Acta Analytica

, 23:337

Species as a Relationship



DOI: 10.1007/s12136-008-0029-x

Cite this article as:
Tanner, J. Acta Anal (2008) 23: 337. doi:10.1007/s12136-008-0029-x


The fact that humans have a special relationship to each other insofar as they belong in the same species is often taken to be a morally relevant difference between humans and other animals, one which justifies a greater moral status for all humans, regardless of their individual capacities. I give some reasons why this kind of relationship is not an appropriate ground for differential treatment of humans and nonhumans. I then argue that even if relationships do matter morally species membership cannot justify a difference in moral status. This has important implications because it removes one barrier to giving animals greater moral status.


AnimalsMoral statusReciprocityRelationshipSpecies

1 Introduction

Many argue that members of our species deserve special consideration that other species do not because of the special relationship we stand in to members of our species, a relationship we do not have to members of other species.1 Species membership can be viewed as a relationship insofar as it can be seen as an agent-dependent reason for assessing a being’s moral status and/or how to treat it. Species can, however, also be viewed as a property that has moral relevance in its own right (for example, that humans are the kind of being that has moral status because they are rational or something similar).2 I have discussed species as a property elsewhere (2005, 2006). The purpose of this paper is to discuss species as a relationship.

I will start by outlining why relationships are thought to be the source of moral concern in section I and give some examples. In Section II, I will offer some criticisms. First, if relationships are viewed as reciprocal, the reciprocity requirement does not apply between normal and marginal humans. Second, not all humans have the ability care and/or form relationships of the type required. Third, marginal humans do not have strong relationships. Fourth, we do have relationships with animals that they can reciprocate. Fifth, relationships should not ground ethics. Sixth, if relationships are a source of moral concern, this justifies racism and sexism. Finally, species is not the right kind of relationship to ground moral concern. I will argue that merely belonging to the same group as those with moral status is not a morally relevant relationship. Such a relationship cannot therefore be a justified basis for extending moral consideration to marginal humans or withholding it from animals.

1.1 Relationships as the Source of Moral Concern

Numerous philosophers have argued that relationships are the source or moral concern. Benson, for instance, says:

to think of oneself as human is not to think of the biological classification…but…a point in a network of overlapping relations…with other individuals…concern for other people begins with natural affection towards kin, friends…and is extended by recognizing other human beings as potential reciprocating objects of the same affections… The…attempt to eliminate partial affections…may remove the source of all affections. (1978, pp. 536–7)

The idea is that our relationship to other humans is what separates animals from marginal humans (those humans who are not rational–this includes those who are pre, post and non rational). Narveson argues that marginal humans (unlike animals) are:

invariably members of families, or…other groupings…the object of love and interest…[of]…members of the moral community. (1983, p. 58)

Those who think relationships are what matter in morality tend to think we do not have to give equal consideration to all, but that we should prioritise those close to us: family, friends, neighbours, etc. Some argue that animals cannot enter relationships at all because they are incapable of reciprocating in the right way. Others argue that our relations to animals give rise to obligations to animals, but only to those we live with, and even then the obligations are minimal. I shall outline a couple of relational views.

