Commentary on passages from Wittgenstein on Rules and Private Language, pp. 92-113


  • Of course if we were reduced to a babble of disagreement . . . there would be little point to the practice just described. In fact, our actual community is (roughly) uniform in its practices with respect to addition.  Any individual who claims to have mastered the concept of addition will be judged by the community to have done so if his particular responses agree with those of the community in enough cases, especially the simple ones . . . .  (K, pp. 91-2).


    COMMENTARY:  Once again, KW seems to be blissfully ignorant of the fact that mere agreement is not enough to warrant claims about someone being an adder, or not being an adder, as the case may be. If the community responses are themselves bizarre (by our lights, of course; are there others?!) then the community may very well fail to count as an addition community, i.e., a community of adders. Clearly, we must assume something about the nature of the responses beyond mere agreement in order to be satisfied that the community contains adders, or itself makes use of addition.

    This passage and others like it point out the limitation of KW's so-called sceptical solution. Despite its self-professed task of looking at our actual practices to see what we find there, it fails, miserably, to do justice to our rule following practices. As far as KW is concerned, agreement is the bottom line in rule following. Since agreement requires more than one individual, ICI's can't be rule followers, only community members can be. But the fact is that agreement is not the bottom line. As many have noted, if I wake up tomorrow and everyone in my community tells me that addition requires us to say, "68 +57 = 5", this does not make it so.

    Clearly, what we need from KW here is an explanation for the agreement. Why is our practice uniform? Why are we able to both spot and reject those who aren't adding (as opposed to, as on the KW account, to spot and reject those who aren't in agreement with us)? I think the answer is not terribly complicated, viz., addition is a pretty simple rule and its use in so-called new cases is pretty obvious. We are also taught the rule in ways that make the getting of certain answers part and parcel of following the rule.

    Although Kripke later seems to express recognition of the fact that agreement is not the bottom line, the fact is he is damned if he does and damned if he doesn't. That is, if Kripke sees Wittgenstein as taking agreement to be the bottom line then KW is offering an inadequate account of our practice. But if KW admits that agreement is not the bottom line in our rule following, KW's case against private language is seriously compromised.



  • We say of someone else that he follows a certain rule when his responses agree with our own and deny it when they do not; but what is the utility of this practice?  The utility is obvious and can be brought out by considering again a man who buys something at the grocer's. [See K, pp. 75-6].  The customer, when he deals with the grocer and asks for five apples, expects the grocer to count as he does, not according to some bizarre non-standard rule; and so, if his dealing with the grocer involve a computation, such as '68 + 57', he expects the grocer's responses to agree with his own.
    . . . When we pronounce that a child has mastered the rule of addition, we mean that we can entrust him to react as we do in interactions such as that just mentioned between the grocer and the customer.  Our entire lives depend on countless such interactions, and on the 'game' of attributing to others the mastery of certain concepts or rules, thereby showing that we expect them to behave as we do.  (K, pp. 92-3).


    COMMENTARY:  Once again, KW makes the silly claim that bare agreement or disagreement is sufficient for us to say whether someone does or does not follow a particular rule. As for the remainder of this passage, I have long thought that KW is guilty of confusing the utility of counting with the utility of meaning attributions. I don't think I have once in my life had an expectation about a shopkeeper's "counting as I do".

    More importantly, perhaps, I doubt any shopkeepers had expectations about my counting but I am pretty certain that if they thought about it at all they couldn't care less how I was inclined to count, for their particular way of counting was going to hold sway! If nothing else, it's pretty clear that all that matters in transactions between buyers and sellers is that both possess the ability to assess each others "countings" and to arrive at agreement about what is right and what is wrong in particular cases.

    But this ability to arrive at agreement gives the lie to KW's idea that there is nothing more to correct rule following than mere agreement (or disagreement) in responses. The truth is that expectations about others behaving the same as we do, even if we do have them, are by the way. (For all I know you may be quussing and I plussing, but if the case at hand is not one where plus and quus differ, the difference is of no account at all). What keeps our use of rules running smoothly is our ability to assess our answers for agreement or disagreement with our rules and our ability to communicate disagreements about particular answers. All of this implies much more than mere agreement can give us. And without these abilities we would have a breakdown of the practice.


  • We can [state] this in terms of a device that has been common in philosophy, inversion of a conditional.  For example, it is important to our concept of causation that we accept some such conditional as:  "If events of type A cause events of type B, and if an event e of type A occurs, then an event e' of type B must follow. . . . [H]ow do [Humeans] read the conditional? Essentially they concentrate on the assertability conditions of a contrapositive form of the conditional. It is not that any antecedent conditions necessitate that some event e' must take place; rather the conditional commits us, whenever we know that an event e of type A occurs and is not followed by an event of type B, to deny that there is a causal connection between the two event types.  If we did make such a claim, we must now withdraw it.  Although a conditional is equivalent to its contrapositive, concentration on the contrapositive reverses our priorities.  (K, p. 94).


