§30. So one might say: the ostensive definition explains the use—the meaning—of the word when [only when] the overall role of the word in language is [already] clear. Thus if I know that someone means to explain a colour-word to me, the ostensive definition "That is called 'sepia'" will [now] help me to understand the word.—And you can say this, so long as you do not forget that all sorts of problems attach to the words "to know" or "to be clear".
One has already to know (or be able to do) something in order to be capable of asking a thing's name. But what does one have to know? [This he tries to answer in the next section]
He continues his criticism of what he calls the 'Augustinian' picture, but is really his own Tractatus. If we are not tempted by this picture, these passages are less relevant. (Also there is no point in trying to find out what else Augustine said, and whether he is being misrepresented) It is, he now feels, extraordinarily naive. It isn't just that it doesn't include uses of language such as "Ow!", shrugs, and so on. It is that it doesn't even fit with how the object language works - or, more particularly, how it gets set up, and indeed how it is taught. The idea that we could start with a blank slate, and learn our first language by our parent pointing to things in the world, uttering the relevant sounds (writing the appropriate words on labels), so that we have 'red', 'mug', 'table', and on', and now can say '"Red mug on table", and, indeed, "That thing's name is 'table'", is mistaken.
Ostensive teaching (A teaches B - Master actually teaches Pupil; parent actually teaches child) compared to Ostensive definition (Something more abstract). He does not make clear the distinction between the two, but that is fine, given his general attitude to clarity of distinction. Furthermore, we doubt that there is much of a distinction for him, since on his non-metaphysical, anthropocentric, approach, no abstract connection (association) exists - somewhere, somehow - between the word 'table' and that thing → ∏ . Teaching B how to use language does not consist (simply) of teaching B these abstract connections (because the abstract connections give us the language). There is no abstract connection, and, in this sense, there is no 'ostensive definition1'. What there is is the development of a way for person B to use 'table'; which is done by a kind of teaching involving pointing and saying 'table' - and other things - and hoping that B does not get the wrong end of the stick. 'Definition2' is achieved when B appears to use 'table' in the same way that the rest of the community of users use it.
Sticking a physical label - a Post-it note - on that object, with 'yellow' written on it, saying 'Yellow', pointing at the mug, and getting B to say "Yellow", is fine as a stage in ostensive teaching of the use of 'yellow in our language; but it is not providing any kind of abstract 'ostensive definition1' of 'yellow'.