From Auguste Comte, The Positive Philosophy (translated and condensed by Harriet Martineau), Vol. 2 (New York: D. Appleton & Co., 1854), 68-74 and 95-110.

If we look with a philosophical eye upon the present state of social science, we cannot but recognize in it the combination of all the features of that theologico-metaphysical infancy which all the other sciences have had to pass through. . . .

If we contemplate the positive spirit in its relation to scientific conception . . . we shall find that this philosophy is distinguished from the theologico- metaphysical by its tendency to render relative the ideas which were at first absolute. This inevitable passage from the absolute to the relative is one of the most important philosophical results of each of the intellectual revolutions which has carried on every kind of speculation from the theological or metaphysical to the scientific state. In a scientific view, this contrast between the relative and the absolute may be regarded as the most decisive manifestation of the antipathy between the modern philosophy and the ancient.

Men were long in learning that Man's power of modifying phenomena can result only from his knowledge of their natural laws; and in the infancy of each science, they believed themselves able to exert unbounded influence over the phenomena of that science . . . We see the metaphysical school . . . attributing observed events to chance, and sometimes, when that method is too obviously absurd, exaggerating ridiculously the influence of the individual mind upon the course of human affairs . . . It represents the social action of Man to be indefinite and arbitrary, as was once thought in regard to biological, chemical, physical, and even astronomical phenomenona, in the earlier stages of their respective sciences . . . There is no chance of order and agreement but in subjecting social phenomena, like all others, to invariable natural laws, which shall, as a whole, prescribe for each period, with entire certainty, the limits and character of political action--in other words, introducing into the study of social phenomena the same positive spirit which has regenerated every other branch of human speculation. Such a procedure is the true scientific basis of human dignity; as the chief tendencies of man's nature thus acquire a solemn character of authority which must be always respected by rational legislation; whereas the existing belief in the indefinite power of political combinations, which seems at first to exalt the importance of Man, issues in attributing to him a sort of social automatism passively directed by some supremacy of either Providence or the human ruler . . .

The last of the preliminary considerations that we have to review is that of the scientific prevision of phenomena, which, as the test of true science, includes all the rest. We have to contemplate social phenomena as susceptible of prevision, like all other classes, within the limits of exactness compatible with their higher complexity. Comprehending the three characteristics . . . we have been examining, prevision of social phenomena supposes first, that we have abandoned the region of metaphysical idealities, to assume the ground of observed realities by a systematic subordination of imagination to observation; secondly, that political conceptions have ceased to be absolute, and have become relative to the variable state of civilization, so that theories, following the natural course of facts, may admit of our foreseeing them; and, thirdly, that permanent political action is limited by determinate laws, since if social events were always exposed to disturbance by the accidental intervention of the legislator, human or divine, no scientific prevision of them would be possible. Thus, we may concentrate the conditions of the spirit of positive social philosophy on this one great attribute of scientific prevision . . .

The next step . . . is to examine . . . the means of investigation proper to Social Science . . . We may expect to find in Sociology a more varied and developed system of resources than in any other, in proportion to the complexity of the phenomena, while yet this extension of means does not compensate for the increased imperfection arising from the intricacy. The extension of the means is also more difficult to verify than in any prior case from the novelty of the subject; and I can scarcely hope that such a sketch as I must present here will command such confidence as will arise when a complete survey of the science shall have confirmed what I now offer.

As Social Physics assumes a place in the hierarchy of sciences after all the rest and therefore dependent on them, its means of investigation must be of two kinds: those which arise from the connection of sociology with the other sciences; and these last, though indirect, are as indispensable as the first. I shall review . . . the direct resources of the science.

