Misinterpreting Kripke's Wittgenstein: A Reply to Byrne

I. Introduction
In his "On Misinterpreting Kripke's Wittgenstein", Alex Byrne claims that Saul Kripke's Wittgenstein on Rules and Private Language " has . . . been widely misinterpreted."¹ According to Byrne, the misinterpretation is reading Kripke's Wittgenstein (hereafter, KW) as being a nonfactualist about meaning attributions (i.e., statements of the form, "Jones means such-and-so by 'X'"). As Byrne sees things, this view "cannot be right" (B, p. 341). I will show that Byrne's view rests on several misinterpretations of Kripke's text. I conclude not only that KW can be a nonfactualist (and so vindicate the prevailing interpretation), but that this is also the most textually satisfying interpretation.

II. The Alleged Misinterpretation
In section I of his paper, Byrne notes that KW considers and rejects a variety of candidates for a fact showing that I mean addition by 'plus' [rather than quaddition by 'plus'], drawing the interim conclusion that "there can be no such thing as meaning anything by any word" (K, p. 55).² Byrne says that KW could take this result "at face value, and adopt an "error theory" with respect to meaning-facts: all attributions of meaning are false, and that's that." (B, p. 339). But, says Byrne, KW does not take this option, raising the question: "What does he do instead?" (B, p. 339).³ Byrne claims that on "the prevailing interpretation, [KW] says that, contrary to appearances, attributions of meaning are not factual claims at all. They look like statements of alleged fact, being declarative sentences, but really they have some entirely different role." (B, p. 339). Somewhat more specifically, the prevailing interpretation is alleged to strap KW with the view that meaning attributions "have no truth conditions" (B, p. 340). Although Byrne allows that a nonfactualist KW is a "natural reading" of much of what Kripke says, nonetheless, "it cannot be right" (B, p. 340), for "KW cannot possibly be a non-factualist" (B, p.341).

III. Byrne's Conception of Nonfactualism
Now, as noted above, Byrne's initial characterization of a nonfactualist is someone who holds:
(a) meaning attributions (or, other kinds of sentences) "are not factual claims at all, but have "some entirely different role",
or
(b) meaning attributions have no truth conditions. (Hereafter, 'meaning attributions' will be abbreviated to MAs).
However, because of Kripke's sceptical solution, both (a) and (b) are threatened by ambiguity. For example, if 'truth conditions' is short for 'TLP-style truth conditions' then KW must be a nonfactualist since he clearly does reject the Tractatus' truth conditional model of meaning. In a personal correspondence, Byrne has informed me that he agrees that KW is a TLP nonfactualist, so this is not, for us, a controversial claim.⁴ Since he agrees that KW is a TLP nonfactualist, it's clear that Byrne is assuming a different notion of truth conditions and facts than is found in TLP, and so alleging that KW cannot possibly be someone who denies that MAs have these sorts of truth conditions or state these kinds of facts. Clearly, we must get beyond (a) and (b) above, if we are to get to the bottom of Byrne's account.

Nicely enough, Byrne takes us beyond (a) and (b) in his argument to show that the nonfactualist KW of the prevailing interpretation "cannot be right". Byrne's case against reading KW as a nonfactualist relies (almost entirely) on a passage from p. 86 of Kripke's book. There Kripke tells us that since Wittgenstein accepts a redundancy theory of truth, his sceptical solution is compatible with our saying things like, "It's true that Jones means plus by '+'" and "It's a fact that Jones means plus by '+'". Byrne takes this passage as showing that KW "accepts a deflationary account of truth aptness (intimately conjoined with a deflationary account of truth). . . " (B, p. 341). Byrne contends that this is "incompatible with non-factualism" (B, p. 341) and concludes that "the "non-factualist" interpretation [has Kripke ignoring] a fairly obvious objection to the internal coherence of Wittgenstein's position, as he construes it." (B, p. 341). Byrne also says that p. 86 shows that KW "holds that [a meaning attribution] does (or could) state a fact, in the perfectly ordinary sense of 'fact'" (B, pp. 341-2). According to Byrne, this shows that "Kripke's Wittgenstein cannot possibly be a non-factualist".

Summing up, Byrne's case against a nonfactualist KW is the following slight affair:
1. Nonfactualism is incompatible with deflationism about truth aptness and truth, both of which are espoused by KW;
2. No one who holds that meaning attributions state facts "in the perfectly ordinary sense of 'fact'" can possibly be a (meaning attribution) nonfactualist, and KW holds this view.

