- The sceptical paradox is the
fundamental problem of Philosophical
Investigations. If
Wittgenstein is right, we cannot begin to solve it if we remain in the
grip of the natural presupposition that meaningful declarative sentences
must purport to correspond to facts; if this is our framework, we can only
conclude that sentences attributing meaning and intention are themselves
meaningless . . . . The
picture of correspondence-to-facts must be cleared away before we can
begin with the sceptical problem. (K, pp. 78-79).
COMMENTARY: Many commentators reject the idea that the sceptical paradox set out in Kripke's book constitutes the fundamental problem of PI. They also reject the idea that the paradox spoken of by Wittgenstein at §201 constitutes the fundamental problem of PI.
I have already suggested, to put it mildly, that I do not agree that KW's sceptic succeeds in showing that meaning/language is impossible, even for a truth conditional notion of meaning. As such, I disagree with KW that the correspondence-to-fact picture must be cleared away before we can “begin” with the sceptical problem. (One suspects that Kripke’s ‘begin’ here is short for, “begin to solve”).
As my previous commentary makes clear, what needs to be cleared away is the bogus idea that an account of meaning is inadequate (and so is to be rejected) if it is unable to show how we can be justified to say that someone meant plus rather than quus, or some other bent function. Drop this bogus idea and the sceptical paradox never gets off the ground. Curiously enough, It’s clear to an astute reader (to borrow a line from the next Kripke passage below) that KW, in some sense, does just this, namely, gives up (or simply ignores) this bogus idea. For he nowhere confronts the assertion conditions account with the “rather than” question, that is, he never asks, let alone answers, this question. Had he done so, he would have appreciated that the assertion conditions that license us to say, “Jones means plus by ‘+’, also license us to say, “Jones mean quus by ‘+’. As such, the assertion conditions account, like the truth conditions account, has no good answer to the “rather than” question. Which means that, so long as the “bogus idea” is in play, the assertion conditions account is just as inadequate as the truth conditional account. By the same token, if one eschews the bogus idea, if one doesn’t ask the “rather than” question, a truth conditions account will do just as well as an assertion conditions account.
KW, however, is blissfully ignorant of the fact that his assertion conditions account holds no advantages over the truth conditional account vis-à-vis the sceptical challenge or its sceptical solution. His insistence on, and discussion to try to show, the contrary is due, in the end, to his not holding these accounts to the same requirements. It’s sad to think that KW (or Kripke) could fail to appreciate this, and thereby be led to offer a quite bogus discussion extolling the power of an assertion conditions account to save us from semantic nihilism. Sadder still is the thought of people reading this discussion and being convinced by it.
- We have not yet looked at
the solution of the [sceptical] problem, but the astute reader already
will have guessed that Wittgenstein finds a useful role in our lives for a
'language game' that licenses, under certain conditions, assertions that
someone 'means such -and-such' and that his present application of a word
'accords' with what he 'meant' in the past. It turns out that this role,
and these conditions, involves reference to a community. They are inapplicable to a single
person considered in isolation.
Thus, as we have said, Wittgenstein rejects 'private language' as
early as §202. (K, p. 79).
COMMENTARY: To begin with, this passage contrasts, I believe, with the passage from p. 71 and p. 73, where Kripke first lays out the two questions which comprise Wittgenstein's view of language in the Investigations. In the previous passage, KW was said to replace questions about the truth conditions of a sentence with questions about the conditions under which a sentence may be appropriately asserted (denied) and about the sentence's "utility". The present passage, however, suggests that KW is out to show us something about the language -game of meaning attributions, e.g., that such a language-game is useful and that the language-game licenses meaning attributions, "under certain conditions".
Furthermore, it is also suggested here that the language-game's role and the conditions it (viz., the language-game of meaning attributions) sees as licensing meaning attributions "involves reference to a community" and are "inapplicable to a single person considered in isolation". But what prevents fans of truth conditions or meaning facts from echoing KW here? Surely truth conditionalists could claim to find a useful role for a language-game that licenses, under certain conditions, not only the assertion of meaning attributions but also licenses claims that such and such a meaning attribution is true because of facts about the subject of the attribution or his behavior, broadly construed. This shows, once again, that a truth conditional account is really a species of an assertion conditions account, as Kripke conceives of them anyway.
