Do Moral Violations Require A Victim?

If you’ve ever been a student of psychology, chances are pretty good that you’ve heard about or read a great many studies concerning how people’s perceptions about the world are biased, incorrect, inaccurate, erroneous, and other such similar adjectives. A related sentiment exists in some parts of the morality literature as well. Perhaps the most notable instance is the unpublished paper on moral dumbfounding, by Haidt, Bjorklund, & Murphy (2000). In that paper, the authors claim to provide evidence that people first decide whether an act is immoral and then seek to find victims or harms for the act post hoc. Importantly, the point seems to be that people seek out victims and harm despite them not actually existing. In other words, people are mistaken in perceiving harm or victims. We could call such tendencies the “fundamental victim error” or the “harm bias”, perhaps. If that interpretation of the results is correct, it would carry a number of implications, chief among which (for my present purposes) is that harm is not a required input for moral systems. Whatever cognitive systems are in charge of processing morally-relevant information, they seem to be able to do so without knowledge of who – if anyone – is getting harmed.

Just a little consensual incest. It’s not like anyone is getting hurt.

Now I’ve long found that implication to be a rather interesting one. The reason it’s interesting is because, in general, we should expect that people’s perceptions about the world are relatively accurate. Not perfect, mind you, but we should be expected to be as accurate as available information allows us to be. If our perceptions weren’t generally accurate, this would likely yield all sorts of negative fitness consequences: for example, believing you can achieve a goal you actually cannot could lead to the investment of time and resources in a fruitless endeavor; resources which could be more profitably spent elsewhere. Sincerely believing you’re going to win the lottery does not mean the tickets are wise investments. Given these negative consequences for acting on inaccurate information, we should expect that our perceptual systems evolved to be as accurate as they can be, given certain real-world constraints.

The only context I’ve seen in which being wrong about something could consistently lead to adaptive outcomes is in the realm of persuasion. In this case, however, it’s not that being wrong about something per se helps you, as much as someone else being wrong helps you. If people happen to think my future prospects are bright – even if they’re not – it might encourage them to see me as an attractive social partner or mate; an arrangement from which I could reap benefits. So, if some part of me happen to be wrong, in some sense, about my future prospects, and being wrong doesn’t cause me to behave in too many maladaptive ways, and it also helps persuade you to treat me better than you would given accurate information, being wrong (or biased) could be, at times, adaptive.

How does persuasion relate to morality and victimhood, you may well be wondering? Consider again the initial point about people, apparently, being wrong about the existence of harms and victims of acts they deem to be immoral. If one was to suggest that people are wrong in this realm – indeed, that our psychology appears to be designed in such a way to consistently be wrong – one would also need to couch that suggestion in the context of persuasion (or some entirely new hypothesis about why being wrong is a good thing). In other words, the argument would need to go something like this: by perceiving victims and harms where none actually exist, I could be better able to persuade other people to take my side in a moral dispute. The implications of that suggestion would seem to, in a rather straight-forward way, rely on people taking sides on moral issues on the basis of harm in the first place; if they didn’t, claims of harm wouldn’t be very persuasive. This would leave the moral dumbfounding work in a bit of a bind, theoretically-speaking, with respect to whether harms are required inputs for moral systems or not: that people perceive something as immoral and then later perceive harms would suggest harms are not required inputs; that arguments about harms are rather persuasive could suggest that harms are required inputs.

Enough about implications; let’s get to some research 

At the very least, the perceptions of victimhood and harm appear intimately tied perceptions of immorality. The connection between the two was further examined recently by Gray, Schein, & Ward, (2014) across five studies, though I’m only going to discuss one of them. In the study of interest, 82 participants each rated 12 actions on whether they wrong (1-5 scale, from ‘not wrong at all’ to ‘extremely wrong’) and whether the act had a victim (1-5 scale, from ‘definitely not’ to definitely yes’). These 12 actions were broken down into three groups of four acts each: the harmful group (including items like kicking a dog or hitting a spouse), the impure group (including masturbating to a picture of your dead sister or covering a bible with feces), and the neutral group (such as eating toast or riding a bus). The interesting twist in this study involved the time frame in which participants answered: one group was placed under a time constraint in which they had to read the question and provide their answers within seven seconds; the other group was not allowed to answer until at least a seven-second delay had passed, and were given an unlimited amount of time in which to answer. So one group was relying on, shall we say, their gut reaction, while the other was given ample time to reason about things consciously.

