Why do we sometimes de-humanize our fellow humans? Some preliminary reflections

(This post has been jointly written by Harvey WHITEHOUSE and Justin BARRETT)

Whatever else it may entail, de-humanizing involves the de-activation of our capacity to empathize. There is now a substantial body of research in experimental psychology showing that humans have highly developed capacities for empathizing (i.e. for assessing the mental and emotional states of others). Even though the nature and origins of human empathy continue to be contested, there exists extensive and robust evidence as to the psychological mechanisms involved, the way these emerge in childhood, how they operate at the neurological level, how certain pathologies affect the operation of these mechanisms, and how human empathizing abilities compare with those of other animals (especially other primates). Although humans are naturally sensitive to the feelings and intentions of other agents it has long been recognized that empathy can be switched off for special purposes, for instance when doctors seek to treat gross physical injuries, when military leaders engage in the strategic deployment of troops, and when battery-hen farmers are calculating their meat stocks. Under such circumstances reasoning about other agents ceases to be empathetic; indeed in many such cases it simply will not be relevant to consider what the agents in question are thinking and feeling.

Some researchers have begun to investigate non-empathetic ways of reasoning about other agents. For instance, speciation (the tendency to classify our fellow humans as if they were natural kinds with essentialized heritable qualities) may be necessary for various types of reflective ideas about human types, such as racial categorizing or attribution of charisma or religious specialization (witches, shamans etc. who are thought to be inherently different from other people). Or to take another example, teleological reasoning (the tendency to view our fellow humans as instruments with specialized functions, just like tools and weapons) seems to be entailed in certain types of strategic decision-making (e.g. the idea that foot soldiers can serve as cannon fodder in a strategic advance or that civilians can serve as a human shield).

What we currently know very little about is how various kinds of ecological factors promote non-empathetic and potentially de-humanizing responses towards others. In the hope of stimulating further discussion we propose the following candidates.

Role specialization. If the division of labour is weak, and everybody undertakes more or less the same repertoire of tasks and roles, we might predict that speciation of person categories (at least within the community) will be correspondingly limited or absent but, with a strengthening of the division of labour, roles and offices can eclipse individuality, leading to increasing speciation of social categories. Exploration of this topic might produce new insights into the nature and causes of modern forms of alienation, based on comparison across a wide range of societies past and present: from hunter-gatherer bands to kinship based polities and from traditional states to modern bureaucracies.

Domestication of animals. The hunting of animals often entails extensive speculation on the mental states of individuated quarry whereas the farming of livestock can encourage a more generic view of the way species as a whole think and behave. In human prehistory the shift from dependence on wild meat to the management and breeding of a range of animal species may have been revolutionary in cognitive as well as economic terms. At an extreme, the more mechanized farming methods of modern times appear to make empathetic reasoning entirely irrelevant in animal management: animals become more like a crop to be harvested than a type of agent. It may be instructive to track the effects of different forms of animal domestication on the role of empathy in human-animal relations, both in relation to the archaeological record and using extant sources. Such investigations could also open up new approaches to understanding of changing of social trends in such diverse areas as conservationism, animal warfare, consumer preferences, tourism, and even pet ownership.

Scope and effectiveness of conflict-resolution mechanisms. In resolving interpersonal conflict people must draw on empathizing capacities. But such efforts can break down, for instance on the grounds that an adversary cannot be reasoned with. When efforts to empathize fail, an adversary may be de-humanized. Such patterns can emerge when negotiations between warring groups reach an impasse, when estranged lovers divorce, when work colleagues shun each other, when kin groups feud, etc. Presumably the tipping point for this is regulated to some extent by institutional arrangements. Where conflict resolution mechanisms are highly effective or incentives to cooperate are very strong the tipping point might be seldom reached. Comparison across a wide range of institutional settings would be required to investigate these issues.

Urbanization. Whereas in rural settings it might seem that everybody knows everybody else, in urban environments encounters with strangers are routine. Since it is practically impossible to acquire rich biographical data on every person encountered we are obliged to fall back on simple categorical heuristics for dealing with strangers. This simple fact may help to explain the phenomenon of bystander apathy. The anonymity afforded by urban environments may well provide a cover for criminal behaviour just as everybody-knowing-everybody-else may deter anti-social behaviour in rural settings. But perhaps urban criminals are more confident of evading prosecution not only because they are harder to identify but also because observers routinely de-humanize those observed, regarding the latter's behaviour as ‘not their business' unless a narrow range of considerations encompassed by human kind heuristics may be said to apply.