Very imperfect and even vicious notions prevail at present as to what Observation can be and can effect in Social Science. The chaotic state of doctrine of the last century has extended to Method; and amidst our intellectual disorganization, difficulties have been magnified; precautionary methods, experimental and rational, have been broken up; and even the possibility of obtaining social knowledge by observation has been dogmatically denied; but if the sophisms put forth on this subject were true, they would destroy the certainty, not only of social science, but of all the simpler and more perfect ones that have gone before. The ground of doubt assigned is the uncertainty of human testimony; but all the sciences, up to the most simple, require proofs of testimony: that is, in the elaboration of the most positive theories, we have to admit observations which could not be directly made, nor even repeated, by those who use them, and the reality of which rests only on the faithful testimony of the original investigators; there being nothing in this to prevent the use of such proofs, in concurrence with immediate observations. In Astronomy, such a method is obviously necessary; it is equally, though less obviously necessary even in mathematics; and, of course, much more evidently in the case of the more complex sciences. How could any science emerge from the nascent state--how could there be any organization of intellectual labor, even if research were restricted to the utmost, if every one rejected all observations but his own? The stoutest advocates of historical skepticism do not go so far as to advocate this. It is only in the case of social phenomena that the paradox is proposed; and it is made use of there because it is one of the weapons of the philosophical arsenal which the revolutionary metaphysical doctrine constructed for the intellectual overthrow of the ancient political system. The next great hindrance to the use of observation is the empiricism which is introduced into it by those who, in the name of impartiality, would interdict the use of any theory whatever. No logical dogma could be more thoroughly irreconcilable with the spirit of the positive philosophy, or with its special character in regard to the study of social phenomena, than this. No real observation of any kind of phenomena is possible, except in as far as it is first directed, and finally interpreted, by some theory: and it was this logical need which, in the infancy of human reason, occasioned the rise of theological philosophy, as we shall see in the course of our historical survey. The positive philosophy does not dissolve this obligation, but, on the contrary, extends and fulfils it more and more, the further the relations of phenomena are multiplied and perfected by it. Hence it is clear that, scientifically speaking, all isolated, empirical observation is idle, and even radically uncertain; that science can use only those observations which are connected, at least hypothetically, with some law; that it is such a connection which makes the chief difference between scientific and popular observation, embracing the same facts, but contemplating them from different points of view: and that observations empirically conducted can at most supply provisional materials, which must usually undergo an ulterior revision. The rational method of observation becomes more necessary in proportion to the complexity of the phenomena, amid which the observer would not know what he ought to look at in the facts before his eyes, but for the guidance of a preparatory theory; and thus it is that by the connection of foregoing facts we learn to see the facts that follow. This is undisputed with regard to astronomical, physical, and chemical research, and in every branch of biological study, in which good observation of its highly complex phenomena is still very rare, precisely because its positive theories are very imperfect. Carrying on the analogy, it is evident that in the corresponding divisions, statical and dynamical, of social science, there is more need than anywhere else of theories which shall scientifically connect the facts that are happening with those that have happened: and the more we reflect, the more distinctly we shall see that in proportion as known facts are mutually connected we shall be better able not only to estimate, but to perceive those which are yet unexplored. I am not blind to the vast difficulty which this requisition imposes on the institution of positive sociology--obliging us to create at once, so to speak, observations and laws, on account of their indispensable connection, placing us in a sort of vicious circle, from which we can issue only by employing in the first instance materials which are badly elaborated, and doctrines which are ill-conceived. How I may succeed in a task so difficult and delicate, we shall see . . ., but, however that may be, it is clear that it is the absence of any positive theory which at present renders social observations so vague and incoherent. There can never be any lack of facts; for in this case even more than in others, it is the commonest sort of facts that are most important, whatever the collectors of secret anecdotes may think; but, though we are steeped to the lips in them, we can make no use of them, nor even be aware of them, for want of speculative guidance in examining them. The statical observation of a crowd of phenomena can not take place without some notion, however elementary, of the laws of social interconnection: and dynamical facts could have no fixed direction if they were not attached, at least by a provisional hypothesis, to the laws of social development. The positive philosophy is very far from discouraging historical or any other erudition; but the precious night-watchings, now so lost in the laborious acquisition of a conscientious but barren learning, may be made available by it for the constitution of true social science, and the increased honor of the earnest minds that are devoted to it. The new philosophy will supply fresh and nobler subjects, unhoped-for insight, a loftier aim, and therefore a higher scientific dignity. It will discard none but aimless labors, without principle and without character; as in Physics, there is no room for compilations of empirical observations; and at the same time, philosophy will render justice to the zeal of students of a past generation, who, destitute of the favorable guidance which we, of this day, enjoy, followed up their laborious historical researches with an instinctive perseverance, and in spite of the superficial disdain of the philosophers of the time. No doubt, the same danger attends research here as elsewhere: the danger that, from the continuous use of scientific theories, the observer may sometimes pervert facts, by erroneously supposing them to verify some illgrounded speculative prejudices of his own. But we have the same guard here as elsewhere--in the further extension of the science: and the case would not be improved by a recurrence to empirical methods, which would be merely leaving theories that may be misapplied but can always be rectified, for imaginary notions which can not be substantiated at all. Our feeble reason may often fail in the application of positive theories, but at least they transfer us from the domain of imagination to that of reality, and expose us infinitely less than any other kind of doctrine to the danger of seeing in facts that which is not.

