Jeffrey Stout’s Ethics after Babel is, in his own phrase, a ‘philosophy of moral diversity’ – of the sheer foreignness to some people and societies of the moral thinking of some other people and some other societies. Once satisfied that moral diversity is a fact, many philosophers despair of ethics: becoming either moral sceptics, and doubting that true answers to questions about right and wrong can be found by taking thought, or moral nihilists, and denying that moral questions have true answers. Others like Plato, conclude that ethics remains possible, but only by rising above human moral diversity to ‘a God’s-eye view’ of right and wrong. Stout’s point of departure is that to aspire to such a view is vain: it is the ancient sin symbolised as the building of the Tower of Babel. Yet he does not despair. Clearing Babel-like projects out of the way, he directs us to what, in his mind, is an acceptable alternative to them.
Stout on the sin of Babel is reminiscent of the Christian Gospels on the sin against the Holy Ghost: by leaving obscure what the sin is he makes worse the bad news that there is no forgiveness for it. Aquinas asserted that the whole of what he called ‘the natural law’ is embodied in the Ten Commandments revealed by God to Moses, and that human reason can both ascertain what its principles are, and derive from them its more general subordinate precepts. He also described the natural law as ‘a participation in the eternal law by rational creatures’ – the eternal law being the law by which God rules all things. Did he aspire to a God’s-eye view of right and wrong, and so unwittingly commit the sin of Babel? Not according to Stout; for he is chosen (along with Jefferson and Martin Luther King) as an example of a moralist who has gone about his work in a non-Babel way. On the other hand, Kantians and utilitarians are branded as sinners, although it is left dark how they offend if Aquinas did not.
Stout sheds a little light on this question in elucidating what it is for a moral proposition to be true: ‘I have no trouble,’ he writes, ‘with the idea of a culture-transcendent Moral Law if it commits us merely to such notions as ... that there are moral truths ... The problems come when we appeal to the Moral Law, in our explanations or criteria of moral truth, in a way that requires denying that we are necessarily employing culturally embedded categories as we do. For then the implied sense of the phrase “culture-transcendent Moral Law” is: “a set of moral truths one could know or understand without making use of a moral vocabulary”.’ Stout simply does not see how you can be a Kantian or a utilitarian unless you believe that moral propositions are true or false according to whether they correspond or fail to correspond to a set of moral facts one could know or understand without making use of a moral vocabulary.
Why does he imagine that Kantians and utilitarians are committed to believing that? As far as I can tell, it is because he assumes, without explicitly saying so, that if you believe that there are moral truths (that is, reject moral nihilism) you have only two options: either to posit language-transcendent entities for those truths to correspond to, or to embrace a pragmatist doctrine he proceeds to put forward – a variant of the conception of truth as warranted assertibility. According to Stout, all there is to explain about the concept of truth is explained by the fact that ‘truth, for us, here and now, is always warranted assertibility.’ Unlike less sophisticated pragmatists, however, he denies that truth can be formally defined as warranted assertibility. It would be better, he submits, to have no theoretical definition of it at all. Since all Kantians and utilitarians known to me reject this variant of pragmatism, if he is right about their options, they can do no other than posit language-transcendent entities.
It is a truism that, wherever in culture and history we may find ourselves, we form our beliefs solely in view of our ‘epistemic context’: the ‘reasons and evidence’ there available to us. Stout’s position is that what we are warranted in asserting here and now is relative to that context, although he has failed to notice that neither we nor others know exactly what it is. Consider a thinker X, who did not make an advance which a thinker Y made a little later. Certainly X’s epistemic context did not make Y’s advance available to him as an advance already made, but comparison of his thinking with Y’s shows that it did make it available to him as a possibility. Yet X was not aware of that possibility, nor would later historians be aware of it unless Y or somebody else had taken advantage of it. Unless every possibility in a given epistemic context will one day to be taken advantage of, nobody can ever know what all the resources of that context are. Yet Stout claims that if I am justified in accepting a proposition, then ‘the proposition, my context and I’ are related in a way that is ‘as objective as can be’ and ‘not subject to worrisomely arbitrary subjective manipulation’. On the contrary, if truth, for us, here and now, is always what we are warranted in asserting by our epistemic context, then it is in principle unascertainable.
Stout does not notice this, because he identifies our epistemic context, not with the possibilities to be found in our real situation by using our wits, but with those historians may afterwards conclude were available in it. He then proceeds to treat this manipulable historical construction as an epistemic barrier limiting what we can think. That would be bad enough: but worse may follow. Although we cannot know what future historians may judge to be available to us, we may try to anticipate it. In so trying, since we cannot simultaneously think and study our thinking historically, we may take as a working hypothesis that the difference is negligible between what is available to us now and what was available to us in our recent past, which we can so study. And so we may persuade ourselves that we can never do better than ask what, in view of what we believe was available in our recent past, we are justified in thinking now. Instead of releasing us from the ivory tower of Babel into the world of action, a pragmatism on these lines would imprison us in the manipulable constructions it calls our epistemic contexts.
