An Introduction to the Study of Experimental Medicine

The ideas and theories of our predecessors must be preserved only in so far as they represent the present state of science, but they are obviously destined to change, unless we admit that science is to make no further progress, and that is impossible. In this connection, we should perhaps make a distinction between mathematical sciences and experimental sciences. As mathematical truths are immutable and absolute, the science of mathematics grows by sim-r pie successive juxtaposition of all acquired truths. As truths in the experimental sciences, on the contrary, are only relative, these sciences can move forward only by revolution and by recasting old truths in a new scientific form.

In the experimental sciences, a mistaken respect for personal authority would be superstition and would form a real obstacle to the progress of science: at the same time, it would be contrary to the examples given us by the great men of all time. Great men, indeed, are precisely those who bring with them new ideas and destroy errors. They do not, therefore, respect the authority of their own predecessors, and they do not expect us to treat them otherwise.

This non-submission to authority, which the experimental method regards as a fundamental precept, is by no means out of harmony with the respect and admiration which we bear to the great men preceding us, to whom we owe the discoveries at the base of the sciences of to-day.* In the experimental sciences, great men are never the promoters of absolute and immutable truths. Each great man belongs to his time and can come only at his proper moment, in the sense that there is a necessary and ordered sequence in the appearance of scientific discoveries. Great men may be compared to torches shining at long intervals, to guide the advance of science. They light up their time, either by discovering unexpected and fertile

* Claude Bernard, Cours de medecine ea^perimentale, legon d'ouverture {Gazette m4d., April 15, 1864.) phenomena which open np new paths and reveal unknown horizons, or by generalizing acquired scientific facts and disclosing truths which their predecessors had not perceived. If each great man makes the science which he vitalizes take a long step forward, he never presumes to fix its final boundaries, and he is necessarily destined to be outdistanced and left behind by the progress of successive generations. Great men have been compared to giants upon whose shoulders pygmies have climbed, who nevertheless see further than they. This simply means that science makes progress subsequently to the appearance of great men, and precisely because of their influence. The result is that their successors know many more scientific facts than the great men themselves had in their day. But a great man is, none the less, still a great man, that is to say, — a giant.

There are, indeed, two sides to science in evolution: on the one hand, what is acquired already, and on the other hand, what remains to be acquired. In the already acquired, all men are more or less equal, and the great cannot be distinguished from the rest. Mediocre men often have the most acquired knowledge. It is in the darker regions of science that great men are recognized; they are marked by ideas which light up phenomena hitherto obscure and carry science forward.

To sum up, the experimental method draws from within itself an impersonal authority which dominates science. It forces this authority even on great men, instead of seeking, like the scholastics, to prove from texts that they are infallible and that they have seen, said or thought everything discovered after them. Every period has its own sum total of errors and of truths. Certain mistakes are, in a sense, inherent in their period, so that only the subsequent progress of science can reveal them. The progress of the experimental method consists in this, — that the sum of truths grows larger in proportion as the sum of error grows less. But each one of these particular truths is added to the rest to establish more general truths. In this fusion, the names of promoters of science disappear little by little, and the further science advances, the more it takes an impersonal form and detaches itself from the past. To avoid a mistake which has sometimes been committed, I hasten to add that I mean to speak here of the evolution of science only. In art and letters, personality dominates everything. There we are concerned with a spontaneous creation of the mind, that has nothing in common with the

noting of natural phenomena, in which the mind must create nothing. The past keeps all its worth in the creations of art and letters ; each individuality remains changeless in time and cannot be mistaken for another. A contemporary poet has characterized this sense of the personality of art and of the impe]*sonality of science in these words, — "Art is myself ; science is ourselves." The experimental method is the scientific method which proclaims the freedom of the mind and of thought. It not only shakes off the philosophical and theological yoke ; it does not even accept any personal scientific authority. This is by no means pride and boastfulness ; experimenters, on the contrary, show their humility in rejecting personal authority, for they doubt their own knowledge also and submit the authority of man to the authority of experience and of the laws of nature.

