The following discussion presents considerations favouring three proposals, which I begin by listing:

1. 1 While facts correspond to some truths, there are not facts corresponding to all truths; facts are not just true propositions.

2. 2 A kind of noncognitivism is a plausible view of ethical and aesthetic judgment, a noncognitivism based on epistemological, not logico-semantic considerations.

3. 3 These same considerations turn out also to apply to the assertion and justification of mathematical propositions.

My proposals and arguments grow out of refection on minimalism about truth and out of epistemology, broadly considered; ‘broadly considered’ because I am taking epistemology to be concerned with the justifications given in response to queries and challenges. Queries, challenges and justifications differ in different regions. I will suggest that while we always claim truth when we make assertions, we do not always claim knowledge, do not always represent ourselves as knowing. Flat, unqualified assertion is, typically, claiming to know or representing oneself as knowing. But only typically. A tight fit between assertion and claiming to know is only completely comfortable when the subject matter is what I call ‘plain facts’, and when the asserting is telling or informing others of plain facts.

Minimalism about truth seems to put paid to noncognitivism as consisting in the failure of evaluative judgments to be truth-evaluable. I am going to suggest that if instead we construe noncognitivism as the absence in evaluative discourse of knowledge claiming, a kind of noncognitivism is sustainable in ethics and aesthetics. If we allow truth values for evaluative judgments or assertions, while denying them the force of knowledge claims, we enhance our understanding of such discourse. The upshot is that ethics and aesthetics (and mathematics) are best understood as domains in which truth is not correspondence to fact. Some support for recognizing a difference between truths and facts lies in our everyday practice of treating beliefs as of truths (or, when mistaken, of falsehoods) but knowledge as of facts. It is odd to talk of believing or disbelieving the facts, while knowing or not knowing the facts is idiomatic. It is also odd to speak of ‘necessary’ facts, as opposed to necessary truths, though ‘contingent’ is a happy adjective either way. Some further considerations from ordinary usage will be mentioned later. My proposals may only amount to a novel way of understanding the difference between contingency and necessity; I am, in effect, proposing that we restrict talk of facts so that only what can (not merely might) be or could have been (not merely might) have been) otherwise is the matter of fact. I am aware that ‘fact’ is a word in our language that has uses that are not as I am proposing. But the word ‘fact’ has served in diverse metaphysical regiments and been returned to mufti in uncertain shape, if indeed ‘returned’ is the right word for it all.

To assist my proposals' progress, I rely on the view put forth in Wittgenstein's Tractatus, that only propositions that have true/false polarity are fully-fledged and proper propositions. The rest are things we assert all right but, may I say, somewhat shady. Wittgenstein says of such propositions that they lack sense. But that only seems to mean not possessing true/false polarity, not being nonsensical, which clearly they are not, even if Wittgenstein himself and some of his exegetes say they are. Notoriously, Wittgenstein distinguishes in his book between ‘lacking sense’ (sinnlos) and ‘nonsensical’ (unsinnig). To make things even worse, he then, at the end of the book, seems to be saying that even the merely sinnlos things he has had to say are ultimately unsinnig. Making things even worse, we are left in deep obscurity about the business of what cannot be said but only shown. Are all or only some of the unsinnig and sinnlos sentences in his book cases of trying to put into declarative sentences what can only be shown? I fear we shall never know.

Despite all that, true/false polarity is a perfectly clear idea and it is the key thing; it is a matter of it being possible for P either to obtain or fail to obtain. Such possibility is a matter of what could, not merely might, be so. ‘It might be that P’ says only: ‘Nothing I know rules out P’, while ‘It could be (have been) that P’ speaks of real alternative possibility, which is to say, contingency.

Expressing possibility is quite a different business for a sentence than is avoiding nonsensicality. It is through understanding, making sense of, some form of words that we grasp that the words succeed or fail at representing a possibility, and which possibility that is. Making sense is the correlate of understanding, another way of speaking of having meaning, of not being nonsensical. As such, it must not be confused with being possible. It makes perfectly good sense for someone to say either ‘There are more than 4 primes between 10 and 20’ or to say ‘There might be more than 4 primes between 10 and 20.’ Such statements make sense though there are only 4 primes there and could not be more or fewer. To say there are more than 4 is to express belief in something impossible. To say there might be more or fewer is to express uncertainty. ‘It might be that P’ is well understood as another way of saying, ‘Nothing I know rules out P’. It is a further question whether it could be that P.