Most importantly, however, the imaginary reply of truth conditionalists helps bring out a serious problem in the passage above. The problem is that the utility of a particular language game presupposes its existence. No one can deny that our language-game of meaning attributions is useful. The sceptic’s question is whether it is possible, that is, the question is whether we can find a way of licensing our meaning attributions "under certain conditions". As suggested above, not any old thing someone wants to say about another's (or one's own) meaning, at any old time, in any old circumstance or given any old conditions, can be accepted as a licensed meaning attribution. The trick is to hook up the right meaning attributions with the right conditions. KW's sceptic alleged that this couldn’t be done, so long as we assume the meaning attributions state facts and so must be hooked up with facts in the world, be they public or private. Of course, we must, as I have stressed throughout this commentary, keep in mind the argument the sceptic uses to establish that we can't hook up meaning attributions with "facts". The alleged problem is that the facts license multiple and allegedly conflicting meaning attributions (the “rather than” problem).
Thus, echoing the sceptic, what we want from KW is not just some way of licensing meaning attributions but some way of licensing them uniquely. Otherwise, we've been swindled. I contend that we have been swindled because KW fails to show us that his way of legitimating meaning attributions (viz., by opting for assertion conditions rather than truth conditions) can beat sceptical worries about uniqueness.
For unsuspecting readers, it will be useful to be told here that the phrase, 'a single person considered in isolation' does not mean a person all alone on an island, nor a person considered in such conditions. For Kripke allows that physically isolated individuals, or persons considered to be physically isolated, can be said to follow rules. As such, it's terribly unclear just what Kripke or KW does mean by the notion of an "individual considered in isolation". For more details, see K, p. 110 for Kripke's not entirely coherent or cohesive song-and-dance about just what sort of isolated individual can or cannot be said to have a language.
- Wittgenstein's sceptical
solution concedes to the sceptic that no 'truth conditions' or
'corresponding facts' in the world exist that make a statement like
"Jones, like many of us, means addition by '+'" true. Rather we
should look at how such assertions are used.
Can this be adequate? Do we not call assertions like the one just quoted
'true' or 'false'? Can we not with propriety precede such assertions with
'It is a fact that' or 'It is not a fact that'? Wittgenstein's way with
such objections is short. Like many others, Wittgenstein accepts the
'redundancy' theory' of truth: to affirm that a statement is true (or
presumably, to precede it with 'It is a fact that . . . ') is simply to
affirm the statement itself, and to say it is not true is to deny it: ('p'
is true = p) . . . . We call
something a proposition, and hence true or false, when in our language we
apply the calculus of truth functions to it. That is, it is just a
primitive part of our language game, not susceptible of deeper
explanation, that truth functions are applied to certain sentences. (K, p. 86).
COMMENTARY: I'm not sure about the point of this passage, or why Kripke thinks it does any work at all. Clearly Kripke thinks it answers an objection that might be raised against his Wittgenstein. But upon reflection, it doesn't seem that it really gets to the heart of the matter. That Wittgenstein is a redundancy theorist about truth (or a deflationist about truth) means that, for him, assertions that add 'it's true that' or 'it's a fact' to sentences come to the same thing as asserting the sentences themselves, without the 'it's true that' or 'it's a fact'. However, someone critical of KW's agreement with his sceptic about the nonexistence of meaning facts or "truth conditions" to make, "Jones means addition by '+'", true, takes the view that KW's sceptical solution, if accepted, would leave us unable to give any substance to our meaning attributions themselves. In short, meaning attributions are threatened with meaninglessness in a world without meaning facts. As such, KW's critics are hardly going to be answered by being told that KW is a redundancy theorist. For telling these critics that, "It's a fact that Jones means plus by '+'" says no more and no less than, "Jones means plus by '+', is inadequate given that the critics see KW undermining the legitimacy of asserting the latter expression.
The bottom line then is as follows: While we do indeed call meaning attributions true or false, it is also the case that we assert meaning attributions as well. But the sceptic alleges that the lack of meaning facts undermines the legitimacy of asserting meaning attributions. Thus, those who question the adequacy of Wittgenstein's sceptical solution, are not merely asking how KW can hope to explain how we are able to call meaning attributions true or false in the absence of meaning facts but also how we are able to justify asserting them at all. Such people are unlikely to be appeased, indeed, they ought not be appeased, by Kripke's claim here that Wittgenstein is a redundancy theorist about truth.
We can, as redundancy theorists, agree that if meaning attributions can be legitimately asserted then so too can claims that such attributions are true, since these claims come to the same thing. But we still haven't been told how meaning attributions can be legitimately asserted in the absence of meaning facts.
