1 Introduction

One of the most interesting questions within contemporary metaethics is about whether or not moral facts or truths are mind-independent, that is, whether or not the human mind is constitutive of these truths and facts. Metaethical constructivists are deeply opposed to the mind-independence of morality and have offered several arguments against mind-independence theories (henceforth MI-theories). In this paper, I aim to resume their line of arguments against MI-theories while presenting an argument that stems from the considerations of moral phenomenology. My claim is that while MI-theories fail to make sense of at least one central component of moral phenomenology, constructivists are in a very good position to account for it.

My contribution thus is built around the notion of moral phenomenology and the rarely discussed role of phenomenological investigation in metaethical theorizing in general. Hence, the first step of my argument introduces the investigation of moral phenomenology and elaborates its central role for what it means to do metaethics in the first place. I argue that we should understand moral phenomenology not only in terms of the first-personal experience(s) of moral agents forming moral judgments or being subject to moral obligations, but also to include the various phenomena of common moral practice and discourse. As I further show, it is exactly this broader understanding that lies at the heart of doing metaethics. Accordingly, if metaethical theories failed to accommodate the phenomena of moral practice, this failure looms large.

Secondly, I argue why MI-theories cannot explain and make sense of one central phenomenon of moral practice concerning our often employed relativist understanding of morality. Arguably, moral relativism, which I take to include subjectivism as well as intersubjectivism (Sayre-McCord 1988), is not a very attractive view; however, many moral agents do think and speak in a relativist fashion. The problem with MI-theories is that they cannot explain or even make sense of such thinking and speaking, because advocates of these theories are committed to the claim that any relativist understanding of morality demonstrates a severe misunderstanding and even confusion about the very concept or nature of morality as such.

This diagnosis, however, appears inadequate, for not only do many agents conceive of morality in a relativist way, but what is more, there seems to be nothing absurd about it. After all, the relativist interpretation of morality can be supported by a long-standing and sophisticated philosophical debate concerning its possible truth (cf. Harman 2001). In addition, this debate also rests on results stemming from different scientific areas such as anthropology and psychology. Therefore, it does not seem correct to think that people who deal with morality in a relativist fashion are conceptually or otherwise confused or that they are incompetent in dealing with morality; rather, they are employing a genuine way of spelling out moral truth-conditions (Sayre-McCord 1988, 14 f.).

Constructivism, on the other hand, fares better in explaining any relativist understanding of morality. As I ultimately show, even Kantian objectivists can make sense of this understanding without thinking of common moral agents as fundamentally mistaken from the outset.

Assuming the central importance of accommodating moral phenomenology, the ability to explain and make sense of the phenomenon of the relativist understanding of morality counts against the adoption of any MI-theory but in favor of constructivism. Or so I will argue.

The paper will proceed as follows. In the second section, I introduce the constructivist position in metaethics and clarify its notoriously ambiguous relation to moral realism. In the third section, I discuss the feature of moral mind-independence in more detail. The fourth section explores the investigation of moral phenomenology and elaborates its central role in metaethical inquiry. In the fifth section, I show why MI-theories are at odds with moral phenomenology by considering the phenomenon of the often employed understanding of morality in a relativist fashion as well as the reasons that count in favor of it. Finally, I discuss three objections against my argument and provide a conclusion.

2 Constructivism, realism, and moral mind-independence

What is constructivism in metaethical debate? According to the standard definition, it is the theory that moral truths and facts are not discovered or tracked, but rather constructed by human, rational agents (Bagnoli 2013, 1). To be more precise, constructivists believe that there is a certain procedure in virtue of which moral agents can determine moral truths (Darwall et al. 1992, 140) and that, independent of the application of such a procedure, “there are no moral facts” (Rawls 1980, 519).

But what is this procedure about? First-order ethical constructivism holds that in order to determine certain truths about, say, social justice, the procedure starts with the acquisition of certain intuitions or convictions about justice claims, which are then brought into a reflective equilibrium (Rawls 2005, §9; Street 2010). Metaethical constructivism, on the other hand, reflects on human attitudes, thoughts, and outlooks more generally (Shafer-Landau 2003, 14; Street 2010). We can therefore state that:

Metaethical constructivism is the view that moral truths and facts are not discovered or tracked, but are the output of a hypothetical procedure that is applied to some proper input in terms of the practical standpoint, where this is understood in terms of human attitudes, thoughts, outlooks, emotions, etc.

