Corps de l’article

1. INTRODUCTION

It is a commonplace that humans’ overall relationship with animals is highly “ambivalent”[1]. Aside from the fact that there is disagreement among humans about how animals ought to be treated, there seems to be inconsistency at the intrapersonal level. Across different contexts, individuals practice, tolerate, and support ways of treating animals that do not seem to reflect a consistent set of attitudes. It seems obvious, for instance, that there is a stark contrast between the treatment of some companion animals and the treatment of animals in the food industry or in experiments. When this kind of inconsistency is made salient in public debate, those who hold the widespread view that animals are entitled to some moral consideration, but less than humans, will typically say that, on reflection, both ends of the apparent spectrum are excesses: while there exists excessive cruelty in some areas, there is also excessive regard given to certain individual animals. However, common wisdom might have “ambivalence” all wrong. Maybe it is not the case that humans tend to have some very good and some very bad relationships with animals and that this is overall rather paradoxical. In fact, those who argue for the equal moral consideration of animals will not need much convincing to agree that even relationships that look benign and involve positive subjective feelings on the part of humans are shot through with the same kinds of views and tendencies that underlie obvious human disregard towards animals. Attitudes on the positive end of the supposed spectrum may not be so good, after all. Ambivalence may affect our dealings with animals on a different level, and perhaps the relevant ambivalence exhibited by humans does not result in animals having some very good and some very bad experiences with humans. For animals, the outcomes of their interactions with humans may typically be downright bad, although the severity of the harm inflicted on them may vary.

This paper explores the idea of ambivalent speciesism—speciesism that expresses itself both in hostile and benevolent attitudes and behaviours, but that remains, overall, harmful to, disrespectful or inconsiderate towards members of certain species. It thus explores a view of speciesism that is barely beginning to be addressed in empirical research (Altınal and Tekdemir, 2020; Caviola and Capraro, 2020), but which is made highly plausible by insights into the ambivalence found in, for instance, sexism (Glick and Fiske, 1996) and by everyday experience. It has long been acknowledged that phenomena such as racism and sexism are marked by ambivalence. Hostile stereotypes tend to go together with benevolent ascriptions of positive characteristics and with benign behaviours. For instance, women might simultaneously be seen as “incompetent at work” (Glick and Fiske, 1996, p. 494) and endowed with a “superior moral sensibility” (Glick and Fiske, 1996, p. 500), and positive perceptions might prompt, for instance, “intimacy-seeking” behaviors (Glick and Fiske, 1996, p. 491). While hostile speciesism—from animal experimentation to the use of animals as food—dominates our conceptualization of the problem, a moment’s reflection reveals that we should expect speciesism to exhibit patterns indicative of ambivalence as well. Nonhuman animals are often seen as incapable of reasoning, yet endowed with a sensitivity to changes in their natural environment that is not only superhuman, but in some descriptions has the appearance of supernaturalness as well (they mysteriously “sense” things—for instance, they always know about natural disasters first). They are seen as amoral, yet loyal; dumb, but cute; unfit for social coexistence with humans, but perfectly adapted to and capable of flourishing in their respective “habitats,” and so forth. This paper argues that positive views of animals form an important subset of genuinely speciesist views. It also outlines some apparent patterns of ambivalence in speciesism. In the course of this demonstration, two questions arise that this paper can only touch upon, but not attempt to resolve. One concerns the precise nature of the interconnection between subjectively positive and negative speciesist views and behaviours. Do we recognize positive views, attitudes, or behaviours as elements of speciesism because they are caused by or because they are supportive of negative views, or both? Or is the connection a different one entirely? This complex question is not unique to the phenomenon of speciesism, and while it must be noted, it cannot be answered within the limits of the present investigation. The second question is whether thinking of speciesism as potentially ambivalent might have conceptual repercussions. Speciesism can be conceptualized as a type of discrimination—the unjustified lesser consideration or worse treatment of beings that are classified as members of certain species (Horta and Albersmeier, 2020, p. 4). The existence of benevolent speciesism is one reason to reconsider whether (and, if so, how) a definition in terms of discrimination (with its focus on lesser consideration or worse treatment) is able to capture the phenomenon that is speciesism in all its forms. The present paper will raise this conceptual question, but it cannot offer a definite answer. Primarily, it undertakes the prior task of demonstrating the need to account at all for what appear to be benevolent forms of speciesism. These often-overlooked forms of speciesism contribute in their own way to the subordination and exploitation of nonhuman animals, and we can expect them to multiply and intensify as efforts for the recognition of animals’ moral entitlements become more successful. We need to address them conceptually, theoretically, and practically in working against speciesism.

Section 2 considers findings of research into the ambivalence in racism and sexism and defends the idea that investigations into speciesism should take cues from this research. Section 3 anticipates the potential conceptual repercussions of recognizing speciesism’s probable ambivalence. In section 4, I explore the phenomenon of benevolent speciesism which together with its hostile forms, makes speciesism as a whole ambivalent. Patterns indicative of ambivalence and their overall negative impact on nonhuman animals are outlined. The conclusion (section 5) demonstrates that the notion of ambivalent speciesism creates a much more unified picture of humans’ interactions with animals and indicates how this picture can be filled in by future research.

2. AMBIVALENT PREJUDICE

If speciesism is anything like racism, sexism, and other such phenomena, it will exhibit a certain level of ambivalence. Before looking more closely at what ambivalence is in these other kinds of contexts, a concern about the antecedent of the previous sentence must be addressed. After all, it has been doubted that speciesism has meaningful similarities to racism or sexism at all. The term “speciesism” has been deliberately modelled after the terms “racism” and “sexism” (Ryder, 1975; Singer, 2009 [1975]). The analogy itself has attracted some criticism, the harshest version being the complaint that the coining of this term is a “deliberately devised verbal parallelism that is utterly specious” (Cohen, 1986, p. 867).

However, the proponents of the term have presented arguments for why the mentioned phenomena should be recognized as being analogous. They have shown that these phenomena are connected because all are types of discrimination. For speciesism to be recognized as one such type, it had to be shown that members of other species are morally considerable and that, therefore, there exists meaningful potential for failing to consider them adequately—that is, there exists the real possibility that they are being discriminated against[2]. In the wake of the contributions that first introduced the notion of speciesism, the entire field of animal ethics has constantly dealt with the question about the grounds and extent of the moral considerability of nonhuman animals. There are now well-worked-out arguments in favour of animals’ moral considerability from a variety of schools of thought within normative ethics. Rights-based views (Cavalieri, 2001; Regan, 2004), contractarian (Rowlands, 2009), Kantian (Franklin, 2005; Korsgaard, 2018), egalitarian (Horta, 2016), and virtue-ethical views (Hursthouse, 2006)—as well as different approaches to including animals and their entitlements in political thinking (Donaldson and Kymlicka, 2011; Cochrane, 2018)—all argue that sentient beings other than humans can be morally harmed, wronged, or treated unjustly. Also, they all make the case that, in reality, this potential is frequently actualized. In addition, it has been argued that speciesism, just like racism and sexism, is simply another manifestation of tribalism—that is, it is a product of one and the same evolved tendency to draw unjustified moral distinctions between perceived in-groups and out-groups (Jaquet, 2022). All these different views imply that if we care about social justice, we must recognize and worry about speciesism as a major form of discrimination. Consequently, one can recommend at least taking cues from research into other types of discrimination in order to make progress in understanding and tackling the phenomenon that is speciesism.

