The Thinkableness of All Thoughts and the Irreplaceability of Pictures: Cora Diamond on Religious Belief

Abstract

Under the ideas of ‘hinges’ and ‘pictures’, as these relate to deep disagreement, Wittgenstein’s view of religious belief is a multifaceted challenge to conceptions of thought-world relations. In this article, I discuss Cora Diamond’s analysis of this challenge. Diamond herself is not particularly interested in hinges; I try to understand why. I first bring in a discussion between Michael Williams and Duncan Pritchard on how to read On Certainty. This allows me to identify Diamond’s perspective on deep disagreement and pictures: she concentrates on making sense, and not directly on knowledge. To further clarify her perspective, I introduce Hilary Putnam’s reading of the Lectures on Religious Belief, which proposes a cognitivist view of religion as ethics, centering on the notion of picture. Although Diamond is close to Putnam, for her, the most important challenge posed by religious belief lies not with epistemological issues of rational versus arational grounds of belief, or cognitivism versus non-cognitivism in ethics, but rather in making us drop the Fregean (and Tractarian) idea of the thinkableness of all thoughts, making place for ‘irreplaceable pictures’. I end by suggesting that Diamond’s analysis sheds light on often uncontested assumptions about the natures of thought and communication.

1. Introduction: Cora Diamond and the Philosophy of Religion—Starting with an Example from the Lectures on Religious Belief

Cora Diamond’s approach to the philosophy of religion is highly idiosyncratic. Her main interest lies not in any of the usual topics of philosophy of religion, such as God’s existence, the nature of religious experience or whether religious belief is warranted, but in the nature and limits of thought and language.1 The investigation of thought and language that Diamond pursues around religious belief (Diamond 2005) results in two proposals. I will call these the unthinkableness of ‘all thoughts’ (i.e., a rejection of the idea, which accompanies Frege’s conception of thought (Gedanke), that there is a fixed totality, or set, of all thoughts) and the irreplaceability of pictures in thinking. In this article, I trace Diamond’s path as she puts forward each one of these proposals by reading passages of Wittgenstein on religious belief.

The following well-known example, from the Lectures on Religious Belief, is particularly important in Diamond’s project:

Suppose someone were a believer and said: “I believe in a Last Judgement,” and I said: “Well, I’m not so sure. Possibly”. You would say that there is an enormous gulf between us. If he said “There is a German aeroplane overhead,” and I said “Possibly. I’m not so sure”, you’d say we were fairly near.

(Wittgenstein 1966, Lecture One, p. 53)

In Wittgenstein on Religious Belief: The Gulfs Between Us (Diamond 2005), Diamond uses the example as a guiding thread. The first thing to say about Wittgenstein’s ‘Last Judgment man’ example2 above is that it illustrates disagreement in our ongoing use of concepts. What Wittgenstein calls the ‘enormous gulf between us’ is what epistemologists and philosophers of language often call deep disagreement (Fogelin 1985), one suggestion being that the acknowledgement of such deep disagreement between parties motivates relativism.3 What does Wittgenstein mean, though, by this ‘enormous gulf between us’? He is certainly calling our attention to the fact that there are two quite different situations in the exchange between the two interlocutors. In the first case, when the man who believes in the Last Judgment speaks about the Last Judgement with the man who does not believe in such things, an argumentative path between them seems absent. In the second case, when the same two interlocutors speak about the German airplane, there clearly is an argumentative path between them: there seems to be agreement on what would constitute evidence and would allow the issue to be decided. In both cases, the interlocutors do speak to each other; in both cases, each interlocutor, in a way at least, understands the words the other one uses. Only, in the first case they are not talking about the same thing. In both cases, one might think one interlocutor is saying to the other: ‘Perhaps’, meaning ‘I hear your words, I understand them. Maybe things are as you say.’ According to Diamond, what Wittgenstein points out is that despite such similarity between the two situations, the interlocutors are not facing each other in the same way in both (see Diamond 2005, p. 100, Four ways we may be distant from each other). In other words, Diamond thinks that it is not sufficient that two people speak the same language, to hear and, in a way at least, understand the same by the same words. What else is at stake then? This is what interests Wittgenstein.

The example above comes from transcripts of lectures that the philosopher gave in Cambridge in 1938. The lectures were officially on religious belief. It was not, however, only religion, or above all religion, that interested Wittgenstein in these lectures. He was interested in belief in general, in the ways we think things about things and use language to express what we believe. That is also what interests me here. Naturally, one reason why Wittgenstein’s example above is so significant is that the phenomenon illustrated by the example above, the example of the exchange between the religious and the non-religious men speaking about the Last Judgment, can happen in various other areas. Doing science or discussing ethics, Wittgenstein’s ‘gulfs between us’ are often present. Interlocutors may speak the same language, exchange words and yet not, as it were, meet in thought. Another example through which Diamond herself pursues, in ethics, the issue of the Wittgensteinian gulfs between us are confrontations, in 19th century America, about whether slavery is unjust and insupportable—where one side thinks it definitely is (unjust and insupportable) and the other side thinks it definitely is not.4 When such gulfs are present, what is at stake is more than well-conducted arguments. One immediate, although not sufficient, way to put the contrast is in terms of ordinary disagreement versus deep disagreement. The case of the men talking about the Last Judgment, or, the example of one person telling another, as they say goodbye and leave for China, ‘We might see one another after death’,5 are not cases of ordinary disagreement. Ordinary disagreements concern argumentative exchanges in which peers dispute the truth of a specific proposition (such as ‘Benfica (a Portuguese football club) was the national champion in 2004–2005’ or ‘The capital of Ivory Coast is Abidjan’). They dispute the truth of the proposition, but they agree on what would constitute evidence for or against the controversial claim. In deep disagreements, though, a distance in thought is present that goes beyond argumentative disagreement. Diamond thinks that such distance concerns different ways of seeing reality. That is what she explores reading Wittgenstein on religious belief.

It is not just in the Lectures on Religious Belief that Wittgenstein is trying to understand the phenomenon of a distance in thought, amounting to different ways of seeing reality. He is concerned with it also in On Certainty (Wittgenstein [1969] 1972) and in Culture and Value (Wittgenstein 1984), speaking of pictures, world pictures (Weltbilder), and of the inherited background (überkommene Hintergrund) that we appeal to when distinguishing between true and false (Wittgenstein [1969] 1972, p. 94). He is concerned with it when he introduces the terminology of what is now hinge epistemology: the hinges (Angeln), that stay put when what is around them moves, and the riverbed (Flussbett), that contrasts with the moving waters.

We may think of deep disagreements as occasions where different Weltbilder clash. Because a Weltbild is a tacit, implicit, pre-rational framework for thinking, it is not easy to understand what it means that Weltbilder clash. One thing we know: world-pictures do not change easily. In fact, resistance to rational change is part of what is at stake in Wittgenstein’s concept of hinges, as it is interpreted differently by hinge epistemologists such as Annalisa Coliva, Danièle Moyal-Sharrock, or Duncan Pritchard.6 In On Certainty (Wittgenstein [1969] 1972), Wittgenstein uses the terms hinges, framework, and riverbed to name particularly fixed and non-explicit aspects, aspects not subject to doubt in our epistemic practices, perhaps beliefs which remain ‘immobile’ and untouched during the normal process of adjustment and revision of beliefs that characterizes rational inquiry. Some of Wittgenstein’s examples are ‘I have two hands’ (Moore’s example), ‘The world did not begin to exist five minutes ago’, or ‘I am a human being’. According to some epistemologists (e.g., Duncan Pritchard), we could think of ‘God exists’ in a similar way: religious beliefs are hinges too. Wittgenstein’s point is that we can be rational, go on with our uses of concepts, and go on with belief revision, while hinge beliefs stay put. One important point then is that religious belief shows us something of general interest for epistemology—that there are hinges in our thinking in general, and that they have certain features and impose particular constraints onto rational processes.

