It has taken me eleven chapters to explain my view of what dehumanization is, how it works, what its function is, and why we should take it very seriously. But the journey is not yet over. There is another whole layer of dehumanization that needs explaining, and this requires adding some new and very important elements to the story that I have so far told.
My own thinking on these matters was stimulated by the need to address two important objections to the theory of dehumanization presented in my 2011 book Less Than Human. One of these, which I call the problem of humanity, has been discussed by several writers. It is well known and is often regarded as a knock-down refutation of the claim that dehumanization is real. The other, which I call the problem of monstrosity, has been largely unrecognized. After grappling with these issues for several years, I came to realize that they are both solved by a single theoretical stroke—one that dramatically deepens our understanding of what dehumanization is and how it works.
The problem of humanity is this: people who think of and talk about others as less than human also think and talk about them as human. This implies that, whatever is going on when people ostensibly dehumanize others, it is not accurate to say that they consider those they dehumanize to be nothing but subhuman animals. I addressed this problem in passing in Chapter 1, in the discussion of Winthrop Jordan’s claim that White slaveholders did not literally regard enslaved Black people as subhumans. Jordan’s incredulity about the animalization of Black people turned on the fact that slaveholders implicitly acknowledged the human status of those they held in bondage. I noted in that discussion that an even stronger argument can be made. The White people who described Black people as animals also often referred to them not merely implicitly, but explicitly, as human beings. This is typical of dehumanizing discourse. Nazis referred to Jews both as vermin and as criminals, but the category “criminal” is applicable only to human beings and not to vermin. And Islamophobes say that Muslims are both bloodthirsty beasts and religiously motivated terrorists, but only humans can be terrorists. The practice of referring to dehumanized people as people suggests either that there is something missing from theories of dehumanization or that the whole notion of dehumanization is seriously misguided and that people do not really conceive of other people as less than human.
The earliest statement of the problem of humanity dates back to Morgan Godwyn’s writings, in particular his 1680 work The Negro’s and Indians Advocate, which I cited in Chapters 4 and 8. The word “dehumanization” did not exist during Godwyn’s lifetime. It was introduced early in the nineteenth century. But Godwyn certainly had a concept of dehumanization; he referred to it using terms such as “brutifying” and “soul murdering,” and he wrote of Black people being “unsoul’d” and “unman’d.” Godwyn was not interested in dehumanization as such. His mission was to show that it is factually, morally, and spiritually wrong to think of Black people as subhuman creatures. In the service of this goal, he sought to demonstrate that the White people who claimed that Black people were nothing more than “brutes” contradicted themselves by also implicitly affirming that Africans were human beings. Godwyn had a remarkably sophisticated understanding of the dehumanizing process. For instance, unlike many people today, he recognized that dehumanization is not a “failure” to see the humanness of others or merely a natural response to “difference.” He was aware that dehumanizing beliefs about Black people are ideological constructions for legitimating oppression, and charged that slaveholders “infer their Negro’s Brutality, justifie their reduction of them under Bondage; disable them from all Rights and Claims; even to Religion itself, pronounce them Reprobates; and upon a sudden (with greater speed and cunning than either the nimblest juggler or witch) transmute them into whatsoever substance the exigence of their wild reasonings shall drive them to.”1
Godwyn notes that despite this seeming transmutation, slaveholders were unable to fully relinquish their awareness of the humanity of the people whose souls they tried to murder. If Black people were not human, then
why should they be tormented and whipt almost (and sometimes quite) to death, upon any, whether small or great Miscarriages … were they (like Brutes) naturally destitute of Capacities equal to such undertakings? Or why, should their Owners, Men of Reason no doubt, conceive them fit to exercise the place of Governours and Overseers to their fellow Slaves, which is frequently done, if they were mere Brutes? … It would certainly be a pretty kind of Comical Frenzie, to imploy Cattel about business, and constitute them Lieutenants, Overseers, and Governours, like as Domitian is said to have made his horse a Consul.2
And what about the implications of their rape and sexual exploitation of enslaved people at the hands of their White overlords?
If slavery hath such a faculty or power to turn men into beasts, or if all Negro’s be naturally such, may we not be bold to demand what will become of those Debauches, that so frequently do make use of them for their unnatural Pleasures and Lusts? Or such of our People who have intermarried with them? Surely they would be loth to be endited of Sodomy, as for lying with a Beast. It would therefore be convenient for them to renounce that Beastly opinion; or else that the law may have its free Course and be let loose upon them.3
Finally, Godwyn points out that Whites’ claim that slaves are livestock does not pass muster, because slavers are guilty of “treating their Slaves with far less Humanity than they do their Cattel. For they do not use to starve their Horse, which they expect shall both carry and credit them upon the Road; nor to pinch the cow of her fodder, by whose milk their Families are sustained: Which yet (to their eternal shame) is too frequently the lot and condition of these poor People, from whose labour their Wealth and Livelihoods do wholly arise.”4
Godwyn anticipated most of the points raised in present-day discussions of the problem of humanity. For instance, we find the philosopher Stanley Cavell writing nearly three centuries later that “it is sometimes said that slaveowners did not see or treat their slaves as human beings, but rather, say, as livestock; some slaveowners themselves have been known to say so.… But does one really believe such assertions?”5 His answer is “no,” which he justifies thus:
When he wants to be served at a table by a black hand, he would not be satisfied to be served by a black paw. When he rapes a slave or takes her as a concubine, he does not feel that he has, by that fact itself, embraced sodomy. When he tips a black taxi driver … it does not occur to him that he might more appropriately have patted the creature fondly on the back of the neck. He does not go to great lengths either to convert his horses to Christianity or to prevent their getting wind of it. Everything in his relation to his slaves shows that he treats them as more or less human—his humiliations of them, his disappointments, his jealousies, his fears, his punishments, his attachments.6
Leaving aside the disturbing claim about Black taxi drivers, whom Cavell seems to regard as slaves, and leaving aside the fact that he does not draw on any historical evidence about the beliefs and attitudes of real White supremacists (in contrast to his story about a “more or less mythical slave owner”), it is clear that Cavell denies that racists regard Black people as subhuman animals. In his view, the slave owner really just believes that the slave is “indefinitely different” from himself, and that “what he really believes is not that slaves are not human beings, but that some human beings are slaves.”7
Philosopher Kwame Anthony Appiah voiced related concerns in his book Experiments in Ethics nearly a decade after Cavell. He argues that the claim that genocidaires think of their victims as subhuman animals is “not quite right” because “it doesn’t explain the immense cruelty … that are their characteristic feature. The persecutors may liken the objects of their enmity to cockroaches or germs, but they acknowledge their victims’ humanity in the very act of humiliating, stigmatizing, reviling, and torturing them. Such treatments—and the voluble justifications the persecutors invariably offer for such treatment—is reserved for creatures we recognize to have intentions, and desires, and projects.”8
Appiah points out that dehumanizers say that their victims deserve punishment, but that the notion of “deserving punishment” is applicable only to human beings—not to vermin. However, he does not move from this to the conclusion that genocidaires really think of their victims as human beings and not subhuman animals. If a conception of dehumanization is “not quite right,” it is partially or largely right, and therefore incomplete rather than entirely mistaken. As you will see, I think that Appiah was on the right track.
