What Can a Constructivist Say About Animal Ethics — Or Any Other Normative Question, for That Matter?

Abstract

In Fellow Creatures, Christine Korsgaard claims that human beings ought to treat all sentient animals as ends in themselves. However, in this article, I argue that Korgaard's method goes beyond what a coherent constructivist conception allows, and I claim that we should therefore adopt a Humean rather than a Kantian version of constructivism. I believe that such a conception permits us to hold substantial ethical positions about non-human animals without having to compromise our ontological commitments.

Résumé

Dans Fellow creatures, Christine Korsgaard soutient que les êtres humains auraient l'obligation morale de traiter les animaux sensibles comme des fins en soi. Cependant, cet article tente de démontrer que la méthode korsgaardienne dépasse ce que permet une théorie constructiviste conséquente et soutient que nous devrions opter pour une version humienne plutôt que kantienne du constructivisme. Selon moi, une telle conception permet tout à fait de soutenir des positions éthiques substantielles sur la question animale sans avoir à compromettre ses engagements ontologiques.

Introduction

In her most recent book Fellow Creatures: Our Obligations to the Other Animals, Christine Korsgaard presents in a unified way her arguments for non-human animal rights. The author, best known for her constructivist interpretation of Kant's ethics, argues, contrary to classical conceptions of Kantian philosophy, that human beings have a moral obligation to treat sentient animals as ends in themselves (Korsgaard, 2018, p. xi). Although in Fellow Creatures Korsgaard reiterates her commitment to the constructivist thesis that everything has value for someone (see Bagnoli, 2017, Section 7.2; Desmons, 2018; Korsgaard, 1996, Lecture 3; Street, 2010, Section 3), she nevertheless sets out to demonstrate that such a position would categorically imply that every sentient creature should be granted a right to what is good for them (Korsgaard, 2018, pp. 10–11). Thus, she presents an attractive position for those who want to avoid the metaphysical implications of moral realism while seeking to directly justify our moral obligations to non-human animals. It is therefore a work with a scope simultaneously metaethical and normative. Thus, the aim of Fellow Creatures is twofold: first, to defend a Kantian constructivist approach to normativity; second, to argue that such a conception categorically implies substantial moral obligations towards non-human animals.

In my view, however, a critical analysis of Korsgaard's approach raises significant issues. For example, for an author such as Sharon Street, the Korsgaardian method can be interpreted as exceeding what a consistent constructivist theory allows. For Street, Korsgaard goes too far when she argues that the practical point of view of any moral agent implies substantial and categorical obligations. Although Street's criticisms are directed at earlier work, in this article, I will attempt to adapt her arguments to demonstrate that the approach employed by Korsgaard in her most recent book still implies favourable presuppositions about the intrinsic value of certain normative principles, namely that a creature for whom things can be good or bad must value itself as an end-in-itself, and that the functional good of these sentient creatures must have a categorical priority over our ends. After having exposed the theses of Fellow Creatures in Section 1, my critique will then examine, in Section 2, the aspect of her argument that can be described as “metaethical” — that is to say, regarding the possibility conditions of our normative evaluations on the animal question — and not its applied elements. In Section 3, however, a new problem will have to be tackled in light of the criticism I have made. If it is true that Korsgaard's thought could be considered to go beyond what a consistent constructivist theory allows, but that we agree with her thesis of “tethered values” (Korsgaard, 2018, p. 9), then it seems that we should opt for a Humean version of constructivism, as Street proposes. There could nevertheless be an important objection, namely that we would go against the purpose of Korsgaard's approach and expose ourselves to the impossibility of defending direct moral duties towards non-human animals — an objection that I will present in the form of the perfectly coherent Caligula argument. I will conclude, however, that this objection is misguided because it is based on a misunderstanding of the normative implications of a Humean constructivism, since its proponents are rather quite capable of defending substantial ethical positions on the animal question — or any other normative question, for that matter.

Ultimately, in the same way that Korsgaard's work pursues two main objectives, the aim of this article is also twofold. First, I intend to criticize Korsgaard's Kantian constructivism by arguing that her approach exceeds what a consistent constructivist approach to normativity allows — and this, beyond its extension to the animal question. Second, I will attempt to demonstrate that a coherent constructivist approach does not have to commit to the existence of categorical and universal moral obligations in order to justify substantive moral positions —such as a commitment to the value of non-human animals. The end of this article is therefore not normative in nature, in the sense that I will not attempt to defend a particular position on the question of animal rights, but rather address the metaethical question of the modality of the justifications of our substantive moral positions in the context of current constructivist theories.

