On the Philosophy of Discovery, Chapters Historical and Critical

65. With such questions before us, as have now been suggested, I can see nothing but a most mischievous narrowing of the field and enfeebling of the spirit of scientific exertion, in the doctrine that "Deduction is the great scientific work of the present and of future ages;" and that "A revolution is peaceably and progressively effecting itself in philosophy the reverse of that to which Bacon has attached his name." I trust, on the contrary, that we have many new laws of nature still to discover; and that our race is destined to obtain a sight of wider truths than any we yet discern, including, as cases, the general laws we now know, and obtained from these known laws as they must be, by Induction.

66. I can see, however, reasons for the comparatively greater favour with which Mr. Mill looks upon Deduction, in the views to which he has mainly directed his attention. The explanation of remarkable phenomena by known laws of Nature, has, as I have already said, a greater charm for many minds than the discovery of the laws themselves. In the case of such explanations, the problem proposed is more definite, and the solution more obviously complete. For the process of induction includes a mysterious step, by which we pass from particulars to generals, of which step the reason always seems to be inadequately rendered by any words which we can use; and this step to most minds is not demonstrative, as to few is it given to perform it on a great scale. But the process of explanation of facts by known laws is deductive, and has at every step a force like that of demonstration, producing a feeling peculiarly gratifying to the clear intellects which are most capable of following the process. We may often see instances in which this admiration for deductive skill appears in an extravagant measure; as when men compare Laplace with Newton. Nor should I think it my business to argue against such a preference, unless it were likely to leave us too well satisfied with what we know already, to chill our hope of scientific progress, and to prevent our making any further strenuous efforts to ascend, higher than we have yet done, the mountain-chain which limits human knowledge.

67. But there is another reason which, I conceive, operates in leading Mr. Mill to look to Deduction as the principal means of future progress in knowledge, and which is a reason of considerable weight in the subjects of research which, as I conceive, he mainly has in view. In the study of our own minds and of the laws which govern the history of society, I do not think that it is very likely that we shall hereafter arrive at any wider principles than those of which we already possess some considerable knowledge; and this, for a special reason; namely, that our knowledge in such cases is not gathered by mere external observation of a collection of external facts; but acquired by attention to internal facts, our own emotions, thoughts, and springs of action; facts are connected by ties existing in our own consciousness, and not in mere observed juxtaposition, succession, or similitude. How the character, for instance, is influenced by various causes, (an example to which Mr. Mill repeatedly refers, ii. 518, &c.), is an inquiry which may perhaps be best conducted by considering what we know of the influence of education and habit, government and occupation, hope and fear, vanity and pride, and the like, upon men's characters, and by tracing the various effects of the intermixture of such influences. Yet even here, there seems to be room for the discovery of laws in the way of experimental inquiry: for instance, what share race or family has in the formation of character; a question which can hardly be solved to any purpose in any other way than by collecting and classing instances. And in the same way, many of the principles which regulate the material wealth of states, are obtained, if not exclusively, at least most clearly and securely, by induction from large surveys of facts. Still, however, I am quite ready to admit that in Mental and Social Science, we are much less likely than in Physical Science, to obtain new truths by any process which can be distinctively termed Induction; and that in those sciences, what may be called Deductions from principles of thought and action of which we are already conscious, or to which we assent when they are felicitously picked out of our thoughts and put into words, must have a large share; and I may add, that this observation of Mr. Mill appears to me to be important, and, in its present connexion, new.

XI. Fundamental opposition of our doctrines.—68. I have made nearly all the remarks which I now think it of any consequence to make upon Mr. Mill's Logic, so far as it bears upon the doctrines contained in my History and Philosophy. And yet there remains still untouched one great question, involving probably the widest of all the differences between him and me. I mean the question whether geometrical axioms, (and, as similar in their evidence to these, all axioms,) be truths derived from experience, or be necessary truths in some deeper sense. This is one of the fundamental questions of philosophy; and all persons who take an interest in metaphysical discussions, know that the two opposite opinions have been maintained with great zeal in all ages of speculation. To me it appears that there are two distinct elements in our knowledge, Experience, without, and the Mind, within. Mr. Mill derives all our knowledge from Experience alone. In a question thus going to the root of all knowledge, the opposite arguments must needs cut deep on both sides. Mr. Mill cannot deny that our knowledge of geometrical axioms and the like, seems to be necessary. I cannot deny that our knowledge, axiomatic as well as other, never is acquired without experience.

