An analysis of the sense-data theories of Moore, Russell, and Broad
Schlagel, Richard Harold
MetadataShow full item record
I When I look at my hand, am I seeing my hand, or am I seeing sense-data of my hand? When one looks at a penny from an oblique angle, does he see an elliptical appearance of the penny, or does he see the actual surface of the penny? From a distance, an object will look small and less differentiated. Does this mean that one is not seeing the object, but seeing instead an appearance of the object? What is the object? and what is the appearance? The qualitative similarity between delusive and veridical perceptions and the apparent incompatibility between the way an object looks under different conditions has led to a distinction between the object, per se, and the appearance. That is, since the way an object looks can be made to vary independently of the object (by changing the lighting, for example), a distinction is made between the variable looks, or appearances. and the object itself. Then, on the view here adopted, either linguistic habits or the demands of ontology encourages treating the appearance as an entity. This, in turn, alters the cognitive status of the physical object, and raises ontological and epistemological questions concerning the status of sense-data and their relation to the physical object and minds. The purpose of this dissertation is to examine this approach to the problem of delusive and varying perceptions to determine whether it is the most adequate approach consistent with the data of perception. The problem is to determine whether varying and delusive perceptions require explanation by a theory of perception incorporating the notion and language of sense-data. The sense-data theories of Moore, Russell, and Broad are taken as test cases. II The dissertation begins with an analysis of Moore's "A Defence of Common Sense," in which Moore does two main things. First, he claims to know with certainty the truth of ordinary propositions expressing fundamental facts (e.g., spatial, temporal, physical, and psychological) about the world and experience. Secondly, analyzing such propositions as "this is a hand," he concludes that such propositions refer to, or have as their "ultimate" subjects, sense-data. Further, Moore believes that there are good reasons for thinking that sense-data are never actual parts of the surface of the physical object, and discusses three theories attempting to explain the relation between the two. Moore finds "very grave difficulties" in each theory. However, since Moore's analysis of ordinary propositions undermines the truth of such propositions, there are also grave difficulties in Moore's position. For if the proposition "this is a hand" does not designate a hand, but refers to sense-data which may not be identical with the hand, then it is difficult to see how the proposition could be true with certainty as it is ordinarily understood. Moore concedes this difficulty in his paradoxical admission that he is "both feeling sure of and doubting the very same proposition at the same time." Thus an analysis of propositions into sense-data leads one to doubt the truth of such propositions, the very point against which Moore wrote his "Defence." A further difficulty in Moore's argument is his definition of sense-data. At first Moore seemed to think that sense-data could be picked out, as any object might, but later he defines sense-data in such a way that "seeing" a physical object logically entails "directly seaing" a sense-datum. A sense-datum is "directly seen" in the way an after-image or hallucinatory datum is seen. However, it is not at all clear that when one is seeing a physical object he is also "directly seeing" an image or datum of some kind, in spite of Moore's definition. III Russell's theory of neutral monism is an attempt to reconstruct the world in terms of sense-data. Russell continues the long line of distinguished British Empiricists which accepts the indubitable data of sense as the foundation for the interpretation of reality. Three fundamental reasons motivate Russell in this attempt. (1) He believes that the similarity of sensory expariences requires accepting all sensed data as equally veridical, and necessitates reinterpreting expariance in terms of such data. (2) He believes that the scientific distinction between primary and secondary qualities also indicates the primacy of sensory experience for a theoretical reconstruction of tha world. (3) He believes that the recant methodological developments in the philosophy of mathematics and in logic point to the elimination of inferred or postulated entities in favor of a more empirical and economical system based on immediate sensory data. However, in spite of its theoretical simplicity, Russell's theory of neutral monism provides little insight into the structure of physical objects and experience. Rather, it offers a reinterpretation of the physical object and experience from which it is extremely difficult, if not impossible, to deduce the common characteristics of both. For example, it is difficult to account for the unity of the physical object on Russell's view that the physical object exists at every place whare it has an effect. In addition, Russell's elimination of inferred entities tends to a depreciation of scientific theories which lead to an understanding and control of physical phenomena not provided by philosophical theories such as Russell's. Furthermore, the scientific distinction betwean primary and secondary qualities should not be taken as evidence against the reality of ordinary objects of perception. That bricks are red and grass is green is not disproved by the physiologist's discovery that color vision depends upon certain chemicals in the cones. Nor does the physicist's description of physical objects as sub-microscopic molecules or atoms prove that perceived macroscopic objects do not have the shapes and properties they are perceived as having, since no theory can change the data of perception. Rather, scientific theories of perception and of matter provide both an explanation and a theoretical model for interpreting ordinary experience which can be verified by perception. IV Broad's acceptance of sense-data follows from his analysis of perceptual situations. According to Broad, since all of the characteristics of a physical bbject are never presented in a single perception, the physical object as such is never "literally contained" in the perceptual situation. What is contained is an "objective constituent" which at most could only be a spatial-temporal part of the physical object. However, Broad concludes that the "objective constituent" can never be identified with the physical object. According to Broad, the necessary and sufficient conditions of perception lie in the immediate area of the percipient. Consequently, the physical object is only an incidental cause in perception. Such evidence as (a) delusive perceptions, (b) the physiological conditions of perception, and (c) the temporal disparity between the emission of light and the perception of distant objects lead Broad to this conclusion. He discusses three philosophical theories, the "Theory of Multiple Inherence," the "Multiple Relation Theory of Appearing," and the "Sensum Theory," which attempt to account for the so-called discrepancies between varying perceptions. A close analysis of the three theories reveals, however, that they do not explain why different perceptions of the same object are not incompatible, but merely restate the facts of perception. As such, they are not as scientific theories, ad hoc explanations of phenomena, but are merely linguistic alternatives. But they are not innocuous alternatives for they imply that ordinary perceptual objects such as chairs and stones do not exist with the properties we perceive them as having. Instead, a perceived object is said to be a colored segment of absolute space, or a sense-datum. The difficulty is that neither segments of space nor sense-data (which are private, transient, and unsubstantial by definition) have the qualities or properties we verify perceived objects as having. Furthermore, science offers explanations as to why different appearances of an object are not incompatible; namely, because the conditions under which the object is perceived are different. There is no incompatibility in an object appearing elliptical from one point of view and round from another when one takes into account the angle of reflection of the light. And the physical object is not incidental to perception since removal of the perceived object eliminates perception of it. V In conclusion, then, delusive and variant perceptions do not require. a philosophical explanation terms of sensedata, but can be explained scientifically in general agreement with common sense. However, such aberrant perceptions do expose the limitations and conventionality even of veridical perceptions, and the language based on such perceptions. The language and point of view of common sense cannot be accepted as an infallible oracle. In fact, their utility depends upon an over-simplification of the conditions of perception. When a quality or property is predicated of an object the conditions facilitating this predication are ignored. However, marginal cases of delusive perceptions expose the conditions of perception, and stimulate scientific explanation. Thus the temporal disparity between the existence and perception of a distant object reveals the function of the emission of light in perception. Color-blindness led to an investigation of the physiological conditions of perception, and microscopic observation introduced other dimensions of perception. Such data exhibit the conditions and limitations of ordinary experience on which ordinary language is based. But they do not require befurcating experience into appearances and reality. Reality, if anything, is perceptually multi-dimensional. This suggests a stratification of languages to avoid pseudo questions and unnecessary epistemic and ontological commitments.
Thesis (Ph.D.)--Boston University
RightsBased on investigation of the BU Libraries' staff, this work is free of known copyright restrictions