To What Extent Do Ethical Theories Agree About What We Owe Animals?

Original text Luciano Carlos Cunha · Volume XI, p. 67 of the PDF edition · English translation

6.1. Introduction

Discussions in normative ethics usually emphasize the disagreements among the various approaches, especially among consequentialist, deontological, and character-centered theories. Paying attention to these disagreements is important for assessing the different foundations and practical implications of these theories. However, there are also points of convergence. Below we will see that one possible point of convergence is the equal consideration of all sentient beings, that is, of all beings capable of having experiences, regardless of species, degree of cognitive capacities, etc. This is so for two reasons, explained below.

(1) There are general arguments for equal consideration that do not depend on the prior adoption of any specific ethical theory/school. If these arguments are sound, every ethical theory must incorporate the equal consideration of all sentient beings.

(2) There are specific arguments, coming from the most varied contemporary ethical theories, in support of the equal consideration of all sentient beings. Some of these theories necessarily imply equal consideration, and those that do not are nonetheless compatible with equal consideration.

In section 6.2, we will look at the main arguments that have been used to defend the prevailing position, which gives little or even no moral consideration to nonhuman animals. In section 6.3 we will look at the main argument for the view that the good of any and every sentient being matters in itself. In section 6.4 we will look at the basis on which the idea that every sentient being should receive equal consideration is founded. Having seen these general arguments, which do not depend on the prior adoption of a specific normative theory, the following sections present the main theories of contemporary normative ethics and examine their implications for the moral consideration of nonhuman animals. At the end, we will look at practical implications specifically regarding the situation of exploited animals and of those living in the wild.

6.2. Arguments that attempt to justify disregard for animals

Animals are exploited for the most diverse purposes: food, as test models, clothing, etc. This use causes death and suffering to a gigantic number of animals. Despite this, it is a widely accepted practice. Moreover, animals are also harmed by natural processes, such as malnutrition, hunger, thirst, disease, weather events, natural disasters, etc. The standard view is that we should not help them. However, if the victims, whether of exploitation or of natural processes, were human, the standard view would be the opposite. Can this disparity be justified?

Among the most common claims made in an attempt to justify the standard view are⁴⁴:

(1) “Only humans matter morally⁴⁵.”

(2) “Animals were not made in the image and likeness of a creator⁴⁶.”

(3) “Animals do not have (nor have the potential for) certain capacities⁴⁷ (such as reason, language, freedom, moral agency, a sense of justice, autonomy, culture, the ability to claim rights, etc.) or certain relationships with us⁴⁸ (affective relationships, relationships of mutual solidarity, political relationships, etc.).”

(4) “Animals do not belong to the same species as the individuals who have those capacities or relationships⁴⁹.”

(5) “Disregarding animals is a natural or traditional attitude⁵⁰.”

Below we will look at an argument that points out a problem common to all of these attempts to justify the standard position.

6.3. The argument from relevance

The argument from relevance⁵¹ starts from the trivial premise that, for a criterion to be justified, it must be relevant to what is at stake. For example, what is relevant in assessing a candidate for a medical team is their medical skills, whether they are punctual, whether they treat patients well, etc. By contrast, in assessing whom to attend to in an emergency, what is relevant is knowing who is injured, who is seriously ill, who is at risk of dying, etc.

Now imagine if it were required that, to be accepted to work on a medical team, one had to be ill, and that, to be seen by a doctor, one also had to be a doctor. This would be applying criteria that are relevant in one context to another in which they are not. The argument from relevance points out that the defenses of disregard for animals commit a similar confusion. This will be explained in more detail below.

What is at stake in questions of moral consideration is knowing how our decisions would positively or negatively affect those they reach. Giving moral consideration to someone means avoiding harming them and seeking to benefit them. If so, then the appropriate criterion for knowing whom to give moral consideration to is knowing who can be harmed and benefited.

The species to which someone belongs; whether or not they were created in the image of a deity; whether or not they have certain capacities or relationships (or the potential for them); whether or not a behavior is natural, etc.—none of this determines who can be harmed or benefited. Therefore, these are not relevant characteristics for knowing whom to give moral consideration to.

It is the possession of sentience (that is, the capacity to have experiences) that determines who can be harmed or benefited. This is so for two reasons:

(1) It is sentience that makes there be someone, an individual inhabiting a body, and not merely a living, “empty” body. It is what separates a human, a dog, a fish, a shrimp, and a bee, on the one hand, from a plant, a bacterium, and a fungus, on the other.

(2) It is what makes individuals capable of experiences (which may be positive, neutral, or negative), of desires (which may be satisfied or frustrated), and of enjoying items on a list of objective goods. Therefore, whatever theory of well-being one holds, one must recognize that sentience is what makes there be well-being at all—that is, it is what makes it possible for someone to be harmed or benefited.

Disregard for animals is commonly defended by claiming that we have duties only toward those capable of recognizing duties⁵²; that we should protect with principles of justice only those who have a sense of justice⁵³; and that only those who understand what rights are should have rights⁵⁴. This is a confusion analogous to thinking that, for someone to be entitled to be seen by a doctor, they must also be a doctor. These capacities are relevant for knowing whom to hold responsible, but not for knowing whom to consider morally. This is already recognized in the case of babies and of humans with cognitive impairments: we recognize that we should consider them morally, even though they are incapable of being held responsible. However, this also implies that every sentient being, regardless of species, should receive direct moral consideration—that is, their good matters in itself.

6.4. Some consideration or equal consideration?

A common view is that we should give some consideration to animals, but less than we should give to humans. There is, however, a principle, called the principle of equal consideration⁵⁵, which shows that this attitude is biased.

Equal consideration is already implied by the argument from relevance, which we saw in the previous section. If we have reasons to consider someone because that someone can be harmed and benefited, then the strength of these reasons must depend on the magnitude of the harms and benefits —and not on the species to which they belong.

In an unbiased approach, harms and benefits of similar magnitude receive the same weight, and greater harms and benefits receive greater weight. This suggests that not

only is there no justification for excluding animals from the sphere of moral consideration: there is also no justification for assigning less weight to their good.

It might be objected that there must be something wrong with the idea that the strength of the reasons to avoid harming and to seek to benefit should depend solely on the magnitude of the harms and benefits. According to this objection, if whoever suffers the greater harm has, for some reason, done something to deserve the harm they suffer, then we should not prioritize eliminating that harm (indeed, we should not even eliminate it, since it is deserved).

The reply to this objection involves pointing to three things:

The first is that the idea of desert can easily be incorporated into the principle of equal consideration. For example, the principle can be adapted to something like: “prioritize avoiding the greater harm, unless the potential victim of the greater harm has done something to deserve the harm they would suffer” and “prioritize promoting the greater benefit, unless the potential recipient of the greater benefit has done something not to deserve the benefit they would enjoy.”

The second is that the idea of desert actually depends on equal consideration to make sense. This is because, for someone to deserve the harm they suffer (and not to deserve the benefit they enjoy), they must first have acted unjustly toward someone—that is, they must first have violated the principle of equal consideration.

Finally, the third is that including the notion of desert in moral reasoning could never justify diminishing the strength of the reasons to avoid harming and to help nonhuman animals. This is because, for someone to deserve the harm they suffer, they must be a morally responsible agent. Since nonhuman animals do not have cognitive capacities above the level at which it makes sense to hold someone responsible, it makes no sense to say that they deserve the harms they suffer, whatever these may be.

In light of this, it might be objected that, for the same reason, only beings with cognitive capacities above a certain level can deserve benefits, and that this would justify prioritizing normal adult humans.

