The exchange which concludes the first section of the Investigations has troubled me for a long time. So has the fact that the elements of the shopkeeper example prompting that exchange are repeated, in variants of or additions to the builders' language-game of §2, in remarks extending all the way to §86.

It's plain enough that one voice in the exchange, the one that asks questions and that we tend to read as opposed to Wittgenstein's (even if internally so), can be associated with the idea of meaning given in §1b (a 'philosophical' idea, according to §2 and later remarks). But the construction of §1, and the variety of relations between its parts, have the unhappy result that the slightly censorious tone of the concluding exchange seems cast retroactively over every part of the remark, suggesting that the voice asking after meaning in the shopkeeper example neglects use, sees no difference between words, is held captive by Augustine's (or 'Augustine's', quoted, since it is deliberately not attributed directly to him) picture, and so on. That all errors are the same, all criticism aimed at the same fundamental point.

The simplest way to indicate why I think that suggestion is misleading, is just to draw a diagram of the structure of §1:

The double arrows are meant to indicate that Wittgenstein evidently introduces some distance between Augustine's words, and what (it seems to Wittgenstein) 'we' 'get' from them, the picture of the essence of human language. And then similarly—but perhaps by way of a different path or procedure or process—with the idea of meaning whose 'roots' we can find in that picture. (It grows out of it, in other words.) The fact that these two elements of §1 play an important role in the structure of the book as a whole seems like a good reason to be sensitive to their distinctness. Meaning is a regular topic of contention in the remarks to come, rarely going unmentioned for more than a short stretch. 'Pictures', in the sense given here, go mostly unmentioned until the logical and metaphilosophical remarks begun at §89 (there is a quick reference to a picture in the relevant sense in §59, suggesting that the 'simples' passages introduced by the passage from Plato may involve unstated consideration of a 'picture' paralleling the one drawn from the Augustine passage), and do not really receive any discussion in their own right until Wittgenstein breaks off metaphilosophizing and resumes his usual remarks, starting around §134.

Likewise, the single arrows indicate simply that Wittgenstein's remarks relate directly to Augustine and his words. In §1c he talks about what Augustine does and does not say, and what he might be believed to have had in mind—to have thought, or meant—by what he did say. §1d provides an example of what Augustine did not mention in two senses: first, by involving a word of a different kind than the ones he's thought to have had in mind (nameish and nounish words, basically), and by stressing its different use as the main reason for that difference in kind. Later references to Augustine tend to confirm the impression that the picture, the neglect of different kinds of word, and the inattention to the uses of words come together, as far as Wittgenstein's critical work is concerned. And after a long lull, when Augustine is mentioned again, at the start of the logical remarks (§89), it is in a related role, essentially as a reminder about the need to call things to mind, to remember them well, to remind ourselves or to need to be reminded.

But the exchange which concludes §1, in the last several sentences of §1d, can be seen as directed not at Augustine (or someone who says the kind of thing he says, a la §1c's qualification), but just at the voice whose concern links it with the idea of meaning from §1b. The voice cannot be totally inattentive to use; different uses of different kinds of word have just been put before it. And what the voice asks about, meaning, was just (for us) explicitly distinguished from the picture associated with Augustine.

This voice is, though, asking its first question—which could sound skeptical—just about a person, the shopkeeper.

If this sort of effort to separate the strands (the very ends of them, that will run through the book) of criticism begun in §1 is on the right track, then one way of understanding the repetition of the features of the shopkeeper example in remarks §§2–86 might begin by noticing that in §2, Wittgenstein knots the strands together, but not without drawing out one more aspect of Augustine's words from §1a: that something about the way they conceive of us, of our language, our using language, is or can be regarded as primitive. That he should draw this out by first introducing the idea of a language-game as well suggests not primarily something about a level or degree of development or sophistication (though those are clearly relevant) of a conception of language, but something about the point of language so conceived (as §5 says).

'… like growing a tree without roots.'

'… notebooks which promised, deceitfully, on the blank front page, to be the place, finally, where life and writing would converge.'

'Their speech will be confounded; they will leave off the building of their own city.'

New Hope for the Dead

'It's "personal" because you pour yourself into it—your thoughts, your programming.'

'The actual magic trick is making the worker disappear.'

'… a spiritually corrosive quality…'

Snow.