On the Philosophical Distinction
Between Knowledge and Opinion

By Bernard Bosanquet

I am privileged to speak this evening before a society of philosophical students in the city which has been called the modern Athens. It appeared to me, therefore, that I might not inappropriately lay before the Society some thoughts on that central question by the treatment of which in ancient Athens the first foundations were laid of a European philosophy.

The question, "Is there such a thing as knowledge, and, if there is, by what features may we recognise it?" had, I take it, a far more radical bearing in Plato’s time than in our own. For us, it is a matter of extreme scientific and also of ethical interest to define the grounds and principles on which, and subject to which, human thought can claim to apprehend the nature of things. The idea of the unity of the world is vital even to those who think that they deny it. But, except in some remote theoretical sense, no one does, or can deny it today. The great inheritance of science and philosophy is to logic, as civilized law and religion to moral reflection, or as the fine art of the world to the perception of beauty. If anything bewilders us in the proceedings of nature, we set it down, as a mere matter of course, to our own ignorance. Nor -182- does anyone seriously dispute the main content of civilized morality, or the universal value of beauty. Our theories are tested by these things, and not these things by our theories. But in Plato’s time these great objective supports were largely wanting to philosophy; though doubtless the civilization which he knew seemed much larger to him than, by comparison, and owing to our ignorance, it now seems to us. In the way of systematic knowledge, we think there was only a little mathematics; in the way of moral consensus, only the institutions, and the not very stable convictions of his own small country, and to some extent of the Hellenic world; in the way of realized beauty, the products of the short-lived maturity of one only, though that the most gifted, among nations. I cannot but think that the suggestion that these principles and activities belonged to no coherent unity, and possessed therefore no absolute and universal validity, was in his day a natural and probable suggestion to a degree which we cannot for a moment imagine. If now, for example, the mysterious debility were to strike Great Britain, which has struck other nations that in their time have led the world, we should look, I suppose, with confidence to Europe and to America for successors who could carry on the torch of science and of civilization. But if in Plato’s time the educated and politically civilized society of Hellas, and more especially of Athens, was to be crushed, or, as he clearly foresaw, to deteriorate, where was the philosopher to look for the hope of humanity?

Therefore, it seems to me, we should consider Plato’s account of scientific knowledge, although drawn from the acutest analysis of experience, as in part a prophecy, which the later history of the world has wonderfully accomplished and defined. To complain that Plato did not say, and did not indeed know in precise detail, what he meant by his dialectic, -183- is to complain of a philosopher for possessing the genius that could lay down the universal conditions of a science for which the actual materials did not in his time exist. He had to work with only a few fragments of organized experience, and in face of a world apparently relapsing into moral chaos; but perhaps the difficulty thus occasioned is compensated not only by his genius, but by the burning reality which the questions of philosophy thus acquired for him.

Of these burning questions the chief and typical one was that which I mentioned: Is there such a thing as knowledge, and what are its distinctive features?

We all know how the question is introduced in the fifth book of the Republic. Politics, Plato says in effect, are a science; you will never get government properly organized till it is in the hands of people who have some grasp of principles. And in support of this suggestion he goes on to explain where the distinction lies between the mind as grasping a unity of principle, and the mind as wandering through a variety of particulars. I will not follow the discussion in Plato’s sense, but will merely mention what throws light on the question before us. Plato draws many contrasts between the world of opinion and the world of science; but the central contrast which is the focus of all the others is this, that opinion may make a mistake, but science is infallible. And the fundamental question which I should like to discuss this evening is what we mean by any such conception as that of the infallibility or necessity of science, and what limitations we must observe in applying it

In the main, we shall not improve much upon Plato. According to him Opinion was liable to err because, in fact, it constantly contradicted itself. And it contradicted itself because its content was relative, but not defined. And its -184- content was not defined, because it was merely an aggregate of perceptive or traditional judgments, which no attempt had been made to analyse or to reconcile. The "many" or "manifold," which is constantly recurring in Plato, as the characteristic content of popular opinion, obviously means not merely separate objects or sensations, but isolated and therefore conflicting judgments. Thus we hear of the many popular formulae, νομιμά, of "beauty and justice," and again of "the many justices," that is, cases of justice regarded as rules of justice; and so of the "many beauties," i.e. conflicting standards of beauty. He is thinking of minds filled with unrationalised instances which appear as fluctuating standards and conflicting judgments.

