1 Introduction

Over 40 years after the publication of Animal Liberation, Peter Singer still finds himself explaining to his critics “[w]hy speciesism is wrong” (Singer 2016), which—at least according to his original definition—amounts to explaining “why prejudice is wrong.” This is a remarkable and unfortunate situation. It has arisen because the term ‘speciesism’ has often been used in a purely descriptive sense, to refer to the view that species membership is (directly or indirectly) morally relevant, but without thereby discounting this view as mere prejudice. This paper argues that this terminological choice severely impoverishes the ethical vocabulary, and obscures the considerable common ground among theorists with differing views on the relevance of different properties for moral consideration. It is rarely ever recognized that with only a few exceptions, even those who defend the preferential treatment of members of certain species (typically, of humans) have reason to agree with this evaluation: they, too, should find the lack of agreement on the normative content of the concept speciesismFootnote 1 problematic. There is a palpable need for a descriptive concept to pick out, but not condemn as unjustified, ethical stances on the moral relevance of species membership. However, such a concept has to be expressed by a different term than ‘speciesism.’

This paper explains why normative and applied ethics need both, the normative concept speciesism that can be used to address prejudice or discrimination and a distinct descriptive concept to classify certain positions in the debate about criteria for receiving moral consideration.Footnote 2 The paper does not propose a novel definition of speciesism. It rather offers a more detailed defense of a normative account of speciesism than has been given so far and presents the case for a clear terminological distinction between it and a separate, purely descriptive concept—speciescentrism—for the mere classification of ethical positions. Section 2 lays out the basic distinctions between competing concepts of speciesism and identifies the most important pragmatic requirements for the use of the term ‘speciesism.’ Section 3 points to problems that result from reducing the normative concept of speciesism to its descriptive content and proposes a terminological alternative that links up with existing ethical vocabulary. Section 4 replies to some possible objections to the proposal. Section 5 concludes with a brief outlook on the potential of the normative notion of speciesism to be useful in research into discrimination.

2 Arguing Semantics

Originally, ‘speciesism’ denoted a kind of prejudice or discrimination, where both these referents were understood in a normative way. That is to say, the discrimination that speciesism was said to be was not just differential treatment—but unjustified differential treatment, cf. section 2.1. The use of ‘speciesism’ as a merely descriptive technical term for classifying positions in normative ethics therefore involved a crucial deviation from its original meaning. Arguments concerning nonhuman animals’ claims to moral consideration have often been bound up in intricate ways with a struggle for interpretative authority over the meaning of the word ‘speciesism.’ The choice of either the normative or descriptive concept is often not made explicitly and there has been little debate about the preferable choice in philosophical contexts (but see Jaquet 2019 for a defense of a descriptive account). Overall, little attention has been paid so far to the fact that the appropriation of the term ‘speciesism’ as a piece of purely descriptive philosophical terminology comes with costs to the ethical discourse. The term ‘speciesism’ has been used to cover a range of different positions. This section brings out two key distinctions within the pool of competing speciesism concepts. Each of the two distinctions is sometimes alluded to in the literature, but their interplay is largely disregarded.

2.1 The Normative and the Descriptive Concept

The first distinction we must attend to is the one between a normative and a descriptive concept of speciesism. Speciesism was first introduced as a normative concept. When Richard Ryder coined the term ‘speciesism,’ he intended it to denote a failure to recognize a presumed moral continuum between humans and nonhuman animals (Ryder 1992, 170–1), a failure that results in discrimination and that he later described as a form of prejudice, much like racism:

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, 16)

Ryder’s term ‘speciesism’ is rather inclusive: it applies to all sorts of unjustified beliefs analogous to racial prejudice as well as the resulting behavior. Singer took up Ryder’s term, but focused his definition on the broadly epistemic basis of discriminatory behavior: he characterized speciesism as “a prejudice or attitude of bias in favor of the interests of members of one’s own species and against those of members of other species” (Singer 2009 [1975], 6, emphasis added). The reference made to “interests” here is noteworthy. By virtue of this reference, speciesism is not construed as just any kind of unequal consideration, but as unequal moral consideration.

Elsewhere, when highlighting the peculiarity of speciesist beliefs, Singer is not speaking of a “prejudice” anymore. He simply says that “‘speciesism’ refers to the view that species membership is, in itself, a reason for giving more weight to the interests of one being than to those of another” (Singer 1990, 10, emphasis added). By calling speciesism a “view” rather than a “prejudice” or “bias,” he provides a purely descriptive account that does not build reference to some epistemic shortcoming or a lack of moral justification into the concept.