Physics and chemistry, as established sciences, offer us the independence and impersonality which the experimental method demands. But medicine is still in the shades of empiricism and suffers the consequences of its backward condition. We see it still more or less mingled with religion and with the supernatural. Superstitution and the marvellous play a great part in it. Sorcerers, somnambulists, healers by virtue of some gift from Heaven, are held as the equals of physicians. Medical personality is placed above science by physicians themselves; they seek their authority in tradition, in doctrines or in medical tact. This state of affairs is the clearest of proofs that the experimental method has by no means come into its own in medicine.

The experimental method, the free thinker's method, seeks only scientific truth. Feeling, from which everything emanates, must keep its complete spontaneity and all its freedom for putting forth experimental ideas ; reason also must preserve that freedom to doubt, which forces it always to submit ideas to the test of experiment. Just as, in other human actions, feeling releases an act by putting forth the idea which gives a motive to action, so in the experimental method feeling takes the initiative through the idea. Feeling alone guides the mind and constitutes the yrimum movens of science. Genius is revealed in a delicate feeling which correctly foresees the laws of natural phenomena ; but this we must never forget, that correctness of feeling and fertility of idea can be established and proved only by experiment.

"We have so far dealt with the influence of the experimental idea. Let us now consider how the method, while always forcing upon reason the dubitative form, may guide it more safely in the search for truth. We said elsewhere that experimental reasoning is practised on observed phenomena, or observations ; but it is really applied only to the ideas which the phenomena have aroused in our mind. The essence of experimental reasoning, then, will always be an idea which we introduce into a piece of experimental reasoning in order to submit it to the criterion of facts, i.e., to experiment.

There are two forms of reasoning: first, the investigating or interroorative form used by men who do not know and who wish to learn; secondly, the demonstrating or affirmative form employed by men who know or think they know, and who wish to teach others. Philosophers seem to have differentiated these two forms of reasoning under the names of inductive reasoning and deductive reasoning. They also accept two scientific methods: the inductive method or induction, proper to the experimental physical sciences, and the deductive method or deduction, belonging more particularly to the mathematical sciences.

It follows that the one special form of experimental reasoning with which we must deal here is induction. Induction has been defined as the process of moving from the particular to the general, while deduction is the reverse process moving from the general to the particular. I certainly shall not presume to engage in a philosophic discussion which would here be out of place and beyond my competence; only in my capacity as experimenter I shall content myself with saying that it seems to me very difficult, in practice, to justify this distinction and clearly to separate induction from deduction. If the experimenter's mind usually proceeds by starting from particular observations and going back to principles, to laws, or to general propositions, it also necessarily proceeds from the same general propositions or laws and reaches particular facts which it deduces logically from these principles. Only, when a principle is not absolutely certain, we must always make a temporary deduction requiring experimental verification. All the seeming varieties of reasoning depend merely on the nature of

the subject treated and on its greater or less complexity. But in all these cases, the human mind always works in the same way, with syllogisms ; it cannot behave otherwise. Just as man goes forward, in the natural movement of his body, only by putting one foot in front of the other, so in the natural movement of his mind, man goes forward only by putting one idea in front of another. In other words, the mind, like the body, needs a primary point of support. The body's point of support is the ground which the foot feels; the mind's point of support is the known, that is, a truth or a principle of which the mind is aware. Man can learn nothing except by going from the known to the unknown ; but on the other hand, as science is not infused into man at birth, and as he knows only what he learns, we seem to be in a vicious circle, where man is condemned to inability to learn anything. He would be so, in fact, if his reason did not include a feeling for relations and for determinism, which are the criteria of truth ; but in no case can he gain this truth or approach it, except through reasoning and experience.

It would be incorrect to say that deduction pertains only to mathematics and induction to the other sciences exclusively. Both forms of reasoning, investigating (inductive) and demonstrating (deductive), pertain to all possible sciences, because in all the sciences there are things that we do not know and other things that we know or think we know. When mathematicians study subjects unfamiliar to them, they use induction, like physicists, chemists or physiologists. To prove this point, I need only cite the words of a great mathematician.

