Knobe describes a number of studies in which, he claims, subjects' moral judgments influence their views about whether the actions of others were intentional, about whether an agent did something or merely allowed it to happen, and about whether an agent caused an undesirable consequence. He concludes that the exercise of competencies that humans use in making what might seem to be purely factual judgments about the world – such as judgments about causes and judgments about other agents' mental states – is “suffused with moral considerations from the very beginning” (sect. 5.3, para. 3).
Knobe suggests, very plausibly, that people's judgments about “the cause” of an event depend on a selection of relevant alternatives. His experimental evidence supports the conclusion that in some cases moral considerations partly determine this selection, although it remains an open question how wide this range of cases is. The same may well be true of judgments distinguishing between “doing” and “allowing.”
In this comment, however, I will focus on Knobe's claims regarding judgments about intentional action. Here his conclusions seem to me not to be supported by the evidence he describes, because there is an alternative interpretation of his experimental results that is more plausible than the one he proposes.
The use of intentional and its cognates involves a well-known ambiguity (see Anscombe Reference Anscombe1958, p. 9; Scanlon Reference Scanlon2008, p. 10). One sense of “intentional” is the one opposed to “unintentional.” An agent does something “intentionally” in this sense if he or she realizes that this is what he or she is doing – call an action that is intentional in this sense belief-intentional. An agent's intention in the other sense is what he or she aims at in so acting. What an agent does “intentionally” in this other sense is opposed to what he or she sees as a mere side-effect of so acting – call what is intentional in this sense aim-intentional.
The effects on the environment of the policies adopted by the chairmen in the two experiments Knobe describes are belief-intentional: the description of the cases makes clear that they are aware that the policies they choose will have these effects. But these effects are not aim-intentional: the descriptions make clear that the chairmen are indifferent to these effects, and are concerned only with profits. Given that these facts are made clear in the presentation of the scenarios, it is reasonable to believe that the subjects in each case have the same beliefs about the chairman's mental state: that the bringing about of these effects is belief-intentional but not aim-intentional. The differing answers that the subjects give to the question of whether the chairmen harmed or helped the environment “intentionally” is indeed due to moral considerations, but not in the way that Knobe suggests.
The important moral fact here is that agents are commonly open to moral criticism for bringing about bad effects when they know that these effects will occur even if they do not aim at these effects – that is to say, when they do so belief-intentionally, even if not aim-intentionally. But agents are generally held to merit moral praise or credit for bringing about good consequences only if they do so aim-intentionally. Given that the subjects see harm to the environment as a bad thing, when they are asked whether the chairman in the first scenario harmed the environment intentionally, what they are likely to ask themselves is whether the chairman's action was intentional in the sense relevant to moral criticism for bringing about such an effect (that is to say, whether it was belief-intentional). In the other case, since the subjects are likely to view helping the environment as a good thing, when they are asked whether the chairman helped the environment intentionally, what they are likely to ask themselves is whether what the chairman did was intentional in the sense relevant to moral praise or credit (that is to say, whether it was aim-intentional). What the shift from harming to helping does is not to change the subjects' interpretation of the chairman's mental states in the respective scenarios, but rather, to change the question about those mental states to one that seems to the subjects to be relevant.
This interpretation of the subjects' responses seems to me extremely plausible. It is also supported by some of the further details that Knobe mentions. For example, he reports that when subjects are asked whether “the chairman intended to harm the environment,” answers are moved strongly in a negative direction (sect. 3.2, para. 4). This is to be expected on the interpretation I propose, because the verb intend suggests (aim) intention more strongly than does the adverb intentionally.
My interpretation also explains why subjects disagree with the claim that an agent was “in favor of” a morally good outcome but are neutral on the question of whether the agent was “in favor of” a morally bad outcome (sect. 3.2, para. 5). This is because the agent fails to favor the morally good outcome in the way relevant to moral praise or credit; but, even if he or she does not actively favor the morally bad outcome, an agent who is perfectly willing to bring about that outcome for some other reason is more favorably disposed toward it than he or she should be, and therefore open to some criticism on this score.