Yet this, as far as I can tell, is the descensus Averni Stout has followed. He has found, in the slang of the French intelligentsia, the perfect word for what it leads to: the word bricolage. He glosses it as meaning ‘what bricoluers do with ... collections of assorted odds and ends, namely, put some of them together to serve the purposes of the moment’. Whenever a question arises in any branch of inquiry, bricoleurs will consider what odds and ends of information, theory and method are available to them from the thought of the recent past, and will put them together to yield an answer in whatever ways they foresee will win the approval of their contemporaries. As fashion changes, styles will oscillate from reactionary to revolutionary, and the period of oscillation will vary. My description of bricolage has probably betrayed my distaste for it. Yet Stout relishes it, even though his own descriptions of it closely resemble mine: ‘the creative intellectual task of every generation,’ he tells us, ‘involves moral bricolage’.
In the hands of a decent and studious exponent, as Stout is, moral bricolage will yield little that most readers of the LRB will object to. His own ethical projects are those of a person of liberal mind, who has both studied recent work in the mainstream social sciences and pondered radical criticisms of how contemporary Western societies function. Effective social criticism, he declares, must go into ‘the uneasy relation between [value-generating] social practices and such institutions as the capitalist market-place and large-scale bureaucracies’ and he urges that social practices and institutions be considered together. In societies with liberal political institutions (even though such institutions are fragile and call for ‘unflagging vigilance’ on the part of their citizens), he believes it possible to bring the capitalist market-place and bureaucracies into sufficient harmony to increase the production of wealth, while fostering value-generating social practices. Of course, no such harmonisation will be final. The social fabric is a ‘coat of many colours, constantly in need of mending and patching, sometimes even recutting and restyling’. In carrying out these repairs, he adds, ‘we can make good use of Aristotelian and civic republican talk about the virtues and politics as a social practice directed toward the common good without supposing that this sort of moral language requires us to jettison talk of rights and tolerance. We can use this talk by thinking of liberal political institutions as oriented toward a provisional telos – a widely-shared but self-limiting consensus on the highest good achievable ... But this telos justifies a kind of tolerance foreign to the classical theological tradition.’ I have no quarrel with this.
Unfortunately, moral diversity is lost to view in these worthy prescriptions for social criticism. Since for the most part Stout avoids issues to which, in Western societies, morally diverse approaches are in fact taken, the whole of his treatment of social criticism throws less light on what moral bricolage can do to resolve such diversities than his brief comment on one of them: ‘Disagreement over an issue like abortion,’ he remarks ‘might ... be compared with disagreement among cosmologists over the origin of the universe. It’s not clear what, if anything, is going to settle these disagreements, although neither one seems impossible in principle to resolve by rational means.’ What is revealing about this comparison is its falsity. Cosmologists disagree about the origin of the universe, not because they disagree about the relevant physical and chemical theory, but because they do not have the observational data they need. In radical contrast, both sides in the disagreement about abortion are confident that the necessary principles have been established, and nobody to my knowledge has even suggested that additional medical information (physical or psychological) will make any difference. There is nothing to be done except directly to investigate the moral principles at issue. But that would be the sin of Babel. The odds and ends to which a bricoleur is confined are the principles and arguments of both sides in the dispute as it has been conducted; and since some of them are contrary to others, nothing put together out of them would be warranted by them as a whole.
If, as I have tried to show, Stout’s variant of pragmatism will not do, are non-sceptics and non-nihilists compelled to accept the only other option he will allow them: that of positing linguistically transcendent entities for moral truths to correspond to, like G.E. Moore’s non-natural facts? Not at all. True thoughts are expressible in true sentences; and just as the sentence ‘It is true that snow is white’ is logically equivalent to ‘Snow is white,’ so it is for all true or false sentences. As used to be said in the Sixties, you speak truly if you ‘tell it how it is’. The reality with which we are concerned is one we interpret through language. No sentence can be either true or false unless it reaches what is, and tells either how it is or how it isn’t. As Donald Davidson has pointed out in Inquiries into Truth and Interpretation, ‘in giving up dependence on the concept of an uninterpreted reality ... we do not relinquish the notion of objective truth, quite the contrary ... Truth of sentences remains relative to language, but that is as objective as can be. In giving up the dualism of [conceptual] scheme and world, we do not give up the world, but re-establish unmediated touch with the familiar objects whose antics make our sentences and opinions true and false.’ For this reason, as Davidson has also tirelessly argued, your belief in a sentence’s truth is evidence (not by itself conclusive) of what you take it to mean. Hence, if we want to know what a utilitarian means when he utters such a sentence as ‘It is wrong to do what will not promote the greatest happiness of the greatest number,’ the way to find out is to inquire how he thinks things would be if it were true, and how if it were not.
