Keywords

A popular twenty-first century t-shirt announces, “Hello. I am a person. What are you?” With globalization—immigration, collaboration, cyber-networks, intermarriage, multiculturalism—categorizing another person just got much more complicated. Although categorical biases are not gone, they are daily diluting, ambiguating, subtyping, contextualizing (Bodenhausen and Peery 2009). As categories multiply, subdivide, and recombine, individuation might seem likely (Swencionis and Fiske 2013). But more complicated categories also offer more opportunities for dehumanization. For example, when Italian Fascists encountered mixed-race people, they merely pitched them into an allegedly disgusting outgroup trash-bin (Durante et al. 2010). Which is it, then: Do multiple categories defy dehumanizing processes, or do they provide just another opportunity for bigotry? The answer is, “Both, and it depends.”

Recognizing or denying another’s humanity varies by degrees, along simple, predictable, and apparently universal dimensions. In making this argument, we are indebted to the storied Nebraska Symposium series and some of its previous contributors. As Donald Campbell (1965) noted, when nonspecialists in human motivation are invited, they must broaden the symposium by examining the relevance—however remote—of motivation to their own research. Our approach shares with Campbell the premise that people are motivated to protect their ingroup, thereby derogating outgroups. We argue that two dimensions adequately describe how they divide ingroup from a variety of outgroups: societal structures of interdependence (cooperative versus competitive) and status.

This chapter first provides background on how social cognition generally varies along these two predictable dimensions, then concentrates on evidence that outgroup dehumanization does too. Finally, as an antidote to the toxicity of dehumanization, the chapter turns to ingroup humanizing processes, exploring their what, who, why, how, and when, with a view to expanding people’s view of who exactly counts as human.

Warmth and Competence in the Stereotype Content Model

Consistent with fundamental motives for survival, the Stereotype Content Model (SCM; Fiske et al. 2002, 2007) argues that other people in our environment offer unique survival opportunities and challenges decided by their intentions. People are autonomous agents, so their motivational relevance derives from how their goals mesh with our own goals. Thus, we first need to learn their intentions for good or ill. Does this person or group intend to cooperate or compete with me and my group? If cooperative, they are warm, trustworthy people. If competitive, not. The others’ apparent warmth (trustworthiness, friendliness) stems from perceived interdependence (cooperation/competition), reflecting inferred good or bad intentions, respectively. Whether the other is an individual in a dark alley or a new immigrant group arriving at our borders, we need to know, like the sentry: “Halt, who goes there? Friend or foe?”

If intentions are key to survival, then we next want to know whether the others can act on their intentions. After all, if the others are incapable and ineffective, then their intentions matter less than if the others are capable and effective. Our judgment of their competence (capability) stems from perceiving the others’ apparent status and prestige. Status apparently reflects the ability to enact intentions.

Evidence for the SCM

People’s lay theories of groups’ warmth and competence derive from their understanding of social structure. Interdependence, measured as control over both tangible and symbolic resources, predicts perceived warmth, measured as a combination of friendliness/sociability and morality/trustworthiness (Fiske et al. 2002; Kervyn et al. under review).

On the status-competence dimension, although less-than-obvious at first glance, the relationship between ascribed status (prestigious jobs, economic success) and perceived competence is robust and substantial. Meritocracy rules (people get what they deserve), the world over, according to people’s reports (Cuddy et al. 2009; Durante et al. 2013).

The SCM structure-to-stereotypes pattern— status → competence and interdependence → warmth— fits into a chain of perceptions supported by both surveys (Cuddy et al. 2007; Fiske et al. 2002) and vignette experiments (Caprariello et al. 2009; Oldmeadow and Fiske 2007) (see Fig. 1). Together, the warmth x competence stereotypes predict emotional prejudices unique to each quadrant. Using survey responses to common societal groups, people report how society reacts to these groups, and subsequent data show that the reported societal responses also reflect individual reports in private.

Fig. 1
figure 1

Stereotype content model (SCM) chain of perceptions

Full size image

Figure 2 shows the clusters from an American random sample survey. Starting in the high-warmth, high-competence quadrant, societal defaults, and reference groups (such as middle class, Christians, and Americans, ingroups for many respondents) elicit pride and admiration. They also receive both active help (protection) and passive help (association, going along).

