In the target article “Person as Scientist, Person as Moralist,” Joshua Knobe has devised an innovative model where moral appraisals play a fundamental role in how people make sense of agents and their actions. According to Knobe, people's intuitions depend on a comparison between the action under consideration and an alternative possibility, which he calls the “default position.” The default position falls somewhere along a continuum, but experimenters fail to designate its exact location. In this commentary, we contend that the default position must be fixed and clearly articulated. Otherwise, some of the subjects' intuitions seem incoherent. We agree with Knobe that there seems to be a default position against which people judge whether or not some action under consideration is favored. But we believe that his approach may be made more precise than it is by specifying clearly what the default position is.
According to Knobe's competence model, moral considerations figure into how subjects make a comparison between the action under consideration and certain alternative possibilities. It seems people who view an action as morally bad uphold an attitude at least slightly toward the con side, and people who view an action as morally good tend to have an attitude at least slightly toward the pro side. An action is favored when “the agent's attitude falls sufficiently far beyond the default” (sect. 5.2, para. 5). The core of Knobe's explanation has it that “moral judgments affect [people's] intuitions by shifting the position of the default” (sect. 5.2, para. 6, emphasis Knobe's).
Knobe's competence model has done a nice job of explaining why a majority of subjects answered the harm and help scenarios as they did (sect. 3.1). When a majority of subjects (82% according to Knobe Reference Knobe2003a) compare the chairman's attitude to the default position that harming the environment is morally bad, they favor the response that the chairman intentionally harmed the environment. Likewise, when a majority of subjects (77% according to Knobe Reference Knobe2003a) compare the chairman's attitude to the default position that helping the environment is morally good, they do not favor the response that the chairman intentionally helped the environment. Although Knobe's competence model has succeeded in explaining the majority's intuitions, his model may not succeed in explaining the intuitions of subjects who gave the minority view.
Some subjects responded that the chairman did not intentionally harm the environment (18%) or that the chairman did intentionally help the environment (23%) (Knobe Reference Knobe2003a). These represent a minority response in the harm case and help case, respectively. If Knobe's competence model is correct, then the minority's default position for the harm scenario is that harming the environment is a morally good thing. The data also suggest that the minority's default position in the help case is that helping the environment is a morally bad thing. These views are unusual and the result of applying Knobe's competence model.
Knobe may object to this assessment. Subjects receiving the harm scenario may hold that harming the environment is morally bad but the chairman's indifference does not constitute that he intentionally harmed the environment. These subjects may refrain from saying that the chairman intentionally harmed the environment because the chairman did not want to harm the environment. Subjects who responded that the chairman intentionally helped the environment may uphold the default position that helping the environment is morally good. Since the chairman knew that the program would help the environment, subjects chose the response that he intentionally helped the environment. If this is correct, Knobe is able to show why the competence model explains the minority's intuitions.
The problem with this response is that one of us (Ulatowski) collected data where two-thirds of subjects given both the harm and the help case chose minority responses (Nichols & Ulatowski Reference Nichols and Ulatowski2007). Of the two-thirds, half responded that:
(1) The chairman intentionally helped the environment and the chairman intentionally harmed the environment.
or that:
(2) The chairman did not intentionally harm the environment and the chairman did not intentionally help the environment.
On response (1), if the competence model is correct, then respondents judged that not only is harming the environment morally bad but helping the environment is morally bad, too. On response (2), if the competence model is correct, the default position for subjects is not only that helping the environment is morally good but that harming the environment is morally good, too. Subjects' default positions seem to be inconsistent and, therefore, incoherent.
We suggest that the default position be clearly defined to avoid incoherence of subjects' intuitions. In a series of experiments testing whether the distinction between doing and allowing depends on moral appraisals, we specified an alternative possibility against which subjects should compare the agent's action (Ulatowski & Johnson Reference Ulatowski and Johnson2010):
Five people are in imminent danger of death, and you are a part of a team that is taking a special train to rescue the five people. Every second counts. You have just taken over from the driver, who has gone to the back of the train to check on something. Since the train is on automatic control, you don't need to do anything to keep it going. But you can stop it by putting on the brakes. You suddenly see someone trapped ahead on the track. If you don't do anything, he will be killed (though the train will be able to continue on its way). But if you do stop, and then free the man, the rescue mission will be aborted. So you let the train continue.
We asked subjects, “Since you could have stopped the train, did you kill the man on the track?” We stipulated the default position: to stop the train. We believe that by specifying the default position, it may prevent an incoherent interpretation of people's intuitions.
Our aim in this commentary has been to expose the incoherence in subjects' responses when an experiment fails to stipulate the default position. Ultimately, we cannot assume that we know what the subjects' default position is.