A source of confusion in contemporary ethics is that different utilitarians answer this question differently. One, a follower of J.S. Mill, may say that if it is true that it is wrong to do what will not promote the greatest happiness of the greatest number, then things are this way: not doing what will promote the greatest happiness of the greatest number is repugnant to the feelings of the majority of cultivated people who associate themselves with the direction in which modern civilisation is moving. Another utilitarian, influenced by the Kantian tradition, may object: ‘No, it is not true unless things are this way: not doing what promotes the greatest happiness of the greatest number is contrary to what human beings who think practically and competently unconditionally prescribe’. Other answers have been given – including ones that Stout might fairly object to – but these suffice to show that neither utilitarianism nor Kantianism as such is guilty of the sin of Babel.
Most of the Kantians and utilitarians whom Stout denounces for not perceiving that ethics is moral bricolage think of their principles as truths not about a linguistically uninterpreted reality, but about linguistically interpreted objects in the world – human beings either in so far as they have feelings of a certain kind, or in so far as their rational capacities are practical. Instead of impiously aspiring to divine knowledge of transcendent reality, they claim only to find good reasons for certain beliefs about earthly things. They are not vainly building a Tower of Babel. Even so, as Aquinas pointed out, if God exists, he cannot well be ignorant of good human reasons for moral beliefs, any more than he can be ignorant of good human reasons for believing that 2+2=4.
In denouncing sin it is hard not to depict it as more evil than it is. Stout has not merely reprobated the innocent, his fervour in doing so has carried him away. Even if it were a sin not to be a moral pragmatist/bricoleur, he should not have embroidered his denunciation of that sin by two further charges, both plainly false, and the second distinctly odd.
The first is the charge of ‘foundationalism’. Referring expressly to my version of Kantianism and Kai Nielson’s of utilitarianism, Stout implies that no moral system in which specific precepts are derived from a fundamental principle or principles can ‘avoid vicious circularity’ unless those principles are ‘established on grounds that do not presuppose the acceptability of some or all the system derived from [them]’. In short, the principles of a moral theory cannot be morally fundamental unless they are epistemically fundamental. The absurdity of such criticism becomes obvious as soon as it is applied to a non-ethical deductive system – say, that of Newtonian mechanics. It is intellectually impossible to get on in elementary mechanics without assuming what Newton called his ‘axioms or laws of motion’, and applying them to new cases; yet those cases are part of the evidence by which Newton’s system as a whole must in the end be justified.
The circle in reasoning here is not vicious. When a deductive system like Newton’s (or I would say Kant’s) has shown that it can make sense of a domain of phenomena in a way that its rivals cannot, you are entitled provisionally to assume its truth, and apply it to new cases. Yet if for any reason the result of such an application is called in question, the system itself is called in question. How the question will be resolved – whether by accepting the questioned result, finding a mistake in application or by correcting the system – will differ from case to case. Kantian moral theory is no more foundationalist than Newtonian mechanics, although the evidence pertinent to deciding its truth, since it is about operations of practical reason and not about the motions of bodies, is obtained by studying how competent human beings think practically, and not how bodies move. Hence it is an epistemological mistake that deductive moral systems must be foundationalist. Morally, specific precepts derive from principles: but epistemically, a moral system setting out how those precepts are derived from those principles is shown to be true as a whole by its capacity to make sense of the domain it is about.
Stout’s second (and odd) supplementary charge grows out of a joke first cracked in his third chapter, which is regarded as good for a laugh in the dull spots thereafter. The diverse varieties of moral judgment are expressed, in any natural language, by fragments distinguishable in it. There are comic possibilities in referring to such fragments as ‘moral languages’. Keeping a straight face, Stout speaks of Mario Puzo’s readers as learning a new ‘moral language’ when in The Godfather ‘they encounter for the first time the fragment of English by which his Mafia characters express their Mafia-style moral judgments.’ He then plays for laughs by arguing that since a ‘moral language’ containing terms of art must be an artificial language, an ‘Esperanto’, studying Kant’s moral theory is learning a ‘moral Esperanto’, and Kant himself is a ‘moral Esperantist’. I assume that readers are expected to credit him with knowing that the fragment of Kant’s German needed for expressing his ‘metaphysics of morals’ cannot be a language, because the words in that fragment have no sense independently of the senses of German words that are not in it; and also to credit him with knowing that Esperanto is a language into which Ethics after Babel could be translated. The cream of the joke, however, is how the audience reacts: the spectacle of Kantians like myself explaining why talking of such things as human dignity and ends in themselves is not talking Esperanto. Well, I hope it has been funny.