    COMMENTARY:  Kripke once again betrays a misunderstanding of Hume's sceptical worries about causation. Anyone who understands Hume's discussion in the Enquiry, where he talks of sceptical doubts, a sceptical solution to them, as well as providing several definitions of 'cause', knows that Hume REJECTS the conditional above (and so cannot be said to accept its contrapositive or to "concentrate" on the assertability conditions of such a contrapositive). For Hume rejects the idea that there is any basis for saying that one event must follow another. For Humeans the conditional above (and so too its contrapositive) gives expression to a faulty conception of causation.

    What "Humeans" accept is something like the following: A cause is "an object followed by another, and where all the objects, similar to the first, are followed by objects similar to the second." (Or: "An object followed by another, and whose appearance always conveys the thought to that other."). Neither definition requires any need to take the "inversion of the conditional", nor requires us to "reverse our priorities". For neither definition commits us to "belief in a nexus" which "necessitates" one event following another.

    While it is true enough that we can deny that something is a cause if we see the alleged cause occur but it is not followed by its usual effect, clearly there are other possibilities to consider here. Kripke's "contrapositive" case above is particularly unHumean. For Hume, causal claims arise after a period of habituation or accustoming. We say that events like a cause events like b after a long process of conditioning. Given this process, our mind is led to expect a b- type event whenever we see an a - type event. Now, what are we to say when the expectation is not fulfilled? That is, what are we to say if there has been a constant conjunction between events in the past but it fails in some new case?

    Clearly, we need not say that a- type events don't cause b- type events. We could instead say that, despite appearances, what we took to be an a- type event really wasn't one. (E.g., if we see someone fall over dead after eating a piece of bread, we don't conclude that bread doesn't cause nourishment. Rather, we look to see how the stuff eaten differs from bread, e.g., it may contain poison. Of course, if we can't find anything about the poisoning stuff different from the stuff that had nourished in the past, and if we continue to experience bread eaters dropping dead, we would say that the causal relation between bread and nourishment, which existed in the past, had lapsed. Bread would in this case still be a cause, it would simply be a cause of poisoning rather than nourishment).

    The upshot is that one alleged "violation" of constant conjunction is not enough to warrant rejecting a claim about event types being causally related. Nor two cases, nor three or four. For we must first ensure that the alleged failures are in fact failures. That is, we must ensure that we really had an a- type event occur in the first place. But if did convince ourselves of this, it would still be sometime before we would entertain the idea that bread doesn't cause nourishment. So, just as a single case doesn't warrant making a causal claim, a single case cannot warrant rejecting a causal claim. This suggests, rightly, that the alleged reversal of priorities is a chimera. We don't really get something from the contrapositive of the conditional that we can't get from the conditional itself.

    A moment's reflection ought to make any logician suspicious of Kripke's attempt to reverse priorities by appeal to contrapositives of conditionals. One must be even more concerned with the particular examples Kripke gives of reversal of priorities. (See footnote 76, pp. 93-4). All of them are reversals of "because" statements. None are contrapositives of conditionals. Needless to say, it is bizarre to say the least, that Kripke should suppose that we can use contrapositives of conditionals (rather than "because" statements") to reverse priorities.


  • A similar inversion is used in the present instance.  It is essential to our concept of a rule that we maintain some such conditional as "If Jones means addition by '+', then if he is asked for '68+ 57', he will reply '125'."  . . . As in the causal case, the conditional as stated makes it appear that some mental state obtains in Jones that guarantees his performance of particular additions such as '68 + 57' -- just what the sceptical argument denies. Wittgenstein's picture of the true situation concentrates on the contrapositive, and on the justification conditions.  If Jones does not come out with '125' when asked about '68 + 57', we cannot assert that he means addition by '+'.  (K, pp. 94-5).


