Berkeley's Argument for Immaterialism - page 2

Written by A.C. Grayling.

Ideas, Perception and Mind

A key concept in the foregoing is that of ideas. Berkeley uses "idea" to mean "any immediate object of sense or understanding", but as already noted he is careful to distinguish this from what, in the second paragraph of P, he had described as "such as are perceived by attending to the passions and operations of the mind", which he later calls notions. The distinction is as follows. Ideas are always sensory; they are either the content of states of sensory awareness, or the copies of these in memory and imagination. Notions on the other hand are concepts of spirit – of self, mind and God – and have a more complex origin. As regards self-knowledge, notions originate in immediate intuition; as regards other minds, in interpretation; and as regards God, in "reflexion and reasoning" (P42, 140-2, 3D232).

Two features of ideas are crucial for Berkeley: their inertness and their mind-dependence. They are the latter simply in virtue of being ideas. Their being the former is a more intricate matter. Anticipating Hume, Berkeley argues that there are no necessary connections between ideas; they are individual entities "with no power or agency included in them. So that one idea or object of thought cannot produce or make an alteration in another" (P25). We verify this by introspecting, which reveals, says Berkeley, that "there is nothing in [ideas] but what we perceive," and we perceive no power or activity in them (ibid.). We have a "continual succession" of ideas, some arising and others disappearing; but because they are causally inert, they are not themselves responsible for these changes, so there must be some other cause of them (P26). The only candidate remaining for this role is spirit or mind. Since my mind is causally responsible for very few ideas and their changes, there must be "some other spirit that produces them" (P29).

Berkeley gives the name perception to any way of having ideas and notions before the mind, in sensing, conceiving, imagining, remembering, reasoning, and the rest. It is accordingly a generic term, and is not restricted to sensory perception alone. "Perceiving" denotes a causal relation: minds perceive ideas either by causing them (as when finite minds imagine or dream, and as when the infinite mind wills the existence of the universe) or by being causally affected by them (as when finite minds receive the ideas caused by God, = encounter the physical world).

Any inference to the nature of the spirit that is causally responsible for ideas and their changes must start from the nature of those ideas and their changes. "The ideas of sense," says Berkeley, again anticipating Hume, "are more strong, lively, and distinct than those of imagination; they have likewise 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" (P30).

These "set rules or methods" we call "Laws of Nature; and these we learn by experience, which teaches us that such and such ideas are attended with such and such other ideas, in the ordinary course of things" (ibid.). From this Berkeley concludes that God, the "Author of Nature", is the ultimate source of ideas and their connections.

From this in turn it follows that although everything that exists is mind-dependent, it is not dependent on particular or finite minds, but has an objective source and structure, namely, the eternal, ubiquitous and law-like perceiving of an infinite mind. This is the sense in which Berkeley is a realist; the world exists independently of the thought and experience of finite minds (2D166-7)  -  which explains what he means by claiming to defend common sense, for common sense holds that grass is green and the sky is blue whether or not any of us happen to be looking at either, whereas Locke and the corpuscularians held otherwise - grass has powers to make us see green, but it is not itself green; indeed, on the Lockean view the world is colourless, odourless and silent until perceived, when it produces in the perceiver visual, olfactory and auditory experiences. But for Berkeley the world is just as we perceive it to be even when we are not perceiving it, because it is always and everywhere perceived by the infinite mind of a deity.

The deity perceives the universe by thinking it, that is, causing it to exist by conceiving it. In a letter to the American Dr Samuel Johnson Berkeley remarks that his view differs only verbally from the theological doctrine that God maintains the universe in existence by an act of continual creation. So the ideas which constitute the world are caused by the deity, and appear in our consciousnesses as the effects of his causal activity: this is the metaphysical way (level 3) of describing what, in ordinary terminology, we describe as seeing trees, tasting ice-cream, and so forth. The latter way of describing the facts is not incorrect; Berkeley's argument is that the ordinary and the metaphysical ways of describing reality are alternative descriptions of the same thing.

A significant feature of this account is its view of causality. Locke had argued that the empirical basis for our concept of causality comes from our own felt powers as agents, able to initiate and intervene in trains of events in the world. This sense of our own efficacy we "project" onto the world to explain chains of events in it, imputing to events we describe as "causes" an agency or power on analogy with our own. For Berkeley the projective move is empirically ungrounded. We indeed have experience of causal agency as spirits, which are the only active things we know. But although it is a convenience to impute causal agency to things (ideas) in our ordinary way of talking, they are inert, and apparent causal connections between them are ultimately owed to the regular, consistent, law like causal activity of God.

