Metaethics Explored 9781847600493, 9781847600677

Citation preview

Running Head  

http//www.humanities-ebooks.co.uk

The World is all that is the case

Philosophy Insights General Editor: Mark Addis

Metaethics Explored Paul Davis

For advice on use of this ebook please scroll to page 2

Publication Data © Paul Davis, 2007 The Author has asserted his right to be identified as the author of this Work in accordance with the Copyright, Designs and Patents Act 1988. Published by Humanities-Ebooks.co.uk Tirril Hall, Tirril, Penrith CA10 2JE

Reading Options * To use the navigation tools, the search facility, and other features of the toolbar, this Ebook should be read in default view. * To navigate through the contents use the hyperlinked ‘Bookmarks’ at the left of the screen. * To search, expand the search column at the right of the screen or click on the binocular symbol in the toolbar. * For ease of reading, use to enlarge the page to full screen * Use to return to the full menu. * Hyperlinks appear in Blue Underlined Text. To return from an internal hyperlink use the ‘previous view’ button (more than once if need be). * For a computer generated reading use Read out loud>

Licence and permissions Purchasing this book licenses you to read this work on-screen and to print one copy for your own use. Copy and paste functions are disabled. No part of this publication may be otherwise reproduced or transmitted or distributed without the prior written permission of both the copyright owner and the publisher. Making or distributing copies of this book constitutes copyright infringement and would be liable to prosecution. Thank you for respecting the rights of the author.

ISBN 978-1-84760-049-3

Metaethics Explored Paul Davis

Philosophy Insights. Tirril: Humanities-Ebooks, 2007

Contents A Note on the Author Acknowledgements Chapter 1. Introducing the Issues 1.1 Some Everyday Moral Utterances 1.2 What are they about? 1.3 The Attractions of a Popular Picture 1.4 Rules: A Patch on Morality’s Surface? 1.5 From Angry Young Men and Women to …? 1.6 Metaethics and Normative Ethics Revisited 1.7 Summary and Reflection Chapter 2. Cognitivism and Noncognitivism 2.1 The Basic Distinction 2.2 Moral Realism and Moral Irrealism 2.3 Error and Success Theories 2.4 Moral Knowledge Again 2.5 The Truth Conditions of Moral Statements 2.6 Critical Reflections on the Proposed Truth Conditions of Moral Statements 2.7 Divine Command Theory of Morality 2.8 Kantianism Chapter 3. Objectivist Realism under Siege 3.1 A Philosophical Health Warning 3.2 David Hume 3.3 Hume’s Arguments Considered 3.4 Mackie’s Argument from Queerness 3.5 So Where are We? 3.6 Emotivism

Metaethics Explored  

Chapter 4. Out on the Street 4.1 The Death of God 4.2 The Death of Deference 4.3 Moral Disagreement 4.4 Multiculturalism 4.5 Rejection of Specific Moral Beliefs 4.6 Seedy Underbelly of Conventional Morality 4.7 Decline in Communal and Symbolic Life 4.7 Reluctance to be Judgemental 4.8 Postmodern Irony Chapter 5  A Naturalist Objectivist Realism? 5.1 The Basic Features 5.2 Moore’s Open Question Argument 5.3 Empirical Reasoning and Moral Facts 5.4 Morality and Objectivist Naturalism 5.5 Ethics and Natural Disciplines Bibliography

A Note on the Author Paul Davis is tutor in philosophy at the University of Edinburgh, and formerly lecturer in the philosophy of sport at the University of Wales Institute, Cardiff. His doctoral thesis is on free will, and he has published on consciousness, ethics, and ethical and aesthetic issues in sport.

Acknowledgements For help and encouragement with this text, I would like to thank Bobby Davis, Gillian Davis, Emily Brady, Pauline Phemister, Dory Scaltsas, David Cromwell, Joe Fodey, Lyndon Clarke, Rachael Atherton, Frank Rae, Rob Muir, John Divers, Lizzie Eldridge, and Richard Gravil. I would especially like to thank Robert and Kathleen Davis for all they have done for me.

Chapter 1. Introducing the Issues 1.1 Some Everyday Moral Utterances ‘He’s a good person.’ ‘That was a terrible thing to do.’ ‘She’s basically honest.’ ‘He can be mean.’ ‘Hitler was evil.’ ‘One should love one’s neighbour.’ ‘She brought a lot of good into the world.’ ‘Slavery is wrong.’ ‘That was a courageous thing to do.’ ‘It’s wrong not to keep appointments.’ ‘Suicide is never permissible.’ ‘Others must always be treated with respect.’ ‘One should never tell a lie.’ ‘Euthanasia can sometimes be justified.’ ‘You ought to apologise for that.’ ‘Eating animals is fine.’ ‘Her behaviour was a bit “off”.’ ‘Treating people like that is not acceptable.’ ‘You ought to return your library books on time.’ ‘I must try to be more sensitive to the feelings of others.’ ‘That would be a nice thing to do.’ ‘That wasn’t a very nice thing to do.’ ‘The poverty and starvation in the world is unacceptable.’ ‘That was callous.’ The above are examples of moral judgements. There seems nothing exotic about any of them. Some of them might be wrong, and disagreement seems possible in many or even all cases. But they seem like ordinary things to say.

Metaethics Explored 

1.2 What are they about? But what sort of judgements are moral judgements? How can we set about discovering whether they are true or false? Indeed, are they true or false? Are there moral facts? Is moral argument just a species of rational argument? If not, then what is going on in moral argument? In short, what is this thing, morality? This book will explain and discuss a range of answers to the above cluster of questions. Questions such as these are the territory of metaethics. The current metaethics scene is rich and diverse, and characterised by the same flat divergences of view that can be found in any area of philosophy. Michael Smith offers a brisk summary of six of these divergences. First, some hold that moral practice presupposes moral facts that don’t exist. Others deny that moral practice involves any such presupposition. Second, some hold that moral facts exist and that they are ordinary facts. Others hold that moral facts exist, but that they are facts of a unique kind. Third, some hold that moral facts exist and are part of the causal explanatory network. Others hold that there are no moral facts at all. Fourth, some hold that there is an internal or necessary connection between moral judgement and the will. Others hold that there is no such connection. Fifth, some hold that moral requirements are requirements of reason. Others hold that it is not necessarily irrational to act immorally. Sixth, some hold that morality is objective, that there is a single, ‘true’ morality. Others hold that there is not a single true morality. We will be discussing some of these divergences in the course of this book. Metaethics was once sharply distinguished in philosophy from normative ethics. The popularity of the distinction accompanied a broader feeling that philosophy might be nothing but the analysis of language. Metaethics was supposed to be concerned with the meaning and justification of moral judgements (e.g. are moral statements about facts?), whilst normative ethics was more concerned with substance, e.g. right and wrong, good and bad, ought and ought not. However, some subsequent philosophical arguments tended to undermine the distinction by suggesting that questions of substance cannot be separated from questions of meaning. The six points of divergence set out above should be sufficient to suggest that we cannot confidently do normative ethics without thinking about metaethics. And, indeed, much recent work in moral philosophy is impossible to classify as exclusively metaethical or normative. For instance, Mackie’s landmark book on ethics prescribes principles of conduct that Mackie explicitly regards as continuous with the metaethical treatment he offers   M. Smith, The Moral Problem (Blackwell, 1994), 3–4.

Metaethics Explored 10

first. Basic textbooks in moral philosophy, similarly, tend to contain reflections on both metaethics and normative ethics. This book, again, will have things to say on what would previously have been parcelled out to ‘normative ethics’. Metaethical questions might seem not to matter. We can all think of people that we regard as morally good and whom we think are liable to have thought fleetingly at most about any of them. Some philosophers have suggested too that very different answers on our cluster of metaethical questions can coexist with identical moral views and responses. However, Smith’s six points of divergence alone suggest that metaethical questions cannot be dismissed. Furthermore, it is worth pausing a little upon significant consequences of some of the questions we have already seen. Moral Facts, Moral Truth, and Moral Knowledge We have seen, for instance, that the status of ordinary moral utterances as judgements is controversial. Smith includes the question of whether moral facts exist among his six points of metaethical divergence. According to Emotivism, for instance, moral statements are, despite appearances, not genuine judgements (see 3.6). They do not comment on facts, but express emotions or seek to create an influence. Therefore, they are neither true nor false, not even approximately so. There are considerable consequences if an account such as this is correct. How are moral ‘views’ formed or altered? Not by looking at moral facts, for sure. What is moral disagreement about? The belief that we attempt to morally persuade others by citing moral features is mistaken. There are no specifically moral phenomena. And if moral statements cannot be true, there must be significant consequences, at the least, for moral knowledge. How, if at all, is moral knowledge possible? Some other accounts, however, assert the existence of moral facts. Furthermore, we can know these, and therefore have moral knowledge. There are different ideas about what moral facts consist in, as we will see especially in the next chapter. And these different answers only reinforce the significance of metaethics. For instance, different accounts of what moral facts consist in result in different accounts of what moral disagreement is ultimately about. Indeed, when the push of ethical disagreement comes to shove, we might regularly find that a difference of metaethical outlook is in play, casting suspicion on the suggestion that metaethics needn’t make a significant difference to moral views.

  J. L. Mackie, Ethics: Inventing Right and Wrong (Oxford University Press, 1977).

Metaethics Explored 11

Amoralism Another point of impact that metaethics has is the question of amoralism. What is an amoralist? An amoralist is someone who recognises moral features but is wholly unmoved by them. He might even be expert in knowing what is good, bad, right, wrong, commendable, deplorable, courageous, nice, callous, a bit ‘off’, wicked, etc. But he is indifferent to this knowledge. It does not affect his motivation a jot. On one theory, known as Internalism, this is not possible. (There are in fact subtly different versions of internalism. The version I am talking about is sometimes called ‘Motivational Internalism’. ‘Internalism’ is safe here.) We noted on page 9 that, for some, there is a necessary connection between morality and the will. This is the position of the internalist, for whom moral properties are intrinsically motivational. That is, one cannot recognise moral features and be indifferent to them. If one is genuinely indifferent, then one doesn’t genuinely know the moral features (though one might think one does). Externalism is the denial of internalism. Here, it is consistent to fully know moral features and be indifferent to them (in other words, amoralism is possible). In Plato’s dialogue, Republic, Glaucon and Adeimantus quiz Socrates on why one should behave morally. The internalism/externalism debate is a large one, and we will touch on it again in the course of the book. However, one theory already mentioned entails an answer. Can you work out which one it is and which answer it gives? It is emotivism, and the answer it entails is internalism. For the emotivist, moral utterances express emotion and/or try to create an influence. Motivation is therefore built into moral utterance. Therefore, internalism is true, and amoralism in turn impossible. Our Very Happiness? Finally, some philosophers have pointed out that metaethics might even make a difference to our happiness, since there is a world of difference between desiring things because of the value we take them to have and attributing value to things because we desire them. If we believe that the things we desire are intrinsically valuable (valuable whether anyone in fact desires them), then we are liable to believe that the result of satisfying these desires is a richer happiness than that on offer from things whose value depends on being desired. Anything, after all, could be desired. On a metaethical ticket such as emotivism, for instance, it is not clear how anything has value in itself, and therefore not clear that we can legitimately desire things because of their intrinsic value.   Plato, The Republic (J. M. Dent & Sons Ltd., 1976), 35–46.

Metaethics Explored 12

1.3 The Attractions of a Popular Picture What might people reply if asked what morality is? What would you say? One quite popular suggestion is that morality is ‘rules for how we should treat one another.’ If correct, then it seems that moral statements are true or false (they have ‘truth value’), and that their truth values depend upon how they cohere or fail to cohere with moral rules. (The statement might state a moral rule itself, or its contradiction. We will see, in 3.2, David Hume’s consideration of the suggestion that the morality of actions consists in their ‘relation to a prior rule of right’). In the meantime, it is helpful to engage in some reflection upon this suggestion. In philosophical explanation and evaluation, it is often helpful to begin by noting the apparent attractions of a position (‘sympathetic habitation’, as it is sometimes called). Why do people think what they do? What reasons do they give or have? Are there other reasons, not usually appreciated, for holding the position? Only after such sympathetic habitation is it possible to evaluate and, if necessary, propose alternatives. The position might be a high-powered philosophical theory or a popular notion like that just given. There is a family of apparent attractions in the idea that morality is a set of rules for how we should treat one another. Morality versus Inclination One of these is the hunch that the requirements of morality essentially run counter to our inclinations. Our earliest initiations into a moral universe feature the thwarting of our inclinations by authority figures, most obviously parents. This swiftly mutates into articulated generalities: one is not allowed to go here, do this or do that, and one must do this and do that. Rules seem to articulate prohibitions and compulsions, and many people seem to find it in turn natural to conceive morality as a set of prohibitions and compulsions. Morality seems conceptually connected to the thwarting of inclinations, in turn generalised in prohibitions and compulsions, which get formalised in rules. This picture seems to get reinforcement beyond childhood. Our inclinations seem to be a huge and simple part of who and what we are. They seem to be something that we straightforwardly discover or feel, and that motivate us in a similarly straightforward way. The alien externality of morality seems, in contrast, writ large. This simple picture might be incorrect, but it seems ‘from the inside’ a natural one for people in our moral culture to hold.

Metaethics Explored 13

Duty This externality seems emphatically confirmed in a notion that seems to figure centrally in morality: duty. Duty seems something very different from inclination. It seems to be something that imposes upon us from without. It is something that we have to learn, often with difficulty. It seems natural, again, to think of it as typically framed in rules. We must, for instance, tell the truth, keep our promises, and love our neighbour. Few people, if any, are constantly inclined to behave in these ways, and most of us are regularly inclined not to. Duty, and what it enjoins us to do, seems to reinforce key connections between morality, externality, and rules for the treatment of others. Immanuel Kant, as we will see in the next chapter (2.8), counterposes duty to inclination, and exults the former with extraordinary reverence. Law Heavy connotations of the law also figure in the modern picture of morality. In fact, one philosopher, Elizabeth Anscombe, famously complained that moral philosophy itself has assumed this flavour. She argued that notions such as duty and obligation are, in the absence of God, like laws without a lawgiver. The residual legalistic overtones add reinforcement to the notion of morality as rules from without that slap down our inclinations. (Anscombe recommended that moral philosophers abandon their whole obsession with moral rules.) Kant, again, speaks reverentially of the moral law. … And God Omission of God would be an elementary oversight. For some people, morality is ultimately sourced in the will of God. And this picture is typically accompanied by the belief that God’s will is manifested in the promulgation of rules, e.g. prohibitions on stealing, lying, adultery, and the worship of false gods. (Believers sometimes take exception to the emphasis upon God’s rules of prohibition. So we should acknowledge the sort of rules that give instructions to act, e.g. the instruction to honour thy father and mother.) Indeed, the cliché that we live in a secular society is undermined by the fact that clerics make up our official premiership of moral experts. Divine Command theory of morality is a subject for the next chapter. We will see in Chapter Four, too, that the Command picture of morality continues to exercise a tacit but sig  G. E. M. Anscombe, ‘Modern Moral Philosophy’, Philosophy, 33 (1958), 1–19.

Metaethics Explored 14

nificant influence over many people who claim to dismiss God. Just Look It Up Parental induction, duty, and the accompanying legalistic and religious overtones are a potent mix that helps sustain the notion of morality as a set of rules from without that tell us what we can do to (or with) one another, thwarting our inclinations in the process. And this mix is perhaps seasoned with something psychologically deeper. Many people might feel attracted to external rules of conduct because they would seem to do the work for them. That is, many people have an awareness that they can’t do just anything to anyone, and rules of conduct would seem to prevent the need for reflection, let alone dilemma or tragic conflict. We simply consult the rulebook. The appeal of this moral picture is connected to the belief that morality’s rules provide fixed and visible cut-off points that show us clearly what is off-limits. For instance, the rulebook tells us that we cannot lie, and we have in turn no trouble distinguishing lies from truth telling. So we know through and through how to behave. Jean-Paul Sartre pointed out that there is a perverse liberation in taking one’s self to have no choice in what one does. And the idea of morality as a set of rules of conduct might seem to provide such liberation. The binding and therefore liberating character of these rules is reinforced by the threat of punishment held out by parents, the law, and God. It is not difficult, then, to understand why the notion of morality as a set of rules for how we should treat one another tempts people when asked about morality in the abstract. The idea has clear attractions. Furthermore, it would be very rash to say that these attractions are entirely illusory. Morality seems to regularly require that our action is counter to our inclinations. And these requirements seem often to be articulated in rules. Rules do seem to be, at the least, a significant part of morality. You have already been introduced to the emotivist picture of morality (see pp. 10– 11). The picture of morality just sketched, with its heavy connotations of externality, would seem to rule it out. This picture seems to allow no moral purchase to emotion or anything like it. You might see that as something in its favour. Or you might see it as a reason to reject it. Or you might not be sure.

  Jean-Paul Sartre, Being and Nothingness (Methuen & Co. Ltd., 1958), 47–70.

