What is his idea of grammar?

For a very basic understanding (and because it was easily accessible):

Grammar, usually taken to consist of the rules of correct syntactic and semantic usage, becomes, in Wittgenstein’s hands, the wider—and more elusive—network of rules which determine what linguistic move is allowed as making sense, and what isn’t. This notion replaces the stricter and purer logic, which played such an essential role in the Tractatus in providing a scaffolding for language and the world. Indeed, “Essence is expressed in grammar … Grammar tells what kind of object anything is. (Theology as grammar)” (PI 371, 373). The “rules” of grammar are not mere technical instructions from on-high for correct usage; rather, they express the norms for meaningful language. Contrary to empirical statements, rules of grammar describe how we use words in order to both justify and criticize our particular utterances. But as opposed to grammar-book rules, they are not idealized as an external system to be conformed to. Moreover, they are not appealed to explicitly in any formulation, but are used in cases of philosophical perplexity to clarify where language misleads us into false illusions. — SEP article

There is also Baker and Hacker's account of Wittgenstein's use of 'grammar':

The use of a word, Wittgenstein averred, is determined by the rules for the use of that word (AWL 30). For using words in speech is a rule-governed activity. The rules for the use of a word are constitutive of what Wittgenstein called 'its grammar'. He used the expression 'grammar' in an idiosyncratic way to refer to all the rules that determine the use of a word, i.e. both rules of grammar acknowledged by linguists and also what linguists call 'the lexicon' and exclude from grammar - i.e. the explanations of meaning (LWL 46f.). To grammar belongs everything that determines sense, everything that has to be settled antecedently to questions about truth. The grammar of an expression, in Wittgenstein's generous use of 'grammar', also specifies the licit combinatorial possibilities of the expression, 'i.e. which combinations make sense and which don't, which are allowed and which are not allowed' (ibid.; emphasis added). 'What interests us in the sign', he wrote, 'the meaning which matters for us, is what is embodied in the grammar of the sign.... Grammar is the account books of language' (PG 87). Wittgenstein contended that the questions 'How is the word used?' and 'What is the grammar of the word?' are one and the same question (ibid.). The use of a word is what is defined by the rules for its use, just as the use of the king in chess is defined by the rules (AWL 48). The meaning of a sign lies in the rules according to which it is applied, in the rules that prescribe its use (MS 114 (Vol. X), 4r). Two words have the same meaning, he said, if they have the same rules for their use (AWL 3). — G. P. Baker P. M. S. Hacker

I very much dislike readings of Wittgenstein which equate him with saying that language is a 'rule governed activity'. There is a sense in which this is the case, but it's a sense that must be so heavily qualified and so massively underwritten by conditionals that I think it does far more to obscure and mislead than clarify the issues. One of the things Witty does in the PI is to expose the differential nature of rules, the fact that rules can and do play different roles in language (e.g. §54), so to say something like "For using words in speech is a rule-governed activity" is not so much wrong as simply empty - this says nothing in particular. Furthermore, Witty's constant refrain about rules governing rules ad infinitum (e.g. §84, §86) - and the ridiculousness of such an idea - also shows, to me anyway, what little stock he put in the idea of rules 'governing' language.

And really - just go through the PI: every mention of 'rules' is saturated with a kind of scepticism and cynicism that the whole enterprise comes across warning against any kind of hypostatization of rules. Rules are everywhere spoken about with suspicion, if not outright derision (§81-§86 is full of just such skepticism, you can almost breathe it, reading those passages).

I'm reading Stanley Cavell right now, and his take on Witty and 'rules' seems much more apropos:

"In order for there to "be" such things as rules, we have to agree in our judgment that a rule has been obeyed (or not). (The rule itself is dead.) In order for there to be such things as (what we call) measurements, we have to agree in our judgment that a particular thing turns out to have such-and-such measurements. It is one thing to know that you measure length by successive layings down of a stick; it is some thing else to know that this object is just under fourteen sticks long. (The stick itself is dead. It doesn't tell you where to begin laying it down; what counts as succession; and when, and what to do if, the last laying down goes just over.)" (The Claim of Reason, p.36, my emphasis): the rule itself is dead - this seems to me far more in the Wittgenstinian spirit than 'language is governed by rules'.

