[Phronesis] is of the ultimate particular, of which there is not scientific knowledge but perception—not sensory perception, but like the perception whereby we perceive that the triangle is the ultimate particular in geometry.
Aristotle, Nicomachean Ethics 1142a27-29Knowing is seeing.
John Locke, Of the Conduct of the Understanding, s.241. Ontological Minimalism
The problem with moral realism is widely supposed to be its ontological extravagance. So suppose we begin with maximal ontological parsimony: say, with a metaphysics like the one expressed in Democritus' ‘nothing but atoms and the void’.Footnote 1 Or even, since Democritus apparently allowed there to be infinitely many atoms, suppose we begin with something more minimal still—the hyper-Democritean thesis that nothing exists except the void, and in it, the following sixteen-dot matrix:
What ontology could be more minimal (except, presumably, one with fewer dots)? But even this hyper-Democritean ontological minimalism sustains an indefinite multiplicity and complexity of patterns. If these sixteen dots exist thus arranged, then it is immediately not the case that only these sixteen dots exist. There also exists every pattern that these sixteen dots constitute. After all, a capacity which perceived the sixteen-dot matrix as four four-dot squares would be picking up a pattern that is genuinely there in the matrix, whether you notice it or not. So would a capacity that perceived it as a four-dot square inside a twelve-dot square; and a capacity that perceived it as a twelve-dot cross against a background sketched in by the four corner-dots; and one that perceived it as three vertical or three horizontal corridors… and so on. There is no limit to the patterns that we can find in the diagram—beyond those set by our ingenuity, imagination, or mathematical/geometrical aptitude. And, to say it again, find, not create, seems to be the right word: the patterns would still be there even if we didn't see them. In the words of Daniel Dennett: ‘Other creatures with different sense organs, or different interests, might readily perceive patterns that were imperceptible to us. The patterns would be there all along, just imperceptible to us.’Footnote 2 (Though he claims to be uninterested in the labels ‘realist’ and ‘instrumentalist’, Dennett has no qualms about talking of ‘the brute existence of pattern’ (op. cit. p. 51), and of patterns as ‘existing’ and ‘real’—which all sounds like realist, not instrumentalist, talk to me.)
Now, plausibly, the world is a good deal more ontologically complex—even on the simplest and most reductive reading of what counts as ‘the world’—than a sixteen-dot matrix. (Even on Democritus' reductive account it is an infinitely-many-dot matrix.) A fortiori, then, the world contains far more patterns than those in the sixteen-dot matrix. And these patterns are really there in the world: or at any rate, they are just as much really there in the world as the dot-patterns listed in the last paragraph are really there in the sixteen-dot matrix.
The first moral of this little tale is that it's not as easy as it looks to be ontologically parsimonious. No doubt we could say, if we chose, that all these perceived patterns in the sixteen-dot matrix are just in the head, or are just interpretations of that matrix, so that the matrix is all that really exists. Such claims may keep our ontology nice and sparse, but they look unattractive in almost all other respects. To cite just the most obvious difficulty: since mathematics seems to be largely the study of the logical behaviour of patterns like these, a parsimonious attitude to the ontology of such things as dot-matrices apparently commits us to saying that mathematics is all in the head. Whatever there might be to be said for fictionalism about mathematics and allied positions (and that is not my topic here, so I will not pursue it), this is not, intuitively, a welcome consequence. It's not hard to see why we might have more sympathy with the more ontologically generous attitude described above, and stick with the intuitively appealing affirmation that the patterns really are there no matter whether we see them.
So much for the first moral. The little tale has other morals too. Two in particular: patterns are properties of such matrices; and the recognition of such patterns is a kind of perception. I take these points in turn in sections 2–3.
2. Patterns and Properties
Perhaps not all properties are patterns. Perhaps not all patterns are properties. But certainly some patterns are properties: for instance, as already pointed out, the patterns in the matrices described in section 1 are properties of those matrices. At least we may say this: the marks of a pattern and the marks of a property are, when one thinks them through, remarkably similar. Consider eight marks.
