He was also an idealist, in the sense that he thought that the fundamental "stuff" of reality consisted in ideas (in the Lockean sense). The easiest way to describe it to a modern audience is that he thought reality - the "real" world in which we all actually live - was like a giant MMO, where God plays the role of the server. What makes things "be real" is "being perceived," it's just that God perceives all in order to keep things running. To prove this idea that reality is fundamentally ideas instead of fundamentally stuff, he set out to establish that the idea of a "physical" object is incoherent, insofar as it entails a contradiction (a "manifest repugnancy," as Berkeley wrote). He wrote a bunch on this, but he presented one "Master Argument" on which he said he was "content to put the whole," as in wager everything. Unfortunately for him, he trades on an ambiguity - depending on what exactly he meant, it might be either equivocation, conflation, or amphiboly. I think this threefold nature of the ambiguity is what made it go undetected for centuries.
Either that, or nobody really feels obligated to comb through the old ramblings of Irish bishops except for philosophy undergrads. Y'know. One or the other.
Bishop George Berkeley’s “Master Argument” is a thought experiment meant to show that it is absurd to believe in mind-independent physical objects. While the argument itself is clear and appears well-structured, it rests on an underlying ambiguity that Berkeley may well not have noticed. Additionally, the link between inconceivability and impossibility is not clearly established in this case, and breaks down on closer examination. Despite these flaws, Berkeley’s physical arguments still show something interesting about the link between the world and what human minds think of it, but it is not the illustration that Berkeley imagines it to be.
Berkeley begins by asking the reader to consider whether he or she can conceive of an object that is not being perceived by anyone, and proceeds to argue that such a belief entails a contradiction (Berkeley, p. 193-4). Laid out propositionally, the argument proceeds along these lines:
1. Consider whether you can conceive of an unperceived object.
2. If you can, then you perceive it with your intellect.
3. Therefore the object in question is not unperceived.
4. This (1 & 2 & 3) is a contradiction.
5. Therefore you cannot conceive of an unperceived object – such a thing is inconceivable.
6. Therefore unperceived objects are impossible.
7. Therefore mind-independent physical objects are impossible.
Almost right away, in premise 2, we have a problem. The connection between 1 and 2 responsible for the contradiction in 3 (pointed out explicitly in 4) hinges on the matter of whether the two usages of “perception” are equivalent. They are not. The question in 1 is as to whether I can imagine an object that is not currently impinging upon the sense organs of a conscious being. I can do this with minimal effort. Premise 2 then says that I am now perceiving it with my intellect – but this is not the same as the kind of perception we were dealing with in 1. “Perception with the intellect” is simply conception or imagination, and is different from perception with the senses. We may modify the argument and propose 1’: “Consider whether you can conceive of an object without conceiving of it.” Proposition 2’ follows: “If you can, then you are conceiving of it.” But this is a trivial point, when stated explicitly in the only way it genuinely holds. Indeed, Berkeley is setting up the reader for failure – no one would take this bet, upon which Berkeley is “content to put the whole” (Berkeley, p. 193).
This is, in all likelihood, an innocent mistake and not a deliberate equivocation on Berkeley’s part. The early Moderns conflated many mental operations together; John Locke, for instance, defines an “idea” as any conscious state (Locke, p. 47). With this in mind, we can imagine the following modifications to premises 1 and 2, as follows:
1’’. Consider whether you can have an idea of an object, of which object no conscious
mind has an idea.
2’’. If you can have such an idea, then your conscious mind has an idea of the object.
This works in the way that Berkeley wants it to, insofar as the act of imagining what he asks the reader to imagine does in fact entail imagining a contradiction. However, this is not what Berkeley needs for his Master Argument to work; he needs the idea of an object that is not currently perceived by the senses to show that the concept of a mind-independent physical object is absurd – an uninteresting tautology about whether an imagined object is imagined by a mind is utterly unhelpful. The only way to accomplish this is to either equivocate between the two senses of “perceive,” or to conflate intellectual perception with sense perception. It seems most likely, as well as most charitable, that Berkeley was operating innocently under the latter error. After all, when I perceive an object with my senses, I often explicitly conceive of it as well; but it does not follow from this that every instance of intellectual conception is also an instance of sense perception – that would be affirming the consequent. In Berkeley’s defense, it does stand to reason that his God does not rely upon sense organs for perception as we do, and so in the specific case of God these really would be the same.
Alternatively, we can consider this a problem of parsing the first proposition. Consider the following two versions of proposition 1, where emphasis indicates a contiguous phrase:
1A. Can you think of an object x existing without your thinking about object x?
1B. Can you think of an object x existing without your thinking about object x?
Clearly 1A is impossible, but 1B is trivially easy by positing a world with object x and zero conscious minds. Berkeley invites us to consider 1B, but then acts as though we have attempted 1A, when these are two distinct considerations: 2 follows from 1A, but not from 1B, which indicates only that I am conceiving of some possible world in which no conscious mind is conceiving of object x.
Without this connection between premises 1 and 2, 3 does not follow, and so neither do 4 and the rest. Berkeley might at this point insist upon his own idealism, and fair play to him: on Berkeley’s view, such “intellectual perception” is at root the only kind of perception that we have. In a way, this does accurately describe how the mind is only able to entertain its own activity – it’s just that some of that activity depends on external stimulus. Let us grant him this point and see where it goes. I shall borrow here from Robert Cummins’ distinction between indication and representation:
“Indication is transitive, representation is not. If S3 indicates S2, and S2 indicates S1, then S3 indicates S1… A representation of the pixel structure of a digitized picture of my aunt Tilly is not a representation of my aunt Tilly’s visual appearance, though, of course, it is possible to recover the later from the former. …a representation of the pixel structure is an encoding of my aunt Tilly’s visual appearance.” (Cummins, p. 3)
In the matter at hand, this amounts to the difference between thinking of a thing itself and thinking of a thing’s label explicitly as a label. Going back to premise 1, I may now say, “Yes, I can conceive of an unperceived object: I am now thinking of the unperceived objects beneath my couch.” Having not examined my sub-couchal regions in several months, I have no idea what the world has in store for me, so I cannot possibly be conceiving of the objects themselves – I must resort to a label. Yet whatever is in fact under my couch, whether coins or chimeras, that is what I denote by the phrase “the unperceived objects beneath my couch.” (As it happens, there were two bottlecaps and a pen.) Not only may I do this any number of times, I may even do it with explicit objects like tables and chairs – the only difference is that I must be honest when I report that I am thinking explicitly of the words, “an unperceived chair,” and not any particular chair.
This brings us to Berkeley’s refutation of Locke’s account of abstract ideas, which Berkeley thinks undergirds the idea of physical substance (Berkeley, p. 186). While Locke argued that we note the commonalities of individual chairs in the world and subtract their differences to arrive at the abstraction of a chair (Locke, p. 159), Berkeley points out that we think of no such “abstract chairs” but instead imagine particular chairs when we think of “a chair” (Berkeley, p. 171 & 186). We may think of one chair, or many, but every chair is a particular chair, and every idea of a chair is a particular idea of a particular chair. Note that this is not a claim about how we abstract, but simply a claim that we do not think explicitly of our abstractions. This prevents, say, John Locke from side-stepping Berkeley’s argument by saying, “I conceive an unperceived abstract chair,” since we do not conceive of abstract objects. But the inability to think of abstractions themselves in no way prevents us from thinking of object labels without thinking of objects themselves.
And so we see that even though it is impossible to conceive of a particular object and at the same time have that object not be conceived of by anyone, we may of course conceive of representations of objects without indicating those objects with our thoughts. But even if we could not, the Master Argument would only highlight a limitation of the human mind, for while conceivability is often a good test of possibility, there is no necessary connection between the two. The human mind has limitations, to be sure, but the Universe is not necessarily under any obligation to conform to our expectations. While square-circles are both inconceivable and impossible, they are inconceivable only to those who understand the properties of squares and circles, and impossible only when we hold those meanings fixed – but this impossibility follows from the meanings of the terms, and not at all from the shortcomings of the human mind. Berkeley wants to have it the other way, but even if it were the case that humans could not think of object labels without thinking of the referents themselves, this still would not show that it is impossible for mind-independent physical objects to exist, for we could simply imagine beings who could perform exactly the sorts of mental exercises described above, thereby dissolving the contradiction.
Berkeley’s Master Argument, as we have seen, fails on two counts: first, the contradiction it claims to exploit does not follow from the argument without a crucial threefold ambiguity, whether an equivocation, a conflation, or amphiboly; and second, the alleged inconceivability of mind-independent physical objects only holds for beings who cannot confine their imaginations to conceive only of an object’s label and not the object itself, which in no way establishes their logical impossibility. But when we take a step back and look at the argument in the context of Berkeley’s idealism, we can still see that the naïve conception of physical substance is untenable (Berkeley, p. 185 & 187-8), and the idea of what physical substance must then be is, at least, bizarre. And that is a conclusion to which anyone acquainted with modern physics would readily agree.
Berkeley, Short Works of GeorgeBerkeley.
Cummins, Representation and Indication, https://netfiles.uiuc.edu/rcummins/www/HomePage/Papers/RepAndInd.pdf