Knobe describes a number of studies in which, he claims, subjects' moral judgments influence their views about whether the actions of others were intentional, about whether an agent did something or merely allowed it to happen, and about whether an agent caused an undesirable consequence. He concludes that the exercise of competencies that humans use in making what might seem to be purely factual judgments about the world – such as judgments about causes and judgments about other agents' mental states – is “suffused with moral considerations from the very beginning” (sect. 5.3, para. 3).
Knobe suggests, very plausibly, that people's judgments about “the cause” of an event depend on a selection of relevant alternatives. His experimental evidence supports the conclusion that in some cases moral considerations partly determine this selection, although it remains an open question how wide this range of cases is. The same may well be true of judgments distinguishing between “doing” and “allowing.”
In this comment, however, I will focus on Knobe's claims regarding judgments about intentional action. Here his conclusions seem to me not to be supported by the evidence he describes, because there is an alternative interpretation of his experimental results that is more plausible than the one he proposes.
The use of intentional and its cognates involves a well-known ambiguity (see Anscombe Reference Anscombe1958, p. 9; Scanlon Reference Scanlon2008, p. 10). One sense of “intentional” is the one opposed to “unintentional.” An agent does something “intentionally” in this sense if he or she realizes that this is what he or she is doing – call an action that is intentional in this sense belief-intentional. An agent's intention in the other sense is what he or she aims at in so acting. What an agent does “intentionally” in this other sense is opposed to what he or she sees as a mere side-effect of so acting – call what is intentional in this sense aim-intentional.
The effects on the environment of the policies adopted by the chairmen in the two experiments Knobe describes are belief-intentional: the description of the cases makes clear that they are aware that the policies they choose will have these effects. But these effects are not aim-intentional: the descriptions make clear that the chairmen are indifferent to these effects, and are concerned only with profits. Given that these facts are made clear in the presentation of the scenarios, it is reasonable to believe that the subjects in each case have the same beliefs about the chairman's mental state: that the bringing about of these effects is belief-intentional but not aim-intentional. The differing answers that the subjects give to the question of whether the chairmen harmed or helped the environment “intentionally” is indeed due to moral considerations, but not in the way that Knobe suggests.
The important moral fact here is that agents are commonly open to moral criticism for bringing about bad effects when they know that these effects will occur even if they do not aim at these effects – that is to say, when they do so belief-intentionally, even if not aim-intentionally. But agents are generally held to merit moral praise or credit for bringing about good consequences only if they do so aim-intentionally. Given that the subjects see harm to the environment as a bad thing, when they are asked whether the chairman in the first scenario harmed the environment intentionally, what they are likely to ask themselves is whether the chairman's action was intentional in the sense relevant to moral criticism for bringing about such an effect (that is to say, whether it was belief-intentional). In the other case, since the subjects are likely to view helping the environment as a good thing, when they are asked whether the chairman helped the environment intentionally, what they are likely to ask themselves is whether what the chairman did was intentional in the sense relevant to moral praise or credit (that is to say, whether it was aim-intentional). What the shift from harming to helping does is not to change the subjects' interpretation of the chairman's mental states in the respective scenarios, but rather, to change the question about those mental states to one that seems to the subjects to be relevant.
This interpretation of the subjects' responses seems to me extremely plausible. It is also supported by some of the further details that Knobe mentions. For example, he reports that when subjects are asked whether “the chairman intended to harm the environment,” answers are moved strongly in a negative direction (sect. 3.2, para. 4). This is to be expected on the interpretation I propose, because the verb intend suggests (aim) intention more strongly than does the adverb intentionally.
My interpretation also explains why subjects disagree with the claim that an agent was “in favor of” a morally good outcome but are neutral on the question of whether the agent was “in favor of” a morally bad outcome (sect. 3.2, para. 5). This is because the agent fails to favor the morally good outcome in the way relevant to moral praise or credit; but, even if he or she does not actively favor the morally bad outcome, an agent who is perfectly willing to bring about that outcome for some other reason is more favorably disposed toward it than he or she should be, and therefore open to some criticism on this score.