Ever since Aristotle's Nicomachean Ethics, there has been debate over the extent to which there is separation between morality and cognition. We applaud Knobe's modern effort to integrate the investigation of these areas. There are three main reasons, however, to doubt his thesis according to which moral evaluations affect the ordinary understanding of social and psychological phenomena “from the very beginning” (target article, sect. 5.3, para. 3).
First, moral and non-moral evaluations of the social world do not always work together. In particular, in support of the position that Theory of Mind reasoning is not theory-like and does not proceed in terms of a process that can be characterized in terms of “child-as-scientist” (Leslie et al. Reference Leslie, Friedman and German2004), preverbal infants appear to possess basic mind-reading skills (e.g., Surian et al. Reference Surian, Caldi and Sperber2007). No evidence suggests that such acquisition depends on input from moral competencies. Moreover, children with selective impairments of mind-reading skills appear to have an intact ability to make some basic moral judgments (Blair Reference Blair1996; Leslie et al. Reference Leslie, Mallon and DiCorcia2006b).
Second, even when moral evaluations appear to shape ordinary intuitions about the social world, non-moral considerations are a necessary input to the shaping of these intuitions. We have found that both adults (Pellizzoni et al. Reference Pellizzoni, Girotto and Surian2010) and preschoolers (Pellizzoni et al. Reference Pellizzoni, Siegal and Surian2009) attribute intentionality to a negative side-effect produced by an agent who was not aware of it. By contrast, participants did not do this when the agent was described as having a false belief about the negative side-effect. When the side-effect was positive, participants judged that it had been produced unintentionally, regardless of whether the agent believed that it could occur or not. Thus, evaluative considerations interact with belief ascription in determining intentionality judgment.
Third, counterfactual thinking affects moral evaluations, rather than vice versa. We have attributed the above-described results to a differential availability of plausible counterfactual alternatives that undo the negative side-effect. When individuals read about an agent who did not know that his action could produce a negative side-effect, they could easily think, “Had he made an inquiry, he would have discovered the side-effect and made a different choice.” Indeed, when readers undo the negative outcome of a story, they alter the protagonist's choices (Girotto et al. Reference Girotto, Legrenzi and Rizzo1991). But when a misinformed agent had no reason to anticipate a negative side-effect, individuals could not easily imagine a plausible alternative (e.g., ”Had he imagined that he was misinformed, he would have made a different choice”).
Counterfactual thinking seems to play an important role in Knobe's thesis, too: Individuals attribute intentionality to a negative but not to a positive side-effect because they tend to construct alternatives that are morally right (“If the agent had chosen differently, he might have produced a positive side-effect”) rather than morally wrong (“If the agent had chosen differently, he might have produced a negative side-effect”). The problem with Knobe's interpretation that moral evaluations determine the selection of counterfactuals is that it does not explain how counterfactuals are generated or how counterfactuals affect other mental activities, including moral judgment.
His interpretation neglects the finding that individuals do construct morally dubious alternatives. For example, they imagine breaking the constitutive rules of a game in order to undo a failure (e.g., Girotto et al. Reference Girotto, Ferrante, Pighin and Gonzalez2007). When applied to intentionality attribution, Knobe's interpretation appears to confuse morality with normality. According to Knobe, individuals who read the positive side-effect story do not imagine the chairman damaging the environment because this possibility is morally wrong. We would say that they don't do so simply because this possibility alters normal events, that is, the normal tendency of chairmen to seek to make profits (Kahneman & Miller Reference Kahneman and Miller1986; Uttich & Lombrozo Reference Uttich and Lombrozo2010). With regard to the effects of counterfactual thinking, Knobe's interpretation neglects the finding that moral evaluations often depend on the availability of counterfactual alternatives. For example, individuals attribute more compensation to the victim of an accident (and more responsibility to the perpetrator) when it is preceded by exceptional rather than by normal circumstances (Macrae Reference Macrae1992). Finally, Knobe's interpretation cannot easily explain our results: The possible alternatives evoked by the non-informed agent version (e.g., ”Had he made an inquiry…”) were not morally different from those evoked by the misinformed agent version (e.g., ”Had he imagined that he was misinformed…”). Yet, only in the first case did individuals attribute intentionality to the negative side-effect (Pellizzoni et al. Reference Pellizzoni, Girotto and Surian2010).
To investigate the relations between the moral and non-moral facets of naïve psychology remains a high priority for future research. However, in this connection, it is not necessary to postulate that moral evaluations play a pervasive role in the ordinary understanding of intentional actions.
