Wittgenstein on the Grammar of Unshakeable Religious Beliefs

Abstract

The paper offers a reinterpretation of Wittgenstein’s 1938 lecture on religious belief and challenges a prominent view that it commits Wittgenstein to a form of non-cognitivism and/or that it reflects a lack of understanding of religious practices. It further argues that the lecture is not in tension with Wittgenstein’s later views on the nature of philosophy.

1. Introduction

Wittgenstein’s remarks on religion have been a source of inspiration for many in their own thinking about issues in the philosophy of religion. However, philosophical positions such as fideism, (non-)cognitivism, and relativism have also been attributed to Wittgenstein himself.1 However, there is reason to believe that Wittgenstein neither attempted to provide a substantial philosophy of religion nor tried to describe the complete grammar of religious statements. The former contradicts his ’metaphilosophy’ most clearly formulated in his Philosophical Investigations, while the latter makes Wittgenstein unnecessarily ignorant of religious practice (Richter 2001).

My goal is limited—I do not wish to evaluate everything Wittgenstein has said about religion but rather how he treats (a certain form of) religious beliefs in a lecture given in 1938, Wittgenstein (1966).2 This lecture has been used to argue both that Wittgenstein has a substantial philosophy of religion and/or that he attempts to describe the complete grammar of religion Carroll (2014). My aim is to show that both approaches are misguided. I think the lecture is particularly interesting because it is one of the few instances where Wittgenstein addresses the topic as a philosopher speaking to an audience.

There is no good reason to believe that Wittgenstein deviates from his general philosophical project, most clearly expressed in Philosophical Investigations. Wittgenstein’s focus is not directly on the metaphysics or epistemology of religious beliefs but on their grammar. It is how we talk about religious beliefs that occupy him.3

Even if one agrees with this, there are still several pitfalls; Wittgenstein does not claim that how we talk about religious beliefs requires them to be unshakeable or contentless—concepts we will examine later. In a good Wittgensteinian spirit, it is important that we are sensitive to the example we are investigating and do not hastily move to a generalized theory.

It is worth noting that the language game Wittgenstein discusses in this lecture appears to represent a language game that he found particularly attractive.4 This can make it difficult to separate the philosopher from the religious thinker—a distinction I will claim is crucial for interpreting Wittgenstein’s remarks. One should distinguish between Wittgenstein’s personal thoughts on the possibility of religion and his philosophical remarks (Richter 2001). Overlooking this distinction can easily lead to misunderstandings.

Understanding Wittgenstein’s claims is not only interesting for exegetical reasons but for the insight it provides into the topic itself. It has consequences not only for how we should understand certain expressions of religious beliefs but also for expressions of beliefs more generally. By making us realize how one language game with beliefs functions, Wittgenstein gets us to recognize the multifaceted aspect of how we speak about beliefs (Citron 2012). For example, it is wrong to claim that all beliefs necessarily aim at truth.5 To say that they must do this would be to go against the grammar of beliefs.

2. Lecture on Religious Beliefs

I will begin by providing a review of some central aspects of the lecture.

A person says, “I believe in Judgment Day.” He says this because he always has a thought of judgment before his mind—it regulates everything he does. It is the backdrop for his actions, giving his life focus and meaning. When he believes in Judgment Day in this sense, does it mean he believes that a phenomenon, Judgment Day, will occur? Well, maybe he does, maybe he does not. The essential aspect of his linguistic expression “I believe in Judgment Day” is not this factual aspect, but the role that thoughts of Judgment Day play in his life. In this way, the expression plays a completely different role than if he says, “I believe that plane in the sky is German.”

In Wittgenstein’s own words:

Suppose somebody made this guidance for this life: believing in the Last Judgement. Whenever he does anything, this is before his mind. In a way, how are we to know whether to say he believes this will happen or not?

Further:

But he has what you might call an unshakeable belief. It will show, not by reasoning or by appeal to ordinary grounds for belief, but rather by regulating for all in his life.