    COMMENTARY:  The tendentious character of Kripke's exposition here is, or ought to be, fairly obvious. First, notice that the conditional alleged to be essential to our concept of a rule, is stated in the indicative mood. It is not, in particular, normative in character. And yet, chapter 2's case against the dispositional account of meaning rested on the alleged normative character of meaning. Something is very, very wrong here indeed. If we buy the essentialness of this conditional, whither the normative character of meaning? I suggest that Kripke is wrong and the conditional that is essential to our concept of meaning or rule following, is not simply that people who mean addition will say 68 + 57 = 125. Rather, we contend that people who mean addition ought to say that 68 + 57 = 125, all other things being equal. Though one would expect Kripke to say this, the reason he doesn't say it is because the "ought" version does not "make it appear that some state of Jones guarantees" some particular answer in this case. Clearly, a normative claim in this case suggests that there is no guarantee that someone who means addition will say one thing rather than another. It simply says something about what an adder ought to say. But then the need for the inversion of the conditional lapses. Trouble indeed. Furthermore, a normative claim is also going to do justice to what we do in actual cases whereas Kripke's inverted conditional does not. For it is patently not the case that someone's coming out with something other than '125' is enough to lead us to claim that s/he does not mean such-and-such. The normative claim, and only the normative claim, expresses a recognition that adders need not always say what they ought and yet still count as adders.
    These problems are, believe it or not, by the way, since Kripke's claim fails to distinguish between the ICI and community members. For it is clear that we can say of an ICI, no less than any community member, that s/he does not mean addition by '+'. There is nothing to prevent us from recognizing the ICI's failure to say '125' and nothing to prevent us from using the conditional (or the contrapositive and its justification conditions) to deny his/her status as an adder.
    It is now possible to say something general about KW's so-called sceptical solution. The key to the solution is not, as Kripke suggests, the shift from truth conditions to assertion conditions. The key to the solution is KW's shifting the task from one of justifying the claim that someone means such-and-such to that of justifying the claim that someone doesn't mean such-and-such. KW no where shows how to justify saying of anyone, ICI or community member, that s/he means such-and-such. Rather, his sceptical solution shows us how to rule out someone meaning such-and-such. However, KW's conjuring trick is to claim that ICI's can't be said to be rule followers and then claiming that the community changes everything. But instead of showing us how to make sense of community members meaning such-and-such, KW shows us how to say of a community member that s/he does not mean such-and-such. And the truth is that nothing prevents us from saying the same of an ICI. So much for the idea that the community changes everything, or is necessary to give substance to rule following talk.


  • On Wittgenstein's conception, a certain type of traditional -- and overwhelmingly natural -- explanation of our shared form of life is excluded.  We cannot say that we all respond as we do to '68 + 57' because we all grasp the concept of addition in the same way, that we share common responses to particular addition problems because we share a common concept of addition.  . . . For Wittgenstein, an 'explanation' of this kind ignores his treatment of the sceptical paradox and its solution. There is no objective fact -- that we all mean addition by '+', or even that a given individual does -- that explains our agreement in particular cases. Rather our license to say of each other that we mean addition by '+' is part of a 'language game' that sustains itself only because of the brute fact that we generally agree.  (K, p. 97).


    COMMENTARY:  There are many problems here. First and foremost, KW has not yet told us how or why we are justified to say of each other that we mean addition by '+'. Rather, his whole discussion of the sceptical solution up to this point has focused on showing how or why we are justified in ruling out someone as an adder (or quadder, presumably). But he has not shown us what licenses us to say of each other that we mean one thing rather than another. The sad truth is that the sceptical solution is smoke and mirrors. KW no where explains why or how we are justified in saying that someone, whether community member or no, means one thing rather than another. As such, the sceptic's challenge is not solved, even sceptically.

    Rather, what we get from KW's celebrated sceptical solution is an explanation, and not a very convincing one at that, of why we are able to deny that someone means such-and-such. But surely we could have told the sceptic way back in chapter 2 that particular answers (e.g., saying 68 + 57 = 5) at least allow us to rule out someone as an adder. This leads me to offer my own "impossibility argument": It's impossible to see KW's sceptical solution as in any way solving the sceptical problems laid out in chapter 2.

    My earlier remarks should suffice to cast doubt on the adequacy of KW's claim that our agreement on particular cases is a brute fact. That is, upon reflection, one of the most absurd claims made in recent philosophy. For this reason, I think it is obvious that KW is not LW.


  • On Wittgenstein's sceptical solution to his problem depends on agreement, and on checkability -- on one person's ability to test whether another uses a terms as he does.  In our own form of life, how does this agreement come about?  In the case of a term like 'table', the situation, at least in elementary cases, is simple. A child who says "table" or "That's a table" when adults see a table in the area (and does not do so otherwise) is said to have mastered the term 'table': he says "That's a table", based on his observation, in agreement with the usage of adults, based on their observation.  That is, they say, "That's a table" under like circumstances, and confirm the correctness of the child's utterances.  (K, P. 99).


