1. Philosophy as a discipline

Like Bernard Williams, I think philosophy can be a humanistic discipline and hope to do it that way. Also like him, I don’t think that that is the only thing philosophy can be. The question of what kind of philosophy has most affinity to the humanities will be one main topic of this paper. The other main topic (though reached only at the end) will be how, or whether, the humanities can fit into an idea of science which Williams called ‘the absolute conception’. Here my conclusion will be that they cannot fit.

Before we get to that, however, I want to reflect briefly on some remarks that Williams makes about the nature of philosophy. At the beginning of ‘Philosophy as a Humanistic Discipline’ he disclaims ‘essentialism’ about philosophy. This view, he says, takes philosophy to have

Philosophy, he says,

I agree that ‘making sense of ourselves and of our activities’ is something that philosophy can and should aim to do. So can the other humanities, as Williams says. His phrase is a good way to characterise the ‘wider humanistic enterprise’ as such, the enterprise to which all the humanities contribute. Still, one wants to know how philosophy differs from the others, and further, whether ‘making sense of ourselves’ is philosophy’s only task. What about making sense of the world?

Philosophical ideas are no more timeless than scientific or historical ideas. They develop, they multiply, some of them die. Yet some philosophical questions are timeless, even if they are asked in different ways in different places and times. The disciplines and methods of history and science are dictated by their appropriateness to the questions they pursue; so too philosophy’s disciplines and methods are dictated by the questions it pursues. And there is, as Williams often emphasised, such a thing in philosophy as ‘getting it right’.Footnote ¹ Where there are questions and right or wrong answers a discipline emerges as the natural product of any serious effort to answer them. True, it does not yet follow that philosophy is a distinct discipline. One might hold that philosophy is a discipline but not a distinct discipline; that its disciplinary methods are in fact simply those of science, so that answers to its questions should be pursued in the same way and by the same methods as answers to any other scientific questions. Williams rightly dismisses this as scientism (Williams Reference Williams and Moore2006: 182). Philosophical questions are not scientific questions. But it would be equally unacceptable to hold that the methods of philosophy are simply those of history, so that answers to philosophical questions should be pursued in the same way as answers to any other historical questions. By analogy with ‘scientism’, that would be ‘historicism’.Footnote ²

There is a more serious twentieth-century view, associated with Wittgenstein and the Vienna Circle and not currently fashionable but still challenging: philosophy cannot be a discipline whose essence is determined by its own distinct questions because there are no distinct philosophical questions. Philosophy’s proper task is to eliminate itself, by showing that its supposed questions, the ‘fundamental problems of philosophy’, dissolve. Insofar as there are ‘methods’ in philosophy they simply consist in whatever activity best enables us to see that. In Wittgenstein’s hands at least this is not purely negative therapy. His investigations serve the ancient ascetic goal, which mattered so much to him: spiritual freedom (see Skorupski Reference Skorupsi and Mata2017).

But, contra Wittgenstein, deep philosophical questions arise from resilient tensions that pattern our thought. They cannot be dismissed as factitious linguistic confusions. For there seems to be no perfect way of resolving, let alone dissolving, these tensions – or if there is it has not yet been found. Every attempt at comprehensiveness and overall coherence leaves some difficulty unanswered – a different difficulty for different attempts, even though each attempt goes on being refined and deepened within its limits. So a new question arises: why should this be? If the fly can find no way out of the fly bottle – nor yet just die there – what does that show us about ourselves, our thinking, and its relation to the world? Can there be a unified, stable way of grasping this relation? And if not, what does that mean? These questions are essentially philosophical. Thinking about them is not the only philosophical activity, but awareness of them, and sensitivity to them, belongs to philosophy’s essence.

2. Metaphysical v. critical

Which brings us to philosophy as a humanistic discipline. To be sure, there is an important level at which philosophy’s status as a humanistic discipline is not in doubt. It is a part of a liberal education. It belongs with the humanities, inasmuch as it teaches the disciplines, skills and virtues of a liberal education – honest argument, good judgment, self-awareness, sensitivity to terminology, attention to the meaning of texts etc. It is no accident that philosophy – all philosophy – requires and thus fosters these liberal qualities. There is no intellectual discipline to which they are more important. To describe philosophy as one of the humanities in that classificatory sense should not be controversial.

