CHAPTER I
The Nature and Scope of Political Philosophy
Until the beginning of the present century philosophy was generally regarded as a source of knowledge which transcended, both in scope and certainty, the discoveries of natural science. Science, it was agreed, marked an advance on the uncritical and often unrelated beliefs of ordinary life, yet it was itself based on the observations of the senses and consisted of the uncertain generalizations based upon them; whereas philosophy was assumed to answer questions about such subjects as the existence of God, the nature of knowledge, and the authority of the moral law upon which sense-experience, from its very nature, could throw no light. On such subjects, it was believed, reason was alone competent to pronounce and, when it did so, its conclusions were characterized by a logical and universal certainty which the generalizations of natural science could never claim.
That philosophical knowledge is certain and indubitable is a claim which, in a broad sense, all philosophers have made, or at least implied; and if a short and simple definition of philosophy were sought the title of the late Professor Deweyâs Gifford LecturesââThe Quest for Certaintyââmight serve as a starting point at least. For all philosophers have claimed, or at least implied, that philosophical knowledge not only is, but must be, true. But this general agreement has not prevented fundamental differences of opinion regarding the nature and scope of such knowledge; and since these differences are reflected in the application of philosophy to the problems of political theory it is important to be aware, however generally, of their nature.
The different conceptions of philosophy ultimately depend upon different conceptions of the nature of indubitable knowledge. The propositions of mathematics are usually cited as typical illustrations of such knowledge. For example, the proposition âTwo plus two equals fourâ is said to be necessarily and universally true on the ground that, once we have grasped its meaning, we recognize that it must be necessarily and universally true, and because further instances of its truth do not increase our certainty that it must always be true. Its falsity, in other words, is inconceivable. On the other hand, there are numerous propositions of which the falsity is perfectly conceivable. It may be true that âThe cat is blackâ or that âPoliomyelitis is caused by a virusâ, but these propositions are not necessarily true. On the contrary, their falsity is perfectly conceivable, even if observation appears to confirm their truth.
Analytic and Synthetic Propositions
The distinction just illustrated is variously referred to as the distinction between rational and empirical knowledge, or between a priori and a posteriori knowledge, or between truths of reason and truths of fact. And it is generally true to say that all philosophers have claimed, or at least implied, that their theories are rational and a priori. Where they have differed is in their view of the scope of such knowledge. And the main difference has been that some have held that rational knowledge is always analytic, while others have held that it is sometimes synthetic.
The difference between analytic and synthetic propositions was defined by the German philosopher Immanuel Kant (1724â 1804) as follows: Analytic propositions, he said, âadd nothing through the predicate to the concept of the subject, but merely break it up into those constituent concepts that have all along been thought in it, although confusedlyâ, while synthetic judgments âadd to the concept of the subject a predicate which has not been in any wise thought in it, and which no analysis could possibly extract from itâ.1 The difference is, in short, that the predicate in an analytic proposition is contained within the meaning of the subject, while in a synthetic proposition the predicate is not contained within the meaning of the subject but adds something related to it. Kant illustrated the difference by the two propositions âAll bodies are extendedâ and âAll bodies are heavyâ. The former, he thought, is analytic, because the concept of âextensionâ is part of the meaning of âbodyâ, while the latter is synthetic because the concept of âheavinessâ is not part of the meaning of âbodyâ, but only a quality which it acquires when it is placed in a gravitational field.
