Hinge commitments and trust

Abstract

According to the innovative account of the structure of rational evaluation offered by Wittgenstein in his final notebooks, published as On Certainty, our rational practices necessarily presuppose arational hinge commitments. These are everyday, apparently mundane, commitments that we are optimally certain of, but which in virtue of the ‘hinge’ role that they play in our rational practices cannot themselves enjoy rational support. Granted that there are such hinge commitments, what is the nature of the propositional attitude in play? Many commentators have described this propositional attitude as a kind of trusting, on account of how our hinge commitments are effectively a groundless kind of presupposition. In contrast, I want to push back against this way of thinking about hinge commitments and argue instead that it is crucial to our understanding of Wittgenstein’s proposal especially in terms of its implications for radical scepticism to realize that hinge commitments are not presuppositions and that the hinge propositional attitude is not one of trusting.

1 Introductory remarks

In his final notebooks, published as On Certainty (= OC), Wittgenstein (1969) persuasively describes how certain basic commitments function like ‘hinges’ that stand fast in our practices of rational evaluation. Since they form the framework for rational evaluation, these hinge commitments are themselves immune to rational evaluation. Even though our rational practices function against this background framework of arational hinge certainty, our ordinary beliefs are nonetheless rationally in order as they are. Amongst other things, this radical account of the structure of rational evaluation is meant to help us resolve the sceptical problematic by enabling us to reject the faulty philosophical picture that is generating this puzzle.

The question that concerns us, however, is how we are to understand the distinctive nature of our certainty in these hinge commitments, given that it is essentially arational. Many commentators have suggested that we should think of this propositional attitude as being a kind of (propositional) trust, a trust that is reasonable even if not straightforwardly rational.¹ This is because they regard our hinge commitments as a kind of groundless presupposition. I want to push back on this idea. Properly understood, hinge commitments are not presuppositions and the kind of propositional attitude involved in these commitments is very different from trust. I also explain why this exegetical point is important: we miss out something very interesting in Wittgenstein’s proposal by casting it in terms of trusting groundless presuppositions, not least in terms of misunderstanding the nature of his response to the problem of radical scepticism.

2 Wittgenstein on hinge commitments in outline

Let’s start by outlining the main features of Wittgenstein’s account of hinge commitments. Wittgenstein’s focus is the kind of everyday commonsense certainties that Moore (1925; 1939) and others have noted, such as that (in normal circumstances) one has hands (e.g., OC, § 1), what one’s name is (OC, § 425), and the language that one is speaking (OC, § 158).² These are the kinds of commonsense claims that one is ordinarily optimally certain of.³ But what epistemically underwrites this certainty? Wittgenstein’s answer is quite extraordinary at first blush, in that he claims that this certainty is by its nature arational. The reason for this is that Wittgenstein maintains that it is part of the very structure of rational evaluation that it presupposes a backdrop of certainty that enables rational evaluations to take place. As he puts it in a famous set of passages (OC, §§ 341 − 44), this backdrop of certainty is the ‘hinge’ on which the activity of offering rational evaluations turns. Given that this framework of certainty is required in order for rational evaluation to occur, it cannot be itself rationally evaluated. This is why the specific hinge commitments that manifest this overarching backdrop of certainty our commonsense commitments to various mundane claims about one’s hands, the language one is speaking, one’s name, and so on are thus arational. It follows that at “the foundation of well-founded belief lies belief that is not founded.” (OC, § 253)

This proposal is so striking because one would naturally suppose that claims that one is optimally certain of should possess a high degree of rational support, at least if that certainty is well-grounded anyway. Conversely, to learn that these everyday certainties are arational might seem to suggest that one’s system of beliefs as a whole lacks a rational grounding. After all, if in normal circumstances one’s commitment to the proposition that one has hands lacks a rational grounding, then how are any of one’s beliefs rationally grounded? Allowing that our system of rational evaluation has arational hinge commitments at its heart seems to lead directly to radical scepticism, whereby none of our beliefs enjoys an adequate rational grounding.