But we can go further to ask what kind of philosophising displays, or serves, the humanistic interest – ‘making sense of ourselves’. In what follows I shall address that question by discussing three lines of thought that make an appearance in Williams’s paper. Two of them Williams considers rather in passing, somewhat enigmatically. Yet they are important elements in his philosophical thought. There is the idea of science conceived as pursuing (and as the only way to pursue) an ‘absolute conception’ of the world. And there is (in his words) ‘the fundamentally Kantian insight that there simply is no conception of the world which is not conceptualised in some way or another’ (Williams Reference Williams and Moore2006: 185). The third line of thought is the most important of all: it is the idea of ‘making sense’ itself, which plays a role throughout Williams’s paper. I shall connect it with another Kantian idea that has had great influence in the humanities and social sciences, the idea of an ‘intelligible standpoint’ which we can take on ourselves and others. And I shall ask how the intelligible standpoint fits with the absolute conception of the world.

As a first step it will be helpful to keep in mind two broad kinds of philosophical interest that are often in tension. They are by no means the whole of philosophy; they are neither exhaustive nor exclusive. However each one is a significant current presence. I shall call them the Metaphysical and the Critical. My suggestion will be that philosophy, when it is undertaken as part of Williams’s humanistic project of ‘making sense of ourselves and our activities’, is characteristically aligned with the Critical interest. I do not suggest that philosophy’s ‘true’ vocation is to express or vindicate humanism, or that this is its only task. How important the Critical interest can or should be partly depends on the times, and is mainly itself a philosophical question. To me it is very important. But it’s quite possible to pursue significant metaphysical paths from which the concerns of humanism seem no more than a distraction.

The Metaphysical interest focuses on the world and its structure, on what absolutely exists and how. It is cosmocentric, and/or theocentric. Its method is a priori. If, from this standpoint, there are to be objective norms, in particular epistemic norms, at all (as against revelations of faith), they too must somehow be (as it is put) in the world. On this approach ontology, the inquiry into being, is primary. Epistemological issues are secondary. Giving metaphysics primacy need not flow from a religious stance; however a religious stance may well take metaphysics rather than epistemology as fundamental, as it has done in the past. Indeed any ambitious metaphysics of reality as a whole, whether or not it flows out of religion, may look askance at the limited, anthropocentric, interests of mere humanism.

The Critical approach rejects the primacy of the ontological. It focuses on the subject, (individual or social) and on what the subject can know, whether of the world, of the realm of norms, or of itself. It seeks to make sense, to produce clarity of understanding. The contrast, it may be said, is already implicit in Socrates and Plato. However in modern times an especially important turning point is Kant’s idea of a critique of reason. Kant puts before us the idea that a critique of reason – an exploration of its powers and its limits – must inevitably lead to rejection of metaphysics. And in a characteristically humanistic development, he sees critique as emancipatory.Footnote ³ The continuous rethinking of the basic anti-metaphysical idea (ontology subordinate to epistemology) constitutes the Critical tradition in philosophy; I take it to be broader than Kantianism as such, so that the Vienna Circle or Wittgenstein can be read as Critical. Critique denies that knowledge of reality as it is in itself can be acquired by a priori thinking. Crucially, it adopts some version of Kant’s ‘Copernicanism’ – his thesis that knowledge is possible for us only if its objects conform to our concepts.Footnote ⁴

3. Williams and Putnam

Given this broad framework, can we see Williams as a representative of the Critical stance in philosophy? In ‘Philosophy as a Humanistic Discipline’ he pursues a debate with Hilary Putnam which throws some light on this.Footnote ⁵ Putnam had suggested that Williams ‘views physics as giving us the ultimate metaphysical truth’. He based his suggestion on Williams’s well-known suggestion that science aims to give us ‘an absolute conception of the world’. In reply, Williams sharply rejects Putnam’s metaphysical reading of the absolute conception:

He continues,

Well, it can hardly be denied that Williams’s ‘jargon’: the ‘absolute conception’, ‘the world as it is in itself’, ‘anyway’, sounds metaphysical. If we put aside metaphysics, however, as Williams tells us to do, what point is made by the absolute conception?

A first answer is that it expresses a realism about science, but a realism that is empirical rather than metaphysical. It takes science to be an empirical investigation into the nature of things, with at least the aim of being a fully comprehensive account. It accepts that it legitimately uses any of the normal tools of empirical inquiry, and it asserts that these tools are tools of genuine discovery. But it denies that science gives ‘ultimate metaphysical truth’, or at least it makes no such claim.