Fig. 2
figure 2

Example SCM data, illustrating warmth and competence stereotypes, emotional prejudices, and discriminatory tendencies

Full size image

Diagonally opposite the societal ingroups are the most extreme societal outgroups, allegedly low on both warmth and competence. These groups commonly include poor people (of any race), drug addicts, immigrants, and homeless people (who are three standard deviations from the overall mean, hence outside the cluster analysis). These groups reportedly elicit more disgust and contempt than any other quadrant, and they receive both active harm (attack) and passive harm (neglect). The contrast of extreme ingroups and outgroups has been a staple of intergroup relations research for decades.

What the SCM adds is ambivalence. Groups viewed as high warmth but low competence—such as older people and people with disabilities—elicit pity and sympathy. Benign as these emotions might seem, they reinforce the status hierarchy because pity is directed downward. The mixed behavioral tendencies reflect this ambivalence: active help and protection (help them cross the street) but neglect and ignoring (don’t go for coffee afterwards). Many such groups are institutionalized, consistent with taking care of them but also isolating them.

The other type of ambivalence describes stereotypically high-competence but low-warmth groups, such as rich people (all over the world), and in the U.S. currently, Asian and Jewish people. Elsewhere this quadrant likewise includes other outsider entrepreneurs. They elicit envy and jealousy, which communicate respect for their competence but resentment because they are not “one of us.” They receive a volatile mix of passive association (shopping at their stores) and active harm (attack and fight, when the chips are down).

Generality of the SCM

Data from dozens of countries, across lab and field, many subgroups, and now, even computers, corporations, and creatures also array along warmth-x-competence space. Across 36 countries on six continents, in their native languages (Cuddy et al. 2009; Durante et al. 2013), pretest samples nominated their society’s more salient 15–25 social groups, then the main sample rated them along dimensions of interdependence, status, warmth, and competence. The groups always spread out across the space, with interdependence predicting warmth and status predicting competence. Granted, cultures vary not just in the groups they name, but in the superiority of the ingroup (Asians are more modest; Cuddy et al. 2009), and in the relative usage of the ambivalent quadrants (more equal countries use them less; Durante et al. 2013). The SCM usefully describes the social geography in each instance.

SCM dimensions also fit historical descriptions of stereotypes. Italian Fascists depicted several racial and ethnic groups in their magazine “defending the race.” Independent coding of the content spontaneously generated the SCM quadrants, with idealized Italians and Aryans, envied and resented English and Jews, as well as despised Black and mixed-race people (Durante et al. 2010). Closer to home, Princeton students have described 10 national and ethnic groups every 15–20 years since Katz and Braly’s (1933) classic study. Modern judges coded the 80 paradigmatic adjectives by warmth and competence, producing a map of American stereotypes that works over time (Bergsieker et al. 2012).

Besides content analysis reproducing the warmth-x-competence space, multidimensional scaling of groups’ paired similarity ratings also reproduces the space (Kervyn et al. under review). Likewise, comparisons to previous research-generated models suggest both compatibility and distinctiveness. Osgood’s semantic differential dimensions (Osgood et al. 1957) feature evaluation and potency-activity in person perception, operating at a 45° angle to the SCM. Evaluation runs from low–low to high–high warmth and competence (it’s good to be either one); potency-activity runs from a low of warm-but-incompetent groups to a high of cold-but-competent groups, suggesting threat (Kervyn et al. under review). Other social cognition models are even closer to the SCM: trait ratings (Rosenberg et al. 1968), self and other person perception (Peeters 2002; Wojciszke 1994, 2005), political candidate perception (Abelson et al. 1982), and gender roles (Abele 2003). Independent invention also confirms generality (see Fiske et al. 2007 for others). These two dimensions constitute a principle that deserves to be true.