If this is right, then it is a quick step to the first thesis above. Facts correspond to true propositions only where the corresponding states of affairs can fail to obtain. This comes to saying that only contingent truths have facts corresponding to them. I shall argue further below that withholding factuality from getting it right in ethics, aesthetics and mathematics is the right way to think about these domains. There is some support for this in the oddity of the expression ‘a mathematical fact’ as opposed to ‘a mathematical truth’, just as, as I remarked above, necessary truths are better off than necessary facts. Moreover, our practice of leaving matters of fact to juries, detectives, pathologists, poll takers, and the like, and more general matters of fact to science, does not include leaving mathematics to those agencies. It is also to the point that the excellent advice not to argue about facts (encouraged by the presence of the Guinness Book of Records in many barrooms and pubs) is indeed sound advice. But who would advise that we not argue about what it is most reasonable to believe or what it is right or wrong to do?

Whether those observations meet with agreement is not so important as the point, which should readily be agreed to, that the negations of true arithmetical propositions are necessary falsehoods. There aren't any contingent truths in mathematics. It is less recognized that there aren't any in the heart of ethics either, that is in judgments of right and wrong, good and bad that we pass on specific actions or activities. Such judgments, when true, are necessary truths; their negations do not represent possible ethical states of affairs.

That sounds more radical than it is. Take an ethical judgment as indisputable as such judgments can get: The mass arresting and killing of Jews, Gypsies and mentally retarded people by the Nazis during World War II was wrong. It could not have been right or all right. Specific actions that are wrong or unjust could not instead have been right or just. Only if we speak of the actions in general terms, such as ‘What the Nazis did to those people in Auschwitz … ’, could it be said that what they did could have been all right, and not wrong. But that does not make it contingent that what they actually did do was wrong. It just says they could have done something else that was all right, like isolating those groups for their protection from bigoted neighbours in comfortable accommodation during the war. They could hardly have done what they actually did do (starving, gassing, hanging, shooting, using as human guinea pigs, and incinerating all those people) and those very actions have been good rather than bad, right rather than wrong.

A (perhaps any) terrible type of action may take place in circumstances (no doubt dreadful ones) where we are (despairingly) convinced that it is the right thing to do, circumstances where we speak of the lesser evil. But this does not undermine the claim that judgments about specific deeds, in specific circumstances, are, if true, necessarily true. If a horrible action in dreadful circumstances really is the lesser evil, then it (non-contingently) is the right thing to do. That in such cases we speak of what has to be done or what must be done, and not merely of what ought to be done is indicative of the necessity involved.

If this is right then our central moral judgments lack bipolarity; they are not such that their affirmations and their negations both express genuine possibilities. It is this feature, their negations being impossible, even though their affirmations are true or false which argues for my suggestion that there aren't facts corresponding to truths in mathematics and ethics. Ethics is a large territory, however, and I am not suggesting for a moment that all ethically significant judgments wear their truth or falsity necessarily. If some person is a courageous or cowardly, it seems clear enough that this could have been otherwise. Moreover a judgment that says of a particular act that it was generous, allows that that very act could have been otherwise motivated, e.g., by the desire to purchase affection, or the desire to make a show of one's wealth, or to gain a reputation for generosity. If the class of ethical judgments is allowed to include claims about character and motivation, then many ethical judgments will be bipolar. I shall later note that though ethical judgments, if taken to include talk of character and motive, vary in the matter of contingency and necessity, they all share what contrasts them most sharply with plain fact: we do not pass their content on in our own voice, without preface, when we have heard them and accepted them from someone or somewhere else. This striking feature is also present in aesthetic judgment and in mathematical statement. Later in the discussion, this point will be developed in detail.

