24.00: Problems of Philosophy Prof. Sally Haslanger December 10, 2001 (Meta-)Ethical Subjectivism (or Non-cognitivism) For the past couple of weeks we have been focusing on the following questions: i) Which acts are right and which are wrong? Which acts ought we to perform? ii) What makes an action right or wrong? What about the action determines its moral status? Our third question has received less attention: iii) How do we know what is right and wrong? This last question is especially pressing for those who maintain that morality is an objective matter, that there are objective moral facts that serve as a basis for moral evaluation. Several questions immediately emerge for such a view: if there are objective moral facts, why is there so much moral disagreement? What sort of things are these moral facts supposed to be? Unlike ordinary physical facts, moral facts are not knowable by observation: we don't see or taste that lying is wrong, and such facts are not part of our physical theories. But then where are such facts supposed to reside and how do we gain access to them? The moral relativist might seem to have a strategy for responding to these concerns: moral facts are just facts about what societies approve/disapprove of. One version of cultural relativism is: An action is morally right iff it is permitted by the ultimate mores of the society in which it is performed. (608) Shafer-Landau considers an individualistic form of relativism he calls normative subjectivism: An action is morally right iff the person judging the action approves of it. (605) So, for example, attending lecture regularly is morally right (judged by me) iff I approve of attending lecture regularly. Correlatively, attending lecture regularly is not morally right (judged by you) iff you do not approve of attending lecture regularly. An apparent advantage of these relativistic approaches is that the "spookiness" of "moral facts" seems to dissipate. Moral facts are nothing more than sociological/anthropological or psychological facts about us as groups or individuals, so they are no less mysterious than these other facts; moreover, there is no special worry about how we could come to know them. We considered several objections to moral relativism previously, and Shafer-Landau's main arguments against normative subjectivism should be familiar from that discussion. A further question, however, is whether the relativist strategy of reducing moral discourse to talk simply about what we/I approve of does justice to the normativity of ethical life. Put simply, the concern is that a description of an individual's (or culture's) attitudes (or practices) does not capture the force that moral claims carry in specifying not just what is, but what ought to be. E.g., someone might affirm that: American culture permits the production and sale of hate propaganda and violent pornography. But also maintain that: The production and sale of hate propaganda and violent pornography is wrong and should not be permitted. In other words, given a factual description of some state of affairs (including descriptions of social and individual attitudes), the question still remains whether the state of affairs described is good or right or appropriate. It is helpful now to introduce a distinction between normative ethics and metaethics. Normative ethics is an inquiry into questions (i) and (ii): it attempts to tell us what is morally right or wrong and why. Metaethics, however, steps back from the various answers to these questions to ask about the nature of ethical claims and moral theories more generally. Metaethics asks, for example, whether moral claims can be true or false, and if so, on what grounds; it also asks on what grounds we could be justified in holding a particular normative moral theory as opposed to another. Let's call a moral theory cognitivist if it holds that moral claims are like ordinary assertions in being truth-apt, i.e., they are the sorts of claims that can be true or false. All the views we've considered so far, although they have disagreed on a number of key points, have been cognitivist. The relativist, egoist, utilitarian, Kantian, and particularist all agree that claims such as: It is always right to keep one's promises. is either true or false (though, for different reasons, most of them would claim it to be false). A non-cognitivist moral theory denies that moral claims are truth-apt, i.e., non-cognitivists claim that moral utterances are neither true nor false because they aren't ordinary assertions. (Note that Shafer-Landau uses the term "meta-ethical subjectivism" for the view I'm calling "non-cognitivism". To see their point, note that we do a lot of things with language besides assert truths (and falsehoods); we ask questions, make commands, express shock or pain, etc: Where is my glove? (question) Shut the door! (command/imperative) Ouch! (exclamation) Non-cognitivists understand moral utterances on the model of these other forms of speech. So, for example, an emotivist take a moral utterance to be nothing more than the expression of an emotional response towards the kind of act in