The human reason has two ways of ascertaining truth. The first of these is that of reasoning from particulars to generals in an ascending line, until we come at last to one of those invariable uniformities that we call laws of nature. This method we call the inductive or a posteriori. But when we have reached what we feel sure is a law of nature, and as such true in all times and places, then an easier and more powerful method of ascertaining truth is open to us — the method of reasoning in the descending line from generals to particulars. This is the method that we call the deductive, or a priori method. For knowing what is the general law, the invariable sequence that we call a law of nature, we have only to discover that a particular comes under it to know what is true in the case of that particular.
Chapter 11 — Methods of Political Economy
In relation of priority the two methods stand in the order in which I have named them — induction being the first or primary method of applying human reason to the investigation of facts, and deduction being the second or derivative. So far as our reason is concerned, induction must give the facts on which we may proceed to deduction. Deduction can safely be based only on what has been supplied to the reason by induction; and where the validity of this first step is called in question, must apply to induction for proof. Both methods are proper to the careful investigation that we speak of as scientific: induction in its preliminary stages, when it is groping for the law of nature; deduction when it has discovered that law, and is thus able to proceed by a shortcut from the general to the particular, without any further need for the more laborious and, so to speak, uphill method of induction, except to verify its conclusions.
There is a further method of investigation, which consists in a combination of these two original methods of reason, and which has been found most effective in the discovery of truth in the physical sciences. When our induction is so pointed to the existence of a natural law that we are able to form a surmise or suspicion of what it may prove to be, we may tentatively assume the existence of such a law, and proceed to see whether particulars will fall into place in deductions made from it. This is the method of tentative deduction, or hypothesis. Where the application of the inductive method was really needed in what is now called by the "new lights" the "classical" political economy was to test the premises from which its deductions were made, and to clear them of what had no better warrant than the disposition to use political economy to justify existing social arrangements. It was not needed to take the place of the deductive method, where that was applicable. For the deductive method, when applied to the further extension of what has already been validly ascertained, constitutes the most powerful means of extending knowledge that the human mind can avail itself of.
In its use of the deductive method after its premises had been settled, the classical political economy was not in error. The error that gave insecurity to its whole structure lay deeper still, in the insufficient inductions on which those premises rested. But, instead of addressing themselves to these flaws in its accepted premises, the various schools of economists generally classed as inductive have denied that there were any general principles that could with certainty be laid down as the basis for deduction. Thus, if such a question be asked them as, does free trade or protection best promote the general prosperity? Or, what is the best system of land tenure? Or, what is the best system of taxation? Or, what are the limits of governmental interference with industry, or trade union regulations? No general answer can be given. It can only be said that one thing may be best in one place and time, and another in another place and time, so that the matter can be determined only by special investigations.
But to me it seems clear that if political economy can be called a science at all, it must as a science, that is to say from the moment the laws of nature on which it depends are discovered, follow the deductive method of examination, using induction only to test the conclusions thus obtained. For the particulars which are included in its problems are too vast and too complex to admit of any hope of bringing them into order and relation by direct induction.
Now the law of nature which forms the postulate of a true science of political economy is not, as has been erroneously assumed, that men are invariably and universally selfish. As a matter of fact, this is not true. Nor can we abstract from man all but selfish qualities in order to make as the object of our thought on economic matters what has been called the "economic man," without getting what is really a monster, not a man.
The law of nature which is really the postulate of a true science of political economy is that men always seek to gratify their desires with the least exertion, whether those desires are selfish or unselfish, good or bad.
That this is a law of nature we have the highest possible warrant, wider in fact that we can have for any of the laws of external nature, such for instance as the law of gravitation. For the laws of external nature can be apprehended only objectively. But that it is a law of nature that men seek to gratify their desires with the least exertion, we may see both subjectively and objectively. Since man himself is included in nature, we may subjectively reach the law of nature that men seek to gratify their desires with the least exertion, by an induction derived from consciousness of our own feelings and an analysis of our own motives of action; while objectively we may also reach the same law by an induction derived from observation of the acts of others.
Proceeding from a law of nature thus doubly assured, the proper method of a political economy which becomes really a science by its correct apprehension of a fundamental law, is the method of deduction from that law, the method of proceeding from the general to the particular.
Thus, in the main, the science of political economy resorts to the deductive method, using induction for its tests. But in its more common investigations its most useful instrument is a form of hypothesis which may be called that of mental or imaginative experiment, by which we may separate, combine or eliminate conditions in our own imaginations, and thus test the working of known principles. This is a most common method of reasoning, familiar to us all, from our very infancy. It is the great working tool of political economy, and in its use we have only to be careful as to the validity of what we assume as principles.