Thus Euler expresses himself in a memoir entitled: De inductione ad plenam certitudinem evehenda: ''Notum est plerumque numerum proprietates primum per solam inductionem observatas, quas deinceps geometrae solidis demonstrationibus confirmare elaboraverunt ; quo negotio in primis Fermatius summo studio et satis felici successu fuit occupatus." ^ The principles or theories which serve as foundations for a science, whatever it may be, have not fallen from the sky; they were necessarily reached by investigation, inductive or interrogative reasoning, as we may choose to call it. It was first necessary to observe

* Euler, Acta academiae scientiarum imperialis PetropoUtanae, pro anno MDCCLXXX, pars posterior, p. 38, Par. 1. something which happened within ourselves or outside of us. From the experimental point of view there are ideas, in the sciences, which we call a priori, because they are a starting point for experimental reasoning (see page 27 and the following pages), but from the point of view of ideogenesis they are really a posteriori ideas. In a word, induction must have been the primitive, general form of reasoning; and the ideas which philosophers and men of science constantly take for a priori ideas are at bottom really a posteriori ideas.

Mathematicians and naturalists are alike when going in search of principles. Both use induction, make hypotheses, and experiment, that is to say, make attempts to verify the accuracy of their ideas. But when mathematicians and naturalists reach their principles, then they part company. Indeed, as I have already said elsewhere, the mathematician's principle is absolute, because it is not applicable to objective reality just as it is, but to relations between things considered in extremely simple conditions which the mathematician chooses and, in some sort, creates in his mind. Now, as he is thus sure that he need not introduce into his reasoning other conditions than those which he has defined, the principle remains absolute, conscious, adequate for the mind, and his logical deduction is equally certain and absolute : he no longer requires experimental verifications; logic is enough.

A naturalist is in a very different position ; the general proposition which he has reached, or the principle on which he relies, is relative and provisional, because it embodies complex relations which he is never sure that he can know. Hence, his principle is uncertain, since it is unconscious and inadequate for the mind; hence, deductions, though quite logical, always remain doubtful, and so he must necesarily appeal to experiment to verify the conclusion of his deductive reasoning. The difference between mathematicians and naturalists is capital in respect to the certainty of their principles and of the conclusions to be drawn from them ; but the mechanism of deductive reasoning is exactly the same for both. Both start from a proposition; only the mathematician says: Given this starting point, such and such a particular case necessarily results. The naturalist says : If this starting point is correct, such and such a particular case will follow as a consequence.

When starting from a principle, the mathematician and tlie naturalist, therefore, both use deduction. Both reason by making a syllogism; only, for the naturalist the conclusion of the syllogism is doubtful and requires verification, because its principle is unconscious.^ Such experimental or dubitative reasoning is the only kind that we can use when reasoning about natural phenomena; if we wished to suppress doubt and if we dispensed with experiment, we should no longer have any criterion by which to know whether we were in the wrong or in the right, because, I repeat, the principle is unconscious, and one must therefore appeal to our senses.

From all this I should conclude that induction and deduction belong to all the sciences. I do not believe that induction and deduction are really two forms of reasoning essentially distinct. By nature man has the feeling or idea of a principle that rules particular cases. He always proceeds instinctively from a principle, acquired or invented by hypothesis; but he can never go forward in reasoning otherwise than by syllogism, that is, by proceeding from the general to the particular.

In physiology, a given organ always works through one and the same mechanism ; only, when the phenomenon occurs under different conditions or in a different environment, the function takes on a different aspect ; but fundamentally its character remains the same. In my opinion there is only one way of reasoning for the mind, just as there is only one way of walking for the body. But when a man goes ahead on solid flat ground, by a straight road whose whole extent he knows and sees, he advances toward his goal at an assured and rapid pace. On the contrary, when a man follows a winding road in the dark and over unknown hilly ground, he dreads precipices and goes forward cautiously, step by step. Before taking a second step, he must make sure that he has placed his first foot on a spot that is firm, then go forward in the same way verifying experimentally, moment by moment, the solidity of the ground, and always changing the direction of his advance according to what he encounters. Such is the experimenter who must never go beyond fact in his searching, lest he risk losing his way. In the two preceding examples the man goes forward over different ground and in varied surroundings, but he goes forward none the less by the same physiological method. In the same way, when an experimenter simply deduces relations from definite phenomena by means of known

•i.e., Not a postulate and not exclusively an affair of the mind. Translator's note. and established principles, his reasoning develops in a secure and necessary way, while, if he finds himself in the midst of complex relations and with the support only of vague, provisional principles, the same experimenter must then go forward cautiously and must submit to experiment each one of the ideas which he successively puts forward. But, in both these cases, the mind still reasons in the same way and by the same physiological method, only it starts from a more or less binding principle.

