How human are the dehumanised?

Developmental psychologist Paul Bloom recently published an article in The New Yorker about dehumanisation. He argued – drawing on research from many subfields in philosophy, psychology, anthropology and sociology – that the way we often think about things like slavery, genocide and misogyny is in some respects upside down. The problem isn’t that people sometimes see others as not human, it’s that they see others as very human indeed – with all that that entails.

In a key passage that elaborates on this idea, Bloom quotes philosopher Kate Manne:

“In being capable of rationality, agency, autonomy, and judgment,… [others] are also someone who could coerce, manipulate, humiliate, or shame you. In being capable of abstract relational thought and congruent moral emotions, they are capable of thinking ill of you and regarding you contemptuously. In being capable of forming complex desires and intentions, they are capable of harboring malice and plotting against you. In being capable of valuing, they may value what you abhor and abhor what you value. They may hence be a threat to all that you cherish.”

In other words, far from seeing others as not human, what Bloom and others are saying is that perpetrators of dehumanisation take very seriously indeed the humanness of others: their capacity for distinctly human behaviour. And so what we typically call dehumanisation may actually be – from a cognitive, evolutionary, and naturalistic point-of-view – an extreme means of trying to manage social relations. (None of this is, of course, any form of justification. The science is orthogonal to the morality.) This is all in clear contrast to most other approaches – even other evolutionarily- and cognitively-minded approaches, such as in David Livingstone Smith’s book Less Than Human – which tend to take the word dehumanisation at face value, as the removal of humanness from others.

Bloom and Manne’s thesis invites, I think, an interesting comparison with evolutionary theories of morality.

Humans are a massively social species. One distinctive feature of our sociality is the extent to which we cooperate. Many evolutionarily-minded scholars have, in this light, studied morality as an cognitive adaptation to an environment in which we compete to be chosen as partners for mutually beneficial interaction.

What happens when humans, equipped with these adaptations, encounter other humans and believe – with or without good reason, it doesn’t matter which – that they (the focal individual) are very unlikely to ever find themselves entering into cooperative endeavours with these others? There is in such circumstances no competition to be chosen and recruited as partners and these other humans are, ex hypothesi, a source only of possible costs and problems, with no commensurate upside. Such beliefs could in turn create space for the characterisation of others as, for instance, parasites, and hence for treating them very poorly indeed. The demonisation of immigrants in contemporary political discourse is but one modern manifestation of these dangers, and Bloom’s article describes several others.

To speculate: might the opposite be possible too? If and when we see others as sources only of possible benefits, with no downside risks, do we thereby create space for idolatry and worship? Are, in other words, dehumanisation and idolatry two sides of the same coin? Are we equally prone to each? If not, why not?

One of Theodore De Bry’s illustrations for Bartolome de Las Casas’ Short Account of the Destruction of the Indies.

9 Comments

  • comment-avatar
    Hugo Mercier 4 December 2017 (22:00)

    Even humans we don’t cooperate with behave like humans
    Even if our morality modules might not get very activated when we think of people with whom we can’t have cooperative relationships, our ToM should still be activated, right? If they can inflict costs on us, or provide us with benefits we’re still better off understanding them. If that’s true, then we’re not really (or fully) dehumanizing them. So I guess I’d side with Bloom and Manne. If they’re right, is your question still relevant?

    I guess you could restrict dehumanizing to the turning off of the morality modules. Maybe a more realistic context for that to happen would be one of strong hierarchy. One can’t really cooperate with superiors who are very far above and much more powerful than ourselves. So the morality modules shouldn’t be very useful (as in your situation of pure benefit, but for somewhat different reasons, since there are still risks — indeed, much higher risks than in standard cooperation, but of a different nature). If you take dehumanizing to only mean that the morality modules are not (or less) activated, then I’d say that could indeed apply to such interactions, which become purely strategic, rather than cooperative.

