MORAL DIVERSITY AND CAUSALLY POWERFUL MORAL PROPERTIES IN CORNELL REALISM*

DIVERSIDADE MORAL E PROPRIEDADES MORAIS CAUSALMENTE PODEROSAS NO REALISMO DE CORNELL

ABSTRACT

There are two important objections against moral realism: The objection of moral disagreement and the objection of the explanatory irrelevance of moral facts. Both objections aim to reject objective moral facts, so any realistic position must consider these objections. The ways in which these objections are constructed as well as the possible responses that can be given to them vary depending on the kind of moral realism in question. For this text, I consider how the two objections can be mobilized against Cornell’s naturalist realism, as well as the responses available.

Keywords:
Moral realism; Cornell realism; Explanatory irrelevance; Moral disagreement; Moral Twin Earth

RESUMO

Há duas objeções importantes ao realismo moral: a objeção do desacordo moral e a objeção da irrelevância explanatória dos fatos morais. Ambas as objeções visam rejeitar a existência de fatos morais objetivos, portanto qualquer posição realista deve levá-las em consideração. As formas pelas quais essas objeções são formuladas, bem como as possíveis respostas que podem ser oferecidas a elas, variam de acordo com o tipo de realismo moral em questão. Neste texto, examino de que maneira essas duas objeções podem ser mobilizadas contra o realismo naturalista de Cornell, assim como as respostas disponíveis.

Palavras-chave:
Realismo moral; Realismo de Cornell; Irrelevância explanatória; Desacordo moral; Terra Gêmea Moral

Introduction

Realism is, in general, a claim of existence and ontological independence of objects in a certain domain. For the moral case, realism typically endorses the theses that typical moral assertions as “ϕ is wrong” or “ϕ is good” when literally constructed, are either true or false and that at least some of them are true. Moreover, the truth of these sentences is, broadly speaking, independent of linguistic practices or conceptual schemes, but reflects objective moral fact.

The nature of moral facts, however, is an issue that goes beyond realism itself and leads to a metaphysical problem concerning the nature of what exists. It is possible for a realist to both commit to the existence of sui generis moral facts and properties – simple parts that integrate our reality – and, on the other hand, explain these facts and properties from a naturalistic perspective.

As a general thesis, naturalism is the ontological position that the natural world is the foundational unit of reality and that, therefore, we should only commit to the existence of natural properties. Consequently, any discourse that is not obviously referring to facts and properties of this category must be naturalized, and if that is impossible, eliminated. Specifically for the moral case, naturalism is the thesis that, if there are moral properties, they must be natural properties and not properties of a special kind.

There are many ways to be a naturalistic moral realist, one of which was developed by Richard Boyd, David Brink, and Nicholas Sturgeon in the mid-1980s at Cornell University. By its initial proponents, this version of realism is called Cornell realism.

Cornell realism commits to the following theses:

1. There are moral facts constituted by the instantiation of moral properties (Boyd, 1988, pp. 182).

2. Moral statements describe moral facts and, therefore, are true or false; furthermore, since moral facts exist, some moral sentences are true (Boyd, 1988, pp. 182).

3. The truth-conditions of moral statements are largely independent of current moral opinions or theories (Boyd, 1988, pp. 182).

4. Moral properties can be included in a naturalistic perspective of reality in which only natural properties exist (Sturgeon, 1985, pp. 239).

5. Moral properties supervene on certain natural properties. Supervenience is the thesis that there cannot be a difference in the moral characteristics of an object without a difference in its non-moral characteristics (Sturgeon, 1985, pp. 247).

6. Moral properties supervene on certain natural properties related to human well-being and flourishing, necessary conditions for survival, and other needs associated with basic well-being, desires of various kinds, and distinctly human needs (Brink, 1984, pp. 121-122).

7. Moral properties are irreducible to natural properties, with one property being reducible to another only if they are the same property (Brink, 1989, pp. 179-180). Therefore, moral properties are not identical to natural properties of a different nature.

These theses characterize Cornell realism as a realistic, naturalistic and non-reductive version of metaethical theory.

There are some objections that can be raised against the proposal. In this paper, I explore two objections against moral realism as developed against Cornell realism: The explanatory irrelevance of moral facts, and the objection of moral disagreement.

