Review

Citation preview

Review: [Untitled] Reviewed Work(s): Locke, Berkeley, Hume: Central Themes. by Jonathan Bennett Charles E. Marks The Philosophical Review, Vol. 83, No. 1. (Jan., 1974), pp. 126-131. Stable URL: http://links.jstor.org/sici?sici=0031-8108%28197401%2983%3A1%3C126%3ALBHCT%3E2.0.CO%3B2-8 The Philosophical Review is currently published by Cornell University.

Your use of the JSTOR archive indicates your acceptance of JSTOR's Terms and Conditions of Use, available at http://www.jstor.org/about/terms.html. JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use. Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www.jstor.org/journals/sageschool.html. Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact [email protected].

http://www.jstor.org Tue Apr 24 02:54:37 2007

BOOK REVIEWS

LOCKE, BERKELEY, HUME: CEJlTRAL THEMES. By JONATHAN BENNETT. New York, Oxford University Press; Oxford, Clarendon Press, 197 I . PP. X, 36 1 . $ I 1. 2 5 (cloth) ; $4.50 (paper). Professor Bennett's book is a series of vigorous engagements with Locke, Berkeley, and Hume on the topics of meaning, objectivity, and causality. The book begins with a discussion of Locke's theories of meaning and classification and Berkeley's criticisms of them. Next, Bennett treats Locke's theory of substance, his "veil-of-perception" doctrine, and his distinction between primary and secondary qualities. Berkeley's attacks on these are assessed as is his idealism. Two chapters are devoted to Berkeley's argument for the existence of God based on the passivity of our perceptions and the doctrine that causes are agents. Hume's theory of meaning, his dichotomy of matters of fact and relations of ideas, his treatment of causaiity, and "Of Scepticism with Regard to the Senses" are the subjects of the last third of the book. Since Bennett's main interest is in the topics of meaning, objectivity, and causality themselves, he attacks, emends, and supplements Locke, Berkeley, and Hume. Often he defends a position of his own. Four representative episodes should convey something of the philosophical flavor of Bennett's book. ( I ) Locke held that "to attach meaning to an utterance is to make it 'stand as a mark' for one or more 'internal conceptions' or 'ideas' in one's own mind, and language's main task is to transfer ideas from one mind to another" (p. I ) . Although Berkeley and Hume differ from Locke on the nature of general ideas and on the necessity of each meaningful word's standing for an idea present in the mind of the speaker, they are essentially Lockean about meaning. Throughout the book, Bennett traces various philosophical errors to this common ground. Bennett's first objection to Locke's theory is that "standing for" is left unexplained. He also suggests that an explanation of meaning in terms of standing for is question-begging, presumably because "stands for" is a close cousin of "means." Although Locke's notion of standing for needs explanation, Bennett's suggestion seems wrong since the relation between a word and the idea it primarily signifies must be unlike any familiar semantic relation. Bennett's second objection is

BOOK REVIEWS

directed against all "synchronous-act" theories of meaning of which Locke's is a species. Suppose that meaning something by what one utters involves doing or undergoing something else while one is speaking. That something else must be an inner activity-that is, one that "does not meet the eye of the unskilled though alert bystander" (p. 5). Since we do not know that our fellow human beings engage in inner acts of the relevant kind, we do not know that they ever mean something by what they say. But we do have such knowledge. So every "synchronous-act" theory is wrong. If a "synchronousact" theory is simply a theory about what meaning is, Bennett's argument does not work. But he assumes that a theory of meaning is an account of the meaning of "meaning." So, if "q is part of the meaning off" entails that it is impossible to know f without knowing q, the argument turns upon Bennett's assumption about the proper concern of a theory of meaning. (2) Bennett contends that Locke's view that secondary qualities are "nothing in the objects themselves but powers to produce various sensations" provides the basis for a correct account of the distinction between primary and secondary qualities. Locke's view is explicated as a compound of (i) the Analytic Thesis, that "any statement attributing a secondary property to an object is equivalent to a counterfactual of the form: 'If x stood in relation R to a normal human being, the human would have a sensory idea of such and such a kind' " (p. 94) and (ii) the Causal Thesis, that "in a perfected and completed science, all our secondary-quality perceptions would be causally explained in terms of the primary qualities of the things we perceive" (p. 102). Bennett argues that the Analytic Thesis is true but the same thesis for primary qualities is false. If we attempt to imagine a case of visual and tactual "size blindness" patterned on cases of color blindness, we find that a "size-blind" man, unlike a color-blind man, could not remain ignorant of his defect if his other sensory capacities were normal. A thing's size and its other primary qualities are connected in many familiar ways with its modes of interacting with other things; a thing's color and its other secondary qualities are not so richly and familiarly connected with its modes of interacting with other things. Such connections as there are between a thing's color, for example, and the ways it interacts with other things do not insure that an otherwise normal man would discover his own color blindness. Since a thing's color is not richly connected with its ways of interacting with other things, we could imagine giving our color terminology a "purely