Commoditization. Whereas gifts establish empathetic obligations between exchange partners, commodity transactions may be portrayed as ‘strictly business' such that once a transaction is completed the parties ideally have no further obligations. Commoditization may foster a de-humanizing view of others, as generic incumbents of roles (clients, investors, customers) rather than as individuals with personal histories. Research in this area might help us to solve some of the more enduring riddles of economic anthropology, such the origins of the obligations to give, receive, and repay gifts. It could have commercial implications too, for instance in helping industry to predict more accurately the changing dynamics of gifting in capitalist economies.


  • David Berreby 23 December 2008 (14:30)

    I deeply sympathize (and empathize!) with the approach described in this interesting post. But I wonder if the goal should be to understand how “reasoning about other agents ceases to be empathetic.” I’d argue that dehumanization does not involve the switching “off” of activity that would otherwise be “on.” Rather, it seems to me, dehumanization involves, in the minds of “dehumanizers”, a constant struggle. Empathy wrestles with the cognitions that inhibit it. Some evidence for this view comes from the fact that abstract reasoning itself is often a significant inhibitor of empathy. (Which btw raises the question: Is dehumanization a special case that requires explanation, or a consequence of the mind’s structure?) On that point, I’d argue that “executive function” should be added to the list of topics above. When human beings (a) reason abstractly and (b) use reason as a license to suppress their emotions, then empathy itself becomes dehumanized: It is easily depicted and experienced as the impulse of those whose comportment is closer to animals than that of the most human of the humans — those who act with intelligence, discipline and consistency. I think here of the 19th century settler in Oregon who encountered native people who were starving and begging for mercy and for food. He said it “hurt my feelings” to know that they were to be killed. But those were the orders, “so we did the work.” Similarly, General Curtis LeMay, architect of massive bombing campaigns conducted by the U.S. military during World War II, had no trouble vividly imagining the horrors caused by his missions. But all pilots in that position, he wrote, must put aside the image of the ruined house and the injured child crying for her mother — “if you intend to do the work your nation expects of you.” The importance of these kinds of self-reports by professional dehumanizers is that they form a record of how “executive function” was deliberately employed to suppress empathy in the name of God, the People, the nation, science, the honor of the guild, etc., etc.. And in that record, we don’t often find that empathy was “on” until it was turned “off.” Instead, these people more often describe constant restling with contradictory thoughts and feelings. Secondly, there is, in settings in which people are dehumanized, often an aspect of “protesting too much.” Why is it never enough to treat people as if they were livestock or blocks of wood? Why, instead, must the perpetrators *announce* that they are doing so, in theatrical demonstrations? Adam Gopnik, the essayist and critic, made this point a couple of years ago (and it’s at hand, so I’ll just quote him here):”It is often said that terror of this kind is possible only when one has first ‘dehumanized’ some group of people-aristocrats, Jews, the bourgeoisie. In fact, what motivated the spectacle was exactly the knowledge that the victims were people, and capable of feeling pain and fear as people do. We don’t humiliate vermin, or put them through show trials, or make them watch their fellow-vermin die first.” The exaggerated spectacle of dehumanization might be another sign that the line is not constant between thinking of others as objects and thinking of them as people.

  • Dan Sperber
    Dan Sperber 9 January 2009 (19:13)

    Here is a side remark on what Harvey and Justin call “non-empathetic and potentially de-humanizing responses towards others” (twisting their meaning, I admit). There is a type of activity where humans are approached in an at least partially non-empathetic way, I mean the sciences. Even the most interpretive of anthropologists and historians would (or at least should) agree that their scholarly task is not fully carried out by means of empathy (or [i]Verstehen[/i]). Ecological factors unknown to the people studied, for instance, are often relevant to understanding their culture. On the other hand, it can be argued – I, for one, have – that no approach to human social and cultural life can be carried out without involving some interpretive work. This is true even of the most would-be naturalistic approaches, which ignore their interpretive dimension at the risk of being much less genuinely naturalistic than claimed. The empathetic approach is often presented as intrinsically respectful of others’ humanity while the non-empathetic or non-interpretive ‘hard-science’ component of the social sciences is often seen as indeed “potentially de-humanizing,” but is this really so? In fact, an interpretive approach can be de-humanizing, as when racists or anti-Semites interpret the mentality they attribute to people they despise so as to show them to be sub-human. And a naturalistic approach with a very limited interpretive (in the [i]Verstehen[/i] sense) component can throw light on what is specifically and even admirably human, as has done, for instance Chomsky’s naturalistic approach to language and its creative aspects.