Where there is bipolarity it is easy to see how a fact and its corresponding proposition can come to exist independently of each other. It is obvious that many states of affairs obtain and have obtained without the propositions that would be true in virtue of that obtaining, themselves existing. It helps here to remember another point from the Tractatus, that a proposition is itself an obtaining state of affairs, a concatenation of logico/syntactic elements. For a proposition to come into being, the constitutive concepts have to come to being. The concepts that are interwoven in the making of a proposition can, of course, be thus interwoven whether the proposition is true or false, and whether the proposition is asserted or occurs unasserted.

Truths dichotomise in into two different ways. The first way is where bipolarity holds sway and where it does not. Another is where states of affairs can obtain without the existence of the concepts that constitute the relevant propositions, and cases where this is not so. The dichotomies are not isomorphic, since propositions such as that someone voted for the winning candidate or checkmated his opponent are bipolar, while the concepts involved must exist if the states of affairs themselves obtain. Some such facts have been called institutional. The failure of bipolarity for ethics and mathematics reflects a requirement that there be concepts of right, wrong, good and bad and concepts of a sum, a product, a prime, etc., if anything is to be any of these. Whether this should lead us to dub ethical and mathematical truths ‘institutional’ needs further reflection, which I eschew for now. It may be that institutionality requires rules, not just concepts. That would have interesting implications for elucidating meaning as a matter of rules.

If we accept that there are facts only where there is bipolarity, we gain a vantage point from which to explain the tenacity of noncognitivism and the dichotomy of fact and value. I shall now go on to provide reasons for construing noncognitivism not as the absence of truth-evaluability from evaluative discourse, but as having to do with the character of claiming and querying there. I will note a contrast between how the words of others bear on plain factual discourse as opposed to evaluative discourse. A further contrast will be observed between the role of the words of others in ethical as opposed to aesthetic discourse. That will spell out what I meant when I said ethics and aesthetics are one, but not entirely one.

Let me begin this part of my discussion by formulating The Epistemic Principle of Hearsay: If b hears from a that P and has no reason to distrust a, b will take him or her self to have found out that P; and, having so found out, b will be willing henceforth to assert that P in his or her own voice. (I am using the phrase ‘own voice’ to cover all of ‘flatly’, ‘unqualifiedly’, ‘without preface’ (more on prefaces later).)

Let ‘hearing from’ do work for a variety of transactions, including being told or informed individually by a, hearing a lecture by a, overhearing a conversation between a and c, reading newspaper, a weekend events notice in the town centre, or other text, etc., It is a good bet that most of what we claim to know we came to know via the words of others. Most relevant to this discussion is the case of one person telling or informing another of something.

The Hearsay Principle applies most clearly and simply when P is plain fact. Further along, I will suggest ‘surrogate’ principles that operate in the passing on of other than plain fact and in the acquisition by b of states of mind short of knowledge or belief. In many transmissions what b acquires is only good reason to believe or accept P. What it is to be plain fact and what it is to be knowledge transmissible via hearsay involve a kind of circle. When P is plain fact, b will, as I said, be willing to claim flatly, in his own voice, that P. He will, if queried, be happy to justify his claim by saying that he has heard that P from a, or been told that P by a. Of course, unless a knows that P, b does not, even though he thinks he does, come to know that P. Whether b's false belief was a justified false belief depends on whether his trust in a was justified. But hearsay is surely a main source of our justifiable (excusable?) false beliefs. Hearsay is so prominent as a means of acquiring knowledge and of acquiring false beliefs as well, that it shows patently that it is not a requirement of knowing that we must know that we know.

When a does know that P and b is told P by a or hears it from a, b does come to know that P and is henceforth justified in making knowledge claims. As I just said, it is likely that the bulk of what we flatly assert, claim to know, we have obtained via hearsay from sources we have no reason to distrust. We get a mix of knowledge and error this way, but the errors are among our justified false beliefs, which we carry around with us for indefinite periods. Such borrowed authority depends, of course, on a being one who knows that P and presents himself in asserting that P as one who does know. We take it to be knowledge that is transmitted and acquired and trust our sources as a matter of default. The default case is a matter of those from whom we hear these things putting themselves forth as knowing and our receiving their words as claims to know.