After ‘speciesism’ was picked up as a technical term in ethical debates, the restriction to the concept’s descriptive core has become common for almost obvious pragmatic reasons. Defenders of the preferential consideration of humans adopted the term ‘speciesism’ for naming their own views but could clearly not include the judgment of a mere prejudice in their accounts (e.g., Holland 1984; Cohen 1986; Chappell 1997). In seriously engaging with the defenses of favoring humans, their critics no longer relied on the normative notion that characterizes speciesism as unjustified by definition, either. The need for a purely descriptive concept that serves the classification—and not stigmatization—of positions is palpable in debates about criteria of moral relevance. While still speaking of “speciesism,” friends and foes of the preferential consideration of humans alike often eliminate evaluations conveyed by terms such as ‘prejudice’ and ‘discrimination,’ substituting for them expressions like ‘view’ and ‘difference in treatment.’ In this adjusted version, speciesism is customarily defined as “difference in treatment based on an appeal to species membership” (Graft 1997, 107), the belief “that membership in a particular species is morally relevant” (Bernstein 2004, 380), and “the belief that, as a matter of fact, the division between humans and other animals is morally significant” (Holland 1984, 291). Speciesists are accordingly described as “believers in the justifiability of favoring one species over another “(Pluhar 1995, 126). Speciesism is thus construed as a superordinate concept to anthropocentrism. But the two terms have also been treated as synonyms, next to less common expressions such as “homocentrism” (Pluhar 1995, 10), “homo-sapiens-centered speciesism” (Jamieson 2008, 108), “human speciesism” (Bernstein 2004, 380), but also an established term like “humanism” (Frankena 1979), which carries rather positive established connotations. This multiplication of terms of art is, however, not the gravest problem with the transformation of ‘speciesism’ from a term that expresses contempt into a descriptive technical term in ethics, as I will show in section 3.

2.2 The Wide and the Narrow Concept

Both Ryder and Singer later specified what kind of (prejudiced) belief they took ‘speciesism’ to denote. Ryder explains that there is a specific sense of ‘speciesism’ that ties the term to defenses of discrimination appealing “solely” to species membership, and he suggests calling this type of position “strict speciesism” (Ryder 1998). Singer gives the same kind of clarification: “The term ‘speciesism’ refers to the view that species membership is, in itself, a reason for giving more weight to the interests of one being than to those of another.” (Singer 1990, 10, emphasis in original) Singer explicitly distinguishes speciesism from the view that there are differences in humans’ and nonhumans’ capacities which are morally relevant and therefore justify the unequal consideration of humans’ and nonhumans’ interests. He insists that this appeal to factual differences is not adequately called “speciesism,” but constitutes a “very different position” (10). The literature that has taken up this originally provocative neologism has for the most part not adhered to Singer’s differentiation. The term ‘speciesism’ has been applied to the view that species membership is in itself morally relevant as well as to the alternative view according to which there is a correlation between species membership and a morally relevant characteristic that grounds humans’ claim to preferential (or exclusive) moral consideration. Table 1 shows some of the labels that have been used to mark the distinction (for a critique of some of the terminological options, see Horta 2010, 251–3). Notably, these labels come with a variety of characterizations of the relation between the two views.

Table 1 Terminological distinctions: direct and indirect appeals to species membership

According to Singer, everyone who has defended “indirect speciesism” (LaFollette and Shanks 1996) has just talked past him. On an abstract level, his focus on “strict” or “bare” speciesism can seem quite plausible: whatever the proponent of a certain set of characteristics may believe about their distribution, if she does not claim species membership to be directly morally relevant, why then call her a “speciesist”? The answer to this question seems to depend on whether a wide notion of speciesism is interpreted in a normative or a descriptive sense.

2.3 Normative Content plus Wide Scope

In this subsection, we will first look at the reasoning behind defining speciesism in a wide and normative way. The descriptive concept will be discussed in sections 3 and 4. The following shall be our working definition of a wide and normative concept of speciesism:

Speciesism is the unjustified comparatively worse consideration or treatment of those who are not classified as belonging to a certain species (or group of species) whose members are favored, or who are classified as belonging to a certain species (or group of species) whose members are disregarded. (Horta and Albersmeier 2020, 4)

This account defines speciesism as unjustified and it does not restrict it to behaviors or ways of considering others that are based on an appeal to species membership. It is thus normative and wide. This section defends the idea that the term ‘speciesism’ should be reserved for this kind of concept. For further details on the claimed advantages of the definition, see (Horta and Albersmeier 2020).

Of those who have opted for such a wide and normative notion of speciesism, only some have explicitly addressed the dissent over the concept’s scope. Joan Dunayer, as one example, has criticized Singer’s restriction to the narrow concept for jeopardizing the originally intended analogy between racism, sexism, and speciesism:

It’s racist to give greater weight to the interests of whites than nonwhites, sexist to give greater weight to the interests of males than females, and speciesist to give greater weight to the interests of humans than nonhumans for any reason. (Dunayer 2004, 3, emphasis in original)