One of the things Witty does in the PI is to expose the differential nature of rules, the fact that rules can and do play different roles in language (e.g. §54), so to say something like "For using words in speech is a rule-governed activity" is not so much wrong as simply empty - this says nothing in particular. — StreetlightX

If Wittgenstein shows that "rules can and do play different roles in language", then clearly there are "rules...in language". These rules must govern the language, given that is the purpose of rules.

Also, I don't really see much difference in the roles played by rules at §54. Wittgenstein identifies a particular rule being used (i) as an aid in teaching the game, or (ii) as a tool of the game itself. Otherwise, he notes that an observer might be able to detect the rules of the game "from the way the game is played". That's not a whole lot of different roles, and the role of (i) is only to prepare the pupil for the role of (ii) when they start playing the game.

Furthermore, Witty's constant refrain about rules governing rules ad infinitum (e.g. §84, §86) - and the ridiculousness of such an idea - also shows, to me anyway, what little stock he put in the idea of rules 'governing' language. — StreetlightX

Wittgenstein is right to ridicule "rules governing rules ad infinitum". But you have provided no evidence that he somehow views "rules 'governing' language" in a similar fashion.

The rule itself is dead. — StreetlightX

This appears to be a Cavellian, not a Wittgensteinian view. There is some evidence in the text that Wittgenstein considers language to be governed by rules, even if you think that this is a heavily conditional claim to the point of being empty. For example:

But then the use of the word is unregulated — the ‘game’ we play with it is unregulated.” —– It is not everywhere bounded by rules; but no more are there any rules for how high one may throw the ball in tennis, or how hard, yet tennis is a game for all that, and has rules too. (§68)

133. We don’t want to refine or complete the system of rules for the use of our words in unheard-of ways.
For the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear.

To follow a rule, to make a report, to give an order, to play a game of chess, are customs (usages, institutions).
To understand a sentence means to understand a language. To understand a language means to have mastered a technique. (§199)

There can be no debate about whether these or other rules are the right ones for the word “not” (I mean, whether they accord with its meaning). For without these rules, the word has as yet no meaning; and if we change the rules, it now has another meaning (or none), and in that case we may just as well change the word too. (Boxed note (b), p. 155)

558. What does it mean to say that the “is” in “The rose is red” has a different meaning from the “is” in “Two times two is four”? If it is answered that it means that different rules are valid for these two words, the retort is that we have only one word here. — And if I attend only to the grammatical rules, these do allow the use of the word “is” in both kinds of context. — But the rule which shows that the word “is” has different meanings in these sentences is the one allowing us to replace the word “is” in the second sentence by the sign of equality, and forbidding this substitution in the first sentence.

Furthermore, if Wittgenstein does not consider language to be rule-governed, then why does he spend much time discussing rule-following (§185-242)? Why does he mockingly ask: "Are the rules of the private language impressions of rules?" (§259)

Perhaps further clarification can again be provided by the Stanford Encyclopedia:

How do we learn rules? How do we follow them? Wherefrom the standards which decide if a rule is followed correctly? Are they in the mind, along with a mental representation of the rule? Do we appeal to intuition in their application? Are they socially and publicly taught and enforced? In typical Wittgensteinian fashion, the answers are not pursued positively; rather, the very formulation of the questions as legitimate questions with coherent content is put to the test. For indeed, it is both the Platonistic and mentalistic pictures which underlie asking questions of this type, and Wittgenstein is intent on freeing us from these assumptions. Such liberation involves elimination of the need to posit any sort of external or internal authority beyond the actual applications of the rule. — SEP article

To grammar belongs everything that determines sense, everything that has to be settled antecedently to questions about truth. — G. P. Baker P. M. S. Hacker

Rules are everywhere spoken about with suspicion, — StreetlightX

There can be no debate about whether these or other rules are the right ones for the word “not” (I mean, whether they accord with its meaning). For without these rules, the word has as yet no meaning; and if we change the rules, it now has another meaning (or none), and in that case we may just as well change the word too. — W

Rules of use, (grammar) determine sense. Without these rules a word has no meaning, meaning is use.