Although Diamond pays much attention to the general interest of Wittgenstein’s remarks about religious belief, she herself is not particularly interested in hinges. I will try to understand why. In the following section (Section 2. How to read On Certainty: Is it hinges? Is it epistemology?), I first take an indirect route for that: I bring in a discussion between Michael Williams and Duncan Pritchard on how to read On Certainty. I want to free the space in which Diamond approaches deep disagreement and pictures in our use of concepts: she concentrates on making sense, and not directly on knowledge. In Section 3 (How to read the Lectures on Religious Belief: Is it ethics? Is it about ‘being in the grip of a picture’?), I introduce Hilary Putnam’s reading of the Three Lectures as presenting a cognitivist view of religion as ethics, centering on the notion of picture, to further clarify Diamond’s perspective. In Section 4 (Neither epistemology nor ethics: Diamond on the thinkableness of all thoughts and the irreplaceability of pictures), I argue that for Diamond, the most important challenge posed by religious belief lies neither with epistemological issues of rational versus arational grounds of belief, nor with cognitivism versus noncognitivism in ethics, but rather in making us drop the Fregean (and Tractarian) idea of the thinkableness of all thoughts, making place for what she calls ‘irreplaceable pictures’. Finally, in Section 5 (Conclusions—Diamond on religious belief and the natures of human thought and human communication), I suggest that Diamond’s view sheds light on assumptions about the nature of communication which are seldom contested.

2. How to Read on Certainty: Is It Hinges? Is It Epistemology?

Let us suppose then that at least some religious beliefs, such as ‘God exists’, are hinges. But what are hinges, as Wittgenstein conceives them? Hinge epistemologists and other interpreters of Wittgenstein clash around interpretations of On Certainty and of what Wittgenstein means by hinges therein. In what follows, I consider the interpretations of Duncan Pritchard and Michael Williams.

Michael Williams (2019) acknowledges that Duncan Pritchard’s reading of On Certainty and his own share a lot. There are, however, important differences. According to Pritchard (2016), Wittgenstein’s main purpose in On Certainty is to show that epistemic justification is local, i.e., that rational inquiry does not call into question the totality of beliefs. Yet according to Williams, what Wittgenstein is really interested in in On Certainty is showing that meaning, truth, and proof are circumstance-dependent. In this, Williams claims, Wittgenstein is quite close to J. L. Austin (1962). The dispute boils down to the interpretation of the following paragraphs of On Certainty, where Wittgenstein says:

That is to say, the questions that we raise and our doubts depend on the fact that some propositions are exempt from doubt, are as it were like hinges on which those turn. (§ 341)

That is to say, it belongs to the logic of our scientific investigations that certain things are in deed not doubted. (§ 342)

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. (§ 343).

(Wittgenstein [1969] 1972, p. 44^e)

Williams and Pritchard agree that Wittgenstein is claiming here that rational evaluation of our beliefs takes place against a background of certainties. Knowledge and justification depend on basic certainties. These are called hinges, or hinge commitments. Examples of hinge commitments could be: ‘I have two hands’, ‘or ‘I have parents’, ‘I have a brain’, or ‘I live in Porto’.7 Such commitments provide, as it were, a framework for our epistemic practices of inquiry and justification. This may seem surprising, since such hinge commitments are not basic in any traditional epistemological sense, the sense epistemologists might have in mind. They are not, e.g., empirical beliefs about sense data, nor do they, say, concern mental self-acquaintance. They are rather quite ordinary beliefs, connected to life. Yet they seem to have a sui generis epistemic character, namely, they do not respond directly to rational considerations. They are not open to the question ‘How do you know?’ in the same way as other beliefs are. Imagine that I have been away from my country, in an isolated place, and I believe that the current Prime Minister of Portugal is António Costa. But then a German friend tells me: ‘But António Costa is no longer prime minister of Portugal. The current prime minister is Luís Montenegro’. I would promptly revise my belief, without any further problems. The same would not happen, however, if my friend tells me that, contrary to what I believe, the world began to exist five minutes ago, or that I do not have parents, or that I do not really have a brain, or that I do not live in Porto. How am I to change my beliefs based on these discoveries?

Williams (2019) brings in Austin to formulate his objections to Pritchard’s reading of On Certainty. So, here is an example built around Austin. Let’s say that Austin knows that he lives in Oxford. He is sure of that. Still, the person next to him may, when he says that he lives in Oxford, very well doubt that he does. She may think he is lying or pretending. She doubts him—yet Austin himself cannot doubt that he lives in Oxford. Or can he? What form could Austin’s doubt take? How could he go about proving to himself that he lives in Oxford? What could constitute evidence? Does the fact that he receives correspondence in his name, with an Oxford address, constitute proof, for him, that he lives in Oxford? Is knowing that he lives in Oxford something he concludes from such evidence? For example, he receives mail addressed to him in an Oxford address and goes through the process of thinking: ‘Well, I must live in Oxford, since this mail is addressed to a Mr. J. L. Austin, and as far as I know J. L. Austin is me and this place I am in is Oxford. This means I must live in Oxford’. This would be quite strange. His being certain that he lives in Oxford must somehow have a different ‘shape’. Yet clearly there is nothing self-evident, or basic, about Austin’s knowing that he lives in Oxford, not as we might think that there is something basic about someone thinking ‘This is red’. How can something like ‘I live in Oxford’ be knowledge then?

According to Wittgenstein ([1969] 1972, p. 166), that there are such hinges and that we generally do not call them into question marks the ‘groundlessness’ of our beliefs. Because hinge propositions are at work in rational inquiry, the very idea of a general, thorough, rational assessment of our beliefs seems incoherent. Pritchard and Williams agree on this too. At the heart of their dispute lies something different. First, from the outset, there is the question of whether hinge propositions constitute knowledge. According to Pritchard, the answer is negative: hinge-propositions are not (in his term) ‘knowledge-apt’. Strictly speaking, hinges are not beliefs. Hinge commitments are arational. They have an ‘animal character’. Given such animal character (etwas animalisches is Wittgenstein’s expression, Wittgenstein [1969] 1972, p. 359), hinge commitments are not optional. Our animal commitments are not optional (what would that mean, for them to be optional?). That goes with the fact that they are neither acquired through rational processes nor respond to rational considerations. As Pritchard puts it, nobody teaches us that we have two hands; they may teach us how to do certain things with them. Rational inquiry does not reach to such basic and fundamental commitments to evaluate them.