Kantâs definition drew attention to an important difference between analytic and synthetic propositions, although not all analytic propositions naturally fall into the simple subject-predicate form which his examples illustrate. The essential characteristic of an analytic proposition is that it defines the meaning, or part of the meaning, of its subject and does not describe unessential features which may, or may not, belong to it. A cube of iron has a certain weight at sea level, a smaller weight at the top of a high mountain, and no weight at all at a certain point between the earth and the moon; but these differences are not essential elements in the meaning of the description âcube of ironâ. It is clear, on the other hand, that if the cube of iron had no extension it would not be a cube of iron, since extension is an essential part of the meaning of the phrase âcube of ironâ. In other words, to deny an analytic proposition is self-contradictory since that is simultaneously asserting and denying the same thing. It is, to borrow Bertrand Russellâs example, like saying âA bald man is not baldâ.1
Modern philosophers have devoted much attention to the study of analytic propositions, and many would agree with Professor Ayer that âa proposition is analytic when its validity depends solely on the definitions of the symbols it containsâ,2 and that this is so because analytic propositions âdo not make any assertion about the empirical world. They simply record our determination to use words in a certain fashion.â3 They are, in other words, tautologies; and the reason why we think it worth while to assert them and sometimes, as in mathematics, to draw elaborate deductions from them, is that our reason is too limited to recognize their full significance without going through these complex verbal processes.
These considerations may appear to be extremely abstract and their connection with what is commonly understood as âpolitical philosophyâ far from obvious; but in fact this connection is both simple and fundamental. For philosophy is the âquest for certaintyâ, and if certainty is a characteristic of propositions, then an inquiry into the nature and scope of certain, i.e. a priori, propositions must be the essential task of all philosophy. If, in
other words, the general object of philosophy is to discover the nature and implications of rational thinking, then an enquiry into the nature of the propositions by which rational thinking is expressed is necessarily one of the most important tasks of philosophy so understood.
All philosophers who have recognized the distinction between analytic and synthetic propositions have agreed that analytic propositions are necessary and a priori. Controversy has centred on the question whether synthetic propositions may also sometimes be a priori. And the different answers given to this question have determined very different conceptions of the scope and purpose of philosophy. For if the propositions of philosophy must always be a priori, and a priori propositions must always be analytic, it follows that the propositions of philosophy must always be analytic.
Now one important class of proposition which is never analytic is the class of existential propositions, i.e. propositions asserting something of the real world. While it is necessarily true that 2 plus 2 equals 4, it is not necessarily true that there are four distinguishable objects in the real world. For example, if I have ÂŁ2 in one pocket and ÂŁ2 in another, it necessarily follows that I have ÂŁ4. in both pockets, but it is for empirical observation to ascertain whether in fact I have ÂŁ2 in one pocket and ÂŁ2 in another pocket. This simple example illustrates the important principle that analytic propositions apply only in a hypothetical sense to the real world. No analytic proposition of the form âA is Bâ can be asserted categorically of the real world. It can only be asserted in the hypothetical form âIf X (some existing thing) is A then it must be B.â But the proposition asserting that X is in fact A is synthetic and cannot be necessarily true unless synthetic propositions can be a priori.
Thus if a priori propositions are always analytic, philosophy will be unable to demonstrate the truth of any proposition about the existing world except in so far as it is logically implied by an existential proposition whose truth has been establisbed (if it can be established) by empirical observation. The function of philosophy, in other words, will be to examine the implications of propositions and not to demonstrate their truth.
As already mentioned, however, it was widely believed until some fifty years ago that philosophy could establish facts about the existing world quite independently of experience. Philosophy was, indeed, often looked to for a rational justification of beliefs, such as religious or moral beliefs, already held on non-rational grounds, and it was assumed that this justification could be given independently of experience. But during the present century there has been a strong reaction from these methods and a growing acceptance of the alternative view that the function of philosophy is to clarify rather than to extend the content of human knowledge.
The theory that a priori thinking can never by itself establish a truth about the existing world is known as Empiricism, since it always asserts that such propositions can be established only by empirical observation. The alternative theory that a priori thinking can by itself establish truths about the existing world is known as Rationalism. And it is clear from the preceding discussion that Rationalism can be defended only if synthetic a priori propositions are possible. For if such propositions are not possible no proposition about the existing world can be established a priori, and some form of Empiricism must therefore be accepted.