The Argument's Resilience to Objections

It is obvious that Berkeley's theory rests upon a vital and very debatable assumption, borrowed unquestioningly from Locke who equally unquestioningly borrowed it from the Cartesians, namely, that the place to begin philosophical enquiry is among the private data of individual consciousness, that is, among the ideas constituting an individual's experience. If one accepts this Cartesian super-premiss (a large "if") the early steps of Berkeley's argument appears persuasive, as may be seen by considering a proposed objection to it, namely, that it commits the elementary error of identifying sensible qualities and sensory ideas; for  -  says the objection  -  there is a large difference between "the table is brown" and "the table looks brown to me", because the truth-conditions of the two statements differ. The table could be brown without it seeming so to me, and vice versa; so Berkeley's argument collapses.

But this argument begs the question against Berkeley by assuming that claims about what qualities an object possesses are independent of claims about how they can be known to possess them; which amounts to the claim that there are observation-independent facts about the qualities of objects which can be stated without any reference to experience of them. But this claim is exactly what Berkeley rejects, on the grounds that any characterisation of a sensible quality has to make essential reference to how it appears to some actual or possible perceiver. How, he asks, does one explain redness, smoothness and other sensible qualities independently of how they appear? So the objection fails by premissing a seems-is distinction which is precisely what Berkeley opposes on the grounds that it leads to scepticism.

To deny that there is a seems-is distinction is just another way of asserting that sensible objects (things in the world) are collections of sensible qualities, and hence of ideas. So Berkeley takes the contrast he wishes to resist to be one between (a) sensible objects, which as collections of sensible qualities are what is immediately perceived, and (b) objects existing independently of perception but causing it. This is not the same contrast as (c) sense data in the sense of uninterpreted contents of sensory states, and (a) sensible objects. It is important to note this because for Berkeley what is immediately present in experience is the sensible object, not some mediating representation (or collection of representations) different from the object. We do not, he says, infer from colour patches and other sensory data to the existence, in a world beyond them, of books and trees; what we see (and touch etc.) are, immediately, books and trees.

This however prompts another objection, this time that Berkeley is having things both ways: he says that we immediately perceive such familiar objects of sense-experience as books and trees, while at the same time saying that what we immediately perceive are colours and textures. To see what is involved here, consider an argument advanced in more recent philosophy. This says that books and trees are interpretations of, or inferences from, the sensory data of experience, and that in speaking of books rather than colour-patches we are going beyond what talk of colour patches strictly licenses. This is because we take physical objects to exist independently of particular perceivings of them, to be publicly available to more than one perceiver at a time, and so on  -  none of which is true of the sensory ideas from which they are inferred. So we have to keep (a) and (c) strictly separate.

Berkeley can be defended against this objection by appealing to the distinction of levels. At level 1 we immediately perceive colours and textures, while at level 2 we immediately perceive books and trees. The latter consist wholly of the former, and it is only if one disregards the distinction of levels that one might fall into the mistake of thinking that when one perceives a smooth red book, one perceives redness and smoothness and a book, as if the book were something additional to the sensible qualities constituting it. Just such a view is forced by the materialist view, in which something inaccessible to sensory awareness constitutes the underlying causal origin of the sensible qualities we perceive.

Some critics object that having thus argued that all perception is immediate, Berkeley promptly proceeds to admit a species of mediate perception by inference or "suggestion". The passage cited is the one where Berkeley says, "when I hear a coach drive along the streets, immediately I perceive only the sound; but from the experience I have had that such a sound is connected with a coach, I am said to hear the coach" (1D204). This might count as a case of mediate perception if Berkeley did not immediately add, "It is nevertheless evident, that in truth and strictness, nothing can be heard but sound: and the coach is not then properly perceived by sense, but suggested from experience". The same applies to our practice of saying elliptically that one sees that the poker is hot; again, one does not see heat, one sees that something is hot, that is, one infers on the basis of experience that when something looks like that, it will feel a certain way if you touch it. These are not cases of mediate perception, but of experience-based inference, to which Berkeley gives the name 'suggestion": the ideas of one sense suggest the ideas of another.

The foregoing shows that as long as certain of Berkeley's premisses are accepted, and as long as discussion of the main plank of his views (the notion of God and his metaphysical activity) is deferred, his views are resilient to objection. If we reject the Cartesian super-premiss  -  that the place to start is the data of individual experience  -  his views are not so resilient.

These remarks touch upon one set of objections to Berkeley's views. Others, as remarked, more threatening to his position, concern its underpinning, namely, the infinite mind to which a central metaphysical role is allotted. This is discussed below.

Matter and Materialism

The concept of matter is redundant, Berkeley's argument purports to demonstrate, because everything required to explain the world and experience of it is available in recognising that minds and ideas are all there can be. But Berkeley adds to this argument-by-exclusion a set of positive anti-materialist considerations.

An important argument for materialism is that use of a concept of matter explains much in science. Berkeley summarises the view thus: "there have been a great many things explained by matter and motion: take away these, and you destroy the whole corpuscular philosophy, and undermine those mechanical principles which have been applied with such success to account for the phenomena. In short, whatever advances have been made ... in the study of nature, do all proceed on the supposition that corporeal substance or matter doth really exist" (P50). Berkeley's reply is that science's explanatory power and practical utility neither entail the truth of, nor depend upon, the materialist hypothesis, for these can equally if not better (because more economically) be explained in instrumentalist terms. Instrumentalism is the view that scientific theories are tools, and as such are not candidates for assessment as true or false, but rather as more or less useful. One does not ask whether a gardening utensil such as a spade is true, but whether it does its intended job effectively  -  and not merely effectively, but, as required by Ockham's Razor, as simply and economically as possible.

Berkeley expressed his early version of instrumentalism as a "doctrine of signs", in which the regularity and order among our ideas reflect the steady will of God, which is so reliable that we can represent the connections thus observed as laws. He writes, "the steady, consistent methods of Nature, may not unfitly be styled the language of its Author, whereby he ... directs us how to act for the convenience and felicity of life" (P109). Science is thus a convenient summary, for sublunary purposes, of what at the metaphysical level of explanation would be described in terms of the activity of infinite spirit.

This is a rejoinder to an attempted "appeal to the best explanation" on behalf of the materialist hypothesis. It is at the same time a rejoinder to a closely allied argument, an "appeal to the simplest explanation." This says that postulating the existence of matter simplifies the account we give of the world. The rejoinder consists in the same slash of Ockham's Razor; it is that since experience could be exactly as it is without matter existing independently of it, the materialist hypothesis is not the simplest explanation after all.

But the key point for Berkeley is that whatever else matter is, by definition (a) it is non-mental, and as such cannot be the support of qualities, because qualities are ideas, and ideas can only exist in a thinking substance; and (b) it is inert, that is, causally inactive, and so cannot produce change, motion, or ideas.

For Locke and others among Berkeley's predecessors the concept of primary qualities was important because, they held, experience of them puts us most closely in touch with independent reality. Berkeley rejects their view on the ground, already mentioned, that "nothing can be like an idea but an idea" (P8). Materialists hold that primary qualities are "resemblances" of "things which exist without the mind" (P9); but since primary qualities are ideas, and only ideas can resemble ideas, it follows that "neither they nor their archetypes can be in an unperceiving substance" (P9). Moreover, as also already noted, the primary-secondary quality distinction, understood in terms of a supposed difference between the way each kind of quality relates to mind, involves a specious abstraction of one kind from the other; for since one cannot conceive such primary qualities as motion or number apart from such secondary qualities as colour, both are on a par in the way they relate to mind, viz. by being essentially dependent upon mind for their existence as ideas.

Some of Berkeley's critics think he failed to separate the question of material substance from that of the primary-secondary quality distinction, since one can reject materialism while retaining the distinction. But this in fact is what Berkeley does, for he does not deny that there is a distinction between primary and secondary qualities - he recognises that the former are available to more than one sense at a time, the latter available to one sense only; that the former are measurable, the latter not (or not so straightforwardly); and so on - but he points out that in the crucial respect of their relation to mind, they are on a par in both being sensible and hence mind-dependent.

The Mind-Idea Relation

One charge levelled at Berkeley is that his account of the crucial relation between minds and ideas is contradictory or at very least confused. At P2 he says that the mind "is a thing entirely distinct from [ideas], wherein they exist, or, which is the same thing, whereby they are perceived" (my emphases). At P5 he adds that it is not possible to conceive "any sensible thing or object distinct from the sensation or perception of it," and in the same paragraph he remarks, "is it possible to separate, even in thought, my ideas from perception? For my part, I might as easily divide a thing from itself." These assertions appear to commit him to three principles which are together inconsistent, but which each play an important part in his argument.

The three principles have been called the Distinction Principle, asserting that minds and ideas are distinct from one another (P2, 27, 80, 142), the Inherence Principle, asserting that ideas exist only in the mind (P2, 3 and passim), and the Identity Principle, asserting that ideas are not distinct from perceivings of them (P5, 1D195 ff). The second and third are consistent; the first and third appear to contradict each other; and the relation between the first two at very least demands explanation.

Most critics think that the best solution is to abandon the Distinction Principle. One reason is that it appears to commit Berkeley to an act-object analysis of perception, whereas the Identity Principle commits him to an adverbial analysis. The first describes perceiving as an act of mind directed upon an object, rather as the beam of a torch is directed upon something we wish to illuminate; the object is independent of the act, which can be repeated with different objects (the act of looking, like the beam of the torch, can have as its successive objects a book, a cat, a desk). The adverbial analysis has it that perception is a modification of the mind, so that, for example, to see a cat is to have one's mind shaped or modified into a "catly-perceiving state". Here there is one event - the modification of one's mental states in a certain way - whereas on the act-object analysis there is the mental act and something independent of it, viz. its object. Since this analysis demands the independence of the objects of perception, which on Berkeley's theory are ideas and hence incapable of independence from mind, the Distinction Principle seems to be the obvious candidate for rejection.

But the principle is crucial to Berkeley; the very plan of P depends on it: "Human knowledge [reduces] to two heads, that of ideas and that of spirits" (P86). And this is no surprise, since if the principle were rejected, minds would be identical with their ideas, but this cannot be, for Berkeley insists on the differences: minds are active, ideas inert; ideas are dependent entities, minds substantial. In C Berkeley had considered and rejected the notion (yet again anticipating Hume) that minds are just bundles of ideas, on the good ground that "a colour cannot perceive a sound, nor a sound a colour ... therefore I am one individual principle, distinct from colour and sound; and, for the same reason, from all other sensible things and inert ideas" (3D234).

The Distinction Principle, therefore cannot be abandoned. But neither can the others; the Inherence Principle, after all, is simply a version of esse est percipi, and the Identity Principle follows from the attack on abstraction, which tells us that we cannot abstract ideas from perception of them. Is there a solution?

There is; and it is to be found in recognising that the expressions "Inherence" and "Identity" are misleading. For Berkeley does not hold that ideas "inhere" in the mind, as attributes are said to inhere in substance, nor that ideas and perception of them are identical. I take each point in turn.

The "Inherence" Principle states that "ideas exist in the mind". The formula "in the mind", as already noted, is to be understood as "with essential reference to mind", in the sense that the existence of an idea is dependent upon its being perceived - actually, not just possibly, perceived: recall that in Berkeley's theory everything that exists is actual. The sense of "dependence" here is that in which (to adapt an obstetric example of Plato's) an embryo is dependent on a womb: it exists in it, and cannot exist without it, but it is nevertheless distinct from it. "Inherence" is an adverbial notion, whereas Berkeley holds that ideas and minds stand in the internal causal relation denoted by the generic concept of perception: "there can be no substratum of ... qualities but spirit, in which they exist, not by way of mode or property, but as a thing perceived in that which perceives it" (3D237).

As for the "Identity" Principle, it is a straightforward mistake to construe Berkeley's anti-abstractionist view - namely, that any account of ideas cannot be abstracted from an account of perception - as amounting to an assertion of the identity of ideas with perception of them. The assertion that one cannot "conceive apart" any "sensible thing or object distinct from the perception of it" (P5) is not a claim that these are numerically the same thing. Consider an example: bread and the process by which it is baked are internally related; there cannot be one without the other; but bread is not numerically identical with the baking of it. The same kind of internality characterises the relation of minds and ideas. As one would expect, that merely iterates the point made about the "Inherence" Principle.

The question of the mind-idea relation is important because it is the only major threat to the internal coherence of Berkeley's theory. These comments show that there is not after all such a threat. As before, it remains that the Cartesian basis of the project, and its linchpin metaphysical thesis that an infinite mind perceives everything always, are the two real problems with Berkeley's theory. It is time to consider the second of these.

Spirit as Substance

Berkeley took his arguments to amount to a new and powerful argument for the existence of a God. As such, as noted, they are a contribution to natural theology; nothing turns on revelation or traditional conceptions of deity, beyond that such a being has to be infinite and omnipotent. Indeed Berkeley's arguments require no more than a metaphysical god thus conceived. Whether such a being is a person, or whether it is interested in what it has created, is neither here nor there, so long as it fulfils its function of making the universe exist.

The nub of Berkeley's argument for God is that since everything that exists is either mind or ideas, and since finite minds, even in concert, could not perceive all the ideas that constitute the universe, there must be an infinite mind which perceives everything always and thereby keeps it in being.

The classic statement of the argument occurs in the second of the Three Dialogues (2D212-4). From the proposition "that sensible things cannot exist otherwise than in a mind or spirit" Berkeley concludes, "not that they have no real existence, but that seeing they depend not on my thought, and have an existence distinct from being perceived by me, there must be some other mind wherein they exist." This conclusion is a weaker one than that there is a single infinite mind which perceives everything always; it establishes no more than that there is "some other mind" - who might for all we know be the next door neighbour. But in the very next sentence Berkeley adds, "As sure therefore as the sensible world really exists, so sure is there an infinite omnipresent spirit who contains and supports it." This is quite a leap. The missing step is provided later (2D215): "I perceive numberless ideas; and by an act of my Will can form a great variety of them, and raise them up in my imagination: though it must be confessed, these creatures of the fancy are not altogether so distinct, so strong, vivid, permanent, as those perceived by my senses, which latter are called real things. From all which I conclude, there is a mind which affects me every moment with all the sensible impressions I perceive. And from the variety, order and manner of these, I conclude the Author of them to be wise, powerful, and good beyond comprehension." This "Author" Berkeley a few lines later describes as "God the Supreme and Universal Cause of all things." The missing step is, accordingly, a version of the teleological argument for the existence of a God.

The argument in fact has two stages. The first argues that things are causally dependent on mind for their existence, and therefore, since I cannot think of everything always, there must be mental activity elsewhere carrying out the task. The second stage says that one can infer the character of that mind by inspecting the nature of its ideas: since the universe is so huge, beautiful, intricate, and so on, it must be a "wise, powerful," and so on, mind.

The first thing to note is the inadequacy of the teleological argument here co-opted as the second stage. The appearance of design, purpose or beauty in the universe does not entail that it was designed; and even if it did entail this, it does not thereby entail that it was designed by a single mind, or an infinite mind, or a good mind. (What if – regarding this last point – we reflected on the cruelty in nature, and the disease, waste, pain and other evils abundant there? What picture of a creating mind would this suggest?) In any case there are more economical ways to explain the teleological appearance of the universe, the best being evolutionary theory.

What of the first stage? The most it establishes is the conclusion that what exists can do so only in relation to mind. The relation in question needs to be explained; Berkeley is committed to saying that it is a causal relation, but it is exactly this which pushes him to the unpersuasive second stage of the argument for an infinite mind. An alternative resource might be to say that there is no account to be given of the world which does not make essential reference to facts about thought or experience of it, and this might furnish the starting point for views like Kant's or those of certain contemporary "anti-realists".

Although Berkeley does not need the God of traditional theology but only a metaphysical being causally competent for its task, his employment of teleological considerations mixes tradition with metaphysics to the injury of the latter. There is certainly no shadow of an argument why the mental activity to which the existence of the whole universe is referred has to be a single mind or an infinite one. A committee of finite minds might seem an even less palatable option, but nothing in the argument excludes it.

Upon inspection, accordingly, the argument for the metaphysical linchpin of Berkeley's theory does not work. As remarked, if this does not entail the collapse of the project, it will be because there is some other way of substantiating the idea that what exists stands in an internal relation to thought or experience of it. On that score, philosophy is not without resources.

Concluding Remarks

Other points in Berkeley's views repay further examination, for example his concept of conceivability, his finitary realism, the character of his idealism, his views of time, and–as just suggested–the metaphysical implications of his arguments considered independently of their theistic basis. But enough has been said to suggest reasons for his influence on later thinkers, not least among them the Phenomenalists and the Logical Positivists. It is in particular both interesting and philosophically important to understand why phenomenalism, as a version of Berkeley's theory in which the bare concept of "possibilities of perception" ("possibilia") has been substituted for the concept of a deity, is arguably less cogent than the original. From the point of view of acceptability there is little to choose between a metaphysics of possibilia and a theological basis for the universe, neither of which is especially attractive. But at least in Berkeley's thesis, everything that exists is actual.