Sketch of My Account of Dehumanization

Research into dehumanization has been and is almost entirely in the field of social psychology. Philosophers have mostly been latecomers to the table. As I have explained elsewhere (Smith Reference Smith2021, Reference Smith2023) there is not a consensus amongst psychologists about what dehumanization is, much less about its causal structure. Three main positions have been adopted by these psychologists. The most influential thesis is that to dehumanize others is to conceive of them as lacking what Haslam calls “uniquely human” or “human nature” attributes, rendering them animal-like or machine-like respectively (Haslam, Reference Haslam2006, Reference Haslam and Kronfeldner2021). Another prominent theory is that to dehumanize others is to regard them as having diminished mentality (Waytz and Epley, Reference Waytz and Epley2012), and a third is that dehumanized people are people who are denied the capacity for agency and experience (Harris and Fiske, Reference Harris and Fiske2006). I consider all of these accounts to be inadequate insofar as they fail to comport with paradigmatic examples. For instance, Harris and Fiske have it that dehumanized people are regarded as lacking warmth and competence. They write that this account “predicts that only extreme out-groups, groups that are both stereotypically hostile and stereotypically incompetent (low warmth, low competence), such as addicts and the homeless, will be dehumanized.” However, Nazis regarded Jews as conspiring to rule the world, and thus they were granted a high degree of agency by those who dehumanized them (for a thorough critique of standard social psychological conceptions of dehumanization, see Smith [Reference Smith2023]). I am also uncomfortable with pluralistic accounts of dehumanization (Kronfeldner Reference Kronfeldner. and Kronfeldner2021, Reference Kronfeldner2024) because they invite confusion about an already contested concept by grouping empirically disunified phenomena under a single semantic umbrella.

In contrast to these psychological accounts, as well as alternative philosophical accounts (e.g. Mikkola Reference Mikkola2016; Thomason, Reference Thomason, Thomas and Jackson2019, de Ruiter Reference de Ruiter2023, Phillips Reference Phillips2023, Killmister Reference Killmister2024; Kronfeldner Reference Kronfeldner2024, McDonald Reference McDonald and Popa-Wyatt2024; Tarasenko-Struc Reference Tarasenko‐Struc2025), I conceive of dehumanization as the attitude of conceiving of others as less than human creatures, where “less than human” should be understood as “essentially having less intrinsic value than is accorded to human beings.” That dehumanized people are regarded as having less-than-human intrinsic value is compatible with their being thought to possess greater-than ordinary powers such as extraordinary physical strength, prodigious sexual appetites, and even preternatural intelligence. I understand dehumanizing attitudes to be beliefs or belief-like states. I am agnostic about the exact nature of these representational states (for criticism of the claim that dehumanizing representations are beliefs, see Schwitzgebel [Reference Schwitzgebel2020]). I distinguish dehumanization from sexism, racism, misogyny, ableism, transphobia, and xenophobia simpliciter. That is not to say that dehumanization is unrelated to these attitudes and practices. In fact, both racism and certain kinds of ableism are quite closely related to it, as I will illustrate below. But a relationship, however intimate, is not an identity; dehumanization is, on my view, a phenomenon sui generis.

Although I characterize dehumanization as a psychological phenomenon, I do not think that it can be properly understood in exclusively psychological terms. Dehumanization is a psychological response to social forces, and is not explicable without taking those forces into account. A purely psychological account of it is inevitably partial and impoverished. By analogy, although a tennis player moving their body to the right can be described in terms of neuromuscular events occurring in their body, it is impossible to make sense of a tennis player repeatedly moving their body to the right if one does not know that the ball is being hit in that direction and that the rules of the game—both of which are non-physiological facts—require that the player try to hit the ball back across the net. This said, the emphasis in the present article is on the intrapsychic components of dehumanization, rather than on the sociopolitical ecology that elicits it.