To be sure, racism, sexism, and other types of discrimination among humans are different from one another. Speciesism is bound to be different from all of them in its own ways. However, there are recurring themes and patterns that it would be negligent to ignore when we are trying to understand speciesism, as well as these other phenomena. Research on ambivalent sexism, for instance, itself took its cues partly from research into racism (Dovidio and Gaertner, 1986), which had found that “racism had become (a) more subtle and (b) more ambivalent” “in response to changing social norms” (Glick and Fiske, 2011, p. 531). The fact that racism and sexism are marked by ambivalence is important to take into consideration when we are trying to understand the phenomenon of speciesism. Racism and sexism are by now understood to encompass not only negative, but also subjectively positive beliefs about and attitudes towards the stereotyped group (cf. Dixon et al., 2012). Sexism, for example, has been described as including “hostile” and “benevolent” attitudes (Glick and Fiske, 1996), where benevolent sexism is

a set of interrelated attitudes toward women that are sexist in terms of viewing women stereotypically and in restricted roles but that are subjectively positive in feeling tone (for the perceiver) and also tend to elicit behaviors typically categorized as prosocial (e.g., helping) or intimacy-seeking (e.g., self-disclosure).

Glick and Fiske, 1996, p. 491

The resulting overall picture of women shows them as “nice but incompetent.” Benevolent sexism is thought to be a compensatory or legitimizing counterpart to hostile sexism (Glick and Fiske 1996, pp. 491-492). Both appear to be complementary and have been shown to predict gender inequality across cultures (Glick and Fiske, 2011). Likewise, positive stereotypes appear to be crucial elements of racism. Commendatory stereotypes such as the one linking African Americans to athletic excellence seem to enhance thinking of group differences as being grounded in a fundamental natural distinctness. They also facilitate the application of negative stereotypes (such as the one linking African Americans to crime) (Kay et al., 2013).

In his introduction to the topic of racism, Ali Rattansi explains:

Stereotypes, like other views, reveal contradiction and ambivalence rather than completely invariable contempt or hostility or admiration towards other groups. The attributes of other groups tend to be split between “good” and “bad” ones. Attitudes towards Asians in Europe and the US, for instance, reveal admiration for supposed community unity, thrift, ambition, hard work, respect for education, and “family values,” but also hostility for insularity, suspicion regarding their loyalties to the Western nation-states in which they have come to live, and a sense of superiority towards their more “backward” cultures, especially in relation to religion, the status of women, and so forth.

Rattansi, 2007, p. 128, emphasis in original

A general pattern underlying ambivalent prejudice is theorized to centre around two traits that are mutually exclusive in ambivalent prejudice: warmth and competence. These are conceptualized as two fundamental categories of social cognition, representing social capacities or benevolence (warmth) and intellectual or technical capacities (competence), respectively (Fiske et al., 1999):

The warmth dimension captures traits that are related to perceived intent, including friendliness, helpfulness, sincerity, trustworthiness and morality, whereas the competence dimension reflects traits that are related to perceived ability, including intelligence, skill, creativity and efficacy.

Fiske, Cuddy, and Glick, 2007, p. 77

In ambivalent prejudice, the two dimensions are pitted against each other: persons who are seen (bi-dimensionally) ambivalently are perceived as either warm or competent (Fiske, Cuddy and Glick, 2007; Fiske, 2012). The notion of ambivalence here refers to the fact that a group that is an object of prejudice is not viewed in a purely negative way, but that, on the contrary, negative views are typically accompanied by positive views about the same group, where these positive views also deserve to be counted as instances of prejudice. While “ambivalence” in everyday language might be used primarily to refer to contradictory attitudes towards the same object, “ambivalent prejudice” refers to a set of views that are compatible but have different valences.

It is believed that benevolent stereotypes have a compensatory function. As Glick and Fiske argue, a prevalent tendency to agree with benevolent sexist statements like “women have a superior moral sensibility” or “women have a quality of purity few men possess” (Glick and Fiske, 1996, p. 500) “helps to pacify women’s resistance to social gender inequality” (Glick and Fiske, 2001, p. 109). At the same time, benevolent prejudice is required as a balance to hostile prejudice by the prejudiced person who still somehow values the stereotyped group. When uniform outright disparagement is not an option, hostile prejudice has to be complemented and counterbalanced by benevolent attitudes and views, allowing the prejudiced subject to uphold a comfortable self-image (Glick and Fiske, 1996, p. 492; 2001). This is coherent with the finding that positive stereotypes emerge when the social status of the stereotyped group is alleviated (Czopp, Kay, and Cheryan, 2015, p. 452)—an important insight to keep in mind when we approach ambivalence in speciesism. Given animals’ overall lesser social standing, we should expect the benevolent forms of speciesism to be less pronounced compared to hostile forms. But we should also expect benevolent speciesist tropes to multiply and be disseminated as the movement for the recognition of nonhuman animals’ entitlement to moral consideration makes progress.

Awareness of ambivalence in speciesism is important because benevolent prejudice is, for several reasons, seen as problematic by those who study it (Glick and Fiske, 1996), not only because of its compensatory function, which ultimately amounts to a facilitating function for subordination. While benevolent prejudice can come with some preferential treatment of the targeted group, it is viewed as overall harmful. Negative effects can be very specific. For instance, it has been found that benevolent sexism increases victim blaming in cases of acquaintance rape (Abrams et al., 2003; Yamawaki, 2007), possibly because it comes with beliefs about what constitutes good conduct for a woman, and supposed transgressions lead to the perception that the woman is “responsible for anything unfortunate that may happen to her” (Abrams et al., p. 2003, p. 121). Also, in individuals who are high in benevolent sexism, other effects that lead to increased victim blaming are enhanced (Halicki, Hauser, and Wänke, 2023). Not paying attention to benevolent speciesist prejudice can lead us to overlook similar mechanisms that come into play when benevolent speciesist stereotypes are “disappointed,” as when, for instance, supposedly “loyal” dogs are punished for defensive behaviour that is construed as aggression. It has also been found that exposure to a positive stereotype can be a negative interpersonal experience in itself, due to the de-individualizing nature of stereotypes (Siy and Cheryan, 2013). It can make one expect that one is also the target of negative stereotypes (Siy and Cheryan, 2016), and internalized positive stereotypes can cause distress and inhibit help-seeking (Gupta, Szymanski and Leong, 2011). Reminders of positive stereotypes can impair performance, both in positively stereotyped (Cheryan and Bodenhausen, 2000) and in negatively stereotyped domains (Kahalon, Shnabel and Becker, 2018).