IV. A Verbal Dispute?
Though Byrne's claims help to clarify his notion of nonfactualism, it's worth noting that the notion of nonfactualism which Byrne is using presumably does not include TLP-nonfactualim. For as noted above, Byrne allows that KW is a TLP-nonfactualist (i.e., a rejecter of TLP-style truth conditions and TLP-style facts). That is, for Byrne, KW does claim that MAs do not have TLP-style truth conditions and that MAs do not state TLP-style facts. Presumably then, being a TLP-style nonfactualist is compatible with deflationism about truth aptness and truth and also does not preclude one from holding that MAs state facts "in the perfectly ordinary sense of fact". I will assume then that not all nonfactualisms whatsoever are incompatible with deflationism about truth aptness and truth.
It should be clear then that just because the prevailing interpretation's KW is said to be a nonfactualist, this is not enough to warrant the conclusion that he holds something that is incompatible with deflationism about truth aptness and truth. Frankly, I am not sure why Byrne supposes that the prevailing interpretation's account of KW's nonfactualism is incompatible with deflationism about truth aptness and truth. But that he does assume this is clear from his paper. By my lights, the only sort of nonfactualism that is obviously incompatible with deflationism about truth aptness and truth is one which denies that we can legitimately affirm MAs in the absence of "meaning facts", (i.e., facts of the sort denied by KW's sceptic).

LONG ASIDE; added 9/22/99: Questions about what sort of 'meaning-facts' are denied by the sceptic arise here, as well as questions about the need for such facts in order to legitimately affirm MAs. Everyone apparently agrees that KW at least denies the existence of TLP-style meaning facts, whatever those are. Presumably, everyone would also agree that KW denies that we can legitimately affirm MAs in the absence of TLP-style meaning-facts, SO LONG AS MAs ARE UNDERSTOOD a la TLP; again, whatever such an understanding involves. But notice that there appears to be room for KW to propose a non-TLP understanding of MAs that allows them to be legitimately affirmed in a world without TLP-style facts. The question then, is: Does KW's non-TLP understanding of MAs count as factualism or not? I would urge that the answer to this question depends not only on how one understands KW's understanding of MAs but also on what sort of facts one takes KW's sceptic, and KW, to deny. For KW is supposed to agree with his sceptic about the nonexistence of "meaning facts". By my lights, the key to understanding KW's position involves appreciating that his case against the existence of meaning facts is more of a case against finding facts that can justify one MA over another than it is a simple case of denying facts per se. As such, it's not so much that there are no facts to support some particular MA over all others. Nothing in the sceptic's story shows that there are no facts to support the MA, "Jones means plus by '+'", over the MA, "Jones means minus by '+'". Clearly there are such facts. Rather, KW and his sceptic claim that there are no facts to support asserting one particular MA over other MAs concocted by the sceptic. And this result is thought by the sceptic to mean that no MAs can be legitimately affirmed. And now the question is whether the demand for a unique justification, i.e., the claim that we can't legitimately affirm an MA so long as we cannot rule out legitimately affirmed other MAs in the same circumstances, is something to be found in a TLP understanding of MAs and if so, if it it is possible to have a viable conception of MAs which drops this demand for uniqueness?
It's clear that if one drops the requirement that MAs be given a UNIQUE justification by facts then the sceptic's case against meaning facts can be rendered compatible with the legitimate affirmation of MAs. This, I believe, is all that KW does, viz., drops the requirement that MAs have a UNIQUE justification in order to be legitimately asserted. If this is correct, then it's not clear that KW gives, or is required to give, a non-TLP interpretation of MAs, or even to offer an analysis or interpretation of MAs at all. It all depends on whether a TLP understanding of MAs demands facts to provide "unique justification".
This leaves several question. First, how or why does KW's rejection of the requirement that the legitimate assertion of MAs requires a UNIQUE justification by facts constitute a rejection of a TLP understanding of MAs? That is, does a TLP understanding of MAs commit one to the uniqueness requirement? It's not clear to me that it does. But even if it does, we can still ask whether it is possible to reject the uniqueness requirement and yet provide a viable account of meaning or of MAs?
END OF LONG ASIDE.

Perhaps the salient feature of the prevailing interpretation's KW (and KW himself), is his contention that MAs can be legitimately asserted in the absence of "meaning facts". Indeed, the raison d'être of KW's sceptical solution is to show us how we are able to legitimately affirm MAs while yet accepting the sceptic's denial of meaning facts. In short, whatever else KW's account comes to, it is supposed to be compatible with the legitimate assertion of MAs. This cannot be denied. According to Kripke, the compatibility is made possible by a shift from a truth conditional (TC) account of language to an assertion conditions (AC) account. (Of course, we can have our doubts about the success of KW's claim that his sceptical solution is a plausible or successful account of what is involved in the legitimate assertion of MAs. Most importantly, we can have our doubts that KW's account of the assertability conditions shows that ACs allow us to successfully affirm MAs whereas TCs do not).