Wittgenstein’s claim, however, is that recognising that our everyday certainties are arational hinge commitments is key to resolving the sceptical problematic. The general contour of the anti-sceptical thought in play—we will be getting into the details in due course—is that we misdescribe our epistemic practices by supposing that our commitments can be rationally grounded en masse. Instead, when we examine our ordinary epistemic practices more closely and therefore recognise the essential role that arational hinge commitments play in these practices—we realise that the philosophical picture of the structure of rational evaluation that the sceptic (and, for that matter, the traditional anti-sceptic) is working with is untenable. In particular, Wittgenstein’s claim is that it isn’t a merely contingent matter that our rational practices are structured like this as if we could dispense with our hinge commitments if only we were smart enough (or more thorough, say) but rather that it is in the nature of rational evaluation that it takes place relative to a backdrop of arational hinge certainty.^4,5

3 Hinge commitments and trust

We will say more about Wittgenstein’s account of hinge commitments below, but this outline will suffice for us to understand why some commentators might characterise the propositional attitude involved in our hinge commitments in terms of trust. For aren’t one’s hinge commitments essentially presuppositions or assumptions the obtaining of which, since they necessarily lack grounds, is simply down to trust? After all, if one can have no rational basis for their truth (and it’s possible to recognise this fact), then it is hard to see how they could count as beliefs, since to believe something is to believe it to be true, and yet one has no basis for the truth of what is being committed to when it comes to one’s hinge commitments. It follows that our hinge commitments look like presuppositions (or something in that general vein) that we are obliged to trust.

One finds interpretations of hinge commitments along these lines in many places, but perhaps the most influential version of this tendency is found in the work of Crispin Wright. Here, for example, is how Wright glosses the key upshot of Wittgenstein’s appeal to hinge commitments in On Certainty:

I suggest that the principal message of On Certainty is that scepticism embodies an insight which Moore missed: the insight that to be a rational agent, reflectively pursuing any form of cognitive enquiry, means placing trust in suppositions which—at least on the occasion—are not themselves the fruits of such enquiry and are therefore not known. (Wright, 2004a, 305).

Here we get the three essential ingredients of the kind of construal of On Certainty that I have in mind. First, that radical scepticism reveals an important truth (an ‘insight’) about our cognitive limitations, such that we simply cannot discharge all of our presuppositions. Second, that our hinge commitments constitute fundamental presuppositions or assumptions (in this passage: ‘suppositions’) that we cannot discharge. And third, that we must therefore regard the propositional attitude involved in our hinge commitments as being essentially one of trust. On the third point, Wright is also explicit that we do not believe our hinge commitments, on account of their essentially arational nature.⁶ This propositional attitude of trust is thus one that excludes belief. This indicates that Wright is following philosophical orthodoxy in treating trust as more than just a bare reliance on the truth of the target proposition, as trust in this thin sense would obviously be compatible with belief.⁷ In particular, he is concerned with the specific kind of reliance that is involved in being groundlessly committed to a presupposition, a type of commitment that is disconnected to reasons and truth in ways that belief cannot be. (We will be revisiting the notion of belief in play here below).

Given how Wright is understanding our hinge commitments whereby the sceptic has exposed our cognitive limitations, limitations that require us to groundlessly trust our most fundamental presuppositions it thus becomes essential to show that our trusting of them is somehow reasonable even while being strictly arational. The alternative would be that the view simply collapses into radical scepticism. The radical sceptic also contends, after all, that at the heart of our practices is a groundless trust in our presuppositions that we cannot overcome due to our cognitive limitations. But the sceptical point is that this reveals that none of our beliefs enjoy an adequate rational standing.

Wright aims to show that our trust in our hinge commitments is reasonable by appealing to the notion of entitlement. As he puts it:

Trusting without evidence can still be rational or not. Entitlements are warrants to trust […]. (Wright, 2004b, 204, italics in the original).

The thought is thus that we shouldn’t concede to the radical sceptic that since the trust that attaches to our hinge commitments is groundless that it is therefore unreasonable. There is, rather, a particular kind of groundless trusting of one’s presuppositions that can be rational, and where that applies one has an entitlement to this propositional attitude.

The basic idea behind Wright’s entitlement proposal is that where a commitment is required in order for one to be a rational agent at all and where there are no specific reasons to doubt the proposition that is subject to this commitment (which there obviously won’t be in the case of hinge commitments) one is entitled to trust this commitment. In short, trusting is in these conditions reasonable, even if it is not supported by reasons that indicate the truth of what is being presupposed in the target commitment.

Consider this passage, which is worth quoting in full since Wright not only sets out the entitlement framework but also in the process draws an illuminating parallel between our epistemological predicament and the Serenity Prayer:

[The moral …] to be taken from On Certainty is that the concept of warranted belief only gets substance within a framework in which it is recognised that all rational thought and agency involves ineliminable elements of blind trust. Since rational thought and agency are not an optional aspect of our lives, we are entitled—save when there is specific evidence to the contrary—to make the presuppositions that need to be made in living out our conception of the kind of world we inhabit and the kinds of cognitive powers we possess.

