¹ In-text references to Berkeley's works are abbreviated as follows: A Treatise Concerning the Principles of Human Knowledge (PHK); Three Dialogues Between Hylas and Philonous (TD); New Theory of Vision (NTV); Theory of Vision Vindicated (TVV). For Correspondence with Samuel Johnson I will use Library of Liberal Arts editions edited by Colin Turbayne. For Alciphron (AL) I use the edition by Berman, David (London: Routledge, 1993)Google Scholar. Except for Three Dialogues, I use Berkeley's section numbers.

² Winkler, Kenneth P., Berkeley (Oxford: Clarendon Press, 2001), p. 157Google Scholar, and Pappas, George, Berkeley's Thought (Ithaca, NY: Cornell University Press, 2000), pp. 168–78.Google Scholar

³ By “extensional” here I mean that any referentially equivalent description of what I perceive preserves truth-value. If the apple is John's dessert, I perceive John's dessert by perceiving the apple, though knowing nothing about John. Perceiving that an object is an apple is intensional, or referentially opaque. Referentially equivalent descriptions of what I perceive do not necessarily pre- serve the truth-value of sentences in which they occur.

⁴ Winkler, Berkeley, pp. 159-60.

⁵ George Pappas, personal correspondence.

⁶ Call immediate perception of physical objects “IPPO.” Pappas thinks PHK, (pp. 34-35, 84) and TD (p. 168) support the contention that Berkeley countenances IPPO. The passages in PHK, in my opinion, are aimed simply at contrasting what is real from what is imaginary. Although expressed in an Aristotelian mode (evidence of the existence of a glove is that he [Berkeley] sees it, feels it, wears it), TD (p. 168) essentially contrasts substance as substratum with individual substances as congeries of ideas. Gloves are indeed seen and felt, but that is best interpreted as the claim that after association of elements in the congeries named “glove” we do by sight or touch recognize gloves. This is mediate perception. Pappas also quotes PHK (p. 84) where Berkeley writes: “If at table all who were present should see, and smell, and taste, and drink wine, and find the effects of it, with me there should be no doubt of its reality.” Is this, as Pappas believes, an example of IPPO? I doubt it since “all at table” most likely possess and make use of the concept of wine. Perception here is conceptual.

⁷ Berkeley's own discussion of the conventional nature of numerical identity is in TD (pp. 193-95). Daniel Flage stresses the importance of Berkeley's conventionalism about numerical identity in “Berkeley, Individuation, and Physical Objects,” in Individuation and Identity in Early Modern Philosophy, edited by Barber, Kenneth F. and Gracia, Jorge J. E. (Albany, NY: SUNY Press, 1994), pp. 133–55.Google Scholar

⁸ Pappas correctly notes that the prefix “proper” in “proper and immediate” directs attention to immediate perception of sense qualities or ideas. Of course this is not evidence that Berkeley thinks there is immediate perception of physical objects.

⁹ Pappas (Berkeley's Thought, p. 179) faults George Pitcher for not noting the real significance of prefixes like “strictly” before “immediate.” See Pitcher, George, “Berkeley on the Perception of Objects,” Journal of the History of Philosophy, 24, 1 (January, 1986): 99–105.CrossRefGoogle Scholar

¹⁰ There are good discussions of the bundle theory in Loux, Michael J., Metaphysics (London: Routledge, 1998), pp. 97–116Google Scholar; Simons, Peter, “Particulars in Particular Clothing: Three Trope Theories of Substance,” Philosophy and Phenomenological Research, 54, 3 (September, 1994): 553–76CrossRefGoogle Scholar (reprinted in Metaphysics: Contemporary Readings, edited by Steven Hales [Belmont, CA: Wadsworth, 1999], pp. 397-413); and Van Cleve, James, “Three Versions of the Bundle Theory,” Philosophic Studies, 47 (1985): 95–107Google Scholar (reprinted in ibid., pp. 376-84).

¹¹ He also says, “Take away the sensations of softness, moisture, redness, tartness, and you take away the cherry … a cherry is nothing but a congeries of sensible impressions or ideas perceived by various senses” (TD, p. 195). The question becomes whether you can touch, taste, see, or smell a congeries. The sense data mentioned are not strictly ways the cherry appears but elements of the congeries named “cherry.” See discussion in text.

¹² The house example seems an odd illustration of the issue of numerical identity across persons. The problem (posed by Hylas) in the text is that, assuming ideas are possessed by minds, how many can be said to see the numerically same object (TD, p. 193). Whether the house is the same with and without its rooms illustrates rather the puzzle of an object's numerical identity over time.