When any sort of phenomenon strikes us in nature, we work out our idea of the cause determining it. Man in his primal ignorance imagined divinities connected with each phenomenon. To-day men of science acknowledge forces or laws : it is they that govern phenomena. An idea that comes to us at the sight of a phenomenon is called a 'priori. Now we shall later easily show that this a priori idea, which rises in us a propos of a special fact, always contains implicitly and, in some sort, without our knowledge, a principle to which we tend to refer the special fact, so that when we think that we are moving from a special case to a principle, i.e., making an induction, we are really making a deduction; only the experimenter guides himself by an assumed or provisional principle which he alters moment by moment, because he is searching in almost total darkness. In proportion as we gather facts, our principles become more and more general and more secure; so we gain the certainty that we deduce. But nevertheless, in the experimental sciences, our principle must always remain provisional, because we are never certain that it includes only the facts and conditions of which we are aware. In short, our deductions are always hypothetical until verified experimentally. An experimenter, therefore, can never be in the position of the mathematician, precisely because experimental reasoning, by its very nature, is always dubitative. If we wish, we can call the experimenter's dubitative reasoning induction, and the mathematician's afiirmative reasoning deduction ; but the distinction will then apply to the certainty or uncertainty of our starting point in reasoning, not to the way in which we reason.

I will summarize the preceding paragraph by saying that there seems to me to be only one form of reasoning : deduction by syllogism. The mind, even if it wished, could not reason otherwise, and if this were the place for it, I might try to support my proposition by physiological arguments. But to find scientific truth, we, after all, have little need to know how our mind reasons; it is enough to let it reason naturally, and in that case it will always start from a principle to reach a conclusion. All we need do here is to insist on a precept which will always forearm the mind against the countless sources of error that may be met in applying the experimental method.

This general precept, one of the foundations of the experimental method, is doubt: it is expressed by saying that the conclusion of our reasoning must always remain dubitative when the starting point or the principle is not an absolute truth. We have seen that there is no absolute truth apart from mathematical principles; in all natural phenomena the principles from which we start, like the conclusions which we reach, embody only relative truths. The experimenter's stumbling block, then, consists in thinking that he knows what he does not know, and in taking for absolute, truths that are only relative. Hence, the unique and fundamental rule of scientific investigation is reduced to doubt, as great philosophers, moreover, have already proclaimed.

Experimental reasoning is precisely the reverse of scholastic reasoning. Scholasticism must always have a fixed and indubitable starting point; and, unable to find it either in outer things or in reason, it borrows it from some irrational source, such as revelation, tradition, a conventional or an arbitrary authority. The starting point once settled, scholastics or systematizers deduce logically all the consequences, even invoking as arguments observation or experience of facts when they are favorable ; the one condition is that the starting point shall remain immutable and shall not vary with their experiences and observations, but on the contrary that facts shall be so interpreted as to adapt themselves to it. Experimenters, on the contrary, never accept an immutable starting point ; their principle is a postulate, all of whose consequences they logically deduce, but without ever considering it absolute or beyond the reach of experiment. The chemists' elements are elements only until proof to the contrary. All the theories which serve as starting points for physicists, chemists, and with still more reason physiologists, are true only until facts are discovered which they do not include, or

