Greg Detre
Monday, 21 January, 2002
Dr Tasioulas
The meta-ethical landscape is dotted with different boundaries, signposts and enclaves, huddled around a few central questions: Are the sentences we use when we make moral claims capable of being true or false? If so, are any such moral sentences actually true? Would such moral claims be objective? Or would they be relativised according to biology, culture or whim? To what extent do different cultures agree on fundamental first order principles? Why do moral beliefs have a motivating force?
Before going further, I am going to try and consider what the word �objective� means. The simplest way of characterising objectivity is universality. We might say that a moral claim, e.g. that �torturing babies is wrong� is objectively true, really only to mean that everybody in the world agrees with it. The problem with this universality definition is that just because everyone agrees with it doesn't make it right, in the way that we want things to be right. We might imagine that if 99% of the world was killed in a nuclear holocaust, leaving only the obscure Brazilian Cocoabutter tribe whose DNA is mangled by the fallout such that every single member of the tribe from then on was to take singular delight in torturing babies, we would not agree (from the grave) with Cocoabutter philosophers who considered that �torturing babies is right� because everybody says so.
Perhaps we could throw the net of objectivity a little wider, and appeal to anybody (i.e. any rational being) capable of understanding the statement we�re making for agreement. This is closer to Nagel�s iterative definition of objectivity � in order to �acquire a more objective understanding of some aspect of life or the world, we step back from our initial view of it and form a new conception which has that view and its relation to the world as its object�. Objectivity is defined negatively in terms of subjectivity � take a subjective viewpoint � our own, for example. Step back from it slightly, form a conception which wholly contains and widens that viewpoint, e.g. other members of Corpus Christi College. As we broaden to the entire human population, and the entire population of rational beings, we hopefully lose the most misleading and myopic aspects of our own perspective, and approach a view which incorporates such a wide variety of physiologies, priorities and perspectives that we approximate towards a �view from nowhere�.
Moral objectivists tend to characterise their propositions in terms of truth. Truth seems to be beckoning towards some optimal objectivity, the objectivised view that one would have if one could incorporate the perspectives of the entire space of (entirely) rational beings together. It is for this reason that truth is often described in terms of the point on which a perfect inquiry converges. Of course, pragmatists and relativists like Rorty and Putnam are desperately keen to rid us of just this sort of notion of truth, but that comes later.
Lastly, I want to distinguish absoluteness from objectivity. �It can be objective that an interpretation or an explanation is the correct one, given the interests which are relevant in the context�[1]. The way I imagine this is to consider a black box, which we�ll call the Ethicomatic 9500. The Ethicomatic 9500 contains the rules by which moral features supervene on natural features. You place a moral agent inside, complete with the details of his choices, circumstances, context, culture, interests and values, and out the other side, it prints out on ticker tape what the right action for that agent should be. Our hypothetical moral objectivist, when given one of these for Christmas, excitedly taps in the details for various dilemmas that he and his friends have argued over, and disappointedly finds that sometimes, the �right� act will be different for different people, even when the extraneous circumstances are identical. He tries to take it back, claiming that it�s not giving him an objective answer. What the shopkeeper will tell him is that he is confusing objectivity with absoluteness. The answer it gives may be the objectively right answer, but it may be different according to the particular circumstances, culture, interests and values of the agent. This corresponds with Wiggins� first characterisation of relativity (i.e. contextualised relativity, or �interest relativity�).
Following Michael Smith[3], I am going to start by outlining three broad meta-ethical swathes along cognitivist grounds:
moral realism:
The sentences we use when we make moral claims (about the rightness or wrongness of an act) are capable of being either true or false � they are �truth-apt� (� this is the central tenet of any form of cognitivism)
Some such sentences are actually true
nihilism (error theory):
Nihilists accept that there can be sentences expressing moral claims about the way the world is which have truth value
But all such truth-apt moral
sentences are false
non-cognitivism (including expressivism and emotivism):
There are no truth-apt moral sentences
More specifically, emotivists/expressivists hold that sentences making moral claims are not used with the intention of saying something that is capable of being either true or false � we do not use to them in an attempt to make claims about the way the world is. Rather, we use moral sentences to express our feelings about acts, people, states of the world etc. (e.g. �Boo for torturing babies!�)
If one holds that sentences ascribing rightness or wrongness to actions are capable of being true or false, one is left with one of the top two objectivist/cognitivist positions. According to Smith, believing that there are such truth-apt moral sentences, and that some really are true defines the moral realist.