1.1.1 Nel Noddings’s Ethics of Care and Weak Reciprocity

Nel Noddings argues that only those individuals we care about, and who can reciprocate that care, can have moral claims on us.3 Noddings thinks moral status is a function of the emotional relationship ‘caring’ (1984). Receptive to the feelings and needs of the ‘cared-for’, the ‘one-caring’ is motivated to meet those needs. But “both parties contribute to the relation; my caring must be somehow completed in the other if the relation is to be described as caring” (1984, p. 4). Noddings does not have the contractual sense of reciprocity in mind; the cared-for is not expected to care in return and does not need to be one-caring (1984, pp. 69–71). Nor does she have the stronger sense of reciprocity in mind, that there is a correlation between duties and rights; only those capable of reciprocal relations with free, rational beings can have rights (see for example McCloskey 1979, 1987; Narveson 1987, p. 42; Rawls 1963, p. 72; Watson 1979, p. 99). For Noddings the cared-for need not reciprocate fully, but they are called upon to complete the caring relationship (if they do not the relationship is incomplete). The reciprocity Noddings has in mind can be characterised as ‘weak reciprocity’. The cared-for must recognise and respond to the one-caring (1984, p. 72). They may share “aspirations, appraisals, and accomplishments” with the one-caring (1984, p. 72). The cared-for reciprocates by “receiving the efforts of one-caring, and this receiving may be accomplished by a disclosure of his own subjective experience in direct response to the one-caring or by a happy and vigorous pursuit of his own projects” (1984, pp. 150–1). The caring relationship must be completed by the cared-for. Ethical caring depends “completely…upon both our past experience in natural caring and our conscious choice” (1984, p. 157). We can, therefore, decide whom to enter into a caring relationship with.

Animals cannot be the one-caring nor, Noddings argues, can they be cared-for. Despite animals’ being unable to enter into a caring relationship, we have obligations not to cause them pain (1984, p. 150). She also thinks we have some extra obligations to our pets, obligations we have acquired. But, Noddings argues, animals cannot complete the caring relationship in the correct way. A cat is:

a responsive cared-for, but clearly her responsiveness is restricted: she responds directly to my affection with a sort of feline affection–purring, rubbing, nibbling. But she has no projects to pursue. There is no intellectual or spiritual growth for me to nurture. (1984, p. 156)

To rats, however, we have no obligations. We cannot enter into a reciprocal relationship with rats–we cannot receive rats as one-caring. There are significant differences in individual humans’ capacity to respond to animals emotionally so there is no universal obligation to care for them. Any obligations we have to animals (aside from that not to cause them pain) we have as a result of forming relationships with them. Because animals cannot reciprocate in the appropriate way, the relationships we have with them are never as strong and are incomplete.

1.1.2 Becker, Virtue and Social Distance

Human interests, Becker argues, should be prioritised over animals’ because:

There are certain traits of character…people ought to have–traits constitutive of moral excellence or virtue. Some of these traits order preferences by “social distance”–that is, give priority to the interests of those “closer” to us…Animals are typically “farther away” from us than humans. Thus…people ought (typically) to give priority to the interests of their own species. (1983, p. 225)

Becker argues that because of their social distance animals’ interests will be less important than those of humans who are usually closer to us. Becker acknowledges his arguments do not show that humans are morally superior to animals. He does not deny we should give consideration to animals’ interests, or that their interests can sometimes override humans’.

1.2 Criticisms of Relationships as the Source of Moral Concern

1.2.1 Reciprocity as Fairness

Sapontzis mounts an argument against the strong understanding of the reciprocity requirement. On the strong understanding there is a correlation between duties and rights; only those capable of performing duties are owed rights (Sapontzis 1985, 251). Thus, on this interpretation animals do not have rights. But, he argues, our common practice suggests that rights do not depend on the ability to recognize and act on duties:

Infants and the severely retarded, brain damaged, and senile are not regarded as resources for fulfilling the interests of normal humans (even though they could provide outstanding material for medical research). In spite of their inability to recognize and act on duties, their interests are protected by moral rights. (1985, p. 252)

Sapontzis argues that the reciprocity argument has intuitive appeal because it appears to be fair (1985, p. 252). The idea is that it is only fair that if I give you moral consideration, you do likewise.

But this type of argument has an “Achilles’ heel”; it cannot provide a basis for “obligations of the powerful to the powerless” (1985, 252). If someone is not in a position to affect you there is no reason to give that individual moral consideration. If reciprocity is all that is required for moral status, only the strong have moral status. This is not how we usually understand rights. It is a boundary condition of an acceptable moral theory that it protects the weak. When the strong are dealing with the weak, we need to inhibit their power or give the weak additional power against the strong. If the reciprocity requirement does not apply between strong and weak, it does not apply between normal and marginal humans. Thus, the reciprocity requirement does not distinguish animals from all humans; some marginal humans are no more able to reciprocate than animals.