And because its content is unrationalised, therefore the world of opinion tends to coincide with the world of sense, and is, of course, spoken of as the world of the things that are seen in contrast with the world of the things that are understood. The Greek expressions δόξα and δοκει ("seeming," and "it seems to me") lend themselves to this distinction. I do not suppose that these words indicate sensuous appearance, as do ϕαίνεσθαι and ϕαντασία, but they do indicate a contrast with active thought, a sort of personal acceptance as opposed to a universal conviction. And Plato, in the Republic, as we know, sweeps into the category of opinion or fallacious appearance even the representations of fine art, because they can be considered as images or imagery, and therefore as sensuous.

Science, or knowledge, on the contrary, was infallible, in the sense that its content was single, and its inmost nature therefore excluded the possibility of contradiction or fluctuation. Not that its content was other than relative, but then, being relative, it was defined. Of course there is no confusion or contradiction in relativity when you know that your terms -185- are relative. Relativity in this sense is the root of scientific necessity.

And thus, moreover, being defined, the content of science was necessarily intellectual. It is impossible to have a connected system of conditions in the shape of unanalysed perception or traditional judgments. And so the object or content of science was spoken of as the world of things understood, in contrast to the world of things perceived by sense. We are not here concerned with any materialising misconceptions, of Plato’s or of our own, respecting that intelligible world. There is no question whatever that the unseen world which Plato was labouring to describe was the world of science and of morality — the connected view which gives meaning at once to nature and to human life.

I suppose that the account which we should now accept of this distinction between knowledge and opinion would be essentially founded on that of Plato; but the conditions of modern thought have driven home one or two important points on which his language is not and could not be absolutely unambiguous.

“Opinions are a sort of debris of antiquated science and political or theological tradition, of general maxims and half-understood principles.”

In the first place, we must be very cautious in accepting the opposition between the world of science and the world of sense. We have not in exclusive use the convenient Greek term, "it appears to me." We recognise no peculiar connection between opinion and sense. We speak without a blush of "scientific opinion" and even of "scientific authority." Our opinions are a sort of debris of antiquated science and political or theological tradition, of general maxims and half-understood principles. They have not, we are inclined to think, enough immediate touch with the world of sense-perception. Their fault is rather intellectual confusion than imperfect abstraction from sense. -186-

Our science, on the other hand, seems closely bound up with sense-perception. Nor, again, should we ever dream of ranking Fine Art among unreal illusions, because it is, and must be, largely sensuous. The extremes of our mental world seem to have met, and even to have crossed. Our chaotic opinion is intellectualised, and our coherent science is materialised. If we try to distinguish the world of things seen from the world of things understood, where are we to bestow that act of seeing which a distinguished microscopist begins by describing as "an act of the pure understanding"?

The fact is, that there are correlative misapprehensions attaching to this idea of a world of sense-perception, which we must take care to avoid. Sensation, we are too apt to say, is illusory or false. This is incorrect. What we ought to mean is that sensation is not true; but for the same reason for which it is not true, it is also not false; for it is not a judgment at all, and nothing but judgment can be true or false. On the other hand, if we mean to say that sense-perception, such as human seeing or hearing, is illusory, as Plato too often appears to imply — that mayor may not be; these activities are judgments, and may no doubt be false, but also may be true.

There is, indeed, a secondary difficulty affecting the truth of any perceptive judgment which is bona fide a singular judgment,because its subject is to some extent unanalysed, and therefore not accurately conditioned. But when you are once fairly started on the continuous evolution of judgment, you will find it very hard to draw any intelligible line between judgments affected by this secondary difficulty, and judgments which are not so affected, or are so in a less degree. And granting that judgments affected by it may intelligibly be called "nearer to sense," it still remains quite untrue that judgments dealing with the determinate concrete objects of our perceptive -187- world are necessarily judgments thus near to sense. As Plato says, our primary remedies against sensuous illusions are number and measure; and so-called sensuous objects, as they exist for civilized and for scientific minds, are penetratingly determined at least by measurement and enumeration.