Now, according to Kripke, Wittgenstein's deflationism about truth aptness and truth means that, 'It's a fact that Jones means plus by '+'"', says nothing more nor less than "Jones means plus by '+'". Thus, anyone that allows the legitimacy of saying the latter also allows the legitimacy of saying the former, and vice versa. For the phrase, 'It's a fact that', is simply redundant. Most importantly, deflationism about truth aptness and truth allows KW to hold that insofar as, "Jones means plus by '+'", can be legitimately asserted despite the nonexistence of facts, so too can the result of prefacing it by, 'It's a fact that'. As such, the nonfactualism attributed to KW by the prevailing interpretation is, by Kripke's lights anyway, compatible with deflationism about truth aptness and truth. Indeed, Kripke himself is arguing that since the legitimate assertion of MAs does not, for KW, require or commit us to the existence of meaning facts, neither does asserting the result of prefacing MAs with 'It's a fact that', courtesy of KW's deflationism about truth aptness and truth. If there is a problem with Kripke's argument then Byrne should attack Kripke, not the prevailing interpreters, for they have read KW correctly. KW is claiming that since MAs can be legitimately asserted despite the nonexistence of meaning facts, the same goes for statements like, "It's a fact that Jones means plus by '+'", since, courtesy of deflationism, 'It's a fact that' is strictly redundant.

The bottom line here is that if Byrne wishes to say that KW cannot be a nonfactualist simply because KW is a deflationist about truth aptness and truth, his complaint against the prevailing interpreters may be merely verbal. That is, Byrne and the prevailing interpreters may mean two different things by "nonfactualism". In particular, Bryne sees nonfactualism as something incompatible with being a deflationist about truth and truth aptness whereas the prevailing interpretation does not. The crucial issue then is whether Byrne regards KW as denying the existence of meaning facts, because it seems to be the only nonverbal issue they disagree about. For it is clear that Byrne and the prevailing interpreters agree that KW allows us to legitimately assert MAs, as well as the result of prefacing them with, 'It's a fact that'.

Similar remarks apply to claim 2, i.e., only the most severe of nonfactualisms is incompatible with it. Recall that Byrne's sole justification for claiming that KW holds that MAs state facts in the perfectly ordinary sense of 'fact' is the passage from p. 86 of Kripke's book, i.e., the passage where Kripke reminds us about Wittgenstein's deflationism about truth. As such, Byrne is identifying:
(i) KW's allowing MAs to be prefaced by redundancy uses of 'It's a fact that',
with
(ii) KW holding that MAs state facts in the perfectly ordinary sense of 'fact'.
Though I have serious doubts about the legitimacy of this identification (both as an independent claim and as a proper interpretation of KW) I will not challenge it here since I am not prepared to say anything substantive about our ordinary use of "It's a fact that Jones means plus by '+'".⁵ However, acceptance of the identification of (i) and (ii) still allows the prevailing interpretation's KW to be seen as holding that MAs state facts in the perfectly ordinary sense of 'fact'. This is because, once again, the only sort of nonfactualism that must deny that MAs state facts in Byrne's perfectly ordinary sense of 'fact' (assuming, again, that this is to be identified with allowing MAs to be prefaced by redundancy uses of 'It's a fact that') is one that denies that we can legitimately affirm MAs. But we have already seen that this severe and extreme nonfactualism is not the sort attributed to KW by the prevailing interpretation. For their KW does allow us to legitimately affirm MAs.⁶

V. A Substantive Disagreement between Byrne and the Prevailing Interpretation
At this point, doubts can be raised about there being a real dispute between Byrne and the prevailing interpretation. While it is true that KW can be called a "factualist" insofar as he allows us to legitimately assert, "It's a fact that Jones means plus by '+'", and since it's obvious that KW being a deflationist about truth aptness and truth means that he does allow that it's legitimate to assert statements like, "It's a fact that Jones means plus by '+'", it's clear that KW can be called a factualist. And insofar as KW cannot be both a factualist and a nonfactualist (unless, horror of horrors, he's inconsistent), the fact that he is a factualist seems to warrant claiming, as Byrne puts it, that "KW cannot be a nonfactualist". The problem with this little argument is that what is here labeled 'factualism' is not obviously incompatible with the nonfactualism of the prevailing interpretation. For the prevailing interpretation takes KW as holding that the legitimate assertion of statements like, "It's a fact that Jones means plus by '+'" , by no means requires or states the existence of "meaning facts". Indeed, the raison d'être of KW's sceptical solution is to disabuse us of the idea that legitimate assertions involve, require, or state facts. The bottom line then is that there is no obvious incompatibility between someone who holds that we can legitimately assert, "It's a fact that Jones means plus by '+'" and someone who denies the existence of meaning facts.
Admittedly, it sounds odd (and then some) to say that KW is both factualist and nonfactualist, but once it is made clear what it says, viz., KW allows us to preface MAs with redundancy uses of, "It's a fact that", while also denying that there are any facts for MAs to state facts (and, a fortiori, MAs don't state facts) there should be no difficulty in accepting it.

In particular, the prevailing interpretation takes KW to be nonfactualist because, given his apparent denial of "meaning facts", there seem to be no appropriate facts for MAs to state (nor any substantive, i.e., non-redundant, truth conditions for them either. It is also worth noting that p. 86 of Kripke's book says nothing at all about whether there are any such facts or truth conditions, nor whether they are what is "stated by" MAs.⁷). Indeed, the whole point of the passage on p. 86 seems to be at odds with Byrne's reading, for it seems designed to show us how KW's nonfactualism (i.e., his denial of meaning facts and so his denial that MAs state such nonexistent facts) can be made compatible with the acceptance of ordinary statements like, "It's a fact that Jones means plus by '+'", and the like. According to Kripke, Wittgenstein's redundancy view has him holding that asserting "It's a fact that Jones means plus by '+'", asserts no more nor less than "Jones means plus by '+'. Since the latter can be legitimately asserted despite the nonexistence of meaning facts, so too can the former, according to Kripke, for they come to the same thing, courtesy of Wittgenstein's redundancy theory. In other words, KW's assertion condition account of MAs allegedly shows us how we are able to legitimately assert MAs in a world without meaning facts. A fortiori, the assertion condition account shows that MAs are both legitimately assertable and not fact-stating, for there are no facts for them to state. Since MAs don't state facts neither does the result of prefacing them with "It's a fact that". ⁸

Byrne himself seems alive to the problems just laid out. After telling us that the passage from p. 86 of Kripke's book "makes clear" that KW holds that MAs state facts, "in the perfectly ordinary sense of 'fact'", Byrne says:

So what on earth is going on? Why does Kripke apparently flatly contradict himself by saying that "Wittgenstein holds, with the sceptic, that there is no fact as to whether I mean plus or quus". ([K,] pp. 70-1).⁹

What's going on, I contend, is that Byrne misses a perfect opportunity to correct his misreading of p. 86 of Kripke's book. The just quoted passage makes clear that KW's denying the existence of a fact as to whether someone means plus or quus is incompatible, by Byrne's own lights, with the way Byrne reads p. 86.¹⁰ Now insofar as there really is a substantive factualism expressed on p. 86 (i.e., a factualism that is genuinely at odds with the nonfactualism of the prevailing interpretation), I agree with Byrne that p. 86 is incompatible with KW holding with his sceptic that there is no fact as to whether I mean plus or quus. For as suggested above, KW's apparent denial of such a fact is perhaps THE reason for the prevailing interpretation's nonfactualist reading of KW.

But given the explicitness of the passage from pp. 70-1, along with my previous remarks concerning the compatibility of p. 86 with the nonfactualism of the prevailing interpretation, the proper conclusion seems to be that p. 86 does not express a substantive factualism. Byrne, of course, disagrees. Since he is convinced of the correctness of his reading of p. 86, Byrne sets himself the task of showing why the passage from pp. 70-1 cannot be taken at face-value, i.e., why it does not, its explicitness notwithstanding, express the view of KW. The remainder of my paper will be an examination and critique of Byrne's attempt to explain away the passage from pp. 70-1. I will show that Byrne's attempt to do this rests on a misreading of Kripke's text and so is unsuccessful. I conclude that the best way to read KW is to accept the passage from pp. 70-1 "as is". If this is right, then, contra Byrne, we ought to see p. 86 as expressing a weak, trivial sort of factualism, which is compatible with the nonfactualism of the prevailing interpretation.

VI. Byrne's argument deflated
Before examining the details of Byrne's account, it should be noted that Kripke's claim on pp. 70-1 appears too explicit to be a slip of the tongue or fingers, nor does it seem to contain any sort of hedge or ambiguity. In short, since Kripke, at pp. 70-1, has pretty explicitly told us what Wittgenstein holds, it's unclear how we can avoid taking this as being one of the views held by the entity known as Kripke's Wittgenstein. If Kripke says, "Wittgenstein holds that p" then surely we can and ought to say, "KW holds that p".

Also, as noted above, the claim on pp. 70-1 is hardly an isolated passage. Similar passages are present throughout Kripke's book (see pp. 38-9 and p. 86 especially). Thus, if anything is certain about KW's view it would seem to be, first, that his sceptic denies the existence of any fact as to whether someone mean plus or quus, and second, that KW agrees with his sceptic that there is no such fact. (Recall that this agreement is the starting point of KW's sceptical solution). At the least then, we should be sceptical about Byrne's contention that Kripke's claim at pp. 70-1 ought not be accepted "as is". Now, on to Byrne's account.

