Philosophy 23 Lecture Notes: Berkeley

Berkeley is best known for espousing the "immaterialist hypothesis" in contrast to Locke's "corpuscularian hypothesis" concerning the origin of our ideas from without. He claimed that the origin of these ideas is not mind-independent bodies, but a spirit (God). The doctrine of mind-independent bodies was subjected by Berkeley to a number of different arguments, which will be described below. Several important parts of his arguments relied on the rejection of what Berkeley took to be Locke's doctrine of abstract ideas, which was attacked by Berkeley in the Introduction to the Principles of Human Knowledge.

Abstract Ideas
In our readings on Locke, there is only one section which discusses the mental operation of abstraction (II,XI,9). The function of abstraction is to allow the use of general terms to apply to things of the same kind: "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 those ideas." How this works, exactly, is a matter of contention, and here I will only spell out the theory as seen through Berkeley's eyes.

Consider Locke's example of the use of the general term 'white' to apply to chalk and snow (today) and milk (yesterday). The word is universal, applying to many objects, and corresponding to it is whiteness, a universal idea, a single color in all the complex ideas of white things. Locke called the universals "precise, naked appearances in the mind," and they are the result of the operation of abstraction (or at least so it seems).

Berkeley's picture of Locke's account of the semantics of general terms proceeded along these lines. He advanced a battery of criticisms against this picture. First, he claimed that the operation of abstraction is psychologically impossible. That is, one is unable, by any effort, to separate (or "precind") the naked appearance whiteness from the other qualities making up the complex idea of a thing. Every color is the color of some expanse with a determinate figure.

A follow-up criticism is that the result of such an operation is unintelligible. This can be stated in terms of a dilemma. Either the whiteness is a determinate whiteness or it is not. For whiteness to serve as a universal, it must be the same for all shades of white. If it is a determinate shade of white, say the white of the milk I saw yesterday, then to represent the different shade of white in the chalk, it must also be that shade, and so both shades, which is impossible. On the other hand, if it is neither shade, then we do not have a definite idea. One cannot imagine a white which is no determinate shade of white. Here Berkeley reveals a key assumption, that the abstract ideas would have to be a sensible idea rather than some kind of idea comprehended by reason alone.

The final criticism is that the postulation of abstract ideas is not necessary to account for our use of general terms. In the first place, Locke had claimed that the production of abstract ideas is difficult, to which Berkeley replied that the use of general terms by small children could not be accounted for.

But more importantly, there is a way of giving a semantics for general terms without appealing to intermediary ideas at all. When I use the term 'white,' it can refer indifferently to any of the ideas of milk, chalk or snow. I need have only one of them in mind and let it do duty for other ideas. Adapting the example of Sec. 16, we can say that I may consider the chalk merely as white, without attending to the other qualities making up the complex idea of the chalk. Note that this nominalist solution requires that considering a thing as of a certain sort. But this raises anew Locke's original question, namely, how we rank things into sorts.

Immaterialism Berkeley attacked Locke for his "materialism," and proposed an immaterialist hypothesis in its stead. It is important to recognize that Berkeley understood materialism as the thesis that matter exists, not that the mind is material, which is its current meaning.

Let us say that an object is corporeal if it has all the qualities of body, extension, figure, motion and rest, solidity, and whatever other qualities a body has. A material thing is a corporeal thing which is mind-independent. Thus materialism is the thesis that there exist mind-independent corporeal things, and immaterialism is its denial.

Berkeley propounded a number of arguments against materialism in the opening sections of the Principles of Human Knowledge. I have identified eight distinct arguments, which I will now discuss in more or less detail.

I. Objects of Perception (§§ 1,4)
  1. The qualities of bodies are perceived by the senses.
  2. What is perceived by the senses is ideas.
  3. Therefore, the qualities of bodies are ideas.
This argument trades on an explicit equivocation made by Locke, who distinguished ideas in the immediate sense (the sensations in our minds) and in the mediate sense (the qualities in bodies which cause these sensations). See II,VIII,7-8. Thus Locke could accept the conclusion so long as it is understood that 'ideas' is taken in the mediate sense.

II. Meaning of 'exist' (§ 3)
  1. When we say a body exists we mean that we perceive or would perceive it.
  2. What we perceive or would perceive are ideas.
  3. Therefore, when we say a body exists we say something about ideas.
This argument's first premise is an instance of a broader verificationist notion of meaning. For Berkeley, a sentence referring to a sensible object is meaningful just so far as there are empirical conditions under which its truth can be confirmed. But once again, so long as 'ideas' is interpreted in the mediate sense, there is nothing opposed to Locke's view. On the other hand, if it is interpreted in the immediate sense, Locke would object. He believed that unperceivable corpuscles exist, for example.
III. Abstraction (§ 5)
  1. It is impossible to separate in thought the things we see and feel from their perception.
  2. Bodies are objects we see and feel.
  3. Therefore, it is impossible to separate in thought bodies from their perception.
Here, Locke would say that it is possible to separate in thought the things we see and feel from their perception. This of course cannot be done for ideas in the immediate sense, but it can for ideas in the mediate sense.

