image/svg+xmlALF

The Moral Status of Non-Human Animals in the Social Contract - Possibilities of an Anti-Speciesist Contractarianism

Lara Biehl, 25.01.2024

Contractarian moral theories face challenges in granting direct moral status to non-human animals. The contracting parties in the original position are self-interested and rational, deriving moral principles to advance their own interests. As a result, non-human animals are typically given, at most, indirect moral status. Additionally, social contract theories seem to prioritize the principle of reciprocity, which requires mutual recognition of obligations. Since non-human animals cannot fulfill duties, they are not considered direct bearers of moral rights. However, in his paper "Contractarianism and Animal Rights" (1997), Mark Rowlands attempts to challenge these prejudices against contractarian moral theories. He argues that contractarian theories, which aim to provide a fair environment for the derivation of moral principles, should ensure impartiality and objectivity by disregarding morally arbitrary characteristics. Based on this premise, Rowlands suggests that contractarian theories can also defend the direct moral status of non-human animals.

Rowlands argues that the belief that animal rights cannot be derived through contractarianism is based on an inaccurate and incoherent interpretation of John Rawls's principle of intuitive equality. If the parties in the original position must abstract from all prejudices, then properties such as rationality, species membership, and moral agency should also be abstracted. As a consequence, once the hypothetical scenario of the original position is resolved, the parties would no longer know whether they could be a human or a non-human animal. Because the parties are rational and self-interested, they would then derive moral principles that ascribe moral rights to non-human animals.

There are a number of objections to Rowland's approach. The main problem identified by Rowland's opponents concerns the claim that the original position cannot be described in such a way that we can derive a direct moral obligation towards non-human animals. The objection is that there is a necessary connection between contractarianism and reciprocity. If reciprocity is necessarily related to contractarian moral reasoning, then it is not possible to protect non-human animals directly through contract-derived principles.

In the first part of the paper I will focus on the question of whether the principle of reciprocity follows from the structure of contractarian systems. Clarifying this question is helpful because it is the starting point for all contractarian theories that seek to accommodate the interests of non-human animals. In doing so, I will discuss counter-arguments by mainly David Svolba and Peter Carruthers, since they are well-known proponents of an "orthodox" contractarianism that excludes non-human animals from the scope of moral principles derived by the contract. As the paper progresses, it will become apparent that the main conflicts which arise between Rowlands and his opponents are based on "pre-contractarian" intuitions. Drawing on Tom Regan and the argument from species overlap, I will show that opponents of an anti-speciesist contractarianism rely on arbitrary concepts and terms in their description of the original position. I will then argue that a distinction between moral agents and moral patients can be successfully implemented in the original position. In the final part, I will address unresolved issues in Rowland's anti-speciesist contractarianism.

1. Basic concepts of contractarianism: John Rawls, Mark Rowlands, and the moral status of non-human animals.

1.1 John Rawls and the veil of ignorance.

In A "Theory of Justice", John Rawls sets out a contractarian framework to justify basic social and political principles of justice. According to Rawls, the principles of justice are the result of a decision-making process on which rational, self-interested and free agents would agree under fair conditions. These principles are identified through a hypothetical scenario called the "original position". The agents in the original position do not know what properties or social positions they will have outside this scenario. If fair principles are to emerge from the original position, the circumstances in which rational agents agree on them must also be fair. Rawls tries to ensure that all parties are equal through the concept of the veil of ignorance.1 If contractors know what characteristics and position they will have outside of the original position, they could use this knowledge to gain an advantage in the distribution of social goods.2 Since all agents are biased in some way and want to optimise their own interests based on their social and economic situation, they would be tempted to derive principles that are, for example, racist, classist or sexist. To ensure that this does not happen, and that decisions are based on justice, the contracting parties must, behind the veil of ignorance, abstract from any arbitrary individual characteristics and social positions that might give them an unfair advantage over others. Thus, contractors are forced to put themselves in the shoes of the best- and worst-off members of society and can be expected to formulate principles that benefit the worst-off.

Rawls set of morally arbitrary properties: {race, gender, social status, wealth, talents, degree of intelligence, etc.}

Therefore, agents know only those facts that are necessary to derive just principles. For example, they know the circumstances of justice, they know general facts about human society, they know how politics and social institutions work, and they have an understanding of economic theories, human psychology, and scientific concepts. They also have a "sense of justice".3

The principles of justice are established through Rawls' concept of "reflective equilibrium", which involves a hypothetical process of reflection. Rational individuals, guided by a sense of justice, make "considered judgments" that are based on reason and are free from obvious bias. While these judgments may initially be inconsistent with their own or others' judgments, they are still categorised under general principles in the original position. Inconsistencies in the different judgements lead to a re-examination and possible adjustment of the individual judgements to bring them into line with the general principles or vice versa. This process of reflection continues until a harmonious balance between considered judgments and general principles is achieved.4

1.2 The moral theory of contractarianism

The purpose of the social contract is to establish principles that ensure the fair distribution of primary goods and to legitimise certain political and social institutions. But contractarianism can also serve as a theoretical basis for deriving and justifying moral principles. Mark Rowlands, in contrast to Rawls, is a proponent of a contractarian moral theory. Rawls' reflections highlight the significance of the veil of ignorance in fostering impartiality and discouraging decisions based on morally irrelevant factors. Consequently, when examining the justification of a moral norm established behind the veil of ignorance, we can assert its validity based on the agreement reached by rational agents operating under fair and unbiased conditions.5 Contractarian frameworks offer a distinct advantage in their ability to establish impartiality and objectivity with relative ease, as previously mentioned. This is achieved by eliminating irrelevant properties from consideration. As the principles are derived through deduction, the process by which they are formulated becomes more straightforward and comprehensible.6

1.3 The moral standing of non-human animals in the social contract.

As already noted, the contracting parties must be rational individuals. Rawls' understanding of rationality is instrumental and self-interested. Agents derive fair principles deductively from their beliefs and interests.7 According to Rawls, the rationality of agents involves two capacities. On the one hand, they must have a conception of the good (a rational plan of life), according to which each agent tries to negotiate those principles which implement this conception as perfectly as possible and give them the largest possible share of social goods.8 The other capacity is a moral one: contracting parties need to have a sense of justice that helps them to understand and apply fair terms, and to be motivated to act on them.9 Beings endowed with this kind of rationality are called moral agents. Conversely, beings who have the capacity to experience suffering, who have interests and intuitions, but who lack the necessary level of rationality to make moral judgments and fulfil duties towards others, are not considered moral agents. Consequently, they cannot participate in the formation of the social contract. Since the social contract is established through a hypothetical decision-making process based on the rationality and self-interest of the agents involved, and since no non-human animal has the capacity to actively participate in the formulation of principles or to hypothetically consent to a resulting agreement, they are not accorded direct moral status.