This contrast is also found in Kant, in his distinction between ‘empirical’ and ‘transcendental realism’. As is familiar, Kant claims to save the former by rejecting the latter. He agrees that natural science pursues an inquiry into how things empirically are. And he holds that there is no metaphysical route that can substitute for, go beyond, or set limits on empirical inquiry. On this he and Williams agree.

Williams further agrees with Kant, indeed emphasises, that any representation of the world must be couched in some concepts, concepts which belong to the cognitive community that constructs it. Highlighting this insight, he somewhat surprisingly says, was actually the point of introducing the idea of an absolute conception in the first place:

Obviously, if this insight is to be more than a tautology the emphasis must lie on the words ‘conceptualised in some way or another’. The point, if I am not mistaken, is to leave open the possibility that the world could be conceptualised in more than one way. True, there are no a priori limits on what ‘our’ empirical science can discover about the world – how complex and detailed its representation of the world can become – nonetheless, though the representation is open-ended it remains our representation, conceptualised by us. We cannot exclude the possibility of other representations, developed from a framework of different concepts. Furthermore, for all we know, some of these representations could tie in first place for simplicity and predictive accuracy – and this seems to raise the question of what the world is really like. Kant would say that the question is intelligible, but unanswerable. Some twentieth-century philosophers e.g. Carnap – and Williams? – would in contrast say that this use of the word ‘really’ has no sense.Footnote ⁷ What is real is an intra-theoretical not an extra-theoretical question. Both these views of the question are ‘Critical’, as I understand the word, insofar as (i) they endorse the anthropocentric ‘Kantian insight’ and (ii) they reject metaphysical as against empirical realism about science (what Kant called ‘transcendental realism’). Williams, far from being a metaphysical realist about science, seems to agree with (i) and (ii). To that extent he takes the Critical stance. This certainly does not settle the question of what his response to Kant’s ‘Copernicanism’ was or should have been – what he took the significance of the ‘fundamentally Kantian insight’ to be. But let us put that aside for the moment. The question I want to pursue now is about the affinity between Critical philosophy and humanism.

4. Intelligible explanation

One could argue that the Critical attitude grows out of the humanistic interest. This is the interest in making ourselves and others intelligible. I have in mind the specific sense of that word Kant introduces when he distinguishes two possible standpoints on human beings – the ‘empirical’ and the ‘intelligible’:

Kant deploys this contrast between empirical (we could say natural-scientific) and intelligible standpoints in a striking account of freedom. He introduces a ‘causality of freedom’ which operates in any autonomous choice. The causality of freedom is distinct from natural causality; it operates at the noumenal, not the empirical level. Freedom as autonomy, it thus seems, is possible only if autonomy is explicable in a certain very distinctive way, which can apply only to a transcendental subject.

In contrast, I believe that Kant’s concept of freedom as autonomy stands firm without any appeal to transcendental idealism.Footnote ⁸ The ‘laws’ of the intelligible world are indeed ‘independent of nature’ and ‘have their foundation merely in reason’. That is perfectly true, inasmuch as they consist in nothing other than a priori truths about reasons. To act autonomously is to act from awareness (at some imputable level) of some particular normative truths.

We become aware of their truth because they are true. This ‘because’ is at the heart of the ‘causality of freedom’. Kant’s use of the term ‘causality’, however, is misleading. The relation he refers to in this way, namely, the relation between a normative truth and its recognition by an agent, is certainly explanatory, in that if it obtains the agent’s judgment is free of ‘determination by alien causes’Footnote ⁹ – sources other than recognition of normative truth. But it is not causal.

Suppose, to illustrate, that I become aware that someone is in difficulty and should be helped. My thought includes awareness of a normative truth: that the person should be helped. That normative truth does not cause my awareness of it in the way that the cup on the table causes (is one causal antecedent of) my awareness of the cup. Yet I am conscious of the normative truth and my consciousness of it can cause me to act – in this case to provide the necessary help. The normative truth itself, unlike awareness of its truth, has no causal power – nevertheless its truth explains my awareness of its truth and thereby plays an essential role in what we can call intelligible explanation. Awareness of and response to normative reasons is the so-called ‘causality of freedom’ to which Kant draws attention. We should no doubt be wary of the phrase. But we should not reject the insight.