IV. Likeness (§ 8)
  1. Ideas can only be like other ideas.
  2. Therefore, our ideas of bodies can only be like ideas.
  3. Therefore, if bodies have qualities resembling our ideas, then those qualities are ideas.
Berkeley has given us no reason to accept the first premise. Later, he will try to show an essential respect in which ideas in the immediate sense cannot be like ideas in Locke's mediate sense.

V. Primary and Secondary Qualities (§§ 9-14)
Locke had thought he had shown that the ideas of secondary qualities do not resemble their originals. Berkeley argued that the same arguments which purport to show this also apply to primary qualities. So ideas of primary qualities do not resemble their causes.

VII. Unknowability (§ 18)
It has already been shown that bodies are not known through the senses. They are not known through reason, either, since this would require the establishment of a necessary connection between ideas and their originals. But there is none to be found.

VII. No Explanatory Advantage (§ 19)
No one can explain how bodies can act on spirits so as to produce ideas in the mind, so there is no explanatory advantage to the materialist hypothesis over a hypothesis which posits another kind of cause.

VIII. "Master" Argument (§§ 22-23)
Any attempt to conceive the unperceived existence of bodies must fail. Therefore, it is impossible to conceive of a body existing unperceived. This is called in the literature the "master" argument because Berkeley professed to stake his whole case on it alone.

Primary and Secondary Qualities
I want to return to the argument concerning primary and secondary qualities. It is one of the best-known and central arguments in the group. Berkeley concluded from it that if material things exist, we cannot know that they do -- a skeptical conclusion. He would still need a reason to rule out the existence of things independent of the mind.

Recall that for Locke, a material substance is a group of primary qualities united by an unknown support. Locke had claimed that these qualities (most notably extension and solidity) are such that we are unable to conceive a body without them. He claimed further that our ideas of these qualities resemble the qualities themselves. The primary qualities are the source of powers in bodies to change things and receive change. Most importantly for present purposes, they can cause sensible ideas which do not resemble them; these are the ideas of secondary qualities.

Another way to distinguish the two is that the primary qualities are absolute, while the secondary are relative to perceivers. The color of a thing may vary from perceiver to perceiver in a case where the object is the same. Berkeley noted that this phenomenon holds for primary qualities as well. My idea of an object and yours may vary with respect to the size they represent the object to be, say large and small.

This variation in perceived size would not faze Locke. He would claim that the idea of a body resembles the body respect of there being some extension. The exact extension is not found in the idea itself, but it need not be for the quality to be primary. The body simply has to have the quality in some degree no matter what state it is in.

Berkeley's response would be that to say that a thing has extension in general without assigning it a specific degree is to attempt to use an abstract idea. But in the Introduction, he had claimed that there are no abstract ideas of this sort. Even if there are not, it is still possible to claim that my idea of, say, a ball of a determinate size can resemble a ball of any size. Berkeley himself had claimed that the figure of a triangle of determinate size and shape can be used as representative of all others.

Inertness and Activity
Now I want to move on to Berkeley's dogmatic argument, whose conclusion is that material substances cannot exist. The reason is that they are said to be causes of our ideas, but if they resemble our ideas, they could not be causes. Berkeley claimed we may observe that our ideas are passive, having nothing of activity in them, but a cause is an active thing. So the active cause of our ideas cannot be anything like an idea. A further consideration is that the materialists themselves admit that matter is inert, and hence cannot serve as a cause.

But it was pointed out in class that this is not a fair characterization of the position of the materialists. An object may cause a change in another object without being in one sense inert. That is, a material thing does not originate any change in itself, but merely passes along motion received from other things. But a billiard ball struck by a cue is nonetheless caused to move by the motion of the cue. If this is so, then there is room for resemblance between purely passive ideas (which pass in what Berkeley calls a "train") and matter.

Spirits
At any rate, Berkeley concluded that the only permissible explanation of the origin of those ideas we do not cause ourselves is another spirit. It was asked in class how Berkeley knows that he does not cause some of his ideas, and the only response is that we observe this to be so. This answer depends on the assumption of the ˘transparency÷ of the mind, that what goes on in the mind can be discovered by self-observation. Leibniz and others have challenged this assumption. Another difference between ideas said to be caused by another spirit and those which are self-caused is a difference in vivacity.