Along similar lines, Oscar Horta highlights that racism and sexism include “all kinds of unjustified disadvantageous consideration or treatment”—whether they are defended by an appeal to physical traits or characteristics presumably correlated with those traits (Horta 2010, 246). On this view, there is a commonality that warrants labelling both as “speciesist:” appeals to species-specific characteristics and “strict speciesism” somehow result in the same kind of discrimination—discrimination along species lines. We might call this kind of reasoning in favor of a wide notion of speciesism “outcome-oriented.” However, it is unpersuasive to all who question the merit of upholding the analogy between racism, sexism and speciesism to begin with. To have any appeal for those who are unconvinced that giving “greater weight to the interests of humans than nonhumans for any reason” actually qualifies as discrimination, more “source-oriented” considerations would have to point to some flaw in the justification offered for the position or behavior in question—a flaw that bears a suitable relation to the feature “species membership.” “Source-oriented” considerations focus on the reasons for the lack of justification in a speciesist position. It might seem that a straightforward way to argue for a wide notion of speciesism based on source-oriented considerations would be to claim that “strict” and “indirect” speciesism both involve some ultimately unjustified or materially false claim concerning species membership and therefore deserve to be subsumed under ‘speciesism.’ For strict speciesism, it is the claim that species membership is itself morally relevant (which is deemed implausible). For indirect speciesism, it would be some unjustified empirical claim about the lack of a morally relevant characteristic in members of certain species.

The insight that many views on moral considerability lack justification due to wrong empirical claims about the distribution of certain characteristics is what the ‘argument from marginal cases’ or ‘argument from species overlap’ (ASO) is supposed to evoke. The ASO is directed against the idea that there is some property (which might even be a relational property, cf. Horta 2014, 145) that can be cited as a plausible necessary condition for being morally considerable and that this property is found in all and only humans. By pointing to humans who lack the supposedly relevant property (e.g., some advanced degree of rationality), and/or by showing that there are nonhuman animals who do possess it (thereby revealing a ‘species overlap’ regarding the possession of the proposed property), the ASO undermines possible bases for affording humans primary or exclusive moral consideration.Footnote 3

However, the justification for an indirectly speciesist view can fail not only because of a false empirical claim about the distribution of some characteristic, but also because its choice of a morally relevant characteristic is deemed implausible. In that case, the reference to species-membership that would warrant the label ‘speciesism’ seems to be lost. Suppose, for example, that humans were the only primates with opposable thumbs. If it were claimed that opposable thumbs were necessary for being morally considerable, the source to which the supposedly resulting speciesism could be tracked down would not be a false claim concerning species membership, but an implausible claim about the moral relevance of opposable thumbs, which does, however, not mention species at all.

Still, source-oriented reasoning may show that the label ‘speciesism’ is warranted if individuals’ species membership crucially bears on the moral reasoning in some way. This will be the case whenever it can be argued that the choice of the supposedly morally relevant characteristic itself has to be explained by a prior (conscious or unconscious) commitment to securing a view that favors members of certain species. In these cases, the view in question would be identified as speciesist not because the arguments that support it feature a premise that mentions species membership and is false or unjustified, but because the argument is on some level affected by a speciesist bias. This reasoning requires that speciesism is given wide scope in yet another sense. Again contrary to what Singer’s definition states (see section 2.1), the bias that is speciesism cannot only be a bias against interests, i.e., it cannot only have scope over the way individuals think about others in moral terms. A speciesist bias is not only a distortion in moral reasoning, but a distortion in any kind of reasoning regarding members of certain species. This lines up with how supposedly analogous terms like ‘racism’ and ‘sexism’ are widely used, even though in the case of ‘speciesism’, ethical debate has so far usually maintained the narrow focus on consideration of interests, i.e., moral consideration. When this limitation is overcome, it is easy to see why apparently indirectly speciesist views ought to be counted as speciesist, i.e., why the moral notion of speciesism ought to be wide.

The difference between outcome- and source-oriented considerations in favor of the wide notion may appear to be a mere difference in emphasis, because outcome-oriented views could never be free from source-oriented considerations. Since applying the label ‘speciesism’ at least requires determining that justification is lacking, there always have to be considerations of the grounds of justification of a given speciesist position. However, the two perspectives could at some point come apart.

Source-oriented perspectives seek to identify the influence of some form of “bare” or “strict” speciesism on what appears to be a speciesist position. They rule out honest, non-systematic mistakes (where non-systematic is supposed to mean not even affected by an implicit bias) as ultimate sources of speciesist positions. An entirely outcome-oriented perspective would not do that. It would subsume under ‘speciesism’ ways of treating members of certain species which are based on beliefs about them that just happen to be wrong, but which individuals might be epistemically justified in holding. There may be reasonable disagreement about this issue, but here, I will assume that a wide and normative notion of speciesism should be based on source-oriented considerations in order to exclude such “contingent discrimination.” This idea requires some illustration, but it is notoriously difficult to come up with an actual case in which a group is disadvantaged for obviously contingent reasons. So, the following example should be read as a hypothetical—a thought experiment that is based on an actual instance of disadvantageous treatment. The basis for the thought experiment is the fact that women tend to receive worse care than men when they suffer heart attacks. This fact can be cited as a manifestation of a sexist bias in medicine (cf., e.g., Schiebinger 1999, ch. 6). However, we could stipulate an alternative history of medicine during which there was a time when the best available diagnostics and general medical knowledge—generated by the best, non-biased research—just would not have suggested to the best doctors that female heart attacks were on some level the same kinds of events that caused men to experience excruciating chest pain etc. We would have to assume that the doctors’ resulting ignorance would not be attributable to a less careful examination of female patients or some other epistemic neglect disadvantaging them. We might say that in this case, we should not count the misdiagnosis and mistreatment of female patients—though in effect systematic and disadvantageous—as an instance of sexism, because it was just the unfortunate outcome of a general lack of knowledge. In this sense, the apparent resulting discrimination would have been “contingent.” Whether we would decide to still call it “discrimination” or not, source-oriented considerations nevertheless would suggest that the respective disadvantageous treatment does at least not single out women in the right way to justify labelling it “sexist” discrimination—as there is no lower-level disadvantaging of women qua women to be identified. In this way, source-oriented considerations rule out “contingent discrimimation.” From this perspective, the unfortunate disadvantaging of members of certain species that comes about by fully honest, non-systematic mistakes should likewise not be labelled “speciesist.”