For Williams, in contrast, hinge commitments certainly are knowledge—they are in fact the very prototype of what it is to know something—and Williams thinks that Wittgenstein is with him in seeing hinge-propositions as knowledge. In On Certainty, Wittgenstein says:

We know, with the same certainty we believe any mathematical proposition, how the letters A and B are pronounced, what the color of human blood is called, that other human beings have blood and call it ‘blood’. (§ 340)

(Wittgenstein [1969] 1972, pp. 43–44^e)

Granted, such objective certainties are tied neither to conclusive reasons nor to self-evidence. But this means, according to Williams, that they can simply be linked to the absence of reasonable doubts in a particular situation. For example: I know that I live in Porto. Does it make sense for me to doubt this, here and now? Williams claims it doesn’t. Naturally, he then must account for what ‘making sense’ means here. (Williams 2019, p. 73). There seems to be an alternative: either it makes no sense to doubt because doubting that I live in Porto would be unreasonable, or it makes no sense to doubt because doubting would be completely unintelligible. Williams takes both. Both aspects are involved in someone, and something, making or not making sense. I will return to this issue of making sense later, since it is crucial for Diamond. What Williams wants to insist on anyway (as he did in Williams (1991), Unnatural Doubts) is that doubts also need grounding. There must be specific reasons for doubting if we are to doubt something. It could simply be that in a given circumstance everything speaks in favor of p (for example, in favor of my thinking here and now that I have two hands, or that I live in Porto) and nothing speaks against it. In these circumstances, I have no reason to doubt those certainties. According to Williams, Wittgenstein and Austin converge around this.

In epistemology, scenarios for doubt are often brought in for discussion in terms of so-called relevant alternatives. That idea has its origin in Austin’s Other Minds (Austin 1970), where the need of reasons for doubting is also put forward. A well-known example of Austin is ‘I know there’s a goldfinch in the garden’. Austin remarks about this example that to say that ‘I know that there is a goldfinch in the garden’ is to say that, for present purposes and in the present situation, there is no pertinent alternative. This does not mean, though, that it is sufficient to definitively establish that what I see is not, for example, a stuffed goldfinch (and if that which I see is a stuffed goldfinch, then I do not in fact know that there is a goldfinch in the garden). Knowing that p ‘is what it is and is not many other things’, Austin says. He suggests anyway that I count as knowing that there is a goldfinch in the garden even if I don’t know that that which I see is not a stuffed goldfinch, and if it is a stuffed goldfinch then it is clearly not a goldfinch.

At stake here is how we stand regarding the skeptic; William’s claim is that Wittgenstein and Austin take very similar stances in answering the skeptic. In contrast, Williams and Pritchard have very different views. I will go back to Austin, to illuminate the difference. According to Pritchard, with the goldfinch example above Austin is simply highlighting how different our ordinary practices of epistemic evaluation are from the radical evaluation of beliefs that is demanded by the skeptic. This is all that Austin does (Pritchard 2016, p. 68). So, for Pritchard, the skeptic goes further than Austin; he asks for something different and much more important. But Williams thinks that it is Austin who is on the right track about what really matters in skeptical discussions, and that what Austin is claiming is quite close in spirit to Wittgenstein’s claims in On Certainty.

Let us go back to the idea of reasonable doubts. A fundamental thrust of Williams’s work in epistemology is the idea that reasons to doubt are not as easy to find as the skeptic thinks. This means for instance that ‘fairy tales’ (this is William’s term for what other people prefer to call ‘thought experiments’) are not enough to bring forth relevant doubts (“fairy tale imaginings do not do it”, he says). For Williams, scenarios such as the Evil Genius or the Brain in a Vat, which for other philosophers present radical skeptical challenges, are precisely fairy tales, or, in other words, mere stories. What does this mean? Radical skeptical arguments such as the Evil Genius or the Brain in a Vat may be formulated as an Argument from Ignorance (AI). Let SS be a skeptical scenario, like the Brain in the Vat scenario, for a proposition p like ‘These hands are mine’:

• (AI-P1) I know that p;
• (AI-P2) I don’t know that not-SS;
• (AI-C) If I don’t know that not-SS, then I don’t know that p.

But just compare, Williams says:

That Wile E. Coyote, in a cartoon, runs off a cliff and hangs in the air until he looks down is no reason to doubt the law of gravity [Wile E. Coyote is a character in a Looney Tunes cartoon, S.M.]

(Williams 2019, p. 76)

Williams’ observation boils down to a point about context. Contextualist epistemologists often assume that one can produce a change of context merely by considering hypotheses of error in thought, that is, by creating skeptical scenarios. The recipe of attributor contextualism for considering possibilities of error is Metaphysical Possibility plus Closure in a dialectical context, i.e., in a linguistically formulated, and thus ‘created’, context (see Pritchard 2016). Yet for Williams, mere dialectical considerations of the possibility of error such as these cannot be an obstacle to knowledge: the reason is that ‘fairy tale scenarios’ simply miss the actual world, the real world. But why is the real world so important? This is where the crucial point about context comes in. From an Austinian point of view, the objects of epistemic evaluation are utterances, i.e., claims which are advanced, or commitments which are assumed, by specific speakers who find themselves in particular circumstances in the world. Whether an utterance constitutes knowledge depends on what the claim is, who is claiming it, to whom the claim is directed, where, and when. For Williams, being in a position to know that p involves all the above-listed aspects of a person’s epistemic circumstances, not just a linguistic exchange with an interlocutor with whom one agrees, for example, to play the skeptical scenario game. Context is not just a conversational or dialectical matter—this is an unjustified idealization (I will go back to this point in the final section). All aspects of the agent’s situation in the world matter for context, not just the linguistic exchange. This notion of context, which Williams sees as an Austinian, multidimensional, and circumstance-dependent notion of context, is behind Williams reading of On Certainty and his view of hinge propositions (he, in fact, prefers the term framework for formulating his position on circumstance-dependence).8

The importance of circumstance-dependence can be better understood if we consider Williams and Pritchard discussing whether hinge commitments are knowledge or not. One serious problem with Williams, according to Pritchard, is that he thinks that hinges can be (re)moved, that they are optional. Remember, though, that Williams sees hinges not as visceral commitments, as Pritchard does, but as even possibly enabling wissenschaftlichen Untersuchungen, i.e., scientific investigations. So, the way Williams sees things, Pritchard is, on the one hand, dismissing hinges as arational, and on the other hand, availing himself too easily of a notion of knowledge in which knowledge-claims are circumstance-independent (again, idealized). In other words, Pritchard assumes he starts from the good view of what a knowledge claim is, and that that is a purely epistemological matter: there is no need to go into speculations about truth and meaning, claims and contexts. For him, questions of truth and meaning, which are, granted, controversial, clearly belong in the philosophy of language. They should be left aside when we do epistemology:

While it might ultimately have been Wittgenstein’s intent [in On Certainty] to motivate his anti-skepticism by appeal to a highly context-sensitive account of meaning, we should be wary of saddling ourselves with this philosophical baggage if we can avoid it. For to take this kind of line (…) is to make oneself hostage to wider philosophical claims, extending well beyond epistemology that may not be themselves plausible in the final analysis.

(Pritchard 2016, p. 86)

Unlike Pritchard, Williams does not see epistemology as a self-standing enterprise. He thinks epistemological discussions cannot bypass questions of language and making sense. He also believes this very same conviction lies at the core of Wittgenstein’s On Certainty, as of Austin’s Sense and Sensibilia, and Other Minds. So, Williams’ general point against Pritchard’s many and diverse elaborated positions in epistemology in Pritchard, 2016, is simply that they rest on a blind-spot: the assumption of the circumstance-independence of knowledge claims. In contrast, for Williams, at a fundamental level knowledge and certainty are grounded in conditions for meaning; there is no such things as matters of evidence and proof which are solely epistemological. Therefore, the phenomenon Wittgenstein is interested in in On Certainty is not just the locality of epistemic justification. What Wittgenstein is interested in are the conditions required for making sense in a multidimensional context. A multidimensional context includes the concrete current situation of the putative knower; dialectical factors are just one part of this multidimensional epistemic context. Contexts (e.g., skeptical contexts) cannot merely be created conversationally, as if by fiat. Under this light, Pritchard’s view of hinges appears in Williams’s eyes as, as it were, purely epistemological and only superficially contextual. While trying to come to terms with discussions of contextualism in epistemology, Pritchard simply ignores any considerations about the meaning and truth of assertions. Williams, in contrast, claims:

I have never thought that one could treat epistemology as a kind of self-standing subspecialty, in the way some people do today. I don’t believe there are clear lines between epistemology, philosophy of language, philosophy of mind, and metaphysics.

