Truth and A Priori Possibility: Egan’s Charge against Quasi Realism:

Simon Blackburn

 

This is the preprint of an article whose final and definitive form will be published in the Australasian Journal of Philosophy 2009; the AJP is available online at: http://www.tandf.co.uk/journals/

 

Abstract: In this journal Andy Egan argued that, contrary to what I have claimed, quasi-realism is committed to a damaging asymmetry between the way a subject regards himself and the way he regards others. In particular, a subject must believe it to be a priori that if something is one of his stable or fundamental beliefs, then it is true. Whereas he will not hold that this is a priori true of other people. In this paper I rebut Egan’s argument, and give further consideration to the correct way to think about our own fallibility.

 

1. Setting the Scene

In a passage in the Appendix to Ruling Passions, I wrote:

 

The problem comes with thinking of myself (or of us or our tradition) that I may be mistaken. How can I make sense of fears of my own fallibility? Well, there are a number of things I admire: for instance, information, sensitivity, maturity, imagination, coherence. I know that other people show defects in these respects, and that these defects lead to bad opinions. But can I exempt myself from the same possibility? Of course not (that would be unpardonably smug). So I can think that perhaps some of my opinions are due to defects of information, sensitivity, maturity, imagination and coherence. If I really set out to investigate whether this is true, I stand on one part of the (Neurath) boat and inspect the other parts. (Blackburn 1998: 318)

 

In a lucid, and I understand influential paper, Andy Egan makes the initially surprising claim that the way I deal with thoughts about my own fallibility in this passage ends up committing me to an asymmetry between how I must regard myself, and how I must regard others (Egan 2007). I must, he claims, believe that it is an a priori truth that my own stable beliefs are true, whereas I will not believe this of others. A stable belief of a person is one that is not changeable by anything that he himself would regard as an improvement. Someone’s fundamental beliefs are those that are in this way stable(2007: 212). Egan writes that if someone’s fundamental beliefs are erroneous, then ‘this isolates him from the truth in a particularly serious way’, since it is impossible for him to arrive at the truth about the matter in question by engaging in any process of belief revision that he’d endorse as legitimate. The truth would be what we might call internally inaccessible to him, using the qualification of ‘internally’ to remind us that this is not a causal notion. It is not a claim about where a subject could in fact get to: there is no claim about de facto accessibility or inaccessibility. Knocks on the head or processes of conversion might get a subject anywhere. But not by processes that he would regard, antecedently, as about to improve his position.

                  There is a possible ambiguity in internal accessibility as this explains it. There is one question, whether there may be a process leading the subject to a changed position, such that at each stage of the process, the subject regards himself as having improved. That would be a question of the post hoc judgments made as time goes by. A different notion is one of antecedent judgment, that is, of whether from the prior position, before a change takes place, a contemplation of the change also endorses it. These need not be the same, but let us consider the post hoc judgement first. If the Ghost of Christmas Still to Come unveils the future me, I might blanch in horror at the character I have become. Perhaps I see a complacent unfeeling conservative, and this character’s self-congratulatory belief that each of the processes was an improvement, that led to him being as he is, merely increases my pallor. Indeed any spiral of decline might typically be accompanied by the complacent thought that now I am getting things right at last: such complacency would itself be a symptom of the decline. If some dreadful process leads me to join the Moonies, then a particularly horrible thing about it may be that at each point in the process I believe that I am becoming more enlightened, and from the endpoint believe that joining the Moonies has made me a better person.

                  The prior or antecedent judgment is more tricky. Clearly, imagining a step that gains current approbation is not necessarily imagining an improvement either. If someone’s current position is sufficiently sunk in sin, she may not be able to recognize genuine improvement for what it is, and she may accept changes as improvements when they are not. If we are ourselves sufficiently submersed in sin it would follow that we could not now appreciate any genuine improvement of ourselves for what it is, and this gives rise to what we might call the Christian paradox, that if there is no health in us then one of the things we would be unable to recognize is that there is no health in us. Even Christianity, therefore, had to exempt some little bit from its general gloomy diagnosis of the human condition, a little glimmer of grace that enabled us to recognize our fallen state. That bit had to have its head above water, had to be fairly healthy after all.