You have
Access
In the target article “Person as Scientist, Person as Moralist,” Joshua Knobe has devised an innovative model where moral appraisals play a fundamental role in how people make sense of agents and their actions. According to Knobe, people's intuitions depend on a comparison between the action under consideration and an alternative possibility, which he calls the “default position.” The default position falls somewhere along a continuum, but experimenters fail to designate its exact location. In this commentary, we contend that the default position must be fixed and clearly articulated. Otherwise, some of the subjects' intuitions seem incoherent. We agree with Knobe that there seems to be a default position against which people judge whether or not some action under consideration is favored. But we believe that his approach may be made more precise than it is by specifying clearly what the default position is.
According to Knobe's competence model, moral considerations figure into how subjects make a comparison between the action under consideration and certain alternative possibilities. It seems people who view an action as morally bad uphold an attitude at least slightly toward the con side, and people who view an action as morally good tend to have an attitude at least slightly toward the pro side. An action is favored when “the agent's attitude falls sufficiently far beyond the default” (sect. 5.2, para. 5). The core of Knobe's explanation has it that “moral judgments affect [people's] intuitions by shifting the position of the default” (sect. 5.2, para. 6, emphasis Knobe's).
Knobe's competence model has done a nice job of explaining why a majority of subjects answered the harm and help scenarios as they did (sect. 3.1). When a majority of subjects (82% according to Knobe Reference Knobe2003a) compare the chairman's attitude to the default position that harming the environment is morally bad, they favor the response that the chairman intentionally harmed the environment. Likewise, when a majority of subjects (77% according to Knobe Reference Knobe2003a) compare the chairman's attitude to the default position that helping the environment is morally good, they do not favor the response that the chairman intentionally helped the environment. Although Knobe's competence model has succeeded in explaining the majority's intuitions, his model may not succeed in explaining the intuitions of subjects who gave the minority view.
Some subjects responded that the chairman did not intentionally harm the environment (18%) or that the chairman did intentionally help the environment (23%) (Knobe Reference Knobe2003a). These represent a minority response in the harm case and help case, respectively. If Knobe's competence model is correct, then the minority's default position for the harm scenario is that harming the environment is a morally good thing. The data also suggest that the minority's default position in the help case is that helping the environment is a morally bad thing. These views are unusual and the result of applying Knobe's competence model.
Knobe may object to this assessment. Subjects receiving the harm scenario may hold that harming the environment is morally bad but the chairman's indifference does not constitute that he intentionally harmed the environment. These subjects may refrain from saying that the chairman intentionally harmed the environment because the chairman did not want to harm the environment. Subjects who responded that the chairman intentionally helped the environment may uphold the default position that helping the environment is morally good. Since the chairman knew that the program would help the environment, subjects chose the response that he intentionally helped the environment. If this is correct, Knobe is able to show why the competence model explains the minority's intuitions.
The problem with this response is that one of us (Ulatowski) collected data where two-thirds of subjects given both the harm and the help case chose minority responses (Nichols & Ulatowski Reference Nichols and Ulatowski2007). Of the two-thirds, half responded that:
(1) The chairman intentionally helped the environment and the chairman intentionally harmed the environment.
or that:
(2) The chairman did not intentionally harm the environment and the chairman did not intentionally help the environment.
On response (1), if the competence model is correct, then respondents judged that not only is harming the environment morally bad but helping the environment is morally bad, too. On response (2), if the competence model is correct, the default position for subjects is not only that helping the environment is morally good but that harming the environment is morally good, too. Subjects' default positions seem to be inconsistent and, therefore, incoherent.
We suggest that the default position be clearly defined to avoid incoherence of subjects' intuitions. In a series of experiments testing whether the distinction between doing and allowing depends on moral appraisals, we specified an alternative possibility against which subjects should compare the agent's action (Ulatowski & Johnson Reference Ulatowski and Johnson2010):
Five people are in imminent danger of death, and you are a part of a team that is taking a special train to rescue the five people. Every second counts. You have just taken over from the driver, who has gone to the back of the train to check on something. Since the train is on automatic control, you don't need to do anything to keep it going. But you can stop it by putting on the brakes. You suddenly see someone trapped ahead on the track. If you don't do anything, he will be killed (though the train will be able to continue on its way). But if you do stop, and then free the man, the rescue mission will be aborted. So you let the train continue.
We asked subjects, “Since you could have stopped the train, did you kill the man on the track?” We stipulated the default position: to stop the train. We believe that by specifying the default position, it may prevent an incoherent interpretation of people's intuitions.
Our aim in this commentary has been to expose the incoherence in subjects' responses when an experiment fails to stipulate the default position. Ultimately, we cannot assume that we know what the subjects' default position is.
ACKNOWLEDGEMENTS
We would like to thank Dave Beisecker for helpful conversation leading to a draft of this commentary, and Elijah Millgram for comments on an earlier draft.