For instance, the "marginal producer" is being translated into the "bulk-line" producer as figures of costs of production make inductive studies possible. We used to be told: "Price is governed by the cost of production of the marginal producer." Who is he? "He is the producer who makes no profit: that is, whose cost of production is equal to the price." Such a statement needs no verification and tells nothing which could be verified. In place of this, inductive studies make their statements in some such form as this: "Price is commonly such that from ten to fifteen percent of the output is produced at greater cost and the rest at less." 1

Or, we are told that price is fixed at the point where supply and demand become equal. Supply turns out to mean the volume of goods that holders are willing to sell at any given price (implying a demand at that price) and demand is the volume of goods that buyers are willing to take, if offered at a given price (implying that someone is willing to sell that much at that price; in other words, implying a "supply" equal to the "demand"). Such a proposition is self-evident, and of no use for purposes either of prediction or verification. It is only as common sense or induction adds descriptions of the behavior of "supply" and "demand" that the proposition has meaning attached to it, by a sort of associative process, and becomes a vehicle or symbol of truths which its literal formulation does not categorically denote.

Some forms of the doctrine of marginal utility are even more clearly tautological. Our old friend, the "economic man" is becoming very self-conscious and bafflingly non-committal. Instead of introducing himself to his readers with his old time freedom, he says: "I may behave one way and I may behave another, but what is that to you? You must take my choices as you find them: I choose as I choose and that is all you really need to know." The poor thing has been told that his psychology is all wrong, and he is gamely trying to get on without any and still perform as many as possible of his accustomed tasks. He has become a symbol, rather than a means of description or explanation. Yet this non-committal treatment of economic choices seems to be of immense import to those who have become accustomed to attach much of their thinking to this symbol. Human behavior in economic life is so many-sided that if we insist on summing it up under simple formulas and also insist that

1 This concept is developed in articles by Taussig, "Price Fixing as Seen by a Price Fixer," Quart. Jour. Econ. Feb., 1918; "Price Fixing and the Theory of Profit," Quart. Jour. Econ. Nov., 1919 and by Kemper Simpson, Quart. Jour. Econ. Feb., 1921, and May, 1923.

these formulas shall always be one hundred per cent accurate, the only ones that can satisfy these requirements are in the form "whatever is, is"; preferably camouflaged into a semblance of meaning. A letter-perfect description of the economic world is impossible. And not only that: it would also be largely useless as a form of general economics, because it would not generalize enough to assist the mind to an understanding. General economics must simplify in order to interpret; otherwise its description will be just as unwieldly and baffling as the world itself. Thus theoretical economics must steer a course somewhere between what is futile and what is impossible. It will be a never-ending search for generalizations that are significantly true, and for that very reason are often neither one hundred per cent accurate, nor universally applicable.

II

A PRAGMATIC VIEW OF ECONOMIC TRUTH

1. SELECTION AND EMPHASIS

This simplification lays upon the theorist a great burden of responsibility for the character and influence of the kind of selection he sees fit to make. It robs him of his defense of scholarly detachment, when critics assail him, for instance, on the ground that his doctrines are out of touch with movements of reform and surgings of unrest. He may answer that he is not supposed to be in touch with them, in his capacity as a scientist; he is supposed merely to furnish a wholly objective description of things as they are. But he is open to the rejoinder that, so far as his description is selective, it is not wholly objective. He selects things that are significant, and significance is at bottom a matter of some underlying

purpose.

Some things are significant for one purpose, and some for another. A careful thinker always wants to know for what purpose a definition is to be used before he will admit that it expresses the essential and significant features of the thing that is to be talked about. "Value" (as actually used) means one thing for purposes of taxation and another thing for purposes of regulation, and there are similar varieties of meaning covered by the terms cost, wealth, capital and income. Each term covers a related family of ideas, so that it comes to be of the essence of sound method to choose the particular idea that is appropriate to the purpose in hand, and to

avoid all conclusions based on a merging of this idea with others in the same family.

This sort of selection is guided by something different from the abstract and disinterested search for pure truth or pure accuracy of description, which so many theorists would like, in the interests of scientific standing, to achieve. There are always vastly more aspects capable of analysis, vastly more distinctions capable of being used as the basis of reasoning, than the theorist can possibly organize into a systematic body of thought, and more than the hearers can grasp as a connected whole.