    COMMENTARY:  On several occasions while writing these commentary pages I've come across passages from Kripke which seem so off the mark or just plain USELESS (vis-à-vis the issues in the text) that I fear that some readers will wonder whether I am quoting Kripke faithfully. Such is the case with the passage above. Surely THE key element in our agreeing with each other is the fact that children are taught to utter words in certain circumstances. Far from being a brute fact, agreement is insured by the teaching process. This is Quine 101. But teaching a child presupposes a teacher who understands the terms being taught. It is this understanding which makes possible the teaching of the child. For it allows the teacher to assess, criticize or accept the responses of the child as betraying misunderstanding, being mistaken or being correct respectively. If there is a brute fact here it is this understanding, not agreement.

    Once again, it is worth noting that the notion of agreement here means only agreement on responses. But as we know from chapter 2, agreement in responses can mask disagreement on meanings. So agreement on responses cannot serve to justify meaning attributions. By the same token, "checkability" is also just checking of responses, not checking on meaning. At best then we have the ability to justify claims that someone doesn't mean such and such but that is all.

    Finally, let us also appreciate that if I am able to determine that your responses are at odds with those I am inclined to give then I can also tell if and when I give a response contrary to my understanding of some rule or other. I am not, as many seem to think, forced to accept that anything that comes out of my mouth or pen is ipso facto the right way to follow the rule. If I was, I could never be corrected by another nor bother correcting another. Instead, I'd say: You don't add like me but everything you say is correct rule following.


  • In the text we argued that for each particular rule, if conditionals of the form:  If Jones follows the rule, in this instance he will . . . , are to have any point, they must be contraposed.  If the community finds that in this instance Jones is not doing . . . , he is not following the rule. Only in this 'inverted' way does the notion of my behavior as 'guided' by the rule make sense.  Thus for each rule there must be an 'external check' on whether I am following it in a given instance.  (K, p. 103, fn. 83).


    COMMENTARY:  Since Kripke repeats his contraposition claim, I'll repeat my claim that if all we needed to answer the sceptic was to show that, whatever the status of his complaints, he cannot deny our ability to reject someone as following a particular rule (not, however to reject them as a rule follower at all) then the sceptical problem and its so-called sceptical solution turn out to be little more than a curious joke.

    Also, Kripke's "thus" in the last line here is befuddling. To begin with, what does it follow from here? Second, it suggests that the contrapositive move is unneeded. For if I am not not following the rule in a given instance, then the community will count me as following the rule. But then whither the import of contraposition? Last, but not least, this passage gives us a view precisely of the sort allegedly ruled out by the sceptic, and accepted as such by KW. For in the end, we take someone's responses in particular cases as our only "check" on whether s/he is or is not following the rule.  But isn't that exactly what the sceptic said we couldn't do in chapter 2, viz., say that someone was following plus simply because s/he gave "plus responses" in particular cases?!


  • . . . [Consider] Wittgenstein's remark that "Finitism and behaviorism are quite similar trends. Both say, but surely, all we have here is . . . .  Both deny the existence of something, both with a view to escaping from a confusion." (Remarks on the Foundations of Mathematics, p. 63 [II, §61])  How are the two trends 'quite similar'?  The finitist realizes that although mathematical statements and concepts may be about the infinite (e.g., to grasp the '+' function is to grasp an infinite table), the criteria for attributing such functions to others must be 'finite', indeed 'surveyable' . . .  .  Similarly, though sensation language may be about 'inner' states, the behaviorist correctly affirms that attribution to others of sensation concepts rests on publicly observable (and thus on behavioral) criteria . . . .  Mathematical finitists and psychological behaviorists, however, make parallel unnecessary moves when they deny the legitimacy of talk of infinite mathematical objects or inner states . . . .  Such opinions are misguided: they are attempts to repudiate our ordinary language game.  (K, pp. 106-7).


    COMMENTARY:  This passage seems to offer us a position at odds with what appears to be KW's official pronouncements concerning meaning. In particular, Kripke here suggests that although our statements can be and are "about the infinite", when we attribute a grasp of infinite functions to ourselves or others, the criteria for such attributions "must be finite" or "surveyable". By my lights, this entails that there is no problem with us attributing to each other access to, or understanding of, or grasps of, infinite tables. But this seems to be at odds with Kripke's claims in chapter 2, e.g., claims that deny the existence of infinite tables in our heads. (Cf. p. 22).

    I also do not understand the 'must' in Kripke's claim that the criteria for attributing to someone a grasp of an infinite function "must" be finite or surveyable. It seems to me Kripke is guilty of transforming an "is" into a "must".