Byrne's case for why we shouldn't take Kripke's claim from pp. 70-1 at face value begins with the following claim (B, p. 342):

The key to interpreting [Kripke's claim that Wittgenstein holds, with
the sceptic, that there is no fact as to whether I mean plus or quus]
lies a few pages earlier, where it is apparent that [KW] is only concerned
to deny the existence of a "'superlative fact' . . . about my mind that
constitutes my meaning addition by 'plus'." But the lack of a superlative
fact certainly does not mean, according to [KW], the lack of any fact
at all. That is a "philosophical misconstrual" (p. 65).¹¹

Before going further, it should be noted that Byrne's last two lines betray an illegitimate reading of p. 65 of Kripke's book. On p. 65, Kripke is discussing his Wittgenstein's sceptical solution. The solution begins, we are told, with Wittgenstein "agreeing with the sceptics [sic] that there is no 'superlative fact' . . . about my mind that constitutes my meaning addition by 'plus' and determines in advance what I should do to accord with this meaning" (K, p. 65). Kripke then says, in the name of his Wittgenstein, that "the appearance that our ordinary concept of meaning demands such a fact is based on a philosophical misconstrual . . . of such ordinary expressions as 'he meant such-and-such'', 'the steps are determined by the formula', and the like." (K, pp. 65-6). It's clear then that Kripke is telling us that his Wittgenstein believes that there has been a philosophical misconstrual of certain expressions, which in turn leads us to suppose, mistakenly, that without a superlative fact, meaning is impossible. In short, our mistake, according to KW, is supposing that the lack of a superlative fact means a "lack of meaning".

Now it is surely one thing to say that a philosophical error leads us to suppose, incorrectly, that our ordinary concept of meaning demands a superlative fact, and quite another to say that it's a philosophical error to regard the lack of a superlative fact as meaning the lack of any fact at all.¹² Since it is obvious that KW can hold that the lack of a superlative fact does not mean a "lack of meaning" without taking any sort of stand on whether the nonexistence of superlative facts means or does not mean the lack of "any fact at all", it's clear that there is nothing on p. 65 to warrant Byrne's claim that, for KW, it's a "philosophical misconstrual" to take the lack of a superlative fact as meaning the lack of any fact at all. In spite of the lack of textual warrant, Byrne does, of course, strap KW with the view that it's a mistake to suppose that the lack of a superlative fact means the lack of "any fact at all". The reason he does so is because he assumes, question-beggingly, that KW is a substantive factualist, i.e., someone who is committed to the existence of (some sort of) "meaning facts", which facts are "stated by" MAs.

If KW holds that the lack of superlative facts means the lack of "any fact at all", then, given his denial of superlative facts, he cannot be a substantive factualist. For a substantive factualist cannot get by without facts. To put it another way, a substantive factualist KW, who holds that there are no superlative facts, cannot also hold that the lack of superlative facts means the lack of "any fact at all". For this would force KW to deny the existence of facts for MAs to state. Of course, the problem with concluding from all of this that KW does not believe that the lack of superlative fact means the lack of "any fact at all", is that such a conclusion rests on the assumption that KW is a substantive factualist. But recall that Byrne is in the middle of trying to show why the passage from pp. 70-1 is not to be taken at face-value, which he must do in order to defend his claim that KW is a substantive factualist, (and so cannot be a nonfactualist)! Clearly then, Byrne's interpretation of p. 65 cannot assume that KW is a substantive factualist without begging the question. Contra Byrne then, KW does not, on p. 65, tell us that it's a philosophical error to suppose that the lack of a superlative fact means the lack of any fact at all.

The lack of textual warrant for Byrne's claim that KW believes that the lack of a superlative fact does not mean the lack of any fact at all, seriously damages his case against taking pp. 70-1 of Kripke at face value, which continues as follows (B, p. 342):

Now Kripke appears to take issue with the irenic view --attributed by him to Wittgenstein-- that our ordinary notion of meaning does not embody a commitment to "superlative facts." Such claims of the irrelevance of philosophical scepticism to the beliefs of the "ordinary man," Kripke finds "almost invariably suspect" (p. 65; see also p. 66).

So it would appear to be Kripke's position that if the "sceptical argument" is sound, and hence there is no superlative fact, then there is no fact simpliciter. That is why Kripke (misleadingly) says that "Wittgenstein holds, with the sceptic, that there is no fact as to whether I mean plus or quus." More carefully, this is a thesis that Kripke believes that Wittgenstein (given the soundness of the sceptical argument) ought to hold. But as Kripke makes perfectly plain in the passage from p. 86 . . ., he believes that Wittgenstein does not hold it.

According to Byrne, Kripke's disagreement with KW about whether our ordinary conception of meaning requires superlative facts shows that Kripke believes that if there are no superlative facts then there are "no facts at all". And because Kripke believes this, he sees the sceptical argument as establishing not only the nonexistence of superlative facts but also the nonexistence of "any fact at all" as to whether someone means plus or quus. So, Kripke believes that KW is (or ought to be) committed to denying the existence of any fact as to whether someone means plus or quus. Thus, Kripke straps his Wittgenstein with this view on pp. 70-1. But, according to Byrne, Kripke's claim is "misleading" (one would think false would be more appropriate) presumably because Kripke arrived at it by leaning on a principle which KW himself rejects, viz., that the lack of a superlative fact means the lack of any fact at all.