What concerns me in the following is how to think about constructivism in the moral realism/anti-realism debate. For it is often thought that constructivism is an anti-realist view (Brink 1989, 16), or that it is “premised on a rejection of realism” (Street 2008, 220, my italics).

But the truth of these claims largely depends on how one understands the term “moral realism”, for there is some ambiguity in how we commonly refer to it (Sayre-McCord 1988). Thus, on one rather prominent interpretation, to be a realist about morality means to endorse a certain view about the status of moral statements that I will call strong cognitivism. As Sayre-McCord claims:

In every case, what marks off some particular terrain as the realist’s remains the same […], it is the view that some of the disputed claims literally construed are literally true. Wherever it is found […], realism involves embracing just two theses: (1) the claims in question, when literally construed, are literally true or false (cognitivism), and (2) some are literally true (ibid, 5).

Others consider a rather different understanding of moral realism and readily identify it with a metaphysical view. Brink thus argues that “[m]oral realism is a kind of metaphysical thesis about the nature and status of morality and moral claims” (1989, 14). The metaphysical understanding of realism entails strong cognitivism because it assumes that there are moral facts (ibid.), but offers a distinctive explanation or “truth-making theory” (Asay 2012) of why there are facts in the moral domain. Accordingly, realists, who endorse the latter, metaphysical interpretation of moral realism, argue that there are moral truths because there are mind-independent facts that render moral judgments true or false (cf. Brink 1989, 14).Footnote 1

Constructivists are deeply opposed to the idea that moral truths or facts were independent of the human mind and thought. They thus reject any form of mind-independence realism (IMR) because they hold that these truths and facts are the outcome of a certain (constructive) procedure. This means that moral facts spring or come into existence only after the procedure has been carried out successfully. Therefore, on the constructivist understanding, the existence of moral truths depends on two things: what is entailed in the practical standpoint, i.e. our human thoughts, outlooks, and attitudes, and what rational reflection reveals to be morally true. Moral facts are therefore constituted by rational reflection (Brink 1989, 15).

We can thus conclude that constructivism is a form of moral realism in virtue of its endorsing strong cognitivism, but that it is anti-realist in the sense that it rejects the view that these truths and facts are independent of all human thoughts and attitudes. So while constructivism “endorses the reality” of the moral domain (cf. Shafer-Landau 2003, 14) because it holds that there are moral truths and facts, these facts are not thought to be mind-independent because they are the outcome of the procedure of rational reflection applied to the practical standpoint.

Note, however, that while the feature of moral mind-independence is commonly ascribed to IMR, recent developments in the non-cognitivist camp show that some non-cognitivists like quasi-realists also believe that moral truths are independent of the (human) mind. Quasi-realists are commonly (although not necessarily) projectivists who aim at making sense of the truth-stating character of moral discourse (Blackburn 1984, 180; 1999). While Blackburn’s quasi-realism thus starts with the idea that moral agents project – or spread – their sentiments onto the world and hence begins as non-cognitivist expressivism (Blackburn 1984), in the end it wants to make sense of moral truth and objectivity just as IMR does (Blackburn 1984; 1993). One feature it adopts so as to be capable of “saving the appearances” of truth and objectivity as embedded in common moral language and thought (Asay 2013, 213; Ridge 2006) is moral mind-independence. As Blackburn argues:

[…] it is not my sentiments that make bear-baiting wrong; it is not because we disapprove of it that mindless violence is abominable; it is preferable that the world should be a beautiful place even after all consciousness of it ceases (1993, 153).

And:

The counter-factual “If we had different attitudes it would not be wrong to kick dogs” expresses the moral view that the feature which makes it wrong to kick dogs is our reaction. But this is an absurd moral view, and not one to which the projectivist has the least inclination. (Blackburn 1984, 217).

The aim of the quasi-realists then is not simply to mimic moral realism, understood in the strong cognitivist sense, but rather to mimic IMR in virtue of adopting the feature of moral mind-independence. Consequently, if constructivists are opposed to IMR because of its adoption of the feature of moral mind-independence, they must also be opposed to quasi-realism.

But what exactly does it mean to endorse the mind-independence of morality? The quote from Blackburn points us in the right direction, for mind-independence in general is a compound of two theses:

  1. 1.

    It is not the case that if I/we think that X is bad then it is bad

and

  1. 2.

    It is not the case that if X is bad then I/we think it is (cf. Zangwill 1994, 206; Warenski 2014, 826).

Let me consider two features that will prove to be important for my further argument: (i) a clarification concerning in exactly what respect moral truths and facts are supposed to be independent from the mental; and (ii) in what way MI-theories grant this feature a special status with regard to how we want to deal with morality in general.

Let us take (i) first. There are many ways in which the independence-claim of the feature of moral mind-independence can be understood (Kramer 2009). When MI-theories endorse the moral mind-independence, however, it is the existential mind-independence that they typically have in mind, where “[s]omething is existentially mind-independent if and only if its occurrence or continued existence does not presuppose the existence of any mind(s) and the occurrence of mental activity” (Kramer 2009, 25).

In order to better understand existential mind-independence, let us consider IMR and constructivism again. Advocates of IMR maintain that it is no part of the wrongness of, say, lying that we have certain thoughts or attitudes about lying (Jenkins 2005, 206), meaning that the human mind is not constitutive of the wrongness of lying. Constructivists, on the other hand, as we have seen, are anti-realist not because they reject the idea that there could be moral facts, but because they claim that the moral realm (only) results from the exercise of rational reflection on human attitudes, thoughts, and outlooks. In this respect, constructivists reject the existential mind-independence because they hold that mental activity is a precondition for moral truths and facts to come to exist in the first place.

But there is also the second aspect (ii). An often neglected idea entailed by the feature of moral mind-independence concerns its special status as regulating how we want to deal with morality in general. Zangwill, for instance, claims that moral mind-independence cannot be understood as a substantive moral commitment, for these commitments can always be challenged, questioned, and debated (Warenski 2014, 863). The feature of moral mind-independence, by contrast, must be interpreted as an evaluative standard concerning the competences of moral reasoning as such (Zangwill 1994, 211). Consequently, according to Zangwill, if someone were to challenge and question the feature, it would mean that she did not understand some essential ingredient of what it means to deal with moral requirements, judgments, and so on. If the very idea of something being a conceptual truth is that “we attain the knowledge in question by following out the implications of what we must know in order successfully to deploy a concept” (Zangwill 1994, 211–12), then any questioning of the mind-independence of morality counts as a heavy failure of not having correctly understood some central feature of morality, a failure that expresses a deep confusion about morality or an incompetence to deal with it (ibid., 217). While Zangwill notes that something being a conceptual truth does not mean that it cannot be rejected at all (ibid., 212), nevertheless there is a difference between some first-order moral requirement and something being a claim about moral competence as such, because only the latter concerns the correctness of one’s very understanding of morality. No such major mistake is involved in adopting or advocating some first-order moral statement or principle which – even if false – does not amount to having not understood a general requirement of how we want to understand and talk about morality in the first instance.

At this point, however, some may want to object that Zangwill’s analysis of the conceptual status of the feature of moral mind-independence is (a) not explicitly endorsed by (all) IMRealists, and (b) even incompatible with some MI-theorist’s understanding of the feature, such as in the quasi-realist interpretation. Thus, whenever we read about IMRealists discussing mind-independent moral facts, there is no suggestion that the feature should be understood in terms of a conceptual truth.Footnote 2 In addition, Zangwill’s argument about the conceptual status of the feature is especially designed in order to show where some MI-theories such as quasi-realism have gone wrong, namely in their failing to acknowledge the status of the feature as a conceptual status. Consequently, it may appear that Zangwill’s understanding of the feature is not widely endorsed after all and that there is thus reason not to consider it as an evaluative feature concerning morality after all.

