SOURCE: “The Attack on Matter” and “Immaterialism and Common Sense,” in The British Empiricists, Oxford University Press, 1992, pp. 106-38.

[In the first essay that follows, Urmson gives an overview of Berkeley's conflict with John Locke in the area of the definition of matter. In the second, he examines Berkeley's contention that, in order for his theory of immaterialism to be considered true, it must coincide with common sense.]

THE ATTACK ON MATTER

George Berkeley was born on 12 March 1685, near Kilkenny in Ireland. His ancestry was English, but his grandfather, a royalist, moved to Ireland at the time of the restoration. Berkeley considered himself to be an Irishman; he referred to Newton as ‘a philosopher of a neighbouring nation’ (P [Principles of Human Knowledge] 110) and, commenting sarcastically in his private notebook on what he regarded as a philosophical absurdity, he wrote: ‘We Irishmen cannot attain to these truths’ (C [Philosophical Commentaries] 392). He was sent to Kilkenny College, which Congreve and Swift had recently attended, and then, aged only fifteen, to Trinity College, Dublin, in 1700. There he studied mathematics, languages, including Latin, Greek, French and Hebrew, logic and philosophy; the philosophy course was up to date and included the study of Locke, the French philosopher-theologian Malebranche, and other very recent and contemporary thinkers. He graduated B.A. in 1704 at the age of nineteen and then remained at Trinity College studying privately until he was elected to a fellowship in 1707. He was ordained deacon in 1709 and priest in 1710; he continued to hold his fellowship until 1724, when he resigned to become Dean of Derry, though extended leaves of absence took him first to London and then to Italy.

Though he continued to write for many years after leaving Trinity College, Dublin, it was during the tenure of his fellowship that he wrote the works for which he is now famous. His first work, An Essay towards a New Theory of Vision, appeared in 1709, when he was twenty-four; it is a work as much of experimental psychology as of philosophy, and in it he principally discusses how we perceive by sight the distance, size and position of objects. The next year, in 1710, he published A Treatise concerning the Principles of Human Knowledge, usually known simply as The Principles; this is the most important of all Berkeley's writings and contains the most complete account we have of the philosophical position which he was never to abandon. In 1713 he published the Three Dialogues between Hylas and Philonous, a more popular exposition of his view in which Philonous (whose name means ‘lover of mind’) vanquishes in argument and converts to his view Hylas (‘matter’), the materialist. In 1712 he published his Latin essay De motu [On Motion] which contains by far the fullest account we have of Berkeley's view of the nature of the natural sciences.

Those are the main works published by Berkeley from which we may learn his basic metaphysical and epistemological doctrines, bearing witness to a youth of extraordinary intellectual activity. But, in addition to his published writings, we also have available to us a unique and most interesting document. In 1705, soon after graduating, Berkeley started to write a series of notes on philosophical topics in which he worked out the basis of his new and revolutionary views and made strategic plans for their publication. These notes contain ideas which he immediately rejects and theories which are silently abandoned in his published works, as well as the essential of his position; one cannot therefore safely attribute to him views found only there and not in the published work. But these notes are most valuable as an aid to the correct understanding of the published works and shed fascinating light on the creative processes of a philosopher of genius. They were written in a quarto notebook which was unknown until it was discovered by A. C. Fraser and published by him in 1871 under the title Commonplace Book; later editors disliked this name and the notes are now generally called The Philosophical Commentaries.

Berkeley's main targets for attack were to be John Locke and Sir Isaac Newton; he was well acquainted with the works of both and admired both of them greatly. Of Newton he wrote: ‘I have taken as much pains as (I sincerely believe) any man living to understand that great author, and to make sense of his principles … So that, if I do not understand him, it is not my fault but my misfortune’ (L [The Works of George Berkeley, Bishop of Clogne,edited by A. A. Luce and T. E. Jessop] iv 116). He also referred to Newton as ‘a philosopher of a neighbouring nation whom all the world admire’ (P 110), and as ‘an extraordinary mathematician, a profound naturalist, a person of the greatest abilities and erudition’ (L iv 114). Of Locke he wrote in his notebook: ‘Wonderful in Locke that he could, when advanced in years, see at all thro’ a mist; it had been so long agathering and was consequently thick. This more to be admired than that he did not see farther’ (C 567). He is, of course, using the word ‘admire’ in its usual sense at that time of ‘wonder at’.

But this respect was not extended to all those whom Berkeley called the mathematicians; in his private notebook he wrote: ‘I see no wit in any of them but Newton. The rest are mere triflers, mere Nihilarians’ (C 372). But these words were for his eyes only; he realised that in a publication, as a young man attacking the establishment, he must exhibit a more conciliatory attitude. So in the same notebook he wrote: ‘Mem. Upon all occasions to use the utmost modesty—to confute the mathematicians with the utmost civility and respect, not to style them Nihilarians, etc. N.B. To rein in ye satirical nature’ (C 633-4). There are many self-addressed memoranda in the notebooks, often significant, often amusing, and only rarely opaque like: ‘Mem. Story of Mr. Deering's aunt’ (C 201).

So the principal objects of attack were Newton and Locke. But there was only one main element in their views that he repudiated, that being the doctrine of matter. Berkeley shared their view that ideas are the sole object of the human mind, for example. It is most vital to remember the historical context in which he was writing; often the modern reader finds Berkeley assuming, without more than perfunctory argument at most, things which to him are obscure and doubtful; these are generally things that at the time were accepted by most men and certainly by his opponents, who would have found it tedious and uncalled-for if he had argued them at length. It is also important to be aware that Berkeley never questioned the value of Newton's scientific work, which he believed to be independent of the doctrine of matter. But this doctrine he regarded as an unintelligible and totally superfluous element in his views and one that could be excised without damage to what was left. He attacked it because he believed that it was false, and he attacked it with passion because he believed that it had evil practical consequences. The full title of the Principles, the work in which he first attacked the doctrine of matter, is significant: A Treatise concerning the Principles of Human Knowledge wherein the chief causes of error and difficulty in the sciences, with the grounds of scepticism, atheism, and irreligion, are inquired into.

So there are at present four main takes before us: first, to understand why Berkeley thought matter unintelligible; second, to understand why he thought matter superfluous; thirdly, to discover why he thought belief in it dangerous; fourth, to see what account of the world Berkeley was able to give without incorporating matter into it.

THE UNINTELLIGIBILITY OF MATTER

To understand why Berkeley thought that matter was unintelligible, that the word ‘matter’ was without meaning in the way it was used by Newton and Locke, we must first realise that he was in a certain respect an extreme empiricist. The concept of empiricism is itself vague, and different philosophers have understood it differently; but luckily we need not go into this problem, for the way in which Berkeley was an empiricist becomes clear from the very first sentence of the Principles of Human Knowledge. This sentence reads as follows:

It is evident to any one who takes a survey of the objects of human knowledge, that they are either ideas actually imprinted on the senses; or else such as are perceived by attending to the passions and operations of the mind; or lastly, ideas formed by help of memory and imagination—either compounding, dividing, or barely representing those originally perceived in the aforesaid ways.

By ideas imprinted on the senses he means visual images, heat, smells, tastes and all what Locke would have called ideas of primary and secondary qualities; ideas obtained by attending to the passions include introspective awareness of love, hatred, joy, grief and the like—we can think of these passions because we have experienced them. By attending to the operations of the mind we acquire the ideas of thought itself, of memory, of imagination, and so forth. The ideas gained in these various ways can be recalled in the memory, and in the imagination we can compound and divide to produce what are in some sense new ideas—we can imagine a dragon or a mermaid. But this is composition and division; we divide our ideas of the human body and the body of a fish, and by compounding the top end of a human body and the lower end of a fishy body we make the idea or mental picture of a mermaid. So all the materials of thought are such as are supplied by experience, and we cannot think about or have knowledge about what is not the kind of thing we have experienced.