(1) Intensionality: There are the same sort of reasons for treating both patterns and properties as intensional entities. Just as the same extension (the man Benjamin Franklin) can coincide with more than one intension (the inventor of bifocals, the first US postmaster-general), so the same ‘instantiation base’ can simultaneously support more than one pattern, and more than one property. (The sixteen-dot matrix simultaneously supports the pattern of four four-dot squares, and the pattern of a twelve-dot cross with four dots for corners.)
(2) Abundance I (instantiation): We have parallel reasons for saying that as many patterns as can be seen in a given state of affairs are actually present in that state of affairs, and for saying the same about properties.
(3) Abundance II (the modality principle): Some people say that any property exists if it is metaphysically possible for it to be instantiated. Just the same claim might be made about patterns, and the methods of rejecting such a claim, if we wanted to, would be very like the methods of rejecting the modality principle for properties.
(4) Uninstantiated existence: The questions whether properties exist if they have no instances, and whether patterns exist without instances, clearly turn together. An intuitive answer to the question about patterns is that a pattern can be real whether or not it has exemplars. If we accept that, we should accept the corresponding answer to the question about properties.
(5) Spatiotemporality: Properties can (in principle) appear anywhere and at any time, given a suitable instantiation base; so can patterns. One and the same property can be in multiple places at multiple times; so can one and the same pattern.
(6) The ‘one over many’ argument: Often when there is a positive answer to the question of what a number of things have in common, the natural answer is that those things all exemplify the same pattern, just as it is often that those things all exemplify the same property. To the alternative (‘nominalist’) replies that those things show patterns or properties which are merely similar to each other, and not actually the same, there are closely parallel objections (what makes them similar, if not their exemplification of the same properties/ patterns?).
(7) The basicity of instantiation: The concept of the instantiation of properties does not seem to admit of much further substantive analysis. The same seems true of the concept of the instantiation of patterns.
(8) Substrates: For all that, there is one thing we can say: for a property to be instantiated is necessarily for it to be instantiated in something. The same again is true of patterns.
At the very least, the moral is that patterns give us a useful model of how some (many? most?) properties work. Some may accept a stronger moral, that properties are the very same thing as patterns. Some putative examples of properties look recalcitrant for this thesis. For instance, if spatiotemporal position, or orientation, or the charges of electrons or the flavours of quarks, are to be understood as properties, it may not seem very obvious how instances of these properties can also be instances of patterns in any sense except a weak one. But I do not need this stronger thesis; it is enough for my purposes if patterns are at least a model for interestingly many properties. I turn now to pattern-recognition and perception.
3. Patterns and Perception
Perhaps not all perception is pattern-recognition. Perhaps not all pattern-recognition is perception. But certainly some pattern-recognition is perception. Here are four examples.
(1) Seeing a face. To see something as a face is to represent it perceptually as falling under a pattern that many other things can or do exemplify: the pattern of the human visage. Developmental psychology suggests that face-recognition is ‘hard-wired’ in humans: new-born infants have a capacity to lock on to examples of this pattern as soon as they are born. Certainly the representational perception of faces, and of the subtle and complex ways in which they display their owners' thoughts and feelings, is an absolutely standard part of the adult human's perceptual repertoire. Faces are things that we ‘just see’, simply and directly and without reflection or inference. We can barely imagine ourselves into the position of someone (an alien, perhaps) who has to work out that a given array is a human face, or what expression it wears. To us, these facts are given by immediate observation.
(2) Seeing a chair. To see something as a chair is to represent it perceptually as falling under a pattern that many other things can or do fall under: the property of being a particular kind of artefact, designed for sitting on, that is found in most human societies. For those of us who have been acculturated into any chair-using society, there is rarely any question of needing to work out that something is a chair. Exceptions are of course possible, as when we are confronted by a radical-modernist design of chair, or as when we need to decide whether something can be used as a chair. As any visitor to stately homes knows, some things that are undeniably chairs cannot be so used. Such experiences are entertaining, and enlightening, because they remind us what it would be like to be from a non-chair-using society. It would mean that we were put in a position where we could not just see things as chairs, and had to decide what counted as a chair by the very different method of inference.