A striking difference between the plainly factual and the evaluative can now be exhibited. The Hearsay Principle does not apply with evaluative Ps. We do not pass on in our own voice, do not flatly assert, evaluative Ps even if we have accepted them from others. It is not even clear what exactly we acquire though acceptance; nor is it clear why I should speak of acceptance rather than belief (more on this below); and we are constrained, in passing on what we have acquired, to preface what we say with such as:

Such citations are readily offered, after unprefaced assertion of the plainly factual, as responses to queries. But we do not typically preface our plainly factual assertions with them. In plainly factual matters, queries may assume any of the following forms:

By contrast, here is a list of some forms of query and challenge we encounter when making evaluative claims:

Let's imagine a few cases: I tell someone that the local gallery has a Cezanne exhibition starting tomorrow, and my interlocutor asks me how I know or if am I sure, or how I found out. I say that I have been to the gallery, or cite a notice board, a newspaper, etc. But if I tell her that the exhibition contains several of Cezanne's finest still lifes, my interlocutor will be entitled to infer that I have seen them for myself. I might not have said flatly that the Cezannes were among his finest; I might, instead of flatly asserting what I did assert, have said: ‘Apparently … ’, and backed that up with: ‘According to Robert Hughes … ’. I might do that because I trust Hughes' judgments about such paintings and am sure that, upon seeing them for myself I will reach the same judgment.

Consider this in the light of the Hearsay Principle. If the trust I have in Hughes amounts to holding that he knows these things and if I am confident that I will respond with comparable approval and pleasure to the works, why then is this not—and it is not—a case in which I have come to know these Cezannes are fine ones and come to that knowledge by believing Hughes when he, knowing, has said so?

On the say so of Hughes, I may resolve to go to the gallery, anticipating viewing the Cezannes, quite sure that I will enjoy and admire them. I have thereby surely done as much as can be done in the way of accepting Hughes' authority. Why then must I see them for myself to be justified in saying, in my own voice, that they are among Cezanne's finest? We do not pass on, in our own voices, as knowledge, things we have, in this way, accepted from critics, even critics we deeply trust. There clearly is such a thing as accepting the authority of another to the point of confidently acting on it; but it is an acceptance that is short of knowledge, even short of belief. I go to the exhibition, confident that I will respond favourably to the works and I do that on Hughes' say so. But I do not come to know from Hughes that the paintings are fine ones. I can't know that without seeing the works and appreciating them for myself, responding to them for myself. Whether knowledge is or is not the state I reach, even when I work it out for myself, I am not intelligibly grounding my claims if I do not back them up with something about the work and how I experience it. There is here an activity here of claiming and defending claims that has its own grammar as indicated by the differences that the forms of query exhibit.

So aesthetic judgments cannot be slotted into the Hearsay Principle, at either end. Coming out flatly, in my own voice, with aesthetic judgments, without prefaces such as ‘Apparently … ’, ‘According to … ’, etc. licenses the inference that I have experienced the work for myself and made my own judgment; or, if the words of the judgment are ones I first heard from elsewhere about that work (or other item), I need to have, as it is said, made that judgment my own. When I do make a judgment of my own, flatly, unqualifiedly, I may be queried. And the querying comes in the forms I have listed above.

We sometimes find such queries daunting, but we recognize their appropriateness. Art confront us with difficulties in explaining our responses and their depths. It doesn't help much to resort to the old saw about knowing what we like but not knowing what is good. We can be just as daunted by the query: ‘What do you like about it?’ as by: ‘What's good about it?’ We may only be able to meet such queries by urging our interlocutors to look for themselves or to notice something we think they may not have noticed, to contemplate the painting, listen more carefully to the music, read the poem again, etc.

What is particularly noteworthy is that there is kind of response— though not really—to query or challenge that is incoherent, unintelligible. To the query ‘What makes you say that?’ or its like, the reply ‘Robert Hughes says so’ is nonsense. You can say that if you have said: ‘Apparently …’, and are queried about that remark. You could thereby register your inductively based confidence in Hughes' taste and judgment. ‘Apparently …’ in such situations, is roughly the same as ‘According to a source of reliable judgment on these works … ’. If when I say for myself that the works are superb, someone says that I am saying so only because Robert Hughes' has said so, I shall be insulted. There is, however, nothing problematic or self-deprecating about citing Hughes' authority in an explanation of why, though the work doesn't seize me straight off, I work harder at appreciating it. That is what critics are for.