                  We need now to consider how from the subject’s point of view, a change might both be an improvement and result in her abandoning something to which she is now committed. On the face of it this might seem paradoxical, not just in ethics, but everywhere. If I believe that p, how can I at the same time suppose that coming to believe that ¬p might be an improvement? Don’t I have to regard it, from the current standpoint, as a fall? The reason not is that I can have a commitment to processes of belief formation and adjustment that outweigh my commitment to p. I might be fairly certain that the film starts at 8.00 p.m., but know well that I could turn up, and find that it is earlier or later. Turning up puts me into what, from my present perspective, I am certain would be a better position to know the starting time than I am in at present, and that commitment outweighs my attachment to 8.00 p.m. as the starting time. In ethics it is no so clear, since there is little with the same clear-cut, decisive effect of being confronted with the time of the film. But the processes of improving information, coherence, maturity, imagination, insight, and humanity, that I gestured at in my account are intended to have a similar procedural status. Of course, according them that status is making them fairly central to my moral outlook.  But it is not making them a priori, nor immune to possible inspection from other parts of the boat.

                  Why not? I say that a proposition is only to be deemed a priori when we cannot see how anybody who understands it could intelligibly deny it (Blackburn 1987). I can imagine someone intelligibly querying any of the processes I select: indeed, in the same work I turned over the question of whether increasing coherence is always a good thing, and others have done the same for increased information and even culture and imagination (Blackburn 1998: 310; Gibbard 1990: 18—22; Dworkin 1996: 87). Nasty pieces of work might deny the value of maturity and humanity. They might say outright that these are bad things, but it is usually futile to try to reverse the polarity of positive words, and far more likely they will take processes which to the rest of us might look like increases in maturity and humanity, and redescribe them, for instance as merely fanciful or sentimental. 

                  Someone might take the other tack. They might say that it is a priori true that increased humanity, for instance, improves one’s moral outlook. They might try to derive this from views about the purpose of moral thought, and by forging from that an a priori connection with human well being. That is fine, but it does not open any door to the asymmetry Egan wants. If I decide it is a priori that increased humanity would improve my moral outlook, then I will regard it as equally a priori that it would increase yours and theirs.

                  In any event, Egan’s argument does not concern individual principles, de re. It is not about particular principles, which any of us may hold, and which perhaps deserve an a priori status. Rather, he argues that with the account I gave in the quoted paragraph, each of us will have a de dicto a priori guarantee that whatever they may be, any stable or fundamental beliefs of our own are true. It is as if we had an a priori proof of a proposition of the form ‘if I said something on Saturday morning, then it is true’ and such a proof must clearly be fishy, whether or not everything I said on Saturday morning was a theorem of first-order logic. If quasi-realism is convicted of offering us a parallel piece of snake-oil about ourselves, then it is indeed in trouble. Since we will not believe this a priori of other people, it certainly violates my own ‘no smugness’ requirement, which requires me to think of myself and other people as symmetrical, so far as any a priori guarantee against error goes. However, from my point of view, the de dicto a prioricity would be worse than just being a preface to an asymmetry between how I regard myself and how I regard you. It would straightforwardly violate my views about a prioricity. For it is obviously not unintelligible for someone to suppose that something may be one of my most fundamental and stable commitments, but not at all likely to be true, and certainly not on that account. If this is an intelligible position for anyone to be in, and it clearly is, then the conditional claim that if something is one of my most fundamental and stable commitments, then it is true, must not itself be a priori.