(Couto and Corti 2020, Contextualism in Context: Interview with Michael Williams)

I believe Diamond would concur with Williams. That is one important reason why she does not read Wittgenstein’s views on religious belief as an epistemology of hinges. Could it be more illuminating, then, to discuss Wittgenstein’s proposals on hinges and pictures in the context of ethics?

3. How to Read the Lectures on Religious Belief: Is It Ethics? Is It About ‘Being in the Grip of a Picture’?

Hilary Putnam does exactly that in Chapter 7 of Renewing Philosophy, Wittgenstein on Religious Belief (Putnam 1992). Putnam read Wittgenstein’s Lectures of Religious Belief in the context of clarifying what his own Jewish faith meant for him (Putnam 1992, 2002, 2004, 2008; Miguens 2020). Along the way, he depicts the Lectures as a model of what philosophical investigations should be. He sees Wittgenstein in the Lectures as doing ‘serious and honest research of the most difficult kind’ (Putnam 1992, p. 141), in fact contrasting Wittgenstein’s struggling and tentative way of doing philosophy in the Lectures with, e.g., David Lewis’ idea of working out consequences of metaphysical positions, since ‘there are countless other ways things could have been’ (Putnam 1992, p. 145). Putnam cherishes Wittgenstein’s lectures because «there Wittgenstein goes against simplistic views of religious belief and religious language. He never accepted the “facile idea that religion is essentially conceptual confusion”» (Putnam 2008, p. 11), or saw religious discourse as pre-scientific or primitive, something that has strangely managed to survive in the age of digital computers (Putnam 1992, p. 143). He sees Wittgenstein as not ‘saying the standard things about religious language’, i.e., not only that religion belongs with a pre-scientific conception of the world but also that religious discourse is non-cognitive, that it is emotive, and that it is incommensurable with science. What is Wittgenstein saying then?

According to Putnam’s reading of the three Lectures, Wittgenstein is claiming that religion is ‘largely ethics’. Putnam concurs; for him, religion is largely a view of the ethical life. For claiming this, Wittgenstein’s idea of a picture, in the Lectures, is key. The religious person ‘uses a picture’, Putnam says; the unshakeability of religious belief is manifest not in reasoning but in regulating a life. This is what a picture does, and what a picture does does not exclude doubt. In fact, it can even alternate with it—in this, religious belief plainly contrasts with ordinary belief (Putnam 1992, p. 145). As Putnam puts it, if I stop believing that the right way to build a bridge is the way I am building it, I stop doing it. I will halt the construction, run tests, and make calculations. This is clearly not so with religious belief.

Also, what is at stake in religious belief is not knowing something theoretical (e.g., about God, whether God exists, or whether some particular dogma holds). For Putnam, ‘theorizing about God is besides the point’ (Putnam 2008, p. 6). In other words, for Putnam, what matters in religious belief are not the questions at the center of much analytic philosophy of religion (think, e.g., of Alvin Plantinga’s work), whether God exists, whether religious belief is warranted or not. In fact, although Putnam’s global reading of the Three Lectures starts from problems posed by the gulfs between us (the same problems regarding contradiction and disagreement that Diamond discusses), his interpretation ‘ends’ with a version of Cavellian moral perfectionism and religion’s place in it. What matters for Putnam in religious belief (and this is also his view on why religious belief matters) is that it concerns how we see our lives and how we transform them. Is this a cognitivist view, one according to which religious propositions have cognitive content and truth-value? Or a non-cognitivist view, according to which religious propositions merely express emotions and pro-attitudes? Perhaps the question cannot be that direct. Putnam approves Wittgenstein’s rejection of non-cognitivism (he even reminds us of Stanley Cavell’s observation that Charles Stevenson, the father of emotivism, wrote ‘as if he had forgotten what ethical arguments sound like’, Putnam (1992, p. 153)). But Putnam’s view of religion as a view of the ethical life comes after (what he calls) the collapse of the fact-value distinction, and a fact-value distinction is needed to make space for non-cognitivist positions, e.g., emotivism, or expressivism. According to non-cognitivist positions, utterances like ‘Licorice tastes good’ do not express truth-conditional contents; rather, they express evaluative attitudes. Tractarian Wittgenstein may be seen as holding a non-cognitivist view about our saying ethical things about things, or religious things, but later Wittgenstein cannot. Reading the three Lectures, Putnam is interpreting post-Tractarian Wittgenstein. In post-Tractarian Wittgenstein, the collapse of the fact-value distinction is taking place. Is Putnam’s own position on ethics a form of cognitivism then? Can we draw a parallel with Wittgenstein on religion? Putnam explicitly claims that «The philosophical doctrine of non-cognitivism does not help us understand what religious discourse is really like any more than the philosophical doctrine of incommensurability does.» (Putnam 1992, p. 153). How should we see Putnam’s view then?

One clue is that in his own work on ethics, Putnam attacks non-cognitivism. In Miguens (2020, p. 406), I proposed the following summary of Putnam’s view of ethics: «From his pragmatist, Deweyan, viewpoint, Putnam sees ethics as a “system of interrelated concerns, which are mutually supporting but also in partial tension” (Putnam 2004, p. 22). It is from this perspective that he goes about deconstructing noncognitivist views of value judgments traceable to logical positivism, or criticizing the all-too-frequent restriction of options, when it comes to ethical discussions, to Kantianism and (empiricist) utilitarianism. In his two books on ethics, Putnam (…) welcomes the collapse of the fact-value dichotomy (Putnam 2002) and sets out to explain how ethics can be objective without being metaphysical (Putnam 2004). Following the line of pragmatists, Putnam takes the fact-value dichotomy, assumed in many understandings of the nature of moral judgement to be as dogmatic as the analytic-synthetic distinction [and related to it, one should add]. In our thinking in general, and not just in our ethical thinking, description and evaluation are intertwined. Also, our thinking of practice is “messy” in that it involves many ways of valuing and describing: some valuings are descriptions, and some valuings are not descriptions.» Clearly, not all approaches to ethics focus on the shape of the ethical life, which is what Putnam proposes. But to get from views of ethics centered on rightness of overt action to a view of ethics centered on ethical life, both pragmatism and religion were important for Putnam (Miguens 2020, p. 407).

What matters for me here, is that, for Putnam, religion was important in criticizing noncognitivism. But how? What does Putnam himself do, in coming to terms with the epistemic status of religious belief? Putnam’s wider reflections on religious belief include, in Jewish Philosophy as a Guide to Life (Putnam 2008), not only a reading of Wittgenstein’s Lectures but also interpretations of the philosophies of Franz Rosenzweig, Martin Buber, and Emanuel Levinas. There he is not just doing hermeneutics of the texts of ‘Jewish philosophers’. He is also describing and analyzing religious beliefs and attitudes, namely his own. These exercises exemplify the difference a religious picture might make in people’s lives.