The philosopher Kate Manne is the most prominent contemporary critic of theories of dehumanization. She argues that dehumanization does not exist, and what seems like dehumanization is really something else. Manne’s main target is a view that she calls humanism, a central component of which is that “when we recognize another human being as such … then this is not only a necessary condition for treating her humanely, in interpersonal contexts, but also strongly motivates and disposes us to do so.”9 She points out that recognizing others as human is entirely compatible with hostility toward them, and that seeing that they are human often involves recognizing that they are threatening in distinctively human ways. Human beings can be malicious, cruel, oppressive, envious, rivalrous, and so on—therefore, recognizing others as human does not translate into treating them kindly. What, then, is the purpose served by referring to them as subhuman? Here is Manne’s answer: “One simple point is that dehumanizing speech can function to intimidate, insult, demean, belittle, and so on … since it helps itself to certain powerfully encoded social meanings. And given that human beings are widely (if erroneously) held to be superior to nonhuman animals, denying someone’s humanity can serve as a particularly humiliating kind of put-down. When a white police officer in Ferguson called a group of black political protesters ‘fucking animals’ … he was using this trope to demean and degrade the protesters and reassert his own dominance.”
Manne points out that the police officer’s use of an animalistic slur to degrade and humiliate the protestors indicates that even though he called them “animals,” he regarded them as human. If we assume otherwise, this scenario would simply not make sense, Manne argues, because “such put-downs would hardly be apropos when it comes to actual non-human animals, who could neither comprehend the insult nor be successfully put down by having their nonhuman status correctly identified. This requires human comprehension, not to mention an incipient human status to be degraded from. There is nothing to object to in being called a rat if, in fact, you are one.”10
Let me put Manne’s point in a slightly different way. Dehumanization is normally diagnosed on the basis of what people say—usually, their use of animalistic slurs. Such slurs are intended to derogate others—to humiliate them, to assert their inferiority, and so on. But people who slur others in this way could not possibly have such an intention if they really believed that the objects of their hostility were subhuman animals. The use of animalistic slurs presupposes that the speaker recognizes both that the person whom they are targeting is a human being and also that that person is aware of their status as a human being.
It should be evident that many of Manne’s critical points are not applicable to the theory of dehumanization that I have set out in this book. I do not advocate humanism as she defines it, as I do not hold that recognizing others as human makes one favorably disposed toward them. I do not claim that the use of animalistic slurs always, or even mostly, indicates that the speaker has dehumanizing beliefs. And I reject the idea that dehumanization is a failure to notice the humanity of others. Nevertheless, Manne’s characterization of the problem of humanity deserves serious attention, if only to address the problem of distinguishing slurs that express genuinely dehumanizing attitudes from those that do not.11
So far, I have explained dehumanization as the attribution of a subhuman essence to others. From the dehumanizer’s perspective, the other person might appear to be human on the “outside,” but they are actually subhuman animals on the “inside.” However, there are many examples of dehumanization that do not fit this model. Consider the victims of lynching described in Chapter 1, who were regularly described in racist literature of the day not simply as predatory animals, but also as monsters and fiends. And consider the putatively Satanic Jews described in Chapter 10, or the description of Jews in the SS booklet Der Untermensch, which stated, “Inside of this creature lies wild and unrestrained passions: an incessant need to destroy, filled with the most primitive desires, chaos and coldhearted villainy.… The subhuman thrives in chaos and darkness, he is frightened by the light. These subhuman creatures dwell in the cesspools, and swamps, preferring a hell on earth, to the light of the sun.”12 Taking this line of enquiry further, it is not necessary to use words like “monster” or “demon” to refer to and conceive of others as demonic or monstrous. Words like “superpredator” and “terrorist” can do exactly the same job. So, when considering the problem of monstrosity, we need to look below the surface of the manifest message.
Monstrousness does not fit into the theoretical framework that I have so far described, because monsters are not subhuman animals. They are unnatural entities: outsiders to the Great Chain of Being. The tenth-century Liber Monstrorum de Diversis Generibus (The Book of Monsters of Various Kinds) captures this point perfectly: “Made as they were, the order of creation must keep them on the outside.”13
Thinking of others as monsters is quite different from thinking of them as animals. An animal might be frightening or disgusting, but monsters are horrifying, and they are felt to be much more dangerous than any animal because of their extraordinary powers. Monsters are malevolent and uncanny, whereas nonhuman animals are neither. And most importantly, monsters have no place in the order of nature. They embody its subversion.
