I think most books treat his early and late theories as inconsistent theories, in a sense that one can agree on either his early or late theory, but not both. However, I think the two theories are supplementary, and it is possible to agree on both of his theories.

Here is what I mean. I think his early theory deals exclusively with descriptive sentences, such as:

  • The man is fighting with a lion.
  • The mother saved her child from a burning house.
  • The warrior defeated his enemy.

Each of these sentences does indeed represent a fact, and the relationship between a descriptive sentence and a fact is explained by picture theory.

On the other hand, the meanings of abstract words are given by pragmatic usages based on descriptive sentences. This is where Wittgenstein's later theory comes in. For example, the meaning of 'courageous' is given by its usage in descriptive sentences like:

  • The man fighting with a lion is courageous.
  • The mother who saved her child from a burning house was courageous.
  • The courageous warrior defeated his enemy.

Therefore, his early and late theories are supplementary, and both his theories are needed to clearly define how language works. We first understand the world by picture theory, and build up more complex concepts according to language game theory. Is this interpretation of Wittgenstein valid?

  • 'Complementary' - fitting coherently together. Woudn't 'inconsistent' or 'incompatible' better convey the sense of your question as the opening para. explains it? Just a suggestion.
    – Geoffrey Thomas
    Aug 22, 2021 at 9:56
  • @GeoffreyThomas I took your suggestion, thank you!
    – Dimen
    Aug 22, 2021 at 9:58
  • You can see Hintikka's Wittgenstein Aug 22, 2021 at 10:24
  • Sometimes I think of an analogy to special relativity and general relativity. The Tractatus was about a specific case of language where PI was the general case.
    – Allan
    Aug 22, 2021 at 12:20
  • 2
    I think the continuity of Wittgenstein's thought is indeed underappreciated, but your view is still overly optimistic. Wittgenstein himself begged to differ, "Language is much more complex than logicians and the author of the Tract. Log. Phil. have imagined" he wrote in 1936, and in PI confessed himself "forced to recognize grave mistakes... in that early book". But he also held the Tractatus to be a proper background for PI, and Kuusela in Development of Wittgenstein's Philosophy traces the threads connecting them.
    – Conifold
    Aug 22, 2021 at 20:09

4 Answers 4


The first person we'd have to convince of this in Wittgenstein himself, who took a somewhat dismissive, self-deprecating tone towards his earlier work in his later work (see the oblique self-reference in Philosophical Investigations §23).

I think it's more appropriate to see this as an evolution in Wittgenstein's thinking, not (1) mutually inconsistent theories or (2) independent but supplementary theories. Up and through the Tractatus Logico-Philosophicus Wittgenstein was fully aligned with the intent of (Russell's) Analytic Philosophy. He wanted to define philosophy as the application of a rigorous mathematized philosophy: something where we could translate freely between logical propositions and tangible real-world phenomena without getting lost in subjective or metaphysical maundering. As he says at the end to the Tractatus:

The right method of philosophy would be this. To say nothing except what can be said, i.e. the propositions of natural science, i.e. something that has nothing to do with philosophy: and then always, when someone else wished to say something metaphysical, to demonstrate to him that he had given no meaning to certain signs in his propositions. This method would be unsatisfying to the other — he would not have the feeling that we were teaching him philosophy — but it would be the only strictly correct method.

In short, here he wants to maneuver around anything that cannot be reduced to a 'natural sciences' type propositions by claiming that one of the 'signs' (terms, words) has no meaning (no clear and observable referent).

However, Russell's version of Analytic Philosophy ran into significant difficulties, mainly around the problem of 'denoting': of defending the unambiguous reference between a word (sign) and a unique real-world phenomenon. See Frege's "On Sense and Reference" and Russell's "On Denoting". While Russell et al were focused on trying to shore up the definitiveness of referencing, Wittgenstein went the other direction and began wondering whether he could rebuild what he wrote in the Tractatus using references that were purely conventional: determined only and entirely by use in language. (see PI §2, and the subsequent discussion). We can even see the beginnings of this line of thought in the Tractaus itself (TLP §§3.33-3.334), where he makes Russel's Paradox disappear by asserting that:

No proposition can say anything about itself, because the propositional sign cannot be contained in itself

...effectively distinguishing between two linguistic states that he would later come to call language games.

