Bertrand Russell: Problems of Philosophy Chapter 2 The Existence of Matter

Summary: Russell continues to consider the possible difference between appearance and reality. In the past, we instinctively believed that sense-data were equivalent to objects, but observation, discovery, and dreams have at least at times shown them to be disparate. Given this, we can wonder whether or not there is in the general sense "matter" : something behind the absolutely certain sense data that continues to exist when no one is looking or if no minds were to exist at all. Russell thinks that if we cannot ascertain a matter with independent existence, then we may lapse into solipsism. If others are only known via their bodies, and the independent existence of bodies is doubted, then the existence of other minds also becomes debatable. Russell admits solipsism as a possibility, but he thinks we have no reason to suppose its actuality. Next, Russell reviews the Cartesian methodology of systematic doubting anything not shown to be clearly and distinctly true. While Descartes believes the method leads him to the certainty of his own existence, Russell doubts this. Particular experiences have far more certainty than personal existence. The sensation of a color is far more "a color is being seen" rather than "I am seeing a color." Even though an experiencer may be presupposed, this does necessarily involve any "I." The main payoff of the Cartesian method is to show that primitive thoughts, feelings, and experiences are the most certain, even if the sensations are of illusions.

Can we infer from these primitives the existence of something else? Clearly, our common sense says yes. When we put a tablecloth over a table such that the sense data attributed to the table is hidden from view, we do not believe that the table has disappeared and that a levitating tablecloth remains in its place. However, common sense has often been wrong. We might next refer to our public commerce that involves the supposition of "public neutral objects," objects that are the same for all parties even though the sense data received by each person is privately received. We believe this due to the similarity of sense data being received. Russell, however, believes that this reasoning begs the question because other people are being used to prove that there actually are other people. If we are to make headway, we have to look at our private experiences and show that they tend to show that there are others. Russell doubts that this can be shown certainly (the dream sequence is always a logical possibility), but it can be shown probabilistically. Russell's mainly argues that the hypothesis that there are independent objects is simpler. Several observed regularities such as the motion, the hunger, and language of sets of sense data attributed to objects require less explanation if they belong to real objects. Russell admits this conclusion is weaker than we might hope, but he argues it is typical of philosophical arguments. He believes knowledge is based upon instinctive beliefs, so that if all are rejected, nothing is left. After all, we cannot have reason to reject a belief except on the ground of some other belief. Nonetheless, instinctive beliefs can through philosophical reflection be organized into an hierarchy that forms a harmonious, coherent system that is "most possible." Yet it remains possible that the whole system is mistaken, so a healthy bit of doubt should be applied to all beliefs.

NOTE: To relate Russell's tangent at the end of the chapter back to the case in point, Russell implicitly seems to be asserting that we can compare an explanation asserting the existence of independent objects and one that does not, and that while both may be coherent, the latter will require far more propositions to do so. Since it requires the acceptance of more propositions to explain sense data, it is "less possible," since we can imagine far more possible universes in a system that has less propositions (like the independent object explanation). If we place our bets with the more possible, we have a greater chance of being right if we rolled the cosmic dice. While this may be valid, it is debatable whether or not the non-independent object universe requires more propositions. Furthermore, even if this is the case, it may not give Russell what he wants to say here: that it is more reasonable to believe in what is more possible, and therefore it is more reasonable to believe in a world with matter when it comes to explaining sense data.