To be entitled to accept a proposition in this way, of course, has no direct connection with.

the likelihood of its truth. We are rationally entitled to proceed on a basis of trust merely because (or when) there is no extant reason to believe it is misplaced and because, unless we do so, we cannot proceed at all. An epistemological standpoint which falls back on a conception of entitlement of this kind for the last word against scepticism needs its own version of (what is sometimes called) the Serenity Prayer: in ordinary enquiry, we must hope to be granted the discipline to take responsibility for what we can be responsible, the trust to accept what we must merely presuppose, and the wisdom to know the difference. (Wright, 2004a, 305).

As the appeal to the Serenity Prayer makes explicit, Wright conceives of the Wittgensteinian way of responding to radical scepticism as being essentially concessive, in that it exposes hitherto unrealised cognitive limits that we need to work within as rational subjects. We might have hoped to have shown that our system of commitments doesn’t simply rest, ultimately, on trusting groundless presuppositions (“blind trust”), but since this is not in the offing one must present an account of the structure of rational evaluation that embraces this feature. Here is Wright articulating this concessionary aspect of his Wittgensteinian response to radical scepticism:

[The Wittgensteinian line …] concedes that the best sceptical arguments have something to teach us that the limits of justification they bring out are genuine and essential but then replies that, just for that reason, cognitive achievement must be reckoned to take place within such limits. The attempt to surpass them would result not in an increase in rigour or solidity but merely in cognitive paralysis. (Wright, 2004b, 191, italics in original).

Since radical sceptical arguments teach us something, they can’t simply rest on a mistake, as in that case we could reject them wholesale. Indeed, Wright is elsewhere quite explicit that aiming to respond to the sceptical argument by showing that it rests on a mistake is a hopeless endeavour:

But there is no disguising the fact that the exercise comes as one of damage limitation. That will disappoint those who hanker after a demonstration that there was all along, actually, no real damage to limit that the sceptical arguments involve mistakes. Good luck to all philosophers who quest for such a demonstration. (Wright, 2004b, 206-7, italics in original).

The upshot is that while we might have antecedently hoped to be able to establish a much stronger anti-sceptical conclusion, one that actually shows that the radical sceptic is guilty of error, all we can in fact secure is the much more modest claim that there is a reasonable way of forming one’s beliefs even granted the essential cognitive limits that the radical sceptic has (rightly) exposed.

As I’ve previously noted, Wright is certainly not the only commentator to describe our hinge commitments in terms of trusting our presuppositions. Even commentators who are critical of Wright’s entitlement reading will nonetheless follow him in characterising our hinge commitments in this fashion.⁸ As I hope the foregoing makes clear, this should not be surprising, as this is a very natural way of interpreting the notion of a hinge commitment. Nonetheless, I want to suggest that this way of thinking about hinge commitments should be rejected. Our hinge commitments are not presuppositions (or assumptions, hypotheses, suppositions, etc.,), and our propositional attitude towards them is not one of trust. As I hope to show, understanding why this is the case is crucial to getting clear about what Wittgenstein’s response to radical scepticism is meant to be.

4 Hinge commitments: a closer look

In evaluating whether hinge commitments should be understood in terms of trust, a good starting-point is that Wittgenstein is quite clear at a number of junctures in the On Certainty notebooks that our hinge commitments are not to be thought of as assumptions or presuppositions (or anything in the vicinity of these propositional attitudes, like hypotheses, suppositions, and so on). Indeed, he even makes this point in the very passage where he introduces the hinge metaphor:

But it isn’t that the situation is like this: We just can’t investigate everything, and for that reason we are forced to rest content with assumption. If I want the door to turn, the hinges must stay put. (OC, § 343)

Wittgenstein’s point here is quite emphatic and stands directly opposed to the claim that we saw Wright advancing in the last section, whereby our hinge commitments are effectively assumptions that are forced upon us.

More generally, throughout the notebooks that make up On Certainty Wittgenstein is keen to stress that the propositional attitude involved in our hinge commitments is of a highly distinctive kind, and certainly very different to trust. He emphases the “primitive”, “animal”, visceral nature of the certainty we have in our hinge commitments. (OC, §§ 475 & 359) This is an absolute certainty that he argues is rooted in one’s actions rather than being the result of ratiocination. (OC, § 475) This is why they are “not reasonable (or unreasonable)” but simply there “like our life.” (OC, § 559) Wittgenstein is thus describing a kind of brute certainty that is quite unlike a trusting. For one thing, one can be agnostic about the truth of the claims that one trusts, just as one can be agnostic about the truth of the claims that one presupposes, assumes, hypothesizes, and so forth. But agnosticism is not an option when it comes to one’s hinge commitments; only all-out certainty is applicable. Furthermore, one’s trusting can be rational or irrational, just as one’s employment of presuppositions (assumptions, hypotheses, etc.,) can be assessed for rationality. But Wittgenstein’s point is that when it comes to the propositional attitude involved in our hinge commitments rational evaluation is simply inapt.