69. Perhaps ordinary readers may despair of following our reasonings, when they find that they can only be made intelligible by supposing, on the one hand, a person who thinks distinctly and yet has never seen or felt any external object; and on the other hand, a person who is transferred, as Mr. Mill supposes (ii. 117), to "distant parts of the stellar regions where the phenomena may be entirely unlike those with which we are acquainted," and where even the axiom, that every effect must have a cause, does not hold good. Nor, in truth, do I think it necessary here to spend many words on this subject. Probably, for those who take an interest in this discussion, most of the arguments on each side have already been put forwards with sufficient repetition. I have, in an "Essay on the Fundamental Antithesis of Philosophy," and in some accompanying "Remarks," printed²⁹⁰ at the end of the second edition of my Philosophy, given my reply to what has been said on this subject, both by Mr. Mill, and by the author of a very able critique on my History and Philosophy which appeared in the Quarterly Review in 1841: and I will not here attempt to revive the general discussion.

70. Perhaps I may be allowed to notice, that in one part of Mr. Mill's work where this subject is treated, there is the appearance of one of the parties to the controversy pronouncing judgment in his own cause. This indeed is a temptation which it is especially difficult for an author to resist, who writes a treatise upon Fallacies, the subject of Mr. Mill's fifth Book. In such a treatise, the writer has an easy way of disposing of adverse opinions by classing them as "Fallacies," and putting them side by side with opinions universally acknowledged to be false. In this way, Mr. Mill has dealt with several points which are still, as I conceive, matters of controversy (ii. 357, &c.).

71. But undoubtedly, Mr. Mill has given his argument against my opinions with great distinctness in another place (i. 319). In order to show that it is merely habitual association which gives to an experimental truth the character of a necessary truth, he quotes the case of the laws of motion, which were really discovered from experiment, but are now looked upon as the only conceivable laws; and especially, what he conceives as "the reductio ad absurdum of the theory of inconceivableness," an opinion which I had ventured to throw out, that if we could conceive the Composition of bodies distinctly, we might be able to see that it is necessary that the modes of their composition should be definite. I do not think that readers in general will see anything absurd in the opinion, that the laws of Mechanics, and even the laws of the Chemical Composition of bodies, may depend upon principles as necessary as the properties of space and number; and that this necessity, though not at all perceived by persons who have only the ordinary obscure and confused notions on such subjects, may be evident to a mind which has, by effort and discipline, rendered its ideas of Mechanical Causation, Elementary Composition and Difference of Kind, clear and precise. It may easily be, I conceive, that while such necessary principles are perceived to be necessary only by a few minds of highly cultivated insight, such principles as the axioms of Geometry and Arithmetic may be perceived to be necessary by all minds which have any habit of abstract thought at all: and I conceive also, that though these axioms are brought into distinct view by a certain degree of intellectual cultivation, they may still be much better described as conditions of experience, than as results of experience:—as laws of the mind and of its activity, rather than as facts impressed upon a mind merely passive.

XII. Absurdities in Mr. Mill's Logic.—72. I will not pursue the subject further: only, as the question has arisen respecting the absurdities to which each of the opposite doctrines leads, I will point out opinions connected with this subject, which Mr. Mill has stated in various parts of his book.

He holds (i. 317) that it is merely from habit that we are unable to conceive the last point of space or the last instant of time. He holds (ii. 360) that it is strange that any one should rely upon the à priori evidence that space or extension is infinite, or that nothing can be made of nothing. He holds (i. 304) that the first law of motion is rigorously true, but that the axioms respecting the lever are only approximately true. He holds (ii. 110) that there may be sidereal firmaments in which events succeed each other at random, without obeying any laws of causation; although one might suppose that even if space and cause are both to have their limits, still they might terminate together: and then, even on this bold supposition, we should no where have a world in which events were casual. He holds (ii. 111) that the axiom, that every event must have a cause, is established by means of an "induction by simple enumeration:" and in like manner, that the principles of number and of geometry are proved by this method of simple enumeration alone. He ascribes the proof (i. 162) of the axiom, "things which are equal to the same are equal to each other," to the fact that this proposition has been perpetually found true and never false. He holds (i. 338) that "In all propositions concerning numbers, a condition is implied, without which none of them would be true; and that condition is an assumption which may be false. The condition is that 1 = 1."

