The Basis of Induction

Lachelier, Jules. “The Basis of Induction.” The Journal of Speculative Philosophy 10 and 11 (January 1876 and 1877).

Induction is the operation by which we pass from the knowledge of facts to that of the laws which govern them. The possibility of this operation is doubted by none; and yet on the other side it seems strange that some facts, observed in a time and place thus determined, should suffice to establish a law which may be applicable to all places and to all time. The best experience teaches at most only how phenomena connect themselves under our eyes; but that they should connect themselves in the same manner always and everywhere—that, no experience can teach us, and yet we do not hesitate to affirm this. How is such an affirmation possible, and upon what is it founded? This is the question, equally as difficult as it is important, which we now mean to essay to solve.

Apparently the most natural solution consists in pretending that our mind passes from facts to laws by a logical process, which does not confound itself with deduction, but which rests as deduction does upon the principle of identity. Without doubt a law is not logically contained in any portion, be it small or great, of the facts which it regulates; but it seems as if it might be contained at least, in the whole of these facts, in their totality —and we might even say that it does not in reality differ from this totality, of which it is only the abridged expression. If this should be so, induction might be subject to some practical difficulties, but it would be in theory the simplest thing in the world. It would suffice to form, by force of time and patience, a complete collection of facts of each species. These collections once made, each law would establish itself by the institution of one term for several and would then be above the shadow of all contestations.

This opinion seems to be that of Aristotle, if we judge him according to the celebrated passage of the Analytics, where he represents induction under the form of a syllogism. The ordinary syllogism, or at least that of the first form, consists, as everybody knows, in the application of a general rule to a particular case; but how is this rule to be demonstrated, when it is not itself contained in a still more general rule? It is here that intervenes, according to Aristotle, the inductive syllogism, whose mechanism he explains by an example. It is proposed to demonstrate that animals without galls live a long time. We know, or are instructed to know, that man, horses and mules, are the only animals without galls, and we also know that these three sorts are long lived animals. We can reason therefore thus: Man, the horse, and the mule, live a long time. Now, the only animals without galls are man, the horse and the mule, therefore, all the animals without galls are long lived.

This syllogism is irreproachable and does not differ essentially from ordinary syllogisms of the first form; but it differs in matter, in that the middle, instead of being a general term, is a collection of particular terms. Now it is precisely this difference which expresses the essential character of the inductive conclusion; because this conclusion consists, contrary to the deductive conclusion, in drawing from the complete collection of particular cases a general rule, which is only a summary of the whole.

Whatever may be the bearing of this passage, it is easy to show that laws are not for us the logical result of a simple enumeration of facts. In truth, not only do we not hesitate to extend to the future laws which would represent at most under this hypothesis the totality of past facts; but a single fact carefully observed appears to us a sufficient basis for the establishment of a law, which at once embraces both the past and the future. There is then no conclusion properly so called, from facts to laws; hence the extent of the conclusion will exceed, and in most instances exceeds infinitely, the premises. Otherwise, each fact is contingent, considered in itself, and any sum of facts, however great, presents always the same character. A law, on the contrary, is the expression of a necessity, at least presumed; that is to say that it carries with itself the sequence that a certain phenomenon should follow or accompany such another, if always understood that we are not to take a simple coincidence for a law of nature. To conclude then from facts to laws, would he conclude not only from the particular to the universal, but yet more, from the contingent to the necessary, it is therefore impossible to consider induction as a logical operation.

As to the authority of Aristotle, it is much less decisive upon this point than it appears at first to be. It is evident, in fact, that Aristotle did not seriously admit that man, the horse and the mule were the only animals without galls, nor that it was possible in general to arrange a complete list of facts, or of individuals of a determined species; the syllogism which he describes supposes therefore, in his thought, a preparatory operation, by which we tacitly decide that a certain number of facts roof individuals may be considered as representatives of an entire species. Now it is evident from one side that this operation is induction itself, and from the other that it is founded not upon the principle of identity, since it is absolutely contrary to this principle to regard some individuals as the equivalent of all. In the passage cited, Aristotle preserves silence upon this operation; but he has described it in the last page of his Analytics, with a precision that leaves nothing to be desired. "We perceive," he says, individual beings, but the object proper to perception is the universal, the human being, and not the man called Callias. “Thus, from the avowal even of Aristotle, we conclude not from individuals to the species, but we see the species in each individual: the law is not for us the logical content of the fact, but the fact itself, seized in its essence, and under the form of universality. The opinion of Aristotle upon the passage of the fact to the law, that is to say, the essence itself of induction, is then directly opposed to that which we are disposed to attribute to him.

We are thus obliged to abandon the proposed solution, and to recognize that induction is not founded upon the principle of identity: this principle is, in truth, purely formal, that is to say, it truly authorizes us to announce under one form, what we have already announced under another, but it adds nothing to the contents of our knowledge. We have need, on the contrary, of a principle in some sort of material, which adds to the perception of facts, the double element of universality and necessity, which appears to us to characterize the conception of laws. To determine this principle, we shall make now the object and end of our research...

The existence of a special principle of induction has not escaped the notice of the Scotch school; but this school does not appear to have clearly seized the character and the value of it: "In the order of Nature," says Reid, "that which shall come, will probably resemble what has already come, under similar circumstances. “This declaration is inexact, and "probably" is superfluous. For it is perfectly certain that a phenomenon which produces itself under certain conditions, will produce itself continually, whenever all these conditions shall be reunited afresh. It is true that the vulgar deceive themselves nearly always about these conditions, and that science itself has great difficulty to assign them exactly; from thence it comes that our attempts are so often deceived, and that we know perhaps no law of Nature which does not suffer from some exceptions.

In fact, induction is always subject to error; in law (droit) she is absolutely infallible. For if it is not certain that the conditions which determine to-day the production of a phenomena, will determine it to-morrow^ the foresight founded upon an imperfect knowledge of these conditions would not even be probable. Royer Collard is happier when he founds induction upon two judgments, of which one announces the stability and the other the generality of the laws which govern the Universe: but scarcely has he posited this double principle, before he compromises it, or rather destroys it by the strange commentary heads to it. According to him, in truth these two judgments are neither necessary nor evident by themselves; the stability and generality of the laws of Nature are a fact for us, which we believe because it is so, and not because it would be absurd or impossible for it not to be so. But then who guarantees for us the existence of this double fact? Is it universal experience, or may it be, by chance, an induction anterior to that which it is requisite to explain? No, replies to Royer Collard, it is our nature herself. It is difficult to imagine a more complete confusion of ideas. Our nature cannot instruct us a priori of a fact of experience now outside of the experience of facts, there are for us only the truths of reason, of which the opposites are absolutely impossible. A judgment which is empirical, without being nevertheless necessary, is a veritable monster, which has no place in human intelligence. Reid seems to doubt his own principle. Royer Collard does not hesitate to pronounce, himself, the condemnation of his.