A first problem with this objection is that it could not prioritize benefiting all normal adult humans, but only those who had done something to deserve some extra benefit (and, even with respect to these individuals, prioritizing them would not be justified in any and every situation, but only in those in which they had done something to deserve some extra benefit—once they had received the extra benefit in question, their priority would be canceled until they again did something to deserve another extra benefit).

The second (and greater) problem is the following. It makes sense to say, of two individuals who have cognitive capacities above a certain level, that it is justified to give an extra benefit to one and not to the other, because one did something to deserve the extra benefit and the other could have done so but chose not to. However, it makes no sense to say, comparing an individual who has cognitive capacities above a certain level with another who does not, that it is justified to give an extra benefit to one and not to the other because the former did something to deserve the extra benefit, given that the latter does not have that capacity. For example, if we have two normal adult humans who could have cooperated in a situation, but one of them chose not to (and could have done so without sacrificing anything of greater importance), it makes sense to say that only the one who cooperated deserves a certain benefit. However, it makes no sense to say the same if we are comparing a normal adult human with, for example, a baby. The baby should receive all the benefits it needs without having to do anything to deserve them (since it does not have that capacity). But the same holds for nonhuman animals.

Thus, the notion of desert, although it can be incorporated as an addendum to the principle of equal consideration, would not justify giving less weight to the good of nonhuman animals, since the species to which someone belongs, like different degrees of cognitive capacities, are the results of the natural lottery, and not of merit or demerit.

Accepting the principle of equal consideration does not commit us to a specific ethical theory/school⁵⁶. This is because this principle does not prescribe what the right action is (that is done by the substantive principles of each ethical theory, which we will see in the following sections). The principle of equal consideration only prescribes that, whatever substantive principle we adopt, there is no justification for applying it in a biased way—that is, there is no

justification for giving different weight to the good of individuals on account of arbitrary factors such as their species, their race, their gender, etc.

6.5. The moral consideration of sentient beings from the standpoint of various ethical theories

The full moral consideration of all sentient beings has been defended from the standpoint of the most varied theories of normative ethics⁵⁷. In the following sections, we will look briefly at each of these theories, and then see how they would have to be applied with regard to the moral consideration of nonhuman animals.

In all the examples we will see, the letters represent individuals, and the numbers represent their respective levels of well-being. We will not assume here any specific view as to which aspects are relevant for assessing individuals' well-being (whether positive/negative experiences, satisfied/frustrated preferences, items on an objective list, etc.). We will leave this open and speak only of individuals' well-being, since the whole discussion here holds equally whatever factors we believe are relevant for assessing well-being.

6.6. Utilitarianism

The goal of utilitarianism⁵⁸, in its classical version, is to maximize the aggregate total sum of the well-being of the individuals affected by the decision. Let us look at an example:

If our only options are S1 or S2, utilitarianism will prescribe choosing S2, since it has the greatest overall total balance (that is, the positive total minus the negative total).

6.7. Number of beings with positive well-being, and average utilitarianism

Another possibility is to claim that the number of individuals with positive well-being is something valuable in itself (and not merely indirectly, in relation to the goal of maximizing the total balance). Below is an example to illustrate how this view would determine the right action, and also to illustrate the difference between it and average utilitarianism. Average utilitarianism is a variant of utilitarianism that prescribes that we choose not the situation with the greatest total sum, but the one with the highest average (that is, the total sum divided by the number of individuals):

For utilitarianism, whether classical or average, both situations are equally good, since the total sum is the same (+4), and so is the average (+1). However, it could be said that S2 is better in at least one respect: in it, there are two individuals with positive well-being, whereas in S1 there is only one. Now consider the following situations:

Classical and average utilitarians will say that S1 is better than S3, since in it both the total sum and the average are greater. By contrast, someone who holds that the more individuals with positive well-being the better will say that S3 is better in this respect, even though the total sum and the average are lower, since in it there are two individuals with positive well-being, whereas in S1 there is only one. Whether they will consider S3 better all things considered will depend on how much weight they assign to the aspect “number of individuals with positive well-being” compared with the aspects “total sum” and “average.” But they will consider it better at least in one respect.

6.8. Maximin

Maximin is directly concerned with individuals' levels of well-being (regardless of what the sum, the average, or the number of individuals with positive well-being is). It prescribes choosing the course of action in which the minimum level of well-being is highest (hence the name maximin, which comes from maximizing the minimum⁵⁹). Let us look at an example:

In this example, classical and average utilitarians, and even those concerned with the number of individuals with positive well-being, would have to say that both situations are equally good. However, proponents of maximin would say that S2 is better, since in it whoever is in the worst condition (C=-4) is better off than whoever is in the worst condition in S1 (C=-5).

Again, as happens in conflicts between any other aspects, in situations that are better with respect to the minimum level but worse in other respects, the situation will or will not be considered better all things considered depending on the weight given to maximin compared with the other aspects.

6.9. Prioritarianism

6.9.1. What characterizes prioritarianism?

Prioritarianism⁶⁰ has differences from and similarities with maximin, utilitarianism, and egalitarianism (the last of which we will see in section 6.10). Let us begin by comparing it with maximin. Consider the following two situations:

According to maximin, S1 is better, because in it whoever is worst off (B=-49) is better off than whoever is worst off in S2 (B=-50). That is, S1 has the highest minimum level of well-being. However, one objection to the conclusion that S1 is therefore better all things considered is that, although in S2 the minimum level of well-being is lower, there is a considerable improvement in the well-being of one of the individuals. In S1 we have two individuals suffering greatly, and in S2 at least the suffering of one of them is considerably less than it would be in S1.

One way of trying to ground the claim that S2 is better is by appealing to the utilitarian principle: in S1 the aggregate total is -97, whereas in S2 it is -80. However, another way is to appeal to the prioritarian principle. Prioritarianism, like utilitarianism, also seeks to maximize the aggregate total balance. However, unlike utilitarianism, in which each additional unit of well-being has the same value regardless of how well or badly off the one who would receive that additional unit was, in prioritarianism each additional unit of well-being has greater value the worse off the one who would receive it is.

Thus, unlike maximin, which is concerned only with maximizing the minimum level, prioritarianism is also concerned with the amount of benefit that can be produced by each course of action. And, unlike utilitarianism, which is concerned only with maximizing the total benefit, for prioritarianism it matters whose well-being would be improved (the worse off someone is, the more value there is in improving that someone's well-being).

To better understand this difference from utilitarianism, we will now compare the prioritarian principle with the principle of diminishing marginal utility, found in utilitarianism.

6.9.2. Differences from the principle of diminishing marginal utility

The prescriptions of utilitarianism, in a wide range of cases (but not in all), coincide with favoring those who are in the worst situation. However, this happens in utilitarianism not because it values benefiting those in the worst condition in itself, but because doing so is usually more efficient for maximizing the total balance, owing to the principle of diminishing marginal utility, explained below.

The principle of diminishing marginal utility states that the smaller the amount of a certain resource someone has, the more well-being an additional unit of that resource generates for that someone. For example, if someone has only one kilo of rice, giving that person an additional kilo will make a greater positive difference to them than if we gave the kilo to another person who already has ten kilos of rice. Thus, what this principle says is that resources have diminishing marginal utility.

For example, suppose that initially we have A=+10; B=+1 (total = +11), and that we have a resource that, if given to A, will increase A's well-being by two points (A=+12, B=+1, total = +13), and that, if given to B, will increase B's well-being by four points, since B is in a worse situation (A=+10, B=+5, total = +15). Thus, utilitarianism would prescribe giving the resource to B because the outcome will then have a greater aggregate total balance.