    (On the topic of idolatry, I really enjoyed that paper:

    https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2301392
    )

  • comment-avatar
    Burt 5 December 2017 (02:01)

    There is the matter of identityI think a key phrase in the article is “they may be a threat to all that you cherish.” We tend to identify with groups and roles, taking on the symbols, practices, and expectations involved with that identification. Others who may have different identities (as established both by their particular memberships and such, and by the way that our identifications depict those others) can be interpreted as threatening simply because they are different, hence giving an example of the fact that our ways are not the only ways. This can be felt as an identity threat and we will react accordingly. If our culture has been supporting our exploitation of others this can be exacerbated.

  • comment-avatar
    Thom Scott-Phillips 5 December 2017 (09:50)

    @hugo
    Thanks for the comment Hugo. But it leaves me worried that I may have been unclear!

    Did you read my post as disagreeing with Bloom, Manne and others? I don’t mean to. Moreover, I’m also *not* suggesting that the cognitive processes that underpin morality aren’t activated. On the contrary! What I’m suggesting – in the form of a tentative enquiry – is that those moral processes are activated, and that they can, under some circumstances, produce as output the belief that some groups of others are necessarily a source of costs and problems alone.

    The paper you suggest looks interesting. I will read in due course.

  • comment-avatar
    Thom Scott-Phillips 5 December 2017 (09:59)

    @Burt
    Thank you, Burt, for the comment. I agree that group identity is clearly important.

    However I am not sure the causality is as straightforward as you suggest. You write that others “can be interpreted as threatening simply because they are different”. I see the link between perception of difference and perception of threat as a puzzle – something to be explained – rather than an explanation. After all, sometimes we see different others as allies, not threats.

    So that invites the question: under what particular circumstances do we see different others as threats? (And under what circumstances do we see them as potential allies?) My post suggests a tentative answer to such questions.

  • comment-avatar
    Hugo Mercier 5 December 2017 (10:38)

    @Thom
    I though your post relied on the idea that our moral psychology evolved for cooperation. Shouldn’t it follow, then, that these mechanisms shouldn’t be fully activated, or even activated at all, when dealing with people we “are very unlikely to ever [enter] into cooperative endeavours with”?

  • comment-avatar
    Radu Umbres 5 December 2017 (23:24)

    Different kinds of humans
    Thanks Thom for this provoking comment on Paul Bloom’s article. I too side with the idea that apparent dehumanisation does not shut down all mental machinery dedicated to social interaction, and even builds upon it.

    If I get what you are saying, moral reasoning starts from the representation that certain people are not worthy or willing to cooperate, and will never be chosen as cooperative partners. But you add something which should be emphasised more: they are seen as costly, dangerous, predatory. I think this is a crucial step, and just lack of cooperation does not do it. Many people live with others that will never be their cooperative partners. Some are forbidden or dissuaded to do so, such as in the case of Indian castes, feudalism or social classes. But they are not dehumanised, and social separation can be even legitimated as a part of social existence. In contrast, dehumanised groups are perceived as spongers, thieves, unfair competitors, dirty or violent.

    I don’t think strong hierarchy as mentioned by Hugo is enough if meant only in quantitative terms. You can have cooperative relationships across social levels, and a king may feel a moral obligation to treat his subjects fairly, and vice-versa, and could a CEO and her employees. The distinction must be qualitative, a different kind of person, not just placed higher or lower among the same kind of people. For Nazis, a Jewish millionaire or a Jewish beggar were similarly despised and dehumanised. And not merely different, as Burt suggests, but different in a bad way. Japanese or Americans were quite different, but they received a compatible moral identity. A representation of others as morally-defective may do the job of switching off moral inclinations towards others, but also design them as essentially different and inferior.

    I think Thom’s question on idolatry blends in nicely with Maurice Bloch and David Graber’s work. Bloch writes about the representation of transcendental entities as givers of life and prosperity, such as kings or ancestors, a gift which can never be repaid in full and makes people perpetually indebted receivers. The Marxist term of fetishisation comes to mind.

    Finally, there is something interesting about moral mental modularity in the moral anthropomorphisation of the non-human. In my Romanian fieldwork as I think in many other places, animal breeders interpreted the moral character of some animals in human terms (stingy, sly, avenging). In Medieval Europe, there were cases of animal given a trial (and sometimes acquitted of charges). Is this an extension of the actual domain whose proper domain would be human beings, while dehumanised people are not included in the moral domain?