Thus, we must clearly realize that knowledge and opinion both exist in the medium of judgment, that is, of thought. That marvelous dialectician, common language, forces us most commonly to say, "I think," when a Greek would have said, "It seems to me." And though one may be tempted in a moment of irritation to exclaim with Dr. Whewell, "Do you call that thinking?" yet, philosophically speaking, if a judgment is made, it is thinking, and we must be quite clear that our distinction between science and opinion is a distinction within the world of thought, which is a single world, and to which the objects of human sense-perception emphatically do belong.

Then, in the second place, and as a consequence of this generic oneness of our world, we must guard ourselves against finding the differentia of knowledge in any isolated principle which may seem to commend itself to us as peculiarly intellectual in origin or by contrast. We shall do no good by comparing one isolated judgment with another in order to accept that which is more remote from experience or concrete reality. We need not hope, that is to say, to distinguish part of knowledge as the content from part as the form of thought, or to enumerate a list either of innate or of a priori principles of the mind. It has been said, and by an illustrious Idealist thinker, "Two pure perceptions, those of time and space, and twelve pure ideas of the understanding, were what Kant thought he had discovered to be the instruments with which the human spirit is furnished for the manipulation of experience. Whence these strange numbers?" Directly you mention -188 them, you feel that you cannot insist on them in that rigid form. And if, or in as far as Plato meant that science was in the long run to hit upon an abstract ultimate first principle, or principle external to its content, from which knowledge was to be suspended as a coat hangs from a peg, then, and so far, he wavered in his conception of the nature of truth. It might be questioned, for example, what he had in his mind when he said of the mathematical sciences, "How can the whole system amount to knowledge, when its beginning, middle, and end are a tissue of unknown matters? It is, in fact, no more than an elaborate convention." We should of course say, and should have expected him to say, that in any conceivable system of knowledge the beginning, middle, or end are only known by being in the system, and ipso facto become unknown, if regarded in abstraction from it. And I do not think this would be at variance with what he had in his mind. Probably his difficulty was that, as he constantly hints, the greater whole of knowledge was beyond his power to construct; there was, therefore, a saltus or discontinuity round the edge of the mathematical sciences relatively to the whole of knowledge. It was not that he expected to find some law of Causation, or law of Uniformity, or Principle of Identity to which they could all be attached. He evidently was convinced that "the truth is the whole."

Thus we must look for the infallibility or necessity which distinguishes knowledge from opinion, not in the distinction between intellect and sense, nor in the distinction between an empirical and a necessary judgment (unless explained in quite a peculiar way), but in the degree of that characteristic which makes it in the first place thought, and in the second place, knowledge, at all.

All thought is determination, or connection, or definition; -189- but popular thought is insufficient determination, and for that reason is self-contradictory. Every judgment determines a unity by a relation; but as every unity is a centre of relations, it is plain that until the unity has been exhaustively analysed, all its different relations will seem to conflict, because each of them will claim to include the whole of it. And the only remedy for such conflicts is, accordingly, further determination, as Plato explains with unsurpassed clearness in the seventh Book. As determination progresses, then, the unity of thought is maintained; but its differences, which were at first merely found together, come to be systematically arranged, and to have their reciprocal bearings quite precisely defined. So then every part of the system becomes charged with the meaning of the whole, and the relativity of the different elements becomes a source of necessity, instead of a source of confusion. Two terms are relative in this scientific sense when you can tell what form the one will have, by looking at the form of the other. Plato is apt to allude to the apparent contradiction between the appearance of an object seen at a distance, and that of the same object seen close at hand. But of course to an educated eye there is no such contradiction; the one appearance under one condition necessarily involves the other under another condition. The contradiction would arise if the angle subtended by the object were the same at two different distances. The estimate of real size, as formed by an educated eye, is a consequence or combination of the various appearances combined with other evidence, and does not vary with the distance at which the object happens to be seen. We do not judge a man to be very small when we see him a long way off", nor to grow bigger as he comes nearer. In fact, we more generally make too much allowance for distance, and think a man taller at a distance than he really turns out to -190- be when we see him near. The various angles subtended at different distances do not contradict but confirm one another, because their conditions are made explicit; if we confuse their conditions, they will contradict one another. A railway engine coming towards one at full speed does seem to swell, because one has no time to adjust the perception of distance to the angle subtended by the object, i.e., to distinguish the perception under one condition from the perception under another.