There are several forms of dehumanization, but they all have something in common that I call metaphysical threat. Things that are perceived as metaphysically threatening superimpose representations of mutually exclusive natural kinds in a single, contradictory entity. Specifically, dehumanized people are represented in the minds of their dehumanizers as wholly human and wholly subhuman. To briefly explain and recapitulate arguments I have made elsewhere (Smith Reference Smith2021), members of our species are highly attuned to indications of humanness in others. Under normal circumstances, when confronted with another member of our species we automatically “see human.” However, epistemic deference is also a key feature of human life. We place certain individuals in the social role of “expert”—persons who are supposed to know. When such experts tell us that the world is not as we believe it to be, we normally defer to them, even when their description does not comport with the verdict of our senses. This is often warranted. When a physicist tells me that the seemingly gapless chair that I am sitting on is actually an aggregate of miniscule particles floating in empty space, I accept this even though it seems to contradict what my eyes tell me is the case. Likewise, we are inclined defer to genuine or merely putative experts who tell us that some group of seemingly human others are not really human at all. In such cases, then, we both regard some others as human (on the evidence of our senses) and as subhuman (on the evidence of testimony of an epistemic authority), resulting in two simultaneous, contradictory representations of them. Dehumanized people are thus not just seen as “lower” animals, but as metaphysically threatening, monstrous human animals, and therefore regarded as a different and more disturbing subhuman than animals are. I am using “animals” here in a conventional, vernacular sense which differentiates humans from animals and situates the former as higher on a hierarchy of value than the latter.

The concept of metaphysical threat owes a great deal to Noel Carroll’s (Reference Carroll1990) analysis of horror fiction. Carroll argues that the monsters that populate horror fiction are so disturbing because they violate metaphysical categories that are regarded as mutually exclusive. Consider the difference between a wolf and a werewolf. A wolf might be threatening, but a werewolf—a monstrous composite of wolf and human—is horrific. I argue that the contradictory cognitive state experienced by dehumanizers elicits a peculiar mental disturbance in them. I am not sure there is a completely satisfactory label for this mental state, though “horror,” “creepiness,” and “uncanniness” are all in its neighborhood. Dehumanized people are disturbing because they seem to transgress the boundaries that are thought to demarcate natural kinds from one another, and are, like the monsters of horror fiction, felt to violate the order of nature. The notions of nature, naturalness, and unnaturalness are therefore key components of my theory of dehumanization.

David Hume somewhat hyperbolically remarked in his Treatise of Human Nature, that of all our concepts, “there is none more ambiguous and equivocal” than “nature” (Hume 1978: 474). Hume proceeded to identify several notions of the natural: what is natural is not miraculous, or is not artificial, or is what is ordinary. But there is another conception of the natural that Hume does not discuss, a notion of the natural that is at once metaphysical and normative (see Mill Reference Mill1874)—that naturalness corresponds to the proper order of the world. Things that are deemed to be natural, in this sense, are regarded as wholesome and pure (think of “natural foods,” “naturopathic medicine,” “the harmony of nature,” and so on). Things regarded as departing from this natural order are considered to be not merely fictional, unusual, or artificial, but unnatural—unwholesome, pathological, perverse, and monstrous. This notion of the natural is a folk-metaphysical construction that is entrenched in our culture and commonly lurks in the prereflective background of our thoughts, and is easily elicited. Dehumanized people are felt to be unnatural in precisely this sense.

Because dehumanization involves the compresence of the human and the subhuman in a single entity, it is unstable. The contradictory representations are both conjoined with one another (because they are representations of the same thing) and cognitively repel one another (because they are contradictory). This generates a peculiar state of psychological tension, with the dehumanizer alternating back and forth between thinking of the other as human, thinking of them as subhuman, and thinking of them as a monstrous hybrid of the two. A counterintuitive entailment of this analysis is that when people seemingly regard others as nothing more than animals, they are probably not dehumanizing them.

The kind of dehumanization that I have researched most extensively is what I call demonizing dehumanization. People who are dehumanized in this way are regarded as both physically threatening (rapists, murderers, etc.) and also metaphysically threatening (in virtue of being regarded as being human and subhuman simultaneously). This psychological cocktail of physical and metaphysical threat mixes fear with uncanniness, terror with fascination.