The pattern of local benefits and global disadvantages might differ for animals—they could incur more or less harm from benevolent speciesism than humans experience as a result of intrahuman prejudice. Animals lack some opportunities for benefiting from potential positive effects of benevolent prejudice. For instance, while interpersonal and intergroup effects of benevolent prejudice in the human case are overall negative (Czopp, Kay, and Cheryan, 2015, pp. 455-459), among the mixed intrapersonal effects is the chance that exposure to benevolent stereotypes leads to performance increase (Czopp, Kay, and Cheryan, 2015). It seems unlikely that there could be an analogue in the case of speciesism. For animals, practical treatment is more important than (subtle) verbal treatment. Conversely, this means that they will also be immune to the harms of some types of (verbal) displays of prejudice. However, confrontation with prejudice can very much happen in practical ways: de-individualizing treatment can be practical rather than merely verbal, and at least then it can impact animals in negative ways—for example, when treatment based on speciesist stereotypes does not meet individuals’ specific needs. Animals are not only members of their species and, while species membership is a useful heuristic of what will be important for individual animals, it also is insufficient to consider only species-typical rather than individual behaviours, capacities, interests, and needs. I will explore some of the potential patterns of harm caused by benevolent speciesist prejudice in section 4.

3. ACCOMMODATING AMBIVALENCE—CONCEPTUAL CHALLENGES

In those areas where it has been studied so far, the ambivalence of prejudice has already posed terminological challenges. For instance, in a paper that investigates ambivalence in prejudice against gay men, the term “heterosexism” is rejected as a replacement term for “homophobia”—which is avoided because it mischaracterizes the problem as a phobia rather than prejudice—on the grounds that “heterosexism” has been given definitions that equate it with hostility, which would make it a problematic term when benevolent prejudice against homosexual men is under consideration (Brooks et al., 2020, p. 4). Beginning to see speciesism as ambivalent might have further-reaching conceptual repercussions. The task of accounting for ambivalence raises questions concerning the adequate characterization of the link between discrimination and prejudice.

In the previous section, I spoke of ambivalent “prejudice” and “stereotypes”—terms that dominate the discourse in considerable parts of the empirical research into racism, sexism, and the like. However, the definition of speciesism given above does not mention either prejudice or stereotypes; it instead characterizes speciesism as unjustified lesser consideration or worse treatment—that is, discrimination. In expressing its concept exclusively in terms of discrimination, it thus differs from speciesism’s original definition. When coining the term, Richard Ryder said:

I use the word “speciesism” to describe the widespread discrimination that is practised by man against the other species, and to draw a parallel between it and racism. Speciesism and racism are both forms of prejudice that are based upon appearances.

Ryder, 1975, p. 16

Unfortunately, though, Ryder did not indicate what precisely the relation between prejudice and discrimination was. If anything, the definition suggests that the two are one and the same. The prospect of benevolent speciesism sheds light on the need to further clarify the notion of discrimination (unjustified lesser consideration or worse treatment) and its relation to prejudice and stereotypes. It is at least not obvious that the definition presented above (§ 1) can account for forms of speciesism that we could classify as benevolent.

Prejudice, most generally, is a preconceived view. Discrimination is sometimes described as being based on prejudice:

Prejudice refers to a preconceived judgment, opinion or attitude directed toward certain people based on their membership in a particular group. It is a set of attitudes, which supports, causes, or justifies discrimination. Prejudice is a tendency to over categorize.

Rouse, Booker and Stermer, 2011, p. 1144

In legal philosophy, the similar view that unequal treatment is discrimination when it is based on prejudice can be found in the works of Ely (1980) or Dworkin (1985; cf. Altman, 2020, § 4.1).

In other cases, discrimination is conceptualized as a behavioural component of prejudice. The APA Dictionary of Psychology, which defines prejudice in the most general sense as “any preconceived attitude or view, whether favorable or unfavorable,” characterizes it more specifically in the following way:

a negative attitude toward another person or group formed in advance of any experience with that person or group. Prejudices include an affective component (emotions that range from mild nervousness to hatred), a cognitive component (assumptions and beliefs about groups, including stereotypes), and a behavioral component (negative behaviors, including discrimination and violence).

American Psychological Association, 2023

Likewise, in the SAGE Handbook of Prejudice, Stereotyping and Discrimination, Dovidio et al. distinguish between prejudice, stereotypes, and discrimination as “three forms of social bias,” where prejudice is an “attitude reflecting an overall evaluation of a group,” stereotypes are “associations” and “attributions of specific characteristics to a group,” and discrimination is “biased behavior toward, and treatment of, a group or its members” (Dovidio et al., 2010, p. 5). Prejudice is then said to include beliefs (and, thus, stereotypes) as its “cognitive component” and a “behavioral predisposition to behave negatively toward the target group” (hence, to discriminate against members of the group) as its “conative component” (Dovidio et al., 2010, p. 5).

In contrast, if prejudice is a preconceived view, then this seems to suggest that it should rather be seen as the result of discrimination—insofar as it is based on an inadequate consideration of the evidence pertaining to beliefs about the targeted group, that is, on cognitive discrimination (Lippert-Rasmussen, 2014, pp. 41-42). If this were the case, the definition of speciesism expressed in terms of discrimination would already account for prejudice—because it mentions the grounds of prejudice.

The problem of adequately characterizing the link between discrimination and prejudice exists even when ambivalence is not considered—and it exists not only with respect to the phenomenon of speciesism, but also with respect to racism, sexism, and other such items whose definitions remain contested matters. As we are interested in speciesism here, the question of the precise link between prejudice and discrimination presents itself anew once ambivalence in speciesism is considered. This consideration poses the challenge of finding a place for instances of speciesism that would count as “benevolent.” It does seem that it is easier to accept that a prejudice can have positive valence (because it can have commendatory content) than it is to make sense of benevolent/preferential consideration or treatment as a manifestation of discrimination.

If discrimination were a way of treating or considering others that is based on prejudice, we could say that some individuals can be treated worse or be given lesser consideration based on a benevolent prejudice (as when individuals are seen as more independent and more capable than there is reason to believe they are, and are treated negligently as a result). Conversely, if prejudices were based on discrimination in the sense that they were based on poorer consideration of evidence relating to views about a group, we could say that it is possible to form a benevolent prejudice based on such cognitive discrimination (as when an inflated view of some individuals’ capability and competence is being formed due to a lack of consideration).

In contrast, it seems more difficult to accommodate the idea that discrimination could assume benevolent forms: the discrimination equivalent of “benevolent prejudice” would seem to have to be “preferential” rather than “lesser” consideration or “better” rather than “worse” treatment. Assigning lifeboat seats to women first based on the view that they are more worthy of protection then men would be a behavioural equivalent to a benevolent prejudice. Yet it would not be captured by the standard notion of discrimination, because what we would have to express if we wanted to say that this was a benevolent manifestation of sexist discrimination would be that this was discrimination against women (not against men and not simply between the sexes). This is because the benevolent or benign behaviour is directed at women. To express this, we would have to say that discrimination could consist in preferential consideration or treatment. We would then have overthrown the very concept of discrimination, which includes as a necessary condition the imposition of “some kind of disadvantage, harm, or wrong” to the ones who are being discriminated against (Altman, 2020, § 1.1). According to the standard view, to treat someone preferentially is discrimination not against but “in favor of” the preferentially treated individual (Thomsen, 2018, p. 22). Such discrimination is “discrimination against” only with respect to those who are, as a result, treated disadvantageously. This characteristic of discrimination would suggest that we can make sense of positive manifestations of prejudiced beliefs or attitudes, but not in the same way of their behavioural consequences, components, or presuppositions.