There are several problems with Byrne's argument here but the primary difficulty is that my remarks above undermine Byrne's assumption that Kripke leans on a principle rejected by KW. For we saw above that there simply is no textual warrant for taking KW to believe that the lack of a superlative fact does not mean the lack of any fact at all. If this is right then Byrne's contention that Kripke's claim on pp. 70-1 misleads readers is without any support at all. For Kripke could very well be expressing KW's views in assuming that the lack of a superlative fact does mean the lack of any fact at all, and so rightly saying that Wittgenstein holds with the sceptic that there is no fact as to whether someone means plus or quus.

Another problem is Byrne's final suggestion that the passage from p. 86 of Kripke's text shows that Kripke believes that Wittgenstein does not hold that there is no fact as to whether someone means plus or quus. This means that, according to Byrne, although Kripke knows that KW doesn't hold that there is no fact as to whether someone means plus or quus, nonetheless he's going to say he holds it anyway! I find this suggestion too remarkable to be acceptable.¹³

But there is a bigger problem still with Byrne's final suggestion, viz., the appeal to p. 86 begs the question. For Byrne, taking Kripke's claim on pp. 70-1 at face-value is inconsistent with his (i.e., Byrne's) account of p. 86. So, Byrne appeals to p. 65 in an attempt to convince us that the passage from pp. 70-1 does not really express the views of KW, but only Kripke himself. But Byrne's final line above tells us that we can be sure that KW doesn't hold the view attributed to him by Kripke on pp. 70-1, because the passage from p. 86 makes it "perfectly plain" that he doesn't!

I conclude that there is no textual warrant for not taking Kripke's claim from pp. 70-1, "as is". If this is right then KW as a nonfactualist and Byrne is wrong that p. 86 shows KW to be a substantive factualist. By my lights, the point of p. 86 is to show how KW can hold the views that Byrne contends cannot be rendered consistent, viz., allowing us to say, "It's a fact that Jones means plus by '+'", while at the same time agreeing with his sceptic about the nonexistence of a fact as to whether someone means plus or quus. These views can be reconciled, says Kripke, by appreciating Wittgenstein's deflationary views about 'true' and 'fact'. As such, to say, "It's a fact that Jones means plus by '+'" is to say, "Jones means plus by '+'". But as long as we have an assertion conditions account of meaning (which account turns its back on facts in favor of conditions that legitimate the assertion of a sentence; see K, pp. 77-8), we can legitimately say the latter, and so the former, in the absence of any facts about whether someone means plus or quus.

The compatibility of p. 86 with pp. 70-1 can also be seen as follows. KW holds that our ordinary conception of meaning does not require a superlative fact (K, p. 65). This allows him to agree with his sceptic that there are no superlative facts and still maintain that MAs can be legitimately asserted. Similarly, a nonfactualist KW could maintain our right to affirm MAs, despite the nonexistence of any fact as to whether someone means plus or quus, by holding that our ordinary conception of meaning can get by without the existence of any such fact. Such, I contend, is pretty much what KW does.

'Superlative fact' is really just Kripke's shorthand expression for any fact that can answer the sceptical challenge (see K, pp. 9 ff.), which means, among other things, any fact that shows whether someone means plus or quus. If this is right, then every fact is at least potentially a superlative fact, i.e., every fact gets its chance to satisfy the sceptic. Most facts never really get to the plate of course, because they are simply irrelevant to, and so obviously unable to answer, the sceptical challenge. The blue sky, the twining ivy, Darwin's theory of evolution (for noncreationists only, of course), etc., are all of them unsuccessful candidates for a fact as to whether someone means plus or quus. Other facts, e.g., the fact that someone says, "2 + 2 = 4, 2 + 3 = 5, . . ., 34 + 16 =50, 34 + 17 = 51 . . . , etc.", or the fact that someone soberly and sincerely alleges to be plussing, are relevant to, but no more successful at, answering the sceptic than the blue sky and the twining ivy. Still other "facts", i.e., facts whose factual status is in question, e.g., infinite tables in my head, a Platonic form of plus in Plato's heaven, ineffable flashes of insight, etc., are also unsuccessful candidates for a fact as to whether I mean plus or quus. In short, no fact answers the sceptical challenge, i.e., there is no fact that shows whether someone means plus or quus.

It should be obvious that the sceptic's case against there being a superlative fact does not threaten the existence of any of the facts mentioned above. It leaves any "existent fact" standing, albeit exposed as a nonsuperlative fact. It's possible then that KW could take one or more of these existent facts to be stated by, or to correspond to the truth conditions for, MAs. Unfortunately for Byrne, I don't see any textual evidence for such a view. On the contrary, there appears to be evidence against such a reading. So far as the text is concerned, on KW's assertion conditions account of meaning, a meaning attribution like, "Jones means plus by '+'", does not state any fact, and certainly does not state the fact that Jones agrees with us in his responses to particular problems, nor the fact that Jones' sincerely and soberly contends that he is adding, etc. (See K, p. 111). It appears that, for KW, once the sceptic shows us that there is no fact as to whether someone means plus or quus, the only solution to meaningless MAs is to opt out of a fact-stating account of meaning.