Metaethics Explored 15

1.4 Rules: A Patch on Morality’s Surface? However, the idea that morality just is a set of rules regulating conduct has deep limitations. Suppose I ask you what morality is and you answer that it is rules for how we should treat one another. There are two questions I might ask. First, is moral action entirely a matter of behaving in accordance with rules? Second, where do the rules come from? Even if it is true that moral requirements are universally grounded in rules, it seems odd to think that the rules are not in turn grounded in something. Moral rules cannot be constructions out of nothing. Similarly, they cannot have any old content. Rules might well be on the surface of the moral soil. But if they are genuinely moral rules, they surely have to be the result of something more deeply in the soil. The remainder of this opening chapter will offer reflections that seem to strengthen the notion that rules are a patch on the surface of morality. The Major Normative Theories The major normative theories themselves illustrate the notion that moral rules are, at most, the result of something more basic to morality. For utilitarianism, that more basic feature is happiness. Actions are right or wrong insofar as they respectively promote or diminish happiness. (There are different versions of utilitarianism, but happiness is, in the theory’s original form, the fundamental element.) Utilitarianism need not involve rules at all, and in its simplest form, Act Utilitarianism, it does not. In the act utilitarian’s universe, one attempts to carry out happiness calculations as one goes along. The obvious impracticality of this generates Rule Utilitarianism. However, it should be obvious that moral rules emerge only because of the connection they are taken to have with the fundamental moral variable of happiness. That is why they exist as moral rules. Another of the major normative theories, Deontology, seems to involve rules in a much more direct way than rule utilitarianism. For the deontologist (or at least a common kind of deontologist) an action is right if and only if it is in accordance with a moral rule. There are, again, different versions of deontology, and each attempts to specify what makes a rule a moral one. One possible version of deontology is Divine Command Theory. Here, a moral rule would be a rule laid on us by God (see 2.7). More current versions have their roots in Kant, and cast moral rules as those that are required by rationality, or that would command universal rational acceptance (see 2.8). Again, the key feature shared by the different deontological approaches is that moral rules are underlain by something more basic, such as the will of God or

Metaethics Explored 16

reason. The third of the major normative theories, Virtue Ethics, has room for moral rules too. They are admittedly ‘thin’ in form, and involve instructions generated by the virtues, e.g. ‘Act justly’, ‘Act courageously’. However, the virtues are not themselves parachuted from outer space, but are underlain by a notion of human flourishing. A virtue is a character trait that a human being needs to flourish or live well. Therefore, ideas of what it is for humans to flourish determine what counts as a virtue and the content, in turn, of the virtue ethicist’s rules. Furthermore, proper virtue requires that one’s action is accompanied by the right feeling (usually pleasure) and is grounded in habit. It is probably universally accepted among moral philosophers that the morally basic concepts involved in the above sketches are less than transparent. What counts as genuine happiness, for instance, is contestable. There is ample scope for debate, similarly, about the content of God’s will, the nature (if any) of human flourishing, and the rules required by reason. (Kant’s attempts to show how specific moral rules flow from reason are widely thought uncompelling. See 2.8.) Utilitarians, deontologists, and virtue ethicists dispute among themselves no end, to be sure. But the difficulties here are not material to the point being made. This point is that the notions at issue figure in the respective theories as the basic source of moral rules. For all three theories, moral rules are derivative, i.e. they result from something more fundamental. (We will see in 3.2 that Hume similarly regards whatever moral rules there are as underlain by sentiment.) Disagreement about Rules Disagreement about moral rules seems commonplace. For instance, you might think that one should never tell a lie or that one should never commit suicide. I, on the other hand, might believe that one can tell a lie in such-and-such circumstances or that one can commit suicide in such-and-such circumstances. If morality just is rules of conduct, then there seems to be no scope for us to begin to discuss this disagreement, let alone settle it. If moral rules are not grounded in something more basic to morality, then what could such discussion be about? You would have your rule, I would have mine, and that would be the end of it. But we do discuss such disagreements, and argue for rules that we favour by appeal to those features that we think ground   Therefore, the virtue ethicist flatly rejects the preceding notion that morality essentially runs counter to our inclinations.

Metaethics Explored 17

them. Those features might include the will of God, reason, happiness, flourishing, or something else. But the reasons we give for our rules don’t take the form of yet more rules. And if we did give such a reason, could we then be asked to give a reason for that rule? Changes of Mind Similarly, it would seem very difficult to account for moral changes of mind if rules are the alpha and omega of morality. You believed before that suicide is never acceptable, but now you think that in a certain narrow range of circumstances it is acceptable. Your view has changed from ‘Suicide is never acceptable’ to ‘Suicide is acceptable only if ….’ But the change in your view is surely not an arbitrary or whimsical affair. The rule you favour has changed because you have thought further about the features that you regard as relevant to the rule, e.g. happiness or a worthwhile human life. There are reasons for your change of mind. And, again, these reasons don’t seem to involve just more rules. Conflicts between Rules Again, conflicts between what is prescribed by moral rules might be intractable if the rules were all there is to morality. Imagine that one has, for instance, a rule never to tell a lie and a rule never to knowingly hurt another’s feelings. The clash between motivations such as these is well known. Your partner makes a snap decision to get a new haircut and breezes in to ask what you think of it. You believe s/he will be hurt if you tell what you think is the truth. But you should never tell a lie. If rules were the end of the moral story, then you would be doomed to flap helplessly or toss a coin. But you are not so doomed. If you are a morally reflective person, you might have considered before now why we are generally guided by the sorts of rules that now seem to clash. What is it about us that inspires belief in the rules against lying and knowing hurting? That enquiry might have suggested that the relevant rules as you would usually articulate them are not fully adequate. Perhaps they need to be refined. (And it might be that no refinement you try seems to you adequate.) Perhaps you judge that you should tell the truth and therefore hurt feelings, so long as it is done sensitively and with regret (a virtue ethicist could well say this). Or perhaps you conclude that you should tell a lie because the hair will soon return to normal and your partner is then very likely to conclude that it is better that way (if not, then there is always Plan B). The fact that these responses seem possible does not mean that there is generally

Metaethics Explored 18

an easy way out of clashes that moral rules sometimes seem to generate. Nor does it mean that there are not horrible and even tragic moral dilemmas. It suggests only that there are moral resources that lie deeper than rules and that can be tapped when moral rules seem to clash. Moral competence arguably involves the capacity for quickfire moral perception in situations such as the above. Indeed, one’s application of what one takes to be an unconditional moral rule might involve this too. Consider what is sometimes involved in loving one’s neighbour or in merely respecting others. We should therefore heed Aristotle’s instruction not to expect moral greatness in the young or inexperienced. Whatever the Consequences? Some of our Everyday Moral Utterances seem in fact to state rules, e.g. ‘Suicide is never permissible’ and ‘One should never tell a lie’. However, this gives rise to another serious question: are there any unconditional moral rules? Kant argued that one should never tell a lie, even when one’s intention is to deceive someone who wants to commit a murder. He argued also that one should never commit suicide. However, it is less than certain, despite Kant, that such unconditional conclusions are sustainable. Take lying. Many reject unconditional prohibition because of consequences liable to follow from slavish adherence. (Oddly, Kant’s absolutist argument involves the premise that we cannot know for sure that telling the lie won’t result in the murder.) In a less dramatic spirit, can we say of the earlier example that it is never permissible to lie to one’s partner about the haircut? Again, those who think it sometimes permissible will point to consequences, e.g. the opportunity for one’s partner to see the light without the relative brutality of hearing the truth moments after exit from the hairdresser. Similarly, those who reject the unconditional prohibition of suicide are liable to point, as Hume did, to beneficial consequences that suicide can sometimes have. Generally, the consequences of our actions seem to often figure as something of moral significance, a figuring that can undermine what might otherwise be considered a moral rule. If this impression is correct, then further tricky questions arise. For   Aristotle, The Nichomachean Ethics (Oxford University Press, 1998), 3–4.   L. W. Beck (ed.), The Critique of Practical Reason and Other Writings in Moral Philosophy (University of Chicago Press, 1949), 346-50.   I. Kant, Groundwork of the Metaphysics of Morals (Cambridge University Press, 1977), 31–2.   S. Copley and A. Edgar (eds.), David Hume: Selected Essays (Oxford University Press, 1993), 322–24.

Metaethics Explored 19

instance, can we articulate what sorts of consequences morally matter, and therefore say when a lie or an act of suicide, for instance, is permissible or even required? Is it simply aggregate happiness, as the unreconstructed utilitarian would have it? If consequences do morally matter, then is the moral nature of an act wholly a function of its consequences, or are some types of act (say lying, suicide, or being late) intrinsically morally significant? Is it possible that what is done and its consequences both matter morally, and that this can sometimes give rise to moral conflict? Any viable approach to these questions would seem difficult to wholly detach from a notion of what is worthwhile, since notions of this seem to play a vital (though frequently inconspicuous) role in the engine-room of morality. It would seem strange for someone to hold a position on the morality of suicide, for instance, without having broader views about what is worthwhile. Is it, for instance, always worthwhile for a human to be merely alive? Or does the quality of a human life need to reach a base threshold before it is worthwhile? Is all life intrinsically worthwhile? Similarly, someone who holds that lying to prevent a murder is permissible or required is influenced by a belief that preservation of innocent human life (perhaps, again, meeting minimal conditions) is worthwhile. More generally, could we make sense of the moral relevance of acts and consequences in the absence of background conceptions of what is worthwhile? Again, would you go to someone who had no views on what is worthwhile, or thought them irrelevant, for advice on a tricky moral matter? Motive Motive seems to be another element that is of unavoidable moral significance. If we say that morality is just rules for the treatment of others, then asking whether someone’s behaviour obeyed the rules would seem to be all there is to moral evaluation. But this does not seem satisfactory. There seems to be a distinction of vital moral relevance between, say, someone refraining from stealing because s/he thinks it wrong and someone similarly refraining through fear of being caught or being punished by God. Each party obeys the rules against stealing and their behaviour is indistinguishable. But their motives are very different, and these different motives would seem to ground very different moral evaluations. Kant invites us to contrast the shopkeeper who is motivated by self-interest not to overcharge with his counterpart who refrains from overcharging out of respect for the moral law. Perhaps the truth value, if any, of some moral statements is the result of a complex mix of act, consequences, and motives.   I. Kant, Groundwork of the Metaphysics of Morals, 11.

Metaethics Explored  20

Our Everyday Moral Utterances Look again at the utterances with which we began. To what extent do we ascertain the truth of them by applying moral rules? This approach could well play a part in some cases, but it is likely that it will often play little or no part. Do we apply or even have rules to ascertain when, for instance, someone’s behaviour is a bit ‘off’, or to ascertain when we are callous, or to ascertain when something would be a nice or a not very nice thing to do? Go through the list for yourself and ask in each case what part, if any, moral rules play or could play in making the judgement true or false. Then try to be more specific about what sort of thing (if anything) does make the statement true or false. Good Persons Some of these utterances consist in the moral characterisation of a person, e.g. ‘He’s a good person.’ What makes someone, say, a good person? Is the good person so because s/he is unfailingly motivated by the right moral rules? Think of the good people you know. Can you really hope to explain their goodness in terms of the following of rules? Is it a matter of just trying hard enough? (And even if so, we could still ask the preceding question about the source of the rules.) I suspect you will find that, however hard you try, the task is monstrously difficult. For sure, unflinching adherence to laudable principle might well be a part of what makes some people good. It might often be what is emphasised when one is lauded as ‘a principled person’. But it is another thing to explain all good persons in terms of following the right moral rules. If you look across the good people you know, it is very likely that you will find, instead, a vast array of qualities contributing to the different goodnesses that you see. Moral rule following might seem to barely figure in some cases. You are liable to find much goodness in qualities of character that cannot be reduced in any useful sense to rule following. The virtues, as has already been noted, do generate rules (e.g. ‘Act justly’). However, this is consistent with a very limited part for rule following in the moral economies of some or even most good persons. First, the rules of virtue are, as was also noted, ‘thin’. They are realisable in a diversity of ways. One person’s just character might manifest itself most at the microlevel of workplace labour, whilst another just character might be most manifest in active commitment to a social or political cause. These differences in manifestation are not the result of rule application. And the precise manifestations of justice might be key to the moral personalities of each

Metaethics Explored  21

and to the ways in which others conceive the justice of each. Second, there are many ways, as indicated in the last paragraph, to have a good character. It is entirely possible that a particular virtue looms large in the make-up of one good person and is peripheral in that of another, equally good person. And, critically, the role that this virtue plays in their respective characters will not result from the application of moral rules by either. Finally, it seems futile to hope that one could uncover rules for attributing goodness to others. Do we work out how good someone really is by the application of rules of moral evaluation? Again, application of rules might conceivably figure in such a judgement, but it will typically play no part. Catchment Area? A question reinforcing the idea that rules are only a patch on morality’s surface is that of who or what is covered by morality’s rules. Who or what morally matters? It is not important that this question is answered now. The important points are that we take morality’s rules to cover our treatment of some classes and not others, and that we regularly disagree about what these classes are. And, again, we try to hammer out our differences by discussion and reasoning that cites what we take to be features of moral relevance. This confirms, again, that we take moral rules to be underpinned by basic sources of moral mattering. To illustrate, consider some candidate classes. It seems obvious that moral rules should cover our treatment of humans. However, what exactly is it about humans which makes them morally matter? And this question overlaps with more precise questions about the application of moral rules. Do moral rules apply to the treatment of all humans? Do they apply to treatment of the comatose, the senile, foetuses, future generations, or the dead? Is there a morally significant distinction between humans and persons? Also, do animals matter morally? If so, why? If they do, do they all matter to the same degree? Should our treatment of a fly be regulated by moral rules just the same as our treatment of a dog? What about the environment? Does it matter morally? Do moral rules extend to our treatment of it? The facts that these questions arise and that people opine about them suggests that we have some notion of qualities that make some things morally matter, and that in turn underlie whatever moral rules we endorse. Note also that the complexity of some of these questions reinforces the feeling that rules are unlikely to be able to guide us unfailingly in what to do. Not only need there be something underlying rules, but they are only a part of what’s on top.

Metaethics Explored  22

1.5 From Angry Young Men and Women to …? As social awareness develops, people tend to query where the requirements of morality come from. A typical and resentful conclusion is that these requirements are meaningless customs or the disguised instruments of self-serving hypocrisy, social and economic stratification, petty sentimentality, and spiteful repression. The ‘rules’ model of morality, with its connotations of prohibition, compulsion, punishment, law, and religiosity, is very congenial to this bitter conclusion. This popular model encourages many intelligent young people to conclude that morality is a rancid illusion. If they are correct, then there are no distinctively moral requirements on our behaviour. ‘Moral’ is a high-minded word for motivations—usually of others—which are at best earthy and at worst sordid. This is a scathing dismissal that explains the social pressures we baptise ‘moral’ as the disguised operations of the basest of appetites. There is probably some insight in this perspective, which will be revisited in Chapter Four. However, as adolescence moves into adulthood, most of us come through the other side of this moral scepticism and drop our surliness, rebelliousness, or irony. But we are liable to continue to struggle if asked to give an account of exactly what lies at the source of morality. This is not because angry young men and women turn into stupid adults. It is because the question is difficult. Philosophers since Plato have suggested and discussed a range of answers. Some of them have already been mentioned. They include God (Divine Command Theory), nature (Naturalism), reason (Kantianism), sentiment (Humeanism), human flourishing (Virtue Ethics), and happiness (Utilitarianism). Some of these will be revisited in this book. And some philosophers have suggested that morality has in fact no unique source of the kind sought by philosophers. 1.6 Metaethics and Normative Ethics Revisited Some further reflection on the comments in the last paragraph reinforces the difficulty of attempting to run a rigid metaethics/normative ethics distinction. Aristotle, for instance, held that virtues are natural qualities required for human beings to fulfil their function and live well. It has already been noted that Virtue Ethics is regarded as a normative ethical theory. Naturalism is the theory that moral facts are natural facts, and is regarded as a metaethical theory. However, if Aristotle is right that virtues are natural qualities, then virtue ethics seems to have the metaethical result that naturalism is true.

Metaethics Explored  23

Utilitarianism, similarly, is regarded as a normative ethical theory. However there would seem at first blush to be nothing more natural than our happiness. This, again, would have the result that if utilitarianism is correct, then naturalism is true. 1.7 Summary and Reflection We have come quite a long way from our Everyday Moral Utterances. We saw that their very status as judgements is a matter of controversy, and that one answer to this question (emotivism) rules out amoralism as well as affecting, perhaps, our very happiness. We noted too that those who think they are genuine judgements do not agree on what these judgements consist in, as we will see more clearly in the next chapter. Consideration of the popular notion tying morality and rules tended to suggest that morality is not wholly explicable (or even nearly so) in terms of moral rules and is liable to be as messy as any other facet of experience or domain of enquiry and discussion. This messiness will be reinforced in the final chapter, and is perhaps already suggested by the diversity of the everyday utterances at the start. But, again, does the institution we call morality involve a messy collection of facts that we can call correctly or incorrectly, or merely a messy collection of responses that are not checkable against a truth? If the latter, then why do moral utterances seem typically to take the form of assertions of fact? And if the former, then what sort of facts are moral facts? The next chapter will begin with some new terminology to help sharpen the treatment of these questions.

Chapter 2. Cognitivism and Noncognitivism 2.1 The Basic Distinction In philosophy, the same word is often used to name different theories. The theories are liable to sprout from the same soil, but that should not disguise the fact that we ought to be highly attentive to exactly what a philosopher intends when s/he talks about such-and-such theory (usually an ‘ism’). Bad philosophy often proceeds through insufficient attention of this kind. The area in which we are engaged is especially liable to this ambiguity and equivocation. Therefore, I will make every effort to spell out the vocabulary I am using. My own basic vocabulary follows Geoffrey Sayre-McCord. Moral Cognitivism is the view that moral statements, literally construed, are literally true or literally false (or approximately so). If moral cognitivism is correct, then moral statements are as straightforwardly true or false as statements such as ‘Today is Saturday’, ‘It is raining just now’, and ‘The Second World War ended in 1945.’ We needn’t get embroiled in the large philosophical question of truth to understand moral cognitivism. All we need recognise is that if moral cognitivism is true, then moral statements are true or false in as plain a sense as the preceding statements. That is the force of moral cognitivism’s insistence on the literal truth or falsity of moral statements. If moral statements are true or false only ‘in a sense’, then moral cognitivism is false. Plain truth or falsity, whatever it consists in, is what moral cognitivism claims for moral statements. (From now, I will refer to moral cognitivism as simply ‘cognitivism’. Also, I won’t repeat that the cognitivist understands the truth or falsity of moral statements to be sometimes approximate.) Moral Noncognitivism (‘noncognitivism’ from now) is the denial of cognitivism. For the noncognitivist, moral statements, literally construed, are neither literally true nor literally false. Moral statements, literally construed, are just not the kind of utterance to which truth value (the property of being true or false) attaches literally. We noted a noncognitivist approach in the first chapter, in the shape of emotivism (p. 10).   G. Sayre-McCord, ‘The Many Moral Realisms’, The Southern Journal of Philosophy, XXIV, Supplement (1986), 1–22.

