So common is the misunderstanding of scientific method that we must attempt here to put the matter ; clearly as possible in a very few words.
First, not even the “exact” sciences like astronomy or physics are exact in the sense of “absolute" or “infallible.” Their firmest laws or uniformities are to be regarded as tentative. They may be upset at any time by further work. But at any given moment they are not to be tampered with unless they prove unreliable in relation to observed facts. In the contemporary revolution in physics, Newton’s laws have not been “disproved”; nor has the principle of indeterminacy been so firmly established as to make all men equal before the game of poker. What has happened in modern physics, as far as a layman can judge, is that the physicist has been sharply reminded that even his neatest uniformities are not absolutes, but are subject to correction, that he is safer in regarding these uniformities as based on observations rather than on the will of God, or the nature of things, or on reality. Or more radically, he should regard these uniformities as his invention rather than his discovery. .
This brings us easily to the second point. Science makes no attempt to study or describe reality-certainly not ultimate reality. Science is not even concerned with truth, in the sense that word has for theologians, for most philosophers, for a good many other people, and perhaps for common sense. The desire to find a final cause, an unmoved mover, a Ding an sich, seems to be so common in men that we have no grounds for believing that this search is not, in one form or another, a fairly constant and in fact essential element in human society. Only, scientists as scientists can have no part in such a search. Eddington, Jeans, even Whitehead, ceased to practice science while they were pursuing theology. Science is based, not on faith, but on skepticism, on a skepticism that will not even worry itself over its status in the universe. And so the scientist works on serenely, undisturbed by the philosopher’s final thrust: that to be constantly skeptical is to believe in doubt, which is after all a form of faith.
Third, the scientist by no means confines himself to “the facts and nothing but the facts.” Dangerous epistemological depths yawn at this point, but we shall have to try and go ahead in spite of them. The popularization of Baconian ideas on induction is probably the chief source of the erroneous notion that the scientist does nothing to the facts he laboriously and virtuously digs up, except to let them fall neatly into a place they make for themselves. Facts themselves are not just “out there" and we should be willing to accept L. J. Henderson’s definition of “fact” as an empirically verifiable statement (italics mine) about phenomena in terms of a conceptual scheme. Actually the scientist cannot work without a conceptual scheme; and though the relation between facts and conceptual schemes is not by any means clear, it is at least clear that a conceptual scheme involves something besides facts, involves, indeed, a working mind.
Let no one be frightened of the term "conceptual scheme." The meaning is really very simple: thunder and lightning impinge on our senses of hearing and sight probably the mere differentiating of this sound and this flash from other sounds and flashes means that we are employing a conceptual scheme. Certainly when we think of Jupiter with his bolts, Thor with his hammer, or the electrical discharge of modern physics, we have clearly arranged our sense-perceptions in accordance with definite conceptual schemes. We possess, indeed, the basic elements of three different theories of thunder and lightning, three differently stated uniformities in these phenomena. But the crucial reason why we should prefer our electrical discharge to Jupiter or Thor as a conceptual scheme is that it is more useful, and that we can by using it get on better also with other conceptual schemes we use for similar purposes. But in the sense which the word true has for the theologian, and most moralists and philosophers, our electrical discharge is not a bit truer than the old notions about Jupiter and Thor.
We may even use two contradictory conceptual schemes, choosing one or the other of convenience, or from habit. We are all of us educated out of the old Ptolemaic conceptual scheme, which saw the sun moving about a stationary earth, into the Copernican conceptual scheme, which sees the earth moving about a stationary sun. Einstein, of course, used a conceptual scheme somewhat different from both of these, but most of us are not yet up to Einstein; in daily life we all, however, contentedly say “the sun rises,” and should be very pedantic indeed if we insisted on saying in Copernican terms “the earth has revolved into sight of the sun.” .
The scientist, then, goes to work roughly in some such fashion as this. He starts with a conceptual scheme of some sort, and with questions, or even hypotheses, which he frames in terms of that scheme. He then hunts for a suitable supply of facts. These facts he seeks to arrange in uniformities or theories which will answer his questions, and 'perhaps suggest other questions. He then immerses himself in the hunt for facts, and emerges with new or ed uniformities. The scientist is not interested in where his conceptual scheme came from, or whether it preceded or followed on facts, or whether it is “subjective” and the facts “objective.” These questions he leaves to the philosophers, who have not settled them yet after two thousand years of debate. But the scientist does, by his recognition that a conceptual scheme is as essential to his activity as are observed facts, emancipate himself completely from self-styled “scientific” materialists, positivists, empiricists, who naively assert that our sense-perceptions filtered through a “mind” are somehow in themselves an orderly and sole reality, or a “reflection” of such a reality. For, note particularly, the facts with which the scientist deals are not phenomena, sense-perceptions, the “external world,” those dear absolutes of innocent positivists, but merely statements about phenomena. A properly verifiable statement about Cromwell or Lenin is then as much a fact as the reading of a thermometer in a laboratory. We cannot here go into the thorny problem of what satisfactory verification is; the practicing scientist, the practicing historian, the practicing judge (and, one hopes, jury) have their own well-tried craft-methods of verification.
