1(8) Contextualism in Ethics Gunnar Björnsson In more than one way, context matters in ethics. Most clearly, the moral status of an action might depend on context: although it is typically wrong not to keep a promise, some contexts make it permissible. More radically, proponents of moral particularism (see PARTICULARISM) have argued that a reason for an action in one context is not guaranteed to be even a defeasible reason in every context; whether it counts against an act that it breaks a promise or inflicts pain might depend on the particulars of the situation. In moral epistemology, Timmons (1999: Ch. 5) argues that whether a moral judgment is epistemically responsible depends both on the basic moral outlook of the moral judge and on whether the context of judgment is one of engaged moral thinking, or one of distanced, skeptical reflection. In the former, the judge’s basic moral outlook can serve to justify the judgment; not so in the latter (see EPISTEMOLOGY, MORAL). Our focus here, however, will be on forms of metaethical and, more precisely, semantic contextualism in moral discourse and moral thinking. According to these forms of contextualism (henceforth “metaethical contextualism,” or just “contextualism”), the meaning or truth‐conditions of a moral judgment or moral assertion depend not only on the properties of the act it concerns, but also on features of the context in which the judgment or assertion is made, such as the standards endorsed by the moral judge or the parties of the conversation. If metaethical contextualism is correct, it might be that when two people both judge that abortions must be banned, one judge might be correct whereas the other is mistaken, because they accept different fundamental norms. This would undermine the idea that there are unique correct answers to moral questions. Metaethical contextualism is supported from three directions. First, what is expressed by terms such as “good” and “ought” seems to be context‐dependent when used outside ethics, varying with speakers’ interests and concerns. One might therefore expect similar context dependence when these terms are used to express moral judgments, assuming a corresponding variety of moral interests and concerns. Second, many have thought that deep moral disagreements show that the interests and concerns behind moral judgments do vary in this way. Finally, contextualism promises to make sense of what seems to be an intrinsic yet defeasible connection between moral judgments and moral motivation, by tying the meaning or truth-conditions of moral judgments closely to interests and concerns of moral judges. At the same time, contextualism faces two broad kinds of challenge: to make sense of the seemingly categorical or objective pretensions of moral claims, and to explain why the parties to deep moral disagreement often behave as if they were disagreeing about substantive issues rather than talking past each other. In the sections that follow, we look closer at both sources of support and problems for contextualism. Draft for the 2019 revision of the International Encyclopedia of Ethics (ed) Hugh LaFollette. Wiley. CONTEXTUALISM IN ETHICS 2(8) The Context Dependence of “Good” and “Ought” It is well known that the content of what we say is determined not only by the words we choose, but also by the context in which the words are used. Clearly, what claim is made by an utterance of “I am here now” depends on the time and place of the utterance, and on who is making it: “now,” “here,” and “I” pick out different times, places, and individuals in different contexts. Similarly, the judgments expressed when we say of someone that she is “tall” might variously measure her against the length of others her age, or others her age and sex, or against that of, say, professional basketball players. Interestingly, context‐sensitivity also seems to be a feature of expressions used to make moral or normative claims, such as “obligation,” “right,” “wrong,” “must,” “ought,” and “good.” We shall focus on “ought” and “good,” beginning with the latter. Consider a sentence such as: (1) “The weather is good.” If it is uttered by tourists considering a day at the beach, it seems to say that the weather answers to the requirements for a pleasant day at the beach; if uttered by farmers considering crop yields, it would instead be saying that the weather answers to the requirements for a high crop yield. Since these are different requirements, one utterance of the sentence might be true while the other is false, even if the utterances concern the weather at the same time and same place. In light of these and other examples, it seems clear that “good” means, roughly, “such as to satisfy requirement R,” where R varies across contexts (see Mackie 1977: 55–6; for similar suggestions, see Finlay 2014: Ch. 2; Thomson 2008: Chs. 1–4). The relevant requirement, R, is often clear from the kind of thing called “good”: “a good conversation” typically means one satisfying typical requirements on conversations, and “a good knife” means one satisfying the requirements