But the objection is premature because it neglects that the MI-theorist understanding of the special status of the feature does not hinge on whether it is understood as a conceptual truth in particular. Thus, whether or not Zangwill’s interpretation in the end is correct does not change the way in which advocates of MI-theories understand the feature itself, namely as a feature that concerns how one ought to deal with morality in general. This is so because even quasi-realism, as Zangwill explicitly objects, does not consider the feature as a conceptual truth (ibid., 213), it nevertheless takes it as regulating our correct handling of morality:

It is not because of the way we think that if kangaroos had no tails they would topple over. We discover such facts, we do not invent them. […] The “mind-independence” of such facts is part of our ordinary way of looking at things (Blackburn 1984, 217, my italics).

Hence, granted that IMRealists and quasi-realists have different opinions concerning the exact status of moral mind-independence (i.e. in terms of a conceptual truth or general feature of moral reasoning), they nevertheless agree that the feature structures our way of dealing with morality, that it forms part of our “ordinary way of looking at things”, etc. Accordingly, while both the IM- and quasi-realist may object to someone who conceives of moral facts as mind-dependent that she either is conceptually or otherwise confused, the crucial point is that both think that neglecting the feature indicates not only a mistake, but profound confusion about morality.

3 Moral Phenomenology

Constructivists have presented many (different) arguments against IMR, and they have done so mainly because of its adoption of moral mind-independence (Bagnoli 2002; Street 2006). In the remainder of the paper I resume the line of the constructivist criticism by arguing that constructivism fares better than MI-theories in accounting for moral phenomenology. At this point, the reader might think that this sounds familiar; after all, Carla Bagnoli argued, in her 2002 paper “Moral Constructivism: A Phenomenological Argument”, that both IMR and quasi-realism are at odds with moral phenomenology. There are, however, crucial differences between Bagnoli’s argument and my own, most importantly in respect of how Bagnoli understands the term “moral phenomenology” and how I deploy the term here. Bagnoli refers to its (more) traditional understanding according to which to engage in moral phenomenology means to analyze and interpret the “what-it-is-like features of concrete moral experiences” (Horgan and Timmons 2005, 58), such as making a moral judgment, being the subject of moral obligations, and so on. Importantly, then, moral phenomenology is a first-person study that is concerned with the various experiential aspects of (our) moral life (Drummond 2008, 35).

However, even though the interpretation of moral phenomenology as the investigation of the “first-personal” aspects of moral life is widely shared, there are problems with this understanding, problems that demand a more inclusive, broader understanding of it. One worry is that there are many different ways of explaining the first-order experiences concerning our moral life with the resources of one’s preferred metaethical framework. So, let’s propose, for the sake of argument, that all agents experience moral demands as originating outside of them, that is, as being part of the world. At first glance, there is only one metaethical theory that can make sense of this feeling: IMR. Thus, it is no surprise that advocates of IMR explicitly refer to moral phenomenology in order to prove the plausibility of IMR:

The simple form can be found in Mackie’s view that we take moral value to be part of the fabric of the world; taking our experience at face value, we judge it to be experience of the moral properties of actions and agents in the world. And if we are to work with the presumption that the world is the way our experience represents it to us as being, we should take it in the absence of contrary considerations that actions and agents do have the sorts of moral properties we experience in them. This is an argument about the nature of moral experience, which moves from that nature to the probable nature of the world (Dancy 1986, 172).

As Dancy’s (further) argument shows, he thinks that what gives rise to our moral judgments is that there are moral facts “out there in the world” (Kirchin 2003, 249). On closer inspection, however, our experiences do not support the IMRealist view in the way Dancy thinks, for there are other metaethical theories available that can make perfect sense of these experiences while not counting as a brand of IMR. Thus, let us return to Blackburn, who starts his quasi-realist theory by introducing projectivism (Blackburn 1984). According to projectivism, moral agents (also) experience moral properties as originating outside of themselves, but not because they are somehow built into the fabric of the world, but rather because agents tend to “spread” their own sentiments “onto the world” (ibid.). Thus, the projectivist argues that we “project some attitude or habit or other commitment which is not descriptive onto the world, when we speak and think as though there were a property of things which our sayings describe which we can reason about” (ibid., 170–1, my italics). Consequently, the projectivist theory allows Blackburn to offer a non-cognitivist explanation of the exact same feeling that IMRealists take to count in favor of their own view. It follows from this that our reasons for favoring IMR on the grounds of phenomenological analysis are not as strong as its advocates would like to think.