This is different from having improving the situation of the worst off as a goal. If, in another situation, giving the resource to someone who is already better off is more efficient for maximizing the total sum, utilitarianism will prescribe favoring whoever is already better off. Let us look at an example.

Let us imagine the same situation: A=+10; B=+1 (total = +11). Suppose this time that A, even though already better off than B, would still have their well-being increased by more points if they received the resource in question. Imagine, for example, that A is a peanut fan and that B likes peanuts, but not as much as A, and that if we give a bag of peanuts to A their well-being will increase by three points (A=+13, B=+1, total = +14), and that if we give the bag of peanuts to B their well-being will increase by two points (A=+10, B=+3, total = +13). In this case, utilitarianism would prescribe favoring whoever is already better off, since this would be more efficient for maximizing the total sum.

In short, according to the principle of diminishing marginal utility, an additional unit of a resource is more efficient for maximizing the aggregate total if it is given to whoever has less of that resource. According to prioritarianism, by contrast, an additional unit of well-being has more value the lower the level of well-being of whoever would receive that additional unit. Parfit (1995, p. 105) sums up the difference this way: “just as resources have diminishing marginal utility, so utility has diminishing marginal value.” Let us look at an example:

Let us imagine that, from the initial situation, we can increase the well-being of one of the two individuals by +5. In S1 we increase by +5 the well-being of A, who is already at +20. In S2 we increase by +5 the well-being of B, who is at -10. According to utilitarianism, both situations are equally good, since the total is the same (+15). Prioritarianism, in turn, would say that S2 is better, since in it the benefit of +5 goes to whoever is worse off. That is, prioritarians would say that increasing by +5 the well-being of B, who is in a worse situation, has more value than increasing by +5 the well-being of A, who is already better off.

6.9.3. Objections to prioritarianism

One possible objection to prioritarianism is to point out that, depending on how much greater the benefit that could be given to someone already much better off were, it would still have to prioritize benefiting the one already much better off, even while giving greater weight to units of benefit in proportion to how badly off someone is. A theory that apparently does not have this implication is egalitarianism. Below we will see the differences and similarities between prioritarianism and egalitarianism, as well as the debate surrounding egalitarianism.

6.10. Egalitarianism

6.10.1. What characterizes egalitarianism?

Consider the following situations:

Classical utilitarianism would say that S1 and S2 are equally good, since the total balance is the same. Maximin would say that S2 is better, since the minimum level of well-being is higher. Prioritarians would say that S2 is better because the +5 that B gains has more value than the -5 that A loses, since B's situation is much worse. Egalitarians⁶¹ would also say that S2 is better, but for a different reason: in S2 the level of inequality is lower.

In the example we have just seen, the total balance is the same in both situations. However, egalitarians would also say that certain situations with a lower total balance can be better all things considered than others with a greater total balance. Let us look at an example:

Again, according to maximin S3 is better than S1 because the minimum level is higher. Prioritarians would say that S3 is better because, since B is in a worse situation, the +9 B gains is more valuable than the +10 A loses. Egalitarians, in turn, would say that S3 is better because the level of inequality is much lower than in S1.

6.10.2. The leveling down objection (version 1)

One of the main objections to egalitarianism is the leveling down objection. Below we will discuss two versions of this objection. Consider the following situations.

The first version of the objection claims that egalitarianism would have to say that S2 is better than S1, since in S2 there is no inequality, and in S1 there is a little inequality. But, the objection concludes, this is absurd, since in S2 no one is better off than in S1; in fact, everyone ends up in a much worse condition.

Before we see how egalitarians might respond to this objection, it is important to note that it does not affect prioritarianism, maximin, or sufficientarianism (which we will see in section 6.11). Thus, the leveling down objection offers no reasons to reject reductions in inequality when doing so improves the situation of the worst off.

Egalitarians could respond to this first version of the objection by denying that they must claim that S2 is better all things considered than S1. What egalitarians must hold is that S2 is better in a certain respect than S1 (in S2, the total good is better distributed). However, claiming that a situation is better in a certain respect does not imply having to claim that it is necessarily better all things considered (that is, taking into account everything that matters). This is because egalitarians hold that equality matters in itself, but they do not hold that it is the only thing that matters, nor that it always matters more than anything else. Egalitarians would agree that S2 is worse all things considered. They would only say that S2 is better in terms of equality (that is, the total value is better distributed) and that, therefore, one way to improve S1 would be to redistribute the total good as follows: A=+14.5 and B=+14.5.

6.10.3. The leveling down objection (version 2)

The second version of the leveling down objection, for its part, denies that there is any respect in which S2 is better than S1. This second version holds that, for one situation to be better than another in some respect, it must benefit someone. Since in S2, compared with S1, no one is in a better condition (both are in a worse condition), there is no respect in which S2 is better than S1, the objection concludes.

This second version of the objection is usually directed at egalitarianism by prioritarians. Let us look at two replies egalitarians have given to this second version of the objection:

Temkin (2000, p. 137-146) argues that a situation can be better than another in a certain respect without benefiting anyone. For example, if one of the two (or both) deserves the harm they suffer. We will return to the aspect of desert in section 6.14.

Horta (2010a, p. 140-142), for his part, argues that there is a respect in which a situation such as S2 can be said to be, in a certain respect, better for the individuals affected. To defend this thesis, he distinguishes between relational and non-relational dimensions of value. Let us look at an example. Let us imagine that there are two worlds and that the letters below indicate the well-being levels of individuals belonging to two groups in each world.

World Group A Group B W1 +40 +20 W2 +19 +19

If we are born in W1 and in group B, we can say that we were unlucky to be born in group B and not in group A. But we cannot say that we were unlucky to be born in world W1 (given that even the worst off in W1 are still better off than the best off in W2). Now, if we are born in W2, things go the other way around. That is, if we are born in W2, we can say that we were unlucky to be born in W2 and not in W1 (since even the worst off in W1 are still better off than the best off in W2), but we cannot say that we were unlucky to be born in the group we were born into (given the equal distribution in W2).

Thus, there would be two dimensions of value: one would concern the situation someone is in (in the example, the worlds W1 or W2); the other would concern the position someone occupies relative to others, given the situation they are in (in the example, groups A and B). It is in this second dimension that it would make sense to say that S2 (where A=+1, B=+1) is better in a certain respect for the individuals affected than S1 (where A=+14, B=+15): once someone has the bad luck to be in situation S2, there is no risk of being in a worse situation than that of the other individuals who will be in S2 (that is, the total is better distributed).

6.11. Sufficientarianism

Sufficientarianism⁶² has as its goal that all individuals have a sufficient level of well-being. This theory therefore sets a threshold of well-being above which life is sufficiently good⁶³. With regard to those below the threshold, sufficientarianism sets priority according to three criteria: (a) how far each individual is from this threshold; (b) how many individuals are benefited; and (c) the size of the benefit that can be produced⁶⁴. This shows that sufficientarianism has several similarities with prioritarianism and egalitarianism, the main difference being that, with regard to those above the threshold, sufficientarianism sets no priority criterion at all⁶⁵.

Sufficientarianism also has similarities with and differences from the view discussed in section 6.7, based on concern for the number of individuals with positive well-being. The similarity is that both share that same concern. The difference is that someone can have a minimally positive life and still not have enough for it to be good. Thus sufficientarianism, unlike the view discussed in section 6.7, will aim to reduce inequality and to prioritize the worst off when they are below the threshold.