                  I say that Egan’s is a surprising argument, because the Neurath’s boat model is usually thought to consort with general anti-foundationalist sentiments, hospitable to fallibility or to open-ended processes of self-criticism and possibly temporary resting places, where we hope that we have achieved reflective equilibrium. But if Egan is right, then there is this big, bad, bug underlying the whole story: an a priori self-confidence and corresponding asymmetry in my attitude to myself and others. Perhaps, however, the argument gets some credibility from its cousinship with the familiar thought that only ‘metaphysical realism’ allows Cartesian scepticism, the complete dissociation between thought and truth, to make sense, whereas anti-realisms of various hues must rule it out. By extension, there may arise the idea that only something called ‘realism’ about values and obligations allows us to entertain the idea that I, or we, may be totally dissociated from truth about these things, whereas anti-realisms must hold it a priori that I, or we, must be roughly on the right track. I return to this resemblance later.

                  Menwhile, the argument is perhaps also surprising in the width of its embrace. Because although in his paper Egan concentrates on what I have said, we can expect others to be alongside me in the target area. For example ‘secondary quality’ sentimental theories, such as those of John McDowell or perhaps David Wiggins, however their authors may take them to differ from quasi-realism, share with it a Neurath’s boat account of what makes for moral reflection and self-doubt. Reflective equilibrium accounts, and other constructivisms, may also stand in the same arena. Ronald Dworkin, for example, has the view that moral and legal truth alike are the endpoint of processes of interpretation, and although he talks much of the objectivity of those processes he also allows that while we can suppose that physical truths are ‘just there anyway’, opening up the possibility that we systematically get them wrong, in the case of moral and legal truth this is not so (this was a message of his valedictory lecture in 2008 at University College, London). Here as well, then, there is the threat of an a priori guarantee that if something is a fundamental principle of my (or our) interpretive practices, then it yields the truth, and there will be a corresponding asymmetry between stable elements in our own interpretive repertoires, and those in others. So smugness is just as great a problem.

                  I conclude this exposition by mentioning that Egan’s argument is not affected by the possibility that it is not a priori that any of my commitments are, in fact stable. It may be that my situation is wholly internally improvable, that is by steps that antecedently I endorse. To see how this might happen, consider the parallel of artistic taste. Suppose I start off with a fairly sickly and unpractised taste for crude and sentimental paintings of a chocolate box variety. I recognize, in a kind of abstract way, that I might learn more about art by going to the National Gallery, so I do, and I find my tastes changing. After years of practice and patience, I bear no resemblance to my original self: not a single painting, or way of judging painting need survive the process. Whereas at the beginning I used the National Gallery as a touchstone, by the end I may be using Tate Modern. Yet at each stage I do something which has an antecedent merit in my own eyes. And, of course, at each stage the antecedent expectation is the the process is going to improve me, and the post hoc judgment is that at last I am beginning to see properly. In principle this structure could apply to ethics. But this does not affect Egan’s argument, which requires only that it is a priori that if a commitment of my own is stable, then it is true. He does not require that any commitment actually is stable, and this is a strength of his argument.

 

2. Egan’s Argument

So the charge of first-person, de dicto, conditional a prioricity is grave. But does it stick, or is there a hidden error? The crucial passage in Egan’s paper reads as follows:

 

For me to be fundamentally in error, I need to have some moral view that’s (a) stable, and (b) mistaken.  But given Blackburn’s account of moral error, this can’t happen.  For my moral belief that P to be stable is for it to be such that it would survive any improving change (or course of improving changes).  For my moral belief that P to be mistaken is for there to be some improving change (or course of improving changes) that would lead me to abandon P.  So on Blackburn’s account of moral error, a moral belief is mistaken only if it’s not stable.  So for me to be fundamentally in error, I’d need to have some moral view that was (a) stable, and (b) not stable, which I pretty clearly can’t have. 

                  So if I’m a reflective quasi-realist, I can know in advance, just by thinking about what moral error is, that I can’t be fundamentally morally mistaken (Egan 2007: 214) .

