9
Sullivan’s paper is characteristically rich and thought-provoking, and I cannot hope to do justice to it in these comments. But there is, as Sullivan says, a comparatively straightforward disagreement between us about where the later Wittgenstein stands in relation to transcendental idealism, and I hope at least to cast some light on why I still see this matter differently from how Sullivan sees it.
Prima facie Sullivan and I are diametrically opposed in our views of where the “two” Wittgensteins stand in relation to transcendental idealism. In the cartoon sketch that Sullivan gives, I see the early Wittgenstein as embracing transcendental idealism and the later Wittgenstein as enabling us to resist it, whereas for Sullivan it is the other way round. It looks as though we could not be further apart. But, of course, it is not as simple as that. For one thing, the cartoon sketch is, precisely, a cartoon sketch; and when it comes to providing a somewhat more realistic portrait, with all the obvious caveats about what it even is to attribute any kind of position to Wittgenstein on the strength of either of his two principal texts, then Sullivan and I look far closer to each other than the cartoon sketch suggests. Indeed there have been times in the past when the focus of our disagreement, at least as far as the early Wittgenstein is concerned, has been at the meta-level: on the question whether there is any ground-level disagreement at all. (And I am pleased to add parenthetically that, with respect to that meta-level disagreement, I now think that Sullivan was right and I was wrong—Sullivan having always maintained, correctly as I now see it, that we do disagree.1) But also, more importantly, as Sullivan emphasizes in his essay, we are very close even in according transcendental idealism such significance in the exegesis of Wittgenstein. Each of us, if pressed on the question of what the relation is between the Wittgenstein who is ultimately hostile to transcendental idealism and transcendental idealism itself, would concede that, while Wittgenstein would, at that stage in his thinking, repudiate transcendental idealism as nonsense, it is a kind of nonsense whose attractions his own work engenders, and to whose attractions he himself, at more than one level, is susceptible. This is a point of contact, I think, that makes the differences between Sullivan and me minimal compared with the differences between both of us and commentators who deny that there are any interesting questions about the relation between Wittgenstein and transcendental idealism. Still, the fact remains that Sullivan sees transcendental idealism in the later work where I do not, and I want to say a little about why I do not.
First, a couple of preliminaries. I need to say something about David Bell’s remarkable contribution to this debate. I have already conceded that I cannot hope to do justice to Sullivan’s essay in these comments; and I am afraid that his excellent discussion of Bell is one part of his essay about which I am going to say virtually nothing. But I do need to say something, partly because Sullivan finishes his essay by expressly putting a question to me about Bell, and I owe him an answer; but also because my answer is in any case importantly relevant to what I shall go on to say. Sullivan’s question is about a suggestion I have made, namely that there is a deep affinity between some of Bell’s ideas and some ideas of my own.2 In particular, Sullivan wants to know whether I intend this suggestion to extend to the issues on which he (Sullivan) focuses in his own discussion of Bell. Concerning what I have called ‘the Feeling of Unity,’ do I take such unity to be, in Sullivan’s terms, a pre-condition of combination through concepts, or do I take it to be the combination itself? It looks as though this is an awkward question for me. This is because it looks as though the former alternative is what is required for me to stay in line with Bell, while the latter alternative is what is required for me to keep transcendental idealism at bay—in the way in which I think it should be kept at bay and, significantly in the current context, in the way in which I take the later Wittgenstein to keep it at bay.
My answer is the latter. That is, I take the unity in question itself to be combination through concepts. And I am consequently less confident than I was that there is the convergence in Bell’s ideas and mine which I earlier suggested there was. I now think that I may have got Bell importantly wrong. If so, then not for the first time—and I am sure not for the last time either—so far from my putting Sullivan right (see p. 184),3 I have reason to thank him for putting me right.
Now to the second preliminary. This is a much less weighty matter which I raise almost in a spirit of frivolity, but I raise it because here too there is something that connects with what I shall go on to say. Early in his essay Sullivan quotes that familiar section from Philosophical Investigations—§3814—where Wittgenstein raises the question, “How do I know that this colour is red?” and replies, “It would be an answer to say that I have learnt English.” Sullivan quotes this passage disapprovingly. Wittgenstein’s reply misfires, he suggests; for it alludes, in the face of a request for a justification, “to all that could be found in the place where further justification might be sought” (p. 173). Yet later, when discussing a case of basic rule-following, in which the rule is:
and the premise is:
PREMISE: a is red,
Sullivan says that PREMISE cannot be understood, let alone judged to be true (and therefore, we may conclude, let alone known to be true), independently of RULE. What does this amount to? It seems to amount to the claim that there can be no judging (or knowing) that something is red without knowing that the English word “red” is predicable of such things—a clear echo of what Wittgenstein says at §381. So has Sullivan committed the very gaffe that he claims to find in Wittgenstein?