¹³ Winkler, Berkeley, pp. 158-59. Winkler's attribution to Berkeley of a non-conceptual or non-epistemic take on immediate perception of objects is meant to answer the question (ibid., pp. 159-60).

¹⁴ On the other hand Winker (ibid., p. 160) suggests that for Berkeley we need to perceive enough of the qualities of an object to perceive the object as he quotes from the third dialogue, “When I see, and feel, and taste in sundry certain matters, I am sure the cherry exists, or is real.” However, the passage, as Winkler notes, concerns when, for Berkeley, we can be certain what we immediately perceive exists, a claim consistent with Winkler's view quoted above that perceiving one member of a collection is sufficient to perceive the collection.

¹⁵ Pappas considers the relation member to collection to be distinct from the relation part to whole (Berkeley's Thought, p. 198). See again Flage, “Berkeley, Individuation, and Physical Objects.”

¹⁶ Jackson, Frank, Perception: A Representative Theory (Cambridge: Cambridge University Press, 1977), pp. 6–8Google Scholar, referred to in Winkler, Berkeley, p. 152. In the same way if redness is perceived as the surface of an object, then I do perceive the object by perceiving the surface. But even perceiving a red patch as the surface of an object would seem, in Berkeley's view, to be mediate perception. Pappas uses an example like Jackson’s: “by immediately perceiving some of the parts of the building, one thereby immediately perceives the building” (Berkeley's Thought, p. 198).

¹⁷ There is an interesting exception. I might simultaneously and immediately apprehend all the component sense data that constitute an object, say, if I am hit in the face with a cherry. However, for Berkeley, establishing a group of sense data as a congeries normally involves experiencing each datum seriatim, for example, having visual and later tactile experiences in gaining a concept of distance. I owe the cherry (or perhaps chocolate) example to Richard Glauser from the 2003 Berkeley conference in Rennes.

¹⁸ Pappas, Berkeley's Thought, p. 199 (my emphasis).

¹⁹ I leave the notion of “direct” somewhat loose here, but meaning at least that perception is not mediated by mirrors, telescopes, microscopes, etc.

²⁰ A point noted by Lennon, Thomas M. in his “Berkeley on the Act-Object Distinction,” Dialogue, 40 (2001): 651–67, at p. 661.Google Scholar

²¹ See Berkeley's Correspondence with Samuel Johnson. Johnson asks: “The curious structure of the eye, what can it be more than you admit of, between that and vision?” (Correspondence I, 5).

²² Obviously if God possesses concepts of trees, rocks, gloves, etc., it is not by psychological association. Fortunately, that is not an issue here. One virtue of Berkeley's theory is that it answers a question all bundle theories of objects (or minds) face; what is the glue that binds bundle members together? In Berkeley's conception individual minds, themselves unities, link sensible ideas together, linkages that are the basis of concept formation. Both bundle elements and bundles inhere in—though, are not strictly properties of—minds.

²³ See Loux, Metaphysics, p. 107.

²⁴ See Pappas, Berkeley's Thought, p. 175, and Winkler, Berkeley, p. 151.

²⁵ There are many places where Berkeley identifies ideas and sense data (PHK, pp. 1, 38; TD, pp. 135, 145). In TD (p. 156) Philonous says: “The things, I say, immediately perceived are ideas or, sensations, call them what you will.” Admittedly “idea” and “sensation” are not synonymous for Berkeley. He holds there are three types of ideas, “ideas imprinted on the senses, or else such as are perceived by attending to the passions and operations of the mind, or lastly ideas formed by the help of memory and imagination—either compounding, dividing, or barely representing those originally perceived in the aforesaid ways” (PHK, p. 1). There are indeed places where Berkeley does suggest con- geries themselves are ideas, but I take him in those cases to be speaking loosely. For example Philonous rebuts Hylas's suggestion that ideas are caused by “certain motions in the brain,” by asking, “do you conceive this brain or no? If you do, then you talk of ideas imprinted in an idea causing that same idea, which is absurd” (TD, p. 151). The brain, as all physical objects, is indeed mind dependent. However, given Berkeley's classification of ideas, it is not an idea.

²⁶ To claim knowledge here is obviously too strong, but not an issue I need to deal with.

²⁷ I want to thank participants at the 2003 Rennes conference on Berkeley where a version of this article was read. Also thanks to Steve Hales, Kurt Smith, and George Pappas who made comments on an earlier draft.