Now, some take these worries to mean that we should rethink the importance of phenomenological inquiry for metaethical theorizing (Kirchin 2003). In contrast, I propose that there is a sense in which moral phenomenology can be considered of vital importance to metaethics that, at the same time, avoids the above-mentioned difficulties. This sense is not totally new, but it does require a broader understanding of phenomenological investigation, one that goes beyond its more common interpretation. Horgan and Timmons present a proposal that gestures in the right direction:

In metaethical inquiry, talk of “moral phenomenology” is used very broadly to include such deeply embedded phenomena as: (1) the grammar and logic of moral thought and discourse; (2) people’s “critical practices” regarding moral thought and discourse (e.g., the assumption that genuine moral disagreements are possible); and (3) the what-it-is-like features of concrete moral experiences (Horgans and Timmons 2005, 58).

At first the argument is surprising because, following the more traditional description of moral phenomenology, only (3) is at the heart of phenomenological analysis. On the other hand, it is often this latter sense that philosophers have in mind when they argue that some moral phenomena count in favor of their position and against the adoption of others. For instance, some claim that to “save the ethical appearances”, inter alia, means to offer a plausible account of the truth-stating character of moral discourse (Ridge 2006), and others maintain that moral “pluralism and dilemma come onto the scene as purported facts of moral experience” (Railton 1992, 720, my italics). Arguably, this is not the sense in which we originally have introduced the notion of moral phenomenology. But, nevertheless, it is exactly these phenomena that often stand at the center of metaethical discussion and to which, as we have seen, theorists refer as moral phenomena.

Now, having spelled out how to understand phenomenological investigation in metaethics, there arises the question of how important it is for metaethical theorizing in general. As I want to argue in the following, there are indeed very strong reasons to engage with moral phenomenology and to care about the results it yields on account of what it means to do metaethics in the first place. Thus, consider Hussain and Shah, who give us the textbook definition of metaethical inquiry:

Meta-ethics is the second-order investigation of the practice of making moral judgments – the everyday practice and the more philosophical practice of normative ethics. Meta-ethical investigations […] try to understand what is going on when we engage in the practice of making moral judgments (Hussain and Shah 2013, 86).

Often, when we think about metaethics, we primarily think about moral metaphysics or epistemology. Accordingly, the vast majority of metaethical theorists are preoccupied with these issues. However, there is a more general reason as to why we are concerned with these themes in the first place, and this is precisely because we are trying to understand what is going on in common, everyday moral practice, for as a metaethical theorist “one wants to be able to accommodate those commonsense assumptions” that underlie our everyday moral practice (Timmons 1999, p. 11, in Sinclair 2012, 162).

Therefore, the reason for why we ought to care about accounting for the phenomena of morality is because the question whether or not a metaethical theory succeeds in accounting for it inter alia decides about the theory’s very status as a metaethical theory. For if it is true that metaethics is first and foremost about accounting for what is going on in (common) moral practice, then not being able to offer an account that is in accordance with that very practice means overlooking something that is crucial to metaethical theorizing as such. Accordingly, failing to account for moral phenomenology in some sense means to fail as a metaethical theory.

Now, in the following, I want to present an argument against MI-theories that relies on moral phenomenology as outlined in this section. I claim that common moral practice entails a set of phenomena concerning our rather common understanding of moral matters in a relativist fashion and that only some metaethical positions such as constructivism can make sense of these phenomena, while others, MI-theories in particular, cannot do so.

4 The Possibility of Relativism

The claim that I defend in this paper is that MI-theories have problems in accounting for moral phenomenology, while metaethical constructivism does not encounter such problems. So where does the problem lie?

If we are convinced that moral practice aims at mirroring moral truth, the central question pertains to how we are to interpret this truth. We can thus adopt either

  1. a)

    objectivism

or

  1. b)

    relativism, where this includes subjectivism as well as intersubjectivism (Sayre-McCord 1988, 14–15).

