An Analysis of Knowing

John Hartland-Swann

• Publisher: Routledge
• Year: 2015
• Print ISBN: 9781138917897
• eBook ISBN: 9781317411871
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy History & Theory

First published in 1958, this book focuses on the meaning, interpretation, and use of the verb 'to know'. In our daily lives we are often claiming to know this or not to know that; and it is not therefore surprising that the verb has played a major role in philosophical speculation from Plato down to Bertrand Russell. This book analyses the varying meanings of 'know' in its different operational roles: knowing Jones seems to have a different sort of logic from knowing French or from knowing what to do – and equally from knowing that the earth is round and from knowing how to read music. Knowing something is also different from merely believing it.

The main purpose of this book is to elucidate, in a new and original way, this whole question of the logical behaviour of 'know'; but its further and no less important purpose is to show how, once we have grasped the way in which certain key 'know'-statements function, a number of philosophical disputes may be discussed more fruitfully and settled more expeditiously. Some of the analyses offered will be regarded as controversial and will undoubtedly provoke discussion. The style is lucid and economical and technical terms are reduced to a minimum. This work is intended not only for the professional philosopher and the university student, but also for the general reader who is interested in the methods of modern philosophical analysis.

CHAPTER TWO Knowing That: Its Function in Discourse

DOI: 10.4324/9781315685342-2

The general view which I am going to put forward is that knowing, if it is knowing that, involves deciding, or accepting a decision, that something is, was or will be the case. This will be the key-point in my argument and we shall see that, if this view is sound, various important consequences ensue.

I

When we set out to examine the logical behaviour of ‘know’—in whichever of its various uses we may select—we find that two interrelated questions are involved: (i) How does ‘know’ function in discourse? (ii) When are we justified in using the word? To ask these questions about a word such as ‘eat’ would be a waste of time, since the answers are tolerably clear; but the history of philosophy, and more particularly current philosophical discussions, have shown that it is far from a waste of time to ask these questions about ‘know’—as I have already implied in the Introduction.

To illustrate this point, and at the same time introduce my own analysis, I propose to consider, as a start, the three following examples of knowing that:

1. I know that my daughter will marry you.
2. I know that Gandhi is dead.
3. I know that gold will dissolve in aqua regia.

In what sort of circumstances do statements such as these come to be made? Let us look at each in turn.

a. I know that my daughter will marry you

Imagine the following situation. I am courting Angela, the daughter of Mrs Smith. I am uncertain about my prospects, and too nervous to put a direct question as yet to Angela herself, so I say to Mrs Smith: ‘Do you think Angela will marry me?’—to which Mrs Smith replies: ‘I know she will.’ I am still not altogether convinced and riposte: ‘But how do you know? Did Angela say she would?’ ‘No,’ replies Mrs Smith, ‘but all the same I know she will.’ Now suppose that Mrs Smith considers she has good reasons for using the expression ‘I know’ rather than ‘I believe’. She may, for instance, have noticed her daughter ceasing to pay attention to other suitors, she may have seen her studying leaflets about cruises to the Balearic Isles, which I had declared to be an ideal place for a honeymoon, she may have been consulted about wedding clothes and whether a certain salary, the same as my own, would suffice to keep two people in reasonable comfort—and so on. Suppose, further, that when I at last come to the point and ask Angela the vital question, she says Yes.

Now, according to my analysis, what Mrs Smith has done is this. She has decided, on the basis of evidence which seemed to her both reliable and sufficient, that her daughter will marry me and she has expressed this decision by saying ‘I know she will’. This is not a complete analysis of Mrs Smith's ‘know’-statement; for it would be clearly unsatisfactory to translate that statement into ‘I have decided that my daughter will marry you’. One reason, though not the only reason, for this is that the translation tends to be ambiguous: it might convey the impression that Mrs Smith, in saying she had decided that her daughter would marry me, was hinting that she was going to order her daughter to marry me, or would ensure that her daughter married me. This ambiguity follows from the operational versatility of the verb ‘decide’ itself; for its most usual role is to convey that we have decided to do or bring about something—and frequently the use of deciding that can clearly imply this role, e.g. I decided that the gardener must be dismissed. Nevertheless, we are all agreed that people can decide that so-and-so is the case without intending to do or bring about anything at all, and it is only this sense of ‘decide’ which I am here invoking. The decision-analysis, then, is offered merely as the essential basis of Mrs Smith's ‘know’-statement—or, as will appear shortly, of any other ‘knowing that’-statement—and that is why I restricted myself, at the outset, to the claim that knowing, if it is knowing that, involves deciding. Having said this, I shall from now on assume that my use of the term ‘analysis’, where it refers to what I have called the decision-analysis, will be interpreted in this way.

Another question now arises, which does not stem from my particular analysis, namely, Was Mrs Smith justified, or not, in using the expression ‘I know’? Ought she to have said merely ‘I believe’ or ‘I am certain’? The immediate point to be noted is that this question cannot be settled by any appeal to that nebulous authority called Common Sense. If we did make such an appeal, we should receive conflicting answers: some people would admit, and others deny, that Common Sense sanctions the use of ‘know’ in circumstances such as these; and equally some would pay regard to, and others deny the relevance of, what might have happened in the end, i.e. Angela saying No. To obtain an unambiguous answer, then, it would seem that we must appeal to some special theory of knowing; and this is the course I shall in fact adopt a little later on. For the moment, however, I am more concerned with analysis, so let us pass on to our second example.

b. I know that Gandhi is dead

What do I imply by saying ‘I know that Gandhi is dead’? There is of course no ‘one and one only’ answer to this question, because the speaker may have uttered this ‘know’-statement either on the basis of personal experience or on the basis of what he had read or been told. But in my own case when I say ‘I know that Gandhi is dead’ I imply that I have accepted the concordant decisions of eye-witnesses and newspapers to the effect that Gandhi was murdered by an assassin and is therefore dead. These decisions were based on evidence, direct and indirect, which was regarded as both reliable and sufficient by those who made them; and it was these decisions which, when widely publicized, enabled both myself and others to say without hesitation ‘I know that Gandhi is dead’. When we inspect this situation closely, we see that there was in fact a whole chain of decisions involved, some actually forming and all capable of forming the basis of correlative ‘know’-statements. Let us assume that Abdul the reporter was on the spot near where Gandhi was standing. He saw the pistol raised, heard the shots and watched Gandhi fall; later he questioned the doctor who was summoned to the scene and who pronounced Gandhi to be dead, having so decided on the basis of his medical experience. He, the doctor, may have expressed this decision to Abdul by saying ‘It's no use hoping for a miracle; I know he is dead.’ Abdul accepted this decision of the doctor, and would have expressed his acceptance of it, if questioned by a suspicious bystander, by echoing the doctor's ‘I know he is dead’. Abdul then went to a telephone booth and rang up The Times of India office; he recited his evidence to the sub-editor, who decided to accept Abdul's report as true. The report was duly printed, under the headline ‘Gandhi Assassinated’. A shopkeeper in Delhi read the printed report, was asked by an already disquieted customer, ‘Is Gandhi really dead?’, and replied, ‘I know he is; I have just read it in The Times of India.’ The shopkeeper had thus decided to accept the newspaper's report and expressed this decision by saying ‘I know, etc.’. When the news was cabled to England, I read it in The Times. The Times, let us assume, made it clear that it regarded the report as true; it had decided to accept it without reservation. I am in the habit of accepting reports of The Times on matters of fact, and here was a case where The Times was in effect saying ‘We know that Gandhi has been murdered and is therefore dead’; so, after carefully reading all the evidence, I myself decided that Gandhi was dead and from then on, in response to any queries on this matter, I should have replied ‘I know that Gandhi is dead’.

When we come to the question of justifying my ‘know’-statement here, some philosophers would say ‘Although I do not necessarily accept your analysis, your “know”-statement here was justified—which Mrs Smith's “know”-statement was not—because you could prove, i.e. adduce conclusive evidence, that Gandhi was in fact murdered’; whereas other philosophers would say ‘I agree that your “know”-statement was justified here, but I don't admit that you can never use “know” properly unless you can state conclusive evidence’. The first alternative represents what would be Professor Woozley's view; the second what would be Professor Moore's—and in due course we shall have to examine both these views about the role played by ‘conclusive’ evidence in the justification of ‘know’-statements. Meanwhile let us pass on to our third example.

c. I know that gold will dissolve in aqua regia

As before, this statement involves a decision which might be conveyed by the sentence ‘I have decided to accept the teaching of reputable chemistry textbooks to the effect that gold will always dissolve in aqua regia’. Behind this decision of mine there lies, once again, a chain or series of previous decisions. To put the matter somewhat elliptically, chemists A, B and C who experimented with acids and metals brought into contact with each other unanimously decided that gold dissolved in aqua regia and expressed this collective decision by saying ‘We know that gold dissolves (= always has dissolved and always will dissolve) in aqua regia’. This decision was then accepted by the writers of chemistry textbooks, who duly enshrined it by writing ‘gold will dissolve in aqua regia’, thereby implying that they know it will. I then read a chemistry textbook, or talk to a qualified chemist, and learn from either of these sources that gold will always dissolve in aqua regia. Having little acquaintance with chemistry myself, but being at the same time respectful of the authoritative and unanimous pronouncements of reputable chemists, I decide to accept the official teaching, which is merely the official decision, with regard to the behaviour of gold in aqua regia and to express this by saying ‘I know that gold will dissolve in aqua regia’.

Now, when it comes to the question whether I was justified in my use of ‘know’ here, we may note that whereas the layman would have little hesitation in allowing that my statement was legitimate, most philosophers would be prone to veto it on the ground that I am using ‘know’ about the future. They would not consider it relevant that my decision was based on the unanimous decisions of reputable chemists that gold always has behaved and always will behave in a perfectly stable manner when immersed in aqua regia, whereas Mrs Smith's decision was based on her calculations about the behaviour of human beings, which is generally agreed to be relatively unstable. In their view, both statements, since they equally refer to future events, are on a par. But this equation obviously calls for investigation, since it is by no means clear that it is in fact logical to treat the two statements alike. This, however, must be undertaken when we go into the whole question of justifying or vetoing ‘know’-statements. For the present, I am more concerned to develop and strengthen my arguments for the thesis I have been maintaining, namely, that when you say you know that S is, was or will be P, you are in effect announcing that you have made or accepted some decision. Let us therefore turn our attention to a familiar dispositional analysis of knowing that, where ‘know’ is translated into ‘able to state correctly’.¹ Now it is essential for me to examine the status of ‘correctly’ here, for my claim that all ‘know’-statements in-volve a decision, or the acceptance of a decision, is closely bound up with the logical behaviour of this adverb.

¹ I am aware that this analysis does not meet with universal assent and I shall consider arguments against its adequacy in a later chapter.

II

To start off in a dogmatic vein, I will say that the force of ‘correctly’ in this dispositional analysis of ‘know’-statements, and I think in any context, is ‘in accordance with some decision’, and usually ‘in accordance with a dominant decision’. I shall consider later the obvious query ‘But surely a dominant decision can sometimes be incorrect?’ and likewise its counterpart ‘Does a new dominant decision on a specific topic render the former dominant decision incorrect?’ At present, I want to examine how far my analysis of ‘correctly’ ties up with my analysis of ‘I know that Gandhi is dead’ and ‘I know that gold will dissolve in aqua regia’. I analysed the first of these, it will be remembered, as ‘I have decided to accept the concordant decisions of eye-witnesses, as reported in the newspapers, that Gandhi is dead, and I express this acceptance by saying “I know that Gandhi is dead”’. Suppose, now, I give the (provisionally accepted) dispositional analysis of this statement; it will then run ‘I am able to state correctly that Gandhi is dead’—and this analysis gives no prima facie indication that any element of decision is involved. But the decision-element becomes immediately obvious if we apply the logical probe to ‘correctly’ and analyse it as ‘in accordance with a dominant decision’. But do we have to analyse ‘correctly’ in this way? Could we not just say that ‘correctly’ means ‘in accordance with the facts’ or ‘in accordance with what is the case or the truth’? There are long and short answers to this query, and the long answers commit one to discuss conflicting theories of truth, whether empirical propositions ‘picture’ or ‘correspond to’ reality and what not. I do not propose to embark on any such discussion; nor is it necessary for me to do so, since a relatively short answer will solve our particular problem. Suppose someone says to me ‘The earth is round and not flat’ and I say ‘Are you stating a fact or a hypothesis?’ The reply will be ‘I am stating a fact’, by which is meant ‘I am asserting that so and so is the case’. But this ‘fact’, like a host of other empirical ‘facts’, can only be plausibly stated to be such just because various people, considered qualified to decide, have actually decided that the earth is indeed round and not flat. Suppose again that someone says to me ‘I can translate the Dunciad into idiomatic Greek’, and I say ‘Are you stating a fact?’ To establish whether this person is stating a fact a decision has got to be made, either by myself or (better) by some more competent Greek scholar. Suppose, finally, that someone says to me ‘I have a pain in my knee-cap’ and that, facetiously perhaps, I say ‘Are you stating a fact?’ The expected reply—‘Of course I am’—can only be made because, owing to certain neural reactions, the person concerned has decided that he has a pain in one particular place, namely, his knee-cap. So what I am claiming is that all ‘correct’ empirical statements and sense-statements¹ —i.e. those which are ‘in accordance with the facts’—are statements based on some decision, either personal or what I have called dominant. Notice by the way that, except in the case of sense-statements, the proper use of ‘correctly’—and by ‘proper’ I simply mean what people in general will regard as such—is always dependent, directly or indirectly, on a dominant decision in the end. Thus, if I decide, contrary to the dominant decision of qualified tropical zoologists, that Malayan cobras are sometimes non-poisonous and claim that I am correctly stating a fact about Malayan cobras, I can only expect the world at large to admit that I am stating correctly if and when the qualified zoologists come round to my view and thus alter their previous decision.

¹ I use the expression ‘sense-statement’ to cover statements of the type ‘I am seeing (or sensing) a blue patch’ or ‘I have (or feel) a pain’. This convention was, I believe, initiated by Professor Malcolm and is frequentl...

Table of contents

• Cover Page
• Half-Title Page
• Title Page
• Copyright Page
• Original Title Page
• Original Copyright Page
• Dedication
• Prefatory Note
• Table of Contents
• Introduction
• Knowing That: Its Function in Discourse
• Knowing That: The Problem of Justification
• Knowing How and Knowing That
• Other Uses of ‘Know’
• Knowing and Being Aware Of
• Knowing and Believing
• Our Knowledge of the World
• Two Final Problems
• Index of Proper Names