‘There are no moral facts.’ Discuss This claim is the rejection of moral realism and cognitivism. The central issue is whether our moral statements are capable of being true or false and of whether we can possess moral knowledge. Those who answer ‘yes’ are moral realists. They maintain that there are moral facts and that our moral statements are capable of being true or false in virtue of the moral facts. That is, the realist claims there are facts concerning the rightness or wrongness of an action or state of affairs. Moral realists are also moral cogntivists. They hold that we can have knowledge of the moral facts. Realists hold that moral facts are objective. What is right is not determined by what I or anybody else thinks is right. I shall explain that there is a set of powerful challenges to realism and cognitivism. These point to the problems the realist faces in explaining the nature of moral facts and our epistemic access to them. Furthermore, the critic argues that realism is unable to account for the motivational force of moral judgements – when one judges that something is, say, good or bad, virtuous or vicious, then one is thereby motivated to act in the appropriate way. In the face of these challenges to realist metaphysics, epistemology and moral psychology, anti-realists such as Hume and Ayer argue that we should regard our moral talk as expressions of attitude or emotion. In examining this expressivist challenge I shall argue that realism can be articulated in a way that defuses the force of the challenges while preserving the core motivations for realism. 1 The immediate metaphysical worry is to explain what kind of fact a moral one is. If we take a fact to be state of affairs – how things are – then moral facts are those which possess some feature or property in virtue of which they have an evaluative or normative element. Realism has two options. The first realist option is naturalism. This analyses moral properties and facts as natural facts or combinations of natural facts. That is, as fully analysable as features of the world that can be specified in terms of the natural sciences. The goodness of an act or state of affairs (say helping someone cross the road) is identical to or reducible to some (combination of) natural fact – say the promotion of well-being or happiness. One way of spelling out naturalism is utilitarianism. Here the moral property of goodness is identified with the natural property of happiness. However, any attempt to identify moral properties with natural ones is ruled out by the second realist option, non-naturalism. A moral property such as goodness is not part of or analysable in terms of the objects and properties picked out by (what we today call) the natural and social sciences. The goodness of a person or the rightness of an action is a simple sui generis property which belongs intrinsically to the thing in question. A highly influential version of non-naturalism is developed in the work of G.E. Moore known as Intuitionism. Moore believed that much of the thinking about the nature of moral judgements rested on what he called ‘the naturalistic fallacy’. Simply put the ‘fallacy’ is committed in the identification of the simple, non-natural property of goodness with some natural property – the sort of property natural sciences and 2 psychology refer to. According to Moore most philosophers have conflated the property of goodness with the things that possess it or with some other property(s) that good things have. However, Moore explains that we cannot define ‘good’ in more basic terms because it is a simple and unanalysable property. Drawing an analogy with our knowledge of colours whereby you know what ‘yellow’ is through direct acquaintance with yellow objects, Moore proposes that one knows a good action when one encounters it. The key element in Moore’s criticism of naturalism is his Open Question Argument (‘OQA’). Suppose, Moore says, goodness were identical with some other property. Let us say that goodness is identical with the promotion of happiness (his own example is what we desire to desire). Now ‘good’ and ‘happiness’ are synonymous and every competent speaker would therefore know the following naturalistic identity claim: NI goodness is the promotion of happiness It now follows that to ask ‘is that which promotes happiness good?’ demonstrates a lack of understanding or sense on the part of the enquirer. It is not a real or live question. For, given our definition of good, it is just the same as asking ‘is that which is good, good?’ However, Moore explains that it is always an open question to ask of some act or state whether it is good or not. To ask if promoting happiness is good is always a live or open enquiry. So, good does not just mean the promotion of happiness (or whatever natural property(s)) and NI is false. Moral realism is therefore committed to an understanding of moral facts as non-natural. 