The first objection, the explanatory irrelevance of moral facts, rejects the instantiation of moral facts constituted by objective moral properties. The second critique, concerning the moral disagreements, can be constructed against Cornell realism in two ways: the first according to the Moral Twin Earth argument, which culminates in the rejection of the semantic proposal of Cornell realism. In its second form, the critique mobilizes the variety of moral discourses as evidence of the absence of objective moral properties.

The explanatory irrelevance of moral facts

There are different versions of the critique regarding the explanatory irrelevance of moral facts, but generally, this objection highlights that the postulation of objective moral facts does not contribute to enhancing the explanatory power of a metaethical theory. Subscribing to this argument is a principle of ontological economy, which suggests that theories with fewer ontological commitments are preferable to more extravagant theories — unless these ontological additions increase the explanatory power of a theory. Therefore, moral realists must offer a defense of the existence of moral properties as being significant for the explanatory power of a theory. However, according to authors like Gilbert Harman, moral properties seem dispensable for that purpose.

According to Harman, “if you round a corner and see a group of young hoodlums pour gasoline on a cat and ignite it, [...] you can see that it is wrong” (Harman, 1988, p. 120), but nothing in this situation definitively indicates that the observation made depends on the presence of a distinctively moral property – wrongness. It is possible that the observation of the action as wrong is just a reflection of a certain moral sense of the observer developed from certain moral principles that the observer previously assumed (Harman, 1988, p. 120). Thus, as it is the case that the existence of any moral property is not necessary to determine the observation, it seems to be completely irrelevant to our explanation whether your intuitive immediate judgment accurately describes a state of affairs (Harman, 1988, p. 122).

There is more than one way to understand this conclusion: either it concerns the irrelevance of moral properties, or the irrelevance of moral facts. Taken as an argument about the irrelevance of moral facts, the argument is somewhat controversial. Sturgeon, for example, questions whether

[…] is it true, as Harman claims, that the assumption that the children are really doing something wrong is “totally irrelevant” to any reasonable explanation of your making that judgment? Would you, for example, have reacted in just a way, with the thought that the action is wrong, even if what they were doing hadn’t been wrong, and could we explain your reaction equally well on this assumption? (Sturgeon, 1985, p. 246, original emphasis).

If moral facts are irrelevant, what the argument claims is the possibility of the existence of a situation very similar to the one previously described involving teenagers, but with the only distinction that it is not wrong – i.e., with the absence of the moral fact. This case, however, would not be compatible with the thesis of the supervenience of the moral on the non-moral.

According to the supervenience thesis, there cannot be a difference in the moral characteristics of an object without a difference in its non-moral characteristics. Supervenience is one of the commitments of Cornell realism and does not need to be taken as true by Harman. However, it seems quite counterintuitive that the moral value of an event can vary in an arbitrary and incoherent manner. Minimally, if two situations appear indistinct in their nonmoral aspects, they should be indistinct in their moral aspects, and in this case, for the moral situation to be different, there must also be a difference in the non-moral situation. For it not to be a wrong action, something in the action of the teenagers must be, in some way, different.

Being “morally wrong” in the teenagers’ case consists of being an action of intense, gratuitous cruelty against a defenseless being. For the action not to be incorrect, it must differ in this crucial aspect. But, if it were the case that the teenagers were engaging in a different enough action that it was not wrong – perhaps kindly playing with a cat, or offering it some treats, for example – it would not be reasonable to consider this action wrong in the same way (Sturgeon, 1985, p. 247). Unless the observer is mistaken in their perception of the non-moral facts – that is, if they correctly understand what is going on – this counterfactual situation seems implausible. Thus, it seems that the action being wrong is an important part of explaining the moral judgment made.

If moral facts are relevant to explaining moral practices, then the argument from irrelevance is not that appealing. However, this is not the only interpretation of Harman’s argument from irrelevance. It is still possible to interpret this argument as applying to the irrelevance of considering moral facts as instantiations of distinctively moral properties. In this interpretation, it is not feasible for an observer, as in the narrated hypothesis, to consider an action that is not wrong as wrong when observer errors are excluded. If the teenagers were merely kindly playing with a cat, this would not be wrong, and the observer would form a distinct moral judgment or none. However, the observation of a wrong action does not necessarily demand the instantiation of the wrongness property. The core of this problem of irrelevance against moral realism is that what makes an action of cruelty against the cat constitute a negative moral fact does not need to be that the action of cruelty instantiates the objective property wrongness.