    Finally, I have a difficult time understanding why or how Kripke's last two lines here, (where he tries to explain what is wrong with finitism or behaviorism by Wittgenstein's lights), do not serve to indict his own account of Wittgenstein. After all, KW pretty clearly challenges the legitimacy of talk of infinite mathematical objects or inner states. Indeed, he certainly denies that we can appeal to either one to help us justify our meaning attributions. At the very least, we can say that KW's sceptic has offered us a "misguided attempt to repudiate our ordinary language game".  If this is right, KW ought not agree with his sceptic at all. Rather, he should reject the sceptic's case against meaning facts as a misguided attempt to reject our ordinary language game (instead of agreeing with his sceptic but then ignoring the sceptic's claims by switching from TCs to ACs).


  • Let me, then, summarize the 'private language argument' as it is presented in this essay.  (I) We all suppose that our language expresses concepts -- 'pain', 'plus', 'red' -- in such a way that, once I 'grasp' the concept, all future applications of it are determined (in the sense of being uniquely justified by the concept grasped).  In fact, it seems that no matter what is in my mind at a given time, I am free in the future to interpret it in different ways -- for example, I could follow the sceptic and interpret 'plus' as 'quus'.  (K, p. 107).


    COMMENTARY:  I will deal with Kripke's summary in two parts. The question raised by this passage is: Who denies that we are free to interpret our concepts in different ways? Yes I can interpret 'plus' as 'quus' (or, better, interpret 'plus' to refer to the function quus). But we can also recognize the difference between the interpretations, provided we're considering a case where the two functions require different answers. Most importantly, the fact that we are free to interpret whatever is in our mind, "in different ways", does nothing to show that there is a problem with being able to justify applications of the concept in future cases. As I have urged on many occasions here, simply because a quusser can justify saying 2 + 2 = 4 by citing the function quus, does nothing to show that a plusser is not justified to give the same response.

    The key word in Kripke's claim above is not "justified" but rather "uniquely". That is the term that should have been italicized. However, that we are unable to provide a unique justification of our responses to computation problems, i.e., unable to show that all and only those who say that 2 + 2 = 4 are plussers, does nothing to show that we have not grasped the concept plus nor that we have no idea what concept of '+' we are using. Contra KW's sceptic, the sceptical problem does nothing here to threaten our language games of function talk or meaning attributions respectively, precisely because it does not and cannot remove our ability to detect differences among different interpretations of the terms of our language.

    Finally, we should balk at the suggestion here, which is not to be found in KW's arguments in chapter 2, that we are to identify, (i) "future applications are determined by the concept" with (ii) "future applications are uniquely justified by the concept".  Throughout chapter 2, KW (or his sceptic) suggests that it is justification for particular responses simpliciter that is impossible. Here, for the first time, Kripke acknowledges that KW's complaint is that we are unable to provide unique justifications for our responses.  And this would be a problem just in case any and all interpretations were indistinguishable. (E.g., we know that someone who says 2 + 2 = 4 does not mean minus or times by '+').  This is not, however, the case.  The alleged sceptical problem then is at best a case of Quinean style indeterminacy of translation, and nothing more. At worst, it's nothing more than a curious joke.


  • (2) The paradox can be resolved only by a 'sceptical solution of these doubts', in Hume's classic sense.  This means that we must give up the attempt to find any fact about me in virtue of which I mean 'plus' rather than 'quus', and must then go on in a certain way.  Instead we must consider how we actually use: (i) the categorical assertion that an individual is following a given rule (that he means addition by 'plus'); (ii) the conditional assertion that "if an individual follows such-and-such a rule, he must do so-and-so on a given occasion" (e.g., "if he means addition by '+', his answer to '68 + 57' should by '125'").
    . . . (3) As long as we consider a single individual in isolation, . . . [t]he justificatory element of our use of conditionals such as (ii) is unexplained.  (4) If we take into account the fact that the individual is in a community, the picture changes and the role of (i) and (ii) above becomes apparent.  When the community accepts a particular conditional (ii), it accepts its contraposed form: the failure of an individual to come up with the particular responses the community regards as right leads the community to suppose that he is not following the rule. On the other hand, if an individual passes enough tests, the community (endorsing assertions of the form (i)) accepts him as a rule follower . . . .  Note that this solution explains how the assertions in (i) and (ii) are introduced into language; it does not give conditions for these statements to be true . . . .  In particular, for the conditionals of type (ii) to make sense, the community must be able to judge whether an individual is indeed following a given rule in particular applications, i.e., whether his responses agree with their own.  (K, pp. 108-9).