Thus it results that the possibility of contradiction is removed and turned into confirmation in as far as experience is organized as a single system of determinations. It is in this sense alone that science has a claim to be infallible or necessary.

But now, if this is so, how far does this kind of infallibility take us? To what extent does it justify us in even asserting that we have knowledge at all?

To begin with, we cannot show, strictly speaking, in this way or in any other, that it is impossible for a change of relations to occur without a change of conditions. We can only say that the suggestion is unmeaning to us, as it involves the saying and unsaying, or being and not-being of the same matter in the same relation. To do and undo is for us simply to leave nothing done; we therefore disregard this contingency; in other words, we assume the unity of reality, which assures us that what is once true is always true, and that what turns out to be not true never was true. Our problem is, how can we be assured that we are making no mistakes? We are powerless if it is suggested that we may be making no mistake, and yet may be in error. That falls outside our discussion tonight.

But there are difficulties more relevant to our problem.

The necessity of science does not provide against our -191- determinations being insufficient, as is plain from the progressive character of science.

There are at first sight two degrees of this insufficiency, though ultimately they may have the same root.

First, there is confusion of conditions. That is to say, you may lay down a connection between condition and consequent, in which by some error of identification you have simply placed one condition or consequent where you ought to have placed another. I will give two examples, one of a more or less debatable case, the other of an extreme case.

The old Wage-Fund theory said, as I understand it, that the wages of labour with a given population depended on the total amount of capital available, and destined in the minds of capitalists, to be paid in the shape of wages. In one sense, this is a truism, i.e., on a given pay-day the whole amount paid divided by the number of persons to whom it is paid, gives the average wage. But in the more real sense, viz., that this fund is a pre-existent fixed quantity, the amount of which actively decides the rate of wages, the doctrine is now disputed, and generally held, I believe, to have been overthrown. Wages are paid out of the produce of labour, and not out of a preexisting fund, and the capitalist very likely gets his hands on the produce actually before he parts with the wages, which therefore are not limited by the amount of a pre-existing fund. The old Wage-Fund theory perhaps rested on a confusion between the truism which I first mentioned, and the very real connection that exists in various ways between the amount of plant or stock in a country, which is Auxiliary capital, not Wage-Fund capital, and the productivity and general employment of its labour, which in their turn affect the rate of wages.

Now how, if at all, does the necessity of science maintain -192- itself in the overthrow of this doctrine? In some such way as this, that the postulates and conditions which made such a doctrine necessary to the scientific system, will continue after its overthrow to be fulfilled by some more or less cognate doctrine, liberated from the confusions which disfigured this one. We shall still speak, I suppose, of the importance of saving. We shall still be aware that an undertaking like the Forth Bridge could only be carried out by a country with an enormous command of accumulated wealth, and that, with a given population, the best chance of raising wages lies in increasing the amount of capital productively employed. Only we must not restrict capital to wage-fund capital, but must include in it, for example, machines and materials.

Now the necessity of the science consisted in the demand for a representation of all these relations and conditions, which, as their determinations advance in accuracy, mould and remould the doctrine that is to satisfy them, but without sacrificing its identity of content or function. The alteration of such a doctrine is like the transformation of gills into lungs, or the substitution of a Westinghouse continuous brake for a hand-brake on a train. You pass from one fulfillment of certain organic demands to another.

As an extreme case, where the connection seems quite irrational, I will just mention what Swift wrote, that once when he was half-asleep, he fancied he could not go on writing unless he put out some water which he had taken into his mouth. He was confusing between writing and speaking, of course. There really is a necessity in the background even there.

In the second place, a science may be precise as far as it goes, but may omit some entire sphere or branch of fact, as Euclidean geometry is now said to omit certain kinds of space. Against this possibility there, prima facie, is no theoretical -193- resource except in a postulate of exhaustiveness, viz., that our knowledge bears some appreciable proportion to the whole of Reality. I incline to think that we take this postulate on ethical grounds, i.e., we are convinced that Reality will not so far dwarf our knowledge as to annihilate our life or wholly frustrate our purposes. It ought to be mentioned, too, that probably a science which is not complete cannot be truly systematic.

