting phenomenon, can account for the empirical findings in question without compelling rejection of a theory-theory characterization of naïve psychological reasoning. That is, just because we bring to our reasoning something like a base-rate assumption, or perhaps even a “core knowledge” assumption (à la Spelke & Kinzler Reference Spelke and Kinzler2007) concerning prosocial intentions and behavior, it does not follow that we are essentially unscientific or irrational in the reasoning that ensues.
Even if we begin with a default assumption about its nature, we may nevertheless view human behavior through a theory-like framework that can be rationally revised through experience (e.g., Gopnik & Wellman Reference Gopnik and Wellman1992). In essence, we suggest that, despite the arguments raised by Knobe against motivational and conversational pragmatics hypotheses, there is some such alternative view that can account for the empirical findings without invalidating a theory-theory characterization of naïve psychological reasoning.
Knobe's point regarding the pervasive nature of observed asymmetries in psychological reasoning, at least in reasoning that invites moral judgment, is well supported. But the asymmetries may simply reveal that people are poised to seize on moral failings and treat them as focal. They are less inclined to view prosocial activities as warranting explanation and evaluation. In keeping with the widely popular notion that we have a deep-seated intuitive sensitivity regarding harm (e.g., Haidt Reference Haidt2001; Hoffman Reference Hoffman2000; Turiel Reference Turiel and Eisenberg2006), perhaps we exhibit what a theory-theorist might consider a core knowledge bias when reasoning about others. Specifically, we view good intentions and behavior as the norm and are consequently hypersensitive to deviations from it. When asked about people who act badly, we assume the deviation is what we should attend to and explain (in a Gricean fashion), and respond accordingly.
On this view, when asked whether the chairman of the board has intentionally harmed the environment by instituting a new program that he knows would create harm in addition to increasing profits, people respond affirmatively. Such a response, we think, reflects vigilant attention to someone knowingly acting in a way that causes harm. The same interpretation applies to the findings reflecting variations on the terms (“deciding,” “desire,” “advocating,” etc.). In these cases, the unacceptable, yet fully cognizant, behavior is simply assumed to be the topic of conversation. But when asked whether the chairman intentionally helped, rather than harmed, the environment, people respond differently – because helping is not a weird and worrying thing that commands vigilant attention. A question about whether helping was intended is interpreted as a query about whether the chairman had that outcome as a goal. In a world where prosocial aims are assumed to be the norm, we are not inclined to give the chairman moral credit for the outcome because the original goal, according to the narrative, was to increase profit. These asymmetrical responses do not reflect irrational reasoning; they reflect different assessments of what is the focal issue, or curiosity, to be addressed and explained. It may be that such asymmetries are universal because there is a core knowledge presupposition about the prosocial nature of people, possibly from very early in development (see Hamlin et al. Reference Hamlin, Wynn and Bloom2007). This assumption, in itself, does not mean the reasoning is illogical or unscientific, but it does influence the topic presumed to warrant explanation and discussion.
How does this interpretation account for people's judgments of causation? When both an administrative assistant and a faculty member are said to have removed pens from the receptionist's desk, the former with permission and the latter illicitly, people indicate that the professor caused the problem. This response may also reflect vigilant attention to the anomalous bad behavior, the event that requires explanation, and a corresponding shift in conversational focus. Knobe astutely notes that it is not just that the behavior is strange or unusual (see, e.g., Roxborough & Cumby Reference Roxborough and Cumby2009) but specifically that the behavior is morally bereft. So it seems it is moral culpability in general that is regarded as anomalous, not the specific actions of a certain person or even a group of people. Moreover, it does not matter whether in fact the final outcome is actually good (Hitchcock & Knobe Reference Hitchcock and Knobe2009, cited in the target article); the focus continues to be on the “bad” behavior. So – and here we think we agree with Knobe – it is not accurate to characterize what we bring to these reasoning tasks as purely a base-rate assumption regarding any specific action; rather, we simply assume that people will usually have benign intentions and act in a prosocial fashion.
In one respect, our interpretation resembles Knobe's. We agree that moral considerations are being taken into account in the reasoning reported in the literature. And it may be objected that our interpretation is simply another version of the motivational and conversational pragmatics hypotheses refuted by Knobe. But we think it is different to view people as bringing to their reasoning core content assumptions about human nature, assumptions that influence the presumed focus of conversations. The fact that people have expectations about others' prosocial and antisocial proclivities does not in itself mean that reasoning about causality in the social realm is irrational, impervious to experience, or otherwise at odds with scientific theorizing.
We question Knobe's thesis that in acknowledging the asymmetries observed in reasoning associated with moral judgment we must perforce abandon a “theory-theory” characterization of naïve psychological reasoning. Certainly there are pervasive asymmetries, at least when the situation affords opportunity for moral evaluation, but such asymmetries may not, in themselves, invalidate a theory-theory perspective.
We suspect that underlying the asymmetries is a tendency to view prosocial dispositions as the norm, a tendency that makes us especially vigilant about antisocial dispositions. In other words, we firmly expect others to have prosocial intentions and behavior. When asked about actions that involve harm-infliction or rule breaking, we assume these anomalies are the events that require explanation and attention. This vigilance, which recognizes moral culpability as a conspicuous and interesting phenomenon, can account for the empirical findings in question without compelling rejection of a theory-theory characterization of naïve psychological reasoning. That is, just because we bring to our reasoning something like a base-rate assumption, or perhaps even a “core knowledge” assumption (à la Spelke & Kinzler Reference Spelke and Kinzler2007) concerning prosocial intentions and behavior, it does not follow that we are essentially unscientific or irrational in the reasoning that ensues.
Even if we begin with a default assumption about its nature, we may nevertheless view human behavior through a theory-like framework that can be rationally revised through experience (e.g., Gopnik & Wellman Reference Gopnik and Wellman1992). In essence, we suggest that, despite the arguments raised by Knobe against motivational and conversational pragmatics hypotheses, there is some such alternative view that can account for the empirical findings without invalidating a theory-theory characterization of naïve psychological reasoning.
Knobe's point regarding the pervasive nature of observed asymmetries in psychological reasoning, at least in reasoning that invites moral judgment, is well supported. But the asymmetries may simply reveal that people are poised to seize on moral failings and treat them as focal. They are less inclined to view prosocial activities as warranting explanation and evaluation. In keeping with the widely popular notion that we have a deep-seated intuitive sensitivity regarding harm (e.g., Haidt Reference Haidt2001; Hoffman Reference Hoffman2000; Turiel Reference Turiel and Eisenberg2006), perhaps we exhibit what a theory-theorist might consider a core knowledge bias when reasoning about others. Specifically, we view good intentions and behavior as the norm and are consequently hypersensitive to deviations from it. When asked about people who act badly, we assume the deviation is what we should attend to and explain (in a Gricean fashion), and respond accordingly.
On this view, when asked whether the chairman of the board has intentionally harmed the environment by instituting a new program that he knows would create harm in addition to increasing profits, people respond affirmatively. Such a response, we think, reflects vigilant attention to someone knowingly acting in a way that causes harm. The same interpretation applies to the findings reflecting variations on the terms (“deciding,” “desire,” “advocating,” etc.). In these cases, the unacceptable, yet fully cognizant, behavior is simply assumed to be the topic of conversation. But when asked whether the chairman intentionally helped, rather than harmed, the environment, people respond differently – because helping is not a weird and worrying thing that commands vigilant attention. A question about whether helping was intended is interpreted as a query about whether the chairman had that outcome as a goal. In a world where prosocial aims are assumed to be the norm, we are not inclined to give the chairman moral credit for the outcome because the original goal, according to the narrative, was to increase profit. These asymmetrical responses do not reflect irrational reasoning; they reflect different assessments of what is the focal issue, or curiosity, to be addressed and explained. It may be that such asymmetries are universal because there is a core knowledge presupposition about the prosocial nature of people, possibly from very early in development (see Hamlin et al. Reference Hamlin, Wynn and Bloom2007). This assumption, in itself, does not mean the reasoning is illogical or unscientific, but it does influence the topic presumed to warrant explanation and discussion.
How does this interpretation account for people's judgments of causation? When both an administrative assistant and a faculty member are said to have removed pens from the receptionist's desk, the former with permission and the latter illicitly, people indicate that the professor caused the problem. This response may also reflect vigilant attention to the anomalous bad behavior, the event that requires explanation, and a corresponding shift in conversational focus. Knobe astutely notes that it is not just that the behavior is strange or unusual (see, e.g., Roxborough & Cumby Reference Roxborough and Cumby2009) but specifically that the behavior is morally bereft. So it seems it is moral culpability in general that is regarded as anomalous, not the specific actions of a certain person or even a group of people. Moreover, it does not matter whether in fact the final outcome is actually good (Hitchcock & Knobe Reference Hitchcock and Knobe2009, cited in the target article); the focus continues to be on the “bad” behavior. So – and here we think we agree with Knobe – it is not accurate to characterize what we bring to these reasoning tasks as purely a base-rate assumption regarding any specific action; rather, we simply assume that people will usually have benign intentions and act in a prosocial fashion.
In one respect, our interpretation resembles Knobe's. We agree that moral considerations are being taken into account in the reasoning reported in the literature. And it may be objected that our interpretation is simply another version of the motivational and conversational pragmatics hypotheses refuted by Knobe. But we think it is different to view people as bringing to their reasoning core content assumptions about human nature, assumptions that influence the presumed focus of conversations. The fact that people have expectations about others' prosocial and antisocial proclivities does not in itself mean that reasoning about causality in the social realm is irrational, impervious to experience, or otherwise at odds with scientific theorizing.