that we typically have when using knives, i.e., cutting and handling well. At other times, R is clear from our specific knowledge of the interests and concerns of the speaker: a scenographer talking about “a good knife” might have in mind one fitting a certain scene well, whether or not it cuts well. When the kinds of requirements in question are otherwise unclear, they are often made more explicit: “good for the crops,” “good for a day at the beach,” “good for you but not for me.” “Ought” seems to display similar patterns of context dependence. In general, to say that something ought to be the case or that someone ought to do it seems to express that it is “favored” among some set of relevant alternatives. What varies from context to context are the considerations that select the relevant alternatives and favor one of them. For example, depending on context, an alternative might be favored because it is probable given the evidence, or makes the achievement of some end most probable, or conforms better to some ideal: (2) “Since she left almost an hour ago, she ought to be here soon.” (3) “[Given what we knew,] It ought to have dissolved; I wonder why it didn’t” (Mackie 1977: 73). 3 (4) “To look more casual, Jill ought to wear blue jeans.” (5) “To open the safe, Jill ought to use a Phillips screwdriver” (cf. Wedgwood 2007: 117). As with “good,” context might make the italicized explications of the relevant considerations unnecessary: watching Jill struggling to open the safe, we can utter (5) without the explication, to the same effect. Notice that “ought” judgments that relate to goals or ideals need not express the speaker’s endorsement of the goal or ideal in question: (4) and (5) might be uttered as pieces of conditional advice, or as matter‐of‐fact information about dress codes or locksmithery. Similarly, “ought” judgments can be made relative to a restriction of alternatives that we do not endorse, as in: (6) “Since you are not going to stop shooting up heroin, you ought at least to use clean needles” (cf. Jackson 1985: 181–2; Wedgwood 2007: 119). Judgments can also be made both relative to evidence that we ourselves possess, and relative to evidence possessed by an agent who we think ought to do something. For example, looking down from above on someone making his way through a maze, we might be correct in saying both of the following: (7) “He has no way of knowing it, but he ought to turn left at this point.” (8) “Given what he knows, he ought to turn right at this point” (Wedgwood 2007: 118; my italics). Again, although explicit indications of the relevant states of evidence often help, the right conversational context makes them redundant. (For different ways of understanding the context dependence of “ought” judgments, see Mackie 1977: Ch. 3; Wedgwood 2007: Ch. 5; Price 2008: Ch. 2; and Finlay 2014: Ch. 3. For criticism, see Thomson 2008: Chs. 10–11.) From Semantic Context Dependence to Metaethical Contextualism The discussion in the previous section left open numerous questions about the exact analyses of “ought” and “good,” and said nothing about how context determines the content of claims involving these terms. What seems clear, however, is that what utterances using those terms express does depend on context, and more precisely on interests, goals, and ideals that are relevant in that context. Our question now is whether this extends to uses of “good” and “ought” in ethical contexts, such that two people who are considering whether an action is morally good, or whether it ought to be performed, might be asking different questions. In other words, does the contextual variation support metaethical contextualism? We have already seen one possible example of this, illustrated by (7) and (8): apparently some “ought” judgments mean to identify the alternative that is morally or rationally ideal in relation to the knowledge or beliefs of the agent, whereas other judgments also take into account information that the agent lacks (Jackson 1985; Finlay CONTEXTUALISM IN ETHICS 4(8) 2014). If such information relativity is possible, perhaps differences between those who tie the moral status of actions to their actual consequences and those who tie it to their expected consequences stem from a focus on different moral questions. That there are different such questions is further suggested by thoughts about what one ought to do in cases of normative uncertainty: (9) “I know that the correct theory is either utilitarianism (in which case I ought to choose policy A) or prioritarianism (in which case I ought to choose policy B). But since I have no idea which of these two theories is right, I ought to choose policy C. For on either theory, policy C is a close second.” (Pittard and Worsnip 2017: 168). This line of reasoning seems to presuppose that the ought-claim of the conclusion is compatible with the ought-claims of utilitarianism and prioritanism. The apparent intelligibility of this reasoning, whether sound or not, thus suggests that we readily recognize different ought-questions. More contentious and radical than