It should also be noted that the original sceptical problem concerned the JUSTIFICATION for, or legitimacy of, saying that someone meant plus, or for giving one response rather than another to "new" computation problems. Do we really think that an appeal to how people actually USE such expressions is sufficient to solve the sceptical problems? Simply knowing how people actually use a bit of language is not enough to show that the use is justified or legitimate. Indeed, even knowing what people actually take as justifying or legitimating bits of language is not enough to show that the bit of language in question is justified or legitimate. For sometimes people can make mistakes about the legitimacy of their attempted legitimizations. Or, what we actually do in our language games can hardly be deemed as self-justifying. (See my next commentary).
- Finally, we can turn to
Wittgenstein's sceptical solution and to the consequent argument against
'private' rules. We have to see under what circumstances attributions of
meaning are made and what role these attributions play in our lives . . . . [W]e will find out what circumstances actually license such assertions
and what role this license actually
plays. (K, pp. 86-87).
COMMENTARY: Pardon my ignorance but it sounds like Kripke is suggesting that solving (albeit sceptically) the sceptical problem is simply a matter of looking at "real live" cases of meaning attributions to discover the circumstances under which they are made and also discerning the role such attributions "play in our lives" (recall Goldfarb's complaints about the sterility of the latter notion). But surely this is a perverse approach to trying to uncover the assertion conditions for meaning attributions, at least given the sceptical challenge. For such an approach assumes that anytime someone asserts a meaning attribution, either about him/herself or about another person, that it is a legitimate assertion, no questions asked!
But it must be admitted that sometimes people make meaning attributions in circumstances under which the meaning attribution is not really legitimate. As such, we cannot just read off the assertion conditions for meaning attributions from "real live" cases. We must first ensure that the real live case is a case in which the meaning attribution can be legitimately asserted. Of course, I think we can and do know the conditions under which meaning attributions are, or can be, legitimately asserted. However, such knowledge involves an acquaintance with, or understanding of, the language game of meaning attributions.
But there's a fly in the ointment here, viz., the sceptic. Surely the sceptic showed that all is not well with our usual language game of meaning attributions. Surely the upshot of the sceptical challenge is that the conditions legitimating the assertion of, "Jones means plus by '+'" are also conditions legitimating the assertion of "Jones means quus by '+'". Once again it seems that we are faced with the question of whether there is any way to make the world safe for one meaning attribution over another. Kripke's suggestion here that we are to look at actual cases, or look to see under what circumstances meaning attributions are actually made, suggests that the sceptical paradox is a result of not looking to the actual cases. However, this does not seem to be the problem at all.
On the contrary, the problem is that there seems to be nothing in the world, nothing in our minds, (and so nothing no where) which would legitimate asserting the plus meaning attribution without also legitimating the quus meaning attribution. If this is the sceptical problem, and I think it is, then solving the sceptical problem, whether sceptically or straight, cannot be a matter of simply looking to actual, "real live", cases of meaning attributions. For real live cases of meaning attributions obviously ignore sceptical hypotheses of the sort proffered by the sceptic. But those who are troubled by questions about the justification and legitimacy of our meaning attributions are hardly going to be content with being told that since the sceptic is ignored in real live cases, we can with perfect right assert that Jones means plus by '+', despite the fact that for all we can tell, Jones might very well mean quus by '+'. By simply appealing to what we actually do in our language game, Kripke or KW doesn't solve the sceptical problem, he ignores it.
- First, consider what is true
of one person considered in isolation. The most obvious fact is one that
might have escaped us after long contemplation of the sceptical paradox.
It holds no terrors in our daily lives; no one actually hesitates when
asked to produce an answer to an addition problem! Almost all of us
unhesitatingly produce the answer '125' when asked for the sum of 68 + 57,
without any thought to the theoretical possibility that a quus-like rule
might have been appropriate! And we do so without justification. Of
course, if asked why we said '125', most of us will say that we added 8
and 7 to get 15, that we put down 5 and carried 1 and so on. But then,
what will we say if asked why we 'carried' as we do? Might our past
intention not have been that 'carry' meant quarry; where to 'quarry' is . . . . ? The entire point of the
sceptical argument is that ultimately we reach a level where we act
without any reason in terms of which we can justify our action. We act
unhesitatingly but blindly. (K,
p. 87).
COMMENTARY: This passage constitutes the first step in Kripke's official story about KW's sceptical solution. Notice its opening line, viz., "consider what is true of one person considered in isolation". I have long been perplexed by its presence here, since the remainder of the passage seems not merely to be about someone "considered in isolation". For example, the passage speaks of our being asked for things, presumably by others, e.g., asked for a sum, asked why we answered '125', asked why we carried as we do, and asked about our past intention regarding 'carry'. It's also said that there are no terrors for us "in our daily lives". Needless to say, in our daily lives we are not isolated, nor can we be considered in isolation in our daily lives. It is difficult then to see that any of what is said in this passage is true only of one person considered in isolation. At the least, even if all that is said here is true of one person considered in isolation, it is pretty obvious that it is no less true of people who are not considered in isolation as well.
