§4. These limitations to the use of the deductive method in Political Economy appear to me obvious and incontrovertible: and I have endeavoured always to keep them in view throughout the discussion of the laws of Distribution in Book II. I must admit, however, that they have not always been duly recognised by deductive economists; who have in consequence been led to make somewhat too sweeping assumptions as to concrete facts. I think that writers of the opposite school have done good service in criticizing these assertions, and the confident and dogmatic tone in which they have been enunciated. But I cannot accept the conclusion which some of them have proceeded to draw, that the traditional method of English Political Economy is essentially faulty and misleading. I quite admit that the direct utility of the deductive method, as a means of interpreting and explaining concrete facts---though not its validity, so long as it is regarded as merely abstract and hypothetical---depends on its being used with as full knowledge as possible of the results of observation and induction. But its indirect utility, as a means of training the intellect in the kind of reasoning required for dealing with concrete economic problems, depends to a far less degree on such empirical knowledge; and I cannot see that this indirect utility is materially affected by any divergences that have been shown to exist between the premises of current deductions and the actual facts of industry. On the other hand, I think that both the validity and the utility of the current deductions have been somewhat impaired by a want of thorough explicitness as to the assumptions on which these reasonings depend, and by a want of clearness in the cardinal notions employed in them. In order to guard against this latter defect, I have been led to perform with rather unusual elaborateness the task of defining the cardinal terms of Political Economy. The precise advantages that I have hoped to gain by this are explained in the second chapter of the following book, in which the task is commenced: I trust that I shall convince the reader that the process, however tedious, is absolutely indispensable to that exact treatment of economic questions, to which alone the epithet `scientific' ought to be applied.
Here I may notice the discussion that has recently been raised on an issue still wider than that debated between the advocates of the ``à priori'' economics, and the Inductive or ``realistic'' school; viz. on the pretensions of Political Economy to be a science at all. I certainly think the language sometimes used by economic writers, suggesting as it does that the doctrines they expound are entitled in respect of scientific perfection to rank with those of Physics, is liable to be seriously misleading. But I am not disposed to infer from this that we ought deliberately to acquiesce in treating Political Economy unscientifically. My inference would rather be, not that we ought not to aim at being as scientific as we can, but that we ought to take care not to deceive ourselves as to the extent to which we have actually attained our aim: that, for instance, so far as we are treating Political Economy positively, we should avoid mistaking a generalisation from limited experience for a universal law; and so far as we are treating it hypothetically, we should take care not to use words in different meanings without being aware of the difference, nor suppose our notions to be quantitatively precise when they are really indefinite. The endeavour to be scientific in this sense will not lead to hasty and mistaken dogmatism; on the contrary, it will, I hope, deliver us from the hasty and mistaken dogmatism, caused by loose and confused thinking, to which `common sense' or `natural intelligence' is always liable.[Back to:]