    COMMENTARY:  A careful examination of this passage is nearly enough to reveal what is wrong with the main ideas of Kripke's book. For starters, is it not the case that we actually use meaning attributions to state facts about someone? And if we do, doesn't that spell trouble both for the claim that there is no fact about me in virtue of which . . . . , as well as the claim that an appeal to actual use is the way to a "sceptical solution" of the paradox. Second, as I have noted throughout this commentary, KW is guilty, in his sceptical solution, of ignoring the question of how it is that we are able to say of someone that s/he means plus rather than quus.

    This is made clear here, where he explicitly says we must consider how we actually use the claim that someone means addition by 'plus'. But surely the sceptic either did nothing to cast doubt on our ability to say of someone that s/he means addition by 'plus', in which case the sceptical problem is of no account, or else he does cast doubt on our ability to ascribe "meaning addition" to someone by questioning our ability to justify saying that someone meant plus rather than quus. If the latter is the case then KW's sceptical solution never addresses the main problem, viz., how can we justify saying that someone means plus rather than quus. And if the response to this is that we give up the need to do this once we shift from truth conditions to assertion conditions, (a response, by the way, which Kripke himself never makes), we need an explanation for why this shift allows us to ignore the possibility that someone who has given answers in accord with both plus and quus, means quus rather than plus.

    For the assertion conditions for meaning quus are surely identical to the assertions conditions for meaning plus up to the very case which the sceptic considers, viz., a case where the numbers in the computation are greater than 56. The claims following (3) and (4) above are both false. Or at least, if the claim following (3) is true, it is no less true for community members. As such, it is unclear how or why the community makes any difference at all in the ability to introduce or assert or use claims like (i) and (ii). Indeed, Kripke argues that for the ICI, any dispositions to respond, as well as any feelings of confidence s/he may have about his/her responses, "could be present, . . . even if he were not really following a rule at all, or even if he were doing the 'wrong' thing." What's flabbergasting about this argument is that Kripke is unaware that it applies just as well to any "community member"!

    It's also clear here that the sceptical problem, viz., what justifies saying that someone means plus rather than quus, has been ignored. Kripke here speaks of our "endorsing assertions of the form (i)", when we find that an "individual passes enough tests", i.e., that an individual has answered in accord with plus or with the community's notion of plus in enough cases. But surely this is where the sceptic could step in and ask us how we know the individual's responses have been plus responses rather than quus responses. Clearly, we know nothing of the kind.

    Also, Kripke's claim that the solution "explains how the assertions in (i) and (ii) are introduced into language; it does not give conditions for these statement to be true", is a red herring. Surely KW owes us, and regards himself as providing, more than a story about the introduction of claims like (i) and (ii). Surely he purported to be providing justification conditions for these claims, i.e., showing us why or how we could be justified to say that someone means plus in the absence of meaning facts. (N.B. The conditionals (assertions of kind (ii)) are themselves used, according to KW, to justify saying of someone that s/he doesn't mean such-and-so in cases where the consequents are violated. As such, the question of their truth or of their justification is never really discussed by KW. Instead he appeals to them as a way of justifying assertions that someone does not means such-and-such. KW never does tell us then how such conditionals come to be introduced in our language games).


  • . . . [F]ollowing [Philosophical Investigations] §243, a 'private language' is usually defined as a language that is logically impossible for anyone else to understand.  The private language argument is taken to argue against the possibility of a private language in this sense.  This conception is not in error, but it seems to me that the emphasis is somewhat misplaced.  What is really denied is what might be called the 'private model' of rule following, that the notion of a person following a given rule is to be analyzed simply in terms of facts about the rule follower and the rule follower alone, without reference to his membership in a wider community.  (K, p. 109).


    COMMENTARY:  While it's nice to see Kripke claiming that standard ideas about Wittgenstein's private language argument (hereafter, PLA) are "not in error", it's difficult to accept his claim that the standard account of the PLA is guilty only of a misplacing of emphasis. Surely the differences between Kripke's understanding of the PLA and that of the standard view come to more than a difference of emphasis!  Surely a language that is logically impossible for anyone else to understand is not all that KW is denying.

    Indeed, Kripke explicitly says on the very next page that it's quite possible for us to consider an isolated and solitary individual as a rule follower, as doing and saying what we do and say.  Clearly then, the "language" of an ICI, whatever else it is, is not going to be logically impossible for anyone else to understand.  Apparently, all it takes to understand the "language" of an ICI is a community, an appeal to which of course undermines the status of an ICI as an ICI.  As such, KW's PLA differs considerably from the standard version of the PLA, for the ICI does not speak a language that is logically impossible for anyone else to understand.  At best, an ICI is such that another cannot understand him/her, because to speak of another person understanding an ICI is, ipso facto , to longer speak of an ICI. Again, it's clear that this is not part of the standard conception of Wittgenstein's PLA.


