Wittgensteinian views of religious language

Wittgensteinians lay stress on the idea that

1. One cannot understand central worldview concepts without living as part of a community that operates with these concepts.

The non-Christian cannot understand the Christian concept of the Trinity; the Christian and the atheist cannot understand the Jewish concept of God's absolute unity as understood by Maimonedes; the theist cannot understand the concept of a completely natural world; and the non-Fascist cannot understand the concept of the Volk. It is only by being a part of a community in which these concepts are alive that one gains an understanding of them.

Often, a corollary is drawn from this, that while internal critique or justification of a worldview tradition such as Christianity, naturalism or Nazism is possible, no external critique or justification is possible. In fact, there is an argument for this corollary.

2. (Premise) One's evidence set cannot involve any propositions that involve concepts one does not understand.
3. (Premise) Necessarily, if a proposition p uses a concept C, and a body of propositions P is evidence for or against p for an agent x, then some member of P involves C.
4. If x is not a member of the community operating with a central worldview concept C, then x does not have any evidence for or against any proposition involving C. (1-3)
5. (Premise) External critique or justification of a worldview of a community is possible only if someone who is not a member of the community can have evidence for or against a proposition involving a central worldview concept of that community.
6. Therefore, external critique or justification of a worldview of a community is not possible. (4 and 5)

This is a particularly unfortunate result in the case of something like Nazism, and may suggest an unacceptable relativism.

The argument is valid but unsound, and I think unsalvageable. I think that (5) is false, and on some plausible interpretations of (1), (2) and (3) are false as well.

First of all, people successfully reason with scientific concepts that they do not understand, like the concept of a virus or of gravity. They inherit the concept from a scientic community that they are not a member of, and while they do not understands the concept, they get enough about the inferential connections involving the concept that the concept should become useful. Thus, even if I do not really understand the concept of a virus, my evidence set can include facts about viruses that I know by virtue of testimony[note 1], and inferential connections with other facts, such as that if x has the common cold, then many viruses are present in x's body. Thus, (2) is false.

As for (3), I don't know for sure if it's false, but seems quite possible that while C does not occur in one's evidence set, it might occur in one's rules of inference. And there does not seem to be anything wrong with having a concept in one's evidence set that one does not understand.

But perhaps you are not convinced by the critique of (2) and (3). I suspect this is because you take (1) to be more radical than I do. The "cannot understand" in (1) is understood as entailing "cannot operate with"—even the weak sort of grasp that the layperson has of scientific concepts is denied to non-members of a community in the case of central worldview concepts. On this interpretation of (1), (2) and (3) are false. I am inclined to think that this interpretation of (1) is the incorrect one because it renders (1) false. The central worldview concepts of a community do not seem to be significantly different from the central concepts of a scientific community. Still, I see the force of such a beefed-up (1), at least in the case of the concepts of the Christian faith (not so much because of the need for community membership as such, but because of the need for grace to enlighten one's understanding).

In any case, (5) is false on either understanding of (1). The reason is simple. To support or criticize a position, one does not need evidence for or against a position. One only needs evidence for or against the second order claim that the position is true. Often, this a distinction without much of a difference. I have evidence that

7. there is life on Mars

if and only I have evidence that

8. the proposition that there is life on Mars is true.

However, this is so only because in (8) I refer to the proposition under the description "that there is life on Mars." But take a different case. I go to a mathematics lecture. Unfortunately, as I shortly discover, it's in German. I sit through it uncomprehendingly. At the end of it, I turn to a friend who knows German and ask her what she thought. She is an expert in the field and says: "It was brilliant, and I checked that his central lemma is right." I still don't know what the speaker's central lemma is, but I know that it is true. I do not have evidence for the lemma, and it could even be (say, if the talk is in a field of mathematics I don't know anything about) that I don't have the requisite concepts for grasping the lemma, but I have evidence that the lemma is true.

Likewise, it is possible to have evidence for and against the claim that the community's central worldview propositions are true, without grasping these propositions and having evidence for or against them. For instance, while I may not be able to understand what the members of the community are saying in their internal critiques, but I may understand enough of the logical form of these critiques and of the responses to them to be able to make a judgment that the critiques are probably successful. Moreover, even if I do not understand some concept, I may grasp some metalinguistic facts such as that if x is a Gypsy, then x is not a part of anything in the extension of the term "the Volk", or that if the only things that exists at w are the particles of current physics, and at w their only properties and relations are those of current physics, then w is in the extension of the term "completely natural". Given such facts, I can gain arguments for or against the thesis that the central worldview claims of the community are true. Thus, (5) is false.

There is a hitch in my argument against (5). External evaluation of the community seems to require that while I have no grasp of particular central terms, I have some grasp of the larger grammar of sentences used by members of the community and I understand some of the non-central terms in their language. But what if I don't? This could, of course, happen. The community could speak an entirely foreign language that I am incapable of parsing.

I can make two responses. The first is that (5) is a general claim about communities whose central worldview concepts I do not understand, and that general claim has been shown to be false. There could be some radical cases where the outsider's lack of understanding is so complete that external critique or justification is impossible. But such cases do not in fact occur for us. Humans share basic structures of generative grammar and a large number of basic concepts due to a common environment.

The second response is that the behavior of members of the community can provide evidence for and against the correctness of their central ideas. If their airplanes keep on crashing, there is good reason to think their scientific concepts are bad. If they lead a form of life that does not promote the central human goods, there is good reason to think that their ethics is mistaken, while if they lead a form of life that does promote the central human goods, there is good reason to their ethics is sound. Now, of course, I could be wrong. Maybe for religious reasons they want their airplanes to crash and design them for that. Maybe they abstain from some central human goods for the sake of some God-revealed higher good. Maybe they are a bunch of hypocrites, and they aren't really achieving the central human goods. However, such possibilities only show that I cannot be certain in my external evaluation. But the claim that external justification and critique is possible is not the claim that one can achieve certainty in external justification and critique. What I've said shows we can achieve high probability even in cases where the community's language is radically not understandable.

Things might be different if we're dealing with an alien species of intelligent beings. But I suspect we could still come to probabilistic judgments, just somewhat less confident ones.

I think the above considerations not only show that the argument (1)-(6) fails, but that we're unlikely to get any successful argument along those lines.