(In George Eliot's Middlemarch Dorothea, in Rome with Casaubon, is brought to realize with dread the mistake she has made in marrying the man. Moving through galleries, and herself often moved by what she sees, she eagerly anticipates comment from her husband. All she gets is information about what this or that art historian has said about the provenance and excellence of the works. Casaubon never once really speaks for himself of what is before them.)

We have not mastered the use (the grammar) of aesthetic predicates if we apply them to works of art unqualifiedly and then, when queried, mention the testimony of others; but the grammar of predications of plain factuality is such that we unhesitatingly, unqualifiedly, pass on information we have got from others, not fretting that we learned of what we say from the words of others.

The Hearsay Principle does not work when the Ps are aesthetic; but a sort of variation on it applies to aesthetic authority: The Aesthetic Hearsay Principle: If b hears from a that P and b accepts a's authority, then b becomes willing (eager, depending on P) to seek experience (or avoid it) of the relevant work, more or less confident that he will come to judge that P for himself.

DIGRESSION

Why do I speak of ‘hearsay’ and not of ‘testimony’? Initially, reflecting on these matters I formulated what I called the Epistemic Principles of Testimony', which reads: ‘If a knows that P and tells b that P, and b believes a, then b comes to know that P.’ Having thus come to know that P, b is henceforth justified in claiming, in his own voice, without qualification, to know that P.'

This formulation, in effect, combines the hearsay principle with an explicit condition that a knows that P. That may be fair enough, but it does not make explicit that the default situation is that b accepts as knowledge what he is told. He hears a as claiming knowledge that P because a, in matters of plain fact, presents himself as knowing by unqualifiedly asserting.

Now I am hesitant to regard such interaction as testimony given and heard. Testimony primarily goes on in courtrooms and is delivered by people such as eyewitnesses and experts. Further, jurors are not expected, in the case of testimony to take themselves as having come to know that P just because they have heard it said and have no reason to doubt the witness. This may work when a witness is (merely) confirming the defendant's alibi of plain fact, such as that the he was in the pub, watching the cricket between 9 and 11 PM last Saturday night. But when such a witness is available there usually won't even be a trial, the prosecution having eliminated the cricket watcher from their enquiries.

When expert witnesses are involved, testimony is received not as knowledge that P, but as reason to believe that P. A juror or someone in the audience may come firmly to believe what the expert says, but he will hardly leave the court willing flatly, without qualification, to assert any of the Ps he picked up there. Or perhaps better to say, when testimony does produce subsequent flat, unqualified assertion by those who heard it, it will be only a special case of hearsay, special in that hearers are provided, by the formality of the context, with explicit assurance of reliability. Testimony is not, like ordinary hearsay, a matter of default. It would be blatantly—sociologically one might as well say—mistaken to say that we get the bulk of what we think we know and claim to know from testimony rather than from hearsay.

END OF DIGRESSION

Querying and responding to queries in ethics is similar to such querying and responding in aesthetics. Again let us imagine a case: a student whose judgment I trust tells me that a new colleague, Jones, is arrogant and condescending in the classroom. I decide to discuss the matter both with this colleague and perhaps with the Head of Department. I will not flatly say (to Jones): ‘You are arrogant and condescending in your classes and it should stop’ nor will I say (to the Head) ‘Jones is arrogant and condescending in his classes.’ I will preface what I say with ‘I have heard that …’ or, ‘Apparently …’. I may even say that ‘According to Smith (my source) …’ or ‘Smith tells me that …’.