 

Now unfortunately this is not quite right, and the extent to which it misses being an accurate formulation of anything the quasi-realist ought to accept is critical. It is not quite right that for a belief that p to be stable is ‘for it to be such that it would survive any improving change (or course of improving changes)’. This gives a criterion of stability in terms of whatever is an improving change. Whereas as we have seen, officially stability is a matter of surviving anything that the subject would regard as an improving change, either antecedently, or post hoc. Without this conflation, the result that a moral belief is mistaken only if it is not stable does not follow, and the contradiction does not follow either. To see this, let us distinguish the two ideas more carefully. It is the difference between:

 

(M) If something is entrenched in my outlook, in such a way that nothing I could recognize as an improvement would undermine it, then it is true.

(I) If something is entrenched in my outlook, in such a way that nothing that is an improvement would undermine it, then it is true.

 

The first of these is the gun Egan would point at the quasi-realist: the a prioricity of (M) would deliver a kind of first-person smugness that I was concerned to avoid. The second is very different. It talks of immunity to actual improvement: something at least close to Crispin Wright’s notion of superassertibility. And it may be a priori that if something is immune to all actual and possible improvement, then it is true—given a natural connection between improving and getting towards the truth. But (I) is not a problem, for it introduces no asymmetry between myself and others. If I hold (I) to be a priori true, I should equally hold its impersonal version to be so:

 

(Ię) If something is entrenched in anyone’s outlook, in such a way that nothing that is an improvement would undermine it, then it is true.

 

For that matter, neither version of (I) introduces anything specific to ethics: it may be a priori that if someone’s belief that the film starts at eight o’clock is immune to improvement, then it is true. For if it were false, then an improvement is clearly on the cards, namely replacing it with the truth. Deflationism about truth is quite compatible with (Ię). A deflationist will interpret it as a generalization corresponding to the schema ‘if p then an outlook which includes ¬p is capable of improvement’ and under  any acceptable interpretation of improvement, this will be something to be asserted.

                  So a swift rebuttal on my part would be simply to reject the conflation between (M) and (I). I obviously would want to reject the parallel conflation in the case of others. As I have already said, things that you might regard as improvements may be the reverse. What you regard as an improvement might be a deterioration, and vice-versa. And if I can think this about you, then equally you will be able to think it about me, and it cannot be a priori that you are wrong. Indeed, I should be able to acknowledge the possibility myself. The Neurath’s boat methodology seems to imply as much, for why should my own standards for improvement be themselves exempt from critical scrutiny?

                  Of course, while I am deploying particular principles, I am not at the same time scrutinizing them. But they may come under the microscope at any time. I do know in advance that if there are principles or commitments, whichever they may be, that are stable, they will in fact survive anything I can currently recognize as a process of improvement. That is the proper definition of stability. But why should I have any kind of conviction, let alone an a priori conviction, that if anything is, in this way stable, then it deserves to be so, by being true? A very small tincture of self-doubt would entail silence, on my own part, about whether everything that is an improvement is something that I regard as an improvement, or conversely. And an even smaller tincture is enough to enable me to interpret others as intelligibly supposing that although something is one of my stable commitments, it is untrue, which is enough to dismiss an a priori status for the denial of this.

                  To sustain his charge Egan must now try to deny that a quasi-realist should be able to say the things I have just said. Notice his claim, in the quoted paragraph, that ‘on Blackburn’s account of moral error, a moral belief is mistaken only if it is not stable’. So the reformatted charge would be that on my account of moral error, what I just called a conflation, the conflation between (M) and (I), is no conflation at all. Perhaps only realists of some stamp, or at any rate persons far from me, can say that it is. But why should that be right?

