An Intelligent Person's Guide to Philosophy

by Scruton, Roger

  All discourse and dialogue depend upon the concept of truth. To agree with another is to accept the truth of what he says; to disagree is to reject it. In ordinary speech we aim at truth, and it is only on the assumption of this aim that people make sense. Imagine trying to learn French in a society of monolingual Frenchmen, without making the assumption that, in general, they aim to speak truly. Of course, not everything we say is true: sometimes we make mistakes, sometimes we tell lies or half-truths. But without the concept of truth, and its sovereign standing in our discourse, we could not tell lies; nor could we have the concept of a mistake.

  Truth is sovereign too in rational argument. From the beginning of history people have needed to distinguish valid from invalid arguments, and no word in the language is more smooth from the touch of human need than ‘if’ — the sign that discourse has shifted from statement to hypothesis, and that a deduction has begun. ‘If p then q; not-q therefore not — p.’ Such is our paradigm of valid inference, and only a lunatic would reject it. But what do we mean by ’valid’? Surely, an argument is valid when it is impossible that the premises should be true, and the conclusion false. Validity is defined in terms of truth.

  It is an odd fact that logic, which ought to be the most scientific part of philosophy, is in many ways the most controversial, and also the slowest to change. Aristotle summarized and classified the valid ‘syllogisms’, and gave a subtle account of truth and inference. But nobody built on his achievement until modern times. Although Leibniz made some important advances, the knowledge of logic among philosophers actually declined during the nineteenth century. The greatest nineteenth-century philosopher - Hegel - wrote a book called Logic which contains only invalid arguments. It was not until the work of two philosophical mathematicians, George Boole (1815-1864) and Gottlob Frege (1848-1925), that the subject began to make progress. And it is testimony to the scientific nature of logic that there was progress to be made. (You don’t make progress in art, literature or religion.)

  The structure of language and rational argument can be understood, according to Frege, only if we make a distinction between the sense of our words, and their reference. ‘The Morning Star’ has a different sense from ‘The Evening Star‘, but it refers to the same thing. The sense of a phrase is what we understand when we understand it. The reference is the object or concept ‘picked out’ — in this case the planet Venus, the star which appears first in the morning and last at night. The distinction between sense and reference runs through all language. Names and descriptions, predicates and relational terms, prepositions and connectives - all have both sense and reference, as do sentences themselves. We can apply the distinction to sentences, Frege argued, by recognizing the deep relation between language and truth. If we assign to each sentence a ‘truth-value’, according to whether it is true or false, then we find that, from the point of view of logic, the truth-value stands to the sentence as the object stands to its name. We understand a sentence when we know the difference that would be made to the world, were the sentence to be true: in other words, the sense of a sentence is given by the conditions for its truth. Truth-value, and truth-conditions, give the two dimensions of sentence-meaning.

  The emphasis on truth provides a clue to the structure of language. When we join two sentences with the word ‘and’, we form a new sentence which is true when its component parts are both true, otherwise false. That is how we grasp the word ‘and’. It refers to an operation defined in terms of truth-values. The same goes for other words which form new sentences from old ones - including ‘if’, ‘not’, and ‘or’. Seeing language in this way, we begin to make sense of its structure. We see how it is that, from a finite array of words, infinitely many sentences can be constructed and understood. We begin to distinguish the valid from the invalid arguments, the well-formed from the ill-formed complexes, and the different functions of the parts of speech. For example, we can begin to describe the real logical difference between names, which refer to objects, predicates, which refer to concepts, and ‘quantifiers’ like ‘some’ and ‘all’, which have a logical role of their own.