Gowans adds an important further epistemological qualification: that we are able to know these moral propositions, that is, that we have the capacity to justify a correct moral belief. Otherwise, living in a world with truth-apt and true moral propositions that we have inadequate/incomplete knowledge of would probably leave us as quietists, unable to defend or persuade one another about our moral views.
Mackie[4] provides early versions of the two major arguments against moral objectivism that I am going to consider, which he terms the argument from relativism (though I prefer the label, �the disagreement argument�) and the argument from queerness.
The disagreement argument can be stated in various ways. Incorporating Gowans� approach, it starts from the empirical observation that there �are widespread and deep moral disagreements that appear persistently resistant to rational resolution�. Then, he reiterates the definition of moral realism which has �the capacity for being justified in our moral beliefs� as a prerequisite. Finally, �agreement between well-informed, reasonable persons is a key indicator of the extent to which we have the capacity for being justified in our (moral) beliefs�. Taken together, these suggest that the persistent lack of agreement indicates a limited capacity for being justified in our moral beliefs, undermining moral realism. The strength of this conclusion rests on the combined strength of each of these premises: how widespread, deep and persistent these disagreements are; how key an indicator of justification of belief a lack of agreement is considered to be; whether moral objectivism requires us to be wholly justified in all our beliefs.
An important rebuttal concerns the extent to which these cross-cultural variations can be explained as arising out of the application of general principles to different concrete circumstances. This is the hope motivating the moral system builders like Mill.
further rebuttals to the
disagreement argument
Mackie�s argument from queerness confronts objectivism at the metaphysical and epistemological level, though it requires substantial revision to be as forceful as it can be.
The epistemological half of his argument leads him to claim that any moral objectivist is committed to some variant of intuitionism, since values would be entities or qualities of a very strange sort inaccessible to us by naturalistic means, and so requiring some sixth �moral�/intuitive sense. Certainly, if such a sense existed, we would expect it to have led to a far more painless convergence on ethical views than even a charitable survey of disagreements within and between communities would reveal. Of course, Mackie mentions Price�s criticism that much of our fundamental non-moral knowledge (such as essence, number, identity, solidity, infinite time and space, causation etc.) is seemingly subject to the argument from queerness when put like this. Fortunately of course, they can be accommodated easily within a naturalistic framework, as I will discuss shortly.
The metaphysical half of his argument questions what sort of ontological status such objective values would have. He contends that they would have to be something like Plato�s Forms, in order to marry an abstract objectivity with a motivation to act accordingly. He is unable to conceive what possible necessary connection could link natural features and moral features � following Hume, he knows that continual coincidence is not enough to replace semantic or logical necessity in harnessing the moral and the naturalistic. Attempts to naturalise ethics have gained in popularity since Mackie�s criticisms were made, and to an extent he could be accused of attacking a straw man. G. E. Moore is amongst the only philosophers to have advocated such a fundamental realm of extra, sui generis, non-natural ontological [moral] properties. As well as suffering from all of the problems listed above, he was unable to explain how or why these non-natural moral properties supervened on particular natural properties, and whether this link between the moral and natural was contingent. However, a strong case can be made for ethical naturalism, as I will try and show.