It is now clear that Social Science requires, more than any other, the subordination of Observation to the statical and dynamical laws of phenomena: No social fact can have any scientific meaning till it is connected with some other social fact; without which connection it remains a mere anecdote, involving no rational utility. This condition so far increases the immediate difficulty that good observers will be rare at first, though more abundant than ever as the science expands; and here we meet with another confirmation of what I said at the outset . . .--that the formation of social theories should be confided only to the best organized minds, prepared by the most rational training. Explored by such minds, according to rational views of co-existence and succession, social phenomena no doubt admit of much more varied and extensive means of investigation than phenomena of less complexity. In this view, it is not only the immediate inspection or direct description of events that affords useful means of positive exploration; but the consideration of apparently insignificant customs, the appreciation of various kinds of monuments, the analysis and comparison of languages, and a multitude of other resources. In short, a mind suitably trained becomes able by exercise to convert almost all impressions from the events of life into sociological indications, when once the connection of all indications with the leading ideas of the science is understood. This is a facility afforded by the mutual relation of the various aspects of society, which may partly compensate for the difficulty caused by that mutual connection: if it renders observation more difficult, it affords more means for its prosecution.

It might be supposed beforehand that the second method of investigation, Experiment, must be wholly inapplicable in Social Science; but we shall find that the science is not entirely deprived of this resource though it must be one of inferior value. We must remember . . . that there are two kinds of experimentation--the direct and the indirect: and that it is not necessary to the philosophical character of this method that the circumstances of the phenomenon in question should be, as is vulgarly supposed in the learned world, artificially instituted. Whether the case be natural or factitious, experimentation takes place whenever the regular course of the phenomenon is interfered with in any determinate manner. The spontaneous nature of the alteration has no effect on the scientific value of the case, if the elements are known. It is in this sense that experimentation is possible in Sociology. If direct experimentation had become too difficult amidst the complexities of biology, it may well be considered impossible in Social Science. Any artificial disturbance of any social element must affect all the rest, according to the laws both of co-existence and succession; and the experiment would therefore, if it could be instituted at all, be deprived of all scientific value, through the impossibility of isolating either the conditions or the results of the phenomenon. But we saw . . . that pathological cases are the true scientific equivalent of pure experimentation, and why. The same reasons apply, with even more force, to sociological researches. In them, pathological analysis consists in the examination of cases, unhappily too common, in which the natural laws, either of harmony or of succession, are disturbed by any causes, special or general, accidental or transient; as in revolutionary times especially; and above all, in our own. These disturbances are, in the social body, exactly analogous to diseases in the individual organism: and I have no doubt whatever that the analogy will be more evident (allowance being made for the unequal complexity of the organisms) the deeper the investigation goes. In both cases it is . . . a noble use to make of our reason, to disclose the real laws of our nature, individual or social, by the analysis of its sufferings. But if the method is imperfectly instituted in regard to biological questions, much more faulty must it be in regard to the phenomena of Social Science, for want even of the rational conceptions to which they are to be referred. We see the most disastrous political experiments for ever renewed, with only some insignificant and irrational modifications, though their first operation should have fully satisfied us of the uselessness and danger of the expedients proposed. Without forgetting how much is ascribable to the influence of human passions, we must remember that the deficiency of an authoritative rational analysis is one of the main causes of the barrenness imputed to social experiments, the course of which would become much more instructive if it were better observed. The great natural laws exist and act in all conditions of the organism; for as . . . in the case of biology, it is an error to suppose that they are violated or suspended in the case of disease: and we are therefore justified in drawing our conclusions, with due caution, from the scientific analysis of disturbance to the positive theory of normal existence. This is the nature and character of the indirect experimentation which discloses the real economy of the social body in a more marked manner than simple observation could do. It is applicable to all orders of sociological research, whether relating to existence or to movement, and regarded under any aspect whatever, physical, intellectual, moral or political; and to all degrees of the social evolution, from which, unhappily, disturbances have never been absent. As for its present extension, no one can venture to offer any statement of it, because it has never been duly applied in any investigation in political philosophy; and it can become customary only by the institution of the new science which I am endeavoring to establish. But I could not omit this notice of it, as one of the means of investigation proper to social science.