An illustrious savant of our day has formulated the fundamental axiom of Induction, in saying, that among living creatures as well as among bodies of dead matter, (corps brut), the conditions of existence of all phenomena are determined in an absolute manner. This expression is as just as it is precise and explains perfectly how our minds can pass from facts to laws; for if each phenomena produces itself under conditions absolutely invariable, it is clear that it suffices to know what these conditions are in any case, in order to know by that only, what they should been all. Only there is perhaps in nature room to distinguish two sorts of laws; the one applies to simple facts, as that which states that two equal and opposed forces will form an equilibrium; the others on the contrary announce between phenomena relations more or less complex, as that which declares that among living creatures the like will engender like. Nothing is less simple than the transmission of life, and it is certain that the formation of a new being demands a concourse of a prodigious number of physico-chemical actions. It is certain also that these actions do not always act themselves in the same way, because sometimes monsters are born from them. Now if we know only a priori that the same phenomena take place under the same conditions, we should confine ourselves to affirming that the product of each generation will resemble its authors, if all the conditions requisites are reunited; and whenever we pronounce contrarily, in absolute terms, that like engenders like, we evidently suppose, in virtue of some other principle, that all the conditions are reunited, at least in the majority of instances. It is this secondary principle which M. Claude Benard has, in some sort, personified in physiology, by calling it the directing or organic idea (idee directrice, ou organique); but it appears equally indispensable in brute matter as in organized beings. There is not, in fact, a single chemical law, which does not suppose, amidst the sensible phenomena whose relations it proclaims, the intervention of insensible phenomena whose mechanism is absolutely unknown to us; and to believe that this mechanism will act always in a way to produce the same results, is to admit in nature the existence of a principle of order which watches, as -we may say, over the existence of chemical species, as well as over that of living species. The conception of the laws of nature, with the exception of a small number of elementary laws, seems to be founded, therefore, upon two distinct principles: one in virtue of which the phenomena make a series in which the existence of the antecedent determines that of the successor; the other in virtue of which these series make in their turn, systems, in which the idea of all determines the existence of the parts. Now a phenomenon which determines another in preceding it, is what has been called from all time an efficient cause, and a whole which produces the existence of its own parts is, according to Kant, the true definition of a Final Cause. We are able to say then, in one word, that the possibility of Induction rests upon the double principle of Efficient Causes and Final Causes.

So far, we have limited ourselves to the search after the principle in virtue of which we pass from the knowledge of facts to that of laws. Now that we think we have found it, it is needful to establish that this principle is not an illusion but may lead us to a veritable knowledge of Nature. In a word, it is necessary that the establishment of the fact should follow the demonstration of the law. To demonstrate a principle may seeming truth rather a bold enterprise, and it is one which the Scotch Psychology has not accustomed us to undertake. They say, not without appearance of reasonableness, that proof cannot go as far as the Infinite, and that we must indeed come to a certain number of truths absolutely first, which are the basis even of our thought, and which impose themselves upon us in virtue of their own self-evidence. But without speaking of the difficulty which one has always found in determining the number of these first truths, what right have they to affirm that a proposition absolutely denuded of proofs, is a principle which expresses the constitution of the mind and of things, and that it may not be a pure prejudice the result of education and of habit! They allege the impossibility in which we are of conceiving the contrary of these truths; but the question is always that of knowing if this impossibility belongs to the nature of things or to the subjective disposition of our thought; and the skeptics of to-day reply reasonably that there has been a time when nobody believed that the earth turned around the sun. Without doubt it is absurd to suppose that principles may resolve themselves into other' more general principles which may serve them for proofs; for, either this resolution would go on to infinity, and the demonstration of principles would never be achieved, or it would end in a certain number of indemonstrable propositions, which would then be the veritable principles. But it is not necessary that all demonstration should proceed from the general to the particular; for even when the knowledge is most general in all, it remains still to be explained how this knowledge is found in our minds, and to be established also that it represents faithfully the nature of things. Now there is a means of resolving these two questions at once. It is to admit that our thought begins only generalities and abstractions; and to seek on the contrary, the origin of our knowledge in one or more concrete and singular acts, by which the thought constitutes itself by immediately seizing the reality. Either our science is but a dream, or the principles upon which it is founded are in their turn the expression of a fact, which is the fact even of the existence of the thought. It is then in this fact, and not in a primitive axiom, that we should essay to solve the principle upon which Induction rests.

It remains now to learn in what this first step consists, by which the thought enters into commerce with reality; and we are not able, it seems, to represent it to ourselves except in two ways, since contemporaneous philosophy admits only two definitions of reality itself. Either, in fact, reality consists exclusively of phenomena, and all knowledge is in the last analysis, a sensation; or reality is, in some sort, divided between phenomena and certain entities inaccessible to our senses, and in these cases human knowledge ought to burst forth at once from the sensible intuition of phenomena, and by a sort of intellectual intuition of these entities. We will go on then, adequately, in demonstrating the principle of Induction—from Experience, strictly so-called, to the intuition of things in themselves (choses en soi).and it is only in the event of discovering that neither of these two "ways will lead us to the conclusion sought for, that we will deem ourselves authorized to try a third way.