  • comment-avatar
    Burt 6 December 2017 (16:57)

    Difference as Threat/AttractiveThom, I agree with the way you state the question: when is difference seen as a threat and when as a potential ally or as something attractive (the old saying, opposites attract). Sometimes though, I think, the attraction itself can evoke a threat response. In part, I see this as going with the structure of beliefs that a person has and whether a perceived difference can be assimilated into that structure rather than requiring an accommodation of it as a change that could potentially threaten the entire belief structure itself (which, in my view, comes to be entangled with personal identity so that it can co-opt biological survival mechanisms in its defense). This goes to the distinction between (relatively) open and (relatively) closed cultural systems, the rigidity or adaptability of the system either externally or in individual cognition. Unfortunately I’m traveling at the moment so don’t have access to some of my files on this. I do have a book that while it doesn’t present a scientific view, does give what I consider an excellent description relating to this theme: Doris Lessing’s book of radio lectures: Prisons We Choose to Live Inside. A brief quote: “In old Southern Rhodesia the white attitudes towards the blacks were extreme: prejudiced, ugly, ignorant. More to the point, these attitudes were assumed to be unchallengeable and unalterable…, it was not permissible for any member of the white minority to disagree with them. Anybody who did faced immediate ostracism; they had to change their minds, shut up, or get out. …Also, the rules of this particular game demanded that it was not enough to say: ‘So and so disagrees with us, who are the possessors of evident truth.’ It had to also be said: ‘So and so is evil, corrupt, sexually depraved,’ and so on.” (This rings a bell regarding current American politics.) The mechanism seems to _require_ that disagreement be associated with moral failings. Perhaps as a means of self-justification.

    Interestingly, this seems part of an orienting reaction where the brain automatically orients to novelty, in a sense asking “is this dangerous or not; if not, how can it be of benefit?” In the Lessing quote, I suspect that a good bit of the rigidity of attitude related to an unacknowledged understanding of the instability of the social system so that people felt they were defenders of the moral universe. This goes with a sociological observation (again, apologies for lack of a reference) that when a culture comes under pressure and feels attacked by a materially more powerful culture one of the first reactions is a hardening of boundaries and a rise of fundamentalism. I’m tending to ramble a bit here, but this is a topic I’ve had a long time interest in.

  • comment-avatar
    Hal Morris 6 December 2017 (18:20)

    Words, words, words … “Show Me” (tiny.cc/dsxbpy)
    We tend to use phrases like “to [not] treat X as a human being” like they mean something definite when they are actually a sort of cloud of all the meanings and associations they have for everyone.

    We have a sort of “gavagai” problem here. In arguing the indeterminacy of language translation, W.V.O. Quine gives the case of a “native speaker” pointing to a rabbit and saying “gavagai” to, it would seem, an anthropologist new to some hunter-gatherer-ish culture. Should the anthropologist open her notebook and wrote “gavagai = rabbit”? What if “gavagai” means “animal” or the soul of the native speaker’s grandfather who inhabits the rabbit?

    Unfortunately, I’ve never seen this anecdote followed up with situations in which it might shed light on the familiar, nor is it said to apply to translation between the minds of two people who “speak the same language” (a very problematic phrase as well). Most people, *perhaps* not political scientists, say “right wing” or “left wing” and mean some arbitrary collection of baggage (even more arbitrary in the case of “conservative” vs “liberal”). To many, it is “left wing” to have liberal attitudes towards sex, but a very strong force in what is called American “conservatism” comes from certain varieties of libertarian, who think prostitution and all drugs should be legalized, but speak with utter contempt of “left wingers” or “liberals”. We have the gay venture capitalist Peter Thiel praising Trump in the 2016 GOP presidential convention. In discussing a Gorbachev era soviet politician, the word “conservative” often meant someone who adhered to strict Marxism/Leninism. Our two-party system is an obstacle to clarity on such matters.

    To take another broad example, which I think is closer to where I observations like “The Cheyenne[/Yanomamo/…] word for themselves really means ‘human being'” have long been a staple of pop anthropology, and at some time must have been made by anthropologists. Is this even a coherent thing to say? If you asked me to define “human being”, I might say something like a bipedal primate with massively developed intelligence, language, and sociality. I could come up with a dozen more plausible “definitions” none of which I could explain to a Yanomamo.

