30

Berkeley, Principles, Introduction

Abstract Ideas

Early modern philosophy was as influenced by developments in early modern religion as by developments in early modern science. This is nowhere as much the case as it is in the philosophy of George Berkeley, later Anglican Bishop of Cloyne, in Ireland. Berkeley lived at a time when religion in Britain was split along the lines that were described in the chapter on Bayle. At one extreme was a group of radical religious rationalists, referred to as the Deists. The Deists believed that religion ought to be based on reason alone, and that only what could be proven by reason should be included in the body of religious beliefs. The earliest Deists had written before Locke, and Locke’s own remarks about the relation between reason and revealed truth may have been influenced by them. But Locke had maintained that it is possible to have a revealed religion that goes beyond what reason can tell us on its own, as long as the revelation does not directly contradict reason and reason can authenticate the divine origin of the revelation. The chief Deistical writers, in contrast, thinkers like Anthony Collins, John Toland, and Matthew Tindal, were not willing to make such compromises. They wrote books with titles like, Christianity not Mysterious, and A Discourse of Free-Thinking, that denied revelation and the authenticity of the Biblical miracles supporting it, and rejected the Trinity, the divinity of Christ, and even the Resurrection. Most of the Deists affirmed the existence of a divine Creator, and they maintained that the Creator is worthy of worship, that the proper form of this worship consists in obedience to the laws of morality as revealed by reason, and that there will be an afterlife and a final judgment, though their views were by no means uniform. For many at the time, including Berkeley, their stripped-down religion was no longer recognizably Christian. It was tantamount to Atheism.

At the other extreme were the Fideists, who held that the Holy Spirit compels the elect to believe the Christian doctrines, however contrary to reason. In the Britain of Berkeley’s day they were represented by the more radical elements of the Presbyterian church in Scotland. They had had their heyday during the time of the English civil war. At that time, a Puritan movement in England, allied with the Scottish Presbyterians, had beheaded a King who was viewed as inclining too far towards Catholicism, and temporarily set up a commonwealth. These were the people that Locke attacked in his chapter on enthusiasm as dangerous fanatics, whose willingness to accept their own passionate beliefs as divine revelation and throw down the sovereignty of their own reason was mirrored in the violence with which they treated others and the turmoil into which they had plunged the state.

However, anyone who looked at the work of a Fideist like Bayle (and almost everyone did, as the Dictionnaire was one of the most popular and widely read works of the time) could readily see that there were other things that made fideism dangerous besides its pernicious political tendencies. The scepticism Bayle had so effectively employed as an inducement to faith was a dangerous weapon that could just as readily be turned against faith itself to produce a spirit of indifference and agnosticism. That spirit seemed to be only too well exemplified by many of Bayle’s own writings. Fideism seemed as dangerous to religion as Deism.

The full titles of Berkeley’s major philosophical works,

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,

Three Dialogues between Hylas and Philonous, the Design of which is plainly to demonstrate the Reality and Perfection of Humane Knowledge, the Incorporeal Nature of the Soul, and the Immediate Providence of a Deity: in Opposition to Sceptics and Atheists

make his concern to find a middle way between the extremes of scepticism and atheism quite evident.

At a fairly early point in his life — his late teens or early twenties — Berkeley appears to have concocted a radical plan for combating the extreme views on religion of his day. One need simply deny that there are any material things in existence. This would at once deprive the rationalists of their chief source of pride and security and the Fideists of their chief sceptical argument.

Berkeley seems to have reasoned along the following lines:

The modern philosophers, like Locke and Descartes, had supposed that what we first and most certainly know are just our own ideas. They would have liked to think that these ideas are caused in us by extended bodies existing outside of us, but they ran into serious problems when they tried to prove that these bodies exist. That had given Bayle and the Fideists an excuse to launch their sceptical attack and claim that reason is inadequate to give us knowledge of the material world. But had Locke and Descartes not tried to prove that there is material world — had they instead insisted that the only things that exist are minds and their ideas — then they would not have opened themselves to attack, and the sceptics would have had no reason to claim that we do not have knowledge of the “real” world, since it is hardly possible to deny that we know our ideas, and much more difficult to deny that we have reason to suppose the existence of other minds.