Unlike the problem of humanity, the problem of monstrosity is hardly ever mentioned in the dehumanization literature as a problem. Although many writers point out that members of dehumanized groups are often “demonized,” they do not offer any explanation for this, and do not seem to notice that it does not comport with theories of dehumanization. This overlooking has large implications, because the challenge that it poses is not exclusive to theories like mine, which draw on the concept of the Great Chain of Being.
Researchers into dehumanization have not given demonization its proper due. None of the social psychological theories of dehumanization that I have discussed in this book have a place for it. On Nick Haslam’s account, we dehumanize others by attributing machine-like or animal-like characteristics to them. But demons are neither machine-like or animal-like. And theorists who think of dehumanization as the denial of mental states to others also have no place for the demonic. Demons are evil, but they are not mindless. Others causally refer to demonization as a kind of dehumanization, without giving any account of what this amounts to or how it occurs. Things are no better in social political science. It is of course recognized that demonization plays an important role in international relations, but such analyses take the concept of the demonic as given, rather than subjecting it to analysis.14
As I reconstruct it, the argument that moves from the premise that dehumanizers recognize the humanness of those whom they seem to dehumanize to the conclusion that they do not really think of these others as nonhuman proceeds like this. The first premise states that some people refer to some other people as subhuman creatures, but they also seem to regard them as human beings. The second premise states that it is logically impossible, and therefore inconceivable, for any being to be both human and subhuman, because these two categories exclude one another. The third premise states that, given the truth of the first and second premises, those who speak of others as human and as subhuman must be conceiving of them as either human or subhuman, but not as both. The fourth premise states that it is much more likely that these people conceive of those whom they seem to dehumanize as human than as subhuman. This premise might be justified in all sorts of ways, for example commonsensically (they don’t look human) or through a sophisticated argument like Manne’s. And all of this leads to the final conclusion that it is overwhelmingly likely that perpetrators do not really believe that those whom they describe as animals are really less than human.
This is a flawed argument. Its problem lies in the transition from the second premise to the third one. The second premise is a special case of the general truth that nothing can simultaneously possess a property and lack that very same property. This is true as a matter of logic. It cannot be true, for example, both that I am over six feet tall and that I am less than six feet tall, or that the tumbler on my desk both contains whiskey and does not contain whiskey. But the third premise, which is supposed to follow on from it, is a psychological claim rather than a logical one. It would only follow from the second premise there was another, intermediate premise stating something like “Human psychology always conforms to the rules of logic.” That would be a ridiculous claim to make, because it is a truism that human psychology can and often does defy logic. We are adept at irrationality, and are often able to entertain contradictory beliefs side by side. So, unless there is good reason to think that beliefs about humanity and subhumanity are insulated from this sort of irrationality, the premises of this argument fall short of establishing its conclusion.
Versions of this argument that rely on animalistic slurs as their sole evidential base also suffer from another weakness. It is true that inferences about dehumanization rest largely on what people say. And, as I have pointed out, we cannot legitimately conclude that a person conceives of another as a subhuman just because they use this sort of speech against them. In Manne’s example, the White cop who called Black protesters “fucking animals” may not have actually thought of them as animals, and may have said this just to put them down, just as she argues. But the evidential base for the reality of dehumanization is much broader than examples like this. There are many, many instances of people claiming that other people are subhuman, where this is meant to be taken literally as a statement of fact. I have supplied quite a few of them in this book. Here is yet another, from the anthropologist Hugh Raffles, who points out, drawing on historical documents, that “in early modern France … ‘since coition with a Jewess is precisely the same as if a man should copulate with a dog,’ Christians who had heterosexual sex with Jews could be prosecuted for the capital crime of sodomy and burned alive with their partners—‘such persons in the eye of the law and our holy faith differ[ing] in no wise from beasts.’ ”15
Even if animalistic slurs never express dehumanizing beliefs (which is surely not the case), that does not undermine the evidential base for the reality of dehumanization. And once we accept that there are examples that, by any reasonable criterion, point to the existence of true dehumanization, it becomes an open question whether the dehumanizing mentality lurks behind any given episode of animalistic derogation.
However, the problem of humanity remains. Why is it that those who conceive of others as subhumans also conceive of them as humans? What does this tell us about the nature of dehumanization? Consider the passage that I just quoted. Even though the statements about the subhumanity of Jewish women were clearly meant to be taken at face value, the quoted text states that “such persons” differ in no wise from beasts. It seems strange to say that there are persons that are no different from beasts. If they are persons, they are not beasts, and if they are beasts, they are not persons.
Look again at the flawed argument that I gave a few paragraphs back. The argument can be repaired by eliminating the dubious premise “So it must be the case that those who speak of others in this way think of them as either human or subhuman, but not both” and adding in its place “So it must be the case that those who speak of others in this way believe them to be both human and subhuman.” Once that is done, the rest of the argument falls away, leaving us with the conclusion that when people dehumanize others they regard them as being human and subhuman simultaneously. This should not be interpreted to mean that dehumanizers conceive of others as partly human and partly subhuman, like mermaids or centaurs. It should be understood to mean that they conceive of dehumanized people as completely human and completely subhuman, both at once. I am aware that this sounds bizarre, and perhaps even unintelligible. But I will do my best to dispel this aura of strangeness.