Rowlands summarizes the argument that follows as the «argument from orthodox contractarianism»:

P1The parties forming the social contract are rational agents.
P2Rational agents are responsible for deriving moral principles from the original position.
P3The contract and the principles derived from it cover only rational agents.
P4Non-human animals are not rational agents.
P5The contract and the principles derived from it do not cover non-human animals.
P5The only beings who have direct moral status are those who participate in the social contract and are affected by the principles derived from it.
KNon-human animals do not have a direct moral status

The argument shows that, according to orthodox contractarianism, non-human animals have no direct moral value. This does not mean, however, that they would be completely unprotected, since contracting parties could derive indirect duties towards non-human animals. Namely, if the parties were concerned about the poor or unregulated treatment of non-human animals, they could lay down some principles of how they should not be treated. However, their moral status would depend not on their intrinsic value, but on the preferences of moral agents. Grounding indirect duties to non-human animals in contractarianism has not been promising in protecting their interests. As long as the economic or personal interests of moral agents are greater than their interest in protecting non-human animals, contractarianism still justifies killing animals for food, fashion and fun.11

1.4 Mark Rowland's anti-speciesist contractarianism

Rowlands believes that the orthodox contractarian argument is wrong. He argues that if the veil of ignorance is coherently applied, non-human animals must necessarily be included in the social contract. Rowlands does not doubt P1. He agrees that the contracting parties must be rational agents. He locates the fallacy of the argument in the move from P2 to P3; in infering that just because the contracting parties form the contract, only they are protected by it.12 In order for the veil of ignorance to function as a legitimate means of deriving just principles, the original position must already be described in such a way that it creates just conditions. This can only be achieved by applying a principle prior to the decision-making process that produces it. Rowlands calls the principle - which already underlies the social contract - the "principle of intuitive equality" (IEP).

IEP: If a property E is unearned in the sense that the individual I who possesses it has contributed nothing to its possession, then I is not morally entitled to benefit from E's possession.13

Rawls makes the mistake of treating properties such as rationality and moral agency as morally relevant, and thus falls prey to what he calls an unreflective intuition - namely, that non-human animals cannot be within the scope of the social contract. But once we recognise that the anthropocentric description of the original position is inconsistent with our reflected intuitions identified by IEP, it must be modified.14

Unreflective intuition: An unreflective intuition is an attitude formed by tradition or otherwise, socially or personally, but which on closer examination turns out to be bigoted and unjustified.

According to the IEP, traits such as rationality, moral agency, or species membership are unmerited because they can generate advantages for their possessors, even though the possession of these traits is accidental. Thus, it is arbitrary for a human being to gain special privileges simply by virtue of being a member of a species, since neither humans nor non-human animals can help being a member of a particular species. Since the IEP underlies the contractarian system, it follows that we must abstract from more than the things found in Rawls' set behind the veil of ignorance:

Rowland's extended set of morally arbitrary properties: {race, gender, social status, wealth, talents, degree of intelligence, degree of rationality, moral agency, species membership, etc.}

How does Rowland's conclusion change the hypothetical scenario? If the IEP is applied coherently, it will produce principles of morality that give direct moral status to non-human animals. If properties such as species membership and rationality are abstracted, the parties cannot know whether they are human or a non-human animal when the veil is lifted. They will therefore base their judgements on self-interest, so that non-human animals are not harmed by the decisions and lifestyles of moral agents.15 According to Rowlands, contractors will then be guided by the question: Can I rationally care about being an entity X?

2. Objections

The most common objection raised against Rowlands concerns the rejection of the reciprocity principle (RT), which, according to David Svolba, necessarily follows from contractarianism.17 RT entails reciprocal obligations between agents: whoever receives moral rights should also be able to fulfil obligations. Svolba argues that RT follows from the social contract because the social contract is a freedom-restricting agreement between different rational and self-interested agents. Because they are self-interested, they would not agree to distribute moral burdens unequally among members of a society, while distributing the benefits and protections of justice equally.18

Rawls also makes a similar argument for the validity of RT: in order to build a fair and just society, agents must cooperate with each other. But cooperation always requires the sacrifice of freedom. For this to be fair, agents must be able to benefit from the loss of freedom, and this is done by requiring others to give up some of their freedom as well.19 Svolba compares non-human animals involved in the social contract to "free riders"; someone who enjoys an equal share of the benefits without agreeing to provide the same benefits to others. He concludes that no contractor would agree to this:20 In a society, it would be unjust if only a portion of the population adhered to specific duties, such as the duty to respect others' property rights. If one group followed this duty while the other group took advantage of the situation by stealing and enriching themselves through theft, it would create an unfair and imbalanced scenario. The individuals who uphold their duty would suffer losses, while those who engage in theft would unjustly benefit from the stolen property.

The fact that only moral agents receive direct moral status in Rawls' sense is not arbitrary, since they are the only ones obliged to reciprocate. Because non-human animals cannot perform duties, they would be akin to free riders, and if the social contract is to be fair, they do not receive direct moral status. The contract form is not suitable for including non-humans because a contract is essentially an agreement between parties who are capable of cooperation. Svolba now argues that Rowland's claim that Rawls follows an unreflective intuition is misguided.

Svolba hence advances a principle alternative to Rowland's IEP intended to more adequately capture these formal presuppositions of contractarianism: the intuitive-equality-of-persons principle.

IEPP: If P possesses property Y and Y is undeserved because P is neither responsible for Y or has done nothing to merit Y, then the possession of Y cannot justify P's receiving an advantage over other persons in society S.