Demonizing dehumanization is closely tied to race, because groups that are demonically dehumanized are usually first racialized. Demonizing dehumanization is also highly gendered, mainly targeting male members of the racialized group. We see this in the demographics of lynching. Of the more than four thousand people lynched in the southern United States, at least 97 percent were African American men. These men were regularly described in the media of the day, as well as in the broader racist literature, as subhuman beings (see Smith, Reference Smith2011, Reference Smith2021; Curry, Reference Curry2017), just as, more recently, the term “superpredator” was reserved for characterizing young, African American men (Haberman, Reference Haberman2014). This pattern repeats in other episodes of racialized mass violence (see also Jones Reference Jones2004). For example, although there are many examples of Jewish men portrayed as subhuman creatures in Nazi propaganda, I am not aware of a single example of a Jewish woman depicted as less than human. Jewish women are portrayed as ugly and disgusting, but not as monstrous or demonic beings with subhuman attributes. Demonizing dehumanization mainly targets men for a reason. Demonizing dehumanization is biased toward men because it involves both physical threat and metaphysical threat, and it is primarily men rather than women that are deemed to be physically threatening. There is an intimate relationship between racialization and physical threat. When a group of people become racialized by the dominant group, the men of the racialized group are typically experienced as physically dangerous. They are deemed to be murderers, rapists, terrorists, and so forth (see, for example, Berkowitz Reference Berkowitz2007 and Muhammad Reference Muhammad2019).

There are other kinds of dehumanization that do not involve physical threat. I group these under the heading “enfeebling dehumanization.” People who are dehumanized in this way (for example, some people with disabilities) are felt to be metaphysically but not physically threatening. Finally, dehumanization is not an accidental state—a mere oversight or failure to recognize the humanity of others. Rather, it has the function of disinhibiting violence. All social species must be equipped with inhibitions against violence against members of their communities. In the human case, these inhibitions extend beyond the immediate breeding group to those beyond it. Consequently, violence—especially lethal violence—does not come easily to us (Collins Reference Collins2009, Grossman Reference Grossman2009). However, over time we have developed cultural technologies to selectively disable inhibitions against violence. Dehumanization is one of these (others include the use of intoxicating substances, mind-altering rituals, religious ideologies, and the development of long-distance weapons).

There is much more to my account of dehumanization than I am able to summarize here, but I hope that the foregoing is sufficient to make the rest of this paper comprehensible to those who are unfamiliar with my work.

Problems of Interpretation

Work on dehumanization addresses morally weighty empirical questions that standard empirical methodologies do not, and perhaps cannot, properly address. It is concerned with messy, morally urgent matters of fact that are implicated in enormous human suffering. A methodological constraint that anyone working on dehumanization must inevitably encounter (whether they are aware of it or not) is that inferences about dehumanization (or ostensible dehumanization, so as not to beg the question) require one to interpret the ways that people represent others, in speech, in thought, and in action. In doing this, one inevitably swims in murky hermeneutic waters; for any pattern of human behavior, there are always alternative interpretive strategies available, and therefore always alternative interpretive conclusions to be had. One might seek epistemic relief in the proposal that all of these alternatives, provided that they are coherent, are equally credible—that there is no singular fact of the matter about whether or not dehumanization, in the sense that I have laid out, is occurring. I am uncomfortable with an easygoing relativism that supports the view that whether or not dehumanization is occurring is dependent on one’s interpretive preferences. So, in what follows I will presume that whatever one takes dehumanization to be, it is a factual matter, rather than just a matter of hermeneutical taste, whether or not, in any given case, it is occurring.

Some readers acquainted with my work may have noticed that the title of my paper “Paradoxes of dehumanization” (Smith Reference Smith2016)—the paper that initiated my ongoing exchange with Kate Manne—echoes the title of Donald Davidson’s (Reference Davidson, Wollheim and Hopkins1982) “Paradoxes of irrationality,” the paper in which Davidson mounted a measured defense of Freudian psychology. This was no accident. From the start of my research into dehumanization, nearly twenty years ago, Freud and Davidson have been major influences on my thinking. Most recently, the differences between their perspectives about the character of mental states have helped me understand my disagreement with Kate Manne and other dehumanization skeptics more deeply than before.

