You are here

Chapter I: Knowledge and Certitude

§ 1

The candidate sat staring at the sheet of questions placed before him by the examiner, but not writing anything. ‘What ails you?’ asked the examiner. ‘The questions are too difficult,’ replied the candidate; I cannot answer them. ‘Well,’ said the examiner, ‘put down what you know.’

Hearing this told, it occurred to me that perhaps I also ought to put down what I know, but the attempt made me realize afresh how limited my knowledge is, nor did I find it by any means easy to define these limits. Hence arose the following reflections.

Many of the questions put to me, and not least those I am accustomed to put to myself, I just cannot answer at all, and many others I can answer with only a low degree of assurance. But it would seem that to know is to be quite sure. You would laugh if I said I know he is a Frenchman, but I am not quite sure‘, or’ I know there was a frost last night, but I'm not certain’. Knowledge, then, implies certitude, complete assurance. On the other hand, this does not exclude the possibility of probable knowledge in the sense of knowledge of probabilities. There is no contradiction in saying that I can be certain of a probability or a likelihood; for instance, that of the next thousand babies born in this city some are likely to be boys and some girls; or that if I plant my vegetable garden at the usual time this year, some at least of the seeds are likely to germinate; or that of the next hundred men I ask a question, some are likely to be more strictly truthful in their answer than others. Furthermore, however, I should hold that such knowledge of the probable behaviour of things implies some certain knowledge, however limited, of the things themselves. If this were not so, we would have to say that the natural sciences do not yield any knowledge of nature, and so to exclude them from the realm of knowledge altogether (as indeed Plato wanted to do for a somewhat different reason); for it is generally agreed that a high degree of probability is the most that can be claimed for any scientific result. Again, to say that knowledge implies certainty does not mean there are not different degrees or levels of knowledge, for there is clearly a difference between knowing something less well and knowing it better. If you ask me whether I know John Smith and the town he lives in, I may answer: ‘Yes, I know both, but I know neither very well.’ What is implied is only that if I have no certitude at all in regard to either, I cannot be said to have any knowledge of them.

I have spoken both of certitude and of certainty, and perhaps it is well to follow Cardinal Newman in maintaining, as far as may be, a distinction in usage between the two words, applying the former to the state of mind of the knowing subject and the latter to the propositions he enunciates.1 Yet, if we do this, we must remember that neither word has any meaning except with reference to the knowing subject. Certitude would mean such an assurance as leaves no room for doubt in the inquiring mind, while certainty would mean the character accruing to a proposition through its being supported by such evidence as makes it impossible for the inquiring mind to doubt it. The realities with which the inquiring mind is confronted are themselves neither certain nor uncertain; they simply are.

Plato did us a great service in insisting on the difference between assured knowledge and mere opinion, and this service remains even when we are unable to agree with him as to the relative incidence of the two. Yet I fear I still introduce many of my utterances with the words ‘I know that’ or ‘I am sure that’, when I have no real right to do so. Plato's Greek word doxa, which I here render as opinion, really means seeming, being cognate with the expression ‘It seems to me that’; and there is a multitude of sentences which I have full right to introduce with that phrase but which fall far short of assured knowledge. It is my duty as well as my necessity to form opinions. But to opine is not to know.

Some indeed have said that all pretended human knowledge is mere opinion, that there is nothing of which any of us has a right to be sure. It was even said of one such that ‘the consciousness of a single certainty would have lain as an intolerable burden upon his mind’. Such a view was, as we know, widely canvassed in a late period of Greek thought, not least in Plato's own Academy after his death. To be consistent, of course, all that such sceptics could affirm would be, not that there is no certainty, but that it is not certain whether or not there is; and the best of them were sufficiently clear-minded and self-critical to realize this. The same view is still frequently reaffirmed, as by Professor Ayer in his Language, Truth and Logic, where he writes that ‘There are no objects whose existence is indubitable’2 and that ‘It is only tautologies that are certain’.3 But since these statements are clearly not tautologies the writer has no right to be certain of them, and perhaps should not affirm them with such dogmatic assurance as he appears to do.