If we want to be true to Wittgenstein’s aims in On Certainty, it is thus crucial that we don’t interpret hinge commitments as being merely presuppositions or assumptions that one trusts. Indeed, the propositional attitude seems to be sui generis, in that there doesn’t appear to any related propositional attitude that we are familiar with that fits the bill. They are not acceptances, for example, as acceptances can be akin to presuppositions and such like in terms of both being compatible with agnosticism about the truth of the target proposition and being rationally evaluable.⁹ Or, to take another propositional attitude that is in the vicinity here, they are unlike aliefs in crucial respects too.¹⁰ This is because one’s aliefs can be in tension with what one believes, and yet, as we will see, it is in the nature of one’s hinge commitments that this cannot be so.

Wright is at least partly alert to this point about the distinctive nature of the propositional attitude involved in our hinge commitments when he notes that they are not beliefs. As we noted above, Wright argues that beliefs are constitutively aimed at the truth and hence cannot be held in an explicitly arational manner like hinge commitments. I would like to finesse Wright’s point about hinge commitments and belief, however, by distinguishing between the folk sense of belief and a narrower sense of belief that is more of interest to epistemologists. The latter notion is essentially that propositional attitude that is a constituent part of knowledge, which explains why it has the essential conceptual connections to reasons and truth that Wright is alluding to when he says that our hinge commitments cannot be beliefs. Call this notion K-apt belief.¹¹ One important property of K-apt belief is that one cannot K-apt believe that p while recognising that one has no rational basis for the truth of p, as that would offend against the essentially truth-directed nature of this property. It is, however, pivotal to hinge commitments that one can come to recognise that one has no rational basis for their truth and yet one will continue to be certain of them regardless. More precisely, Wittgenstein’s point is that while we might, in a philosophical mood, maintain that we doubt our hinge commitments, our actions will reveal that one’s certainty in these propositions remains unaffected by our recognition of their groundlessness. This is one of the ways in which this certainty is immune to rational considerations. The upshot is that our hinge commitments are not K-apt beliefs.¹²

Nonetheless, it is consistent with this claim to hold that our hinge commitments are beliefs in the folk sense. Indeed, maintaining that they are beliefs in the folk sense is also consistent with treating our hinge commitments as involving a sui generis propositional attitude. The reason for this is that the folk notion of belief is highly permissive and hence spans a range of propositional attitudes. Think of the very different kinds of propositional attitudes that get counted as beliefs in the folk sense, such as deep religious conviction, ‘balance of evidence’ scientific claims, a detective’s hunches, and even pathological commitments (such as delusions).¹³ The folk conception of belief thus contains multitudes. It is hard to say what all these propositional attitudes have in common, but we can at least identify a sufficient condition for the folk conception of belief, which is that if one sincerely endorses p, then one counts as believing it in this sense. All the propositional attitudes just listed are thus beliefs in this sense, even though they may have very different properties (e.g., balance of evidence scientific endorsement may be better captured as acceptance).¹⁴ Crucially, however, this clearly doesn’t suffice for K-apt belief. Indeed, this distinction between folk belief and K-apt belief allows us to grant that hinge commitments may be beliefs in the former sense (since they clearly concern propositions that one would sincerely endorse if called upon to do so) even though, for the reasons we set out, they wouldn’t qualify as K-apt beliefs.

An important feature of one’s hinge commitments to specific propositions is the manner in which they are essentially manifestations of an overarching hinge certainty in one’s general worldview. It is this general arational certainty in the veracity of one’s worldview that leads to one’s hinge commitments in specific propositions. Wittgenstein’s idea is that certainty in one’s worldview cannot be rationally grounded since one needs to have a worldview in place in order to engage in rational evaluations. But once one has a worldview, then that structures one’s rational evaluations, thereby exempting certain specific propositions from rational evaluation given that they are essentially core nodes of this worldview. (OC, § 162) This is why one’s most commonsense certainties, such as that one has hands, are hinge commitments, as to allow doubt here would be to doubt one’s worldview as a whole. But such a wholesale doubt of one’s worldview (as opposed to a piecemeal doubt that merely targets aspects of one’s worldview) goes beyond the limits of what it is coherent to doubt. As Wittgenstein puts it: “A doubt that doubted everything would not be a doubt.” (OC, § 450).