1. A Kantian Defence of the Intrinsic Value of Animals

Although Fellow Creatures has a total of 12 chapters in which Korsgaard addresses a significant number of interrelated topics and issues, she acknowledges that her position on the animal question develops from two main ideas (Korsgaard, 2018, pp. xi and 156). First, in Chapters 1 and 2, she develops her understanding of what value is and demonstrates how it implies that it is not possible to say that humans are more important (in absolute) than animals. Then, in Chapter 8, she argues that it is possible to interpret the Kantian argument for the value of humanity as supporting the idea that animals are ends in themselves worthy of moral considerations.

1.1. Against the Egocentric Predicament

Korsgaard takes as the starting point for her reflection an intuition she considers fundamental among opponents of animal rights, namely the idea that humans are simply more important than non-human animals.¹ For Korsgaard, such a conception is probably due to a misunderstanding of what value is as such. According to her, we must recognize that all value is necessarily tethered to the point of view of a sentient creature, that is, something is ultimately always important to or for someone.² It should be noted that Korsgaard is not falling into value relativism or even denying that something can be “absolutely” important. Rather, she simply agrees with the constructivist principle that no value can be independent of the practical point of view of an individual for whom things can be good or bad. Therefore, she argues that it would be virtually impossible for us to reject the fact that non-human animals are creatures with a point of view from which things can be evaluated as good or bad, given that they experience their own condition, that they feel pleasures and pains that they seek or avoid, and consequently that they have a “valenced” experience of the world. According to her, it is absurd to assert that humans would be more important than animals in absolute terms, because it would then be necessary to ask: “more important for whom?”³ This implies that, for Korsgaard, one cannot establish an “absolute ranking” of the value of humans and the other animals, because what is important from an animal point of view does not have to be compared to what is important for a human, unless one falls into an “egocentric predicament” where it would simply be presupposed that the human point of view has greater value.

However, it is still possible, according to Korsgaard, to assert that something is “absolutely important” or “absolutely good” to the extent that it could be demonstrated that it would be important, or good, for all of us, that is, for any individual for whom things can be good (Korsgaard, 2018, p. 10). Borrowing the notion of “functionality of living organisms” from the Aristotelian tradition, Korsgaard develops a conception according to which the function —the end — of any living creature would be the preservation of her form and the pursuit of the well-functioning of her organism (Korsgaard, 2018, p. 19).⁴ Thus, since it can be considered that the end of any action taken by an organism would be the proper functioning of that organism, and since it has been established that an animal is a creature for which things can be good or bad and which can “seek” her own good, it appears that this “well-functioning” becomes something good in itself for her. Based on this principle, Korsgaard draws a distinction between functional and final good: a functional good is anything that can be evaluated in terms of “good” or “bad” on the part of a creature, whereas a final good would rather, in a teleological sense, be what would be for this creature worthy of pursuit in itself (for its own sake) (Korsgaard, 2018, p. 17). Anything contributing to her well-functioning is, therefore, according to Korsgaard, a final good for any entity experiencing her own functional condition, thus meeting the requirements of what can be considered absolutely good. Subsequently, Korsgaard argues that “it is absolutely good, good-for us all, that every sentient creature get the things that are good-for her, and avoid the things that are bad-for her” (Korsgaard, 2018, p. 10).

1.2. Animals Are Ends in Themselves

Based on what has just been established, Korsgaard undertakes in Chapter 8 to demonstrate that, contrary to what is generally supported by classical interpretations of Immanuel Kant's philosophy, one cannot be satisfied with the idea that rationality is a property necessary for the recognition of a creature's intrinsic value (Korsgaard, 2018, p. 132). In fact, Kantian ethics can still be compatible with the recognition of direct moral obligations towards animals. Suffice it to say, Korsgaard undeniably agrees with the fundamental idea that our obligations to other individuals are justified by the fact that we must recognize them as ends in themselves (Kant, 1994, p. 108). However, she distinguishes between two senses of that concept. First, she identifies an active sense, possessed by rational human beings in that they are able, through practical reason, to legislate for themselves and others and thus place themselves under a mutual obligation to respect each other's autonomy (Korsgaard, 2018, Sections 7.3 and 7.4). Second, she distinguishes a passive meaning to the extent that individuals must be considered an end in themselves if we are compelled to treat their ends, or at least what is good for them, as absolutely good — that is, good for any individual for whom things can be good (Korsgaard, 2018, p. 141).⁵ While conceding that for Kant we possess the status of end in ourselves in the passive sense by virtue of the active sense, Korsgaard instead considers these two conditions as sufficient and not necessary to obtain such a status (Korsgaard, 2018, Section 8.5).