So, when the wide and normative notion is used based on source-oriented considerations, the charge of speciesism always comes with the charge that there is something underlying an “indirectly” speciesist view, such that on some level, there is direct influence of the criterion “species membership.” The charge of speciesism could then amount to a charge of intellectual dishonesty—when it seems “that membership in the human species has been the real criterion” for moral consideration all along (Waldau 1998, 321). It could also amount to the charge of a speciesist disposition, i.e., a tendency to adopt unjustified positions or false beliefs that inform views about how members of certain species may be treated. Finally, it could be an accusation of culpable ignorance or the blameworthy adherence to an unjustified moral view (depending on the evaluation of what chances the speciesist might have had to correct for her speciesist bias).Footnote 4

2.4 Pragmatic Requirements

So far, we have seen the differences in a wide and a narrow as well as in a normative and a descriptive concept of speciesism. We have worked through some of the ways in which the two sets of alternatives interrelate, and explored the commitments regarding the role of species membership that a wide and normative concept of speciesism might involve. Before we can see the problems that arise from the disagreement about the normative content of speciesism, we need to take stock of the pragmatic functions fulfilled by different (normative and descriptive) speciesism concepts:

  1. i.

    We have seen earlier that there has been demand for speciesism as a descriptive concept that picks out moral positions or ethical theories and serves their (mere) classification. Depending on whether speciesism is conceptualized in a wide or narrow way, this concept picks out positions that either assign moral relevance directly to species membership or arrive at respective conclusions about the group of morally considerable beings via some empirical assumption. In any case, positions thus classified can still be considered open to justification or refutation. Providing orientation within that discussion is all the descriptive concept does.

  2. ii.

    This distinguishes the descriptive concept from the one that is normative in the moral sense. As a moral concept, speciesism picks out unjustified moral views and morally wrong behavior and is applied to convey criticism, to condemn or denounce the view or behavior in question as wrong, unjustified, or blameworthy.

  3. iii.

    Speciesism can also work as a normative epistemic concept. In this sense, the term ‘speciesism’ refers to all kinds of (non-moral) prejudice against members of certain species. We have just seen how this consideration can serve to ground the wide moral notion of speciesism in the narrow (epistemic) notion.

  4. iv.

    Concepts like sexism and racism are supposed to pick out cognitive biases, certain behavioral dispositions or their social effects. Accordingly, speciesism might also be useful as a diagnostic concept that serves the identification of these phenomena for further investigation or explanation.

There is obviously a profound difference between concepts that fulfill function i. on the one hand and concepts that fulfill functions ii. and iii. on the other (we will get back to function iv. in the conclusion). Both, classification (i) and condemnation (ii & iii) serve legitimate purposes, but one and the same term cannot be used to for both. In the next section, I will detail why the term ‘speciesism’ should be reserved for condemnation rather than mere classification, i.e., why it should be used to express the normative and not the descriptive concept.

3 The Conceptual Predicament

From Carl Cohen’s explicit avowal of speciesism in his defense of animal experimentation (“I am a speciesist” Cohen 1986, 867)Footnote 5 to Shelly Kagan’s recent provocative question, “What’s wrong with speciesism?” (Kagan 2016), philosophers have repeatedly questioned or rejected the normative meaning of the term ‘speciesism’ itself. It usually goes unnoticed that this terminological choice comes with severe costs to our ethical vocabulary, conceptual scheme, and, therefore, our moral reasoning.