BOOK REVIEWS

visual" basis if our visual determination of a thing's color persistently conflicted with known truths about colors. If we did so, our color terminology would still do most of the "work" it now does for us. More generally, we can imagine that our sensory determinations of any secondary quality persistently conflict with scientific and commonsense knowledge about that quality. If we choose to give the secondary quality term a purely sensory basis, it would still do most of its present "work." This, I think, is intended to support the Analytic Thesis. O n the other hand, there would be "no point" to giving a primary quality term a purely sensory basis if our sensory determinations of the quality conflicted with what we know of that quality. This shows that the Analytic Thesis cannot be repeated for primary qualities. One immediate difficulty facing this complex argument is that some secondary qualities-for example, temperature (or heat)-are more richly connected with a thing's ways of interacting with other objects than color. Another is that it is not obvious that, in the possible world in which the visual determination of a thing's color conflicts with our scientific account of color, a decision to give our color terminology a purely visual basis would yield true color ascriptions. If it does not, I do not see how Bennett's case supports the Analytic Thesis. (3) Locke and Berkeley believe, as does Bennett, that our ultimate evidence for the way the world is consists of facts about our own sensory states. (Sensory states are the states one is in when he perceives or seems to perceive something.) Could any facts about sensory states be good evidence for the way the world is? Do we have an adequate basis for our belief that there is a n objective realm? Bennett claims that any empirical argument from sensory evidence to a conclusion about the objective realm presupposes "an unquestioned acceptance of the existence of an objective world about which we know a good deal" (p. 67); further, someone who conjectures that perhaps there is no objective world is misusing the distinction between appearance and reality or else is giving it an unordinary sense. Since Locke took such a conjecture at face value and tried to refute it by empirical argument, he was bound to fail. The sole support offered for Bennett's two crucial claims is the fact that our ordinary procedures for determining whether something is the way it appears to us to be rely on our accepting yet other appearances as reliable guides to reality. Locke's theory of perception can be taken as a n attempt to answer the skeptical questions by providing a legitimate means of arguing from facts about our sensory states to facts about the objective realm; so viewed, Bennett labels it the "veil-of-perception" doctrine. Bennett

BOOK REVIEWS

joins Berkeley in rejecting this role for Locke's theory of perception on the ground that the causal inference involved is legitimate only if we have independent access to the putative causes of our sensory states. Bennett also thinks that Berkeley's objection refutes anyone who answers the skeptical questions by asserting that the existence of an objective realm is posited as part of a (mainly causal) explanation of our sensory states. Berkeley's objection, it seems to me, needs to be elaborated and defended. An analogous objection would rule out theoretical entities as causes unless we have access to them independently of inferences from their supposed effects. Both Bennett and Berkeley believe that there is a necessary connection between appearances and reality: "any statement about the objective realm has a meaning which could be expressed in statements about 'ideas' or appearances or sense data" (p. 136). They differ on how the meaning of such a statement is to be expressed. According to Berkeley, objects are collections of ideas. Bennett discusses an impressive array of difficulties for Berkeley's idealism. If his idealism is true, there cannot be unperceived objects, one object cannot be perceived on separate occasions, either one cannot perceive a whole object or objects are not interpersonally perceivable, two ideas belonging to different senses cannot be members of the same object, and an adequate account of the distinction between appearance and reality is impossible. According to Bennett objects are logical constructions out of sense data: "any statement about the objective realm has a meaning- which could be expressed in some set . . . of stat,ements about sense data, . . . a set which will be long and complex and will contain members of the form 'If it were the case that . . . , then suchand-such sense data would be had' " (p. 136). Bennett does not defend his phenomenalism against the familiar arguments to show that the required equivalences are impossible to come by; one would like to know how he handles these and other standard objections to phenomenalism. Berkeley flirted with phenomenalism and, as Bennett argues, his defense of idealism against the charge that it obliterates the distinction between appearance and reality commits him to phenomenalism. Why didn't he accept phenomenalism? Bennett answers that Berkeley's theory of meaning demanded a completion of "A table is . . ." in the language of ideas. Phenomenalism does not yield such a completion. (4) Hume's opponent holds that there is a necessary connection between causally related events. This necessary connection is supposed to entitle one "to say, on a non-inductive basis, things of the form