Indeed, if there are moral properties, judgments like the one made by the observer in Harman’s example could be correct or true, to the extent that they adequately described the presence or absence of these properties. However, even if such moral properties do not exist, an observer could make the same moral judgment – making it entirely irrelevant, in this sense, whether the judgment is true because it describes an objective state of affairs. It is possible that moral facts are the result of certain affective interactions or human conventions with non-moral facts, making it feasible for the observer to consider the action wrong, whether or not objective moral values exist (Harman, 1988, p. 122).

It could be the case that moral facts result from moral education that presents certain rules for using terms like ‘right’ and ‘wrong,’ making them descriptive of certain types of events. This condition would meet the thesis of moral supervenience – it is true that changes in perceived moral value reflect changes in non-moral aspects of events, and it would not be possible for an observer to evaluate identical circumstances in non-moral aspects with distinct moral values. However, supervenience would not be due to an essential relationship between moral and non-moral properties, because it can be explained as a demand for coherence in the use of moral terms, or even a certain regularity in the sensitivity of moral subjects.

Because this explanatory possibility is available, an objection of metaphysical extravagance can be made against moral realism. The extravagance would consist in unnecessarily postulating the existence of moral properties, since moral phenomena could be perfectly understood even if such properties do not exist. If moral properties are extravagant postulations, the argument from explanatory irrelevance affects the first three theses of Cornell realism, which assert that there are moral facts constituted by the instantiation of objective moral properties.

Although the argument from explanatory irrelevance challenges the nature of what constitutes moral facts, it does not prove that moral facts are irrelevant to moral observation. In fact, moral facts play a crucial role in explaining how agents perceive moral situations. The argument mainly disputes the specific nature of these facts. Some philosophical positions might assume a heavier ontological load by positing moral facts as requiring special properties. However, Cornell realism operates within a naturalistic framework, suggesting that if moral properties exist, they are not considered special or extraordinary. Therefore, under Cornell realism, moral facts involve moral properties, but this addition is only quantitative, not qualitative, in terms of properties. Consequently, the objection that Cornell realism is ontologically extravagant becomes less compelling.¹

It is still a possible argument that moral properties might be dispensable, because it is true that non-realist theories are viable explanatory hypotheses. Since non-realist theories provide explanations for moral phenomena that do not postulate the existence of moral properties, these theories are more ontologically economical. However, for this to be an argument against moral realism, it is also necessary to show how the most optimized and complete explanation for the moral phenomenon is anti-realist, which would culminate in the thesis that objective moral properties are, in fact, dispensable for best explanations. As the debate between realist and non-realist options remains open, there seem to be no definitive reasons to endorse this premise. Therefore, although realism posits a greater number of properties, this is not enough to dismiss Cornell’s realists.

On the other hand, considering the argument from ontological parsimony, the viability of Cornell realism is bolstered by its naturalist thesis. To accommodate naturalism, Cornell realism needs to provide an explanation of how moral properties are natural properties. On this topic, according to Cornell realists, there are natural properties on which moral properties supervene, which essentially connect the domains. However, supervenience is more of a description of the type of relationship obtained than an explanation about that relationship. To be constituted as a naturalist proposal, Cornell realists need some kind of definition for moral properties that fits them into a naturalistic framework.

A naturalist definition, as necessary, can be provided, for example, if moral terms rigidly designate some specific configuration of the natural world (Timmons, 1999, p. 58). Rigid designators are terms capable of selecting a single referent in any possible world where the referent occurs and never selects anything else. When two terms rigidly designate the same referent, it is possible to establish a metaphysical identity between these two terms, and therefore, a definition of one from the other. “Hesperus is the same as Phosphorus and both are the same as Venus,” or “water is H₂O” are classic examples of definitions that employ identities of this nature.

Essentially, as the names in these definitions: <‘Hesperus’; ‘Phosphorus’; ‘Venus’>, <‘water’; ‘H₂O’> are rigid designators, and as each set respectively names the same object or substance, it is possible to establish an identity between the terms. Since rigid designators name the same referent in all possible worlds, the identity established is also true in all possible worlds – it is necessary. Necessary identities are sufficient for definitions, so it is a definition for water that it is H₂O, and the same holds for other cases.