6.12. Negative versions of utilitarianism, prioritarianism, and egalitarianism

Egalitarianism, prioritarianism, and utilitarianism also have negative versions. Below we will compare the negative versions of these theories with their traditional versions.

6.12.1. Negative utilitarianism vs. classical utilitarianism

The goal of negative utilitarianism is to bring about the situation in which there is the lowest aggregate negative total, and, unlike classical utilitarianism, it holds that no amount of positive well-being can compensate for negative well-being. Let us look at an example:

Classical utilitarians would say that S1 is better, since the total balance is greater. Negative utilitarians, by contrast, would say that S2 is better, since in it the negative total is lower.

6.12.2. Negative egalitarianism and prioritarianism vs. negative utilitarianism

Like negative utilitarianism, negative prioritarianism and negative egalitarianism also aim to reduce negative well-being, and also hold that positive well-being cannot compensate for negative well-being. But, unlike negative utilitarianism, for negative egalitarianism and negative prioritarianism what matters is not only the total sum of negative well-being avoided: what also matters is how negative well-being is distributed among individuals. Let us look at an example to illustrate:

Negative utilitarians would say that there are no reasons to prefer S3 or S4, since the negative total is the same. Negative prioritarians and negative egalitarians, by contrast, would say that S4 is better. Negative prioritarians would say that S4 is better because there is a considerable improvement for whoever would be worse off. Negative egalitarians, in turn, would say that S4 is better because in it there is less inequality: in S3 we have one individual with mild pain and another suffering terribly; in S4 the burden is shared, and we have two individuals with more moderate pain.

6.12.3. The debate over negative theories

The central objection to all negative theories points to the implications of their not taking into account what is positive for individuals. For example, if the goal is only to avoid what is negative for individuals, then all these theories seem to imply that it is always better to die (since that eliminates everything negative), even when the other option is to have a wonderful life, but with a single moment of very mild suffering.

One way to respond to this criticism is to acknowledge that it may make sense when one is talking about happiness versus suffering within the same individual, but that this does not affect such a comparison across different individuals. That is, even if what the criticism points out is correct, it still makes sense to hold that no amount of happiness for some individuals can compensate for the suffering of other individuals.

One way of adapting such theories to get around the criticism with regard to assessments of happiness vs. suffering within the same individual is to argue that pursuing what is positive is also important, but that avoiding what is negative is more important without amounting to a trump. In this way, depending on the amounts of happiness and suffering involved, it could be worth going through a certain amount of suffering. How much suffering would be tolerated would obviously depend on how much happiness was at stake, but each variant of this kind could set different weights for avoiding suffering versus promoting happiness. The more weight they give to avoiding what is negative, the closer they will come to the negative versions. The closer the weights given to producing what is positive versus avoiding what is negative, the closer they will come to the classical, non-negative versions of these theories.

6.13. Why all these theories imply equal consideration

All the theories we have seen so far have in common the fact that they aim at a goal. What distinguishes them from one another is the goal they aim at. However, all of them necessarily imply the equal consideration of all sentient beings. The reason is simple: if, in some case, someone gives less weight to the good of a certain sentient being when applying one of these theories, then in that case they will no longer achieve the goal the theory in question aims to achieve. This will be illustrated by the following example.

We will look at an example with three of these theories (utilitarianism, maximin, and egalitarianism), but the same conclusion applies to all the other theories we saw earlier. Consider the following situations:

The different theories would disagree about which of the situations is better (and that is what distinguishes the goals they aim at). For utilitarianism, S1 is better, since the total is greater (+30). According to maximin, by contrast, S2 is better, since the minimum level is higher (+14). Egalitarianism would agree that S2 is better, but because in it the level of inequality is lower (-1).

In what sense would all these theories have to accept equal consideration? In the following sense: if the individuals swapped places, they would have to maintain the same decision:

If someone refuses to defend the same decision if the individuals swap places (for example, because they always want to prioritize the human individual), they will not achieve the goal pursued by the theory in question.

6.14. The aspect of desert

The theories we have seen so far take different aspects into account in determining the right action. However, they all share a common basis: for them, all that matters is how much individuals will be harmed or benefited.

However, as we already saw in section 6.4, one possible objection to this foundation is to argue that the aspect of desert is important. Note that desert is an aspect and not a theory, since there are no theories holding that this is the only relevant aspect. What happens is that this aspect can be added to the theories we saw earlier. Thus, there could be a desert-sensitive egalitarianism, a desert-sensitive prioritarianism, a desert-sensitive utilitarianism, and so on.

The aspect of desert may concern deserving a harm, a benefit, or both. We will look at each of these possibilities below.

As regards deserving harms, the reasoning would be as follows: if someone has done something to deserve the harm they suffer, we should not prioritize eliminating that harm (indeed, we should not even eliminate that harm, since it is deserved), even if the individual who deserves the harm is in a worse condition than others who did nothing to deserve the harm they suffer.

As regards deserving benefits, the reasoning would be as follows: if someone has done something that makes them deserving of a certain benefit, and other individuals could have done the same thing but did not, then we should give that benefit to the deserving individual, even if other individuals (who could have become deserving but did not) need that benefit to a greater degree.

Note that whether an individual is deserving (whether of a harm or of a benefit) depends not only on what they did but, in addition, on their degree of choice. Let us first look at an example concerning deserving a harm. It makes sense to prevent a human baby or a nonhuman animal from hurting other individuals (regardless of whether those they would hurt were human or nonhuman, and of their levels of understanding). However, it makes no sense to punish a human baby or a nonhuman animal for reasons of desert, since that would require them to be morally responsible and able to choose deliberately to act in one way or another. Thus, only responsible agents, such as normal adult humans, can deserve a harm. In light of this, one might think that, then, every time a normal adult human causes harm to innocents, it makes sense to punish them for reasons of desert. However, this is mistaken, since it does not take into account another crucial factor: assessing whether the individual in question was in a position to have avoided doing what they did, or whether this was beyond their control.

Let us now look at an example concerning deserving a benefit. Let us imagine that two individuals are equally able to cooperate on an important task, but only one of them decides to cooperate (while the other decides simply to wait for the first one's hard work so as to reap the fruits later). In this case, it makes sense to say that only the one who cooperated deserves the fruits of the work in question (even if the second needs those fruits more than the first). Note, however, that, again, it makes sense to claim this only because the individual who cooperated understood what was at stake (that is, was a responsible individual) and was fully able to cooperate. It is for this same reason that it makes no sense to claim that human babies or nonhuman animals do not deserve a certain benefit because they did not cooperate to obtain it (since they are not morally responsible agents and, even if they were, they are generally not in a position to cooperate to obtain it).

With regard to the grounds for including the aspect of desert, the different views that include it may regard it as relevant in itself and/or relevant only indirectly⁶⁶. For example, a common view is to hold that reprimanding and punishing individuals who harm others, and praising and rewarding those who benefit others, is important only because doing so encourages individuals to avoid harming and to seek to benefit others. On this position, the aspect of desert is seen as only indirectly relevant to minimizing harms and maximizing benefits. However, another possibility is to regard the aspect of desert as relevant in itself. In this case, it would be argued that it is important to reprimand and punish individuals who harm others, and to praise and reward those who benefit others, because that is what they deserve for the choices they made in the past as rational agents, regardless of whether this encourages them to avoid harming and to

seek to benefit others. A third option is to hold that both grounds are important, that is, that we have reasons to reprimand and punish individuals who harm and to praise or reward those who benefit, both because that is what they deserve given their past choices as rational agents and because doing so encourages them to avoid harming and to seek to benefit others. On this third option, the aspect of desert is seen as being both important in itself and important indirectly when it comes to determining the right action.

