[A reply to this article can be found here]

The following comments were written in response to an unsigned editorial in The Pragmatist (volume 14, no. 2). The editorial appeared under the title 'The Necessity of Pragmatism' and defended a number of theses about the libertarian philosophy of personal freedom. The body of the editorial is devoted to the concept of truth or rather to the question whether the concept of truth is adequately defined by what is called 'the standard dictionary definition'. The latter is identified as the statement that truth is the agreement of ideas with reality. It is later referred to as the ordinary-language, consensus, common sense and lexicographical meaning of the word 'truth'. The author argues that this definition is incoherent and that only the pragmatism of William James really resolves that incoherence. He goes on to suggest that pragmatism is a 'very adequate foundation for libertarianism' and that it is a critically enhanced common sense philosophy. Consequently its critics must be either uncritical superficial people or else producers of metaphysical phantasies.

I shall try to make clear as briefly as possible why I do not think that James' pragmatism posits a serious challenge to the standard definition of truth and why I do not see it as a sound foundation for libertarianism.

The editorial starts with a delightful story about William James and a group of hikers discussing a scene in which a person moves around a tree in a vain effort to catch a glimpse of a squirrel that manages to keep out of sight by always hiding on the other side of the tree. The question arises whether the man moves around the squirrel as well as the tree. Some deny that he does because the man never gets behinds the squirrel's back. When James appears, he remarks that the answer to the question depends entirely upon what one means by 'going around'. And there's no doubt it does. Unless one understands a question or a statement, one cannot begin to answer the one or assess the truth-value of the other. To say, for example, that an author is skating on very thin ice may be to make true statement, even if there is no ice anywhere near the author.

James' intervention makes it clear to the hikers that their discussion is the result of a misunderstanding: each of them believes the others to understand the question in exactly the same way as he does, whereas in fact they assign different meanings to the words used to express that question. They are not giving different answers to the same question, but to different questions. Consequently there is no 'real disagreement' in which one person says that something is true and the other says that the same thing is false.

Assessed in this way, the story about the hikers and the squirrel-problem is an instructive one. Its moral is that at least some disputes are evidence of terminological confusion rather than of disagreement about the facts.

As the editorial notes with obvious approval, 'James generalised the method he had employed to keep the peace on this camping trip, and called the method pragmatism'. What is that method and where does its generalisation take us? It certainly does not take us to the point where we should say that all disagreements are always about words and never about facts or 'real differences'. However, the editorial mentions another lesson we should remember from James' story, namely that 'one can best resolve such non-terminological disputes if one asks what practical difference the distinction makes'. Again that is wise advice. If the answer to a question does not make any practical difference whatsoever to you, why bother with it to the point of making enemies? Of course, most of us know why that advice, wise as it may seem to people who have no axe to grind with respect a particular question, usually fails to impress people who do have a strong interest in it. And I don't mean merely a strong intellectual interest. Their self-esteem, vanity and reputation, to say nothing of their income and power, may depend very much on it. For them, the practical differences of resolving the dispute in one way or another may be enormous.

I have no quarrel with James' advice as far it goes. However, what bearing does it have on the question of the nature of truth? The answer: none whatsoever! The author of the editorial fails to see this because he apparently assumes that to define truth as an agreement of ideas and reality is to make a metaphysical statement. Furthermore, he appears to assume that the only way to check the truth of that statement is by first specifying what the word 'agreement' means and then to check whether there is something in reality that 'agrees' with that idea of agreement. Hence his conclusion that this procedure gets us nowhere. However, it does not support his criticism of the standard definition of truth.

There is nothing logically wrong with saying that truth is an agreement or correspondence between what is said or thought and reality. However, as the editorial amply demonstrates, the word 'agreement' (or 'correspondence' or 'mapping' or some other word used in its place) may mislead some people who fail to grasp its role in such a statement. Consequently it may be better to avoid it. That is indeed possible and easy to do. We do not need the word 'agreement' to describe or explain the concept of truth. The statement that I am bald is true if and only if I am bald. If I make the statement in French, saying 'Je suis chauve', it is still true if and only if I am bald. The statement that the man in the story goes around the squirrel (in one sense of going around) is true if and only if the man goes around the squirrel (in that particular sense of going around). To state a truth is to state what is so. That is all there is to the concept of unqualified truth.

