The GFP embodies a particular conception about the understanding of propositions, which can be revealed through analysis. All propositions are truth-functions of elementary propositions (TLP 5) which tell you what is the case if they are true. These express the thought that simple objects are concatenated in a certain way. However, “language disguises thought” which means that as the multiplicity of thought is hidden, we cannot see the connection between the proposition and reality. When completely analysed, however, “a thought can be expressed in such a way that elements of the propositional sign correspond to the objects of the thought” (TLP 3.2). In this way, the proposition manages to line up with the thought it is expressing. Given that “A logical picture of facts is a thought” (TLP 3), the proposition reveals itself as a picture of a situation in reality, where there is a one-to-one correspondence between the elements of the picture and the objects in the world.
For the Tractatus, then, language, thought and the world are all constructed according to a common logical pattern (TLP 4.014). If these are made to line up, there is no further question we can ask about the proposition’s logical relationship to reality. All we can do is see whether the proposition is true or not. This is the view criticised in the Investigations:
Thought, language, now appear to us as the unique correlate, picture, of the world. These concepts: proposition, language, thought, world, stand in line one behind the other, each equivalent to each. (But what are these words to be used for now? The language-game in which they are to be applied is missing?) (PI 96)
What is the complaint here? What he is pointing to as a weakness of the Tractatus is what we above highlighted as a (seeming) strength. On my interpretation, the strength and weakness derive from the same source: the presence or absence of the ability to elucidate the sense of our ordinary propositions. For the Tractatus there is no need to express it in this way as we could express the same proposition with any sign. Indeed, a propositional sign with the right multiplicity may have no use in our language (we wouldn’t know what to do with it). However, if the sort of analysis envisaged in the Tractatus is feasible, then it reveals the content of the proposition as the terminus of such an analysis (as envisaged by the GFP). On the other hand, if we can’t say how it applies to our propositions, then what is the use of accepting the GFP? What was positive is now negative: no further questions can be asked. It stops us saying anything about the relationship between the proposition and reality or what makes it true.
Indeed, it is the inability to see how such an analysis might proceed, that made Wittgenstein think his earlier position was empty. Consider the following:
I spoke as if there was a calculus in which such a dissection would be possible. I vaguely had in mind something like the definition that Russell had given for the definite article... At the root of all this was a false and idealized picture of the use of language. (PG p.211)
In PG he says that we can elucidate the sense of certain expressions through the use of definitions. In such instances it will help us see the connection between different concepts, see the logic of that language-game and remove misunderstandings. However, firstly, it would be a mistake to say that this could apply to all of language. Secondly, to the extent that one can dissect a proposition into logically more basic ones, it becomes a ‘problem of calculation’ as to whether the proposition is elementary or not. By this he means we must have a method of discovery whereby we could discern/calculate whether it is further analysable or not. It is precisely because we don’t have such a method for the majority of language, that it is misleading for the GFP to talk about thought being disguised or hidden. As the quote testifies, he thought something like Russell’s definition applied to the whole of language. In relation to this he says, “I saw something far away and in a very indefinite manner, and I wanted to elicit from it as much as possible”.
Given what I have said above we have to ask why he was so confident of it earlier. Was he ignoring ordinary language and did he simply suppose that we could break down propositions into elementary ones? Would it have bothered him that we can’t see how such an analysis would go or would have merely stated it must be like that? Consider the following:
The strict and clear rules of the logical structure of propositions appear to us as something in the background- hidden in the medium of the understanding. I already see them (even through a medium): for I understand the propositional sign, I use it to say something. (PI 102)
He is referring to such passages as “the understanding of general proposition palpably depends on the understanding of elementary propositions” (TLP 4.411). We are able to understand all sorts of propositions, including ones we have heard for the first time, even though we can’t spell out its projective relation to reality. Equally, we didn’t have to have any new ‘logical experience’. As such, palpably the ingredients were there for understanding the proposition and I know how to put them together to form the thought. Here, my understanding can be brought to light but not in the sense of telling me something I didn’t already know. Instead, in terms of using the medium of language in such a way that logically makes clear what I already know.
In these passages we don’t see analysis as far off and distant but which we must say is there anyway. On the contrary, they seem directly relevant and they feel as if they are contained within my understanding of ordinary propositions. As I argued in the previous section, analysis is only possible because we already know the meanings of the parts of thought. In this way, at each stage of the analysis, the proposition will recognisably be the proposition being analysed. As such, it would certainly be a concern if there was a disconnect between analysis and our understanding, if there was no method of discovery, or if we couldn’t see how it would be applied in the situations in which we use the proposition. The logic of propositions are not ‘out of sight’ but seen through the medium of language, in its application to the world.
The argument then is that it wasn’t dogmatism, as such, about logical analysis or a focus on an ideal language that led him down the path to the GFP. Instead, it was caused by the very real way in which we explain the sense of a proposition to someone. That is, we either point to the situation, point to a picture of it, or explain how things are in that situation. Through this, the person being explained to manages to grasp the sense and is then able to use it appropriately. This requires the person to understand the parts and how they are put together in the sentence. As far as it goes, that account is fine. However, the philosopher then takes that picture, image of the situation or whatever, and says to himself “This is true”, “This is how things are”. Then s/he starts to believe in something like Russell’s theory of descriptions where either we know the elements of the picture by acquaintance or by description.
Now, I said that it was fine in as far as it goes because i) we do understand a proposition because of the words that make it up and ii) pointing to something (for instance) can help us grasp the sense of a sentence or the meaning of a word. However, this is where the Augustinian picture comes in with a philosophically naive conception of what it is to ‘grasp a sense’. We begin to feel that what is grasped is something given in the proposition, or by the explanation or through the pointing. However, we don’t grasp the meaning individually or exclusively through any of these things. These things only manage to do against the background of the rest of language. If we look at the wider context of language, it is harder to see how digging down below its surface helps us become clear about how we are using the propositions.
The GFP, in theory, is simply what is shown through the analysis of language and the uncovering of what we already know when we understand of a proposition. What we will find will differ with each proposition and can only be discovered as the end-result of a process of analysis. However, the kind of answer is given once-and-for-all in advance of such an elucidation. He later believes that if we look at the conditions under which our propositions make sense, we will no longer believe that all propositions can be elucidated in the same way. One reason he had earlier thought it could is because of a prior conception of what understanding a proposition consisted in, brought about through a particular way they are explained. Once we read his critique of the Augustinian picture of language, this motivation is undercut.
 Wittgenstein to Waissman as cited by (Monk, 1990) p. 183