For Korsgaard, all sentient creatures therefore possess this status of end in themselves in the passive sense since, as I have just suggested, as creatures possessing a final good, humans and non-human animals all necessarily take what is good for them as absolutely good and worthy of pursuit. Therefore, all sentient creatures are themselves ends in themselves because we are obliged to recognize that they necessarily consider themselves to be ends in themselves and that what is good for them is therefore worthy of pursuit in itself. Korsgaard summarizes this point as follows: “animals necessarily take themselves to be ends in themselves in this sense: that is simply animal nature, since an animal just is a being that takes its own functional good as the end of actions” (Korsgaard, 2018, p. 146). In other words, since all sentient creatures, human and non-human, seek the well-functioning of their organism and must be recognized as necessarily valuing themselves as ends in themselves, we are obliged to treat their ends, or at least what is good for them, as good absolutely. For Korsgaard, it is then only reasonable to concede that there is a moral requirement to treat all sentient beings as ends in themselves. Since animals are ends in themselves, she argues, it follows that there is an important set of substantial moral obligations towards them, such as respecting their right to possess what contributes to their well-functioning (what is good for a creature), which implies their lives as an essential condition for their functionality (Korsgaard, 2018, pp. 21, 136, 137, 141, 144, and 145).

We can schematize Korsgaard's argument as follows:

1. 1. Constructivist premise: nothing has value independent of a singular point of view; a thing is only of value to a creature that values it.

2. 2. Factual premise: sentient animals have a valenced experience of the world and all value the well-functioning of their organism.

3. 3. Universalization lemma: if something has value for all valuing creatures, it is “absolutely good,” and then that something must categorically be respected.

4. 4. Reflexivity lemma: if a creature values something absolutely, then it necessarily values itself (is an end-in-itself in the passive sense of the term).

From these premises and lemmas, it is deduced a) that we must categorically respect the well-functioning of sentient organisms; b) that non-human animals are ends in themselves.⁶

Clearly, the present analysis of the arguments developed by Korsgaard does not claim to be perfectly exhaustive. Of her own words, the Kantian argument she supports takes at times “complicated twists and turns” (Korsgaard, 2018, p. 145). However, the presentation of her position here has focused on the most fundamental aspects of her reflection, which she summarizes as follows:

I have argued that animals have moral standing because animals, including humans, have a good in the final sense of good (8.5), and that we have a good because we have valenced responses to the things that affect the functional goodness of our own condition (2.1.7). (Korsgaard, 2018, p. 156)

2. Critique of Korsgaard's Argument

In my view, when we compare Korsgaard's argument in defence of rights for non-human animals with some of her earlier work where she defended a similar thesis for human rights, one particular problem seems to be recurrent: the argument depends on certain normative presuppositions that go beyond what a metaethical constructivist position allows. Indeed, Sharon Street argues that Korsgaard's Kantian approach is problematic insofar as she only inconsistently follows its implications — specifically, nothing has value independent of a singular point of view and that a thing is only valuable to a creature that values it (Korsgaard, 2018, p. 9).

2.1. I Have to Value Myself to Be Able to Value Anything

Indeed, I believe that Korsgaard's argument for the recognition of animals as ends in themselves borrows somewhat the same form as her Kantian argument for the idea that substantial and categorical moral obligations to humanity derive from the practical position of any moral agent. It is this precise point that differentiates her position from other forms of constructivism, and more particularly from Humean constructivism, according to which it is impossible to affirm the existence of such categorical obligations (Street, 2010, p. 370).

In The Sources of Normativity, Korsgaard argues that human beings can be seen as creatures capable of distancing themselves from their desires and impulses, and capable of asking themselves how and why they should act (Korsgaard, 1996, p. 121). Thus, being human, according to Korsgaard, would be characterized by the fact that we need reasons to act, a need that would be answered, according to her, by the fact that we possess different practical identities (Korsgaard, 1996, Section 3.4.7).⁷ However, given the reflexive aspect of our nature, such practical identities should in turn be justified normatively, justification finding satisfaction in the fact that we value ourselves as human beings and that this humanity would thus become an end-in-itself for any agent. Korsgaard summarizes her argument as follows:

Since you are human you must take something to be normative, that is, some conception of practical identity must be normative for you. If you had no normative conception of your identity, you could have no reasons for action, and because your consciousness is reflective, you could then not act at all. Since you cannot act without reasons and your humanity is the source of your reasons, you must value your own humanity if you are to act at all. (Korsgaard, 1996, p. 123)⁸

For Korsgaard, such an argument would then have substantial and categorical normative consequences for every human being since all rational agents would be forced to recognize that they necessarily value their own humanity as an end-in-itself.

One thing that may seem surprising here is that, in her works preceding Fellow Creatures, Korsgaard supported the idea that the reflexive nature of human beings was the source of all moral thought. Some might wonder how this position is compatible with the idea that non-human animals, which are not endowed with reflexivity, should be included in our moral community (Korsgaard, 2018, p. 148).