No. I am teasing. Sullivan makes clear in a footnote5 that the quotation marks in RULE are not important. RULE is not meant to express a contingency about the English language, and cannot be assimilated to Wittgenstein’s own explicit reference to English (or rather, of course, in the original, to German—though, incidentally, the fact that a good translation of the original should involve a reference to English, rather than to German, is not without significance, as we shall see). But this reminds us that there is no simple rule linking the use of quotation marks to the use of some linguistic item to draw attention to that very linguistic item. Just as it is possible to put quotation marks round an occurrence of the word “red,” and still intend to be drawing attention to redness rather than to the word itself, so too it is possible to use the word “red” without quotation marks and intend to be drawing attention to the word rather than to redness. And that, surely, is Wittgenstein’s point, or at least part of his point.
To explain. The question “How do I know that this colour is red?” strikes no obvious chords when taken out of context. We can of course imagine its being asked in contexts where it would be a perfectly reasonable, unmysterious thing to ask. But we can also imagine its being asked in contexts, in particular philosophical contexts, where it would continue to strike no obvious chords. Now: it would be an answer to say, “I have learnt English.” In other words: one way of taking the question would be as equivalent to “How do I know that this colour is called ‘red’?” Quite likely this would not be what was intended, especially if the context really were a philosophical one. But then replying in that way would just be an invitation to say more about what was intended. It is a familiar rhetorical device, and I dare say it is one that you have used yourself.
We are reminded of something that is in any case obvious, namely that whether an answer to a question is a sensible answer depends on what exactly is being asked. Obvious or not, this provides a good cue for me at last to explore my principal disagreement with Sullivan.
Let me first say something about how Sullivan sees transcendental idealism as arising in the later work. Think once again about rule-following. Sometimes, when I am following a rule, I can justify what I do by articulating the rule. In such a case somebody else may seek further such justification—for instance, because the justification that I have proffered shows me to be following some subsidiary rule which is itself relatively complex and whose application to this case is unclear or unstraightforward. Eventually, however, their repeated requests for further such justification must issue in a blank. “Then,” Wittgenstein’s suggestion seems to be, “I am inclined to say: ‘This is simply what I do’” (§217). And we can see why this gives Sullivan pause. There seemed to be a chain of justifications, each of which issued in a request for a further justification of the same kind, until eventually no further justification of that kind was available and the question, “What justifies you in doing that?” was met instead with this appeal to “what I do.” The questions hitherto had been of a piece and the answers had been of a piece. They had been questions and answers about reasons, and, to that extent, about limits. But when an answer of that kind was no longer available there seemed, instead, to be an appeal to some brute contingency admitting of alternatives. That is to say, there seemed to be an appeal to a limitation. But what was needed, surely, to show that justification of the relevant kind had given out, was not this violent shift of direction but something more or less Platonic, something still expressing limits rather than limitations, but in such a way as to deflect the question.—“What justifies you in doing that?” “Oh, well, that’s just what the rule requires.”
Wittgenstein does seem to be guilty of some fundamental error, then. He does seem to suggest that we need a shift of direction here; that we need to appeal to limitations rather than to limits; that we need, in yet another metaphor, to dig beneath bedrock and say, as Sullivan nicely puts it, what would occupy the place of a further justification if only there were such a place. And, in suggesting these things, Wittgenstein does seem to cast the limits in question, which depend on the limitations in question, as themselves limitations. In sum, he seems to embrace transcendental idealism.
Here is another way of approaching the same matter. Wittgenstein asks at the beginning of §217, “How am I able to obey a rule?” and immediately registers that this question can be taken in two ways, as a question about causes and as a question about justifications. He further intimates that he is concerned with the question as taken in the second way. But the appeal to “what I do” seems to be more relevant to the question as taken in the first way, which is why its almost immediate appearance suggests the very shift of direction of which Sullivan complains. Crudely speaking, it is as if Wittgenstein fails to keep causes and justifications separate.
But can this be? Wittgenstein is surely the last person to be guilty of any such conflation. Again and again he adverts to this distinction in its various different guises and to its importance (§§189 and 460 are both worth consulting in this connection). And indeed one of the views that he is especially keen to undermine—not Platonism, with which, on many construals, he arguably has no quarrel, but rather what Sullivan helpfully calls “hyper-Platonism”—is a view that rests on this very conflation. What is misguided in hyper-Platonism is not, say, the idea that there are rails to infinity, which, as Sullivan reminds us, can be understood in an entirely innocuous way, but rather how that idea is construed, so that, for instance, these rails constrain us in the same way as physical rails. That is a confusion of justifications with causes.