Here are some examples from Putnam’s Jewish Philosophy as a Guide to Life.9 Some people, like Rosenzweig, think that revelation is an ongoing process in their life; that revelation is an event between God and man; that they are always in the presence of God; that the only commandment is to love God; that one should pray to meet the demand of the day; that one should meet the demand of the day with courage and confidence. Some other people obviously do not think so. Some people think, after Buber, that one comes to God by entering into an I-You relationship, and that an I-You relation is never a matter of knowledge. Although it is impossible to theorize about God, one can address God. Man receives not a content but a presence, a presence as strength. Such presence includes three elements: the abundance of reciprocity, of being admitted; the inexpressive confirmation of meaning (the question of the meaning of life vanishes); the idea that this meaning is not a meaning of another life but of this life, this world (Putnam 2008, pp. 55–67). Some people go around thinking it and some other people do not think that at all (they might very well be, for instance, analytic philosophers of religion). There is also Putnam’s profound and sympathetic reading of Levinas. Some people see the “trace” of God in the face of the Other. They think that an infinite willingness to be available to the suffering and neediness of others is demanded of them; that they are obligated with respect to the Other, that they are to be infinitely more demanding of themselves than of others. Many others simply don’t think this.

The religious person uses a picture, lives by a picture, and the presence of such pictures in some persons’ lives, while they are completely absent in other persons’ lives, might amount to gulfs between them. Like Diamond, Putnam is interested in Wittgenstein’s actual description of such gulfs. He approves Wittgenstein’s view of the matter: «it is not that the believer makes a claim and the atheist asserts its negation. It is as if religious discourse were somehow incommensurable, to employ a much-abused word. But there are many theories of incommensurability, and the problem is to decide in what way Wittgenstein means to deny the commensurability or homophony of religious and non-religious discourse.» (Putnam 1992, p. 143). That is why Putnam concentrates his attention on Wittgenstein’s idea of a picture, and the way a picture transforms a person’s life. His incursion into religious belief ends with a view of religion as one’s life being organized by a picture, (in Wittgenstein’s words, «the whole weight might be in the picture», Wittgenstein (1966, p. 72)). So, what matters is the connection between, say, talk of the Last Judgement and what we do, i.e., the connection between language and action—no connection is made between the assertion of belief in a Last Judgement and, e.g., a historical narrative of some kind (as if the Last Judgement man were claiming that something will come about in this same time we live in which will be God’s Judgement, say, in one thousand years). A view of pictures illuminates the fact that some people can do something with the words ‘There will be a Last Judgement’ while some others cannot. As Diamond will stress, in the Lectures, Wittgenstein acknowledges that he himself cannot do anything with the words ‘I will be alive after my death’. He can understand the words, has read the same books the religious man who utters them has, but cannot do anything with those words. For Putnam, ‘doing something with the words’ is made clear by the idea of a picture—a picture guiding a life.

But if the weight lies in the picture, and on action and life, what exactly happens to issues concerning disagreement and incommensurability? Diamond goes one step further than Putnam here, reading the same text. For her, the way exchanges between interlocutors who do not share a conceptual framework go on show important things regarding the nature of thought and language in their relation to world. This too needs to be made clear.

4. Neither Epistemology nor Ethics: Diamond on the Thinkableness of All Thoughts and the Irreplaceability of Pictures

The ideas of religious hinges and of religion as a view of the ethical life exemplify the use of Wittgenstein’s views on religious belief to do epistemology and to do ethics. Interesting insights arise in each case. I want to claim, though, that for Diamond, the most important challenges posed by Wittgenstein’s views on religious belief lie not with the issues of rational or arational grounds of belief on which hinge epistemology centers, nor with discussions around cognitivism and non-cognitivism in ethics, but elsewhere. They lie in making us drop the idea of the ‘thinkableness of all thoughts’ (Diamond 2005, p. 100) and then trying to work out where we stand in thinking about thinking; it is this that opens a place in our thinking for what Diamond calls, after Wittgenstein, ‘irreplaceable pictures’. It is under such shape that Diamond’s reading of Wittgenstein on religious belief illuminates often uncontested assumptions about the nature of human linguistic communication.

Right at the start of Wittgenstein on Religious Belief—The Gulfs Between Us, Diamond gives a name to what, for her, is most challenging in the Three Lectures. She says that a challenge is posed there to an idea of Frege, an idea which is also present in the Tractatus (Wittgenstein [1961] 1974), with a different form. Frege’s idea is (in Diamond’s terms) that there is a stack of thoughts which are available to us all, as thinking beings. A thought «does not belong to me or to some or other community. New thoughts do not become available to us as language develops; and, although languages are different form each other, they are not different in their underlying logical capacities» (Diamond 2005, p. 100). They all allow us to think the same thoughts. This is Frege’s idea of the thinkableness of all thoughts. In Tractarian Wittgenstein, Frege’s idea of the thinkableness of all thoughts becomes that of a single logical space. In the Tractatus, the logical space is the space of possibilities for the existence and non-existence of states of affairs. According to Tractarian Wittgenstein, each proposition determines a place in logical space. The idea of logical space does much work in the Tractatus, going as far as the conception of the ‘world’ (The facts in logical space are the world (1.13)). A challenge to the thinkableness of all thoughts for all, or to the idea of a single logical space, is, thus, a serious challenge. Diamond speaks of it as ‘a challenge to the idea of a common intellectual life of humanity’ (Diamond 2005, p. 100). Without the thinkableness of the same thoughts for all thinking beings, a ‘common intellectual life’ of humanity seems at risk.

Speaking of a ‘common intellectual life’ of humanity, Diamond is introducing a way to think of relativism. We could think of relativism along Rortyan lines: relativism is then the idea that an assertion being warranted is a matter of which community we feel solidarity with. We could think of relativism in terms of objects and universes of discourse: relativism is then the idea that A or B (say, dinosaurs or genes) are social constructions; they are not out there apart from our needs and interests; what there is, and what we may talk about, depends on universes of discourse. But relativism may also be formulated, in Diamond’s terms, as the idea that there is no common intellectual life of humanity.

Let us look closer at the idea of a common intellectual life of humanity. As Diamond sees it, it means the following: if it is possible for person A to believe that p, and to believe that not p, then p, and not p, are simply thinkable; if they are thinkable, then they are thinkable for all thinkers. It is such connection between the idea of a thought and its availability to ‘all thinkers’ that makes Diamond speak of the idea as Fregean—for Frege, a thought (unlike a Vorstellung) does not belong to a bearer, i.e., to a particular, private, mind; a thought is a thinkable which any thinker may think. Only thoughts, not Vorstellungen, bring in, according to Frege, the question of truth. The problem is that this idea seems to rule out the crucial examples in the Three Lectures. Cases such as those of the religious and the non-religious man talking together about the Last Judgement, or about ‘seeing each other after death’, should not exist, since in these cases there seem to be thoughts that are thinkable by one interlocutor but not by the other. Diamond stresses that such unthinkability is not a psychological matter. The question is not that someone lacks mastery of some complex topic, as in when a person knows much more calculus or quantum mechanics than another person. Wittgenstein’s examples are examples of situations where someone believes something whereas others cannot either believe it, or deny it, or consider its truth or falsity as not yet determined. This is the actual nature of Wittgenstein’s ‘gulfs between us’ as seen by Diamond. Now, as Diamond puts it, if Frege is right about the common intellectual life of humanity, then there simply should not be such gulfs. She herself wants to keep the idea of a common intellectual life of humanity. But is it possible to do it, given such cases? Is there a possible path to keeping the idea of a common intellectual life of humanity, if we follow Wittgenstein on certainty, hinges, and pictures? This is the question that interests her.