It seems clear to me that the 'early' and 'late' Wittgenstein are merely Wittgenstein developing his philosophy over time. People still working in AP or its descendants (e.g., Popper's school) tend to like the Tractatus, but to dismiss the Philosophical Investigations as an incomprehensible curiosity; they are the source of the 'two Wittgensteins' ideation.


Others here have already referenced Wittgenstein's own views on the matter, which I think are at least part of the basis for the conventional distinction in the secondary literature between "early Wittgenstein" and "later Wittgenstein." I do think there's value in highlighting specific continuities between the Tractatus and the Philosophical Investigations (e.g. both hold some form of "the limits of my language are the limits of my world"; both held some zen-koan view of philosophy as something to be resolved/dissolved and done away with rather than built upon), if only because the convention is to highlight their differences. But I would resist reading TLP as a theory of descriptive statements and PI as a compatible theory of other types of language use. Firstly, because PI isn't just about other uses of language; he has plenty to say about descriptive statements in PI. The main difference in perspectives, in my opinion, is that in TLP he was committed to having one picture (namely, the picture theory of meaning), and in PI he explored the idea that every way of conceptualizing the relation between language and reality is just one of many possible pictures (or metaphors, or analogies, or cognitive constructions, or tools for understanding, or whatever picture one chooses to conceptualize the conceptualization of the relation between language and reality). In TLP, his goal is a theory that will solve philosophy once and for all; in PI, his goal is a method that will dissolve philosophy once and for all. And I think later Wittgenstein's method is way more valuable than any theory that can be extracted from early or later Wittgenstein. That said, I still think the Tractatus is a brilliant work and still of philosophical (not merely historical) value, even though I disagree with most of it.


I don't think Wittgenstein did say the picture theory of language is fundamental. Instead he tested what happens if you assume that, and used that to see how we actually use language.

I'm always amazed this passage is so widely ignored by those who insist there's inconsistency between early and late Wittgenstein:

"My propositions serve as elucidations in the following way: anyone who understands me eventually recognizes them as nonsensical, when he has used them—as steps—to climb beyond them. (He must, so to speak, throw away the ladder after he has climbed up it.)

He must transcend these propositions, and then he will see the world aright."

-6.54, TLP

I really like this articlr on the method here, and linking it to other thinkers: Nāgārjuna, Nietzsche, and Rorty’s Strange Looping Trick.

Strictly speaking The Private Language Argument is derived from Wittgenstein's work, rather than being fully expressed in it. As I see it the TLP helped to understand the vast scope of what cannot be pictured, only pointed towards, and later how the understanding there involves shared modes of life, and the kinds of examples we have learned of different 'games' of pointing. Pointing is a game from doing that, not something learned by pointing at pointing in the world.

"Propositions can represent the whole of reality, but they cannot represent what they must have in common with reality in order to be able to represent it – logical form.

In order to be able to represent logical form, we should have to be able to station ourselves with propositions somewhere outside logic, that is to say outside the world."

-TLP 4.12

I hold to the Resolute Reading between early and late Wittgenstein, not because I see him as having advocated a picture theory, but because he showed what happens if you take such seriously - and said in the TLP that insisting it is how we see the world alone, can only fail to see things 'aright'.

The apparently simple 'basis' of language in picturing fails, because it cannot begin without a mind-to-mind agreement on what we are doing first, with pictures only following the arrival of logical form, that has it's basis in intersubjectivity. This is a fundamental shift from language as denoting or reference, to it arising from how we use. So we can acknowledge logical form, while also knowing we can reformulate it, towards our goals with it rather than as some kind of noumena constantly eluding us.

“The confusions which occupy us arise when language is like an engine idling, not when it is doing work.”

-PI 132


To use Aristotelean terminology (which Wittgenstein would not be happy about since he proudly claimed to have never read a single page of Aristotle), the content / matter of the theories changed, but the form stayed the same to some extend.

Late Wittgenstein rejected most of his early conclusions, but a lot the building blocks he used to reach those conclusions stayed the same. For example, he rejected his theory of atomic propositions representing atomic facts, but he still discussed propositions, sense and meaning in his later philosophy. He was interested in the same set of questions his whole life, and he gave 2 very different sets of answers to those questions.

Most philosophers would disagree, but there is quite a lot of overlap between early and late Wittgenstein.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .