Playing offense: The rules of the blame game

Appears in Baited Area #3

I. Foreseeing offense

Here’s an unfortunate coincidence: Neige, the most prominent filler word in Mandarin Chinese, equivalent to “like” or “you know” in English, sounds exactly like “nigga.” After moving to China, I was shocked to hear people peppering their speech with the word. I knew that sooner or later I was bound to see this coincidence result in a cultural train wreck. In late 2020, the train derailed when USC business professor Greg Patton used this word as an example of a filler word in a communications course.[1] Predictably, the shit hit the fan, offended students calling for censure. It’s clear that Patton caused offense; no one is denying this. The controversy, however, surrounds a different question: Is Patton blameworthy for the offense he caused?

Whatever your gut reaction to this question, my aim here is to convince you that your intuition regarding this case is not all too trustworthy. Our psychological processes for attributing blame are systematically skewed by cognitive biases. In what follows, I’ll describe the processes by which negative emotions can lead us to rationalize either blame or excuses for offense, independent of evidence.

Blame judgments rely crucially on judgments of ability and foresight. This nuance is often lost in public debates on the matter, such debates often devolving into rehashed arguments over intent vs. impact. Some, on the one hand, will argue that offenders should not be blameworthy because they did not intend the offense. Take, for instance, Pamela Taylor, a West Virginia non-profit director who in 2016 called Michelle Obama an “ape in heels” on Facebook. According to Taylor, “My comment was not intended to be racist at all.”[2] Even supposing we believe her, is this lack of intent enough to absolve her of blame? Clearly not. People can be held blameworthy for all kinds of harm they didn’t intend. With the exception of the mohawked pig in Captain Planet, few polluters actually intend to pollute; rather, their pollution is a negligent byproduct of their desire to make duckets at all costs. Hence, malintent is not a necessary condition of blameworthiness.

Others, on the other hand, will retort that offenders are blameworthy simply in virtue of the negative impact of their offense. Causing harm by itself, however, is not sufficient to render blame, as blameworthiness is not independent of the offender’s mental states. To take an extreme case, suppose I’m walking through the grocery store, minding my own business, when suddenly I step on a button carefully concealed by a supervillain that kills 1000 people in Hoboken. Despite the negative impact of my actions, I clearly can’t be held to blame. Similarly for offense, it would be pretty hard to blame a Chinese tourist causing offense for saying neige. By itself, causing harm is not enough for blameworthiness

This Scylla of intent and Charybdis of impact offers a false choice. While intent is not necessary and impact is not sufficient to ascribe blame, the ability to have done otherwise is routinely judged as necessary for a person to be blameworthy.[3] In other words, if you are blameworthy for something, you must have been able to have done otherwise, a principle traceable back to Aristotle.[4] For this reason, mental incompetence is an acceptable legal defense; if mental incompetence renders a person unable to choose their actions, that person cannot be held to blame for their actions. When the evil villain tricks me into stepping on a button that kills people, I can’t be held accountable for my actions if I couldn’t have done otherwise.

More germane to offense, the ability to foresee is a special case of the type of ability required for blameworthiness. If a person does not have sufficient evidence to predict an event, then that person can’t be expected to act as to bring about or avoid that event. First, this Principle of Foresight explains why we tend to hold Taylor accountable for calling Michelle Obama an ape in heels. Even if she did not intend to cause harm, it’s pretty implausible to think that she could not have foreseen her comment causing offense. Critics would ask what rock she’s been living under to render her ignorant of the historically racist connotations of apes. It’s the ability to foresee, rather than the intent, that is relevant to blame. This principle also explains why we typically wouldn’t blame Chinese tourists for saying neige. How the hell would they know that their filler word sounds like a racial slur in English?

Hence, neither intent nor impact are deciding factors in the neige case. Rather, we want to know whether Patton could have foreseen the offense caused. If, in conjunction with other factors, he could have foreseen the offense, then he should be held to account, regardless of whether he intended to cause offense. If he could not have foreseen the offense, then he cannot be held accountable.

If the ability to foresee harm is a necessary condition of being blameworthy for causing harm, then there is a straightforward inferential path to blame. A little logic is in order. To say that Q is a necessary condition of P is to say that if P is true, then Q must be true. If someone is to blame for causing harm, then they must have been able to foresee the harm. The inference to ability requires an inference known as modus tollens: If P then Q, and not Q, so not P. Thus, if someone could not have foreseen harm that they caused, we can infer that they can’t be held to blame. The inference would look something like this:

1. If a person is to blame for the consequences of their actions, then this person must have been able to foresee the consequences of their actions. (Foresight Principle)

2. Greg Patton was not able to foresee causing offense by saying “neige.”


3. Greg Patton is not to blame for causing offense by saying “neige.”

Even if we disagree about the truth of premise 2, we can all agree that this is a valid argument: Were premise 2 true, the conclusion would follow. We also tend to agree with premise 1, the Foresight Principle.

