First, to think about the value of philosophy, one needs to have some idea what philosophy is. There are two broad ways one might do that. One might describe some paradigmatic examples of philosophy and say that that sort of inquiry is what one means by ‘philosophy’. Or one might try to give an account of how philosophy works, how it is possible, what its method is.
Russell’s chapter on the value of philosophy comes at the end of a book in which he has discussed some central problems of philosophy and so he implicitly uses the first approach. But it is possible to say some general things about his view of the nature of philosophy. It is based on reasoning and argument. It is often concerned with the nature of language and the analysis of concepts. It does not draw on empirical findings: it is generally a priori. So with those characteristics in mind one can go on to ask the value of such a form of inquiry.
But one might want to think a little more about how philosophy is possible in advance of asking that question. If so, one route is to ask how philosophy could tell us anything interesting or substantial whilst being mainly a priori. One answer, popular in the 1930s and still quite attractive, is to start from the division of kinds of truths set out in Hume’s fork (between relations or ideas and matters of fact). In fact there are three intuitive distinctions relevant:
Epistemological: a priori vs a posteriori
Metaphysical: necessary vs contingent
Semantic: analytic vs synthetic
Although these are three quite distinct principles for sorting truths, there is an argument that they sort truths into the same groups. If one knows a truth a priori, one can know it without knowing which possible world we live in. If one can know it without knowing that, it must be because it holds in all possible worlds. Thus a priori truths are also necessary. But how can a truth be true in all possible worlds and knowable without experience? Because it does not depend on the state of the world itself at all: that is it is a truth in virtue of meaning only: analytic.
This argument gives a view of philosophy as a priori, leading to necessary truths and working through conceptual (not empirical) analysis. As a view of philosophy it suffers two defects.
Not all philosophy in the past fits it. (Hume’s appeal to our customs and habits looks a posteriori, synthetic and contingent.) So is that so much the worse for such examples of philosophy or for the account?
The sorts of truth may not align after all. Kant thought that maths was not analytic. (No amount of thinking of the meanings of 7, 5, and plus gives one the concept of 12, he argues.) And some necessary truths (such as that gold has an atomic number of X) are a posteriori.
Still, at least it is a first stab at an account of how philosophy is possible.
Now for the value of philosophy. Russell suggests that philosophy aims at knowledge but largely fails. If the reason for this is his first suggestion: that when there is progress, it is no longer called philosophy, then that is not a worry. We want knowledge and if philosophy gives us that then that is fine, whether or not we continue to call it ‘philosophy’. But he seems to suggest that it is a more fundamental problem and that makes the value of philosophy more problematic.
Russell then gives two broad answers:
The value of philosophy is in its reduction of dogmatism. Philosophy teaches us to doubt and question and that is a helpful thing for generating empirical knowledge. This line of thought suggests that philosophy has an instrumental value in promoting empirical knowledge which itself has an instrumental value through technology.
The value of philosophy lies in our union with the universe which is our highest good. Now this value looks intrinsic. It is simply a good thing to stand in a contemplative union with the universe even if we cannot acquire much knowledge this way. But do we really understand what Russell means by this value? And does it fit the philosophy we typically do?
The first answer looks to be the better one (not that we have to pick; Russell holds them both). But if so, the value of philosophy lies in the thinking and reasoning skills it gives us, skills that we can use elsewhere in other walks of life. We might add that such skills include positive abilities such as to analyse and argue, to thinking things through, to frame one’s thoughts in a logical structure etc.
What of the actual subject matter of philosophy? The philosophy of action, of mind, of science, ethics etc? (we could teach philosophical skills whilst teaching quite different philosophical content. Do the subjects of philosophy matter?) Here I think that there are two kinds of answer leading to two different routes to an instrumental value:
We do make some progress in all these areas. (Russell is overly pessimistic.) There has been a development in the philosophy of science since the 1930s even if the models of science that philosophers now hold are very much more complex and less attractive than earlier simpler ones. Russell’s theory of descriptions really is a useful idea in the philosophy of language etc etc. We have much more spohisticated models of the relation of mind and brain now.
We make no progress but that is not philosophy’s aim. On a therapeutic view of philosophy (cf the later Wittgenstein), philosophical puzzles arise out of misunderstanding our everyday language and thought and trying to explain and simplify it. The proper task of philosophy is then to dissolve philosophical problems not to answer them in their own terms. So there is no progress or advancement of philosophical knowledge because there is nothing to be known. But we can avoid errors and mistakes that people are constantly prone to.
In either model, we are at least less mistaken about science, mind or language through a study of philosophy (either because we have learnt substantial truths; or because we have removed natural confusions to which we are prone) and so philosophy has the value of knowledge in general.