Mysticism and Logic, Chapter VI

“On Scientific Method in Philosophy,” pages 97-124

People typically are led to philosophical contemplation either by religious/moral concerns (Plato and Hegel, say), or by scientific interests (Hume and Locke, for example) – or by a healthy mixture of these motives (e.g., Aristotle, Kant). “Herbert Spencer, in whose honour we are assembled to-day, would naturally be classed among scientific philosophers…[p. 97].” Nonetheless, religion and ethics are also central to Spencer’s thought, and to his attachment to the notion of evolution.

Despite the creativity spawned through ethical and religious motives, their net impact on philosophy has been negative. Science, at least the kind that is divorced from those same sorts of religious motives, should be the driving force in the future of philosophy. Within the realm of science, however, it is the methods of inquiry, and not the cutting-edge results, that are positioned to be profitably borrowed by philosophers.

The pull of religion and ethics on philosophy is evident in the amount of cogitation on the “universe” and on “good and evil.” All this talk of the universe is a leftover from a pre-Copernican mindset, where the key position of mankind derived from the notion that the earth itself is central to the cosmos. A related vestige is the quick resort to claims of a oneness to the universe. Post-Copernicus, we should recognize “that the apparent oneness of the world is merely the oneness of what is seen by a single spectator or apprehended by a single mind [p. 99].”

[Russell (page 100) initiates the first of two numbered but untitled sections -- largely via a lengthy quoted passage from William James about how the ability to conceive of different universes and to collect them under a common name does not imply any connection between them, any “oneness”.] Nevertheless, there are two smaller unities, one involving that subset of existence experienced by an individual consciousness, and a second concerning the constancy of scientific laws in the small piece of the cosmos known to us. But neither of these unities holds any external legitimacy, permits any valid generalizations beyond their sphere. General laws in physics are unavoidable, as there is a finite amount of particles and thus their complete data forms a sort of (degenerate) general law; “what is surprising in physics is not the existence of general laws, but their extreme simplicity [p. 102].” The rules are so simple that even we can discover them. But again, there’s no reason to expect that those rules remain simple outside of sample; indeed, the rules we have identified must perforce be non-complex, otherwise we would not have discovered them.

More generally, the data that we have collected are not necessarily representative of all that exists -- data are selectively encountered. And those general scientific results we possess are rather infirm, particularly likely to be overturned as we learn more. We must ensure that any philosophical insights that we deduce from scientific results are those that largely will withstand the likely future modifications to those results. We should be wary, for example, of being too dependent on the supposed conservation of energy or mass. Both mass and energy are proving to be more complex than previously thought, and while within the traditional realm of physical sciences the old results are sturdy, out-of-sample generalizations requiring the conservation of some measure of energy or mass are unfounded.

Philosophies developed around evolution, whether older (Hegel, Spencer) or more modern (Pragmatism, Bergson), involve a normative bias, a notion that evolution is progress. [Bergson and progress came up earlier in “Mysticism and Logic”.] The evolutionary philosophies (though not Hegel’s) tend to use biological evolution as their guide. But our biological sample is quite limited, and even within it, the claim of progress (“from the protozoon to the philosopher [p. 106]”) is made by those who think of themselves as, to date, the apex of this supposed upwards movement. More generally, the ethical ideas employed by those whose philosophical inquiries are morally motivated are human-centered, and constitute “an attempt, however veiled, to legislate for the universe on the basis of the present desires of men [p. 107].” Hopes get in the way of facts.

Ethics are really about social affiliation, offering justifications for the actions of the group to which one belongs. The fact that ethical schools support some (seemingly) socially desirable behaviors, such as self-sacrifice, does not undermine their ultimate, action-laundering purpose. The lack of neutrality in ethics is what renders moral considerations unsuitable as a complete basis for philosophical reasoning – even though there is much of practical value in some ethically-based philosophical approaches, such as that of Spinoza. A scientific philosophy must proceed upon facts, not upon hopes for human progress.