Understanding ambivalence is crucial for making sense of the contradiction that lies at the heart of the dehumanizing mind. Swiss psychiatrist Eugen Bleuler coined the term “ambivalence” in 1910. Bleuler identified three kinds of ambivalence: “affective ambivalence” (conflicting emotions toward the same person), “ambivalence of the will” (the desire to pursue some course of action coexisting with the desire not to pursue it) and “intellectual ambivalence” (belief in contradictory propositions). Freud picked up the term in 1912, and it soon became a mainstay of the theoretical vocabulary of psychoanalysis.16
Freud’s major discussion of ambivalence appears in the second of four essays published in 1912–1913 that were pulled together as the book Totem and Taboo. Totem and Taboo is Freud’s main foray into speculative anthropology. It is a work in which he attempted to theorize the phenomenon of taboo—forceful and often puzzling restrictions on certain kinds of behavior often found in traditional societies. As Freud put it, “Behind all these prohibitions there seems to be something in the nature of a theory that they are necessary because certain persons and things are charged with a dangerous power, which can be transferred through contact with them, almost like an infection.… The strangest fact seems to be that anyone who has transgressed one of these prohibitions himself acquires the characteristic of being prohibited—as though the whole of the dangerous charge has been transferred over to him.”17
Freud argued that taboos arise from ambivalent attitudes that become sedimented into a culture, and he used the ambivalent attitudes of individuals to throw light on them. In particular, Freud was interested in the striking similarities between taboos and the strange private rituals and prohibitions found in cases of obsessional neurosis (nowadays known as obsessive-compulsive disorder). He treats these cultural taboos as obsessional neuroses writ large, extrapolating as best he can from one to the other. “Obsessional patients,” he writes, “behave as though … persons and things were carriers of a dangerous infection liable to be spread by contact onto everything in their neighborhood.” He continues, comparing the ritual purifications associated with cultural taboos with the private rituals of those suffering from obsessional neurosis: “Obsessional prohibitions involve just as extensive renunciations and restrictions in the lives of those who are subject to them as do taboo prohibitions; but some of them can be lifted if certain actions are performed. Thereafter these actions must be performed. They become compulsive or obsessive acts, and there can be no doubt that they are in the nature of expiation, penance, defensive measures and purification.”18
Freud then goes on to argue that, in both cases, it is ambivalent attitudes that causally underwrite beliefs about a dangerous, contagious power that must be met by social isolation and rituals of purification. This attitude is not just a matter of mixed feelings, desires, or beliefs. If that were the case, the conflict between them could be resolved by deliberation bringing the incompatible ideas into relation to one another. But resolution is impossible because the incompatible attitudes are mentally segregated from one another. As Freud says, “The conflict between these two currents cannot be promptly settled because—there is no other way of putting it—they are localized in the subject’s mind in such a manner that they cannot come up against each other.”19
I think that something like what Freud describes goes on in episodes of dehumanization. There are two contradictory beliefs or mental representations of the dehumanized person present in the dehumanizer’s mind. One of these is a conception of that person as a human being and the other is a conception of them as a subhuman animal. These are walled off from one another and cannot interact.20
The next task is to explain how and why this occurs. Consider the ways that we normally come to categorize things. We usually do so on the basis of what our senses tell us. We think something is blue if it looks blue, we think something is a porcupine if it looks like a porcupine, we think someone is our best friend if they look like our best friend. But there is also another basis for categorization: expert opinion. Experts are people, or groups of people, whom we regard as having authoritative knowledge of some domain. They are the ones who are supposed to know.
We often accept as true what the experts tell us about the world even if what they tell us contradicts what our senses tell us. For example, solid objects like the chair on which you are sitting look and feel gapless. But physicists tell us that such objects consist mostly of empty space. Even though our eyes tell us that solid objects are gapless, we defer to the physicists because, in our culture, they are supposed to know.
There is no necessary connection between having the status of expert and having genuine knowledge. This is because being an expert is ultimately a political status, a position of power and authority. Sometimes, those who are supposed to know do not really know, or they conceal what they really know in the service of some political agenda. But because we defer to those who occupy the cultural role of expert, we are likely to trust even these claims and accept them as true.
I have been talking about scientists as experts, but there are many other figures who can and do occupy this role. Academics, members of the clergy, celebrities, athletes, politicians, motivational speakers, and even radio talk-show hosts may all be accorded the status of expert. Today, in the age of the internet, anyone with an attractive enough message can be presented as an expert, and accumulate followers who regard them as one. And purported expertise is often distributed through the whole of a society rather than being lodged in certain individuals. In such cases, pervasive, taken-for-granted ideological beliefs have the status of “common knowledge” that defines reality, even when these beliefs do not correspond to the deliverances of our senses.
However, the configuration of the human mind sometimes presents an obstacle to adopting expert opinion. “The Party told you to reject the evidence of your eyes and ears,” wrote George Orwell in 1984, but sometimes expert testimony does not cause us to relinquish what our senses tell us. Sometimes, we are unable to abandon these perceptual beliefs, even though we also accept what the experts claim. Consider again the example of the solid objects. Even though you and I both accept what physicists tell us about solid objects, it is impossible for us to stop regarding the chair and other such objects as gapless. We hold both beliefs—the belief in the chair’s gaplessness and the belief in its gappiness—simultaneously, even though they are contradictory. Optical illusions are a great illustration. Consider the Müller-Lyer illusion. The illusion consists of two horizontal lines, one placed above the other. The two lines are exactly the same length, but the top line terminates with arrows at each end, and the bottom line terminates with inverted arrows at each end. Even though you know that the two horizontal lines in the illusion are precisely the same length, you cannot help seeing the one at the bottom as longer than the one at the top. Even though you know that what you are “seeing” is not really there, you just can’t help seeing it that way.
This is how dehumanizing beliefs come about. At its heart, dehumanization involves a contradiction between the perceptual categorization of others as human and the more “theoretical” categorization, acquired from others, that they are subhuman.