We recall also that the most representative philosopher of modern times, Immanuel Kant, declared that the purpose of his greatest work was ‘to abolish knowledge’, and not only so, but to do this ‘in order to make room for Glaube’.4 Now this word is German for belief, and Plato's word which I have rendered as opinion (doxa) has likewise often been rendered as belief. Was Kant's purpose, then, the anti-Platonic one of renouncing knowledge in order to rest content with mere opinion? By no means; and the confusion is to be explained in the following way. Kant was here speaking of our apprehension, not of the things of sense, but of things beyond sense. He was not unwilling to allow that our apprehension of the things of sense could be spoken of as knowledge. Plato (as we have said) had denied this, holding that of the things of sense we could only have opinion; but he claimed that we could have assured knowledge of the super-sensible realm, and he pointed out a way by which he thought this could be attained. Something of this way had been followed by Kant's predecessors in the medieval period and after it, chiefly under the name of natural theology, but Kant himself denied that this way was viable. He therefore said that we have no knowledge of super-sensible realities, but only belief. Yet the belief he had in mind was not mere opinion but faith; and the confusion is partly caused by the fact that the German Glaube covers both our English words belief and faith. Now this concept of the kind of belief which is faith is something that came into Western thought long after Plato's time, presenting the problem in an entirely new light. It represents the specifically Christian view of the way in which super-sensible realities are apprehended. It is the correlate of the Christian idea of revelation, though for Kant himself its meaning had been greatly curtailed through the influence of the rationalism of the Enlightenment.

It is, however, a mistake to set faith, as Kant did, in contrast to knowledge. To do this is to mingle two different terminologies, the Greek and the Christian (or we might even include a third, that of modern empiricism) in a most confusing way. In the Bible there is no such simple contrast between faith and knowledge. The Biblical contrast is rather between faith and sight. ‘Faith is… the evidence of things unseen.’5 ‘We walk by faith, not by sight.’6 Tennyson speaks like Kant when he writes that

We have but faith, we cannot know;

For knowledge is of things we see.7

But while from the Christian point of view Tennyson is right in holding the unseen world to be apprehended by faith, he is wrong in denying the name of knowledge to such apprehension. In the New Testament8 to know God and to have faith in him are often hardly more than two ways of saying the same thing. And yet this knowledge that is faith is not the best kind of knowledge. The concept of faith always contains both the idea of knowing and the idea of not knowing fully.

§ 2

No Christian, then, can say that he knows nothing. What he must say is that such knowledge as he has is severely restricted both in extent and in kind. None of us can claim to have been granted more than certain slender divinings of truth, which is otherwise surrounded by clouds and thick darkness. ‘Now I know in part’, writes St Paul. Nor was it only that his knowledge was small in extent, but what he had of it was also dim. ‘Now we see as in a mirror, as in a riddle’, he says in the same verse.9 And when the New Testament gives to our present knowledge the name of faith, this is part of the meaning. The Apostle is as little satisfied with the quality as with the quantity of his knowledge. He longs for a clearer apprehension which, using the same visual figure, he calls ‘seeing face to face’, and one of the things he claims to know by his present faith-knowledge is that the better kind of knowledge, in which faith is transcended, will hereafter be his. He has faith that one day he will be granted something better than faith.

We must also remember that there is a difference between knowing, and knowing that we know; or, what is the same thing, a difference between knowing that we know (or do not know) and merely thinking or opining that we know (or do not know). If there were no such difference, my endeavour to set down what I know would have been far less difficult than I actually found it to be. It seems to require strenuous mental effort on my part to distinguish my real from my putative knowledge—and hardly less to distinguish my real from my putative ignorance. There might not indeed be any such difference, were I what the psychologists call a completely integrated personality or what the theologians call sinless, but in fact the part of me which recognizes truth is not always identical with the part of me which acknowledges my recognition of it. That is why, in any tally I make of things I assuredly know, I shall be likely to include some things that should have been left out; and also, and equally, likely to leave out some things that should have been put in. To follow a makeshift terminology which I have used elsewhere,10 I should say that there are doubtless some things which ‘in the bottom of my heart’ I know full well to be true, but whose truth I have never fully acknowledged ‘with the top of my mind’. Perhaps therefore we all know both more and less than we think we know. This comes partly from intellectual indolence, but partly also from a failure in intellectual honesty; for there are some things we so much want to be true that we stifle our doubts concerning them, and other things the acknowledgement of which would make such unwelcome demands on us, or entail so inconvenient a readjustment both of our professed outlook and of our habitual conduct, that we succeed in suppressing or ‘repressing’ what would otherwise be a fully assured conviction of their truth. Neither the gnostic not the agnostic is always fully honest with himself.

There can thus be no indefectibility in my tally of the things I know. All human thinking is defectible. I may indeed believe that there is an indefectible authority on which I can lay hold, but I still remain defectible in my way of laying hold upon it. Whatever is of divine revelation must be infallible as it issues from God, but we can make no such claim for it either as it is transmitted to us or as it is received by us; for the human element always enters both into the transmission and into the reception. Or even if we believe with the Roman Catholic that the transmission is no less infallible than the original emission, the reception by us at least remains fallible. Omne quod recipitur, says Aquinas, recipitur per modum recipients—which is always a fallible mode. Moreover, even if the ground appear solid enough beneath our feet once we have accepted our authority as infallible, so that Newman could say that after he had been received into the Roman Church in 1845 he never again ‘had one doubt’,11 yet that initial step of accepting the authority, of choosing this from among the several authorities in the world that claim infallibility for themselves, must be taken on our own responsibility. We should hope indeed that in exercising this responsibility and making this choice we were aided by divine grace, but we should not claim the plenary guidance of the Spirit of God, plenary inspiration, not to say infallibility, for our exercise of it. Newman is quite clear about this. The credentials of the Roman claim, he says, must be examined by the ordinary rules of evidence before the claim is admitted. We must

confess that there is no ultimate test of truth besides the testimony borne to truth by the mind itself, and that this phenomenon, perplexing as we may find it, is a normal and inevitable characteristic of the mental constitution of a being like man on a stage such as the world.12

Moreover, if our reliance is on Holy Scripture instead of on the Church, the same considerations apply. Its teaching, says Archbishop William Temple, is

so inextricably human and divine in one that no single sentence can be quoted as having the authority of an authentic utterance of the All-Holy God.13


The Christian will believe that he has an infallible authority in the mind of Christ; but he has no infallible means of ascertaining its application to given circumstances.14

And the fallible elements in its transmission are such that

there is no single deed or saying of which we can be perfectly sure that He said or did precisely this or that.15

§ 3

But, it will be asked, does this mean that after all there is no such thing as certitude and therefore no such thing as knowledge? If it meant this, then indeed we should have to say with George Meredith,

Ah, what a dusty answer gets the soul

When hot for certainties in this our life.16

But I shall contend that it does not mean this. The most it means is that, while all authentic experience has some certitude in it and is, as it were, transfused with certitude, it is never possible to distil this certitude in its complete purity into the particular theoretical affirmations we make. We are convinced we are in touch with reality, we do know something assuredly, but when we try to express in theoretical terms what we know and are sure of, we never have the same assurance that we have got our answer quite right. We know we are thinking something that is certainly true, but there is always the risk of error in our way of thinking it. As the late Lord (then Mr A. J.) Balfour wrote in 1894,

Though the fact is apt to be hidden from us by the unshrinking definitions with which alike in science and in theology it is our practice to register attained results, it would… be a serious mistake to suppose that any complete correspondence between belief and reality was secured by the linguistic precision and the logical impeccability of the propositions by which beliefs themselves are communicated and recorded.17

Consider first the case of the natural sciences. None of their results can be regarded as final in the form in which it is at any time envisaged and formulated. They are all corrigible and in fact subjected to constant correction; and it is or should be the very pride of natural science that this is so. Yet in each result something of indubitable truth is contained. There are of course many instances of the outright reversal of scientific conclusions, but even in these instances something of true apprehension was present in the approach to the conclusion, though the conclusion itself was wrongly drawn. What usually happens, however, is rather in the nature of a revision—it may be only a deepening—of the original insight. The scientist knows that he is seeing something, has hit on something, but always there may be something he has not yet hit upon which may greatly modify his understanding of what he has already hit upon. If no element of certain knowledge entered into the probabilities of science, science could not be progressive. But we are certain that it does make progress, that its later conclusions, however far they may be from the final truth, are at least nearer it than those which they have superseded. The Copernican cosmology may in its turn be departed from no less than the Ptolemaic, but it is certain that it makes an advance upon the Ptolemaic.

Consider again our moral convictions. These are governed, as Kant rightly said, by a sense of unconditional obligation, which is to say that there is about them something of absolute certainty. Yet it is notoriously difficult, and much more difficult than Kant supposed, to distil this sense of obligation into specific moral rules or commands for which the same unconditionality can be claimed. Kant insisted that it was our unconditional duty always to speak what we believe to be the truth, but most of us would wish to qualify this by the mention of a number of conditions such as would apply in certain cases. We are indeed under absolute obligation in respect of truth-speaking, but our efforts to codify this obligation are always to some extent provisional. The obligation itself is absolute, but the particular duty to which we are obliged is manywise relative. As has been well said:

The ought is unconditional. If I ought to do X, I ought do it without qualification or reserve. Any qualification falls on the side of the content of duty, not on the side of the obligation.18

Then there are our religious convictions. Here, if anywhere, we find that certitude is claimed. The Christian has full assurance of being in authentic touch with the unseen. He has been visited by God's revelation of himself, of his mind and will in Jesus Christ our Lord, and divine revelation can be no other than infallible. What it proclaims is absolute truth, subject to no qualification or revision, ‘irreformable’. On the other hand we have said that the Christian can never perfectly capture into his own thinking this infallible revelation, this absolute truth. He tries by means of the most suitable concepts his mind can frame, and the best language he can command, to grasp what it portends, but in this he never achieves anything like complete success, so that his theological formulations are always ‘reformable’, subject to correction, revision and development. Yet it is precisely and only because he has been visited by infallible divine revelation that such correction, revision and development are possible; and as these processes are carried forward, they are controlled and guided by this revelation, and ought to be controlled and guided by nothing else. This means that all his thinking has been invaded, and continues to be pervaded, by an infallibility, an absoluteness, and therefore a certainty, which he nevertheless remains unable to hold securely in his own very human grasp or, to vary the figure, to domesticate into the household of his own very human mind.

And this, according to Kierkegaard and those theologians who have come under his influence, is why theology must always be ‘dialectical’. God, they say, addresses himself to us in revelation; and we, if we face towards him and respond to this address, receive this revelation in faith. But when we try to think just what it is that has been revealed to us, we have to proceed by means of human concepts, abstract concepts each of which takes away (abstrahit) something from the concrete reality which has been revealed. Each such concept therefore grasps no more than one aspect of the truth which it regards in isolation from the rest; and thus misleads at the same time as it enlightens. This situation can be rectified only by allowing each concept to correct the others. When the revealed content is subjected to human reflection, it is, as it were, diffracted by our thought in several different directions, each leading to a result which is invalid until complemented by all the rest. Yet this process of complementation can never be completed without the appearance of those antinomies which must always continue to characterize finite reflection on the infinite. From a truly dialectical theology the element of paradox can never be eliminated. This has been illustrated as follows:

The attempt to put our experience of God into theological statements is something like the attempt to draw a map of the world on a flat surface, the page of an atlas. It is impossible to do this without a certain degree of falsification, because the surface of the earth is a spherical surface whose pattern cannot be reproduced accurately upon a plane. And yet the map must be drawn for convenience’ sake. Therefore an atlas meets the problem by giving us two different maps of the world which can be compared with each other. The one is contained in two circles representing two hemispheres. The other is contained in an oblong (Mercator's projection). Each is a map of the whole world, and they contradict each other to some extent at every point. Yet they are both needed, and taken together they correct each other. They would be either misleading or mystifying to anyone who did not know that they represent the surface of a sphere. But they can serve their useful purpose for anyone who understands that they are intended to represent in handy portable form the pattern covering the surface of this round earth which he knows in actual experience.19

Especially in the writings of Professor Paul Tillich I find many excellent statements of the point I have been trying to make. In one of his essays, for example, he is defending the Christian conviction that the ultimate ordering of our destiny is not in the hands of a demonic fate, but in the hands of God who stands above fate:

Without this certainty, which is the inmost kernel of Christianity, we should be thrown back to the Greek situation, and should have to begin to traverse the whole fateful path of philosophy over again. But this eternal truth, this logos does pulsate through all our thinking; There can be no act of thought without the secret presupposition of its unconditional truth. But this unconditional truth is not in our possession. It is the hidden criterion of every truth that we believe we possess. There is an element of venture or of risk in every statement of the truth. Yet we can take this risk in the certainty that this is the only way in which truth can reveal itself to finite and historical beings.20

With this we may compare what Dr Karl Barth has to say when, speaking of such knowledge as we have of God and his Word, he speaks of its inevitable Welthaftigkeit, which I can only translate as ‘involvement in the world’:

God alone conceives of Himself, even in His (revealed) Word. Our concept of Him and His Word can be no more than a pointer to the limits of our conceiving.… Above all God's speech is and remains God's mystery in respect of its involvement in the world.… When God speaks to man, this happening is never so marked off from the general run of what happens that it could not easily be interpreted as being a part of it.… The veil is thick. We do not possess the Word of God otherwise than in the mystery of its world-involvement. But that means that we always have it in a form which as such is not the Word of God, and as such does not even betray the fact that it is the form which encloses God's Word.… Its form is not a suitable but an unsuitable medium for God's presentation of Himself. It does not correspond to the content but rather contradicts it. It does not unveil it but rather veils it. The ‘world-involvement’ of the Word of God does not mean merely that it meets us in the dress of creaturely reality; for we must say also that, because this creaturely reality is that of fallen man, it meets us in a form which is not that of a pure nature.… Hence our knowledge of God's Word is not mediated to us by a reason that has somehow remained pure and can consequently see through the creaturely reality to God's mystery, but is always mediated to us by our fallen reason.21

It is for such reasons that, instead of claiming certainty or finality for our particular thoughts about God and the unseen world, I must content myself with claiming that certainty ‘pulsates through all our thinking’ or that our experience in this realm is everywhere ‘transfused with certainty’.

§ 4

But the difference between these two claims calls for further elucidation, and I should wish to elucidate it by distinguishing between two kinds of knowledge—knowledge of truth and knowledge of reality. As we have seen, there are those who claim that we have no knowledge, because no certainty, of either kind. I have already quoted Professor Ayer as saying that only tautologies are certain and that there are no objects whose existence is indubitable. Professor Tillich appears to agree. ‘Knowledge of reality’, he writes, ‘has never the certitude of complete evidence.… Every knowledge of reality by the human mind has the character of higher or lower probability.’ There are, he says, two types of knowledge which yield complete certitude, our knowledge of the propositions of logic and mathematics (Professor Ayer's ‘tautologies’), and our apprehension of immediate sense-data concerning which he writes, ‘He who sees a green colour sees a green colour and is certain about it. He cannot be certain whether the thing which seems to him green is really green. He may be under a deception. But he cannot doubt that he sees green.’22 Neither of these knowledges, however, is knowledge of reality. It should at once be added, however, that Professor Tillich saves himself from this apparent scepticism by claiming that there is a certitude of faith, which is of another kind. But to this faith he will not allow the name of knowledge—any more than Kant would have done.

