Virtually everyone, including both professional and lay evaluators of human behavior, agrees that to praise or blame an agent, the agent must have acted intentionally, with foresight of the consequences, and must have caused the outcome. In a perfect evaluative world, assessments of intentionality, foresight, and causation would be made independently of the judge's expectations, affective and attitudinal reactions, and moral beliefs and predilections. By now, however, there is abundant evidence that such factors have a powerful and pervasive influence on intentionality, foresight, and causation judgments. In his target article, Knobe offers an alternative to the view that such influences are motivated by the desire to justify praising or blaming an agent who evokes positive or negative evaluative reactions.

Knobe has argued that the reason that concepts such as intentionality, causation, and foresight are influenced by moral considerations is because these concepts are suffused with moral considerations (i.e., moral considerations figure into the underlying competencies). To illustrate his position, he uses his well-traveled example of an executive who knows that, as a side effect of initiating a certain program, the environment will be helped or harmed, but whose only concern is to increase profits. Knobe assumes that in the “help” scenario, the normative (or “default”) expectation would be to have at least a moderately pro-attitude toward helping the environment. The default expectation in the “harm” scenario would be for a moderately anti-attitude (or in Knobe's term, con-attitude) toward harming environment. In the “help” case, the executive doesn't meet the threshold required for having a pro-attitude; whereas in the “harm” case, the executive's failure to endorse a con-attitude places him in the range of plausibly having a pro-attitude. Thus, in the latter case but not the former, the executive is thought to have acted intentionally because he apparently has a pro-attitude toward the harmful outcome. Ascriptions of intentionality are thus dependent on the sort of attitude an evaluator thinks an agent should have about a particular outcome. Intentionality (and presumably other concepts such as causation and foresight) is applied when the agent's presumed attitude crosses the evaluator's normative threshold.

Our alternative to Knobe's position – the culpable control model of blame (Alicke Reference Alicke1992; Reference Alicke2000; Alicke & Zell Reference Alicke and Zell2009; Alicke et al. Reference Alicke, Rose and Bloom2010) – can explain the environmental harming/helping findings without positing that concepts such as intentionality are inherently moral. Further, it can explain cases other than the specialized side-effect scenarios that are the focus of Knobe's theory. The culpable control model assumes that positive and negative evaluative reactions – which are judgments of right or wrong, good or bad, or approval or disapproval – to the people involved in an event, their actions, and the outcomes that ensue can induce social perceivers to process information in a “blame validation” mode. Blame validation involves interpreting the evidence regarding intentionality, causation, and foresight in a way that justifies praising an agent who elicits positive evaluations or blaming one who arouses negative evaluations.

The culpable control model explains Knobe's findings by assuming that social perceivers view the environment-harming executive as a major jerk (i.e., one who arouses strong negative evaluations), but view the environment-helping executive as only a minor one. Imputing intentionality to the environment-harming executive, therefore, validates social perceivers' negative evaluative reactions, and in turn, supports a blame attribution. The culpable control model, therefore, does not require the assumption that concepts such as intentionality, causation, and foresight are suffused with moral considerations. Rather, the influences of these evaluations can be explained in terms of the desire to blame an agent whose actions arouse strong disapproval.

It is important to note that most of Knobe's examples apply to cases where foreseen but unintended side-effects occur, which narrows the application of the theory. Furthermore, Knobe's example is not an optimal one for considering the relative merits of his position and the culpable control model because his default assumption regarding an agent's attitude is confounded with positive and negative evaluations of the agent's goals and actions and the outcomes that occur. For example, in the “harm” scenario, the agent's indifference toward harming the environment diverges from the attitude we would expect most agents to have in this situation; but this indifference, as well as both the agent's decision to let the environment be harmed and the fact that the environment ultimately is harmed, also provides a basis for negative evaluative reactions. So, we need some other case to differentiate these two views.

In an early set of studies (Alicke Reference Alicke1992, Study 1), participants learned that a young man was speeding home to hide either an anniversary present or a vial of cocaine from his parents before they arrived home. A car accident occurred under somewhat ambiguous circumstances: It could have been due in part to his speeding, but also to environmental impediments such as a partly obscured stop sign. The study's results were clear: When the driver's motive was undesirable (i.e., to hide cocaine), his driving was cited as far more causal than the environmental obstacles. However, precisely the opposite was true when his motive was to hide an anniversary present. This case would be very difficult for Knobe's theory to explain: Why would anyone assume that the driver who was speeding home to hide cocaine had a pro-attitude towards causing car accidents, and that the driver who was speeding to hide an anniversary present did not? The more plausible alternative, based on the culpable control model, is that negative evaluations of the driver whose motive was to hide cocaine induced participants to skew the evidential criteria for causation to support their desire to blame him.

In sum: Knobe's position is plausible in cases involving foreseen but unintended side-effects, but it has some trouble explaining cases outside of this narrow scope. The culpable control model can explain many cases involving side-effects, as well as most cases that do not involve side-effects, and can do so without claiming that concepts such as intentionality, causation, and foresight are suffused with moral considerations.