Tuesday, February 27, 2007

Attn: Peter Hacker

Today I saw something on an Amazon page that made me just about fall out of my chair laughing. Unfortunately, this puts me in a dilemma. I can just reproduce it without comment, but then only about five people will understand why I thought it was funny. Or I can explain it, killing the humor entirely. You already know what I'm going to do, so let me get to it. (In my defense let me say that I was eventually going to talk about this stuff anyway, and this seems like a good spur.)

I begin my explanation thus. In recent years there has raged in the scholarly teapot of Wittgenstein interpretation a bitter tempest indeed. Like many twentieth-century philosophers, including especially his logical-positivist followers, the early Wittgenstein rejected "metaphysical" questions as nonsensical pseudo-propositions. This much is clear; but how exactly does his argument go, and what exactly is it supposed to do? This is an important question even if one's main concern is with the later Wittgenstein, as it affects one's account of the continuity, or lack thereof, in the transition to the later view.

Nowadays, "metaphysical" is not the dismissive term it once was, at least to my ear. Any time you're talking about objectivity or truth or reality or reference or any kind of mind-world relation, you're "doing metaphysics"; and of course this is something everyone has to do, even if what you say about these things is that we should reconstrue them completely in order, well, not to "do metaphysics" in (what continues to be) the "bad" sense (for philosophers of certain persuasions, including but not limited to mine). What sense that is we can infer from the famous lines by Hilaire Belloc, who extols (extols, mind you) those Dons
With hearts of gold and lungs of bronze,
Who shout and bang and roar and bawl
The Absolute across the hall ["Lines to a Don"]
... as opposed, that is, for the record, to the dyspeptic nobody who "dared attack [Belloc's] Chesterton." Dyspeptic or not, those of us who reject "metaphysics" in this sense are put off less by shouting and banging (or, again, with truth and objectivity properly construed) than with what Kant called the "push to the unconditioned," i.e. positing (and attempting to describe, by means of a priori philosophical reflection) an Absolute Reality underlying or "grounding" or transcending the ("mere") contingency of our worldly experience.

Even after Kant's attempt to cut metaphysics down to size, in Russell's time the halls of Cambridge still rang with such bawling, e.g., in the voices of the British Hegelians and their followers (not sure which of them were at Cambridge exactly, but we speak here of the likes of McTaggart, Bradley, and T. H. Green). To Russell, the refugee from Hegelianism, this was all pretentious nonsense; but how can you prove that something is nonsense? It seems, as Kant had noted, that in order to draw a line between sense and nonsense one would have to be able to think both sides, which is supposedly just what one can't do (which leads Kant to take a different tack in his rejection of "transcendental illusion" – but we're not here to talk about Kant).

Russell's "young engineer" Wittgenstein took on this task with a vengeance. Convinced by Russell and Frege that "whatever can be said at all can be said clearly," in the Tractatus Wittgenstein lays out a theory of meaning and reference which delineates, from within as it were, the limits of intelligibility (c.f. the preface: "it will therefore only be in language that the limit can be drawn, and what lies on the other side will simply be nonsense" (p. 3)). The book ends with the famous pronouncement that "whereof we cannot speak, thereof we must be silent."

Here's where the trouble begins. For one standard translation (i.e. of Wovon man nicht sprechen kann, darĂ¼ber muss man schweigen) reads instead: "What we cannot speak about we must pass over in silence." This translation of the enigmatic sentence implies that even if language can describe only the world of empirical contingency, there are mystical truths which transcend our language – truths beyond its reach, yet no less real and pressing for all that. The point of saying this, presumably, is that when metaphysics tries to articulate these truths, it necessarily fails, so it should stop trying; yet, again, they remain truths all the same. On this reading, Wittgenstein is saying that metaphysics is nonsense (non-sense), but it's "substantial" nonsense, not gibberish: it gestures at important truths beyond the reach of our cognition (they're there all right, but we must "pass over" them in silence). Indeed, Wittgenstein explicitly and repeatedly (although not so much in the Tractatus itself) proclaims his conviction that the "ethical" (by which he means concern with value) is intensely important for him, and it is not the ethical itself, but only philosophical pseudo-inquiry thereinto that he rejects.

One side of the scholarly controversy accepts this reading, but the other side regards it as committing Wittgenstein to exactly that which (on this alternate reading) he was concerned, in the closing sections of the Tractatus, to reject: the idea of ("ethical" propositions as) "substantial" nonsense, as opposed to mere gibberish. Instead, say these "New Wittgensteinians" (led by Cora Diamond and James Conant), nothing "transcends" our language in this sense: there's what we can say, and that's it. "Beyond" language there lies nothing over which we must (perhaps wistfully) pass in silence, and sentences which purport to describe "the transcendent" (in this sense) are nothing but mere nonsense: "the essence of any entity resides in its substantial form" is no different from "karvo sotok skebanzulane." It's not that they (necessarily) fail to capture the thought at which they aim, but instead that there is no thought there in the first place to be captured. There's sense and there's nonsense. The former describes the world; the latter does nothing except sow confusion (sense talks; nonsense walks). If that were all there were to it, this reading would leave Wittgenstein sounding like a positivist (and of course the Vienna Circle positivists took themselves to be avid Wittgensteinians, with the Tractatus as their bible); but of course there's plenty more, most of which we won't get into. (Short version: there's no reason to see this "Jacobin" view of nonsense as, for example, in any way downplaying Wittgenstein's sense of the importance of ethics.)