Fourth, though the scientist is very careful indeed about matters of definition, and is as disdainful of sloppiness as any historian and of bad thinking as any logician, be distrusts rigidity and attempts at perfection. He is interested less in beauty and neatness of definition than in having his definitions fit not his sentiments and aspirations, but the facts. Above all, he does not dispute over words. He is less interested in the accurate theoretical distinction between a mountain and a hill than he is in making sure that he is ‘ dealing with concrete elevations on this earth. He does not expect class terms to be perfect, mutually exclusive; when he distinguishes between a plant and an animal, he is not at all offended if you call his attention to a living thing that seems to belong to both his classifications at once. He sets to work studying the living thing and will, if necessary, modify his class terms. But he is quite willing also, if it proves more convenient, to set up a new class term of borderline plant-animals. This simple willingness to be guided by convenience is of course one of the amazing things about the scientist and one of the most difficult for those who have not had a scientific training to adapt ourselves to. Most of us are early trained to prefer our opinions to our convenience.
Fifth, perfectly respectable scientific work can be and is constantly done in fields where the kind of controlled experimentation classically associated for instance with physics and chemistry is not possible. We may call this sort of scientific work, based indeed on experimental work auxiliary to it, but not in itself a series of controlled experiments, clinical. The clinician is best known in the medical sciences, where he appears very early in fifth-century Greece with Hippocrates. The clinician works through the case method. His data are amassed, not through experiments he can control, but through a series of cases which he observes and compares. The clinician, again, is not sloppy; but he can rarely be rigorously exact. He is helped greatly when he can draw on experimental sciences-organic chemistry, for example. But in his own right the good clinician is a good scientist. Obviously the social sciences can depend but to a limited extent on actual controlled experimentation; but they can be clinical sciences depending more on observation than on experimentation.
Finally, scientific thinking cannot be, except perhaps in suggesting problems, what nowadays most of us know well enough as wishful thinking. The scientist’s own hopes and fears, his own standards of what he would like to have prevail on this earth must be kept as far as possible out of his work, and especially out of his observations of, or dealing with, facts. How far such hopes and fears and standards enter into his choice of conceptual schemes, how far they influence the kind of questions he asks, are difficult problems we may perhaps be permitted to dodge. Sufficient that the techniques of most of the established sciences provide a very effective check on the cruder forms of wishful thinking. History, which because it has been so long an art and a craft is perhaps the most respectable of the social sciences, provides in the technical training undergone by most professional historians a surprisingly effective and not wholly dissimilar check on the more violent forms of partisan writing and thinking.
All in all, there is no reason why we should feel that the natural scientist uses methods, set standards, forever quite unattainable by the social scientist. Natural science as the more innocent materialists of the last century saw it-exact, infallible, a cosmos built on what in modern folk-usage is called induction-must seem remote to a struggling economist or sociologist. But natural science as it has always been understood by its ablest practitioners and is now widely understood-natural science as expounded methodologically by a Poincare-is no such thin substitute for Divine Providence, no such metaphysical abstraction.
Only God is exact, infallible, omniscient, unchanging, and modern science has been content to leave the search for God to disciplines fitted by long success for such a search.
Of the bare elements of scientific thinking-conceptual scheme, facts, especially “case histories,” logical operations, uniformities-the social sciences in general come out well on the score of facts. Even in the field of history, where neither laboratory nor questionnaire methods of research are available, our existing supply of facts is surprisingly good. .You cannot draw Cromwell back to life, but neither can you call the dinosaurs back to life. What we know about Cromwell from the written record is in many ways as reliable as what we know about dinosaurs from the fossil record. To say that history is a fable agreed upon, or a set of tricks played upon the dead, is to slander, or at least to misjudge, the great body of industrious and sober workers who have carried on the study of history. Notably the last century or so has seen the formation of a body of research workers in history who, with all their faults, maintain standards comparable in some ways to those maintained by similar groups in the natural sciences. These research workers do not indeed uncover the simple raw material of facts. The humblest antiquarian arranges the facts he digs out of his documents into some kind of pattern.