The judgment in this story, one touching on character as well as conduct, shares with aesthetic judgment the feature that if I say it flatly, off my own bat, without qualification, inference is licensed that I have sat in on Jones's classes or overheard them and arrived on my own at the harsh judgment. This is what I meant above when I said that Wittgenstein was, in a way, right to say that ethics and aesthetics are one. Ethical and aesthetic judgments are alike in not slotting in to the Hearsay Principle. We do not acquire the confidence to assert flatly, unqualifiedly, in our own voices, the content of ethical and aesthetic judgments we have heard from other, no matter how confidently we accept what others say. It is necessary to characterize what we acquire in some way short of knowledge or certainty. Perhaps some kind of probability is in the offing. Respecting the judgment and circumspection of my informant in the case of arrogance in the classroom, I may well decide that some action is necessary; I have good reason to believe that Jones is arrogant in the classroom. What is striking is that even if I take myself to have found that Jones is arrogant in his classes, I will not make assertions to that effect without preface. The preface may be ‘I have good reason to believe …’.

Similarly with aesthetic judgments; having great confidence in the critic whose review I have just read, I may act accordingly, act with good reason to believe I will find the play, movie, painting, whatever, superb. I may, in such cases even say the work must be very fine, inferring this from the admiration manifest in the review. Such an inference will be straightforwardly inductive. But when hearsay works as it does for plain factual material, inductive inference seems not to be involved.

What we feel constrained to preface in the various ways I have indicated does not need to be so serious as in that story. Suppose my wife and I are especially fond of fresh tomatoes, as are all our friends and acquaintances in the neighbourhood. I arrive home from work and say: ‘The tomatoes at Anderson's are fabulous today’. If I am empty-handed, my wife may say: ‘Too bad you didn't pick some up; I didn't get out for any shopping today', inferring that I passed by the shop. If I had merely been told by one of the neighbours about the tomatoes, then, even though we have all come to rely on each other in these matters, I would still have prefaced my comments with, e.g, ‘Tom told me just now …’. This might go differently; if Tom is our regular and reliable source of information about the quality of local produce, makes a point of it every day to inform us as we walk by him on his porch with his late afternoon beer, I might, but just might, then speak without preface of the excellence of today's tomatoes. What this probably shows is that of all evaluating, grading things like apples and tomatoes is about as close as it gets to being plainly factual.

It is implausible to think that critics and works of art could ever get into the relation Tom and local tomatoes have to us in the story above. Critics can achieve positions of such authority that we will readily go to a film, play, exhibition, or concert, buy a book, etc., if they have praised it, go relishing what is ahead for us, or, alternatively never bother with anything they have rubbished. But we still do not reach the position of saying outright, in our own voice that the works involved are fine or good, or rotten, or whatever, without experiencing for ourselves the works in question. Prefaces prevail until we find out for ourselves, in the only way we can, by experiencing the works.

DIGRESSION

Given ongoing improvement in the quality of prints of paintings and photography of carving and sculpture, especially if a work is small enough so that in the print or photograph it is life-size, there can be experience of the work via the photograph or print which does adequately ground some aesthetic claims about it. Even so, supposing the photographs of the tympani and capitals at, let us say, Vezelay, to have completely convinced me that these works must be beautiful I should still hesitate to say flatly, as I do having seen them, that the carvings of Vezelay are ravishingly beautiful. I would thereby license the inference that I have been to Vezelay. Some kind of reproduction, documentary film perhaps, may eventually bring about something like what I am told holds with music: an accomplished musician can, I am assured, be moved deeply by reading a score, astutely remark the excellences and weaknesses of the composition. I have heard that Stravinsky preferred to read scores because he feared that live performances would always botch the music. And Brahms is supposed to have once said: ‘Go to the concert! Why? I have read the score!’

END OF DIGRESSION

Ethical judgments are like aesthetic ones with respect to how the authority of others plays its part. One's own judgment is required in both domains, the importance of judging for yourself and backing up judgment with your own reasons; or having the candour to acknowledge that the work just grips you or the action disturbs you. One of the forms of query: ‘What is F about it?’ is most understandable and appropriate in ethics and aesthetics. If I identify for you some animals in the zoo as hyenas, it would be very peculiar for you to ask what is hyena about them. But something like this can happen if you are relying on my zoological expertise to help you understand that those animals there are reptiles when the only reptiles you are acquainted with are snakes. ‘What makes them reptiles?’ can be a perfectly good form of query, and so can ‘What is reptilian about them?’ But in the terrain of the plainly factual, there are not words which fit readily and frequently into the form: ‘What is F about x?’ as do, ‘right’, ‘wrong’, ‘good’, ‘bad’, and many other predicates.

