What then if we reject the view that simple signs were elucidated via ostensive definition? One should certainly agree with Kenny that in the Tractatus “he is saying is that the understanding of names and the understanding of propositions stand or fall together.” That is, you cannot have facts without things, or things without facts. This much I agree with but it doesn’t explain the import of elucidations. How do they explain the meanings of simple signs? Let us return to the dilemma from Philosophical Remarks where the elucidation “This is A” is either a proposition or a definition. Most commentators take Wittgenstein to be endorsing the first option only: elucidations are fully fledged, true-false propositions. For example, McGuiness says the following:
..teaching can only be carried out by means of complete propositions or complete thoughts. The learner has to grasp these as a whole, and, when he has done that, he will have an understanding of the primitive signs contained in the proposition.
We cannot explain the meaning by pointing at the object, but have to provide ‘illustrative examples’ (White’s translation of Erlauterung) of propositions with the name in. In taking this view of elucidations they firmly ‘bite Wittgenstein’s bullet’ in saying the following: “We must then leave it to chance whether the other catches onto the meanings of those sentences, which is something that can only be done by grasping the meaning of the name.” 
This accords with my view in that 1) one can learn whole propositions without first learning the meanings of the simple signs and 2) in learning a proposition you will ‘grasp’ the objects involved in understanding it. As such, if there was already a perspicuous sign-language with the right logical multiplicity, and elucidations were there to help you to understand or speak that language, one is taught complete propositions and it is left to chance whether you catch on.
Everyone in the debate about elucidations has been focusing on a learning situation where we come to learn what a word or a proposition means. It is as if we were learning language for the first time and we had to ‘grasp’ what was being said. Given this, it has been seen as an obstacle to explaining the signs that “they can only be understood if the meanings of the signs are already known” (TLP 3.263). Either an ostensive definition is needed to connect language and reality or it is left to chance whether the learner catches on. However, this is to ignore the context of the passage about elucidations. What is of interest is not how we came to be acquainted with simple objects or have the ability to form elementary propositions in the first place. To the extent that we can speak language at all, we have those experiences and that ability. What is of interest is how we come to know what objects are involved in a proposition as the end result of analysis. We can see this from the fact that 3.263 is a commentary of “In a proposition a thought can be expressed in such a way that elements of the propositional sign correspond to the objects of thought” (TLP 3.2).
3.263 starts from the fact that we already understand how to use the propositional sign and that with it, we can already express that particular thought. Ex hypothesei, if we know the thought we know the objects of thought. Given that a thought is a picture of a state of affairs, we not only know the objects but how they are related in that state of affairs pictured. The objects are the ‘logical co-ordinates’ through which the situation is projected into the propositional sign (NB p.20). Given this, the last line of the passage is seen as an advantage: it is because we already know the meanings that 3.2 is possible. The point is that we do not need any extra information in order to lay out the sense in which a proposition is used. We fully understand a proposition if we use it correctly but the thought can be expressed in such a way that it is clear that it is a logical picture of a fact. In line with the argument throughout the essay, the Tractatus is looking at how we can display the logical features of a proposition in a way that is logically perspicuous.
In making the above argument I am making a distinction between knowing that a simple sign ‘A’ has [object] A as its meaning, and knowing the meaning itself (i.e. knowing A). Of course, if we understand an analysed proposition understanding the simple sign is to know its meaning. In such situations, it would be nonsense to ask which object is the meaning of ‘A’ (‘A’ is the same sign as ‘A’). However, the distinction needs to be drawn for the following reason. We can know the meaning of a sign (i.e. be acquainted with the object) without a) knowing that in a particular notation, ‘x’ has that object as its meaning b) having any sign in my notation (i.e. English) which specifically names it. It is this kind of implicit knowledge of objects and understanding of elementary propositions that I’m claiming are necessary for analysis. However, until the analysis (philosophical elucidation) itself is performed, I am unable to name the objects. Equally, I would be unable to use the analysed proposition without it being explained which object is named by the signs.
How the last stage of analysis is supposed to be achieved- where somehow we are brought to recognise the simple sign names an object we have grasped all along- is as mysterious as any other part of the analysis. That we are able to (at least theoretically) express the proposition so it lines up with the objects of thought was seen as a demand of logic.
As Hacker finds ostensive definitions of simple objects to be found in the Tractatus he believes that the Investigations’ criticisms of ostensive definition hit their target, and as Kenny doesn’t find it there he thinks the Tractatus has been misrepresented. However, why, in looking at the passage, must we take “This is A” an elucidation of a Tractarian simple? Would it not be more likely to be something like “This is red”, “This is a ball”, “This is Neil”? Nowhere in his earlier work does he talk about observation statement in relation to simple objects. Where he does talk about ‘pointing’ in the Notebooks, it is about the kind of things above which we are obviously acquainted with. These may indeed be the simple elements of representation, (i.e. simple signs in not further analysable propositions) if the later Wittgenstein is correct. However, whether such propositions are fully analysed is precisely what is up for dispute.
In both periods, ostensive explanation can help us grasp the way ‘red’ is used but for different reasons. Earlier he believed that ostensive explanation revealed our ability to pick out a state of affairs as being ‘red’. However, given that the state of affairs may not exist, the ability does not consist in ‘red’ being a simple element of representation that refers to something red. If, despite this, it helps me grasp the contribution of the word to the sense of a proposition saying that ‘something is red’, it must be analysable by other terms I understand. One is able to elucidate it in this way because the objects of thought are already latent in my understanding of such sentences. After all, I understand the meaning of the sign as I can use it in propositions to assert the truth and falsity of states of affairs.
Here ostensive explanation doesn’t provide the elucidation, it is what is in need of elucidation. It gives me an intuitive grasp of its meaning: it means just that. It needs to be elucidated to yield its objective content. The point of the GFP is, however, that whatever the rules of the logical structure of language are, we are already in command of them. I think this better explains the following: “Logical analysis and ostensive definition were unclear to me in the Tractatus. I thought at the time there is a “connection between language and reality”. Ostension made it seem that there was a connection between language and reality but didn’t reveal what it was; logical analysis reveals that connection. Later he believed both were flawed. “This is ‘red’” is either a proposition or definition. If the former, we could pick it out because we already knew that to be red (because we know how red is used in language). If the latter, then it becomes is part of grammar rather than by revealing reality.
 (Kenny, 1974) p.5
 (McGuiness, 1981) p.70
 (White, 2006) p.61
 I’m not putting all three in the same boat except to say understanding the names come with understanding the proposition. McGuiness thinks it absurd that we could simply be acquainted with a singular object as we only ever sense a concatenation of objects. For Kenny and White it is possible that we can point to a thing.
 Wittgenstein in 1932 as cited in (Hacker, 1975) p.608