You have
Access
Ever since Aristotle's Nicomachean Ethics, there has been debate over the extent to which there is separation between morality and cognition. We applaud Knobe's modern effort to integrate the investigation of these areas. There are three main reasons, however, to doubt his thesis according to which moral evaluations affect the ordinary understanding of social and psychological phenomena “from the very beginning” (target article, sect. 5.3, para. 3).
First, moral and non-moral evaluations of the social world do not always work together. In particular, in support of the position that Theory of Mind reasoning is not theory-like and does not proceed in terms of a process that can be characterized in terms of “child-as-scientist” (Leslie et al. Reference Leslie, Friedman and German2004), preverbal infants appear to possess basic mind-reading skills (e.g., Surian et al. Reference Surian, Caldi and Sperber2007). No evidence suggests that such acquisition depends on input from moral competencies. Moreover, children with selective impairments of mind-reading skills appear to have an intact ability to make some basic moral judgments (Blair Reference Blair1996; Leslie et al. Reference Leslie, Mallon and DiCorcia2006b).
Second, even when moral evaluations appear to shape ordinary intuitions about the social world, non-moral considerations are a necessary input to the shaping of these intuitions. We have found that both adults (Pellizzoni et al. Reference Pellizzoni, Girotto and Surian2010) and preschoolers (Pellizzoni et al. Reference Pellizzoni, Siegal and Surian2009) attribute intentionality to a negative side-effect produced by an agent who was not aware of it. By contrast, participants did not do this when the agent was described as having a false belief about the negative side-effect. When the side-effect was positive, participants judged that it had been produced unintentionally, regardless of whether the agent believed that it could occur or not. Thus, evaluative considerations interact with belief ascription in determining intentionality judgment.
Third, counterfactual thinking affects moral evaluations, rather than vice versa. We have attributed the above-described results to a differential availability of plausible counterfactual alternatives that undo the negative side-effect. When individuals read about an agent who did not know that his action could produce a negative side-effect, they could easily think, “Had he made an inquiry, he would have discovered the side-effect and made a different choice.” Indeed, when readers undo the negative outcome of a story, they alter the protagonist's choices (Girotto et al. Reference Girotto, Legrenzi and Rizzo1991). But when a misinformed agent had no reason to anticipate a negative side-effect, individuals could not easily imagine a plausible alternative (e.g., ”Had he imagined that he was misinformed, he would have made a different choice”).
Counterfactual thinking seems to play an important role in Knobe's thesis, too: Individuals attribute intentionality to a negative but not to a positive side-effect because they tend to construct alternatives that are morally right (“If the agent had chosen differently, he might have produced a positive side-effect”) rather than morally wrong (“If the agent had chosen differently, he might have produced a negative side-effect”). The problem with Knobe's interpretation that moral evaluations determine the selection of counterfactuals is that it does not explain how counterfactuals are generated or how counterfactuals affect other mental activities, including moral judgment.
His interpretation neglects the finding that individuals do construct morally dubious alternatives. For example, they imagine breaking the constitutive rules of a game in order to undo a failure (e.g., Girotto et al. Reference Girotto, Ferrante, Pighin and Gonzalez2007). When applied to intentionality attribution, Knobe's interpretation appears to confuse morality with normality. According to Knobe, individuals who read the positive side-effect story do not imagine the chairman damaging the environment because this possibility is morally wrong. We would say that they don't do so simply because this possibility alters normal events, that is, the normal tendency of chairmen to seek to make profits (Kahneman & Miller Reference Kahneman and Miller1986; Uttich & Lombrozo Reference Uttich and Lombrozo2010). With regard to the effects of counterfactual thinking, Knobe's interpretation neglects the finding that moral evaluations often depend on the availability of counterfactual alternatives. For example, individuals attribute more compensation to the victim of an accident (and more responsibility to the perpetrator) when it is preceded by exceptional rather than by normal circumstances (Macrae Reference Macrae1992). Finally, Knobe's interpretation cannot easily explain our results: The possible alternatives evoked by the non-informed agent version (e.g., ”Had he made an inquiry…”) were not morally different from those evoked by the misinformed agent version (e.g., ”Had he imagined that he was misinformed…”). Yet, only in the first case did individuals attribute intentionality to the negative side-effect (Pellizzoni et al. Reference Pellizzoni, Girotto and Surian2010).
To investigate the relations between the moral and non-moral facets of naïve psychology remains a high priority for future research. However, in this connection, it is not necessary to postulate that moral evaluations play a pervasive role in the ordinary understanding of intentional actions.