It hardly needs saying that both Freud and Davidson thought a lot about interpretation. Davidson was concerned with what ordinarily goes on when we ascribe mental states to others, using the commonsense framework often described as “belief-desire psychology” or “folk psychology.” Davidson argues that we interpret the behavior of others in such a way that it makes sense. As he put it:

Freud did not dispute the claim that our ordinary interpretive practices rely on assumptions of coherence, rationality, and consistency. However, he argued that human psychology is inherently contradictory, and that interpretations of behavior that are based on criteria of coherence, rationality, and consistency are often misleading. Both clinically and theoretically, psychoanalysis is predicated on a view of the mind that contrasts starkly with the commonsense psychological one that is the core of Davidson’s approach. Freud consistently emphasized mental incoherence, irrationality, and inconsistency, writing (to give just one of many examples):

As the philosopher Israel Levine summed up, “we might say that the unconscious system has no logic” (Levine Reference Levine1923: 127). Freud used the term “compromise” in a technical sense for a single mental representation that serves as a proxy for two antithetical ones. He refers to this phenomenon as “condensation” in the context of his theory of dreaming, and elsewhere as “ambivalence”, a term that he borrowed from Swiss psychiatrist Eugen Bleuler (Reference Bleuler1910), who distinguished between affective, volitional, and cognitive forms of it (Riklin, Reference Riklin1910). Freud deepened Bleuler’s conception, especially his notion of cognitive ambivalence. Bleuler held that cognitive ambivalence (or as he called it, “intellectual ambivalence”) occurs when one simultaneously believes a proposition and its negation. For Freud, ambivalence occurs when contradictory, emotionally-charged mental representations coexist in a single mind, segregated from one another. As he put it in Totem and Taboo, “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” (Freud, Reference Freud and Strachey1953: 30). From this perspective, being ambivalent is the very opposite of having “mixed” feelings: the contrary attitudes are prevented from mixing. Both exert an influence on the person, but they do not enter into a rational relation with one another.

Davidson suggests that when some piece of behavior fails to comport with an existing interpretation, we adjust that interpretation to restore coherence, rationality, and consistency. Although, strictly speaking, this way of looking at things does not rule out the core Freudian tenet of psychological contradictoriness, it favors interpretations of others as coherent overall (for a different assessment of the Freudian/Davidsonian interface, see Hopkins [Reference Hopkins, Wollheim and Hopkins1982] and Evnine [Reference Evnine1989]). Even though Freud was probably an anti-realist about propositional attitudes (Smith Reference Smith1999), he would agree that Davidson’s account captures certain of our ordinary interpretive practices. This is evident in his discussion of the use of the problem of other minds as a basis for inferences about mental incoherence. In a paper that strikingly anticipated elements of Davidson’s (Reference Davidson, Wollheim and Hopkins1982) argument published nearly seventy years later, Freud proposed that we infer the existence of other minds by drawing “an analogy from their observable utterances and actions, in order to make this behavior of theirs intelligible to us” (Freud Reference Freud and Strachey1957: 169). The procedure that we deploy to make psychological sense of others can also be applied to oneself to infer mental incoherence.

Freud rejected the view that renegade mental states that do not cohere with the bulk of the subject’s mental states—that is, those mental states that are refractory to Davidsonian norms— should be understood as belonging to a person within the person. In doing so he ultimately rejected the “dissociationist” approach (Smith Reference Smith1999) that was common in his day. Dissociationists believed that all mental states and processes are conscious states and processes. But they also believed that consciousness can be divided into various, quasi-independent sub-consciousnesses. As William James, summarizing Pierre Janet’s view, put it, “An hysterical woman abandons part of her consciousness…. The abandoned part meanwhile may solidify into a secondary or sub-conscious self” (James, Reference James1950: 210). This is why in rejecting dissociationism, Freud consistently avoided the term “subconscious,” which belonged to the dissociationist paradigm, in favor of “unconscious.” The dissociationists’ informal procedure for distinguishing a person’s primary consciousness from one or more secondary consciousnesses was based on what we nowadays think of as Davidsonian interpretive principles: those psychological elements that fail to cohere with the subject’s main consciousness were assumed to belong to a secondary or tertiary consciousness of which the main consciousness is unaware. While agreeing that we ordinarily apply the principle of charity when inferring other minds and attributing content to those minds, Freud firmly rejected an analogous charity-based approach to inferences about unconscious mental phenomena on the grounds that it leads to the absurd idea of an unconscious consciousness, that it licenses an indefinitely large number of subsidiary consciousnesses, and—perhaps most importantly—that all such ascriptions fall foul of the non-logical character of unconscious mentation.