This is a very much stronger fact—foregoing pleasures, always appealing to this picture. This is [sic] one sense must be called the firmest of all beliefs, because the man risks things on account of it which he would not do on things which are by far better established for him. Although he distinguishes between things well-established and not well-established.

Lewy: Surely, he would say it is extremely well-established.

First, he may use “well-established” or not use it at all. He will treat this belief as extremely well-established, and in another way as not well-established at all.

If we have a belief, in certain cases we appeal again and again to certain grounds, and at the same time we risk pretty little—if it came to risking our lives on the ground of this belief.

There are instances where you have a faith—where you say “I believe”—and on the other hand this belief does not rest on the fact on which our ordinary everyday beliefs normally do rest.

Wittgenstein (1966, p. 54)

One might be tempted to think that beliefs can be compared by looking at the reasons a person is willing to give in their defense. However, it is unclear whether rational reasons based on evidence are relevant for a belief of this type.6 Rather, what characterizes such a belief is not the reasons one is willing to give but the risk one is willing to take on their behalf. And, as Wittgenstein notes, the risk one is willing to take is not necessarily aligned with the strength of the evidence one might have for the belief. One might be willing to sacrifice one’s entire life for a belief in this, in the normal sense, unfounded belief in Judgment Day, which one is not willing to do on behalf of the best-established scientific claim.

Note Wittgenstein’s delimitation: “There are instances where you have a faith.” Wittgenstein’s main focus in the lecture is such a case, where a person has an unshakeable belief in Judgment Day.

Wittgenstein wants to show that one can think very differently about a phenomenon without it being obvious that one believes the opposite of the other:

Suppose someone is ill and he says: “This is a punishment,” and I say: “If I’m ill, I don’t think of punishment at all.” If you say: “Do you believe the opposite?”—you can call it believing the opposite, but it is entirely different from what we would normally call believing the opposite.

I think differently, in a different way. I say different things to myself. I have different pictures […].

Here believing obviously plays much more this role: suppose we said that a certain picture might play the role of constantly admonishing me, or I always think of it. Here, an enormous difference would be between those people for whom the picture is constantly in the foreground, and the others who just didn’t use it at all.

Wittgenstein (1966, pp. 55–56)

Even though it is tempting for an atheist, linguistically speaking, to say that he does not believe in Judgment Day, this is a misleading way of speaking if his conversation partner uses his belief in Judgment Day in the aforementioned sense, namely, as regulating his life.7 Not only do we get an analysis of why debates between religious and non-religious individuals often appear unsatisfactory—they talk past each other—but we also get an analysis of what it means to actually understand the expression “I believe in Judgment Day.” In our language game, it is an expression of an unshakeable belief that is not essentially about believing that a phenomenon, Judgment Day, will occur.

Wittgenstein also wants to give us a reason to believe that what makes the aforementioned case religious is not primarily its content. We can imagine a person having a belief with the same content, understood in the normal sense—not as a set of background images and relationships and a life-regulating goal but as expressing a proposition—without it necessarily being religious:

The point is that if there were evidence, this would in fact destroy the whole business.

Anything that I normally call evidence wouldn’t in the slightest influence me.

Suppose, for instance, we knew people who foresaw the future; make forecasts for years and years ahead; and they described some sort of a Judgement Day. Queerly enough, even if there were such a thing, and even if it were more convincing than I have described, belief in this happening wouldn’t be at all a religious belief.

Suppose that I would have to forego all pleasures because of such a forecast. If I do so and so, someone will put me in fires in a thousand years, etc. I wouldn’t budge. The best scientific evidence is just nothing […].

We don’t talk about hypothesis, or about high probability. Nor about knowing.

In a religious discourse we use such expressions as: “I believe that so and so will happen,” and use them differently to the way in which we use them in science.

Wittgenstein (1966, pp. 56–57)

The grammar of “I believe” in such a religious discourse is of a different nature.

