This essay was written for a class, History of Analytic Philosophy Part I, at University of California, Irvine. The course was taught by Professor Jeremy Heis in the Winter Quarter of 2019.

Simplicity and Sense-Data

Bertrand Russell, in The Problems of Philosophy [1912] (TPOP) and in Our Knowledge of The External World as a Field of Scientific Method in Discovery [1914] (OKEW) wrestles with the problem posed by the Early Modern Idealists: Should we really say that there is physical matter in the world which is distinct from our minds?  The meaningfulness of this question depends on the skeptical distance between our access to knowledge of the world caused by the sense-data theory.  The theory claims that when I say that I sense a table in front of me, I’m not really sensing the table, I’m sensing some seemingly intermediary data which seems to relay information from continuously existing table-matter to me.  We only have direct knowledge of logic and sense-data, but not of the physical world, which means we can doubt the existence of material objects.

In both TPOP and OKEW, Russell takes the sense-data theory to be accurate, but he changes his mind about whether we should posit that there exist material objects.  In TPOP, he believes it is simpler to say that physical things exist as a way to explain the consistency of what we call “things” across experience, while in OKEW, Russell thinks the simplest theory is not to posit their existence because he thinks he can describe what we consider a “thing” to be based only on a logical construction of our sense-data (and that doing that is metaphysically simpler than introducing a new metaphysical entity into his theory).  When responding to the skeptical challenge, Russell argues about what we should do in terms of what makes the overall theory of the external world simpler and consistent, and he states in both books that he does not think that we can prove the existence or nonexistence of material things.

~

Let’s begin with The Problems of Philosophy.  In the first four chapters of the book, Russell lays out his reasoning for subscribing to the sense-data theory, and then deals directly with the skepticism of the material world that results from such a theory.

Russell argues for the existence of sense-data in two ways: an argument from illusion, and an argument from perspectival variation.  The former is easily explained through an analogy with dreams, which we take to be sensory experiences of ‘things’ with no corollary in the ‘material’ world; why shouldn’t all of our sensory experiences be like those that we experience in our dream, sensible but with no origins in objectual reality?[1]  The latter comes from the observation that it seems that we can change our sense perception (let’s say, of a table) by walking around it and noticing a change in shape in our visual field, even though what we infer to be the physical table does not seem to have gone under any physical changes.[2]  Russell does not doubt that we have a direct, subjective psychological reality,[3] but believes that these arguments reveal a gap between what is sensed and the object we take to be being sensed; the object might not exist at all if everything is like a dream, and there is a difference between what we sense of the table and the table’s seemingly objective existence that is separate from the arbitrary perspective of observers.

The question Russell then poses is: “Granted that we are certain of our own sense-data, have we any reason for regarding them as signs of the existence of something else, which we can call the physical object?”[4]  Russell sees that there is no logical inconsistency in taking the skeptical stance that “the whole of life is a dream, in which we ourselves create all the objects that come before us.”[5]

However, just because we cannot disprove the skeptical stance, Russell thinks that it is “a less simple hypothesis ... as a means of accounting for the facts of our own life, than the common-sense hypothesis that there really are objects independent of us, whose actions on us causes our sensations.”[6]  This is a revealing sentence of his view: although it is possible to doubt, Russell chooses not to because he values simple theories and believes that doubting is more complex than not.  Positing physical objects as the cause of our sense-data is simpler, he argues, because it allows us to understand how one thing can seem to continue existing outside of our perception long enough to re-enter our perception, and how things can change even when they are not being directly observed by us.[7]  A public physical reality of these objects which is independent of our private sensory experience quickly explains those two observations, and also how different people are able to observe the same thing (albeit from different perspectives).[8]  A lot more work would have to be done to show how certain collections of sense-data are linked across these gaps in experience, and how they can seem to undergo change independent of whether we are observing the or not—thus, it seems like the skeptical route is unnecessarily more complex.

~

What changes in the two years between The Problems of Philosophy and Our Knowledge of The External World is not Russell’s adherence to sense-data, nor his preference for simple theories.  Instead, it seems like the basis for Russell’s new view might be a development in his understanding of what it means for a theory to be 'simple’.  Instead of a theory being more complex if more explanatory work is required, Russell shifts the notion of simplicity to the principle of Occam’s Razor, which asks us not to posit unnecessary metaphysical entities when what we’ve already posited can already explain the observations.[9]

In TPOP, Russell said that observations about apparent consistency across discontinuous sets of sense-data would be “odd” if understood without positing physical objects.[10]  The idea of an independent physical reality to account for these cases seems “natural.”[11]  But even in TPOP, he does not claim that such an explanation is impossible, and going by the definition of ‘simple’ from Occam, the extra work required to define ‘thing’ as a specific set of sense-data would still result in a simpler theory than positing the existence of a physical reality that is separate from that which is sensible.

In short, in the fourth chapter of OKEW, Russell claims that “a ‘thing’ will be defined as a certain series of aspects, namely those which would commonly be said to be of the thing.”[12]  This doesn’t change what is scientifically verifiable, it is only a shift in language “as to avoid an unnecessary metaphysical assumption of permanence.”[13]

Getting to and arguing for a clearer definition of ‘thing’ takes up the rest of that chapter.  That is, what characterizes the aspects that are commonly said of things?  Eventually, we land on the definition: “Things are those series of aspects which obey the laws of physics.”[14]  More specifically, a ‘thing’ in OKEW is the set of continuous transitions which are not distinguished with spatio-temporal gaps, and which follow the laws of physics.

I will briefly outline how Russell comes to that conclusion and supports it. First, Russell notices that continuity might be a useful principle—but continuity not in perception (because the gaps in perception are what we are trying to account for) but in the way it can be described by physics (which can help explain change within the gaps in our perception).[15] This is still not enough, as we still need to explain how we distinguish between one thing and another to apply the causal laws to. Here, Russell goes into arguments that describe points of space and instances in time, which he thinks cannot be sensed directly in sense-data, but can be logically constructed out of sense-data, given that some senses of space can encompass other senses of space (e.g. one part of my field of view contains a smaller, narrower field of view within it) and that events in time can encompass shorter events within them (e.g. the event which is my year at school encompasses this month of March).[16] Russell links the causal laws of physics to the psychological experience of sense-data, and this constitutes his response to the Idealist skeptics in Our Knowledge of The External World.

Footnotes

[1] Russell, The Problems of Philosophy (1912, Reprint, New York: Dover Publications Inc., 1999), 9, 11.

[2] Ibid, 3-4.

[3] Ibid, 11.

[4] Ibid.

[5] Ibid, 13.

[6] Ibid.

[7] Ibid, 13-14.

[8] Ibid, 12.

[9] Russell, Our Knowledge of The External World (London: George Allen & Unwin LTD, 1915), 112.

[10] Russell, The Problems of Philosophy, 13.

[11] Ibid.

[12] Russell, Our Knowledge of The External World, 112.

[13] Ibid.

[14] Ibid, 115-116.

[15] Ibid, 114-115.

[16] Ibid, 119-120.

Bibliography

Russell, Bertrand. Our Knowledge of The External World as a Field of Scientific Method in Discovery. London: George Allen & Unwin LTD, 1915.

Russell, Bertrand. The Problems of Philosophy. 1912. Reprint, New York: Dover Publications Inc., 1999.