Because RT is true and follows from the structure of contractarianism, we are not meant to abstract from all undeserved properties, but only from those that can lead to disadvantage or injustice between rational persons.22 Since rationality and moral agency - and thus, indirectly, species membership - are necessarily related to or constitutive of the notion of rational persons, these properties are not - contrary to Rowland's inferences from IEP - arbitrary.23 Rowland's integration of non-human animals is therefore not grounded in procedures that take place in the original position and is thus purely subjective.21

In summary, Rowlands has to comment on two propositions:

  • I) Reciprocity is the only principle that can be integrated into the social contract.
  • II) Rowlands' argument is based on subjective pre-existing value judgments.

3. Speciesism, reciprocity, and preexisting value judgments: Rowland's defense

Rowlands now has several ways of countering these objections. In order to better understand his defence, it is helpful to consider the principle on which Rowlands bases contractarianism. Rowlands relies on Tom Regan's distinction between "moral agents" and "moral patients".24 Regan defines a moral agent as someone who has higher cognitive abilities and is able to take impartial perspectives, establish moral principles, and be held accountable. Regan does not deny that there should be reciprocity between moral agents. Mutual cooperation between them is necessary to build a just society. But this does not mean that all non-persons automatically receive only indirect moral status (if any). Regan contrasts the moral agent with the moral patient. A moral patient can act, is capable of suffering and intention, but is incapable of the mental tasks listed above. Moral patients are not exclusively non-human animals. Mentally handicapped people, babies and young children, people suffering from dementia, and comatose people do not have the mental capacity of a moral agent - but they are still considered to deserve moral consideration. Regan believes that human moral patients are given moral status not just because they are human, but because they share many other diverse capacities with moral agents; they can suffer, have desires, pleasures, beliefs, emotions and preferences. Moral patients can be harmed and have a subjective sense of when something bad is happening to them.25

The question arises as to why beings with the capacity to fulfill duties are permitted to harm and jeopardize the well-being of those who lack this capacity. After all, if RT were true, not only would the status quo be able to be maintained because moral agents prioritize trivial needs such as taste or fashion over a lifetime of suffering and usually cruel death for non-human animals, but it would also allow them to exploit some human people not capable of reciprocity. This outcome is unacceptable because it entails legitimizing the killing and exploitation of human beings, which is morally objectionable. Additionally, it appears arbitrary to protect moral agents from potential suffering while allowing moral patients who are similar in their capacity for suffering to be harmed.

Based on this intuition, Regan argues that the moral scope of principles of justice should not be limited to those who are sufficiently rational. Rowlands can now use this argument for his version of the social contract. The purpose of the veil of ignorance is to establish impartiality and avoid prejudice and bias. Regan identifies impartiality with the formal principle of justice, which says that we must treat similar cases similarly. The principle is formal in that it does not specify what characteristics a creature must have in order to be considered "similar" to other creatures.26 If we describe the original position in such a way that only moral agents derive principles for other moral agents, we are not upholding the principle of formal justice, and so cannot speak of an impartial description of the original position. Proponents of RT portray fairness as requiring moral agents to give up freedom in order to obtain rights. However, the restriction of freedom that a moral agent would have to give up in order to refrain from oppressing non-human animals is much less severe than the freedom that a non-human animal would lose if they were forced to give up any satisfaction of basic needs and/or their entire life. In fact, in Regan's words, this would mean that we treat similar cases differently.

Rowlands can thus counter to I.) by arguing that impartiality is a precondition of contractarian theories and it is violated by arbitrary exclusion of a group of living beings. Therefore, reciprocity cannot derive from the structure of the social contract.

Rowland's counterthesis to I.): Reciprocity as a necessary condition for the description of the original position violates the principle of impartiality.

The problem is that Regan's formal principle of justice does not refute IEPP and RT. Because the principle is formal, it is also not clear what the relevant similarity must be. Svolba could simply appeal to IEPP and argue that "being sufficiently rational" is a morally relevant similarity because only most people are capable of reciprocity, and consequently only similar parties will be treated similarly. Moreover, another objection takes on greater significance: Rowland's inclusion of Regan's claim that non-sufficiently rational beings should also be treated morally by virtue of their sentience already presupposes moral considerations that do not arise from a reflective equilibrium behind the veil of ignorance. If moral principles had to follow only from reflective equilibrium, Rowlands would not be able to derive animal rights or their direct moral consideration.

But this objection is misguided and based on a misinterpretation of how the relevant criteria are formed behind the veil of ignorance. Rawls has defined the description of the original position itself as a reflective equilibrium, which is not static, but can or must be adjusted again and again.27 Rawls introduced the notion of revising the description of the original position due to the acknowledgement that the social contract, although hypothetical, can sometimes be grounded in flawed moral intuitions. Historically, individuals who supported or engaged in enslavement would justify their actions by arguing that skin color is a legitimate criterion and would devalue individuals of different races as inferior or lacking agency. Contractarian theories have been employed to derive racist principles or to dismiss racial injustices altogether.28 If Rawls did not allow the description of the scenario to be adapted by contemporary and reflective intuitions, it would be easy to derive principles that are sexist or racist. In this way, it becomes clear that Rawls' set of unmerited characteristics also includes pre-existing value judgments.:

If X, which constitutes a set of unmerited properties that must be abstracted from, is entered into the social contract, then principles are derived from X that are unrelated to the arbitrary properties contained in the set. If X is equated with «being sufficiently rational» in Svolba's sense, there will be moral norms derived from the contract that will protect rational people - no matter their other properties. If X is identified with sentience, then sentient beings are protected by the contract. No matter what value X contains, it already contains various value judgments that will influence decision making. Accordingly, it is wrong to only accuse Rowlands of implementing preexisting value judgments to the contract; rather, it seems to be a feature of contractarianism.

Rowland's counterthesis to II.): The description of the original position always resorts to pre-existing value judgments, but they are legitimate only if they are reflected.

