Paul asked: Possibly this question is too broad, but what does it mean to say that ethical propositions do not refer in relation to Metaethics. Could you give a concise outline? Answer by Matthew Sims One thing that the claim ethical propositions, such as ‘X is wrong’ do not refer could mean is that the predicate ‘is wrong’ fails denote or pick-out an objective or real quality in the world. Now, in order for this kind of explanation to be of any value at all we must also analyze the notion of objectivity. Let’s for simplicity sake borrow from Michael Dummett’s definition of ‘real’ which crudely put states that something is ontological objective if it is such as to exist independently from our attitudes (our beliefs, thoughts, etc.) and from our theories or methods of verification. When we predicate, say ‘squareness’ to some object, such a property obtaining is not dependent upon our having certain beliefs about ‘squareness’; the quality is such that if some object indeed were square, it’s being so would not be affected by, say, the disappearance of all humans. That just wouldn’t matter. Furthermore, the property of being a shape that has four and only four right angles, could be referred to by some other word besides ‘squareness’ (think of the German ‘viereckig’) and yet this does not prove that such a quality is dependent upon our beliefs or attitudes but only that such an objective property can be referred to by many varying syntactic conventions or symbol constructions. Other properties that are normally not questioned as ‘real’ or ‘objective’ are those which Locke deemed ‘primary’: extension, quantity, shape, motion, and solidity. More controversial are the objectivity of Locke’s secondary qualities: colour, sound, taste, temperature, and smell. So the claim that moral propositions do not refer might be just to say that moral predicates are not made true by objective properties — if truth apt at all. Such a claim immediately could be said to conflict with our intuitions. One may object, ‘It is beyond doubt that the act of burning babies is wrong and anyone who would see such an act would also see the wrongness.’ Is it really the case however that we see the wrongness like we see squareness? One might ask ‘where is the wrongness’ and demand that we point it out to them. In doing so, it seems difficult to discern what the wrongness consists in — particularly if it is to be seen with the eyes or heard with the ears. Is the look upon the offenders face, his ghastly smile or the sound of the screaming infant as it roasts. Would a person born blind never perceive the wrongness of such an act as they never perceive redness? Would they be limited to being informed about the wrongness of such an act? This seems highly unlikely if just not false. The same kind of thought experiment could show that a deaf person (or a person with the bare minimum sensory capacity necessary to detect the occurrence of the action) would nonetheless in this situation ‘sense’ the wrongness. But if this is the case, then wrongness is not seen, heard or sensed in any way similar to other objective properties that we come into contact with. How then is ‘wrongness’ (or rightness for that matter) sensed if not with our sensory organs? Here the moral realist, one who claims that moral properties are objective and that at least some moral propositions are true, faces a problem. If she claims that we have some kind of moral sense organ that is above and beyond our other sensory organs, then seemingly the need to posit such organs for the sake of arguing for the reality of moral properties dulls Ockham’s razor; there are certainly other theories that are available that do not involve positing strange moral sense organs that might explain our attributing moral properties to states of affairs. An epistemological argument very similar in nature was famously put forth by J.L. Mackie as part of his ‘Argument from Queerness’. His argument was in short that in order to become aware of moral properties, if indeed they are like other objective properties, we would need very special sensory organs other than those we do possess. An obvious reply to this is that might run, ‘It was only recently in our history that we’ve come to recognize proprioceptive and kinesthetic functions as kinds of perceptual features that are not tied to particular sensory organs. So why couldn’t there be a moral sensory feature of perception independent of any sensory organ?’ The answer to this is simply that both proprioception and kinesthetics play an independent explanatory role in various theories whereby their existence as perceptual features are justified and coherent with the features provided by particular sensory organs. However the mere hypothesizing of special sensory organs or that sensory organs are unneeded for sensing moral properties to merely justify the reality of moral properties is a move that should be advanced only if there is independent evidence for those objective moral properties other than our intuitions and linguistic habits. Additional to the epistemological tier of Mackie’s Argument from Queerness, is an ontological tier. Given the normative nature of moral claims, if moral properties were objective, then they would have a certain ‘built-in do be pursuedness’ that would necessarily accompany them anytime one would come into contact with such a state of affairs. This kind of moral motivational internalism can be dated back to Plato’s conception of the Form of Goodness. Mackie’s argument runs that it is just not the case that we meet such properties as evidenced by the fact that sometimes people do what they judge to be morally wrong and do not do what they judge to be good. One response to this