  Modern logic emphasizes the distinction between syntax and semantics. Language is built from a finite vocabulary, according to ‘syntactic rules’ which tell us which strings of words are acceptable and which are deviant. But these rules are incomplete without the rules of semantics. Semantic rules assign ‘values’ to the terms of a language: in other words, they assign an object to each name, a concept (or class) to each predicate, a function to each connective, and so on. And they show how to evaluate a complete sentence in terms of the values of its parts. Semantic rules can be constructed only in the way sketched by Frege: by assuming that sentences are assessed in terms of their truth-value. Without this assumption syntax is arbitrary, assertion pointless, and rational discourse impossible to explain. Since philosophy begins from rational discourse - and in particular, from the question ‘Why?’ — philosophy is committed to at least one, all-important claim: namely, that there is a real distinction between the true and the false.

  ‘What is truth? said jesting Pilate, and would not stay for an answer.’ The famous words which begin Bacon’s essay on truth remind us that it is not only in moments of tranquillity that philosophy dawns. Pilate’s question continues to haunt us. Even if language aims at truth, does it ever reach its target? And how do we know? We spontaneously think of truth in terms of reality. A belief, thought or sentence is true if it corresponds to reality. But what is reality, and how do we know it? Here is one of the places where philosophy is apt to go in circles. My desk is part of reality; so too is the colour brown: but what of the brownness of my desk? What makes that a part of reality? Surely, the fact that my desk is brown. Wittgenstein wrote that ‘the world is the totality of facts, not of things.’ (Tractatus Logico-Philosophicus, 1.1.) What makes it true that my car won’t start is not my car, which is quite innocent in the matter, but the fact that my car won’t start. Propositions are made true by facts, and each true proposition identifies the fact that makes it true. Only by dividing reality into facts, do we arrive at the entities to which true propositions correspond.

  But what precisely are facts? And how does one fact differ from another? What is the fact that makes it true that my car is red? Surely, the fact that my car is red. There are as many facts as true propositions, and vice versa. But in that case, what is the difference between them? Why speak of truths and facts, when one and the same thing - a proposition, introduced by the word ‘that’ — is used to identify both of them? Why assume that facts exist, independently of the truths that express them?

  But are we forced to identify facts through propositions? Do we not have ways of attaching our words to the world, for example, by pointing to what we mean? Could I not show what it is, that makes it true that my car is red, by pointing to the redness of my car? Well yes; but pointing is a gesture, and its meaning must be understood. Suppose I point my finger at the car. What leads you to suppose that I am pointing at the car, rather than the house behind my shoulder? After all, you could have read the gesture in another way, from the finger back to the shoulder. The simple answer is that we read the gesture as we do because there is a rule or convention which guides us. This is how we understand it. Moreover, the convention says only that the gesture points to the thing in front of me; further conventions have to be invoked, in order to know which fact about the thing I am singling out for your attention. Pointing belongs to language, and leans on language for its precision. It is only when we are able to read the gesture as an expression of thought that we can use it to anchor our words in reality. But that raises the question what thought? Why, the thought that my car is red! Indeed, no other thought would do: only this would serve to convey the fact that we have in mind, when referring to whatever makes it true that my car is red. We are once again back where we started. All attempts to pass from a thought to the reality described by it come round in a circle. The path from thought to reality
leads in fact from thought to thought.

  This is not surprising, according to Hegel: since thought is all there is. Or rather, not thought exactly, but something called ‘spirit’ (Geist) of which thought is the conscious expression. The difficulties over the concepts of truth and reality disappear, just as soon as we adopt the position of ‘absolute idealism’. This tells us that the world is not some inert array of facts, standing outside and opposed to thought as its passive target, but thought itself, made real and objective through its own internal energy. If we wish to speak of the ‘truth’ of a thought, then we should use this term to refer, not to its correspondence with some unthinking reality, but to its coherence with the system of thought which identifies the world.