Rowlands now has to show that his approach is more plausible than those of representatives of an orthodox contractarianism. Consequently, the task is to strengthen Rowland's counter-thesis to I.) as much as possible. The idea is to show that the morally relevant properties between moral agents and moral patients can be reduced to properties such as sentience. One way to do this is to expose Svolba's and Rawls's person-based position as arbitrary. Svolba holds that it is wrong to call rationality and moral agency arbitrary properties because they are constitutive of the concept of person and personhood is relevant to the social contract (a person, according to them, is only someone who is also a moral agent). Rowlands can charge Svolba that it is precisely this unreflective use of the strict concept of person that leads to the presupposition of reciprocity and makes his argument circular. The problem with Svolba's and Rawls' notion of a person is not that they prescribe descriptive necessary criteria for "personhood", but that many theorists implicitly draw normative conclusions from the distinction between persons and non-persons. Rowlands can now argue that RT follows as a necessary principle from contractarianism only because contractarians are already defined as the only group of living beings worthy of moral consideration - but this conclusion is unfoundedly speciesist and circular.

P1The contracting parties are persons.
P2Only persons should be granted moral status when the veil is lifted.
KThe contracting parties should have moral status when the veil is lifted.
P3Non-human animals are not persons.
K2Non-human animals should not be granted moral status when the veil is lifted.

The argument is circular because the conclusion - that only contracting parties have moral status - is already assumed in the premises, and P1 is trivial because contracting parties must be persons, since only they are capable of contracting. Although P1. is true, P2. - the already normative interpretation of the concept of persons - must be justified by orthodox contractarians, otherwise they will fall victim to arbitrariness and outright speciesism. Svolba, however, is unable to provide any substantive justification other than the already criticised RT, but merely notes that Rowland's approach is generally too controversial for a contractarian theory and that most contractarians would therefore reject the IEP:

«[...]the principle [IEP] would be roundly rejected by those who — like Rawls — subscribe to the traditional and still pervasive moral view that persons, by virtue of their distinctive capacities for rationality and moral agency, are owed special moral consideration and respect.»29

This is not an argument against IEP, however, since Rowlands would probably not deny that this view is prevalent. But he can argue for the fact that the view that it is acceptable to exploit non-human animals for trivial purposes reflects an unreflective intuition - even if it is widely held. The distinction between moral agents and moral patients may manifest itself as a consequence of a way of thinking, also found in the empirical sciences, that defines the difference between humans and non-human animals not as a strict separation but as a gradual difference that has resulted from years of evolution of different species. While it is true that non-human animals do not exhibit a sense of justice in the Rawlsian sense, it could be argued that this criterion is not necessary for a concept of personhood or for moral reflection.

There is a growing criticism that all traditional concepts of personhood that are linked to moral significance are likely to be underpinned by unreflective speciesist intuitions. This is based on the observations listed in reference to Regan concerning the often neglected consideration of sentience, intentionality and emotionality. Representatives of the orthodox position point to certain higher mental capacities as crucial for moral status, although many of these capacity arguments can be exposed as circular, as shown above, because the morally relevant capacities are always identified in such a way that they can only be applied to cognitively normal humans.

Conceptions of personhood, such as those of Peter Singer30 or Tom Regan, include, in Singer's case, higher mammals and, in Regan's case, all (adult) mammals, birds and probably fish. Regan, for example, considers a person to be someone who is a «subject-of-a-life». A living being, in Regan's sense, is a person or subject-of-a-life if they:

  • have desires and beliefs,
  • can perceive and remember things
  • have a sense of the future or events that will occur in the future,
  • have an emotional life that includes sentience and emotion,
  • show preferences and interest in their own well-being,
  • have the ability to perform actions and act on the basis of the goals he/she has set,
  • have a psychophysical identity
  • and possess individual well-being, logically independent of whether that living being's life or experiences have any benefit to others.31

If the difference between humans and non-human animals is only gradual and not strict, then it seems reasonable to abstract from "being human" and the cognitive capacities that characterise the human species. Thus, we should likewise use a notion of personhood that captures, as Regan does, those properties that it takes for a moral agent to be rationally concerned about being an entity X. Rowlands can argue that while the formal principle of justice cannot directly refute Rawls and Svolba's approach - its extension to moral patients is more reasonable and intuitive because the capacity to suffer, emotionality, and intentionality, for example, are the most fundamental common properties that unite rational and non-rational beings. This approach also minimises the risk of violating the principle of formal justice.

Rowland's argument can be further strengthened: There is a well-known counter-argument against speciesist contractualist theories - the argument from species overlap - which attempts to show that the precondition of a demanding notion of person cannot be presupposed by contractarian theories - if they wish to be plausible.

3.1 Not all "moral patients" are non-human animals: contractarianism and the problem concerning the capacity argument

Direct moral status for "moral patients" can follow from contractarianism because moral agents are able to put themselves in the position of non-sufficiently-rational or differently-rational living beings to some extent. Regan, for example, has rejected contractarianism as a moral theory because, if RT is required, not only non-human animals but also some humans are denied direct moral status. The mentally handicapped, babies and infants are incapable of reciprocity and so would not be given direct moral status behind the veil of ignorance. Regan believes that no one can accept this consequence, since it would make their moral consideration dependent only on the interests of moral agents. If inadequately rational people had only indirect moral status, there would be no reason why rational agents could not use them for their purposes. For example, a person suffering from dementia might be allowed to be sacrificed in order to have their organs harvested and donated to rational agents. This would then not be categorized as morally wrong because a particular individual has been harmed, but because the well-being of a rational agent is at stake.