In Fellow Creatures, Korsgaard leaves aside the concepts of reflexivity, practical identities, and the value of humanity for obvious reasons. To put too much emphasis on these criteria might hinder our ability to value non-human animals worthy of moral considerations.⁹ While the reflexive aspect of human beings remains obviously significant for the recognition of the value of humanity, Korsgaard nevertheless suggests in her most recent work that it cannot be the only sufficient criterion to recognize a creature as worthy of categorical moral obligations.

Thus, by admitting the distinction between the active and passive senses of the notion of “end-in-itself,” Korsgaard makes it possible to conceive of an argument similar to the one she makes for the categorical value of humanity, but this time explicitly inclusive of non-human animals. Indeed, as I mentioned earlier, for Korsgaard, it is not only humans who, by virtue of their rationality, value themselves as ends in themselves, but all creatures for whom things can be good or bad by virtue of the fact that they take their ends as absolutely good. ¹⁰ Korsgaard's arguments thus follow the same form in the sense that, starting from the fact that, as creatures, we value certain things, she concludes that we must necessarily value ourselves as ends in themselves, and that then substantial and categorical moral obligations follow. Hence, by the very performativity of their functionality, all animals, human or non-human, consider the well-functioning of their organism as an end in itself, and their own person as a final good. Therefore, every moral agent should recognize the obligation to respect all sentient creatures as ends in themselves.

In other words, as I understand them, the two arguments follow the same development that could be summarized as follows: For every individual A, if A values X, then A must value themselves as an end in themselves (reflexivity of value thesis) and if A values themselves as an end in themselves, then A is an end in themselves for all other creatures (universalization of value thesis). In the case of the argument for the value of humanity, Korsgaard argues that any human who values anything must necessarily value their humanity as an end-in-itself. Because of their reflexive nature, all humans value certain things, so they all value their own humanity as an end-in-itself. All moral agents therefore have categorical obligations to human beings. For the case in favour of the value of non-human animals, Korsgaard instead argues that all sentient beings who value anything must necessarily value their own functionality as an end-in-itself. Because of their own condition, all sentient animals value certain things, so they all value their own functionality as an end-in-itself. All moral agents therefore have categorical moral obligations towards sentient animals. We thus see that it is by relying on two fundamental principles that Korsgaard comes to be able to assert that humanity and the functionality of sentient organisms have categorical values. First, she insists on the idea that if a creature values anything, it must necessarily value another thing as an end-in-itself. Second, she states that the fact that a creature values a certain thing as an end-in-itself implies categorical obligations for all moral agents.

2.2. Sticking to the Constructivist Paradigm

Although Street generally recognizes that the Kantian argument has the advantage of offering a strong form of moral objectivity, she argues that when Korsgaard defends the categorical value of humanity, it deviates too far from the constructivist paradigm, according to which there is nothing moral outside the practical point of view of an agent (Street, 2010, p. 370. See also Desmons, 2018, p. 480; Rawls, 1993, p. 78). For Street, to ask whether an agent has reasons to judge something as valuable or counting as normative is to step outside the framework that can make sense of the question itself. Recall that, for constructivists like Korsgaard and Street, the fact that there is a value X simply means that an agent A takes X as having value; that is, X has value for A. Therefore, when Korsgaard asserts that we must consider ourselves as an end to justify the rest of our values, Korsgaard is guilty of being inconsistent with the way agency works. In other words, for Street, to value something is already to consider oneself to have reasons in favour of that thing. In this sense, for Street, Korsgaard's question, “Do I have reasons to value something?” is like asking the question, “Is the Empire State Building taller?” where the context necessary for the question to make sense is omitted, that of a substantial practical identity of an agent already taking certain things as valuable (Street, 2012, pp. 49–50).

In the context of her argument for the moral value of non-human animals, Korsgaard seems to me to be making the same mistake. Thus, to affirm that a creature, because it values certain things, necessarily values itself as an end-in-itself, and that we must then categorically value that creature and what is good for it intrinsically seems to be a normative assertion made outside a referential framework in which it could make sense, that is, a situated practical point of view. With Street, I thus oppose the idea that a “pure practical reason” would necessarily commit us to specific normative values, for a constructivist approach to morality should stick to the idea that only the specific position of an agent and the set of normative values and judgements it entails can provide such a substance. In other words, the validity of normative principles such as “if a creature values anything, it must necessarily value something else as an end-in-itself” or “the fact that a creature values a certain thing as an end-in-itself implies categorical obligations” can only depend on their ability to withstand scrutiny from the perspective of a specified set of particular normative values and judgements (Street, 2008, pp. 208–214). To take up Korsgaard's own argument,¹¹ it would be necessary here to be able to ask the question, “For whom does the fact that a creature seems to have the well-functioning of her organism as the end of her actions imply substantial and categorical obligations?”