Elsewhere I have described this overarching certainty in one’s worldview that one’s overall way of understanding the world is not fundamentally in error as the über hinge commitment. Note that the suggestion is not that subjects must have an occurrent thought with this specific content. The idea is rather that the über hinge commitment is manifest in the primitive certainty that one exhibits in one’s day-to-day practices; it is this overarching certainty in action that must be in place in order to be a rational subject who occupies the space of reasons at all. (Wittgenstein approvingly quotes Goethe in this regard: “In the beginning was the deed.” (OC, § 396) One’s hinge commitments to particular propositions are thus manifestations of one’s über hinge commitment in the specific conditions that one occupies. This is why, for example, in normal conditions it is a hinge commitment that one has hands, but that such a hinge commitment might be absent in abnormal conditions where this proposition no longer expresses a commonsense node of one’s worldview. Awaking in hospital after a serious car accident, for instance, it might be perfectly coherent to wonder whether one has hands, and moreover to ground one’s K-apt belief that one has hands on one’s sight of them. As one’s circumstances change, so one’s K-apt beliefs will change as a consequence and this will have an effect on which specific hinge commitments are manifestations of the über hinge commitment. There is thus an important sense in which one’s specific hinge commitments are determined by one’s set of K-apt beliefs, in that they are held fast by what lies around them.¹⁵ (OC, § 144)

This way of thinking about our hinge commitments to specific propositions has some important consequences. For example, it follows that not just any brute certainty in a proposition will count as a hinge commitment, since it needs to stand in the right kind of relationship to the über hinge commitment. Relatedly, while one’s hinge commitments are not part of one’s K-apt beliefs, they cannot be in tension with one’s K-apt beliefs as they are essentially manifestations of them, given that they are core nodes of one’s worldview. These properties set hinge commitments apart from types of extreme delusion that are characterised by pathological certainty, as often the certainty in play in delusions is concerned with claims that are explicitly in tension with the subject’s K-apt beliefs. For example, a subject suffering from Cotard’s delusion who is convinced that they are dead, but otherwise acts as normal (including, for instance, in doing such things as keeping up their life insurance payments), is not exhibiting a hinge commitment.¹⁶

I’ve argued elsewhere that this way of understanding hinge commitments is crucial to capturing how this proposal is meant to deal with the problem of radical scepticism.¹⁷ In particular, Wittgenstein is offering us a way of responding to what is these days known as closure-based radical scepticism. This is a form of radical sceptical argument which makes use of an inference from one’s everyday knowledge to the entailed anti-sceptical conclusion that one is not the victim of a specific sceptical hypothesis (e.g., that one is not a brain-in-a-vat). The principle of closure in play states that one ought to be able to generate knowledge of the entailed proposition in this way. Conversely, if that isn’t possible—if it isn’t possible to know the denials of radical sceptical hypotheses then it seems the everyday ‘knowledge’ isn’t genuine after all.

It is often argued that accepting the Wittgensteinian notion of a hinge commitment requires one to give up the closure principle at issue in this sceptical reasoning. If one’s hinge commitments manifest one’s overarching certainty in the veracity of one’s worldview, then it seems that one is hinge committed to the denials of radical sceptical hypotheses by default. That means that they lack a rational basis and so cannot amount to knowledge. Accordingly, if one’s everyday knowledge is bona fide, it seems that closure must go.¹⁸ The problem with this style of anti-scepticism, however, is that it seems to exacerbate the paradoxical nature of the sceptical problem rather than resolve it, for isn’t the closure principle itself a highly plausible claim, at least as the radical sceptic employs it anyway?

In any case, I don’t think that this is the right way to understand the Wittgensteinian proposal. To begin with, we should note that it is these days widely noted that the closure principle in play in this reasoning is only plausible insofar as it is interpreted as a competent deduction principle, whereby the subject is forming their belief in the entailed proposition on the basis of a competent deduction from known premises.¹⁹ This point dovetails with our previous observation that our hinge commitments are not K-apt beliefs, as the type of belief that needs to be formed on the basis of the closure-based reasoning must obviously be a K-apt belief (given that one is meant to have knowledge of the entailed proposition). The crux of the matter is thus that the closure principle, properly understood, is perfectly fine. What is objectionable in the sceptic’s reasoning is the idea that one can form K-apt beliefs in one’s hinge commitments via the employment of this rational principle—if Wittgenstein is right, then this is a conceptual impossibility.²⁰

More generally, what the Wittgensteinian should object to in the sceptical reasoning is the thought that there can be fully general rational evaluations, whether of a positive or negative (sceptical) kind. That all rational evaluation presupposes hinge commitments entails that such an idea is simply incoherent. What one is trying to do when undertaking the sceptical closure-style inference is precisely to rationally evaluate one’s commitments as a whole, but this is something that cannot be done. It is thus not closure that’s the problem here, but rather the underlying picture of rational evaluation whereby fully general rational evaluations are coherent undertakings, as if they were simply harmless extensions of the kind of localised rational evaluations that occur in everyday contexts.