If moral terms and terms for natural properties are rigid designators, there is the possibility of constructing metaphysical identities as explored also in the moral case, provided that a moral term and a term for a natural property necessarily designate the same referent. This is probably the most common characterization of Cornell realism, which would have been an extension of the historical-causal semantics in the style of Kripke-Putnam to include moral terms as rigid designators of natural properties. This characterization was especially disclosed as a possibility to respond to Moore’s open question argument, which contends that there are no analytically true natural definitions for moral terms. I do not explore the open question argument in this paper; however, there is an important objection to the thesis that moral terms are rigid designators: the Moral Twin Earth argument (Horgan; Timmons, 1992).

Moral Twin Earth Argument

The Moral Twin Earth argument is a thought experiment that posits the existence of a planet very similar to Earth – Moral Twin Earth – where beings are very similar to humans and similar social organizations are found. Notably, among the natives of Moral Twin Earth, there is a vocabulary that functions very similarly to the moral vocabulary of Earth, providing initially good evidence that the twin moral terms might be translated or identified with Earth’s moral terms.

Assuming that the moral terms are rigid designators of certain natural properties, the meaning of each moral term is exhausted by the specification of this referent, meaning that the identification of moral terms with their twins depends on the identity of the referents. However, it is metaphysically possible that, despite apparent similarities, the uses of the twin moral terms are regulated by different natural properties than those regulating the uses of moral terms on Earth. It could be the case that the moral theories endorsed on Earth and Moral Twin Earth are different and have led to distinct historical-causal processes, such that a fundamental moral term on Earth like ‘wrong’ might capture the natural property D, while its twin term ‘wrong_T’ refers to a distinct property C. Properties C and D would both be capable of performing the role associated with the term ‘wrong’ or ‘wrong_T’ within each theory, but since C and D are distinct properties, ‘wrong’ and ‘wrong_T’ would not have the same meaning (Timmons, 1999, p. 62).

The consequence of distinct meanings for terms that perform the same role would be a kind of interplanetary relativism. Essentially, moral judgments made by Earthlings would have different meanings from those made by inhabitants of Moral Twin Earth, so the same objection normally applied to relativist positions would also be applicable to Cornell realism: The impossibility of genuine agreements or disagreements among speakers who adopt different moral theories.

A genuine disagreement is characterized as a dispute over the literal content expressed by the speakers, that is, a dispute over the truth value of propositions. For it to be a genuine disagreement, therefore, the statement – ‘a’ – must be opposed to a statement ‘not-a’. Similarly, a genuine agreement involves an agreement on the truth value of the same proposition, meaning both speakers affirm ‘a’ or both affirm ‘not-a’. However, if moral terms are rigid designators of natural properties and twin moral terms designate different properties, the content expressed by a moral judgment on Earth does not contradict or reaffirm a moral judgment from Moral Twin Earth.

For example, while on Earth “it is wrong to lie” expresses the proposition that ‘the action of lying instantiates property D’, on Moral Twin Earth, as the twin moral terms capture a distinct property, the judgment that “it is not wrong to lie” when articulated by a twin speaker expresses the proposition that ‘the action of lying does not instantiate property C’. These propositions are not contradictory because while one speaker refers to the presence of a certain property D, the other refers to the absence of a certain property C (Horgan; Timmons 2000, p. 143) and both propositions could be true. Thus, although there appears to be a substantive moral dispute about the status of the action of lying, actually, there is just a case of confusion between homonymous terms.

To be fair to the intuition that there is a moral dispute between speakers from Earth and their twins, it is necessary to assume that the twin moral terms are linguistic equivalent to the moral terms. However, despite differences concerning the natural properties involved, the terms are equivalent, this indicates that these terms are not rigid designators of these properties. Nevertheless, in this case, it is not possible to construct metaphysical identities like those authorized by historical-causal semantics. Ultimately, without the identities, the naturalistic definitions are lost, and Cornell realism cannot demonstrate how natural properties are included in a naturalistic framework (Timmons, 1999, p. 62).