73. Mr. Mill further holds (i. 309), that it is a characteristic property of geometrical forms, that they are capable of being painted in the imagination with a distinctness equal to reality:—that our ideas of forms exactly resemble our sensations: which, it is implied, is not the case with regard to any other class of our ideas;—that we thus may have mental pictures of all possible combinations of lines and angles, which are as fit subjects of geometrical experimentation as the realities themselves. He says, that "we know that the imaginary lines exactly resemble real ones;" and that we obtain this knowledge respecting the characteristic property of the idea of space by experience; though it does not appear how we can compare our ideas with the realities, since we know the realities only by our ideas; or why this property of their resemblance should be confined to one class of ideas alone.

74. I have now made such remarks as appear to me to be necessary, on the most important parts of Mr. Mill's criticism of my Philosophy. I hope I have avoided urging any thing in a contentious manner; as I have certainly written with no desire of controversy, but only with a view to offer to those who may be willing to receive it, some explanation of portions of my previous writings. I have already said, that if this had not have been my especial object, I could with pleasure have noted the passages of Mr. Mill's Logic which I admire, rather than the points in which we differ. I will in a very few words refer to some of these points, as the most agreeable way of taking leave of the dispute.

I say then that Mr. Mill appears to me especially instructive in his discussion of the nature of the proof which is conveyed by the syllogism; and that his doctrine, that the force of the syllogism consists in an inductive assertion, with an interpretation added to it, solves very happily the difficulties which baffle the other theories of this subject. I think that this doctrine of his is made still more instructive, by his excepting from it the cases of Scriptural Theology and of Positive Law (i. 260), as cases in which general propositions, not particular facts, are our original data. I consider also that the recognition of Kinds (i. 166) as classes in which we have, not a finite but an inexhaustible body of resemblances among individuals, and as groups made by nature, not by mere definition, is very valuable, as stopping the inroad to an endless train of false philosophy. I conceive that he takes the right ground in his answer to Hume's argument against miracles (ii. 183): and I admire the acuteness with which he has criticized Laplace's tenets on the Doctrine of Chances, and the candour with which he has, in the second edition, acknowledged oversights on this subject made in the first. I think that much, I may almost say all, which he says on the subject of Language, is very philosophical; for instance, what he says (ii. 238) of the way in which words acquire their meaning in common use. I especially admire the acuteness and force with which he has shown (ii. 255) how moral principles expressed in words degenerate into formulas, and yet how the formula cannot be rejected without a moral loss. This "perpetual oscillation in spiritual truths," as he happily terms it, has never, I think, been noted in the same broad manner, and is a subject of most instructive contemplation. And though I have myself refrained from associating moral and political with physical science in my study of the subject, I see a great deal which is full of promise for the future progress of moral and political knowledge in Mr. Mill's sixth Book, "On the Logic of the Moral and Political Sciences." Even his arrangement of the various methods which have been or may be followed in "the Social Science,"—"the Chemical or Experimental Method," "the Geometrical or Abstract Method," "the Physical or Concrete Deductive Method," "the Inverse Deductive or Historical Method," though in some degree fanciful and forced, abounds with valuable suggestions; and his estimate of "the interesting philosophy of the Bentham school," the main example of "the geometrical method," is interesting and philosophical. On some future occasion, I may, perhaps, venture into the region of which Mr. Mill has thus essayed to map the highways: for it is from no despair either of the great progress to be made in such truth as that here referred to, or of the effect of philosophical method in arriving at such truth, that I have, in what I have now written, confined myself to the less captivating but more definite part of the subject.

CHAPTER XXIII.
Political Economy as an Inductive Science

(Moral Sciences.)—1. Both M. Comte and Mr. Mill, in speaking of the methods of advancing science, aim, as I have said, at the extension of their methods to moral subjects, and aspire to suggest means for the augmentation of our knowledge of ethical, political, and social truths. I have not here ventured upon a like extension of my conclusions, because I wished to confine my views of the philosophy of discovery to the cases in which all allow that solid and permanent discoveries have been made. Moreover in the case of moral speculations, we have to consider not only observed external facts and the ideas by which they are colligated, but also internal facts, in which the instrument of observation is consciousness, and in which observations and ideas are mingled together, and act and react in a peculiar manner. It may therefore be doubted whether the methods which have been effectual in the discovery of physical theories will not require to be greatly modified, or replaced by processes altogether different, when we would make advances in ethical, political, or social knowledge. In ethics, at least, it seems plain that we must take our starting-point not without but within us. Our mental powers, our affections, our reason, and any other faculties which we have, must be the basis of our convictions. And in this field of knowledge, the very form of our highest propositions is different from what it is in the physical sciences. In Physics we examine what is, in a form more or less general: in Ethics we seek to determine what OUGHT to be, as the highest rule, which is supreme over all others. In this case we cannot expect the methods of physical discovery to aid us.

