Mysticism and Logic, Full Time

[The Mysticism and Logic Halftime report, on Chapters I through VI, is here; what follows in this post, the Full Time review, focuses on Chapters VII through X.] I was snookered. When I embarked on Reading Bertrand Russell thirteen years ago (precisely today), I noted that my embryonic interest in Russell’s oeuvre did not extend to his serious philosophical and mathematical work: it is the public intellectual (and activist), not the professional philosopher/mathematician, that draws me to Russell. The first half of Mysticism and Logic lodged within my comfort zone, sometimes snugly and sometimes with a little effort. But starting in Chapter VII, sense-data and sensibilia and particulars and denotations came to the fore, and I found myself adrift. Having survived this (uh, three-year) storm, I now will go the full cognitive dissonance, and claim (as I believe) that the journey was worth the cost, and that I have emerged from the deluge somewhat cleaner – though ex ante, I might have preferred a hot (or even a cold) shower. (And ex post, I do not find myself running with undo haste towards Our Knowledge of the External World.)

Chapter VII ("The Ultimate Constituents of Matter") began intriguingly enough: object persistence is questionable, our senses are untrustworthy (of course), and a table is similar not to a trombone but to the role that a trombone plays in a symphony, where what is fundamental is the relationships among all the instruments. In the end, Russell provides a sort of dual view of physical objects, one perspective that is appropriate for physics, and a connected one that is appropriate for psychology, but where we need not fear that the psychological view can contaminate the physical view. Further, the psychological perspective challenges our commonsense notion of object permanence, and indicates that sense-data do not require a mind to have a physical presence.

Chapter VIII (“The Relation of Sense-Data to Physics”) doubles down on the abstruse philosophical investigation. The set-up is superb, obvious in retrospect but new to me: We directly experience sense-data, and from these data, and them alone, we attempt to make inferences about the external world. But as we only will ever directly experience the sense-data, we face a problem in knowing when our inferences are correct. What would be ideal would be to do away with inferences, to replace them with constructions – like mathematicians, starting from an imagined concept of irrational numbers, were able to replace (or fortify) their imaginings with a constructed version.

The road that Russell takes beyond the set-up of Chapter VIII is not an easy one, at least for me. It comes to some comfortable conclusions, however – when the appearance of an object changes because I squint, the object itself does not change. (Though if perspectives arbitrarily close to the object did change through squinting, then the object would change: objects are a sort of limit of a multitude of appearances that converge upon them. Squinting changes one appearance, but not the limit of the set of appearances.) Objects are physical, not simply matters of (potentially misleading) sense-data. Hallucinations or dreams (or squinting?) that sever the usual connection between sense-data and objects present a problem, but one that Russell’s approach easily overcomes. As objects are collections of correlated appearances, the problem with hallucinations is not that the sense-data generating them are any less real, but that they lack the usual correlations with other appearances that characterize really-existing matter. Indeed, we need to compare the hallucinator’s reports with those of others before we realize that the hallucination is just that.

Having solved(?) the question of the nature of matter in chapters VII and VIII, Russell tackles causality in Chapter IX ("On the Notion of Cause"). And actually, the solutions are similar, resting on correlations – between appearances emanating from sense-data (or proto sense-data) in chapters VII and VIII, and between antecedents and effects in Chapter IX. These correlations are less than fully reliable, we can have intervening events undoing the usual connection between a “cause” and an “effect” (and, as noted, we can have hallucinations disconnecting one appearance from the usual object). Fortunately, the advance of science does not require any invariant cause-and-effect relationships. Philosophy, too, can thrive with less-than-assured, but time-tested, inductive causal relationships. Science itself essentially invokes a sort of inductive claim, that nature is largely uniform, so that relationships we see at one place or time are likely to exist at other places or times.

The question of the existence of free will is not answered by Russell’s discussion of causation. But Russell’s analysis does suggest that our feeling of the possession of free will is consistent with a deterministic system.

Chapter X ("Knowledge By Acquaintance and Knowledge By Description") applies to the realm of logic our lack of direct experience with everything other than sense data, again addressing the question of what we know. And once again, it is partly through somewhat exceptional cases (like those hallucinations or images produced by squinting) that Russell can test his ideas. Thus we meet “the present King of France [p. 225],” whose lack of existence does not hinder our ability to judge propositions that concern him. [Russell dwelt on the present King of France in multiple venues.] My suspicion is, much as I have struggled to make heads or tails out of Chapter X, that it has become pretty standard fare for mathematically sophisticated logicians.

Chapter X, like much of the second half of Mysticism and Logic, brought to my mind the stanza (number XXVII, it turns out) from the Fitzgerald translation of Omar Khayyam’s Rubaiyat, namely, 

Myself when young did eagerly frequent
 Doctor and Saint, and heard great Argument
 About it and about: but evermore
 Came out by the same Door as in I went.

I am OK with the door, it is a fine door – and maybe, though the door remains unchanged, I nonetheless have slightly been altered? [I recently learned that Russell once titled a book review that he wrote “The Same Door,” drawing on the identical stanza of the Rubaiyat; somehow, this has cheered me. Russell’s review appeared in 1919, one year after Mysticism and Logic.] 

As I (or I 2.0) retreat out of the same door, I will, as always, take along some Russellian nuggets. One is Russell’s motivation (p. 129) to see his first movie: he wanted to test Bergson’s claim that mathematicians look at the world as a sort of film, where what appears to be a permanent object is really a series of closely connected instantaneous images. Also, not yet the third Earl Russell, Bertie slipped in a nice jibe at royalty, in talking about the unnecessary persistence of the law of causation, which survives “like the monarchy, only because it is erroneously believed to do no harm [p. 180].” And Bertie’s point (p. 102) that of course the known laws of nature are quite simple, because otherwise we wouldn’t know them, is surely right, though perhaps in danger of becoming less so: do big data and machine learning promise the uncovering of significantly more complex laws?

In many parts of Mysticism and Logic, I was reminded of Flatland, and the difficulties facing “someone” in a two dimensional space trying to understand three dimensions. Russell is aiming at something similar, to understand the universe while inside it, and only being able to access it through sense-data. (For what sounds to me like a compatible, more modern take, see this article in Quanta.)  But I was also reminded of Flatland when thinking about my own situation, a non-philosopher trying to fully understand Mysticism and Logic! I fear I did better on the mysticism part. Happy 148th Birthday, Bertie.