The Moral Twin Earth objection is the most well-known critique of Cornell realism, but it relies entirely on the assumption that moral terms are rigid designators which do not need to be endorsed by Cornell realists. On the contrary, the identity between moral and natural properties in this format implied by rigid designation is rejected by Cornell realists due to the possibility of multiple realizability of moral properties:

Moreover, the multiple realizability of moral facts and properties may provide another reason for resisting the identification of moral and natural properties. A plausible claim about a variety of property types and tokens (properties and property instances) is that they could have been realized in a variety of different ways. Functionalist theories of mind, for example, are based partly on this kind of claim about the one-many relationships between mental states and physical systems […] A similar claim seems plausible about moral properties (Brink, 1989, p. 158).

Essentially, there are many ways for something to be right, wrong, good, or bad, so a single moral property can be related to different natural properties. Therefore, moral properties are instantiated in events and circumstances composed of distinct natural properties. However, because they are multiply realized, even though there is a connection between particular instantiations of moral properties and some configuration of natural properties, these properties cannot be identical, and there are two reasons for that: First, because identity is a transitive relation, so x = x₁, x₂, x₃ … is only true if x₁ = x₂ = x₃… As the properties that multiply realize a moral property are distinct, an identity in these terms would characterize a misuse of the identity relation. Second, because identity relations imply reductions – when two properties are equal, they are reducible to each other.

The first of those difficulties could be remedied by considering not particular properties but sets of properties. However, the second difficulty could only be avoided if Cornell realism were a reductive version of naturalist realism. Therefore, Cornell realism in no way depends on the thesis that moral terms rigidly designate natural properties, and there are no reasons to present it as a version of the historical-causal semantics applied to moral terms.

Since moral terms are not rigid designators of natural properties related to the realization of moral properties, the regulation of the use of moral terms only can depend on moral properties themselves (Boyd, 1988, pp. 186). However, this implies that the Moral Twin Earth argument, which suggests that there are no genuine disagreements because moral terms and their twins have distinct meanings, is ineffective. If moral terms indeed designate moral properties, then both the terms and their twins should share the same meaning, leading to genuine disagreements. Therefore, Cornell realism faces no counterintuitive conclusions regarding disagreements with Moral Twin Earth inhabitants.

However, even if the Moral Twin Earth argument is not decisive against Cornell realism, there are still two difficulties to be addressed: The first is given that naturalistic definitions are not guaranteed by a semantic theory concerning moral terms, Cornell realists need to offer some explanation for the connection between certain natural properties and moral properties. Even if it is true that moral properties supervene on natural properties, it is still necessary to explain why there is an essential dependency of moral properties on the natural properties of their base and why moral properties, even though intrinsically related to their natural bases, are distinct properties. Secondly, even if it is assumed that disagreements involving different moral evaluations are semantically genuine, it is still necessary to clarify on what aspect the disagreement resides; after all, if moral judgments describe the instantiation of objective properties – wrongness, goodness, rightness, etc. why is not there a unified theory about moral value? These two objections are discussed in the following sections.

The connection between moral and non-moral: the “is” of composition

Cornell realism is a non-reductive version of naturalism, meaning that moral properties, if they exist, are natural properties distinct from other natural properties. However, even though they are irreducible, moral properties are intrinsically related to certain natural properties through the thesis of supervenience. Additionally, because Cornell realists are naturalists, moral properties depend on the presence of some natural properties, such that it is not possible for something to have a moral property without a natural property. However, how can we explain these connections between distinct properties?

According to Brink, “[moral] properties, so construed, are […] composed by organized combinations of natural and social scientific facts and properties” (1989, p. 159), in the same way that:

[…] tables are constituted by certain combinations of microphysical particles, large scale social events such as wars and elections are constituted by enormously complex combinations of smaller scale social events and processes, biological processes such as photosynthesis are composed of physical events causally and temporally related in certain ways (Brink, 1989, p. 159).

Common to these composition relationships is that they constitute organized complexes of parts that form a whole. In this sense, compound events are nothing more than a certain organization of smaller parts. Just as physical objects like tables are nothing more than certain energetic arrangements, moral properties are specific arrangements of natural properties typically associated with human well-being and flourishing.