                  The prior question is whether quasi-realism offers an “account” of moral error at all. What I do offer is an account of a state of mind and a process—the state of mind of worrying whether one is oneself in moral error, and of the process of seeking to root out any hidden error. But if some theorist bent on finding truth-conditions asks me what my account of moral error itself is, then I am not very forthcoming. What would she expect? It is much more in the spirit of quasi-realism (or expressivism, or indeed pragmatism in general, of which quasi-realism is but a small corner) to avoid such formulations. This is not an ad hoc move, but an integral part of the package. The pragmatist works not by the Fregean or Moorean method of tussling with the proposition in question itself, and more or less vainly chasing after reductions or analyses, but by expanding his gaze to a whole practice: in this case, the practice of thinking about what to endorse and reject in connection with ethics and morality. What is moral error itself? is a Fregean question. The pragmatist response is to think about what we are up to when we impute it, or fear it, or confidently dismiss it as a possibility.

                  Since this is crucial, let us first consider the flip side, moral correctness, or in other words moral truth. Does the quasi-realist set about saying what this is? No. That is just the kind of approach that is completely off his menu. Just as the quasi-realist avoids naturalistic reductions, so he avoids saying what it is for a moral claim to be true, except in boring homophonic or deflationary terms. The only answer we should recognize to the question ‘what is it for happiness to be good?’ is happiness being good. Thus far, as Allan Gibbard and I have also emphasized, Moore was right: every proposition is what it is and not another thing (Gibbard 2003: 21—37; Blackburn 1998: 86). Of course, having said that much, we can go on to ruminate over what it is in virtue of which happiness is good, and deploy our standards and values to pursue this ethical question. But according to expressivists, that is a different enterprise.

                  As with truth, so with error. Moral p is mistaken or erroneous if and only if ¬p— and that is all that should be said. ‘What it is’ for it to be a mistake to think that happiness is always good, is happiness being sometimes indifferent or bad. What it would be for it to be a mistake to think that slavery is always wrong, would be slavery being sometimes just fine.

                  I have come to realize that many philosophers can do little more than gape at this point. Surely expressivism and quasi-realism present themselves as theories, as more than mere deflationary platitudes? Of course they do—but not by finding ‘truth-conditions’ for moral statements. The extra theory is a theory of the overall practice and the emergence of content. We try to explain the emergence of the propositions that serve as the counters of thought, out of the necessary conditions for this practice to take place. It is only a twentieth-century prejudice that took the whole task of philosophy to be ‘analysis’, or the generally unsuccessful search for metapahysical or normative ‘facts’ making true our sayings. The pragmatist or quasi-realist is not so blinkered. He does not suppose that our linguistic behaviour with semantic terminology exhausts the evidence, or circumscribes the methods of philosophy. He looks to the whole practice or ‘stream of life’ in which the sayings are embedded.

 

3. A Parallel Case

I was more convinced of the merits of this deflationary approach to truth than by the time of the later book, and some later essays, than I had been in Blackburn 1984. There, I flirted with an account of moral truth as membership of a maximally improved set of attitudes, parallel to Peirce’s account of scientific truth as whatever belongs to science at the limit of possible enquiry. But even thenI did not endorse this as a reductive account, but expressed significant reservations about it (1984: 249–50). I already had pragmatist leanings, and much preferred the idea of truth as a kind of ‘focus imaginarius’ which we use to think about the enterprises of improvement. I was already fond of quoting Nietzsche on there being no need to insist that a process has an endpoint (Blackburn 1980b: 22). The procedural value of saying ‘I may be wrong’ is no more than the value of wondering how to deploy better information, or how to iron out incoherence, or improve in those things I admire, such as maturity and imagination and cultivation.

                  Perhaps it would help here to digress through a parallel case that guided me from graduate student days (see also 1980a: 89). Can we make sense of singular case probabilities? It is very tricky to give them truth conditions, such as frequency of occurrence of ‘similar’ events in either finite or infinite sequences of trials—just for a start, there is the obstacle of identifying similarity, and of course in reality there may be no such sequences. And it is all too easy to think they must all be either 0 or 1 even if we do not know which, but that makes no sense of our practice as good empirical gamblers (a bookie that only offers odds corresponding to 1 or 0 on horses in a race or candidates in the current American election will not be long in business).  It might seem better to ‘relativize’ them to an evidential basis, supposing like Keynes that we are identifying the degree of support some given body of evidence offers to the proposition that the event will happen. But as A. J. Ayer convincingly argued a long time ago, even if there exist such support relationships, which is doubtful, such a relational account does not do justice to the way in which we think that an improvement in evidence also improves our probability judgment (Ayer 1963). It does not just change the context but may have us saying that our previous judgment was wrong (an argument that strikes me as anticipating much of the debate in recent revivals of contextualism).