Furthermore, Wittgenstein claims that there is something ridiculous about a person who bases this religious belief on reasons:

But I would ridicule it [making the religious belief reasonable], not by saying it is based on insufficient evidence. I would say: here is a man who is cheating himself. You can say: this man is ridiculous because he believes, and bases it on weak reasons.

Wittgenstein (1966, p. 59)

There is something amiss by trying to justify or undermine a religious belief by evaluating it by the standards of rationality. We are tempted to do this because we are misled by how the linguistic expression of belief functions in more ‘regular’ cases of belief. But a religious belief of this type is different; it belongs to a different language game. Given how Wittgenstein expresses himself here, it seems prima facie that he is no longer speaking exclusively about our Judgment Day example but claims that, in all religious discourse, belief takes on a different role than in science. This is not to say that all religious beliefs are unshakeable. All he says is that it is a different language game, where our usual use of words like hypothesis, high probability, and knowledge does not play a central role.8

We are dealing with a well-known Wittgensteinian insight; what we believe is essential for some linguistic expressions—such as understanding and rule-following—turns out to be not (exclusively) a mental state but equally dependent on elements outside the person and how he acts and orients himself in life. By being sensitive to language use, Wittgenstein helps us realize that the linguistic expression of belief is a more multifaceted expression than it initially appears.

3. The Grammar of Religious Belief

Despite the preceding, it is my opinion that it is wrong to claim that Wittgenstein in the lecture claims that a religious belief must be unshakeable. Given the lecture as we have it recorded, it is an understandable reading that Wittgenstein claims that having a religious belief is to believe something unshakeably.9 However, upon closer examination, this is a hasty conclusion.

From our earlier quote, he says that an unshakeable belief—its grammar—is characterized by three elements:

• It does not rest on normal grounds for belief.
• It is totally regulating for one’s life.
• One is willing to take great risks based on it.

I see little reason to attribute to Wittgenstein what appears to be, at best, a linguistic surprise, and at worst, an overly demanding and narrow understanding of the religious.10 This is while recognizing that the example we are discussing involves a person who has an unshakeable belief. However, we should not jump from this example to an uncharacteristic generalization of everything we would call religious. This is despite Wittgenstein seeming to claim that in all religious discourse, our usual use of words like hypothesis and knowledge has no central role.

To advance the discussion, I will introduce a set of distinctions intended to clarify what Wittgenstein is—and is not—saying. These distinctions are not meant to establish precise boundaries that Wittgenstein himself would endorse. On the contrary, as I will argue, a central point is that Wittgenstein does not commit himself to any fixed or even loosely defined boundary for what constitutes religious belief—not even in terms of unshakeable belief. That said, this does not preclude the possibility that, within a particular language-game, it may be entirely appropriate to say that a given expression is used with sharp boundaries. The aim of these distinctions, then, is not to impose a rigid framework but to illuminate the flexibility and context-dependence of Wittgenstein’s approach.11

Here are the distinctions:

LIFE-REGULATING: A belief that regulates your entire life.

RISK: A belief you are willing to take great risks based on.

GROUNDLESS: Evidence is irrelevant to the language game the belief is part of.12

UNSHAKEABLE: LIFE-REGULATING + RISK + GROUNDLESS

RELIGIOUS: ?

It may well be that any belief satisfying UNSHAKEABLE would also be recognized as a religious belief. However, it is insufficiently motivated for textual reasons to attribute to Wittgenstein the stronger claim that a belief must be unshakeable in order to count as religious.13 Such a reading would additionally risk imposing a conceptual rigidity that Wittgenstein’s later philosophy resists.