However, importantly, “[...] we should distinguish between the ‘is’ of identity and the ‘is’of constitution” (Brink, 1989, pp. 157). So that, even though a table is made up of smaller microphysical events, it may not be identical to these events. Similarly, moral properties that are constituted by natural properties of another kind may not be identical to the properties of their base.²

Because they are not identical, the possibility of multiple realizability of moral properties is easily accommodated by Cornell realism. However, since the natural properties that compose a moral property have this role only contingently, it is not possible to define the moral property extensionally from the natural properties that compose it (Boyd, 1988, p. 217). Instead, a moral property is defined by Cornell realists based on a certain role that this property plays in a moral theory (Brink, 1989, p. 197). Similarly, what makes a natural property relevant for the composition of a moral property is its role in helping to compose the structure of causal powers that are relevant in a moral theory.

According to Cornell realism, the identification of moral properties is guided by a broadly consequentialist moral theory.³ Consequently, moral rightness is related to the promotion of human goods associated with wellbeing and human flourishing:

There are a number of important human goods, things which satisfy important human needs. Some of these needs are physical or medical. Others are psychological or social; these (probably) include the need for love and friendship, the need to engage in cooperative efforts, the need to exercise control over one’s own life, the need for intellectual and artistic appreciation and expression, the need for physical recreation, etc. (Boyd, 1988, p. 203).

Exactly which human needs are relevant to well-being and flourishing is itself “a potentially difficult and complex empirical question” (Boyd, 1988, p. 203). Nevertheless, Boyd argued that these goods – or, more precisely, instances of their satisfaction – tend to be homeostatically clustered across a wide variety of circumstances.

For the purposes of this paper, the most important implication of this view is that the natural properties composing a moral property perform the functional role associated with a moral concept while the structure of the cluster itself – that is, the relations among those natural properties – constitutes a condition for the realization of that functional role (Boyd, 1988, p. 203).

From this characterization, it follows that different properties may be capable, in certain arrangements, of producing the same causally relevant effects and, therefore, of composing a moral property. Since the causal powers that characterize a moral property are realized through the interaction among other properties, rather than mere co-instantiation, it is possible to maintain that moral properties are distinct from the natural properties of their bases – they constitute a distinct set of causal powers.⁴

On the other hand, since the causal powers that characterize a moral property depend on the powers and relationships between natural properties, there is an essential relationship between the properties at both levels, which underpins, for example, the covariation thesis endorsed by the supervenience of the moral on the natural. In general terms, moral changes reflect some type of change at the non-moral level because if a moral property ceases or begins to exist, this reflects a change in natural properties. On the other hand, this does not mean that any change in the non-moral properties of an event results in a change in the moral aspect, because more than one possible arrangement of properties can realize a moral property. Thus, changes in moral aspects always reflect a change in non-moral aspects, but not vice versa.

Finally, mapping the relevant changes, or the properties most associated with the composition of moral properties, is an empirical task. Then to determine what makes something wrong or right, or when there is a change in the moral status of an event, it is necessary to investigate empirically (Boyd, 1988, p. 202). As the target of empirical investigations, it is expected that they will constrain the observations of moral agents. However, if it is the case, “then what explains […] the persistence of cultural variability in moral beliefs?” (Boyd, 1988, p. 185).

Moral Disagreements and Hard Cases

To explain moral diversity, it is necessary to differentiate the kinds of divergence that can compose morality. Some moral divergences can be explained as simple observational errors – misunderstanding a situation in its non-moral aspects can lead to an incorrect moral judgment. Divergences caused by errors like these are not a challenge to moral realism because they do not necessarily reflect a divergence in moral values. There are also some mechanisms that can distort moral judgments, leading to different interpretations of the same situation and contributing to the existence of disagreements. For example, political or religious ideologies, personal interests, or intense emotions can be external mechanisms that obscure moral judgments. Although this type of disagreement is more difficult to resolve than those involving simple misunderstanding, the existence of this type of divergence is also not especially problematic for moral realism. As external factors can distort the perception of even the most indisputably objective aspects of reality, this type of divergence does not indicate the absence of objective moral properties.

However, it does not seem that every moral disagreement is the result of distortions or mistaken perceptions. In some situations, it is reasonable to assume that both interlocutors in a discourse have valid reasons for supporting the moral judgments they make. There are also other situations where the reasons and available data do not seem sufficient to definitively determine the moral value of an action, or what should be done. In situations of the first type, we have persistent moral disagreement; for the second type, we have cases of moral dilemmas or hard cases.