[Page 110 begins the second of Russell’s two numbered but untitled sections.] After spurious ideas about the universe and the good are expunged, we see that philosophical propositions must not depend on any specific worlds, but apply to all possible worlds: such are the sort of general propositions found in logic. While such propositions apply to all things separately, they say nothing about any universe, any collection of these separate things. “The philosophy which I wish to advocate may be called logical atomism or absolute pluralism, because, while maintaining that there are many things, it denies that there is a whole composed of those things [p. 111].” Philosophical claims must be a priori, incapable of being proven or disproven through empirical data: arguments built around the path of history, for instance, are not of this nature. In brief, “philosophy is the science of the possible [p. 111, italics BR’s],” or the general.

By this reckoning, philosophy and logic are identical. Logic is general: note the use of variables in stating propositions; further, logic identifies the forms of propositions that can apply to these general facts, it yields “an inventory of possibilities [p. 112].” Perhaps surprisingly, specific areas of inquiry, such as those concerning space and time, are hindered by a lack of understanding of logical forms; that is, particular sciences can still be helped by developments in logic.

Science has made definite progress over the centuries, but the same cannot be said for philosophy. Every philosopher starts out afresh, with new fundamentals, and then the shortcomings in those fundamentals render the entire chain of reasoning to be erroneous: there are no partial truths in this style of philosophy, nothing that can serve as a starting point for later advances. “A scientific philosophy such as I wish to recommend will be piecemeal and tentative like other sciences; above all, it will be able to invent hypotheses which, even if they are not wholly true, will yet remain fruitful after the necessary corrections have been made [p. 113].” Like science, philosophy can then make better and better approximations to the truth. Philosophers should not dream up grand systems; rather, they should look to break current conundrums into smaller questions that individually are susceptible, with the correct logical forms, to solution.

Consider the question of space as promulgated by Kant’s Transcendental Aesthetic [good explanatory lecture pdf here]. Kant’s problem is comprised of three separate questions in different areas of inquiry: logic, physics, and the theory of knowledge. The theory of knowledge issue is the thorny one, the one we are furthest from solving.

The logical problem involves recognizing that what is key about the geometry of space is not so much the underlying axioms that serve as the foundation of a specific geometry, but rather how points in the space can be (partially) ordered by a “betweenness” criterion, where it can be ascertained if point B lies between points A and C. There are many geometries that share the same betweenness criterion, and geometrical reasoning takes place in a strictly logical fashion at this more general level of relations, as opposed to acting on some underlying, and less general, set of axioms.

The physical problem is how we connect real-world objects to mathematical entities such as points and planes – after all, mathematical physics has proven itself to teach us quite a bit about actual objects, even though those objects do not meet the definitions employed in the mathematics. A. N. Whitehead has shown the way, to understand a point, for instance, as the class of all physical objects that contain the point (p. 117). This approach does not require that we assume that objects are made of points, but helps explain how theories based on points nevertheless have real-world relevance.

Kant’s concern about how we can have a priori knowledge of geometry is softened (or eliminated) when we distinguish the logic of geometry from its physical manifestations. We can have a priori knowledge of the logic, but our physical knowledge is synthetic, and only approximates the logical constructs. Kant’s worry about how we can have synthetic, a priori knowledge of geometry is answered by recognizing that we lack such knowledge. We don’t know what happens when we look at actual parallel lines in space, so we shouldn’t act as if we have a priori knowledge of them: those who claim such knowledge have taken a very constrained view of the nature of space.

A similar style of analysis can help clarify the reality of what we perceive and if that reality is independent of the observer. The objects we perceive might be like the supposed noise in a forest, non-existent when not perceived – and there is no way we could tell if this is the case. Independence is as ineffable as reality: we can always identify multiple channels of causation of events, so that in the end, there is only correlation, not causation. “The view which I should wish to advocate is that objects of perception do not persist unchanged at times when they are not perceived, although probably objects more or less resembling them do exist at such times…[p. 123].” [Russell points to his 1914 book, Our Knowledge of the External World, for elaboration.]

So a scientific approach to philosophy requires that we farm out a subset of questions to allied sciences; that we accept that answers to some other questions are beyond human capabilities; and, that with the analytic method of being careful with definitions and decomposing large questions into bite-sized chunks, philosophers can make the slow and steady progress that marks science more generally.