Throughout the lecture, Wittgenstein has tried to show that the content—understood as a proposition—is not (exclusively) what makes the belief RELIGIOUS. This is what the example of having evidence that the phenomenon of Judgment Day will occur is meant to illustrate. This also illustrates that LIFE-REGULATING and RISK are not enough either. We can easily imagine that such a belief about Judgment Day could/would greatly regulate your life if you had, say, scientific evidence that something like this would occur. We can also imagine that you would be willing to take great risks based on this compared to other beliefs you have.14 However, we still do not have UNSHAKEABLE because we lack GROUNDLESS.15

There is clearly something in our linguistic practice that coincides with calling something religious if it satisfies UNSHAKEABLE. Perhaps one might claim that:

RELIGIOUS = UNSHAKEABLE + specific content

Only linguistic statements containing certain content, concerning God, immortality, and the like, will qualify something as RELIGIOUS,¹⁶ this despite it not being correct to attribute the belief to the person because he has a specific mental content that, in our example, the phenomenon of Judgment Day will occur.

On the other hand, one might argue that we should not add a requirement for specific content, as we can imagine linguistic situations where we, for example, would say that a person treats a political issue or whatever it may be as something religious and, consequently, that a content-based delimitation would be difficult and perhaps artificial.

Whatever one might think about this, my goal here is to understand Wittgenstein’s position, and it seems incorrect to attribute to him the claim that RELIGIOUS = UNSHAKEABLE.17

One gets the impression that Wittgenstein on a personal level can understand cases of the religious as UNSHAKEABLE.18 Nevertheless, we must not conclude from this that Wittgenstein believes that the religious language game can only be played in this way: “I would definitely call O’Hara unreasonable [that tries to make religious belief reasonable]. I would say, if this is religious belief, then it’s all superstition.” Wittgenstein (1966, p. 59). Even though Wittgenstein believes such a belief would be superstition, he does not say that it therefore would not qualify as RELIGIOUS.19 Rather, he seems open to the fact that it can be something we would call a religious belief.

It is important that we have this distinction clear—the distinction between what we are interested in in this text, which is Wittgenstein’s analysis of the grammar of religious beliefs, and on the other hand, what Wittgenstein himself saw as a possible religious belief in his own life. This is a crucial distinction not only for understanding what Wittgenstein says but also for evaluating criticism that has been directed at Wittgenstein’s ideas. For example, it is common to use Wittgenstein’s notes on religion, where he himself reflects on the possibility of the religious, as an expression of what he believes is the only possible grammar.20 I believe it is poorly motivated to attribute to Wittgenstein the position that he thinks UNSHAKEABLE is a prerequisite for any linguistic description of religious beliefs. Rather, it is an expression of one way Wittgenstein himself can see the possibility of having a religious belief.21

In Child (2011, pp. 224–5), we get the following passage, motivated by a common criticism that Wittgenstein misunderstands the nature of religious beliefs:22

Wittgenstein is plainly right that religious belief involves the things he says it does: the use of a set of distinctively religious concepts in describing and thinking about the world; a commitment to lead one’s life in a certain way; and a certain pattern of evaluations of oneself and others. But is he right that that is all there is to religious belief? In particular, is he right to deny that religious belief essentially involves a host of factual beliefs[?]

Child (2011, p. 224)

The question is based on a misunderstanding. Wittgenstein has not said anything like this (in the lecture) meant as a general analysis of the linguistic expression of religious belief. Even though he himself would call such a belief superstition, he does not say that, in our vocabulary, UNSHAKEABLE is essential for any description of religious belief or that one cannot reject GROUNDLESS in one’s religious statements.

[I]t seems plausible to think, his account does not accurately capture the character of the beliefs that most religious believers have actually held, either now or in previous generations.

Child (2011, p. 225)

It is difficult to understand why Wittgenstein would disagree. If we return to the lecture, Wittgenstein’s claim is that in our example of Judgment Day, this expresses an unshakeable belief. Wittgenstein has never said that we cannot imagine a person who also believes that the phenomenon of Judgment Day will occur and that this is part of what he believes or that there are other forms of religious beliefs. That Wittgenstein has shown that some religious beliefs are UNSHAKEABLE and that we can have such a belief without attributing to the person that he believes a specific proposition is not to be told that it cannot be otherwise. Perhaps it can even be essential in other language games and situations.