Of course you may cut the knot of all these discussions by saying that sciences which make mistakes are not science. But this would not help us, because then we should say that our question is, how far the sciences are characterized by science.

“Scientific authority is a contradiction in terms. Unhappily the scientific mind itself often forgets this.”

In the third place, the systematic character of scientific necessity is in itself a limitation on the extent and application of that necessity. For if and in as far as the systematic character is lost, then and so far the necessity is lost too. It has been said of political economy, that if you do not know it all, you do not know it at all. This is true in strict theory of every science and of all science. So that the scientific judgment, transferred by the help of language into a mind not equipped with the body of knowledge, is science no longer. It has become mere opinion, mere authority. This explains the curious contempt which practical men have, as a rule, for the evidence of scientific experts. Scientific authority is a contradiction in terms. Unhappily the scientific mind itself often forgets this, and offers, like Thrasymachus, to put its doctrine into men’s souls by physical force. But this is impossible. Knowledge can only be communicated as knowledge. You cannot claim the necessity of science for a scientific conclusion torn from its organism and hurled into the sphere of opinion. Think of the popular interpretations -194- of any such propositions as, "The soul is a substance," or, "Sensation is subjective."

But here we have arrived at the end of our negatives, and the balance begins to turn.

If, for the reason just stated, Knowledge cannot refute Opinion; neither, for the same reason, can Opinion refute Knowledge. An individual judgment and a universal judgment cannot be contradictory in the strict sense. The judgment, "If A is B, then C is D," is not affected by the judgment, "C is not D." They are in different planes, and do not meet. Before you can bring the two into relation, you must ascertain how A and B are behaving in the case, when it is alleged C is not D. Then we shall find, in proportion as the hypothetical judgment belongs to a thoroughly organized body of science, that it is easy to incorporate the new determination in the old system. I will once more take an example from political economy. The economical doctrine says that prices determine rent, and rent does not determine prices. But of course it is a common opinion that a tradesman in a fashionable street is compelled to charge higher prices than a tradesman in a less fashionable street, in order to recoup himself for the higher rent which he has to pay. If we put out of sight the alternative of his obtaining a larger sale, I should suppose that this might be the fact, although one would imagine that he would have fixed his prices so as to obtain the greatest profit, even if his profit was not to go in rent. But waiving this argument again, and admitting the alleged fact, what does it amount to? What made his landlord ask for that high rent, and what made the tradesman contract to pay it? Why, that both of them thought that the prices necessary to pay this rent could be got out of the public in that locality. You cannot -195- put up your prices just as you please; and if you cannot get the prices necessary to pay your rent, why, then you cannot pay your rent out of the proceeds of the business, and the rent must come down; and, no doubt, if you are under a lease, or competing for houseroom with other occupations that pay better, you may say to the public, "Really I am forced to try to keep my prices up." But strictly the reason for this is not that the prices do not determine the rent, but that they obviously do, and the tradesman is crushed between two determinations of his rent, legal and economical; only, being unable to revise his bargain, he may try to hold the prices up with both hands, so to speak; and with a friendly circle of customers, or a circle who need him in their district, to some extent he may succeed. But in some such way as this the relation between the scientific doctrine and the popular opinion is not, I think, very hard to see, when you look at the matter all round. And of course it does modify the doctrine a little bit; but on the whole, when you analyse the alleged case, it joins on pretty easily to the science. The science, of course, primarily considers what a man will freely bargain to do; it never denies that a man may have a loss thrown upon him by a bad bargain which he cannot revise, and that so far his rent, which is naturally the consequent, will become for him the condition, because he cannot alter it.

If, then, we try to state the positive value of the so-called infallibility of science, it appears to reduce itself to this — that the organization of a province of experience is an affirmative or actual achievement, which may be subsequently modified or transformed, but cannot be lost or cancelled. We cannot guarantee the particular formulation of an isolated principle; but then we know that identity does not depend on particular formulation, but on continuity of function. I should be very -196- sorry to predict in what precise terms the Principle of Sufficient Reason may be stated by philosophers a hundred years hence. But that the determinate relativity of the parts of experience will be embodied in some principle or other, is as certain, I think, as that there will be science at all.