Not only the intergroup (SCM) and the interpersonal (as just cited) levels, but also the subtype level seems to fit these two dimensions. SCM studies have mapped subtypes of women and men (Eckes 2002), gay men (Clausell and Fiske 2005), immigrants (Lee and Fiske 2006), and African Americans (Fiske et al. 2009). That the target possesses intentionality is key, as objects do not easily reproduce the space (Harris and Fiske 2008; Harris et al. 2007), but animal species do (Sevilliano and Fiske under review). Dogs, cats, and horses are warm and competent, like us, and we keep them; dirty/slimy or cold-blooded species (rats, mice, snakes, lizards, fish) are disgusting and we exterminate them; predators (lion, tiger, bear, leopard) are awesome and we respect them; whereas cows, pigs, ducks, rabbits, and hamsters are pitiful and some of us eat them. What’s more, human organizations are seen as having intent and capability (Kervyn et al. 2012); corporations may not be people, but people comprise them, so we make sense of them the way we make sense of people. Companies can be all-American “us” (Johnson and Johnson, Campbell’s, Hershey’s), or they can be disgusting and contemptible (BP, AIG, Marlboro). They can also be enviable (Rolex, BMW, Porsche) or pitiful (USPS, VA, Amtrak). People react accordingly.

Generality over time, cultures, subgroups, organizations, and species suggests that “universal” is not too bold a reach for the SCM. Arguably, “universal” suggests “adaptive.” If inferring others’ intentions and effectiveness—warmth and competence—has survival value, then maybe we would have neural signatures for friends and foes, up and down the hierarchy.

Varieties of Dehumanization in Mind, Brain, and Behavior

Neuro-imaging and EMG data do reveal distinct neural and muscular signatures of responses to distinct kinds of ingroups and outgroups. Our lab team has been working our way around the SCM quadrants, tackling each one to discern its patterns in mind, brain, and behavior. But first, consider one study focused on the status-competence dimension because of its centrality in human social life (Fiske 2011). We will end this section by focusing on the other dimension, interdependence-warmth, a more optimistic view. Between these book-ends of the two dimensions, we visit each quadrant in turn.

The Trolley Study: Valuing People’s Lives by Their Status

Focusing first on the status dimension, in one study (Cikara et al. 2010), we borrowed the trolley dilemma from experimental philosophy: Joe is standing on an overpass spanning a train track along which is speeding an out-of-control trolley, toward five people, who will be killed by it. Joe could push another bystander off the overpass, stopping the trolley and saving the five people, but killing the one person sacrificed. Assuming that he does this (and cannot jump himself), is this morally acceptable? People (80–90 %) overwhelmingly answer “no.” This default—moral revulsion—is established in the relevant literature.

In our variant, we specified who was on the overpass and who was on the trolley track, creating all combinations of the SCM quadrants for the person sacrificed and the group of five saved. We first obtained and pretested 128 photos, 32 for each SCM quadrant (16 instances of each of two prototypic groups); each photo pretested as an instantly recognizable instance of its group (e.g., an American with a flag, a disabled person, a homeless person, a rich person). Students rated the acceptability of all combinations of sacrificing and saving, while having their brains scanned.

The results were stark; the ratings data reversed the default findings. People found it more acceptable to sacrifice one low-status (incompetent) person (69 %), regardless of who was on the tracks, or to save five high-status (competent) people (77 %), regardless of who was sacrificed. And they found it most acceptable (least unacceptable) to sacrifice one low–low (e.g., homeless) person to save five ingroup (e.g., generic American) people (84 %).

This interaction also manifested in a neural pattern characteristic of complex social decision making, implying that people did not make this moral trade-off automatically. In particular, the medial prefrontal cortex (mPFC) is implicated in dispositional attribution (Harris et al. 2005), forming impressions of people versus objects (Mitchell et al. 2005), and meta-analyses of social cognition in neuroimaging studies (Amodio and Frith 2006; van Overwalle 2009). Here, the mPFC activated when people considered the most acceptable (least unacceptable) trade-off, sacrificing one member of the most extreme outgroups (a homeless person) to save five of the ingroup (Americans).

To us, this utilitarian approach to valuing people’s lives seems both normal (insurance companies value people’s lives by their future earnings) and shocking. In the abstract, most people find it morally unacceptable to trade-off a generic person’s life even against five other generic people. But when the person sacrificed and the people saved become specific members of valued and devalued societal groups, participants reluctantly agree to the trade-off. In effect, they agree to lessen the worth of the person sacrificed, a form of dehumanization. But perhaps valuing the people saved solely in terms of their status is a form of super-humanization, some being more worthy than others. The next study supports the dehumanization interpretation.