What I think has been shown so far can be put by saying that ethical and aesthetic Ps do not slot into the Hearsay Principle. Even when we happily accept the authority of another as a basis, indeed find ourselves willing to act, if action is called for, in accordance with the transmitted content, what is transmitted is not knowledge. In aesthetics, as I have said, we resolve to experience (or avoid experiencing) the works for ourselves, and we also give works further attention when we find our initial responses out of line with our source. We are willing to suppose we have missed something, not looked hard enough or long enough. We are also willing to pass on, with prefacing, the content of our authority's judgment. So there is a kind of surrogate of the Hearsay Principle, which I have suggested above.

With such a surrogate, we have to leave out talk of a knowing and b coming to know and note that what b comes to believe (even if very firmly) is that he will find the work a good one, or perhaps that it must be good, inferences from the evidence of the critic's claims and remarks.

In ethics, at least since autonomy has come to be thought central, (more about which shortly), things are similar. Someone, a, tells me that Jones has just behaved unscrupulously in some interaction, a kind of interaction I may be about to engage in with Jones. I do not come to know that Jones has behaved unscrupulously if that means, as it would if ‘Jones has acted unscrupulously’ were a plain matter of fact and my informant a trusted one; rather, in believing my source, I do not come to know what he told me to be so; rather I accept his claim in the sense that I now avoid Jones or enter warily into interaction with him. The most natural thing to say is that I acquire good reason to be wary of Jones, but reason that still falls short of making me willing flatly to offer as reason for avoiding him that he is unscrupulous, rather than the weaker reason that I have heard from someone reliable that Jones has acted unscrupulously in certain dealings. I may pass on advice to be wary of Jones, but it will be prefaced. What has been transmitted falls short of allowing me to pass on in my own voice, the proposition that Jones is unscrupulous.

Relevant here is the dubious idea of moral expertise. If the family are discussing, with or without his participation, whether euthanasia is what should be done with aged and senile Uncle Bill, it just doesn't do for one of us to say: ‘We should do it’ and respond to ‘What makes you say so?’ with ‘I called Peter Singer on the phone this morning and he assured me that it is the right thing to do'. I have to have understood Singer's reasons and made them my own. By contrast, if we had been discussing which car to buy and I said ‘A Toyota is the way to go; that's for sure.’ And was asked why I think so, I could say that I was talking to a top NRMA motor critic just this morning and he raved about Toyotas, saying no better middle-priced car has yet been marketed'. I do not, in such a case have to have an understanding of the factors that went into the expert's judgment; I do not at all have to make his reasons my own. Cars are like tomatoes in this respect.

This is not to say I cannot consult authorities or take advice in moral matters; of course I can, and am often well advised to do so. While moral expertise is a bizarre notion, moral authority is not, though the modern emphasis on autonomy raises problems about moral authority. I fail to be autonomous if, having done the deed, assisted Uncle Bill's suicide (say), I justify my action by saying that Peter Singer told me it is right or all right.

But here I want to share my puzzlement about something. If we look a little further at giving reasons for our ethical judgments and decisions, an interesting difference crops up between ethics and aesthetics. We have seen that it is incoherent to cite authority or testimony as backing for my claim that a work is marvellous when you query me with ‘Why do you think so?’, or ‘What is marvellous about it?'. ‘Robert Hughes says it is marvellous’ is out of court as a response to such a query. It is perfectly in order as an answer to the question: ‘Why are you spending so much time looking at that one?’ In ethics, however, there is no clearly analogous incoherence. We (I, some of us?) find it objectionable or irresponsible, when someone, having said that abortion is wrong, gives as his response to our query ‘Why?’ or ‘What is wrong with it?’, the answer: ‘My religion forbids it’ or ‘The Pope has decreed it sinful’. Some such reply as ‘It is the destruction of human life in order to make your own life easier’ may be argued about; but it is the sort of thing we must be offered as a reason if the opponent of abortion is to exhibit the autonomy we (I, some of us) have come to expect in ethical argument.