information relativity is the idea that the content of moral judgments varies with the ideals operative in the context of judgment. Here, proponents of metaethical relativism have argued that deep moral disagreement and differences in moral outlook between cultures, or between, say, liberals and conservatives or consequentialists and deontologists, make it implausible that the truth‐ conditions of their moral judgments coincide (see RELATIVISM, MORAL; DISAGREEMENT, MORAL). If sentences involving “good” and “ought” generally express different claims depending on the requirements, ends, and ideals that speakers take to be relevant to their judgments, this would seem to provide a natural fit for the moral relativist, accommodating both individual and cultural moral variation. However, some who have stressed context dependence have simultaneously rejected metaethical relativism. Even if one thinks that the terms used to express moral judgments are context‐dependent, one might also think that, in moral contexts, our judgments relate to the same fundamental requirements (see Mackie 1977; Wedgwood 2007). In the following section, we consider some of the concerns that people have raised against metaethical contextualism. Categoricality, Motivation, and Disagreement One common objection to metaethical contextualism is that it fails to capture the practical relevance or normativity (see NORMATIVITY) of moral judgments. If moral “ought” judgments are always relative to some set of considerations, and if there are numerous such sets that favor different alternatives, morality seems neither to offer practical guidance nor demand any particular action (see, e.g., Montminy 2007). What the contextualist can say, however, is that when we are asking ourselves what we ought to do, then, insofar as our question has a determinate content, we are already relating that question to one particular set of considerations. Moreover, insofar as these are considerations that we ourselves endorse acting from, the judgment we arrive at will be practically relevant for us. 5 Even so, one might think that metaethical contextualism fails to capture the sense in which we intend our judgments to relate to the correct set of moral considerations, not just the set that we happen to endorse. But it is unclear how this would be a problem for contextualism. Insofar as one’s sense that considerations should be correct has reasonably determinate content, the contextualist can understand it as constituting an abstract or higher‐order way of specifying the considerations that one endorses. As long as standards of correctness vary across contexts, the upshot would still be a form of metaethical contextualism. Moreover, deep moral disagreement does seem to involve variation not only in what parties count as relevant moral considerations, but also in standards of correctness. A related but importantly different objection focuses on the connection between moral judgments and moral motivation. According to metaethical contextualism, our motivation to act on a particular “ought” judgment depends on something external to the judgment itself, in particular on whether we are motivated by the set of considerations invoked by the judgment. This might seem highly plausible for some explicitly relativized or conditional “ought” claims (such as the claim that, in order to look more casual, I ought to wear blue jeans). But in normal cases that lack explicit relativization, thinking that one ought to do something and being at least somewhat motivated to do it do not seem to be two separate states: the motivation seems internal to the judgment. Similarly, to think that an action is “good” often seems to already involve being in favor of its performance. Many have taken this to suggest that moral judgments are either beliefs with contents that are somehow intrinsically motivating or desires or preferences of some sort rather than beliefs (see INTERNALISM, MOTIVATIONAL; COGNITIVISM; NON-COGNITIVISM). Though contextualism might seem incompatible with an intimate connection between “ought” judgments and motivation, some have instead argued that contextualism is particularly well placed to account for this connection. While denying that the content and nature of “ought” and “good” judgments guarantee motivation, Dreier (1990) and (in more detail) Finlay (2014: Ch. 5) suggest that, in identifying the content of a given “ought” sentence, we normally assume that the set of considerations that it relates to is one that motivates the speaker. (Similarly for the requirements that a “good” sentence relates to.) The reason for this is that, in contexts where it is not otherwise clear to what considerations an occurrence of “ought” relates, the audience can reasonably assume that the considerations are ones that matter to the speaker: otherwise, why make the claim? (This assumption is especially reasonable in contexts where the speaker is engaged in practical deliberation.) Making an “ought” claim in such contexts thus creates a presupposition that the speaker cares about the relevant