R.E. Ewin and Alan Tapper
Julius Kovesi was a moral philosopher contemporary with Alasdair MacIntyre, and dealing with many of the same questions as MacIntyre. In our view, Kovesi’s moral philosophy is rich in ideas and worth revisiting. MacIntyre agrees: Kovesi’s Moral Notions, he has said, is “a minor classic in moral philosophy that has not yet received its due”. Kovesi was not a thinker whose work fits readily into any one tradition. Unlike the later MacIntyre, he was not a Thomistic Aristotelian, nor even an Aristotelian. Nor was he in any obvious sense a liberal. He saw his viewpoint as Platonic, or perhaps more accurately as Socratic. Unlike MacIntyre, his writings have little to say about justice. However, he did offer a theory of practical reason. His main contention was that all human social life embodies a set of concepts that govern and guide that life, concepts without which that life would be impossible. These include our moral concepts. Moral concepts are not external to, but constitutive of social life.
Moral Notions is a short book, and, while it is about moral concepts, one might fairly say that there is in it not much discussion of any particular moral concepts. Those of us who think it an important book have the task of showing how its tersely-made arguments can be expanded and applied, with fruitful results. This is what we will attempt to do in this paper, drawing upon some well-known work by MacIntyre, and showing what we think is both Kovesian in spirit and intellectually constructive in MacIntyre’s work. We will suggest that in places these discussions could with profit be even more “Kovesian” than they already are. We will take two examples of moral or partly-moral concepts, in each case beginning with MacIntyre’s discussions: patriotism and lying. Naturally, we are not pretending to give a complete analysis of these concepts. Our point is to show in what direction we think further discussion needs to go.
Kovesi’s theory of meaning emphasised that concepts are to be understood in terms of their points, or, in his terminology, their formal elements. The point or formal element of a concept is the reason or reasons we have for forming and using that concept. The reason or reasons play a role in our socially-shared lives. This role will vary as circumstances vary. In one context concept X will apply in such-and-such a way; in another context it will apply in a very different way. The applications are what Kovesi called the material elements of the concept. His key point was that a concept will have one and the same point or formal element in all of its applications. What counts as a table or a murder may well vary dramatically from case to case, while remaining a table or a murder in each case. To see the sameness across diverse cases requires us to think in terms of the point of the concept. If we fail to do that, we see only the diverse material elements and thus will too easily conclude that there is no general concept.
Kovesi’s account of concepts was in some ways familiar to philosophers of his time. Something of this theme was perhaps to be found in Wittgenstein’s account of language games, though Kovesi shows how Wittgenstein himself failed to get his own point when he formulated the idea using the “family resemblance” metaphor. (Wittgenstein, Kovesi thought, “is still looking for empirical similarities between A and Z though it is not one thread that runs from A to Z”.) More to the point, J.L. Austin taught his students (Kovesi amongst them) that we understand a concept fully only when we grasp a variety of examples and instances. Kovesi himself thought that his own theory of concept-formation is to be found in Socrates and Plato. Plato’s Theory of Forms, he thought, was an attempt to talk about how concepts can remain one and the same while also having various subdivisions and many diverse instantiations. This was not a matter of the so-called problem of universals – the problem of how the one concept can apply to many instances. It was a different problem altogether – the problem of how one concept can have many diverse applications. It was this that was central to Kovesi’s thinking.
Kovesi distinguished between ‘complete’ and ‘incomplete’ moral concepts, giving murder and cheating as examples of the first, and lying as an example of the second:
when x is a moral term, the judgment ‘x is always (or sometimes) good’ tells us about the logical and conceptual features of the term x; it tells us whether the term specifies an act from the moral point of view and to what extent it does this. When the term is a complete term, complete from the moral point of view, then ‘right’ and ‘wrong’ function like reminders, they signify that our term has been formed from the moral point of view. When our term is incomplete, or open to further specifications from the moral point of view, then we use ‘right’ and ‘wrong’ to discriminate and distinguish from the moral point of view between different instances of the act referred to by the incomplete term.
We take patriotism and lying to be incomplete moral concepts. Thus we will be applying and defending Kovesi’s distinction, one not yet sufficiently recognised by moral philosophers, MacIntyre perhaps included.