Humeâs Empiricism
Before the present century, when the doctrine has received wide support, the most celebrated exponent of Empiricism was the Scottish philosopher David Hume (1711â1776), now generally recognized to have been one of the greatest philosophers of all time. Hume held that the only propositions which are certainly true are those which describe ârelations of ideasâ, by which he meant analytic relationships in the sense defined above. Those which describe âmatters of factâ, i.e. synthetic propositions, cannot be rationally justified, although they can be accepted as true in so far as they are justified by direct observation. But of course the great majority of synthetic propositionsâin particular, the socalled âlawsâ of scienceâgo far beyond this and make assertions which cannot be justified by experience.
Thus Hume argued that the belief in the universal truth of scientific laws follows repeated observations of the sequences which they describe; but he denied that there is any necessity in these sequences, or even in the occurrence of the belief that they are universal and necessary. If I infer that, because all observed samples of arsenic have proved to be poisonous, therefore all samples whatsoever are poisonous, no logical justification of this inference can, according to Hume, be given. It is just a fact that, following on the observation of numerous samples of arsenic which prove to be poisonous, everybody believes that all samples whatsoever will prove to be poisonous. But there is, according to Hume, no rational justification for this belief; it just happens to occur following on experience of the effects of arsenic in a limited number of instances, and just happens to have proved a reliable guide in practice. There is no guarantee that it will prove to be true of all instances whatsoever. Thus there is nothing âreasonableâ in the belief in the a priori sense.
Hume reached the same sceptical conclusions about the general propositions of morality. He thought it obvious that these propositions are synthetic, and argued that they cannot therefore be a priori. Such propositions as âJealousy is evilâ or âLying is wrongâ are, he thought, obviously synthetic in that their predicates are not part of the meaning of the subjects. And such propositions cannot be a priori, for no necessary connection can, in his view. be discerned between the subject and the predicate. Hence the basis for these moral generalizations must be the same as the basis for the generalizations of natural scienceâthe observation of a limited number of instances. And this is not a rational ground for asserting them.
Having denied that moral generalizations have any logical necessity, Hume set himself to analyse the empirical evidence on which they are based. He reached the conclusion that the basis of such generalizations is a peculiar type of sentiment or feeling. When I say âHonesty is goodâ I am, according to Hume, saying, in a rather specific sense of the word âlikeâ, âI like honestyâ. I am, in fact, describing not an inherent quality of honesty but a feeling excited in me by the contemplation of honesty. This feeling Hume called the âpleasing sentiment of approbationâ. He thought that moral disapproval in the same way expresses a sentiment of disapprobation. Thus Hume concluded that there is nothing ârationalâ or âlogicalâ in morality and that it is impossible to show, on a priori grounds, that moral propositions are true or false. Their truth or falsity depends on the purely empirical question whether they are or are not accurate descriptions of the feelings to which they relate.
Humeâs scepticism is therefore of a revolutionary character, for it implies that neither the principles of natural science nor the laws of morality have any universal necessity, and that practical thinking is of an essentially irrational character. Our belief in such generalizations as âArsenic is poisonousâ or âLying is wrongâ is not, according to Hume, arrived at by any logical argument, but is simply a natural belief which occurs in certain situations, and whose occurrence cannot be explained on a priori grounds. All attempts to show that such beliefs are necessarily true must, in his view, completely fail.
The consequences of Humeâs scepticism are most striking in the sphere of morality., for they imply that there cannot be what Kant was later to call a âcategorical imperativeâ. This is the principle of unconditional obligation to do what is right. All historical codes embody this principle, for they are composed of âlawsâ of the type set forth in the Decalogue, which imply an unconditional obligation to do or to refrain from doing certain acts. But on Humeâs theory a moral judgment is the assertion that something excites a certain feeling, and there is no reason why this feeling should be universally experienced, nor any sense in saying that it ought to be universally experienced, for the word âoughtâ is, on Humeâs theory, itself an expression of a feeling.
Hume carried his attack on conventional assumptions about morality even further, for he argued that, even if mo...