This result is implausible and counterintuitive. When we learn that a non-human animal has been "unnecessarily" tortured and mistreated, we feel sorry for the animal and judge the act of cruelty to be wrong because we know that something has happened to the animal that has caused them negative subjective experiences - not because a moral agent feels that their well-being has been affected by the act of cruelty.21

This poses a dilemma for orthodox contractarian theorists: If reciprocity is a necessary condition for the original position, then insufficiently rational people do not receive moral status. If, on the other hand, these people are to receive moral status directly through the social contract, and moral agents must therefore be able to imagine themselves as not equally rational creatures, there is no non-arbitrary reason why the scope of the contract should not be extended to non-human animals - especially since non-human animals usually have higher cognitive capacities than human marginal cases.32 Peter Carruthers, one of the most prominent proponents of orthodox contractarianism, attempts to resolve the dilemma by arguing for the plausibility of indirect moral status for insufficiently rational people. Carruthers argues that the exploitation of marginal human beings is not legitimised by the social contract because it would have a socially destabilising effect, and that moral agents are better advised to implement moral rules that are "psychologically supportable" for them. His conclusion is based on the premise that there are certain things that are deeply embedded in human nature - such as caring for one's children or elderly relatives. Allowing people who are not sufficiently rational to be exploited could lead to social instability, because many moral agents have an emotional or "natural" attachment to human moral patients.33 Accordingly, contracting parties would recognize this risk in the original position and derive moral protections for non-sufficiently-rational humans. RT, he follows, is not compromised because human marginal cases do not receive direct moral status. Moreover, the contracting parties do not have to be able to imagine themselves as a not equally rational being, but they only have to imagine as to what the consequences would be for their well-being if non-sufficiently rational beings close to them were exploited.

Carruther's social instability argument is unconvincing. First, it is unclear what applies in cases where no emotional relationship can be identified between moral and non-moral agents. A cognitively underdeveloped orphan baby who has no relatives would be at risk of exploitation - unless Carruthers can somehow infer from the social instability argument that moral consideration is automatically transferred to all human beings; even if they have no relatives to care for them. It is questionable whether he can do this, since a reference to mere biological kinship alone would no longer relate to his original social instability argument.

Another point worth noting is that Carruthers himself states that members of certain cultures have killed physically deformed or cognitively disabled children after birth because of cultural or religious ideals - but this killing has had no destabilising effect on society.34 His explanation for this historical fact is that these societies remained stable only because killing was based on religious beliefs or unreflective intuitions (e.g. "a mentally handicapped child is possessed by the devil"). Behind the veil of ignorance, however, it is necessary to abstract from such beliefs. Non-human animals, he argues, are not taken into account at all, in contrast to human marginal cases, because the emotional bond that rational agents have with non-human animals is merely culturally shaped and does not arise naturally - unlike the bond between parent and child or relative.35

Carruther's argument is misleading and invalid in several ways. Firstly, his last assertion is a purely empirical assumption, which he would first have to verify, since one can also argue that humans have had emotional attachments to members of the animal kingdom since the emergence of our species. Moreover, the prevailing indifference to the (mis)treatment of "farm and laboratory animals" may be as much the result of cultural conditioning as was the indifference to the killing of mentally disabled or physically deformed babies in earlier eras.

It is also not true that the exploitation of non-human animals cannot have a destabilizing effect on society. Groups such as the Animal Liberation Front or the anti-vivisection movement SHAC have very much had a destabilizing effect on society and it is imaginable that as awareness grows about the contemporary treatment of non-human animals and their suffering, the exploitation of non-human animals will lead to protests, disruptions and illegal actions in the future. Carruthers dismisses this argument by claiming that animal rights activists are not motivated by a "natural" bond with non-human animals, but by unjustified moral convictions.36 However, Carruthers appears to contradict himself in this instance. The so-called "natural" bond to a child has not been universally present in all cultures or among all parents. On the other hand, moral recognition of humans who are not sufficiently rational has been achieved through education and the acknowledgment of reflected intuitions. If this is the case, it is not implausible for a similar shift in belief to occur regarding non-human animals. The belief that the killing and exploitation of animals is morally acceptable may stem from an unquestioned intuition that, although still prevalent in society, lacks justification. Carruthers must either abandon his argument of inconsistency or revise it in a way that grants indirect moral status to certain non-human animals as well.

Based on Carruther's argument, it seems difficult to grant indirect status to human marginal cases, but not to non-human animals. Svolba takes a different approach and tries to grant direct moral status to all humans. He suggests that rational agents must abstract from their rationality behind the veil to the extent that they can imagine falling victim to certain laws of nature that might cause them to have cognitive deficits - be it through genetic factors, an accident, or various diseases. Svolba then attempts to exclude non-human animals from the contract by arguing that moral agents can only identify with human marginal cases and therefore only they have direct moral status.

According to Svolba, this identification is not a "technical" one, but presupposes the ability of persons to "see" themselves in others. Svolba gives the example that I can imagine myself as a pharaoh in ancient Egypt or as an alien, while imagining myself as a cat always requires a humanisation of identification: "we succeed only in imagining a rational mind trapped in a cat shaped body"(Bauer, Svolba, S. 59) Once again, it is unclear why this does not apply equally to non-sufficiently-rational humans, since moral agents are just as unable to imagine being a cognitively disabled human and we would, in trying to identify with her, imagine our rational minds "trapped in a non-rational human shaped body". The core difference Svolba now makes between these two cases is that in the original position all that is needed is a counterfactual projection, applicable either within person-like (e.g., aliens) or within species-like groups. The latter is because we can imagine that we might have serious accidents or genetic defects, and thus carry cognitive defects, whereas we cannot imagine that we might be born as a cat.37

Svolba presupposes a separate form of distinction between moral agents and moral patients in order to ascribe direct moral status to non-sufficiently rational humans. But this leads us to the question of whether Svolba still holds RT at all, or whether, on his account, it must follow from contractarian theories. Svolba claims that RT must still be a presupposition insofar as human marginal cases could have been "persons" (in the narrow sense) if they had been treated differently by fate. In this respect, it is still true that moral agents derive principles for moral agents - but they also recognise that they can lose their status as moral agents, or could never have formed it. In the case of children, it can be argued that they are potential persons (children have moral standing because they are becoming persons, and growing up to be persons must be protected) - but it is harder to see how RT comes into play when we include the mentally disabled, since they - just like non-human animals - have no potentiality for moral agency.38

The problem with Svolba's approach can best be illustrated by underlining again which types of identification Svolba distinguishes:

Qualitative identification (1): Identification or imaginative projection exists when rational agents can recognize themselves in other rational agents.

Quantitative identification (2): Identification or counterfactual projection exists when rational agents can imagine that they could have been human non-persons.