  • Does this mean that Robinson Crusoe, isolated on an island, cannot be said to follow any rules, no matter what he does?  I do not see that this follows.  What does follow is that if we think of Crusoe as following rules, we are taking him into our community and applying our criteria for rule following to him.  The falsity of the private model need not mean that a physically isolated individual, cannot be said to follow rules; rather that an individual, considered in isolation (whether or not he is physically isolated), cannot be said to do so . . . .  Our community can assert of any individual that he follows a rule if he passes the tests for rule following applied to any member of the community.  (K, p. 110).


    COMMENTARY: This distinction between those who are merely "physically isolated" and those who are "considered in isolation", threatens to make KW's claims and arguments trivial or false. For starters, the distinction itself is by no means clear. In normal parlance, if someone asks you to "consider a single person in isolation", more likely than not you imagine something like a Crusoe figure, alone on an island, isolated from any other human beings. Kripke's passage suggests, however, that the idea of a Crusoe figure alone and isolated from all other persons, is not exactly what a person considered in isolation comes to. A Crusoe figure is alleged to be merely physically isolated but not necessarily to be a person considered in isolation. What more then is needed to turn a Crusoe figure from being merely physically isolated to being a "single person considered in isolation"?

    There seem to be at least two answers. First, Jones is a single person considered in isolation just in case our assessments of Jones' (alleged) rule following behavior are based on nothing more than Jones' own justification conditions (whatever that means; cf. Kripke, p. 89. At the least, Kripke seems to allow us access to Jones' behavior as well as his "psychological states"; as a Quinean, I have my doubts if access to psychological states is going to add anything here). A clearer answer (perhaps!) is drawn from the passage above and it suggests that Jones is a single person considered in isolation just in case our assessments of Jones' (alleged) rule following behavior avoid "applying our criteria for rule following" to Jones. For according to the passage above, once we apply our rule following criteria to Crusoe he ceases to be a person "considered in isolation" and is merely a "physically isolated individual".

    Utilizing this latter criterion it seems trivial to say that we can say nothing more about an ICI than that s/he is licensed to apply the/his rule in the way it strikes him/her (i.e., we can't say s/he is a rule follower). For if we are to say anything substantive about anyone's (alleged) rule following behavior, we must make use of our standard criteria for such talk. Otherwise, we can say nothing at all. But according to Kripke, once we apply our standard criteria to someone, s/he ceases to be an ICI and becomes merely "physically isolated". It should be obvious from this that KW's case against "private language" is based on nothing more than a semantic quibble. I say this because it is clear that KW is not claiming that it is only by direct confrontation and interaction with others that one can become a rule follower. Clearly the community is allowed to act at a distance in making Crusoe a rule follower. Furthermore, since we must make use of our rule following in order to assess anyone as a rule follower, it's clear that KW's ICI fails to be a rule follower, by definition, and so without any need for argumentation. Not exactly a stirring or profound result.

    This is not to say that some have not found KW's result to be profound, or at least important. They seem to think that it is a deep result to be told that in order to say anything about someone we perforce make use of our criteria for making such claims and that without being able to make use of such criteria, we can't say anything at all. My main complaint with those who take this reading of KW as yielding important insights is that such a reading of KW is sympathetic to say the least. In short, the problem with this account of KW is that it seems at odds with Kripke's text. Surely Kripke's text suggests throughout that Wittgenstein is going to show us that confined to a solitary individual's behavior and mental states, WE are not going to be able to say anything more about this individual's (alleged) rule following behavior than that s/he can apply the/his rule "in the way it strikes him/her".

    But if this "WE" is to have any significance here, and consequently if the claim is to have any import at all, it must refer to beings with a language, and all that that implies, who are being denied the ability to say anything substantive about the solitary individual's rule following. Instead, KW gives us the notion of an individual "considered in isolation", a notion which has the interesting property of robbing anyone who is going to talk about such an individual of his/her language. As such, it's not really WE who are unable to say anything substantive about an ICI. Rather, it's beings without a language who cannot do so. For my money, it's not terribly interesting to be told that beings who are unable to make use of their usual language cannot say anything of substance about someone. Indeed, it's trivial.

    (For the record, there are many who think, or thought, that KW is arguing, and arguing successfully, that no physically isolated individual can be said to follow rules. I do not think they should be faulted for reading KW in this way, for as I have suggested here, Kripke certainly gives the impression that this is what Wittgenstein is claiming. Couple this with the fact that such a result, viz., showing that no physically isolated individual could be a rule follower, or designated as one by us, would be a significant result, and one can appreciate why many have taken KW to make such a claim. This passage from p. 110 however, shows that KW not only has not shown that no PII can be a rule follower or speaker, but that KW is not even trying to establish such a result).


