An architectural requirement

Tommi Uschanov referred in comments the other day to a couple of passages from On Certainty and to Philosophical Investigations 217. So I re-read 217 and was surprised to see the part about architecture, which I had forgotten. It provides a link to Wittgenstein's views on Heidegger and the influence on his thinking of Adolf Loos (see below). So I would like to investigate all this, starting by looking only at 217. Here it is:

217. “How am I able to follow a rule?” – If this is not a question about causes, then it is about the justification for my acting in this way in complying with the rule.
     Once I have exhausted the justifications, [then] I have reached bedrock, and my spade is turned. Then I am inclined to say: “This is simply what I do.”
     (Remember that we sometimes demand explanations for the sake not of their content, but of their form. Our requirement is an architectural one; the explanation a kind of sham corbel that supports nothing.)

(This is Hacker's and Schulte's translation, but I have inserted the word 'then' for the German 'so' to make it clear that "I have reached bedrock" is not meant just as another way of saying "I have exhausted the justifications'.)

The demand for explanations here is fairly clearly regarded by Wittgenstein as a bit of bad taste: he wasn't a big fan of sham corbels (those are corbels above), and this remark recalls the "Dictation for Schlick" (see pp. 75-77), in which he discusses the various requirements of style that are felt to be required at different times, and says that what he says has been influenced by Loos (it's here, too, that he makes an implicit reference to Heidegger). But he doesn't say it is bad taste; he just notes that it is a matter of taste. Anyway, back to the start of the passage.

The first paragraph distinguishes two possible meanings of a question such as "How is it possible to follow a rule?" We could inquire into the biological features of human beings, perhaps, that make it possible for us to follow rules. This would be a causal, scientific inquiry. It is not for philosophers. Or we could want to understand how doing this could be justified (or made the right thing to do) by such-and-such a rule. How does the rule translate, or get translated, into certain acts and not others? What makes this and not that the right thing to do, given the rule?

The second paragraph suggests that multiple answers can be given, but gives no examples. I suppose we can imagine any rule we like, and then imagine explaining it to a child or other person unfamiliar with the rule and its application. After some time you run out of things to say to explain the rule, and then you are likely to be inclined to say, "That's just what you do," or something similar. Wittgenstein does not say that we are right to be so inclined, nor that the inclination leads us to say something true. But he doesn't deny this either.  

Then in the third, parenthetical paragraph he suggests that the question and the answer are pretty meaningless. We ask for an explanation or justification not because any such thing is needed but only because that is what our taste requires. And the explanation(s) offered (which end in "This is simply what I do") does nothing, except satisfy our taste for such things. It isn't an explanation at all, we might say, and so the demand for an explanation is sort of empty too. It is a request for something that can never (not just not yet) be provided. "How is rule-following possible?" seems to be a bit like "Why is murder wrong?" and "Why are we here?" It looks like a question, but it has no answer. So one might say it isn't really a question at all. (Although the point is not to be dogmatic about what is and what is not a real question. Call it a question if you like, as long as you see that it is the kind of question that can never be answered.) So Wittgenstein's view of rule-following would appear to be that it's a mistake (and one in bad taste) to ask about it, unless you do so in a scientific spirit.

That's my initial take on this passage anyway.