As I explained in Chapter 11, when we encounter other human beings we tend to rapidly and reflexively perceive them as human beings.21 So, when we are told by someone that certain people are less than human, and we accept this on their epistemic authority, this does not prevent us from also seeing them as human beings. Under these circumstances, we think of these others as fully human in one “part” of the mind and as fully subhuman in another “part” that is segregated from the first. As Freud wrote, “They are localized in the subject’s mind in such a manner that they cannot come up against each other.”
This way of looking at things resolves the problem of humanity. Dehumanizers implicitly or explicitly regard those whom they dehumanize as human beings because it is impossible for them to shake that belief, which sits side by side with their belief that these others are subhuman creatures. Even though the mind of the dehumanizer harbors both beliefs, only one of them can be salient at any given time. And when one is in the mental foreground, the other one retreats into the background. This is why dehumanizing discourse tends to alternate between characterizing the other as human and characterizing them as subhuman. The historian of American racism David Brion Davis observed this, although he did not theorize it. I quoted him in Chapter 2 as arguing that “since the victims of [lynching] are perceived as ‘animalized humans,’ this double consciousness would probably involve a contradictory shifting back and forth in the recognition of humanity.”22 An interview with a woman who participated in a lethal attack on Roma residents of the town of Hadereni, Romania, in which a man was burned to death and many homes were destroyed by arson, provides a very clear example of Davis’s “contradictory shifting back and forth.” The interviewee told a journalist, shortly after the pogrom, “On reflection … it would have been better if we had burnt more of the people, not just the houses.… We did not commit murder—how could you call killing Gypsies murder? Gypsies are not really people, you see. They are always killing each other. They are criminals, sub-human, vermin.”23 Notice that she starts off by saying that she regrets not having burned more of the people (that is, human beings) to death, but next says that killing Roma is not murder because Roma are not people (they are not human beings). Then she says that they are criminals (only human beings can be criminals) but asserts right afterward that they are subhuman vermin (once again, not human).
If I am right, the problem of humanity is not really a problem at all. Given the way that the human mind works, it is an expectable aspect of the dehumanization. But there are further ramifications of this idea—ramifications that both throw a great deal of unexpected light on the phenomenology of dehumanization and pave the way for a solution to the problem of monstrosity.
I begin with a point that is obvious but that is all too often overlooked. Consider the Nazis’ claims that Jews are creatures rather like rats. People have lots of different attitudes to rats. My grandmother, who spent her teenage years in a rat-infested tenement building, was terrified of them. But my daughter adored rats as a child, and kept two of them as pets. Psychologists who use rats in maze-running experiments regard them dispassionately, as do medical scientists who use them to test the effects of experimental drugs. My point is that there is nothing intrinsically disgusting about these rodents. But the Nazis did not say that Jews were simply rats. That would have been delusional (Jews are not small furry creatures with naked tails). Instead, they described them as human vermin. For all their ideological ardor, committed Nazis could not help seeing Jews as human beings. But at the same time, they held fast to the ideological belief that Jews were Untermenschen. So, Jews were not just rats in the eyes of true believers of Nazi ideology. They were rat / humans. Understanding this is vital for theorizing dehumanization, because there is an important research literature spanning several disciplines that suggests that contradictory beings like rat / humans elicit a distinctive and highly disturbing psychological response. They are felt to be what in German is called Unheimlich, a word that is conventionally (if imperfectly) translated as “uncanny.” Freud pointed out in an article that he wrote on this topic that the uncanny “is undoubtedly related to what is frightening.… Yet we may expect that a special core of feeling is present which justifies the use of a special conceptual term. One is curious to know what this common core is which allows us to distinguish as ‘uncanny’ certain things which lie within the field of what is frightening.”24
Freud had his own ideas about how feelings of uncanniness come about, but it will be more productive for us to turn to the work of his contemporary, a German psychiatrist named Ernst Jentsch. In his 1906 paper “On the Psychology of the Uncanny,” Jentsch set out to determine exactly what it is to experience something as uncanny, what sorts of things elicit this response, and why they do so. He argued that uncanny things produce a sense of disorientation and uncertainty. But he was aware that not all uncertainties generate the uncanny feeling, or produce it to the same degree. There must be a special sort of uncertainty involved: “Among all the psychical uncertainties that can become a cause for the uncanny feeling to arise, there is one in particular that is able to develop a fairly regular, powerful and very general effect: namely, doubt as to whether an apparently living being really is animate and, conversely, doubt as to whether a lifeless object may not in fact be animate—and more precisely, when this doubt only makes itself felt obscurely in one’s consciousness. The mood lasts until these doubts are resolved and then usually makes way for another kind of feeling.”25
Skillfully crafted human figures in a wax museum are one of his prime examples. Even though one knows that these are inanimate figures, they are so lifelike that one cannot help regarding them as real people. The outcome is seeing them as both animate and inanimate. But nothing can be wholly animate and wholly inanimate, and it is this that produces the uncanny feeling. To the extent that the uncanny feeling persists, he conjectures, “it is probably a matter of semi-conscious secondary doubts which are repeatedly and automatically aroused anew” or “the lively recollection of the first awkward impression lingering in one’s mind.”26
Having got to this point it is helpful to push back at the translation of Unheimlich as “uncanny.” The English word does not quite capture the state of mind that Jentsch is gesturing toward. Something “uncanny” can be simply odd or astonishing—even enjoyably fascinating, as is exemplified by open-mouthed wonder at the “uncanny” feats of a champion athlete. But Jentsch is clearly talking about a disturbing quality of experience, one that sends chills down your spine and makes your blood run cold. To find a good English equivalent, just ask yourself how uncanny things such as wax figures in a dimly lit room strike you. One word that is likely to come to mind is “creepy.” Uncanny things are creepy things, and the state of mind that they produce is the state of being “creeped out.” Now, imagine that the wax figures in the dimly lit room begin to move. They turn their heads toward you, open their mouths, and blink their waxy eyes. How would that make you feel? Here is how Jentsch discusses this kind of scenario, using the example of humanoid automata:
This peculiar effect makes its appearance even more clearly when imitations of the human form not only reach one’s perception, but when on top of everything they appear to be united with certain bodily or mental functions. This is where the impression easily produced by the automatic figures belongs that is so awkward for many people. Once again, those cases must here be discounted in which the objects are very small or very familiar in the course of daily usage. A doll which closes and opens its eyes by itself, or a small automatic toy, will cause no notable sensation of this kind, while on the other hand, for example, the life-size machines that perform complicated tasks, blow trumpets, dance and so forth, very easily give one a feeling of unease.