2.3 Do We Necessarily Think of Ourselves as an End in Ourselves?

More specifically, two explicit problems arise here (see also Brown, 2018, p. 586). First, I agree with Street when she argues that the fact that a creature values certain things does not have to be conceived of as necessarily implying that it does so to the extent that it values itself as an end-in-itself.¹² In “Coming to Terms with Contingency,” Street considers the fictional example of an alien entity with traits similar to those of an ant to illustrate this position. The creature, highly intelligent and demonstrating a reflexive and coherent character, would occupy the role of a worker within a society organized in the manner of an ant colony led by a queen. Experiencing her own functional condition in the same way as earthly creatures, she would value several things, that is, several things would be important to her. However, she would not consider herself and her existence as an end-in-itself, but would see her person and her functional good as having only trivial and purely instrumental importance: only the well-being and survival of her queen and her colony would have real importance to her and would be the ultimate end of her actions (Street, 2012, pp. 53–54). Without dwelling on the details of Street's example, we can see that it is our own values that determine what is valuable (to us), and that does not imply that we ourselves are valuable (to us) (Street, 2012, p. 54). Street reminds us that things are ultimately valuable only because we value them; in other words, that something has value only means that an agent values that thing, no more, no less. It is therefore by no means a logical necessity for an agent who values something that they themselves have value, or for them to consider themselves as having value. In this way, that an animal creature seeks or flees certain things by virtue of her valenced experience, or that certain things seem to be important to her, simply means that these things may have value for her, but in no way implies that her person is necessarily an end-in-itself in her own eyes. To repeat the formulation used above, for constructivists, “X has value” implies only that an agent A values X. That another object Y must have value for any agent that values X cannot therefore be categorically inferred without being guilty of attributing to oneself an omniscient point of view on what has value.

2.4 The Normative Force of the Status of End-In-Itself

Nevertheless, even if it could be conceded to Korsgaard that it is reasonable to assert that the fact that a creature values certain things necessarily implies that she does so from a position where she values the well-functioning of her organism as an end-in-itself, a problem still persists. Indeed, it remains unclear in what way the fact that this position is inescapable implies that it is normatively more important than any other practical identity or end that a creature may have and that categorical obligations arise from it for any moral agent (see Brown, 2018, pp. 583 and 586). Let us remember, for Korsgaard, sentient creatures possess a final good because they have valenced reactions to things that affect the functional good of their own condition. From the natural fact that sentient creatures have valenced responses to things that affect the functional good of their own condition, Korsgaard advances the normative conclusion that every moral agent must categorically act in such a way as to respect what contributes to the well-functioning of every animal creature. According to the constructivist paradigm advocated by Street, for it to have the normative force it claims to have, such a principle would require an impersonal point of view from which it would be possible to determine which ends it is acceptable to pursue, or which ends would be more important. Thus, the principle then turns out to be a naturalistic presupposition implicit in the Korsgaardian approach (see Brown, 2018, p. 587; Korsgaard, 2018, p. 168). However, such a naturalistic presupposition is inconsistent in the context of a constructivist theory of value since it is opposed to the fundamental idea that there is nothing moral outside the practical point of view of an agent. Let us remember that it is precisely for this reason that Korsgaard rejects the validity of an egocentric predicament where it would simply be presupposed that the human point of view would have greater value than the animal point of view.¹³

In other words, to assert that “as the purpose of every action of an animal creature is the functionality of her organism, the functionality of that creature is therefore a final good and has value in itself for any moral agent” seems to go beyond what allows the tethered theory of value that Korsgaard defends by taking the form of a naturalistic reasoning stipulating an ought from an is. It is indeed possible to contest the premises of such reasoning: what is the proper functioning of an animal? From what point of view can we establish such criteria? A philosopher like Street could suggest that what constitutes the “proper” functioning of a creature is likely the result of contingent forces such as those of natural selection. Determining whether this functionality should take precedence, be more valuable than others, or be pursued for its own sake is indeed possible only from the practical standpoint of a being who already values certain things and already has, for example, certain normative conceptions about the value of these contingent factors that have formed the functionality of our organisms. In short, these questions only demonstrate the contingent and normative nature of the criteria of what Korsgaard considered to be absolutely good, that is to say good for all.

Finally, we see how the criticisms I have just elaborated lead to rejecting the universalist and categorical claims of Korsgaardian theory. Indeed, I have tried to demonstrate two things. First, Korsgaard's approach presupposes that a creature for whom things can be good or bad must value herself as an end-in-itself. Second, it also presupposes that the functional good of these sentient creatures must have a categorical priority over our ends. I have thus argued that these principles, if we want to be consistent with a constructivist approach to normativity, cannot be attributed categorical and universal value since it would imply a “point of view from nowhere.”