Delving more deeply here into the convergences and contradictions between the Freudian and Davidsonian approaches to interpretation, interesting though they are, will lead readers away from my aim in invoking them. My point in raising them is to identify two interpretive stances: the view that mental ascriptions should maximize coherency, rationality, and consistency, versus the view that much of our mental life is incoherent, irrational, and inconsistent, and is therefore not capturable in the Davidsonian net. I do so because the tension between Davidson and Freud is pertinent to the theory of dehumanization generally, and particularly because it reveals the meta-hermeneutical foundation of my disagreement with Kate Manne.

To see how these alternative interpretive strategies affect how we assess ostensibly dehumanizing discourse, consider this excerpt from an interview published in The Independent newspaper. The interviewee is a person named Maria, who was a perpetrator in the 1993 Hadareni pogrom—mob violence that resulted in the murder of three Roma men and the destruction of 13 Roma homes in Hadareni, Romania.

Notice how Maria alternated between representing Roma people as human and representing them as subhuman. “Gypsies” are “not really people,” but they are “criminals”—a label that is reserved for human beings. But then again, they are “sub-human,” “vermin.” What are we to make of these blatantly incompatible claims? This question should be central to any account of dehumanization, because representations of the other as both human and subhuman is typical of dehumanizing speech. Here are just a few examples. Nazi propaganda described Jews both as Untermenschen (subhumans) and as criminals (Berkowitz, Reference Berkowitz2007). Radical Buddhists in Myanmar described Rohingyas both as criminals and as reincarnations of vermin (Smith Reference Smith2020). Sylvia Wynter told us: “[P]ublic officials of the judicial system of Los Angeles routinely used the acronym N.H.I. to refer to any case involving a breach of the rights of young Black men who belong to the jobless category of the inner-city ghettoes. N. H. I. means ‘no humans involved’” (Wynter Reference Wynter1994: 42), but only humans are subject to the judicial system. President Trump has remarked that “illegals” (that is, undocumented immigrants to the United States) are “not human” and are “animals,” but nonetheless calls for their arrest and deportation (one does not arrest and deport members of an invasive animal species, although one might capture and release them elsewhere). The historian of slavery David Brion Davis noted the same phenomenon in a discussion of the dehumanization of African Americans during and after slavery, writing, “I would only add that the victims of this process [of dehumanization] are perceived as ‘animalized humans,’ this double consciousnessFootnote ¹ would probably involve a contradictory shifting back and forth in the recognition of humanity” (Davis 2018: 15).

One way of making sense of examples like that of Maria’s testimony is to interpret them in such a manner as to make them seem coherent. The most attractive way of doing this, for a number of reasons, is to interpret Maria’s claims about Roma subhumanity non-literally. One might propose that Maria recognized that Roma people really are human beings, and that her statements about their subhumanity should be interpreted as derogatory speech that does not license the inference that she mentally represented them as subhuman entities. This is an aspect of Manne’s position, as well as that of other dehumanization skeptics.

The Davidsonian approach to Maria’s discourse is attractive, for several reasons. One is that real dehumanization does not or would not comport with our ordinary interpretive practices. My account of dehumanization would have it that Maria both recognized that Roma are human beings and conceived of them as not really human. But this seems prima facie impossible. The Davidson-style interpretation is supported by the fact that characterizations of others as less than human are often nothing more than derogatory speech that is intended to harm or humiliate them (Smith Reference Smith2020) without expressing the belief that they are literally subhuman. In contrast, my interpretive stance veers more towards the “Freudian” end of the spectrum. I take the position that Maria’s assertions that Roma are human and her assertions that they are not human both reflect her genuine beliefs about Roma people, and that we should not interpret her as being in a coherent mental state (Smith Reference Smith2020, Reference Smith2021).