It is not necessary that we should essay to make for ourselves an empirical demonstration of the principle of Induction. This demonstration has already been given by Mr. Stuart Mill in his System of Logic, and as we could not possibly hope to do this better than he has done it, we will content ourselves with the examination of this. We must recognize in advance that an enterprise of building upon sensible experience a proposition which pretends to -the title of a principle, does not promise great chance of success, in spite of the skill of Mr. Mill; but the demonstration, even if insufficient, of a principle, after making all allowances against it, is of more value, and attests a thought more philosophic than the complete absence of all demonstration. For the rest, it is easy to infer that the principle demonstrated by Mr. Mill is not precisely that which we formulated above and presents neither exactly the same elements nor the same character. Egregiously speaking, there should be no more questioning the philosophy of experience, of efficient causes than of final causes. For, if our senses do not teach us that a series of phenomena may be directed to a certain end, neither can they teach us anymore, that each term in the series exerts upon the succeeding one any influence whatsoever. There is nothing to be astonished at in Mr. Mill's keeping absolute silence upon the finality we believe that we have discovered in phenomena; but in what sense can he say that one phenomena is cause of that which follows it, and thus found Induction upon what he calls the Law of Universal Causality? There is here a singular compromise between the exigencies of his system and the scientific tendencies of his mind. For, on one side he rejects as an illusion, all idea of a necessary connection, and in consequence all true causality: and, on the other, he does not hesitate to preserve the word and up to a certain point, the thing, in admitting between phenomena an order of succession absolutely invariable. Which constitutes, in fact, the most inflexible Determinism. He does not fear extending the empire of Determinism even so far as the human will; but he assures us at the same time that he does no wrong by this to free will, since the causes of our actions limit themselves to preceding them invariably, without exerting upon them any real influence. As to the character of the principle of Induction, there is evidently nothing in experience which could teach him that all phenomena should or must have an invariable antecedent, and his law of causality can only be the expression of fact; but, fact or law, as it may be, what must we think of the universality which Mr. Mill attributes to it? We find here a second compromise stranger than the first, between the needs of science and the logic of Empiricism. The law of causality is valuable, not only for our planetary system, but also for the group of stars of which our sun forms a part; it will be still in vigor not only in a hundred thousand years, but according to appearance, in a hundred million years; but beyond these limits, it may well be, that it will have the fate of the particular laws for which it serves as a basis, and that phenomena may succeed each other—as Mr. Mill expressly says—at hazard—that is an order of succession, contingent and limited to the phenomena upon which our thought can exert itself reasonably. Behold here definitely, all that the principle includes whose demonstration remains for us to examine. This demonstration seems to be very simple. "We only know facts immediately, and the sole means through which we can distinguish general truths from these facts (that may be contained in them) is induction; the principle of induction then must be in itself the result of an induction, without there being a circle to apprehend in this. In fact, there are two sorts of induction; the one is the scientific induction, which consists in erecting into a law one single fact, well instanced, and which supposes evidently that every fact is the expression of a law; the other is vulgar induction, which proceeds by a simple enumeration of examples, which supposes nothing before itself, and which consequently may very justly serve as a basis for the principle which serves in its turn to justify the first. It is true that since Bacon, this latter form of inductions abandoned as a process without value; and it is certain that it wants in confidence when it concerns particular laws of nature, because here the enumeration can never be complete, and one hundred examples confirming it does not exclude the possibility of one hundred contrary examples. But it is not the same when it concerns the law of Universal Causality. As there is not a single case in which it may not be applicable, there has not been a single fact, since man has watched Nature, which is not called upon either to confirm or contradict it; and as it has been confirmed by all without being contradicted by a single one, it rests upon a complete enumeration, and possesses an irrefragable certitude.

If there is not a circle in this demonstration, there is at least a begging question so manifest, that it is necessary to look twice before attributing it to a mind so penetrating as that offer. Mill. The enumeration of examples, they say, is never complete for the particular laws of nature. Is it any more so for the laws of Universal Causality? Can we assure ourselves that this law may never be contradicted, even within the limits already so narrow, of human experience? Have not men believed for a long time, following Mr. Mill himself, in a sort of partial and intermittent reign of chance? But in all these cases, the enumeration which he speaks of can only affect the past? Now it is needful to know whether the law of causality is valuable for the future, since this law should serve as a foundation for Induction, and that induction consists practically in a conclusion from the past to the future. We establish to-day a relation of succession between two phenomena, and we wish to know if the same relation will occur to-morrow. Yes, they say to us, because the phenomena have observed until now an absolutely invariable order of succession. But who knows whether they will be able to preserve it tomorrow? And if the particular laws of nature need to be guaranteed by the law of universal causality, in what superior law shall we search for the guarantee of this law itself?

But we take ill, perhaps, the thought of Mr. Mill. He has not perhaps believed that the inference of the future from the past, illegitimate and impossible in itself, in each particular case, becomes possible and legitimate in virtue of a general rule, founded upon a similar inference. He is persuaded, on the contrary, that man makes the induction spontaneously, and without the aid of any principle. He declares expressly that the law of universal causality, far from preceding in our minds the articulata’s of nature, follows and supposes them; and it is from these laws themselves that it draws, according to him, the authority which it needs in order to guarantee them. The spontaneous inductions which would suggest to the first men the regularity of the most ordinary phenomena, would not inspire them, really, with more than a mediocre confidence. They might believe, without being very sure of it, that all fire would burn and that all water quenches thirst; and when they are advised to reunite all these provisional laws under a common title, they have believed, without being more sure of it, that general phenomena are subjected to laws. But their confidence accrues naturally in the measure that experience confirms the result of their first inductions; and every fact which comes to confirm a particular law, deposes by that much in favor of the law of causality, which gathers thus to herself many favorable testimonies, as there are others collected. There is therefore nothing astonishing in that this law finishes by being invested with an absolute certitude, whilst others only attain by themselves to a degree of probability more or less elevated; and it is quite simple also that this certitude would react, in some sort, upon each one of these particular laws, of which the law of causality is at once the resume and the sanction. The principle of induction reposes then, neither upon a sterile accumulation of past facts, nor upon a system of laws capable of sufficing to themselves; it is the last utterance of a spontaneous induction, whose results, more or less probable whilst they remain isolated, become certain in being concentrated in a single one. It is the key to the arch which crowns and sustains at once the edifice of science.

Thus understood, the theory of Mr. Mill contains neither circle nor a beginning of the question (petitio principii); bat it reduces itself to two arbitrary suppositions, of which the second is (what is more important) contradictory. ^Ye do not see, to begin, how the result of spontaneous induction, only probable, if you choose, in all that touches upon the particular laws of nature, can become certain when it concerns the law of universal causality. This law, it is said, governs so many phenomena, and therefore it is confirmed by experience more often than all the rest put together. Admit that the probability of induction increases by virtue of success and in ratio of it, the number of proofs of causality favorable to the law, will always be finite, and therefore not able to clear the infinite distance which separates probability from certainty. To say that this law succeeds in all cases, is the abuse of an equivocation; because this expression can only be extended evidently to the past, and in order that it may include all cases without restriction, it would have to be certain that there would be no more facts ever to come, and consequently no further inductions to make. In the second place, what is this spontaneous induction, and what place does it occupy in a system where experience is presented as the unique source of our knowledge? Is it then one and the same thing to observe the production of a phenomena, and to judge that the same phenomena will reproduce itself in the same circumstances? But this is not all: in supposing that from the first observation (for the hundredth will not teach us any more on this point) men have a right to conclude from the past to the future, but is it that this conclusion was only probable at first? From two things come really one: either at the moment of this first observation, their minds contain nothing mortal the perception of an external fact, and there is nothing in this perception which could suggest the lightest anticipation of the future: or, to this perception they add, drawing apparently from their own recesses, the conception of a durable nexus between phenomena, and this conception, as all a priori judgment had an absolute value, which the ulterior results of experience can neither add to nor diminish.