In order to be able to rely on a definition of speciesism in terms of discrimination and still reconstruct positive behaviours towards some individuals as speciesism that is directed against those individuals, one could construe the relation between discrimination and speciesism in the opposite direction. One would have to say that such positive behaviours are speciesist when they are based on prejudice, where prejudice is the result of the worse consideration of evidence regarding a group or individual, that is, cognitive discrimination. This approach initially seems reasonable, precisely because prejudices are standardly described as involving inadequate consideration. In his seminal work on prejudice, Gordon W. Allport characterizes it as a way of “thinking ill of others without warrant” (Allport, 1954, p. 6), “based on a faulty or inflexible generalization” (Allport 1954, p. 9). Similarly, in her work on epistemic injustice, Miranda Fricker describes prejudice as “a judgement made or maintained without proper regard to the evidence,” and that is, as something that typically involves some degree of epistemic culpability (Fricker, 2007, pp. 32-33). On a view like this, there are at best sometimes “mitigating circumstances” such that “the subject’s patterns of judgement are influenced by the prejudices of his day in a context where it would take a very exceptional epistemic character to overcome those prejudices” (Fricker, 2007, p. 33). According to Fricker, by counting such a person as prejudiced, we are subjecting them (as we should) to “‘circumstantial’ epistemic bad luck” (Fricker, 2007, p. 33). Locating the negative aspect of prejudices in the way they are adopted or maintained allows Fricker to make room for benevolent prejudice, because it is not the negative content that is needed to identify a prejudice as something problematic. On this view, “prejudices are judgements, which may have a positive or a negative valence, and which display some (typically, epistemically culpable) resistance to counter-evidence owing to some affective investment on the part of the subject” (Fricker, 2007, p. 35). While they can be acquired in a way for which the subject is not strictly speaking culpable (although we may still hold the subject accountable), prejudices are characterized by an epistemic vice—namely, “resistance to counter-evidence,” which seems to correspond to Allport’s relating prejudice to “faulty or inflexible generalization” (Allport 1954, p. 9).

However, the conceptual link to faulty reasoning or stubbornness (potential modes of cognitive discrimination) can be challenged. Endre Begby (2013) points out that prejudiced beliefs may often be justifiably held on to in an apparently “inflexible” way. Far from being “rare exceptions” (Fricker, 2007, p. 33), those cases are the norm. This much at least is plausible, if prejudiced beliefs are typically not universal generalizations, but generic judgments about groups of individuals with a shared typical characteristic. For instance, the prejudice that women are less capable of abstract reasoning (an example both Begby and Fricker adopt from Arpaly, 2003) will most likely not be the view that all women have a lower aptitude for abstract reasoning than all men. It will rather be the view that the normal distribution of cognitive capacity in women has some overlap with that of men’s capacity, but that it is situated closer to the lower end of the scale. The prejudice is that most women are less gifted in this area than most men are. This view is much less vulnerable to counterevidence, because it predicts that some women will surpass many men in cognitive ability. Thus, the typical form of prejudice makes it questionable whether the feature of resistance to counterevidence provides a convincing characteristic of prejudice (Begby, 2013). More generally, Begby observes that

as we move toward a “non-ideal” form of epistemological theorizing—emphasizing the intrinsic capacity limitations of the human mind, as well as the particular and highly contingent informational constraints that ordinary agents are forced to operate under—we might find that there is no good grounds for saying that prejudiced belief, simply in virtue of being prejudiced, lacks epistemic warrant, and that prejudiced believers are always manifesting some form or other of epistemic irrationality.

Begby, 2021, p. 2

This consideration calls into question whether the kinds of views that strike us as prejudiced can be captured by a concept of prejudice that implies epistemic culpability or at least underperformance, and that can therefore be connected to prior cognitive discrimination (a target-specific epistemic underperformance).

Even when we are not swayed to revise our concept of prejudice by the consideration of the “intrinsic capacity limitations of the human mind” and the “contingent informational constraints that ordinary agents are forced to operate under” (Begby, 2021, p. 2), there remains a more general problem with grounding prejudice in discrimination. The problem is that discrimination is a comparative concept, in that it picks out lesser consideration or worse treatment, whereas prejudice is not comparative in that way. When the concept of prejudice is based on an epistemic shortcoming, this does not presuppose that this shortcoming comes to bear only on the consideration of evidence regarding certain groups and not others. The sense in which prejudice can be thought of as being a worse judgment than others is a different one from the sense in which discrimination is lesser consideration or worse treatment of someone. In the case of discrimination, the comparison is with the way the agent relates to some other objects of treatment or consideration (actual or hypothetical). In the case of prejudice, the comparison is with the kind of judgment the subject has the capacities to form. Prejudice is a deviation from what the epistemic subject can do as a knower. If prejudice is based on epistemic underperformance, the inferior way of reasoning could be universally applied. As preconceived judgments, prejudices could, in principle, be adopted with respect to every epistemic object—being prejudiced against everyone is not conceptually ruled out. People could fail to do what they were capable of, epistemically, across the board—they could be prejudiced against everyone and everything. In contrast, treating everybody badly would not fit the concept of discrimination. Someone who, in a paranoid way, suspects anyone they encounter—irrespective of group membership—of wanting to cheat them could be considered prejudiced against everyone. In contrast, it would be difficult to construe the verbally aggressive behaviour of a man towards women as an instance of discrimination if the same man were equally verbally aggressive towards other men.

The problem of accounting for the connection between discrimination and prejudice is not unique to the particular topic of (ambivalent) speciesism. In regard to this particular topic, however, the problem presents itself thus: if speciesism is basically a kind of discrimination, and if discrimination is lesser consideration or worse treatment, then accommodating benevolent or positive forms of speciesism would require showing how they are connected to some negative way of engaging with (evidence regarding) members of certain species. Speciesist prejudices could be such positive elements because the contents of prejudices can have positive valence as long as prejudice itself could be characterized by a different defect—by being based on cognitive discrimination. However, considering the intrinsic limitations of epistemic subjects and also the difference in the comparative versus noncomparative definitions of discrimination and prejudice, it is questionable whether prejudice can in fact be (smoothly) traced back to prior cognitive discrimination in this way.

For the purposes of this paper, I will cut the discussion of this problem short at this point. In principle, based on the view that the phenomenon of speciesism includes (positive and negative) prejudice, there are three conceivable solutions for aligning the definition of the concept given above with the phenomenon of speciesism. The first option would be to argue that whatever can count as speciesism—be it benevolent or hostile, be it a view or a kind of treatment—has to be based on the disadvantageous consideration or treatment of members of some species. The definition given above would remain unchanged, and speciesism would be the unjustified lesser consideration or worse treatment of beings that are classified as members of certain species. This account would eventually have to be supplemented by a theory of how different items in the speciesism domain (i.e., prejudice, stereotypes, and so on) actually do come down to disadvantageous consideration or treatment. Plus, it is faced with the problem that prejudice, if it is to be a component of speciesism, would seem to require a redefinition in comparative terms.