How should we go about evaluating the truth of premise 2? Deciding whether one could have foreseen the consequences of one’s actions, however, is not as easy as it may seem. Even supposing there is a fact of the matter, we tend to have conflicting intuitions on foresight, depending on factors we’ll soon discuss. Two different individuals may be presented with the same facts and come to two radically different conclusions. One individual may look at logically identical yet superficially dissimilar cases and come to two radically different conclusions.

To illustrate, let’s look at two similar cases. In late 2018, Republican gubernatorial candidate Ron DeSantis asked Florida voters not to “monkey this up” by electing his opponent, Democrat Andrew Gillum.[5] As Gillum is Black, many took offense to this comment, given the historically racist usage of the word “monkey.” Several months later, Minnesota Democratic Representative Ilhan Omar tweeted “It’s all about the Benjamins baby” in reference to Israel.[6] Considering stereotypes concerning Jews and money, many took this tweet to be anti-Semitic. Discounting the mind-reading necessary to accuse them of intentional dog-whistles, could either DeSantis or Omar have foreseen the offense they caused? It isn’t at all clear. Much of the uproar in this case involved clashing intuitions on what DeSantis or Omar could have foreseen. To some, DeSantis clearly should have been able to put two and two together, connecting the word “monkey” with its racist connotations. To others, it’s absurd to expect people to make these obscure semantic connections. The same with Omar.

So far, we’ve seen that the Foresight Principle can be used to determine when a person cannot be held blameworthy for their actions, causing offense in particular. We first start with information concerning a person’s ability to have foreseen the consequences of their actions. We can then use this information to make a modus tollens inference, determining when a person cannot be held blameworthy.

It would seem a tad loony, however, to drive the inference in the other direction, an inference known as modus ponens: If P then Q, and P, so Q. Applying this inference to the Foresight Principle, the reasoning would go as follows:

1. If a person is to blame for the consequences of their actions, then this person must have been able to foresee the consequences of their actions.

2. Greg Patton is to blame for causing offense by saying “neige.”


3. Greg Patton must have been able to foresee causing offense by saying “neige.”

We don’t think it particularly rational to first judge whether a person is blameworthy and as a result infer that this person must have been able to foresee the consequences of their actions. It makes little sense to say that I could have foreseen the consequences of my actions because I’m to blame for them. The inference should go the other way: I’m not to blame because I couldn’t foresee the consequences. As it turns out, this is exactly what we do.

II. Blame biases

All too often, we do make the inference from judgments of blame to judgments of ability, including foresight. This sort of inferential process is known as blame validation, a process that features a “proclivity to favor blame versus nonblame explanations for harmful events and to de-emphasize mitigating circumstances.”[7] This psychological process consists of two parts. First, we evaluate a harmful event, such as offense caused. One of several biases may then lead us to attribute blame to someone before evaluating evidence of ability or foresight. Suppose someone steps on your foot. You fly into a rage, immediately casting censure upon the transgressor. In the second step, we rationalize how our object of blame could have had the necessary ability or foresight to prevent the harm caused. In your blind rage, you tell yourself that the transgressor surely could have seen your foot. You think of all the ways that this person could have avoided your foot and ignore any evidence that that this person might not have been able to see your foot. Your anger, in short, drives your blame, which in turn creates a need to judge that a person could have foreseen certain consequences.

Inferences from judgments of blame to judgments of ability are driven by cognitive biases, defined as “cases in which human cognition reliably produces representations that are systematically distorted compared to some aspect of objective reality.”[8] This process begins with emotional judgments, many of which are flying under our conscious radar. We start off making spontaneous evaluations, or “affective reactions to features of harmful events and the people involved that influence blame attributions.”[9] In other words, we make judgments of blame in response to our emotional evaluations of events.

One such bias is the outcome bias. With respect to blame, an outcome bias occurs when the outcome of a person’s action influences our judgments of blame independent of the person’s ability to influence or foresee the outcome. In one study, participants read two different vignettes about Peter Garness, a man who comes home to find what he believes to be an intruder.[10] After finding the alleged intruder in his daughter’s room, Garness shoots and kills the intruder. In one version of the vignette, the intruder turns out to be his daughter’s boyfriend. In another version, the intruder turns out to be a dangerous burglar. Even though Garness had the same evidence in both conditions, participants that read the boyfriend condition tended to blame Garness, whereas those that read the burglar condition did not. This study provides evidence that we are biased to blame actors when the outcomes of their actions arouse negative affect,[11] independent of the evidence the actor had at the time.