6.15. Pluralist possibilities

One of the central points of disagreement among the various theories of normative ethics is the fact that some do not take into account aspects that others consider relevant. The fewer aspects a theory considers relevant, the greater the chances of its being criticized by other theories for not taking into account the aspects the others consider relevant. That is why this kind of criticism is much more commonly directed at monist theories, since these take into account a single aspect. This means that one way of trying to deal with this problem is to take multiple aspects into account when determining the right action. The more aspects a theory takes into account, the less exposed it will be to the charge of leaving out important aspects.

For example, one could construct a pluralist theory that took into account many (or even all) of the aspects we have covered so far in this chapter. With regard to this point, however, there are at least two important difficulties, which we will look at below.

The first concerns the reasons for including a given aspect. A first possibility is for a theory to include many aspects because it really regards each of them as relevant to the morality of the decision. A second possibility is for the theory to include many aspects because of the disagreement among different people as to which aspects are relevant and as to their weight (leaving open whether or not they are really relevant). A third possibility would be for the theory to include many aspects because of the high degree of uncertainty as to which of them are relevant.

The second difficulty concerns the work of reasoning on the basis of many aspects. The more aspects a theory includes, the greater the possibilities of conflict among the various aspects, and thus the greater the difficulty of assessing what the right decision is (since it would involve assessing which aspect weighs more in each circumstance and why). In this sense, proponents of monist theories could say that their advantage is that, since they work with a single aspect, they do not face this problem.

We can sum up the issue as follows. The fewer aspects a theory includes, the greater its advantage in terms of theoretical simplicity, but the greater the possibilities of accusing it of leaving out important aspects. On the other hand, the more aspects a theory includes, the smaller the possibilities of accusing it of leaving out important aspects, but the more complicated the reasoning based on it will be.

6.16. Approaches not characterized by a goal

All the approaches we have seen so far are characterized by pursuing a goal, in the sense of bringing about certain states of affairs in the world. As we have seen, the various theories we have covered are distinguished from one another because they aim at different states of affairs.

The theories we will see in sections 6.17, 6.18, and 6.19, by contrast, are characterized much more by the kind of procedure they prescribe for decision-making than by the state of affairs aimed at by the decision. The theories we will see in section 6.20, in turn, are characterized by prescribing the kinds of character traits we should cultivate.

Of course, in a broad sense, it could be said that all these theories also have a goal (whether following certain procedures or developing certain character traits). However, here they have been grouped separately from the approaches we have seen so far because, while those we have already seen aimed at states of affairs in the world (maximizing well-being, reducing suffering, reducing inequality, prioritizing the worst off, treating individuals according to what they deserve, etc.), the approaches we will see below are founded on factors that concern either the way agents make decisions or agents' character traits.

In light of this, one might think that the approaches we have seen so far are necessarily consequentialist, and that those we will see below will be, respectively, deontological and character-centered. This will be so in most cases, but not necessarily, as explained below.

The approaches we have seen so far are best known for their consequentialist variants, but there are (or it is always possible to construct) deontological variants of them. Take, for example, the approaches whose goal is to reduce inequality. Let us imagine that, if we do not cause a great inequality now, the complete history of the world will contain more inequality than it otherwise would. A consequentialist egalitarianism would prescribe causing that inequality now. A deontological egalitarianism, by contrast, could prescribe not causing that inequality now (however much more inequality there would be as a result of our deciding not to cause inequality now—if it regarded the prescription “do not cause inequality” as an absolute trump; or only if the resulting inequality were below a certain point—if it regarded the prescription “do not cause inequality” as a non-absolute trump). The same holds for any other aspect we have seen so far.

As for the approaches we will see below, those we will see in sections 6.17, 6.18, and 6.19 (respectively, Kantianism, discourse ethics, and contractualism) are usually classified as deontological (however, as we will see, some interpret contractualism as not necessarily implying a deontological approach). The approaches we will see in section 6.20 (virtue ethics and the ethics of care), in turn, are usually understood as character-centered ethics.

6.17. Kantianism

Immanuel Kant (1724-1804) characterized the categorical imperative as a rational principle that we must always follow, regardless of our inclinations to the contrary. According to Kant, it is from this rational requirement that all moral duties derive⁶⁷.

Kant formulated the categorical imperative in several ways. For what we are going to discuss, the first and second formulations are the important ones. The first states that we should act in such a way that we can rationally will that the maxim (that is, the principle) behind

our action become a universal law. That is, it is right to do something only if we can rationally will that everyone else do it too (and on the basis of the same maxim). The second, which Kant believed to be derived from the first, prescribes treating all rational beings as ends in themselves, and never as mere means.

Kant believed that, because nonhuman animals are not capable of applying the categorical imperative, our duties toward them would be merely indirect duties toward humans. That is, according to Kant, the wrong in harming an animal would lie in the potential for harming humans (whether because it would be practice for harming other humans, or because one would harm oneself through brutalization), and not in the harm done to the animal itself.

Contemporary Kantians such as Korsgaard⁶⁸ and Franklin⁶⁹ have defended Kant's approach but disagreed with the idea that only rational beings should be treated as ends in themselves. They have argued that, if we were in an animal's place, we would not consider it acceptable to be disregarded as they are (first formulation) and that, therefore, all sentient beings (and not only rational beings) should be considered ends in themselves, even if they are not capable of understanding and applying the categorical imperative (second formulation, reformulated).

Similarly to what the argument from relevance points out, Franklin argued that Kant confuses the subjects of the categorical imperative (who must be rational) with the objects to which it applies (which need not be rational, only sentient). He reinterpreted the second formulation of the categorical imperative as follows: “act in such a way that you always treat sentience, whether in your own self or in that of any other, never simply as a means, but always at the same time as an end⁷⁰.” For Franklin, rather than Kant's formulation, it is this one that is truly connected with the first formulation.

Other authors have also developed theories based on the Kantian position. For example, Gewirth⁷¹ argued that all agents, by the mere fact of acting, are assuming that they have the right to act, as well as other rights necessary for acting (such as those necessary for survival), and that, to be consistent, we should respect the rights

of others. This idea was developed by Pluhar⁷², who argued that the same should apply to our actions that affect nonhuman sentient beings, since they too have the interests and needs that we must have in order to be protected by rights.

6.18. Discourse ethics

Discourse ethics holds that we should follow only principles that would be mutually acceptable to all those affected by the consequences of their application, and that such principles can be defined only through public debate, in which all the arguments for and against the different points of view must be examined and compared⁷³.

One might think that, because nonhuman animals do not have discursive competence, discourse ethics would exclude them from moral consideration. However, we should not confuse the following essential categories of discourse ethics⁷⁴: (1) moral agents are those individuals capable of acting morally; (2) moral debaters are the individuals capable of taking part in moral discourse; and (3) moral subjects are the individuals who, regardless of whether or not they are moral agents or debaters, can be harmed or benefited⁷⁵.

For the points of view of all those affected by the decision to be considered, the discussion must involve as many participants as possible. According to discourse ethics, when this is not possible, the participants should accept only the norms that those affected by them would agree to. This implies considering not only those who cannot take part in the debate for logistical reasons, but also those who cannot because they lack discursive competence (such as babies, humans with cognitive impairments, and nonhuman animals⁷⁶).

6.19. Contractualism

The question underlying contractualism is this: “How can we reach agreement on the norms we should follow?” According to contractualism, the norms we should accept are those that would be approved by self-interested and rational agents in a hypothetical contract. However, as we will see below, although there are similarities between the classical and contemporary versions of contractualist theory, there are disagreements about the conditions under which the contract would have to take place for the resulting norms to be legitimate.