3. Opting for a Humean Constructivism

If we accept Korsgaard's tethered theory of value, but recognize the validity of the criticisms that have just been formulated, then it would seem that we should side with the Humean constructivists. They argue “that a state of the world or an action is judged right or wrong, or better or worse, in light of our other assessments already in place” (Maclure, 2018, p. 505, translation mine), excluding the possibility from an impersonal point of view of a “pure practical reason” involving substantial and categorical normative obligations for any agent (see Street, 2010, p. 370). Rather, these philosophers argue that a constructivist approach to morality should stick to the idea that normative truths derive only from the specific position of a historically located agent and from the set of normative values and judgements it entails (Maclure, 2018, p. 507). Street certainly agrees with Korsgaard when she argues that value has “entered” the world with animals (Korsgaard, 2018, p. 21). However, according to Street, we must accept the contingent aspect of our practical position and recognize that the value of things emerges only with agency, that is, with the fact that a creature values certain things, and that there is no other reason in itself to value these things (Street, 2012, p. 58). For Humean constructivists, the mere fact of our humanity or of the functional good of sentient creatures would therefore not be sufficient to justify the existence of substantial and categorical obligations arising from the practical position of any agent.

3.1. The Impossibility of Opposing Caligula's Practices

Does this mean that we must abandon Korsgaard's goal in Fellow Creatures of defending the intrinsic value of non-human animals, or even any substantial moral position? Indeed, one of the main criticisms Humean constructivism faces is that it has simply unacceptable normative consequences, such as the conclusion that it would no longer be consistent to hold certain moral positions as objectively false. To support this idea, some authors suggest that it would be impossible for us to oppose to the practices of a perfectly coherent Caligula.¹⁴ Indeed, according to Street, we have to recognize that a consistent constructivist position would imply the theoretical possibility of the existence of a being who would value, above all else, torturing people — or non-human animals — and maximizing their suffering and that, if it were understood that this individual was perfectly consistent with all the normative values and judgements involved in their practical position, and perfectly informed about non-normative facts, it would be clear that this individual would have strong normative reasons for doing so. According to Street (2016, p. 325), the objection would culminate in the idea that, in the words of Ronald Dworkin, there would ultimately be “no moral objection to exterminating an ethnic group or enslaving a race or torturing a young child, just for fun, in front of its captive mother” (Dworkin, 1996, pp. 117–118). As such, for many, Caligula's example should embody the very idea that certain values are inherently irrational or that valuing certain things is simply a normative error (Street, 2009, pp. 273–274). In this way, it is understood that, for many like Korsgaard, for example, the fact that a “Caligulan” agent would not recognize the value of non-human animals would certainly be due to an error of reasoning.

Yet, Humean constructivists consider this objection to be unconvincing for several reasons. First, although they recognize that a “perfectly coherent Caligula” is, in theory, conceivable, that such an individual exists would nevertheless be highly unlikely to the extent that a human being who would have “Caligulan reasons” for torturing people for pleasure would either be presumably incoherent and irrational, or so different and distant from what our practical position implies, that it would be impossible for us to agree with their normative positions (Street, 2016, pp. 330–332. Also see Maclure, 2018, pp. 518–519). Thus, let us recall, Humean constructivism nevertheless recognizes the possibility of a certain form of normative truths. It is then possible for us to affirm that an individual like Caligula is repugnant to us, and that if such a person were ever to exist, we would have normative reasons to want to avoid their behaviours, to defend ourselves against them, to imprison them, and to want to change this person, even if we do not think we are speaking from an “absolute” point of view. Subsequently, Street has always argued that Humean constructivism does not lead to a nihilistic view of values, but that on the contrary the constructivist can assert that it is necessary to prevent a being like Caligula from torturing people since such a thing would logically and instrumentally derive from their practical position (Street, 2009, 2010, 2012, 2016). Thus, although for the Humean constructivists normativity takes its source in the individual practical point of view, this does not allow us to neglect the collective and institutional aspects of morality, because “agents are always inserted in a moral context that surpasses them, made of norms whose authority is imposed on them” (Maclure, 2018, p. 518, translation mine).¹⁵

3.2. What a Humean Constructivist Can Say About the Animal Question

In the case at hand, namely the question of the treatment of animals, however, answering the objection seems less obvious: to what extent can we be dealing with individuals who do not value non-human animals if they are perfectly consistent with all the normative values and judgements contained in their practical position and perfectly informed about non-normative facts? It even seems legitimate to ask: but what can a Humean constructivist say about animal ethics — or any other normative question, for that matter? Again, it seems that the answer is that we are allowed to defend the normative truths that follow coherently from our practical standpoint. For the constructivist, the principles of justice are in reality only the reflection of the conceptions posited by the practical position of individuals, and are therefore often determined by their initial presuppositions (Desmons, 2018, p. 477). It is therefore possible to conceive of a practical position that would imply sincere empathy towards the condition of non-human animals, although an opposite position is also conceivable. However, this in no way implies that the conceptions posed by the practical position of an individual are immutable, quite the contrary. A moral agent, even one theoretically perfectly coherent and informed, is obviously always in a dynamic relationship with her environment and her practical position is therefore inevitably subject to change (Maclure, 2018, pp. 514–516). However, let us emphasize that the very idea that a human being can be perfectly coherent or perfectly informed about non-normative facts remains in itself highly unlikely. Our values thus change continuously under the influence of normative and non-normative facts that come our way, and our normative positions must therefore be subject to periodic consistent reassessments.