Unsurprisingly, there appeared to be a connection between harm and victimhood: the directly harmful scenarios generated more certainty about a victim (M = 4.8) than the impure ones (M = 2.5), and the neutral scenarios didn’t generate any victims (M = 1). More notably, the time constraint did have an effect, but only in the impure category: when answering under time constraints in the impure category, participants reported more certainty about the existence of a victim (M = 2.9) relative to when they had more time to think (M = 2.1). By contrast, the perceptions of victims in the harm (M = 4.8 and 4.9, respectively) and neutral categories (M = 1 and 1) did not differ across time constraints.

This finding puts a different interpretive spin on the moral dumbfounding literature: when people had more time to think about (and perhaps invent) victims for more ambiguous violations, they came up with fewer victims. Rather than people reaching a conclusion about immorality first and then consciously reasoning about who might have been harmed, it seems that people could have instead been reaching implicit conclusions about both harm and immorality quite early on, and only later consciously reasoning about why an act which seemed immoral isn’t actually making any worthy victims. If representations about victims and harms are arising earlier in this process than would be anticipated by the moral dumbfounding research, this might speak to whether or not harms are required inputs for moral systems.

Turns out that piece might have been more important than we thought

It is possible, I suppose, that morality could simply use harm as an input sometimes without it being a required input. That possibility would allow harm to be both persuasive and not required, though it would require some explanation as to why harm is only expected to matter in moral judgments at times. At present, I know of no such argument having ever been made, so there’s not too much to engage with on that front.

It is true enough that, at times, when people perceive victims, they tend to perceive victims in a rather broad sense, naming entities like “society” to be harmed by certain acts. Needless to say, it seems rather difficult to assess such claims, which makes one wonder how people perceive such entities as being harmed in the first place. One possibility, obviously, is that such entities (to the extent they can be said to exist at all) aren’t really being harmed and people are using unverifiable targets to persuade others to join a moral cause without the risk of being proved wrong. Another possibility, of course, is that the part of the brain that is doing the reporting isn’t quite able to articulate the underlying reason for the judgment well to others. That is, one part of the brain is (accurately) finding harm, but the talking part isn’t able to report on it. Yet another possibility still is that harm befalling different groups is strategically discounted (Marczyk (2015). For instance, members of a religious group might find disrespect towards a symbol of their faith (rubbing feces on the bible, in this case) to be indicative of someone liable to do harm to their members; those opposed to the religious group might count that harm differently – perhaps not as harm at all. Such an explanation could, in principle, explain the time-constraint effect I mentioned before: the part of the brain discounting harm towards certain groups might not have had enough time to act on the perceptions of harm yet. While these explanations are not necessarily mutually exclusive, they are all ideas worth thinking about.

References: Gray, K., Schein, C., & Ward, A. (2014). The myth of harmless wrongs in moral cognition: Automatic dyadic completion from sin to suffering. Journal of Experimental Psychology, 143, 1600-1615.

Haidt, J., Bjorklund, F., & Murphy, S. (2000). Moral dumbfounding: When intuition finds no reason. Unpublished Manuscript. 

Marczyk, J. (2015). Moral alliance strategies theory. Evolutionary Psychological Science, 1, 77-90.

2 comments on “Do Moral Violations Require A Victim?

  1. chris on said:

    “It is true enough that, at times, when people perceive victims, they tend to perceive victims in a rather broad sense, naming entities like “society” to be harmed by certain acts.”

    By society they could mean social cohesion or social capital.

    • Jesse Marczyk on said:

      I’m not quite sure what they mean by it, to be honest. My best guess is just that things would, in some way, be worse for many or most people were some act- like masturbating to a dead relative – condoned