As for the third of those methods, Comparison, the reader must bear in mind the explanations offered, in our survey of biological philosophy, of the reasons why the comparative method must prevail in all studies of which the living organism is the subject; and the more remarkably, in proportion to the rank of the organism. The same considerations apply in the present case, in a more conspicuous degree; and I may leave it to the reader to make the application, merely pointing out the chief differences which distinguish the use of the comparative method in sociological inquiries.

It is a very irrational disdain which makes us object to all comparison between human society and the social state of the lower animals. This unphilosophical pride arose out of the protracted influence of the theologico-metaphysical philosophy; and it will be corrected by the positive philosophy, when we better understand and can estimate the social state of the higher orders of mammifers, for instance. We have seen how important is the study of individual life, in regard to intellectual and moral phenomena--of which social phenomena are the natural result and complement. There was once the same blindness to the importance of the procedure in this case as now in the other; and as it has given way in the one case, so it will in the other. The chief defect in the kind of sociological comparison that we want is that it is limited to statical considerations; whereas the dynamical are, at the present time, the preponderant and direct subject of science. The restriction results from the social state of animals being, though not so stationary as we are apt to suppose, yet suceptible only of extremely small variations, in no way comparable to the continued progression of humanity in its feeblest days. But there is no doubt of the scientific utility of such a comparison, in the statical province, where it characterizes the elementary laws of social interconnection, by exhibiting their action in the most imperfect state of society, so as even to suggest useful inductions in regard to human society. There can not be a stronger evidence of the natural character of the chief social relations, which some people fancy that they can transform at pleasure. Such sophists will cease to regard the great ties of the human family as factitious and arbitrary when they find them existing, with the same essential characteristics, among the animals, and more conspicuously, the nearer the organisms approach to the human type. In brief, in all that part of sociology which is almost one with intellectual and moral biology, or with the natural history of Man; in all that relates to the first germs of the social relations, and the first institutions which were founded by the unity of the family or the tribe, there is not only great scientific advantage, but real philosophical necessity for employing the rational comparison of human with other animal societies. Perhaps it might even be desirable not to confine the comparison to societies which present a character of Voluntary cooperation, in analogy to the human. They must always rank first in importance: but the scientific spirit, extending the process to its final logical term, might find some advantage in examining those strange associations, proper to the inferior animals, in which an involuntary cooperation results from an indissoluble organic union, either by simple adhesion or real continuity. If the science gained nothing by this extension, the method would. And there is nothing that can compare with such an habitual scientific comparison for the great service of casting out the absolute spirit which is the chief vice of political philosophy. It appears to me, moreover, that, in a practical view, the insolent pride which induces some ranks of society to suppose themselves as, in a manner, of another species than the rest of mankind, is in close affinity with the irrational disdain that repudiates all comparison between human and other animal nature. However all this may be, these considerations apply only to a methodical and special treatment of social philosophy. Here, where I can offer only the first conception of the science, in which dynamical considerations must prevail, it is evident that I can make little use of the kind of comparison; and this makes it all the more necessary to point it out, lest its omission should occasion such scientific inconveniences as I have just indicated. The commonest logical procedures are generally so characterized by their very application, that nothing more of a preliminary nature is needed than the simplest examination of their fundamental properties.