Second, one could adopt the view that the definition cited above is primarily one of speciesist discrimination, but not of the more general concept of speciesism that covers phenomena other than discrimination as well. The resulting definition would state that speciesism was discrimination or some other phenomenon x. An obvious concern about such a disjunctive definition is that an account of some unifying feature would remain a desideratum: one would want to know why one should regard the different phenomena as sufficiently unified to subsume all of them under the term “speciesism.”

The third option would be to reverse the causal connection option 1 allows for and posit that different kinds of items can count as speciesism in virtue of their contribution to discrimination. Speciesism would thus be defined as discrimination or x, provided that x plays a causal role in the occurrence or maintaining of that discrimination. While this would be a way to meet the challenge to option 1, which is to explain the connection among the different manifestations of speciesism, it risks ending up with an overinclusive notion of speciesism. The challenge would be to make option 3 more distinctive so that not all types of however marginal causal factors that feature in the occurrence of speciesism become instances of speciesism.

Ultimately, we might end up finding that supposed instances of benevolent speciesism typically check both boxes: that they are results of as well as contributors to speciesist discrimination. In this paper, I cannot argue extensively for one solution, but in order to proceed with the prior task of demonstrating the need to account for what appear to be instances of benevolent speciesism, I must adopt some preliminary understanding. In accordance with option 1, I will not change the definition of speciesism. In the rest of the paper, I will rely on the following preliminary characterization of benevolent speciesism:

Benevolent speciesism consists in beliefs, attitudes, or behaviours that involve positive emotions towards, ascriptions of valued characteristics to, or subjectively benign interactions with members of certain species, but that systematically fail to do justice to the nature, needs, and moral considerability of the targeted beings, manifesting some form of disregard.

This definition is somewhat at odds with the definition of speciesism, since benevolent speciesism—like prejudice—is not a comparative concept (because disregard is an absolute). More work would be needed to overcome this problem, but I will proceed with this tentative characterization of benevolent speciesism, assuming that it will suffice for the purpose of indicating that there is a phenomenon than needs addressing in the first place.

4. AMBIVALENT SPECIESISM

Speciesism has begun to attract the interest of empirical researchers as an object to be studied in its own right (Caviola, Everett and Faber, 2019; Everett et al., 2019; Caviola and Capraro, 2020). It has also been implicitly targeted by research into some specific behaviours such as meat-eating and related attitudes towards animals (Bastian et al., 2012; Piazza and Loughnan, 2016; Wang and Basso, 2019; McGuire, Palmer and Faber, 2022). Ambivalence in speciesism has not been studied much yet, even though the notion of ambivalent sexism has been brought to bear on research into human exceptionalism and its relation to sexist attitudes (Roylance, Abeyta and Routledge, 2016).

Scholarly attention to ambivalence in speciesism is only beginning to emerge. One study by Berk Efe Altınal and Göklem Tekdemir addresses “ambivalent speciesism” in these terms. Altınal and Tekdemir set out to develop an Ambivalent Speciesism Scale (2020)[3]. Their understanding of ambivalent speciesism differs from the one proposed here, though. They focus on what they call “protective speciesism,” modelled after one of the dimensions of benevolent sexism as defined by Glick and Fiske—namely, protective paternalism (Glick and Fiske, 1996). Protective paternalism occurs because women are seen as incompetent, yet “men are dyadically dependent on women (because of heterosexual reproduction) as wives, mothers, and romantic objects.” Consequently, men are led to take on “the protector and provider role” (Glick and Fiske, 1996, p. 493), and both men and women adopt such views as “a good woman should be set on a pedestal,” “women should be cherished and protected by men,” and “men should sacrifice to provide for women,” or agree with the notion that, in disasters, women should be rescued first (Glick and Fiske, 1996, p. 500). Altınal and Tekdemir take their cue from this conception but develop it in a distinct way. The hallmark of their definition of “protective speciesism” is agreement with the goals of the welfarist approach to the protection of the interests of animals (Altınal and Tekdemir, 2020, p. 495). This approach has it that animals may be used for human ends—they may be confined and killed, their bodies and body products may be consumed by humans, and so on—but use must be “humane.” That is, it must harm animals as little as is instrumentally required given the desired levels of use. There clearly is a certain ambivalence involved in the pursuit of “animal protection” thus construed: on the one hand, animals are seen as worthy of protection; on the other hand, they are seen as consumable.

However, analyses of ambivalent speciesism in terms of an endorsement of “animal protection” or “animal welfare”/“welfarism” differ crucially from inquiries into ambivalence as it pertains to sexism or racism. For one thing, Altınal and Tekdemir’s scale is not focused on isolating benevolent speciesism (as a separate counterpart to hostile speciesism); rather, it is mostly concerned with inherently ambivalent positions and plainly hostile ones. Items on the scale include such presumably ambivalent views as “the moral action regarding animals such as cows and chickens raised for food is to kill them in the most painless way possible when the time comes” and “anyone who cares about animal rights violations in animal experiments should demand laws and safeguards to regulate these experiments rather than oppose them.” While they include plainly hostile views such as “animals exist to serve people” and “people can use animals as they wish,” they include only one benevolent item: “animals should be given rights such as the ‘right to life’” (Altınal and Tekdemir, 2020, p. 499).

The focus on the endorsement of the doctrine of animal protection makes the Ambivalent Speciesism Scale quite different from the Ambivalent Sexism Inventory. Apparently ambivalent views such as “animals ought to be killed in the most painless way possible” could be further unpacked and do not make up the whole of ambivalent speciesism by themselves. Benevolent sexism, by comparison, is conceptualized as a distinct category, complementing hostile sexism and including an “idealization of women” as “pure creatures who ought to be protected, supported, and adored” (Glick and Fiske, 2001, p. 109). There is no analogue to these cherishing and glorifying, apparently purely positive views and behaviours on the Ambivalent Speciesism Scale. The Ambivalent Sexism Inventory (Glick and Fiske, 1996) consists of two sets of eleven items each, one covering benevolent sexism and the other covering hostile sexism. When trying to understand ambivalence in speciesism, we should also look for separate or separable benevolent and hostile manifestations.

A closely related point is that the program of animal protection/welfarism is not all that ambivalent in itself. Welfarism does claim to strike a balance between interests in using animals and interests in protecting animals. And the limit it imposes on use is in principle not arbitrary, but is supposed to give animals the level of protection that they are due—no less, but also no more. Whatever ambivalence there appears to be in this doctrine is actually reconciled in unified precepts regarding how animals are to be treated. Consequently, the aspect of compensation is absent from the ambivalence that marks animal protection. The welfarist position promotes not more than average protection (as in paternalism), but supposedly adequate (though in reality far more limited) protection for animals. The view that animals are worthy of some protection is not adopted so that it will function to compensate for the view that they may be used. Furthermore, no interrelation of views from different dimensions (warmth and competence) comes into play here, either.

Ultimately, focusing on protective paternalism as the manifestation of benevolent speciesism places overly narrow constraints on our view of the phenomenon. Among the different types of benevolent speciesism, there is an especially influential one that does not fit this frame at all: the attribution of competence, where this attribution underlies the view that humans need not care for animals. This type of benevolent speciesism recommends negligence rather than paternalism and is probably rampant in attitudes towards wild animals, as well as some companion animals.