Diamond calls our attention to how Wittgenstein’s positions on thought-language-world relations developed. He starts (in the Tractatus) from a conception of a logical space as a space where all propositions are interconnected. She thinks it is precisely such conception of the logical space that is undergoing criticism in the Three Lectures. That is in fact, for Diamond, what makes the Lectures particularly significant; their importance clearly goes beyond philosophy of religion. What steps does Wittgenstein take, then, toward conceiving the alternative to the singleness of ‘logical space’?

Diamond suggests that in the Three Lectures, Wittgenstein claims that there are different ways to stand before the man who believes in the Last Judgement.10 It is not the case that we are faced with the alternative of either seeing him as merely articulating word-sounds or seeing him as being fully involved with meaning and engaged in argument with us. It is helpful to bring Putnam back in at this point. In fact, interpreting Wittgenstein’s Three Lectures, Diamond herself is engaging with Putnam, and not just with Putnam on religion (see Diamond and Gerrard (2003), How Old Are These Bones? Putnam, Wittgenstein and Verification).

It is important to keep in mind that both Putnam and Diamond are concerned with whether Wittgenstein’s position on language-thought-world relations is realist or idealist, eventually relativist. The case of the old bones, which Diamond brings into the discussion, is «a case invented by Hilary Putnam in the course of his long-running philosophical debate with Richard Rorty. Back in the seventeenth century, strange bones have been dug up at Whoozie, and someone wonders how old the bones are. We now know that they are over a million years old; we have used twentieth-century techniques to establish their age. In Newton’s time, there were no such techniques. But suppose someone to have speculated about the age of the bones. Putnam says that if this person had entertained the idea that the bones were a million years old, he would have been right, and if he had rejected it as absurd, he would have been wrong. [Diamond tells us that since she refers to this speculator many times, she has given him a name: Leibniz.] Putnam was responding to a suggestion by Rorty that we should redefine ‘true’ to ‘chime with’ Heidegger’s claim that Newton’s laws became true through Newton’s work, and that before Newton’s discovery they were neither true nor false. In developing Heidegger’s point, Rorty had said that, if the Latin sentence which Newton used in the seventeenth century to state the principle of inertia had been uttered by someone in the tenth century, it would not then have been a truth-value candidate. It became a candidate for being true or false when there developed a set of coherent and useful practices within which there could be embedded uses of that sentence to make assertions. So, if we follow Rorty’s recommendation, we should say that the sentence ‘The bones found at Whoozie are a million years old’ became a candidate for being true or false during the twentieth century; and it seems then that we should reject Putnam’s claim that, if Leibniz had said that the bones were a million years old, he got something right.» (Diamond and Gerrard 2003, pp. 99–100).

Illuminated by the history of science context above, these questions in the Three Lectures are brought together with the question what counts as thinking the same (whether something does), even if theories, or frameworks, change profoundly. The example discussed in the article involves considering whether a person (the person Diamond calls ‘Leibniz’) thinking in the 17th century that ‘These bones are one million years old’, could mean by that what we ourselves mean now. What is at stake it “meaning the same” and thinking the same? Does the question How Old are these Bones? make sense ‘across time’ and scientific change? In Diamond and Gerrard (2003), the authors side with Putnam when he argues against those philosophical views which tie the content of truth-claims to the methods of investigation and verification available at a particular time. Putnam thinks that such views threaten ‘our idea of human communication’. Our idea of human communication involves, according to him, the possibility of communication across different cultures and across periods of time.

Like Putnam, Diamond is critical of the idea that people coming from temporally distant cultures cannot mean, by their words, the same thing that we do, even if they do not share modes of investigation and verification with us. So she spells out what she thinks could be making us think that they cannot mean, by their words, the same thing that we do (this is, in other words, the way of thinking that lies behind Rortyan relativism). She describes this as ‘a way of arguing Wittgenstein can be seen as committed to’. It is a way of arguing that she herself wants to reject. Such way of arguing starts from two premises:

(1)

Modes of investigation and verification belong to the grammar of the language.

(2)

The grammar of the language determines what we are thinking about.

The conclusion to draw from such premises is that if people do not share modes of verification, then they cannot be thinking the same. They cannot share a belief, or hypothesis. So if we start with such assumptions, we will think that ‘Leibniz’ in the XVIIth century and ourselves are not thinking the same when we think, and say, ‘These bones are one million years old’, as we are facing these bones. If we think along these lines, then we accept that coincidence in the shape of words (‘saying the same’, in English, or in German, e.g., ‘these bones are one million years old’, or ‘diese Knochen sind eine Million Jahre alt’) does not per se indicate that what the people uttering the words are thinking is the same.

Diamond does not accept this supposedly Wittgensteinian way of thinking. She agrees with Putnam on the bones example: both Putnam and Diamond claim that we can identify what those people (‘Leibniz’, in this case) are speaking about and saying by using our own way of speaking. We can do it even if our ways of speaking depart significantly from the ways of speaking available to the people whose thoughts we are describing.

Although the questions above belong in a larger context of discussion, they form the background of Diamond’s Wittgenstein on Religious Belief—The Gulfs Between Us, as she reads the text. In fact, they signal where Diamond wants to go. I will now go back to the three lectures and to religion, and to Wittgenstein, as he is read by Diamond. We know that Diamond does not want to drop Frege’s idea of a ‘common intellectual life of humanity.11 She claims that we can identify what other people are speaking about using our own way of speaking, even when our ways of speaking depart significantly from what would be available to the people whose thought we are describing. This being so, she needs to replace Frege’s picture of the stack of thoughts, or the Tractarian idea of the singleness of the logical space, with something else. She finds what she needs in Franz Rosenzweig. He provides her with an alternative metaphor to the Fregean picture of the stack of thoughts: that of the movement of intertwining of words in human lives (Diamond 2005, pp. 114, 134). It is such intertwining of words in human lives that allows us to account for the sense made by the Last Judgement man in the Lectures, still holding on to the idea of a common intellectual life of humanity.

What Diamond actually does to pursue this idea while reading Wittgenstein on religious belief, is to look at exchanges between interlocutors one by one, i.e., in the terms of Kinzel and Kusch (2018), to look at exchanges in a non-idealized way (I will develop this in Section 5). She acknowledges that, in each case, interlocutors may stand in different relations towards the words uttered, including their own words. This is a recurring theme in Diamond 2005. Making a judgement which one backs up by evidence, in a situation where both interlocutors converge in what counts as evidence, is but one sort of relation to the words of others, and to our own words. It actually presupposes (to go back to the former metaphor) that interlocutors are placed at a very ‘near location’ in the space of judgement. This is illustrated, for example, by the case of the German airplane in the three Lectures. Yet that is not the only position in which we stand towards the use of words by others, and by ourselves, when we are engaged in actual exchange and in (linguistic and conceptual) ‘negotiation’.

I will now move on to another very important example in Diamond’s interpretation of the Lectures. Someone says to their interlocutor “We will see each other after death” (the person is leaving for China and they don’t know if they will see again the interlocutor they are addressing, Wittgenstein (1966, pp. 72–73)). The example marks the place in the Lectures where Wittgenstein discusses pictures.12 It is one of the examples around which Diamond tries to articulate the connection between the absence of a single logical space and the ‘essentiality’ of pictures (another example which appears near this one is ‘God sees everything’). She suggests that in his exchanges with Smythies in the third lecture, Wittgenstein is doing exactly this: exploring the essentiality (or irreplaceability) of some pictures and not others. She analyses this in terms of what we do with our words as we go on using them.