The same kind of point can be made with regard to Noddings’ ‘weak reciprocity’. Those who are unable to fulfil the requirements of the caring relationship are those who stand most in need of being cared for. This is, for an ethic of care, decidedly uncaring. This is a serious failing for a theory that ostensibly has caring at its centre. Remedying that failing would mean accepting that the cared for need not reciprocate (even weakly). If the caring relationship does not require marginal humans to reciprocate, it cannot demand that animals reciprocate. Thus, as with the stronger understanding of reciprocity, Noddings’ ‘weak reciprocity’ view fails to distinguish animals from all humans.

1.2.2 The Ability to Care/the Ability to Form Relationships

What if a relationship only works one way? For example, a mother cares for her son, but he cares nothing at all for her. On Noddings’s account, the mother has no moral duties to him (the relationship is incomplete). But the mother would presumably reject this. The point carries across to other marginal humans and animals with which we may have very loving one-way relations in which they do not fully reciprocate.

It is a common and fundamental principle of most theories, and an intuition of most moral agents, that we have moral obligations to all humans. For theories that rely on relationships or caring to be consistent with this, they must say we have the potential, at least, to care for/form a relationship of the right kind with all humans. It is not obvious all humans are capable of such caring or forming such relationships. There may be some who are incapable of caring for others or forming the appropriate kind of relationship with them. The perpetrators of rape, murder and torture do not care for their victims nor do they enter into the kind of relationship that Becker et al. have in mind. On such views rapists and torturers are without moral obligations to their victims because they do not care for them or have a relationship with them. Given that this principle is unable to satisfactorily account for the moral status of moral agents, we should be very wary of accepting it.

1.2.3 Marginal Humans Do Not Have Strong Relationships

Some marginal humans are excluded by the relationships/reciprocity requirement because they are not capable of having the type of reciprocal relationship required (see McMahan 2005, p. 368). Some marginal humans do no have the cognitive capacities to form relationships and never will. For example, many autistic children are “mindblind” (i.e. they are not aware of other minds) (Baron-Cohen 1995). They cannot form relationships because they cannot recognise other individuals in the way necessary to form a relationship with them. If they are not aware of other people existing, they cannot form the kind of relationships that are thought to be morally relevant.

Francis and Norman answer this objection thus:

Babies are the objects of…people’s emotional attachments–most usually…parents…To take the life of a young baby…would…cause extreme emotional distress [to them]…This is an indirect reason for protecting the lives and interests of human babies, but an extremely powerful one. (1978, p. 510)

But the problem with this reply is that it makes the moral status of marginal humans indirect. Scott Wilson defines the difference between direct and indirect moral status thus:

a being belongs to the class of beings with direct moral status if and only if we can have at least some moral reasons to treat that being in a certain way due to the nature or intrinsic properties of the being itself. A being belongs to the class of beings with indirect moral status if and only if all the moral reasons we have to treat that being in a certain way arise due to the relational properties of that being. (2001, p. 137)

Direct moral status is contingent upon the intrinsic properties of the being itself. Indirect moral status is, however, entirely contingent upon the subjective relationships a being happens to have–relationships that could quite easily have been different or not existed at all. This type of contingent moral status is not the type of status most people have in mind when they think of marginal humans, especially babies and young infants. Most of us, philosophers and non-philosophers alike, want to say that a baby has moral status regardless of the relationships it happens to find itself in. A lack of such relationships would not, for most, be a reason to deny them moral status.

But even those who would be satisfied with only indirect moral status for marginal humans should not be satisfied with this argument. Some marginal humans, such as the severely mentally retarded, spend their whole life in institutions with little or no contact with their families. Staff at these institutions have some kind of relationship with them, but it is not the kind of reciprocal relationship those who oppose the argument from marginal cases have in mind.4 Some marginal humans are unable to reciprocate. That marginal humans are not capable of reciprocating is not a problem for those who place the emphasis on having a relationship with marginal humans because relationships can be one sided (but animals can also be passive receivers). Reciprocity does require both parties to reciprocate. If Benson et al. are to be consistent, they will have to admit some marginal humans do not have moral status of the kind in question, or if they do, some animals do, too.