Metaethics Explored  25

2.2 Moral Realism and Moral Irrealism Cognitivism does not entail a single moral truth. Why? Recall what cognitivism claims. It claims that moral statements, literally construed, are literally true or literally false. It is therefore consistent with the falsity of every possible moral statement. If every possible moral statement is false, then cognitivism is true. Enter Moral Realism (‘realism’ from now). Realism is cognitivism plus the claim that some moral statements, literally construed, are literally true. (Strictly, one true moral statement would suffice to make realism true.) The denial of realism is called ‘antirealism’ or ‘irrealism’. I will use the latter. An irrealist cognitivist believes that all moral statements are false. All noncognitivists are also irrealists, since they too deny that there are any true moral statements. Ask yourself, again, how you would explain the difference between the noncognitivist and the irrealist cognitivist. 2.3 Error and Success Theories It seems that most people, to put it no higher, speak much of the time as though making what they believe to be true moral statements. For instance, it would seem puzzling for someone to make one of our Everyday Moral Utterances yet deny that she is making a statement she thinks is true. If the irrealist cognitivist is correct, then it is an illusion to think that true moral statements are ever made. One never makes a true moral statement, since all moral statements are false. The irrealist cognitivist therefore holds what has been called an ‘error’ theory of moral statements. The (widespread) error is the belief that some of them are true. The core argument of the irrealist cognitivist (or error theorist) is that certain conditions must be satisfied in order that a moral statement is true, and these conditions never are satisfied. (Conditions of a statement’s being true are sometimes called ‘truth conditions’. I will use this term from now.) Therefore, all moral statements are false. The error theorist must tell us what these truth conditions are and why they are never satisfied. He is also under pressure to explain how people have gone wrong in believing that some moral statements are true, and this is something error theories typically attempt. Different error theorists needn’t agree on these questions. There might be different notions of the truth conditions of moral statements, why they are never satisfied, and how people succumb to realist illusion. The noncognitivist believes, too, that people err if they believe any moral state-

Metaethics Explored  26

ment is true. But the mistake is not connected here to truth conditions, since moral statements have, for the noncognitivist, no truth conditions to be satisfied or identified. For that reason, the noncognitivist is not considered an error theorist. Whilst the error theorist believes that all moral statements are false, the noncognitivist believes this is a mistake too. The realist counterpart of the error theorist is sometimes called a ‘success’ theorist. That is because he holds that the truth conditions built into moral statements are sometimes satisfied. Some moral statements are true. 2.4 Moral Knowledge Again There are different kinds of knowledge. Procedural knowledge is knowledge of how to do something, e.g. drive a car. Acquaintance knowledge is knowledge by acquaintance, e.g. knowledge of London. Propositional knowledge is knowledge that is expressible in sentences. The classic definition of propositional knowledge is true, justified belief. The sufficiency of this definition has in fact been challenged. However, no one has denied that true, justified belief is necessary for propositional knowledge (‘knowledge’ from now). This makes knowledge undetachable from truth. For instance, if I know that today is Saturday, it follows that ‘Today is Saturday’ is true. However, if irrealism is correct, then no moral statement is true. Statements such as ‘Others must always be treated with respect’ and ‘That was a terrible thing to do’ are either false or not the bearers of truth value. I cannot, for instance, know that others must always be treated with respect. I cannot have any of the knowledge claimed in our Everyday Moral Utterances, since none of the statements involved is true. Realism, on the other hand, asserts that some moral statements are true, and therefore upholds the possibility of moral knowledge. If it is possible that there is moral knowledge that is not propositional knowledge, then irrealism might yet be compatible with moral knowledge. This should not be ruled out, but is a question we will not examine here, since moral irrealism is a theory with huge consequences no matter. We tend to speak as though we have moral knowledge that is articulable in clear, simple statements such as those with which this book began. Moral irrealism makes that belief false. 2.5 The Truth Conditions of Moral Statements Let’s recap. Cognitivists hold that moral statements, literally construed, are either

Metaethics Explored  27

literally true or literally false. Noncognitivists hold that moral statements, literally construed, are neither literally true nor literally false. Realists hold that some moral statements, literally construed, are literally true. This allows for moral knowledge. Irrealists hold that no moral statements, literally construed, are literally true. This rules out moral knowledge (insofar as it is expressible in true statements). The realist is a success theorist. The irrealist cognitivist is an error theorist. The noncognitivist is an irrealist, since he holds that no moral statements, literally construed, are literally true. But he is not an error theorist, since he holds that no moral statements, literally construed, are literally false. A critical component of all of the positions above is the notion of moral statements literally construed. What one takes the literal construal of moral statements to consist in crucially affects one’s position. To know whether moral statements, literally construed, are true or false, or universally false, or neither true nor false, we need to know what moral statements consist in. Only then is it possible to know what conditions, if any, make moral statements true (their truth conditions). And only then is it possible to know if the relevant conditions are ever satisfied. We have already seen that truth conditions of moral statements just don’t enter the picture for the noncognitivist. Cognitivism, on the other hand, contains three broad accounts of the truth conditions of moral statements. Subjectivism The first cognitivist account of the truth conditions of moral statements is Subjectivism. Here, the truth of moral claims depends on the subjective states of individuals. Whether something is good/bad, right/wrong, etc. depends on someone’s attitude towards it. This approach is capable of surprising degrees of sophistication. On one version, judgements of value are relative to the desires, preferences, and goals of the judger. For instance, the claim ‘x is good’ is elliptical for ‘x is good-for-me’ or ‘I approve of x.’ The truth value of the first claim depends on the desires, preferences, and goals of the person doing the judging, and the truth value of the second claim depends on the approval or disapproval of the person doing the judging. If we apply this approach to our everyday moral utterances, we find that ‘He’s a good person’ means ‘I approve of him’ or ‘He is good for me.’ (This version is sometimes called Simple Subjectivism.) Another account takes a more discriminate approach. On the Ideal Observer theory, the moral character of something depends on the response of some observer

Metaethics Explored  28

ideally constituted and situated. The ideal observer is standardly defined as someone who knows and appreciates all the relevant facts, who has no mistaken beliefs about the situation, and who is disinterested and impartial. ‘Disinterested’ means that his own interests play no part in his judgement of the situation, and ‘impartial’ means that his judgement is not biased through a connection with anyone involved in it. On this account, ‘Slavery is wrong’ means that a disinterested, impartial observer who knows and appreciates all the relevant facts and holds no relevant, mistaken beliefs would disapprove of slavery. The wrongness of (say) slavery is its disposition to cause disapproval in an ideal observer. Intersubjectivism Intersubjectivism allows, like its subjectivist counterpart, that some moral statements are true. However, it differs in that the truth of moral claims does not depend on facts about particular individuals. It depends, instead, on the conventions, practices, and discourses of groups of people. There are, again, different versions. The simplest is Conventionalism, which treats moral statements as being about the practices and conventions in force in the relevant society. Benedict, for instance, has said that, “Mankind has always preferred to say ‘It is morally good’, rather than ‘It is habitual’ but historically the two phrases are synonymous”. On this account, ‘One should never tell a lie’ means ‘It is the convention of our group to never lie.’ Another version treats the truth of moral claims as a function of the hypothetical conventions or practices of hypothetical people. Some approaches hold (echoing subjectivism’s Ideal Observer theory) that the truth of moral statements depends on what appropriately idealised agents would agree to under certain specified conditions. There are differently nuanced notions of the idealised agent. However, the idealised agent of intersubjectivism mirrors her subjectivist counterpart insofar as her input is not influenced by personal interests and biases. Here, ‘One should never tell a lie’ means ‘Appropriately idealised agents negotiating under appropriate conditions would agree to never lie.’ A third and currently influential version treats morality as an indefinitely evolving form of life, consisting of habits, practices, institutions, and (critically) language. Language does not get to ‘how things are’ independent of language. Instead, language is a social device for expressing and working out our needs. It is by redescription that our moral horizons change. We can always reject the descriptions of others.   R. Benedict, ‘Anthropology and the Abnormal’, The Journal of General Psychology, 10 (1934), 73.

Metaethics Explored  29

The slave, for instance, can reject the descriptions given by his master and recast experience in his own terms. Such redescription, if sufficiently arresting, can make new moral statements true, so long as we recognise that to call a sentence ‘true’ is only to commend it very heartily and not to say that it ‘captures reality’ ahead of its rivals. Morality is therefore an ever-evolving and unpredictable conversation. A rough and ready literal literal construal of ‘Suicide is never permissible’ is ‘Current social descriptions of suicide don’t allow us to condone it.’ There are heavyweight exponents of this approach, the best known of whom is Richard Rorty. He has counselled that ‘we cease to think of morality as the voice of the divine part of ourselves and instead think of it as the voice of ourselves as members of a community, speakers of a common language.’ There is flexibility, according to context, in what the relevant community and language consists in. It might be very local or it could be global. A pithy summary of this intersubjective approach can be found in some words of Annette Baier: ‘There is no room for moral theory … which is not simply an account of our customs and styles of justification, criticism, protest, revolt, conversion, and resolution.’ Objectivism Objectivism is the view that the truth conditions of moral statements make no reference to anyone’s subjective states nor to the conventions, practices, or discourses of any group of people. In illustration, objectivist Ross asserts that ‘If I delight in contemplating a virtuous action, for instance, it is because I think the action to be good, and not vice versa.’ Therefore, ‘That is a good thing to do’ means ‘That action has the property of goodness, where goodness is a property that exists independently of anyone’s subjective states and of the conventions, practices, or discourses of any group.’ There are, again, different versions of objectivism. It is worth distinguishing immediately between a ‘hard’ and a ‘soft’ version. In the former, moral facts obtain independently of human cognitive capacities and conceptual schemes. In the latter, moral facts are grounded in qualities such as human nature, the point of morality, and circumstances. In each case, moral facts are independent of anyone’s subjective states and of intersubjective agreements (real or hypothetical) and practices (which includes conventions and discourses).   R. Rorty, Contingency, Irony and Solidarity (Cambridge University Press, 1989), 59.   A. Baier, Postures of the Mind: Essays on Mind and Morals (University of Minnesota Press, 1985), 232.   W. D. Ross, The Right and the Good (Oxford University Press, 1930), 31.

Metaethics Explored  30

The Major Normative Theories Revisited It should perhaps be noted too that two of the three major normative theories noted in the first chapter (pp. 15–16) seem, prima facie, to give rise to objectivist truth conditions of moral statements. Utilitarianism entails (in its simplest form) that moral truth is discoverable by finding out the happiness levels that would result from the options under consideration. Happiness levels do not depend upon the subjective states of (say) real or hypothetical judgers or the agreements or practices of real or hypothetical groups. The version of deontology that explains moral rules as requirements of rationality also looks robustly objectivist, since rationality seems to be independent of the subjective states of real or hypothetical judgers and the practices, etc. of real or hypothetical groups. We will see before the end of the chapter how this approach manifests itself in Kant’s moral theory. (Divine Command Theory might provide an intriguing exception from within deontology, as we will see in 2.7.) Virtue ethics is a less clear case, since virtuous behaviour might sometimes be so only because certain qualities of character are deemed virtuous by one’s moral culture (consider chastity, for instance). If this is the case, then virtue ethics might be at least partly a form of intersubjectivism. This is a standard worry about virtue ethics that we will consider no further. (Before going any further, try to think of examples of how different views about the truth conditions of moral statements could give rise to different views on moral topics, as suggested on p. 10.) 2.6 Critical Reflections on the Proposed Truth Conditions of Moral Statements Someone proposing truth conditions of moral statements has not demonstrated that moral realism is true, since she has not demonstrated that there are any true moral statements. What is needed to get from any proposal of truth conditions to the conclusion that there are true moral statements? Two conditions need to be satisfied. The first is that the truth conditions proposed must sometimes be satisfied. And the second is that the proposed truth conditions must give the correct account of the literal construal of moral statements. This section will focus briefly on the second of these conditions. Is there reason to believe that any of the preceding proposals gives the correct account of the truth conditions of moral statements?

Metaethics Explored  31

Subjectivism The subjectivist has her work cut out if she is to show that she has the correct account of moral statements. Consider the first version that we saw in 2.5 (simple subjectivism). First, if I attribute a moral character of any kind to anything, it is less than obvious that this character is necessarily tied up with my own desires, preferences, and goals. It seems, indeed, that two people can agree on the moral value of something and have quite different desires, preferences, and goals (I will refer to these from now as inclinations). According to a key strand of ordinary moral awareness (heavily sponsored in Kant), one feature that makes a judgement a moral one is precisely its independence from the inclinations of the judger. Simple subjectivism seems to put the cart before the horse. That is, my inclination towards an action is not what makes it good or right. It seems instead (echoing Ross) that I prefer it because I judge it to have the quality of rightness or goodness. The parallel observation applies to the construal of ‘x is good’ as ‘I approve of x.’ It does not seem to be my approval of x that makes x good. Instead, I approve of x because I take it to be good. I feel that I ought to approve of it, and indeed that everyone ought to approve of it. What happens if we consider not only good/bad and right/wrong, but also the richer range of moral properties that finds some expression in our Everyday Moral Utterances? For instance, if I judge so-and-so’s action to be a courageous one, it is difficult to see how my inclinations or approvings are constitutive of that judgement. Again, I might prefer or approve of an action because I take it to be courageous. But it is impossible to see how the courage of an action arises from my preferences or approvings. That is not to deny that dispute about whether an action is courageous (or cowardly) is possible, nor that judgements in this area might be inscribed with elements of one’s broader outlook (for instance, what I see as courageous you see as rash). Prior to the recent revival of virtue ethics, a quality such as courage might not have figured in the literature as a moral quality. But inclusion of the virtues seems to exacerbate the difficulties for simple subjectivism. Again, recognition of the range of qualities that figure in our moral lives throws up another problem for simple subjectivism. If all positive moral properties arise from my approvings, say, and all negative moral properties from my disapprovings, then how can I distinguish between the different properties on each side of the fence? This is an example of a broader feature of emotion. It is a common mistake that emotions consist exclusively of feelings, and are therefore classified by examining these feelings. Instead, there is a critical ‘externality’ to emotion. How do we distinguish

Metaethics Explored  32

jealousy, anger, and indignation? Looking inward to the subjective excitations would make that task impossible. Instead, we identify these emotions through the sorts of events in the world with which they are bound up. For instance, we know that it is jealousy we feel because the excitation is a response to the fact (or impression) that someone else has something that we badly want. Similarly, we know that it is indignation, and not mere anger, because we feel we have been treated in a way that shows insufficient recognition of our merit or status. Consider the following moral descriptors: shady, sneaky, a bit ‘off’, not very nice, dishonest, and wrong. On the simple subjectivist ticket, how can we distinguish what they pick out? We disapprove of each, presumably. Is it a matter of looking inward to find out if we feel a ‘shadiness’ disapproval, or a ‘sneakiness’ disapproval, or a ‘bit-offish’ disapproval, etc.? No. To ascertain that it is the precise quality of shadiness of which we disapprove, we look outward from our excitation, and find smoke and mirrors, obfuscation, tactical silence, etc., in the behaviour of which we disapprove. And again, it is because of these (shady) qualities that we disapprove. The parallel story could be told about how we distinguish, say, honesty, generosity, niceness, and rightness. There are, finally, two overlapping objections to the first version of subjectivism. Moral disagreement does not seem to be possible on this account. You and I might seem to disagree on a moral topic. But I am only reporting my inclinations or approvings and you are only reporting yours. You can agree what mine are and I can agree what yours are. And subjectivism seems to make us morally infallible, on the premise that we cannot be wrong about our approvings or inclinations. So how can we have any genuinely moral dialogue? The Ideal Observer version of subjectivism seems an immediate advance on simple subjectivism, for several reasons. First, its invocation of a disinterested and impartial judger banishes the influence upon judgement of the judger’s inclinations. Second, it seems to go a long way to dealing with the problems of moral disagreement and moral infallibility that beset simple subjectivism. If I approve of something and you disapprove, the explanation might well be that one of us fails to know and appreciate the relevant facts, or fails to be disinterested and impartial. When we discover the response of someone who does fit the bill, then we know which of us is correct. However, this approach also faces difficulties, some of which have parallels in simple subjectivism. There are reasons why characterisation of the ideal observer might prove tricky. One is that this seems very hard to do this without invoking the very moral notions that the ideal observer is intended to explain. Recall that the ideal observer is someone who knows and appreciates all the relevant facts. It is this

Metaethics Explored  33

feature of the ideal observer that seems impossible to detach from moral notions. Imagine that I know that you tortured a cat without benefit to anyone, and I disapprove. Imagine that I express my disapproval to a neighbour of normal moral awareness, who suggests to the culprit that, since I am someone who knows and appreciates the relevant facts (and is impartial and disinterested), your behaviour is confirmed to be wrong. As someone of normal moral awareness, the neighbour has a decent notion of how to distinguish what is relevant from what is irrelevant. For instance, if I told him I disapprove of your behaviour because it was Tuesday or because you wore a green sweater, he would reject my considerations as irrelevant and conclude that my disapproval is morally empty. But I have instead told him I disapprove because you caused suffering to an innocent creature for no good reason. He concludes that I am someone who knows and appreciates the relevant facts, and that my disapproval is therefore morally significant. But how do we distinguish the relevant facts from the irrelevant? Why is cat suffering and innocence relevant, whilst the day and the clothing of the culprit are irrelevant? It seems impossible to answer other than by saying that the former facts are morally significant and the latter are not. Circularity therefore seems to threaten. The theory tries to account for moral properties in terms of the responses of an ideal observer, but the ideal observer is (partly) constituted in terms of the ability to distinguish properties of moral relevance. Useful reinforcement of the last point is provided by the fact that moral disapproval is not the only kind of disapproval there is. How do I know whether my disapproval is moral or, say, aesthetic? The most plausible answer is one that invokes what I take to be specifically moral qualities in the object of my disapproval. Again, if the responses of the Ideal Observer are confined to simple approval and disapproval, then the theory posits a very thin moral agent. If, on the other hand, this observer might feel more differentiated moral responses, then the preceding point about the ‘externality’ of emotion is reinforced, and with it the problem of accounting for relevant facts without slipping in moral notions. For instance, how does the ideal observer distinguish his moral outrage from mere moral anger, or his moral admiration from mere approval? Isn’t it by looking outwards to the morally relevant causes of his excitation? Finally, can we be confident that ideal observers will converge in their responses? If they don’t, should we conclude that there is, in the given case, no moral truth? In conclusion, it is doubtful that subjectivism can give a convincing account of the literal construal of moral statements.

Metaethics Explored  34

Intersubjectivism Does Intersubjectivism fare any better? The simplest version, Conventionalism, is a non-starter. Moral statements surely cannot amount to statements about social habit. This would make it impossible to intelligibly raise moral objections to these habits. The second version noted faces clear difficulties too. The most obvious parallels a difficulty for Ideal Observer subjectivism. How do we settle upon the features of appropriately idealised agents without begging the question? That is, could we get near this without invoking moral notions (the idealised agent knows and appreciates the relevant facts …)? And how can we know that idealised agents would agree? Could moral truth really be built on such fragile foundations? The third version is probably the most compelling intersubjectivist approach. Since it rejects a final vocabulary and, instead, sees morality as organic, it has scant room for the fragile concept of the ideal observer. Similarly, it differs from conventionalism in that it has conceptual room for immoral conventions. A convention might be immoral, not because it fails to measure up to moral requirements finally independent of us, but because it fails to measure up to our indefinitely evolving procedures of justification. But there are difficulties too. It seems that any viable moral theory must allow that we could all get something morally wrong at some point, and realise only later, or indeed not realise at all. It is not clear how this particular intersubjectivist approach can satisfactorily account for this. It does not seem to be a matter of mere redescription that we now (generally) morally reject slavery, apartheid, child labour, sexism, racism, and indifference to animals, for instance. Social redescription (arresting metaphor, etc.) might well be involved in generating changes in moral outlook. But these redescriptions are believed by their proponents to connect with truths that had been previously obscured or distorted. Among these putative truths are the arbitrariness and plain error involved in previous immoral practice. Sexism and racism, for instance, often involve discrimination on irrelevant grounds (arbitrariness). Child labour and animal abuse involves the belief that the welfare and suffering of these groups matters little (arbitrariness). Slavery, and various manifestations of racism and sexism, involves plainly false beliefs about the constitution and capabilities of certain groups (error). The changes in these areas are not a mere matter of the social conversation taking a different turn by dint of tactical redescription. Instead, the redescriptions   For instance, it has been believed—perhaps universally at times— that the reproductive systems of women would be damaged by vigorous physical activity, and that their nurturing capabilities would be undermined by involvement in the public sphere.

