Ordinary language is all right.
One could divide humanity into two classes:
those who master a metaphor, and those who hold by a formula.
Those with a bent for both are too few, they do not comprise a class.
—Frozen in place lately by the thought that, if the core of Wittgenstein's response to his former self occurs somewhere around §§81–142, and that response differs most significantly with the author of the Tractatus over the status of skepticism, then why shouldn't the metaphilosophical remarks in §§109–133 have just as much to do with skepticism as (I've been saying) those from §§81–88 and §§134–142 do—so that there is a sense in which the remarks on logic and philosophy do not really interrupt this sequence?
Maybe this is a perspective from which it's easier to answer the question: Why is Wittgenstein's later philosophy also a metaphilosophical critique of all philosophy?
The strange form of 'doubt' Wittgenstein seems to acknowledge the standing possibility of in §87—exemplified by someone asking, of a particular explanation of what is meant by some word, what is further meant by individual words used in the explanation, and the words used in response, and so on, down even to very basic words for e.g. the names of colors, thus, in general, withholding a shared understanding of what is meant by these words, of what things are called—is one which philosophy, in its tractarian form, evidently aims to answer, to rule out, to discipline.
Does that mean that philosophy is any less skeptical? No: not, it would seem, if we treat that skeptical question, 'But what do you mean by…?', like a kind of characteristic gesture. Philosophers will ask this no less often—particularly among themselves. And they will ask it in a more terrible, sublime way: 'But what do you mean by "mean", by "meaning"?'. They will get themselves into a frame of mind in which they feel perplexed as to how anyone can mean anything at all, how words mean things. They aim to ask and answer questions like these in such a way that the instigating doubts, the ones that looked fairly innocent, prove on reflection not to be (in the end) disastrous.
That they do this, that they profess a wish not to doubt, and purport to devise or discover reasons why one need not, does not mean that they're not starting out from that same possibility of doubt, as it were conceding it, letting it confirm their fears. They may differ from skepticism as initially voiced, but for Wittgenstein, they do so mainly in their dogmatism, in the emptiness of their assertions, in their illusion of meaning certain things by them. And in the sophistication of their development of something which it is already possible for ordinary people to do—something which we already sometimes do, have, feel: doubt.
If elsewhere Wittgenstein voices doubts of various sorts, directed in various ways, in order to give them a hearing, get them heard, but also to dismiss them, concede them, confirm them, give form to them, rule them out in a variety of ways, and he is regarding (traditional, or his former) philosophy too as a (sophisticated) form of skepticism, then we can expect him to treat it similarly: to voice its doubts, to answer those doubts as voiced, but to go on without supposing that those answers have eliminated all doubt.
What skeptical exchanges appear in the philosophy remarks? In §§116–121, as in the 80s, philosophy's doubts concern the ordinary uses of words—from the other side, as it were, from the far side of a supposed departure from the ordinary. In §§122–129, philosophy's inability to ground the actual use of language (cf. §81) is linked to a supposed reason to need to do so: the technical problems of another discipline, mathematical logic.
In §§130–133 skepticism is almost explicit, fully voiced: despite laying down no theories, seeking no regimentation of language by logic, etc., Wittgenstein does seek to bring a kind of order to 'our knowledge of the use of language'. What threatens that knowledge, in the form of philosophy? Skepticism, it seems:
'Die eigentliche Entdeckung is die, die mich fähig macht, das Philosophieren abzubrechen, wann ich will. – Die die Philosophie zur Ruhe bringt, so daß sie nicht mehr von Fragen gepeitscht wird, die sie selbst in Frage stellen.'
—A philosophy tortured by questions that call philosophy into question is a philosophy consumed by skepticism.
But being able to stop philosophizing needn't mean quelling doubt, putting an end to it (say, as a phenomenon, as an abiding or recurrent human experience). It can mean not giving in to it.
Thinking of joining a social network again so I can fill in the religion box on their profile form w/ 'Marquee Moon'.
So, like an accountant?
—The interrupted new material in the 80s is complemented by material awkward in its own way at §§134–142. And doubt is again a theme in §142: 'It is only in normal cases that the use of a word is clearly laid out in advance for us; we know, are in no doubt, what we have to say in this or that case…'. Then, from §143 on, we're off: series continuation, 'reading', rule-following, private language, the diarist…
Why the interruption? Not just a side trip to lodge some criticism of the Tractatus (criticism of unclear substance, at that). To make direct contact with skepticism, with the existence or removal of doubt, Wittgenstein has to be in the vicinity of ideas about assertion, truth. But, curiously, these ideas track passages from the Tractatus about 'logic and its application' (just after, significantly, 'Alle Sätze unserer Umgangssprache sind tatsächlich, so wie sie sind, logisch vollkommen geordnet…', 5.5563):
5.557: 'Die Logik darf mit ihrer Anwendung nicht kollidieren.'
§141: 'Können nun Bild und Anwendung kollidieren? Nun, sie können kollidieren, als uns das Bild eine andere Verwendung erwarten läßt; weil die Menschen im allgemeinen von diesem Bild diese Anwendung machen.'
There were some references before (e.g. §87c) to normality, but it seems like this bit from §141 twists things decisively in that direction—makes the right terms available.
And acknowledging the possibility of conflicting applications seems like it has to be part of that.