Lying as an incomplete moral concept
In his two part discussion of lying and truth-telling MacIntyre’s method is to start from moral theories, those of Kant and Mill. Towards the end of the discussion, he offers an account of the ethics of lying that, he says, contains elements of Kantian morality while also agreeing with elements of J.S. Mill’s views on the subject. We are not here wishing to discuss Kant or Mill. We will focus on MacIntyre’s proposed account, which we see as similar to Kovesi’s.
Truthfulness is good and lying is wrong, MacIntyre says, for three kinds of reasons:
- lack of commitment to truthfulness can corrupt and destroy the integrity of rational social relationships;
- truthfulness is necessary because it makes trust possible, especially in the giving and receiving of criticism of existing social practices;
- and truthfulness is a virtue that helps us to hold in check the power of phantasy, which “can be and often is used to disguise and to distort our activities and our relationships and has the effect of deforming them” (137).
We may wonder whether the second and third reasons are really distinct. But, that aside, the point being made can be summed up in the idea that “in any relationship in which the goods of rational persons are to be achieved, the truthfulness of those participating in that relationship will be of crucial importance” (137). The point will therefore extend to relationships with strangers.
MacIntyre observes that this account of the evil of lying relates the evil to other evils: “it is a salient characteristic of moral evils in general that they are destructive of rational relationships” (138). He then contends that this account of the evil of lying also explains why lying is sometimes justified. Here he introduces two examples. One – a Dutch woman shielding a Jewish child from Nazi investigators – is an example of justified lying. Another is the case of justified killing, in which a mother is protecting her child from an aggressor. The cases illustrate the same moral point: truth-telling is good for the same reason that lying is sometimes justified – both rest on the value of rational social relationships. Lying is not wrong when it is necessary to protect truthful relationships against aggressors. In those circumstances, lying is a duty if the relationship being protected is one that I have a duty to protect, such as a mother to her child, or a guardian to her ward. This account, he thinks, is not at all ad hoc. It is such that it “both generally and indeed almost always prohibits lying and yet requires it on certain normally rare types of occasions” (134).[1]
To explain the good of truthfulness, MacIntyre situates it in social practices, including the practice of social criticism. He thus provides what Kovesi mostly took care not to provide. In Kovesi’s terms, MacIntyre names the formal element of the concept of lying, and of morality more generally. The moral is the protection of rational social relationships. Social practices, critically considered, are what morality is designed to promote and defend. (See Ewin on this.) Thus, we might suggest that Kovesi provides the theory of meaning that helps make sense of MacIntyre’s moral argument.
Compare Kovesi’s approach:
The sense in which a complete moral notion provides us with a principle is the sense in which it enables us to say that the following two are examples of the same act: saying what is not the case in order to bring punishment on someone for an act for which he is not responsible, and saying what is not the case in order to gain a benefit to which I am not entitled; but on the other hand saying what is not the case in order to save the life of an innocent from a maniac is not an example of the same act. (MN, 80)
Here the ‘principle’ at work involves ideas of responsibility and entitlement – that is, presumably, rights and duties.
Kant’s rigorism about lying was famously challenged by Benjamin Constant, who counter-argued that it is permissible to lie to a murderer because the murderer has no right to the information he seeks. On this view, the morality of lying is determined by the rights and duties of the situation. MacIntyre holds a similar view of the permissibility of lying. The morality is determined by the duties of the situation. Both Constant and MacIntyre appeal to “Kantian” concepts of rights and duties. Putting their claims together, we might say that the morality of lying rest on the justice of the situation. One must not lie when to do so is unjust. One may lie when to do so is not unjust. One must lie when to do so is justly required of one.
As in the passage above, Kovesi focuses on the formal element of the concept of lying. Reflection on examples shows us that it is somewhat complex. Lying is commonly used to cover up wrongs already done and to facilitate the doing of future wrongs. In these sorts of cases lying is plainly wrong. Its formal element is injustice, and it is morally speaking in the same category as the wrongs with which it is associated, such as murder, etc. Clearly, perjury and slander are sub-categories of this kind of lying. Lying differs from common injustices in that it involves falsehood, but the falsehood only makes the lying possible; it doesn’t add to the wrongness.