Let us first consider whether it is really true that there are no objects (within which category we can for our present purpose include other subjects) whose existence is indubitable. It is well known that in an essay entitled ‘A Defence of Common Sense’ published in 192523 Professor G. E. Moore protested strongly against such a view, repeating and developing this protest in a number of subsequent papers. He claimed complete certainty for such affirmations as that he himself was a bodily existent, that this body of his had existed for many years, that there are and have been other living bodies not unlike his own, and that these, like his own, have been the bodies of persons who in their turn enjoyed a like certitude regarding other existents. Here, then, is roundly challenged the view that the direct and only indubitable objects of our knowledge are sense-data (noises, patches of colour etc.) and not anything ‘out there’ (exsistens). Bertrand Russell had argued that even so simple an affirmation as ‘I am sitting, as I write this, in a chair before a table’ requires ‘much careful consideration before we can be sure that we have stated it in a form which is wholly true’.24 But Professor Moore insisted that this confuses the meaning of the affirmation, which is quite clear and of whose truth it is possible to be quite certain, with the philosophical analysis of that meaning, which may be very hard to come by.

The view that our direct and indubitable knowledge is only of sensations or so-called sense-data, and not of existent objects, has been further and most interestingly challenged in a recent little book by Dr E. L. Mascall. It has been assumed by Locke and Berkeley, and has continued to be assumed by the contemporary logical empiricists, that the perceptive element in sense-experience consists simply of sensations, of the registration of sensible particulars like colours, sounds and tactual resistances, and that from these latter the intellect then deduces, by way of inference from effect to cause, the existence of extra-mental existents.25 Against this Dr Mascall contends that, ‘while there is no perception without sensation, the sensible particular or sense-datum is not the terminus of perception, not the objectum quod… but the objectum quo, through which the intellect grasps, in a direct but mediate activity, the intelligible extra-mental reality, which is the real thing’.26 He thus sums up his position:

First, that although perception normally takes place through the medium of sensation, its essence is not sense-awareness but phenomena as an objectum quo through which it passes to the apprehension of the objectum quod which is the intelligible trans-sensible being. Hence, in the second place, the intelligible object is not something whose existence is deduced from that of the sensible phenomena, as Locke thought, nor is it something mentally constructed out of the sensible phenomena, as many modern empiricists have held, but something grasped through them. Thirdly, in order to penetrate beneath the sensible phenomena to the real intelligible things that support them, we need, not an attitude of detachment, ratiocination and attention to the phenomenal surface of things, useful as this is for certain purposes, but an attitude of involvement, contemplation and penetration into their intelligible depths.27

So also Professor Macmurray protests against the assumption ‘that what is given in immediate sense-experience is a sense-datum, not a physical object’,28 and insists that we have certain and indubitable knowledge of particular extra-mental existents. If I say that I know John Smith and also the village in which he lives, these statements, unless I am lying, are certainly true in the sense in which I mean them.29