(3) Seeing four cows. To see something as a group of four (four anythings—cows will do) is to represent it perceptually as falling under a pattern that many other things can or do fall under: the property of being a four-numbered collection of things. For those of us who are fully numerate, and where the cows are symmetrically arranged thus ::, it will normally be possible for us to see that there are four cows in the group by direct observation. We will not need to count the cows; we will see in one look that there are four of them. Contrast what a typical five-year-old will do: she will work out that there are four cows not by direct observation, but by pointing and counting up to four. Contrast also how most adults would deal with a group of thirty-eight cows, which would be by the toddler's method with the four cows: in a word, by inference, not perception. (‘Most adults’: there are a gifted few who can just see that a group of cows are thirty-eight, or do similar feats for even larger numbers.Footnote 3)
(4) Reading. If you are a literate Anglophone, then when you see the word BREAD you will take it in in one look, and know immediately what it says. You will not need to spell it out letter by letter, and work out what it says by inference from its elements plus principles of combination. You will not be so favourably placed when you are confronted by the array Xлeб, unless you know enough Russian to know that this is the word khlyeb, the Russian for ‘bread’. Similarly, you will struggle to read the phrase mikä su ongelmas sitte o? unless you are sufficiently fluent in Finnish to know that this means ‘What's your problem, pal?’.
In these cases, unpractised readers are likely to have to do something more laborious than just seeing what these inscriptions say. They will have to work them out, one step at a time. They will not read them, because they will not ‘directly’ perceive what is written: they will infer it from other things that they do perceive. For to spell out the word BREAD, letter by letter, and infer from what its letters are plus knowledge of the rules of phonetics that ‘bread’ is what the inscription says, is not to read it.
Dennett (op. cit. p. 34) adds a fifth example, chess: ‘Expert chess players can instantly perceive (and subsequently recall with high accuracy) the total board position in a real game, but are much worse at recall if the same pieces are randomly placed on the board, even though to a novice both boards are equally hard to recall… expert chess players, unlike novices, not only know how to play chess; they know how to read chess—how to see the patterns at a glance.’
To see the cows as four, the chair as a chair, the face as a face, or the inscription Xлeб as the word khlyeb (as ‘bread’-in-Russian, if you are fluent enough in that language), or to ‘read’ the chess-board accurately; these are all cases where competent observers directly pick up a pattern that is present in the relevant perceptual field. In these cases the relevant perceptual field is the visual field, but that is of no essential importance; similar phenomena can occur in all the different sensory modalities. There are also patterns that we pick up through more than one sensory modality. Think of the multi-sensory ways in which you might perceive that you have been insulted at a party, or that the England cricket team have just regained the Ashes.
In cases like the five just described, perception is pattern-recognition in that perception is the capture of a Gestalt. No doubt not everything that gets called perception is pattern-recognition of this sort; after all, we can distinguish mere seeing from seeing-as. But a lot of perception is like this.
4. Perception and Inference
This sort of perception can be contrasted with what I am calling inference. Inference is active, perception is passive. In inference, I set my mind to work something out, whereas in perception, something ‘just comes to me’—I am subject to an occurrence that I do not make happen, except in the minimal sense in which I can, e.g., choose in which direction I look. Inference is experienced as structured, perception as simple: to perform an inference is, normally, to run through a number of steps of reasoning, whereas perceiving something is a step-less, instantaneous whole. So perception is quick where inference is slow: perceiving something can happen instantaneously, inferring something normally takes time. (Dennett on chess again (op. cit. p. 42): ‘the scale of compression when one adopts the intentional stanceFootnote 4 towards the two-dimensional chess-playing computer galaxy is stupendous: it is the difference between figuring out in your head what white's most likely (best) move is [and] calculating the state of a few trillion pixels through a few hundred thousand generations.’)