3.1 What Is Wrong with Asking for a Justification

The question “What is wrong with speciesism?” can be given two different readings, an explicative and a normatively challenging one. The former asks for an identification of the wrong-making features of speciesist beliefs or behavior. The latter questions the wrongness of such beliefs and behavior. On the explicative reading, the question could be paraphrased as “What is the thing that is wrong with speciesism?” On the normatively challenging reading, it would have to be paraphrased as “Is there really anything wrong with speciesism?” It is this version of the question that will be shown to be problematic here. In asking “What is wrong with speciesism?” in a normatively challenging fashion, one necessarily commits to a non-normative reading of the term ‘speciesism.’ It is only possible to meaningfully consider the justifiability of speciesism if the term does not, by definition, denote something that is unjustified (cf., e.g., Kagan 2016, 2; Chappell 1997, 96). Given that the conversation one appears to join by speaking of ‘speciesism’ was originally built around the normative concept, one effectively changes the subject by switching to the descriptive one. Now, detrimental effects on the debate as a whole (due to confusion) might be avoided or mitigated by being explicit about the change of topic. But there is a more serious downside to the choice of a purely descriptive notion of speciesism. Committing to the descriptive concept essentially means to waive the evaluative functions fulfilled by the alternative, normative concept (see 2.4.). By opting for the descriptive instead of the normative concept, one foregoes a means of issuing a normative judgment. Speculating on the justifiability of speciesism thereby results in a loss of normative expressiveness, which itself might even result in a loss of power of judgment. When it is given the purely descriptive reading, ‘speciesism’ is no longer available as an expression for what is conceptualized as an unjustified moral view or morally wrong behavior. This loss, I shall argue, is a loss to the greater share of the participants in the debate by far, and it is not outweighed by any benefit of gaining the term ‘speciesism’ as an expression for the descriptive concept.

3.2 Cui Bono?

Who benefits from the kind of change in the conceptual landscape that occurs through eliminating speciesism’s normative content? Or rather, the question should be: who doesn’t lose anything by relinquishing the term ‘speciesism’ as a means of expressing criticism? We might need to distinguish between gross and net losses at this point and say that a gross conceptual loss incurred by relinquishing the normative concept speciesism constitutes a pro tanto reason for keeping the normative concept. If that loss is not outweighed by other conceptual gains, it constitutes a net loss and an all-things-considered reason for sticking with the normative meaning of ‘speciesism’. Remarkably, the only ones who do not even incur a gross loss in giving up the normative term are those who Dunayer calls “old speciesists” (Dunayer 2004), i.e., those denying that members of other species are morally considerable at all (in contrast to “new speciesists”, who hold that members of certain species count for something, but for less than members of some privileged species). Those who believe that there is no level of disadvantageous consideration that could be unjustified—that members of other species cannot possibly be treated unfairly or wronged—have no use for a term denoting a bias against others’ interests or a kind of discrimination. Only those who hold this extreme view disagree with the widely shared assumption that the distinction between justified and unjustified treatment can apply to animals (cf. Horta and Albersmeier 2020, 7).

It is often overlooked that all others—defenders of the preferential treatment of one’s fellow species members and their critics alike—incur a gross loss by giving up the evaluative term and have at the very least a pro tanto reason to hold on to the normative notion. And usually this should translate into an all-things-considered reason for them. Not only do their views render its application conditions satisfiable, but most of them should also be able to agree that these conditions are actually frequently satisfied. Almost no one views actions affecting nonhuman animals as morally indifferent and at least some practices involving harm to animals are widely conceived of as seriously morally wrong. Therefore, discussing the justifiability of speciesism results in undermining a remarkable extent of common ground. Most participants in any serious debate about the moral standing of nonhuman animals should want to be able to address rationalizations of practices that are abusive of them as “speciesist.” To them, the question, “Is speciesism really unjustified?” should indeed sound comparably awkward to how an analogous question about racism or sexism would sound. To them, the unavailability of ‘speciesism’ as a term for passing judgment should constitute not only a gross loss but a net loss, as there is little to outweigh the conceptual deprivation which comes with changing the meaning of the word ‘speciesism’. The group of those who can agree on the very basic notion that there is moral wrongdoing, injustice toward or morally relevant harm done to nonhuman animals is large and heterogeneous. Many in this group hold views that involve some moral species hierarchy and they can therefore see an advantage in tying ‘specieism’ to a purely descriptive concept, as it would meant to eliminate a pejorative term that could be used for attacking the views they are defending themselves. However, as in the case of racism and sexism, there is still the option of having the relevant conversations about the justification of specific instances of disadvantageous treatment as conversations about “whether x is speciesist” rather than by classifying x as speciesist and then debating its justifiability. And so, it is not the case that there is an actual conceptual need here that outweighs the conceptual loss that is the abandonment of the normative concept of speciesism. We may see the failure to appreciate the usefulness of the term ‘speciesism’ as a manifestation of a speciesist bias itself and address it as a kind of “metadiscrimination,” i.e., unjustified disregard for a type of discrimination (Horta and Albersmeier 2020, 6; cf. Horta 2018).

The fact that there are convinced, self-declared “speciesists” (Cohen 1986) is not an indication that the larger linguistic community should accept a descriptive definition of speciesism any more than the fact that there are self-declared racists should push anyone to accept a non-normative notion of racism. Anyone who acknowledges that there can be unjustified disadvantageous treatment of people viewed as belonging to a group identified by “racial” features can simply refuse to use the descriptive notion of racism—thereby ruling out the normatively challenging version of the question “What is wrong about racism?” as conceptually confused. Almost everyone has good reason to treat speciesism the same way. Attempts to redefine speciesism as something that is not wrong by definition are epistemically and morally harmful, all things considered, for the same reasons attempts to redefine racism and sexism in this way are. In each case, descriptive definitions deprive us of a way to call a moral (and epistemic) problem by a name.