Sixty-four years after Jentsch’s paper, Japanese roboticist Masahiro Mori published a paper that drew the same conclusion. Although very short and highly speculative, Mori’s paper has been immensely influential, especially in the cognitive sciences. It is titled “Bukimi No Tani,” which was translated into English by the art critic Jasia Reichardt as “The Uncanny Valley.”27 Like the German Unheimlich, the Japanese bukimi can be rendered as “creepy”—so the title of Mori’s paper can equally well be translated as “The Valley of Creepiness.” The thrust of the article is straightforward. Mori speculated that as robots become more and more human-like, we will feel more and more comfortable with them until technology reaches the point where robots are almost, but not quite, indistinguishable from human beings. At that point, he suggested, there will be a precipitous drop in the feeling of affinity, and the humanoid robot will be experienced as bukimi. He called this “the uncanny valley.” Mori made the same prediction about prosthetic limbs. An artificial hand that is not quite indistinguishable from a flesh-and-blood hand will, he supposed, produce feelings of repugnance. And just like Jentsch, Mori argued that adding movement will only augment the disturbing effect. “Since the negative effects of movement are apparent even with a prosthetic hand,” he wrote, “to build a whole robot would magnify the creepiness. This is just one robot. Imagine a craftsman being awakened suddenly in the dead of the night. He searches downstairs for something among a crowd of mannequins in his workshop. If the mannequins started to move, it would be like a horror story.”28
Mori’s use of the expression “horror story” is significant, because there is a quantitative dimension to the uncanny. Uncanny things are on a spectrum extending from the merely creepy (for example, a prosthetic hand) to those that elicit feelings of sheer horror (for example, the moving mannequins). The feeling of horror is not the same as the feeling of fear. As we proceed, I will articulate more clearly what it is that separates horror from fear. For now, though, I want to focus on just one element—one that applies in equal measure to all uncanny things. Things that are uncanny, whether creepy or horrifying, have a peculiar allure that distinguishes them from things that are merely repulsive or terrifying. This is nicely illustrated by Plato’s story about Leontius in the Republic:
Leontius, the son of Aglaeon, was on his way up to the town from the Piraeus. As he was walking below the North Wall, on the outside, he saw the public executioner with some dead bodies lying beside him. He wanted to look at the bodies, but at the same time felt disgust and held himself back, but at the same time he was disgusted and turned away. For a time he struggled and covered his eyes. Then, desire got the better of him. He rushed over to where the bodies were, and forced his eyes wide open saying “There you are, curse you. Have a really good look. Isn’t it a lovely sight?”29
I do not think that “disgust” is the right word for what drove Leontius’s rubbernecking impulse. Truly disgusting things do not draw the eye toward them, as these corpses did. Instead of being repelled, Leontius behaved like a person watching a horror film who covers their eyes with their hands when things get intense, but then just can’t resist peeking out through the gaps between their fingers.30
For all their insightfulness, neither Jentsch nor Mori got to the bottom of the uncanny. To get closer to the core of the kind of uncertainty or ambiguity that makes things seem uncanny, consider this chillingly evocative passage from Arthur Machen’s novel The House of Souls: “What would your feelings be, seriously, if your cat or your dog began to talk to you, and to dispute with you in human accents? You would be overwhelmed with horror. I am sure of it. And if the roses in your garden sang a weird song, you would go mad. And suppose the stones in the road began to swell and grow before your eyes, and if the pebble that you noticed at night had shot out stony blossoms in the morning?”31
Machen’s examples are at the horror end of the uncanniness spectrum. He was, after all, a writer of horror fiction. Consider the first example—that of a talking pet. Being confronted with a talking dog would certainly be disorienting, but this would not be because you are uncertain whether this entity is a dog or a human being. Rather, it would be because the talking dog has properties that are unique to dogs (its canine appearance) as well as properties that are unique to humans (the ability to speak). You are not wondering whether this creature is a dog or a human. Your reaction is driven by the fact that it seems to be both, but being a dog is incompatible with being a human. It is not uncertainty that elicits the uncanny feeling, it is contradiction. Similar considerations apply to singing roses, and the examples of stones that grow and shoot out blossoms similarly involve an impossible combination of the mineral and the botanical.
Machen’s passage points to several fundamental features of uncanny things. The first is that they all involve categorical contradiction. Uncanny things seem to transgress the categories that we use to make sense of the world. Whereas Mori’s paper was about the categorical boundary between humans and robots (and human limbs and prosthetic limbs), Jentsch had a more expansive view of the sources of uncanniness, and allowed that other kinds of categorical contradiction can elicit this disturbing feeling. This more general conception of categorical contradiction has received experimental support from a study conducted by Eva and Patrick Weiss showing that cognitive conflict occurring at category boundaries need not always involve the category “human,” but also that ambiguity about whether or not an entity is human produces the most pronounced effect.