Recognising this point also highlights an important aspect of the Wittgensteinian claim that our hinge commitments are unknown. For while this is technically true, it is also apt to mislead, as baldly stated it leaves open the possibility that one could know one’s hinge commitments, as if this were simply a cognitive limitation on our parts. Wright certainly seems to embrace this possibility. Recall that he describes Wittgenstein’s hinge proposal as enabling us to recognise how the sceptic has exposed our cognitive limitations (while giving us a way, nonetheless, to resist the sceptical conclusion). But properly understood there is no cognitive limitation on display here. The point is rather that our hinge commitments are simply not in the market for knowledge. While our hinge commitments are unknown, it is not that we are ignorant of them, as if we could know them if only we were more thorough, smarter, better appraised of the facts, and so on.²¹ Our inability to know them is no more an indication of our cognitive limitations than our inability to imagine a circle-square is an indication of our imaginative limitations.

This point is important for our purposes. Recall that it is central to Wright’s treatment of hinge commitments that not only is it hopeless to try to convict the radical sceptic of making a mistake, but that radical scepticism also contains a deep insight about the nature of our cognitive limits. This is why this approach is so concessive to radical scepticism. The point is to find a reasonable way to proceed within the cognitive limits that the radical sceptic exposes, and not to attempt to show that the radical sceptical argument could be rejected as unsound. As the foregoing makes clear, however, Wittgenstein was clearly aiming to highlight how the radical sceptical reasoning was based on a faulty philosophical picture that should be abandoned.

In particular, radical scepticism purports to present us with a paradox, in the sense of identifying deep inconsistencies in our ordinary ways of thinking about knowledge. According to Wittgenstein, however, all it in fact reveals is how a particular philosophical picture concerning the coherence of universal rational evaluations—a picture that is not at all present in our ordinary epistemic practices—generates paradox. Dubious philosophical theory is thus masquerading under the guise of commonsense. The goal of Wittgenstein’s remarks is therefore to reveal to us the nature of our ordinary epistemic practices and thereby show us how we are being misled by such theory. We will then be in a position to reject the dubious theoretical picture that was generating the supposed paradox. Contra Wright, the Wittgensteinian account of hinge commitments does show that radical scepticism rests upon a mistake. Relatedly, in embracing hinge commitments we are not thereby coming to recognise a sceptical truth about our cognitive limitations but rather gaining an understanding of the essential nature of rational evaluation.²²

5 Hinge commitments and trust: take two

With Wittgenstein’s account of hinge commitments brought into sharper relief, it should be clear why the kind of propositional attitude that Wittgenstein had in mind is very different to trust. We noted earlier that the notion of trust that’s operative when one describes hinge commitments in terms of trust is more than just a mere reliance on the truth of the target proposition (which could also apply to belief, including K-apt belief). To trust that p in the sense that we are interested in is rather a particular kind of reliance, of a sort that would be relevant to presuppositions that one is groundlessly committed to, which as we saw is how certain theorists regard our hinge commitments.

The exact properties of trust have been hotly contested by contemporary philosophers, but we can side-step many of these disputes. This is because there are two core properties of trust that are relevant for our purposes, and both are widely endorsed in the literature. The first property is that trust is optional at least in the minimal sense that one could in principle withhold it. One’s trust can be let down and frustrated, in which case that trust is lost.²³ This might occur with presuppositions when we discover that they lead us astray, or that there are better, more useful, presuppositions available. In general, presuppositions can in principle be jettisoned, including by coming to know them so that they are no longer presuppositions or by simply rejecting them altogether. The second property is that trust is compatible with doubt and hence is incompatible with complete certainty. More specifically, trust is compatible with agnosticism about the truth of the target proposition.²⁴ We see this very clearly in the case of groundless presuppositions which are recognised as such, where agnosticism about their truth would be the entirely natural—and, indeed, rational way of regarding them.