In the following subsections, I will outline some possible patterns of ambivalent speciesism, distinguish ambivalence from complex negativity, and indicate benevolent speciesism’s potential to harm animals.

4.1 Compensatory recognition

The patterns of ambivalence in speciesism can be expected to differ from those in other forms of discrimination, just as patterns of ambivalence in racism and sexism differ from one another, because the pattern of humans’ social interaction with and subordination of nonhuman animals is itself unique. For instance, since the relationship between humans and nonhumans is not “complementary” in the same way as heterosexual intrahuman relationships are construed, there might be no strict analogue to “complementary gender differentiation,” a pattern underlying the attribution of warmth to women and the attribution of competence to men. Yet there might be closely related patterns (see §4.4, infra), including a pattern of compensatory recognition, where animals are ascribed certain capacities and denied others, and where this pattern is consistent with assigning animals an overall low social status.

Consistent with this social positioning is the denial of reason in animals. It is because animals are not considered rational that they supposedly have no interest in running their own lives, having a say in how communal life should go, and being free from constraint and coercion. In as much as animals are, nevertheless, somehow valued, there is a motivation for humans to counterbalance this negative view with benevolent views and attitudes. If the resulting ambivalence towards animals likewise is bidimensional ambivalence, these positive views would, as in the case of sexism, relate to the warmth dimension. One manifestation of this compensatory recognition of a positive feature is the perception of certain animals as “cute” and “lovable.”

Viewing someone as cute itself is not actually a prejudice—it is the result of a match between some physiological features and certain aesthetic preferences. However, the lovability that comes with perceived cuteness is not naturally paired with respectability. The love and adoration for “cute” animals are often possessive and looking down on animals. Calling someone “cute” belittles them, albeit in an affectionate way. It is no coincidence that it is one of the first examples of benevolent sexism cited in Glick and Fiske’s 1996 paper: “A man’s comment to a female coworker on how ‘cute’ she looks, however well-intentioned, may undermine her feelings of being taken seriously as a professional” (Glick and Fiske, 1996, p. 492). Even without the sexual overtones that exist in the human case, the focus on cuteness can be detrimental to considerate attitudes towards animals as well. There is an accidental hint at this problem in the animated Disney movie Zootopia, where, as in many fictional works, the role of animals is a merely metaphorical one. A literal interpretation would, in this case, be very revealing. In the film, police officer Judy Hopps, a rabbit, on her first day at the police station, corrects the friendly cheetah receptionist Benjamin Clawhauser, when he calls her cute: “You probably didn’t know, but a bunny can call another bunny cute, but when other animals do it, it’s a little...” She needs to say no more, for Clawhauser immediately realizes his mistake and apologizes for “stereotyping” Hopps. When we do not view this as an allusion to human stereotypes, we come across a noteworthy warning that filmmakers unwittingly have put in Judy’s mouth. Maybe animals would want to object to being viewed as “cute.”

In Zoopolis, Donaldson and Kymlicka criticize the tendency of some abolitionists to view juvenile features as signs of an unnatural or undignified existence for domesticated animals (Donaldson and Kymlicka, 2011, pp. 82-86). Appeals to naturalness and dignity in fact seem unhelpful. Yet abolitionists might be on the right track, for a focus on juvenile—cute—features might play a role in maintaining the benevolent complements of hostile speciesism, by presenting animals as “cute but dumb.” The resulting pattern of social interaction is one of affectionate subordination.

Given the severity of the harms involved in discrimination against nonhuman animals, benevolent speciesist attitudes other than enthusiasm for some animals’ cuteness may not come to mind easily. Awe at the “surprising” performances of animals in behavioural studies seems like a candidate. However, on second thought, this surprised type of awe is simply the mode in which a negative prejudice reveals itself—that is, the prior underestimation of animals’ capacities.

Yet, something very close to this kind of surprised awe does play a role in speciesist compensatory recognition. Consistent with the denial of rationality in animals is the compensatory attribution of specific forms of technical excellence. In contrast to bidimensional ambivalence, the pairing of rational incapacity with technical excellence is an incongruity within the competence dimension. It exploits the fact that competence can be very compartmentalized. An animal might have some specific capacity but lack another (overall more valued) capacity. For instance, while animals supposedly lack rationality, they can have highly distinguished sensory capabilities. Even in humans, there are more than the classical five senses, but there are even more in other animals, who experience the world, inter alia, through echolocation, magnetoreception (sensing the earth’s magnetic field), electroreception (sensing electrical fields), hygroreception (sensing changes in moisture in the environment), and infrared sensing. When these sensory capacities are alluded to outside academia, they are sometimes mystified as animals’ “sixth sense.” The conceptualization of a sense not possessed by humans as “the sixth sense” is an incidence of perspectival anthropocentrism in itself. But beyond that, it also pushes the respective capacity into the realm of the supernatural. Man remains the measure of all things natural, and the seemingly inexplicable perceptions of animals are to be met with awe and wonder. This puts animals on a pedestal but also out of the reach of understanding. Meanwhile, this idealization happens in a circumscribed realm that is ultimately deemed irrelevant for a being’s social standing. It is a recognition that humans can afford to extend without jeopardizing their claim to dominating animals. The attribution of these awe-inspiring sensory powers might well be a compensatory or legitimizing counterpart to hostile speciesism—not appeasing animals, but humans themselves, who hold animals in some regard while maintaining the belief in the moral justification of species hierarchies. The pattern of attributing rational incapacity, but technical excellence to animals results in their exclusion from consideration while avoiding their total disparagement.

4.2 Complex negativity

Compensatory recognition must be distinguished from a pattern that manifests an altogether different sort of ambivalence than the one we are interested in here. In the context of an overall negative evaluation, some ascriptions of seemingly pro tanto positive traits do not create ambivalence in the relevant sense. Rather, they make the predominating negativity more complex. This occurs when a positive attribute is invoked to explain the overall badness of the animal, as in the following example, taken from an online forum for hunters:

Shrewd critters them magpies...they are downright rummaging in bushes and trees for songbirds’ nests. On the one hand they are supersmart and on the other hand unbelievably stupid. When you shoot down a magpie and there are others in the immediate proximity you just have to wait, and a few curious ones will gather at the site to see what happened with [sic] their conspecific :-)) I once got 6 in one go that way[4]. (my translation)

European magpies incur hate from some humans for being “nest robbers,” with more esteemed “songbirds” as their victims (actually, magpies are themselves classified as songbirds, but not by the folk notion of what a songbird is). In this example, the despised animals are given some credit for a technical skill in finding nests, but this competence is highly compartmentalized and does not weaken the overall disparagement but is rather instrumental to it. The birds are seen neither as warm nor as overall competent. The little skill (competence) they are recognized to possess works only to explain why they are wholly bad animals—it is a vicious excellence. Even within the dimension of competence, it is, however, overshadowed by the incompetence to which their behaviour in the face of a hurt or dead conspecific is attributed.