While I allow that the sceptical paradox "holds no terrors in our daily lives", I would add, contra Kripke, that this fact has not escaped us and ought not escape us. Also, Kripke is correct to speak of "the theoretical possibility" that we might be following a quus-like rule. For that is precisely what it is, at best. I would also add that there are many theoretical possibilities that we have no way of ruling out but nonetheless ignore as being too fanciful to be taken seriously. The seven-minute old universe for one, the theft of all our furniture and its replacement by exact duplicates, for another. Needless to say, if the quus scenario is no better than the seven-minute-old universe case, then it really isn't worth taking seriously.
One serious point of exegesis concerns Kripke's claim, "[a]nd we do so without justification". What is it that we do without justification, (i) ignore the theoretical possibility that we might be quussing rather than plussing, or (ii) produce the answer '125' when asked for the sum of 68 + 57? Perhaps it is both (i) and (ii) that Kripke thinks we do without justification. However, whether Kripke is speaking of (i) or (ii), or both, his claim is a bit overstated. As I remarked earlier in my commentary, it is wrong to say that '125' is an unjustified answer to the sum question. It is also wrong to say that the claim that I mean/meant plus is unjustified. If anything is unjustified here, it is, or would be, my claim to mean plus rather than quus. What we need to appreciate (and which Kripke doesn't whereas Quine does) is that being equally justified is not to be confused with not being justified. Plus and quus are equally justified by my behavior (both inner and outer) to be the function I meant by '+'. But whereas they are justified, the functions minus and times are not justified as the function I meant by '+'. This takes care of (ii). As for (i), I think there is justification for ignoring the possibility that we're quussing. As I noted earlier, if nothing else, simplicity considerations suffice to justify ignoring it.
For similar reasons, Kripke's penultimate line here, (viz., "The entire point of the sceptical argument is that ultimately we reach a level where we act without any reason in terms of which we can justify our action"), is the sort of claim to which a Yes and No response is appropriate, depending on how "justify" is understood. If justifying our action requires finding something that justifies one action and only one action (e.g., the action of saying 68 + 57 = 125 rather than that of saying that it equals 5 or 20 or 56) then the sceptical argument shows that our act of saying 125 to 68 + 57 is unjustified. But if justifying our action does not require "unique justification" then the sceptical argument cannot be said to be successful, i.e., cannot be said to have shown that ultimately we act without justification.
However, the bottom line in all of this is that Kripke's penultimate line gives rise to a kind of dilemma. If the sceptical argument is successful, i.e., if in fact we reach a level where we act without any reason in terms of which we can justify our action, then where are we to go from here? Isn't the only conclusion to draw from this argument as follows: the attempt to find a justification for one answer rather than another is hopeless? Furthermore, since KW allegedly accepts the sceptical argument, then he too is committed to saying that ultimately we act without justification. And yet presumably, the sceptical solution is supposed to show us how to justify the act of saying 68 + 57 = 125 rather than 5. But if this can be shown, what becomes of the success of the sceptical argument? For such a solution would show that we do not in fact reach a point at which we act without any reason it terms of which we can justify our action.
There is, I think,
only one way out of this dilemma, although it doesn't strike me as completely
satisfactory. The way out is to say that the sceptic is successful in showing
that ultimately there is no justification but nonetheless justification of
actions or meaning attributions is possible at something less than an
"ultimate level". Frankly, I think something like this is precisely
what KW offers us. That is, he agrees with the sceptic that, ultimately, there
is no justification but claims that such a lack is no threat to our ordinary
game of meaning talk because we simply don't demand ultimate justification. My
main complaint with this rather facile solution to the problem is that it fails
to appreciate that the case against meaning facts was based on their inability
to deliver an "ultimate justification". If we are going to give up on
the demand for an ultimate justification then the case against meaning facts
must be reassessed.
- It is part of our language game of speaking of rules that a
speaker may, without ultimately giving any justification, follow his own
confident inclination that this way (say, responding '125') is the right way to respond, rather than
another way (e.g., responding '5'). That is, the 'assertability
conditions' that license an individual to say that, on a given occasion,
he ought to follow his rule this way rather than that, are, ultimately,
that he does what he is inclined to do. (K, pp. 87-8).