Classical contractualism

In Hobbes's⁷⁷ classical contractualist theory, the central idea was that life in the state of nature was usually brutal and short, whether because of scarcity of resources and disease or because of vulnerability to attack by others. Hobbes characterized this state as “the war of all against all.” He argued that, to escape this situation, rational and self-interested agents would voluntarily join together to adopt a system in which each would give up the power to harm others (and might even agree to help others), but only if everyone else did the same. In this way, each agent's chances of maximizing their own interests would increase. That is, according to contractualism, the best way for each agent to maximize the fulfillment of their interests is not to act in a purely self-interested way, but as if there were a hypothetical contract in which each agent would take the interests of others into account.

However, in the system Hobbes envisioned, there is no device to ensure impartiality in the hypothetical contract. This allows the contracting parties to approve biased principles. For example, if the only reasons for agents to agree to take part in the contract are self-interest and the perception that in the state of nature they would be vulnerable to more powerful agents, then the most powerful agents of all have no reason to respect weaker agents, since the latter pose no threat to them. In response, it might be objected that the most powerful agents always have reasons to accept the contract, since weaker agents can band together and, collectively, have enough power to harm the most powerful. However, there are many situations in which powerful agents would realize that the weaker agents would very probably be unable to organize, or, even if they could, would still be unable to pose a threat. In such situations,

Hobbesian contractualism would hold that the most powerful agents would have no reason at all to respect weaker agents. Moreover, in any and every situation, weaker agents could end up having to accept norms that unfairly favored the stronger, since they would realize that they could end up in an even worse situation if they did not accept them.

Contemporary contractualism

In order to get around this central problem of Hobbesian contractualism, contemporary contractualism, influenced mainly by the work of Rawls⁷⁸, incorporates a concern with impartiality. This method consists, again, in imagining rational agents who aim to maximize their own well-being, but who have not yet been born into the world in which they will live. This is the so-called original position. In it, the agents must decide which moral and political principles will be in force in the future world in which they will live. But they must decide behind a veil of ignorance, under which they do not know what characteristics they will have after they are born. They will not know their race, gender, nationality, whether they will be born rich or poor, healthy or with some illness, whether they will have some physical or psychological limitation, what abilities or difficulties they will have, etc. All of this will be determined by lot—thus, there are equal probabilities of being born as any one of the individuals. The central idea is that whatever rational and self-interested agents would consider just/unjust under such conditions would in fact be just/unjust.

What kind of theory is contractualism?

Contractualism is usually interpreted as a deontological theory. On this interpretation, this would be so because the rules we should follow are those that would be approved by the agents in the original position, even if this does not result in the best consequences. However, some disagree with this interpretation. For example, Harsanyi⁷⁹, in 1955 (that is, 16 years before the publication of Rawls's work), developed a method very similar to the veil of ignorance. However, according to Harsanyi, the principles that agents would approve under conditions of impartiality would not be deontological, but consequentialist.

Noting the compatibility of contractualism with normative approaches from various schools at the factoral level, authors such as Kagan⁸⁰ have interpreted contractualism as a theory at the foundational level, and not at the factoral level. On this interpretation, contractualism is more a theory about why to accept certain principles than a theory about which principles to accept. In this sense, the contractualist grounding would be a way of explaining why there are better reasons to accept a given view of normative ethics at the factoral level (whether deontological, consequentialist, or character-centered): because it would be the view approved under conditions of impartiality (this would be the foundational level).

Contractualism, nonhuman animals, and humans with cognitive impairments

Rawls believed that the veil-of-ignorance method was not applicable to our decisions that affect animals, since he believed that, for it to make sense to protect someone with principles of justice, that someone must also have a sense of justice (since this is necessary in order to decide in the original position)⁸¹. According to Rawls, the same holds for humans who lack a sense of justice (as is the case with babies, children up to a certain age, and even adults who, because of illness, accident, or a congenital condition, do not have cognitive capacities developed to the point of understanding principles of justice). According to Rawls, duties toward nonhuman animals and toward these humans would be merely indirect duties toward the humans who do have a sense of justice (for example, the duty not to kill a child would be an indirect duty toward the child's parents).

One might think that this view is implied by Rawls's contractualism because of the very essence of contractualism, which consists in imagining self-interested agents. In other words, if the ground for not harming others is self-interest, then we have no reasons to recognize direct duties toward beings incapable of taking part in this bargain with us. However, Rawlsian contractualism introduces a condition of impartiality, through the device of the veil of ignorance, and this is introduced as a matter of justice, to prevent biased attitudes (it is therefore not founded on the agents' self-interest).

In response, it might be objected that each agent does have self-interested reasons to accept the introduction of the veil of ignorance, since it protects each agent from the biased attitudes of other agents. However, the device of the veil of ignorance is introduced precisely to avoid the implication, present in Hobbesian contractualism, that the most powerful agents would have no reasons to respect those incapable of posing a threat to them. Therefore, the most powerful agents of all have no self-interested reasons to accept the introduction of the veil of ignorance. That being so, if Rawlsian contractualism itself introduces a key element that is founded not on self-interest but on ensuring justice (the veil of ignorance), it does not seem to make sense to hold that nonhuman animals and humans with cognitive impairments should be excluded from the benefits of justice simply because there is no advantage for normal adult humans in respecting them.

However, it might be objected that nonhuman animals and humans with cognitive impairments should be excluded from the benefits of justice not because there is no advantage for normal adult humans in respecting them, but because they lack a sense of justice. On this view, for someone to be protected by principles of justice, they must have a sense of justice. This reasoning would make sense if, in order to be treated unjustly, one had to have a sense of justice. But imagine, for example, that a baby receives less food than the other babies in a nursery because of the color of its skin. The fact that the baby lacks a sense of justice does not make the treatment it received any less unjust: its well-being was given less weight on the basis of an arbitrary factor (skin color).

In response, it might be objected that in this case the treatment the baby received was unjust because one day, when it is an adult, it will develop a sense of justice. One problem with this reply is that the injustice of the treatment in question comes from the arbitrary criterion chosen (skin color), and this is independent of whether or not the baby in question has the potential to develop a sense of justice someday. For example, if the baby in question had been born with a condition that prevented it from ever developing a sense of justice throughout its life, giving it less food would remain just as unjust. In short: for someone to be held responsible if they commit an injustice, they must have a sense of justice; however, it is not necessary to have a sense of justice in order to suffer an injustice. Thus, the lack of a sense of justice in nonhuman animals does not make the disregard they typically receive just.

Another defense of the claim that the veil of ignorance is compatible with excluding nonhuman animals was made by Carruthers⁸². He argued that we have no reasons to avoid harming animals because, if agents are applying the veil of ignorance, they can assume they will be human after the veil is lifted, since animals are not capable of applying this procedure. However, as VanDeVeer⁸³ pointed out, the agents in the original position also know that they understand very complex moral and political principles, but this does not imply that in the original position they can assume they will retain this high level of rationality after the veil is lifted, since, if they could, they would be tempted to favor the most rational individuals. For the same reason, in the original position they should not know which species they belong to.

This idea was explored and developed by Rowlands⁸⁴. He observed that one of the purposes of the veil of ignorance is to rule out undeserved advantages and disadvantages, which are the result of luck: that is why the veil of ignorance excludes knowledge of the natural properties one would have. However, Rowlands observes, membership in a species, like all other natural properties, is an undeserved result of the natural lottery, and, therefore, the benefits and harms that result from it are also undeserved. For this reason, knowledge of which species we will belong to must be excluded in the original position: we must imagine what we would decide if we did not know whether we would be born as humans or as nonhuman animals.