This pattern is evidence that, in the context in which this description is published, there is no need to compensate for a negative ascription to an animal, because a negative view of the animal prevails. In contrast to cases of ambivalent speciesism, no truly positive feature is attributed here, as the prima facie positive attribution is but a part of the negative characterization itself: when bad magpies are smart, they are not intelligent, but sly—or “shrewd.” Such complex negativity draws on pro tanto positive features (or capacities that, while they are often admired, are not necessarily positive independently of how they are put to use), turning them into vicious manifestations of excellence. This negativity is to be kept apart from the ambivalent speciesism that interests us here.

4.3 Self-serving recognition

The role of the benevolent element in ambivalent speciesism is not always best described as compensatory, but rather as justificatory. This is true of the phenomenon one could refer to as deprioritizing or self-serving recognition. It occurs when a positive feature is ascribed to an animal, where recognition of this positive feature would lend rational support to taking a dismissive attitude towards animals’ potential needs and entitlements. When an animal is credited with a feature from the competence dimension, this can underpin the rejection of calls for assisting that animal. An attribution of greater competence also has a downside: men, for example, are less likely than women to receive help (Glick and Fiske, 1996, p. 492; cf. Eagly and Crowley, 1986). When animals are said to have a “sixth sense” that allows them to know about impending disasters before humans, this provides a reason to deprioritize or entirely exclude them when it comes to rescue efforts.

In a similar way, benevolent speciesism might play a role in opposition to interventions on behalf of animals in nature. In his reconstruction of the supposed problem with resistance to such interventions, Oscar Horta focuses on negative speciesist attitudes, arguing that resistance often expresses “concerns that are not really taken seriously when human beings are involved…. If humans were suffering and dying and altering an ecosystem would help them, nobody would object that it was too dangerous or that we had no right to do it” (Horta, 2010, p. 84). Suspecting that assistance to animals in the wild is opposed because “while anthropocentric and environmental ideals are seen as important, the wellbeing of nonhuman animals is seen as completely irrelevant,” he concludes that dismissal of the need to help wild animals is thus based on “speciesist positions” (Horta, 2010, p. 84). As a further reason for opposing assistance to wild animals, Horta considers the belief “that nature is a rich source of value because of the existence of nonhuman sentient animals who have happy lives” (Horta, 2010, p. 75). The idea that animals have happy lives is not discussed, though, as an expression of speciesism in its own right. However, if speciesism can be ambivalent, this would suggest that viewing wild animals as living happy lives can itself be a result of (benevolent) speciesism. Underlying this idea is the representation of animals as capable and well equipped for their lives in their respective ecological niches. The trope portraying animals as capable can be a manifestation of the compensatory technical excellence, rational incapacity pattern, but it can also be a self-serving attribution, where the hostile from of speciesism—the dismissal of the claim for help—flows directly from the benevolent form. The resulting attitude of negligent admiration diminishes the fact that animals have a far smaller arsenal of methods, instruments, and material for addressing individually suffered harms (no antibiotics, no sick leave, no safe spaces, no equal-opportunities officer, and so on).

4.4 Incapacity-derived recognition

Another pattern of ambivalence occurs where benevolent attributions are based on or logically derived from hostile views. In contrast to cases of complex negativity, they are marked by the subjectively benign feeling that is characteristic of benevolent prejudice.

There are good reasons to believe that most people perceive some animals to be emotionally warm, not just on account of their cuteness and lovability, but also on account of their trustworthiness and perceived emotional availability. This is evident, for instance, in the commendation of animals for use in therapy, especially in child therapy. The use of nonhuman animals for therapeutic purposes (commonly euphemized as animal-assisted therapy, AAT) exploits the perception of animals as being trustworthy and not judgmental, with the animals’ trustworthiness often intended to spread to the therapist (cf., e.g., Jones, Rice, and Cotton 2019), as the following report explains:

It is clear to me that David has been able to express himself through AAT due to his presence in the world of animals: the security, warmth and acceptance provided by Mushu, David’s identification with Cuddly along with his experience of building a mutual understanding with Cuddly, the play with and caring for the rats and hamsters, and finally his perception that I am part of that warm, accepting, trusting, respectful world, which allows him to trust me also.

Parish-Plass, 2008, p. 24

Two mechanisms could be at play in situations like this. First, what could (and indeed seems to) be crucial here is that a genuine capacity is ascribed to animals—or at least to Cuddly, who partakes in “mutual understanding” with the child. On this interpretation, the benevolent attribution would fit within a larger pattern of compensatory recognition, where animals’ lack of rationality (competence) is counterbalanced by something like an emotional sixth sense—a special emotional availability and intuitive understanding (warmth) that ultimately extends trustworthiness to the therapist: “The relationship between the therapist and the animals makes the therapist less threatening and more trustworthy in the eyes of the at-risk child, who has good reason not to trust adults” (Parish-Plass, 2008, p. 27).

Second, the benevolent characterization of animals as trustworthy might often also or rather alternatively be predicated on a hostile view—namely, on the view that animals supposedly lack cognitive capacities (competence) themselves, which enables animals to count as warm. On this type of view, animals’ trustworthiness is a direct consequence of their incompetence: animals are too dumb to have bad intentions. They do not deceive, because they wouldn’t know how and why. It is not necessary for animals to compensate for their supposed lack of intelligence by possessing an unrelated trait from another dimension of personality, since lack of intelligence has a direct effect on traits included in the warmth dimension. One pole of the ambivalent speciesist view supports the other: animals are dull; therefore, they are trustworthy. Underlying the subjectively positive perception that animals are trustworthy can be the common view that they are simply dull. Humans can “entrust” their secrets to them, and animals will not judge them—they don’t understand a thing. They don’t even know how to judge. This type of incapacity-derived recognition results in a kind of patronizing admiration or condescending commendation that can acknowledge animals’ merits in one dimension without risking laying the grounds for any entitlements for animals.

This type of ambivalence in particular points to an apparent asymmetry in the hostile and benevolent elements of ambivalent speciesism (sexism, racism, and so on). Sets of entirely negative or hostile views about or attitudes or behaviours towards members of certain species would register as speciesist even in the absence of any counterbalancing, derived, or otherwise related positive or benevolent views, attitudes, or behaviours. The inverse is not true. For example, if David himself simply viewed Cuddly as trustworthy, without basing this in any way on a presumption that Cuddly was dull or otherwise incompetent, he would not open himself to the criticism of being speciesist. It appears that only an adult who, observing and adopting David’s trusting attitude, explained Cuddly’s trustworthiness with reference to the animal’s lack of cognitive sophistication would even invite the suspicion of being speciesist. This would mean that whether a benevolent view (or behaviour) counts as an element of speciesism would depend on its integration into a larger set of an agent’s views (or behaviours) that is already apparently speciesist—where the independent characterization of this set of views as speciesist must ultimately refer to some hostile view or negative way of considering or treating members of certain species.

4.5 Local benefits, global disadvantage

Benevolent speciesism can come with some local benefits for some animals—such as access to the outdoors for “independent” cats and consideration as a family member for “loyal” dogs. But not only can these local benefits turn out to be harmful themselves—as when “outdoor cats” get run over by cars, accidentally locked up in a neighbour’s garage, or abducted, or when dogs’ human families happen to be abusive towards their members—but there are more ways in which benevolent speciesism can entail harm.