COMMENTARY: Another puzzling and troublesome passage. To begin with, what is the opening sentence saying? It's true enough that we may do lots of things in our language games. I may shout "Lillibulero" at the top of my lungs during a lecture on Plato's cave scene and no one bats an eye. Does this mean that my shouting is a legitimate step in this particular language game? The question here is whether all the things we may do in a language game are kosher or beyond dispute or beyond criticism. Is KW here claiming that anything anyone wants to say about "the right way" to respond to a particular computation problem, or anything anyone wishes to call the "right way to respond" is, ipso facto, unobjectionable? If so, it is clear that the claim is false. For if my daughter, a 2nd grader, tells me that 5 is the right way to respond to "68 + 57 = ?", I am going to challenge her claim. And she better be able to say more than that she is following her confident inclination about how '+' runs.
Perhaps she'll say that she believes that '+' requires the answer '5' in any case in which the numbers in the problem are greater than 56. And if she says this, then I will say that she in fact is following her understanding of '+' in the right way. However, it is clear that she would then have given me more than an appeal to her "confident inclination" in support of her response. I would also tell her that other people use '+' differently from her and then show her how other people understand '+'. I would show her, in short, why or how people come to take '125' as the right way to respond to 68 + 57. (NOTE: It's clear that so long as someone gives what we all regard as the right answer, no one will ask for any justification. Particular answers are, in some sense, their own justification. Such is the case for the answer 125 to the question, "68 + 57 = ?", but not the case for the answer 5 to the same problem. Answering 5 requires an appeal to the function quus or some similar bent-function, for example. Again, though, do not confuse being "self-justified" and not being justified! ).
Of course, it is true enough that someone may claim to be following some rule and insist that his response to a problem is the right way of responding. It's also true that challenging someone's claim to be following a rule correctly can be a pointless task, so long as the challenger and the alleged rule-follower are using the same sign in different ways. For it cuts no ice at all to say to a quusser, who claims that 68 + 57 = 5, that '+' requires 125 as its answer. For '+' doesn't require anything until it has been tied to some function or other. Tie it to the function quus, and the quusser is right to say 5; tie it to the function plus, and the plusser is right to say 125. And furthermore, there is no right answer to the absurd question: "What function ought '+' to be tied to?" If this is all Kripke or KW is saying then of course, he is correct. But it seems too trivial to do justice to what KW WANTS TO SAY.
It's also a mystery to me how the two sentences of this passage can be equivalent, although it is clear that Kripke believes them to be so. Otherwise, the "that is" would be out of place. The second sentence, in particular, is a bit more substantive than the first sentence, but unfortunately, it appears to be false. In particular, the second sentence speaks of assertability conditions that LICENSE an individual to say something about how he ought to respond to "his rule". The first question here is who accepts or believes that the assertability conditions laid out here as licensing an individual to say, e.g., that he ought to say 68 + 57 = 5, are simply that, ultimately, responding with 5 rather than 125 is what he is inclined to do? I do not. As the case of my daughter above makes clear, I at least expect some appeal to a rather odd function if I am going to accept my daughter's claim that 68 + 57 = 5.
In answering computation problems, one's inclinations are not what we use to legitimate our answers. Again, my daughter may have an inclination to quus rather than plus, and I the opposite inclination. But still, her answer is not justified by her inclination but rather by the function quus. The sober truth here is that because she is inclined to quus rather than plus, she takes 5 to be the right answer rather 125. But this is still not justifying a response by bare inclination. It's still justifying a response by appeal to a function. As such, the assertability conditions that license an individual to say that on a given occasion, he ought to follow his rule this way rather than that, are, ultimately, that what he says about the way he ought to follow his rule is in fact the way he ought to follow his rule! In plain English, the ACs that license an individual to say that he ought to follow his rule this way (by saying 5) rather than that (saying 125), are, ultimately, that his rule requires him to say 5 rather than 125. And unless the rule follower can tell us a coherent story about a rule that requires the answer 5 in the case at hand, we will dismiss him as a crank.