Both a) and b) seem to be genuine ways of spelling out moral truth-conditions:

[Moral] [t]ruth-conditions are “subjectivist” if they make essential reference to an individual; “intersubjectivist” if they make essential reference to the capacities, conventions, or practices of groups of people; and “objectivist” if they need make no reference at all to people, their capacities, practices, or their conventions (ibid.).

Brandt (2001) argues that moral relativism cannot be interpreted in this way because relativism directly leads to nihilism or presupposes the falsity of moral realism (Carson and Moser 2001, 3–4). But this is not true, for there is no general problem in arguing for the possibility that moral truths should be construed as being relative to a given parameter (Scanlon 2001, 142) or to a “system of moral coordinates” (Harman 1996, 13), and that one of these parameters or coordinates may make “essential reference” to individuals or collectives. One can even go a step further and claim that moral relativism counts as a form of realism if the truth-conditions of both subjectivism and intersubjectivism are actually satisfied (Sayre-McCord 1988, 15).

What is true, however, is that moral relativism stands in direct opposition to MI-theories because of the way in which people and their mental states figure in the truth-conditions of moral judgments (ibid, 14–15). For relativism entails that the truth-conditions of moral judgments must make reference to the mental states – such as beliefs, desires, and aims – as well as the attitudes of moral agents or groups of such agents. MI-theories thus entail the falsity of relativism because these latter theories argue that “moral judgments or standards are objectively true independently of the beliefs and attitudes of human beings (even ideally rational beings) about those judgments or standards” (Carson and Moser 2001, 3, my italics).

But not only do MI-theories merely imply the general implausibility of moral relativism, it is in fact one of the direct aims of MI-theories to reject moral relativism right away by adopting the moral mind-independence. Thus Warenski claims that the negation of conditions (1) and (2) that define moral mind-independence would lead to a subjectivist understanding of morality because it would mean that moral facts would change whenever our attitudes change (Warenski 2014, 862). And Shafer-Landau readily contrasts his version of IMR with moral relativism in order to “get a better understanding” of the IMRealist “commitments” (Shafer-Landau 2003, 16). Finally, we should keep in mind that not only IM- but also quasi-realists endorse the existence of mind-independent moral facts, and it shows that Blackburn also shares the IMRealist aim of rejecting any relativistic interpretation of morality when he maintains that quasi-realism gives “a complete defense against relativism […]” (1999, 218).

Now, the problem with MI-theories is not that they stand in opposition to any relativistic understanding of morality; after all, Kantian constructivists also consider themselves to be objectivists. Rather, the problem is that the MI-theorist endorses moral mind-independence as a conceptual or general feature of moral reasoning and how we (ought to) deal with morality in principle. One consequence of this understanding with regard to the different interpretations of moral truth-conditions then is that every relativist understanding of moral matters necessarily amounts to a conceptual or general confusion about morality, or counts as an incompetence to deal with it. For any relativist interpretation of moral matters is heavily at odds with the feature of mind-independence if relativism means that the truth-conditions of moral judgments are determined by people’s mental states. Hence, if moral mind-independence is a fundamentally necessary feature of having correctly understood what it means to reason in moral terms, then thinking and speaking in a relativist fashion must amount to a very general confusion. It follows from this that anyone, no matter her reasons, who thinks and speaks as if relativism were true must be heavily mistaken about morality as such.

Now I think that we have good reasons to be skeptical about the appropriateness of this verdict. After all, not only do people often treat morality in a relativist fashion, but this treatment can indeed be supported by many different factors. For instance, there is a long-standing and ongoing philosophical debate about the (possible) truth of moral relativism. Such debate also rests on results from many other scientific areas such as psychology and anthropology. Thus, psychologists have analyzed in detail children’s development and their learning of how to deal with morality, whereby they discovered that children tend to proceed to engage with morality in an objectivist fashion and then, as they grow up, tend to think and speak in a more relativist way (Schmidt et al. 2017). In addition, there is anthropological investigation of the phenomena of moral pluralism and diversity that are often cited when arguing for the correctness of relativism (Rachels 2001; Wellmann 2001).Footnote 3