The second feature of Machen’s examples is that they all involve things that we think of as natural kinds (although artifacts can figure in them, as is shown by the wax museum and robot examples). And third, they all involve living things (although nonliving things can figure in them, as is shown by the example of the sprouting stones). Now, putting these three elements together, we get the following specification: a thing is uncanny if and only if it is a contradictory living (or once living) thing that violates the boundaries that we take to demarcate biological natural kinds from one another.
A moment’s reflection reveals why this might be. As I explained in Chapter 4, we tend to essentialize natural biological kinds, and because essences do not come in degrees—a thing cannot have more or less of them—we take essences to demarcate those kinds absolutely. So, if a single entity is categorized as belonging to two different natural kinds, these representations are irreconcilable. A being classified as an insect and as a human being is not felt to be “sort of” an insect and “sort of” a human being—an insect-like human or a humanoid insect—but is represented as completely an insect and completely a human being. Dehumanized people are experienced as uncanny by their dehumanizers, because they violate the human / subhuman boundary. They are conceived as wholly human and as wholly subhuman, but these two representations of the dehumanized person cannot be reconciled with one another. The dehumanizer’s mind is pulled in two directions at once, and it cannot settle on either of the two mutually exclusive alternatives. The dehumanizer’s consciousness oscillates between them, thereby giving rise to the problem of humanity.
Why are these contradictory, metaphysically transgressive representations so disturbing? To answer that question, we need to draw on the work of a different thinker, who approaches the same conceptual territory from a different direction.
Mary Douglas was a British anthropologist who wrote an extraordinarily influential book, published in 1966, entitled Purity and Danger. The book is an anthropological study of ritual uncleanliness. Its basic insight is that whatever does not fit into the framework of categories that one’s culture uses to order the world is felt to be dirty, abominable, and polluting. “Dirt is never a unique, isolated event,” Douglas writes. “Where there is dirt there is a system. Dirt is the by-product of a systematic ordering and classification of matter, in so far as ordering involves rejecting inappropriate elements. Dirt is the by-product of the systematic ordering and classification of matter.… In short, our pollution behavior is the reaction which condemns any object or idea likely to confuse or contradict cherished classifications.”32
Every culture has some conception of the natural order: a framework of categories that are used to make the world intelligible. And these concepts of the natural order are used to underwrite the social order. Normally, we think of claims as either descriptive or normative. They state either how things are or how they should be. But the form of thinking that Douglas describes does not conform to this pattern. The natural order is how the world is arranged, and it is how the world should be arranged. But there is also a realm of the unnatural. These are unnatural things—things that exist but are outside the natural order. They are harbingers of pollution and chaos. The idea of the unnatural inevitably accompanies systems that purport to describe the natural order, because there are always things that have no proper place in the framework. These anomalous things are experienced as powerful and dangerous, and must therefore be segregated, marginalized, controlled, or destroyed.
Douglas’s formulation goes significantly further than those set out by Jentsch and Mori. Unlike theirs, it attends to the social and political underpinnings of classificatory schemes and does not limit the disturbing effect to clashing perceptual (primarily visual) signals. For Douglas, abomination is primarily a matter of how we classify things rather than how they appear to us. This comports very well with my theory of dehumanization. Dehumanized people are regarded as anomalous beings, but this is not because of how they appear. We classify them as human on the basis of their appearance, and as subhuman on the basis of what we have been told. And it is this double consciousness of the dehumanized other that derives the most toxic consequences of the dehumanizing process, because it turns people into monsters.
Addressing the problem of humanity in this way goes well beyond the problem itself, and opens the door to a much deeper understanding of facets of the dehumanizing process. Because dehumanizers conceive of those whom they dehumanize as simultaneously human and subhuman, they also think of them as creepy, horrifying, defiling, and this helps to explain why they are driven to dominate, or exterminate, them and why these people are so strongly associated with filth and disease. But it also tells us a lot more, because the solution to the problem of humanity also turns out to be a solution to the problem of monstrosity.
As I mentioned in the preface to this book, there are at least two kinds of dehumanization. Sometimes, dehumanized people are experienced as docile creatures, as beasts of burden or objects of ridicule. We can see it, to give but one example, in the Nazis dehumanization of disabled people. Johann Chapoutot begins his book The Law of Blood: Thinking and Acting as a Nazi with the example of eighteen Nazi physicians who were put on trial for murdering fifty-six disabled children as part of the Nazi euthanasia program. The charges against them were dismissed by the Hamburg regional court in 1949. Chapoutot writes, “The hospital’s director, Dr. Wilhelm Bayer, objected strenuously to the charge of ‘crimes against humanity.’ Such a crime, he asserted, ‘can only be committed against people, whereas the living creatures that we were required to treat could not be qualified as “human beings.”’ ” Another Nazi physician, Werner Catel, who also participated in the Aktion T4 euthanasia program was interviewed in 1964 in the German magazine Der Spiegel. When the journalist pointed out that the death penalty had been abolished in West Germany, Catel responded, “But don’t you see that when a jury makes a decision it is always judging human beings, even if they are criminals? We are not talking about humans here, but rather beings that were merely procreated by humans and that will never themselves become humans endowed with reason and a soul.”33
I call this kind of dehumanization “enfeebling dehumanization.” When people are dehumanized in this way, they are not considered to be horrifying, dangerous monsters (the sense of “monster” that I have been using in this book). This qualification is important, because there is a long history of referring to disabled people as “monstrous.” I have not had much to say in this book about enfeebling dehumanization. This is not because it is unimportant. It is because it has a different phenomenology and different dynamics from the kind of dehumanization that I have concentrated on here, which I call “demonizing dehumanization.” When people are dehumanized in this way, they are regarded as sinister and malevolent. It is in these cases that the problem of monstrosity arises.