considerations, whatever they are. Since this presupposition precedes any assignment of a specific content to the “ought” claim, it will naturally seem that the motivation is inseparable from the judgment expressed. If this pragmatic account is correct, it seems to capture the practical relevance of moral judgments without postulating intrinsically motivating contents or taking on the CONTEXTUALISM IN ETHICS 6(8) difficulties of a non‐cognitivist analysis of moral judgments. Moreover, it leaves room for cases suggesting that motivation is external to “ought” judgments: cynics, amoralists, or apathetic people who realize that they ought not to do what they are doing, but who are unmoved by this thought (see AMORALIST). Contextualism allows that such characters make judgments in relation to the requirements and considerations that they were once moved by, or that other people in their community are moved by. Contextualist analyses of “ought” might also provide a natural fit for the idea that what an agent ought morally to do must be an action that the agent could be (rationally) motivated to do (see RATIONALISM IN ETHICS). In many contexts where we make judgments about what someone ought to do, our judgments seem to have a rational practical point, being intended to identify an action that the agent could (in principle) be rationally convinced to do, or has internal reasons to do (see REASONS, INTERNAL AND EXTERNAL; cf. Finlay 2006). If the truth‐conditions of “ought” judgments depend on our concerns in making them, as examples of the context dependence of “ought” suggest, the truth of judgments with this particular point could well depend on whether the action in question is favored by an ideal or end that the agent could be rationally compelled by. Furthermore, if agents’ varying subjective desires determine what ideals or ends agents can be rationally compelled by, they would determine what agents ought to do, in the sense expressed by these judgments. The upshot would be a version of normative relativism, though one restricted to practical “ought” judgments with a rational practical point. Perhaps the most common and serious objection to metaethical contextualism is that it misrepresents agreement and disagreement (see Lyons 1976 for a classical statement; cf. Dreier 2009). For example, it seems natural to say not only that those on the other side of seemingly intractable moral controversies about abortion, animal rights, ownership rights, etc. ultimately relate to different moral considerations, but also that we disagree with their judgments and, often, that they are wrong or mistaken. In saying this, we seem to do two things. First, we attribute moral judgments to those on the other side, understanding that these judgments are based on their fundamental moral outlook. Second, we assess the correctness of their judgments based on our moral outlook. This two‐part practice seems hard to reconcile with the contextualist assumption that the truth‐conditions of moral judgments depend on the considerations that moral judges relate to when making their judgments. Given this assumption, our assessments seem blatantly insensitive to the truth‐conditions of the judgments assessed. Also, this insensitivity would contrast starkly with how we assess judgments involving some other context‐dependent expressions. If Alexander says of 7‐year‐old Beth that “she is tall,” clearly comparing her with other children her age, we would not naturally say that Alexander is wrong merely because Beth struck us as short compared to adults. If he is wrong, it seems, it would be because he got his comparison wrong, not ours. Metaethical contextualists seem to have two broad kinds of options in trying to explain these phenomena. The first sort of explanation would be that, in attributing disagreement and making 7 insensitive assessments, we make a mistake. Perhaps we make a semantic mistake, failing to understand that the truth‐conditions of moral judgments depend on the considerations that are relevant in the context of judgment. Given that such semantic blindness seems constrained to some specific context‐dependent domains (including moral discourse, but excluding attributions of tallness), this explanation would seem plausible only if something about these domains would make this mistake particularly likely (cf. Francén Olinder 2013). Or perhaps, in spite of the deep differences between parties of intractable moral disagreements, we make a pragmatic mistake, failing to take into consideration the possibility that the parties ultimately relate to different considerations in making their judgments. The pragmatic suggestion gets some support from empirical data indicating that insensitive assessments of correctness or incorrectness become less common as the difference between moral outlooks becomes increasingly radical and thus harder to ignore (Sarkissian et al. 2011). If this is right, contextualists need not worry about seemingly insensitive assessments. But data also suggests that we continue to attribute disagreements between parties with