Only (2) is necessary to give direct moral status to non-sufficiently-rational people. (1) represents a strict conception of the notion of identification. (2), unlike (1), is not based on an imaginative identification of persons to «non-persons», but on the hypothetical assumption "I could have been x" combined with statements about biological kinship "humans could only have been humans". However, it is not yet clear why it is necessary to draw a line at the species boundary. Rowlands suggests that the original position has so far been described in a speciesist way, and only those properties have been abstracted that could lead to inequalities between cognitively normal humans. Svolbas (2) suggests that the original state should be described in such a way that, for example, knowledge of one's own DNA is abstracted away, so that contractors have to consider whether they might be born with a genetic defect. It is not trivial why a rational agent can imagine being born with a different DNA, but not in the body of a different species. Svolba sees the original position as a purely theoretical and hypothetical construct, and if he assumes that rational agents can imagine being a moral patient, it is questionable why they can only do so with human moral patients. Svolba would have to assume that agents have some sort of relevant genealogy, or that human history is very relevant in the original position, in the sense that when the veil is lifted they must exist as humans. Rowlands can challenge this view by showing that, while counterfactual identification plays a role in the social contract - as a way for contracting parties to put themselves in moral patients - (1) can also be reformulated as:

*Identification (1)**: Identification or counterfactual projection exists when rational agents can recognize themselves in other agents/beings.

Rowlands can point out that moral agents can indeed identify with non-human animals. When a non-human animal is abused, we feel bad or show compassion because we empathize with the animal and imagine that their mistreatment has caused pain and extreme anxiety and is accompanied by negative subjective feelings. It may in fact be that I could never have been a cat because of my genetics, but it may well be that I can experience the same pain and loss of well-being as they do because of my sentience. Non-human animals are also related to humans and they exhibit similar reactions and emotions to aversive and unpleasant stimuli. Why should the fact that I might have different human DNA be relevant to the original position and why not the possibility that I could have been born in a different body? Why are potentiality, history, or counterfactual scenarios in which I could have been a cognitively disabled human relevant behind the veil of ignorance, but not the fact that I can imagine experiencing the same suffering as a non-human animal? If no strict distinction between humans and non-humans can be made out, Regan's conclusion comes into play again: if contractors should be able to empathize with human non-persons, but non-human animals are excluded, the argument rests on unjustified speciesist assumptions. The veil of ignorance is supposed to guarantee impartiality, and Rowland's (1*) is better able to grasp the principle of formal justice than Svolba's (2) because (1*) identifies identification with shared similar properties, not just shared similar properties within the species boundary.

For contractarianism, this means that when human non-persons are given direct status, RT is not applied in the already defined wording, but is based on a principle that is merely defended in a speciesist way by Svolba. VanDeVeer believes that knowledge of one's own species-internal possibilities in the original position is biased. He assumes that if it is assumed that the contracting parties can imagine themselves to be a non-sufficiently-rational human being, it is only fair to describe the original position in a way that the contracting parties do not know exactly which non-sufficiently-rational being they might be. If the contracting parties already know that they are born human, they will derive speciesist principles, just as if the contracting parties knew that they belonged to a particular race, they would derive racist norms. VanDeVeer places the explanatory burden on the proponent of the argument, suggesting that it must be shown why "belonging to the human species" is a relevant factor.39

According to VanDeVeer, contractors would ask themselves what rights or moral protections they would need if they were a non-sufficiently-rational being. If they had this possibility open to them, they would derive the principles that would give them the greatest possible protection. Rational agents would then be very likely to regard these living beings not as free riders because they are unable to perceive reciprocity, but as living beings they could possibly be and whose needs must be protected as best they can. Thus, counterfactual identification, if it is to be impartial, must capture non-human animals as well as human marginal cases and cannot be applied only within the species.

4. A dilemma for Rowlands.

The discussion of the arguments concerning the possibilities of describing the original position has shown that contractarianism is not necessarily constrained by reciprocity, and that the possibility of including non-human animals seems at least possible. Svolba's final objection, which remains to be discussed, concerns Rowland's set of unmerited properties. Svolba criticises the fact that, because of the IEP, it is no longer clear from behind the veil which properties are relevant for determining an individual's moral claims in the first place. Svolba assumes that the properties that are considered morally arbitrary as a result of Rowland's principle are rejected by most other animal rights advocates, such as Singer and Regan. Svolba's interpretation of the IEP can be understood as follows:

IEP*: A property E is morally arbitrary if it is unmerited and the possession of E is not brought about by the fault of E's bearer.

It follows from IEP* that being a "subject-of-a-life" or having certain interests and preferences is unmerited and does not come about through the fault of the being possessing the properties. Svolba correctly notes that Rowland's guiding value for contractors to follow is to be able to rationally worry about being a certain being when the veil lifts. To be able to worry about being a certain being, however, must be grounded in unmerited properties, since a property is the only reference point for contractors to be able to put themselves into a being they could be after the veil is lifted.

This creates a dilemma for Rowlands: If the contractors had to establish moral principles within Rowland's description of the original position, they would probably decide that a calf should not be separated from the mother after birth and eventually killed, because the calf is sentient and separation from their mother induces a subjective, negative feeling in both individuals. But when Rowlands makes this statement, he ascribes moral status to the calf and their mother on the basis of properties that, by the logic of his own principle, are morally arbitrary and irrelevant, because it's not the calf's fault that they are sentient. Rowlands then has to resort to identifying some unmerited properties, those that determine whether a rational agent can rationally care to be X, as non-arbitrary.40 But then Rowlands has the problem that philosophers like Carruthers can simply insist that rationality and reciprocity are unmerited from a moral point of view, but not arbitrary, and attach the description of the original position to them. Rowlands would then have no way of defending animal rights on the basis of contractarian systems, since it would no longer be clear which properties are unmerited but morally relevant.