  • Aniko Sebesteny
    Aniko Sebesteny 29 January 2009 (04:28)

    Your interesting post made me think, and I have some comments. You will decide if they have any relevance for you… 🙂 I think that if you put the question : why do we de-humanize, then the reverse question should be also examined: why do we humanize ? The question as it is asked in the post suggests that empathy is a modular capacity that would be basically “on” and could be switched “off” or disturbed in certain contexts. According to this, human beings would be basically empathic. I am not sure that it has this irrepressible automatic functioning like the understanding of speech or detecting movement. Empathy could be seen as an ability used when necessary rather than a modular automatic capacity. My intuition is that in ecologic context, people are empathic when needed: when they are forced to be, when it is rewarding, when it is in line with their goal. People would go in the direction of the maximization of the reward and minimization of the effort. So I would suggest the massive use of relevance theory. Cognitive resources will be allocated to empathy if it is relevant. To hunt an animal in the wild, you really have to use empathy to predict its movements. But it would be counterproductive to be empathic with the cattle you are breeding. So the ecologic factors could be thought of as “relevancy-of-empathy” factors. As anthropologist, I cannot help trying some examples. It may be relevant to be empathic with the people who are close to. But not all the time. Let’s take an adolescent girl, Jenny, fighting with her mother. Jenny wants to join her friends in the disco, and her mother would not let her go. They are very close kin, they live together, and still, Jenny doesn’t show empathy neither for her mother’s point of view (that disco is dangerous when one is only 12 years old), nor for her emotional state. She just considers her mother as an obstacle, makes her best to get through, as this is the best way for her, in her opinion, to get to her goal. She will use her empathic capacity to find out how she could threaten her mother (“I will not do my homework any more if you don’t let me go!”), but not to perceive that her mother is sad and anxious due to her behavior. If I am just ignoring the wishes of another person, it may go swiftly. If I have to kill the other person, I probably have to make an effort, to make fake trials for example. (I guess this supports Nicolas Baumard’s findings.) The two kinds of non-empathy seem quite different, and should maybe be conceptually separated. It may even be useful to split the concept of empathy into a whole range of parts. Even with the best intention and cognitive effort, one can difficultly be totally empathic to the overall situation of another person. (And the anthropologist, who tries hard on the field, may become so empathic with everybody that he/she becomes unable to communicate normally or to have an opinion, so to say to be socially functional. Then one drives back the empathy to a “normal level”, and that helps to be integrated.) There may be a whole range of kinds of empathy. Here are some probably ad hoc guesses. One could perceive – the fact that the other is alive (and does not want to be killed) – the physical state of the person – the emotional state – the intentions and desires These different levels could have a different cognitive treatment. Some thoughts about the empathy in rural areas. I have to confess I know better rural Indonesia than rural Europe, as I did fieldwork there. Thinking back, I was astonished by cases of striking lack of empathy between neighbors, even between people claiming common ancestors. A case for instance in Bali when a woman was helping her neighbor in her business of preparing ritual decoration for retail. She was paid close to nothing, though she was spending hours and hours there nearly daily. I cannot give all the details, but the situation was unjust. Though, it didn’t seem to bother the “business-woman”. I guess it would have been counterproductive for her to have empathy for the working neighbor: she would probably have felt obliged to pay her more. It is certainly very interesting to question the implications of empathy in an ecologic environment. If in Bali I know of a friend that she cannot afford to buy a raincoat, and the rainy season is approaching (rain is really heavy there, nothing to do with Oxford rain), then I will feel the obligation of buying her a raincoat. As my financial resources are not infinite, I cannot have a big number of people with whom I empathize to that level. I just cannot afford it.