                  What to do when we cannot find a truth-condition? Think of the practice. So, following Ramsey and de Finetti, we switch to thinking of the activity of exchanging confidences, where confidences are identified in terms of idealized dispositions to buy and sell bets on the singular events in question. This gives us an activity and importantly it is one in which there can be better and worse evidential situations. Better or worse, because the singular case probability judgment has a consequence for how we plan. It has an output: if you accept that the probability is x/y do not buy a bet paying less than y for a stake of x, and do not sell a bet paying more than y for a stake of x. What does it mean to say that estimates of this may be better or worse? The answer is that you do well to pay attention to the evidence, because if your probability estimate is better based, you can buy bets that others would shy away from, or sell them at better prices than your competitors, and still make a profit.  In general, the more your practice is attuned to the way things fall out, the better things go for you: the improved attunement is not just exhibited in the idealized context of buying and selling bets. If you listen to the weather forecast, you tend not to lumber yourself with an umbrella when there is a very small chance of rain, nor with a deck chair when there is little chance of sun. So we have a discourse in which these things are discussed. In such discussion, mistakes are admitted when a particular body of evidence or ‘context’ of evaluation can be relegated as having been partial or misleading, and different probability estimates made some other way are preferred.

                  As in the ethics case, the practical output explains why it is rational to think in terms of, say, ‘the probability of Obama winning’ and to defer to some opinions about that and to reject others. Our discourse then has the hallmarks of a truth-finding discourse. In the probability case there may be tell-tale signs that it only shows these hallmarks as long as it needs to do so. For instance, if we were clear that the probability of Obama winning was high right up until the eleventh hour, and then woke in the morning to find that he had not, do we have to confess a mistake? It seems a little stubborn to go on saying we were right nevertheless, but on the other hand don’t we allow that improbable events sometimes happen, so could not this have been one of them? The brute fact is that after the event we lose interest: the result is in and the bookmakers have closed down. But before the event, the probability acts as a focus imaginarius, on which discourse aimed at improvement is targeted. This approach vindicates the practice, enabling it to resist ambush by error theorists from one side, saying that the whole practice should be abandoned as unsound (in the probability case, this was the proposal of Richard von Mises), and realists on the other, wringing their hands in the conviction that to make it respectable we must pursue the phantom of a fact or truth-condition. 

                  As with ethics, the only direct theory of error, on this account, is that if the probability of Obama winning is 1/3, then estimates increasingly divergent from this are increasingly wrong. There is much more indirect to be said of course, about what should be taken into account in coming to an estimate, and how it should be taken into account. And similarly in ethics, the only direct account of error is the deflationary account already described, while the indirect account says parallel things about what should be taken into account and how it should be taken into account in deciding what is good, or bad, an improvement or a defect. Saying those indirect things we will deploy our standards of course, in the ethical as in the empirical case. 

                  The probability case gives us the same distinction between (M) and (I). Suppose Henry is a dogmatist about some singular case probability. Suppose he has lapped up Al Gore, and thinks that in the near future catastrophic anthropogenic global warming is a racing certainty. He refuses to countenance anything apt to diminish that confidence, dismissing contrary indications as biased and worthless fictions put about by oil companies. Then it may well be that nothing he could recognize (we might say, nothing he could bring himself to recognize) as an improvement in his position would undermine his confidence: were he to live long enough then even as the next ice sheets advance upon him he would go to bed worrying about global warming. But it may also be that a genuine improvement, such as paying attention to actual temperatures instead of filleted reports of them, would undermine that confidence.[1] And who are we to deny, a priori, that on one issue or another—or even this one— we might be a bit like Henry?