To indicate the order of importance of the forms of society which are to be studied by the Comparative Method, I begin with the chief method, which consists in a comparison of the different coexisting states of human society on the various parts of the earth's surface-- those states being completely independent of each other. By this method, the different stages of evolution may all be observed at once. Though the progression is single and uniform, in regard to the whole race, some very considerable and very various populations have, from causes which are little understood, attained extremely unequal degrees of development, so that the former states of the most civilized nations are now to be seen, amid some partial differences, among contemporary populations inhabiting different parts of the globe. In its relation to Observation, this kind of comparison offers the advantage of being applicable both to statical and dynamical inquiries, verifying the laws of both, and even furnishing occasionally valuable direct inductions in regard to both. In the second place, it exhibits all possible degrees of social evolution to our immediate observation. From the wretched inhabitants of Tierra del Fuego to the most advanced nations of western Europe, there is no social grade which is not extant in some points of the globe, and usually in localities which are clearly apart. We shall find that some interesting secondary phases of social development, of which the history of civilization leaves no perceptible traces, can be known only by this comparative method of study; and these are not, as might be supposed, the lowest degrees of evolution, which every one admits can be investigated in no other way. And between the great historical aspects, there are numerous intermediate states which must be observed thus, if at all. This second part of the comparative method verifies the indications afforded by historical analysis, and fills up the gaps it leaves: and nothing can be more rational than the method, as it rests upon the established principle that the development of the human mind is uniform in the midst of all diversities of climate, and even of race; such diversities having no effect upon anything more than the rate of progress. But we must beware of the scientific dangers attending the process of comparison by this method. For instance, it can give us no idea of the order of succession, as it presents all the states of development as coexisting: so that, if the order of development were not established by other methods, this one would infallibly mislead us. And again, if we were not misled as to the order, there is nothing in this method which discloses the filiation of the different systems of society; a matter in which the most distinguished philosophers have been mistaken in various ways and degrees. Again, there is the danger of mistaking modifications for primary phases; as when social differences have been ascribed to the political influence of climate, instead of that inequality of evolution which is the real cause. Sometimes, but more rarely, the mistake is the other way. Indeed, there is nothing in the matter that can show which of two cases presents the diversity that is observed. We are in danger of the same mistake in regard to races; for, as the sociological comparison is instituted between peoples of different races, we are liable to confound the effects of race and of the social period. Again, climate comes in to offer a third source of interpretation of comparative phenomena, sometimes agreeing with, and sometimes contradicting the two others; thus multiplying the chances of error, and rendering the analysis which looked so promising almost impracticable. Here, again, we see the indispensable necessity of keeping in view the positive conception of human development as a whole. By this alone can we be preserved from such errors as I have referred to, and enriched by any genuine results of analysis. We see how absurd in theory and dangerous in practice are the notions and declamations of the empirical school, and of the enemies of all social speculation: for it is precisely in proportion to their elevation and generality that the ideas of positive social philosophy become real and effective--an illusion and uselessness belonging to conceptions which are too narrow and too special, in the departments either of science or of reasoning. But it is a consequence from these last considerations that this first sketch of sociological science, with the means of investigation that belong to it, rests immediately upon the primary use of a new method of observation, which is so appropriate to the nature of the phenomena as to be exempt from the dangers inherent in the others. This last portion of the comparative method is the Historical Method, properly so called; and it is the only basis on which the system of political logic can rest.