Metaethics Explored  35

are successful because it is believed that they expose the arbitrariness and error (for instance) of our previous, legitimating descriptions. (The arbitrariness here is itself a species of error.) In short, it is believed that the redescriptions express truths that the previously dominant ones did not. And we commend the redescriptions—and resultant changes—because of this belief (we don’t baptise redescriptions ‘true’ because we commend them). We must therefore ask what burden is shouldered by ‘simply’ in Baier’s claim (p. 29) of what proper moral theory is. (Beware that much dubious philosophical work is done by words like ‘simply’, ‘merely’, ‘only’, and ‘just’.) Does this word beg the question of what is involved in styles of justification, criticism, conversion, resolution, etc.? This is reinforced by the fact that there are plenty of people who believe that recent moral change involves decline. This belief might be wrong, but it is surely not incoherent. Proponents of this belief reject more recent social descriptions because they believe that they take us away from the truths that were expressed in the old descriptions. In conclusion, it is not clear that intersubjectivism can provide the correct account of the truth conditions of moral statements. Objectivism That leaves objectivism. The discussion to this point suggests that objectivism is the candidate most likely to give the correct account of the truth conditions of moral statements. The versions of subjectivism and intersubjectivism so far met don’t seem to contain a convincing account. However, it is premature to conclude that objectivism is correct. Many philosophers find objectivist accounts unpalatable. It is worth recognising that if one believes that objectivist truth conditions can never be satisfied, and that any viable moral theory must allow true moral statements, then one is bound to reject objectivist accounts of the truth conditions of moral statements. This summarises a quite common position. That is: (i) true moral statements must be possible; (ii) objectivist truth conditions can never yield true moral statements; (iii) objectivist truth conditions are therefore incorrect. The error theorist might accept (ii) but reject (iii), since she does not see that true moral statements need to be possible. For her, objectivism might provide the correct truth conditions of moral statements. However, she will, qua error theorist, believe too that these conditions are never satisfied. If, however, one insists on (i), then acceptance of (ii) entails acceptance of (iii). However, what reason might there

Metaethics Explored  36

be to believe (ii)? Why might it be the case that objectivist truth conditions can never be satisfied? This question will figure in the remaining three chapters, alongside the question of whether objectivist truth conditions are correct. This chapter will finish by looking at two cognitivist theories of morality, each of which is, for different reasons, momentously important. One is robustly objectivist, and the other is usually considered to be so. The latter is the Divine Command theory of morality, and will be considered first. There are three reasons why it is important. First, there is a large number of people who think there is a vital, internal connection between morality and God. Second, there is also, as already noted (p. 13–14), a large number who say sincerely that they reject God but who remain in hoc to the form of the Divine Command model of morality. Third, the first and second reasons give us a powerful reason, in turn, for appreciating the theory’s fundamental deficiencies. The other theory is Kantianism. Kant makes a landmark contribution to moral theory, articulates key elements of ordinary moral awareness, and is a target for opponents of objectivism. 2.7 Divine Command Theory of Morality Divine Command Theory (‘DCT’ from now) is the theory that ‘x is right’ means ‘x is commanded by God’ and ‘x is wrong’ means ‘x is forbidden by God’. (You may find articulations of the theory that run with ‘good’ and ‘bad’ instead of ‘right’ and ‘wrong’. Whilst the good is a significantly different concept from the right, we can sidestep this discussion here. I will talk freely of divine goodness.) The truth conditions of moral statements therefore involve only one thing: the will of God. When we say that something—such as telling the truth—is right, we mean just that God commands it, and when we say that something—say lying—is wrong, we mean just that God forbids it. If God does not command telling the truth, our moral statement is therefore false. If He does command telling the truth, our statement is therefore true. It is important to realise that God is not, on this ticket, a being who is particularly skilled at discovering what is right. He is not the moral equivalent of, say, an expert physicist in whose pronouncements we invest authority. This is because DCT has it that the notion of moral qualities independent of God’s will is empty. For DCT, it is muddled to say that God commands what is right because it is right. Instead, what is right is right because God commands it. God’s commanding something is what makes it right. Take away God and you literally take away morality.

Metaethics Explored  37

Cognitivism DCT is clearly a cognitivist moral theory. (Why, exactly? Try to answer before proceeding.) For its adherents, moral statements are either true or false. Their truth value is determined by how they square with the will of God. It is worth noting at this point that there is nothing intrinsic to DCT requiring that God’s commands take the form of simple rules such as the above example of truth-telling. Sure, DCT tends to be most popularly thought of in this way (the Ten Commandments is probably the most popular exemplar in the West). However, there is no reason by the lights of DCT why God’s commands can’t include, say, imperatives of virtue (e.g. ‘Act courageously’) (see p. 16), or relatively fine-grained commands, e.g. ‘One must never commit suicide unless one has no family left and ….’ Realism/Irrealism and Error/Success Consider now how DCT impinges upon the realism/irrealism question. Think about it before reading on. Remember that realism is the theory that some moral statements are true. For there to be true moral statements on the Divine Command ticket, there has to be a God who issues commands. If there is no command-issuing God and DCT is true, then all moral statements are false. If this thought seems difficult to grasp, it might seem less so if you consider that it reflects what quite a lot of people nowadays really think. That is, a lot of people think that God is the only possible source of moral truth and that God does not exist, therefore all moral statements are false. (This is one theme within existentialism.) Many people are therefore attached to an error theory of morality. Someone who thinks DCT gives the correct literal construal of moral statements and believes also that there is a command-issuing God is a success theorist. For her, some moral statements are true. Whether DCT does in fact give the correct literal construal of moral statements is a question we will consider in a moment. Is DCT Objectivist? Let us pause briefly to ask whether DCT gives a subjectivist, intersubjectivist, or objectivist account of the truth conditions of moral statements. Again, think about it before reading on. The important thing is to know exactly what question is being asked. The answer might be more ambiguous than it seems. It would seem that we can immediately rule out an intersubjectivist characterisation, since DCT entails that moral truth is not a function of any sort of collective. (What of a polytheistic vari-

Metaethics Explored  38

ation, however?) It is tempting to conclude briskly that it is an objectivist account, since it entails that moral claims make no reference to anyone’s subjective states. But is that so obvious? They would clearly make no reference to the subjective states of any human, real or hypothetical. But are God’s commands subjective states? It does not seem outrageous to suggest that, on a Divine Command ticket, moral claims are claims about the subjective states of God, about His desires, preferences, and goals. So it might well be fair to consider DCT a version of subjectivism. Metaethics and Normative Ethics Again It is worth pausing to note that, again, differences of ethical view might result from different evaluations of DCT. For instance, a proponent of DCT who believes that God forbids homosexuality believes that homosexuality is therefore wrong. However, someone who rejects DCT does not believe that the moral character of homosexuality (or anything else) consists in its compliance or otherwise with God’s will. She might indeed believe too that homosexuality is wrong, but she does not believe the wrongness to consist in divergence from God’s will. (She might believe it would be wrong even if God commanded it.) Or she might believe that homosexuality is fine, regardless of what God commands. It is possible, indeed, for someone who rejects DCT to accept that a command-issuing God exists and to regard some of God’s commands as immoral. (Compare: one might not believe that morality, or any part of it, is sourced in the commands of a monarch, and be a subject in a monarchy in which the monarch issues commands.) This point will be sharpened before the end of the section. In the meantime, it illustrates, again, the significance of metaethics for normative ethics. A proponent of DCT ‘reads off’ moral obligations from (supposed) knowledge of God’s will, whilst someone who rejects DCT might reach the same conclusion about what God wills on a certain topic yet hold a different moral opinion on it. Objections to Divine Command Theory There are a number of difficulties with Divine Command theory of morality. Most obviously, it requires God’s existence in order that morality is viable. There are those on either side who are happy with this. That is, there are those who believe that morality is viable because God does exist, and those who (as noted above) believe that morality is not viable because God does not exist. However, there will also be many who are unhappy with this prospectus, e.g. atheists who think morality is viable, and believers or ‘don’t knows’ who think morality viable but not dependent on God. In

Metaethics Explored  39

order for DCT to be correct, then those who are unhappy with its prospectus would have to be mistaken. Another worry is that we need to know the content of God’s will in order to know how to act properly. But how do we know what God commands? Rival moral camps are liable to each invoke the will of God in legitimation of their stances (consider homosexuality and abortion, for instance). This problem is not about the existence of morality, but about knowing what it requires of us. The key philosophical objections to DCT go deeper, however. If DCT is true, then the content of morality is arbitrary. God could command anything at all, and it would therefore be right. If God commanded us tomorrow to murder our neighbours for fun, then that would be the right thing to do. Morality would become radically discontinuous with what we usually take it to be. Concepts such as the right and the good would not be doing their usual work, and there would seem every reason to avoid them. If God’s ‘goodness’ does not connect at all with our normal conceptions of goodness, it is expunged of its character as goodness, and we may as well call it God’s ‘splidgeness’. Similarly, if DCT is correct, then we give no information in saying that God commands what is right. ‘Right’ means ‘commanded by God’. Therefore, ‘God commands what is right’ means ‘God commands what God commands.’ Again, whilst believers tend to praise God for his moral character, it makes no sense to do so if DCT is correct. DCT leaves no independent moral standard against which God can be assessed. How can the attribution of infinite goodness be the attribution of anything if there is no goodness independent of God’s will? On DCT, God is stipulated to be good. The key terms have been defined so that God cannot fail to be good (or right). Conceptual sleight of hand guarantees God’s moral purity. But the cost of this move is that divine purity is a non-quality that it is therefore empty to praise. (Conceptual sleight of hand usually does buy empty triumph.) Finally, whilst there seems ample reason to reject DCT should we confine our moral talk to the right and the good, there seems even stronger reason to do so if we widen the net to include, again, the sorts of moral properties invoked in our Everyday Moral Utterances. How do we account, on Divine Command theory of morality, for moral notions such as the terrible, honesty, meanness, evil, respect, fineness, niceness, and callousness? If these notions have genuinely moral purchase, then they would have to eventually connect with the will of God. Such an ambition is not obviously impossible. But it is stretching. Many virtue ethicists, for instance, would be aghast at the suggestion that moral virtue turns finally on compliance with the will of God.

Metaethics Explored  40

But Might God Morally Help Us? Rejection of DCT is compatible with the aforementioned notion of God as a moral expert. It might be thought that morality is (like high-level physics) very difficult, and that limited creatures like ourselves need the guidance of commands from someone who can understand it better. That is quite probably the position of many believers. But it is not, for reasons that I hope are clear, DCT. And it leads us back to square one: what sort of features is it that a moral expert understands better than the rest of us? If it is not the commands of a deity, then is it, say, convention, factional power, human sentiment, the requirements of practical reason, our evolving language and procedures, natural facts, some combination of these, or something else again? 2.8 Kantianism Though Kant affirmed, by circuitous means, the existence of God, he explicitly rejects DCT. He echoes DCT, however, in his belief that morality does not come down to human inclinations, approvings, agreements, or practices. Kant’s moral theory is rich, dense, and—for many—difficult. We are concerned here only with its cognitivist and objectivist character. For Kant, moral statements are true or false (cognitivism), and this truth value is independent of personal preferences and intersubjective features (objectivism). We noted in the first chapter that, for some, moral requirements are requirements of reason. Kant is the most famous proponent of this view. The truths of mathematics are not a matter of personal preference or any intersubjective feature, and morality is, for Kant, just the same. The former belongs to the domain of theoretical reason, and the latter belongs to practical reason. Kant does not hold the implausible belief that personal preferences do not sometimes skew our awareness of moral requirements. He is all too well aware that our inclinations regularly help us to reach deluded moral conclusions. But the key point is that, for Kant, the inclinations have no proper role in moral deliberation. Proper moral deliberation is a matter of the exercise of unadulterated reason. How does this work? Kant makes an important distinction between a hypothetical imperative and a categorical imperative. The former commands an action only as a means of achieving something else that one wills. An example (not Kant’s) would be ‘If you want to be at the front of the queue, take the early bus.’ Here, the command to   Immanuel Kant, Critique of Practical Reason (Bobbs-Merrill Company, 1956), 128–36.   See p. 9.   Kant, Groundwork of the Metaphysics of Morals, 25.

Metaethics Explored  41

take the early bus depends upon an inclination to be at the front of the queue. If one does not have that inclination (or, strictly, a significantly similar inclination), then there is no practical necessity to take the early bus. One could rationally take a later bus. A categorical imperative, on the other hand, does not depend upon an inclination to achieve something else. It represents an action as in itself good, hence as ‘necessary in a will in itself conforming to reason’. An example of Kant’s is the prohibition of insincere promises, to which we will return in a moment. ‘Make only genuine promises’ is, for Kant, a categorical imperative. This imperative is independent of all ends that agents might have. It does not command that one make only genuine promises except when it suits one to make an insincere promise. The imperative commands absolutely. It is binding on all rational beings, regardless of inclination. And this is how Kant sees morality. The moral law—of which he lovingly speaks—is categorical. It applies universally to rational beings, whatever inclinations individual rational beings have. But how does Kant think that we give content to this categorical affair we know as morality? How do we distinguish imperatives that command absolutely from those that don’t? Kant proposes a condition that he thinks follows from the universality of morality. The condition is known as the Categorical Imperative (or, sometimes, Formula of Universal Law) and is: Act only in accordance with that maxim through which you can at the same time will that it become a universal law. What is a maxim? A maxim is the subjective principle of acting, i.e. the principle according to which the subject acts. The morality of one’s action therefore depends upon whether one could will that one’s principle of action become a universal law. If one could so will, then one’s action is morally sound. If one could not so will, then one’s action is morally unsound. There has been much discussion about what the test of the Categorical Imperative amounts to. What does Kant mean by the idea of being able to will the maxim of one’s action as a universal law? We are not equipped to approach this task without looking at Kant’s illustrations. Kant offers four illustrations, all of which involve maxims that Kant believes fail the test of the Categorical Imperative (and would therefore result in immoral actions). First, he imagines someone who is sick of life and contemplating suicide. His maxim is: from self-love I make it my principle to shorten my life when its longer duration   Ibid., 25.   Ibid., 32.   Ibid., 31–3.

Metaethics Explored  42

threatens more troubles than it promises agreeableness. Kant claims that this maxim contradicts itself because self-love impels ‘toward the furtherance of life’. Therefore, it could not be willed as universal law. Second, Kant considers someone who needs to borrow money and contemplates making an insincere promise to pay it back within a determinate time demanded by the lender. His maxim would then go: when I believe myself to be in need of money I shall borrow money and promise to repay it, even though I know that this will never happen. Kant argues that the attempt to turn this maxim into a universal law would lead to contradiction, since promises would never be believed, making promising itself impossible. Therefore, this maxim could not be willed as universal law. Third, Kant imagines someone with a talent that could make him ‘useful for all sorts of purposes’. However, he is in comfortable circumstances and tempted to let this talent rust. Kant claims that he cannot possibly will this as a universal law, for as a rational being he ‘necessarily wills that all the capacities in him be developed, since they serve him and are given to him for all sorts of purposes.’ Fourth, Kant imagines someone for whom things are going well, and who sees others not so fortunate. Imagine this agent thinks (effectively), ‘So what? I’m alright Jack.’ Kant argues that such a maxim could not be willed as a universal law, since many cases could occur in which one needs the help of others, a help that one has denied one’s self by the universalising of one’s maxim. Shortly after giving these illustrations, Kant tries to clarify the test of the Categorical Imperative by distinguishing a contradiction in conception and a contradiction in the will. Either contradiction is sufficient for a maxim to fail the test. In the former case, universalising of the maxim cannot even be thought of without contradiction. In the latter case, universalising of the maxim would clash with other willings of a rational being. It seems plausible that only the second of Kant’s illustrations can be supposed to involve a contradiction in conception, with the rest looking more like contradictions in the will. Regardless, what matters most for Kant is that application of the test is a requirement of rationality, incumbent upon all rational beings regardless of their inclinations. Kant goes on to offer another two versions of the categorical imperative. His view of the relation between them, as well as the actual relation (or otherwise), are very tricky questions we will not address here. We will note only the second version,        

Ibid., 32. Ibid., 32. Ibid., 33. Ibid., 33.

Metaethics Explored  43

sometimes known as the Formula of Humanity (other labels are Respect for Persons and Persons as Ends): So act that you use humanity, whether in your own person or in the person of any other, always at the same time as an end, never merely as a means. Kant applies this formula also to his four illustrations, and tries to show that it yields the same results as the Formula of Universal Law. Does Kant Succeed? Does Kant succeed in showing that morality is an affair of detached, universal reason, binding upon rational beings qua rational beings? Is this ruggedly objectivist picture of morality plausible? Is moral truth so detached from inclination and social response as Kant tries to make out? It should be said right away that the test of the Categorical Imperative, whatever it involves precisely, surely taps into a key element of our moral awareness. ‘What if everyone did that?’ is a question we ask often, and regard morally salient. If it is right (or wrong) for me to do x in situation y, then it is right (or wrong) for anyone in y to do x. It is the morally relevant features of situation y that make it right for anyone in it to do x. If it is wrong for me to evade my taxes, it is because it is wrong for anyone to enjoy the benefits of government without contributing anything to them. Similarly, the Formula of Humanity seems an immediately recognisable articulation of something morally vital. The formula is, among other things, a check against the exploitation of individuals for broader goals, however laudable the latter might be. This means that it is regularly viewed as a healthy counterpoint to utilitarianism, which, in its simplest form at least, seems to legitimate any treatment of individuals that increases happiness (see p. 15). However, many criticisms of Kant are possible. It is not obvious, for instance, that his test allows us to distinguish moral maxims from nonmoral ones. That is, there might be maxims that could be willed as universal law but which do not seem to be moral in any sense we would usually recognise. One commentator has memorably illustrated this point with the maxim ‘Always eat mussels on Mondays in March.’ Furthermore, the limits on what can count as a maxim are not clear. How specific may a maxim be? Kant rejects lying unconditionally, but is there anything to stop my conceiving ‘Always tell the truth unless to save an innocent life’ as a maxim (that I can will be universalised)? If not, then may I add another clause or two to it? When   Ibid, 38   Ibid, 38–9   A. MacIntyre, After Virtue (Duckworth, 1985), 46.