1.2.4 We Do Have Relationships With Animals That They Can Reciprocate

Those who oppose the argument from marginal cases argue that our relationships with animals are of a different kind or degree from those with marginal humans (e.g. Becker 1983, p. 233; Francis and Norman 1978, p. 508). But, arguments about reciprocity and relationships will succeed only if it can be demonstrated that all humans can but no animals can have relationships and/or reciprocate. But animals do have reciprocal relationships with one another and humans.5 Many people have a stronger relationship to their pet then to strange humans.6 Many prefer animal company (usually a dog or cat), and others devote their whole lives to looking after animals. Pets usually reciprocate affection. If this is so then the supposed difference between marginal humans and animals vanishes.

1.2.5 Relationships Should Not Ground Ethics

Grounding ethics solely on relationships and/or the ability to reciprocate is problematic. It is commonly understood that the moral point of view is impartial.7 Impartiality is an important factor in any moral theory. Impartiality is defined by Regan as “treating similar cases similarly” (Regan 1983, p. 190). Regan considers this to be a requirement of formal justice.8 Similarly, Rawls builds impartiality into the veil of ignorance “[i]f a knowledge of particulars is allowed, then the outcome is biased by arbitrary contingencies” (Rawls 1999, p. 122). Our impartiality is modelled in our ignorance: if you do not know who you are, you have to have equal concern for the interests of everyone.

We need a strong reason to treat those who are alike/similar differently. Showing partiality is to treat someone who is not relevantly different differently due to your partiality for them. Such partiality is almost universally disapproved of.9

If we do not act impartially, we risk taking things that are not relevant into consideration. For example, in awarding the position of surgeon I should not choose my relative (who is not qualified) because I am related to them, I should choose a qualified surgeon. We should not allow our relationships to cloud our judgments about moral status. Personal relationships and affections may sometimes be morally relevant, but we should, for the most part, make our judgements about moral status on impartial grounds.

Noddings argues that marginal humans should be given the same kind of moral consideration as normal humans because “[a]n ethic that forces us to classify human infants with rats and pigs is unsettling. We feel intuitively that something must be wrong with it” (1984, p. 151). In other words she argues that we ought to be partial. But marginal humans, like animals are unable to reciprocate in the appropriate ways. Infants will be able to reciprocate in the future, and those who have lost the ability to reciprocate once had it. But there are some humans who are not, never have been and never will be able to reciprocate. Yet Noddings thinks these humans are morally considerable. If these marginal humans are morally considerable, there is no non-arbitrary reason for excluding relevantly similar animals. Noddings has an answer to this:

Locating our primary obligation in the domain of human life is a logical outgrowth of the fact that ethicality is defined in the human domain–that the moral attitude would not exist or be recognized without human affection and rational reflection upon or assessment of that affection. It is not “speciesism” to respond differently to different species if the very form of response is species specific. (1984, p. 152)

But this does not overcome the problem. If ethics is defined by the caring relationship and animals are excluded because they cannot enter fully into this relationship, then marginal humans must be excluded too. If marginal humans are to be included so must animals.

Relational views are centred around the moral agent (in Noddings’s case the “one-caring”). It may be argued the fact it is centred round the moral agent does not necessarily show it is partial, if the moral agent is what grounds moral considerability. This may be true, but it does gives us reason to be suspicious as it means the moral agent and those they care about are the only ones that matter. Most moral agents, and most theories, think nepotism is wrong (unless there are unusual circumstances), that people should be treated equally. If they cannot satisfy the basic requirement for impartiality among humans, it is unlikely relational theories will be able to satisfy concerns about impartiality with regard to animals (given we have even stronger reasons for denying them moral status).