However, this seems not to be a matter of incoherence, as I have said; it is a moral, not a logical objection to our anti-abortionist. We can persist and ask: ‘What is it that the Pope finds wrong with it?’ But if we do that, our interlocutor is not incoherent if he is unable to give or does not fully understand the Pope's reasons. Obedience is a virtue for the religious (up to a point, it is a virtue for any of us if we are certain positions in certain institutions; in those contexts, it sits alongside loyalty as vulnerable to perversion, but nonetheless a good thing. Obedience's virtue does not depend on a grasp of the reasons behind the demands involved. A rifleman need have no idea exactly why he is suddenly ordered to shoot faster or an artilleryman to alter his direction of fire. It can be, in severe enough circumstances, ours not to reason why.)

I would also note here that we still conduct ourselves autonomously if we can no better than say we just don't feel right about some prospect of action or say, in explaining our reluctance to negotiate or cooperate with someone that we find him creepy. Or if someone says Hitler was worse than Stalin, it can be proper reply to ‘Why do you say so?’ with merely (?): ‘It just strikes me that way.’

While it verges on incoherence to cite Peter Singer or any other secular person as my authority for a decision or judgment, it seems to be coherent, even if objectionable (to me, some of us), to cite a religious authority. There is a parallel in the case of a child saying he mustn't do this or that and replying to ‘Why not?’ with merely (?): ‘Mommy says I mustn't.’ As I said, we may then want further to query the authority's reasons, (or Mommy's) but the religious person we query is not conducting himself incoherently by relying on his church and its authorities as he does, without his own proper grasp of the reasons lying behind a prohibition or an injunction. Maybe this shows that it no accident that Christians address God as ‘Our Father’ and the Church as ‘Mother…’.

This contrast between the aesthetic and the ethical provokes the thought that autonomy in aesthetics has been there from the beginning of the human engagement with art; but that autonomy in ethics is probably a recent development, the aftermath of enlightenment individualism.

I take autonomy to mean working it out for yourself, having your own reasons or, as is said, making the reasons you may have heard from elsewhere your own. That you have to do this in aesthetic judgment seems to be a conceptual matter, a requirement of reason. In ethical life, autonomy seems to enter as itself an ethical value; its absence is morally rebuked. If a query such as ‘Why do you oppose abortion?’ or ‘What is wrong with assisted suicide?’ is met with ‘The Pope has decreed against it’, the querier may object; but the objection can hardly be that reason has been offended. Difference here is ethical conflict.

Finally, I want to note that mathematics, ethics and aesthetics are alike in respect of the importance of working it out for yourself, to have done so or do so if you make flat assertions. A proposition such as ‘There are 4 primes between 10 and 20 does not work in the Hearsay Principle. Even if I believe you when you tell me such a thing, I will still, if I later assert it to another, not respond to the query: ‘How do you know that?’ or ‘What makes you say so?’ by citing a source, as I would with what I have called a plainly factual matter. In making such mathematical claims, easy to back up or difficult, those who assert them are expected to be able to do their sums on their own, to back up what they have asserted. So mathematics is similar to ethics and aesthetics both in this respect and in the respect that the negations of true mathematical propositions are not possibly true. Mathematical propositions lack bipolarity.

There is an apparent exception. There is now a proof of Fermat's Theorem. I know that from the newspapers. But it is a plain fact, or at least a bipolar proposition that there now exists such a proof. Only extreme Platonism will say it has been there eternally, waiting to be discovered. If such Platonism were correct, then ‘There is not a proof …’ would be a necessary falsehood. But it is a contingent fact that a proof has come into existence; it could have failed to happen. And perhaps the proof is still being checked by some sceptics so that it makes sense to say ‘There could, after all, not be a proof.” So whether there is or is not a proof is a matter of fact. It does not follow that the proof, a sequence of mathematical propositions, corresponds to some cluster of mathematical facts. There are no such facts, only truths, and necessary truths to boot.