Kovesi’s account agrees with MacIntyre’s in contending that lying is not always wrong. It is not wrong in the case when the falsehood is used to prevent an injustice, as when it is “saying what is not the case in order to save the life of an innocent from a maniac”. This is Kovesi’s category of “savingdeceit”. The life saved is not that of a wrongdoer being sheltered from police; it is the life of an innocent being sheltered from a wrongdoer. Sheltering a wrongdoer may indeed be a way of saving a life, but it is not the point of Kovesi’s example. The point is the prevention of injustice. The saving is saving a life threatened by injustice. The lie in that case is serving the ends of justice, and that i what makes it not wrong – not even slightly wrong. The liar my regret having to lie, but he should not regret lying, since the act itself is not in the slightest degree wrong.
There is more to the morality of lying than simply the matter of justice and injustice. The potential for lying enters into all aspects of social life and thus into all aspects of morality. For example, lying as an act of kindness can’t be covered by considerations of justice and injustice. In the case of “kind lies” we have already got a counterpart of Kovesi’s “savingdeceit”, in the concept of a “white lie”. The example illustrates Kovesi’s idea of the formal element of concepts. The notion of a white lie looks like a contradiction in terms: wrong if we see it as lying; not wrong if we see it as “white”. But the wrongness is illusory. No injustice is done by a white lie; no unfair advantage is taken. The act is, ex hypothesis, an act of kindness, and kindness is a species of goodness.
The apparent contradiction is not real, but only if the kindness really is kindness, and that presupposes a recipient whose feelings really are vulnerable. Telling a white lie to a person quite capable of hearing the truth is not succeeding in performing an act of kindness. Telling a white lie to a colleague or an authority in the normal course of business – for example, commenting too kindly on their latest draft paper – is to let them down and weaken your relationship. It is therefore not a white lie. Failing to tell a lie when a lie is necessary to prevent a wrong, out of concern for the feelings of the would-be perpetrator, is to be doubly incompetent from a moral point of view. The kindness is misdirected, and is thus not kind; and the consequence is to aid a wrong, which, morally speaking even if not legally speaking, is itself a kind of wrongdoing.
Telling a lie in order to sell a product or a service is neither a white lie not a case of savingdeceit. Often such lies are harmless, but they can deceive and thereby harm the gullible. Sometimes the lies are plainly rudiment, in which case they are injustice. In both types – harmless and fraudulent – the wrongness is plain enough. That such lies are so common in advertising arises from the fact that the consumer can punish the lie-teller only by not buying the product being spruced, but sometimes the product is worth having even despite its deceptive presentation, so the deception often goes unchecked, but tis wrongness, minor though it is, remains wrongness. The wrongness is not minor when the consumer’s health or safety is at risk, and in those cases even small lies can be serious wrongs.
Falsely boosting one’s wares is similar to falsely boasting about one’s abilities. The idle boaster does no wrong if he is seen for what he is. But he can’t be trusted with anything important. And if his boasting is sufficiently skilful he will do wrong, because he will succeed in engaging our trust undeservedly.
To sum up, there are three general points here. Firstly, Kovesi’s idea of savingdeceit is not particularly novel. It is formed from the same pattern we applied in forming the concept of a white lie. The difference is that savingdeceit is governed by justice as its formal element, whereas the white lie is governed by kindness.
Secondly, Kovesi is right that lying is an incomplete moral concept, and right too in pointing to the role of complete moral concepts in our moral life. Again, we already have such concepts in the area marked out, incompletely, by the concept of lying. These are terms such as bearing false witness, perjury, libel, slander, and fraudulence. It is these terms that pick up the ways in which some forms of lying are always wrong, or wrong in themselves, just as murder is always wrong or wrong in itself. “Murder” is a complete moral concept, and so are the terms for these kinds of lying. Given the roles of these complete moral concepts, we should have no need to construe lying as a complete moral concept.
Thirdly, the morality of lying is complex, but its complexities simply mirror the complexities of social lie. There is nothing surprising in this complexity and no special rightness in truth-telling or any special wrongness in speaking falsely. Kovesi’s’ “conceptual functionalism” has one advantage over MacIntyre’s approach to the problem: it takes the focus off what MacIntyre call truth-telling. Cooperative social life involves cooperative talking and telling, and this is what ultimately governs the morality of lying. One can speak falsely while playing a cooperative role, as when acting in a play, or as when deceiving a wrongdoer. Or one can tell the truth non-cooperatively, as in seeking to hurt the feelings of a vulnerable person, or as in telling the wrongdoer where to find his intended victim, or as in randomly reading sentences out of a reliable encyclopaedia. Truth-telling is only distantly related to truthfulness and honesty. It is normally good, as it plays a large part in normal social cooperation. But, apart from when it is playing that part, it is simply neutral. One who walks around idly speaking truths (“five fives are twenty-five”, “‘Sucre is the capital of Bolivia”) is doing nothing admirable, no matter how truth the truths might be. A person is not the less honest or truthful because they tell jokes, sing in the opera, or write poetry. This is the third key point in Kovesi’s account of lying. False-speaking is at most a material element of lying; it in no way explains the wrongness of lying. Likewise, truth-telling is good only in so far as it plays a role in cooperation, and even then it is only a material element of the relevant morality, which itself requires distinctive moral concepts, such as honesty and veracity, to mark it out. From a Kovesian standpoint, one can see that false-speaking and truth-telling are not even moral concepts.[2] These descriptions are morally neutral. The moral issues do not arise at the level of speaking falsely.
This is to explain morality in a top-down way, and it presumes we can give some content to “morally wrong”, so there is the danger of begging the question. But if so, approaching the question by the bottom-up method (as MacIntyre partly does) clearly won’t do either – it presupposes that we can demarcate the moral merely by whittling away at the material elements. To do this we have to have a sense of what we are aiming at – that is, we need what Kovesi called the moral point of view.
Taken overall, MacIntyre’s account is closely similar to Kovesi’s discussion of lying in Moral Notions. There are dissimilarities also, and some of these are minor. Yet the overall similarity is far from obvious, because their methods of argument are very different. This difference makes the comparison and interesting one, and perhaps an illuminating one. Kovesi’s method of argument is to start from the general question of what concepts do, and then of what moral concepts do. We can call this method “conceptual functionalism” just for convenience. (Kovesi’s originality in developing this method was noted by Philippa Foot.)
As noted above, Kovesi held that moral concepts are not external to, but constitutive of social life. MacIntyre contends very similarly that morality is focused on the protection of rational social relationships. However, MacIntyre’s moral philosophy – unlike Kovesi’s – looks in two directions. In one direction, it emphasises the diversity of moral outlooks. Moral theory today, he thinks, must deal with apparently irreconcilable differences. Not only are there such differences, but they extend to the question of how to describe the differences. Thus, there is a strong argument for moral relativism, a relativism that the work of moral philosophers has not diminished but rather deepened. In the opposite direction, he contends that amongst the various moral traditions, there is one and only one tradition that has a strongly plausible strategy for dealing with these moral differences. That is what he has described firstly as the Aristotelian and later as the Thomistic Aristotelian tradition. That tradition, he argues, can provide two vital things missing in rival traditions: a theory of justice, and a theory of practical reason. But if we take the relativistic line, we may cease to have a story of the nature of rights and duties and thus MacIntyre’s account of lying may fall down. One may also question whether the idea of a social practice can be sustained without an account of justice.
From Kovesi’s standpoint, the diversity of moral concepts can have only limited scope. If he is right, then diverse moral stances are possible only because we already have in place a more basic set of moral concepts. It may be that in some cases moral problems are actually insoluble. But what cannot be true is that we lack a set of relevant terms in which to debate our differences. If we lacked those terms we could not even recognise the differences as differences. More basically, we would not even exist in order to have the debates about our differences.
[1] MacIntyre suggests (133) that, for a Kantian, while one may not lie to a would-be murderer, one may distract him or remain silent, or “trip him, knock down, or otherwise hinder” him. These acts, presumably, must be regarded by Kantians as morally neutral, while lying is always wrong. But knocking someone down is assault, and assault is prima facie wrong. In Kovesi’s terms, it is a complete moral concept, like murder, rape, and theft. So if lying is a complete moral concept, and assault is also, why is assault of a would-be murderer not on a par with lying to achieve the same end? MacIntyre’s own view is that lying, assault, knocking down, and even killing are permissible, since killing an assailant to protect one’s child is permissible.
[2] They are not less virtuous even if they intend to deceive us with what they say (contra Kovesi). Sissela Bok remarks that “lying requires explanation, whereas truth ordinarily does not” (3). But random truth-telling makes no sense and relevant truth-telling requires no further explanation, just because it is relevant. Lying, however, requires a special sort of reason take away any culpability. There is no useful contrast between lying and truth-telling, only a contrast between cooperative and uncooperative speech.