radically different moral outlooks even while ceasing to insensitively assess correctness and incorrectness (Khoo and Knobe 2016). Apparently, then, contextualists would still need a nonstandard account of disagreement attributions. Another sort of explanation sees attributions of disagreement and insensitive assessments as perfectly adequate given the practical function of moral judgments. Moral questions are largely shaped by concerns about how to behave, what attitudes to have, and what behavior and attitudes to publicly support, and we normally expect people’s moral judgments to correspond to attitudes guiding their behavior and emotional reactions. Because of this, we have a practical interest in keeping track of when people make moral judgments with conflicting attitudinal or behavioral consequences, as such judgments constitute a kind of practical disagreement. We also have an interest in assessing the correctness of moral judgments of others relative to the considerations that we endorse as publicly upheld guides of conduct. Even given contextualism, this could explain not only attributions of disagreement and why we make insensitive assessments, but also why such assessments are limited when moral outlooks seem to rule out any common guides of conduct. Moreover, it would account for the difference between moral talk and thought and other context‐dependent domains – tallness judgments do not normally have the same intimate connection to attitudes and action. (For developments of this suggestion, see Finlay 2014 and Björnsson 2015.) Whether any of these explanations succeeds is currently an open question, as is the question of whether parties of moral disagreements do have fundamentally different concerns (see DISAGREEMENT, MORAL). See also: AMORALIST; COGNITIVISM; DISAGREEMENT, MORAL; MORAL; INTERNALISM, MOTIVATIONAL; NON‐COGNITIVISM; PARTICULARISM; RATIONALISM IN ETHICS; REASONS, INTERNAL RELATIVISM, MORAL EPISTEMOLOGY, NORMATIVITY; AND EXTERNAL; CONTEXTUALISM IN ETHICS 8(8) REFERENCES Björnsson, Gunnar 2015. “Disagreement, Correctness, and the Evidence for Metaethical Absolutism,” in Russ Shafer‐Landau (ed.), Oxford Studies in Metaethics, vol 10. Oxford: Oxford University Press, pp. 160–87. Dreier, James 1990. “Internalism and Speaker Relativism,” Ethics, vol. 101, pp. 6–26. Dreier, James 2009. “Relativism (and Expressivism) and the Problem of Disagreement,” Philosophical Perspectives, vol. 23, pp. 79–110. Finlay, Stephen 2006. “The Reasons That Matter,” Australasian Journal of Philosophy, vol. 84, pp. 1–20. Finlay, Stephen 2014. Confusion of Tongues: A Theory of Normative Language. New York: Oxford University Press. Francén Olinder, Ragnar 2013. “Moral Relativism, Error‐Theory, and Ascriptions of Mistakes,” Journal of Philosophy, vol. 110, pp. 564–80. Jackson, Frank 1985. “On the Semantics and Logic of Obligation,” Mind, vol. 94, pp. 177–95. Khoo, Justin, and Joshua Knobe 2016. “Moral Disagreement and Moral Semantics,” Noûs, vol. 52, pp. 109–43. Lyons, David 1976. “Ethical Relativism and the Problem of Incoherence,” Ethics, vol. 86, pp. 107–21. Mackie, John L. 1977. Ethics: Inventing Right and Wrong. London: Penguin. Montminy, Martin 2007. “Moral Contextualism and the Norms for Moral Conduct,” American Philosophical Quarterly, vol. 44, pp. 1–13. Pittard, John, and Alex Worsnip 2017. “Metanormative Contextualism and Normative Uncertainty,” Mind, vol. 126 pp. 155–93. Price, Anthony W. 2008. Contextuality in Practical Reason. Oxford: Oxford University Press. Sarkissian, Hagop, John Park, David Tien, Jennifer Wright, and Joshua Knobe 2011. “Folk Moral Relativism,” Mind and Language, vol. 26, pp. 482–505. Thomson, Judith Jarvis 2008. Normativity. Chicago: Open Court. Timmons, Mark 1999. Morality without Foundations: A Defense of Ethical Contextualism. New York: Oxford University Press. Wedgwood, Ralph 2007. The Nature of Normativity. Oxford: Oxford University Press. FURTHER READINGS Beebe, James R. 2010. “Moral Relativism in Context,” Noûs, vol. 44, pp. 691–724. Björnsson, Gunnar, and Stephen Finlay 2010. “Metaethical Contextualism Defended,” Ethics, vol. 121, pp. 7–36. Björnsson, Gunnar 2017. “The significance of ethical disagreement for theories of ethical thought and talk,” in Tristram McPherson and David Plunkett (eds.), Routledge Handbook of Metaethics, New York: Routledge, pp. 275–91. Brogaard, Berit 2008. “Moral Contextualism and Moral Relativism,” Philosophical Quarterly, vol. 58, pp. 385–409. Dowell, Janice L. 2013. “Flexible Contextualism about Deontic Modals: A Puzzle about Information‐ Sensitivity,” Inquiry, vol. 56, pp. 149–78. Enoch, David 2011. Taking Morality Seriously. Oxford: Oxford University Press. Kratzer, Angelika 1977. “What ‘Must’ and ‘Can’ Must and Can Mean,” Linguistics and Philosophy, vol. 1, pp. 337–55. Plunkett, David, and Timothy Sundell 2013. “Disagreement and the Semantics of Normative and Evaluative Terms,” Philosopher's Imprint, vol. 13, pp. 1–37. Silk, Alex 2016. Discourse Contextualism: A Framework for Contextualist Semantics and Pragmatics. Oxford: Oxford University Press.