To say that KW denies that MAs state facts (because there are no appropriate facts for them to state) is not to say that facts play no role in the meaning attribution game. For KW, facts seem to have a role in leading us to assert and accept the MAs we assert and accept. However, to say that certain facts play a role in the assertion of MAs, lead us to assert particular MAs rather than others, is no more incompatible with nonfactualism about MAs than is an emotivist admitting that certain facts (e.g., the fact that people are in pain) lead us to assert particular moral judgments (e.g., "Torture is wrong"). What matters, of course, is one's position on whether the facts that we have are what are stated by our MAs or our ethical judgments. On an assertion conditions account of meaning, to say, "Jones means plus by '+'", is best seen as a way of expressing our acceptance of Jones into our linguistic community (but it certainly doesn't state that Jones has been so accepted, any more than it states that Jones agrees with us in his responses to particular cases, etc.). Although certain facts about Jones (and ourselves, of course) lead us to accept him as an adder, they are not stated by the meaning attribution itself, or so KW appears to hold.

In conclusion, Byrne's claim that it's a mistake, according to KW, to take the lack of a superlative fact as meaning the lack of any fact at all, is without textual warrant. Kripke's notion of a superlative fact, viz., that it's a fact that can answer the sceptical challenge, means that the lack of a superlative fact does indeed mean the lack of any fact at all. Because of this, Byrne's claim that Kripke misleads readers on pp. 70-1 is also without warrant. We should, I contend, take the claim at face-value and so take KW to hold, with his sceptic, that there is no fact as to whether someone means plus or quus. In short, we should take KW to be a nonfactualist. Contra Byrne, none of this conflicts with Kripke's claims on p. 86, the point of which is to show how KW, despite his denying the existence of any "meaning facts", can still allow us to say, "It's a fact that Jones means plus by '+'". We can still say this because, on a redundancy account of 'fact', saying it comes to nothing more than saying, "Jones means plus by '+'". For KW, such a statement can be legitimately asserted in the absence of any meaning facts because of his assertion conditions account of meaning. That is, the fact that there are no "meaning facts" does not and cannot threaten our ordinary use of MAs, or so KW alleges, and Kripke appears to deny.¹⁴

---

Endnotes

¹ Alex Byrne, "On Misinterpreting Kripke's Wittgenstein" in Philosophy and Phenomenological Research Vol. LVI, No. 2, June 1996, p. 339. (Hereafter referred to by 'B'). The alleged misinterpreters include Colin McGinn, Wittgenstein on Meaning (Oxford: Blackwell, 1984); Crispin Wright, "Kripke's Account of the Argument Against Private Language," Journal of Philosophy, 81, 1984, pp. 759-78; G.P. Baker and P.M.S. Hacker, Scepticism, Rules and Language (Oxford: Blackwell, 1984); and Paul Boghossian, "The Rule-Following Considerations," Mind, 98, 1989, pp. 507-49. Also, I will use 'K' to refer to Saul Kripke's, Wittgenstein on Rules and Private Language (Cambridge: Harvard University Press, 1982) and 'TLP' to refer to Wittgenstein's Tractatus logico-philosophicus (London: Routledge & Kegan Paul, 1961).

² Notice the vagueness of Kripke's phrase, "there can be no such thing as . . .". In particular, it leaves unclear whether Kripke is saying that all meaning attributions are false or whether they are all nonsense or meaningless. Kripke doesn't seem to care much about the differences between falsity and nonsense.

³ Boghossian also alleges that KW doesn't give us an error theory of meaning attributions, reading this as the claim that they are false (op. cit., pp. 523-4). However, I do not accept his argument that an error theory of meaning discourse is of "doubtful coherence". The argument leans on the assumption that an error theory takes sentences to be false, as opposed to denying, e.g., that they say what people think they say. Since I allow that error theories (for some set of sentences) include the view that sentences don't say what people think they say (as opposed to being false), I regard Boghossian's argument as a failure.

⁴ He and I also agree that it's not really clear what is meant by TLP truth conditions or facts. But it's clear, we agree, that KW rejects them, whatever they are.

⁵ I am prepared to say that it's one thing to allow for an ordinary use of, "It's a fact that Jones means plus by '+'", and quite another to allow that, "It's a fact that Jones means plus by '+'" states a fact in the perfectly ordinary sense of 'fact'. Whereas the latter forces us to see MAs as stating "ordinary facts", the former commits us to stating MAs, not facts.

⁶ Perhaps Byrne fails to appreciate this because of the ambiguity of speaking of MAs. For we can speak of a Tractarian truth conditional understanding of MAs and an Investigations assertion conditions understanding of them. Taken truth conditionally, the prevailing interpretation does see KW as denying our right to affirm MAs. However, taken assertion conditions-wise, we can and do legitimately affirm them. A similar problem afflicts emotivism. On the one hand, emotivists allow the legitimacy of saying, "X is morally wrong", so long as it is taken as an expression of emotion. However, insofar as it is seen as making a statement about some moral fact or other, it is an illegitimate statement.

