Contrary to a popular reading of his modal epistemology, Berkeley does not hold that inconceivability entails impossibility, and he cannot therefore argue the impossibility of mind-independent matter by appealing to facts about what we cannot conceive. Berkeley is explicit about this constraint on his metaphysical argumentation, and, I argue, does respect it in practice. Popular mythology about the ‘master argument’ notwithstanding, the only passages in which he might plausibly seem to employ the principle that inconceivability entails impossibility are those that (...) argue for the inseparability of primary from secondary qualities. However, an alternative reading of these texts is available that is both consistent with Berkeley's express modal epistemology and credible in its own right. (shrink)

The resemblance theory of representation is the view that one thing represents another by virtue of resembling it. Typically, it is taken as non-controversial that Berkeley accepts the resemblance theory of representation – even if the plausibility of the resemblance theory itself comes under scrutiny. One piece of evidence in favour of this reading of Berkeley is his commitment to the ‘likeness principle’: the view that ‘an idea can be like nothing but an idea’ (PHK § 8). The likeness principle (...) is typically seen as a rejection of the kind of indirect theory of perception that is often attributed to Locke. However, by paying particular attention to Berkeley’s account of how mental representation via the imagination works, Dávid Bartha casts doubt on the consensus view that Berkeley does accept the resemblance theory of representation. (shrink)

Berkeley's philosophy is meant to be a defense of commonsense. However, Berkeley's claim that the ultimate constituents of physical reality are fleeting, causally passive ideas appears to be radically at odds with commonsense. In particular, such a theory seems unable to account for the robust structure which commonsense (and Newtonian physics) takes the world to exhibit. The problem of structure, as I understand it, includes the problem of how qualities can be grouped by their co-occurrence in a single enduring object (...) and how these enduring objects can bear spatiotemporal, causal, and other relations to one another. I argue that Berkeley's solution to these problems lies in his views about language. At one level, human language allows us to exploit patterns in our perceptions to construct a highly structured representation of the physical world which allows us to make accurate predictions at minimal cognitive expense. At a deeper level, these patterns occur in perception because our perceptions themselves form a language in which God speaks to us. (shrink)