But even had Locke and Descartes succeeded in their project of demonstrating the existence of an external world, the result would not have been positive for religion. The external world they envisioned is a world of pieces of extended and perhaps solid matter moving in accord with constant laws and producing all the phenomena of nature, and all our ideas of sensation, as a result of their motions and collisions. Miracles do not happen in such a world, and there is no role for God to play in it, other than to create it and perhaps arrange its parts in the beginning. Thus, the success of new philosophy would tend to support Deism.

Recognizing that material objects do not exist and that the real world can only be a world of spirits and their ideas would forestall this Deistic conclusion at the same time that it blocked Fideism. Berkeley regarded it as a brilliant option.

Of course, arguing for a theory a posteriori, that is, by appeal to the good consequences that would follow were one to accept it, is no way to prove that it is correct — especially if it seems to be as patently absurd as denying that there is an external world. Wishing it were so because if it were all sorts of nice things would follow does not prove that it must be so. We need, therefore, to inquire into the philosophical reasons, as well as the religious motives for Berkeley’s theory.

Berkeley’s main reasons for denying the existence of matter are presented over paragraphs 1-24 of Part I of the Principles. (An originally planned Part II was lost while still in manuscript and never published.) However, these reasons are importantly supported by a critique that Berkeley leveled against the notions of abstraction and abstract ideas. This critique, offered in the Introduction to the Principles, is an interesting reflection on issues in early modern philosophy of language and representation that stands on its own quite apart from any reference to Berkeley’s other doctrines. It is particularly worth studying as a counterpart to Locke’s account of abstraction and abstract ideas in Essay III.

QUESTIONS ON THE READING

1. What is the chief cause of those obstacles and difficulties that have so far prevented us from making any progress in philosophy?

2. What is our “most abstract idea of extension” (Introduction §8) an idea of?

3. What is the one sense in which Berkeley thought it is possible to abstract?

4. What are the “proper acceptations of abstraction” (i.e. the senses in which this ability has been understood according to the tradition Berkeley is attacking)?

5. What are the two arguments against the traditional conception of abstraction that Berkeley had to offer in Introduction §10?

6. What was Locke’s reason for claiming that human beings are able to form abstract ideas?

7. Why did Berkeley find this reason to be inadequate?

8. How can an idea be general without being abstract?

9. What is the argument against the traditional conception of abstraction that Berkeley had to offer in Introduction §13?

10. What is the erroneous supposition about the nature of language that lies at the root of the supposition that we have abstract ideas?

NOTES ON THE READING

Berkeley opened his discussion of abstraction by first outlining the traditional doctrine he proposed to attack. It is important to be clear about when Berkeley was merely describing the views that he proposed to attack, and when he was giving reasons for opposing those views. Introduction 7-9, the first two thirds of 10 and 13, and the first third of 16, in particular, are all devoted to describing positions Berkeley wanted to attack.

According to the traditional doctrine Berkeley set out to attack, abstraction is the ability to separate a thing from its surroundings and consider it on its own. Anyone who has studied chemistry will know that there are certain processes that can cause a chemical compound to precipitate out of a solution. This is the paradigm example of abstraction. The chemist takes something that was originally a mixture and is able to draw something out of it and separate it from the rest.

Berkeley maintained that many past philosophers, paradigmatically Locke, had supposed that the mind is capable of an analogous operation. It was supposed to be able to take an experience and separate it into its component “simple ideas.” As Berkeley understood the views of his predecessors, this mental abstraction involves the same sort separation as chemical abstraction: he understood Locke to have held that we are able to form a simple idea like the idea of redness or the idea of extension apart from having any other simple idea. It is not clear that Berkeley was right in understanding Locke this way. All that Locke may have meant to say is that abstraction involves focusing one’s attention on something that is present in experience or memory while ignoring other things, even though those other things remain there. Forming an abstract idea of red, for example, would not involve forming an isolated, simple idea of red in one’s imagination, but imagining some red shape or other and attending just to the redness and not to the shape. This is not that far removed from what Berkeley himself took abstraction to involve at Introduction 16.

Had he been confronted with this objection, Berkeley might have been happy to grant it. In his view, the principal problem with the doctrine of abstraction is that it leads people to draw certain illegitimate inferences from conceivability to possibility. Thinking that we can conceive one thing without having to think of another, they infer that it must be at least possible for the one thing to exist apart from the other. Most notably, thinking that we can conceive extension without having to conceive colour or any tactile sensation, they suppose that it must be possible for there to be extended bodies that lack colour or any other sensible quality. And thinking that we can conceive colour or solidity to exist apart from being perceived, they suppose that it must be possible for colour or solidity to exist when they are not perceived. Berkeley considered both of these inferences to be illegitimate, and both of them to be founded on the radical doctrine of abstraction, understood as involving the formation of isolated ideas. Had even the proponents of the doctrine of abstraction only meant to say that abstraction involves a kind of selective attention rather than a separate conception, then the arguments for the separability of primary from sensible qualities and of being from being perceived would lose all merit. It is only to the extent that people take the doctrine of abstraction to imply a power of separate conceivability, and separate conceivability to involve the possibility of separate existence, that Berkeley was concerned to oppose it.

Granting that abstraction does involve separate conceivability, there are four cases that Berkeley was concerned with:

i) cases where a simple idea, like red, triangular, rough, stationary, or sweet is supposed to be separately conceived.

ii) cases where we are supposed to conceive a kind of simple idea, like the idea of colour in general, shape in general, extension in general or the ideas of a sensible quality in general or just of being in general.

iii) cases where we are supposed to separately conceive the idea of a kind of substance, like man, woman, cat, dog, gold, lead, turnip, carrot. These are cases where we are supposed to conceive just those simple ideas that the members of the kind exhibit in common, separately from all the other ideas that individuate the members of the kind from one another.

iv) cases where we are supposed to separately conceive a generic kind of substance, like human, animal, or body. These are “yet more abstract ideas” supposedly formed by conceiving just those simple ideas that the different kinds share in common, separately from all the other ideas that individuate the kinds from one another.

There were two other cases that he was not concerned with:

i) cases where ideas given through one sense are conceived separately from ideas conceived through another sense, as when colour is conceived separately from taste or smell.

ii) cases where spatial or temporal parts are conceived separately from one another, as when a hand is conceived apart from an arm or an egg apart from the chick it later becomes.

Berkeley thought that in both of these cases we can form abstract (i.e., separate) ideas. But in the earlier four cases, the formation of abstract ideas is impossible.

Berkeley’s Critique. Berkeley offered two main reasons for rejecting the first four sorts of abstract ideas. The first is given in Introduction 10, the second is added in 13.

Berkeley’s first and main reason appeals to the contrapositive of the claim that what is conceivable is possible: the claim that what is impossible is inconceivable. Anyone who accepts the former claim would have to accept the latter, since the two are logically equivalent. But, Berkeley claimed, everyone agrees that it is impossible for there to be extension where there is no shape, for a shape to exist apart from some sensible quality or other, or for a sensible quality to exist apart from being perceived. There can be no colour that is not spread out and located somewhere. Conversely, whatever is spread out or extended must have some shape. But for there to be shape there must be edges, and there can be no edges where there are no contrasting sensible qualities such as colours or feelings of solidity. Again, if colours and tactile feelings are just sensations, as everyone at the time supposed they are, then there can be no colours or tactile feelings that are not perceived since for a sensation to exist is for it to be perceived. Since it is impossible for these things to exist separately, it follows from the principle that what is impossible is inconceivable that we should not be able to conceive them separately either.

We might object that it is false that what is impossible is inconceivable. Faster than light speeds are impossible. But they are not inconceivable.

Perhaps Berkeley would have tried to defend himself by saying that he had a different sense of impossibility in mind. Faster than light speeds are only physically impossible.

But this raises the question of what other sense of impossibility Berkeley might have had in mind. One other sense of impossibility is logical impossibility. But things are only logically impossible if they involve a contradiction — a claim of the form “Both P and not-P.” It is contradictory to say that something is both coloured and not coloured. But it is not obviously contradictory to say that something is both coloured and not extended. Similarly, it is contradictory to say that something both exists and does not exist. But it is not obviously contradictory to say that something both exists and is not perceived.

Another sense of impossibility is psychological impossibility — an inability on our part to conceive something. But Berkeley could not have appealed to psychological impossibility to prove his point without arguing in a circle. To say that something is psychologically possible is to say that it is inconceivable. But what Berkeley wanted to prove is that it is not possible to conceive extension apart from colour or being apart from being perceived. If his reason for saying this was that these things are impossible in the sense of being inconceivable, then he was assuming what he was trying to prove.

Faced with this result, some commentators have suggested that that Berkeley never intended to argue from impossibility to inconceivability. Instead he only ever meant to assert that, as a matter of fact, abstract ideas of the first four types are inconceivable. This is suggested by large parts of Introduction 10, in which Berkeley simply rested his case on an appeal to his own experience of what he was personally able to, as he put it, “imagine” or “conceive” or “consider” or “form” or “frame” in his mind, as well as on appeal to what he took to be the experience of “the generality of men who are simple and illiterate.” (Of course, in making the latter point, Berkeley did not mean to concede that sophisticated and literate people can abstract ideas. He rather meant to insinuate that, human nature being everywhere the same, what lies beyond the mental powers of the simple and illiterate likely lies beyond the powers of the sophisticated and literate as well, though their commitment to extravagant theories may blind them to these plain facts of their own common sense experience. It is a theme of Berkeley’s thought, already insinuated by Introduction 1-4, that the views of the simple and illiterate are closer to pure common sense, whereas those of the sophisticated and literate have as likely as not simply been corrupted and distorted by false theories.)

But do we really find it impossible to form an idea of colour apart from shape, or of an animal that is not of some particular species? Perhaps it all depends on what we mean by “form an idea.” If forming an idea means something like making a little picture in the mind’s eye, then Berkeley has a very plausible case. I may not be able to “image” or “frame” or picture a colour without giving it some particular shape, size, situation, and motion or rest, or an animal without making it of some particular species. But if we think that the formation of ideas might involve something more like grasping a meaning, then Berkeley’s thesis is far from obvious. Do I picture anything when I grasp the meaning of addiction or adversity? If I do, is what I picture at all adequate to represent what I understand these ideas to mean?

There is, however, another way of interpreting Berkeley’s argument, one that does not involve appeal to the claim that he took ideas to be like pictures or images (plausible thought that claim is). This is to say that he did mean to argue from logical impossibility to inconceivability — but that the logical impossibility he meant to argue from is a richer notion than that provided by analytic truth, where what is logically impossible is what is false simply by definition. Instead he based his notion of logical impossibility on Descartes’s and Locke’s notion of intuitive certainty. What is logically impossible is what is contrary to an intuitively obvious truth, not just what is false by definition. It is not false by definition that orange is more like green than it is like red. But it is intuitively false because, even in the absence of a definition, we can tell, simply by inspecting the ideas involved, that this relation does not hold. Berkeley did take it to be intuitively or at least demonstratively obvious that extension is inconceivable apart from colour or colour apart from being perceived. He might, for instance, have argued as follows: to say that something is extended is to say that it has parts set outside of parts. But for parts to be set outside of parts there must be boundaries or edges where one part ends and another starts. But there can be no boundaries or edges where there are no contrasting colours or feelings of pressure. But colour and tactile feelings are sensations. Sensations are states of mind. So for a sensation to exist is for there to be a mind that is in a certain state, the state of having or, as we say, perceiving that sensation. From these non-analytic but intuitively obvious equivalences it follows that there can be no extension without sensible qualities and no sensible qualities without perception of those qualities.