                  There is an interesting question about how all this bears on the Peircean idea of truth as that on which scientific improvement is fated to converge. What we have discovered instead is that if particularly serious grit gets into the machinery at one stage, it may be immovable by internally recognizable improvement. Only conversion, a kind of leap to a new orbit, will overcome it. Whatever Peirce may have thought, there is no fate about it. All that we can salvage from Peirce’s idea, it seems to me, is that if we started from scratch, or if that is unrealistic, then from a spotless population of beliefs, and used methods that never introduced insensitivity to genuine improvement, and went on long enough, we would get to the truth. Which is often, but probably not always, so.

                  To recapitulate, error is as deflationary as truth. It is imputed from within the practice, from an immersion in the business of making, criticizing, accepting or withdrawing the verdicts whose impact is essentially practical: on stakes and winnings in the probability case, on the way we live in the other. At no point does it give us the certainty, let alone the a priori certainty, that if we hold something, however intransigently, we are right.

                  So Egan’s conflation is just as much a conflation from a quasi-realist perspective as from any other. If something is one of my fundamental commitments, then I will not be able to contemplate its abandonment as an improvement. I will doubtless believe I am right about it, and may simply dismiss dispute, or even open-mindedness about it. But I will have no right to any a priori claim that if this is my state, then my commitment must be true. If I suppose that it is a priori that such a commitment is true, it must be because of a de re virtue it shows, in which case it will be a priori as well that if you hold it you are right. But the guarantee that either you or I is right would have to come from the other, de re, source. 

 

4. Benign Meditations

We may conclude by noticing that there are other interesting questions about modality and a prioricity when we think of our fundamental attitudes. It seems possible to frame principles that we hold to apply in all possible worlds. We really cannot imagine them not to hold. Perhaps a principle that if two situations differ only in that in one of them there is increased enjoyment of the pain of others, then it is worse, is an instance. If someone holds it, they may hold it to govern all possible worlds. It is held as necessarily true. As already argued it should not be supposed a priori in the sense that it is known to be true by anybody who understands the terms it contains. It is all too easy to imagine, or simply to describe, whole societies where it has not been accepted. What goes on here, according to the quasi-realist, is that having a sensibility that sets us against the increased enjoyment side of the comparison, then whatever scenario anyone depicts (whatever possible world they have us contemplate) we remain against it. It is worse here, and would be worse there. Whenever we deploy our sensibility, whatever the input in terms of descriptions of what people and animals are like, this is the output. So it functions to give the same verdict on any possible world. On the other hand, and unfortunately, it is not knowable a priori, and perhaps even cannot be appreciated or shared by those large numbers of human beings who are not the least ashamed of looking forward to sharing St. Thomas Aquinas’s relish in the eternal punishment of the damned, or even just to enjoying tomorrow’s stoning of the adultress. It is interesting that morality consorts much more readily with necessity than with a prioricity, and it is a massive plus for quasi-realism that it explains why this is true, when other theories do not.

                  Earlier on I speculated whether Egan’s approach gains some credibility from its cousinship with the idea that only metaphysical realism can countenance Cartesian scepticism, whereas anti-realisms must rule it out, a priori. It would be pleasantly ironic to find this advanced as a problem, since it is the reversal of a more common charge which was made in earlier times, which was that expressivists put too few constraints on notions like desire, and especially approval, which in turn implies that they can countenance extraordinary and bizarre choices and thence ‘moralities’—including ones which are as far from ethical or moral good sense, or reason, or truth, as the Cartesian sceptic supposes that we might be from any other truth. The expressivist, on this old account, must inappropriately make room for people who suppose that their good lies in possessing a saucer of mud, or whatever it may be (Anscombe 1957: 70).