Daniel Jonah Goldhagen is one of the few writers who tries to explain the relation between dehumanization and demonization. Goldhagen objects to the use of “dehumanization” in very general sense because it elides two “separate conceptual dimensions.” So, instead of subsuming demonization under an umbrella concept of dehumanization, he contrasts the former with the latter:
A belief (really an assemblage of beliefs) exists that can be properly said to be the dehumanizing of others. It is that other people inherently lack qualities fundamental to being fully human in the sense of deserving moral respect, rights, and protection. Such beings are said to lack human capacities and powers and, as a definitional matter, do not need to be treated as humans.… A second belief (also complex and various) is the demonization of others. The belief is about other people’s moral quality, including their moral intentions. It holds the people to be, literally or figuratively, demonic, morally evil.… The dimension of dehumanization is mainly about biological (cognitive, physical, etc.) capacity, held to be impaired. The dimension of dehumanization is mainly about moral character, held to be depraved, or so debased the people might as well be depraved.34
I do not think that this way of characterizing the relation between dehumanization and demonization is very helpful. One reason is that it is purely descriptive, without any underlying theoretical structure—rather like classifying chemical elements on the basis of their color rather than on the basis of their microstructure. Second, it does not work very well descriptively. People who are demonized are not simply seen as having poor moral character. In fact, they may be regarded as amoral. What is crucial is that they are regarded as predatory—as physically threatening. My third worry about Goldhagen’s schema is that the term “demonic” does not have any real theoretical heft. He might just as well have used an ordinary, vernacular term such as “immoral” or “depraved.”
Goldhagen’s framework leads him to miss or misrepresent certain aspects of dehumanization (in my sense of the word): “Dehumanization and demonization do not necessarily go hand in hand. The dehumanized are not always demonized. Whites enslaving blacks in the American South, and later repressing them with Jim Crow, dehumanized them, deeming them subhumans with diminished intellectual (and moral) capabilities, akin to domesticated, semi-wild animals that, when kept in check, were useful but if let loose could be dangerous. Whites did not construe their slaves, or freed blacks, as malevolent demons bent upon harming them.”35 This passage is accurate up until the final sentence. Think back to my account in Chapter 1 of how Black male victims of lynching (and Black Americans generally) were represented in the mass media of the day. Black men were, unequivocally, represented as dangerous beasts, bent on rape and murder—a dehumanizing stereotype that persists to this day. To claim that Black males, conceived as superpredators, were not demonized is bizarre. In the United States, the image of the Black male is the paradigmatic case of demonizing dehumanization.
Goldhagen’s misstep here may be due to his thin, atheoretical notion of the demonic. Unlike Jews, who for centuries have been described as literally demonic, Black Americans have rarely been referred to explicitly as demons. This shows how important it is for a theory of dehumanization to have a theoretically rich notion of the demonic. As I use it, “demonic” is a theoretical term of art that is closely connected to but not exactly coextensive with its vernacular meaning. As I use the terms, the demonic is synonymous with the monstrous, and I therefore use them more or less interchangeably.
To give an analysis of what a demon / monster is that can address the problem of monstrosity, it is helpful to turn to the writings of the philosopher Noël Carroll. Carroll is well known for his work on horror fiction. One of the questions that he investigates, and attempts to answer, is that of what distinguishes monsters—or in his terminology, “horrific monsters”—from other kinds of beings. Carroll draws on Douglas’s theory of uncleanliness to explain what it is for a thing to be a monster. To count as a monster, he argues, an entity must be physically threatening. If an entity does not pose a physical threat to human beings, then it is not a monster. But physical menace is not sufficient, because there are many physically threatening beings that are not monsters. The female anopheles mosquito—a tiny insect that carries malaria—is physically threatening, but it is not a monster. Neither are venomous snakes, or people who commit homicide. All of these beings can be frightening, because fear is the natural response to perceived threat. But they are not horrifying, as monsters are.
Carroll argues that monsters have got to be cognitively as well as physically threatening. What makes a monster cognitively threatening is its subversion of the natural order. A monster is a being that belongs to two or more incompatible natural kinds simultaneously, “a composite that unites attributes held to be categorically distinct and / or at odds with the cultural scheme of things in unambiguously one, spatio-temporally discrete entity.”36 For Carroll, “Demonically possessed characters typically involve the superimposition of two categorically distinct individuals, the possessee and the possessor, the latter usually a demon, who, in turn, is often a categorically transgressive figure (e.g., a goat-god). Stevenson’s most famous monster is two men, Jekyll and Hyde, where Hyde is described as having a simian aspect which makes him appear not quite human. Werewolves mix man and wolf, while shape changers of other sorts compound humans with other species.”37
I prefer the term “metaphysical threat” to Carroll’s “cognitive threat.” “Cognitive threat” suggests a threat to a person’s mental representations of the world, but the kind of threat that monsters pose is a threat to the natural order itself.
Carroll’s analysis of fictional monsters is immensely important for the analysis of dehumanization. The most dangerous and destructive kind of dehumanization transforms others into monsters. Members of the dehumanized group are thought to be dangerous. They are said to be vicious, predatory, cruel, destructive. And they are also felt to be metaphysically threatening because they are seen as both human and subhuman. This makes them seem immensely dangerous in the eyes of their dehumanizers, and explains why it is that although dehumanized people are typically among the most vulnerable members of a population, they are typically regarded as overwhelmingly dangerous. Also, because it is almost always male members of the dehumanized group that are regarded as physically threatening, these human monsters are almost always male.38
It is important to bear in mind that, in my view, making people into monsters is a consequence of dehumanization rather than its aim. The aim of dehumanization is the demotion of others to a lower-than-human metaphysical rank, to disactivate inhibitions against harming them. The aim of dehumanization is to turn people into animals, but because it is next to impossible to completely shut off awareness of their humanity, it turns them into monsters instead.