I find it difficult to escape from these conclusions, but meanwhile let us consider more carefully the distinction I have drawn between the two kinds of knowledge, knowledge of truth and knowledge of reality. There is indeed an exalted use of the word ‘truth’, doubtless of Hebrew origin, which makes it equivalent to reality. I find it stated as a main Biblical usage in Cruden's Concordance that ‘Truth is put for reality’. So we read ‘I am the way, the truth and the life’.30 But when I here speak of knowledge of truth, I mean knowledge of truths, that is, of propositions. Sometimes I say’ I know X’, but sometimes I say I know that X is Y’. This is the distinction between knowledge by acquaintance and knowledge by description to which Lord Russell directed our attention in a famous essay published in 1911: part of his contention (and I am not here concerned with the further logical problem which is perhaps the main concern of the essay) being that the former is logically prior to the latter, because our ability to make affirmations about anything presupposes our acquaintance with it—or, to put it a little differently, that I cannot enunciate truths about anything unless I am already directly confronted with the thing itself. ‘The relation between subject and object which I call acquaintance’, he writes, ‘is simply the converse of the relation of subject and object which constitutes presentation. That is, to say that S is acquainted with O is essentially the same thing as to say that O is presented to S.’31 What is disappointing, however, is that Lord Russell still insists that, apart from ‘universals’, our acquaintance is only with sense-data and not with real objects or with other subjects. ‘We have acquaintance with sense-data, with many universals, and possibly with ourselves, but not with physical objects or other minds.’32

The validity of the distinction (however it may be applied in detail) has frequently been denied, especially by philosophers of the idealist tradition who contend that our only knowledge is of ideas in our minds. ‘All knowledge’, according to G. F. Stout, ‘is of propositions, and of other things only as forming constituents of propositions.’33

I do not at all doubt that what is here called acquaintance actually exists. Without it there can be no knowledge, for if we were not acquainted with some things, we could not know anything.… But it cannot, I think, properly be called knowledge.… How indeed can we know anything, if it is supposed that we know absolutely nothing about it?34

To this I should reply that we cannot indeed know anything without at the same time at least thinking we know something about it, because in the very moment that we are confronted with any reality, so becoming acquainted with it, our minds start to frame certain propositions regarding it, but that it is nevertheless the reality itself, rather than the propositions, which is the prime and direct object of our knowledge. Stout argues as follows:

On Mr Russell's view, acquaintance is not acquaintance with characters or attributes, but with the subject as something distinct from all that can be truly asserted of it in judgements. This is a position which I cannot accept. The subject, taken apart from all its characters and attributes, can only be known, if it can be known at all, as that to which the characters or attributes belong. In other words, it can be known only by description. If we persist in asking what it is in itself and yet refuse to take as an answer any statement of its attributes, we can only say with Locke that it is a ‘somewhat, we know what’.35

Now it has always appeared to me particularly paradoxical to say, as Stout here does and as others have done, that what they nevertheless allow us to call our ‘acquaintance’ with persons ‘cannot properly be called knowledge.’ Nothing could be more contrary to the accepted usage of the term. We habitually speak of our knowledge of persons as being much more fundamental than the knowledge of propositions which may be enunciated concerning them; as when I say, ‘No, I have not the pleasure of knowing John Smith, but I know a good deal about him’, or again when I say, ‘Yes, I know him, but I do not know him very well. I hope soon to get to know him better’—in the latter case recognizing that there are various degrees in our knowledge of persons. To my mind knowledge of persons is the very type and pattern of what we mean by knowledge. Of no other existents is our knowledge so intimate or so direct. ‘When we distinguish between persons and material things’, writes Professor Macmurray, ‘the characteristics we attribute to things are a selection from the characteristics we attribute to a person.’36 And again, ‘We use the term “know” in (the) primary sense when we say that we know our friends and are known by them.’37

Certainly this is the kind of knowledge of which the New Testament speaks and which it so often designates as faith. Ye have known God, or rather have been known by him.’38 Characteristically and primarily faith is faith in God, confidence in him rather than the uttering of judgements concerning him. What Christians do when they say the Creed is to confess their faith in the three Persons of the Godhead, whom then and only then they proceed to identify by ‘descriptive’ phrases such as ‘Maker of heaven and earth’, ‘Who was conceived by the Holy Ghost.…’ Archbishop Temple wrote, ‘I do not believe in any creed, but I use creeds to express, to conserve, and to deepen my belief in God.’39