Berkeley’s second reason for rejecting the first sort of abstract ideas rests on an appeal to the notion that the formation of abstract ideas would involve performing an incoherent or inconsistent task, like drawing a picture of a round square. The task in question is forming an idea that has all of a number of contradictory qualities and none of those qualities at one and the same time. We might think that an abstract idea must do this if it is to adequately specify what things it refers to. The abstract idea of a triangle, for example, would have to refer to triangles of all sizes and colours, with all possible sizes of angles. We might think that it would do that by representing all of those different sizes and colours at once. Yet at the same time, the abstract idea is just an idea of what all triangles have in common. So it could not have any specific size or colour. Berkeley considered this to be an absurdity that follows from the doctrine of abstraction. He drove this point home by capitalizing on some incidental remarks made by Locke, to the effect that the abstract idea of a triangle must represent all and none of the differentiating features of triangles at once. But this was opportunistic. Locke’s remarks are not obviously a sound representation of his position. As Berkeley himself represented the traditional theory over Introduction 7-9, it involves forming an idea by separating just certain components of our experience from the rest of that experience — separating, for example, the features common to all triangles from an experience of a particular triangle. The resulting abstract idea would contain just what all triangles share in common. It would certainly not contain all the features of all triangles at once. Yet it is the latter view of what an abstract idea is and not the former that Berkeley appears to have adopted in Introduction 13.

On a more charitable reading of Introduction 13, Berkeley was just claiming that it is impossible to form an idea of a triangle that does not have angles of any particular size, not one that has angles of all different sizes at once. However, then his claim in Introduction 13 is no different from his main argument in Introduction 10. He was not claiming that abstraction involves forming an idea with inconsistent attributes, but rather just that it involves performing the impossible task of separating certain attributes from their necessarily concomitant particularizing circumstances.

Berkeley’s Theory of Signification. Questionable though Berkeley’s position may be, it is far from implausible, and he further buttressed it by claiming that there is no good reason for adopting the rival view. Berkeley recognized two reasons that might be given for accepting that we form abstract ideas: that such ideas are necessary if we are to account for our use of general terms, which make the preponderance of terms in our language, and that such ideas are necessary for the formulation of scientific laws and principles, which take the form of generalizations from particular experience. In addressing both reasons, Berkeley developed an alternative philosophy of language.

The reason why people have supposed that we have abstract ideas, Berkeley observed in Introduction 11, is that we use general names. We give names to qualities, like redness, that a number of different objects are observed to share in common, and where objects are observed to share a number of different qualities in common, we collect them together into groups and give names to those groups that we then use to refer indifferently to all the objects in the group. We further suppose, Berkeley observed in Introduction 19, that language is used to convey our ideas to others, so that where there is a name there is some idea that is referred to by that name. This fact about our languages and this supposition about the meaning of names lead to the supposition that we must have abstract ideas. For, the argument runs, if there are general names, but all names must refer to some idea, then there must be some idea that our general names refer to. This idea can only be the idea of the set of features that a number of different objects share in common, and since that set of features does not exist on its own anywhere in the outside world, but is always mixed in with other particularizing concomitant features in different, individual objects, it must instead exist in the mind, as a separate, abstract idea that is named by the general term.