Testimony

Most putative examples of dehumanization are ambiguous: we can interpret them in either a “Davidsonian” or a “Freudian” way. So, how can we choose what stance to take in any given case? I think that the best way to resolve this problem is to take the testimony of perpetrators and victims seriously. There are examples of dehumanizing speech that are unambiguous—statements of the sort that, by ordinary epistemic standards, most of us would take to express sincerely held beliefs. Such testimony should be taken seriously rather than interpreted away because it fails to accord with one’s theoretical predilections. Acknowledging the existence of unambiguous cases of dehumanization does not tell us anything about its prevalence, but it does require us to be open to the possibility that ambiguous cases, such as that of Maria, may be genuine examples of dehumanization.

I begin with some examples from perpetrators which readers may find disturbing, but which are crucial for describing dehumanizing attitudes and the actions that often flow from them. The Nazi T4 program, misleadingly described as a “euthanasia” program, began with the murder of children with cognitive disabilities, five to eight thousand of whom were killed by gassing, lethal injection, or starvation in what were called “special children’s wards” in various hospitals in Germany and Austria. The director of one of these killing centers, the Rothenburgsort Children’s Hospital in Hamburg, was a pediatrician named Wilhelm Bayer. After the war Bayer was charged, along with his medical colleagues, with crimes against humanity for his role in the murder of fifty-six children. Here is how Johann Chapoutot describes what happened, in his book The Law of Blood: Thinking and Acting as a Nazi:

Bayer’s defense succeeded, and he was able to retain his medical license and continue his career. A dehumanization skeptic might suggest that Bayer was being disingenuous in claiming that these children were not really human beings. Perhaps, they might suggest, he knew very well that these children were human beings, and that his claim that they were not human was just a ploy to avoid being convicted. This is possible, but implausible. Characterizations of children with cognitive disabilities as subhuman has a long history preceding the Nazi era. Martin Luther, believed that such children were changelings. “Satan has the power to exchange children,” Luther wrote, “placing a devil in the cradle in the place of a child. This devil will suck and eat like an animal…. The parents get no rest from such filthy beasts” (Luther Reference Luther1854a: 41).

Historian Dagmar Herzog notes:

Werner Catel, professor of pediatric medicine at the University of Leipzig, was also affiliated with the T4 program. He was interviewed in Der Spiegel in 1964 (Renner Reference Renner1964). The interview was titled “Killing from Humanity?” (“Aus Menschlichkeit Töten?”). Pressed by the interviewer, he made such comments as: “Believe me, it is definitely possible to distinguish these soulless beings from developing humans,” “The doctor…must discuss the situation with the parents. He must tell them the truth, namely, that this being can no longer be helped, that it will never become a human being,” “We are not talking about humans here, but rather about beings who were merely procreated by humans and that will never themselves become humans endowed with reason and a soul” (Renner Reference Renner1964: 41-47). Catel also continued his medical career after the Nazi era, and continued to advocate killing children with cognitive disabilities. He wrote in his 1962 book Liminal Situations of Life (Grenzsituation des Lebens) that such children are “monsters” and, echoing Luther, “nothing but a massa carnis” (Catel Reference Catel1962: 112).

Arguing that these men did not intend their remarks to be taken literally does not seem credible. However, much of the research into dehumanization pertains not to disabled people but rather to racialized people. So, skeptics might still fall back on the argument that the dehumanization of severely cognitively impaired children is categorically different from the ostensible racialized dehumanization. A hydrocephalic infant is, after all, very different from an African American lynching victim or a Jew exiting a cattle car at Treblinka. So, perhaps dehumanization is real, but operates in a much more restricted sphere than realists like me suppose.

It is true that most examples of racialized dehumanization are not as clear-cut as the medical examples that I have just given. But some are. Consider Morgan Godwyn’s report, published in 1680, that English advocates of slavery told him that “the Negros, though in their Figure they carry some resemblances of Manhood, yet are indeed no Men” (Godwyn Reference Godwyn1680: 3) and that they are “Creatures destitute of Souls, to be ranked among Brute Beasts, and treated accordingly” (Godwyn Reference Godwyn and Brokesby1708: 3) We also have statements by perpetrators such as Elie Ngarambe, a Rwandan genocidaire, who told his interlocutor, Daniel Johah Goldhagen “We did not know that the [Tutsi] were human beings….”. Goldhagen commented, “Ngarambe is emphatic that this was the common view and common knowledge” (Goldhagen Reference Goldhagen2009: 182).