The historical comparison of the consecutive states of humanity is not only the chief scientific device of the new political philosophy. Its rational development constitutes the substratum of the science, in whatever is essential to it. It is this which distinguishes it thoroughly from biological science . . . The positive principle of this separation results from the necessary influence of human generations upon the generations that follow, accumulating continuously till it constitutes the preponderating consideration in the direct study of social development. As long as this preponderance is not directly recognised, the positive study of humanity must appear a simple prolongation of the natural history of Man: but this scientific character, suitable enough to the earlier generations, disappears in the course of the social evolution, and assumes at length a wholly new aspect, proper to sociological science, in which historical considerations are of immediate importance. And this preponderant use of the historical method gives its philosophical character to sociology in a logical as well as a scientific sense. By the creation of this new department of the comparative method, sociology confers a benefit on the whole of natural philosophy; because the positive method is thus completed and perfected, in a manner which, for scientific importance, is almost beyond our estimate. What we can now comprehend is that the historical method verifies and applies, in the largest way, that chief quality of sociological science--its proceeding from the whole to the parts. Without this permanent condition of social study, all historical labor would degenerate into being a mere compilation of provisional materials. As it is in their development, especially, that the various social elements are interconnected and inseparable, it is clear that any partial filiation must be essentially untrue. Where, for instance, is the use of any exclusive history of any one science or art, unless meaning is given to it by first connecting it with the study of human progress generally? It is the same in every direction, and especially with regard to political history, as it is called; as if any history could be other than political, more or less! The prevailing tendency to speciality in study would reduce history to a mere accumulation of unconnected delineations, in which all idea of the true filiation of events would be lost amid the mass of confused descriptions. If the historical comparisons of the different periods of civilization are to have any scientific character, they must be referred to the general social evolution: and it is only thus that we can obtain the guiding ideas by which the special studies themselves must be directed.

In a practical view, it is evident that the preponderance of the historical method tends to develop the social sentiment, by giving us an immediate interest in even the earliest experiences of our race, through the influence that they exercised over the evolution of our own civilization. As Condorcet observed, no enlightened man can think of the battles of Marathon and Salamis without perceiving the importance of their consequences to the race at large. This kind of feeling should, when we are treating of science, be carefully distinguished from the sympathetic interest which is awakened by all delineations of human life--in fiction as well as in history. The sentiment I refer to is deeper, because in some sort personal; and more reflective, because it results from scientific conviction. It can not be excited by popular history in a descriptive form; but only by positive history, regarded as a true science, and exhibiting the events of human experience in coordinated series which manifest their own graduated connection. This new form of the social sentiment must at first be the privilege of the choice few; but it will be extended, somewhat weakened in force, to the whole of society, in proportion as the general results of social physics become sufficiently popular. It will fulfill the most obvious and elementary idea of the habitual connection between individuals and contemporary nations, by showing that the successive generations of men concur in a final end, which requires the determinate participation of each and all. This rational disposition to regard men of all times as fellow-workers is as yet visible in the case of only the most advanced sciences. By the philosophical preponderance of the historical method, it will be extended to all the aspects of human life, so as to sustain, in a reflective temper, that respect for our ancestors which is indispensable to a sound state of society, and so deeply disturbed at present by the metaphysical philosophy.

As for the course to be pursued by this method--it appears to me that its spirit consists in the rational use of social series; that is, in a successive estimate of the different states of humanity which shall show the growth of each disposition, physical, intellectual, moral, or political, combined with the decline of the opposite disposition, whence we may obtain a scientific prevision of the final ascendency of the one and extinction of the other--care being taken to frame our conclusions according to the laws of human development. A considerable accuracy of prevision may thus be obtained, for any determinate period, and with any particular view; as historical analysis will indicate the direction of modifications, even in the most disturbed times. And it is worth noticing that the prevision will be nearest the truth in proportion as the phenomena in question are more important and more general; because then continuous causes are predominant in the social movement; and disturbances have less power. From these first general aspects, the same rational certainty may extend to secondary and special aspects, through their statical relations with the first; and thus we may obtain conclusions sufficiently accurate for the application of principles.