3.3 Differentiating Speciesism from Speciescentrism

Using ‘speciesism’ in a descriptive sense inevitably means to incur a terminological loss. Gaining it as a descriptive term for debatable ethical positions does nothing to remedy this situation. Because there is no adequate replacement for the lost normative term at our disposal, its loss effectively results in a conceptual deprivation. Clearly, two distinct types of concepts are needed for addressing differential treatment along species lines in a normative and a descriptive way. The term ‘speciesism’ should be reserved for the former way. A close terminological alternative for the latter would be a slight variation on the term ‘speciesism.’ We should distinguish from the normative concept speciesism (modeled on racism and sexism) the descriptive concept speciescentrism (modeled on anthropocentrism, ecocentrism, and the like):

speciescentrism: a position that assigns moral relevance to the classification of an individual as a member of a particular biological species.

This concept is defined in a deliberately general wayFootnote 6 and remains neutral on the question which concept of a biological species should be cited to classify individuals. Section 4 will show that ‘speciescentrism’ is the preferable terminological option for this descriptive concept. The remainder of this section will further clarify the concept’s scope.

According to the above definition, speciescentrism might encompass use of species membership as a criterion for moral considerability and significance. The distinction between the two—stemming from the abortion debate—has been drawn by many authors when discussing different kinds of what has so far been called “speciesism.” Whereas “considerability” concerns the question whether an individual gets to be taken into account at all, “significance” relates to the proportionate weight the individual’s interests are to be given with respect to others’ (Goodpaster 1978, 311).

The very notion of moral significance can in fact seem problematic. Talk of “moral significance” treats the importance an individual’s interests ought to be assigned as something that is fixed across contexts by the individual’s properties. It might be more adequate to consider the question who is morally considerable the only one that is answerable on a general level—and the weight of different interests as something to be determined in a context-sensitive, case-specific way (cf. Horta 2017).

Membership in a certain species can be suggested as a a) necessary, b) necessary and sufficient or c) sufficient condition for being morally considerable. The according claims would be that a) only (but possibly not all) members of the favored species are to be considered, b) all and only species members or c) all (but not necessarily only) species members ought to be taken into account. Proponents of the latter view may additionally use species membership as a criterion for moral significance, but criteria for moral significance need not be embraced as criteria for considerability. In fact, the probably most widespread form of speciescentrism is the one combined with sentientism about moral considerability: many people seem to favor sentience as a necessary and sufficient condition for moral considerability but invoke species membership as a criterion of moral significance. They believe that while any sentient being is owed some consideration, humans deserve most. Accordingly, most authors who have dealt with what has so far been called “speciesism” have been concerned with this sort of position.

Speciescentrism pertains not only to the general overarching question of moral consideration, but also relates to specific questions in political philosophy and applied ethics as well. In the debate about moral consideration, ‘moral significance’ can be said to refer to universal, unconditioned rules of precedence—in contrast to context-specific rules of distributive justice. “Moral significance” concerns foremost the allocation of moral consideration itself: it concerns how much we ought to care morally, in general, about certain individuals compared to others. On the present account, proposing the species criterion as a part of a context-specific rule for the allocation of some resource would also be captured by the term ‘speciescentrism.’ So, this term covers a variety of ways of invoking species membership in claims regarding moral consideration. However, we must distinguish speciescentrism from both (1) views on the metaethical relevance of species membership and (2) claims about the species-specificity of morally relevant goods and needs.

(1) Speciescentrism is about the assignment of moral relevance to species membership. In contrast, what might be termed “metaethical speciescentrism” is concerned with the relevance of species membership for moral epistemology, the metaphysics of morality or moral agency. Epistemological speciescentrism, for instance, is exemplified by the claim that there is a human and therefore species-specific point of view from which we necessarily understand and evaluate (morally) what we encounter, i.e., that we cannot escape “perspectival anthropocentrism” (Ferré 1994, 72; cf. Hayward 1997, 56). The claim that morality itself is anthropogenic (Krebs 1999, 23) would be a different type of metaethical speciescentrism, regarding the metaphysics of morality. Such claims about some kind of human-relativity of moral values are to be distinguished from, but can underlie arguments about moral considerability, for example in contractarian-like defenses of speciescentrism (cf., e.g., Goldman 2001) or in something like Bernard Williams’s attack on the normative notion of speciesism, in which he defends primary concern for human beings on the grounds that “[o]ur arguments have to be grounded in a human point of view” (Williams 2011, 131; cf. Grau 2016). These types of arguments rely on or are at least closely connected to the third kind of metaethical speciescentric claim, i.e., the empirical assumption that the capacity for moral agency is an exclusively human one (for criticisms of this view, cf., e.g., Sapontzis 1987, Chapter 3; Bekoff and Pierce 2009). So, metaethical speciescentrism may be the basis for speciescentrism in normative ethics, but the two should be carefully distinguished.