There is a thorny question about how we represent spirits. Spirits are said to be active beings, but ideas are passive. But the argument against materialism just discussed depends on the inability of ideas to represent anything active. Berkeley realized the slip after the publication of the first edition of the Principles, and he ˘corrected÷ it in the second edition (and the Three Dialogues). His emendation was to say that spirits are known through notions, and that we have a notion of a thing when we know the meanings of the words used to refer to it. How we known these meanings is never explained.

Immaterial Bodies
So far, Berkeley had recognized the existence of himself as a spirit, a purely active substance, and his ideas, which are purely passive and dependent on being perceived for their existence. He next distinguished between two kinds of ideas. Ideas can be compared in terms of vivacity, dependence on our will, and coherence. Those ideas which are most vivid, independent of our will, and coherent with one another make up are the components of real things. The others are the product of the imagination (which is also the source of dreams, hallucinations, etc.)

Real things are collections of ideas. An apple is nothing more than a red color, a sweet taste, a spherical shape, a soft texture, etc. This is sometimes called the ˘bundle÷ theory of objects. Most importantly, for Berkeley, things are not material substances. They are completely mind-dependent, and they have no support. (There is a strain of Berkeley interpreters, students of Gustav Bergmann at the University of Iowa, who claim that it is the mind which supports ideas, or in which ideas "inhere.") Berkeley stated that the relation of ideas to the mind is that of perceiver to perceived. Thus although the red color which is part of the composition of the apple is said to be "in the mind," it does not follow that the mind is red.

Berkeley recognized the novelty of his theory, and especially its apparent deviance from common sense. His solution was to urge us to think with the learned (apples are bundles of ideas perceived by the mind) and speak with the vulgar (apples are red, sweet, spherical, soft objects). Indeed, he thought his view accords more with the vulgar (uneducated) point of view, in that unlike Locke, he held that the red we see is really in the apple.

The banishment of material substances from the world of the senses comes at a certain cost, however. There is now no reason to believe that the collections of ideas we call things have any basis in the thing for their being together. Berkeley was a conventionalist in the sense that he thought that the use of names (this apple) is merely a matter of convenience, not significant of any real connection among the properties.

Indeed, Berkeley thought that the connection between the ideas in a collection is looser than anyone had thought. Locke had held that we can observe the same qualities (figure, extension, motion and rest) through both sight and touch. Berkeley broke with this view and claimed that the objects of sight and touch are entirely heterogeneous; they accompany one another but are of different kinds. He came to this view as a result of his work on vision, where he tried to refute the geometrical optics championed by Descartes. In particular, he held that distance is not seen at all, let alone geometrically, but is only estimated by comparison with the background. The moon low on the horizon is compared to the small-appearing mountains, but high in the sky it seems diminished.

This led Berkeley to think of sight as a kind of language or symbol system. Ideas of sight suggest other ideas of sight. They can also suggest ideas of touch, but they do so only imperfectly. We have an idea of a stick in the water; it looks bent. This suggests that on touching it, I would feel something bent. But it feels straight. The standard explanation is that it only appears bent, the distortion being accounted for by the behavior of light in water. But this invokes a geometrical view of light (its rays are bent on an angle determined by the type of medium), and Berkeley thought he had shown that vision is not based on geometrical relations.

God
To return to the main thread, I reiterate that for Berkeley, the difference between a real thing and mere ideas of the imagination is based on his three criteria of vivacity, independence on my will, and coherence. But if these collections of ideas are not dependent on me, on what are they dependent? Berkeley naturally inferred that they must be caused by a spirit, since spiritual substance is the only active being of which we have a notion. The nature of the spiritual cause of our idea is then inferred from the immensity, complexity, beauty, etc. of the ideas comprising our experience. Berkeley concluded that a being capable of causing this kind of pattern of ideas could only be God.

Berkeley's inference is very problematic. How does he know that he is not the cause of all his ideas, though in a manner which he cannot observe? Descartes at least could say that God would be a deceiver if He allowed this to happen, but Berkeley is here trying to prove that God exists based on the premise that he is not the cause of some of his ideas. How does he know that there is a cause at all? (This is Hume's question, asked by Professor Bossart with respect to Locke, who also had no good answer.) Why infer the existence of a single being of the stature of God? As Hume pointed out, there might be many causes acting in a more or less co-ordinated way, to produce our world. Given the imperfection of the correlation between the ideas impressed on our different sense modalities, perhaps a separate spirit produces each kind, with only some relatively loose consultation among them.