If we desire to familiarize ourselves with this historical method, we must employ it first upon the past, by endeavoring to deduce every well-known historical situation from the whole series of its antecedents. In every science we must have learned to predict the past, so to speak, before we can predict the future; because the first use of the observed relations among fulfilled facts is to teach us by the anterior succession what the future succession will be. No examination of facts can explain our existing state to us, if we have not ascertained, by historical study, the value of the elements at work; and thus it is in vain that statesmen insist on the necessity of political observation, while they look no further than the present, or a very recent past. The present is, by itself, purely misleading, because it is impossible to avoid confounding principal with secondary facts, exalting conspicuous transient manifestations over fundamental tendencies, which are generally very quiet; and above all, supposing those powers, institutions, and doctrines, to be in the ascendant, which are, in fact, in their decline. It is clear that the only adequate corrective of all this is a philosophical understanding of the past; that the comparison can not be decisive unless it embraces the whole of the past; and that the sooner we stop, in travelling up the vista of time, the more serious will be the mistakes we fall into. Before our very eyes, we see statesmen going no farther back than the last century, to obtain an explanation of the confusion in which we are living; the most abstract of politicians may take in the preceding century, but the philosophers themselves hardly venture beyond the sixteenth; so that those who are striving to find the issue of the revolutionary period have actually no conception of it as a whole, though that whole is itself only a transient phase of the general social movement.

The most perfect methods may, however, be rendered deceptive by misuse: and this we must bear in mind. We have seen that mathematical analysis itself may betray us into substituting signs for idea, and that it conceals inanity of conception under an imposing verbiage. The difficulty in the case of the historical method in sociology is in applying it, on account of the extreme complexity of the materials we have to deal with. But for this, the method would be entirely safe. The chief danger is of our supposing a continuous decrease to indicate a final extinction, or the reverse; as in mathematics it is a common sophism to confound continuous variations, more or less, with unlimited variations. To take a strange and very marked example: if we consider that part of social development which relates to human food, we can not but observe that men take less food as they advance in civilization. If we compare savage with more civilized peoples, in the Homeric poems or in the narratives of travellers, or compare country with town life, or any generation with the one that went before, we shall find this curious result. . . . The laws of individual human nature aid in the result by making intellectual and moral action more preponderant as Man becomes more civilized. The fact is thus established, both by the experimental and the logical way. Yet nobody supposes that men will ultimately cease to eat. In this case, the absurdity saves us from a false conclusion; but in other cases, the complexity disguises much error in the experiment and the reasoning. In the above instance, we must resort to the laws of our nature for that verification which, taken all together, they afford to our sociological analysis. As the social phenomenon, taken as a whole, is simply a development of humanity, without any real creation of faculties, all social manifestations must be found, if only in their germ, in the primitive type which biology constructed by anticipation for sociology. Thus every law of social succession disclosed by the historical method must be unquestionably connected, directly or indirectly, with the positive theory of human nature; and all inductions which can not stand this test will prove to be illusory, through some sort of insufficiency in the observations on which they are grounded. The main scientific strength of sociological demonstrations must ever lie in the accordance between the conclusions of historical analysis and the preparatory conceptions of the biological theory. And thus we find, look where we will, a confirmation of that chief intellectual character of the new science--the philosophical preponderance of the spirit of the whole over the spirit of detail.

This method ranks, in sociological science, with that of zoological comparison in the study of individual life; . . . the succession of social states exactly corresponds, in a scientific sense, with the gradation of organisms in biology; and the social series, once clearly established, must be as real and as useful as the animal series. When the method has been used long enough to disclose its properties, I am disposed to think that it will be regarded as so very marked a modification of positive research as to deserve a separate place; so that, in addition to Observation, properly so called, Experiment, and Comparison, we shall have the Historical Method, as a fourth and final mode of the art of observing. It will be derived, according to the usual course, from the mode which immediately precedes it: and it will be applied to the analysis of the most complex phenomena.

I must be allowed to point out that the new political philosophy sanctioning the old leadings of popular reason, restores to History all its scientific rights as a basis of wise social speculation, after the metaphysical philosophy had striven to induce us to discard all large consideration of the past. In the foregoing departments of natural philosophy we have seen that the positive spirit, instead of being disturbing in its tendencies, is remarkable for confirming, in the essential parts of every science, the inestimable intuitions of popular good sense; of which indeed science is merely a systematic prolongation, and which a barren metaphysical philosophy alone could despise. In this case, so far from restricting the influence which human reason has ever attributed to history in political combinations, the new social philosophy increases it, radically and eminently. It asks from history something more than counsel and instruction to perfect conceptions which are derived from another source: it seeks its own general direction, through the whole system of historical conclusions.