«My sentences elucidate through this: who understands me recognizes them in the end as nonsensical»
– Wittgenstein, “Tractatus Logico-Philosophicus”, 6.54
«There are three strategies for criticizing Wittgenstein’s conclusions. The first two apply to any philosophical work, while the third aims at the Tractatus specifically.
The first of these strategies is the hardest to pursue and I will not do so in this book. It consists in a so-called immanent critique that takes Wittgenstein on his own terms and probes the soundness of his argument. It asks whether he somewhere committed a mistake that would invalidate his conclusion. This amounts to the difficult task of finding an inconsistency in his argument. While the immanent critics need to establish that it is quite impossible to reconcile some of Wittgenstein’s statements, his defenders have a much easier task (it is not a fair fight): all they need to show is that his statements can somehow be reconciled with more or less ingenuity and with the help of any and all interpretive resources. But though it is least likely to succeed, immanent critique provides the most valuable line of attack because it can teach us the most about Wittgenstein’s philosophical commitments.
The second critical strategy brings external considerations to bear: rather than demonstrate that the Tractatus is somehow flawed on its own terms, one might show that it fails to do justice to the phenomena of language and thought, that it fails to provide a complete picture, that it is seriously limited and takes far too much for granted. The one who pursued this strategy most vigorously and most successfully was Wittgenstein himself, especially in his Philosophical Investigations of 1953. This kind of external critique will add numerous qualifications to the Tractatus, pointing out the narrowness of its approach, determining the limits inherent in its focus on the descriptive language of science. And yet, all of this may leave untouched its core insight and disturbing conclusion. Yes, there is a lot more to be said about our uses of language, but does that in any way diminish the force of the original conclusion, namely that some sentences “work” pretty straightforwardly while many others deceive and confuse us by seeming to work like the straightforward ones? It is because of Wittgenstein’s later recognition of the many uses of language that many have spoken of two Wittgensteins – the early and the late Wittgenstein, Wittgenstein I and Wittgenstein Ⅱ – as if these were different persons, each the author of an important work with a lasting impact.2 For example, Bertrand Russell declared a few years after the death of his former student: “During the period since 1914 three philosophies have successively dominated the British philosophical world: first that of Wittgenstein’s Tractatus, second that of the Logical Positivists, and third that of Wittgenstein’s Philosophical Investigations.”3
Since these first two strategies may fail to directly engage and refute Wittgenstein’s disturbing conclusion, a third strategy combines the others in a manner that is specifically aimed at the Tractatus. This strategy is as old as the book itself. It was first advanced in Russell’s introduction to the Tractatus. After noting that Wittgenstein’s conclusion “grows naturally out of his doctrine,” Russell gives us good reason to quickly dismiss it nonetheless: “What causes hesitation is the fact that, after all, Mr Wittgenstein manages to say a good deal about what cannot be said, thus suggesting to the skeptical reader that possibly there may be some loophole through a hierarchy of languages, or by some other exit.”4
This is an immanent critique in that it discovers a tension within the Tractatus itself; it is external in that it adopts an outside point of view and points out how the work is systematically incomplete by failing to adequately reflect its own use of language. Russell draws our attention not to the aesthetic or stylistic peculiarities of Wittgenstein’s language. He simply asks how the pronouncements of the Tractatus fit into Wittgenstein’s threefold division of sentences. Statements like “The world is all that is the case” or “A sentence is a picture of reality” (TLP 1, 4.021) are unlike scientific descriptions of fact, nor are they logical theorems like the law of noncontradiction. If all other sentences are nonsensical then, surely, this must include the sentences of the Tractatus itself. Indeed, Wittgenstein himself embraced this implication: “My sentences elucidate through this: who understands me recognizes them in the end as nonsensical” (TLP 6.54).
By biting this bullet, Wittgenstein claims consistency for his work. But as we saw from Russell’s introduction, Wittgenstein’s willingness to concede that his own sentences are nonsensical has encountered incredulity from the very start. Where Russell tactfully noted “a certain sense of intellectual discomfort,” others took their gloves off. A. J. Ayer, for example, considered Wittgenstein’s claim “a vain attempt to have it both ways. No doubt some pieces of nonsense are more suggestive than others, but this does not give them any logical force. If the verification principle really is nonsensical, it states nothing; and if one holds that it states nothing, then one cannot also maintain that what it states is true.”5 (…)
In that case, he would be implicated in what is called a performative contradiction. However, if he was not intent to persuade his readers and instead produced empty words that he knew to be nonsensical, why did he bother at all and why should we pay attention to him? In this second case, we would call his efforts moot. And if we are to believe that he set out in good faith only to discover as he went along that his theory renders nonsensical the very statements that are necessary for its formulation, should we not consider this discovery an indictment of the theory? We might then call his attempt to persuade us self-defeating because his own theory deprived his proposed persuasions of their persuasive power. Or, to consider a fourth and last possibility, if the Tractatus actually persuaded us that there are very narrow limits to the sayable, would this not tempt us to take the very existence of this book as a welcome opportunity to escape from these limits through some loophole or some other exit? We might conclude, for example, that Wittgenstein somehow persuaded us and communicated matters without actually “saying” them. In this case, we would begin speculating about this mystically “other” means of communication and turn Wittgenstein’s clear analysis into an obscure flight from reason.
The Tractatus either involves a performative contradiction, or is moot, or self-defeating, or invites mystical speculation – none of these judgments is very flattering, and none of them compels us to accept its provocative conclusion. It therefore seems that by choosing to consider the Tractatus a meaningful philosophical document, we are already indicating that something must be wrong with it – at the very least, it cannot be (literally) true that (all) its propositions are nonsensical. And since for reasons of consistency Wittgenstein was forced to declare that his sentences are nonsensical, far more might be wrong with the Tractatus. (…)
All attempts to provide this defense will encounter a curious phenomenon: though it aims to survey and classify all of language, the Tractatus does not reflect on the sentences in which it is written, save for a few cryptic remarks. The silence it prescribes extends to its own composition. This silence needs to be broken if one wants to save the Tractatus from the charge that it commits a performative contradiction or that Wittgenstein could not have meant what he wrote or that by referring to Wittgenstein’s practice one can somehow escape his disturbing conclusion».