Berkeley's ontology consists of spiritual substances. The supreme substance, God, is the cause of ideas in myself, a finite substance. These ideas, which are what comprise real corporeal things, are to be distinguished from ideas that are the product of my own imagination. There are no material substances, hence the name 'immaterialism' which Berkeley gave to his hypothesis.

The Explanatory Advantages of Immaterialism
We have discussed Locke's materialist corpuscularian hypothesis. According to it, there are spiritual and material substances. The material substances are composed of invisible, intangible corpuscles which transfer motion to one another. Motions are transferred from things to bodies, and then there is an effect on our minds. Descartes had been unable to explain how this could occur, and Locke admitted that he was at a loss. Berkeley trumpeted as an advantage of his hypothesis that it could explain the production of ideas by a supreme spirit, since we are already acquainted from our own experience with the dependence of ideas on spirits. Moreover, Berkeley's hypothesis is ontologically simpler, in that it does not allow the existence of any more than one type of substance.

Rival Accounts of Body
An alternative to the Descartes/Locke materialist view was offered by Nicholas Malebranche, a celebrated post-Cartesian French philosopher. Malebranche's view is known as occasionalism, for the material world exists merely as the occasion for the production of ideas in our minds by God. Insofar as Berkeley shared with Malebranche the same account of the production of some of our ideas, they are often compared. Berkeley himself tried to separate their views by stating that the material world serves no purpose. God does not need anything to prompt Him into acting. Another difference is that Malebranche claimed that we have a rational intuition of essences in the mind of God. In this way he was quite Platonic. As we saw in the discussion of the Introduction to the Principles, Berkeley was a nominalist who eschewed Platonism altogether.

Another alternative was parallelism, which circumvented the interaction problem by denying interaction as well as divine causation of our ideas. On Leibniz's version (Spinoza had another), the ideas we have of things other than ourselves are in us innately, appearing to our consciousness on the proper occasion due to a pre-established harmony. It should also be noted that for Leibniz material things are only phenomena, which are the appearances of immaterial substances which are spirits or at least like them. Berkeley was apparently not aware of this alternative and discussed nothing like it.

Despite his rejection of a "third thing" which plays some role in our having the ideas we do, Berkeley had to recognize that our ideas are copies or "ectypes" of ideas in the divine mind, their "archetypes." The world shows the effects of divine foresight and regulation, which can be accounted for only by postulation of a divine plan. But as Wilfrid Sellars liked to point out, there is a loss of simplicity in this postulation, especially vis a vis pure materialist views which do not rely on the existence of God to explain anything.

One might object to Berkeley's view on the grounds that the role of material things is as the locus of forces, which are an essential ingredient of scientific explanation. This objection was lodged by Leibniz against Descartes, on whose view matter is inert. Can a system of purely passive ideas be adequate as an accounting of the sensible world?

Berkeley would say that it can, and that in fact his system has an edge in this respect. Newton had held that all bodies attract one another in proportion to their mass and the square of their distance apart. However, he could not explain how attraction takes place. That it does take place, he was certain, since bodies move in a way that must be described as an attraction toward one another. On Berkeley's view, the observed phenomenon is all there is to it, so Newton's problem dissolves.

Unperceived Objects
I want to turn now to the question of the status of unperceived objects. A materialist view would seem to have a big advantage in this respect. Material substances exist independently of perception, so there is no difficulty regarding their continued existence unperceived. But the being of ideas is to be perceived, so what am I to make of the book in my study, where no finite spirit, be it human or not, is perceiving it now?

There are two different answers given by Berkeley. The more famous answer is that what is perceived by no finite spirit is perceived by God, who perceived everything all the time. The second type of answer is given in terms of a counterfactual: a finite spirit would receive the object under the appropriate circumstances. I think the second type is less problematic than the first, but twentieth-century philosophers have shown it to have its own share of difficulties. I will discuss the relevant merits of the two accounts next.

The question of the status of unperceived objects (the table in my study when I am not there, for example) is raised by the following argument.

  1. Ideas exist only insofar as they are perceived.
  2. Physical objects are collections of ideas.
  3. Therefore, physical objects exist only when perceived.
If this argument is sound, then Berkeley is forced to find a reason to think that the table in the study is perceived by some spirit even when it is not perceived by me or any other earthly spirit. There two obvious possibilities within Berkeley's ontology for such a spirit. One is God, the other is a lesser spirit, e.g. an angel.

I believe the God option is untenable for the reason that God cannot have ideas. Recall that ideas of sense are passively received, wily nilly, by the human subject. Berkeley believed that God is pure activity: there is nothing passive in God's nature. So God could not be the recipient of ideas. If God "perceives," it would have to be in some sense other than the having of ideas.