A path is made by walking on it; a language is made by speaking thus and not so. If Batman says "Holy Wittgenstein, Robin!" every time he is confused, we will soon enough come to understand what he means.

If Wittgenstein shows that "rules can and do play different roles in language", then clearly there are "rules...in language". These rules must govern the language, given that is the purpose of rules. — Luke

If I state that there is a rule "I must not tread on the cracks in the pavement", but then, on finding my step too ungainly I make another "I must not tread on the cracks in the pavement, except every third step, when I may", are the rules 'governing' my walk? Clearly not. There are rules, that much is evident, but it wouldn't be right to say they 'governed' my walk. What's clearly governing my walk is my desire for it to be gainly. As soon as some rule clashes with that desire it is modified to suit.

The fact that rules have roles is what Wittgenstein is using here to try and show that it is the role, not the rule, that governs language.

And is there not also the case where we play, and make up the rules as we go along? And even where we alter them - as we go along. — PI 83

Exactly. So the satisfaction of some other objective is what truly governs play. If the rules no longer suit it, they are changed. The rules are a convenience, an aide memoir for what worked last time.

Rules of use, (grammar) determine sense. Without these rules a word has no meaning, meaning is use. — unenlightened

This is a muddle. Witty never speaks about 'rules of use'. You won't even find the three words, in that order, in the text. There's good reason for that: specifically, rules do not govern use. Second, rules are not (Wittgensteinian) grammar. They are not synonyms and you won't find that equation in the text either. Meaning is use in a language-game; not, meaning is use governed by rules.

If Wittgenstein shows that "rules can and do play different roles in language", then clearly there are "rules...in language". These rules must govern the language, given that is the purpose of rules. — Luke

That there are rules in language is not under debate. The question is about their role in language. That rules can 'govern language' is also not under debate. The question is whether such 'governance' - another word that appears nowhere in the PI with respect to rules - exhaustively characterizes language, on Witty's view.

--

I mean seriously, if the PI amounted to 'language is a rule governed activity', one wouldn't need to read a jot of it. One would just need to listen to your grade school teacher. As if Witty were just some philosophical enforcer for grammar school disciplinarians. Hopefully we get to the sustained rule following discussions further on in the text, with an eye to these questions. It might clear things up.

Grammar is structural, form without content.
— Fooloso4

This is not Wittgenstein's idea of grammar. — Luke

You are right, but you cited unenlightened saying: "grammar is extracted by pedants from pre-existing communication. It starts as description and becomes prescription - we convene, and from there comes convention." He is not referring to Wittgenstein's idea of grammar and I was responding to this.

The article describes grammar as a:

… network of rules which determine what linguistic move is allowed as making sense, and what isn’t.

Rules and actions are not the same. It is not some network of rules but what we say and do that determines usage. It is not so much a network of rules but a network of beliefs and practices. The difference between the Tractatus and the PI is that in the Tractatus the structure is transcendental, but in the PI the rules are arbitrary, part of a form of life.

When he says that:

“Essence is expressed in grammar … Grammar tells what kind of object anything is. (Theology as grammar)” (PI 371, 373).

He does not mean that there are rules that tell us what kind of thing something must be, what God or soul is or how we must use the terms, but rather that an overview of the way in which the terms are used indicates what is meant by the terms in this or that language game. Seeing how the terms are used requires an overview of the literature, prayers, practices, etc.