Metaethics Explored  44

would it cease to be a maxim? Flanagan combines these criticisms in one passage: Either we keep the maxims that guide our actions at a very general, contextually insensitive level, in which case we may be led to such oddities as not being able to tell a small lie for a terribly important purpose; or we are allowed to universalise maxims that are as fine-grained and contextually sensitive as we want, in which case we can will whatever we please without any inconsistency whatsoever. But this problem of drawing the line on the level of grain of the maxims is just the problem of deciding what features of the world are morally relevant and in what ways. The most general and perhaps most popular criticisms of Kant, though, concern the detached reason that Kant claims is involved in the Categorical Imperative. We can pose two overlapping but distinguishable questions. First, does reason really plough a lone furrow in Kant’s treatment, as Kant thinks? Second, should we properly expect morality to be exclusively an affair of a detached reason? Kant’s commentators do not generally find his four illustrations (pp. 41–2 above) convincing. Yet Kant intends them to demonstrate a resounding ‘yes’ to the first of the questions immediately above. The most famous utilitarian, John Stuart Mill, offers a memorable reaction to this hope. He claims that Kant fails, ‘almost grotesquely, to show that there would be any contradiction … in the adoption by all rational beings of the most outrageously immoral rules of conduct.’ All Kant shows, according to Mill, is that ‘the consequences (Mill’s emphasis) of their universal adoption would be such that no one would choose to incur.’ Of the second question, it is commonly held that the weakness of the Categorical Imperative lies in the unduly compartmentalised reason that it proposes. We might, for sure, contradict ourselves in doing something we would not wish others to do. But the knowledge that we contradict ourselves is not always enough to stop us. Kant recognises this by talking of respect for, and even awe of, the law of the Categorical Imperative. But such respect or awe cannot itself come from reason. It comes from the heart. Those brief reflections upon Kant exemplify a spirit in moral theory that is embodied most famously in Hume, a forerunner of utilitarianism and a central player in the next chapter. This spirit is one of scepticism about the moral purchase of reason.   O. Flanagan, Self Expressions: Mind, Morals and the Meaning of Life (Oxford University Press, 1996), 196.   J. S. Mill, Utilitarianism (Collins, 1962), 254.   Ibid., 254.

Metaethics Explored  45

How much, if anything, can reason do in the sphere of morality? (It is worth reflecting upon this in light of the topics of Chapter One.) If its contribution is limited, then what are the consequences for the nature or even the existence of moral truth? There are many, inside and outside of philosophy, who believe that there are no moral truths on the objectivist’s terms, Kantian or otherwise. That is, they believe that there are no moral truths that do not make reference to real or hypothetical subjective states, or to the agreements, practices, etc. of real or hypothetical groups of people. Some of the many go further, of course, and hold that there are no moral truths at all (irrealism). And some go yet further, and hold that moral statements have no truth value (noncognitivism). These views are the subject of the next chapter and, less explicitly, of the remainder of the book.

Chapter 3. Objectivist Realism under Siege 3.1 A Philosophical Health Warning The end of the last chapter suggests important differences between opponents of objectivist realism. This is an example of something we must be attentive to in philosophy. We might think that a similarity is significant enough to regard those who share it as a grouping. And that thought might be fair. At the same time, there might be very important differences within the grouping. Consider feminism. There are many different feminist outlooks, and whilst it is correct to regard each of them feminist, the differences between some of them might be as important as the similarities. Moral realism is an example from our discussion to this point. The only belief shared by all moral realists is that some moral statements, literally construed, are literally true. Moral realists, as we have seen, show disagreement as to what it is that makes some moral statements true. Similarly, there is no need for them to agree on which moral statements are true. The same note of caution applies regarding the opponents of objectivist realism. What they share is the belief that two claims are not both true. The first claim is that the objectivist account of the truth conditions of moral statements is correct. The second is that these (objectivist) truth conditions are sometimes satisfied. The belief that these claims are not both true is certainly significant enough to regard its adherents as a grouping. At the same time, membership of the group covers very different positions. First, it allows that the first claim is true and the second false. Here, objectivism gives the correct account of the truth conditions of moral statements (objectivism is true), but these conditions are never satisfied (realism is false). This is an example of the position, noted in the first chapter, that moral practice presupposes moral facts that don’t exist. Mackie’s famous error theory is an illustration, and may fairly be called ‘objectivist irrealism’. We will return in this chapter to one of Mackie’s supporting arguments.   See p. 9.   J. L. Mackie, 34–35.

Metaethics Explored  47

Second, it allows that the first claim is false. Here, the objectivist view of the truth conditions of moral statements is incorrect. However, this is neutral between subjectivism, intersubjectivism, and noncognitivism (and there are different versions of each of these). It is also noncommital on whether the truth conditions of moral statements—should noncognitivism be rejected—are ever satisfied. That is, rejection of the objectivist account of the truth conditions of moral statements is neutral between realism, irrealism, and noncognitivism. One could be a subjectivist realist, a subjectivist irrealist, an intersubjectivist realist, an intersubjectivist irrealist, or a noncognitivist. All of these positions affirm that objectivism’s account of the truth conditions of moral statements is incorrect. But there are, clearly, very substantial differences beyond that. The opponents of objectivist realism are therefore a broad church containing many members, a plurality of standpoints, and (very probably) a sizeable constituency who jealously protect their membership cards but are not very clear on exactly what their own metaethical positions are. Their most distinguished philosophical influence is Hume, and it is him to whom we now turn. 3.2 David Hume It tends to be the fate of the great philosopher to be seized by all and sundry as one of their own, and Hume is liable to be claimed as a sympathiser by any opponent of objectivist realism. This might be partly a result of the range, nuances, and (arguably) ambiguities and equivocations within Hume’s arguments. (He might sometimes intend these as playful philosophical devices.) Exegesis of Hume’s metaethics is notoriously difficult. But we need to try to appreciate and examine some of the Humean arguments that have been so influential. There are two major Humean arguments. The first is that morality requires sentiment, in addition to reason. Reason and Sentiment The hypothesis we embrace maintains that … morality is determined by sentiment. It defines virtue to be whatever mental action or quality gives to a spectator pleasing sentiment of approbation (italics Hume’s); and vice the contrary. We then proceed to examine a plain matter of fact, to wit, what actions have this influence. We consider all the circumstances in which these actions agree,

Metaethics Explored  48

and thence endeavour to extract some general observations with regard to these sentiments. It should be stressed that Hume takes reason to play a vital role in morality. The ‘spectator’ of the above passage is not any old spectator, but is the highly attuned spectator of ideal observer subjectivism. This is shown in an earlier passage, in which Hume proposes demanding conditions of ‘proper’ moral sentiment: It is often necessary … that much reasoning should precede, that nice distinctions be made, just conclusions drawn, distant comparisons formed, complicated relations examined, and general facts fixed and ascertained. Hume’s divergence from Kant is that he explicitly regards reason insufficient to generate morality. For Hume, there would be no such thing as morality in a universe of rational beings devoid of sentiment. Hume’s moral ‘spectator’ is defined also by a general sympathy towards others, and therefore approves of actions that promote human happiness and diminish misery, and disapproves of the opposite. However, the passage immediately above emphasises the difficulty we sometimes have in knowing which actions have this quality. And we need to get this right, otherwise our general sympathy might lead us into sentiments that are mistaken. How does Hume make the argument that morality requires sentiment? He splits the judgements of reason into two domains, matters of fact and relations of ideas (this is known as Hume’s Fork), and argues that moral judgements are not properly about either, despite a form that misleads us into thinking otherwise. For Hume, facts are, qua facts, objectivist in nature. Therefore, a subjectivist or intersubjectivist account of the truth conditions of moral judgements would not yield moral facts for Hume. This explains why Hume can endorse ideal observer subjectivism yet reject moral facts. We might be inclined to say, pace Hume, that ideal observer subjectivism entails moral facts that depend upon the approvings and disapprovings of an ideal observer. However, it is a result of Hume’s vocabulary that dependence upon sentiment rules out fact. Therefore, Hume’s rejection of moral fact is not a rejection of moral realism as we have defined it here (some moral statements, literally construed, are literally true). It is a rejection of objectivist moral realism (an example of which we find in Kant). Hume’s rejection of moral fact is a rejection only of the belief that some moral   D. Hume, Enquiry Concerning the Principals of Morals, in L. A. Selby-Bigge, Hume’s Enquiries (Clarendon Press, 1902), 289.   Ibid., 173.

Metaethics Explored  49

statements are true independently of anyone’s feelings and of the agreements and practices of any group of people. Why, in Hume’s view, are moral judgements not properly judgements about matters of (objective) fact? One of his arguments might be called the Argument from Complexity, because it claims a complexity in the circumstances that give rise to moral response. One example is the ‘crime of ingratitude’ (italics Hume’s). Hume challenges us to pinpoint the fact that is ingratitude. He concludes that this is an impossible challenge, because the ill-will or indifference felt by the ungrateful person is ingratitude only when directed towards those who have displayed good will towards us. This ‘complication of circumstances … excites the sentiment of blame (italics Hume’s).’ In another argument, Hume tries to show in one fell swoop that moral judgements are about neither facts nor relations. A ‘speculative reasoner’ concerning triangles and circles, he argues, tries to infer a relation he doesn’t currently know from those relations that he does know. But in the case of morality, there is ‘no new fact to be ascertained; no new relation to be discovered.’ Moral practice consists, not in the discovery of new facts or relations, but in the response of sentiment to all of the relevant facts and relations. In Hume’s illustration, we need to know whether a man was an aggressor in order that we know the morality of the killer’s action. But there is nothing then left for the understanding to do. Instead, it is the job of sentiment to then approve, disapprove, blame, absolve, etc. What other arguments does Hume muster for the conclusion that morality does not rest on relations? He returns to the crime of ingratitude and argues that it cannot consist in the relation of contrariety it involves, because the contrariety found when ill-will or ill-offices is followed by good-will or good-offices is often lauded. Hume sustains the flavour of this argument, again insisting upon the ‘complicated’ character of a moral action, such as ingratitude, which shows (he claims) that its moral significance cannot reside in a relation such as that between two plus three and half of ten. The complicated character of moral qualities means that no such relation exists. In another illustration, Hume asks why we wouldn’t regard it as ingratitude if an oak or elm tree produced a sapling which eventually destroyed the parent tree, when we .

         

D. Hume, Enquiry, 287–88. Ibid., 290 Ibid., 290 Ibid., 288 Ibid., 288

Metaethics Explored 50

deem it ingratitude if a human murders its parent. Hume offers, shortly before the last illustration, a broader argument intended to show absurdity in the idea that vice and virtue consist in ‘relations susceptible of demonstration.’ He argues that we must confine ourselves to four relations, ‘which alone admit of that degree of evidence.’ These are resemblance, contrariety, degrees in quality, and proportions in quantity and number. He argues that since these relations are applicable to inanimate objects, inanimate objects would be ‘susceptible of merit or demerit’ if the essence of morality consisted in these relations. Hume also considers the argument that the morality of actions consists in their relation to a prior rule of right. This argument is close to the cliché discussed in the first chapter, and Hume has a precise response (that finds echoes in that discussion). He rhetorically asks, first, what this rule is, in what it consists, and how it is determined. His second reply bites a good deal more, and takes the form of arguing that this argument gets things the wrong way round. He claims that we examine what actions please and displease us, and try to extract some moral generalities from that. And we have already noted his suggestion that the actions that please and displease us are connected by their respective promotion and undermining of human interests. Hume proposes, also, an analogy between the moral and the aesthetic. He notes that Euclid has fully explained all the qualities of the circle, yet has said nothing of its beauty. This is because beauty is not, in Hume’s eyes, a quality of the circle. It is a sentiment felt by the appropriately constituted mind upon exposure to the circle. Moral qualities are identical in character. Remove the appropriately constituted mind and you literally remove the moral qualities. Moral qualities depend upon sentiment for their very existences. A final Humean argument here concerns reason per se. Reason is essentially ‘cool’, restricted to the discovery of truths which themselves ‘beget no desire or aversion’. Why does Hume think this supports his argument that morality is based in sentiment and not reason? His argument is summarised in one paragraph: Since morals … have an influence on the actions and affections … they cannot be deriv’d from reason; and that because reason alone can never have any such influence. Morals excite passions, and produce or prevent actions.          

Ibid., 293 and D. Hume, A Treatise of Human Nature (Penguin Books, 1969), 518–19. Treatise, 515–16. Enquiry, 288–89. Enquiry, 291–2. Enquiry, 172.

Metaethics Explored 51

Reason of itself is utterly impotent in this particular. The rules of morality, therefore, are not conclusions of our reason. Hume’s picture of morality is therefore a robustly internalist one (see p. 11). The argument of the above passage summarises as follows: morality is intrinsically motivational (internalism); reason cannot motivate by itself; therefore, morality is not based in reason alone. Hume’s Error Theory Like anyone who denies objectivist realism, Hume is obliged to account for the fact that moral discourse seems to involve the assertion of objective fact. Hume’s explanation is that we project our reactions onto the world, ‘gilding or staining all natural objects with the colours borrowed from internal sentiment.’ This idea is sometimes called Projectivism. Ought and Is Hume’s second major argument is that one cannot derive an ‘ought’ from an ‘is’. It is expressed in the following passage: In every system of morality I have hitherto met with, I have always remarked that the author proceeds for some time in the ordinary way of reasoning, and establishes the being of a god, or makes observations concerning human affairs; when of a sudden I am surprised to find that instead of the usual copulations of propositions is and is not, I meet with no proposition that is not connected with an ought or an ought not … As this ought or ought not expresses some new relation or affirmation, it is necessary that it should be observed and explained; and at the same time that a reason should be given for what seems altogether inconceivable, how this new relation can be a deduction from others which are entirely different from it (italics mine). A crisper way of putting Hume’s point is that no amount or sophistication of nonmoral facts ‘gets one to’ a conclusion involving “ought”. The latter is qualitatively and irreducibly different.

  Treatise, 509.   Enquiry, 294.   Treatise, 521.

Metaethics Explored 52

3.3 Hume’s Arguments Considered Complexity Complexity surely does not mean that a domain is not one of objective fact. There are complex, objective facts in many areas. Consider physics, economics, culture, and the weather, for starters. Hume of course proposes the nature of the fact of ingratitude, i.e. ill-will or indifference towards those who display good will towards us. This is not a desperately complex suggestion, and seems quite a decent proposal. But inadequacy would not mean, either, that ingratitude is not a fact. And where the sentiment of blame comes in, it might be that our emotions are, as Ross would have it (see p. 29) engaging with the facts, some of which might be properly considered moral facts. Does Hume have an overly restrictive notion of an objective fact? The parallel response applies to his later Complexity argument that morality does not rest on relations. That relations are complex does not mean they are not genuine relations. Hume contrasts what he thinks is a complex moral affair with a simple arithmetical truth. But there are complex relations in arithmetic (and simple moral relations?) And his example of the oak or elm tree surely does not show that ingratitude is not a relation, though it might force us to work a little in order that we characterise it. The Speculative Reasoner The second argument seems, again, to beg the question. Why is it so clear that, in moral enquiry, we are not trying to ascertain facts we don’t already know? In trying to ascertain the morality of the killer’s action, we might well be trying to ascertain another fact, i.e. a moral fact. The Beauty of the Circle The same allegation of question-begging might well be made against Hume’s analogy with the circle. Whatever Euclid included, is it as clear as Hume presumes that beauty is not a genuine quality of the circle? The status of aesthetic qualities is a tricky question. But it is quite possible that aesthetic (and moral) qualities are robustly ‘out there’. It is possible, too, that beauty is, as Hume suggests, not a quality of the circle, but that moral qualities are genuine facts or relations. Perhaps there is no aesthetic-moral analogy.

Metaethics Explored  53

Moral Demonstration Hume’s argument to the absurdity in the idea that vice and virtue consist in ‘relations susceptible of demonstration’ embodies, perhaps, important mistakes. We must enquire what is meant by ‘demonstration’, and whether Hume is asking the right question. Hume seems to have in mind a very untypical, exacting standard of demonstration, and claims that absurd consequences follow if the only relations that meet this standard are regarded as the essence of morality. He might well be right to think that moral relations are not logically derived, and therefore do not mirror our derivation of half of ten from two plus three. However, when we try to establish the truth in morality, we do not operate with this fierce model of demonstration, which applies perhaps in mathematics and nowhere else. (We return to this point in the final chapter.) Therefore, we are not confined to the allegedly absurdity-generating relations that Hume offers. Hume might be guilty here of the popular malpractice of demanding in morality standards of ‘proof’ that we demand (or get) virtually nowhere else. Moral Rules (again) It is certainly legitimate of Hume to ponder the content of moral rules and how they are determined. We noted in the opening chapter that moral rules must be determined by something more basic, to save them from an arbitrariness that morality cannot countenance. We saw too that the purchase of moral rules might be quite limited. And we saw in the last chapter that few are convinced by Kant’s attempt to determine the content of moral rules. However, it isn’t clear that the picture Hume offers here is faithful to moral experience. Do we really need to examine which actions please and displease us, in order to reach substantive moral conclusions about why they so affect us? Indeed, is the specific pleasure or displeasure psychologically detachable from the qualities in the object that please or displease us? (Again, see the reflections on emotion in 2.6.) In the simple case of a very young child telling another ‘That was a horrible thing to do,’ is it feasible to regard the sentiment of revulsion as psychologically detachable in the way this Humean argument seems to make out? Hume’s argument here seems to depict our acquisition of moral knowledge as akin to the selfdiscovery that might result from visiting new places. And Another Thing … To what extent is it true to suggest that human happiness and unhappiness is the

Metaethics Explored  54

thread that connects (or ought to connect) our respective approvings and disapprovings? Are all moral principles really generalisations about social utility? Is all virtue to be finally explained in terms of this function? Social utility might offer a feasible origin of a substantial set of our moral responses. Indeed, any morality for which social utility is irrelevant would be odd. But mightn’t some moral pressures, such as justice and local interpersonal responses, make no essential nod towards social utility? And what about animals? Do we properly have moral responses towards them? Morality and Motivation Morality motivates by itself; reason does not motivate by itself; therefore, morality cannot be grounded in reason alone. How could this argument be challenged? The conclusion follows from the premises, so the argument is valid. But is each of the premises true? Neither is impregnable. The first states what has already been explained as Internalism. But internalism has been ably challenged and could well be false. Perhaps something over and above morality, such as the psychology of the agent or the content of moral truth, is required in order that morality motivates. (We will not go any deeper into this question.) If it is conceptually possible that one has moral knowledge that does not motivate, then the truth of the other premise—that reason does not of itself motivate—does not entail that morality is not grounded in reason alone. It could be that reason does not motivate, morality does not intrinsically motivate, and morality is a matter purely of reason. However, what of the argument’s other premise, i.e. reason does not motivate? There is no obviously compelling argument provided by Hume, and the motivational ‘impotence’ of reason has been challenged as ably as has his internalism (again, we will not go into the discussion here). Perhaps some of the truths uncovered by reason do entail certain desires. Again, it could be that morality does intrinsically motivate (internalism), reason intrinsically motivates, and morality is a matter purely of reason. Hume’s Error Theory Hume’s projectivism clearly depends upon his argument that there is an error. If moral judgements really are about objective truth, then our moral discourse is not inscribed   Hume robustly addresses apparent counterexamples such as these. See, for instance, Enquiry (183–204), ‘Of Justice’. A key question in the appraisal of the treatment is whether the concept of social utility is stretched to the point where it ceases to do useful work.

Metaethics Explored 55

with the alleged error. Our moral judgements are properly about something objective, instead of merely purporting to be. If there is, on the other hand, such an error, then Hume’s account offers a plausible explanation of it. But the Hume arguments we have explained and discussed are intended to show just that there is an error that needs explanation. And, as we have seen, it is not obvious that these arguments succeed. Before leaving this point, it is worth noting that the claim of human projection is strictly compatible with objectivist realism. It is possible that we project onto the world properties that it in fact has independent of these projections. These properties would exist whether we project or not (and perhaps we would project whether or not the world has these properties). Hume’s belief is that we project onto the world properties that it does not in fact have. Ought and Is A great deal of commentary has been offered on this argument. One obvious objection is that Hume’s allegation of the is/ought gap seems to beg the question by presuming that ‘ought’ statements are not statements about facts. Hume tries to distinguish the results of the ‘ordinary way of reasoning’ from an ‘entirely different’ relation that he suggests cannot be a deduction from those ‘ordinary’ results. But maybe the relation captured in the moral ‘ought’ is an ordinary result of reasoning too. This prospect will be reinforced in the final chapter. (It is worth recalling at this point that statements involving the moral ‘ought’ make up only one—perhaps untypical—kind of moral statement.) Furthermore, the Hume expectation that the moral conclusion should be a deduction from nonmoral premises again seems to smell of the demand for a standard of demonstration that is found perhaps nowhere except in mathematics. And, as we will see in the final chapter, there might well be powerful reasons for believing that moral statements can be derived from nonmoral statements according to normal standards of empirical reasoning. 3.4 Mackie’s Argument from Queerness A well-known modern argument against objectivist realism, combining elements of the preceding Hume arguments, has become known as Mackie’s Argument from Queerness. It was noted earlier that Mackie is an error theorist. He believes that all

Metaethics Explored  56

moral statements are false, since they presuppose nonexistent moral facts. These facts he alleges to be presupposed are ‘objectively prescriptive’. That is, they motivate independently of subjective or intersubjective features. Such facts, Mackie claims, would be radically discontinuous with the furniture of the universe, and their cognisance would require fantastic powers of perception. Therefore, there are no such facts. Mackie’s argument has, again, attracted much comment. It should be clear to you by now that one swift response is to challenge its internalist commitments. Mackie has a problem with what he regards as the intrinsically motivational character of moral facts. He wonders how objective facts could have this property. However, one might counter that moral facts do not have this property. They do not intrinsically motivate. If externalism is true, then this prong of Mackie’s attack is unsuccessful. One might, on the other hand, embrace Mackie’s internalist commitments but question, again, whether the notion of intrinsically motivational facts is so outrageous. However, it is quite possible to reject Mackie’s internalist commitments yet embrace the broader thrust of his Queerness argument. One might allow that moral facts need not motivate of themselves, yet uphold questions about the nature of these (objective) facts and how we can come to know them. A proper refutation of the Queerness argument needs to show that moral statements are about genuine, objective facts that we are capable of knowing. We will see, again, in the final chapter that the obstacles to this might not be as powerful as is often thought. 3.5 So Where are We? It is not obvious how much assurance the opponent of objectivist realism can properly take from the preceding arguments of Hume and Mackie. It is not clear that any combination of them shows objectivist realism to be false. There might yet be moral facts independent of subjective and intersubjective elements. In other words, objectivism might give the correct account of the truth conditions of moral statements, some of which are in turn true. On the other hand, the inconclusiveness of the preceding arguments would not show that objectivist realism is true. In the meantime, we will look at a robustly noncognitivist alternative to objectivist realism. This is emotivism. One emotivist, Stevenson, illustrates Hume’s significance when he says, ‘Of all traditional philosophers, Hume has most clearly asked the questions that … concern us, and has most nearly reached a conclusion that the   J. L. Mackie, 35.

Metaethics Explored 57

present writer can accept.’ 3.6 Emotivism A. J. Ayer is the most famous of the emotivists. This status probably owes more to the broader and influential philosophy underpinning his emotivism than it does to his contribution to metaethics in particular. One conclusion it inspires is that moral statements are simply ‘expressions of emotion which can be neither true nor false.’ Ethical concepts are ‘pseudo-concepts’ (107), which can be used also ‘to arouse feeling, and so to stimulate action’ (108). Different ethical words, such as ‘good’, ‘duty’, and ‘ought’, are intended to express and provoke different feelings (108). Ayer declares that his emotivism removes one of the problems that afflict simple subjectivism. Recall that simple subjectivism allows you and I to each be right when we make what appear to be contradictory ethical statements, since we are each reporting our own feelings (see p. 32). Emotivism removes this difficulty, since on the emotivist ticket we are not reporting our feelings but expressing them (109–10). Here, there is nothing for us to be right or wrong about—not even our own feelings—when we make a moral statement. (Contrast the statement “I am upset”, which can be true or false, with the act of crying, which can be neither true nor false.) Similarly, emotivism can dismiss the charge of infallibility that is made against simple subjectivism. Since emotivism means that no sense attaches to the idea of getting a moral judgement right (or wrong), there is nothing to be infallible about (110). Objection and Ayer’s Reply It seems appropriate to object to emotivism as one would object to simple subjectivism also, i.e. moral dispute is impossible. Moral dispute seems to be a robust fact, but if moral statements are no more than disguised expressions of feeling, then it is impossible to see how this is the case. Ayer’s reply is intriguing. He claims that, contrary to appearances, there are in fact no genuinely ethical disputes (110–13). When we seem to have an ethical dispute, we are arguing about only nonmoral facts. We attempt to show, for instance, that our opponent is mistaken about the facts of the case, the motive of the agent, or the general effect of actions of this type. We hope that the attempt at correction is successful and that a change in attitude follows. However, if the first happens but the second   C. L. Stevenson, Ethics and Language (Yale University Press, 1944), 273.   A. J. Ayer, Language, Truth, and Logic (The Camelot Press Ltd., 1946), 103.

Metaethics Explored 58

doesn’t, there is no more to argue about. We may feel that our own values are superior but we cannot bring forward any arguments to show this. Another way of putting this is that we harbour a misplaced conceit if we regard ourselves as ideal observers and hope to make our opponent more like one. There are no ideal observers. We can strive only to get the (nonmoral) facts agreed and cross our fingers for attitudinal convergence. This argument will undergo some evaluation in the next chapter. Where Does Emotivism End Up? There are, unsurprisingly, more sophisticated versions of emotivism than that just sketched. Stevenson, for instance, offers a substantially richer treatment than Ayer. This involves a picture of nonmoral belief and moral attitude as more intimately interwoven than the preceding picture suggests. It involves, too, a deep recognition, according to Hume, of the complexity of the facts that are sometimes relevant to moral attitude. Some later contributors have in fact suggested that emotivism is ultimately difficult to distinguish from ideal observer subjectivism. Why might this be so? The argument is that it is a part of the very ‘logic’ of one’s moral attitude that one takes it to be correct. And that involves in turn the belief that one’s attitude results from disinterested, impartial appreciation of the relevant facts. It is this quality that would, after all, make one’s attitude a moral one, instead of another sort of attitude. And this seems to mean, finally, that one expects anyone with a disinterested, impartial appreciation of the relevant facts to have the same moral attitude as one’s self. There might seem to be a key difference between emotivism and ideal observer subjectivism. The former, recall, says that moral statements are neither true nor false. The latter says, on the other hand, that moral statements do have truth value, and that this truth value depends upon the responses of an ideal observer. But perhaps this difference need not, on inspection, be so significant after all. As expressions of attitudes or feelings, it seems fair to say that the emotivist’s moral statements, literally construed, are neither true nor false. But, as we have just seen, there is room within emotivism for the notion that the attitudes or feelings can be mistaken. (My expression of anger is neither true nor false, but I might be mistaken to be angry.) And in what sort of circumstances are they mistaken? They are mistaken when they result from an incomplete appreciation of the relevant facts, or a failure to be disinterested and impartial. And if they are not mistaken, then everyone ought to have them. Compare   Stevenson, 273.

Metaethics Explored 59

this with ideal observer subjectivism. Moral statements are true or false. Here, what does the truth of a moral statement amount to? It means that someone who is disinterested, impartial, and appreciates the relevant facts would feel the approval or disapproval that the statement attributes to him. If I say that so-and-so is a good person, then I say that an ideal observer approves of so-and-so. And, clearly, in order to make such a judgement one has to inhabit the mind of the ideal observer to ascertain what would be approved and disapproved by a properly knowledgeable, impartial, and disinterested spectator. The same failure is responsible, it seems, for the mistaken attitudes or feelings admitted by emotivism and the false moral statements within ideal observer subjectivism. That is the failure to successfully inhabit the perspective of the properly knowledgeable, impartial, and disinterested spectator. Similarly, the same cognitive success grounds the emotivist’s rational moral attitudes and the true moral statements of ideal observer subjectivism. Does it really matter so much if we talk about rational/mistaken moral feelings or true/false moral statements? Are emotivism and ideal observer subjectivism so different? Emotivism and ideal observer subjectivism might therefore face the same difficulties, which we have met already. How do we ascertain which facts or beliefs are ethically relevant (whether the final result is an attitude or a statement with truth value)? That is, what facts is the ideal observer attuned to and why? What gives certain nonmoral facts specifically moral significance? The ideal observer disapproves of something, say, because of disapproval-making qualities it has. Can we account for this in a way that doesn’t court the objective moral facts that the opponents of objectivist realism reject? Wouldn’t these qualities have moral significance even if no one in fact disapproved of them? We will touch again on those points of impact in the final chapter. The next and penultimate chapter will articulate and critique some socially popular arguments against objectivist realism.

Chapter 4. Out on the Street 4.1 The Death of God This discussion has been pre-empted in the second chapter (2.7). A fair number of people say that morality depends upon God and that God does not exist. There is nothing substantive to add to the discussion of DCT in Chapter Two. We saw that there are insuperable difficulties for the notion that God is the source of moral truth. We saw too that if He were the source of moral truth, then it is less than clear that we should regard moral truth as objective, and more plausible that we should take moral truth to be grounded in a precise form of subjectivism. Therefore the introduction of God would not seem to support objectivist moral realism. Conversely, objective moral truth cannot be undermined by the absence of God, since it was never dependent on Him to begin with. 4.2 The Death of Deference It is said often these days that people are less deferential than before. They are much less likely to place uncritical trust in apparent experts and authorities such as doctors, lawyers, teachers, parents, politicians, monarchs, philosophers, priests, and imagined gods. But if all these erstwhile experts and authorities have feet of clay, then where can we find moral expertise and authority? Is there such a thing? And if not, what can moral truth amount to? Some people see this argument as grounds for rejecting all talk of moral truth, whilst others think it forces us into a DIY model of the truth in morality. But neither of these conclusions follows. First, the decline in deference might itself be morally mistaken. Perhaps the above ‘experts’ and ‘authorities’ are in fact genuine experts and authorities, and therefore merit deference. But a deeper response to this argument begins by allowing that the decline in deference is legitimate. What follows? The most that is shown is that some people and institutions are not so morally expert as previously thought, and that it is therefore wise to stop presuming that their moral pronouncements are true. Indeed, some brief reflection on one of the reasons

Metaethics Explored  61

for the decline in deference perhaps reinforces commitment to objective moral fact. Many people have concluded that a lot of the moral pronouncements that issue from alleged moral authorities are plainly morally mistaken, and might well have regularly immoral motivations too. For instance, the predominant beliefs of the major churches about sexual relations, sexuality, and sexual politics are now judged by many to be morally in error, and to be (at least partly) motivated by the subjugation of women. The forms of argument generally advanced here do not seem to reflect a subjectivist or intersubjectivist conception of moral truth. For instance, those who cite the subjugation of women as something immoral do not seem to believe that the immorality is a function of subjective feeling or intersubjective agreement. Furthermore, they are quite possibly correct. The subjugation of women was once a universal fact that was sanctioned by subjective feelings and intersubjective agreements (and see pp. 34– 35). However, opponents of female subjugation believe that it was wrong then, and it is most plausible that they are correct. Indeed, some believe that universal female subjugation continues to be an immoral fact that is sanctioned by subjective feelings and intersubjective agreements, and that no species of subjective feeling (e.g. ideal observer response) or intersubjective agreement (e.g. the agreements of idealised agents) could make female subjugation defensible. This belief requires, again, a commitment to objective moral fact. And again, if the belief is true, then there is objective moral fact. It is important to note that the question of who is morally correct on topics such as the preceding is irrelevant to the philosophical point. This is that features of canonical morality are rejected on what seem to be objective moral grounds. The popularity of this argument against objectivist moral realism provides an illustration of how many remain attached to a Command model of morality. Its premise is that moral ‘authorities’ no longer carry the oracular authority they once did. This premise is probably true. However, this can function as an argument against objective moral truth only if objective moral truth needs to be ultimately sourced in edict. And this is an assumption that many find it remarkably difficult to shake off. (Its spectre is manifested in popular questions such as ‘Who says that such-and-such is wrong?’ and ‘Who are you to say that such-and-such is wrong?’) However, the discussion of DCT has shown that this assumption is untenable. Therefore, a decline in moral deference provides no ground for the rejection of objective moral truth. Revealing clarification and strengthening of the last paragraph is provided by the realisation that, with the possible exception of aesthetics, people do not think in this way anywhere other than morality. We do not think, for instance, that mathematical

Metaethics Explored  62

or scientific truth is sourced in the pronouncements of authorities. There are authorities, for sure, but their authority resides in their proven ability to discover truths independent of their pronouncements. If the mathematical community, for instance, suddenly became disillusioned with some of its erstwhile authorities, they would be very unlikely to think that they therefore had ground for rejecting objective mathematical truth. They would think only that they had misjudged the credentials of their putative authorities. Morality suffers a different fate, however. For many, Command is the only objectivist game in town. 4.3 Moral Disagreement A lot of people seem to believe that there is more disagreement on moral issues than there is on other topics. And this belief is frequently combined with the belief that moral disagreement is in principle irresolvable (‘they are never going to agree’). Whilst this is a popular argument against objectivist realism, it isn’t obvious which metaethical stance it is intended to support. Is it proposed as an argument for noncognitivism, i.e. the total absence of truth or falsity in ethics? Or is it supposed to support, say, a simple subjectivist picture of moral truth? Perhaps its fans are not themselves universally clear or uniform in their metaethical conclusions. However, it is clear that the argument is minimally intended to undermine objectivist realism, and it is this aspiration we should now address. The simple factual claim contained in the argument is unconvincing. It is not clear that there is more moral disagreement than other kinds of disagreement. The impression of asymmetry between morality and all else is the result of a failure to compare like with like. Adherents tend to select a controversial moral topic, e.g. abortion, and contrast it with nonmoral topics on which there is little or no disagreement, e.g. simple mathematical or scientific truth. However, the controversy aroused by a topic such as abortion is probably untypical of morality. Much morality goes uncontested. For instance, it would be very hard to find anyone who doesn’t agree that torture for the fun of it is wrong (and cruel, perverse, inhuman, etc.), or that it is rude not to thank someone for a nice (or even mediocre) cup of coffee. Conversely, there are disagreements in maths, science, economics, history, and almost anything one cares to mention. Just as morality is not universally combustible, nonmoral belief is not   We have already noted strong reason (p. 37–8) to doubt that Divine Command theory of morality offers a genuinely objectivist account of the truth conditions of moral statements. The equivalent doubt would seem to arise about any Command account. It therefore seems conceptually muddled to posit Command as the source of objective moral truth.

Metaethics Explored  63

always as simple as calculating two plus three. However, what if we grant the (probably false) premise that there is more disagreement in morality than elsewhere? What follows? It does not follow that there is no objective moral truth. Rampant disagreement on anything is consistent with an objective truth that the disagreement is about. There needn’t be objective truth, of course. The truth that is contested might come down to, for instance, the response of an ideal observer. Here, the disagreement would be about what an ideal observer would approve and disapprove. But it could just as consistently be about something independent of subjective feelings and intersubjective practice. What about the other claim that tends to feature here, i.e. that there is a unique irresolvability to moral disagreement? This claim is much stronger than the last. The last is only that there is a uniquely high level of disagreement in morality, while this claim is that moral disagreements in principle cannot be resolved. If this claim is true, then noncognitivism and simple subjectivism would seem to be the only metaethical positions that could be entertained. However, there is no good reason to think that ethical disagreements manifest a unique irresolvability. First, some apparent moral disputes are not genuine. This is because they reflect the application of the shared moral principles to different circumstances. A good example would be the much-quoted eskimo practices of infanticide and leaving some old people in the snow to die. Contrary to appearance, these practices do not manifest a rejection of the value of human life. Instead, the circumstances of eskimo life make these practices a regrettable necessity that is ultimately affirmative of the value of human life. Any group that values human life would, if placed in eskimo conditions, respond in the same way. Second, many moral disagreements result from reasonable but resolvable disagreements over the nonmoral facts. A moral dispute may in fact come down to disagreement about, say, the consequences for those affected of alternative actions or policies. Examples suggested by David Brink are ‘What redistribution of a certain class of goods would make the worst off person in a society best off?’ and ‘What kind of life would my severely mentally retarded child lead, and how would caring for him affect my family and me?’ To these might be added the previous question (p. 34) of whether women’s reproductive systems are damaged by physically active lives. Answering stretching but tractable questions   D. Brink, ‘Moral Realism and the Sceptical Arguments from Disagreement and Queerness’, Australasian Journal of Philosophy, 62 (1984), 117.   A currently volatile example is provided by Female Genital Mutilation (sometimes called ‘FGM’). There are different moral arguments in support of it. One African tribe, the Dies, believe that the clitoris, if not removed, will grow monstrously until it is coming out of the nose and ears. If this belief were true,

Metaethics Explored  64

such as these is often the key to the resolution of moral disagreement. If we left things at that, then there would seem to be important convergence with Ayer’s claim (see pp. 57–58) that there are no genuinely ethical disputes. Remember that he claims that apparently moral disputes are really confined to the nonmoral facts, and he does not wish to regard these disputes as genuinely ethical in character. Brink and many others are happy to regard as moral disputes those that turn upon the nonmoral facts. You can consider whether disputes such as this are moral or not. What matters, though, is that the locus of dispute we have so far allowed could be endorsed by Ayer, since we have not asserted that there are genuine disputes over moral principles, far less say how they might be resolved. For Ayer, recall, there can be no such disputes, only—say—personal abuse. But there is reason to believe that Ayer is wrong. There is reason to think, instead, that there is dispute over moral principle (not just the nonmoral facts) and that such dispute can admit of resolution. It is a mistake to assume that justification in any area, including morality, proceeds only from general principles to particular cases. (This will be revisited in the final chapter.) Justification can proceed in the opposite direction too, i.e. from particular cases to general principles. In this way, we might move towards resolution of dispute over moral principles. Imagine one person believes, with Kant (see p. 18), that one of our everyday moral utterances is true, i.e. one should never tell a lie. Imagine that her friend rejects it. The latter might persuade the former to reject the principle by flagging up the example where truth telling would be known to result in murder, or the less dramatic case of the partner’s misguided haircut (see pp. 17–18). This might leave them some way from a final, agreed principle, but at least they agree on rejection of the unconditional prohibition on lying (and might be able to arrive at a final, agreed—maybe messy—principle via consideration of further particular cases). Similarly, two people might begin by arguing over whether a kind person will ever knowingly hurt another’s feelings. One thinks kindness can never sanction knowing hurt, and the other disagrees. But the latter might mention the example of an aspiring pub singer who isn’t good enough to make it professionally and whose feelings would be hurt on being told. Does one do the singer any kindness in not telling him the hurtful truth? If the answer is no, then kindness is compatible with knowingly hurting feelings. By arguing in this way from the particular case, one might persuade the other of the general conclusion that knowingly hurting feelings then there would be a strong moral case for the removal of the clitoris. However, the belief is false, therefore this argument in favour of FGM is unsound. The fact that it would be difficult to convince the Dies that this belief is false is irrelevant to the fact that it is false.