Furthermore, a series of inferential moves may be available that explain why I have a given perception. For example, going back to the sixteen-dot matrix, it is the configuration of the dots that explains why I perceive four four-dot squares, or one twelve-dot cross, if I do. However—and this is clearly an important contrast—there is more phenomenologically speaking to either perception than there is to the inference. Inference is, so to speak, topic-neutral. There is nothing in particular ‘that it is like’ to work out why I see a twelve-dot cross in the sixteen-dot matrix. By contrast, there is something ‘that it is like’ to see that cross: seeing the cross in the matrix does have a particular experiential feel.
Perhaps this point generalises to other cases where we might contrast inference and perception. If so then perhaps, for instance, the colour scientist can make out her calculations about green without having any specific sort of experience, but will perceive green only if she has one very particular kind of experience.Footnote 5 In that case, to say that we perceive green will be to say that we grasp instances of this motif, and grasp them as instances of the motif. It will also be to say that the grasping happens by way of direct apprehension in a particular sort of experience (namely, seeing green), rather than indirectly, by way of inference or step-by-step working it out. A colour scientist, if she was sufficiently competent and had enough data about the reflectance of nearby surfaces and other relevant light conditions, could calculate whether something is or isn't an instance of green. Such calculations could give her the knowledge that that thing is green. But they would not enable her to see that it is green: for they would not give her any awareness of the green that comes to her directly, in her experience.
Perception, in this sense, is a quick, analogue, route to many of the same epistemic destinations as can also be reached via the digital route of inference. It gets us there with a useful speed, directness, ‘givenness’ (cp. the activity/passivity contrast just noted), and with a vividness which grounds, and in favourable cases justifies, a sense of certainty. Evolutionarily speaking, it is not difficult to see the developmental advantages of perceptual capacities like these.
What is notable for our purposes is the ontological conservatism of such capacities. As I have argued, we can suppose that such capacities do no more than pick up patterns in some instantiation-base. What that instantiation-base might be is in no way constrained by the notion of perception as a pattern-detecting capacity. It can be anything, provided it is at least generous enough to sustain the instantiation of whatever sort of patterns we are concerned with. But as we have seen, that level of ontological generosity is surprisingly easy to come by.
For all we can tell, then, the instantiation-base of some remarkably complex and recherché types of patterns or properties might be remarkably metaphysically minimal. It might be like this, for instance, with the patterns that form moral properties.
5. Perceiving Moral Properties
David McNaughton writes:
We might suppose that the only properties that can be observed are the ‘proper objects’ of the five senses: touch, shape, and texture; hearing, sound, and so on. If we adopt this austere account of what can be perceived it is clear that not only moral properties but a great many of the things we normally take ourselves to perceive will be, strictly speaking, unobservable. If, on the other hand, we are prepared to allow that I can see that this cliff is dangerous, that Smith is worried, or that one thing is further away than another, then there seems no reason to be squeamish about letting in moral observation.Footnote 6
The upshot of my argument so far is that we should be prepared to allow this. There is no profit in restricting perception to sensation in the way that McNaughton describes. As he notes, this restriction on perception excludes not only moral perception, but many other sorts of perception in the wide sense of pattern-recognition or Gestalt-uptake (representational perception, as we may call it). Nor, I have argued, is that restriction well-motivated by a concern for ontological modesty. If at least some properties are patterns, and at least some perception is pattern-recognition, then there can be perception of properties which has a very modest ontological base indeed: maybe even as modest as my sixteen-dot matrix. Since such perception is so modestly based, it does not face the worry about ontological extravagance. If moral perception is one example of this sort of perception, then there is no good reason why moral perception should face the ontological-extravagance worry any more than any other example of the kind.
But this picture of moral epistemology and metaphysics is not just a bare possibility for the realist. It is an attractive possibility, because it gives him good answers to some of the most familiar problems that he faces. Here, in Mackie's words, are two of the best-known:
If there were objective values, then they would be entities or qualities or relations of a very strange sort, utterly different from anything else in the universe. Correspondingly, if we were aware of them, it would have to be by some special faculty of moral perception or intuition, utterly different from our ordinary ways of knowing everything else.Footnote 7
Mackie sets the realist a puzzle about moral metaphysics: what could moral properties possibly be? And he sets the realist a puzzle about moral epistemology: how we could possibly know about moral properties? On the picture of properties and perception that I have just sketched, neither of these puzzles is very puzzling at all. Moral properties can be just what many other properties are: patterns in reality. And our knowledge of moral properties can be representational perception, in just the sense that, in McNaughton's example, seeing that a cliff is dangerous can be representational perception.