It may be objected that the dilemma Svolba outlines only appears to be one and is based on a misreading of Rowland's IEP. Rowland's wording for IEP is:

«[...] if a property is undeserved in the sense that its possessor has done nothing to merit its possession, then its possessor is not morally entitled to whatever benefits accrue from that possession.»41

Rowlands does not claim that if a property is unmerited, it is morally arbitrary; rather, he highlights that an individual I is not morally entitled to benefit from the unmerited property E if they derive an unfair benefit from it. Rowland's position can be interpreted as stating that properties such as "being the subject-of-a-life" or "having preferences and interests" are unearned because the individuals who exhibit the properties did nothing to have them. However, it is essential for assessing the moral relevance of unearned properties to question whether an agent is able to unfairly benefit from them. Skin color is an unearned property, and it is therefore morally arbitrary because people who have a white skin can gain an unfair and unjustified advantage from it. Most humans are rational and rationality is an unearned trait. To determine whether rationality is morally arbitrary, moral agents must ask themselves whether they could derive an unfair advantage over others because of this property. Since this is the case - because rationality is needed to legitimize the exploitation of non-human animals - it is morally arbitrary. It is more plausible to understand Rowland's IEP in terms of abstracting certain properties as follows:

Principle of unearned properties (Rowlands): A property is unearned and morally arbitrary if its abstraction has no bearing on the fact that a moral agent can rationally worry about being an entity X.

Rowlands does indeed attribute certain properties as non-arbitrary. As accurately described by Svolba, Rowlands' account of the original position leads to the conclusion that it is unjustifiable to kill animals commonly referred to as "farm animals" for food. This is because no rational agent would consent to being killed for trivial purposes when there are alternative options that allow for a healthy life. Moral agents, in this scenario, would be obligated to relinquish certain freedoms, including the freedom to choose their dietary preferences. On the other hand, non-human animals not only endure a predominantly painful death but also lead lives filled with despair and misery due to our entitlement to consume them. The potential suffering experienced by moral agents who adopt a vegan lifestyle by limiting their dietary choices is incomparable to the sacrifices forcibly imposed upon non-human animals. Any self-interested and rational agent positioned behind the veil of ignorance would thus establish moral norms that safeguard non-human animals from such exploitation.42 These considerations that Rowlands attributes to contractors seem to suggest that there is at least one fundamental property that cannot be abstracted from in order to be rationally worried about being X. Rowlands can then be credited with defining the properties to be abstracted from in the sense that they conform to the formal principle of justice, and the common enominator sufficient for identification is identified as that where we are all most alike: in the capacity to have a certain kind of subjective sentience - that our lives can go well or very bad; dependent on the actions of others or the circumstances we live in. The dilemma posed to Rowlands is thus only apparently one.

The final difficulty with Rowland's approach discussed here is the accusation that Rowland's principle is too extravagant, leaving contractors without a clue as to how to derive precise moral principles and norms when they do not know what kind of living beings they will be. If the minimum sufficient common denominator between moral agents and non-sufficiently-rational creatures can be a very primitive form of sentience observed in creatures to which not even other animal rights advocates would ascribe moral status (e.g., insects), Rowland's contractarianism would seem unattractive, counter-intuitive, or impossible as a basis for a moral theory. It is indeed an open question as to how exactly Rowlands resolves this accusation and how exactly he frames contractarian decision making. It is not in my capacity to discuss this in depth, but merely to present some perspectives on what solutions might look like and why Rowland's position may nevertheless offer a promising perspective for answering questions of animal ethics within the contractarian system.

On the one hand, it seems misguided to assume that just because contracting parties are required to imagine themselves as different moral patients it leads to a complete intellectual overload of rational agents. The original position does not preclude contracting parties from having knowledge of what different kinds of creatures, races, genders, etc. there are - they merely do not know their place in the world when the veil is lifted. Rawls allows contractors all information and facts about the world, such as psychological insights into human behaviour or scientific data. In Rowland's system, this knowledge would extend to empirical, psychological and evolutionary data about non-human animals. Since no rational and self-interested agent would agree to the premature separation of a child from an important caregiver, as mentioned in a previous example, moral rules for the use, or rather non-use, of non-human animals could be derived. If it turns out that insects are minimally sentient, so that they can feel or see their surroundings but do not actually suffer, agents will not infer a moral status for them. It is true that the parties to the contract would have to put in considerably more effort in deriving moral principles than if they included only rational agents - but this is a consequence of any moral theory that deals with the moral status of non-human beings. As with other moral theories, the greatest difficulty arises in providing guidance for action in exceptional situations. Can Rowland's theory answer the question of whether to save a dog or a rational agent? Although Rowland's argument about vegetarianism sheds light on the fact that every sufficiently sentient creature must probably be given a right to life, it is still unclear how these rights to life rank in relation to others. In particular cases, when it is a matter of life versus life, should more sentient or intelligent beings be given priority? These are broader considerations that Rowland's anti-speciesist contractarianism must attempt to answer. It should be noted here that, although no conclusive answer can be given in this paper, orthodox contractarians and other moral philosophical systems also have difficulty with such special cases. Can a principle be derived from contractarianism that always gives priority to one's own children, or does this violate the principle of impartiality? Should we, according to Carruthers, always be obliged to favour cognitively more mature individuals over less intelligent ones? Rowland's contractarianism can certainly provide approaches to answering these questions.

Conclusion

Based on Mark Rowland's anti-speciesist approach to contractarian moral theories, several objections to his theory were discussed. The paper focused primarily on answering the questions of whether the formal nature of the social contract even permits conferring direct moral status on a cross-species group of moral patients and what problems arise with Rowland's specific proposition. First, the paper discussed objections raised by David Svolba and Robert Garner, who criticized, on the one hand, that the reciprocity principle is necessarily related to contractarian systems and that Rowlands can only conclude that non-human animals are taken into account by contracting parties by presupposing preexisting value judgments. The main strategy in defense of Rowlands was to strengthen his approach, which claims that Rawls has not applied his own principle consistently and thus reciprocity need not be a necessary condition, and to show that Rowland's view of the content of the principle of formal justice and thus of impartiality is more plausible. As to the question concerning Rowland's reliance on preexisting value judgments, it has been shown that this is an inherent problem of contractarian theories. Because the description of the original position itself is based on a reflective equilibrium, it is inevitable that some moral deliberation must occur prior to the actual hypothetical scenario of decision making. Svolba also attempts to explain the theoretical relevance of rationality and moral agency in contractarianism in terms of already presupposed value judgments. Having clarified this, I have again attempted to present Rowland's position as more plausible than those of orthodox contractarianism. On the one hand, I have done so by resorting to general criticisms of speciesist reasoning - for example, that Svolba's concept of personhood and the moral relevance attached to it is arbitrary and circular - and, on the other hand, by supporting these concerns with the inability of contractarian theories to include human non-persons within its scope. The defender of an orthodox contractarianism is then faced with a dilemma: either reciprocity is presupposed and human non-persons and also certain non-human animals are given no or at most indirect moral status, or the contracting parties are able to imagine themselves as non-sufficiently-rational beings behind the veil of ignorance, according to which an exclusion of non-human animals - just because they are not human - is arbitrary and they would have to be given direct moral status. Lastly, some objections to Rowland's IEP and its practical implementation in the social contract were discussed. Svolba wanted to reduce Rowland's position to absurdity by showing that, according to IEP, any unearned property is ultimately morally arbitrary and thus must also be abstracted from sentience, interests, or concepts such as «being the subject of a life» in the original position. Svolba's interpretation of IEP, however, can be challenged. It seems more reasonable to attribute to Rowlands the view that while certain properties are unearned, they are morally arbitrary only if their bearer can benefit by them over others. From these insights, it can be assumed that contractarian systems have no structural peculiarities that exclude non-sufficiently-rational beings per se from their scope, and that contractarianism can also be used to establish the direct moral status of non-human animals.