Kant distinguishes further between the receptivity of sensible awareness and the spontaneity of thinking – the spontaneity of normative awareness.Footnote ¹⁰ My awareness of the cup requires receptors of some kind – sight, or touch – that inform me of it; but my awareness of the purely normative truth requires no such thing. In this latter case awareness is purely spontaneous. Clearly this raises the question of how spontaneity without receptivity can yield awareness of truth. It is the task of Critical philosophy to address this question without positing human powers that human beings do not and could not have.

5. Intelligible explanation (cont.)

To avoid misunderstanding we should note a few more things about the ‘intelligible standpoint’. First of all, intelligible and causal explanation of course interact. A particular episode of normative awareness, for example, may be caused by my being asked whether something I hadn’t previously considered is true, and my making an effort to work out what I think. If I become aware, as a result of this effort of attention, of the truth of some purely normative proposition N, there is a causal sequence from my thinking about N to my coming to believe that it’s true that N: if I hadn’t thought about N I wouldn’t have come to believe that N. But intelligibility is also present: I come to believe that N because it is true that N and I can see that it is. Causal and intelligible explanations are both present if we say ‘He thought about N and realised that it was true’. There is a causal sequence of reflections and at the same time there is an explanatory but acausal relation between N and awareness of the truth of N.

In a different kind of case – testimony – N is true, and I accept it, but not (sticking to Kant’s sense of the word) spontaneously. We could not say, of belief based on testimony, ‘he believes it because it is true’, as against ‘he believes it because he was told by people he trusts’. In other cases, again, I am influenced by ‘alien causes’ such as indoctrination, wishful thinking, confusion, etc. What I believe in these cases may be true, but I have no sufficient reason to believe it. Still, in the elementary case, in which my awareness of normative truth is spontaneous, what explains and justifies the awareness simply is its truth. Normative truth is the prime mover of autonomous activity.

To be sure, in many cases of human behaviour there may be no intelligible explanation to be found. In some of these cases intelligible explanation still plays a secondary role. The actors may have thought they were acting intelligibly, in which case explaining why they thought that becomes a secondary application of the intelligible model: we look to understand more about their factual or normative beliefs, and their normative setting. In other cases, however, intelligible explanation simply lapses. We know what was done, or believed, but looking for the actors’ reasons, including their beliefs about their reasons, goes nowhere or is just clearly beside the point.

Still, finding the intelligible explanation, if there is one, is the proper aim of the intelligible standpoint. Intelligible explanation is the ideal form of ‘making sense’: of understanding others and of understanding oneself, including making sense of one’s own thinking. That interest in following through, if possible to a full normative clarification of what some actors did – i.e. of the normative adequacy of their aims and their means – is characteristic of intelligible explanation, though there is no guarantee of success. We want to understand, so far as possible, why people, or some particular people, did what they did and why it made sense, at least to them. That includes the causal factors that influenced them, the content and scope of their beliefs, the consequences they expected, and so on. Intelligible explanation blends naturally with causal explanation. But beyond causal explanation we want to understand an action and its motives by grasping for ourselves its rationale. (Why did the First World War break out? What were the reasons motivating the various relevant actors, and do we think they ‘made sense’? If intelligibility is completely or partially absent, as is often the case, and perhaps in this case, the question does not get a fully satisfactory answer and rival interpretations continue.)

Full intelligibility requires that the reasons imputed to the actor align with normative truth as understood by the observer. Alignment is not sufficient, in that it may be alignment on reasons which are actually bad reasons; and it may be absent through inadequate normative understanding on the part of the observer rather than the actor, so that the former, not the latter, needs to change. One of the things anthropology, history, and literature can do is to expand our own understanding by uncovering good reasons which we had not thought of or significances we had not grasped. The drive to make sense has limits; the most imaginative, penetrating attempt to make an actor or a practice intelligible may fail. That said, however, the intelligible standpoint – seeing people as responsive to reasons – is the default standpoint of ordinary discourse, the standpoint within which we understand and communicate with each other. Which may be why the goal of finding an intelligible explanation, as well as appealing to sheer curiosity, has a kind of ethical force.

6. The ‘absolute conception’ and intelligible explanation

We can now go back to the three elements in Williams’s paper which we identified for discussion earlier. Most fundamentally there is the aim of intelligible explanation which we have just considered, and which provides the indispensable foundation of the humanities (‘making sense of ourselves and our activities’). A second element is the ‘Kantian insight’ (we form our concepts) with its suggestion of Kantian Copernicanism. And thirdly, there is the empirical ideal of science as the route to an absolute conception of the world.