There is a means of escape from all these embarrassments; but as this means is not expressly indicated in the work of Mr. Mill, we can only propose it, without knowing whether the illustrious author would have consented to subscribe to it. Suppose first that induction (spontaneous) is not a judgment declared by our thought upon the objective succession of phenomena, but a subjective disposition of our imagination to reproduce them in the order in which they have struck our senses. It may be granted without overleaping the limits of Empiricism, that this disposition, at first purely virtual, would develop in us under the influence of our first sensations; and we conceive at the same time that, feeble in its debut, it would be incessantly fortified by the invariable order in which all our sensations follow each other. Suppose in the second place, that probability consists for us in a powerful habit of the imagination, and certainty in an invincible habit: the passage from probability to certitude has no more, in its turn, anything of the inconceivable, provided that we do not attach to absolute a sense to the word invincible, and that we acknowledge that our belief in universal causality, founded on a prodigious number of impressions (confirmatory), may be shaken in the course of time by a repeated shock of contrary impressions. Logic in this case has nothing more to say; but what becomes of the science, that is to say, the objective knowledge of nature? Will Mr. Mill say that he does not admit the vulgar distinction between nature and our thought, that is to say, between the system of our sensations and a system of things in themselves (chosen soi)? But that which holds the place of nature in his doctrine, is our actual sensations, and not their traces which they leave after them in our imaginations. They are these sensations and not their images, between which science ought to establish the connection and foresee the return. Now because we have adopted the habit of associating in a certain order the images of our past sensations, does it follow that all our future sensations should follow in the same order? This interior nature, whose course does our order itself according to the play of our imagination, does it not escape from us in the same way as the external nature in which the vulgar believe? And the sequel of this theory—is it not pure skepticism, which destroys all reasonable foresight, and leaves us only a mechanical prudence like that of animals

For the rest, whether Mr. Mill desires it or not, it is certain that skepticism is the natural fruit, and the ever-renewed fruit of Empiricism. If nature is only for us a series of impressions, without reason and without connection, we can readily establish these, or rather submit to them at the moment they are produced; but we can neither predict, nor even conceive the future production of them. That which Empiricism calls our thought, by way of opposition to nature, is only a whole of enfeebled impressions which survive of themselves; and to search for the secret of the future in that which is the vain image of the past, is to undertake to discover in a dream what will happen to us during our waking hours. We wish to settle induction upon a solid basis. Do not let us search for her longer in a philosophy which is the negation of science.


It is strange that the school of M. Cousin should have, in general, considered the principle of induction as primitive and irreducible. For the doctrine of this school upon substance and its causes, should offer, it seems to us, an easy means of explanation. If, in truth, phenomena are sustained and produced by entities, abstracted from the vicissitudes of sensible existence, what would be more natural than to search in the uniform action of these entities, the reason for the constant succession of phenomena? And what more satisfactory than to attach the principle which serves as a basis for science to that which is regarded as a basis of metaphysics and of the supreme law of thought? In this school the principle of induction is formulated ordinarily, by saying that there is order in nature: but there is not given perhaps sufficiently precise idea of this order. Do they wish to say that the elementary phenomena which compose the hidden woof of things (la frame cachee des choses) are connected by virtue of an inflexible mechanism, which mechanism ought either to maintain or subvert the exterior and apparent order of nature? Do they mean to say, on the contrary, that nature is engaged in maintaining the harmony of beings, the distinction of species, organizations, life, and the means, in short, which she ought to take in order to attain to this 1' In a word, is order in the means or in the results? This question will no longer be doubtful if one consents to attach the idea of this order to the doctrine of substance and causes. It is believed, generally, that the number of these entities is equal to that of the constant groups of phenomena that we call beings; and their presence appears indispensable above all in organized beings, for which they are a principle, at one time both of unity and of action. Their function is not then to connect each phenomenon to a preceding one by the tie of a blind necessity, but rather to co-ordinate many series of phenomena following one law of agreement and of harmony; if these are not final causes in the sense of Aristotle and of Kant, they are at least causes which act for ends. The conception of universal order is then according to this doctrine, exclusively teleological. Now if it is important to men in power, to count upon the regularity of phenomena more or less complex, upon which their preservation depends, the proper object of science, that which she pursues to-day more ardently than ever, is, on the contrary, to determine the elementary conditions of these phenomena. She has need then of a principle which will guarantee to her the relations of causes to effects, rather than those of means to ends, of a principle of necessity rather than of harmony. If each sensible individual is the work of a thing-in-itself, (chose en soi), which employs his wisdom to conserve it, it suffices to establish by a superficial observation the ordinary results of this secret labor: but it is absurd to follow from experience-to-experience a mechanism of phenomena which will only serve to fetter it, and in which it will vanish even as far as the distinction between individual beings. The principle of universal order, thus understood, is the formal condemnation of science, properly so called.

Whatever may be the insufficiency of this principle, it is interesting to examine if the metaphysic of the school which has adopted it offers it at least a solid foundation. The difficulty does not consist in deducing the notion of universal order from that of things-in-themselves: For, besides that this latter notion is vague enough, all that one believes that he knows of the mode of existence and action of things is so fitted to explain the maintenance of an exterior order in nature, that we are tempted rather to see in it an ingenious hypothesis than a principle certain in itself. But this is not understood in this manner, and the existence of things-in-themselves is regarded as the cornerstone, and nearly as the whole edifice of metaphysics. Let us see how it is proved, and if indeed it is proved.