I want to stress that my commentary here supposes that Kripke's second statement purports to say something of substance. It's possible that he is simply saying that someone can say how he ought to follow his rule without fear of being challenged or refuted by another person. Of course, this is true but trivial. For it is precisely what someone says about how his rule ought to be followed that allows us to determine what his rule is. That is, someone telling us what his/her rule is, thereby EXPLAINS his/her rule or understanding thereof. And we can obviously take one of two different attitudes toward those we ask for such explanations (i.e., those who claim a very odd answer is requisite in a particular case). We can see the explanation as evidence of a mistake (e.g., we could say of someone who is explaining why 5 is the required answer, that s/he misunderstands addition) or we can see it as evidence that a bent rule is being followed. In this latter case, the idea of a mistake is premature and unwarranted. For we cannot criticize a response to a rule until we know what rule is being followed. (Do not, however, confuse this with being unable to determine whether a response is in accord with, or out of step with, any or all rules. See below). And if Kripke is talking about explanations of rule here then it is the rule being followed that is being explained to us. But surely Kripke is not here giving us the assertability conditions for an individual's explanation of how his rule is to be followed. For it would render his remarks here true but trivial.
It should be noted, and noticed, that whereas the first sentence of this passage says nothing at all about a rule that one is using to decide what is the right way to respond, the second sentence says something about "his rule", rather than a rule or the rule. In my paper, "Kripke's legerdemain . . .", I note the importance of appreciating Kripke's slipperiness in either leaving out any mention of rules or in talking about "his rule", as opposed to some particular rule or other, e.g., the rule for plus or quus. While it might be granted that our language game licenses Jones to follow his rule as he sees fit, it is obviously not the case that we are going to license any or all claims which Jones wishes to make about the relationship between his rule and rules like plus or quus. For example, we are not going to license Jones to say that the rule for plus requires one to say that 68 + 57 = 5. His rule for '+' may require 5 as the answer to this case but we do not have to agree that Jones's rule for '+' is our rule for '+'.
As it stands then, KW's claim here about assertability conditions is, at best, true but trivial. For although anyone can say anything they like about their rule, the significant question is whether it is possible for us to judge that someone's response to a computation problem is in step with or out of step with rules like plus or quus. Note also that the weasel word 'ultimately' appears in both sentences. This term also has the power of rendering KW's claims true but trivial. For if I am asked why I use '+' to refer to plus rather than quus, I might very well say, I'm inclined to do so and leave things at that, unable or unconcerned with coming up with any better answer. However, this is not to say that my following of plus or quus is a matter of my inclination. I can follow any rule I'm inclined to follow with impunity but I cannot follow any rule as I'm inclined to follow it with impunity.
- The
important thing about this case is that, if we confine ourselves to
looking at one person alone, his psychological states and his external
behavior, this is as far as we can go. [ . . . ].
There are no circumstances
under which we can say that, even if he inclines to say '125', he should have said '5', or vice versa. By definition, he is licensed to give, without
further justification, the answer that strikes him as natural and
inevitable. (K, p. 88).
COMMENTARY: Perhaps the most puzzling thing about this passage is the use of the metaphorical expression, "this is as far as we can go". I find this bothersome because the paragraphs that precede this line contain the term "ultimately", which suggests, of course, something final, or last, an endpoint of some sort. KW tells us that ultimately, we have nothing more than our inclinations as guides to how to follow rules. And since it's ludicrous in such a case to criticize someone's way of following a rule (people are licensed to give the answers that strike them as natural and inevitable!!), rule following seems to go up in smoke. However, to tell us that so long as we're confined to looking at one person alone, "this is as far as we can go", suggests that our difficulty is illusory only because we haven't gone far enough, and consequently that there is somewhere "farther along" to get to.
But KW seems to neglect that we've been asked to consider the ultimate case, to consider where we get to "ultimately". In the non-ultimate case, we explain our response by appeal to a rule. If pressed for a reason as to why we followed this rule rather than another (or for a reason we interpreted the rule as we did) we ultimately say, that's just the way I'm inclined to go. (Note: This answer says more about the question, viz., that it's a bogus question, than it does about the legitimacy of our rule following). But then whatever solution we find for this alleged problem should not consist of "going farther"; otherwise, we really never reached the ultimate point. And if we did reach the ultimate point, then we can't go farther. Clearly then, the claim that so long as we are confined to one person "this is as far as we can go", is bogus and misleading. We cannot go further and we should not be trying to do so. Rather, what we need to do is to move back from the ultimate position (a position, I remind you, that seems to pose a problem for members of communities no less than for persons considered in isolation) and see if somewhere short of the ultimate position there isn't a way of making sense of rule following.
As such, it's not a flaw of the isolated individual case that once we've gotten to the ultimate position that "this is as far as we can go". Again, the only question here is whether there is someplace short of the ultimate position that allows us to make sense of a person considered in isolation following rules (i.e., to have more than this individual's inclinations to go on in trying to get straight about his/her rule following). Perversely, KW ignores this question altogether.