Dehumanizing the Homeless: Disgust toward the Lowest of the Low

As previously described, the most dehumanized outgroups (low on both warmth and competence) elicit disgust and avoidance, devalued as literally worth-less. In Fig. 1, homeless people fall outside all the other human clusters, and drug addicts are not far behind. We wondered whether people spontaneously consider the minds of homeless and addicted people; failure to consider their minds would suggest treating them like objects, a clear form of dehumanization. Everyday observation suggests the plausibility of this, as pedestrians avoid eye contact and detour around people apparently homeless or drugged, perhaps for fear of contamination or involvement. Whatever the reason, people might not fully consider the other’s mind. Both questionnaire data and neuro-imaging data bear on this issue.

We developed the SCM picture set (Harris and Fiske 2006), later used in the trolley study just noted. Lying in the MRI scanner, students viewed 48 photographs of individuals recognizably associated with stereotypic social groups (4 SCM quadrants × 2 groups each × 6 instances each). For each, they had to indicate whether the photograph made them feel pride, envy, pity, or disgust; the SCM-predicted emotions fit each quadrant as expected. Using the meta-analytic result that social cognition, especially mind perception, reliably activates the mPFC, we compared mPFC activation in the low–low quadrant with all the others. In contrast to the other three quadrants—two of which contain societal outgroups—the mPFC did not activate significantly above baseline to homeless people and drug addicts. The low–low quadrant’s activation effect size was only about two-thirds of the other three, nonsignificant in this study.

Questionnaire data converged with the neuro-imaging data (Harris and Fiske 2011). People indicated on several measures that they had more difficulty attributing a mind and were less likely to interact with the pictured homeless and addicted people from this quadrant. They also rated them on a series of traits as less warm/familiar and less competent/autonomous than the other groups, consistent with their occupying the low–low SCM quadrant.

Together, these data suggest a variety of dehumanization based on disgust, which after all targets objects as well as people (not true for the other SCM emotions, which mainly target people: pride, pity, envy). This disgusted dehumanization resembles Haslam’s type that likens people to animals (Haslam et al. 2013); in particular, vermin also inhabit the low–low quadrant (Sevilliano and Fiske under review).

As discouraging as this dehumanized perception may be, we have found it to be malleable (Harris and Fiske 2007). In response to what we term the “soup kitchen manipulation,” participants had to consider whether each individual would like a certain vegetable pictured immediately beforehand. The vegetable varied, and none were stereotypic of any social group. So simple a change in goal brought the mPFC back on line even for homeless and addicted people. Considering the others’ preferences is the first step to considering their mind and rehumanizing them.

Resenting Investment Bankers: Schadenfreude Toward Envied Outgroups

Now we move to the first ambivalent outgroup cluster, envied groups disliked for seeming cold but respected for competence. People do not want to be them, but they would not mind having what they have (Fiske 2011; Smith and Kim 2007). Hence, envy is volatile; envied outgroups are competent and high status, so we need them and look up to them, going along to get along. But they seem cold, not “us,” so they are resented. This resentment makes us less happy about their good fortune and less concerned about any mishaps that befall them. Malicious glee at someone else’s misfortune is Schadenfreude; people smile when the mighty have fallen.

We aimed to bottle this phenomenon, having noted that people are most reluctant to report envy, out of all the SCM emotions. People do say they envy rich people and business people in suits, but less than they report major emotions in other quadrants. We wanted to go beyond self-reports, so we paired a new set of SCM pictures with pretested everyday events, good (ate a tasty sandwich) and bad (sat on chewing gum). People did typically report feeling good about all groups’ good events, but less so for envied groups; likewise they did typically report feeling bad about all groups’ bad events, but less so for envied groups (Cikara and Fiske 2012).

To get beyond self-reports, we attached electromyographic (EMG) electrodes to measure activity in participants’ smile muscles (zygomaticus major). Indeed, people smiled more to good events than bad ones for all groups except the envied ones. For the rich and the business-suited, people smiled more to their bad event than their good ones. The smiles embodied malicious glee.

This might not seem to matter, except that reported Schadenfreude activates the brain’s reward centers, which correlates with self-reported willingness to harm the envied other. In this case, the envied outgroup is a rival baseball team, and the outcomes are wins and losses, among avid Yankees and Boston Red Sox fans (Cikara et al. 2011a, b). Schadenfreude over an envied rivals’ losses to a third-party team predicts real-world harm, suggesting that envy is consequential.