Testimony from people who are targets of dehumanization is also revealing. In many cases they seem to have had no doubt that their persecutors regarded them as subhuman beings. Mia Bay comments in her book The White Image in the Black Mind, based on interviews with formerly enslaved people in the United States, “Indeed, the highest compliment ex-slaves had for former owners who had been good to them was often simply that these owners had recognized that their black bondspeople were humans, not animals” (Bay Reference Bay2000: 137). When Henry McNeal Turner, African American minister and pioneer of Black nationalism, addressed the Georgia legislature in 1868, saying, “A certain gentleman has argued that the Negro was a mere development similar to an orangoutang [sic] or chimpanzee” we should regard his testimony as credible (Turner Reference Turner, Foner and Branham1868). Likewise, we should accept that when Frederick Douglass stated that “The manhood of the Negro is fiercely opposed” and that, although Whites are granted the inalienable right to life and the pursuit of happiness, “the Negro has no such right—BECAUSE HE IS NOT A MAN!” he accurately represented his opponents’ views (Douglass Reference Douglass1854: 6-7).

Perhaps the hermeneutical dilemma can be resolved by consulting empirical psychological research. There is an extensive literature on dehumanization by social psychologists. This work is, I believe, largely both conceptually and theoretically inadequate (see Smith, Reference Smith2023) and therefore of limited value. But for the purpose of this discussion, it is noteworthy that with (as far as I am aware) only a single exception, when experimental subjects describe some group of others as less than human, they are not asked the key question of whether they mean this literally or figuratively. The single exception, at the time of writing, is a recent study by Landry and his associates (Landrey et. al. Reference Landry2025) who found that a significant proportion of interviewees confirmed that when they characterized members of a target group as subhumans they meant this literally.

The Strangeness Objection

It is worth noting that testimony about dehumanization is typically offered by people whose credibility is routinely cast into doubt. Victims of dehumanization are almost always members of marginalized groups whose views are frequently not regarded as credible by members of the dominant group. Much the same can be said of perpetrators, who are readily accused of seeking exculpation or of being psychologically or morally deranged. However, preemptively dismissing such testimony places the most important source of evidence supporting the reality of dehumanization beyond the epistemic pale, and may sometimes amount to an act of testimonial injustice (Fricker, Reference Fricker2009).

Probably the most significant motive for continuing doubt about dehumanization has to do with its strangeness. Commonsense or “folk” psychology—the kind that we rely on in our ordinary social interactions, and which is typically invoked by philosophers when making psychological claims—does not have the resources for making dehumanization intelligible, much less for explaining its dynamics. Consequently, as long as one views dehumanization through a commonsense psychological lens, its refractoriness seems to favor antirealism about dehumanization. The fact that people often use words such as “pig” or “monster” to denigrate others without regarding them as subhuman, combined with the fact that when people ostensibly dehumanize others, they also refer to them and treat them, implicitly or explicitly, as human beings, prompts skepticism about the reality of dehumanization. From a commonsense psychological perspective, it makes no sense to say that someone conceives of others both as human and as subhuman. From that perspective, then, it is easy to conclude that those who seemingly dehumanize others actually regard them as fully human. Call this the Strangeness Objection. The Strangeness Objection is at the heart of Manne’s critique, as well as that of others, such as Stanley Cavell. A passage from Cavell’s “Skepticism and the problem of others” lays out one version of it quite clearly. Cavell writes, “[I]t is sometimes said that slaveowners did not see or treat their slaves as human beings…some slaveowners have been known to say so…. But does one really believe such assertions?” One should be skeptical, he argues, because:

This objection goes back centuries. A version of it can be found in Morgan Godwyn’s Reference Godwyn1680 book The Negroe’s [sic]and Indian’s Advocate, which is perhaps the first published critique of dehumanizing claims. Godwyn lambasts slaveowners’ claim that enslaved people are subhuman, writing:

I am the first to admit that this argument carries some weight, hence its enduring appeal. However, it is ultimately unconvincing. The logical impossibility on which it turns is compatible with the psychological propensity to entertain contradictory beliefs. Cavell’s “Davidsonian” stance leads him to assume that the slaveowner’s attitude is coherent and stable, which is perhaps why he helps himself to the too-strong claim that everything in the slaveowner’s relation to his slaves shows that he treats them as “more or less” human. Leaving aside the equivocation implied by “more or less,” and leaving aside the fact that dehumanizing attitudes are not necessary features of slavery, even though they often accompany it, it is enlightening to contrast his remarks with those of David Brion Davis, who takes a more “Freudian” stance. In The Problem of Slavery in the Age of Emancipation, Davis recounts the story of a northerner traveling through the American South who was told by a local “I wouldn’t mind killing an African American more than I would a dog” (of course, he didn’t use the term “African American”). Davis asks, “Does this mean that blacks who were treated as animals were seen as ‘only animals’? The answer is clearly no, except perhaps in some extreme cases and for very brief periods of time…” (Davis Reference Davis2015: 15) He goes on to describe dehumanized people as animalized humans, and quotes Appiah’s observation that the thesis that dehumanized people are seen as not at all human “doesn’t explain the immense cruelty—the abominable cruelty” (Appiah Reference Appiah2010: 144) that is meted out to dehumanized people, and insightfully comments, “Clearly…retention of a human element fails to make animalization more humane. Quite the contrary” (Davis Reference Davis2015: 15-17). I do not accept all of the details of Davis’ account, but I think his appreciation of the contradictoriness and instability of dehumanization, and his recognition that it is precisely the contradictoriness of dehumanization that inspires its worst atrocities, are all correct. It is perhaps because of his métier as an historian that he was less prone than many philosophers are to mistake logical impossibility for psychological impossibility.

Another aspect of the Strangeness Objection pertains to what the denial of humanity (or attribution of humanity) amounts to. Cavell’s claim, “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” suggests that humanness is a matter of one’s phenotype, and his claim, “He does not go to great lengths either to convert his horses to Christianity or to prevent them from getting wind of it” suggests that being human is identical to belonging to a particular biological taxon (usually, that to be human is to belong to species Homo sapiens). But this does not correspond to how attributions of humanity really work (see Smith, Reference Smith2021). Having a certain sort of appearance and belonging to the species Homo sapiens can come apart, even though they are normally found together. There are plenty of members of our species that have atypical bodies and behavior, sometimes dramatically atypical bodies and behavior, and yet are nonetheless considered as belonging to our species. The historical record amply demonstrates that when people deny the humanity of others, they typically do not deny that these others are outwardly indistinguishable from human beings, and in cases of the dehumanization of people with atypical phenotypes (for example, Bayer’s infant victims) phenotypic deviance is not regarded as constitutive of their supposed subhumanity. It is not a person’s appearance (black hand versus black paw) that is taken to constitute their humanity, although appearance is usually taken to be (defeasibly) diagnostic of humanness or the lack of it. To give just one example, Nazis never denied that Jewish Untermenschen appear to be human. In fact, Jews were taken to be especially pernicious precisely because they could not easily be distinguished from Aryans, and hence needed to be marked with the yellow star. There is a substantial body of psychological research that indicates that we are disposed to (a) carve the world into natural kinds (including naturalized social groups [Rothbart and Taylor, Reference Rothbart, Taylor, Semin and Fiedler1992; Prentice and Miller, Reference Prentice and Miller2007]), and (b) assume that what makes any individual a member of one or another natural kind is its possession of an “essence”—a deep and unobservable property that all and only members of the kind possess, and which is normally (but not inevitably) causally responsible for the outward appearance that is typical for members of the kind. This disposition is known as “psychological essentialism” (Neufeld Reference Neufeld2022). In light of this, that Cavell’s imagined enslaver expects a black hand rather than a black paw has no bearing on the question of whether or not he regards that enslaved person as human or as less than human. It may be that the slaveowner who described enslaved people as subhuman creatures does not really believe this to be true. Apparent dehumanization is not always genuine dehumanization, but one cannot legitimately conclude that no apparent cases are genuine cases by pursuing the argumentative route that Cavell and his fellow travelers have taken.