In addition to negative outcomes, negative motives can also bias us towards blame. In another study, participants read about John, some schmendrick who gets in a car accident when speeding.[12] In one version of the vignette, John was speeding home to hide an anniversary present from his parents. In a different version, John was speeding home to hide a vial of cocaine. Participants in the cocaine condition tended to heap a greater degree of responsibility on John, even though his motive for speeding had little to do with his responsibility in the accident. Hence, negative motives drive us blame judgments independent of ability or foresight assessments.

Victim attributes may also lead to blame judgments via affective reactions. After learning victims of a shooting were gang members, participants tended to let the shooter off the hook, in comparison with participants who read that the victims were star athletes.[13] In fact, victim information that arouses negative affect can lead to victim blame, including information about victim likeability and demographics. We see victim attribute effects at play in attempts to blame victims of sexual assault when, for instance, people tend to alleviate the perpetrator of blame depending on factors such as the victim’s clothing.[14]

The next inference in the process is from judgments of blame to judgements of ability. After negative affective reactions cause us to blame, we retroactively figure out how a person could have done or known otherwise. Here is how it works: In our heads, we know that blameworthiness entails the ability to have done otherwise. We cannot consistently hold this principle and simultaneously judge that a person is blameworthy for something they could not have controlled or foreseen. When claims are inconsistent, the conjunction of these claims can be used to derive a contradiction by deductive inference. The following three premises are inconsistent:

1. If a person is to blame for the consequences of their actions, then this person must have been able to foresee the consequences of their actions.

2. Greg Patton is to blame for causing offense by saying “neige.”

3. Greg Patton was not able to foresee causing offense by saying “neige.”

If we accept premises 1 and 2, we can derive the negation of premise 3 via modus ponens. Alternatively, we could accept premises 1 and 3 and derive the negation of premise 2 via modus tollens.

To be consistent while retaining our blame judgments, we must figure out how the perpetrator could have done otherwise. In the burglar/boyfriend study, participants in the boyfriend condition where almost two and a half times as likely as those in the burglar condition to judge that Garness “could have acted differently.” In the anniversary gift/cocaine study, participants in the cocaine condition were much more likely to judge that John had a greater degree of causal control over the car accident.

The influence of blame judgments on judgments of ability requires us to reevaluate evidence, perhaps through shifting our standards of evidence or searching harder for information that confirms our beliefs.[15] When victim information creates negative affect, even a small degree of control is enough for us to claim that the victim could have prevented their misfortune. Think of sexual assault victims who are judged responsible because of what they wore. Concerning confirmation bias, participants who read that shooting victims were gang members were much less likely to read testimony that supported the prosecution. To justify our attributions of blame, we sit around figuring out how a person could have had the ability or foresight to have done otherwise.

III. Gerrymandering blame

While the evidence presented so far has concerned blame in general, the same lessons apply to the narrow case of offense. We began by looking at the muddiness surrounding questions of foresight with respect to offense. Could Patton have foreseen that using the word neige would have caused such distress? What about DeSantis and Omar? Even if there is a fact of the matter, ascertaining that fact of the matter is no easy task.

Add to this muddiness the cognitive biases discussed above and you have a recipe for disaster. For any offense caused, you can always retroactively figure out how the offender could and should have foreseen the offense. Start with a negative affective evaluation. You’re offended. In the first step of the process, you lay blame in response to the negative emotion, the offense itself. But you’re not done there. To quell any feelings of inconsistency, you’ll need to convince yourself that they could have foreseen your offense. After all, it’s pretty obvious to you that what they said was offensive.

This process can lead us to hold others to “Superman” standards. Through influences like biased information searches or confirmation bias, we can always figure out how a person could have had the ability to control or foresee an outcome, expanding what is humanly possible to the standard of Superman. When we’re irrationally angry, we sit there and stew, running through all the ways a person could have controlled or foreseen an outcome. Why didn’t they do this? Why didn’t they think of that? Why couldn’t they have flown around the world, spinning it to reverse time? In the grips of affect, all is possible.

This blame process works also in reverse, not to inculpate but to exculpate. When you are being blamed, it’s easy to figure out how you could not have known otherwise: “How could I have known that saying neige would offend people?” Rather than conducting a preferential information search for evidence of you could have foreseen offense, you run an information search for purely exculpatory evidence, evidence that you could not have foreseen offense.