⁷ Another way of putting the same point may be the following, from Michael Dummett: ". . . the correct observation is that the redundancy theory precludes an explanation of the sense of a sentence by stating its truth conditions. The conditions we take as warranting the declarative utterance of a given form of sentence do not exhaust the use we make of it, but they form a central component of it, and must be included in any adequate description of that use, whereas the conditions under which, in general, a sentence of that form is true cannot be called features of its use at all, even if they could be characterized without circularity: they belong to a much higher theoretical level." ("Review Essay", in Journal of Philosophy, Vol. XCIV, Number 7, July, 97, p. 369).

⁸ It could be said that Byrne has at least pointed out something that is perhaps overlooked or ignored by the prevailing interpretation, viz., that there is a weak sort of factualism to be found in KW (we might call it, "deflationary factualism"). Perhaps, but we'll see below that Byrne definitely has a more significant quarry in his sights.

⁹ The passage from K, pp. 70-1 is by no means an isolated one. Throughout Kripke's book, there are similar passages. See especially pp. 38-9, as well as p. 86.

¹⁰ Two things are unclear here, first, whether the prevailing interpreters agree with Byrne about the incompatibility, and second, whether Byrne is right about the incompatibility. I don't believe he is right about the incompatibility (a belief I defend in the remainder of the paper) but if he is then KW has an inconsistent position. For KW clearly does hold that there is no fact as to whether someone means plus or quus. Of the prevailing interpreters, McGinn seems to be the only one who clearly agrees with Byrne that deflationism about 'true' and 'fact' are inconsistent with nonfactualism. Interestingly, Byrne fails to appreciate this and so accuses McGinn of misunderstanding the point of K, p. 86. On p. 71 of his Wittgenstein on Meaning, McGinn, after noting that Kripke's p. 86 expresses awareness of Wittgenstein deflationism about 'true' and 'fact', says, ". . . it cannot be that Wittgenstein really wishes to deny that semantic sentences have truth conditions -- on pain of denying that they express propositions." According to Byrne, McGinn is here claiming "that Kripke does not "draw the right lesson" (that is, the denial of non-factualism) from Wittgenstein's deflationism about truth." (B, p. 341). That is, Byrne contends that McGinn fails to appreciate that Kripke does draw the conclusion that KW cannot be a non-factualist. This is, however, a tendentious reading of McGinn's note. I take McGinn's point to be as follows: Since a redundancy view of truth is incompatible with nonfactualism, Wittgenstein cannot be KW since KW does "really wish to deny that semantic sentences have truth conditions [or state facts]". Thus, McGinn disagrees with Byrne about whether KW really wishes to deny that MAs have truth conditions. Byrne says KW doesn't, and that p. 86 shows as much. McGinn suggests it's clear from p. 86 that KW shouldn't, but other passages (e.g., pp. 70-1) reveal that he does. For McGinn then, p. 86 of Kripke's book just shows that KW has an inconsistent position and provides one more reason to conclude that KW is not Wittgenstein.

¹¹ I will assume that Byrne's "any fact at all" is short for , "any fact as to whether someone means plus or quus", as it must be if his argument is to have any force at all.

¹² For those who don't find it clear, the following should suffice. Suppose a sceptic, after claiming to show that there are no immortal souls, concludes that thinking is impossible. If I resist his conclusion by responding that it's a mistake to suppose that thinking requires an immortal soul, I have as yet said nothing one way or the other about whether the lack of immortal souls means the lack of any souls at all (e.g., mortal souls).

¹³ My conjecture is that the unlikelihood of Kripke so blatantly misrepresenting KW's position is what leads Byrne to try to deduce Kripke's use of the principle,

(P) If there is no superlative fact then there is no fact at all,

from Kripke's disagreement with Wittgenstein about whether our ordinary concept of meaning demands a superlative fact (a deduction which, upon reflection, is basically the same unwarranted move we saw above). If Kripke's use of the "philosophically mistaken" principle can be somehow tied to a disagreement with Wittgenstein, then it becomes plausible that Kripke would strap KW with a view that he knew was arrived at with the use of a principle rejected by KW. For Kripke could take himself to be correcting KW's position. Even so, I find it unlikely that Kripke would knowingly strap KW with a view he doesn't really hold. Still, this is really a moot point given the specious character of the deduction.

¹⁴ I would like to thank Hal Walberg, Dick Liebendorfer and a referee for Philosophy and Phenomenological Research for helpful discussions and comments concerning the topic of this paper. I would also like to thank Alex Byrne for his assistance in helping me understand his view, as well as my own.

---

John A. Humphrey
Minnesota State University, Mankato

Critical Discussions | WRPL Ultimate HomePage | Top of Page