Even in cases with people like O’Hara,23 who attempt to rationalize religious belief based on evidence—rejecting GROUNDLESS—it may still be that we would say they have a religious belief. This is not contrary to Wittgenstein. This is a purely linguistic fact. Nevertheless, he would himself call it superstition.24 It is also wrong to think that he claims that all or most religious people (exclusively) have unshakeable religious beliefs. What we have in this lecture is the description of one feasible language game.

What lesson should we draw from all this? What is the delimitation of religious belief? At this point, it is important to remember a general Wittgensteinian insight: Do not expect a clear delimitation!25 How we talk about the religious is part of natural language, which is intertwined, complex, and connected to many different forms of life and practices. Wittgenstein’s main concern here is precisely to show that the way we talk about religious belief is just as intertwined and layered (Citron 2012).

Principally, Wittgenstein has introduced the enormously important insight that beliefs play many different roles in different language games. In our example of Judgment Day, we have seen that it is possible to express a belief without essentially attributing a truth value to a proposition but rather as an expression of a way of life. I believe most would agree with Wittgenstein’s analysis that, in the example of the person who chooses to regulate his entire life according to Judgment Day, we would say that the person has a religious belief, regardless of whether the person believes that the phenomenon will occur.

4. Concluding Remarks

Wittgenstein’s lecture on religion is not about providing a precise delimitation of the religious, neither metaphysical nor epistemological. It is not even about providing a precise delimitation of the grammar of the religious. It is not about saying that all linguistic expressions of religious beliefs are unshakeable or that all expressions of unshakeable beliefs are religious.

Essentially, Wittgenstein points to the linguistic phenomenon where a belief can play an unshakeable role. If we find ourselves in such a situation, there are other rules than for beliefs that express a proposition that one supposedly has reasons to hold as true.

Recognizing that there exist such linguistic expressions of unshakeable beliefs will enable us, on the one hand, to understand what a person is saying when he expresses such a belief; it is about a backdrop, about relationships, about meaning, and about the risk the person is willing to take. On the other hand, it also enables us to evaluate and understand disagreement and why linguistic communication can often break down when one person speaks about something unshakeable and the other treats it as a regular belief with a truth value. Wittgenstein’s grammatical analysis shows us that there are language games where the grammar of beliefs does not aim at truth or is governed by a norm of knowledge but rather plays a completely different role.

Although not explicitly discussed in the lecture, I believe we are right to think that the grammar of the unshakeable, and the religious, will depend just as much on elements outside the person himself. What we will call a religious or an unshakeable belief will depend on the context we, and the speaker, find ourselves in. Similarities and differences between ourselves and the person speaking, what other things they believe, how they act, what reactions they have, and a whole range of other elements will be crucial for our description.26 This is not meant to be a surprising result in light of Philosophical Investigations. Rather the aim has been to show that it is compatible.

If there is one thing the philosopher of religion should take away from this lecture, I believe it is this: Do not look at the linguistic expression in isolation to determine what makes or does not make it an expression of a religious belief. It is entirely possible that in certain linguistic contexts, one would call a belief religious precisely because it takes certain types of evidence into account.27 In some contexts, even the fact that the expression appears to have a supernatural content may be what makes us say it is religious. Other times, that it is life-regulating despite lacking a supernatural element, such as when one lives based on an ideology or another fundamental mindset that guides one’s entire life, despite or in accordance with the evidence one might have. One can even imagine that a linguistic expression of belief in the importance of scientific evidence itself could function as an unshakeable belief. In the right context, it may even be that we would characterize it as a religious belief in the form of (a certain type of) scientism.

Whether there are substantial truths in religion is not a philosophical concern—at least not for Wittgenstein. The philosopher Wittgenstein is concerned with the grammar of our religious expressions. It is as a religious thinker and as an individual that Wittgenstein speaks about truths in religion and correct approaches and expressions. It is the person, not the philosopher, Wittgenstein, who distinguishes superstition from what he would recognize as a possible religious belief. The philosopher Wittgenstein is neither inconsistent nor ignorant in this lecture.28