§100

We shouldn't be dazzled by what seems to be an ideal of, say, the rules of language or the rules of games. - e.g. Rules lacking any vagueness.

§99

Wittgenstein presents a rationale for his earlier view.

A sentence must have a determinate sense because an indeterminate sense would not be a sense at all.

This is like saying that a boundary that is not sharply defined is not really a boundary. - But Wittgenstein objects to this - a boundary that is not sharp is not much the same as no boundary at all.

§98

Wittgenstein has already suggested that we cannot find the essence of language. This is not because it is too difficult but because there is no essence of language. 'Language' and 'proposition' are family resemblance concepts.

Wittgenstein presents his views from the Tractatus. - There must be perfect order even in the vaguest sentence (determinate indeterminacy).

§97

"Thinking is surrounded by a nimbus." (Does he mean that thinking is cloudy (unclear/vague) or that in thinking about thinking we have to get past cloudiness/vagueness - to get to the order of logic?)

The world does not contain vague objects and logic, as mirror of the world, must then be utterly crystal clear and certain (this kind of reasoning appealed to Wittgenstein when he was caught up in the Tractarian picture/model). - In this passage Wittgenstein is not stating his mature view, he is looking at his own earlier way of thinking.

§96

Wittgenstein sees this idea, of propositions being something remarkable, as being part of a wider conception of language, thought and reality - where all three are isomorphic (share the same logical form - (is that right?))

§95

The misunderstanding referred to in §93 arises from thinking that it is remarkable that one can think what is not the case. (Could the 'problem' be put like this? - How can you refer to something that is not the case? (is it even something?)
Wittgenstein attempted (in the Tractatus) to guarantee no failure of reference by suggesting that the world was made up of simple objects which are indestuctible and which were the constituents of states of affairs (possible concatenations of objects) - which names referred to. - So even states of affairs which didn't exist could be referred to by statements made up of names - all of which had referents.

§94

Wittgenstein suggests that those who say that propositions are remarkable sublimate their account of logic in thinking this. I assume he has his own Tractarian model in mind here.