It is possible to qualify the notion of truth in various ways or to construct analogues of it that define particular concepts of truth as, say, agreement between an idea and some particular object or set of conventions. Thus, the statement that Hamlet was the Prince of Denmark is true within the story dramatised by Shakespeare, regardless of whether it is a true statement about the history of Denmark.

To grasp the concept of truth does not imply that one knows how to determine whether any given statement is true. Nor does it imply that one can state the logical consequences of any particular statement or the causal consequences that would have to occur if it were true. A man may be mistaken about most things, but that does not mean that he has no grasp of the concept of truth.

A definition of a concept is correct if and only if, and to the extent that, it accurately describes the concept. Good definitions do not always come cheap. It is not always easy to disentangle a particular concept from the data of speech and thought. For example, either in stand-alone uses or in combinations with other expressions many words have many meanings, and often different words have more or less exactly the same meaning. However, it is beyond doubt that people hold a statement to be true if and only if they hold that what it says is so. People who use the word 'true' (or its counterpart in some other language, 'vrai', 'vero', 'wahr' and so on) in that way, thereby demonstrate their grasp of the concept of truth. People who do not, thereby demonstrate either that they have failed to grasp the concept of truth or that they wish to use the word 'true' to indicate some other concept. Everybody is free to attach new and uncommon meanings to words in common use. Of course, one does so at one's own risk of being misunderstood. However, what people are not free to do is to say that the new meaning they give to the word is in fact the true or even a better concept of what it means in its common use.

That is where the author of the editorial and pragmatists like him go wrong. He thinks not only and rightly that he is free to give a new meaning to the word 'truth', but also and falsely that his meaning is a better concept of truth than the one that is identified by the common use of the word. Quoting James, he writes that we should 'overhaul the very idea of truth'. With obvious approval he writes that 'James wanted a more complete definition of truth because he wanted one that captured why truth is valuable'. He then quotes James as saying that 'any idea that helps us to deal...with either the reality or its belongings, that doesn't entangle our progress in frustration, that fits...and adapts our life to the reality's whole setting, will agree sufficiently to meet the requirement [of truth]'. To put it crudely, an idea or statement is true if and only if it helps us to get what we want. This pragmatic or opportunistic conception may indeed capture why some ideas and statements are valuable, but it certainly does not agree with the concept of truth. Truth-seekers do value truth and are convinced that it usually pays to base one's actions on truths rather than falsehoods or phantasies, but they will not accept that probable consequence as part of the definition of the concept of truth. It is an open question which truths, mistakes or phantasies are more or less valuable to any given person or class of persons. To argue that we should henceforth use the word 'true' only with respect to statements or ideas that are useful, regardless of whether they are truths or falsehoods, is not to improve our conception of truth. It is to induce us to disregard the overwhelming evidence for the fact that truth is indeed an agreement of an idea or statement with what is really so.

As with truth, so it is with other concepts. Libertarians are familiar with attempts to 'overhaul the very idea of liberty' and to substitute for it a concept that 'captures why liberty is valuable'. They are also familiar with attempts to substitute for it a concept that 'captures why liberty is to be feared and rejected'. They are not sympathetic to such attempts—and rightly so. Liberty is liberty, not equality, necessity, power, wealth, efficient production and fair distribution, pure reason, unconditional goodness or whatever. Liberty is liberty, not licentiousness, callous egoism, selfishness, dog-eat-dog competition or the law of the jungle.

The philosophical problem is to identify the concept of liberty or freedom so that we may be able to criticise particular conceptions of and theories about liberty. It is not the business of philosophers to attach the word 'liberty' to some other concept in order to claim that that concept is the true or a better concept of liberty. That is intellectual fraud, whether those who practise it attach the word 'liberty' to the concept of something else that is generally accepted to be useful or valuable or good, or to the concept of something else that is generally regarded as bad.

Libertarians do claim that liberty is a good thing, but not that it is valuable by definition. They point out that many people value their own and/or others' liberty highly and may supply arguments to the effect that more people should do so. However, they rightly resist the inference made by most of their opponents that if liberty is good or valuable, the 'true meaning of liberty' must be the concept of some other good thing or combination of good things. They know all too well that it will be used by egalitarians or advocates of strong government or whatever to make the claim that only they 'truly understand liberty'.

If libertarian pragmatists agree with the analysis so far, what reason do they have for considering the concept of truth in any other light? And if they do not agree, how are they going to rebut the charges that they play fast and loose with terms such as 'truth' and 'liberty' in order to score easy points in any debate about what is or should be?

Most people, including most libertarians, will not be satisfied with the statement that since liberty is what works, anything that does not work cannot be liberty. If they read The Pragmatist's subtitle 'How liberty works', they expect (and, I suppose, usually get) explanations and arguments about, say, the problems that may arise in a libertarian society and the resources available to people there for dealing with these problems without sacrificing their liberty—or about problems in non-libertarian societies and how they might be avoided or alleviated by a dose of liberty. However, no matter how detailed and persuasive the explanations and arguments may be, they can never by themselves prove to people that the problems they might have in a libertarian society are less serious than those they might experience elsewhere. Nor do they suffice to prove to people that no one in a libertarian society will ever find himself in a situation where the laws of liberty leave him no other option than to learn to live with his problems. A person may very well conclude that liberty does not work at all because he has good reason to believe that how others make use of their liberty will frustrate his own hopes and projects. A libertarian society is no Utopia that comes with a guarantee of a problem-free existence.

I do not believe that libertarian pragmatists are evil persons (which seems to be their perception of how people who do not agree with them must think of them). I do believe, however, that they have espoused a fundamentally incoherent philosophy. I think the root of their error is their view that any plausible answer to the question 'How does liberty work?', is, implies or renders unnecessary an answer to a number of other questions: 'Why should I value my (or any other person's) liberty?', 'Does the fact that my liberty stands in the way of another's plans or hopes, provide him or any other with a sufficient reason to disregard it?', 'Is my liberty a right I can invoke against others because of what I am and they are, regardless of what I think of them and they of me and regardless of, say, the currently fashionable theories about social causation?' and so on.

'In ethics', the author of the editorial writes, 'pragmatism is the view that abandons abstract arguments about the essence of man...and that instead concerns itself with what works and what fails, toward the whole range of actual human purposes.' This, he adds, is 'a very adequate foundation for libertarianism'. Unfortunately, it is not.

Let us admit the 'central insight on which [The Pragmatist] was founded', namely that 'human history contains plenty of data from which we might draw the conclusion that liberty works and slavery fails'. We should immediately recognise that human history also contains plenty of data from which we might infer that slavery works. There is actually very little historical evidence that liberty works, because there are few if any instances of societies that were free in the libertarian sense. Besides, where do we draw the line between 'liberty' and 'slavery'? What is the point of asking whether it is liberty or slavery that works? Surely, no one holds that the abolition of the institution of slavery brings a libertarian society into existence. A person can be unfree without being a slave (or a prisoner) in the common sense meaning of the term.

Where slavery existed as an acknowledged institution, it has usually been defended on the ground that 'it works'. Only where it was challenged on moral or religious grounds have its defenders been led to complement their pragmatic views with moral theories of their own. Indeed, slavery has sustained progressive and powerful civilisations for thousands of years. Compared to the long periods during which slavery was the norm rather than the exception, the time elapsed since the abolition of slavery in the West and especially in the rest of the world is almost insignificant. Where is the evidence that would have convinced the slave-owners of past ages that slavery does not work? Where is the evidence that the experience of an industrial society with rapidly changing technologies, extended markets and a highly developed division of labour can be extrapolated to other types of societies?

Centralised government, with its now immense powers to mobilise men and resources, by sheer force or ruses such as debauching the currency, also has worked very well, at least if judged by the standards of history. It has effectively effaced most traces of more libertarian societies that attempted to oppose the trend towards centralisation. Certainly, many people have tried (often in vain) to escape from its clutches, but many more have attempted to get a piece of the action: If you can't beat 'm, join 'm! That pragmatic rule has worked very well for innumerable individuals, families and other associations, but it would be ridiculous to maintain that it is conducive to liberty.

It is true that in the twentieth century the power to mobilise men and resources by centralised government in the Western warfare-welfare states has proven to be more successful than in the even more centralised fascist or communist totalitarian states. But that is hardly an unambiguous argument for liberty. When libertarians and their opponents see eye to eye, very often the formers' evidence that 'liberty works' is interpreted by the latter as evidence that 'democratic government works'—but liberty is not the same as democratic government. Yes indeed, 'history contains plenty of data', and there is no end to the conclusions 'we might draw from it'.

The problem with pragmatist ethics, from a libertarian or, more generally, individualist point of view, is simple to state: there is no such thing as 'what works and what fails, toward the whole range of actual human purposes'. If the actual human purposes under consideration are the purposes of actual human beings, they are invariably diverse and often in conflict. If they are not constrained in some principled way, what works for some may not work for others. Consequently, even if it were true that everybody should value what works for his actual purposes, it would not follow that there is something that works for everybody, whatever his purposes may be_whether that something is liberty or anything else. Then the pragmatist argument reduces to 'liberty or slavery, whatever, support it if, as long as and to the extent that it serves your purpose'. On the other hand, if the purposes under consideration are constrained, then the dreaded essence of man is brought back in, if not under its own name then under some alias.

Pragmatists are deluding themselves if they believe that liberty works toward the whole range of actual human purposes, whatever they may be. Their confidence in their own arguments probably stems from the conviction that their opinions about what is valuable or not (and hence about what works or not) are shared by all rational persons everywhere. Paraphrasing a famous contemporary of William James, we might say that pragmatists 'seem...to be in that naive state of mind that accepts what has been familiar and accepted by them and their neighbors as something that must be accepted by all men everywhere.' That line, from O.W. Holmes' short but influential article 'Natural law' (1918), was meant to apply to 'those jurists who believe in natural law'_a subject of which Holmes knew virtually nothing, but on which he wrote with sufficient flair to make ignorance of it seem like a mark of intellectual sophistication.

Pragmatists like to ridicule the idea of natural law. However, they should consider how close they are to the naive version of natural law theory if they accept The Pragmatist's statement that 'at this magazine, we believe that we have managed to get to the bottom of the mass of human moral intuitions and to make some sense out of them' — only to discover that most of the magazine's pages are filled with arguments with people who take other 'bottoms' for granted.

I want to conclude with some short remarks on restitution and libertarianism. I fully agree that restitution is a basic pillar of the justice system of a libertarian society. However, it cannot be its only pillar. If a libertarian society is a human society, it will have its share of outlaws as well as of people who try to exaggerate the harm done to them (e.g. in terms of lost opportunities) in order to exploit those who inflicted some harm on them. The outlaws are those who will not volunteer to make prompt and full restitution to the victims of their acts and instead will do (and keep on doing) everything within their power to obstruct or evade the course of justice. A libertarian justice system must be able to deal with crime and to punish criminals. If it cannot do that, only fools and unfortunates will pay restitution. In any case, restitution cannot simply be defined in terms of harm. A distinction has to be made between harm that consists in or is a direct consequence of an infringement of one's rights and harm that is merely manifested in the arousal of, say, feelings of annoyance, indignation or revulsion, or in the frustration of one's hopes and expectations. If no such distinction is made, no sane individual will ever take the pledge to pay restitution. No one will be able to judge whether and to what extent his actions might be harmful to others.

This brings me to the last caveat with respect to pragmatism. The libertarian justice system must be established on a sound and solid basis of law. Law is the order of the human world, the condition of the human world when it is 'in order', i.e. when there is no confusion among persons as to who does or owes what to whom, who participates or cooperates in any undertaking, who does so voluntarily and knowingly or not, and so on. The task of the justice system is to maintain, strengthen and if need be restore that order. It must first of all identify the distinct elements of that order and their relationships to one another. To fulfil its task effectively and efficiently it must in addition develop appropriate rules, methods, procedures and other techniques.

The libertarian idea, I submit, presupposes that the order of the human world (its law) be identified in terms of natural, objective distinctions and not in terms of arbitrary, imaginary or artificial distinctions made for the sake of someone's convenience or some currently fashionable theory of efficiency. It looks upon the human world as a world of natural persons, not of artificial or imaginary 'legal persons'. In short, it adopts the perspective of natural law and not of artificial law. It looks therefore to the natural boundaries among natural persons to determine their natural rights, not to the artificial, conventional or 'socially constructed' boundaries that any 'organisers of society' might wish to impose. It requires a justice system that subjects its rules and procedures to continuous evaluation to determine what works best to maintain the natural law.

I am sure those libertarians who call themselves pragmatists are not fundamentally averse to this perspective, even if it shifts the burden of their argument from the indeterminate 'what works toward the whole range of actual human purposes' to the determinate 'what works toward the whole range of actual human purposes to the extent that it is compatible with natural law'. They should not be taken in by cheap and ignorant dismissals or naive endorsements of natural law produced by people who fail to grasp the distinction between the concept of law (order, which can be natural) and the concept of rule (such as a legal or judicial law, which can never be 'natural' in a literal sense, although it may aim at maintaining, strengthening or restoring the condition of natural law).