  Surprising though it may seem, the dispute that I have just sketched, between the correspondence and the coherence theories of truth, continues unabated. Although expressed in different terms, it remains a real force in intellectual life, even among, perhaps especially among, those who do not really engage in it. The French writer Michel Foucault has invented a new way of doing history, based on the assumption that the truth of a thought is conferred by the system of ruling ideas. The concepts, theories and rationality of an epoch are those dictated by ‘power’; there is no criterion against which to assess them, save those of some rival power which ‘challenges’ their ascendancy. Foucault tells us (in Les Mots et les choses (1966)) that man is a recent invention, and we are understandably startled. Does he mean there were no men around in the Middle Ages? No; he means that the concept: man - as opposed to gentleman, soldier, serf, judge, or merchant - has been current only since the Enlightenment. The implication, however, is that the concept creates what it describes, and that the theories of human nature which burst upon the world in the eighteenth century were theories which created the thing over which they disagreed. Until that time, there was no such thing as human nature.

  All such ideas depend on the observation, in Wittgenstein’s words, that you cannot use language to get between language and the world. Every time you describe reality, you use words; so every time you match a concept with the world, you are really matching a concept with a concept. You are picking out the objects referred to, and using concepts to do so. But this is necessarily so, as the case of pointing illustrates. You cannot pick things out unless you distinguish them from other things, and to distinguish one thing from another is to classify and therefore to apply a concept. Can it be that, from this trivial observation, such momentous conclusions follow, as that the world is nothing but thought, that there is no reality beyond concepts, or that man is a recent invention? Surely not.

  The reason why those ‘idealist’ conclusions do not follow was given by Kant, in the first part of the Critique of Pure Reason. All thinking, he argued, depends upon the application of certain fundamental concepts or ‘categories’ — concepts like those of unity, substance, quantity, and causality, which are not arbitrary classifications but basic operations of thought. These concepts can be deployed only on the assumption of an independent reality: this assumption is built into them, along with the distinction between appearance and reality, seeming and being, which it is their role to elaborate. It is in their nature to ‘aim beyond’ experience to the world which explains it, and everyone who uses these concepts shares that aim, even the idealist who denies that he does so. Even he must use concepts of substance, causality, world and identity, if he is to say what he means; and these concepts commit him to the view that the world exists apart from his thinking.

  So does the world exist apart from our thinking? Or do our concepts merely assume that it does? Kant’s answer is ingenious. If we are to think at all, he argues, we must use concepts. If we use concepts we must deploy the categories. If we deploy the categories, we must assume the distinction between how things are and how they seem. If we make that distinction, we commit ourselves to an objective reality, and aim our discourse towards it. Even to deny the existence of reality is to think, and therefore to assume its existence. We do not need to argue that the world exists; its existence is presupposed in every argument, even the argument that it doesn’t.

  Kant produced many proofs of that kind, which attempt to show how we must think if we are to think at all. Such proofs go beyond rational deduction, to explore what it presupposes. Kant described them as ‘transcendental’, and his own philosophy as ‘transcendental idealism’. A transcendental argument starts from a premise like ‘I think,’ ‘I believe myself to be free,’ or ‘I have the idea of myself’; it then asks, ‘What must be true if there is to be such a thought? What else must I think, and what must the world be like in which I exist, thinking such a thought?’

  Kant’s insight can be approached in another way. Foucault’s conclusion about human nature stems from a failure to distinguish two kinds of concept: concepts which explain the world, and concepts which focus our response to it. The concept: fish is of the first kind; the concept: ornament of the second. When we divide the world into fish and not-fish, we do not regard this as an arbitrary expression of our interest in fish. We believe we are grouping things which belong naturally together, even if we don’t know why. The classification is the first step in a theory, which does not merely describe the class of fish, but also explains it. The concept is exploring the world. At a certain stage we might discover that things which we once classified as fish are not fish at all, since they do not belong with the rest of the kind. This happened when whales were discovered to be mammals.