(2) Species membership is also often appraised as a crucial guide to understanding the practical implications of taking an individual into consideration, because species-specific characteristics are thought to determine individuals’ needs. For example, Martha Nussbaum suggests appealing to a “species-norm” in order to determine what the basic capabilities of individuals are (Nussbaum 2007, 365). In a similar vein, Mary Midgley highlights the relevance of knowing an individual’s species for identifying their needs (Midgley 1983, 105). These ways of referring to species membership when specifying the content of some right are also sometimes attacked as speciesist (e.g., Timmerman 2018, 696–7). In some sense, the species-norm account seems to assign “moral relevance” to species membership. However, this reading of ‘moral relevance’ would overstretch the concept intended here, which concerns the relevance of some feature for the moral considerability of its bearer. Empirical hypotheses about how the wellbeing of some individual already acknowledged as having a claim to consideration is affected by species-specific factors are a different matter. Such claims of species-specificity should not be covered by the term ‘speciescentrism.’

4 Some Objections Answered

Now, let us consider some potential objections to the proposed terminological solution for the descriptive concept speciescentrism.

Redundancy I

One might object that this use of the term ‘speciescentrism’ will be redundant, since all relevant instances of speciescentrism are actually instances of anthropocentrism. On this view, ‘anthropocentrism’ or ‘homocentrism’ suffice to replace ‘speciesism’ in all relevant contexts in which its negative normative meaning is unwanted.

First, this simply does not seem to be true: in ordinary morality as well as in philosophical discussions, anthropocentrism hardly is the only relevant type of speciescentrism. On the contrary, it is striking how much the proposed or actual consideration and treatment of individuals from different nonhuman species differs as well. The assignment of different weight to the interests of, say, chimpanzees or domestic dogs compared to brown rats should also be captured by the term ‘speciescentrism.’ Second, even if it were the case that there was no relevant speciescentrism other than anthropocentrism, the more general concept speciescentrism would still not be redundant. As a superordinate concept to anthropocentrism, it complements the existing conceptual scheme, increasing its coherence. It categorizes positions on moral relevance in a different mode than anthropocentrism. Third, speciescentrism might come in the form of what Bernstein and Jamieson call “indexical speciesism” (Bernstein 2004, 380; Jamieson 2008, 109) or “co-speciesism” (Bernstein 2015, 12), i.e., the view that species membership is morally relevant insofar as members of any species have obligations only toward their fellow species members. In this case, species membership is actually prior to the concept of a particular species. On such a view, being a member of the species Homo sapiens is morally relevant by virtue of instantiating membership in some biological species.

If anything, speciescentrism might not yet be a general enough concept to capture some of the ways in which biological taxonomy is involved in our ethical thinking about individuals. It surely is needed to conceptualize some ways in which the distinction between members of the species Homo sapiens and members of other species has been made morally relevant. But it does not clearly bring out how individuals are often grouped together based on appeals to other taxonomic ranks—as when claims are being made about the considerability of vertebrates (subphylum) or great apes (family). While we certainly can think of some of these categorizations as being about membership in a group of species (cf. Horta and Albersmeier 2020, 3), there certainly are instances of such taxocentrismFootnote 7 (appeals to some biological taxon) that are not adequately described as species-centered. The concept taxocentrism would in fact deserve a more comprehensive treatment than can be given here. For now, we should at least note that speciescentrism is an instance of taxocentrism as anthropocentrism is an instance of speciescentrism.

Redundancy II

It might also be claimed that since there are alternatives to ‘speciescentrism’ already in use (see section 2.1), there is no need for yet another term.

As we have seen, these alternatives tend to be built around cognates of anthropocentric speciesism, not a concept that centers on species instead of the species Homo sapiens. If one wanted to go with another term (that did not include the restriction to anthropocentrism), though, the most important message to take away from the present paper would be that there is, in any event, a need to respect the normative content of speciesism and to carefully and clearly introduce any concept one wishes to use for purposes of mere classification in its place. However, ‘speciesism’ seems to have been an irresistible terminological option for many who wanted to address reliance on species membership for the allocation of moral consideration. I have argued for a term that is close to the original in part because of the original’s continued appeal.

Polysemy

It might be argued that ‘speciescentrism’ is at least as well suited to denote the view that biological species themselves are morally considerable as it is to denote the kind of view we have carved out here. So there is a risk of polysemy and terminological confusion.

The idea that biological species matter morally is a kind of holism, whereas speciescentrism as portrayed here concerns the relevance of the feature species membership for the consideration owed to an individual being. One way of dealing with the fact that the word ‘speciescentrism’ seems to lend itself easily to signifying each of these very different views might be to distinguish speciescentrism from species-centered holism.