  1. Rawls, John, A Theory of Justice. Revised Edition, Harvard 1999, S.102ff.
  2. Ebd., S. 122.
  3. Ebd., S. 129.
  4. Rawls, John, Gerechtigkeit als Fairness, Frankfurt am Main 2003, S. S. 59 – 63.
  5. Garner, Robert, A Theory of Justice for Animals: Animal Rights in a Nonideal World, New York 2013, S. 6f.
  6. Rawls, A Theory of Justice, S. 120.
  7. Ebd., S. 102.
  8. Ebd., S. 124.
  9. Rawls, Gerechtigkeit als Fairness, S. 60.
  10. Garner, Robert, A Theory of Justice for Animals, S. 9f.
  11. Rowlands, Contractarianism and Animal Rights, S. 237., Rowlands, Mark, Animal Rights. Moral Theory and Practice, Hampshire 2009, S. 148.
  12. Rowlands, Contractarianism and Animal Rights, S. 239f.
  13. Ebd., S. 241.
  14. Ebd., S. 242f.
  15. In dieser Arbeit wird vor allem David Svolba stellvertretend als Kritiker von Rowlands Theorie aufgeführt, weil er direkt auf Rowlands Aufsatz geantwortet hat. Es gibt aber weitere Philosophen, die einer Erweiterung der abstrahierbaren Eigenschaften auf Spezies und Rationalität in demselben oder ähnlichem Sinne sehr skeptisch gegenüberstehen: Als weiteres wird später noch Peter Carruthers und Robert Garner aufgeführt, siehe aber z.B. auch: Baxter, Brian, A Theory of Ecological Justice, Routledge 2005, S. 68f.
  16. Svolba, David, Is there a Rawlsian Argument for Animal Rights?, in: Ethical Theory and Moral Practice, Bd. 19, Nr. 4, 2016, S. 973-984, hier S. 979. Svolba bezieht sich bei der Erklärung von RT vor allem auf: Hart, H.L.A, Are There Any Natural Rights?, in: Philosophical Review, Bd. 64, Nr. 2, 1955, S. 175-191, hier S. S. 186, S. 190f.
  17. Rawls, John, Legal Obligation and the Duty of Fair Play, in: Hook, Sidney, Law and Philosophy, New York 1964, S. 3 – 35, hier S. 9f.
  18. Svolba, Is there a Rawlsian Argument for Animal Rights?, S. 980.
  19. Svolba, Is there a Rawlsian Argument for Animal Rights?, S. 977f.
  20. Ebd., S. 978.
  21. Garner, Robert, Animal Ethics, Cambridge 2005, S. 86f.
  22. Rowlands, Animal Rights, S. 162.
  23. Regan, The Case for Animal Rights, Berkeley/Los Angeles 20042, S. 151 – 156.
  24. Regan, The Case for Animal Rights, S. 128.
  25. Rawls, A Theory of Justice, S. 18.
  26. Mills, Charles, The Racial Contract, New York 1997.
  27. Svolba, Is there a Rawlsian Argument for Animal Rights?, S. 981.
  28. Singer, Peter, Practical Ethics, Cambridge 20113, S. 94ff.
  29. Regan, Tom, The Case for Animal Rights, S. 243.
  30. Das Argument ist von Tom Regan abgeleitet, siehe Regan, Tom, An Examination and Defense of one Argument Concerning Animal Rights, in: Inquiry: An Interdisciplinary Journal of Philosophy, Bd. 22, Nr. 1 – 4, 1979, S. 189 – 219, hier S. 196f.
  31. Carruthers, Peter, Animal Mentality: Its Character, Extent, and Moral Significance, in: The Oxford Handbook of Animal Ethics, Oxford 2011, 373-406, hier S. 388f.
  32. Ebd., S. 389.
  33. Ebd., S. 391.
  34. Ebd.
  35. Bauer, Nathan, Svolba, David, Justice at the Margins: The Social Contract and the Challenge of Marginal Cases, in: The Southern Journal of Philosophy, Bd. 55, Nr. 1, 2017, S. 51 – 66, hier S. 58. Dieser Text wurde von Svolba und Nathan Bauer verfasst, der Einfachheit halber werde ich im Folgenden aber nur Svolba nennen, wenn ich auf den Text referiere.
  36. Pluhar hat dafür argumentiert, dass die meisten Potentialitätsargumente nicht erklären können, weshalb von Geburt an geistig Schwerbehinderte moralischen Status erhalten sollten: Pluhar, Beyond Prejudice, S. 146f.
  37. VanDeVeer, Donald, Of Beasts, Persons and the Original Position, in: The Monist, Bd. 62, Nr. 3, 1979, S. 368- 377, hier S. 372f.
  38. Svolba, Is there a Rawlsian Argument for Animal Rights?, S. 983.
  39. Rowlands, Contractarianism and Animal Rights, S. 238f.
  40. Rowlands, Animal Rights, S. 163 – 165.