How these elements fit together raises various questions. The particular question I want to ask is this. If making sense of ourselves and our activities requires taking the intelligible standpoint and seeking the understanding of reasons that that standpoint calls for, how does that fit with the absolute conception, or at least sit coherently alongside it?

It may be said that at least in the ‘neo-Kantian’ way in which I have set up the question – by reference to intelligible explanation – it does not exactly face Williams. The account that I have given of the intelligible standpoint may seem to differ from the account one would expect Williams to adopt, in light of his influential ideas about internal and external reasons. Isn’t Williams’s view more like Hume’s than like Kant’s, in that it leaves no room for the idea that the agent’s awareness of normative truth can be an inherent element in intelligible explanation of his or her response – that normative truth itself can be present in the explanans?

I have argued elsewhere that the central premise in Williams’s account of reasons is best seen not in terms of Hume’s theory of motivation but rather as the principle that a fact is a reason for a person only if that person has the capacity to see for themselves that it is. If a person simply cannot see that some circumstance constitutes a reason to act in some way it is nugatory to insist that it is, nevertheless, a reason for him to act in that way. One could call this the requirement or presupposition of agential insight. It is an internalism about reasons with which the account of intelligible explanation that I have given could agree: you can’t claim that an agent has a reason for action by invoking a normative truth whose truth the agent simply cannot see.Footnote ¹¹ So I don’t think it takes us too far away from Williams to discuss our question in terms of this reading. In any case the question is worth considering.

To return, then, to the compatibility of the absolute conception and the intelligible standpoint. The absolute conception envisages a comprehensive theory of the world, structured by natural causality alone, expressed in a uniform and universal vocabulary that can describe all the facts. Its language, many now think, is the language of physics, as it presently stands or as it will immanently develop in future. In contrast, we have seen, intelligible explanation is not reducible to a view whose sole explanatory category is natural causality. It makes explanatory yet non-causal appeal to normative truths grasped as such, in belief, feeling or action. An immediate worry might therefore be, that unless these normative truths fall within the scope of the absolute conception they undermine its ambition to cover all the facts. This worry, however, overlooks the slipperiness of the word ‘fact’. We use the word in a nominal sense in which ‘the fact that p’ refers to the truth of the proposition that p. We also use it in a substantive sense, the sense in which a fact can have causal powers, known or unknown. (As in ‘The fact that the tyres had perished was the cause of the accident’.) Normative propositions have truth value, in the same sense and under the same limitations as any other proposition has truth value, but they are not propositions about substantive facts. Thus the fact (nominal sense!) that they have truth value is no threat to the ideal comprehensiveness of science.Footnote ¹²

However, intelligible explanation deals not just in normative truths but also in intentional states directed to normative truths. Intentional states are substantive facts with causal power. On the one hand their content is normative, on the other hand they have a causal role. They interact causally with non-intentional states. They don’t just belong in a self-contained domain of discourse with which the absolute conception has no concern.

Given that intentional states are part of the causal network, it follows that they are part of what there is. If so, then on the absolute conception they should be describable in the language of a scientific theory that uniformly describes what there really is. If we take that language to be the language of physics, then prima facie every intentional state must be identical to some physical state: this will be part of the empirical conjecture. If it is found to be a plausible hypothesis it implies no ontological conflict with the absolute conception.

But suppose that, as neuroscience progresses, no ontological identity of intentional and physical states becomes plausible. Would that undermine the intelligible standpoint?

I think it’s clear that this standpoint is one we will never give up. It flourishes in its own right – it is ‘there’, in all its depth, complexity, and inevitability, as the conceptual framework in terms of which we ‘make sense of ourselves and our activities’. All thinking belongs in the intelligible domain; and that domain, the domain of reasons, is in perfectly good epistemic order.

If no physicalist characterisation becomes empirically plausible, even in principle, we may seem propelled to dualism, with all its difficulties. But there is an alternative: some version of ontological relativity, according to which physicalism and the intelligible standpoint are two conceptual schemes that can co-exist, even though the latter is not ontologically reducible to the former. On this view Williams’s idea of an absolute conception can retain its potential comprehensiveness as an account (in terms of its own concepts) of what there is, but not its claim to be, so to speak, conceptually absolute. We operate with two conceptual schemes, whose distinct theoretical entities in each case earn their title to existence by their explanatory power – and in the intelligible case, their inevitability as the medium of our thought. Neither scheme has ontological primacy. This, I suggest, is the Critical stance most suited to Williams’s philosophy.