which contradict them. When these contradictory facts are shown to be firmly established, far from stiffening themselves against experience, like the scholastics or systematizers, experimenters, on the contrary, hasten to safeguard their starting point, to modify their theory, because they know that this is the only way to go forward and to make progress in science. Experimenters, then, always doubt even their starting point ; of necessity they keep a supple and modest mind and accept contradiction, on the one condition that it be proved. Scholastics or systematizers never question their starting point, to which they seek to refer everything; they have a proud and intolerant mind and do not accept contradiction, since they do not admit that their starting point may change. Men of system are also dis^ tinguished from men of experimental science by the fact that the first impose their idea, while the second always give it just for what it is worth. Finally, another essential characteristic, which differentiates experimental reasoning from scholastic reasoning, is the fertility of the one and the sterility of the other. The scholastic who believes himself in possession of absolute certainty comes to naught ; this can easily be understood, since by his absolute principle, he puts himself outside of nature, in which everything is relative. The experimenter, on the contrary, who always doubts and who does not believe that he possesses absolute certainty about anything, succeeds in mastering the phenomena that surround him and in extending his power over nature. Man can do, then, more than he knows; and true experimental science gives him power only in showing him his ignorance. Possessing absolute truth matters little to the man of science, so long as he is certain about the relations of phenomena to one another.

Indeed, our mind is so limited that we can know neither the beginning nor the end of things; but we can grasp the middle, i.e., what surrounds us closely. Systematic or scholastic reasoning is natural to inexperienced, proud minds; it is only by thorough experimental study of nature that we succeed in acquiring the experimenter's doubting mind. That takes a long time ; of those who think they are following the experimental path in physiology and in medicine, many, as we shall see later, are still scholastics. As for me, I am convinced that only study of nature can give scholars a true perception of science. Philosophy, which I consider an excellent gymnastic for the mind, has systematic and scholastic tendencies in spite of itself, which would

be harmful to men of science properly so-called. After all, no method can replace that study of nature which makes true men of science: without that study, all that philosophers have said and all that I myself have repeated after them in this introduction would remain inapplicable and sterile. I do not think, therefore, as I said above, that it is very profitable for men of science to discuss definitions of induction and of deduction, nor, for that matter, the question whether we advance by one or the other of these so-called processes of mind. Baconian induction, however, has become famous and has been made the foundation of all scientific philosophy. Bacon was a great genius, and his great restoration of the sciences is sublime as an idea; we are captivated and carried along in spite of ourselves, in reading the Novum Organum and the Augmentum Scientiarum. We are fascinated by a medley of scientific gleams, clothed in the loftieist of poetic forms. Bacon felt the sterility of scholasticism ; he well understood and foresaw the importance of experiment for the future of the sciences. Yet Bacon was not a man of science, and he did not understand the mechanism of the experimental method. To prove this, it would be enough to cite the hapless attempts which he made. Bacon advises us to fly from hypotheses and theories; we have seen, however, that they are auxiliaries of the method, indispensable as scaffolding is necessary in building a house. Bacon, as is always the case, had extravagant admirers and detractors. Without taking one side or the other, I will say that, while recognizing Bacon^s genius, I believe no more than J. de Maistre ^ that he endowed the human intellect with a new instrument, and it seems to me, as to M. de Eemusat,^ that induction does not differ from the syllogism. Moreover, I believe that great experimenters appeared before all precepts of experimentation, as great orators preceded all treatises on rhetoric. Consequently, even in speaking of Bacon, it does not seem to me permissible to say he invented the experimental method, that method which Galileo and Torricelli so admirably practised and which Bacon never could use.

* Bacon, (Euvres, edition, Francois Riaux, Introduction, p. 30. •M. de Remusat, Bacon, sa vie, son temps et sa philosophie, 1857. Descartes, Discoura sur la m^thode. authority, he gives much more practical precepts for the experimenter than those that Bacon gives for induction. We have seen, indeed, that only doubt promotes experiment ; it is doubt, finally^ which determines the form of experimental reasoning. In connection v^ith medicine and the physiological sciences, however, it is important to determine at what point to apply doubt, so as to distinguish it from scepticism, and to show how scientific doubt becomes an element of the greatest certainty. The sceptic disbelieves in science and believes in himself ; he believes enough in himself to dare deny science and to assert that it is not subject to definite, fixed laws. The doubter is a true man of science; he doubts only himself and his interpretations, but he believes in science; in the experimental sciences, he even accepts a criterion or absolute scientific principled This principle is the determinism of phenomena, which is as absolute in the phenomena of living bodies as in those of inorganic matter, as we shall later assert (page 65).