  • One must bear firmly in mind that Wittgenstein has no theory of truth conditions -- necessary and sufficient conditions -- for the correctness of one response rather than another to a new addition problem.  Rather he simply points out that each of us automatically calculates new addition problems (without the need to check with the community whether our procedure is proper); that the community feels entitled to correct a deviant calculation; that in practice such deviation is rare, and so on. Wittgenstein thinks that these observations about sufficient conditions for justified assertion are enough to illuminate the role and utility in our lives of assertion about meaning and determination of new answers.  What follows from these assertability conditions is not that the answer everyone gives to an addition problem is, by definition, the correct one, but rather the platitude that, if everyone agrees upon a certain answer, then no one will feel justified in calling the answer wrong.  (K, pp. 111-112).


    COMMENTARY:  This passage illustrates Kripke's failure to keep two very different issues straight. These two issues are, first, what makes one response to an addition problem the right response, and the second one is, what shows that someone means plus rather than some other function similar to, but ultimately different from, plus. To be sure, Kripke or KW takes these issues to be intimately related, arguing that the first question depends on the second, since what's right depends on what is meant. But I take it that KW claims that there is ultimately no answering the second question, i.e., ultimately nothing shows that someone means plus by '+' rather than some other function. Or at least the possibility that someone means something other than plus can never be completely ruled out. That we are apparently unable to say that someone means plus rather than some other function, seems to leave us unable to make any sense of "the right response" to various computations.

    The sceptical solution suggests that we can still make sense of something being the right response if and only if we have a community. The argument for this is that without appeal to other responses, anything one says is "right". And since that can't be right (i.e., if anything one says is right then we can't here speak of right), we must have other responses in order to give substance to our talk of the right response. In particular, the sceptical solution suggests that the right response can only be determined by comparing one response with other responses. Roughly speaking, the right response is the response accepted within one's community.

    Now while it is true enough that a community makes it possible to speak of particular responses being in accord with or out of step with "other responses", and so to make sense of saying that one response is not right, it's clear that we do not wish to make the majority response the right response. But that is basically what KW has done. Most importantly however, KW allows that our meaning talk is based on hypotheticals, on conditionals. That is, we can only say that if someone means plus then s/he ought to say that 68 + 57 = 125, or if someone means quus then s/he ought to say that 68 + 57 = 5. But it should be obvious that the use of such conditionals reveals that there is no special problem with determining what the correct responses are to addition and quaddition problems respectively.

    Rather, the only problem, ultimately unsolvable insofar as one seeks a unique answer, is determining what someone means by his/her terms (this is just Quine's thesis of indeterminacy, by the way). Most importantly, there is no real difference here between the so-called solitary case and the case of members of a community. KW's appeal to the community does not allow us to determine someone's meaning to uniqueness.  The ability of members of a community to create hypothetical or conditional statements regarding meaning, can be matched by a solitary individual.  Indeed, they must be so matched, for if no individual can manage the feat of forming conditional statements of meaning or determining when his/her behavior accords with or violates such conditionals, it is impossible that such an ability could arise just in case there are others present. The only difference between a solitary individual and a community member is that a community member can be using a particular sign differently from the way others in his/her community are using it. As such, a community member can be "corrected" by the community, or VICE-VERSA.  Obviously, this cannot happen in the case of a solitary individual, or an ICI. But this ability of a community to "correct" (or create uniformity) has nothing to do with the possibility of substantive rule following or substantive rule following talk. Just here is KW's mistake.

    KW claims, wrongly, that given the sceptical argument, the only way we can make sense of particular responses being right or wrong is to compare them with other responses. That is, being wrong amounts to being out of step and being right amounts to being in accord. Contra KW, the possibility of substantive rule following rests on the ability to create meaning conditionals and to assess behavior as according with or violating such conditionals as the case may be. KW's mistake then comes to the following. KW rightly sees that it makes no sense to speak of a solitary individual using a sign differently from the way others use it. For this we need a community of people.

    However, KW fails to appreciate that this fact has nothing to do with whether a solitary individual is or can be a substantive rule follower. This question is a matter of whether one is able to create meaning conditionals and to assess behavior as according with or violating these conditionals. Nothing in KW's book shows that ICIs are unable to create and assess meaning conditionals. Rather, KW shows only that it makes no sense, in the case of an ICI, to say that his/her use of a sign is or is not out of step with the use of that sign by others.

 


Top of Page | WRPL Ultimate HomePage


Last modified October 2, 2011
JAH, Professor
Dept. of Philosophy