If Wittgenstein shows that "rules can and do play different roles in language", then clearly there are "rules...in language". These rules must govern the language, given that is the purpose of rules. — Luke

Consider the builder's language. How does "slab" come to mean "bring me a slab"? The rule is established in practice. It is the practice that governs the language. When we do as others do it might be said that we are following a rule, but we are simply following along. Suppose we learn what to do by following someone whose actions include a tic. It is not hard to imagine a generation that thinks copying the tic is part of the rules for what to do, and it becomes so.

Rules of use, (grammar) determine sense. Without these rules a word has no meaning, meaning is use. — unenlightened

I don't think this is correct, context is what determines sense, not rules of use. This is what Wittgenstein is getting at, each instance of usage is a particular case, with a specific purpose, and the meaning of the words used is unique to that particular instance of use. That's the basis of "meaning is use".

125 ... The fundamental fact here is that we lay down rules, a technique,
for a game, and that then when we follow the rules, things do not
turn out as we had assumed. That we are therefore as it were entangled
in our own rules.
This entanglement in our rules is what we want to understand (i.e.
get a clear view of).
It throws light on our concept of meaning something. For in those
cases things turn out otherwise than we had meant, foreseen. That is
just what we say when, for example, a contradiction appears: "I didn't
mean it like that."
The civil status of a contradiction, or its status in civil life: there is
the philosophical problem.
126. Philosophy simply puts everything before us, and neither
explains nor deduces anything.—Since everything lies open to view
there is nothing to explain. For what is hidden, for example, is of no
interest to us.
One might also give the name "philosophy" to what is possible
before all new discoveries and inventions.
127. The work of the philosopher consists in assembling reminders
for a particular purpose. — PI

the meaning of the words used is unique to that particular instance of use. That's the basis of "meaning is use". — Metaphysician Undercover

That cannot be true.There must be some consistency of use, to be able to use words at all. Even my lying monkey is using 'ook' ground-wise. 'Slab' means slab, the same every time. it might mean pass a slab, I want a slab, have you a slab, look there's a slab, slabs are the greatest, this is a slab... it always means slab-wise.

{Here is a thing, let me describe it to you. It's somewhat blockblockblock or else somewhat
slab
slab
slab.

Don't drop it on your foot! Let's call it a lintel, it might come in handy for something. } There might be some uniqueness here, but also much conformity.

I very much dislike readings of Wittgenstein which equate him with saying that language is a 'rule governed activity'. There is a sense in which this is the case, but it's a sense that must be so heavily qualified and so massively underwritten by conditionals that I think it does far more to obscure and mislead than clarify the issues. One of the things Witty does in the PI is to expose the differential nature of rules, the fact that rules can and do play different roles in language (e.g. §54), so to say something like "For using words in speech is a rule-governed activity" is not so much wrong as simply empty - this says nothing in particular. Furthermore, Witty's constant refrain about rules governing rules ad infinitum (e.g. §84, §86) - and the ridiculousness of such an idea - also shows, to me anyway, what little stock he put in the idea of rules 'governing' language. — StreetlightX

Good point Streetlight. This is why it's difficult to develop a theory that fits Wittgenstein's ideas, that is, there are too many variables involved. To say that language is governed by rules, is like saying chess is governed by rules. It's tautological, and it doesn't tell us much about the game itself. There is much more to a good move in chess than just understanding the rules. One must be careful not to turn Wittgenstein's words into dogmatic jargon.

Even when we do an exegesis of Wittgenstein's writings the tendency is to interpret his words in a way that violates the very things he is fighting against. In other words, our analysis of what he's saying tends to be the kind of analysis that Wittgenstein is criticizing, not always, but much of the time. We tend to look at his thinking in terms of - is this true or false, but there is a sense where it goes beyond this kind of bipolar thinking.