The moat simple process, if not the surest, is to invoke in favor of this existence, the witness of common sense. Can anyone conceive, it is sometimes asked, a property which does not reside in a substance, an event which may not be determined by a cause? Certainly not: but it is important to know what common sense means by cause or substance. All the world believes that an odor comes from an odoriferous body, and that aa savor belongs to a sapid body: but we should profoundly astonish man, a stranger to philosophic speculation, if we assured him that this body which strikes his eye, and which resists his effort, is itself only a superficial indication of an entity which can neither be seen nor touched. Substance, for the vulgar as for the savant, is synonymous with matter; and the conviction that all reality is material is so profoundly rooted in most men, that only moral or religious reasons can convince him to make an exception in favor of the human soul. As to the word cause, it signifies for them one phenomenon which determines another; they are not of the opinion of Mr. Mill, who admits only a relation of succession, without any real influence, between phenomena, but they are still further off from believing that phenomena appear or disappear at the will of mysterious beings, armed with sort of magic wand. Even examples which they use react against this doctrine, because, when a man has been assassinated, justice searches for the immediate cause of this event in the motion of a weapon pushed by a hand, and it is frightened from the pursuit of an entity which it would have small chance of catching. If one dared to speak in the language of Kant to commonsense, we might say that it believed firmly in substances and in phenomenal causes, but that it had not the slightest suspicion of noumena. If we should renounce the construction of commonsense upon a question which is after all strange to it, there only would remain, it seems, for us to sustain that we know substances and causes by an immediate intuition, analogous to that of sense; for to say we know because we do know without explaining how, is to avow that we know nothing and have nothing today. If we have no intuition of these entities, we have no idea of them, and the word which designates them has no sense; the affirmation of their existence even is without foundation, and the necessity which is alleged can have only a subjective and illusory character. We must leave to the Scotch school these verities of air, which impose themselves upon the mind in virtue of pretended evidence; and it is perhaps because the doctrine of substances and causes has for so long a time preserved this abstract form among us, that it has been judged useless to resolve the principle of universal order into a principle which had not any more solid foundation. On the other side, we must acknowledge that intuition, to which recourse has been had equally, has not so far furnished us with notions that are very precise upon the nature of these entities and upon the manner in which they operate. All that is known upon this latter point is that they develop or manifest themselves, that is simply to say that they contain the reason of sensible appearances; and, as to the first, not only is their essence still unknown, but their number even is so illy fixed, that one often employs the words substance and cause in the singular number; as if a phenomenon could be produced by the general idea of the cause, or as if all the phenomena were the immediate effect of a single and infinite cause. But if intuition scarcely teaches us anything about the substance and cause of a given phenomenon, it is still less fitted to teach us that all phenomena must have a substance and a cause: because it can have respect only to a determined object, and the intuition of a principle, outside of all actual application, is a contradiction in terms. The existence of a thing-in-itself outside of a phenomenon, even if it were given to us to perceive it, would be for us only a particular and contingent fact; and if all things should appear either in succession or at once before the eyes of our intelligence, this experience of a new kind worldly reveal to us a universal fact, not a necessary truth. It is then useless to attempt to found metaphysics upon what is called the principle of substance and the principle of cause; because if knowledge of things-in-themselves is intuitive, it cannot be clothed with the form of a principle, and if it is not, it cannot pretend to any objective value. Lately the influence of Maine de Biran has given birth in the school of M. Cousin to a middle theory, equally distant, it is believed at least, from an abstract dogmatism and from what may be called the empiricism of pure reason. According to this theory, and contrary to the primitive doctrines of the school, we seize immediately, not by reason but by consciousness, a substance and a cause, which is ourselves; and the office of reason limits itself to giving to this primitive knowledge a universal and necessary form, in revealing to us that the phenomena which are strange to us, have not less need of substance and cause than those of which we are the subject. But whether the operation of reason may be either primitive or secondary, it imports equally for us to prove that this operation is legitimate; and if it is demanded of us by what right we extend to all phenomena the conditions of existence of some, we have always to return to the one idea, whether it may be of a science without assignable origin, or whether it may be an intuition like that which is regarded as the exclusive privilege of consciousness. On the other hand, there may be raised some doubts upon the reality, or at least upon the extent of this privilege; and without contesting the original character of the notion of Ego, (moi) it is permitted to demand of oneself whether consciousness puts us in presence of a substance and a cause, in the sense in which these words are taken—that is to say—of a thing-in-itself, distinct from internal phenomena. It does not appear that we are well convinced of this—after all, since the spirituality and immortality of the soul are still continually established by arguments which this hypothesis, if it be verified, render absolutely worthless; and if it is incontestable that the Ego concentrates in its unity and enchains in its identity all diversity submitted to consciousness, perhaps it is just to see in this unity and this identity only the formal conditions of consciousness itself, and not the attributes of a substance charged to explain the apparition of it, and to guarantee the duration of it. It is not doubtful that our actions proceed freely and immediately from our faculty of willing; and from another point, if, as Leibnitz and Kant have taught, the succession of our internal states is not submitted to laws less rigorous than those of physical phenomena, we must acknowledge that we do not find within us, any more than outside of us, the trace of that absolute initiative which seems to characterize the action of a supra-sensible cause. But let us admit that we have consciousness of such an initiative. Is it then upon this model that causes distinct from us must be conceived, and are we able to confide the care of maintaining the order of nature to entities endowed with a liberty of indifference?

