  In philosophy we often say that people wrongly imagine a certain state of afairs, e.g. “they imagine that a law o⌊f⌋ nature in some way compels things to [p|h]appen”, ort “they imagine that it's a question of psychology how a person can know a certain fact whereas it is one of grammar” etc. etc.. But it is necessary in these cases to explain what it means “to imagine this so & so”, what kind of image is it they are using. It often sounds as though they were able to imagine the logically impossible & it is not easy to straighten out our description of the case & to say what ˇin this case they actually imagine.
  E.g.: People treat the question “how do we know what so & so is the case” as a question of psychology which has nothing to do with the sense of the prop which we say is known. But first: where do they take this idea from, how do they come by it? Which ˇreally psychological question are they thinking of?
  There Obviously there is a case in which the question “how does he find this out” is a personal &, perhaps, psychological

one. “How did he find out that N was in his room?” – He saw him through the window or he was hidden under the bed. – “How did he find out that the glas was cracked?” He saw the krack with his naked eye or he saw it through the gl magnifying glass etc. We say he finds out the same thing in different ways & therefore not that what he finds depends upon how he finds it.
When do we say that he finds out the same thing in two ways? Imagine language games: somebody is asked a question “A?” & trained to answer “yes” if he sees a

person A
man

in the next room, “no”, if he doesn't. In He is trained to answer ˇthe question “A?” by “yes” also if he hears A's voice from the next room. “What right have we to ask the same question in these two cases?” or “What right has he to use these two different tests to answer the same question?” Or, suppose someone asked: “Now are ⌊is⌋ th[e|i]se really one ˇ& the same question or do we have two different questions only expressed in the same words?”
Now concider the ostensive definition: “This man is called ‘A’” & ask yourself whether this definition tells us whether ⌊if⌋ we are to regard seeing A from a different

side or in a different position or hearing his voice as criteria of him being there? – Here we are tempted to say: “But surely I just point to this man, so there can't be any doubt what object I am meaning!” But that's wrong though the doubt of course is not whether I mean this → or that ↘ thing.
One may say that the ‘object’ I am inclined to say I am pointing to ˇin the ostensive definition is not determined by the act of pointing but by the use I make of the word defined. And here one must beware of thinking that after all even if the pointing finger pointed to a different object in the sense in which the arrow ⟶

A B
o o

may be said to point to A [&| or] B, so that by a different way of pointing I might have distinguished the cases.
“But we conceive of objects, things, different from our sensedata, e.g. the table as opposed to the views we get of it.” But what does conceiving of this object consist in? Doesn't it is it a peculiar ‘mental act’ occuring whenever, say, we talk about the table? Isn't it using the word table in

the game we do use it? using it as we do use it?

(2015–) Wittgenstein Source Bergen Nachlass Edition (WS-BNE). Edited by the Wittgenstein Archives at the University of Bergen under the direction of Alois Pichler. In: Wittgenstein Source, curated by Alois Pichler (2009–) and Joseph Wang-Kathrein (2020–). (N) Bergen: WAB.