Metaethics Explored  65

is, after all, compatible with kindness. Agreement is reached on previously disputed moral principle. Another, important objection to Ayer is that we might have thought insufficiently about the commitments we make in the leap from nonmoral fact to moral principle. Consider, for instance, someone who tells us he believes that men are superior to women. We ask him why. He tells us that men are generally physically stronger, faster, and more powerful than women. The nonmoral claim he makes is true, and one to which most will readily agree. However, the leap to his moral conclusion requires elevation of physical strength, speed, and power, a presupposition that he might never have subjected to any critical examination. On doing so, he might genuinely wonder (as feminists generally do) whether there is any good reason to award primacy to these qualities, and he might finally conclude that he has, after all, no good reason to hold men superior to women. 4.4 Multiculturalism Many people nowadays are impressed by what they regard as radical and irresolvable differences in moral belief between cultures. It is widely held that what is right in your culture may not be right in mine, and there is no culture-independent standpoint from which the different cultures may be appraised and adjudication reached on points of divergence. This seems to be a form of intersubjectivism that gives rise to multiple moralities, since the truth comes down to whatever principles have been agreed or practices sustained in the culture in question. Most significantly, the argument’s fans take it to show that objectivist moral realism is false. It should be said immediately that much of the social psychology underlying this argument is worth affirming. It is a prescription of moral modesty, intended to correct the moral arrogance to which the white, Anglo-American middle-class has been particularly prone. But this modesty does not entail the absence of objective moral truth. The argument is a version of Conventionalism and is therefore subject to Conventionalism’s problems (see p. 33–34). It leaves it difficult to make sense of reformers from within moral cultures. It might be easily consistent with the reformer who argues that a particular practice, say, is incompatible with moral principles that the culture has agreed. But the project of some moral reformers is not like this. Some reformers, such as Martin Luther King, Nelson Mandela, and Emeline Pankhurst, seek to change basic moral principles of their cultures (in these cases that whites are superior to blacks and men superior to women). It is almost certainly the view of

Metaethics Explored  66

these reformers that the relevant principles are straightforwardly objectively morally mistaken. They might be mistaken here, of course. Perhaps the principles are correct, or their falsity is properly explained on, say, an ideal observer ticket. However, in addition to the general difficulties for subjectivism and intersubjectivism already noted, the most plausible explanation for the belief that the key moral principles are objective is that they are objective. For instance, the falsity of the principles that men are superior to women and whites superior to blacks seems not to be a function of anything subjective or intersubjective. As we noted in the case of female subjugation, there is no inconsistency in the notion that all subjective and intersubjective elements of a culture (or even a world) support a moral position that is incorrect. It is probable that the multiculturalist case against objectivist moral realism is more often stated than believed. It would be hard to find anyone who does not believe that the extension of the vote to women and the end of apartheid in South Africa are objective moral improvements. Most finally believe that the cultures in question previously had things, on the given topic, morally wrong through and through. It is possible that this belief is in fact mistaken. But it seems very difficult to sustain the argument that it is mistaken because the moral beliefs, principles, or practices of a culture are the final court of moral appeal. It more befits our aspirations of moral modesty to appreciate that different cultures might be able to morally learn from each other. Other cultures might be morally superior to our own in some respects. But this prospect entails a culture-independent moral standard against which individual cultures can be morally appraised. This standard might be, for sure, the responses of an ideal observer (say). But it might also be the moral truth of the objectivist. Finally, if it is true that this argument is motivated by modesty, then it is worth enquiring in passing about the status of modesty? Is it an objective moral good? 4.5 Rejection of Specific Moral Beliefs Recent decades have seen a considerable decline in the social popularity of arrangements that previously provided strong moral cues. Marriage is less ‘given’ than before. Stigma no longer attaches to divorce and to procreation outside of marriage. There is greater acceptance of homosexuality. Single parenthood is more readily accepted, as is ‘second families’. Social scepticism about much that once seemed sacrosanct has encouraged the conclusion that morality is either spurious or is no more than the codes of behaviour that a given culture endorses at a given time. What can we properly conclude from this set of modern clichés? The answer is

Metaethics Explored  67

foreshadowed in some of the previous sections of this chapter. We cannot conclude support for the rejection of objectivist moral realism. Again, the said social changes gain their popularity on moral grounds. It is strictly possible, again, that the only legitimate moral grounds of any social change are ultimately subjectivist or intersubjectivist. But it also possible (and plausible) that there are objective moral grounds on which change (or no change) is favoured. And that is surely the way in which most people finally see it. Those who accept, for instance, single parenthood, think the previous rejection of it objectively morally wrong (and those who continue to reject single parenthood deem the previous rejection objectively morally correct). We often hear someone advance the rejection of specific moral beliefs as grounds of the absence of objective moral truth, before arguing in favour of a change in moral belief on grounds that sound suspiciously objective. 4.6 Seedy Underbelly of Conventional Morality There is greater social awareness of the dubious social and ideological functions served—or apparently served—by elements of conventional morality. For instance, conventional marriage is seen as a bulwark of a patriarchal, sexist order, and the work ethic a device that encourages some to work for the benefit of others who profit from their labour. These apparent realisations encourage the conclusion that morality is a ruse for the benefit of socially, economically, and culturally ascendant groups. This argument smells of the perspective of the angry young men and women noted in the first chapter. And there is some insight in it, of an arguably important kind. Some of the allegations it makes might well contain truth of which we should be aware. However, we cannot conclude the falsity of objectivist moral realism. All we can conclude is that the ‘morally correct’ might not be the morally correct. What a culture holds up as just and proper might in fact reek of injustice and perhaps other vice to boot. Radical thinker Terry Eagleton neatly encapsulates what is in fact the insight and error of the angry young person (he doesn’t use any term like this) when he considers the view that ‘morality stinks.’ This view is, he says, correct: ‘morality is the Moral Majority.’ Equally, he warns, it is wrong, since morality is ‘much too important to be handed over contemptuously to the [political] enemy.’ It is significant that those who complain of the preceding features of social reality generally think we have reason to change them. And what sort of reasons are they, if not moral? For instance, it is only because a sexist social order is seen as morally   T. Eagleton, The Ideology of the Aesthetic (Blackwell, 1990), 372.

Metaethics Explored  68

unsustainable (unjust?) that a practice taken to confirm it (e.g. ‘traditional’ marriage) is rejected. And, again, it might well be objective moral truth that is properly invoked in the protests and manifestos for change. 4.7 Decline in Communal and Symbolic Life This section overlaps with the last two. Life seems to have become privatised to a significant extent in much of the western world. Communal life, myth, and ritual seem less important, as people turn increasingly to private pursuits. The extended family has lost its former importance, organised religion and party politics have less attraction than before, and festivals such as Christmas and Easter are heavily expunged of much of their former social and mythological purchase. This transition gained potent reinforcement in the ‘laissez faire’ ideology, embodied in the figures of Margaret Thatcher and Ronald Reagan, that dominated the political and social landscape of the West in the 1980s. The apparent decline in communal and symbolic life has encouraged the feeling that individuals need to morally fend for themselves. If we allow that this sketch is broadly accurate, how much can we conclude? Can we conclude that objectivist moral realism is false? Surely we cannot. First, it is, again, possible that change just sketched is morally in error because it conflicts with objective moral truth. There are very probably plenty that believe this. They might be wrong, but the fact that the social tide has turned against what they regard as objective moral truth does not entail that they are wrong. Second, it is significant that those who think they are wrong often employ supporting arguments that smell like the invocation of objective moral truth. For instance, some argue that the preceding communal and symbolic elements tend to stifle vital, life-enhancing energies of individuals, and are therefore contrary to human flourishing. (Mrs. Thatcher had nothing less than a theory of human nature—homespun, no doubt—to support her laissez faire vision.) This sounds like an argument from objective moral truth. We see, in the change to a more laissez faire environment, not evidence for the end of objective morality, but another moral position. The correctness of this position would itself give no reason to conclude that objectivist moral realism is false. 4.7 Reluctance to be Judgemental There is now greater sensitivity to the social and psychological influences upon behaviour and character. This is especially so with regard to disadvantaged groups

Metaethics Explored  69

and individuals, and members of cultures which seem opaque to us. There is caution in many cases about responsibility and the related condemnation. This is liable to slide towards the different conclusion that the said individuals and groups do not act wrongly when we previously thought they did. It might well be, again, that this change has some healthy motivations and is to be welcomed to a large extent. Insofar as it is a counter to premature moral conclusion, swift condemnation, and harsh punishment, it is surely a good thing. However, it falls well short, again, of showing objectivist moral realism to be false. There is a distinction, often overlooked, between the moral character of an action (say) and responsibility for carrying it out. It is consistent to hold that something is wrong and an agent is not responsible for doing a wrong when he does it voluntarily. Doesn’t it have to be feasible that the agent can come to know that she is doing wrong? For instance, imagine that one regards it wrong to eat meat. In that case, one holds that eating meat is wrong tout court, and that anyone who does it therefore does something that is wrong. However, one might reasonably desist from undifferentiated moral condemnation of all carnivores, on grounds that include the fact that many are born into cultures that make it very difficult to even raise a moral question about meat consumption (other, morally contestable practices could no doubt be substituted). What they do is wrong, sure, but responsibility is a bigger, more complex affair that softens or even eliminates condemnation of individual carnivores. (Perhaps we bear some responsibility ourselves?) There is in fact a frequent whiff of this distinction in the loud unwillingness of some to be judgemental. It often isn’t the unwillingness to judge something wrong—however it might advertise itself—but unwillingness to attribute undifferentiated moral responsibility and therefore to roundly condemn. This is, as already said, a viable and perhaps vital distinction. But it does nothing to show the absence of the objectivist’s moral truth. Before leaving this section, it is worth considering whether some of the anxiety here might be grounded, again, in an inability to extricate ourselves from a Command model of morality. Do we fear that if we explicitly judge others (especially others remote from us) to do wrong, then we assume an oracular authority that isn’t grounded in anything else? (Is this what people mean when they accuse others of ‘playing God’?) 4.8 Postmodern Irony This influence provides further theoretical gloss to the reluctance to explicitly judge

Metaethics Explored  70

others—especially remote others—to be morally in the wrong. It is prefigured in the account and discussion of Intersubjectivism (see pp. 28–29 and pp. 34–35). ‘Postmodernism’ connotes many things, including scepticism about objective truth of all kinds. A flagship theme is that there is no rational necessity to the moral structures that we have. They could have been otherwise. Rorty, a major philosophical sponsor of postmodernism, has said: ‘I do not think there are any plain moral facts out there in the world … nor any neutral ground on which to stand and argue that either torture or kindness are preferable to the other.’ He therefore proposes a stance of ironic detachment towards our moral sensibilities. ‘Postmodern irony’ has in recent decades become woven into the cultural fabric, to the point where many seem to regard it as a criterion of intellectual, social, and cultural credibility. This book is concerned in part with the question of whether there are any plain moral facts, and if so, what form they take. In the meantime, it might be useful to reflect a little on two concepts that appear in the Rorty passage. One is that of a ‘plain fact’. What is a plain fact? The other important concept is ‘neutral ground’. What is neutral ground? Is there such ground, anywhere? There is more to say on these questions in the next and final chapter.

  R. Rorty, 173.

Chapter 5  A Naturalist Objectivist Realism? 5.1 The Basic Features • There are moral facts that are expressible in moral statements and that are independent of moral beliefs, subjective feelings, and intersubjective practices. • Moral facts are natural facts, i.e. moral terms such as ‘good’, ‘just’, ‘right’, ‘wrong’ (and the others figuring in our Everyday Moral Utterances) typically refer to real properties, relations, events, or states of affairs that are neither theological nor transcendental nor in any way queer. • We know many moral facts by normal standards of empirical enquiry and reasoning. The first of the above entails objectivism, cognitivism, and realism. Moral statements have truth value (cognitivism), objectivism gives the correct account of their truth conditions, and some of them are true (realism). The second states naturalism and the third sits comfortably with naturalism. A cumbersome label for the position outlined is Naturalist Objectivist Moral Realism. You will agree this is best avoided, so it shall be called Naturalist Objectivist Realism. (Note that objectivist moral realism needn’t be naturalist. There might be, say, supernatural moral facts that do not come down to the subjective states of a deity or intersubjective arrangements of deities. And naturalism needn’t be objectivist. Subjectivist or intersubjectivist realism could be reasonably held to yield natural moral facts.) If there is much truth in the reflections of the last chapter, the above bullet points are untenable to the majority of people in the liberal democratic world, as they are to a good many philosophers. Any opponent of objectivist realism is bound to fulsomely reject them. This brief text does not pretend to offer a demonstration of naturalist objectivist realism. But the concluding chapter will try to show that some of the obstacles to its acceptance are illusory or exaggerated. These alleged obstacles might involve errors   See p. 37.

Metaethics Explored  72

about the status of objective nonmoral facts and about the nature of argument and demonstration away from the moral. If so, then some of the disanalogies between the moral and nonmoral evaporate. In the meantime, we will consider a well-known argument that moral facts cannot be natural facts. Recognition of its key error will help pave the way for the positive arguments in favour of natural, objective moral facts. 5.2 Moore’s Open Question Argument G. E. Moore wasn’t very comfortable about his idea of the natural, but offers a few suggestions. One is that natural qualities are ones we can know about by ‘empirical observation and induction.’ And he thinks he shows that basic ethical questions are not about such qualities. His official position (perhaps in conflict with his broader treatment) is not only that we cannot argue about basic ethical questions on empirical grounds, but that we cannot argue about them at all. (In this respect, his approach seems identical to that of Ayer. See pp. 57–58.) His arguments against naturalism focus overwhelmingly on intrinsic goodness, and not rightness or any other morally relevant quality. This might be a limitation. However, Moore’s key argument should be considered on its merits, since it might well contain a deficiency that would apply should any other morally relevant quality be substituted for goodness. The strategy of the Open Question Argument (‘OQA’) is simple. It first asks us to consider two natural properties. One is pleasure and the other is what we desire to desire. In each case, Moore claims, it is an open question whether the given property is good. That is, we cannot ‘read off’ goodness from knowing that something is what we desire to desire or that it gives pleasure. ‘Goodness’ does not mean either of these things. Therefore, goodness cannot be either of them. Moore generalises from these examples to all natural properties. For any natural property, it is always an open question whether it is good. Therefore, goodness is not any natural property. Does Open Question Argument Succeed? The most obvious quibble with Moore’s argument is that a couple of examples seem a weak case for such a generalisation. A second objection is that the argument seems to beg the question by presuming that goodness is not itself a natural property, a point to be revisited. We can, however, accommodate this possibility but preserve the spirit   G. E. Moore, Principia Ethica (Cambridge University Press, 1903), 39.   Ibid., 14–17.