A later and profound modification of the doctrine of substances and causes consists in substituting for these two words that of Force, and of saying that we perceive immediately, by a sort of special sense, the conflict of our forces with the external forces. The fact is certainly established, but it is also certain that content with establishing the fact, the principle is not demonstrated: For the sense of which mention is made truly does teach us that our movement is produced by a force, and even makes us indirectly recognize the action of another force in the resistance we meet: but this sense is evidently powerless to teach us that all the movements which are executed in the Universe are produced or arrested by similar forces. Still more, when they speak of forces as things in themselves, we figure to ourselves under this name. I know not what sort of spiritual beings each one of whom is charged with the impulsion of the movement, whether it may be of a living body or of an inorganic mass: NowThis is a supposition which is not only gratuitous, but which is absolutely rejected and contradicted by experience. It may be said that a star in motion is animated by a single force, but it is absurd to represent this force as a simple and indivisible being: For if this star breaks into many fragments each of which continued to go on in its own orbit, we are obliged to acknowledge that the total force which animated it is decomposed into as many partial forces as there are fragments to impel. We know that our muscular effort can, under the influence of our will concentrate itself in one single effort, but we do not know if it proceeds from one single focus, or rather we do certainly know the contrary: because while a part of this- energy remains submissive to our control, another part may determine, in some one of our members convulsive movements which do not in themselves differ from voluntary motions. Thus, not only is there nothing which authorizes us to affirm that the Universe may bae system of forces, but the existence of our own force, in the sense in which the word is taken, is an unsustainable fiction. Force is no more a thing-in-itself, than extension from which it is, for the rest, inseparable, and the particular sensation which attests its presence in us, does not lift us one single step outside of or beyond the sphere of phenomena. Only when we are limited to saying that phenomena repose upon a substratum inaccessible to sense, if they do not give us a precise idea of that substratum, we are left at liberty at least to conceive it a tour will, or rather are determined almost irresistibly to look for the type in our own thought. When we believe on the contrary that we seize immediately this substratum in each voluntary effort it is declared without circumlocution that the tendency to movement proceeds only from itself: The chimerical entities in which itis essayed to realize it, do not linger, but vanish away and leaves definitely in presence of a pure phenomenon charged with explaining itself and also all others. A Metaphysic which looks for its sustaining point {point d’appui) in experience is very near its abdication in the hands of physical science.

The doctrine of substances and causes and that which recognizes nothing beyond phenomena is shipwrecked then equally upon the problem of Induction, but from different reasons". Empiricism attempts vainly to settle itself upon the solid but too narrow ground of phenomena: The contrary doctrine, giving a larger basis to this principle, builds upon empty air, and fails in establishing a necessity of thought, whilst thinking it satisfies it. Substances aid Causes are only a desideratum of the Science of Nature, a name given to the unknown reasons which maintain the order of the Universe, the enunciation of a problem transformed into a solution by an artifice of language. Of the two paths we have followed so far, and between which our choice seemed limited, neither have conducted us to any goal: Does there exist a third? Where shall we find it?


However embarrassing this question may appear at first view, our hesitation cannot be long, because we have absolutely only one part to take. Outside of phenomena and in default of distinct Entities there remains only the thought itself: It is then, in the thought, and in its relation with phenomena, that we should search for the basis of Induction. But before attempting a solution of this kind, let us essay to give a precise idea of it, and to dissipate in advance, the prejudices or hindrances it may awaken.

There are only three modes, possible in which principles may be presented, because there are only three modes of conceiving reality and the act by which our minds enter into commerce with themselves., We may admit with Hume and Mr. Mill that all reality is a phenomenon and that all knowledge is in last analysis, sensation: principles, if there may be question of principles in such an hypothesis, will then be only results, the most general results, of universal experience. We may suppose with the Scotch school and Mr. Cousin, that phenomena are only the manifestation of a world of Entities inaccessible to our senses: and in that case the principal source of our knowledge ought to be a sort of intellectual intuition, which would disarm at once and reveal to us, the nature of these entities and the action which they exert upon the sensible world. But there is a third hypothesis which Kant introduced into philosophy and which merits at least to be taken into consideration: it consists of in pretending that whatever may be the mysterious basis upon which phenomena rest, the order in which they succeed each other is determined exclusively by the exigencies of our own thought. The most elevated form of our knowledge is, in this hypothesis, neither an intellectual intuition nor a sensation, but a reflection, by which the thought seizes immediately its own nature and the relation it holds with phenomena: it is from this relation that we are able to deduce the laws which it imposes upon phenomena, and which are nothing less than principles.

It will be said that this- hypothesis is absurd and destroys itself, since each phenomenon cannot obey as many' different laws as there are distinct thoughts: but it is easy to reply, that here we consider only the faculty of thought in the mind, and that faculty is acknowledged to be identical in all, by the world. When we suppose, for instance, that principles exist in themselves, and outside of all thought or at least beyond the thought of all who inhabit a world like ours, we suppose that all thoughts like ours are equally capable of understanding these principles: It is not therefore wronging their universality to seek a basis in the very faculty through which they are known. But we shall be asked, how can we deny that the existence of principles may be independent of our consciousness, or how shall we conceive that the thought may be able to modify, in some measure the nature of its objects? It is true, that there is nothing impossible in a principle or a thing in general's existing outside of all commerce with our minds: but it will be granted to us, at least, that it is impossible for us to know anything about it, since a thing begins to exist for us only at the moment in which our minds enter into intercourse with it. We willingly grant upon our side, that the existence of principles is independent of our actual knowledge, and that they do not cease to be true because we cease to affirm them internally: but it suffices for that that there should be a reason which will determine us to affirm them every time that we do think of them, and that this reason may be found in our own faculty of knowledge or in things external to our minds. In short, we do not pretend that thought can modify by an arbitrary intervention, the nature of its objects: We assert only this, that in order that these objects should exist for us, they should possess in themselves a nature which would render possible the exercise of the thoughts. It is true, that it remains to know, whether thought is an empty capacity, which may be filled indifferently by all sorts of objects, or if the knowledge which we have of phenomena supposes one or several conditions upon their part: but we could not deny at least that in this latter case, these conditions ought to constitute, for all the phenomena, with which we have any business, the most inflexible of laws.

But the hypothesis which we propose is not only admissible in itself: it is the only admissible one, because it is the only one which permits us to comprehend how we can know a priori the objective conditions of the existence of phenomena. We may speak, it is true, of innate consciousness, which presents itself to our minds under a universal and necessary form: but it cannot be proved that this consciousness connects itself with its objects, and that it is a true knowledge, and not a vain dream. Today that there exists a sort of pre-established harmony between the laws of thought and those of reality is to resolve the question by the question itself: How, indeed are we able to know that our knowledge accords naturally with its objects, if we do not already know both the nature of the objects as well as that of our intellect?” It is needful therefore to recur to the direct intuition of reality, of which at least no one will contest the objective value: but whether this intuition bears upon simple phenomena, or upon things-in-themselves, it is equally certain that it cannot serve as foundation for principles, that is to say for universal and necessary knowledge. Things-in-themselves, which become objects of intuition for us, would be, in fact, only the phenomena of themselves: We might very well say what they were at the instant of appearance, but we could not question what they might be everywhere and always, nor above all could we declare what they could or might not be. But if the conditions of the existence of the phenomena are the conditions of the possibility of thought, we come easily out of this embarrassing alternative: because on one side, we can determine these conditions absolutely priori since they result from the nature of our mind itself; and we cannot doubt on the other hand, that they apply to the objects of experience, since, outside of these conditions, there is for us neither experience nor objects.

Now, how does this hypothesis, if we must call it so, permits to render an account in particular, of the principle of induction? We believe that we should resolve this principle into two distinct laws: one, according to which all phenomena are contained in a series, where the existence of each term determines that which follows it; the other according to which all phenomena is comprised in a system, where the idea of the whole, determines the existence of the parts. These are the two laws which it is needful to establish by showing that if they do not exist, human thought would be impossible: We shall begin with the first of\these.

The first condition of the possibility of thought is evidently the existence of a subject which distinguishes itself from each of our sensations: For if these sensations existed alone, they would entirely confound themselves with the phenomena, so that there would remain nothing that we might be able to call ourselves or our thought. The second is the unity of the subjecting the diversity of our sensations, as well simultaneous as successive: because a thought which was born, and which perished with each phenomenon, would be for us only a phenomenon itself, and we should have need of another subject in order together all these scattered and ephemeral thoughts into the unity of a real thought. Now, how can these two conditions be filled, or how can they represent to us the unity of the subject thinking and the relation it sustains with the diversity of its objects? Shall we say that the subject is a substance, of which the phenomena, or at least the sensations which represent them to us, are the modifications! No, because, after the idea we usually form of substances they only manifest themselves by their modifications, and cannot, in consequence, be distinguished from themes, a subject from an object. Shall we say that we are ourselves in our own eyes, a phenomenon, or rather a durable act, that of voluntary effort, which opposes itself by its duration, and by its active character to the passive and ephemeral modes of our sensibility? No, because this effort which renews itself at every awakening, or rather at every single instant, and which is probably only a bundle of actions exerted separately by everyone of our muscular fibers, does not present the character of absolute unity which appears to us indispensable to the subject of consciousness. Shall we search for the unity of this subject in that of a thought turned in upon itself, which contemplates itself outside of time and of all sensible modification? This hypothesis satisfies better than the preceding, the two conditions which we laid down above; but it seems to us still further removed from satisfying a third condition, which is nevertheless inseparable from the two others. We have indeed established that sensations without subject and without connection cannot constitute of themselves any consciousness: but it is evident that consciousness does not anymore consist in the solitary action of a subject shut up in itself, and external in some sort to its own sensations. It does not suffice therefore, to explain in a more or less plausible manner, how we are able to have consciousness of our own unity: it is necessary to show at the same time how this unity manifests itself, without dividing itself, in the diversity of our sensations, and thus constitutes thought, which is out only the thought of itself, but still more, that of the Universe. Now this is evidently impossible, if the subject thinking is given to itself by an act independent of all sensation and purely special: because not only could this simple and durable act, have not possibly anything in common with the multiple and successive acts which are related to phenomena, but we have no reason to believe that two functions, so strange on to the other, could be exercised by the same mind. The thought would find itself placed then before -its own existence as an insoluble enigma: because it could only exist if our sensations were able to unite themselves in a subject distinct from themselves, and a subject which distinguished itself from them, would seem by that, incapable of uniting them. There is, however, a means of escaping from this difficulty, and there is only this means: It is to admit that the unity which constitutes us, in our own eyes, is not that of an act but that of a form, and instead of establishing amongst our sensations an external and factitious connection, to say that it results from a sort of affinity and of cohesion natural to these sensations themselves. Now the relations natural to our sensations among themselves can be only those of the phenomena to which they correspond: The question then of knowing how all our sensations unite themselves in a single mind, is precisely the same as that of knowing how all the phenomena compose a single Universe. It is true that this latter unity is easier to admit than to comprehend: How, indeed, can several things, of which one is not the other, and which succeed each other, form one thing? Why an infinite number of phenomena, of which each occupies a distinct place in time and space, should be in our eye’s elements of a single world, and not of as many distinct worlds as they are different from each other is difficult to explain. Is it because these places, however different or distinct they may be between themselves, belong all to one single time and one single space? But what prevents our saying that space ends and begins with each of these bodies or rather atoms which occupy it, and that time dies and lives again at each vicissitude of the movements it measures? Space and time, despite the perfect similarity of their parts., are not in themselves one unity but on the contrary, an absolute diversity: and the unity which we attribute to them—far from serving for a basis for that of the Universe, can only repose itself upon the internal links of the phenomena which fill them. The question reduces itself then to the discovery of what makes this relation: and we are only able it seems, to represent to ourselves under this title an order of succession and of concomitance, in virtue of which the place of each phenomenon in time and in space may be assigned by relation to all the others. But always unity which results from such an order is still only a unity of fact, of which nothing guarantees to us the continuance: and we cannot even say that simple relations of time and place establish between phenomena, a veritable unity, in as much as these relations may vary at every instant, and that the existence of each phenomenon rests not only distinct, but still independent from that of others. It is not then in a contingent relation, but in a necessary connection, that we might be able to find at last the unity we look for: because, if the existence of a phenomenon is not only the constant sign, but still more, the determining reason of the other. these two existences are only then, two distinct moments of one existence, which continues itself by transforming the first phenomenon into the second. It is because all these simultaneous phenomena are, as Kant has said, in a reciprocal action, universal, that they constitute one single state of things, and that they are the object of a single thought upon our part; and it is because each one of these states is only, in some sort, but one new form of the preceding, that we are able to consider them as the successive epochs of a single history, which is at once that of thought and that of the Universe. All phenomena then, are submitted to the law of efficient causes, because that law is the only basis that we can assign to the unity of the Universe, and that this unity in its turn, is the supreme condition of the possibility of thought. But the law of efficient causes not only renders possible our knowledge of phenomena; It is also the only explanation which we can give of their objective existence, and that existence furnishes a new demonstration consequently of it.