If they matter, then how malleable are envy and Schadenfreude? During the worst of the 2008 downturn, we primed participants with stories of a laid-off investment banker behaving in one of three unbanker-like (unenvied) ways: volunteering to do local businesses’ bookkeeping (admirable), using the severance package to buy cocaine (disgusting), or commuting with an empty briefcase to Starbucks (pathetic). Primed participants did mitigate their Schadenfreude toward other suits who resembled investment bankers—but not toward all envied outgroups (other rich people), which we interpreted as a form of bounded empathy (Cikara and Fiske 2011a, b).

But is envious prejudice a form of dehumanization? Yes, it resembles Haslam and colleagues’ (2013) automaton version of dehumanization, applied to businesspeople. Investment bankers, female professionals, Asians, and other efficient social groups sometimes appear stereotypically as cold and machinelike. This is the envious form of dehumanization for those viewed as competent but cold.

Pitying the Weak: Prescriptions for the Warm but Incompetent

Reactions to pitied outgroups include disrespect for seeming incompetent, but apparent liking enough for seeming trustworthy and warm. The presumed incompetence carries with it an obligation for the pitied to be nice or forfeit the observer’s pity. Across an array of groups—people with disabilities, older people, and stereotypical women—other people express such pity and sympathy. But that pity is not entirely benign, as it depends on the pitied person remaining low status and incompetent, not high status, autonomous, and agentic. For example, older people who join the grey-panther activists forfeit pity and become obnoxious (Cuddy et al. 2005). Disability-rights activists might expect also to be seen as uncooperative and difficult, agentic but no longer nice. Subservient women who become feminists go from being warm-but-dumb to being smart-but-cold (Glick and Fiske 1996). Pity requires signing the contract to follow the higher status others’ prescriptions to cooperate, defer, and obey, to avoid becoming uppity. What this array of groups share in common is apparent interdependence with outsiders (able-bodied people, younger people, and stereotypical men) charged to care-take them as helpless. The deal they cut is the pitied outgroups acknowledging the status of society’s helpers and cooperating with them, not making trouble. Buying other people’s prescriptions denies a person’s agency and status, arguably one form of dehumanization. Consider three groups as case studies.

Able-bodied People’s Prescriptions for People with Disabilities

People with disabilities fall in the pity cluster because they allegedly are high warmth (“on our side”) but low competence (“below us”). However, they merit pity if and only if they adhere to prescriptions for their stereotypic role. Salient among these prescriptions is that their misfortune not be their fault. Disabled people are penalized for fault, either if they caused the disability or if they neglect its prescribed treatment (Wu, Ames, Swencionis, and Fiske, in preparation). In 32 vignettes (e.g., fell off roof), combined with 32 pictures, disabled people were either at fault for causing it (e.g., recklessly ignored warnings), or not, and either at fault for sustaining it (neglect treatment), or not. Either kind of fault penalized people’s ratings of the disabled person. In this sense, following others’ advice is the prescription to be eligible for pity and a caring response in a lop-sided interdependence with able-bodied people. But the prescription also minimizes one’s humanity because it removes one’s agency to make autonomous choices.

Younger People’s Prescriptions for Older People

More people are interdependent across generations than across disability status. Across the generations with the family, at work, and in society, each side has obligations, but here we focus on the prescriptions for older people (who come up as a stereotyped group more often than do young people). Stereotypes of older people as doddering but dear presuppose that they cooperate with younger people to minimize intergenerational tensions (North and Fiske 2013). Three primary domains of prescribed elder cooperation include appropriate succession (passing along resources such as jobs and wealth), sharing consumption (not using up social security or blocking the road), and identity boundaries (acting one’s own age). Though doubtless there are others, these domains appear as independent factors in ageist attitude scales (North and Fiske in press-a). What’s more, following or violating these prescriptions affects ratings of only older targets but not middle-aged or younger targets, and it is younger raters who react the most strongly (North and Fiske in press-b), consistent with the idea of intergenerational prescriptions. Ageist young people dehumanize old people by denying their agency: viewing them as having no right to their own resources, their share of joint resources, and invading youth identities. Power is all about controlling resources, and prescriptive stereotypes are one way to do this (Fiske 1993).