Instead therefore of saying that we have no knowledge of realities but only of the judgements we make about these realities, we must say that our knowledge of the realities themselves—whether these be the external world or our fellow men or God—is primary, and our knowledge of truths concerning them secondary. The point, then, that I am most concerned to make is that, however difficult we may find it to ascribe certainty to these truths, we may nevertheless enjoy the certitude of having authentic acquaintance with the realities they fallibly seek to describe.40 It is in this way that our experience is everywhere transfused with a certitude which pulsates through all our thinking, and whose pervading presence can on occasion make even of our most speculative theorizings something better than mere fantasy and baseless fabric.

  • 1. A Grammar of Assent (1870), chap. VI. 2.
  • 2. Op. cit., 2nd ed. (1946), p. 121.
  • 3. Ibid., p. 93.
  • 4. Kritik der reinen Vernunft, Preface to 2nd ed.
  • 5. Heb. 11 . 1.
  • 6. 2 Cor. 5. 7.
  • 7. 2In Memoriam, Prologue.
  • 8. ‘This knowledge (whether evpi,gnwsij or gnw/sij) scarcely differs in substance from faith, except that it emphasizes the element of knowing which is contained in the very structure of faith.’—R. Bultmann, The Theology of the New Testament, Eng. trans., Vol. II, p. 128.
  • 9. I Cor. 13. 12.
  • 10. Our Knowledge of God (1939), chap. II. § 5.
  • 11. Apologia pro Vita Sua, chap V.
  • 12. A Grammar of Assent, chap. IX. § 1.
  • 13. Nature, Man and God (1934), p. 350.
  • 14. Ibid., p. 353.
  • 15. William Temple in Revelation, ed. Baillie and Martin (1937), p. 114.
  • 16. Modern Love, XLVIII.
  • 17. The Foundations of Belief, p. 271.
  • 18. W. G. de Burgh, From Morality to Religion (1938), p. 53.
  • 19. D. M. Baillie, God Was in Christ (1948), p. 109.
  • 20. The Protestant Era (Chicago 1948, London 1950), chap. I. iv; italics mine. Karl Jaspers has much to say in the same vein. His position is thus summarized by David E. Roberts, Existentialism and Religious Belief (1957), p. 250: ‘Being-itself, although it transcends every finite mode, is nevertheless encountered in the midst of finitude’ (italics mine).
  • 21. Kirchliche Dogmatik, I. i, p. 171 f.
  • 22. Dynamics of Faith (1957), p. 33 f.
  • 23. In Contemporary British Philosophy, ed. Muirhead, Vol. II.
  • 24. The Problems of Philosophy (1912), p. 11.
  • 25. ‘A material thing is a logical construction out of sense data.’ So C. D. Broad summarizes Bertrand Russell's position. See British Philosophy in the Mid-Century (1957), ed. C. A. Mace, p. 59.
  • 26. Words and Images (1957), p. 33 f.
  • 27. Ibid., p. 70 f.
  • 28. The Self as Agent (1957), p. 112.
  • 29. Compare for example, ibid., p. 101 f., p. 129.
  • 30. John 14. 6.
  • 31. Proceedings of the Aristotelian Society 1910–11; reprinted in Mysticism and Logic (1918), pp. 209 ff.
  • 32. Ibid., p. 231.
  • 33. Studies in Philosophy and Psychology (1930), p. 369.
  • 34. Ibid., p. 392.
  • 35. Ibid., p. 392.
  • 36. Op. cit., p. 117.
  • 37. Op. cit., p. 129.
  • 38. Gal. 4. 9.
  • 39. Nature, Man and God (1934), p. 322.
  • 40. According to J. M. Keynes (Two Memoirs, 1949) ‘Moore had a dream once in which he could not distinguish propositions from tables’.