Principles of Economics by Alfred Marshall (1890)

Book Two: Some Fundamental Notions

Chapter 1, Introductory

1. We have seen that economics is, on the one side, a Science of Wealth; and, on the other, that part of the Social Science of man's action in society, which deals with his Efforts to satisfy his Wants, in so far as the efforts and wants are capable of being measured in terms of wealth, or its general representative, i.e. money. We shall be occupied during the greater part of this volume with these wants and efforts; and with the causes by which the prices that measure the wants are brought into equilibrium with those that measure the efforts. For this purpose we shall have to study in Book III wealth in relation to the diversity of man's wants, which it has to satisfy; and in Book IV wealth in relation to the diversity of man's efforts by which it is produced.

But in the present Book, we have to inquire which of all the things that are the result of man's efforts, and are capable of satisfying man's wants, are to be counted as Wealth; and into what groups or classes these are to be divided. For there is a compact group of terms connected with Wealth itself, and with Capital, the study of each of which throws light on the others; while the study of the whole together is a direct continuation, and in some respects a completion, of that inquiry as to the scope and methods of economics on which we have just been engaged. And, therefore, instead of taking what may seem the more natural course of starting with an analysis of wants, and of wealth in direct relation to them, it seems on the whole best to deal with this group of terms at once.

In doing this we shall of course have to take some account of the variety of wants and efforts; but we shall not want to assume anything that is not obvious and a matter of common knowledge. The real difficulty of our task lies in another direction; being the result of the need under which economics, alone among sciences, lies of making shift with a few terms in common use to express a great number of subtle distinctions.

2. As Mill says: (1) - "The ends of scientific classification are best answered when the obj ects are formed into groups respecting which a greater number of general propositions can be made, and those propositions more important, than those which could be made respecting any other groups into which the same things could be distributed." But we meet at starting with the difficulty that those propositions which are the most important in one stage of economic development, are not unlikely to be among the least important in another, if indeed they apply at all.

In this matter economists have much to learn from the recent experiences of biology: and Darwin's profound discussion of the question (2) throws a strong light on the difficulties before us. He points out that those parts of the structure which determine the habits of life and the general place of each being in the economy of nature, are as a rule not those which throw most light on its origin, but those which throw least. The qualities which a breeder or a gardener notices as eminently adapted to enable an animal or a plant to thrive in its environment, are for that very reason likely to have been developed in comparatively recent times. And in like manner those properties of an economic institution which play the most important part in fitting it for the work which it has to do now, are for that very reason likely to be in a great measure of recent growth.

Instances are found in many of the relations between employer and employed, between middleman and producer, between bankers and their two classes of clients, those from whom they borrow and those to whom they lend. The substitution of the term "interest" for "usury" corresponds to a general change in the character of loans, which has given an entirely new key-note to our analysis and classification of the different elements into which the cost of production of a commodity may be resolved. Again, the general scheme of division of labour into skilled and unskilled is undergoing a gradual change; the scope of the term "rent" is being broadened in some directions and narrowed in others; and so on.

But on the other hand we must keep constantly in mind the history of the terms which we use. For, to begin with, this history is important for its own sake; and because it throws side lights on the history of the economic development of society. And further, even if the sole purpose of our study of economics were to obtain knowledge that would guide us in the attainment of immediate practical ends, we should yet be bound to keep our use of terms as much as possible in harmony with the traditions of the past; in order that we might be quick to perceive the indirect hints and the subtle and subdued warnings, which the experiences of our ancestors offer for our instruction.

3. Our task is difficult. In physical sciences indeed, whenever it is seen that a group of things have a certain set of qualities in common, and will often be spoken of together, they are formed into a class with a special name; and as soon as a new notion emerges, a new technical term is invented to represent it. But economics cannot venture to follow this example. Its reasonings must be expressed in language that is intelligible to the general public; it must therefore endeavour to conform itself to the familiar terms of everyday life, and so far as possible must use them as they are commonly used.