Broadly, naturalism is the view that the world is (entirely) amenable to study through empirical science. The explanatory and predictive richness of empirical science has persuaded many philosophers of the plausibility of such a naturalistic picture of the universe. Smith, for one, is happy to couch his moral realism entirely within a naturalistic framework, arguing that whatever feature of an act that gives it its rightness or wrongness must be constrained by our conception of the world, which is a naturalistic one (at least for the secularists to which we�re appealing). Consequently, an object�s features are exclusively natural features. In order then to make sense of moral sentences� truth-aptitude, moral features must be supervening on natural features of the world. (Supervenience in this way can be defined as saying that any two acts which are identical in all of their natural features must be alike in their moral features as well.)
Unfortunately, Moore devised what has come to be known as the �open question argument�, which would, if sound, lay all naturalising attempts firmly to rest. Let us say that we had come up with a system for evaluating the rightness of an act given knowledge of all the pertinent circumstances involved (i.e. its natural features). To take a simple and cheery example, let us imagine that our theory contends that all �right� acts involve red objects. What it is, on this naturalistic account, for an act to be �right�, is for it to involve a red object � we are positing an analytic (i.e. conceptual) equivalence between the rightness of an act and the redness of the objects nearby.
Moore�s open question argument says that just because acts of a given moral feature always have a certain natural feature, does not mean that the moral and natural feature are the same. In our example, Moore�s argument would be aiming to refute this (as well as for any other natural feature that we chose to put in place of redness). First, we ask ourselves whether �this act involves lots of red objects, but does it involve lots of red objects?� The answer is an obvious, even trivial, yes. The two clauses contain identical (i.e. analytically equivalent) phrases. This is a closed question. Now we ask what should be the same question, substituting in our supposed analytic equivalence, and ask �this act involves lots of red objects, but is it right?� At this point we scratch our heads, and have to subject each other to some reasoned arguments. It is not at all obvious in the same way as the previous question (�this act involves lots of red objects, but does it involve lots of red objects?�) that the answer is yes. This second question is an open question. Thus, rightness and redness of objects cannot be analytic equivalents, i.e. they do not pick out the same feature. This argument can be irritatingly wheeled out to seemingly show that rightness cannot be identical with any natural feature of acts at all. In a naturalistic world, we seem to be left as nihilists with no true moral beliefs intact at all. Given that we weren't very impressed with the idea of non-naturalistic objective moral values that Moore suggested take the place of supervenience on natural features, we can be glad that there is one weak and one strong response to be made to the open question argument.
The weak response tries to argue by analogy with H2O and water. If we swap �water� for �rightness� and �H2O� for �redness�, then the open question argument seems to show that water is not analytically equivalent to H2O (since it is still an open, i.e. non-obvious question whether �this substance is H2O, but is it water?�). We know a posteriori that water is analytically equivalent to H2O, yet Moore�s argument purports to show that they aren't.
This leaves naturalistic moral realists with the problem of showing that the natural and the moral features are a posteriori equivalent. At first sight, this seems to parallel the usual empirical scientific method. But Harman[5] tries to argue that there are hidden difficulties in trying to test and confirm moral principles in the same way as scientific principles. When I am building a theory, whether scientific or moral, I am trying to fit it around observations in the real world. In both cases, I probably lack sufficient evidence to falsify certain hypotheses I have, and in both cases, my observations are themselves biased by my hypotheses and preconceptions to some degree. I think Harman is saying that the difference between scientific and moral theory-building lies in the extra layer between moral observations and the underlying moral theory. A scientific observation informs a scientific theory. A behavioural observation informs a psychological theory, from which we then try to abstract to a moral theory. One�s own moral sensibilities colour behavioural observations in a way that they don't when considering physical observations.
I think that there is a much stronger, and simpler, attack that can be made on the open question argument that focuses on the notion of analyticity. Analyticity may be opaque � it should not be judged on the basis of obviousness, or whether or not an equivalence is open to reasoned argument. If we consider what we mean when we discuss a colour, we are trying to pin down �the property of objects that causes them to look red to normal perceivers under standard conditions�, which is a complex set of constraints involving the light conditions, perceiver�s retinal defects etc. Here we can see that analyticity allows for reasoned argument in determining whether the complex set of constraints represented by the given word is entailed by the proposed analysis. If we accept this account of conceptual analysis, then we can see that the open question argument is simply too stringent in its application of analyticity, and so fails to refute the equivalence between �rightness� and any given natural feature a priori, putting the naturalistic project back on track.