Moral dilemmas and persistent disagreements can be mobilized as an argument against moral realism because they may reflect the absence of objective moral truths (Sturgeon, 1985, p. 229): If moral judgments reflect opinions or become true based on social codes and conventions, it is expected that people who participate in different codes or conventions assign divergent moral values to the same situations and that some dilemmas cannot be resolved rationally. On the other hand, if according to moral realism, moral statements describe objective moral properties, then what explains this kind of disagreement or indetermination?

Initially, it is not the case that all cases of divergence in the moral evaluation of a situation reflect a divergent perception of the involved moral properties. There are actions considered bad or wrong, but which in certain circumstances or in the presence of certain types of mitigating factors are seen as permissible, or more acceptable. The case of abortion, for example, may involve the judgment that, although not good, certain circumstances justify its practice. However, agents and moral codes may differ on what these circumstances are and how much the situation is mitigated in the presence of these factors. This type of disagreement does not involve a direct contradiction between the propositions held by the interlocutors; both can agree that “abortion is wrong,” which would safeguard the possibility of an objective moral property perceived by both sides in the debate.

Similarly, some cases of moral dilemmas may be difficult to resolve not because moral agents cannot determine the moral value of actions, but because, given the situations they are placed in, they must choose between courses of action that are equally good or equally bad (dilemmas seem more common in the latter case). The resolution of moral dilemmas of this nature depends not only on identifying the moral value of the event – whether it is wrong or right – but also on other considerations by the agent, such as comparing the options, other commitments previously made, cultural elements, or other factors.

This kind of situation is not especially problematic for Cornell realism, because moral claims will still be true or false by virtue of the presence of objective moral properties. The cases of disagreement that challenge Cornell realism are those that involve assigning different values to the same type of event, that is, cases where one agent affirms while another denies that a particular event instantiates the relevant moral property.

To accommodate those cases of divergence, it is necessary to revisit the characterization of properties adopted by Cornell. According to Cornell realists, being a moral property means being a property composed of a homeostatic cluster of other causally relevant properties. This means that the natural properties that together constitute a moral property can perform certain functional roles assigned by a moral theory to moral concepts.⁵ The observation of a property, therefore, is made by delineating functional roles according to a moral theory. However, the development of a theory depends on certain background conditions formed by contingent aspects (Boyd, 1988, p. 204).

For example, some values associated with human well-being and flourishing are only recognized if a society has had the opportunity to experience them:

Only after the moral insights gained from the first democratic experiments were in hand, were we equipped to see the depth of the moral peculiarity of slavery. Only since the establishment of the first socialist societies are we even beginning to obtain the data necessary to assess the role of egalitarian social practices in fostering good. (Boyd, 1988, pp. 204-205)

As different societies have undergone distinct historical moments and social movements, it is possible that different societies might recognize different needs associated with human well-being and flourishing. The divergence in the background conditions upon which a moral theory was built can lead to divergences in beliefs and moral judgments (Boyd, 1988, p. 213). In such cases, the source of disagreement lies not necessarily in the absence of objective moral facts, but rather in the different epistemic conditions under which moral theories were formulated. From a realist perspective, further theoretical development and empirical inquiry may lead to the revision or improvement of current moral theories, potentially reducing or even dissolving some of these disagreements.

The claim that moral disagreements can be resolved reflects an optimistic commitment of Cornell realists. However, even under the most optimistic realist expectations, many moral disagreements are likely to persist precisely because they fall into categories discussed earlier and involve factors such as extreme emotions, religious, and political allegiances. In cases of this kind, objective moral truths may exist while nevertheless failing to play a decisive role in resolving the dispute.

On the other hand, moral realism does not require that all moral disagreements be resolved. The crucial point is that the persistence of moral disagreement does not, by itself, constitute decisive evidence against moral realism. For this purpose, it is sufficient for moral realist to provide a plausible explanation of why such disagreements occur and persist despite the alleged existence of objective moral truths. If moral disagreement can be accounted for in terms of differing epistemic circumstances, historical contingencies, and other factors that affect moral inquiry, then the anti-realist inference from disagreement to the non-existence of moral facts loses much of its force (Boyd, 1988, p. 223).

Conclusion

Precisely indicating the core of Cornell realism is not a simple task, as there is no unified work from its main proponents, and many fundamental theses crucial to distinguishing this type of realism are sparsely presented between other discussions. Despite this difficulty, one point on which there seems to be agreement among Cornell realists is the irreducibility of this version of moral realism.