But I see no such confusion in Wittgenstein. Nor do I see any casting of limits as limitations. We earlier considered a chain of questions that can arise when I am following a rule: “What justifies you in doing that?”; “What justifies you in doing that?”; and so forth. And it looked as though Wittgenstein’s appeal to “what I do” was introduced in response to the last question in this chain, when no further justification of the relevant kind was available—as a way of deflecting the question, if not of answering it. But it also looked as though his appeal to “what I do” was meant to be logically of a piece with an answer, providing at least a “quasi-justification” of the relevant kind; indicating what the justification would be if only there were such a thing. At any rate, I think this is how Sullivan sees things. But it is not how I see them. That seems to me to be a misreading of Wittgenstein.
How, then, do I read him? I think that Wittgenstein’s concern is with what it takes to follow a rule; with what the logical conditions of rule-following are. The kind of question to which “This is simply what I do” is pertinent is not “What justifies you in doing that?” but “What is involved in your doing that?” or “How do you know what to do?” or “How are you able to obey a rule?” (see §§211 and 217). At a high enough level of complexity I can answer such questions by citing my justifications. But this is not because I am being asked what justifies me in doing what I do. It is because my having those justifications is, at that level, how I know what to do; is how I am able to obey the rule. (This is a “grammatical” point, not an empirical point.) Eventually, however, my reasons give out, and then I am inclined to say, harmlessly enough in this context, “This is simply what I do.”
Admittedly, the metaphor of blindness, which Wittgenstein also invokes here, is not entirely happy. For indeed I might equally have been inclined to say, once I had reached bedrock, “I can now just see what to do.” Sullivan is absolutely right to draw attention to the limitations (not the limits!) of this metaphor. Nevertheless, I take it that part of the force of the metaphor is to emphasize that, at this basic level, I do not need to exercise any choice; I do not need to reflect on alternatives; I do not need to reckon with any justification.6 Following a rule is, as Wittgenstein observes, analogous to obeying an order (§206). And sometimes my obeying an order is blind in the sense that I do what I am told just because I am told to do it. But in another sense, of course, I can see perfectly well what I am doing. So too with basic rule-following. Sullivan is quite right to dismiss as perverse the suggestion that I need have nothing in mind when I am following a rule. But Wittgenstein himself makes a very similar point in Remarks on the Foundations of Mathematics.7 He there invokes a different metaphor, the metaphor of a mechanism (a metaphor which Bell, interestingly, rejects: see the passage cited by Sullivan on p. 175). Wittgenstein says, “One follows the rule mechanically … ‘Mechanical’—that means: without thinking. But entirely without thinking? Without reflecting” (Pt VII, §60, his emphasis).
But surely transcendental idealism still beckons here? Surely limits are being made to depend on limitations?
Not at all. Wittgenstein’s question is a Kantian question, in as much as it is a question about how something is possible. But it is also a constitutive question. That I do what I do, and that other people do the same, is a pre-condition for there being this or that practice among us, and hence for our having this or that rule. It is not a sub-bedrock quasi-justification for anything we do. It is part of the framework that makes justification for various things we do so much as possible. It is not limits that are being made to rest on limitations, then. It is our grasp of limits. In particular, it is which limits we grasp. If we had been different in various specifiable ways, we would have had different rules. But that is not to say, what would indeed be absurd to say, that the rules we actually have would themselves have been different. The limits are still limits. Wittgenstein’s question is a Kantian question; but his answer is not a Kantian answer.
Notes
1 See my “Was the Author of the Tractatus a Transcendental Idealist?,” in Michael Potter and Peter M. Sullivan, eds., The Tractatus and Its History, Oxford, Oxford University Press, forthcoming.
2 See my “Is the Feeling of Unity that Kant Identifies in His Third Critique a Type of Inexpressible Knowledge?,” Philosophy, 2007, vol. 82.
3 All unaccompanied page references are to Sullivan’s essay in this volume.
4 All unaccompanied section references are to Part I of Ludwig Wittgenstein, Philosophical Investigations, trans. G.E.M. Anscombe, revised edn, Oxford, Basil Blackwell, 1974.
5 See note 33 of his essay above (p. 186).
6 This is a welcome opportunity for me to correct a contrary impression that I give about this metaphor in “Is the Feeling of Unity that Kant Identifies in His Third Critique a Type of Inexpressible Knowledge?,” pp. 476–477. There I characterize the point of the metaphor as follows: “there is, grounding the exercise of concepts involved in making rule-governed objective judgements, something that does not involve the exercise of concepts at all.” But, for reasons that Sullivan makes clear, this is already to flirt with the idea that there is a kind of transcendental idealism in what Wittgenstein is saying.
7 Ludwig Wittgenstein, Remarks on the Foundations of Mathematics, ed. G.H. von Wright, R. Rhees and G.E.M. Anscombe and trans. G.E.M. Anscombe, revised edn, Oxford, Basil Blackwell, 1978.