In other words, our analysis of what he's saying tends to be the kind of analysis that Wittgenstein is criticizing, not always, but much of the time. We tend to look at his thinking in terms of - is this true or false, but there is a sense where it goes beyond this kind of bipolar thinking. — Sam26

Exactly. A problem which has dogged much of this thread. Wittgenstein writes a book which intends to point the way to a method of thinking which goes beyond the finding of facts...and generations of philosophy students immediately spend years trying to discover all the 'facts' in it.

I mean, if one really wants a distillation of the Wittgenstienian approach to rules it really ought to be something like: 'What does the rule do? Look and see, from up close'. Not: 'Here is the a priori thing that rules definitely do in language'. The latter is 'philosophy' in Wittgenstein's pejorative key. To what use is the rule put? That is 'governed' by grammar.

There must be some consistency of use, to be able to use words at all. Even my lying monkey is using 'ook' ground-wise. 'Slab' means slab, the same every time. it might mean pass a slab, I want a slab, have you a slab, look there's a slab, slabs are the greatest, this is a slab... it always means slab-wise. — unenlightened

Not at all. Words evolve into meanings radically different from their original use and they would not be able to do so if they were so constrained.

Chimpanzees use sticks to extract ants, but what they use to do this job could just as easily be used to comb hair. It's just a stick. A screwdriver is actually designed to drive screws. Whilst it could be used to do other things, the purpose is discernable from an analysis of its structure.

Tractatus Wittgenstein saw language as the screwdriver, Philosophical Investigations Wittgenstein sees it as the stick.

Words evolve — Isaac

Yes. In my 2 word monkey-language example, I described such an evolution through the use of a word to deceive. And by that I mean ...

'I don't know what you mean by "glory",' Alice said.
Humpty Dumpty smiled contemptuously. 'Of course you don't — till I tell you. I meant "there's a nice knock-down argument for you!"'
'But "glory" doesn't mean "a nice knock-down argument",' Alice objected.
'When I use a word,' Humpty Dumpty said, in rather a scornful tone, 'it means just what I choose it to mean — neither more nor less.'
'The question is,' said Alice, 'whether you can make words mean so many different things.'
'The question is,' said Humpty Dumpty, 'which is to be master — that's all.'

When things evolve they are related. Have a slab of cake and calm down. Don't cement it into the wall though.

Tractatus Wittgenstein saw language as the screwdriver, Philosophical Investigations Wittgenstein sees it as the stick. — Isaac

In the Tractatus Wittgenstein sees language in very precise terms, that is, there is a one-to-one correspondence between the names within a proposition and the objects that make up facts. So, he sees propositions in very exact terms in the Tractatus, that is, it's more of an a priori investigation. His aim is to justify the vagueness of propositions (Nb p. 70), which is also his aim in the Investigations, although the method in the PI is more of an a posteriori investigation. His goal is the same, but the methods are very different (to repeat).

It's probably the case that teaching children in the early 20's helped him view language in a much more pragmatic way. We know that children have understood a word by how they use the word in a variety of situations and contexts.

It wasn't so much your monkey language as your contention that "'Slab' means slab, the same every time. it might mean pass a slab, I want a slab, have you a slab, look there's a slab, slabs are the greatest, this is a slab... it always means slab-wise."

'Naughty' used to just mean impoverished, now it means badly behaved. 'Clue' used to refer to a ball of wool, now it is some data indicating a conclusion. 'Silly' used to mean blessed, now it means facetious.

These are not words that even continue to mean x-wise. They just mean something completely different, occasionally almost opposite. Nor is this process a gradual evolution. Practically overnight Michael Jackson made 'bad mean 'good'.

In the Tractatus Wittgenstein sees language in very precise terms, that is, there is a one-to-one correspondence between the names within a proposition and the objects that make up facts. — Sam26

Yes, that's kind of what I meant by my analogy. That a proposition, correctly worded, represents one thing and one thing only, like the screwdriver does one thing and one thing only.