which has fueled the contemporary debate on the objectivity of moral properties is that the Argument from Queerness as a whole presupposes that moral properties are analogous to primary qualities. This, it is argued by moral realists such as McDowell, can be conceded consistently with holding that moral properties are analogous to secondary properties; properties which are dispositional in nature but which we accept as being objective ontologically. What could such a theory mean by such a claim? In short, dispositional properties — taking the classical example of colour — like ‘redness’ — in order to obtain require appropriately receptive agents and suitable conditions. Where appropriate receptivity can be understood as being physically suited to causally respond to incoming light waves and having the appropriate brain structure (a functioning visual cortex), suitable conditions are often understood as being normal lighting conditions. If these conditions obtain, a dispositional property of some object will cause a response within the agent. This being the case however is not to say that such a property is not objective, but only that it is a kind of conditional property that ‘would cause a reaction’ if the correct conditions obtain. Similarly the realist claims moral properties are like secondary qualities and thus given suitable conditions such as an appropriately receptive agent, such moral properties are features of reality, poised to be recognized. In McDowell’s terminology, such moral properties ‘merit’ certain moral responses if an agent has been raised with certain Aristotelian sensibilities — in a sensitive to the underlying moral structure of reality. As a result of such a theory’s hold that the meriting relation is only partly causal unlike the purely causal relations which hold between non-moral qualities, it avoids having to posit moral sensory organs. Furthermore, such a theory has no problem with the ‘built-in to be pursuedness’ requirement given that dispositional qualities such ‘goodness’ do have such a built-in motivation and if the agent fails to be motivated then it can be said that they fail to be a suitable or appropriately receptive agent and thus are morally degenerate. One thing to ask immediately might be ‘just what is an appropriately receptive agent amount to?’ ‘What differentiates an agent that has an Aristotelian up-bringing from one that doesn’t?’ In answering this, the realist might just again say that such an agent is morally sensitive to certain objective features of reality. But if asked what such features, the realist cannot say that such an agent is sensitive to ‘goodness’ for if she does, her dispositional definition is forced into a circularity; that of defining moral properties as those that agents who are sensitive to moral properties would respond to. And if I am not mistaken, this is just the kind of answer that McDowell gives and thus such an account fails to convince. Now suppose that the realist who does hold that a secondary quality model is the correct account manages to come up with a way of analyzing the dispositionality of moral properties in such a way that avoids circularity. Is there anything else that might dissuade us from accepting that such an analogy is correct? One thing just might be the very phenomenology of moral experience that the realist attempts to use as justification for the intuitiveness of their position. This objection takes the form of a disanalogy; that unlike moral qualities, the objectivity of secondary qualities is supported by the fact that in order to experience them phenomenologically, there is no requirement in terms of having a pre-existing concept of such a quality. Children and even infants after a certain age experience ‘redness’. On the other hand, moral properties seem to require the possession of pre-existing moral concepts before any agent can experience moral properties (or more carefully — begin to attribute moral qualities to states of affairs). Does such a disanalogy entail that moral properties are not real? No! One of the salient points of McDowell’s analysis is that there is a difference between moral and non-moral properties. However, taken together what it does mean is that we must be on our guard for the kind importance we place upon the justificatory power of analogy as a general method for analysis. In other words, one can’t have it both ways — holding that nearness in analogousness justifies something as being objective and at the same time hold that disanalogy doesn’t matter to something’s being justified as objective. If both the secondary quality model and the primary quality model fails, doe this mean that moral predicates do not refer? Well, yes and no. Yes, in that both of these kinds of properties exhaust the kinds of properties that are accepted in physical theories and thus if moral properties are neither then they are not physical properties. No, firstly in that the realist can put up a fight and claim that moral properties are non-physical properties. Such a fight I am convinced is not worth fighting if she cannot additionally provide a plausible theory of how non-physical normative moral properties are casually efficacious upon a physically closed universe. Secondly, the realist rather than claim that moral predicates refer to properties can tell a story about moral objective facts; pointing categorical facts that act as truth makers for moral statements and avoiding outright ontological claims. Whether such a method works is beyond the scope of this answer. However, if you are interested in this approach, one place to look is in R. Joyce’s The Myth of Morality. The title of leaves very little room to guess the position Joyce hold’s on the matter.