I want now to consider a positive non-objectivist account. I�ll start with a fairly extreme non-objectivist stance, termed �pragmatism� by its proponents. Its most eloquent advocate is probably Richard Rorty, and it relates to the entire metaphysical edifice, swallowing meta-ethics within it. It�s a little bit difficult to fully absorb, and frankly, a little bit Continental. I�m going to try and avoid getting embroiled in details, especially skirting around where prominent contemporary pragmatists disagree with each other, and just sketch out an outline of a pragmatist, non-realist metaphysical picture, and discuss where this leaves meta-ethics, and how it relates to relativism in general.
Rorty wants to attack the very notion of truth as it�s been handed down through the philosophical generations since Plato. I�ll start with some descriptions of the sort of truth he wants to dismiss: �Truth is correspondence to the intrinsic nature of reality� (Plato). Truth as �the accurate representation of reality�. Truth as the single, complete, universal, objective, incontrovertible, possibly inaccessible set of beliefs. As he puts it, �the picture which holds traditional philosophy captive is that of the mind as a great mirror, containing various representations � some accurate, some not�, with philosophy peering into and polishing this mirror in the hope of seeing a better, clearer image of the universe reflected in their own minds.
Instead, he wants us to �treat beliefs not as representations but as habits of action, and words not as representations but as tools�. �There is no point in asking whether a belief represents � either mental or physical reality�. Rather, we should ask, �For what purposes would it be useful to hold that belief?�. To sum up, the argument should �not be about which of us has got the universe right�. �We cannot regard truth as a goal of inquiry�. �The purpose of inquiry is to achieve agreement among human beings about what to do, to bring about consensus on the ends to be achieved and the means to be used to achieve those ends.� �All areas of culture are parts of the same endeavour to make life better. There is no deep split between theory and practice�.[6]
In this way, world-views are like clothing. If things get cold, we wear the warmest clothing we have. If we find a better, warmer fabric, we wear that instead. If the climate changes, we don a new, more suitable garment. There is no single, true clothing (belief/belief-set, world view etc.) best suited to the environment (read universe), no one ideal fabric that we are converging upon, just a series of improving cognitive apparatuses.
In order to see how this relates to ethics, it�s necessary to flesh out this relativist metaphysical picture a little more (excerpted from Rorty�s article on Putnam[7]), although I won't discuss all of these points:
1. �... elements of what we call 'languge' or 'mind' penetrate so deeply into what we call 'reality' that the very project of representing ourselves as being 'mappers' of something 'language-independent' is fatally compromised from the start. Like Relativism, but in a different way, Realism is an impossible attempt to view the world from nowhere� (RHF 28)
2. Our position is of �beings who cannot have a view of the world that does not reflect our interests and values, but who are, for all that, committed to regarding some views of the world - and, for that matter, some interests and values - as better than others� (RHF 178).
3. The contrast between absoluteness and objectivity described above
4. The �supremacy of the agent point of view�. �If we find that we must take a certain point of view, use a certain 'conceptual system', then we must not simultaneously advance the claim that it is not really 'the way things are in themselves'.� (Putnam, The Many Faces of Realism)
5. To say, as [Bernard] Williams sometimes does, that convergence to one big picture is required by the very concept of knowledge is sheer dogmatism ... It is, indeed the case that ethical knowledge cannot claim absoluteness; but that is because the notion of absoluteness is incoherent (RHF 171)
This does not fit obviously into Smith�s tri-partite picture of descending levels of realism. This view sits somewhere amongst the most anti-realist positions in Smith�s tri-partite schema (although pragmatism certainly need not be committed to emotivism/expressivism).