Of course, this is all far from saying that relativism is or must be true. When it comes to the results of anthropology, for instance, one may argue that, ultimately, moral diversity does not count in favor of the relativist theory, or that the supposed diversity of moral beliefs is not as strong as people have supposed it to be (Rachels 2001). The salient point, however, is that the relativist interpretation of morality can be and often is supported by genuine reasons to begin with. Therefore, one may have trouble in accepting a relativist understanding of morality because relativism is a repugnant view, but nevertheless, dealing with morality in a relativist fashion must not rest on a confusion about morality nor does it amount to an incompetence in dealing with moral matters. For nothing about the relativist interpretation of morality appears absurd or off-track. On the contrary. According to MI-theories, however, the relativist understanding of morality is just that: hopelessly mistaken from the very outset.Footnote 4

Let us now turn to constructivism. So, even if my argument is correct so far, one may wonder whether constructivism fares any better. After all, my aim is not just to raise objections against MI-theories, but also to show that constructivists can avoid their difficulty. But are there not Kantian constructivists who endorse moral objectivism too? And does not the same argument I have presented against MI-theories therefore also hold against Kantianism?

I think not. To understand why, let us consider two things: first, the constructivist’s theory; and second, the dispute between constructivist objectivists and relativists. When it comes to the former, recall that constructivism in metaethics is the theory that moral truths and facts are not discovered or tracked, but are the outcome of a procedure applied to the practical standpoint. Now take the quarrel between Kantians and Humeans in the constructivist debate, where the question is whether or not objective moral facts follow from reflection on the practical standpoint (Street 2008; 2010). Humeans claim that they do not, while Kantians think that they do (ibid.).Footnote 5

The dispute can be explained as follows. Both Humeans and Kantians engage in reflection on the practical standpoint. They thus investigate some attitude, outlook, or emotion, and then present arguments for or against the view that objective moral truths are (not) entailed within this attitude. Accordingly, Humeans prefer a relativist interpretation of moral truths because their analysis of some attitude leads them to the conclusion that if moral truths were entailed in this attitude, they could only be understood in a subjectivist sense. Kantians, by contrast, reject subjectivism because their carrying out of the constructive procedure leads them to argue for objectivism.

To show which of these conclusions is correct is not my aim here. What matters is that even Kantian objectivists can make sense of the Humean relativist understanding because their endorsement of objectivism is only the result of having engaged in the constructive procedure. Therefore, Kantians will put to Humeans that they have made some mistake in arguing for the relativist understanding of morality, but this mistake for the Kantian is neither absurd, nor does it demonstrate a confusion or incompetence. It is rather the result of not having correctly carried out the constructive procedure or having misinterpreted its results. This is far from showing that Humeans in general lack the right understanding of morality as such.

On these grounds, constructivism is the more attractive view over any MI-theory because it does not have to claim that every agent who thinks and speaks in a relativist fashion is generally mistaken or does not have an adequate understanding of the very concept of morality. Constructivism thus can accommodate the fact that people often understand morality in a relativist sense and therefore can take seriously and account for our objectivist and relativist understanding of morality.Footnote 6 Constructivists therefore employ an attitude that Harman describes best when he states:

It turns out to my surprise that the question whether there is a single true morality is an unresolved issue in moral philosophy. […] Strangely, only a few people seem to be undecided. Almost everyone seems to be firmly on one side or the other, and almost everyone seems to think his or her side is obviously right, the other side representing a kind of ridiculous folly. This is strange since everyone knows, or ought to know, that many intelligent people are on each side of the issue (2001, 166).

5 Three Objections

In this section, I want to answer three objections against my argument. The first is that I have only considered the possibility of spelling out moral relativism in terms of mental states. However, there might be other reasons that support the relativist interpretation of morality.Footnote 7 For instance, one could argue that only subjectivism must make reference to people’s beliefs and attitudes, while intersubjectivism is explained by referring to human practices and conventions (Sayre-McCord 1988, 15). For this reason one may object that my argument is at best limited because while MI-theories may not be able to accommodate some forms of moral relativism, they may well be able to accommodate others.