question, e.g., Death penalty, BOO!, or Giving to the needy, YEAH! Another form of non-cognitivism, prescriptivism, takes moral utterances to express commands: Don't steal! Keep your promises! In each of these cases the utterance is not an assertion and so shouldn't be considered true or false. First it is worth noting that non-cognitivism addresses the worry that moral claims are not mere statements of fact; it attempts to capture the normativity of moral discourse by locating it in the attitudes we express through our utterances. E.g., non-cognitivisits interpret moral claims not as describing our attitudes but as expressing them: Don't permit the production and sale of hate propaganda and violent pornography! (command) Or: Permitting the production and sale of hate propaganda and violent pornography, BOO! (exclamation) So non-cognitivism avoids this (among other) objections to relativism. Moreover, non-cognitivism appears to have what many view as advantages over the other cognitivist (objectivist) accounts: i) Non-cognitivism explains why moral judgement is intrinsically motivating, i.e., why making a moral judgement motivates one to act in accordance with it. BUT: ? Is moral judgement intrinsically motivating? Isn't it possible to judge that keeping promises is right and still not be motivated to do it? ? In what sense do non-cognitivists make room for moral judgement? ii) Cognitivism postulates unnecessary "entities", viz., moral facts, beyond what we have any evidence for or need for. It is possible to explain everything that happens without reference to moral facts. Because non-cognitivism does not postulate moral facts, it is to be preferred. BUT: ? Can we explain everything without reference to moral facts? iii) Cognitivism is committed to "weird" or "odd" entities: what are moral facts supposed to be anyway? Non-cognitivism avoids this. BUT: ? Ethical naturalism provides a cognitivist account on which moral facts are kind of natural or physical fact, e.g., to say that x is right iff it maximizes pleasure plausibly treats the moral fact as a natural fact that can be verified. (But can ethical naturalism account for normativity?) ? Ethical non-naturalism denies that moral facts can be viewed as a kind of natural or physical fact--moral facts are irreducible--but can suggest different sources of knowledge than empirical observation or scientific inquiry. E.g., some ethical non-naturalists believe that we have a faculty of moral intuition that enables us to know what is right and wrong. iv) Non-cognitivism explains ethical disagreement in terms of emotional disagreement: two people can agree on all the facts but still disagree in their emotional response to the situation. This seems more plausible than other accounts of disagreement. BUT: ? Is this really the best explanation we can give of ethical disagreement? Remember, the cognitivist can say that ethical disagreement occurs either when one party gets the moral facts wrong, or both parties do. However, non-cognitivism faces some serious objections as well: i) Moral discourse has no deeper grounding than expressions of taste. There is no clear basis for saying that one who prefers compassion is better than one who prefers torture; they just have different preferences, different emotional responses. There is no moral truth. And moral knowledge is not possible, since knowledge presupposes truth. ii) Moral considerations are not ultimately subject to rational critique. A non-cognitivist can criticize another's emotional responses as based on misinformation, but the normative component of moral judgement is not subject to rational appraisal. (Though this may be questionable if one believes that emotions are subject to rational appraisal.) iii) Non-cognitivism poses problems for an account of moral reasoning. Consider: 1) Lying is wrong. 2) If lying is wrong, then Clinton shouldn't have lied about his affair with Monica. 3) Clinton shouldn't have lied about his affair. According to non-cognitivism, (1) does not express a proposition that is truth-apt. What about the same utterance in (2)? Can we plausibly interpret both occurrences of "lying is wrong" in the same way? How are we to understand the inference? (Note that the validity of the inference seems to depend on the truth of (1) and the conditional (2). But if (1) is neither true nor false, and the antecedent in (2) is neither true nor false, how are we to evaluate the inference?) iv) How should we interpret occurrences of 'good', 'right', 'wrong', etc. in various other contexts? If good functions equivalently to an exclamation like "hurrah", then should we accept the following interpretations: Do what is good. ? Do what is Hurrah! Hurrah for good people! ? Hurrah for Hurrah people! Questions: 1. What do you think is the point of morality? Why do societies have moral codes at all? Do any of the views we've considered do better in capturing/explaining the point or purpose of morality? 2. What are we looking for, exactly, in a moral theory? What is moral theory attempting to accomplish? Do any of the views we've considered do better than others in accomplishing this?