An alternative response is to argue that some kinds of partiality are justified. Becker thinks the inclination to reciprocate is an element of moral virtue:

the argument I have given is based…on the logic of various elements of moral character–namely the dispositions to reciprocate and to empathize... racism and sexism are not entailed by the logic of virtue…Quite the contrary…they come in part from a culpable failure to reciprocate (1983, p. 240)

This response comes from a virtue ethics point of view. But others are likely to make similar responses about appropriateness. For example Benson argues:

Partiality for our own species, and within it for much smaller groupings, is, like the universe, something we had better accept. That we care at all about the interests of strangers of our own species or animal of other species results from our extending to them by sympathy something of the concern that we feel spontaneously for those with whom we have closer connections. (Benson 1978, p. 536)

First, it is questionable whether prioritising human interests is a necessary part of virtue (Cargile 1983, p. 248). But more importantly these responses do not succeed in saying why species is relevant. If relationships are important, species is not obviously relevant; there is no reason to suppose the moral virtues would distinguish between species. We can reciprocate and empathize with animals. And animals have reciprocal relationships with one another.10

1.2.6 It Justifies Racism and Sexism

If we base moral status on relationships this appears to justify limiting moral concern to our relations, friends, race, sex or nationality.11 A sexist, for instance, judges people of a particular sex by a supposed norm for that group. For instance a sexist may insist women are on average less clever than men. When presented with a woman who not only out performs her “kind” (women) but also out performs all or even most men, the sexist will still insist this woman should be judged as other women are, despite her individual abilities. Similarly, if there is a man less clever than most, or even all women, the sexist will be committed to the ludicrous position that this man is cleverer than this woman despite the fact the woman has out performed the man. What is objectionable about most forms of racism/sexism is that the individual has been judged on the basis of an irrelevant feature (e.g. sex). If we are trying to determine someone’s intelligence we must test their intelligence not their group membership.12 If we want to know whether an individual has moral status we should look at their morally relevant properties (whatever they may be); species is irrelevant.

Midgley argues that racism is “an ill formed, spineless, impenetrably obscure concept, scarcely capable of doing its own work, and quite unfit to generate a family of descendants which can be useful elsewhere” (1983, p. 99). She defines racism as making decisions about an individual’s treatment on the basis of their race (1983, p. 99). On this definition reverse discrimination is, she argues, racist (1983, pp. 99–100). This presents a problem for those who think reverse discrimination is justified (1983, p. 100). I object to racism but not to reverse discrimination. I think these views are consistent because of how I understand racism. My definition is as follows:

To base judgments about and/or treatment of an individual on their race where race is not relevant.

This definition overcomes concerns about positive discrimination because in this case race is relevant. First, it is part of the job to provide a role model. Second, people from ethnic minorities are not competing on a level playing field. Minorities are not treated fairly by would-be employers because they harbour prejudices, some of which they may be unaware of.13 These arguments do not give a full answer to the problems raised by reverse discrimination; they do show a case can be made for it without being racist; whether it is acceptable will rely on whether race is relevant. Just taking someone’s race into account when reaching a decision is not racist: doing so when race is not relevant is.

Bigotry is not uncommon. Some think people of a particular sex, race, culture, sexuality or religion are inferior and as such do not care for them (at least not in the necessary way). Noddings may argue that the bigot at least has potential to care for such people (in the necessary way). This is, at best, psychologically implausible.

1.2.7 Species Is Not The Right Kind of Relationship

Samuel Scheffler makes the following comment in a paper about relationships and responsibilities:

in cases…in which two people are both members of some group but have not themselves interacted in any way, it may be denied that the people do in fact have any special responsibilities to each other. (1997, p. 190)

I think that this point can be taken further. It is, I think, arguable that two people who are both members of a group but have not interacted do not have a relationship with one another. For instance, I am a member of a gym, but I do not have a relationship with someone who also uses they gym (at different times from me) but whom I have never met.

The point I want to make is that there are two types of relationship. The first is having similar properties to another individual such as belonging to the same family or nationality. The second is to have interacted with someone. If relationships have a role to play in morality, it must be the second kind that matters. This is something that those who value relationships would agree to. Those who see relationships as morally relevant all outline the interactions (or potential interactions) with other humans as what makes it relevant.