Stereotypic Men’s Prescriptions for Stereotypic Women

Men and women share a more intimate interdependence than any other intergroup pair, and this intense interdependence especially inspires prescriptions (Fiske and Stevens 1993). Because men have higher societal status, they can demand adherence in return for providing resources (Glick and Fiske 1996). In particular, “benevolent” sexism (paternalistic sympathy, even pity) is reserved for women who adhere to stereotypical gender roles that benefit stereotypical men. Forfeit that protected position, and women become targets of hostile sexism, which sees gender relations as competitive and certain women (e.g., feminists) as trying to control men.

Women who adhere to stereotypic gender roles give up personal agency and status, so this also is one form of dehumanization. We demonstrated this phenomenon by finding photographs of women, fully clothed and bikini-clad, standardized for posture, facial attractiveness, and facial expression (Cikara et al. 2011). People had to pair them with first-person verbs—(I) use, push, squeeze—or third-person verbs—(she/he) uses, pushes, squeezes. Men were faster to pair first-person verbs with bikini-clad women (implying the men’s own agency) and third-person verbs with fully clothed women (implying her agency), compared with the reverse; women showed no such difference. This fits the idea that at least some men tend to deny agency to sexualized women.

We then added bikini-clad and fully clothed men to the mix, showing just the male/female, clothed/bikinied photographs to male participants while we scanned their brains and then tested their memory separately for the faces and bodies of people they had just seen. These men remembered both female bodies and bikini-clad bodies the best, but could not differentiate among the faces shown separately. Remembering a sexualized woman’s body better than her face carries an element of dehumanization.

What’s more, the men’s social-cognition-sensitive mPFC activated less to the bikini-clad women, the higher the men scored on hostile sexism, consistent with the idea that sexual hostility correlates with dehumanizing (or at least dementalizing) sexualized women. The opposite pattern occurred for bikini-clad men: higher hostility toward women correlated with more mPFC activation (mentalizing?) of sexualized men. (Perhaps body social comparison—how did he get those abs—is more at play for hostile sexists.) In any case, a pattern of responses suggest that some men on some measures over-emphasize sexualized women’s bodies, forget their faces, deny their agency, and ignore their minds. These responses fit a form of subordinated dehumanization, perhaps prescriptive: they should have only bodies and no faces, agency, or minds.

Interim Summary

The apparently universal social cognition dimensions of warmth and competence array social groups into recognizable clusters. One dimension, status-competence, reliably predicts people’s valuations of others’ lives in a hypothetical social dilemma that engages the medial prefrontal cortex, part of the social cognition network. More specific experiments compare in turn each quadrant to the other three, describing varieties of dehumanization: dehumanizing the homeless or drug addicted, via disgust toward the lowest of the low; resenting investment bankers, businesspeople, and rich people, via Schadenfreude toward envied outgroups; pitying the weak, via prescriptions for the warm but incompetent disabled, older, or sexualized female persons. We now turn to the fourth quadrant, society’s ingroup.

Individuating the Ingroup

Around the world, indigenous people call themselves the humans, as opposed to those nonhuman Others. Understanding this in light of social adaptation, Donald Campbell’s (1965) Nebraska Symposium chapter noted (pp. 310–311):

an impressive consensus that outgroup threat to an ingroup increases ingroup solidarity…Consideration of the great competitive advantage of social life over solitary life leads to the expectation that biological and sociocultural evolution would have produced in both man [sic] and termite motivational dispositions furthering group life and reflecting its advantages, as in economy of cognition, division of labor, and mutual defense. In some respects at least, man [sic] should be regarded as basically a social animal, with individual dispositions reflecting this fact. Because the wisdom of evolution is retrospective, social motivations such as are found in ethnocentrism may be judged dysfunctional [sic] in a changed environment.

This remarkable analysis has aged well (apart from the generic masculine), noting ingroup social motivations for derogating outgroups and the advantages of humanizing ingroup members. This analysis goes beyond ethnocentrism, in our view, to include sexism, ageism, classism, and other divisions that serve inter-category interests. So, what about ingroup solidarity?

