Wittgenstein returns to the question. "What is the object of a thought?" (e.g. when we say, "I think that King's College is on fire"). (Page 34)
This is, he says, typically metaphysical, because “an unclarity about the grammar of words” is expressed “in the form of a scientific (he has physics in mind) question”.

He identifies four “origins” of the question, which I think are intended to be confusions expressed in it. (Page 35)

1.. ‘One of the origins of our question is the use of "I think/expect" and “I think/expect” in two distinct ways. “I expect him” and “I expect that he will come”. We then compare “I expect him” to “I shoot him”.
He turns his attention to the shadows:-
2 a) Two different sentences can have the same sense, which becomes a shadowy being (unless perhaps it is a material object).
2 b) Shadowy sense to “be a picture” that “cannot be questioned”; it doesn’t need any interpretation.
3) Sometimes a sentence brings images before our mind's eye. These images are “translations” into pictorial language. But they must be similar to, copies of, what they are pictures of.

i) A map of all or part of our planet, he says, is not a picture by similarity or a copy in this sense. (Page 36)
ii) A picture of someone’s face projected it “in such a way that no one would normally call the projection ‘a good portrait of so-and-so’ because it would not look a bit like him”. This is, let us say, a bit complicated.

This gets him to where he wants to be. The sentence itself can do the work of the shadow, and so no shadow is needed. We can explain what the sentence means, perhaps, by an ostensive definition. That’s how words and things can be connected.

He can now go back and correct the path that he started down. (Page 37) The idea of the shadow is deeply rooted, but it is not what we really want to say. What we want to say is that “the fact which we wish for must be present in our wish”. Well, the answer just above cited ostensive definition to explain what “King’s College is on fire”. That works for expectation and wishing, too – after the event. But I could have given a similar explanation before it, as well.

At this point, a summary. How do our thoughts “connect” with the things they are about? The connection is made by means of ostensive definition. Ordinary language makes it seem that the connection must have been made during the act of thinking. But that’s not the case.
He gives us examples that could be taken to mean that thinking or meaning can be regarded as a mental activity, and so leading us to feel that something has been explained and no further questions are appropriate.

Now he returns to imagining and King’s College (Page 38) and subjects this to a cross-examination. I take it that links back to the rather unsatisfactory discussion of imagining earlier. But here, he concludes that our mistake is to think that images and experiences of all sorts, must be present in our mind at the same time.

Finally, a metaphor (Page 39) , - pulling a string of beads being pulled out of a box through a hole in the lid, to persuade us that “We easily overlook the distinction between stating a conscious mental event, and making a hypothesis about what one might call the mechanism of the mind.”

You speak as if “the color system” guarantees a metaphysical space for redness, as though the system enforces an ontological necessity. But the necessity is grammatical, not metaphysical. It comes from how we use color words, not from a hidden structure of reality. — Joshs

I get your point. But eventually realized that the peculiarity of this discussion is precisely that it is conducted, to put it this way, de re and not de dicto. If he asked whether we could abolish the concept of redness, that would have been one question. But he doesn't. He asks whether we could abolish redness, and compares it to destroying a watch. That comparison is a nonsense, to start with.

I keeping reaching for a key that will make all this fall in to place. But it still elude me. The discussion later on page 38 is all very well, but I don't see it clarifies this passage.

@Joshs

Sec 11 Our words’ connection to the world (p. 35-39)

The sentence itself can do the work of the shadow, and so no shadow is needed. We can explain what the sentence means, perhaps, by an ostensive definition. That’s how words and things can be connected. — Ludwig V

Nice work; my thoughts are along the same lines. He is showing us examples** of how we can correct the connection of word and world, as you say, by ostensive definition, or, alternatively, by explanation, demonstration, being an example, by force, etc., but words and the world don’t (usually) need to be (re-)connected because, by default, they just are connected (as you say, “no shadow is needed”). “…the interpolation of a shadow between the sentence and “reality” loses all point” (p.37) [my quote marks]. In the PI he will talk of this as there being no space “to get between pain and its expression”. (#245)

Philosophy imagines we make that connection every time (say, to “our understanding”). But there are events (in time, place) where “language” and the world actually do have a disconnect (along our criteria for judgment), but philosophy interprets the sheer possibility of disconnection, and the difficulty of reconnecting, as if the “problem” is in the activity of (always) connecting which is then just a puzzle to “know”, like a “a queer mechanism” (cue some neuroscience).

But in practice we fall back on the many separate ways we have for straightening things out. Philosophy needs to be shown any of these examples of means of reconnection—shown that language and the world “can be” reconnected—to realize the exception means that the word and world are not always mitigated by some object like “perception” or data, or other “shadow” But it then also follows that there is no “object” for there to be a “fact” of it to communicate. There are not certain, fixed, ever-present objects, as if part of “me”, like, “my understanding”, that I simply put into words.

The best juxtaposition is the difference between “…a thing I am thinking about, not 'that [thing] which I am thinking'.” (P.38) In the first, we are perhaps in a discussion (with ourselves even) considering, remarking on, analyzing, etc. a thing/object. Thinking in the second case is just the description of a thing/object which I have, “my thought”, which I take as a fact (as complete and without any need for context). But, like with the Napoleon example, there is no singular fact that is a certain, unique criteria (there, for identification).

Most importantly, understanding is not “present” during communication. Understanding happens after expression, in coming back to it, e.g., when you have demonstrated that you haven’t understood how to do something, or how to continue a series as expected, or that your expression makes it clear that you do not understand what I was trying to say (apart from disagreeing, etc.). We mostly say things that have already been said in situations similar enough to ours that it doesn’t need more elaboration (mostly). This public nature of language is because it is a record of our history, that “The connection between these words and [the world] was, perhaps, made at another time.” (P. 39)


**Sometimes I feel like his examples here are just terrible. I mean is it just me or waaaaay too unnecessarily esoteric for the point he is trying to make, except that he seems to feel he needs to chase the rabbit all the way down the hole to cover as many senses/analogies in which philosophy might frame our thinking as objects, etc.