We cannot seriously doubt that sensible things exist in themselves and continue to exist after we have ceased to feel them; and, on another side, we cannot understand that there can be a color without an eye to see it, a sound without an ear to hear it, and, in general a sensible phenomenon outside of any modification of our sensibility. It has been believed that the existence of the world might be assured by concentrating it, in some way, entire, within the phenomenon of resistance: but this phenomenon is as relative to what is called justly, the sense of effort, as the other qualities sensible to our other senses; and, if it has the privilege of making us know the distinction of our body from other strange bodies, it certainly has not that of surviving itself or of guaranteeing to us, that these bodies and ours will continue to exist, when we cease to have consciousness of their contact. We may say, at the risk of not comprehending it ourselves, that existence does not belong precisely to phenomena, but to the substances in which they reside. But whether we grant to skeptics that phenomena vanish with our sensations, and in that case, it is useless for us to preserve pretended entities, which are for us, as if they were not; or whether we hold with the vulgar, that the visible sun loses nothing of its brilliancy in quitting our horizon, it is then equally indifferent whether its disc subsists in itself, or reposes upon an entity inaccessible to our gaze. Perhaps by the substance of the sun, one means not, an entity distinct from the visible sun, but the enduring existence which we attribute to the sun itself, and which one wishes to distinguish from the passing impression which he produces upon our senses: but we find ourselves then in presence of the difficulty itself, which it is attempted to solve, and which consists in comprehending how a pure phenomenon can exist in itself and independent of all sensation. For the rest, we shall find looking closer at it, that such an existence is not seriously admitted by anyone: because, when we speak of a phenomenon which produces itself in the absence of all sensible existence either we deprive it of the form under which it offers itself ordinarily to our perception or we become ourselves, in despite of our own supposition, the imaginary spectators. We might be able, then, it seems, to limit ourselves, to the recognition that phenomena, or what is the same for us, our own sensations possess, beyond their actual existence, a sort of virtual existence, that is to say, that, even when we do not experience them, we might experience them, if we were placed in convenient conditions of time and place. One might be able even to suppose, with Leibnitz, that no phenomenon is absolutely excluded from our consciousness, and that not only the smallest parts, or the most distant parts, of the Universe are represented in us by some insensible perceptions, but that the past and the future are in some sort present to us, whether it be by the traces of past perceptions which mingle with our actual perceptions, or whether it be by the germ of future perceptions which an eye more piercing than ours might be able to discover in these very perceptions. We should make then, out of our own thought, according to an expression dear to Leibnitz, a Universe in abridgment; and we should be equally removed from the vulgar prejudice which places sensible things outside of all sensibility, and from the skeptical paradox which admits nothing beyond the grossest and most pronounced sensations, always, however, if we should succeeding procuring thus a sensible, sort of existence for the world, we must acknowledge that this existence is still altogether subjective and relative to our individual sensibility: because we cannot deny that common sense compels us, to distinguish sensible things not only from our actual sensations, but to detach them entirely from ourselves and to assure to them an existence absolute and independent from our own. Shall we say, with Leibnitz, that there exists an infinity of minds, each one of whom represents the same world to himself but under a differing point of view.' But minds which represent bodies, are not bodies; and otherwise, since we have any business only with our own representations, how shall we be able, not only to establish, but even to suspect, that there exist other minds outside of our own? For the rest, whatever the system may be, that is adopted we can never go outside of ourselves; we must then either shut ourselves up in a subjective idealism, very nearly related to skepticism at its best, or find in ourselves a basis capable of supporting at one time, the existence of the sensible world, and that of other minds. Now what can there be in us, which does not depend upon us, and which represents or rather which constitutes, an existence distinct from our own! This cannot be the phenomena themselves, which are only, at least for us, our own sensations. It is not their juxtaposition in space and their succession in time, since time and space seem to be only the forms of our own sensibility, and that it is impossible to assure ourselves that they may be anything else: but, if the place of each phenomenon in space and in time appears to us so determined by those which precede or which accompany it, that it is impossible for us to remove the thought of them, this necessary determination is doubtless something distinct from ourselves, since it imposes itself upon us, and it resists all the caprices of our imagination. Will it be said that this necessity resides itself in us, and that it is not less relative to our understanding than the phenomena themselves to our sensibility? Let there be shown to us then an existence, or in general, a truth pure from all relation to our thought: but let us be permitted to say, in the meantime, that we are, in so much as we are individual, only the whole of our sensations and that a necessity of which our sensations, as such, cannot render any account, constitutes by that itself, an existence as distinct from our own as one could reasonably demand. It is not because we feel certain phenomena, one after the other, that they necessarily link themselves in a chain, but, on the contrary, it is because they should develop themselves in a necessary order under the point of view which is special and particular to them: and, as soon as we recognize that the series of our sensations is only a particular expression of universal necessity we conceive at the least, the possibility of an infinity of analogous expressions, corresponding to as many points of view possible upon the Universe. The necessary determination of all phenomena is then at once for us the existence even of the material world and the only foundation that we can assign for that of other minds; and if one should prefer, in spite of all, to admit without proof existences absolutely external to our own, it is easy to show that one has more to lose than to gain by the change. The supposition of such existences has in truth, nothing impossible in itself: but if it is demanded what they are for us, it will be found that since they are situated outside of us, they can be given to us only by some impression they exert upon our intelligence: they will then only appear as a modification of ourselves and will become absolutely subjective precisely because it is wished that they should be absolutely objective. An existence is only objective for us when it is given to us in itself, and it cannot be given to us in itself except it leaps in some sort out of the bosom of our own existence: between the subjective idealism of Hume and the objective idealism of Kant, it is for common sense to choose.

As to the rest, if the law of efficient Causes explains at once our own knowledge of phenomena and the existence which we attribute to them, these two things are strictly united, and can only form in reality one thing. The property of thought is in reality, to conceive and to affirm the existence of objects: and it is evident that a thing can exist for us, at least, only when itis of the number of objects of thought. But thought is nothing in its own eyes outside of the necessity which constitutes the existence of phenomena. How otherwise would it have consciousness of them, if it is substantially distinct from them, and how will it represent this necessity itself, if not as a sort of blind thought pervading the things! We do not know either what maybe the existence of a thing-in-itself nor what consciousness we may be able to have of ourselves in another life: but in this world of phenomena of which we occupy the center, thought and existence are only two names of the universal and eternal necessity.


Not only does the law of efficient Causes result a priori from the relation of thought with phenomena, but this law permits us to determine in turn, by a new deduction the nature of phenomena themselves.

It is evidently necessary that the laws should be applied to phenomena, since otherwise they would have no signification; and this application could take place only by a simple act of thought, which conceives each law in perceiving the phenomena it governs. But, for this act to be truly simple, it is necessary that it should consist in seizing under two different forms, one only and the same thing, it is needful that the law should be only the abstract expression of the phenomena and that the phenomena should be only in their turn, the concrete expression of the law. Now this correspondence between phenomena and laws may be established in two ways: either the conception of laws is determined by the perception of phenomena, or it must be on the reverse, the perception of phenomena which governs the conception of laws.