The problem with this argument, Berkeley observed, does not lie with its first premise. It is undeniable that we do have and use general names. But Berkeley took the second premise to be questionable. It is not the case that there must always be some idea that a name refers to. Some names do not refer to anything at all but serve rather to rouse emotions, induce us to act in a certain way, or put us into a certain disposition. (Think of the names used to praise or insult someone, or to incite fear and get someone to refrain from performing a certain act.) And other names, Berkeley observed, may simply stand ambiguously as signs for a whole class of things.

This makes sense when considered from the first person perspective. I can readily understand how I might come to observe that a number of objects share certain features in common, be led as a result to think of those objects as all belonging together in a group, and come to use a particular name to refer to that group and to identify all the objects in that group as members of that group. From this perspective, there need be no intermediate, abstract or general idea connecting the name with the group.

But problems begin to emerge when we consider how others would understand this name, or how I myself would understand it at a later time. There is an immediate objection to the claim that some names are signs for whole classes of things: some classes are very large. If, upon hearing a general name, like tree, we had to call to mind all the objects that are referred to by that name, we would be faced with a daunting task. One way to get around this difficulty is to suppose that when we hear a general name we indifferently select just one member from the group of objects signified by that name and form the idea of that individual. But then how does that differ from our not treating the name as a general name at all, but merely as the name of that individual?

In articulating his position, Berkeley was careful to avoid both of these pitfalls. He claimed that when we hear a general name we do “indifferently” form an idea of just one of the objects in the group referred to by that name. But he went on to insist that we take this idea to serve as a sign that stands for all of the other objects in the group. When we do this, the idea is “made general,” though Berkeley insisted that it does not become abstract, but remains an idea of particular object. It is just that this particular idea is made to stand in for all the other objects in the group.

Berkeley gave the example of a geometry teacher who draws a black, one inch long line segment on a piece of paper. By showing the students how to bisect this one segment, the geometry teacher has shown the students how to bisect any line segment. In this case the particular line segment serves to signify any line segment of whatever length or colour, in whatever place or time. An idea of a black line segment of about 40 minimally visible points in length ought similarly to be capable of serving as a sign for all line segments of whatever length or colour. And so it is with all general terms. Hearing the name, “dog,” I form an idea of a particular dog, but I take this particular idea to stand for all dogs whatsoever.

There are obvious concerns that arise with Berkeley’s theory at this point: how does the one particular idea that we form manage to stand in for all the rest? Would there not be some danger that the particular idea, being particular, would not be able to perform its job adequately but would lead us to conceive of only very closely resembling objects as falling into the group? For example, if I form an idea of a particular terrier upon hearing the word, dog, what is to prevent me from only considering other terriers — or other terriers of that particular size or colouration — as alone belonging in the class, dog? Or, alternatively, what is to keep me from considering any four-legged beast, be it a cat or an elephant or an iguana as also being a dog? Berkeley had little more to say and it is hard not to escape the conclusion that he was left having to suppose that something magical happens to make the particular idea signify in just the right way.

Berkeley was somewhat better able to apply his theory to account for the second function of general ideas, their ability to serve as a foundation for the demonstration of general rules and principles. His argument on this score is neatly laid out over Introduction 15-16. According to it, any demonstration that is performed on a “general idea” ought to hold valid for all other ideas that resemble the “general idea” just in virtue of possessing all those features that figure in the demonstration. In this way, a proposition proven by a demonstration performed on a “general idea” can come to be considered valid of a whole range of particulars. Here, it is the features involved in the demonstration that explicitly serve to define the scope of the “general idea” and the sort of problem that arises in connection with the bare communication of the meaning of general terms does not arise.

ESSAY QUESTIONS AND RESEARCH PROJECTS

1. Did Berkeley understand Locke’s position on abstract ideas correctly?

2. Does Berkeley’s rejection of abstract ideas rest on the supposition that ideas are pictures? Is there another way of understanding what ideas are that would evade his argument against forming abstract ideas?

3. Is Berkeley’s account of the meaning of general terms cogent or was he unable to adequately explain how we come to views different things as all belonging to the same group?