There is also a very clear sense in which it is false that an isolated individual is by definition licensed to give, without further justification, the answer that strikes him as natural and inevitable. Perhaps ultimately, this is so, i.e., ultimately we can "justify" our goings-on by nothing more than saying that we find this natural or inevitable. (Once again, though, this is not true only for isolated individuals but for anyone, for the sceptical argument is an equal opportunity argument). However, even in the isolated case, we can make perfect sense of saying that his/her response is out of step with plus or quus or minus or times, etc. Indeed, no one is licensed to say 68 + 57 = 5 when '+' refers to plus, nor that 68 + 57 = 125 when '+' refers to quus.
- No one
else by looking at his [i.e., the person considered in isolation] mind and
behavior alone can say something like, "He is wrong if he does not
accord with his own past intentions"; the whole point of the
sceptical argument was that there can be no facts about him in virtue of
which he accords with his intentions or not. All we can say, if we consider
a single person in isolation, is that our ordinary practice licenses him
to apply the rule in the way it strikes him. (K, p. 88).
COMMENTARY: The simplest way to criticize this passage is by asking how or why does KW think he can get from the first sentence to the second sentence? I myself see no legitimate route. Unless we suppose that an inability to speak of someone's intentions concerning rule following leads, willy-nilly, to licensing any old way someone wishes to follow a rule, we must be dubious of KW's move here. Most importantly however, we need to appreciate that such a supposition would mean that all we can say of anyone at all, not just a single person considered in isolation, is that our ordinary practice licenses them to apply the rule in the way it strikes them to follow it. For the fact is that the sceptical argument shows, for anyone at all, the futility of appealing to someone's intentions to make sense of his/her rule-following. More simply put, we know for certain that however KW ends up making sense of rule following talk, he will not make any references to someone's intentions. (See, K, p. 91).
As such, that we cannot say of someone considered in isolation, "He is wrong if he does not accord with his own past intentions", is true enough but nonetheless irrelevant to whether it is possible to make sense of such a person following rules. For KW will not and does not require us to uncover someone's intentions in order to make sense of his/her following rules. Thus, KW's move from sentence one to sentence two is illegitimate since he rejects the principle under which such a move would be legitimate, viz., that rule following requires access to someone's intentions regarding rule following.
Most importantly, the fact that KW's move here is illegitimate shows that KW never establishes that private language or rule following is impossible for someone "considered in isolation", in any sense in which it is not also impossible for people not so considered. The simple truth is that Kripke has not shown (and cannot show) that we are unable to say of an ICI that his/her way of following a rule is out of step with some particular rule or other. For example, if an ICI says 68 + 57 is 5, we can say of him/her that s/he is not plussing; if s/he says 68 + 57 is 125, we can say s/he is not quussing. We can then say of an ICI much more than s/he is licensed to follow "the rule" in the way it strikes him/her. Furthermore, all of this is included in our "ordinary practice". (For those overly astute readers worried that I am obviously ignoring the fact that my response here obviously requires appeal to a community, (and so vindicating KW's point that it takes a community to speak a language), see my comments to passages from p. 110 of Kripke's book, as well as most of the remaining commentary on this page).
- But of
course this is not our usual
concept of following a rule. It is by no means the case that, just because
someone thinks he is following a rule, there is no room for a judgement
that he is not really doing so. . . . If there
could be no justification for anyone to say of a person . . . that his
confidence that he is following some rule is misplaced, or of a person . .
. that he is no longer in accord with the rule that he previously
followed, there would be little content to our idea that a rule, or past
intention, binds future choices.
(K, pp. 88-89).
COMMENTARY: This passage helps to clarify, and also threaten the legitimacy of, the previous passage, as if the previous passage doesn't have enough problems already. What this passage helps us see is that among the things we allegedly cannot say of a person considered in isolation (but which we must be able to say in order to give substance to our rule following talk in the case of such an individual) is that his confidence that he is following some rule is misplaced or that he is no longer in accord with the rule that he previously followed. But why can't we say such things of an ICI (individual considered in isolation)? Given all that Kripke has provided so far, the only answer would be: Such talk requires access to someone's intentions, something which we can't have in the case of ICIs.
Such an answer is, of course, off the mark for two reasons. One, because we have no such access in the case of anyone at all, which would force us to say that our usual concept of rule following is bankrupt. Two, because it's false, i.e., we do not have to access someone's intentions in order to pass legitimate judgment on their rule following. It follows then, as I noted in the previous commentary, that for all KW shows, we can pass judgment on the success or failure of an ICI's rule following.