If we ask where moral properties are supposed to fit into a scientific world-picture, the answer is that there is nothing in a scientific world-picture to exclude the possibility that the world's constituents should be reasonably perceived as standing in patterns that are not themselves used in scientific explanations. Likewise, if we ask how moral properties relate to the properties that science does use in its explanations, the answer is simply that moral properties and scientific properties are two different types of pattern that are both found in the same world. Of course, there will be no perceiving these patterns without an evolutionary and cultural history that is suitable for the emergence of the requisite perceptual capacities. But given obvious and familiar facts about our own culture, which is easier to say: that we cannot conceive how these perceptual capacities might possibly have developed? Or that just such perceptual capacities apparently have developed—so obviously can have developed?
6. The Supervenience Argument Against Moral Realism
Another notorious difficulty for moral realism is raised by Simon Blackburn:
[I]f A has some naturalistic properties, and is also good, but its goodness is a distinct further fact not following [logically] from its naturalistic features, and if B has those features as well, then it follows that B also is good. And this is a puzzle for the realist, because there is no reason at all, on his theory, why this should follow. If the goodness is, as it were, an ex gratia payment to A, one to which A is not as a matter of logic entitled in virtue of being as it is in all naturalistic respects, then it should be consistent to suppose that although goodness was given to A, it was not given to B, which merely shares the naturalistic features that do not entail the goodness.Footnote 8
Like most moral theorists, the moral realist wants to say that, if A and B share the set N of naturalistic properties, then they must also share the set M of moral properties. But the realist faces the question: why must? The realist cannot say that a minimally rational observer's knowledge of the presence of the naturalistic properties N rationally commits her to knowledge of the presence of the moral properties M. As Blackburn rightly insists, a failure to see that N brings M with it is not a failure in logical competence. So why couldn't it be that A and B both had N, but only A had M? Blackburn thinks that the realist can only leave this mysterious, whereas Blackburn's own projectivist/expressivist theory solves the mystery very neatly. According to the expressivist, the reason why, if A and B share N, then they must share M, is that we are rationally required to be consistent in our attitudes. Hence the contents of the set M are not properties at all, at least not in the same way that the contents of N are properties. They are projections of our evaluations onto the world—projections which are made ‘universalisably’ because they are made in compliance with the rational requirement of consistency.
How might a moral realist meet Blackburn's challenge to explain the puzzling must without recourse to projectivism? The claim that, if A and B share the set N of naturalistic properties, then they must also share the set M of moral properties, comes out on the present line of thought as a claim about the necessary relations of patterns. The claim will take the form: ‘If A and B both exemplify one pattern NP, then they must both exemplify another pattern MP’. And we can explain this sort of ‘must’ without any difficulty at all. We cash it out by an explanation parallel to the one in which we explain how, if anything is a sixteen-dot matrix like the one displayed above, then it must be possible to see a cross pattern in it, and a four-square pattern, and a three-corridor pattern (vertical or horizontal) etc.; or how, if anything is a duck-rabbit, then it must be possible to see a duck in it; and also a rabbit; etc.