Metaethics Explored  73

of Moore’s argument and ask where it leads. Naturalist opponents of Moore generally grant that moral terms cannot be defined, without loss of meaning, by nonmoral terms. If we try to reduce moral qualities to the nonmoral, then we lose something essential to them. We cannot properly analyse ‘morally good’ into ‘pleasure’ or anything else not involving a nonmoral term, in the way that we analyse ‘bachelor’ into ‘unmarried male’. Moral statements cannot be logically derived from nonmoral statements, as was suggested in discussion of Hume (pp. 52–3). But does it follow that moral properties are not natural properties? It is less than clear that it does. Again, there seems something question begging about the argument, insofar as it presumes that the irreducible statements of morality are not statements about natural facts. Facts within chemistry and physics, for instance, might not be reducible to facts within another language, but no one seems therefore tempted to conclude that the facts of chemistry and physics are not natural. It is fair to suspect that this question begging is underpinned by the intellectual elevation of the (so-called) hard sciences. That is, disciplines such as chemistry and physics are believed to deal in the coin of the natural, since they are believed to be concerned with ‘out there’ observables, whilst ethics isn’t (allegedly) about observables and must therefore prove its natural credentials. The fact that moral statements cannot be logically derived from nonmoral statements is considered both a result of and a confirmation of the fact that morality deals in a ‘queer’, nonnatural coin. We will return to this alleged asymmetry between the moral and the scientific. It can be shelved for the moment, because we should note too that there are numerous other disciplines and domains that are acknowledged to be irreducible but that are not generally taken to deal in nonnatural facts. Consider psychology, the social sciences, economics, and culture, for starters. Reduction of any of these domains to, for instance, physics, seems a very silly thought. How could we translate a statement within one of them to a statement within the language of physics (or any other language), without loss of meaning? Conversely, how could we logically derive a statement within one of them from a statement within physics or any other discipline or discourse? But irreducibility in these cases does not lead anyone to think that we are dealing in nonnatural facts. Why should morality be different? Why should the irreducibility of the moral mean that morality does not deal in natural (and objective) fact? Why must moral statements be logically derivable from nonmoral statements in order to count as statements about natural facts, when this criterion of the natural is not generally applied elsewhere? A final and potent illustration is provided by a landmark theory of the mind. According

Metaethics Explored  74

to Anomalous Monism, mental language is irreducible. Mental states cannot be logically derived from non-mental states. And mental states are physical states. By any standard, this would make mental states natural states. Anomalous Monism might be false, but its possibility provides another illustration of how the irreducibility of a language might coexist with natural facts in which the language deals. We can conclude that Moore’s Open Question Argument does not succeed in showing that moral facts are not natural facts. 5.3 Empirical Reasoning and Moral Facts So Moore’s OQA might establish that moral facts cannot be logically derived from nonmoral facts. However, there might be good reason to believe that moral facts can be and often are established by normal canons of empirical reasoning. Consider the following argument for one of the everyday moral utterances at the start: 1. She did a huge amount to help others in need. 2. She helped a lot of people fulfil their potential. 3. She was a dedicated mother and daughter. 4. She brought a lot of good into the world. 1–3 seems to state ordinary (natural) nonmoral facts that are independent of our subjective or intersubjective beliefs and responses, providing powerful grounds for accepting 4. The truth value of 4 seems similarly independent of our subjective and intersubjective beliefs and responses. If 4 is true, then we should believe it and approve of her, but it is not made true by our believing or approving. It is considerations such as 1–3 that make it true. 1–4 seems to follow an ordinary pattern of empirical reasoning. It is true that 4 is not logically entailed by 1–3. We do not derive 4 from 1–3 as we derive (in Hume’s example) half of ten from two plus three. But this is uninteresting. In ordinary empirical reasoning, inferences that are not logically entailed are commonplace. One contributor has noted that the existence of a cell cannot be logically derived from a complete description of its constituents. Similarly, the conclu  D. Davidson, Essays on Actions and Events (Second Edition) (Clarendon Press, 2001), 207–25.   Acts, consequences, and motives could each be relevant to the truth values of 1–3. See, also, pp. 18–19.   p. 49.   J. S. Russell, ‘Moral Realism in Sport’, Journal of the Philosophy of Sport, XXXI(2) (2004), 151.

Metaethics Explored  75

sion that it is spitting rain cannot be logically derived from the fact that I see from my window someone put up an umbrella. Yet I ordinarily infer this conclusion from my observation. The fact that 4 is not logically entailed by 1–3 allows us to preserve the autonomy of morality, but in an unexceptional manner that is consistent with explaining the autonomy of other domains, such as biology from chemistry, chemistry from physics, the mental from the physical, the social from the economic, and so on. We can account in the same way for statements involving the moral ‘ought’, despite Hume’s assertion of an is/ought gap (see pp. 51, 55). Consider, again from our Everyday Moral Utterances, ‘You ought to return your library books on time.’ Here, 1–3 might involve premises about the desires and reasonable expectations of other readers to the borrowed books, the inconvenience caused to library staff by late returns, and the reliance of the system from which you benefit upon timely return. By ordinary empirical reasoning, we infer 4, ‘You ought to return your library books on time.’ But We Don’t Observe 4 or its Effects, do we? We saw in the third chapter that this belief about moral qualities gets expression in Hume (3.2), where it was also suggested then that Hume’s arguments are not fully compelling (3.3). On the one hand, it seems popularly believed that we don’t observe moral properties, and also that moral properties have no observational implications, i.e. we can’t make predictions about what we will observe on the strength of moral properties. But the apparent popularity of these beliefs seems belied by the things that people ordinarily say. Consider, again, the above example. It would seem unexceptionable for one to make any of the following statements: 5. I saw her goodness often. 6. I witnessed the effects of her goodness upon others 7. Her goodness will make her welcome on the panel. The first statement reports observations of the moral property of goodness. The second reports observations of effects of the same moral property. The third predicts an observationally testable effect of it. If (5)–(7), literally construed, are literally true, then the following conclusions are literally true:   Michael Smith has argued that we can logically derive ‘ought’ statements from ‘is’ statements (Ibid., 190–193). The argument is intriguing and, in light of Smith’s whole treatment, not unpersuasive. However, do we need to make this tricky case in order to meet Hume’s is/ought argument?

Metaethics Explored  76

• Moral properties are observed • Effects of moral properties are observed • We can predict and observationally test effects of moral properties Indeed, Nicholas Sturgeon might well be right that the most compelling grounds for ethical naturalism are (i) ethical properties appear to play a causal role in the natural world, and (ii) it is plausible to think that any property that can play such a role must be a natural property. The above bullet points seem to support (i), and seem to follow from notably ordinary moral statements. Of course, the bullet points might be false. Perhaps (5)–(7) are each false and each require, as Mackie believes (see pp. 55–56), nonexistent properties of reality. Perhaps (5)–(7) are true, but only ‘in a sense’, e.g. goodness is not observed, but is (as Hume believed) merely ‘projected’ by observers (see p. 51). But, as we have just recalled, Hume’s metaethical arguments are not obviously decisive. More generally, might there be some oversights typically involved in the denials of the above conclusions? Perhaps one oversight is a mistakenly ‘pure’ conception of nonmoral observation. This conception is implicitly contrasted with what is involved in moral judgements, yielding another doubtful disanalogy between morality and other domains and disciplines. It is correct that a singular moral judgement (‘there goes her goodness again’) is inscribed with background theory or principle, whose approximate truth is presupposed. It is impossible to conceptually disentangle a belief that an act of goodness has been observed from broader, acquired notions about the nature of goodness (and see p. 19). Consider what sort of things would be invoked in a dispute about whether an act was one of goodness. The theoretical background to some moral observations might even be quite messy. But this does not distinguish morality from anything else. Scientific observations would often fail to be the observations that they are in the absence of an acquired theory whose approximate truth is presupposed. For instance, someone who doesn’t accept the relevant atomic theory might not accept that he is observing a proton (‘there goes a proton’) in a cloud chamber. Again, in more complex cases, observations might entail a quite messy theoretical background. Consider some of the more obscure areas of current physics. Consider, again, Freudian psychoanalysis: someone who rejects Freud’s theory of the unconscious might not accept that he is observing patient resistance to therapist suggestions. Even ordinary perception is infused with theoretical assumptions, e.g. object permanence, reliable functioning of   N. L. Sturgeon, ‘Moore on Ethical Naturalism’, Ethics, 113 (2003), 538.   Again, effects of failing to do what one ought to do (such as returning one’s library books on time) can be observed and predicted (e.g. staff and reader irritation).

Metaethics Explored  77

sensory organs, the absence of deception by a demon, etc. If ‘neutral ground’ (see p. 70) is supposed to be a non-theory-dependent standpoint that yields ‘plain facts’, then there is no such ground and there are no such facts, neither in morality nor anywhere else. People tend to overlook the theory-laden character of nonmoral observation, and therefore hold to a spurious asymmetry between morality and everything else. (The popularity of this asymmetry is particularly strong, and particularly dubious, when morality is compared with the physical sciences.) Essential background baggage is no good reason not to regard nonmoral observations as observations of qualities that are natural and independent of our beliefs, responses, agreements, and practices. It is reason only to be mindful of what such observations involve. And might just the same hold for moral observations? Dialectic Indeed, it seems that in both morality and elsewhere, there is an open-ended interplay of observation and theory or principle. This process is often called ‘dialectic’. For instance, the wrongness of child abuse might be a singular moral observation that we are unwilling to give up. Any moral theory or principle that could falsify this observation would have to be rejected. Conversely, the equality of the sexes and races might be a principle that we are unwilling to reject, therefore any apparent moral observation that threatens the principle (e.g. that certain discriminatory treatment is fine) would have to be rejected. However, beyond the theory-limiting observations and the observation-limiting theories and principles, there is plenty scope for amending observation and theory or principle in the light of the other. (One way of putting this is that the relevant observations, theories, and principles are defeasible, i.e. they can be defeated.) Such trade-offs are made in such a way as to maximise initial commitment, overall consistency, and explanatory power (this outcome is sometimes called ‘reflective equilibrium’). Paradigm shifts, furthermore, are possible. We can arguably find these in the shift in attitudes to cruelty and the increasing power of the principle of ‘democracy of the emotions’, according to which previously marginalised groups, such as women, children, non-whites, and animals matter increasingly. The current   It is a familiar point that if empiricism is stretched to its logical conclusion—we genuinely disallow knowledge claims not entailed by experience—then we can properly claim access to nothing other than our own experiences. This worry has led some eminent philosophers to reject the commonsense assumption that physical objects exist independently of all experience and to recast them as potential or actual collections of mental content.   See A. Giddens, Reith Lectures (1999), Lecture 4, ‘Family’: http://www.bbc.co.uk/radio4/ reith1999/lecture4.shtml.

Metaethics Explored  78

moral urgency about the environment is perhaps recommending another. These shifts are possible only against a body of more or less fixed moral observations and principles, e.g. that individual suffering is bad, and that the fulfilment of an individual organism’s potential is good. We see the moral intelligibility of these shifts, or proposed shifts, against our existing equipment. Any proposed moral paradigm shift that failed to connect with any of our existing moral equipment would be unintelligible. We need the rough assumption that there is some truth in moral common sense. Again, this is exactly how it is elsewhere, in the apparent havens of natural, objective fact. If we seem to have sufficient reason to accept a new theory in, say, physics or economics, then some of our previous observations might have to be rejected. If these observations include some that we are independently unwilling to reject, then we will think again and reject the theory. For instance, if quantum mechanics entailed that we reject the existence of the moon, then the theory would be rejected. If not, then we will reject some old observations and affirm the new theory. Paradigm shifts are, again, possible. The shift to quantum mechanics, again, might be an example. Another might be the shift in economics from the Keynesian collectivist model to the Friedman-Hayek free market model. Again, paradigm shifts will not be embraced if they fail to connect with observations and principles we are unwilling to reject. Indeed, such shifts can be entertained in the first place only against such a background. There is, again, no reason to make the shift if it doesn’t connect with anything we already recognise. Foundations Perhaps the impulse, discernible in Hume and Moore, to logically derive moral conclusions from nonmoral premises is underlain by an impulse to ground morality on absolute foundations which solidly support a structure of moral knowledge. And it might seem that definitions such as ‘good’ means ‘pleasure’ could provide such foundations. Whilst Hume and Moore go on to say much about metaethics, many people might think that without absolute foundations morality is spinning in a cognitive void. This thought is almost certainly mistaken and almost certainly involves, again, mistaken notions of procedure and knowledge elsewhere. Not only would the absence of such foundations not have the feared consequences for moral knowledge, but there are no such foundations in other areas, including the physical sciences. Neither in   If some reflections in the last chapter are correct, a significant number of people in the liberal democratic world do believe that morality spins in a cognitive void.

Metaethics Explored  79

science, morality, nor anywhere else can reason go ‘all the way down’. What would such an absolute justification of norms be like? How is it intelligible? The attempt to justify norms in any field will eventually invoke criteria that assume the norms in question. There can be no absolute justification of norms that does not finally beg the question. (This in fact reflects the broader, ancient philosophical problem of the Criterion: how can we know we have a good method of finding the truth unless we already know some truths?) In all areas there is liable to be at any given time a cluster of principles and theories, with no single, overarching principle of operation (and if there were, we could, again, legitimately ask what grounds that). Knowledge, both in morality and elsewhere, might be more like an ecology than a building. It might be messy, piecemeal, uneven, and have at its ‘foundation’ nothing more than the rough assumption that we know one or two things (e.g. that pointless suffering is a bad thing). Again, it seems that the attempt to find a disparity between morality and areas in which we take ourselves to have objective knowledge fails. 5.4 Morality and Objectivist Naturalism Indeed, we can perhaps sustain this idea in the concluding reflections of this book. Recall our original question of what morality is all about. Again, if a crisp answer to that seems tricky, we ought not to worry, since a chemist, physicist, psychologist, economist, or philosopher might struggle to give a crisp answer to the parallel question about her discipline. The territory of each is loosely mapped out, and liable, indeed, to contestation. It is readily believed that the domains of natural, objective truth occupy tightly circumscribed territories, and employ foundational principles and theories grounded in unmediated observations. I hope that enough has been said to convince you that this picture is wildly inaccurate. If morality fails to measure up to it, then it has many companions in guilt. The area that practitioners in a discipline elevate as its true territory is liable to be influenced by their pet theories within it. Therefore, answers to the ‘big question’ about ethics are liable to be influenced by the normative theory one favours, e.g. the Kantian, utilitarian, and virtue ethicist are liable to respectively champion practical reason, happiness, and character conducive to flourishing. However, it is not clear that there need be an all-purpose moral algorithm, such as Kantianism or utilitarianism, in order that there be natural, objective moral fact. It is entirely possible that there is no such algorithm, and that such have been heavily motivated, again, by the fear that without them morality would be spinning in a cognitive void. Perhaps morality

Metaethics Explored  80

involves natural, objective facts within a loosely mapped territory, and an ecology of theories and principles (including universalisation, happiness, motive, character, and the Formula of Humanity) that at once result from these moral facts and help make them the moral facts that they are. Again, the considerations of the first chapter should make the prospect of multiple norms in morality unsurprising. This prospect is ultimately grounded in moral observations and explanations we already accept, e.g. the inherent badness of suffering, and the fact that there are beings for whom things can go well or badly. If we did happen upon a systematic, all-encompassing moral theory, then it would have be grounded in the same way. How else could it be grounded? Moral observations depend on moral theories and principles. Moral theories and principles depend on moral observations. There is perpetual trade-off between the observations and the theories and principles (dialectic). We disagree about the moral facts and about the theories and principles. We disagree about how these affect each other. These disagreements are sometimes resolved swiftly, sometimes more slowly. Sometimes they are not resolved at all. Sometimes there are paradigm shifts that shake up our thinking and responses. If they are legitimate shifts, they hook up in important (but not all) ways with what we think we already know and are compatible with observations and singular judgements we are unwilling to relinquish. It might well be that in these respects morality is in no way odd. Instead, it might be exhibiting the usual features of natural, objective truth. 5.5 Ethics and Natural Disciplines Two objections might be offered. First, if ethics is natural, then shouldn’t there be a natural discipline that deals with it? Second, if ethics is natural, shouldn’t we expect points of continuity and areas of intersection with other natural disciplines? The obvious naturalist answer to the first is that there is a natural discipline that deals with the subject matter of ethics, and that is ethics. The naturalist answer to the second is that we have the expected continuities and intersections. From Plato onwards, ethical questions have been tackled in the context of theories of human nature and the human condition, with the objective of fitting ethics into an accurate overall picture of human life. This objective might suffice, given the complexity of human life, to explain why any all-purpose moral algorithm is liable to fail. And there is no reason why the constituency cannot be broadened, as many philosophers have broadened it, to include animals and the environment. The relevance of the nonmoral for the moral has been recognised several times in this book, specific examples of

Metaethics Explored  81

which were noted in the discussion of moral disagreement (see esp. p. 63–64). There seems every reason to assert important (nonreductive) dependencies between ethics and other natural disciplines, just as we have important (nonreductive) dependencies between (say) psychology and biology, biology and chemistry, and sociology and economics. Some of the above continuities might help support a ‘soft’ objectivist account of the truth conditions of moral statements (see p. 29). Some moral statements are true, and true because of natural features that are independent of anyone’s subjective states and the practices of any group, but these truths require purposive, valuing creatures such as humans.

Bibliography Anscombe, G. E. M. ‘Modern Moral Philosophy’, Philosophy, 33, 1958, 1–19. Aristotle. The Nichomachean Ethics. 2nd Edition. Oxford: Oxford University Press, 1998. Ayer, A. J. Language, Truth, and Logic. 2nd Edition. London: The Camelot Press Ltd., 1946. Baier, A. Postures of the Mind: Essays on Mind and Morals. Minneapolis: University of Minnesota Press, 1985. Beck, L. W. (ed.). The Critique of Practical Reason and Other Writings in Moral Philosophy. Chicago: University of Chicago Press, 1949. Benedict, R. ‘Anthropology and the Abnormal’, The Journal of General Psychology, 10, 1934, 59–82. Brink, D. O. ‘Moral Realism and the Sceptical Arguments from Disagreement and Queerness’, Australasian Journal of Philosophy, 59, 1984, 111–25. Copley, S. and Edgar, A. (eds.). David Hume: Selected Essays. Oxford: Oxford University Press, 1993. Davidson, D. Essays on Actions and Events. 2nd Edition. Oxford: Clarendon Press: 2001. Eagleton, T. The Ideology of the Aesthetic. Oxford: Blackwell, 1990. Flanagan, O. Self Expressions: Mind, Morals, and the Meaning of Life. Oxford: Oxford University Press, 1996. Giddens, A. The Reith Lectures (BBC 1999), Lecture 4, ‘Family’: online, http:// www.bbc.co.uk/radio4/reith1999/lecture4.shtml Hume, D. A Treatise of Human Nature. London: Penguin Books, 1969. Kant, I. Groundwork of the Metaphysics of Morals. Cambridge: Cambridge University Press, 1997.

Metaethics Explored  83

Kant, I. Critique of Practical Reason. Indianapolis: Bobbs-Merrill Co., 1956. MacIntyre, A. After Virtue. 2nd Edition. London: Duckworth, 1985. Mackie, J. L. Ethics: Inventing Right and Wrong. Harmondsworth: Oxford University Press, 1977. Mill, J. S. Utilitarianism. Glasgow: Collins, 1962. Moore, G. E. Principia Ethica. Cambridge: Cambridge University Press, 1903. Plato. The Republic. London: J.M. Dent & Sons Ltd., 1976. Rorty, R. Contingency, Irony, and Solidarity. Cambridge: Cambridge University Press, 1989. Ross, W. D. The Right and the Good. Oxford: Clarendon Press, 1930. Russell, J. S. ‘Moral Realism in Sport’, Journal of the Philosophy of Sport, XXXI(2), 2004, 142–60. Sartre, J-P. Being and Nothingness. London: Methuen & Co. Ltd., 1958. Sayre-McCord, G. ‘The Many Moral Realisms’, Southern Journal of Philosophy (Supplement), XXIV, 1986, 1–22. Selby-Bigge, L. A. Hume’s Enquiries. Oxford: Clarendon Press, 1902. Smith, M. The Moral Problem. Oxford: Blackwell, 1994. Stevenson, C. L. Ethics and Language. New Haven: Yale University Press, 1944. Sturgeon, N. L. ‘Moore on Ethical Naturalism’, Ethics, 113, 2003, 528–56.