Diamond thinks that there can be what she calls ‘different descriptions of a mode of thought and talk shared by a community’ (Diamond 2005, p. 117). She also thinks that there are two notions available of a kind of ‘singleness in human language’ or ‘move toward a shared language’. The first is the Fregean-Tractarian one, which is that of a single logical space. The second is the ‘Rosenzweig’ one, based on the movements of words as these are intertwined in human life (that is, as I speak to you, and you speak to somebody else, and so on). In the second notion, «Our words, one might say, grow, as we make such connections, and their words grow, as they make other connections.» (Diamond 2005, pp. 113–14). In other words, Rosenzweig’s metaphor of the intertwining applies to actual exchanges and negotiations, as these take place in conversations.

One note is needed here: Diamond is quite cautious when she describes spaces and distances in our meeting with others in thought. She says she has not «tried to show whether there is or isn’t a common intellectual life of humanity—just hope to have shown that Wittgenstein’s rejection of the Fregean-Tractarian conception of belief does itself force us to drop the idea.» (Diamond 2005, p. 115). But if we are not dropping it, then we have to de-idealize what is going on in particular exchanges of words between interlocutors. So, imagine we ask: does a non-religious interlocutor understand the words “We will see each other after death” when his religious interlocutor utters them addressing him? Do we understand the words? Do we simply not understand them?

As Diamond puts it, we do not necessarily understand what is being said if we know what ‘alive’ means, what ‘death’ means, and what ‘after’ means (Diamond 2005, p. 109). She proposes that the religious interlocutor himself may take their own words to be ‘untransparent’. The ‘transparency’ case is not the only case possible if we understand, as it were, the meaning of words, even our own words. She connects this with Wittgenstein’s rejection of the idea that a theory of meaning could provide a sort of calculation of what has to be intelligible.13 This is one reason she stands with Putnam when Putnam rejects Rorty’s examples of (supposed) incommensurability (Diamond 2005, p. 113). Use of words, and what we do with words, can be conceived in the terms of ‘calculation’ but it can also be seen as exemplified by the use of words in pictures, as in the examples above. Like Putnam, Diamond thinks Wittgenstein’s idea of a picture is very important for his views on religion. But she thinks the idea is important first in thinking about meaning and understanding—not just, or directly, about ethics. It is in that context that she tries to reconstitute the discussion between Wittgenstein and Yorick Smythies that may have originated the notes that compose the third lecture.

According to Diamond, Wittgenstein is, throughout the third lecture, trying to come to terms with ‘Smythies muddle’. Smythies muddle is based on the idea of a super-picture. A super-picture is a picture that leaves no room for the question of what a picture is of (Diamond 2005, p. 116). There are misunderstandings between Smythies and Wittgenstein: Smythies apparently takes Wittgenstein to think that pictures are replaceable, whereas Wittgenstein in fact ‘thinks higher of pictures’ (these are Diamond’s terms). Thinking higher of pictures amounts to thinking that a picture can being essential, which means not replaceable. Some pictures, according to Wittgenstein, are not accompaniments. They are irreplaceable. But what does that mean?

When Smythies gives Wittgenstein the example of someone saying ‘We will see each other after death’ as an example of the use of a picture, the question arises what that picture is a picture of. Wittgenstein says to Smythies: you gave me two phrases for what goes on after I die: ‘simply ceasing to exist’, and ‘being a disembodied spirit’. He also says: You (Smythies) called them [the two phrases] two ideas, but one does not read off a picture from the phrases qua phrases. Not reading off a picture from the phrases qua phrases, is, for Diamond (and, she thinks, for Wittgenstein), what makes a picture essential, irreplaceable, and not a mere accompaniment. Saying that we cannot ‘read off thoughts what they are of’ is Wittgenstein’s way of going against Smythies view of a super-picture. In the example ‘We might see one another after death,’ this means that the meaning cannot be simply read off the phrase used.

Diamond goes through several examples of pictures at work. She looks for pictures which are irreplaceable pictures in Wittgenstein’s sense (Diamond 2005, pp. 118–19). She asks: what kind of work might a picture do? A picture can be a mere illustration, or a way to convey a message. A picture can be an attitude. A picture can perhaps strengthen dispositions. A picture might belong in a narrative (say, in the narration of the life of Jesus). It is in fact in the context of this discussion that Wittgenstein uses the expression that Putnam retains for thinking of religion as a view of the ethical life (that what marks a picture as essential is that ‘the whole weight might be in the picture’). But according to Diamond, essentiality in the sense that matters here has to do with not knowing immediately what is meant by that use of words. This happens in such way, she claims, that one could not leave open that a whole body of doings, sayings and attitudes could be the same with or without use of the picture.

At this point in her article Diamond brings in thinking about God as this is done, and discussed, by George Berkeley, by Franz Rosenzweig, as well as by theologians Kornelius Miskotte and Helmut Gollwizer. Cases of thinking of the Divine in anthropomorphic terms are important for the points she wants to make. She thinks there is a contrast, in thinking and talking about God, between ‘God’ as a term for any conception of the divinity in any language game (‘any divinity’, divinity in generic terms, devoid of anthropomorphic traits) and the ‘God that has a name’. She is interested in what is in a name, a proper name. She asks us to do the following. Imagine that we loved a person and she dies. Could one replace the person who dies with something with the exact same features? If we answer no, because that does not capture what is lost when someone dies, we are getting at the non-replaceable particularity of that person, which is what is marked by using a proper name (for a person, or for God). Our understanding of what the person was cannot be captured in general terms—what the person was is something we evoke with her name.

Another example of Diamond is that of a person’s thinking of beauty as marked by their acquaintance with George Elliot, and then thinking of beauty differently, given the encounter with such ‘magnificently ugly woman’. The particularity of the encounter transforms the person’s concept of beauty (Diamond 2005, p. 125). Speaking about God by using a name, or using anthropomorphic language, or of Elliot as marking our concept of beauty, are pictures where something cannot be ‘rationally eliminated’. Those pictures are not accompaniments. This, according to Diamond, is what Wittgenstein is after in discussing the essentiality of pictures. (Diamond’s critique of the critique of anthropomorphism in religion, arises as well, in that part of the article, from her accepting that certain pictures are indeed essential and irreplaceable in the sense that matters for Wittgenstein).

To sum up: Diamond tries to help us think about what is at stake when a ‘picture-language’ (her term) is used (for religious belief, but not only for religious belief). She uses the example of thinking about God in a picture-language (e.g., we speak and think of God as having a name, or of God as speaking, or as acting in history). One first thing to acknowledge there is that the words we use are the words we use otherwise, in other circumstances. These are the words that we are able to speak; these are our words. Yet, in her terms, ‘the doings, the sayings and the attitudes could not be the same with or without use of the picture’. Granted, such account of the essentiality of a picture is plainly circular. Yet, according to Diamond, it could not be otherwise: if a picture lies at the base of all of someone’s thinking, as in the case of a religious person, this might be reflected in the fact that a picture-language may be necessary to understand a picture-language.

There are further important consequences of Diamond’s approach to Wittgenstein’s discussion of pictures in this context. One is that she ultimately rejects an absolute contrast between pure ‘picture-less-ness’ in language (say, in science, or in philosophy, conceived as purely rational, and stripped of any pictures) and picture-language. This means, among other things, that she ultimately rejects an absolute contrast between the philosophical language game and other discourses, with which philosophy concerns itself, such as religion. For Diamond, there simply are limits to the usefulness of the picture of a border between such discourses.