It may be objected that we are guaranteeing the whole of knowledge in general, but no element of it in particular, and that this is illusory. To those who cannot conceive a concrete continuous identity I think it is illusory, and ought to be. You cannot, as they would wish, fix and separate any portion of knowledge. Every element of it must take its chance in the systematic development of the whole. Therefore, when speaking of knowledge in general, you can only affirm its self-identity in general. But to any one who can see a meaning in saying, for example, that Christianity today is the same religion that it was 1,800 years ago, this idea of continuous pervading identity will present no difficulty. A substantive identity, we think, can persist through difference, and, can indeed, only be realized in differences.

While, on the other hand, if a certain difficulty attaches to this view, yet it throws an all-important light on the nature of knowledge. It shows us that the necessity of knowledge depends upon its vitality. Axioms and dogmas, traditions and abstract principles, equally with unanalysed perception, are not knowledge but opinion. The life of knowledge is in the self-consciousness which systematically understands, and you cannot have it cheaper. We know not "as much as is in our memory," but "as much as we understand." A science which accepts foreign matter, data to be learnt by heart, is so far not a science. But one who has understood anything, has a possession of which he cannot be deprived.

Anyone who speaks thus confidently is sure to be asked, -197- "What are his metaphysical presuppositions?" It would be more to the point, in my judgment, to ask him if he has obtained any metaphysical results. His only presupposition is, I think, that there is something presented to him which it is worth while to analyse. The principles involved in this analysis, such as the unity of reality, are no doubt operative from the first, but are only established in a definite form by the analysis itself. And any view, more strictly metaphysical, as to the precise ultimate nature of the unity of Reality, would be a still further result, which may or may not be obtained. That mind, in its essence, is one, and that the unity of man with himself and with nature is a real unity, seem to be principles demanded by the facts of science and of society. It is also true that a reality which is not for consciousness is something too discrepant with our experience to be intelligible to us. But whether the human mind will ever form to itself a conception that will in any degree meet the problem of a total unity of Reality, is a question the answer to which must lie in the result of analysis, and not in its presuppositions.

Thus we abide by the position that the characteristic in which Knowledge differs from Opinion is the degree in which, as a living mind, it has understood and organized its experience. The criticism of Goethe’s Mephistopheles on the traditional logic is perfectly just. It is well to take every mental process carefully to pieces; but it is essential to bear in mind that the pieces are elements in a living tissue, in a single judgment, and that in their detachment as "one, two, and three," they are not knowledge. So far from being a mechanical science, logic is perhaps the most vital and scientific of all the sciences. It accepts nothing from perception or from authority, and gives nothing to learn by heart. It depends on no intuition of space, and on no list of elements. Its only task -198- is to understand the process of understanding, the growth and transformations of thought.

This is the conception with which logic began in Plato, and which has never been entirely lost. In the old Dominican Church of Santa Maria Novella, at Florence, there is a series of frescoes illustrative of education, familiar to us through Mr. Ruskin’s description under the name of the Strait Gate. One of these paintings has a peculiar attraction for the student of modern logic. Next but two after the Narrow Gate itself, which indicates the entrance to good life, there is placed over a head of Aristotle the allegorical figure of logic. This beautiful figure is drawn, as Mr. Ruskin points out, with remarkable strength and grace; it is most probably from the pencil of Simone Memmi, of Siena, early in the fourteenth century. In her left hand the figure holds the scorpion with its double nippers, emblem of the dilemma or more generally of the disjunctive or negative power of thought; but in her right hand she holds the leafy branch, symbolising the syllogism, conceived as the organic or synthetic unity of reason. This suggests an ideal worthy of the age of Dante, however little it may have been attained in the explicit logical theory of that time.

It is some such ideal of knowledge that has, as we may hope, been making itself more and more imperatively felt since the revival of letters in Europe; and the view which it involves, of the true distinction between Knowledge and Opinion, is merely one branch of that principle of the unity of mind, which is fraught with consequences of inestimable importance for all aspects of life in the present day. We cannot — such is the lesson we have to learn — we cannot elevate the human mind by any fragmentary treatment, by any communication or assistance which does not stimulate its healthy -199- growth as a single living thing. In fine art, in the province of social rights and duties, in morality, in politics, and especially in the interconnection of all these spheres, it is no less true than we have found it to be in science, that the mind must grow and advance either all together, or not at all.

Top ↑

• Back
• Next