The irreducibility of moral properties combined with naturalism constitutes the most attractive aspect of this proposal because it allows responses to the two major objections presented in this text (and perhaps others that I could not explore). However, the viability of this combination depends on two metaphysically important theses that are not developed by Cornell realists: Firstly, Cornell realism depends on the thesis that compositions bring ontological additions to reality – the “is” of composition must to be different from the “is” of identity; secondly, the thesis that properties can be composed of others – moral properties are composed of homeostatic clusters of other properties related to human well-being and flourishing. A more complete defense of Cornell realism should include some clarifications on these two points.

Despite these aspects still needing clarification, it seems that Cornell realism is a quite viable realist proposal. The main objection to this form of realism, the Moral Twin Earth argument, is not effective against the semantic thesis of Cornell. On the other hand, ontological extravagance arguments, such as those by Harman, may be considered, but they are unlikely to be decisive in the absence of a more economically viable non-relist counterproposal. Finally, cases of disagreement and moral dilemmas can be accommodated despite the objectivity of moral properties.

¹

The argument of ontological extravagance can be applied to the number of properties that exist in cases of regressions or infinite postulation of properties. However, this is not the case with Cornell realism.

²

It may seem quite intuitive that composite events or objects are distinguishable from the smaller events or parts that realize them. However, this is a substantive metaphysical claim that depends on a characterization of the composition relationship that is not ontologically innocent. Not all presentations on compositions consider them as implying ontological additions. A classical characterization of mereology – Classical Extensional Mereology, for example, maintains that composition relations are ontologically innocent (Lewis, 1991).

³

Boyd also argued that the same sort of metaethical position can be formulated on the basis of any other plausible competing moral theory (Boyd, 1988, p. 222).

⁴

Cornell realists do not offer a definition for properties or conditions of identity and difference for them. However, Cornell realists would likely benefit from a theory like that of Sidney Shoemaker (1980), according to which a property is a cluster of causal powers (1980, p. 115). When properties are conceived as causal powers, when X and Y necessarily have the same causal contribution, then X and Y are the same property. Additionally, when X and Y have distinct causal powers, they are not the same property.

⁵

This is not the same as saying that the truth-makers of moral claims are theoretical constructions. A theory offers epistemological access to a reality independent of the theory itself.

Data availability:

All data underlying the results of this study are included within the article.

References

• BOYD, R. “How to be a Moral Realist”. In: G. Sayre-McCord (ed.), Essays on Moral Realism Ithaca: Cornell University Press, pp. 181-228, 1988.
• BRINK, D. O. “Moral Realism and the Foundations of Ethics”. New York: Cambridge University Press, 1989.
• ______. “Moral Realism and the Skeptical Arguments from Disagreement and Queerness”. Australian Journal of Philosophy 62(2), pp. 111-125, 1984.
• HARMAN, G. “Ethics and Observation”. In G. Sayre-McCord (ed.). Essays on Moral Realism Ithaca: Cornell University Press, pp. 119-124, 1988.
• HORGAN, T., TIMMONS, M. “Troubles for New Wave Moral Semantics: ‘The Open Question Argument’ Revived”. Philosophical Papers, 21, pp. 153-175, 1992.
• LEWIS, D. “Parts of Classes”. Massachusetts: Basil Blackwell, 1991.
• SHOEMAKER, S. “Causality and Properties”. In P. Van Inwagen (ed.). Time and Cause Dordrecht: Springer Science Business Media, pp. 109-134, 1980.
• SAYRE-MCCORD, G. (ed.). “Essays on Moral Realism”. Ithaca: Cornell University Press, 1988.
• STURGEON, N. “Moral Explanations”. In G. Sayre-McCord (ed.). Essays on Moral Realism Ithaca: Cornell University Press, pp. 229-255, 1985.
• TIMMONS, M. “Morality without Foundations: A Defense of Ethical Contextualism”. New York: Oxford University Press, 1999.

Conflict of interest:

The author declares that there is no conflict of interest.

Handling Editor:

Mauro Luiz Engelmann

Publication Dates

• Publication in this collection
  02 Oct 2026
• Date of issue
  2026

History

• Received
  01 Aug 2026
• Accepted
  06 Sept 2026