The first excerpt makes immediate trouble for the cognitivist by attacking the idea that reality can be objectively expressed in linguistic (i.e. propositional) form. The cognitivist knows that this is assumption takes him out on a limb a little, but I think it actually makes him much more vulnerable than the attacks from emotivism reveal. Specifically, Putnam is claiming that our cognitive and linguistic processes are so interwoven that a �language-independent� representation of reality is �fatally compromised from the start�. But this is exactly what the cognitivist is trying to do � cognitivism hopes that we can use language as a means of representing moral facts, which we presume are not inherently linguistic. For all we know, syntax, vocabulary and fundamental linguistic form could all be unavoidably imposing upon or anthropomorphising our conception of moral facts.
The second and third points very succinctly, though broadly,
make the case for relativism. Amidst the plurality of possible views that we
could adopt, some might be preferable. The preferred action for a given person
can be considered to be a function of their interests and values (incorporating
culture, biology and circumstance).
discuss Wiggins� five brands of
relativism
It�s very difficult to satisfactorily refute this pragmatist picture, for a number of reasons. One important one is that unless it can be shown to be clearly and irredeemably self-refuting, it can simply accommodate and re-cast any alternative (e.g. more realist) metaphysics as a useful, interim world-view that we can gainfully employ until another comes along which will allow us to function better as people and as a culture.
Moral relativism as a standalone
doctrine is similarly robust. Unlike most moral objectivist doctrines, it is
not seeking a set of moral facts that apply universally.
In conclusion,
a combination of the disagreement
argument, and a reluctance to accept that moral claims are truth-apt
consider Wiggins� discussion of why
we disagree
Nietzsche�s dark and nebulous figure lurks behind mentions of nihilism and immoralism, but in a complicated fashion. I�ve deliberately avoided bringing him into the debate, because it�s very difficult to see where he fits into the classifications that have been discussed. Of course, this is partly because his views evolve and shift radically between (and sometimes within) books. On the one hand, his discussions of morality (in Genealogy of Morals and Beyond Good and Evil) largely take the form of a descriptive or narrative analysis of the master/slave moral systems over past centuries. It would be twisting his words to try and shoe-horn the master morality that he advocates as being an �objective� morality � it is simply what he desires of (and prescribes for) the human race. On the other hand, when he talks of the Overman as the joyous affirmation of mankind, he is subverting any moral/value systems that exist � the Overman lives to create his own values, to be his own system. It is the task of the philosophical labourers to catalogue and describe the moral systems of the past, and the task of the philosophers of the future to �step outside their own time� as part of a continual renewal through the �revaluation of all values�. If anything, this is the ultimate subjectivism.
Nietzsche is interesting here in at least one other major respect � when banishing God, teleology, metaphysics and the rational and moral world order in one triumphant sweep, he also laid bare how empty any non-naturalistic atheistic moral realism must be. When Mackie argues against objective moral values as being �queer�, he is unwittingly gesturing at the space left behind in our metaphysical framework formerly occupied by a deity. God was the mysterious, objective, motivating moral force that lurked above the natural. Without God, some sort of naturalistic basis is our all that remains. Indeed, Derek Parfit[8] sees room for optimism in this. He points out that the vast majority of philosophical thought so far about morality has been based on a theistic metaphysics, and we really have only been pursuing non-religious ethics for a very short time in the grand scheme of things.
[1] Putnam, Realism with a human face, pg 210
[2] Except for specific theories, I will use the terms �moral objectivism� and �moral realism� more or less interchangeably
[3] Michael Smith, �Moral realism� in the Blackwell guide to ethical theory (ed. LaFollette)
[4] John Mackie, Ethics, ch 1
[5] The Nature of Morality
[6] I�ve got references/sources for all of these quotes, if anyone wants to track one down. Honestly, I really have. It�s just that I couldn't keep track of them all when I was trying to mix and match them for a complete narrative. For reading Rorty, a good place to start is the introduction to Philosophy and social hope.
[7] Rorty (1993), �Journal of Philosophy�. The following excerpts are chosen by him from Hilary Putnam, Realism with a human face
[8] Derek Parfit, Reasons and Persons, pp 453