But others of the subjects which I have mentioned, though strongly marked and influenced by this ethical element, are still of a mixed character, and require also observation of external facts of human, individual, and social conduct, and generalizations derived from such observations. The facts of political constitutions and social relations in communities of men, and the histories of such communities, afford large bodies of materials for political and social science; and it seems not at all unlikely that such science may be governed, in its formation and progress, by laws like those which govern the physical sciences, and may be steered clear of errors and directed towards truths by an attention to the forms which error and truth have assumed in the most stable and certain sciences. The different forms of society, and the principal motives which operate upon men regarded in masses, may be classified as facts; and though our consciousness of what we ourselves are and the affections which we ourselves feel are always at work in our interpretations of such facts, yet the knowledge which we thus obtain may lead us to bodies of knowledge which we may call Sciences, and compare with the other sciences as to their form and maxims.

(Political Economy.)—2. Among such bodies of knowledge, I may notice as a specimen, the science of Political Economy, and may compare it with other sciences in the respects which have been referred to.

M. Comte has given a few pages to the discussion of this science of Political Economy²⁹¹; but what he has said amounts only to a few vague remarks on Adam Smith and Destutt de Tracy; his main object being, it would seem, to introduce his usual formula, and to condemn all that has hitherto been done (with which there is no evidence that he is adequately acquainted) as worthless, because it is "theological," "metaphysical," "literary," and not "positive."

Mr. Mill has much more distinctly characterized the plan and form of Political Economy in his system²⁹². He regards this science as that which deals with the results which take place in human society in consequence of the desire of wealth. He explains, however, that it is only for the sake of convenience that one of the motives which operate upon man is thus insulated and treated as if it were the only one:—that there are other principles, for instance, the principles on which the progress of population depends, which co-operate with the main principle, and materially modify its results: and he gives reasons why this mode of simplifying the study of social phenomena tends to promote the progress of systematic knowledge.

Instead of discussing these reasons, I will notice the way in which the speculations of political economists have exemplified tendencies to error, and corrections of those tendencies, of the same nature as those which we have already noticed in speaking of other sciences.

(Wages, Profits, and Rent.)—3. We may regard as one of the first important steps in this science, Adam Smith's remark, that the value or price of any article bought and sold consists of three elements, Wages, Profits, and Rent. Some of the most important of subsequent speculations were attempts to determine the laws of each of these three elements. At first it might be supposed that there ought to be added to them a fourth element, Materials. But upon consideration it will be seen that materials, as an element of price, resolves itself into wages and rent; for all materials derive their value from the labour which is bestowed upon them. The iron of the ploughshare costs just what it costs to sink the mine, dig up and smelt the iron. The wood of the frame costs what it costs to cut down the tree, together with the rent of the ground on which it grows.

(Premature Generalizations.)—4. But what determines Wages?—The amount of persons seeking work, that is, speaking loosely, the population; and the amount of money which is devoted to the payment of wages. And what determines the population? It was replied,—the means of subsistence. And how does the population tend to increase?—In a geometrical ratio. And how does the subsistence tend to increase?—At most in an arithmetical ratio. And hence it was inferred that the population tends constantly to run beyond the means of subsistence, and will be limited by a threatened deficiency of these means. And the wages paid must be such as to form this limit. And therefore the wages paid will always be such as just to keep up the population in its ordinary state of progress. Here was one general proposition which was gathered from summary observations of society.

Again: as to Rent: Adam Smith had treated Rent as if it were a monopoly price—the result of a monopoly of the land by the landowners. But subsequent writers acutely remarked that land is of various degrees of fertility, and there is some land which barely pays the cultivator, if cultivating it he pay no rent. And rent can be afforded for other land only in so far as it is better than this bad land. And thus, there was obtained another general proposition; that the Rent of good land was just equal to the excess of its produce over the worst cultivable land.

Now these two propositions are examples of a hasty and premature generalization, like that from which the sweeping physical systems of antiquity were derived. They were examples of that process which Francis Bacon calls anticipation; in which we leap at once from a few facts to propositions of the highest generality; and supposing these to be securely established, proceed to draw a body of conclusions from them, and thus frame a system.

And what is the sounder and wiser mode of proceeding in order to obtain a science of such things? We must classify the facts which we observe, and take care that we do not ascribe to the facts in our immediate neighbourhood or specially under our notice, a generality of prevalence which does not belong to them. We must proceed by the ladder of Induction, and be sure we have obtained the narrower generalizations, before we aspire to the widest.