Our concern here is the characteristic metaphilosophical strategy Diamond et. al. find in the closing passages of the Tractatus. Statement 6 gives "the general form of a proposition," and in the subsequent commentary Wittgenstein explains how we can see the propositions of logic as tautologies without content, and thus as saying nothing about the world. This means that (6.13) "[l]ogic is not a body of doctrine, but a mirror-image of the world. Logic is transcendental." 6.42 tells us that "it is impossible for there to be propositions of ethics. Propositions can express nothing that is higher."

So far, there's no reason to think anything is amiss with the traditional interpretation. Wittgenstein is summing up his conclusions about the limits of language and the status of ethical pseudo-propositions. Even as late as 6.522, he says that "[t]here are, indeed, things that cannot be put into words. They make themselves manifest. They are what is mystical." A mere half-page later comes the fateful statement 7, and the book is over.

So why do the NWs read the Tractatus in the way they do? Consider what Wittgenstein has to say about philosophy. 4.003 reads: "Most of the propositions and questions to be found in philosophical works are not false but nonsensical." This is just what we might expect him to say, given his rejection of "metaphysics." But isn't the Tractatus itself a work of philosophy? After all, it's hardly a work of natural science (or any other worldly inquiry). But this means that since the propositions of the Tractatus itself do not simply (as a proposition must, if it is to have a sense) picture states of affairs in the world, they too must lack sense. If so, then they can't be true; and if they're not true, then what good are they?

The threat of self-refutation looms. Let's pursue it a little further. At 4.112, Wittgenstein tells us that philosophy is (or should be) "not a body of doctrine but an activity." 4.114 elaborates, in familiar terms: "It must set limits to what can be thought; and in doing so, to what cannot be thought" – which is of course what the Tractatus says it's doing. That the Tractatus is not a body of doctrine might mean that it doesn't matter that its propositions can't be true, as their importance lies not in what they say, but in what they do. But if they don't say anything – if they are utter nonsense, the equivalent of "karvo sotok skebanzulane" – then how could they do anything, except puzzle you? After all, this applies even to (pseudo-) statements like 4.112, philosophical as they are. That something is meant as an "elucidation" (4.112) rather than as a statement of doctrine cannot save it from meaninglessness. It looks like the strategic retreat from doctrine to elucidation cannot help.

The last few remarks of the book (except for 7) are comments on 6.5, which reads: "When the answer cannot be put into words, neither can the question be put into words. The riddle does not exist. If a question can be framed at all, it is also possible to answer it." So where there can be no "answer," there can be no question either – leaving no unanswered questions. The trick, then, is to see this – to overcome our feeling of dissatisfaction with the actual, intelligible questions available to us (answered or not).
6.52: We feel that even when all possible scientific questions have been answered, the problems of life remain completely untouched. Of course then there are no questions left, and this itself is the answer.
We must move from seeing the world in one way to seeing it in another – from seeing it as concealing something essential behind the appearances, to seeing it as it is, i.e., as complete in itself, even if not fully known, given the existence of unanswered yet intelligible empirical questions. When we do this we will cease to scratch where we come to understand that it cannot really itch. But then comes 6.522 (quoted above), in which, as if anticipating the subsequent positivist misunderstanding of what he is doing, Wittgenstein affirms the "mystical" and its importance. We must not confuse hard-won understanding with tendentious and unconvincing behaviorist denial (i.e., of such things as "itches").

So now we know where we're supposed to end up. But how to get there? Even the statements which tell us our goal are still unintelligible by their own lights. As already suggested, the answer has to do with the difference between philosophy-as-doctrine (i.e., as argument) and philosophy-as-elucidation. We think of philosophical arguments as "moving" or "taking" us from one "place" (the premises) to another (the conclusion), by means of the license (or force) provided by logical (deductive) inference. But all they really do, as we shouldn't even need Wittgenstein's analysis of the sense of propositions to tell us, is bring out the implications of what we were already committed to in believing the premises (which of course we then leave in place). In adding a new belief, we merely come to see a new detail in what was already present; or perhaps we find that our commitment to these premises requires that we give up (as unwarranted) some other belief less firmly entrenched than they. In either case (as well as in cases of empirical or inductive inferences) we see our resulting view as an elaboration or correction of our doctrines, such that they now do better what they were already doing: mirroring reality. Thus is philosophical progress made.

But Wittgenstein explicitly denies that this is what he is doing (i.e., what "philosophy [as opposed to science] does"). By doing so even as early as 4.112 (i.e., not simply springing it on us at the end), he has carefully maneuvered us into a position where seeing what he is doing and understanding how he is doing it (that is, how to understand the self-referential nature of his procedure) is the same thing. And to see this is to see how the self-reference of his claims, while it does indeed result in their undermining as claims, is nothing to be feared, but instead holds the key to understanding. We're used to thinking of self-reference in terms of things like the Liar (and the supposed cure for these things, Russell's "theory of types"), i.e., as manifested in sentences like "this sentence is false" – a strict contradiction (for Wittgenstein's response to Russell see 3.331-2). That's not what we're talking about here.