More specifically, in Three Dialogues, Berkeley (through his mouthpiece, Philonous) concluded that a great heat is identical to a pain. But the suffering of pain is a privation, so God has no idea of pain. It follows that God has no idea of heat either. But heat is one of the qualities, the collection of which makes up a hot thing. [I did not mention this in class, but this argument of Berkeley's poses a grave threat to his own view of objects, since pain would have to be one of the ideas making up the collection constituting a hot thing. But, as my teacher Wilfrid Sellars insisted, this is a radically subjective view, reducing physical objects to things like "tickles and itches," as he so often put it to me.]

If God is out, why not angels? After all, Berkeley had explained by appealing to the perception of angels how the Biblical creation story could be literally true. That is, light, the heavens and earth all existed before the creation of spirits, but at this time they were viewed by angelic spirits. Whatever the merits of this picture, it could play no role in Berkeley's philosophy of bodies. He could invoke angels on the assumption of the truth of the Bible, but he could not, and would not, pretend to demonstrate their existence philosophically.

This leaves the other option. Things unperceived exist if they would be perceived under the right circumstances: "If I was in my study, I might perceive [the table]." This counterfactual approach has a certain appeal, in that it apparently needs no backing from any metaphysical doctrine like the existence of God. Twentieth century "phenomenalists" have attempted to flesh out the counterfactuals needed to make the approach work. Most notably, Rudolf Carnap in his Logical Construction of the World tried to do this in mathematical detail.

Unfortunately, the project has never borne fruit. One problem is its sheer magnitude. One would have to specify a set of primitive ideas and provide counterfactuals showing how all objects in the world are constructed from them. But there is a more pressing problem than this.

Consider Berkeley's "If I were in my study, I would perceive the table." This is really an inadequate formulation, since it is couched in terms of things rather than collections of ideas. The consequent would have to contain something like, "I would perceive a certain brown color, hard shape, rectangular figure, ... ." And the antecedent would have to be stated in terms of ideas as well, yielding, "If I were to perceive x, y, z ideas (those which together make up my being in my study), then I would perceive a certain brown color, hard shape, rectangular figure, ... ."

How can we ever justify the counterfactual? It seems always possible that I meet the perceptual conditions of the antecedent but fail to have the ideas specified in the consequent. The only answer Berkeley could give, I believe, is that God guarantees the truth of the counterfactual. This is probably the answer he would give, but note that this answer negates the original appeal of the counterfactual approach, that it seemed to remove the need for a powerful metaphysical backing for the existence of unperceived things.

Mathematics and Science
I turn now to a few remaining consequences for mathematics and science. Berkeley was a maverick in this area, opposing views of Descartes in optics (already discussed) and Newton in mathematics and physics. It turns out that history has sided with Berkeley concerning these matters.

Berkeley's account of mathematical proof (as described in the Introduction to the Principles) had it that it uses particulars (e.g. diagrams) which stand as representative for all others of its type. In considering the parts of things, as in geometry or calculus, we are constrained by certain minima, the smallest discernible objects of vision (and touch). These are the only objects which we are able to admit, so the doctrine of the infinite divisibility of lines goes by the boards. Similarly, Newton's doctrine of the infinitesimal, which underlay his presentation of the calculus, was criticized as "the ghost of a vanishing quantity." Many historians have claimed (though there has been vigorous opposition) that Berkeley's critique of Newton's version of the calculus led to its being superseded by Leibniz's version in England, where Newton at first held sway. Nobody accepts Newton's version of the calculus any more, though not for the reasons Berkeley gave, but because of further refinement of the mathematics, specifically, the notion of the limit.

In science, Berkeley opposed Newton's claim that motion and rest are absolute, rather than being relative to a frame of reference. It had long been recognized, since Galileo, that any object can be described indifferently as in motion or at rest, so long as a suitable frame of reference is chosen. Newton believed that there is, as it were, an absolute frame of reference, which he called absolute space (along with absolute time). This doctrine was very troubling to many philosophers (e.g., Leibniz), who thought either that it is unintelligible or that it gives absolute space and time properties (e.g., infinity) that should be reserved for God.

Berkeley held that space is nothing but a relation between bodies (which, of course, are collections of ideas), and time a succession (or "train") of ideas. Absolute space and time are unperceivable, even on Newton's account, and so have no place in Berkeley's ontology. Further, Berkeley held that the thought-experiment Newton used to establish the existence of absolute motion is a failure. Again, Berkeley anticipated the judgment of history.

Back to Lecture Menu

Eighteenth Century Philosophy Home Page