‘But this response makes the relation of naturalistic and moral properties a matter of entailment.’ Yes and no. Yes, because the relationship is indeed entailment, or something very like it, at the level of the analysis of the patterns. The duck-rabbit pattern contains the duck pattern and the rabbit pattern; the sixteen-dot matrix contains the cross and four-square patterns; and so on—and containment is of course a familiar and fruitful way of understanding at least mostFootnote 9 cases of entailment. No, because the relationship is nothing like entailment at the level of pattern-recognition. It is perfectly possible to perceive the duck-rabbit without perceiving the duck, or the rabbit, or both, and to perceive the sixteen-dot matrix without perceiving the cross, or the four-square pattern, or both. Moreover, you cannot argue anyone into seeing any pattern. All you can do is give them nudges and hints, ‘assemble reminders’,Footnote 10 prompt them with encouraging prods like ‘Do you get the drift?’Footnote 11 or ‘Can you see what it is yet?’ (as Rolf Harris would say)—and hope that they will catch on, and start to experience the relevant sort of Gestalt in their perceptions.Footnote 12
The point is the same as before: moral inference and moral perception go their different ways to the same destination, moral knowledge. The present application of the point is this: working out a complete moral theory would mean so thoroughly explaining the relation that holds between naturalistic and moral properties that that relation would, in the theory, become one of (very complicated and long-winded) entailment. This would disprove Gilbert Harman's claim that, ‘conceived as an explanatory theory, morality, unlike science, seems to be cut off from observation’.Footnote 13 For it would give us a moral theory that explained both why there is moral perception, and also how moral perception can be correct.
However, moral theory is one thing, and moral perception is another. Since moral perception's ambition is not to turn into moral theory, but to get more accurate as perception, there is no reason to say that someone who does not grasp these inferential relations is cognitively defective. Lack of information, like not knowing what time the bus comes, may be a cognitive limitation, but we do not call it a cognitive defect, like not being able to understand a bus timetable even in favourable conditions. Since, on the other hand, the relation between the naturalistic and the moral is something that can in principle be explained by the sort of realistic moral theory that I am developing here, there is no (in-principle) mystery about why the perceptions of certain naturalistic and moral patterns go together. Thus the realist can—in principle—explain what Blackburn says he cannot about the supervenience of the moral on the naturalistic.
7. Moral Realism and Motivation
J.L. Mackie's famous puzzle about the ‘queerness’ of moral properties can be understood differently from the two ways, metaphysical and epistemological, that I took it in section 5. Understood this third way, Mackie is posing a further puzzle, often set for the realist by other authors too, about how any property could be intrinsically motivating. His complaint is that an objectively existing moral property would have to be something that combined existence with ‘intrinsic to-be-pursuedness’; while a perception of such a moral property would have to combine representation with motivation. Mackie does not see how anything could combine these two features.
Plato's Forms give a dramatic picture of what objective values would have to be… [As with a Form], an objective good would be sought by anyone who was acquainted with it, not because of any contingent fact that this person, or every person, is so constituted that he desires this end, but just because the end has to-be-pursuedness somehow built into it. (Mackie op. cit. p. 40)
The moral realist might begin his response to this with an evolutionary point. In evolutionary terms, what is hard to explain is not the representation that motivates, but the representation that does not motivate. As a matter of the history of our species, the (original) point of perceptual capacities in a tough world must usually have been to mandate response rather than to get hold of information for its own sake. Think this way of echolocation: the bat gets hold of a pattern which is the shape of a wall in front of it, and that pattern mandates a response which is an alteration of flight-path. The general schema for such motivating representations will be: Pattern P in context C mandates response R from X. It is not obvious why this schema cannot apply to moral perception in humans just as much as to echolocation in bats.