The problem is this. The (sentences of the) Tractatus tell us that certain forms of words have no sense (are nonsense). Yet they themselves seem to be of that very form. This would be a strict contradiction only if – as in "this sentence is false" – we read them as assertions. But they themselves, taken together (not each by itself, as the Liar), tell us not to do that. Of course we still have a puzzle – a tension between what the sentences (seem to) say and what that which they (seem to) say tells us, about whether that appearance of sense is misleading – but a puzzle is not yet a contradiction. (The "puzzle" of the Liar is that it is a (seemingly sensible) contradiction.)

Turn back again to what the sentences in question say/seem to say. We are to move from seeing the world in one way to seeing it in another, even while recognizing that nothing has changed. It was something we felt to be missing that we must learn to see as – not present, of course, but not "missing" either. What we are actually looking at – (the truth about) the world – remains unchanged. The change is in ourselves, not the world (and not our knowledge of the world). When we see this once, we can see it again (whichever way we see it first). Just as we move from seeing the truth about the world as (empirical/scientific propositions plus philosophical/ethical/transcendent propositions) to (empirical/scientific propositions only), we move from seeing philosophy-as-establishment-of-doctrine to philosophy-as-elucidation. It is obvious how the latter two are different; indeed, most philosophers see the latter as some kind of nihilism and not philosophy at all. But how (as they must be if the parallel is to hold) are they the same? How is the change in ourselves and not in philosophy?

To ask is to answer. The former, while it seemed to move us from one place to another (by forcing on us a belief we didn't already have), actually left us (in simply revealing what was already implicit in the premises) with a new understanding of the status quo ante. The latter, on the other hand, in dealing only with nonsense, seemed instead to leave everything in place, while actually moving us from one way of seeing to another. When we put it like this, the answer jumps out at us: in each case the change is in us and not the (truth about the) world. Let's see how Wittgenstein himself puts it, in the decisive proposition, immediately preceding 7. If 6.54 works, 7 will then strike us in just the right way: as summing up the book – as telegraphed in the Preface, no less – but as not saying anything at all. After all, unlike most "philosophical propositions," it is openly tautologous: what we cannot speak about, we cannot speak about (duh). It is when we do not understand that we look behind or underneath it for some deep meaning – which is of course the point of the entire book. So, finally, here is 6.54:
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 up 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.
Wittgenstein has not mentioned "elucidations" since 4.112 (and the only other reference is 3.263, which in retrospect can be seen to make the same familiar point, only at a different level; (intelligible) propositions "elucidate" the elementary propositions which (conceptually speaking) make them up – but they can only work that way unless if they only say what we already know). We cannot understand what is nonsensical, and Wittgenstein does not intend us to try. Instead we are to understand him. We do so by moving from reading the book's propositions as doctrine to seeing them as nonsensical. That is, we can recognize them as nonsensical when we understand how they can be nonsensical and still have the same effect on us as if (per impossibile) they weren't. After all, they've been telling us for some time that they must be nonsensical (i.e., qua philosophical); the question was how this could be. But we don't do this by simply changing our beliefs about the propositions, as if changing our judgment of the truth-value of these propositions are nonsensical from "false" to "true". Instead, in reading them as doctrine, we use them on ourselves, so that we are then able to see them differently. We "throw away the ladder" only when we have already climbed up it; we do not jump down again once we see it as not really having been able to support our weight (so that we have no "right" to be where we are).

Yet once the propositions have been "transcended," Wittgenstein tells us, we will then be able to "see the world aright." We know now, though, what this can only mean. It's not that we now see the truth where before we were ignorant or mistaken (i.e. our views are "right"). We saw the truth already (that is, to the same extent that we do now, given our continuing non-omniscience about empirical matters); what we now do is understand that that's what we're doing, and that that's all there is to do. That is, now we see rightly what has not changed.

This is tricky, so let me say it again. Before we read the Tractatus, we saw our knowledge of the world as incomplete, and not only in the sense in which it always must be (given our permanent non-omniscience). We wanted answers to our deep questions about (I can't resist) Life, the Universe, and Everything. So we read the book; but it doesn't really give us an answer to our question. It led us on, implying that something in the nature of logic and language and meaning would tell us how to do metaphysics properly, as Kant does (or tries to); but then right at the end it pulls the rug out from under you. Nothing has changed; all we learn is what we already knew. But things are supposed to look different. So we look again.

We look again at ... what? At the only thing we can look at, because it's the only thing there is: the world. If we have understood what (Wittgenstein's) philosophy has done, we will be able to overcome the lingering feeling that something is missing, both in it and in our knowledge, in their joint failure to tell us what we thought we wanted to know, about "the transcendent" or whatever. That is, what is indeed missing from even a complete description of the world is not to be found in a further description of a transcendent reality, to be supplied by "metaphysics" (for this is what it seemed that we needed), but is instead that which we can now see to be implicit in the very idea of a description of the world (i.e. the facts). Keeping in mind its fundamental importance (rather than dismissing it as positivists do), we must now learn to look for it (if that's even how we still want to say what we're doing) in a different way: by seeing the (one and only) world differently. That is, seeing it rightly ("aright") – which of course means that if we see "only" the entire truth about the world, we won't have seen the world "rightly" at all.

Now after all the heavy technical weather of the preceding pages, this moral (that we need to learn to see the world "rightly") may, like that tautological final sentence, seem like not much help. But this is of course Wittgenstein's own verdict as well. Return to the preface:
[t]he truth of the thoughts that are here communicated seems to me unassailable and definitive. [... and the] second thing in which the value of this work consists is that it shows how little is achieved when these problems are solved.
Yet if we don't learn to see the world rightly – or, that is, if we continue to look for "ultimate reality" either in (i.e. as constituted by) the facts about the world or beyond them, then everything we do do, in philosophy or not, will be futile. So even if "little is achieved" in one sense, it's still something that has to be done – or we will continue to search blind alleys for an "answer" which does not and cannot exist.

So, does this mean the "New Wittgensteinians" are right? Not so fast. While the NWs have performed an important service in focusing our attention on the metaphilosophical strategy of the Tractatus as manifested/addressed in 6.5 through 7 (especially 6.54), the battle is far from over. As the traditionalists, led by Peter Hacker, point out, there's plenty of evidence, within the Tractatus as well as in other writings, from before, during, and (importantly) after the period when he was working on the book, that Wittgenstein intended his admittedly nonsensical propositions to communicate several centrally important yet unfortunately incommunicable truths (incommunicable because of what they themselves say about language) – rendering this admission of nonsense disturbingly paradoxical once again. In his contribution to Crary and Read (The New Wittgenstein), sportingly appended to the other contributions as "a dissenting voice" to the NW view, Hacker begins and ends, as we might expect, by stressing 6.522 as an undeniable commitment to the idea of "substantive nonsense" – nonsense that reaches at something transcendent and (necessarily) fails to do so. This renders the threat of self-refutation not merely apparent, but actual – and fatal. Realizing this is what eventually led Wittgenstein to change his views.

My own view ... will be the subject of some other post. Now that I have made it impossible to do so, let us finally laugh together at the knee-slapping irony of the following. It's an excerpt from a reader review at Amazon.com. The book in question is The Sokal Hoax, which is a collection of writing about (drum roll) ... the Sokal hoax, including the original article, the unveiling of its hoaxitudinal nature in Lingua Franca, and part of the subsequent donnybrook. As does a good deal of the book itself, most of the reader reviews chortle at the postmodern folly which was Sokal's target, but some are unamused. The review in question actually seems to be reviewing not The Sokal Hoax at all, but Sokal's later collab with Jean Bricmont, Fashionable Nonsense, with which our reviewer is unimpressed. He decries the method whereby they "cut and paste some random excerpt" of the target writer (say Deleuze) in the hope that out of context, sans explanation, it will look ridiculous – which proves nothing.

You can see where this is going. To show that turnabout is fair play, our reviewer picks a random excerpt of some supposedly rigorously logical analytic-philosophy writing to see how we like it:
Let us apply this operation to a random writer who penned the following statement:

"The elementary proposition consists of names.
Since we cannot give the number of names with
different meanings, we cannot give the composition
of the elementary proposition. Our fundamental
principle is that every question which can be decided
at all by logic can be decided off-hand."

This excerpt -- which any man in the street will tell you is just as nonsensical as any of the excerpts Sokal cuts and pastes from other authors -- was written by Ludwig Wittgenstein in the _Tracticus Logico-Philosophicus_. According to the Sokal/Dawkins argument, this proves that the _Tracticus_ it complete and total nonsense, and that anybody who claims Wittgenstein writes anything more than meaningless gibberish is simply lying.
Isn't that great??

Speaking of "cutting and pasting," that's what I did here – so "sic" the whole thing ("Tracticus"? "Dawkins?"). This is glorious on so many levels. That it sounds like an excerpt from a parody article which some joker tried to sneak into Crary and Read is rich enough, but the implicit idea that self-appointed science 'n' rationality defenders like Sokal et al would be at all reluctant to put Wittgenstein, of all people, in the same category (re: nonsense) as Derrida: priceless. The reviewer thinks that the verdict of the "man on the street" about these lines – that they are "nonsensical" – is a reductio of Sokal & Bricmont's method; but they would turn the tables on him and agree with that verdict ... for the wrong reason! And to top it off, so would Wittgenstein himself, for another reason still! And what that reason is is itself a hot topic of debate in the Wittgenstein community! You can't make this stuff up. Or at least I hope you can't.

For the record, our man may be taking a few liberties here. That is, that excerpt seems to be a bit more random than he lets on. I don't have a machine-searchable text of the Tractatus, and I didn't search exhaustively by hand, but I can't find that selection in my copy, at least as is. The first sentence is from 4.22 (although Pears & McGuinness have "An" rather than "The"). I can't find the other two sentences at all, although they do sound like things the early Wittgenstein would say, esp. the third one (depending on what "off-hand" means).


J. said...

Great post. If anything, the lead-up gave the punchline an even stronger punch.

The quotation is a poor translation:

"5.55 [...] Elementary propositions consist of names. Since, however, we are unable to give the number of names with different meanings, we are also unable to give the composition of elementary propositions.

"5.551 Our fundamental principle is that whenever a question can be decided by logic at all it must be possible to decide it without more ado."

Duck said...

Ah, thank you. I was sure it must have been in there somewhere; I was looking in the 4's, where I found 4.22, and I thought that was his first sentence. I should go through the 5's again – some good stuff in there!

Clark Goble said...

My favorite post of yours yet.

BTW - how influenced are the New Wittgensteinians by Derrida? (I'll confess to not being up on Wittgenstein as I probably should be)

N. N. said...

That review is hillarious. The implication is: 'Of course Sokal is wrong. Otherwise the Tractatus would be nonsense. And who in their right mind would claim that?'

As for the NWs, it's hard to take them seriously. It's simply impossible to be resolute while reading Remarks on Logical Form or Wittgenstein's discussions with Schlick and Waismann.

Duck said...

I see that Blogger's acting up again - I tried to leave a comment and it wouldn't show me the word I was supposed to type into the box. Hope that didn't happen to anyone else. Anyway I have disabled the spam filter for now. Any spam though, and back up it goes.

So anyway, here's what I was going to say.

Thanks Clark, N. N., for your comments.

Clark: I don't think the NWs are influenced by Derrida much at all. Some (but not all) are interested in him though. One of the best articles in Crary & Read is Martin Stone's "Wittgenstein on deconstruction," which is somewhat along the lines of Wheeler's take on W. as a "conservative deconstructor" (you have mentioned reading this book, I think); though Stone is a bit less impressed with Derrida than Wheeler is. Good article - check it out. To my knowledge Diamond and Conant don't mention Derrida. Cavell does, but he's far from an archetypal NW. (Let me take this opportunity to give The Claim of Reason my highest recommendation; there's a key exc. from same in Crary & Read for a taste.)

N. N.: Are you the real N. N., or someone else with the same name? (Never mind.) As I said in the post, I'll try to say something about my own view later on. I actually don't find the NWs any harder to take seriously than Hacker, or the Tractatus itself for that matter. But I have to say they haven't sold me yet. Whoever does convince me will do so by telling me what I want to hear about the Investigations, together with a good story about how we got there from the Tractatus – and the NWs are much closer to doing that than Hacker et al are (not there yet though). On the other hand, as you say, (and as I also mentioned in the post), that extra-Tractarian textual evidence is nothing to sneeze at (Hacker really piles it on in the articles I've seen). Do you have a reference for the discussions with Schlick? One of my advisors was a big (pre-positivist) Schlick fan, and one reason he didn't like Wittgenstein is that he thought W. had corrupted Schlick and turned him into a positivist.

N. N. said...

I'm the N. N. whose name is meaningful even in the event of my death.

The 'discussions' between Wittgenstein, Schlick and Waismann were dominated by Wittgenstein; Schlick and Waismann ask questions here and there. Waismann took notes which are reproduced in Wittgenstein and the Vienna Circle: Conversations Recorded by
Friedrich Waismann
, edited by Brian McGuinness.

Wittgenstein also dictated a massive amount of material to Waismann for the book on Wittgenstein's philosophy that Waismann was to write (Schlick was the primary influence behind the project, and no doubt read everything Waismann recorded). That material has been edited by Gordan Baker in Voices of Wittgenstein: The Vienna Circle: Ludwig Wittgenstein and Friedrich Waismann. Waismann's book was never finished. What he did write was published after his death as Principles of Linguistic Philosophy.

In my opinion, an excellent account of Schlick's development is given by Coffa in The Semantic Tradition from Kant to Carnap.

I guess I don't have any difficulty taking the Tractatus seriously, but that's because I am not attracted by 'resolute' readings at all. In fact, I am violently opposed to them. As for Hacker, I think the second edition of Insight and Illusion is probably the best short volume on Wittgenstein's philosophy that is currently available. It has it's problems, but in Wittgenstein scholarship, that's a given.

Duck said...

Thanks for the references. Now I know which books in particular I should regret not having the time to read (I asked for them, I know).

DR said...

Have you read Denis McManus's The Enchantment of Words? I haven't seen any reviews of it yet, nor have I had the time to think about it very carefully, but a quick read through gave me a very favourable impression. He's more or less a "new Wittgensteinian", although I think he has a footnote saying that one reviewer didn't understand why he thought of his position that way.

As for the external evidence that Hacker sites, it is impressive, but Diamond and/or Conant have/has responded to it somewhere. Not in detail though. They argue that each particular remark would have to be addressed individually, and I think they think they have better things to do with their time than go through them all one by one.

One thing that complicates matters is that it isn't clear what Wittgenstein would have said about the Tractatus if the new reading is right. That is, if it is meant to be nonsense that the reader comes to realize gradually, or at least through hard work, is nonsense, then Wittgenstein would hardly go around telling people that it is all just nonsense. What he would say, I don't know. Which means that quoting him talking to people as if it is not all nonsense does not prove that the new reading is wrong. This will seem like special pleading to some (perhaps to most), but it also suggests that uses of the external evidence beg the question.

Duck said...

DR – thanks for your comment.

I haven't read that book, but I have seen other things by McManus that I liked. I'll add his book to the list (*sigh*). I pretty much agree with your last paragraph, and was going to say something like that in my promised propria persona post (still will, I guess). The key for me is still the relation between the TLP and PI.

I have looked at Insight and Illusion, but I don't own it and it is not to hand. I remember being struck by the number of times Hacker says, in the preface to the second edition, things like "I used to think X, but I have since become convinced that Y" – where Y is quite different (and indeed much better). And I did like his little book in the Routledge The Great Philosophers series. So maybe there's hope for agreement of some kind.

N. N. said...


The difficulty I have with the reply you suggest on the NW's behalf is that most of Wittgenstein's later comments about the Tractatus were not made in conversation. They were written in notebooks, etc. which Wittgenstein had no intention of ever showing to anyone (he instructed Russell to burn his pre-Tractatus notebooks). Given that we don't find a straight forward statement of the 'resolute' thesis in any of the source material, the entire burden of the NW intepretation rests on the 'frame.' In my opinion, that's quite a stretch.

Anonymous said...


Yes, the reply I suggested is at best partial.


surlacarte said...

I quite enjoyed this post (I was referred here by The Valve) and, though I'm not as familiar with Wittgenstein as I ought to be, think I followed it well enough to get the punchline at the end. However, I'm confused by one move in your argument, namely the move from positing to "seeing".

It seems as if the apparent logical contradictions you outline are somehow all sown up by the possibility that the reader of Wittgenstein's Tractatus will come to "see" the text as nonsense. The reference to sight seems to suggest that we've moved to something non-linguistic and thus something no longer subject to the statements about language Wittgenstein makes in the text. Thus, whereas the nonsensical character of Wittgenstein's language could not be stated as such (for such a statement, of the form "this sentence is nonsensical" would itself be nonsensical, by virtue of its lack of a referent), it could somehow be "seen" in a non-linguistic way.

However, insofar as the form of your own statement ("seeing X as nonsense") implies a metaphorical operation, i.e. taking one thing for another, seeing (which is itself a metaphor for some property of the seer) turns out to be basically linguistic.

In other words, I'm not sure I see the difference between seeing a sentence as nonsensical and stating that it is nonsensical, given that:
1) Through the same logic by which we can call the statement "X is nonsensical" nonsensical, we can call nonsensical the visual system which sees X as nonsensical
2) We're not literally talking about the visual system

Again, I emphasize my ignorance on the subject, so if this comment doesn't make any sense or totally misses the point, I apologize. But I think it would significantly help my understanding of Wittgenstein if you could explain how changing the way we see Wittgenstein's language avoids nonsense in a way that changing the way we talk about it doesn't.

Duck said...

Well, surlacarte, I'm not sure exactly what you're getting at, but I'm certainly not ascribing that to any ignorance on your part. It may help to say this: the standard view, to which the NWs are reacting, is the "two-Wittgenstein" view, according to which he came to see that the TLP fails, and so changed his views completely (and indeed, TLP and PI are two very different books). This means that whenever you advance an interpretation of the former, and someone complains that on that interpretation the book fails, you get to say: well, of course – that's why he changed his views; while if you stress the continuity of his views, you can't shrug off problems in TLP so easily. But that doesn't mean you can't acknowledge them at all. Wittgenstein's views did indeed undergo an important change. He himself says in the Investigations that he was forced to acknowledge grave difficulties in it (I'm not looking up the reference, but it's toward the beginning, I think).

So anyway, maybe the problems you see in the TLP as described are quite real – and maybe he himself thought this too – yet the interpretation can still be correct.

Let me try to answer anyway. "Seeing" doesn't seem to me to be "non-linguistic" in the way you suggest. If some form of words is nonsensical, we see this by using our linguistic understanding. The form of words seems to make sense; but when we think about what does make sense we see that it can't – it's an empty form of words corresponding to no thought (and thus to no state of affairs). I don't see why I can't express that judgment in language. ("That form of words seems to make sense, but does not.") After all, linguistics is an empirical discipline like any other. Wittgenstein isn't ruling out any and all semantic or "grammatical" language – just "metaphysical" language; not any and all judgments of nonsense (which after all – the judgments, that is – can be "given a sense" just like anything else), but a priori arguments (as in TLP) about the bounds of sense in general (even from inside). (If this is not right for the early Wittgenstein, then it is indeed something that he changed later on.)

On the other hand, neither does "seeing X as Y" strike me as "basically linguistic" except in the sense that it involves intentional attitudes (which is something, certainly). When we explain how the dog mistook one thing for another (if that's what we want to say), we might very well say that he saw X as Y. It does become harder to ascribe this sort of thing to non-discursive creatures though.

The point of the NW TLP (and PI) is, as they themselves put it, "therapeutical." If it works, a pernicious temptation to utter (a particular kind of) nonsense will be exorcised (if only temporarily). The NW Tractarian version of this "therapy" (not the PI though, on my view) is well described as "Pyrrhonian": as those ancient skeptics said of their views, "they expurge themselves along with the disease" (or something like that). I think when that happens we're not supposed to look too closely at what came up. Instead, we look at the world itself, freed of the temptation to look beyond or beneath.

But I'm still not sure I've answered your question.

surlacarte said...

Thank you for your thoughtful response. As you say, the NWs may be correct to interpret Wittgenstein in this way even if there are grave difficulties in the theory itself as they interpret it. I think I took you, in attempting to render the positions of the NWs coherent, as defending their interpretation of Wittgenstein, and, in turn, I took the NWs, in attempting to render Wittgenstein's positions as coherent, as defending those positions. This may certainly not be the case, so you can file my comment under potential problems in the TLP, as understood by the NWs.

Nonetheless, I'll still attempt to clarify my question. Consider the following sentence from you post:

"Just as we move from seeing the truth about the world as (empirical/scientific propositions plus philosophical/ethical/transcendent propositions) to (empirical/scientific propositions only), we move from seeing philosophy-as-establishment-of-doctrine to philosophy-as-elucidation"

My contention is that, according to the logic of the TLP, as interpretted by the NWs, as interpretted by you, this sentence does not make sense. Part of the reason my reasoning may be confusing is that this contention relies on two assumptions:

1) The propositions "the truth of the world is (empirical/scientific propositions plus philosophical/ethical/transcendent propositions)" and "the truth of the world is (empirical/scientific propositions only)" are nonsensical (when taken as propositions) according to the logic of the TLP as interpretted by the NWs as interpretted by you

This assumption may be based on an overly broad reading of the following paragraph from your post:

Consider what Wittgenstein has to say about philosophy. 4.003 reads: "Most of the propositions and questions to be found in philosophical works are not false but nonsensical." This is just what we might expect him to say, given his rejection of "metaphysics." But isn't the Tractatus itself a work of philosophy? After all, it's hardly a work of natural science (or any other worldly inquiry). But this means that since the propositions of the Tractatus itself do not simply (as a proposition must, if it is to have a sense) picture states of affairs in the world, they too must lack sense. If so, then they can't be true; and if they're not true, then what good are they?

Of course even the intial Wittgenstein quote applies only to "most of the propositions and questions to be found in philosophical works," not all. The source of my confusion may be that I take propositions to which this conclusion does not applies (namely, above the propostions about truth) to be included in this set of nonsensical propositions. So, I ask, are propositions of the form "the truth of the world is (empirical/scientific propositions only)" nonsensical according to the NWs' interpretation of the TLP? If not, then my question has been answered. If so, please continue reading.

2) The propositions "a sees x" and "a see x as y" make sense only if the propositions "x exists" and "x is y" make sense

This is what I meant, I think, by calling "seeing x as y" basically lingusistic. My point was simply that an act of seeing implies a claim about the object, such that if the claim about the object is nonsensical, the act of seeing is as well. Thus I don't see a fundamental difference between stating "dada" and stating "I see dada." If the proposition "dada" does not make sense, how can it possibly make sense for me to see it? It may be a mistake to attribute this position to Wittgenstein, in which case, again, my question would not apply, but it would seem to follow from his definition of nonsense.

If it turns out that the proposition "the truth of the world is (empirical/scientific propositions only)" is nonsensical, and that "a sees x" makes sense only if "x exists" makes sense, then it follows that the phrase "seeing the truth about the world as (empirical/scientific propositions plus philosophical/ethical/transcendent propositions)" does not make any sense. How Wittgenstein's text can result in moving from one nonsensical state to another nonsensical state, well, doesn't make sense. What does it even mean to move from seeing truth one way to seeing truth another way, if truth is a metaphysical concept that doesn't make sense? And, therefore, have the NWs really "interpretted" the TLP if the result of the interpretation is to state something meaningless about the effects of the text? Finally, how can this nonsensical change be called therapeutic if it is impossible (or nonsensical) to say what has changed?

That's why I ask about the difference between seeing and stating. The only difference I see is that if you had stated the same quote about truth in terms of stating rather than seeing, it might have been immediately apparent that it didn't make sense. Since it's harder to see why "seeing truth as x" is nonsensical, since "seeing a dog as a person" does make sense, the switch from stating to seeing seems to help sow up your explanation of the NWs explanation of Wittgenstein, when actually neither make sense.


Assuming, again, that my assumptions are correct, which they might not be. Either way, thanks for you input.

Duck said...


Thanks for your persistence. I think my view of the matter may become clearer (not least to me) after I post a bit more about the Investigations and its criticism of TLP. I hope you will stick around (or return, as this may not happen right away) for that. I may repeat myself a bit (now and later), so be warned.

You are generally right about what I was doing. I was indeed presenting the NW view in what for me would be its best light, sweeping problems (i.e. with the TLP so construed) under the rug for the moment. I think this makes sense: we see what went wrong, if anything, by seeing what would have gone right, if everything had indeed gone right (if that makes sense). I do find problems in TLP; but so does Wittgenstein himself, as well as the NWs. This is not always clear in the NW literature, much of which, not surprisingly, is devoted to defending TLP against standard-view accusations of crude self-refutation, which they take to miss what he was up to. So we may agree in part, but your objections sound to me more like the standard-view kind rather than the NW kind, or mine (which I stlll can't quite call NW).

Hmmm. I'm wondering what to say here and what to say later. [...]

Okay, I started an answer but it got too long and now works better (once finished) as a separate post. Here's my short answer for now. The problem here still strikes me as having to do more with Wittgenstein's aims in TLP than with the details of his strategy (which is, however, where we will indeed, as he does, find some problems). What if (1) and (2) were both true, as you say? Why isn't pointing that out the same thing (putting to one side the self-referential aspect of the Tractarian case) as saying "yes, it's spectacular, but he isn't really sawing that woman in half and 'magically' putting her back together unharmed – the very idea is ludicrous!"? You may spoil the effect, or you may not; but in either case, why bother? Can one "refute" therapy? Why should one try? Shouldn't one at least consider whether therapy is warranted (i.e. the nature of the illness)?

Or consider it this way. You write of Wittgenstein's "positions." Are these his claims, or the beliefs he manifests in (perhaps only appearing) to claim them? What if they're not the same? It's natural to focus on the assertions in a philosophical text; indeed, on the standard conception of philosophy that's all that ever counts in it. But this assumption itself is (one manifestation of) Wittgenstein's target. What he really cares about may be something else entirely. Yet even to convict his (seeming) propositions of falsehood or insincerity (or nonsense) may be to miss the point – that is, as he distinguishes the two in 6.54, not their point (significance) but his point (intent).

I'll say more later. Just think about that for now. In any case I see no point in disputing (1) or (2) rather than their significance (or necessarily affirming them either for that matter, for the same reason). So while I hate (unlike W. himself!) to be enigmatic, at least this rules out certain possibilities, which I hope is some help.

Anonymous said...

Most excellent post, but does Wittgenstein reject a priori truths? Me TLP is not handy, but I seem to recall he didn't, and affirms a somewhat Fregean-platonic view of logical forms as a priori and transcendent--tho' as with most LW's he doesn't really argue; he hands down Die Wahrheit from the heights. And in a sense that is THE central issue to much analytical muckery. Analyticity does not equal a prioricity, right, but without any decent arguments for a priori status of logical entities (mathematics, axiomatic knowledge), there is really not much to a lot of analytical chat--as I think Quine the closet nominalist realized.

Duck said...

Glad you liked the post, but part of its point is that even if your TLP is handy it's not exactly clear what W is "affirming" or "rejecting". He (okay, his propositions) say (or seem to) that logical propositions are tautologies – so if tautologies are a priori truths, then logical propositions are a priori truths – but they also say that logical propositions are (because tautologous) sinnlos. So he "rejects" them in that sense; but at least they're not unsinnig. We don't even know how to use the vocabulary of "affirmation" or "rejection" until we decide what his overall strategy is: that's why 6.54 is the key proposition. Naturally this makes it look like he is waffling, and of course the standard view reads the TLP as straightforwardly self-contradictory. This strikes me as a strength of the NW reading: even if the TLP doesn't work (which I think it does not), it's hard to believe that that's the reason (and that W. didn't see this until much later).

I personally don't make much of the a priori/a posteriori distinction. If you want to make some claim an then go on to say it's a priori too, then go right ahead. Maybe that'll help me see what you mean by it.

Anonymous said...

Russell I am not, but I tend to think the TLP has problems, and those problems seem at least somewhat related to the realism vs. nominalism type of issue. Nominalism seems predicated on a posteriori views of knowledge, whether empirical or what is taken to be axiomatic and rational. Nominalism may not equate to empiricism or to "synthetic" a posteriori, perhaps, but they are all first cousins. TLP appears nominal in regards to syntax, doesn't it? The meanings of names are the objects they refer to or something.

But then he shifts gears when speaking about tautology: logical forms, which are tautologous, are a priori. Deduction is a priori; truth tables, proofs, reductio ad absurdum are a priori. Das stimmt, mein Herr, but only after learning the forms, after aquiring a certain degree of knowledge--that may be "psychologistic" but nevertheless accurate. IS chess a priori too then? And in some sense logic is sort of close to chess.

Personally I think LW's ideas of tautology and truth functionality are not quite correct (tho' my objections are not well formulated); at the very least "its raining or it's not raining" is not really truth functional like T v F is--closer to degrees of rain. Yes, the LEM works, as does contradiction, in the logic game; but not in terms of physical science (which is not so game-like as logic or chess or algebra). That some system has rules doesn't necessarily mean it is a priori.

Anyway, a priori often means something like, we don't know shit about cognitive processes, or how knowledge is aquired, therefore a priori (or innate). Perhaps that seems trivial; it wasn't to the empiricists who denounced scholastics and metaphysicians. I may be incorrect, but I think most of the Vienna Circle (certainly Carnap, who quotes Hume quite regularly) were closer to "physics as ontology" than supportive of Fregean logicism based on "abstract entities"; yet LW a bit more Fregean (tho' the Carrollian BS of the TLP is quite a drawback, ich denke .......

Brommage said...

A bit off-topic, but I thought some interesting continuity. Michael Kremer, in his "The Purpose of Tractarian Nonsense" [Nous 35:1 (2001) p. 39-73], begins by quoting a rather polemical reader review of the TLP. This paper itself is a defense and elaboration of the 'resolute reading' of the Tractatus.

Duck said...

Thanks brommage. I haven't seen that article. I didn't know Nous was publishing such things.

bubba spice said...

its pretty darned neat to see folks thoughtfully consideering the tractarian witty...a lot of folks seem to think it is passe or only useful as a lead-in to the late witty...

a few things that i think really help...
1. remembering the initial asseritions... 'the world is all that is the case...the world consist of facts and not things'-
i think a whole slew of confusion comes from folks NOT taking this not so metaphysical stance seriously

2.taking rather seriously witty's claim that the tractatus is not only an ethical text, but that the important stuff is the stuff NOT written.

it is really pretty darned elegant...the length the dude goes to try to reamain consistent...

3. although it can be risky (as some folks point out) i think it very helpful to rely on the notebooks to help illuminate the tractatus...witty seems a little less concerned with formality and consistency. if you read the tractatus (as plenty of others have advocated) as being as firmly placed in the tradition of tolstoy as it is russell and frege, it helps soften the blow of the last half dozen pages of the book...somehow it minimizes that sense of ,'what the hell!?!?!? you melodramatic bastard! how dare you set me up for this and then try to avoid the self-referrential can of worms you have opened up all over yourself by saying...''hey, forget it,im just showing you stuff..this is all nonsense go through the process...pitch it aside.''

thanks for the interesting post!

Duck said...

Thanks bubba! I'll have to think about your #1 (i.e. that that prop in particular causes problems if misunderstood), but I can certainly agree with your #2. As for #3, I have a copy of the notebooks around somewhere, but it's been a while since I looked at it. I should dig it up. Naturally (as with Nietzsche) that brings in the whole Nachlass issue. And you are right to remind us of LW's interest in Tolstoy etc. Exercise for the reader: explain why (as I think Monk tells us) Tristram Shandy was Wittgenstein's favorite book.