(Correction of them by Induction. Rent.)—5. For instance; in the case of the latter of the above two propositions—that Rent is the excess of the produce of good soils over the worst—that is the case in England and Scotland; but is it the case in other countries? Let us see. Why is it the case in England? Because if the rent demanded for good land were more than the excess of the produce over bad land, the farmer would prefer the bad land as more gainful. If the rent demanded for good land were less than the excess, the bad land would be abandoned by the farmer.

But all this goes upon the supposition that the farmer can remove from good land to bad, or from bad to good, or apply his capital in some other way than farming, according as it is more gainful. This is true in England; but is it true all over the world?

By no means. It is true in scarcely any other part of the world. In almost every other part of the world the cultivator is bound to the land, so that he cannot remove himself and his capital from it; and cannot, because he is not satisfied with his position upon it, seek and find a position and a subsistence elsewhere. On the contrary, he is bound by the laws and customs of the country, by constitution, history and character, so that he cannot, or can only with great difficulty, change his plan and mode of life. And thus over great part of the world the fundamental supposition on which rests the above generalization respecting Rent is altogether false.

An able political economist²⁹³ has taken the step, which as we have said, sound philosophy would have prescribed: he has classified the states of society which exist or have existed on the earth, as they bear on this point, the amount of Rent. He has classified the modes in which the produce is, in different countries and different stages of society, divided between the cultivator and the proprietor: and he finds that the natural divisions are these:—Serf Rents, that is, labour rents paid by the Cultivator to the Landowner, as in Russia: Métayer Rents, where the produce is divided between the Cultivator and the Landowner, as in Central Europe: Ryot Rents, where a portion of the produce is paid to the Sovereign as Landlord, as in India: Cottier Rents, where a money-rent is paid by a Cultivator who raises his own subsistence from the soil; and Farmers' Rents, where a covenanted Rent is paid by a person employing labourers. In this last case alone is it true that the Rent is equal to the excess of good over bad soils.

The error of the conclusion, in this case, arises from assuming the mobility of capital and labour in cases in which it is not moveable: which is much as if mechanicians had reasoned respecting rigid bodies, supposing them to be fluid bodies.

But the error of method was in not classifying the facts of societies before jumping to a conclusion which was to be applicable to all societies.

(Wages.)—6. And in like manner there is an error of the same kind in the assertion of the other general principles:—that wages are determined by the capital which is forthcoming for the payment of wages; and that population is determined in its progress by wages. For there is a vast mass of population on the surface of the earth which does not live upon wages: and though in England the greater part of the people lives upon wages, in the rest of the world the part that does so is small. And in this case, as in the other, we must class these facts as they exist in different nations, before we can make assertions of any wide generality.

Mr. Jones²⁹⁴ classed the condition of labourers in different countries in the same inductive manner in which he classed the tenure of land. He pointed out that there are three broad distinct classes of them: Unhired Labourers, who cultivate the ground which they occupy, and live on self-produced wages; Paid Dependants, who are paid out of the revenue or income of their employers, as the military retainers and domestic artizans of feudal times in Europe, and the greater part of the people of Asia at the present day; and Hired Labourers, who are paid wages from capital.

This last class, though taken as belonging to the normal condition of society by many political economists, is really the exceptional case, taking the world at large; and no propositions concerning the structure and relations of ranks in society can have any wide generality which are founded on a consideration of this case alone.

(Population.)—7. And again: with regard to the proposition that the progress of population depends merely on the rate of wages, a very little observation of different communities, and of the same communities at different times, will show that this is a very rash and hasty generalization. When wages rise, whether or not population shall undergo a corresponding increase depends upon many other circumstances besides this single fact of the increase of wages. The effect of a rise of wages upon population is affected by the form of the wages, the time occupied by the change, the institutions of the society under consideration, and other causes: and a due classification of the conditions of the society according to these circumstances, is requisite in order to obtain any general proposition concerning the effect of a rise or fall of wages upon the progress of the population.

And thus those precepts of the philosophy of discovery which we have repeated so often, which are so simple, and which seem so obvious, have been neglected or violated in the outset of Political Economy as in so many other sciences:—namely, the precepts that we must classify our facts before we generalize, and seek for narrower generalizations and inductions before we aim at the widest. If these maxims had been obeyed, they would have saved the earlier speculators on this subject from some splendid errors; but, on the other hand, it may be said, that if these earlier speculators had not been thus bold, the science could not so soon have assumed that large and striking form which made it so attractive, and to which it probably owes a large part of its progress.