His aim is to justify the vagueness of propositions (Nb p. 70), which is also his aim in the Investigations — Sam26

Yes, but does he not go on to say "It is clear that I know what I mean by the vague proposition", I think the one-to-one correspondence is part of the aim, and I would probably say that makes a significant enough difference.

It's probably the case that teaching children in the early 20's helped him view language in a much more pragmatic way. — Sam26

Yes, that's an interesting idea. Afterall, children are also where one can see what adults consider to be nothing more than a block of wood stand in for anything from a spaceship to a football goal.

Alas, a supposedly humorous reference to British politics - "Brexit means brexit." We are not at all disagreeing. Nothing is fixed forever, and nothing is unconstrained by current practice either. I quote Lewis Carroll above, master of logic, and master of nonsense - because the best nonsense is logical nonsense.

Yes, but does he not go on to say "It is clear that I know what I mean by the vague proposition", I think the one-to-one correspondence is part of the aim, and I would probably say that makes a significant enough difference. — Isaac

If you're asking this in reference to the Tractatus, the answer is, I believe, yes, that is, even a vague proposition in the Tractatus had a very precise nexus to the world of facts.

§104: One predicates of the thing what lies in the mode of representation. We take the possibility of comparison, which impresses us, as the perception of a highly general state of affairs.

§104

I'm slightly stumped on this and wonder if anyone might provide some extra commentary and thoughts: when Witty speaks about the possibility of comparison, what two things are being compared? (1) The thing and (2) The mode of representation? If so, in what does this comparison consist? And when he speaks of 'a highly general state of affairs', my assumption is that a distinction is being made between that and a 'specific' or circumscribed state of affairs. But how to characterise the difference between the two?

Apart from these the general gist I get is something like: don't confuse our use of grammar for the 'way things are'. Or: don't project our grammar, which bears upon our 'mode of representation', onto the things themselves. I'm not comfortable with these formulations which smell too much of metaphysics. And they seem out of keeping with the paragraphs both before and after it. Any thoughts?

https://thephilosophyforum.com/discussion/comment/261068

That cannot be true.There must be some consistency of use, to be able to use words at all. — unenlightened

I don't know about that, "consistency" is something other than simple use. It might be an attribute, a type of use. But meaning is restricted to use here, as the premise, and if consistency enters the picture, then that is another thing. This is the problem of "types" which Wittgenstein exposed earlier. What is exposed is a lack of consistency. If I point to a particular rock and call it "slab", then to maintain consistency, every time I say "slab" I ought to be referring to that particular rock. But that's not the case, "slab" is used to refer to a type, and therefore many different rocks. So the notion of "consistency" is misleading to us, and we ought to move away from that. Now we move to a family of different meanings, different uses, and this difference demonstrates a type of inconsistency. It appears like Wittgenstein is focusing on inconsistency in use, rather than consistency.

I think that's the point of the passage I quoted, 125-127. We get caught up in this idea that everything goes according to rule (consistency), but that's really not the case, as every instance of use is unique.

I'm not comfortable with these formulations which smell too much of metaphysics. And they seem out of keeping with the paragraphs both before and after it. Any thoughts? — StreetlightX

I agree, 104 seems to be right out of context. I'll see if I can make sense of it.

I believe that in this section, 100-108, he begins by referencing the primary elements of the Tractatus. He has already dismissed this idea of material elements of fundamental clarity, upon which language and logic is based, but he still references them as "ideals", which we, for some reason, think must be there. At 100, a game can exist without them. But what kind of "game" could that be? It seems like logic would be impossible -101. The fundamental elements of clarity must be there somewhere, hiding in the background -102. It appears like we cannot dispense the notion, like glasses we cannot take off -103.

I think that what is described at 104 is the primary act of abstraction. Predication is that fundamental act by which we take the particular thing, and say something general about it. This is what allows comparison. To put this in context of the ideal which he is discussing, I would say that "highest generality" implies this ideal. So I think he is saying that we are mislead into believing that this fundamental abstraction, this act of predication is an ideal, the highest form of generality, upon which the strength of logic is based. But really, when he turns things around at 108, it may turn out to be the lowest form of generality. What we thought was the most ideal, the fundamental predication, is really the least ideal.