    While social scientists are often accused of having “physics envy”, philosophers (and others whose ideas tend to be philosophical), are prone to have “mathematics envy”. What’s not to envy? You define a mathematical object and that’s it. The definition is the essence of the object. Everyone who aspires to talk about it *must* start with the definition. The object is a mental construct. There are no physical instances of it to cast doubt on the definition.

    We really need to make statements more like “The Yanomamo are massively ignorant of non-Yanomamo, and generally expect all such (including westerners) to be cannibals and were-jaguars. The anthropologist who has lived among them for years may get the more charitable interpretation of being a defective Yanomamo who also has a pipeline to a realm of magical objects.

    IMO, the further away from examples we get, the closer we usually get to cloud-cuckoo-land, and we must just keep looking at examples where our intuition says “X was dehumanizing Y”. As I understand it, Hutus and Tutsi were closely related peoples, spoke the same language, intermarried, etc. Very often one or two sets of leaders within some pair of groups with differences find extremism politically useful. It can generate wartime situations in which power relations are frozen and amplified. The extremist leaders compare the others to “cockroaches” and the like; something in human nature would seem to make it easy to drum up, in times of fear, which itself easy to generate, a sense of purity of ones own people, and defiledness of the other “any Hutu who ‘marries a Tutsi woman’, ‘befriends a Tutsi woman’, or ’employs a Tutsi woman as a secretary or a concubine’ was a ‘traitor’ to the Hutu people” [John A. Berry and Carol Pott Berry (eds.) (1999). Genocide in Rwanda: A Collective Memory (Washington, D.C.: Howard University Press) pp. 113–115. quoted in Wikipedia].

    Do people then forget what they “knew” about this other group being as human as themselves? Does some brain mechanism suppress anything associated with that? It seems that empathy for any identifiably different group is all to easily turned off. Dehumanization seems to be a way of throwing that switch. It is even conceivable that it switches thought about the “dehumanized” into a module for dealing with other animals (classifying them, linking their recognition to reflexive actions, etc.). This is a candidate for something concrete and perhaps some day testable.

  • comment-avatar
    Thom Scott-Phillips 21 December 2017 (00:46)

    Some responses to comments
    Hi all, and sorry to Hugo, Radu, Burt & Hal for the delay in replying to their comments.

    @Hugo. I suppose, in a way, I am questioning the proper domain (and hence the proper function) of our moral psychology. Perhaps it the proper domain isn’t “people who we might enter into a cooperative relationship with”, but actually just “people”, and its calculus can, correspondingly, produce output beliefs that vary not from neutrality to idolatry (0 to 1), but from disgust to idolatry (-1 to 1). As I suggested in the post, I don’t know if this is right, but it seems to me an interesting idea. (I had not thought about it in these particular terms before, but they feel equivalent to what I wrote in the post. More thought might be needed though.)

    @Radu. I have nothing to add except my agreement. You makes several apposite observations, and extend my thoughts in several interesting directions. Thank you. I’m particularly intrigued by the idea that the actual domain of moral reasoning might sometimes include animals!

    @Burt. You are right, and you show with the Lessing quote, that whatever the original cognitive source, beliefs about the unhuman status of others can be highly stable, especially when encouraged or even enforced by institutional means. In such cases, the chain causal chain from cognition to culture has many criss-crossing links, involving not just individual moral computation, but also beliefs about institutions, and what is mutually manifest to other members of the population.

    @Hal. You ask “Do people then forget what they “knew” about this other group being as human as themselves?”. What Bloom is suggesting – and I think I agree – is that the answer is, perhaps surprisingly, no. Dehumanisers recognising those that they dehumanise as human; indeed, that’s an important reason *why* they dehumanise. If Bloom is right about this, then it’s not so much the case that, as you suggest, dehumanisation is a way to turn off the empathy switch; it’s rather that dehumanisation is a possible outcome when the empathy switch is never turned on in the first place. This seems quite plausible to me, given the long and ugly history of human violence.