Subsequently, it is the view of many thinkers that the phenomenon of human morality can indeed extend its considerations to an ever-increasing number of subjects, including even non-human animals (Anderson, 2004; Dennett, 2017; Donaldson & Kymlicka, 2016; Korsgaard, 2018; Maclure, 2018; Rowlands, 2009; Singer, 2011). Such a thing could be understood by the increasing complexity of our relationships with the various individuals and non-human animals around us and by the changing understanding of our situation's proximity to theirs, thereby transforming our initial assumptions about the value of individuals and sentient creatures. In other words, interacting and understanding the world differently and the beings with whom we evolve certainly has the effect of transforming the way we conceive of our relationships to them and the duties that flow from them. For example, if in the face of new data on animal sensitivity, an individual comes to conceive of non-human animals as sentient beings in the same way as human beings, and if they generally devalue suffering, this could lead consistently to a devaluation on their part of any form of animal suffering.

Therefore, even if one has to renounce moral universalism, there are many options for the Humean constructivist for defending a favourable position on the value of non-human animals, or for any other normative question, for that matter. For example, through dialogue, confrontation, or even conflict, one can simply defend the normative truths that flow from their practical position and try to argue: that the same principles would be consistently implied by the initial presuppositions of their interlocutors; that some of our initial conceptions deserve to be revisited according to a new understanding of a situation and the relationships we have with certain topics; or that it would be favourable to the realization of our common interests to adopt certain normative positions. This is obviously not an exhaustive list of the options coherent with a Humean constructivism to defend our normative positions, but only an overview of what forms a revising process of our moral evaluations could take in light of what it supports (see Maclure, 2018, p. 515).

In the particular case of the treatment of non-human animals, for example, this could result in the following positions. It could be argued that it would follow from principles already accepted by our peers, such as the devaluation of suffering in general and the recognition of the sensitivity of non-human animals, that we should commit to minimizing animal suffering and maximizing their welfare (Singer, 1997). It could also be argued that some representations of animal nature would be undermined by recent studies on their sensitivity and behaviours, or that our ways of conceiving the moral status of these beings deserve to be reconsidered in light of previously underutilized principles, such as the inherent worth of these beings (Regan, 2013). If none of these strategies work, it would still be possible to argue that it would be in our interest, since it would be beneficial to the realization of common causes such as the protection of biodiversity and the environment, to adopt policies to protect some of the fundamental interests of non-human animals (Anderson, 2004). The strategies identified here are obviously not new, but are in fact only a few examples from the abundant literature on the subject of animal ethics. However, it is now clear that each of these — and the very existence of their diversity — takes on a completely coherent meaning within the framework of a metaethical conception such as that of Humean constructivism. Thus, while the purpose of this article is not to develop a substantial “Humean strategy” for the defence of the rights of non-human animals, I nevertheless seek to demonstrate that it is not necessary to claim an “absolute” point of view on morality in order to be able to defend one's normative positions effectively and coherently.

3.3. Possible Objections

Objections could, of course, be made to the position I have just elaborated. Among these, three are more obvious. First, it would seem to many that a Humean constructivist position would require the recognition that a character like the perfectly coherent Caligula would indeed be justified in torturing people for pleasure. Faced with this fact, the Humean constructivist would probably have no choice but to accept this as a theoretical possibility. Indeed, according to Street, anyone who went through the exercise of imagining in detail what a perfectly coherent Caligula would look like would have to come to terms with the fact that he would indeed have such normative motives — although he would also be closer to an alien than to a human being (Street, 2016, pp. 330–332) However, the fact remains that in the case that Caligula would value the suffering of other human beings in a perfectly coherent way, this would likely imply implausible ontological positions, such as the negation of the existence of an inner life in other human beings or radical nihilism — which is nonetheless a possibility. Nonetheless, with regard to the case of individuals devaluing the moral status of animals or simply valuing the fact of eating them, we must nevertheless conceive that this is indeed a common and possibly perfectly coherent position.¹⁶ However, as I have already pointed out, while it is possible to recognize that Caligula does indeed have reasons to act as he does, this does not mean that we have no reason to want to prevent some of his behaviour. The Humean constructivist, if one is consistent, must therefore learn to accept and recognize the diversity of practical standpoints and what it implies, but must possibly also learn to coordinate them. Thus, its theoretical humility is, in my opinion, a strength of Humean constructivism, which has its source in its understanding of normative pluralism and in the recognition of the validity of a diversity of practical points of view.