In living economic theory the investigators find that their best possible efforts cannot exhaust the material that is of vital importance in its effect on the issues of their time and the needs of their age, and they tend to concentrate upon these aspects of things, not because other aspects do not exist, but because they appear less important. In an economic theory that is not living, this selection is governed by tradition, by professionally vested interest in an existing stock of doctrines and methods, and sometimes by the line of least resistance toward deductive studies, because they yield definite results more readily than inductive, and with less labor of gathering data. But the selection is never governed by the sheer search for truth, for the search for truth is not a principle on which one may decide what parts of truth to include and what parts to leave out. Thus the "realistic" painter who paints life in terms of dung-heaps is selecting; life has other things in it than dung-heaps. He may select the dung-heaps because he is disgusted with the other artists who insist on leaving them out. In that case his selection is governed not by what his subject matter is, but by what the other artists have painted, or said, or written, about it.

Take the statement that men seek their strongest interests, combined with a definition which says that men's strongest interests are whatever things they seek. When we come to put this statement in a form that has some meaning, as for example, when one says that every man is a better judge of his own interests than anyone else can be for him, we are instantly faced by the fact that this is not always true, but that the opposite is true in some cases. Now, as far as the practical usefulness of theory is concerned, the important thing is whether or not a generalization like this is likely to be used in cases to which it does not properly apply, or only in cases in which it is true. Such a statement may be safeguarded in two ways. It may be carefully qualified, or the situation may be such that qualification is not necessary.

2. THE IMPORTANCE OF TIME AND PLACE

If all the doubtful issues of the time and place are of such a sort that they can best be settled by assuming that each individual is always the best judge of his own interests, or vice versa, then, so far as that time and place are concerned, one of these sweeping generalizations is wholly true and needs no safeguarding other than the fact that in that time and place only one kind of issue is present. This happy state of things never really comes to pass, but something fairly close to it sometimes is true of the watchwords of strenuous fighting movements. If the existing order is biassed so far in one direction that all that can be accomplished in the other direction will not restore the balance, the party of protest will adopt as a watchword a strong and sweeping statement of the opposite principle, without weakening it by cautious delineations, stating just how far it can properly be carried and what concessions will have to be made. "Men do not follow an uncertain sound into battle.'

On the other hand, the existing order always has some value, if only by the fact that we are used to it, and if the question at issue is whether the existing order shall or shall not be wholly discarded in favor of some untried invention sure to entail enormous injustice and suffering, and uncertain of ultimate success; then sweeping dogmas of a conservative sort come naturally into use. In a way, this is proper as it is natural. So far as concerns the issue raised by a party of wholesale revolution, the most important truth about the existing order may consist, not in minor defects and abuses, and not even in the fact that the best possible system would be very different from the system as it stands, but rather in the fact that it does harness and utilize some very powerful forces and get some very valuable results, and that these would be destroyed by the proposed revolution. The real truth in the defense of the existing order may be, not that this order is ideal, but that the counter proposition would be still worse. Thus Adam Smith did not claim that free enterprise worked perfect results, but rather that the worst evils. sprang, not from free enterprise but from the mistakes of governmental interference. Hence the significant truth of a proposition about the existing order depends, not on that order alone, but on the character of the substitutes which political organization, scientific knowledge and human imagination place within the bounds of serious possibility.

It is probably true that economic conceptions have mostly been

framed with reference to particular errors or evils which they might serve to combat. This obviously is true of the myths and slogans that animate fighting parties, but it is no less true of the most carefully qualified abstractions. The qualifications recognize passively the existence of facts and forces which the abstraction does not actively interpret.

For that reason new issues must bring forth new selections of significant material embodied in new myths, new fighting slogans and new abstractions. It is this which determines the range of the broader generalizations in the "social science of business." Each generation of economists succeeds to a new assortment of practical problems to which its doctrines are to be applied.

It would be vastly interesting, did space permit, to apply this method of interpretation to the ideas of Medieval economists, Physiocrats and Mercantilists. To one who has been trained to regard these ideas merely as examples of the perverse errors of the unenlightened human mind, their appropriateness to their particular environments comes as a startling revelation, tending at once to lessen one's disrespect for these "fallacies" and to weaken one's trust in the absolute verity or eternal adequacy of the doctrines which the peculiar environment of the nineteenth century has brought into a dominant position in economics.

We inherit the tradition of Adam Smith and Ricardo, representing the birth and early infancy of the industrial revolution, when the forces latent in this new economic system were struggling for room to develop, and had hardly begun to foreshadow the miracles they would ultimately bring to pass. For the beneficiary control of these forces, class interests and social systems were already struggling; the combat lying broadly between liberal and humanitarian notions of popular welfare on the one hand, and the landed, mercantile and militaristic interests on the other, by whom the poor were more or less frankly regarded as instruments of production and "food for powder."

The first combat was against the national restrictions of the mercantile systems, together with local and craft restrictions surviving from the smaller-scale medieval economy. Here a group of abstractions were developed, serving the ends of economic freedom, such as the idea that both sides gain in an exchange, that consumable goods are the end of the economic system, that the consumer can exercise