If in the original position we must imagine rational and self-interested agents, the greater the probability of their being assigned by lot to be nonhuman animals after the veil is lifted, the more strongly they would defend the equal consideration of all sentient beings (since, being rational and self-interested, they would not run the risk of being disadvantaged by a system that privileged humans). In the real world, the probability of their finding out, after the veil is lifted, that they would be nonhuman animals is gigantic, since the number of nonhuman animals is vastly greater than that of humans⁸⁵. This suggests that the people who defend speciesism do so only because they know they belong to the

privileged species. If so, then speciesism is not just, since it does not pass the impartiality test.

It might be objected that the device of the veil of ignorance is compatible with defending privileging humans. This would be so because the agents in the original position, even knowing that the probability of being human after the veil is lifted would be extremely remote, might want to take their chances. However, this objection does not hold up because in the original position what we must imagine is what self-interested and rational agents would decide. Since it is not rational to make a bet of this kind (given that failure is almost certain), rational agents would not make it. In fact, knowing the very high probability that they would turn out to be nonhuman animals after the veil is lifted, what they would have reasons to do is to defend prioritizing nonhuman animals. We will see more about this below.

Rational and self-interested agents would have at least two reasons, in the original position, to defend giving priority to nonhuman animals. The first is that, as we have seen, since the number of nonhuman animals is gigantically greater than that of humans, the probability of each agent being a nonhuman animal after the veil is lifted is also gigantic. The second is that the vast majority of nonhuman animals are born either under the regime of exploitation or in the wild. In both situations, their lives usually contain great amounts of suffering, as well as a premature death. On average, the vast majority of nonhuman animals are in a far worse situation than the vast majority of humans. In short, nonhuman animals are usually in the worst situation. These two factors together—the high probability that the contracting parties in the original position will find out that they will be nonhuman animals, and the fact that nonhuman animals are generally in a worse condition than humans—give the contracting parties in the original position strong reasons to defend not only the equal consideration of all sentient beings, but also the priority of improving the situation of nonhuman animals.

Thus, far from there being contractualist reasons to give only indirect moral consideration to nonhuman animals, there seem to be contractualist reasons to give them not only direct moral consideration, but also equal consideration, and, moreover, to defend prioritizing the improvement of their situation.

6.20. Character-centered ethics: virtue ethics and the ethics of care

Character-centered ethics holds that we should act as someone with good moral character would act. Two examples of character-centered ethics are virtue ethics and the ethics of care.

Virtue ethics holds that we should act as someone virtuous would act. Any plausible virtue ethics will regard cruelty and indifference to suffering and death as vices, and will regard compassion, kindness, and sensitivity as virtues. For this reason, it could be argued that the typical treatment of animals displays serious vices of character⁸⁶. Moreover, since insensitivity is not virtuous, virtuous action is not merely avoiding causing harm, but also trying to help whenever possible.

The ethics of care, for its part, holds that we should have a caring character, which implies not only avoiding harming others, but also helping, especially those who are vulnerable and dependent on us. However, some proponents of the ethics of care have rejected equal consideration because it implies giving the same weight to those with whom we do not have a close relationship⁸⁷. In the view of these authors, equal consideration does not give enough importance to our emotions (which are stronger toward those with whom we have a close relationship).

One way to respond to this position is to point out that prioritizing those with whom we have personal relationships does not imply disregarding, or failing to help, those with whom we have no such relationships, since we will not be caring agents if we do not care about those who suffer. If we were to make personal relationships the only basis of care, we would have to neglect even most humans. Therefore, even in an ethics of care that privileged the individuals with whom we have close relationships, we should still try to avoid harming, and to help, those with whom we have no such relationships.

Although the reply above makes an important point, there is also another way to respond: to insist that giving greater weight to those with whom we have a close relationship is a biased and agent-centered attitude (since how much care someone needs is not something

that depends on the relationship the patient has with the agent). For example, if someone we do not know is at serious risk of death, and someone with whom we have a close relationship has a mild pain, it seems that in this case we have reasons to prioritize helping the person we do not know. In this case, the attitude of privileging those with whom we have a close relationship could be accused of being founded on a concern for oneself, and not on a sincere concern for those who need care.

Moreover, the criterion of close relationship is not necessarily implied by the ideal of being caring agents. It is also possible to defend care in impartial terms, rejecting the criterion of close relationship⁸⁸. That is, although the ethics of care is often presented as an ethics that gives weight to close relationships, this is not a necessary implication of the ethics of care.

Thus, even though the ethics of care (if combined with the relationship criterion) does not imply the equal consideration of all sentient beings, it is compatible with equal consideration (provided that only the ideal of care is retained and the relationship criterion is discarded). Moreover, as we have seen, even an ethics of care that gives greater weight to those with whom we have a close relationship still implies that we should avoid harming and also help those who need care (even those with whom we do not have a close relationship), especially the most vulnerable, as is the case with nonhuman animals.

6.21. Implications for animal exploitation

Given what we have seen above, there are both general arguments and arguments specific to each theory of normative ethics for giving direct moral consideration to nonhuman animals, that is, for considering their good as something important in itself.

Moreover, as we have seen, there are also general arguments for giving equal consideration to all sentient beings, and the vast majority of theories of normative ethics necessarily imply equal consideration. And even those that do not necessarily imply it are still compatible with equal consideration. Finally, we have seen that even the theories that tolerate giving different degrees of moral consideration (such as, for example, giving greater

consideration to the individuals with whom we have close relationships) still imply that we should avoid harming, and also help, those who would receive less consideration.

What are the practical implications of all this? A first, very clear implication is that there are strong arguments (both general and theory-specific) for abolishing animal exploitation⁸⁹. That is, the goal should not be to regulate animal exploitation so that animals suffer less: animal exploitation itself is unjustifiable. It is unjustifiable for several reasons. The first, and most evident, is that the harms it causes (gigantic amounts of suffering and death) are gigantically greater than the benefits it produces. The second is that it is biased, since it chooses who will be harmed and who will be benefited on the basis of the species to which they belong (a characteristic that, like race and gender, is a result of the natural lottery, and not of merit or demerit). The third is that it would never be approved under conditions of impartiality (that is, if we did not know whether we would be its victims or would reap its benefits), and, therefore, it violates the most fundamental requirement of justice. The fourth is that it treats sentient beings as mere means to an end.

The objection that utilitarianism is incompatible with abolitionism

One possible objection to the conclusion that all theories of normative ethics have reasons to abolish animal exploitation is to argue that utilitarianism, despite implying equal consideration, is incompatible with abolishing animal exploitation. This objection is raised both by some of its proponents⁹⁰ and by some of its critics⁹¹. This is because utilitarianism requires harming individuals if that is the most efficient way to maximize total well-being (or to minimize total suffering, in the case of negative utilitarianism).

To examine this objection, it is important to note that the utilitarian calculation could be used to assess the institution of animal exploitation as a whole or each individual case of animal exploitation. We will discuss both possibilities below.

As for animal exploitation as a whole, it clearly produces far more harms than benefits. The harm suffered by each animal (years lost through death and a life full of suffering) is much greater than the benefit each human obtains from its exploitation. Moreover, the number of individuals harmed is enormously greater than the number benefited. Thus, the aggregate total balance of animal exploitation is vastly negative.