  Ornaments have only one thing in common: namely, that we use them as ornaments. Our interest forms the class, and the concept and the interest arise and decline together. In an age when people do not distinguish between things which are, and things which are not, ornamental, there are no ornaments. It is perfectly reasonable to conclude, from the fact that the concept of an ornament is a recent invention, that ornaments too are recent inventions. No such reasoning could prove that fish are a recent invention. Concepts like those of fish and man are aimed at reality: they are forensic concepts, tied to explanation, and therefore to the Kantian categories. Their application is determined by the world, not by us, and they lead us on a voyage of discovery. Man could not be a recent invention, even if it is only recently that we have begun to use the concept. For the concept identifies something which preceded its own invention, whose nature is given by laws that have yet to be discovered. In such a case it is not the concept which creates the kind, but the kind which creates the concept. Men form what John Stuart Mill called a ‘natural kind’, as do fish, fleas and water.

  Philosophy exists only because of the question ‘Why?’; Why-questions arise in the context of rational discussion; rational discussion requires language; language is organised by the concept of truth; truth is a relation between thought and reality and reality is objective: neither created by our concepts, nor necessarily well described by them. Such is the train of thought that we have followed. So how do we know that our beliefs are true? Kant’s argument shows that certain assumptions are unavoidable. But does that establish their truth?

  3

  THE DEMON

  Many philosophers think not. An assumption may be both unavoidable and unjustified — like the assumption of honesty in a market. Perhaps the same is true of the assumption that our world is real. Consider the following argument, due to Descartes. An evil demon has control of my experience, and induces all my sensations, thoughts and perceptions, so leading me to believe that I inhabit an objective world. The demon’s deception is systematic: at no point does the experience that he produces deviate from the norm that is familiar to you and me. But the world in which I believe myself to reside is a fiction; I am alone in the universe, with the demon who delights in deceiving me. How, in the face of this possibility, can I be sure that the world of my perception really exists, and really is as I think it to be?

  This famous argument set modern philosophy on its sceptical journey. It does not merely point to the gap between appearance and
reality. It adds that all methods we have for crossing this gap - perhaps all conceivable methods - fail to achieve their purpose. Locked within my illusory experience, I apply the best of scientific tests, distinguish the true from the false among my impressions, develop theories which refer to an underlying reality, and reassure myself, by reading Kant, that the world is there and I am a part of it. But all the while my thought, while imagining that it has ventured forth on a voyage of discovery, has merely patrolled its own perimeter, like a prisoner in a painted cell, who takes the pictures for windows. For all I know, there is nothing beyond my inner life, save the demon who produces it.

  Someone could accept Descartes’ argument as showing that, in the last analysis, the world remains hidden from us, lying beyond the boundaries of thought, but nevertheless believe that the distinctions between the true and the false, the real and the imaginary, the objective and the subjective, are genuine and useful. For, in an important sense, the hypothesis of the evil demon leaves everything unchanged. Whether the hypothesis is true or false, my experience will remain unaltered: indeed, that is the whole point of the story. So too will my concepts, my methods for distinguishing appearance and reality, and the process of scientific discovery which enables me to predict new experiences from old.

  This is an important observation, since many current forms of scepticism give global arguments for local conclusions. Foucault’s argument, considered in the last chapter, is one of them. It assumes that a general attack on the idea of ‘correspondence’ can be used to show that particular beliefs about the world (for example, that human nature is a constant datum) are false. But such general arguments show nothing of the kind; for they are too general. They do not disestablish the distinction between proven truths and mere opinions. For this is a distinction that is made within our scheme of thinking, and relies on no metaphysical picture. A similar instance is provided by ‘deconstruction’, the fashionable philosophy which tells us that, because we cannot use language to think outside language, we can never guarantee the meaning of our words. Hence there is no such thing as meaning, and the decision to attach a particular meaning to a text is always in some sense arbitrary, dictated by politics or power and not by the text itself. This too involves an illegitimate passage from a global theory to a local conclusion. Even if the global theory is true, it leaves everything at the local level unchanged. We still have criteria for distinguishing the meaningful from the meaningless; we still use these criteria in dialogue, and must do so if we are to think or speak at all. And we can still distinguish the true meaning of a text from the private associations of its reader.