5. Conclusions—Diamond on Religious Belief and the Natures of Human Thought and Human Communication

What Diamond does around Wittgenstein on religious belief is complex. In fact, it goes so much beyond what is usually identified as philosophy of religion that it could even be seen as unrelated to current discussions in the field. Analyzing Wittgenstein’s Lectures on Religious Belief, Diamond investigates thought and its expression, as well as assumptions regarding the nature of human communication. She thinks that religion puts the limits of thought and language to test, hence the indirect approach she favors. This position is not new in the history of philosophy. Still, it is rather unique when seen against the background of current philosophy of religion. Its full reach may thus go unperceived. So I will finish by trying to make explicit some connections between Diamond’s claims around religious belief and issues discussed in the analytic literature on disagreement from the last twenty years. I will resort to Carmona and Villanueva (2023), which focuses on intercultural disagreement and intercultural dialogue, in a dialogue with Kinzel and Kusch (2018), which focuses on cases in the history of science (supposedly) motivating relativism. I will resort to these authors because they, like Diamond, seek an answer to the question whether communication is possible between partners who do not share frameworks (Carmona and Villanueva 2023, p. 54).

We saw that reading Wittgenstein on religious belief, Diamond has her eyes on how speakers stand towards one another in conversations and disputes, as well as on where conversations and disputes go when disagreements occur. Diamond’s close look in fact corresponds to Kinzel and Kusch’s proposal to de-idealize disagreement (Kinzel and Kusch 2018, pp. 60–62). Kinzel and Kusch believe de-idealization is needed to counter the diet of examples of philosophical discussions of disagreement. We can see what they mean by idealization by thinking of how, e.g., the disagreement between Galileo and Cardinal Bellarmino is usually depicted within philosophy. Their disagreement is often seen as reducible to something as simple as accepting or not accepting Revelation (as epistemic standard) and thus accepting or not accepting heliocentrism (because heliocentrism contradicts what is written in the Bible). Kinzel and Kusch call attention to the fact that such philosophical account of disagreement contrasts sharply with work on the same case as it is carried on within the history of science: «Open any standard historical account of the episode, and invariably you will find something like the following observation: ‘This was a controversy involving issues of methodology, epistemology, and theology as well as astronomy, physics, and cosmology’ (…) Authors influenced by SSK [sociology of scientific knowledge] add period-specific relationships between patrons and courtiers; traditions of instrument-making; the tensions between different religious orders; the politics of the papal court; the political problems between Spain and the Vatican; the Thirty-Year War; and much else besides». In contrast, in philosophers’ reconstructions, «all of these issues are set aside as irrelevant (…) religion is turned into an epistemic system; epistemic principles are treated as isolated or isolatable entities with fixed implications. And, perhaps most importantly, disagreement as an activity—as an extended series of events, a debate—is reduced to a state: Bellarmine accepts Revelation, Galileo rejects it. There is no discussion of how the two rivals responded to one another, which argumentative strategies they chose to employ to convince one another, or whether and how they modified their strategies over time.» (Kinzel and Kusch 2018, p. 51). This is what idealizing disagreement means.

Kinzel and Kusch’s proposal is thus aimed at de-idealizing disagreement: they suggest that exchanges be considered one by one, and disagreement analyzed as an activity and not as a state. Building on Kinzel and Kusch’s proposal, Carmona and Villanueva isolate specific assumptions about the nature of human communication which they think lead to the idealized view.14 The assumptions are the following: to exchange reasons, epistemic principles, or a set of justificatory standards, must be shared between partners. To be reasonable simply is to defend claims with reasons which are in line with such shared epistemic principles or justificatory standards. Communication between partners always ultimately aims at coordination, even where disagreements occur.15 Given those assumptions, the aim of engaging in conversation and argumentation cannot be other than to defeat the other party. Carmona and Villanueva call this the Disagreement Defeat Model (DDM). According to the Disagreement Defeat Model, there are only two ways to go in linguistic exchanges where disagreements occur: either I stay where I started and the interlocutor comes to me, or I am defeated and join their position. Carmona and Villanueva claim that such Disagreement Defeat Model does crucial work in philosophical views of communication, both in analytic philosophy (e.g., in contextualism in philosophy of language and in hinge epistemology),16 and elsewhere (e.g., in a hermeneutics framework). It is against the Disagreement Defeat Model that they put forward their Situated Judgement Model of Dialogue.

The Situated Judgement Model rests on the fact that communication does take place, all the time, between parties that have totally different epistemic, ethical or aesthetic principles and that do not share a form of life. This happens without any pre-established harmony, or prior consensus. This is illustrated by the initial example in this article (that of Wittgenstein’s Last Judgement man talking to the non-religious interlocutor), as by the Kinzel and Kush’s history of science cases, as by the examples of intercultural understanding in Carmona and Villanueva (2023).17 In order to account for the fact that communication does take place in such circumstances, it is important, according to Carmona and Villanueva, to bear in mind that epistemic agents do not have to be aware of features influencing them, or of principles guiding them. In fact, «Judging together is the only way to build upon differences without any kind of pre-established harmony» (Carmona and Villanueva 2023, p. 54). Evoking § 242 of the Philosophical Investigations,18 they see judgement as a move based on agreement on form of life, and not on explicit argument between parties. Thus, a shared frame of reference neither is needed for communication to start nor needs exist at the end of the exchange. Also, the linguistic exchange taking place between parties should not be viewed as having as its main purpose to defeat (or persuade) the other party. Under this light, disagreements between parties with different standards can be persistent and rational without aiming at the fusion of horizons or the coordinating of standards.

Granted, much further work needs to be done around such alternative idea of communication in order to have to have a counterpart to everything that Wittgenstein had in place in connection with his Tractarian view of the single logical space (which was, after all, a whole conception of thought, language and logic). Naturally, I won’t pursue such a task here. My suggestion is just that Carmona and Villanueva’s Situated Judgement Model, as it is proposed against the Disagreement Defeat Model, is a starting point for articulating the connection between Diamond’s interpretation of Wittgenstein on religious belief and current debates on disagreement in analytic philosophy. Diamond is in fact claiming that a very controversial notion of linguistic communication is simply taken for granted in many areas of analytic philosophy (namely in philosophy of language and epistemology).

In this article, I went some way into explaining why Diamond does not approach religion though textbook problems and arguments, centering on whether God exists or whether religious belief is more or less warranted. Neither does she simply claim, like Putnam, that religion simply is ethics. I showed that there are reasons for all that. One main reason is that Diamond believes that epistemology, or ethics, cannot be isolated from questions regarding uses of language and issues of making sense. Also, for her, there is no separation between issues specific to philosophy of religion and the fundamental issues of philosophy, which concern thought and language and making sense, and eventually coming to new forms of thinking and understanding, and of seeing the world anew (it could be, e.g., seeing goodness in what there is). So, unless we start with an epistemological view which is not Wittgenstein’s, we will see that what is at stake in religious belief and religious thinking is not just, or mostly, a matter of evidence, doubt, and certainty, or rational or arational grounds of belief. Wittgenstein said in the Lecture of Ethics: «Now I am tempted to say that the right expression in language for the miracle of the existence of the world, though it is not any proposition in language, is the existence of language itself.» (Wittgenstein 2014). The existence of thought is the other side of the same miracle. The relation between thought, language, and the world is Diamond’s topic, writing on Wittgenstein on religious belief and putting forward her two proposals: that there is no fixed totality of all thoughts and that pictures are irreplaceable in our thinking.