We are beginning to explore what might be termed hyper-humanization of ingroup members, who are both liked and respected. Focusing on the ingroup and its reference groups, Warmth and Competence R Us, the rest of the chapter focuses on concentrated humanization of the high–high cluster, examining the what, who, why, when, and how of humanizing the ingroup (for an extended analysis, see Swencionis and Fiske 2013).

What is Humanization? Beyond Not-Categorizing

Dehumanization has attracted much research and analysis, as this volume attests. Humanization, less so. The closest research investigates individuation as an alternative to categorization. All concur that individuation requires more effort than category-based processes, and the same holds for humanization versus dehumanization. Let us examine this premise.

As one example, the Continuum Model (Fiske and Neuberg 1990) holds that people’s social-cognition default is automatic categorization, according to easily discerned categories such as gender, age, race, and class. Categories trigger associations to stereotypes, prejudices, and discrimination. But people are no fools: If the category is a poor fit, and if motivated and able, people will consider subtypes and even fully individuating processes. Attention to and interpretation of a person’s attributes allow an individual impression, accurate or not, but at least an individual human. To anticipate, evidence supports these processes (Fiske et al. 1999; for related models, see Brewer 1988; Brewer and Feinstein 1999; Kunda and Thagard 1996). Individuating processes include attention (especially to inconsistent cues), dispositional inference, and attribute-based evaluations (see Fiske et al. 1999, for a review).

Who Humanizes (Individuates)?

Some people are spontaneously inclined to take the trouble to humanize others. Having individuation as part of one’s self-concept predicts individuating processes (Fiske and von Hendy 1992). Being high on Person-, not Thing-Orientation (Graziano et al. 2011) marks people who are intrinsically interested in people, and women on average tend to score higher. People can orient to others for a variety of reasons all measured by the Interpersonal Orientation Scale, which emphasizes affiliation motives (Hill 1987). Of these motives, three seem self-centered—orienting to others for social comparison, emotional support, and approval—so unlikely to motivate meaningful humanization of the other. But the remaining affiliative motive, seeking positive stimulation, seems most likely to yield individuating processes: the motive is measured with items such as, “Just being around others and finding out about them is one of the most interesting things I can think of doing.” And “One of the most enjoyable things I can think of that I like to do is just watching people and seeing what they are like.”

Why Bother?

For someone who is not necessarily a people-person, why undertake the effort to individuate and humanize another? According to another classic Nebraska Symposium chapter, people analyze individuals’ predispositions—that is, make attributions to individual personalities—in order to understand those others, in the service of prediction and control: “The [attribution] theory describes processes that operate as if the individual were motivated to attain a cognitive mastery of the causal structure of his [or her] environment” (Kelley 1967, p. 193). When interpersonal information falls below acceptable levels, people search for information.

Not just sheer amount of information, but its quality, matters. In Gestalt approaches described in yet another historic Nebraska symposium chapter (Heider 1960), people are motivated to achieve a “good figure” of perception, in the visual field, but also in their social impressions, for example, feeling balanced when agreeing with friends.

Also influenced by Gestalt approaches, Asch (1946) posited that people want to form coherent individual impressions, again because of seeking good form. He did not elaborate on the motives but asserted that people aim to find coherence.

While these might seem dry motivations, Fiske (2002) posits core social motives of both understanding/prediction and control, as two of the more social cognitive motives (self-enhancement and trusting others being more social affective ones, not as relevant here). But like all core motives, understanding and controlling operate in the service of an overarching belonging motive to fit in with one’s ingroup. Good reason to bother humanizing/individuating one’s ingroup members.

How Do People Humanize?

The processes of individuation—here taken as a necessary feature of seeing someone as fully human—mainly fit individual impression formation. In the Gestalt tradition, people interpret the meaning of trait terms to fit together into a coherent whole (Asch and Zukier 1984). For example, “cheerful” and “gloomy” become coherent under “moody.” In Kelley’s (1967) theory, people seek behavioral consistency, idiosyncratic behavior not widely shared (not linked to consensus), and a general response (not just one distinct to a particular target); together, these support an overall dispositional inference. People engage these processes to infer the dispositions underlying other people’s intent (Jones and Davis 1965). In Heider’s (1960) terms, all these processes reflect the goal of searching for invariants (consistency) across cues about the other person.