In common use almost every word has many shades of meaning, and therefore needs to be interpreted by the context. And, as Bagehot has pointed out, even the most formal writers on economic science are compelled to follow this course; for otherwise they would not have enough words at their disposal. But unfortunately they do not always avow that they are taking this freedom; sometimes perhaps they are scarcely even aware of the fact themselves. The bold and rigid definitions, with which their expositions of the science begin, lull the reader into a false security. Not being warned that he must often look to the context for a special interpretation clause, he ascribes to what he reads a meaning different from that which the writers had in their own minds; and perhaps misinterprets them and accuses them of folly of which they had not been guilty. (3)

Again, most of the chief distinctions marked by economic terms are differences not of kind but of degree. At first sight they appear to be differences of kind, and to have sharp outlines which can be clearly marked out; but a more careful study has shown that there is no real breach of continuity. It is a remarkable fact that the progress of economics has discovered hardly any new real differences in kind, while it is continually resolving apparent differences in kind into differences in degree. We shall meet with many instances of the evil that may be done by attempting to draw broad, hard and fast lines of division, and to formulate definite propositions with regard to differences between things which nature has not separated by any such lines.

4. We must then analyze carefully the real characteristics of the various things with which we have to deal; and we shall thus generally find that there is some use of each term which has distinctly greater claims than any other to be called its leading use, on the ground that it represents a distinction that is more important for the purposes of modern science than any other that is in harmony with ordinary usage. This may be laid down as the meaning to be given to the term whenever nothing to the contrary is stated or implied by the context. When the term is wanted to be used in any other sense, whether broader or narrower, the change must be indicated.

Even among the most careful thinkers there will always remain differences of opinion as to the exact places in which some at least of the lines of definition should be drawn. The questions at issue must in general be solved by judgments as to the practical convenience of different courses; and such judgments cannot always be established or overthrown by scientific reasoning: there must remain a margin of debatable ground. But there is no such margin in the analysis itself: if two people differ with regard to that, they cannot both be right. And the progress of the science may be expected gradually to establish this analysis on an impregnable basis. (4)


1. Logic, Bk. IV, ch. VII, Par. 2.

2. Origin of Species, ch. XIV.

3. We ought "to write more as we do in common life, where the context is a sort of unexpressed 'interpretation clause'; only as in Political Economy we have more difficult things to speak of than in ordinary conversation, we must take more care, give more warning of any change; and at times write out 'the interpretation clause' for that page or discussion lest there should be any mistake. I know that this is difficult and delicate work; and all that I have to say in defence of it is that in practice it is safer than the competing plan of inflexible definitions. Any one who tries to express various meanings on complex things with a scanty vocabulary of fastened senses, will find that his style grows cumbrous without being accurate, that he has to Use long periphrases for common thoughts, and that after all he does not come out right, for he is half the time falling back into the senses which fit the case in hand best, and these are sometimes one, sometimes another, and almost always different from his 'hard and fast' sense. In such discussions we should learn to vary our definitions as we want, just as we say 'let x, y, z, mean' now this, and now that, in different problems; and this, though they do not always avow it, is really the practice of the clearest and most effective writers." (Bagehot's Postulates of English Political Economy, pp. 78-9.) Cairnes also (Logical Method of Political Economy, Lect. VI) combats "the assumption that the attribute on which a definition turns ought to be one which does not admit of degrees"; and argues that "to admit of degrees is the character of all natural facts."

4. When it is wanted to narrow the meaning of a term (that is, in logical language, to diminish its extension by increasing its intension), a qualifying adjective will generally suffice, but a change in the opposite direction cannot as a rule be so simply made. Contests as to definitions are often of this kind: - A and B are qualities common to a great number of things, many of these things have in addition the quality C, and again many the quality D, whilst some have both C and D. It may then be argued that on the whole it will be best to define a term so as to include all things which have the qualities A and B, or only those which have the qualities A, B, C, or only those which have the qualities A, B, D; or only those which have A, B, C, D. The decision between these various courses must rest on considerations of practical convenience, and is a matter of far less importance than a careful study of the qualities A, B, C, D, and of their mutual relations. But unfortunately this study has occupied a much smaller space in English economics than controversies as to definitions; which have indeed occasionally led indirectly to the discovery of scientific truth, but always by roundabout routes, and with much waste of time and labour.

Contents | next chapter | Political Economy Archive