Further, we can observe some distinctions. First of all, there is a difference between the notion of an intrinsically motivating property (or perception—I won't keep adding this), and the notion of an intrinsically reason-giving property. Mackie's phrase ‘intrinsic to-be-pursuedness’ seems ambiguous between these notions. Is Mackie puzzled about the idea of a property that is intrinsically such as to cause motivation in its perceivers? Or is he puzzled about the idea of a property that is intrinsically such as to give reasons to its perceivers? Or, a different possibility, is he puzzled about the idea of a property that intrinsically motivates because it intrinsically gives reasons to its perceivers? And a further question: how should we take ‘intrinsically’ here? Is an ‘intrinsically to-be-pursued’ property one that, of itself, must cause motivations or give reasons (strong intrinsicness, if you like)? Or is it one that, in and of itself, can cause motivations or give reasons (weak intrinsicness)?Footnote 14
All of these are different possibilities, and it is not clear to me that any of them is bizarre or queer. Agonising pain may well be strongly intrinsically motivating: that is, if you feel it, then you will necessarily be motivated to try and get away from it. Mild pain, or pleasure of any strength, may be weakly intrinsically motivating: if you feel it, then you can be motivated to avoid it (mild pain) or seek it out (pleasure). Many properties that aren't in any obvious way moral properties are widely thought to be intrinsically reason-giving, either strongly or weakly. If you encounter a case of cogent argument, for example, then its cogency necessarily gives you a reason to believe its conclusion (as some would say, going for strong intrinsic reason-giving); or at any rate can give you a reason to believe that conclusion (as others would say, preferring weak intrinsic reason-giving). If there is queerness in any of this, it is hard to see where. Or does the queerness reside in the possibility that some properties might (sometimes) intrinsically motivate because they intrinsically give reasons to their perceivers? But to find that queer, it seems to me, is to boggle at a truism about rationality: that rational agents do at least some of the things that they do because they believe they have reason to do those things. This, of course, is a case of motivation by a belief, not a desire, and that is anathema to orthodox Humeanism. But this sort of motivation seems so well-established and so familiar a phenomenon that I have little to say about that except ‘So much the worse for orthodox Humeanism’.
My discussion of the argument from queerness read in this last way, as posing a problem about intrinsic to-be-pursuedness, produces some interesting provisional conclusions about externalism and internalism about motivation. In particular, it shows how someone could be an internalist about reason-givingness, but an externalist about motivating force. That is to say, he could hold that there are moral properties that are intrinsically reason-giving (in that those moral properties of themselves must or can, always or sometimes, give reasons), without holding that there are moral properties that are intrinsically motivating (in that those moral properties of themselves must or can, always or sometimes, move agents to action). This is my own position.
In sections 4, 5, and 6 I have showed how a moral realism derived from a parsimonious ontology could have the resources to deal with three standard objections to realism. In section 7 I have dealt with a fourth standard objection to realism, though the discussion has, necessarily, ranged rather more widely, and has not been related only to the specific form of moral realism that I have been advocating. The last section, section 8, brings that specific form of realism back into focus. Having argued that these four problems are not the real problem about moral realism, I shall briefly point out what is.
8. Realism and Relativism
It is evident from my discussion of matrices like the sixteen-dot one of section 1 that I admit the reality of all sorts of patterns. This, together with the claim that properties including moral properties are patterns, gets me the result that moral properties are no less real than all the other patterns that we might perceive in reality. Unfortunately it also gets me the result that indefinitely many other sorts of other patterns are there in reality too. Moral relativism will be back with a vengeance if I can't show that there is something special about the patterns, among all these other patterns, that are particularly formed by moral properties. And relativism might seem to be getting back in anyway, since we can hardly avoid admitting the claim that moral-perceptual capacities develop with culture: isn't it obvious that different cultures develop (at least some) different capacities, that different moral observers (at least sometimes) perceive different moral properties, or different relations of salience between the moral properties?
No doubt, one of the best ways of answering this difficulty will be to attack relativism head-on, as incoherent or self-refuting or whatever; but that is another story that I cannot tell here. And no doubt one of the other best responses begins by suggesting that the point of developing capacities for detecting and responding to the patterns that we call the moral properties is given by human interests: both those interests that humans have as the particular kind of creatures they are, and also those interests that humans would have to have no matter what kind of creatures (or rational creatures) they were. The patterns that are salient as the moral properties must, therefore, be the patterns that it is most useful and helpful for us to heed in pursuit of these interests. Moral relativism will only be a serious threat if there is no fact of the matter about which patterns these are. However, we cannot pretend that we have already worked out in full which properties are the salient ones for anything like an optimal normative ethics. It can be perfectly reasonable to see this as a question still in play for further cross-cultural negotiation. But to say this, of course, is to admit that a full answer to relativism of this sort is also another story, and furthermore, that the story is not yet over.
Thus moral relativism remains, at this stage in the dialectic, an unrefuted alternative to moral realism. Hence the moral that I close with: ontological extravagance is not the problem for the moral realist, relativism is.Footnote 15