Besides reconciling inconsistency to find coherence, impression formers show some other predictable patterns. When people go beyond categories, they attend to individual attributes such as traits (Fiske et al. 1987; Neuberg and Fiske 1987; Pavelchak 1989), integrating them into broader dispositional inferences (Erber and Fiske 1984; Ruscher and Fiske 1990). Impression formers mentalize the other person, that is, they infer mind and intent (Ames et al. 2011). These are the mind-perception processes that implicate the brain’s mPFC (Ames and Fiske under review; Mitchell et al. 2005). Similar neural patterns occur in rehumanizing as in impression formation (Harris and Fiske 2007).

Although prediction and control motives imply accuracy, individuating (humanizing) processes are not necessarily accurate (Goodwin et al. 2002; Stevens and Fiske 2000). But the illusion of accuracy is what matters (Neuberg and Fiske 1987). When reality intervenes, such social prediction errors register neurally (Harris and Fiske 2010).

Some predictable errors and biases haunt individual impression formation. The most reliable is that impression formers accentuate the positive in impressions of individuals (Matlin and Stang 1978; Sears 1983).

As a result of positivity’s prevalence, negative information stands out and is diagnostic, being rare. Hence, impression-formers especially weight negative information (Fiske 1980; Ito et al. 1998; Skowronski and Carlston 1989; Taylor 1991). For example, people attend to and weight negative information when they encounter it (Fiske 1980). They then communicate the positive and omit the negative in conveying their impressions (Bergsieker et al. 2012). But receivers are not deceived; omission implies the negativity, and people infer it accordingly (Kervyn et al. 2012a).

The implication of these basic impression formation processes (and there are many others; Fiske and Taylor 2013) is that seeing another person as fully human focuses on inferring a coherent predisposition in the form of perceived intentions, and often presumably positive ones, at least within the ingroup.

When Do People Humanize?

As implied by the idea that most people spontaneously humanize the ingroup, people individuate others on whom their outcomes depend, as social beings. Just as the opening bookend focused on one SCM dimension, status-competence, on the trolley track, the closing bookend on this exploration of SCM’s (de)humanizing quadrants focuses on cooperation-warmth, the primary dimension. When people land on the same team, sharing goals, they attend to each other because their outcomes depend on each other. The individuating processes we uncovered fit well with half a century of intergroup contact research (Pettigrew and Tropp 2006), showing that shared goals are a crucial feature of individuating another person. What our lab contributes is how that process operates: Teammates attend specifically to stereotype-inconsistent information and make individuating dispositional inferences about it. Under outcome dependency, people engage a host of individuating processes (Erber and Fiske 1984; Dépret and Fiske 1999; Stevens and Fiske 2000).

Outcome dependency and individuation also registers neurally, as we discovered when we brought students into the lab to meet two partners, from a rival university. Both were expert in the job at hand, creating educational games for children, using colorful plastic wind-up toys. One partner expected to contribute positively to the joint task with the participant; the other partner expressed serious doubts, as this activity was outside her expertise. Crossing the expectancy manipulation was an outcome-dependency manipulation; the participant expected to meet both of them again, but with one to cooperate on the joint task for a nontrivial prize, and with the other to work independently (and noncompetitively) for an equivalent prize. Each partner brought her teaching evaluations, which were mixed, equal parts positive and negative, therefore consistent and inconsistent with the established expectancy. Participants viewed these evaluations while having their brains scanned. Consistent with past findings that interdependence focuses attention and dispositional inferences on expectancy-inconsistent information, we found selective mPFC activation to unexpected information about another on whom one depends (Ames and Fiske under review).

Why Worry about Humanization?

If we all deplore dehumanization—which we surely do, why else study it?—we ought to be acquainted with the alternative, humanization, warts, and all. Humanizing other people is no panacea; individuated impression formation is not necessarily accurate, fair, or complete. But it seems better than the alternative.

Finale

This chapter has aimed to elucidate some systematic patterns of humanization and dehumanization, drawing on two apparently fundamental dimensions of social cognition and group perception, warmth and competence. Each variety has unique features: predictable stereotypes, emotional prejudices, behavioral tendencies, and neural activations. And in exploring each variety, simple interventions can move observers from dehumanizing to humanizing processes, encouraging people to follow their better natures.