                  My own view is that a prioricity has little to do with it, but nature has a lot. There is nothing a priori impossible about someone wanting a saucer of mud: indeed there is a well-known deficiency syndrome, pica, in which this is exactly what people do want. There is nothing a priori impossible about someone identifying with this want, cherishing it as a value, in the same way that some people cherish and take pride in their ability as wine connoisseurs. Christianity, according to its critics, did a good job of reversing some values, generating its endemic masochism and then ensuing train of ‘monkish virtues’. But all of us, ‘realists’ or not, will find ourselves stymied if asked to work through the idea that our central lists of goods and ills need wholesale reversal, or to work through the idea that our central list of virtues and vices need their polarities switching. We cannot do anything with such suggestions, any more than we could obey the prescription to enjoy distress or love malevolence, which, indeed, would be required for a wholesale reversal of values. Nature and culture have their privileges when it comes to desire and attitude, just as much as when it comes to perception and induction. There may be no truth condition for moral commitments to be found in nature, but natural constraints on what we can admire and what we dislike, for all that. Or, as Wittgenstein put it: ‘“Yes, but has nature nothing to say here?” Indeed she hasbut she makes herself audible in another way.’ (1967:  §364)

                  This completes my commentary on Egan’s complaint. There is indeed a remaining asymmetry between me and you, but it is harmless. I cannot judge a putative change as an improvement in myself without deploying some of my own views. I cannot judge the same as an improvement in you either, without deploying some of my own views. But I can in principle judge the same change as an improvement in you without deploying any of your views. I can do it ‘wholly from the outside’, as we might say, although that would be only in the extreme case, where there is no common ground from which we can work. That is, of course, not peculiar to morality. I cannot assess whether you are getting nearer the truth in your beliefs about anything, without deploying my own beliefs or my own procedures. And I cannot assess the proportion of truth in mine either, without doing the same. But I could in principle assess your progress, without deploying any of your views or your procedures, at any rate unless we have a Davidsonian optimism that many of our views and procedures must coincide, and even then I can assess your progress without knowing whether you would agree with me about it being progress or not. But then, you are in the same position in relation to me, and it is no more surprising than that we each have to hold our own beliefs and see out of our own eyes. [2]

                   

University of Cambridge

Bibliography

Anscombe, Elizabeth 1957. Intention, Oxford: Blackwell.

Ayer, A. J. 1963. The Relational Theory of Probability in The Concept of a Person and Other Essays, London: Macmillan.

Blackburn, Simon 1980a. Opinions and Chances, reprinted in Blackburn 1993.

Blackburn, Simon 1980b. Truth, Realism, and the Regulation of Theory, reprinted in Blackburn 1993.

Blackburn, Simon 1984. Spreading the Word, Oxford and New York: Oxford University Press.

Blackburn, Simon 1993. Essays in Quasi-Realism, New York: Oxford University Press.

Blackburn, Simon 1987.  ‘Morals and Modals’ in Blackburn 1993.

Blackburn, Simon 1998. Ruling Passions, Oxford and New York: Oxford University Press.

Dworkin, Ronald. 1996. Objectivity and Truth: You’d Better Believe It, Philosophy and Public Affairs, 25: 87.

Egan, Andy. Quasi-Realism and Fundamental Moral Error, Australasian Journal of Philosophy 85/2: 205—219.

Gibbard, Alan 1990. Wise Choices, Apt Feelings, Cambridge, Mass: Harvard University Press.

Gibbard, Allan 2003. Thinking How to Live, Cambridge, Mass: Harvard University Press.

Wittgenstein, L. 1967. Zettel, trans. Elizabeth Anscombe, Blackwell: Oxford.

 



[1] As I write (July 2008) best measurements suggest that there has been no global warming since 1998. But I introduce the example not to be polemical on this issue, but only because in some readers it may awaken a vivid sense of what Henry is like.

[2] I am grateful to an anonymous referee for this journal, who suggested a number of valuable improvements to an earlier draft. I am also indebted to Michael Ridge, Daniel Elstein and Cristian Constantinescu for conversations on this matter.