Neither of these properties are applicable to hinge commitments. On the first property of trusting, one’s hinge commitments are not the kind of propositional attitude that one can exercise agency over in the manner that one can with one’s presuppositions. One’s hinge commitments are rather a visceral certainty that one simply has; they are not the product of ratiocination or responsive to such ratiocination. This doesn’t mean that one can’t lose one’s hinge commitments, or that they cannot change over time, but it does mean that whatever the processes are that determine such change they do not involve a simple conscious decision on the part of the subject, much less do they involve rational deliberation.²⁵

On the second property of trusting, one’s hinge commitments involve an absolute conviction in the truth of the target proposition, as manifested in one’s actions. They are thus incompatible with adopting a stance of agnosticism about the truth of the target proposition, much less adopting such a stance in response to rational considerations. One is not certain that one has hands because one has become persuaded of it by considering the reasons in support of this claim; supporting reasons play no role in this certainty. Relatedly, discovering that this commitment is groundless has no effect on the certainty that one has in its truth (much less does it have a rational effect on this certainty), as would be the case when one considers the groundless nature of one’s presuppositions.²⁶

These two differences between the propositional attitude involved in trusting and that involved in hinge commitments are important because they are relevant to the question of whether one’s hinge commitments can be reasonably held. This is because it is the properties of trusting such as these that ensure that this propositional attitude can be assessed for reasonableness. If one retains one’s trust in a presupposition even when it becomes clear that there is no basis for this trust—where, for example, this presupposition has led one astray in one’s inquiries—then one would regard it as an unreasonable propositional attitude to have. Similarly, one can assess whether it is reasonable to trust a presupposition in terms of whether one’s degree of confidence in the truth of that proposition is warranted, given the supporting evidence. In particular, if one recognises that there is no rational basis for the truth of the target proposition, then reason demands that insofar as one continues to trust it one should do so in a manner that is accompanied by a healthy scepticism about its truth. Since neither of these properties of trusting are applicable to hinge commitments, this highlights the extent to which what makes it possible to evaluate trusting as reasonable does not carry over to hinge commitments. It is no wonder, then, that Wittgenstein emphasises that our hinge commitments cannot be assessed in this manner.

6 Exegetical issues

One concern one might have with this debate about how best to understand hinge commitments relates to the fact that we are dealing with unedited notebooks. To what extent can we extract a definitive account of hinge commitments at all? Are Wittgenstein’s remarks on this matter even consistent? This is a fair point, one that I used to have a lot of sympathy for, as the notebooks that make up On Certainty can on the face of it appear quite contradictory in places, in which case aiming for a definitive reading, as opposed to a thesis that is merely inspired by some of the remarks in these notebooks, looks like a lost cause. I now believe, however, that if one pays careful attention to the text it becomes clear that there is far less inconsistency in these remarks than is first apparent.

One reason why the remarks in these notebooks can seem at times contradictory is that while the overriding view about hinge commitments on display is along the lines set out above (an arational primitive certainty, and so forth), Wittgenstein also discusses other common certainties which he evidently regards as rationally grounded and, indeed, as being known. So which is it then: are our hinge certainties primitive arational certainties or rationally grounded knowledge?

The key to resolving the puzzle is to realise that not all the commonsense certainties that Wittgenstein discusses are meant to be the distinctive kind of arational certainties that we are calling hinge commitments.²⁷ One central theme of the notebooks that make up On Certainty concerns the contrast between the particular kind of hinge certainty that plays a special epistemic role in our practices with another kind of everyday certainty that has far less distinctive properties. We can class this second kind of commonsense certainty as common knowledge—viz., the kind of claim that mature agents tend to know, though they can rarely recall exactly how they know it, and which is usually part of the tacit epistemic background of our practices.²⁸ To take one of Wittgenstein’s favourite examples: that water boils when heated. This is a claim that we all tend to know, though we might be unsure about the scientific explanation of why it is true, or even how we came by this belief. Regardless, this is something that we treat as knowledge, and which we recognise is well-supported by the evidence. Our certainty in this claim is thus a grounded certainty, even though it is usually not grounded in specific evidence as opposed to a general sense that this is the kind of claim that enjoys strong evidential support and hence clearly amounts to knowledge.

Once we have these common knowledge certainties in view, then it becomes very clear that in the On Certainty notebooks Wittgenstein is often contrasting this kind of everyday certainty with the more distinctive hinge certainty. Indeed, his ruminations are aiming to get to the bottom of the differences between them, with the distinction between the two kinds of certainty becoming more explicit as we move towards the later notebooks. In particular, by the fourth and final notebook it becomes clear that Wittgenstein is aiming to distinguish between these two familiar kinds of certainty precisely in order to bring the more radical hinge certainty into sharper relief. Consider, for example, this important passage where Wittgenstein compares the common knowledge that water boils when heated with one of his recurring examples of a hinge commitment, his certainty that this really is his dear dying friend in front of him, with whom he has been conversing. Wittgenstein begins by noting that while both are certain, “there is a difference between [the] cases”. Here is how he presents this difference:

If the water over the gas freezes, of course I shall be as astonished as can be, but I shall assume some factor I don’t know of, and perhaps leave the matter to physicists to judge. But what could make me doubt whether this person here is N.N., whom I have known for years? Here a doubt would seem to drag everything with it and plunge it into chaos. (OC, § 613).