This leads us to consider a second possible objection, namely, the question of what it would be possible to do if we lived in a world of Caligulas. As I have already mentioned, Humean constructivism offers seemingly no resources to oppose a morally homogeneous society composed only of “perfectly coherent Caligulas” as such (Maclure, 2018, p. 519). However, this is not a problem, but rather a requirement of the position, insofar as it opposes precisely the idea that there can be a universal and absolute position from which it would be possible to judge the validity of normative principles, or rank the value of certain types of existence. Humean constructivism, as a metaethical position, does not actually imply any particular moral position, unlike Kantian constructivism — which Street denounces precisely for its theoretical inconsistency. A world of Caligulas is therefore not a problem in itself. While it may seem horrible or unjust, this can only be understood by what our own practical position implies, but it also involves conceiving that it would be possible for an individual to cultivate the same feelings or have the same judgements about our own world.

Finally, what has just been established opens the way to a third objection. Clearly, the world we live in is far from being made of morally homogeneous and perfectly coherent beings. Reaching agreement on the norms to be followed is therefore a major problem, sometimes even seemingly insurmountable — especially if we consider that the demands of morality are not categorical and universal. What happens, then, if no agreement exists between the members of a moral community on a specific issue, such as the treatment of non-human animals? First, let us emphasize that this is undeniably an ethical and political problem — not a metaethical one — and that the resources to think about this type of difficulty are abundant. However, I believe there is a general answer. Indeed, it seems implausible that within a moral community there would be no contingent similarities or common aspects in the practical views of its members that could justify some overlap in fundamental evaluative principles. Such tendencies — possibly due, for example, to contingent factors such as historical, geopolitical, and cultural contexts, or even the influence of natural selection, etc. — can often be significant enough to give rise to important convergences in the evaluative attitudes of members of the same community (Street, 2010, p. 370).¹⁷ For example, in the past, by virtue of particular political circumstances, some groups have come to realize how similar they are to other individuals, which has dramatically changed the ways in which they conceived their relationships as well as their moral and political obligations towards them. This is why some thinkers draw parallels between the structure of the debate on non-human animal rights and the evolution of the fight for civil rights in Western democratic countries (Regan, 2013; Rowlands, 2009). So, as I have already mentioned, for many people, we have a growing understanding of how we are similar to non-human animals in the way we experience the world and how we seem to value certain things in a similar way, or how some of our interests obviously converge with theirs. I therefore believe that it is reasonable to be optimistic about the possibility of such agreements on the value of certain normative principles concerning the issue of the treatment of non-human animals, or any other normative issue, for that matter. Thus, while a Humean perspective of morality certainly allows us to make sense of the contingent aspect of our normative positions and of the existence of disagreements between them, it also allows us to think about the possibility of an “evolution” of these and of agreements between them. This is why I believe that not being able to simply categorize normative positions opposed to our own as “categorically false,” but to recognize the contingency of normative truths is in no way, as some argue (Enoch, 2011), a disadvantage of a consistent constructivist position. On the contrary, in light of what has just been argued, such an approach to normativity instead, in my view, allows us to address the complexity of our moral experience in a reasonable, pragmatic, and optimistic way.

Conclusion

Again, the purpose of this article was not to support a substantial normative position on the issue of the treatment of non-human animals or to develop a “Humean” strategy for defending their rights. Starting from an analysis of Korsgaard's metaethical argument in Fellow Creatures, I have instead attempted to demonstrate two things. First, that Korsgaard's approach implies certain presuppositions about the intrinsic value of some normative principles, namely that a creature for whom things can be good or bad must value herself as an end-in-itself, and that the functional good of these sentient creatures must have a categorical priority over our ends. To do this, I have adapted Street's critique to earlier versions of Korsgaard's Kantian constructivism to question the existence of categorical moral obligations for any moral agent. Second, I have argued that such a thing in no way implies value nihilism and that a Humean constructivist could very well defend substantial ethical positions in a coherent manner on any type of normative question. I have also tried to demonstrate that, while it is true that Humean constructivism as a metaethical position implies the impossibility of defending categorical and universal moral obligations, Humean constructivists sensitive to the animal cause nevertheless have many tools to assert, from their own practical positions — and this, even within a society that is indifferent or has little concern for the fate of non-human animals — that the animal question deserves to be reconsidered or that animals receive treatment that contradicts the moral beliefs in force in their society.