BOOK REVIEWS

'There is or has been an F ; Fs are necessarily connected with Gs; so there will be a G.' . . . A non-inductive basis . . . is one which supports a prediction without involving any mere assumption about a matter of brute, future fact" (p. 259). Bennett argues that Hume's opponent is "under tremendous conceptual pressure" to maintain that causal laws are logically necessary. Part of the pressure is generated as follows. Consider a necessitarian who holds that necessary connections are postulated to explain observed regularities: the fact that F events are regularly followed by G events is to be explained by postulating that F events have a Gpower. This fails to provide a noninductive basis for prediction if "This event has a G-power" does not entail "This event will be followed by a G event." But if "This event has a G-power" entails "This event will be followed by a G-event," one cannot explain the fact that F events have always been followed by G events by the claim that F events have G-powers. "All 4s are $" is equivalent to "For somef: all 4s are f and being-f entails being- $." So "one could never explain the fact that all 4s are $ by saying that all 4s have some property their possession of which entails their being $. . . . more generally: whatever proposition P may be, one could never explain the fact that P by saying that something is the case which entails that P" (pp. 268-269). Thus Hume's opponent must claim that "All F-events have been followed by G-events" is itself logically necessary. This result, if true, is impressive and startling. Hume's opponent began talking of necessary connections between causally related events and ends up committed to a thesis about the logical necessity of causal laws. However, the argument just given does not force him into this position. Even if Hume's opponent could be forced to hold that 4s have some property which logically guarantees that they are $, once that property is specified, he is not just claiming that 4s have some property or other which entails their being $. Bennett needs to show that one can never do better than claim that Fs have some property or other which logically guarantees that they are followed by Gs. Bennett rejects Hume's argument that, since the negation of every prediction-licensing conditional is conceivable, none are logically necessary; he rejects as well another argument based on the premise that causal laws, unlike necessary truths, are informative. His own argument that prediction-licensing causal laws are not necessary runs as follows. If they were, there must be entailments in which the entailed, but not the entailing, proposition is about the future. The entailing proposition must have a content that is wholly about the present or

BOOK REVIEWS

past and a logical consequence that is about the future. Assuming that necessary truths are analytic, if P entails Q , there is a chain of sentences S,, S,, S,, . . . , S, such that the first sentence in the chain expresses P, each subsequent sentence expresses all or part of the preceding sentence's meaning, and S, expresses Q. Hume's opponent should want an entailment in which the earliest sentence of the chain which is about the future does not express the content of P-that is, would be unacceptable as an oratio obliqua report of P. But this would not satisfy a necessitarian because it differs in degree only from a case in which S, is explicitly about the future. The sentence expressing P "must have a meaning which in no way reaches forward into the future; it must make an absolutely clean cut across the world at a particular time; for otherwise we cannot know that P has come true when the events which verify or falsify Q still lie in the future" (p. 285). If necessary truths are analytic, then P will always have a meaning, if not a content, which reaches into the future. This argument depends upon the necessitarian opponent adopting a principle to the effect that, if P entails Q, one cannot know that P unless he has verified that (2. I am not sure that this principle need be adopted. Bennett's book is throughout rich in philosophical argument; it will appeal even to a reader uninterested in the history of philosophy. Needless to say, it also contains much fine exegesis. For example, the substance of Bennett's earlier papers on Locke and Berkeley is included. There is a good treatment of intuition and demonstration, including a convincing explanation of why Descartes, Locke, and Hume took our knowledge of our own present mental states to be intuitive knowledge. Bennett's chapter, "Hume on Objectivity," will be welcomed by all who have struggled with "Of Scepticism with Regard to the Senses."

E. MARKS CHARLES University o f Washington

A N ANATOMY OF VALUES: PROBLEMS OF PERSONAL AND SOCIAL CHOICE. By CHARLESFRIED. Cambridge, Harvard University Press, 1970. PP. xiv, 265. $ I 1.00. Professor Fried describes, and traces relations among, certain common values. Conjectures are registered regarding origins and grounds (ways in which common values may have biological or psy-