Note the distinction that Wittgenstein is drawing here between these two certainties. The possibility that the common knowledge certainty is false leads to puzzlement, but no more than that. In contrast, one cannot even make sense of the hinge certainty being false. This is because of the more fundamental role that it plays in one’s rational practices, such that it represents, in those conditions, the overarching über hinge certainty in one’s worldview: to be wrong here would mean that one is wrong everywhere.²⁹

Once we understand that not all the everyday certainties that Wittgenstein discusses in On Certainty are meant to be hinge certainties, then this helps us to make sense of how Wittgenstein uses the notion of trust in these notebooks. While Wittgenstein does use this term when describing everyday certainties, it is clearly just the common knowledge certainties that he has in mind. We are, after all, aware that such certainties are rationally grounded, at least in the social sense that we know that they enjoy this status (even if we cannot offer any specific grounds ourselves). For example, we trust science textbooks and such like because they have ‘proved’ themselves, where trust in such textbooks is ‘backed up’ by experience. (OC §§ 275, 600 & 603) Trust makes perfect sense as a propositional attitude to adopt here, because although one can’t rationally ground these propositions oneself, one can be assured that they are reasonable claims to endorse, given the wider social rational standing that they enjoy. In contrast, trust is not applicable to the hinge certainties. For not only is this not a certainty that one can in any sense choose to adopt (or withdraw), much less in response to rational considerations (which is why it makes no sense to talk of their endorsement as reasonable or unreasonable), it is also not the kind of certainty that, properly understood, enjoys any rational support at all (given that it forms part of the framework that makes the very existence of rational support possible).³⁰

7 Concluding remarks

One may have antecedently thought that this question of how to understand hinge commitments is largely a peripheral matter, of interest only to scholars of the later Wittgenstein. As I hope I have shown, however, understanding this Wittgensteinian notion correctly is important, as it leads to us conceiving of what Wittgenstein is proposing in these notebooks in a completely different way. In particular, recognising that hinge commitments are very different to presuppositions, and hence do not involve the propositional attitude of trust, leads to a much more radical proposal. This is especially relevant when it comes to the anti-sceptical import of hinge commitments.

As we’ve seen, it is a consequence of thinking about our hinge commitments as presuppositions that we trust that they are effectively brought back into the realm of our rational agency, at least in an expanded sense. We can now coherently talk of our adopting an attitude of trusting one’s hinge commitments and this policy can be assessed for reasonableness. Indeed, the whole point of an approach to hinge commitments like that represented by Wright’s account is to explain how it is reasonable to hold them even though they lack direct rational support. In this way, even despite their lack of such support, our hinge commitments can nonetheless be in a sense rational.

Relatedly, on this proposal, the radical sceptic isn’t making any error in their reasoning but rather highlighting a deep truth about our cognitive situation, in that they are exposing a genuine cognitive limitation on our part. Nonetheless, thinking of hinge commitments in terms of trust offers us a way of living within these cognitive limitations. We just can’t offer rational support for all our commitments, but that doesn’t prevent us from being reasonable in the presuppositions that we are obliged to hold.

In contrast, the account of hinge commitments that I am urging (and which I’ve claimed is what Wittgenstein intended) is much bolder. Our hinge commitments are completely unlike presuppositions, and so not amenable to the propositional attitude of trust; indeed, they lie outside of our rational agency altogether. Accordingly, not only are they lacking in rational support (and so unknown), but it’s also the case that they cannot be assessed for reasonableness. As Wittgenstein says, they are not reasonable or unreasonable, but simply there, like one’s life. Unlike the alternative approach just outlined, this way of thinking about hinge commitments is not concessive to radical scepticism at all. Once one rejects the faulty picture of rational evaluation whereby fully general rational evaluations are entirely coherent, then the idea that one’s most fundamental certainties are arational is no longer scepticism-friendly. Since it rests on a faulty philosophical picture, the radical sceptical argument is mistaken, and certainly doesn’t highlight a deep truth about our cognitive limitations. Indeed, far from exposing a cognitive limitation on our parts, accepting that there are arational hinge commitments is simply a consequence of rejecting the faulty picture. In particular, as noted above, that we cannot know our hinge commitments is no more a reflection of our cognitive limitations than the fact that we cannot imagine a circle-square is a reflection of our imaginative limitations.³¹ Just as the idea that our hinge commitments are merely presuppositions that we must trust goes hand-in-hand with making important concessions to radical scepticism, so treating hinge commitments in the very different non-presuppositional way advocated here goes hand-in-hand with a more radical form of anti-scepticism that avoids such concessions.³²