False emphasis

Another complaint might be that by grouping together positions that assign moral relevance directly to species membership and positions that just include empirical assumptions about how certain characteristics correlate with species membership, the wide notion of speciescentrism captures too much (viz. merely contingent speciescentrism) and puts emphasis on contingent factors. This is a complaint in the spirit of Singer’s claim that appeals to (supposedly) factual differences are “very different” from “speciesism.” It calls for a defense of a concept speciescentrism that is both descriptively centered on the attribute species membership and wide, i.e., includes positions that assign moral relevance to other attributes (such as rationality).

The problem seems to dissolve once we look at how the existing conceptual framework in normative and applied ethics works: anthropocentrism, ecocentrism or biocentrism are all characterizations highlighting the extension of morally considerable as construed by each type of position. All put emphasis on conclusions about the group of morally considerable entities—conclusions which will anyhow always be arrived at based on substantial moral and substantial empirical claims. Speciescentrism is analogous to other items in the pertinent conceptual scheme precisely because it highlights conclusions about who is to be morally considered rather than a specific normative premise. This is just what the relevant family of concepts is supposed to do.

Undue valorization

One might worry that translating ‘speciesism’ into a non-normative term gives “speciesists” too much credit and upgrades a biased, unjustified position to a respectable and theoretically important one. Overt attempts to neutralize the very concepts of racism and sexism would seem suspicious. Why seek to neutralize speciesism and make it seem even more acceptable?

In response to this worry, we need to face the state of the respective debates.Footnote 8 We have to acknowledge that, as a matter of fact, the discussion about the justifiability of the preferential moral consideration of humans on a general level is ongoing. The same is not true to the same extent of the assignment of moral relevance (in the sense that links it to moral considerability) to sex and “racial” features. But if the assignment of moral relevance to species membership is as untenable as the former, we need (a) an unequivocally evaluative term to condemn the unjustified treatment of species membership as a morally relevant feature, and (b) the willingness to rationally demonstrate that such treatment is unjustified, i.e., to openly discuss the justifiability of speciescentrism. Precisely because there is a need for a clear, unambiguously normative term, we cannot use that same term for discussions about justifiability. The present suggestion is not that ‘speciesism’ is to be replaced across the board. It is rather that wherever justifiability of differential treatment is seriously under discussion, this differential treatment must not be condemned at the outset as “speciesist.”

Euphemism treadmill

Another concern might be that marking the difference in meaning between the original normative concept and its descriptive substitute by a linguistic alteration is only a temporary solution that equals stepping onto the “euphemism treadmill” (Pinker 2002, 212): soon, the new term will have gained all the negative connotations that made the old one unfit for its intended role as a piece of philosophical terminology.

But critics of speciescentrism need not worry about negative connotations at all. They just need a classificatory term that does not condemn its referent by definition. Defenders of speciescentrism on the other hand should view it as one of their tasks to reduce negative connotations by offering powerful arguments for their position.

5 Conclusion

Disagreement about the semantics of ‘speciesism’ has led to a theoretical and terminological predicament. Authors who opt for the descriptive version of ‘speciesism’ relinquish a valuable means to express due criticism, whereas those who keep with the original normative version cannot consistently use it to merely index ethical positions that share a commitment to the moral relevance of biological taxonomy. The simultaneous use of both versions is detrimental to clarity at best and deceptive or nonsensical at worst. It obstructs the view of the wide range of phenomena theorists with differing views on the criteria for moral considerability can actually agree to classify as unjust toward nonhumans. It additionally impedes the discussion by inviting rhetorical maneuvers that systematically confuse justificatory and semantic issues. The question whether speciesism really is a prejudice only bereaves the discourse of a critical conceptual resource. Therefore, almost irrespective of where one stands on the moral relevance or irrelevance of species membership, there is reason to recognize the need for speciesism as an inherently normative concept as well as the need for a purely descriptive substitute for ethical debates. I have suggested that speciescentrism ought to fulfill the latter function, complementing the established conceptual scheme for positions on requirements for being morally considerable.

Speciescentrism may appear to be the more relevant concept in academic contexts. Due to its condemning nature, it seems that the concept speciesism can only be applied in concluding judgments on positions and behaviors (cf. Jaquet 2019, 455), if that. However, this impression is mistaken. There are important fields of applicability for the normative concept speciesism in science—though even scientists already working on speciesism do not always fully embrace this fact. For instance, in a recent study on “the psychology of speciesism,” the authors hasten to point out that they are relying on ‘speciesism’ in a descriptive reading (Caviola et al. 2019, 1011). But soon thereafter speciesism is equated with a sort of “prejudice” (1012). prejudice—just as Bias—is a normative epistemic concept and this need not be a problem. On the contrary: prejudice and bias are legitimate objects of research. Interest in phenomena like racism, sexism, and speciesism is not adequately characterized as interest in an object matter that is describable in purely descriptive terms. The suspicion of a lack of justification, systematic error or detrimental social effect is constitutive of what drives the respective research programs. Thus, the normative concept speciesism qualifies as a fruitful diagnostic concept (iv.) to rely on in urgently required research into the manifestations and underlying mechanisms of discrimination against members of certain species.