At 105, we don't find the purity of the ideal, here in the fundamental abstraction of the predication, and this dissatisfies us because now we cannot support the rigour of logic. So at 106 he wants to turn way from this highly formal mode of thinking (logic) to everyday thinking because the ideal which was required of logic is just not there. So we need to look at everyday thinking to see what really supports logic. And this is very evident at 107, what was required of logic, that it proceeds from pure, solid principles, is just not there, and our everyday thinking shows us that our thinking is really the opposite of this, we proceed from vague uncertainties, attempting to clarify them. And this is the turning around which he speaks of at 108. Aristotle stated that logic proceeds from the more certain toward the less certain, but maybe he really had things reversed.

The comparison is [edited] a mode of representation - we see the thing in relation to what we are comparing it to. But this is not "the perception of a highly general state of affairs". What we see is the thing as it stands in relation to whatever it is we have compared it with. If we did not make the comparison we would not see it in that way.

It is not a question of whether it is right or wrong to do this. Such a comparison can be misleading but it could also lead to some insight. It is exactly what Wittgenstein does with his multiple examples.

I think 104:

We predicate of the thing what lies in the method of representing it. Impressed by the possibility of a comparison, we think we are perceiving a state of affairs of the highest generality,

should be read in the context of 101->103, and the broader context of Wittgenstein's critique of idealised approaches to language. I think 'we predicate of the thing what lies in the method of representing it' is fleshing out how we 'find' the ideal in language in the sense of 101:

101. We want to say that there can't be any vagueness in logic. The idea now absorbs us, that the ideal 'must' be found in reality. Meanwhile we do not as yet see how it occurs there, nor do we understand the nature of this "must". We think it must be in reality; for we think we already see it there.

our 'absorption' in the problematic of a pure philosophical language of expression, or a universal logic, invites us to 'see' or 'perceive' (note the two perceptual terms in 101 and 104!) the linkage between 'the thing' and 'the method of representing it' as logically prefigured/framed in terms of such an ideal language. Specifically, this ideal logical structure is posited 'of the thing' through 'the method of representing it', and this positing opens (and invites exploration of) a 'background' for us to uncover (under the problematic of idealised logic/language):

102. 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 though
through a medium): for I understand the propositional sign, I use it to say something.

I read it like the 'background' and 'the method of representing it' are collapsed into logical formalism by this 'idealised language' interpretive habit that Wittgenstein diagnoses of philosophers. But the link between the two isn't interpreted as a logical or rational pivoting about a concept, the link is 'seen' -or posited prior to inquiry- through the drive to abstract/ideal logical formalism. Wittgenstein likens this drive towards the ideal in 103 to putting on a pair of spectacles:

103. The ideal, as we think of it, is unshakable. You can never get outside it; you must always turn back. There is no outside; outside you cannot breathe.—Where does this idea come from? It is like a pair of glasses on our nose through which we see whatever we look at. It never occurs to us to take them off

I think a more precise metaphor would be rose tinted glasses, where 'seeing things in a rose tint' is a constraint placed on philosophical discourse to idealise/abstract towards logical formalism capturing the 'essence' of language. The 'essence' of language, the ideal of it, plays the role of the inescapable rose colour implicated in seeing all things; but no argument demonstrates that all is rose tinted, it's 'put there' by one's style of philosophical thought. For one who believed that all philosophy requires arbitrarily donning such glasses; that philosophy begins with a confused and limited apperception;, it is not surprising that they would reject the entire enterprise.