First, because benevolent speciesism can play a compensatory role for hostile speciesism (see §4.1, supra), it ultimately is a facilitating factor for the global disadvantage that speciesism bestows on animals. Insofar as it balances negative views and harmful behaviours and thus lets humans feel more comfortable with hostile speciesism, it plays a crucial role in maintaining the subordination of humans, especially in circumstances marked by heightening sensitivity to animals’ genuine moral considerability. Given this facilitating function, it is no wonder that insights into ambivalence in other contexts imply a warning against tapping into benevolent speciesism in an effort to overcome its hostile forms. As Bergsieker et al. conclude, “persistently emphasizing positive unidimensional outgroup stereotypes may hinder repair of closely linked (but unspoken) negative stereotypes on other dimensions” (Bergsieker et al., 2012, p. 1233). That benevolent speciesism takes on this facilitating function is quite obvious in cases of compensatory recognition (§4.1), and something similar happens in the case of incapacity-derived recognition (§4.4). In the case of the former, the relation between the recognition of a supposed positive trait and the attribution of a negative trait takes the form of an antithesis (“cute, but dumb”); in the case of the latter, the relation is that of an inference (“dull, so trustworthy”). The latter might make for an even more efficient silencing of any suspicion that the overall treatment the animal in question receives could be inadequate.

Second, benevolent speciesism can be more directly harmful. This is the case for self-servingly, negligently benevolent speciesism (see §4.3, supra), as when characterizations of animals as being “capable” or “independent” or having a “sixth sense” underpin withdrawal of care and the lack of a feeling of responsibility. Given animals’ overall low social status, a potential difference from other types of ambivalent prejudice might be that even the seemingly benevolent attributions of some capacities to animals might be more often directly self-serving to humans than in the case of intrahuman prejudice.

Third, benevolent manifestations of speciesism can be harmful because they are merely precariously benevolent—harm is incurred when the positive stereotype is “disappointed.” Simplified positive views that are used as standards for individuals can turn into negative evaluations of the individual who does not fit the mould (as it happens in the context of victim blaming in cases of acquaintance rape; see §2, supra). For instance, the true belief that rabbits are social animals can take the form of a simplified benevolent stereotype according to which a normal rabbit will get along with any other rabbit. This view supports the one-chance approach to animal sociality that many owners of nonhuman animals take because they are not prepared to give animals the opportunity to choose their friends, which would require accommodating rejected social partners. If an individual rabbit does not get along with the one conspecific an owner has provided, it is concluded that “this rabbit does not get along with others” and the apparently unsocial rabbit is kept in isolation. The simplified benevolent view is disappointed, and the conclusion is that the individual is in some sense a defective exemplar of the species—with drastic practical consequences. A similar mechanism is at work when the benevolently speciesist stereotype of “the loyal dog” is disappointed by a dog’s defensive behaviour, where the consequences for the animal can sometimes be lethal.

5. CONCLUSION

Once we identify subjectively benign views of and behaviours towards animals as potential expressions of speciesism, our view of “ambivalence” in the human-animal relation changes, and different aspects of this relation appear much more unified and, unfortunately, defective. As many animal advocates have long since pointed out, the supposedly excessively considerate relationships some humans have with their “pets” are not the polar opposite of the outright domination of animals on farms and in laboratories, but many of these relationships simply represent different flavours of disregard. It is not only or not even typically the case that benevolent speciesism involves excessive caretaking that is meant to offset negligence in other relationships with animals. Instead, both benevolent and hostile speciesism can fulfill the same functions. Both can underlie neglect and disregard. Both can serve as justifications for not giving animals their due. Both can serve to spare us thinking about animals more carefully. The need to disguise some of the dismissive attitudes towards animals as benevolence, however, does open an opportunity for overcoming both benevolent and hostile speciesism, since this need arises from genuine positive attitudes and attention towards animals. At the same time, animal advocates need to work on becoming alert to their own propensity for benevolent speciesism, as it can interfere with their genuine efforts to identify the goals that are worth pursuing on behalf of animals. In view of the phenomenon of ambivalent speciesism, animal advocates also face the challenge of anticipating and managing, as best as they can, the effects of bolstering positive perceptions of animals. For example, countering common perceptions of certain animals as aggressive, dangerous and/or disgusting by presenting them as cuddly and lovable might inadvertently result in introducing benevolent speciesist views that still hinder genuine moral consideration of these animals.

As efforts to improve the situation of animals in a human-dominated world are underway, it is important to prepare for the proliferation of benevolent speciesism. In a society where the social standing of some animals is raised, it is to be expected that benevolent speciesism becomes more frequent. At the same time, negative stereotypes might merely be omitted from public utterances rather than reversed in the minds of human speakers (see Bergsieker et al., 2012). In this situation, benevolent speciesism can be an indicator of remaining problems in humans’ perspectives on nonhuman animals.

As far as the philosophy of different forms of discrimination is concerned, the phenomenon of ambivalent speciesism shines a light on the challenge of producing a coherent account of the relation between different concepts commonly used to explicate phenomena such as speciesism, racism and sexism. It seems that further investigation into ambivalence in speciesism (or rather, into speciesism in general) has to revisit the question of how precisely prejudice and discrimination are interrelated, especially against a backdrop of “non-ideal epistemology” (Begby, 2021).

Research is also needed to test the roles of various forms of benevolent speciesism. We need to better understand whether and how such attitudes contribute to maintaining institutions that are “subjectively benign” for humans, but subordinate nonhumans. As far as the relation of humans with their companion animals is concerned, a higher prevalence of benevolent speciesism in individual humans could possibly be indicative of a lower-quality relationship. The idea here would be that higher levels of benevolent speciesism might predict traits such as less understanding of animal’ needs, less investment in enriching the spaces in which they are kept, less spending on health care, and less reluctance to abandon them. Likewise, benevolent speciesism might play a key role in the underestimation of the need for assistance of animals in the wild and reduced willingness to intervene on wild animals’ behalf. While this paper has dealt mainly with attitudes towards companion animals and animals in the wild, it would also be useful to address the role that benevolent speciesism plays in human-animal relations that are more straightforwardly violent and involve obviously coercive utilization of animals. What role does benevolent speciesism play in those areas where positive attitudes are much more in tension with the overall nature of the interaction with animals, as in the use of animals for food or in experiments? The notion of ambivalent speciesism could complement and possibly further consolidate existing research into the complicated nature of these types of interactions with animals, such as work on the “sacrificial symbolism” practiced in the context of animal experimentation (Arluke, 1988) or research into mechanisms for coping with cognitive dissonance relating to the consumption of animals’ bodies (e.g., Bastian et al., 2012).

The role of benevolent prejudice in maintaining negligent behaviour towards animals is one of the important reasons for seeking to better understand ambivalence in speciesism. That benevolent forms of speciesism so far do not receive much attention is understandable, given the moral significance of hostile speciesism that is rampant and draws attention for good reasons. Also, the nature of benign expressions of prejudice—their subjectively positive feeling—puts them above suspicion for those who are prejudiced as well as for bystanders. However, to better understand what sustains hostile speciesism, we should also look out for its positive counterpart and address speciesism in all its actual ambivalence.