This exculpatory process works not only when you are being blamed, but when people you can identify with are being blamed. Can you see yourself in their shoes? When you read about a sexual assault case, your intuitions can by influenced by whether you are more like the accused or the victim.[16] Those who identify more with the accused may wrack their brains to figure out how the victim rather than the accused is to blame. The victim should have worn different clothes. The victim shouldn’t have been walking alone at night. Here, the victim is held to Superman standards to exculpate the alleged perpetrator.

With these considerations in mind, it is easy to see how two different people may come to the exact opposite conclusions in either DeSantis’ “monkey this up” case or Omar’s “all about the Benjamins” case. Similarly, the same person might look at both cases and come to radically different conclusions. Those with particular political proclivities will look at the DeSantis case and wonder “How could he have predicted that people would have found the word ‘monkey’ offensive?” Others would say the same about Omar and the Benjamins.

From a certain angle, those defending Patton’s use of neige are playing dumb, a little like Pamela Taylor claiming she couldn’t have known calling Michelle Obama an ape in heels would have been found offensive. From a different angle, it’s absurd to expect Patton to foresee that an interlingual coincidence would cause offense. If you can see yourself in the shoes of the offender, your cognitive biases will drive you to exculpate, figuring out how they could not have foreseen the offense. If it is easy to see yourself in the shoes of Greg Patton, it’s easy to convince yourself that there’s no way he could have seen the offense caused. On the other hand, if you can see yourself in the shoes of the offended, your cognitive biases will drive you in the other direction, imbuing the transgressor with “Superman” qualities.

Despite its deficiencies, blame is one of our most crucial social tools. Blame serves as a basis for apportioning censure, the purpose of which is to train people in what not to do. We hold children responsible for their mistakes to prevent them from repeating these mistakes. In social learning, children need to be trained not to cause offense, whether the offense was intended or simply careless. And neither children nor adults can be trained to avoid consequences they can’t foresee. Hence, the Foresight Principle is a valuable device for ensuring that blame serves the purpose of teaching. All too often, however, we use blame not to teach others, but for our own selfish purposes. We also tend to irrationally reject blame to avoid censure. Both of these motivations lead to abuse of the Foresight Principle, appeasing the emotions by gerrymandering what a person could have foreseen. The potential for abusing the Foresight Principle calls for a second regulatory principle, one intended to prevent this abuse. Perhaps we need a principle of charity, requiring us to second-guess our intuitions regarding blame and offense.

[1] Yeung, J. (2020, Sept. 10). USC professor under fire after using Chinese expression students allege sounds like English slur.

[2] McGee, J. (2016, Nov. 14). Official suspended after racist Obama post to return to job. WSAZ News Channel 3.

[3] Buckwalter, W., & Turri, J. (2015). Inability and obligation in moral judgment. PLoS One, 10(8), e0136589.

[4] Nicomachean Ethics, 1109b30

[5] Jacobs, J. (2018, Aug. 29). DeSantis Warns Florida Not to ‘Monkey This Up,’ and Many Hear a Racist Dog Whistle. The New York Times.

[6] Nelson, C. (2019, Mar. 7). Minnesota Congresswoman Ignites Debate On Israel And Anti-Semitism. NPR.

[7] Alicke, M. D. (2000). Culpable control and the psychology of blame. Psychological bulletin, 126(4), 556.

[8] Haselton, M. G., Nettle, D., & Murray, D. R. (2015). The evolution of cognitive bias. The handbook of evolutionary psychology, 1–20.

[9] Alicke, 2000

[10] Alicke, M. D., & Davis, T. L. (1989). The role of a posteriori victim information in judgments of blame and sanction. Journal of Experimental Social Psychology, 25(4), 362–377.

[11] The term “affect” refers to the feeling underlying emotions and moods. For present purposes, it can be treated as synonymous with “emotion.”

[12] Alicke, M. D. (1992). Culpable causation. Journal of personality and social psychology, 63(3), 368.

[13] Alicke, 2000

[14] Janoff-Bulman, R., Timko, C., & Carli, L. L. (1985). Cognitive biases in blaming the victim. Journal of Experimental Social Psychology, 21(2), 161–177.

Grubb, A., & Harrower, J. (2008). Attribution of blame in cases of rape: An analysis of participant gender, type of rape and perceived similarity to the victim. Aggression and Violent Behavior, 13(5), 396–405.

[15] Alicke, 2000

[16] Grubb & Harrower, 2008



Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store