Edit: perhaps another good analogy is this:

$p \wedge (p \rightarrow q) \vdash q$

show that to someone who hasn't learned to process propositions in logical syntax and it wouldn't mean a damn thing. We have to 'learn to see' the connections between natural language argument forms and the modus ponens. The 'representation' of our argument forms (in terms of validity, soundness, truth functionality and so on) consists in fabricating rules for propositional calculi spurned on by real argument patterns, and then we may say that the above formula is modus ponens. Even someone who understood how to argue using the modus ponens syllogism would not necessarily immediately 'map' it to the representation of it in the theory. We have to do a lot of conceptual prefiguring to analogise reasoning with a logical calculus; and if we 'see' language as logical in the sense of a calculus, this consists in a 'comparison' of the thing (modus ponens argument forms in natural language) with its means of representation (the above formula) - a good fit between the two invites us to study more. Even though we had to lose a lot of the structure of arguments to 'reveal' their propositional form. Thus the revelation is, too, of a constrained domain of concepts; reading back the means of representation into the thing we represented using them always suggests 'depth' in the sense of a background to explore; why is it that arguments bear a relation to the notion of truth functionality? Why is it that meaningful language in propositional calculus consists of truth functions of propositions? There is always the danger of forgetting the constraints or simplifications we take to embark upon an inquiry when we discover a 'background' which underlays everything within the constrained conceptual space whose boundaries are those set by our inquiry.

Also consider 'removing the spectacles' as a non-logical operation, that is, it doesn't follow the internal structure of the domain of concepts obtained by peering at a topic through rose tinted glasses, it rather reframes them for other inquiry; which Wittgenstein likens to description or 'looking around'. The end of the first paragraph of 108 fits in well with this interpretation:

(from 108) The preconceived idea of crystalline purity can only be removed by turning our whole examination round. (One might say: the axis of reference of our examination must be rotated, but about the fixed point of our real need)

eg, the idealised language philosophy criticised doesn't 'pivot' around any adequate apperception of the structure in language. In my example, what is determinative for propositional calculi is a filtered representation of the elements of language and their relationships in argument; there's no room for questions, doubts, beliefs, appeals to emotion, reframing attempts and so on, because they are excluded from the 'philosophical grammar' applied to language through the rose tinted glasses. What was excluded from the logical artifice of 'crystalline purity' was the broader structure of language and behaviour; ratiocination in this realm of purity consists in the production of well formed formulae - none of which can be questions, reframing attempts, metaphors, analogies or appeals to structural similarity.

By treating the means of representation of propositions and their logic as the idealised representation of propositional calculi (or the more general one in the Tractatus), much of the world becomes thought of only in category errors - how can any expression have sense if it does not resemble only those things which were posited to?

There's probably an interesting analogy to Kant here, with regard to transcendental illusions.

Kant identifies transcendental illusion with the propensity to “take a subjective necessity of a connection of our concepts…for an objective necessity in the determination of things in themselves” (A297/B354). Very generally, Kant’s claim is that it is a peculiar feature of reason that it unavoidably takes its own subjective interests and principles to hold “objectively.” And it is this propensity, this “transcendental illusion,” according to Kant, that paves the way for metaphysics. — SEP

Wearing the 'rose tinted glasses' or the 'spectacles' invites us to see 'the connection of our concepts' (in the ideal domain of an abstract logical language) as of 'objective necessity in the determination of things in themselves' (the rough ground of ordinary language)- rather than as a limitation of the vantage point we viewed 'the things' (language) from.

In the version I've been using, just below 109 is the comment:

"Language (or thought) is something unique"—this proves to be a superstition (not a mistake!), itself produced by grammatical illusions.

the inappropriate substitution of our conceptual links and tokens within an account for the real structure of the desired topic is exactly the logic of seeking such an 'ideal' and therefore seeing it everywhere; it is not a simple mistake, the conceptual structure of ideal language is very rigorous and its internal links are self consistent - but the endeavour of trying to picture language (and language as a picture) in this way is misguided when dealing with all the ambiguities and subtleties of the huge scope of natural language.