I am willing to grant that this objection is somehow valid. But nevertheless, there are two problems with it. The first is that one could hold that mental states are an essential part of conventions and human practices, thus casting doubt on the idea that one could justify relativism without any reference to mental states like beliefs and attitudes in the first place. But even if this is wrong, the objection will backfire. This is because the MI-theorist’s rejection of moral relativism is, as I have shown, one of the major aims of MI-theories in virtue of their adoption of the feature of (moral) mind-independence. But if the last objection is correct, then one has to ask whether the adoption of moral mind-independence really is such a good way to argue against relativism in the first place. After all, if there were so many other versions of relativism that MI-theories can account for, why think that moral mind-independence counts against relativism at all? What MI-theorists thus would have to explain is why forms of mind-dependence relativism are more troublesome or repugnant than other versions such as an institution-based form of relativism. But I do not know of any such arguments being proposed by advocates of MI-theories.

A second objection against my argument claims that I have neglected a very simple but possible response that advocates of MI-theories could give to my argument, namely that there is no problem at all with claiming that those who are dealing with morality in a relativist fashion just are that: heavily mistaken or confused about morality in the first place.Footnote 8 After all, the participants of common moral practice certainly are not free from error, and many people do make mistakes in the moral domain. Thus, why should there be a problem with saying that those who think and speak in a relativist fashion are not just mistaken and even confused?

This objection also has its problems because it would turn advocates of MI-theories into error-theorists like Mackie and others who argue that common morality rests on a profound mistake (Mackie 1977; Joyce 2007). The trouble is that advocates of MI-theories, especially IMRealists, certainly consider themselves not to be error-theorists about moral practice; on the contrary, they often point out that their theory actually best accounts for what is going on in common moral practice (Dancy 1986; Brink 1989, 24). Hence, for MI-theorists it would be self-contradictory to claim that their theories best account for the phenomena of moral practice on the one hand, while holding that a large part of what is going on in that very practice is profoundly wrong on the other. Therefore, transforming into error-theorists appears neither an easy nor an attractive proposition to advocates of MI-theories.

A third objection might be raised against the importance that I ascribe to moral phenomenology in general. After all, it appears that I could have simply argued that – if my diagnosis is correct – MI-theories imply the falseness of moral relativism and that this is somehow problematic.Footnote 9 So, the objection questions the need to introduce moral phenomenology to begin with.

Nevertheless, while the objection may seem valid initially, the trouble here is that without introducing phenomenological investigation and discussing its importance for what it means to do metaethics, to allow for the possibility of relativism appears ad hoc. Moral relativism surely is not an attractive theory and it even has quite dangerous and repugnant implications. Therefore, there needs to be a reason or rationale for allowing for its truth other than merely allowing for it. Moral phenomenology and its fundamental importance for metaethical investigation give us exactly this rationale on account of two things: first, because accounting for what is going on in moral practice is of central importance for any metaethical theory; and second, because it is true of that practice that people often think and act as if moral relativism was true. Thus it follows that any metaethical theorist has good reason to at least allow for the truth of the moral relativist theory if she wants to present a plausible metaethical view to begin with.

6 Conclusion

In this paper I have argued that we have reason to favor metaethical constructivism over MI-theories (such as IMR and quasi-realism) in virtue of the constructivist being better able to accommodate moral phenomenology. My point was to show that MI-theories fail to make sense of agents’ dealing with morality in a relativist fashion because a consequence of MI-theories is that any relativistic interpretation of morality must be rejected as a mere confusion or sign of incompetence in dealing with morality. However, this verdict is too strong, for not only do many people think of morality in a relativist way; there are, in addition, often genuine reasons for doing so. Thus, there is a long-standing and ongoing philosophical debate about the (possible) truth of relativism and this debate itself rests on results from numerous scientific areas such as anthropology and psychology. Therefore, the relativist interpretation of morality does not seem to be as absurd as the MI-theorists claim after all, for it can be and often is supported by profound reasons. Constructivism, by contrast, can be shown to make perfect sense of moral phenomenology insofar as it leaves room for both the objectivist and relativist understanding of morality.

Given that being able to account for moral phenomenology counts as an important criterion for evaluating the attractiveness and plausibility of metaethical theories, there is good reason to favor metaethical constructivism over any kind of MI-theory, whether in the form of IMR or non-cognitivist quasi-realism.