It might be objected that this makes sense for forms of utilitarianism focused on assessing institutions and general rules, but that those focused on assessing each act could argue that, perhaps in one case or another, exploiting animals would be what best achieves the utilitarian goal. Moreover, it could be argued that the total balance could be better maximized if animal exploitation were reformed so that animals have positive lives, instead of being abolished (since there would then be more animals being born to have positive lives over time, even if they were killed afterward)⁹². But is this really the most efficient way to achieve the utilitarian goal? We will discuss this point below.

Since utilitarianism is a consequentialist theory, in order to assess which option produces the best consequences it must also take into account the influence each decision would have on other agents' decisions from then on. Taking this into account, it seems that advocating the abolition of the use of animals has much greater chances of better consequences over time than advocating a reform of animal exploitation. This is so for two reasons:

(1) Even if, according to act utilitarianism, it is right to kill some to save others in very specific contexts (and this is independent of the species of those harmed and benefited), given the prevalence of speciesism, if this were advocated, the public would understand that it is acceptable to kill nonhuman animals to benefit humans (that is, it would make the judgment on the basis of species), and in contexts in which utilitarianism itself would not approve.

(2) Abolition, by spreading consideration for sentient beings, could motivate people to try to make the world the best it can be for all sentient beings from then on. Advocating the use of animals, on the other hand, reinforces the idea that animals are

mere resources, making it quite unlikely that it would lead people to aim for the best possible world for all sentient beings.

Thus, there are also utilitarian reasons (whether rule or act) for the abolition of animal exploitation.

Above we have seen that, regardless of which theory of normative ethics one finds most plausible, there are strong reasons to advocate the abolition of animal exploitation. This is because, as we have seen, there are both general arguments and arguments specific to each theory of normative ethics for advocating the abolition of animal exploitation.

Below we will look at the practical implications of everything we have discussed so far, but with regard to our decisions that affect animals outside human control, such as animals living in the wild.

6.22. Implications for the situation of wild animals

Programs dealing with wild animals are usually guided by environmentalist values. According to these conceptions, animals matter only as a means to maintaining ecosystems or preserving species⁹³. However, if the argument from relevance is correct, sentient beings matter in themselves.

What would an approach that respected wild animals as sentient beings look like? First, the morality of actions would have to be assessed on the basis of how much they harm or benefit them, regardless of whether or not they realize environmentalist values. This change in attitude would not be limited to preventing human practices that harm them: it would include preventing the harms they suffer because of natural processes⁹⁴.

Animals living in the wild are harmed daily by factors such as hunger, thirst, disease, natural disasters, hostile weather conditions, physical injuries, conflicts, etc.⁹⁵ Moreover, an enormous amount of suffering and extremely high rates of premature

mortality result from the predominant reproductive strategy, which consists in having thousands or even millions of offspring per brood (in stable populations, on average only two per brood survive)⁹⁶. Some things are already being done, on a case-by-case basis, to prevent these harms⁹⁷, but much more could be done⁹⁸.

The general argument for the importance of improving the situation of wild animals who are victims of natural processes is as follows: limiting our concern to the harms caused by human practices is incompatible with truly caring about the good of animals (in fact, it reveals a concern for oneself, with “keeping one's hands clean,” while letting the harm happen). After all, for animals it does not matter whether what harms them originated in human practices or in natural processes: what matters is being free from what harms them.

In addition to this general argument, there are specific reasons, arising from each theory of normative ethics, for regarding this issue as very important. The main specific reasons are the following. (1) Number of victims. The number of victims is so gigantic that it makes even the numbers of animal exploitation nearly vanish by comparison⁹⁹. (2) Suffering per victim. The vast majority of animals born have lives in which suffering greatly predominates (many of them experience only suffering during their lives, and nothing else). These aspects are necessarily considered relevant by utilitarianism (whether classical or average), by the view concerned with the number of individuals with positive lives, by maximin, by prioritarianism, by egalitarianism, by sufficientarianism, by the negative versions of utilitarianism, prioritarianism, and egalitarianism, and by any pluralist combination of these theories. However, as we will see below, they may be considered relevant by several other theories as well.

It might be objected that preventing natural harms is compatible only with consequentialist theories, given their goal of making the world a less bad place. However, this is mistaken, since such a goal is entirely compatible with non-consequentialist approaches. For example, deontological approaches could recognize a general duty to help those in need, regardless of the origin of the harm and of how

the consequences unfold. The difference from consequentialist approaches is that, in deontological approaches, this duty would probably be subordinate to some constraint relative to the agent and to the time of the decision (for example, not violating rights, not harming innocents, etc.). Concern with the issue of natural harms also follows from contractualist approaches, since, behind a veil of ignorance, the contracting parties would very probably argue that resources should be devoted to helping this gigantic number of victims (and perhaps to prioritizing them, given their numbers and the severity of each victim's situation). Finally, helping only victims of anthropogenic harms seems incompatible with a virtuous or caring character. In short, it is false that this concern follows only from consequentialist approaches.

It might be objected that, according to Kantian approaches, we should give less weight to this issue, since it is an imperfect duty¹⁰⁰, as it is not possible to help everyone in need. According to this objection, since it is not possible to help everyone in need, each person would be free to choose to help whomever they pleased and, therefore, no one would have a duty to concern themselves with the situation of animals in the wild¹⁰¹. However, the fact that it is impossible to fulfill all duties of a given type does not mean that there are no criteria for choosing which cases to prioritize. For example, we could say that the situation of wild animals should receive great attention because it involves an enormous number of victims, far greater than all situations of other kinds combined. In short, the fact that it is not possible to fulfill all duties of a given type does not imply that it is morally optional which cases of that type to fulfill, since there may be priority criteria.

Thus, the proposal to improve the situation of wild animals who are victims of natural processes is grounded both in general reasons whose acceptance does not depend on the prior adoption of a specific view of normative ethics and in specific reasons arising from the various theories of normative ethics. In fact, one need only imagine humans who are victims of natural harms to confirm that people, regardless of which ethical theory they believe most plausible, usually recognize that we should do something to help them (or at least are not opposed to helping them). Thus, rejection of the proposal to prevent and minimize the harms that animals in the wild suffer as

a result of natural processes does not arise from someone's adopting this or that school of normative ethics, but from their being speciesist.

6.23. Conclusion

One possible strategy for defending the moral consideration of all sentient beings, regardless of species, is to defend a specific ethical theory and show that it offers reasons to ground such consideration. Indeed, as we have seen, several authors have done this from the standpoint of the most varied theories. However, one risk of this approach is to give the impression that those who believe other theories are better do not need to consider all sentient beings. Thus, it seems more promising to emphasize that there are strong reasons to accept the full moral consideration of all sentient beings, whatever view of ethics we consider most plausible. This is because, as we have seen, there are both general arguments and arguments specific to each theory of normative ethics for defending this consideration. This is a minimal degree of convergence among the various approaches to normative ethics.

Good practical evidence of this minimal degree of convergence is the Montreal Declaration on Animal Exploitation, published on October 4, 2022. It was signed by more than 500 philosophers specializing in ethics and political philosophy from more than 40 countries, and it holds that the current use of nonhuman animals as resources is morally reprehensible and should be abolished. The declaration opens with the following passage:

We are researchers in moral and political philosophy. Our work is rooted in different philosophical traditions, and we rarely agree with one another on anything. We do agree, however, on the need for a profound transformation in our relationships with other animals. We condemn the practices that involve treating animals as objects or commodities.

Should a similar declaration also be made about our having strong reasons to try to improve the situation of wild animals, regardless of whether they are victims of human practices or of natural processes? Given what we have seen in this chapter, everything indicates that it should.