To be similar to something is not significant in the right way, or if it is then we have many more relationships that we must give moral weight to. For example we live on planet earth along with grass so we have a relationship, of sorts, with grass. But this commonality is not enough to secure moral status for grass. We do not value relationships with cousins we have never met the same way we value relationships with friends or partners. The same goes for membership of the group homo sapiens. The fact that I am a human and that you are also a human does not constitute a morally relevant relationship between us. If relationships are what ground morality then there is no valid reason for giving preferential treatment to humans because they are humans because we do not have a relationship with them solely in virtue of their being human. We can only favour those humans (and animals) that we have the right sort of relationships with, that we have interacted with. Similarly, we cannot give animals less consideration because they are not humans.

If we have not interacted with other humans we cannot be said to have a relationship with them; species is not a morally relevant relationship. This removes one obstacle to giving animals equal moral consideration.

2 Summary and Conclusion

I started by outlining why relationships are thought to be the source of moral concern in section I. In section II I offered some criticisms. First, I argued that if relationships are viewed as reciprocal, the reciprocity requirement does not apply between normal and marginal humans. Second, not all humans have the ability to care and/or form relationships of the necessary kind. Third, I argued that not all marginal humans have strong relationships with normal humans. Fourth, I argued we do have relationships with animals that they can reciprocate. Fifth, relationships, I argued, should not ground ethics. Sixth, I argued that if relationships are a source of moral concern this justifies racism and sexism. Finally, I argued that even if relationships are the source of moral status, merely belonging to the same group is not a morally relevant relationship. Such a relationship cannot therefore be a justified basis for extending moral consideration to marginal humans or withholding it from animals. If species is not a morally relevant relationship one obstacle to giving animals equal moral consideration has been removed.


Benson 1978, pp. 536–7; Diamond 1978; Francis and Norman 1978, p. 518; Narveson 1977, p. 176.


I owe this point to an anonymous reviewer of this article.


Benson says something similar (1978, p. 536).


I have discussed the argument from marginal cases at greater length elsewhere (2005, 2006). In brief, the argument from marginal cases says that those capacities that are usually supposed to separate all humans from all animals (such as rationality or moral agency) do not include marginal humans (those humans who are not and/or will never be rational). Thus, consistency demands that either animals have as much moral status as marginal humans or marginal humans have as little as animals. Many arguments have been given to oppose this conclusion. I address some elsewhere (2005, 2006).


Even those who object to the argument from marginal cases admit people have relationships with animals: Fox 1986, p. 60; Francis and Norman 1978, p. 510.


Similar points are made by: Singer 1993, pp. 76–7; Midgley 1983, p. 109). Midgley also says the preference for our own species is neither constant, nor guaranteed, so it cannot be used to draw an absolute line between humans and animals (1983, p. 109).


Dombrowski 1997, p. 112; Regan 1983, pp. 190–1; Sapontzis 1985, pp. 255–6; Singer 1993, pp. 76–7.


Others who object to partiality include: Almeida 2004, p. 31; Dombrowski 1997, p. 23; VanDeVeer 1983, p. 148.


As Midgley points out, it is hard to find an impartial standpoint (Midgley 1983, p. 103). However, this does not mean that we cannot identify those standpoints that are partial.


Francis and Norman acknowledge this point (1978, p. 511).


Similar points are made by: Dombrowski 1997, p. 23; Singer 1993, pp. 76–7.


Rachels calls this “moral individualism”-individuals should be considered as individuals, not evaluated on the basis of characteristics of their group (1990, pp. 173–4, 194–7). Thus, “how an individual may be treated is determined not by considering his group memberships but by considering his own particular characteristics” (1989, p. 101). Others make similar points: McMahan 2005; Cavalieri 2001, p. 72.


Similarly, women’s qualifications are undervalued and under appraised. A woman’s CV that is identical to a man’s is rated lower (Harris and Narayan 1997, pp. 456–7).



I wish to thank an anonymous reviewer for this journal and G. K. Harrison for comments on earlier drafts of this paper.

Copyright information

© Springer Science+Business Media B.V. 2008