This empiricism is much emphasised in Berkeley's private notebooks: ‘Foolish in men to despise the senses. If it were not for them the mind could have no knowledge, no thought at all’; ‘By idea I mean any sensible or imaginable thing’; ‘Pure intellect I understand not’ (C 539, 775, 810). He even sets out his empiricism in five simple propositions:

1 All significant words stand for ideas.

2 All knowledge about our ideas.

3 All ideas come from without or from within.

4 If from without it must be by the senses, and they are called sensations.

5 If from within they are the operations of the mind, and are called thoughts.

(C 378)

Some philosophers of the seventeenth and eighteenth centuries had rejected empiricism of this kind. Thus Descartes and Leibniz claimed that since they had the idea of God, the idea of matter, the idea of soul or spirit, and since these were not obtainable in experience, there must, therefore, be what they called innate ideas, ideas which the human mind was so constituted as to frame and apply to experience without gaining them in experience. Such ideas were not sensible or picture-like. Since Locke had repudiated any such view, Berkeley did not think it necessary to argue against it; as the opening words of the Principles, quoted above, make clear, he himself thought it was evidently wrong. Locke, on the other hand, in spite of having expressly argued against innate ideas, still claimed to have the ideas of God, of substance, of matter and the like. This Berkeley attributed to Locke's recognition of a class of abstract ideas. In addition to the mind's compounding and dividing of ideas, which Berkeley allowed, Locke attributed to the mind a further power of forming ideas by abstraction, and it was this that Berkeley thought to be the cause of what he regarded as Locke's mistake. As for Berkeley, since he was an extreme empiricist he denied that he had any idea of an inherently imperceptible matter, an imperceptible God or even of imperceptible ‘finite spirits’; just how Berkeley managed to combine this empiricism with his undoubted and undoubting Christian orthodoxy we shall obviously have to inquire in due course.

Let us try to understand the basic grounds for the rejection of matter as unintelligible a little more fully before we examine the detail of Berkeley's arguments, which vary greatly in cogency. It is clear that, for Berkeley, what we can think of must be imaginable; for all objects of the mind are ideas and by an idea Berkeley means what is sensible or imaginable, and what is sensible is imaginable. But matter is said by its adherents to be something which is not coloured, not warm or tepid or cold, odourless, tasteless etc.—the ideas of secondary qualities resemble nothing in matter. So if we try to imagine matter it is impossible, since we cannot imagine anything that lacks all secondary qualities. If I visualise something it must have some colour, however dingy and nondescript; it cannot just be a shape of no colour at all if I am to see it. If I imagine myself touching something it must feel hard or soft, and clearly the senses of hearing, taste and smell are concerned exclusively with secondary qualities. So one cannot perceive something having only primary qualities, nor can one imagine such a thing; yet that is what matter is supposed to be. So one cannot have an idea of matter, it cannot be an object of one's mind, it is unthinkable; we can attach no meaning to the word ‘matter’. Perhaps at this stage of the argument some will be inclined to protest that though we cannot imagine matter as defined by Newton and Locke we can conceive of it in abstraction from secondary qualities. But Berkeley is prepared for this move; it involves, he thinks, Locke's doctrine of abstract ideas, and this, he believes, he can show to be absurd.

But let us not yet pursue the details of Berkeley's argument. First we should see that the line he takes is quite persuasive, independently of the particular form in which he presents it. Three-quarters of a century later, Kant will attack Berkeley's presentation of the argument as he understands it; but he too will argue that concepts can be applied only to phenomena and not to things in themselves, that we can think only about objects of possible experience, which certainly do not include matter. In the present century many have argued that of statements that purport to describe the world those alone were meaningful which were capable of verification and falsification by empirical means; but statements to the effect that what we experience is caused by an imperceptible matter are certainly unverifiable experimentally.

The argument may be put, as it was sometimes put by Berkeley, in a linguistic form. If we ask how we can come to understand, say, the word ‘red’, it would seem that, in principle, it can only be by somebody pointing to something red and saying ‘That is red’—in principle, since clearly we learn such words by a less formal version of the same process; perhaps we hear somebody say to somebody else ‘I do hate that shocking pink dress that that girl is wearing’, we take a look at the dress, and say to ourselves ‘So that is shocking pink, is it?’ If one were ignorant of the game of rugby football, one could learn the meaning of the word ‘scrum’ by observing in what circumstances it is proper to say such things as ‘A scrum has been formed’ and so on. Of course, it is much more likely that we shall learn the meaning of such words at least partly by verbal explanation; but to verbal explanation an end must come. Some basic stock of words, it seems clear, must be learnt by confrontation with experience, before verbal explanation can begin.

But language can be abused. In Berkeley's own words, ‘there are many names in use among speculative men which do not always suggest to others determinate, particular ideas, or in truth anything at all’ (P, introduction, para. 19). The abuse may be conscious and in jest, as when a modern philosopher, being teased for requiring more sleep than most people, replied that she did not sleep more than other people, she simply slept more slowly. It may be hard to determine whether a use of words is an abuse; thus Newton spoke of absolute motion and contrasted it with relative motion; Berkeley in the Principles claimed that this was an abuse of words and that no meaning could be attached to the expression ‘absolute motion’. Most scientists and philosophers since have followed Berkeley in this, but what Newton could not see could not be obviously right. But it is very hard to doubt that in some abstract disciplines, such as philosophy, meaning can become so rarified as to be undetectable.

It is clear that the word ‘matter’ has uses in which it is readily intelligible. As Berkeley represents Philonous as saying in the Dialogues: ‘If by material substance is meant only sensible body … then I am more certain of matter's existence than you or any other philosopher pretend to be’ (L ii 237). But the argument is that if matter is to be regarded as something inherently imperceptible then ‘the Matter philosophers contend for is an incomprehensible Somewhat’ (L ii 233). The word is familiar and so we overlook the fact that we have attenuated its meaning until it has vanished into thin air.

There is another argument that Berkeley uses against the intelligibility of matter which is rather more obviously unsatisfactory and rests on a confusion between the notions of substance and of matter. In this argument Berkeley very significantly refers to material substance rather than simply to matter. The essentials of this argument can readily be given in Berkeley's own words.

If we inquire into what the most accurate philosophers declare themselves to mean by material substance, we shall find them acknowledge they have no other meaning annexed to those sounds but the idea of Being in general, together with the relative notion of its supporting accidents. The general idea of Being appeareth to me the most abstract and incomprehensible of all other [, while] it is evident support cannot here be taken in its usual or literal sense, as when we say that pillars support a building. In what sense therefore must it be taken? For my part I am not able to discover any sense at all that can be applicable to it.

(P 17)

In referring to ‘the most accurate philosophers’ Berkeley clearly has in mind principally John Locke and, indeed, the following passage from Locke's Essay on Human Understanding: ‘If any one will examine himself concerning his notion of pure substance in general, he will find he has no other idea of it at all, but only a supposition of he knows not what support of such qualities which are capable of producing simple ideas in us.’ But this is an admission, or, rather, a claim, that the notion of substance is ultimately unintelligible, not an admission that the notion of matter is in any way defective. Locke sees the same difficulty in the notion of substance whether we are talking about bodily or mental substances, as the following quotation shows: ‘we have as clear a notion of the substance of spirit as we have of body: the one being supposed to be (without knowing what it is) the substratum to those simple ideas we have from without; and the other supposed (with a like ignorance of what it is) to be the substratum to those operations which we experiment in ourselves within’.

It is surely clear that the dubious intelligibility of the notion of substance as the mere somewhat that is the bearer of the qualities attributed to it cannot be properly taken to throw doubt on the concept of matter. Matter is not an indescribable somewhat supporting qualities but an imperceptible cause of perceptible ideas; both these views may be unacceptable, but, if so, the unacceptability of matter cannot be derived from the unacceptability of substance. The doctrine of matter is independent of the substance—attribute analysis of things, which appears only in Locke among the authorities we have quoted. By speaking of material substance so often Berkeley allows himself to confuse two distinct issues.

THE SUPERFLUITY OF MATTER

We have seen Berkeley's reasons for wishing to maintain that the corpuscularian philosophers had been unsuccessful in their attempts to make their concept of an inherently imperceptible matter intelligible. It might have been the case that if matter were rejected a gap would be left and it would be necessary to postulate something else to fill the gap thus left. But Berkeley thought that this was not so; not only was matter unintelligible, but there was no gap to be filled: a complete and satisfying account of the world could be given without invoking it or any substitute for it.

If we look at Locke's inventory of the furniture of the world it would seem to include four elements:

1 God, the creator of everything else and the law-maker who determines the character and history of his creation.

2 Matter, which for certain purposes we may divide into that which impinges on our sense-organs to cause ideas and that which constitutes the sense-organs and nervous system on which other matter acts. But there is here no fundamental difference.

3 Ideas, which are mental but caused by the action of matter on the nervous system.

4 The mind that is aware of these ideas and operates on them and with them in thinking.

This ontology, which was shared in all essentials by Newton, most conspicuously lacks the physical object of common, everyday thought and discourse. We can, of course, still speak of chairs and tables; but on the physical side they are but dense collections of elementary corpuscles, and our seeing a chair or table is merely the occurrence of certain ideas caused by the particles emitted by such a collection. A further point to be noted is that the causal links between the elements in Locke's ontology are inexplicable. That the actions and purposes of God are inscrutable we may take for granted; but also the way in which matter acts on mind to produce ideas is scientifically inexplicable. A causal story in mechanistic terms can be given of particles being emitted from those collections of particles which constitute physical bodies; these particles strike on the eye or other sense-organs and by impact on the nerve-endings cause motion in the ‘animal spirits’ which were conjectured to be contained in the nerves; the motions of the animal spirits in the nerves is itself communicated to the brain; but how the motions of the particles that constitute the relevant portion of the brain cause the occurrence of ideas in the mind is not mechanically explicable (and perhaps modern physiology is equally powerless to offer an explanation). Locke gives an example: ‘a violet, by the impulse of such insensible particles of matter of peculiar figures and bulks, and in different degrees and modifications of their motions, causes the ideas of the blue colour and sweet scent of that flower to be produced in our minds; it being no more impossible to conceive that God should annex such ideas to such motions with which they have no similitude, than that he should annex the idea of pain to the motion of a piece of steel dividing our flesh’. So Locke has to appeal to the direct action of God, without any scientific explanation of how the links in the causal chain are joined together.

If we look to see what reasons Locke had to offer for his belief in matter, we are bound to be disappointed. In the chapter entitled ‘Of our knowledge of the existence of other things’ in Book IV of his Essay he tells us that it can be gained only by sensation; but it is far from clear how knowledge of the existence of imperceptible matter could thus be gained. He tells us that our certainty is as great as our condition needs and, assuming that unless we posit matter we must consider waking to be indistinguishable from dreaming, tells the doubter that he ‘doth but dream that he makes the question; and so it is not much matter that a waking man should answer him’. But knockabout humour is scarcely an adequate substitute for argument on such topics as these.

But Locke also says that ‘it is plain those perceptions are produced in us by exterior causes affecting our senses’, and it is this that Berkeley assumes to be the main argument that leads to a belief in matter. God, on this view, created matter to be the causal agency through which ideas are caused in sentient creatures. Thus, in the dialogues between Hylas and Philonous, when Hylas is driven to admit that he cannot sustain the notion of matter as conceived of by Locke and Newton, he falls back on the claim that matter is ‘an instrument, subservient to the supreme Agent in the production of our ideas’, and argues that he has some understanding of matter since ‘I have some notion of instrument in general', which I apply to it.’

To this move Berkeley has a blistering retort, the essence of which can be given in the words given by Berkeley to Philonous:

‘Is it not common to all instruments, that they are applied to the doing those things only which cannot be performed by the mere act of our wills? … How therefore can you suppose that an All-perfect Spirit, on whose Will all things have an absolute and immediate dependence, should need an instrument in his operation, or, not needing it, make use of it? … And the use of an instrument sheweth the agent to be limited by rules of another's prescription, and that he cannot obtain his end but in such a way, and by such conditions. Whence it seems a clear consequence, that the supreme unlimited Agent useth no tool or instrument at all.’

(L ii 218-19)

This argument, of course, assumes that there is a God who is the omnipotent first cause; but, since those he was arguing against conceded it, Berkeley was entitled to make the assumption when arguing against them, especially as he had arguments for the existence of God in reserve. Given this basic assumption, the argument seems good. To postulate both God and matter as an explanation of our sense-experience seems to be superfluous, especially when it is agreed to be totally obscure why the hypothetical matter should cause sensations of the character that they do in fact have. Thus Berkeley at this point seems justified in claiming that the ontology of Locke is redundant and matter is superfluous. In the next two chapters we shall examine Berkeley's positive accounts of science and common sense without the aid of the concept of matter; but first we must consider one more attack on the materialists to which Berkeley attached great importance.

ABSTRACT IDEAS

We have seen that it is basic to Berkeley's attack on matter to hold that to think of something is, ultimately, to imagine it. Matter is unintelligible precisely because one could not see or imagine anything of the kind that matter is supposed to be, having primary qualities but lacking colour, smell, taste or sound. Berkeley is aware that at this stage some of his opponents, then as in our day, will want to claim that Berkeley's criteria of intelligibility are too restrictive. In the jargon of the philosophy of Berkeley's time the objection takes the form that while matter is not imaginable we can none the less have an abstract idea of matter. This Berkeley denies. Before we can consider the controversy we must try to understand the theory of abstract ideas, and this will be the objective of the next few pages.

It was a rarely questioned orthodoxy among the principal philosophers of the seventeenth and eighteenth centuries that thinking was primarily having ideas before the mind. That one can think in words was, of course, recognised, but language was regarded as being actually necessary only for purposes of communication with others. In Locke's typical words: ‘The comfort and advantage of society not being to be had without communication of thoughts, it was necessary that man should find out some external sensible signs, whereby those invisible ideas which his thoughts are made up of might be made known to others.’ One could think privately in words; but words stood to ideas as cheques do to real money, and cheques could be returned to drawer. All agreed that thinking in words could be dangerous—‘Words often used without signification’ is one of Locke's section headings.

With this view of the relation between thinking in ideas and thinking in words, which can indeed be traced back to Aristotle's On Interpretation, Berkeley was in full agreement. Moreover, he was convinced that most of the unintelligible metaphysics that he was attacking resulted from bogus thinking in meaningless words; the last eight paragraphs of his introduction to the Principles are devoted to explaining that this is so. In paragraph 21 of this introduction he writes:

Most parts of knowledge have been so strangely perplexed and darkened by the abuse of words, and general ways of speech wherein they are delivered, that it may almost be made a question whether language has contributed more to the hindrance or advancement of the sciences. Since words are so apt to impose on the understanding, I am resolved in my inquiries to make as little use of them as possibly I can: whatever ideas I consider, I shall endeavour to take them bare and naked into my view; keeping out of my thoughts, so far as I am able, those names which long and constant use hath so strictly united with them.

But what is it to think in bare and naked ideas? It is clear that basically ideas were thought of as mental images, and ‘image’ or ‘picture’ was the normal seventeenth-century meaning of ‘idea’ in non-philosophical writing also. Even the rationalistic Descartes, who was mainly responsible for the wide currency of the word ‘idea’ in philosophy, said that ‘some of our thoughts are like pictures of things, to which alone the name idea is properly suited’. Both Berkeley and Hume clearly regard thinking in ideas to consist in having mental images, and for Locke too, at least when alleged abstract ideas are not involved, it seems that ideas are normally mental images; he can speak of ‘ideas or images’ as if the terms were interchangeable.

But it is not clear what thinking in mental images is supposed to be. Just having a mental image does not seem to be enough. One may, for example, have a mental image which is approximately round and red, but this on its own scarcely seems to be a thought about anything. Let us suppose that this red and round image has a close resemblance to a tomato, since there is no doubt that the philosophers with whom we are concerned believed that resemblance made the idea an appropriate way of thinking about some object. But if the occurrence of the idea is a thought, and indeed a thought having some reference to tomatoes, it remains unclear what thought about tomatoes it is. Let us make a list of possible candidates, expressed in verbal form:

This is a tomato; Tomatoes are red; That tomato was red and round; Some tomatoes are red; The tomato is a fruit; Is this a tomato? Some round objects are red.

Which, if any, of these, and of countless other possible candidates, is the appropriate verbalisation of the thought one thinks when one has before one's mind the round, red image resembling a tomato?

The difficulty can be approached from the opposite direction. Let us suppose that one is to have the thought ‘This is white’ in the form of an idea or mental image. Now if we summon up a white mental image it cannot be just white and have no other characteristics; it must presumably be bright or drab, round or of some other shape, with this or that background. Let us suppose that it is bright and round as well as white, and now the problem is what determines that the occurrence of the idea constitutes the thought ‘This is white’ rather than the thought ‘This is bright’ or ‘This is round’, if, indeed, it constitutes any of them.

These difficulties were to some extent seen, though not with exemplary clarity, by Locke among others. He, relying heavily on The Art of Thinking by Descartes' friend and collaborator, the theologian Arnauld (1612-94), made use of the notion of abstract ideas to try to solve some of them. Most of our thinking requires abstract ideas, according to Locke. He holds that of the words of the English language only proper names signify concrete or particular ideas; all other words which stand for any idea at all stand for abstract ideas. So abstract ideas are as essential to non-verbal thought as words other than proper names are essential to its verbal expression.

What then is an abstract idea? Locke tells us four times in his Essay, in Book II, Chapter xi, Section 9, Book II, Chapter xii, Section 1, Book III, Chapter iii, Sections 6 and 10 and Book IV, Chapter vii, Section 9. Of these the first three are very similar to each other; since Berkeley in his violent attack on abstract ideas quotes only the final and most paradoxical account from Book IV, we should perhaps note what these other concurring accounts say. In Book II Chapter xi we read:

The mind makes the particular ideas, received from particular objects, to become general; which is done by considering them as they are in the mind such appearances, separate from all other existences, and the circumstances of real existence, as time, place, or any other, concomitant ideas. This is called abstraction, whereby ideas taken from particular beings become general representatives of all of the same kind; and their names general names, applicable to whatever exists conformable to such abstract ideas. Such precise, naked appearances in the mind, without considering how, whence, or with what others they came there, the understanding lays up (with names commonly annexed to them) as the standards to rank real existence into sorts, as they agree with these patterns, and to denominate them accordingly. Thus the same colour being observed today in chalk or snow, which the mind yesterday received from milk, it considers that appearance alone, makes it a representative of all of that kind; and having given it the name whiteness, it by that sound signifies the same quality wheresoever to be imagined or met with; and thus universals, whether ideas or terms, are made.

This is not entirely unambiguous, for it is not clear what considering whiteness separately involves. If it is simply a matter of disregarding all other features of the idea, then a white, bright and round image can be used as the abstract idea of white by simply ignoring the other features and treating it as the general representative of everything having this feature. Similarly the same image could have been made the abstract idea of brightness, if we had chosen to disregard the whiteness and other features. Presumably it could have been made the general idea of colour by disregarding the differences between it and red or green images. While this is by no means a full and adequate explanation, we do have here some indication of how having a red and round mental image could constitute the occurrence of a thought and of one particular thought rather than another. In principle, the answer is that it becomes such as a result of some rule-like decision; a red and round image may become the thought of a tomato if we decide to treat this image as the general representative of all things having the characteristics of a tomato.

But there is another possible interpretation of Locke's words, in which the word ‘separate’ is taken more literally. On this interpretation, when Locke speaks of separating the colour of milk from all other circumstances and ideas he is demanding that we should have an image of which the only possible description is that it is white. On this interpretation it would seem that he is demanding the evidently impossible and, since it is better to interpret an author as making sense when it is possible plausibly to do so, we may well prefer the first interpretation.

Berkeley, however, quotes only the perplexing words of Book IV of Locke's Essay, in which the following extract occurs:

Does it not require some pains and skill to form the general idea of a triangle (which is yet none of the most abstract …), for it must be neither oblique nor rectangle, neither equilateral, equicrural, nor scalenon; but all and none of these at once. In effect, it is something imperfect that cannot exist, an idea wherein some parts of several different and inconsistent ideas are put together.

Berkeley has a good deal of fun from this. In his notebooks he had written: ‘Mem. To bring the killing blow at the last, e.g. in the matter of abstraction to bring Locke's general triangle in the last’ (C 687). He refers to it sarcastically as sublime speculations. It seemed to him to be self-evidently absurd.

But, once again, it is possible to interpret Locke here as making better sense, regarding the literal absurdities in his account as rhetoric in a purple passage glorifying the rational powers of mankind. If we ask whether the concept of a triangle includes being isosceles or scalene, the answer is that it includes neither, since it is possible to deny either of them of some triangles; in this sense the abstract idea of triangle is of something neither scalene nor isosceles. If we ask whether the concept of a triangle excludes being isosceles or scalene, the answer is that it excludes neither, since it is possible to assert either of them of some triangles; in this sense the abstract idea of triangle includes both being isosceles and being scalene. But since nothing can be merely triangular without being either equilateral, scalene or isosceles, being triangular cannot be a complete account of any existent thing. Thus Locke's rhetoric may be taken as illustrating the character of a generic concept which covers several species, though it must be owned that his wording invites misunderstanding if I have correctly interpreted him.

Rightly or wrongly, however, Berkeley regards the doctrine of abstract ideas as requiring there to be mental image of an essentially indeterminate nature, and this he regards as absurd. Even if there is the possibility of some degree of indeterminacy of images (if we imagine a speckled hen must it have some definite number of speckles?), still the notion of an image which is merely, say, white and nothing else is indeed absurd, and Berkeley is surely right in rejecting the doctrine of abstract ideas as he interprets it.

Given this view of the theory of abstract ideas, two points now need to be discussed. First, why does Berkeley think that the theory aids and promotes the metaphysics of materialism; secondly, what account does Berkeley give of thinking non-verbally without the aid of abstract ideas?

It is basic to Berkeley's position to claim that matter as described verbally by the corpuscularians is unintelligible since we cannot imagine anything which has primary qualities but including no ideas of secondary qualities. We cannot, for example, visualise a colourless expanse. He also thinks that, while some would claim to think of unperceived objects, it is obviously impossible to imagine an unperceived object, to separate existence from being perceived. What Berkeley is doing in his attack on abstract ideas is to counter the move that, while matter is not imaginable, it is none the less conceivable, since we can have an abstract idea of extension without colour, just as we can have, as Locke explained, an abstract idea of whiteness. In the same way, says Berkeley's imaginary opponent, we can conceive of an object existing unperceived, since we can abstract existence from perception and imagination.

Given Berkeley's interpretation of abstract ideas, which surely does make them indefensible, it is clear that he is right in thinking that the intelligibility of matter cannot be defended by an appeal to them. What I find odd is the claim of Berkeley that the doctrine of abstract ideas was closely connected with the belief in matter. I can think of no place where Locke or any other philosopher made use of the doctrine of abstract ideas to defend the intelligibility of matter. Moreover, if the alternative account of this doctrine which I have put forward is correct, the doctrine could not be used in this way. As I understand Locke, he is explaining not how whiteness, for example, can be conceived of as existing on its own, but how we can think of whiteness as distinct from the other features that any white thing would have; so the fact that we can think of the primary qualities without thinking of colour or taste could not be consistently regarded by Locke as a ground for believing in the existence of colourless and tasteless matter. It is not clear that Locke ever saw the problem of the intelligibility of matter on an empiricist hypothesis, or a fortiori, that he ever defended it. It was one of Berkeley's achievements to see the problem.

So Berkeley will have no truck with abstract ideas. But he does recognise that he must give some account of generality. If, as Berkeley seems to think, we can give an adequate account of thinking of a triangle—some particular triangle—in terms of having an image of it, what account is he to give of the thought of all triangles? He starts to do this in the introduction to the Principles by making a distinction between a special kind of idea, called an abstract idea, which cannot exist, and a general idea which is not a special kind of idea, but simply an ordinary idea or mental image that fulfils a special function. Generality, he holds, is a kind of role, not an internal characteristic. An idea is made general ‘by being made to represent or stand for all particular ideas of the same sort’ (para. 12). He illustrates his meaning with an example drawn from geometry. A geometer will prove some theorem with reference to some particular line; but the proof is general because, though it is one particular line, ‘as it is there used, it represents all particular lines whatever’, and Berkeley takes this proof to be the same as one concerning ‘a line in general’. General ideas are on a par with the drawn line of the geometer.

Berkeley is surely right in claiming that a sign becomes general not by acquiring a peculiar character but by being used in a certain way. A full discussion is impossible here, but one or two difficulties may be mentioned. First, Berkeley says, in a quotation given in our previous paragraph, that the general idea is to represent all particular ideas ‘of the same sort’. But, if I have before my mind some particular idea of, say, a line, how am I to know whether it is representative of all lines or of all short lines, or of all straight lines, or of all geometrical figures, or … ? If it is to be representative of all lines, do I perhaps need to have the general idea of a line already to determine what is of the same sort and so what the line represents? But this clearly leads to a vicious regress.

The second difficulty is one which beset all the theorists who used the terminology of ideas. If we consider verbal thought we can readily distinguish between the words we use, which we may call the thought-vehicle, and the object of thought. If I think ‘The cat is on the mat’, the thought-vehicle is that set of five words, and the object of thought an animal. I can, of course, think about words in words, as when I think of ‘The cat is on the mat’ that it contains five words. But still the words which are the thought-vehicle are readily distinguished from the words which are the object of thought. Now if we regard the objects of our experience as being really ideas and not cats and mats, and if we call thinking the having of ideas, then we call the thought-vehicle and all objects of thought by the same name, ‘idea’. Confusion can easily result, and historically it did.

Will an image in the mind of Berkeley's geometer, working ‘in his head’, be a thought-vehicle or the object of thought? Will the line be the non-verbal equivalent of the word ‘line’? But the proof in words is not about the word ‘line’, and so the proof in non-verbal form should not be about the non-verbal equivalent of the word; a fortiori, the proof in neither case should be about it and everything of the same sort. We have got to distinguish between a line used as an iconic symbol to represent lines and one of the representative lines which it iconically represents. The word ‘represent’ is surely, as illustrated here, dangerously ambiguous in Berkeley's account.

So this chapter must end with two questions. First, does Berkeley succeed in giving an adequate alternative to Locke's account of general ideas in terms of abstraction? Secondly, if we take the alternative interpretation of Locke's theory which was suggested earlier in this discussion, does it not conform to Berkeley's requirement that generality be a question of function rather than of its internal character? Perhaps it is what Berkeley needs, when it is so interpreted. But generality is a very difficult subject which intensive modern study has still left full of difficulties. Nobody in the time of Locke and Berkeley had the conceptual apparatus to give an adequate account of it.

IMMATERIALISM AND COMMON SENSE

So Berkeley has claimed that matter is unintelligible. We cannot imagine anything like it any more than we can observe it, and the invocation of abstract ideas by means of which we may think of the unimaginable involves absurdity. In so far as we allow matter to be vaguely intelligible as a tool or instrument of God we can see that, thought of as such a tool, it is superfluous: ‘to what end should God take these roundabout methods of effecting things by instruments and machines, which no one can deny might have been effected by the mere command of His will, without all that apparatus?’ (P 61). Moreover, even those who, like Locke, postulate the existence of matter acknowledge that they cannot explain the character of our experience by reference to matter; physics cannot explain why certain pulsations of the air cause heard sound or why in certain physical conditions we see one colour rather than another, or, indeed, why we see any colour at all. The most that Locke can plead is that it is ‘no more impossible to conceive that God should annex such ideas to such motions with which they have no similitude, than that he should annex the idea of pain to the motion of a piece of steel dividing our flesh’; but, says Berkeley ‘how Matter should operate on a Spirit, or procure any idea in it, is what no philosopher will pretend to explain’ (P 50). Berkeley also wished to claim that the only causal agency that was intelligible was of the kind we know in acting ourselves, as when we summon up one idea or another; all we witness in the physical world is succession, regular or irregular, so that the causal agency of matter was yet another unintelligible postulate.

So on all counts it seemed clear to Berkeley that in the ontology of Locke, Newton and their associates matter was indefensible and superfluous. They themselves were ready to admit that matter without God yielded no satisfactory explanation of the world. So Berkeley claims that a complete account of the world can be given in terms of the remainder of the Newtonian ontology—God, finite spirits and their ideas. In a famous formulation of this doctrine Berkeley said that to exist is either to perceive (percipere), which constitutes the existence of spirits, or to be perceived (percipi), which constitutes the existence of the inanimate, of ideas. These perceived ideas are objects of the mind which have no existence independent of the mind. So the world is ultimately spiritual; there is nothing beyond minds and their contents.

At least at first sight, this is an outrageous doctrine, one which is incompatible not only with the presuppositions of science but also with common sense. It seems to deny the existence both of the matter which the physical sciences investigate and of the familiar objects of the everyday world, the chairs and tables, the mountains, plains and rivers which surround us and which we all believe to exist independently of ourselves. It is manifestly contrary to common sense to say that bodies do not exist when we do not observe them.

Berkeley was prepared to agree that if his views were incompatible with common sense and made science impossible they would be unacceptable. What makes his philosophy interesting is that he claims that his ontology is perfectly compatible with both common sense and religious beliefs and that, while it admittedly contradicts a metaphysical presupposition of the scientists, he can give a satisfactory account of the nature and value of the sciences without invoking the hypothesis of matter. Moreover, he develops this claim with such ingenuity that it has always been notoriously difficult to refute him.

We shall start by expounding in this chapter Berkeley's arguments designed to show that he can on his basis account for all that common sense requires. In the next chapter we shall look to see what account he can give of the nature of science on his immaterialist hypothesis. Finally we shall consider what account he can give of God and of finite spirits. Since this topic is being so long postponed for full treatment one or two preliminary words on it are perhaps desirable now. One of the first objections to Berkeley which are likely to occur to his readers is that if matter is unimaginable and if for this reason we can have no idea of it, so also is God unimaginable and even finite spirits. So is not God also an unintelligible absurdity on Berkeley's own principles? This is a difficulty; Berkeley was aware of it and he tried to meet it. So the reader is asked for the present not to question Berkeley's acceptance of God and finite spirits but to consider only whether he can give a satisfactory account of commonsense belief and of science in terms of these spirits, their activities and their ideas.

Berkeley consistently claims that one of the great merits of his position is that it is wholly compatible with common sense. In his private notebooks he frequently reminds himself that he must emphasise this in his published work, as is clear from the following quotations:

All things in the Scripture which side with the vulgar against the learned, side with me also. I side in all things with the mob.

I must be very particular in explaining what is meant by things existing.. when not perceived as well as when perceived; and shew how the vulgar notion agrees with mine.

Mem. To be eternally banishing Metaphysics, etc., and recalling men to common sense.

Mem. That I take notice that I do not fall in with the sceptics … in that I make bodies to exist certainly, which they doubt of.

(C 405, 408, 751,79)

In this claim he never wavers. Thus in the first of the dialogues between Hylas and Philonous he even makes conformity to common sense the touchstone of acceptability. ‘Well then,’ says Philonous, ‘are you content to admit that opinion for true, which upon examination shall appear most agreeable to Common Sense, and remote from Scepticism?’ (L ii 172). To this Hylas agrees.

In claiming that he is in accord with common sense Berkeley clearly cannot mean that the common man has always been aware of and has always accepted a philosophical position identical with his own. He himself says that while we should ‘speak with the vulgar’ we should ‘think with the learned’, and he acknowledges that ‘it is indeed an opinion strangely prevailing among men, that houses, mountains, rivers, and in a word all sensible objects, have an existence, natural or real, distinct from their being perceived by the understanding’ (P 4). This view, as distinct from the philosophical doctrine of matter, is clearly attributed to the common man by Berkeley, and with good reason. Rather, it appears, he is making the following twofold claim: first, that Berkeley himself is committed to no positive claims about the world which are not already assented to by common sense, unlike the followers of the corpuscularian philosophy, who have introduced imperceptible matter; secondly, that given any belief of common sense expressed in the common non-technical language of everyday life, he can provide an alternative statement of that belief in terms of his basic hypothesis, equivalent to the commonsense statement in the sense that it would be impossible to imagine a state of the world which would make one of them true when the other was false. Berkeley, it seems clear, was not claiming that the common man was already aware of these equivalent formulations before Berkeley formulated them. As we shall see, he himself had at times considerable difficulty in deciding what reformulation was most acceptable and accurate. He was claiming to be able to provide what such more modern philosophers as Susan Stebbing and John Wisdom would have called a metaphysical or new-level analysis of the statements of common sense, and, as G. E. Moore constantly said, one may know the truth of a statement and understand it without being able to give its analysis.

Berkeley, then, claims that he can give an adequate analysis of commonsense beliefs in terms of an infinite spirit—God—finite spirits and mind-dependent ideas. He is well aware of the apparent difficulties in this position and lists them in paragraphs 34 onwards in the Principles:

First, then, it will be objected that by the foregoing principles all that is real and substantial in nature is banished out of the world, and instead thereof a chimerical scheme of ideas takes place.

Secondly, it will be objected that there is a great difference betwixt real fire, for instance, and the idea of fire, betwixt dreaming or imagining oneself burnt, and actually being so.

Thirdly, it will be objected that we see things actually without or at a distance from us, and which consequently do not exist in the mind.

Fourthly, it will be objected that from the foregoing principles it follows things are every moment annihilated and created anew.

We can learn to understand Berkeley's position no better than by considering why, in his view, it is immune to these objections.

How, then, if there are but spirits and their ideas, can there be a distinction between a real world and the world of dreams, of fantasy and of illusion? Surely the difference is that the world of dreams or fantasy or illusion is indeed only within our minds, and the real world independent of them? But Berkeley denies this. The difference is not between mind-dependent ideas on the one hand and something outside the mind on the other but between ideas having a certain nature, relation to each other and relation to ourselves, and ideas having a quite different nature, relation to each other and relation to ourselves. First, they differ normally in their nature; the ideas of imagination or dreaming or memory are less strong, lively and distinct than those of sense which constitute the real world. Secondly, they differ in their relation to us; the ideas of imagination and memory are brought before our minds by our own decision, whereas the ideas of sense are independent of our will. I can decide to think about trees, for example, but whether I see a tree when I open my eyes is quite beyond my control. Thirdly, the ideas of sense have ‘a steadiness, order and coherence, and are not excited at random, as those which are the effects of human wills often are, but in a regular train or series’ (P 30). In dream, imagination or illusion anything can happen, but in the real world the train of ideas is determined by the laws of nature.

Berkeley cannot claim, nor does he need to claim, that all these three types of difference are always observable; unlike imagination, dream and illusion have a content over which we have no control, for example. But he does want to claim that at least one of these differences will always be present and that awareness of them is the way, the only way, in which reality can be distinguished from illusion or other forms of unreality. How, for example, does Macbeth determine that the dagger that he sees is unreal? Here are Shakespeare's words that he puts into the mouth of Macbeth:

Is this a dagger which I see before me,
The handle toward my hand? Come, let me clutch
          thee!
I have thee not, and yet I see thee still.
Art thou not, fatal vision, sensible
To feeling as to sight? or art thou but
A dagger of the mind, a false creation
Proceeding from the heat-oppressed brain?

Clearly Macbeth determines that the dagger is an illusion, not by finding (how could he?) that it has a status independent of his mind, but by finding that his ideas of sight and touch lack the coherence which would be necessary for the experience to count as experience of reality. Berkeley wants to claim that he can make the distinctions that common sense makes in his own terms and that in fact common sense makes the distinctions in precisely the ways that he describes. So the distinction between reality and the merely subjective is not just detected by the features Berkeley has mentioned, acting as clues; certain trains of ideas count as experience of reality while others do not, ‘by which is meant’, Berkeley claims, ‘that they are more affecting, orderly and distinct, and they are not fictions of the mind perceiving them’ (P 36). To be real and to have these features are, for Berkeley, one and the same thing. Berkeley is giving a reductionist analysis, not giving us hints on how to check up on our experience.

This is an appropriate place to pause and make for the first, but not the last, time the point that there is an important distinction to be made between Berkeley's analysis of commonsense beliefs in terms of ideas and his metaphysics. Berkeley certainly thinks that the metaphysical explanation of why some ideas have coherence, liveliness and independence of our wills is because God, the infinite spirit, causes them. Thus metaphysically reality is to be explained in terms of the activity of God; but the analysis of what we mean when we ascribe reality to something does not contain, as we have seen, any reference to God. This distinction is often ignored and Berkeley is frequently represented as bringing God into his analysis when he does not. Berkeley gives on the whole what may be counted as a positivistic analysis of commonsense beliefs, a fact which is perfectly compatible with the fact that he offers a theocentric metaphysic to explain why our experience has the character that it has. We should not underestimate the importance of God in Berkeley's thought, but we also should not misrepresent Berkeley as holding that talk about chairs and tables is to be analysed as talk about the deity.

Berkeley's treatment of the third objection, that what we see is at a distance from us and therefore cannot be within our minds, is the occasion for us to consider very briefly Berkeley's A New Theory of Vision, in which he is concerned to state how by the sense of sight we are able to judge distance. Berkeley was clear that distance had to be judged and could not be seen: ‘For, distance being a line directed endwise to the eye, it projects only one point in the fund of the eye, which point remains invariably the same, whether the distance be longer or shorter’ (V 2). If we judge distance it is not, he holds, as some have thought, in virtue of the difference of the angles of the eye in binocular vision, for we are not aware of the angles of the eyes when we judge distance, nor do we need to be. Nor do we ultimately judge from sensations in the eye, such as, perhaps, a sense of strain when we look at a thing close up. For we rather learn from experience that we experience the strain when the object is close and so must have another means of determining that the object is close.

Berkeley's own solution to this problem is in terms of a correlation between sight and touch. In the New Theory of Vision he gives this solution as though by touch we were aware of objects outside us in a real physical space, since, as he says in the Principles, he is there concerned with vision alone and not with the fundamentals of epistemology. More accurately, we should speak of a correlation of ideas of touch with ideas of vision. Though we speak of being aware of shape by both sight and touch, Berkeley maintains that the visual idea of, say, a rectangle and a tactual idea of a rectangle are quite distinct and we learn only from experience that given the one idea the other will accompany it. A great controversy raged about a question which the Dublin lawyer Molyneux set in a letter to his friend Locke:

Suppose a man born blind and now adult, and taught by his touch to distinguish between a cube and a sphere of the same metal, and nighly of the same bigness, so as to tell, when he felt one and the other, which is the cube, which the sphere. Suppose then the cube and sphere placed on a table, and the blind man to be made to see; query, whether by his sight, before he touched them, he could now distinguish and tell which is the globe, which the cube?

To this question Berkeley, like Locke and Molyneux himself, gave the unequivocal answer ‘No.’ Now if we suppose that it will take more time, in similar conditions, to get the tactual idea of something round than to get the tactual idea of something rectangular, then the visual idea of the rectangular will be regarded as more distant than the visual idea of the round. This correlation of sight and touch is fundamental, in Berkeley's view. But on this basis we learn to judge distance in terms of the visual appearances of things; thus if one thing having the general look of a ping-pong ball looks larger than something that looks like a tennis ball we shall judge the ping-pong ball to be much nearer.

Thus Berkeley claims that the judgement of distance does not require us to posit an external space. Furthermore the appearance of relative distance from us can be exhibited by objects of imagination, which no one supposes to be in some real space, just as much as by the objects of the senses.

What then of the fourth difficulty that Berkeley mentioned, about the continued existence of ordinary physical bodies like chairs and tables and mountains and rivers? This difficulty has been expressed in a well-known limerick:

There was a young man who said ‘God
Must think it exceedingly odd
If he finds that this tree
Continues to be
When there's no one about in the Quad.’

The author of this limerick, Monsignor Ronald Knox, also produced another limerick which purports to offer Berkeley's solution to the problem and runs as follows:

Dear Sir:
          Your astonishment's odd;
I am always about in the Quad.
And that's why the tree
Will continue to be,
Since observed by
                    Yours faithfully,
                              GOD.

This view, which makes Berkeley locate the continuous existence of bodies in their eternal and continuous perception by God, is, I suppose, the orthodox interpretation of Berkeley on this topic. But it is at best a misleading and inadequate account of Berkeley's view. Certainly Berkeley held that metaphysically the existence of unobserved trees depended on the activity of God; but then he held that metaphysically the existence of observed trees and finite observers depended on the activity of God. But, as has already been said, we must distinguish between Berkeley's metaphysical view that all existence was dependent on God and his analytical views about the meaning of statements. There is very little, if any, evidence that Berkeley believed that, when we assert the existence of some unobserved object, what we mean is something about the observations of God; there is, on the contrary, a lot of evidence that he did not. So the rest of this chapter will be devoted to the discovery of what his views on this topic were. How could Berkeley, the immaterialist, allow the commonsense assumption that bodies exist continuously, whether perceived or unperceived?

This inquiry is both instructive and amusing. For Berkeley's great insight had been the negative one of the non-existence of matter. His initial claim that this insight was compatible with common sense was an act of faith, for he had at that stage no idea what account he should give of bodies. He tried out many theories, some plausible and ingenious, some very crude and unconvincing, before giving his final answer to the question in the third and last of the dialogues between Hylas and Philonous.

We find a number of mutually incompatible and unplausible suggestions about the nature of body in the private notebooks. Here, for example is a particularly ill-conceived one:

We see the house itself, the church itself; it being an idea and nothing more. The house itself, the church itself, is an idea, i.e. an object—immediate object—of thought.

(C 427, 427a)

But, clearly, to identify an object with some single idea will not do. For if I look three times at what common sense would call a house I have, in Berkeley's terminology, three separate ideas; if we identify the house with the ideas then I shall have seen not one but three houses, each of which will have only a fleeting existence.

Here is another suggestion from the same notebooks, one almost as unfortunate as the one just considered:

You ask me whether the books are in the study now, when no one is there to see them? I answer, Yes. You ask me, Are we not in the wrong for imagining things to exist when they are not actually perceived by the senses? I answer, No. The existence of our ideas consists in being perceiv'd, imagin'd, thought on. Whenever they are mentioned or discours'd of they are imagin'd & thought on. Therefore you can at no time ask me whether they exist or no, but by reason of that very question they must necessarily exist.

(C 472)

The ghost of this argument continued to haunt Berkeley in later writing; but he had no sooner written it down than he began to feel uneasy about it. The next note reads as follows:

But, say you, a chimaera does exist? I answer, it doth in one sense, i.e. it is imagin'd. But it must be well noted that existence is vulgarly restrain'd to actual perception, and that I use the word existence in a larger sense than ordinary.

(C 473)

This is, at best, a very half-hearted defence from one who claims to side with the vulgar against the learned.

But, on the whole, these early notebooks incline to phenomenalism. Bodies are to be what they were for Mill in his classical formulation of phenomenalism, permanent possibilities of sensation. Just how honestly one could claim that phenomenalism coincided with common sense did worry Berkeley, as it has worried many other phenomenalists, but he tried to stifle these doubts, as the following quotation shows:

Mem. To allow existence to colours in the dark … but not an actual existence. 'Tis prudent to correct men's mistakes without altering their language. This makes truth glide into their souls insensibly. Colours in the dark do exist really, i.e. were there light, or as soon as light comes, we shall see them, provided we open our eyes; and that whether we will or no.

(C 185, 185a)

What he says about colours he also says about bodies:

Bodies &c. do exist whether we think of 'em or no, they being taken in a twofold sense—

1. Collections of thoughts.

2. Collections of powers to cause these thoughts.

(C 282)

If an explanation of what he means here by powers is needed, he immediately gives it, again in a purely positivistic vein:

Bodies taken for powers do exist when not perceived; but this existence is not actual. When I say a power exists, no more is meant than that if in the light I open my eyes, and look that way, I shall see it, i.e. the body &c.

(C 293a)

Even when Berkeley was writing the Principles he had still not really made up his mind what to say about the continued existence of bodies. Here we have a thoroughly phenomenalistic story:

The table I write on I say exists; that is, I see and feel it; and if I were out of my study I should say it existed; meaning thereby that if I were in my study I might perceive it, or that some other spirit actually does perceive it.

(P 3)

Here ‘some other spirit’ is written without the capitals which Berkeley always uses when referring to God—it surely means merely ‘somebody else’ and not God specifically. But over the page in the sixth paragraph we get the one case that I have noticed of Berkeley putting forward the view assigned to him by Ronald Knox in the limerick quoted above. Having reaffirmed that nothing exists outside a mind, he says of bodies:

Consequently so long as they are not actually perceived by me, or do not exist in my mind, or of any other created spirit, they must either have no existence at all, or else subsist in the mind of some Eternal Spirit.

(P 60)

Here we do have the capital E and capital S. It is unambiguously stated that, when not perceived by finite spirits, bodies, if they exist at all, must be in the mind of God; and this obviously and flatly contradicts the phenomenalistic story he has already given, unless we can claim that Berkeley has without warning moved from analysis to metaphysics.

There are two principal objections to the analysis of statements about physical objects in terms of the ideas in the mind of God, independently of doubts about God's existence. First, it is surely clear that the statement that God does not exist but that there are unobserved physical objects is not self-contradictory, even if it is false; when we speak of physical objects we do not thereby say anything about God. The second objection arises from the Christian theology which Berkeley accepted and needs explaining at a little more length.

It is a commonplace of theology that God is eternal and unchanging. The notion of an unchanging person may be difficult, but for the present all we need to understand by it is that God does not in time come to have ideas or thoughts which he had not had previously. It is another commonplace of theology, based theologically on the first chapter of the book of Genesis but no doubt confirmable by paleontological investigation, that God created the inanimate—the earth, the sea and the heavenly bodies—before he created on earth any living creatures. But if the existence of physical bodies consists in their being objects in the mind of God, then physical objects have existed from all eternity and were not created at all; and the existence of physical objects can also not consist in their being actually ideas of finite spirits, since they were created before finite spirits.

So it would seem that if we accept the doctrine of creation, as Berkeley certainly did, the permanent existence of physical bodies cannot consist in their being actual objects of thought or perception either by an infinite or by finite spirits. It seems that Berkeley did not become fully aware of this problem until he wrote the three dialogues between Hylas and Philonous; there he meets it squarely in the third dialogue, and in doing so comes up with what, given his basic positions, philosophical and theological, must be the most satisfactory answer possible. The solution, stated baldly, is this; God has, indeed, all his ideas from all eternity, and this in no way constitutes the existence of physical bodies. When God created the physical universe his creation consisted in his ordaining that from that time forward ideas of the physical should be available to finite spirits if any; this ordinance was valid even before there were finite spirits, just as criminal laws are valid even if there are no criminals. So the existence of physical objects consists in the relevant ideas being available. Physical objects are permanent possibilities of sensation—that is the analysis—and this possibility exists because God ordains that it should—that is the metaphysical explanation of this possibility.

Thus the phenomenalist analysis prevails in the end, and in a form superior to earlier versions. Berkeley's phenomenalist analysis at the beginning of the Principles, already quoted, starts: ‘The table I write on I say exists; that is, I see and feel it’ (P 3). He goes on to allow for its existence when he is not in his study; but it surely is clear that we never (never) assert that we are actually observing an object when we assert its existence, even if as a matter of fact we are observing it. So as an analysis the account in terms of the possibility of ideas alone is clearly preferable and seems to be about the best version of phenomenalism that it is possible to formulate. On the whole it is true that the Three Dialogues are largely a more popular version of ideas that are more accurately and fully developed in the Principles. But when Berkeley, in the third dialogue, wrote: ‘things, with regard to us, may properly be said to begin their existence, or be created, when God decreed they should become perceptible to intelligent creatures, in that order and manner which He then established, and we now call the laws of nature’ (L ii 253), he was surely making an advance on what he had previously written.

We may close this chapter on Berkeley's view of the physical world with a look at a famous, or perhaps notorious, attempt by Berkeley to prove that it is impossible to conceive of bodies existing independently of minds. He repeats it more than once in very similar language and I quote it now from paragraph 23 of the Principles:

But, say you, surely there is nothing easier than for me to imagine trees, for instance, in a park, or books existing in a closet, and nobody by to perceive them. I answer, you may so, there is no difficulty in it. But what is all this, I beseech you, more than framing in your mind certain ideas which you call books and trees, and at the same time omitting to frame the idea of any one that may perceive them? But do you not yourself perceive or think of them all the while? This therefore is nothing to the purpose; it only shews you have the power of imagining, or forming ideas in your mind; but it does not shew that you can conceive it possible the objects of your thought may exist without the mind. To make this out, it is necessary that you conceive them existing unconceived or unthought of; which is a manifest repugnancy.

This is surely fallacious; certainly it is not possible that a body be conceived of and at the same time both exist and not be conceived of, just as God cannot be conceived of and at the same time exist and not be conceived of, and just as a chair cannot be kicked and at the same time exist unkicked. If one is thinking of a chair then it does not exist unthought of. But this has no tendency to show that the concept of an unthought-of chair has any logical deficiencies.

So Berkeley cannot prove that it is logically necessary to concede that body is ultimately mind-dependent. But he can surely claim to have made a very plausible effort to show that his view gives common sense everything it needs. Our experience of the world, as it actually occurs, cannot be different from how it would be on a Berkeleian view, since our experience of it just is the sequence of ideas of which Berkeley speaks. Any apparent discrepancy will not disprove Berkeley's view in principle, but will merely show that it has been mis-stated. We might perhaps believe that we could catch Berkeley out by taking a photograph of the interior of a room by remote control when there is nobody in the room. But Berkeley is perfectly able to accept that we shall have the requisite series of ideas called looking at a photograph of an empty room in the given circumstances; that is part of the order of nature. Berkeley can well ask what more we want. Occam's razor tells us not to multiply entities beyond necessity, and all that Berkeley is taking away from us, he claims, is a hidden machinery which is a totally superfluous hypothesis, in so far as it is at all intelligible.

If we do imagine that Berkeley speaks to us in some such words as those we may be tempted to reply as follows: even if it be true that the hypothesis is superfluous at the level of everyday life, in science it is not superfluous. So far from the material world being a something we know not what, science is continually discovering more and more about it and, in virtue of our knowledge of its nature, we are able to predict and engineer new marvels which would otherwise be unknown. How could man make the atom bomb if there were no atoms, and how could he predict it from a superfluous and empty hypothesis? We might be tempted to say this because it is a very natural, not to say obvious, thing to say at this stage. Being natural and obvious, the thought has occurred to Berkeley also. Our objection would not have caught him unprepared and he answered it in advance to his own satisfaction.

Note on abbreviations

The following abbreviations are used in references to Berkeley's works:

A: Alciphron

C: Philosophical Commentaries

L: The Works of George Berkeley, Bishop of Clogne, edited by A. A. Luce and T. E. Jessop, 9 vols. Nelson and Sons, Edinburgh, 1948-57

M: De motu

O: Passive Obedience

P: Principles of Human Knowledge

S: Siris

V: A New Theory of Vision

References to L are by volume and page, references to A, M, O, P, S and V by paragraph. References to C follow the numbering (not Berkeley's) adopted in L i.