3 What then is the relationship between the natural and non-natural properties or facts? Here intuitionism appeals to the idea that while the moral facts are not identical to or analysable in terms of the natural facts, there is a form of dependence on the non-moral facts. In two situations which are exactly alike with respect to the natural facts, there can be no difference in moral terms. No two things (e.g. persons, acts, states of affairs) can differ in evaluative terms without also differing in their nonevaluative properties. This relationship is known as supervenience. How are we to account for moral knowledge if moral facts are non-natural? Moore holds that when we are presented with something good our judgement of value is self-evident. This appeal to self-evidence may not strike you as compelling. What if A and B sincerely disagree over what is self-evident? Moore uses the term ‘intuition’ to refer to our direct awareness of goodness. A key objection to non-naturalism is that it cannot adequately explain the way in which we come to have knowledge of the moral facts. To hold that the self-evidence of the goodness of some state of affairs is self-evident in a way analogous to our perception of colour is not an explanation of how we can grasp or intuit some non-natural property. In the end it looks as if we have to appeal to our possession of special moral faculty - and that looks mysterious in the sense that it stands apart from our overall picture of the world as describable in terms of the natural sciences. Realism thus faces a challenge grounded on the strangeness of moral facts and how we could know about them. Perhaps, though, an even more serious set of criticisms is developed by Hume. The problem is not one of spelling out what a moral fact is, 4 but of having any reason to judge there to be any coherent explanatory need for such facts. Let me explain. Hume argued that morality is grounded in sentiment and not reason. When we investigate our moral judgements of something being right or wrong we can never locate some feature of the world – a matter of fact – which we can call vice or virtue, right or wrong. Instead, we can locate in our own psychology feelings of approval or disapproval. There is no moral fact in virtue of which we are forming a judgement capable of being true or false, Rather, we have (complex) sentiments or emotions which direct our judgements and actions about the world. Alongside this analysis Hume makes the logical point that we cannot derive an ought from an is. His point is that realists make a critical error in deducing from a description of the world a moral judgement. Logic is conservative. You can only get out what you put in. In describing the world we talk of what there is. In drawing moral conclusions we talk of what ought to be the case. Yet, as matter of logic a conclusion containing ‘ought’ cannot be drawn from a set of ‘ought-free’ premises’ This has been taken by many to mean that there is an unbridgeable gulf between a description of the world and talk of what we ought to do. As such it has been offered as a decisive reason to block any attempt to analyse moral language in terms of the descriptive language we apply to the world. A third argument takes us into the territory of moral psychology. A widely held thesis is that in forming a moral judgement one is motivated to act appropriately. This is internalism in theories of moral motivation. Hume notes that when we make a moral judgement we are motivated to act, but motivations are never beliefs. 5 Beliefs are states which represent the world; they aim to tell us how things are, to report the facts. As such beliefs and the facts they report are ‘inert’. They do not move or motivate us to do anything. The motivation to act is supplied by desires. To have a desire is to have a mental state that tells you how the world should be. To desire a drink is to be in state directing a (small) change in the world: the world should move from one in which I lack a drink to one in which I have one. Moral judgements cannot be beliefs about moral facts. For, according to Hume’s picture of human psychology beliefs are not motivating. A moral response is a commitment to action. It is in this way intrinsically motivating. A moral judgement must therefore really belong to a different class of mental states – sentiment or emotion or desire. Under the influence of Logical Positivism non-cognitivism finds expression in emotivist or expressivist theories of the twentieth century. For example, A.J. Ayer argued that our moral talk should not be understood as aiming to express truths about right or wrong notwithstanding its propositional form (i.e. when we talk morals we do so in declarative sentences which appear capable of being judged true or false). Instead, moral discourse should be understood as an expression of attitude or emotion. The purpose of a moral judgement is not to express a truth that is somehow out there in the world, but to express my attitude with respect to some act or state of affairs and to influence those of others. 6 The key motivations for emotivism are (a) agreement with Moore that moral facts could not be natural facts, and (b) the Logical Positivist theory of meaning – the Verification Principle. The Open Question Argument rules out the possibility that moral facts could be analysed in terms of natural facts and so be discoverable through experience. This leaves the realist having to maintain that moral facts are non-natural. However, this is ruled out by the Verification Principle. The Verification Principle states that a proposition can be verified (demonstrated to be true or false) either if it is true simply in virtue of the meanings of the term/symbols used (definitions, tautologies) or against observation statements. Moral statements are not analytically true. It is not a mere matter of definition or tautology that cruelty is wrong or courage is a virtue. Furthermore, nor can such moral statements be verified by sense-experience if Intuitionism is true. Intuitionism is the view that moral properties are simple, non-natural properties. That means that the moral properties themselves cannot be described in the empirical terms of the natural or social sciences. There is no empirical evidence which could verify the truth of a statement about a moral property. Moral statements are thus unverifiable and so literally meaningless. If those claims are correct then realism is indeed impaled on the horns of a dilemma. Realism must either describe naturalistic states or non-natural states. Yet, according to Ayer both views are untenable. So, we should reject the realism and its commitment to the view that moral discourse is (as its grammatical form suggests) descriptive and truth-evaluable. Our moral judgements must be expressions of 7 feeling or attitude. To hold that something is right is a sophisticated way of saying ‘hooray!’ I shall finish by suggesting how realism can maintain internalism about moral motivation while defusing the ant-realist criticism. First, non-cogntivosm in its expressivist cannot account for the role of moral judgements in moral reasoning. Consider the following claim. (1) Murder is wrong. Now, (1) has the grammatical form of an assertion and as such it can be evaluated in terms of truth or falsity. In the context of a declarative statement, the expressivist analyses this assertion as an expression of feeling. However, the proposition ‘murder is wrong’ can appear in contexts in which it does not have the role of an assertion. Consider the conditional statement: (2) If murder is wrong, then getting your brother to murder is wrong. In the antecedent (the if-clause) of (2) ‘murder is wrong’ is not being used to make an assertion. In (2) the speaker is not stating that murder is wrong. Instead the speaker is presenting a hypothesis about what follows if it is wrong. In this unasserted context ‘murder is wrong’ it is not used to express disapproval of murder. The Expressivist must allow that we know what is meant by both (1) and (2). Yet, Expressivism must also hold that the semantic function (meaning) in (2) of ‘murder is 8 wrong’ must be different from that given for the apparently straightforward assertion expressed by ‘murder is wrong’ in (1). Now there is a problem in explaining the apparently valid inference: (1) Murder is wrong. (2) If Murder is wrong, then getting your brother to murder people is wrong. Therefore: (3) Getting your brother to murder people is wrong. If the semantic function of ‘murder is wrong’ differs in (1) and (2), then surely one is not entitled to arrive at (3). For the argument to be valid the meaning of ‘murder is wrong’ must be the same in each of (1) and (2). The above argument from (1) and (2) to (3) is apparently no more valid than: (4) My beer has a head on it. (5) If something has a head on it, then it must have eyes and ears. Therefore: (6) My beer must have eyes and ears. This argument is invalid. We are only tempted to think the conclusion follows if we are insensitive to the reliance on the equivocation between the two senses of ‘head’, in (4) and (5) respectively. 9 This suggests that realism may be essential to a coherent and common-sensical account of moral reasoning. It leaves unresolved the unease we face in explaining what a moral fact is. Let us turn to Philosophers such John McDowell and David Wiggins who have (to put things roughly) articulated a version of realism which draws on a parallel between secondary properties and moral properties. Very roughly one may have this line of thought. In the case of colours we do not have to suppose that a yellow object is yellow independently of human response. Equally, the experience of the observer is not just down to her. Seeing the feathers of a canary as yellow is not a matter of personal opinion or taste or decision or practice. It depends on the physical microstructure of the bird in virtue of which its feathers possess a certain reflective quality. Our experience of the bird also depends on the way in which we are constituted so that we respond in certain ways to features of the world. The yellowness of the canary is objective in the sense that it is independent of any singular, particular human response. The yellowness of the bird can only be experienced in its appearance to us, but there is a sense in which the colour can persist unseen. Of course, that sense is not to be cashed out in terms of the experience of perceiving yellowness. Rather the yellowness exists unperceived in the sense that the surface structure of the bird’s feathers possess the dispositional power to produce a particular kind of experience in a certain class of observers. It is, so to speak, waiting for us to notice it. In this respect colour does not exist independently of human responsiveness. 10 When it comes to moral judgements we are disposed to respond to certain features of the world in a particular way. The burning of the dog by a gang of boys elicits in us the judgement – the recognition – of its wrongness. That wrongness is not a wholly independent feature of the world since the wrongness is essentially to do with judgement by us (just as colour is essentially to do with appearance to us). The wrongness of the act is manifest in our judgement that the state of affairs before us furnishes us with a reason to act. If an account of moral properties can be offered to explain the sense in which they have the objectivity and response-dependence of secondary properties, then one may have an account that is sufficiently realist. For then our language is referring to the moral facts understood as reasons for rational creatures like us to act and as reasons they possess the normativity inherent in moral judgement and externally grounded in the moral facts. 11