- If our
considerations are so far correct, . . . if one person is considered in
isolation, the notion of a rule as guiding the person who adopts it can
have no substantive content.
There are, we have seen, no truth conditions or facts in virtue of which
it can be the case that he accords with his past intentions or not. As
long as we regard him as following a rule 'privately', so that we pay
attention to his justification
conditions alone, all we can say is that he is licensed to follow the rule
as it strikes him . . . . The situation
is very different if we widen our gaze from consideration of the rule
follower alone and allow ourselves to consider him as interacting
with a wider community. Others will then have justification conditions for
attributing correct or incorrect rule following to the subject, and these
will not be simply that the
subject's own authority is unconditionally to be accepted. (K, p. 89).
COMMENTARY: Once again KW alleges that an ICI cannot be said, in any substantive sense, to be a rule follower, for "all we can say" of him/her is that s/he is licensed to follow the rule as it strikes him/her. Notice that once again, KW contends that the problem for the ICI lies in the lack of facts about accord with his/her intentions. Once again, I'll claim that KW has conveniently ignored the fact that the case against there being such facts is the sceptical paradox, which is not an argument that applies only to ICIs. Rather, it applies to everyone, ICI or community member, quite equally. As such, if the lack of facts about intentions is incompatible with substantive rule following (i.e., all we can say of someone of whom there are no facts about his/her past intentions is that s/he is licensed to follow the rule as it strikes him/her) then substantive rule following is impossible, not just for ICIs but everyone.
However, this passage does contain an odd maneuver, namely, the claim that regarding someone as following a rule privately requires paying attention only to the alleged rule follower's justification conditions. On the previous page (p. 88), KW said on two separate occasions that considering a person in isolation required us to "confine ourselves to looking at one person alone, his psychological states and his external behavior". But surely KW supposes that an ICI is a private rule follower, and vice versa. If so, we must then suppose that paying attention to someone's justification conditions alone is to be equated with confining ourselves to looking at the psychological states and external behavior of a solitary individual. But this strikes me as problematic, for I see no difficulty at all with criticizing or challenging someone's way of following a rule based solely on access to his or her external behavior and psychological states. Indeed, as I sarcastically note in several papers on Kripke's book, not only would access to someone's external behavior be enough to allow me to criticize someone's alleged rule following, it's unclear that the offer of access to one's psychological states is an offer of any import to determining the legitimacy of another's rule following. (Whatever is of import in the head is capable of being made public).
We should also not neglect the fact that none of us think that the rules we follow, learn, or even create, can be followed any old way someone chooses to follow them, including ourselves. Surely we regard ourselves as beholden to some technique, some results, etc., and hold that correct rule following is a matter of correct following of the techniques for following the rule, not giving answers that conflict with certain results, etc. In short, we do not regard ourselves as free to say anything at all and have it count as a correct way to follow the rule. Indeed, I can no doubt tell you all sorts of answers that I would rule out as being correct ways of following the rule. As such, if someone pays attention to my justifications conditions alone, s/he will find that I do not regard any old answer as acceptable under any old condition.
Of course, if you ask me to justify my way of following a rule, and furthermore require of an adequate justification that it be able to resist sceptical reinterpretation of the symbols used in setting out my understanding of the rule, then of course I will not be able to justify my way of following the rule. But this is just the sceptical paradox and applies to all people everywhere, whether ICIs or not. But even here it is clear that someone confined to my justification conditions alone would be able to say that I do not take 68 + 57 = 5 (and lots of other answers as well) to be a legitimate way of following the rule for '+', whether this can or cannot be justified to the satisfaction of the sceptic. The upshot is that while I am licensed to use '+', as it strikes me to use it (for example, to use it to mean quus or plus), I am not licensed to say 68 + 57 = 5 and claim that I am using '+' to mean plus. (Also, assuming that I have used '+' on numerous occasions so that it accords with using it to mean plus or quus, I am not licensed to say that I've been using '+' to mean minus, or times, etc.).
Finally, the last section of the passage, where KW claims that things are "very different if we widen our gaze to consider our rule follower interacting with a community", is befuddling. KW claims that because others will each have their own justification conditions for attributing correct or incorrect rule following to someone else, the subject's own authority is not unconditionally to be accepted. There are two problems here. One, the mere existence of competing justification conditions is cold comfort. That I can say of Jones that he is not following the rule correctly, would seem to be little more than hot air unless and until I can back up my claim by justifying the justification conditions that I am using to charge Jones with incorrect rule following. But allegedly I, an individual, cannot do this.
My justification conditions are as unjustified as Jones' and he could rightfully accuse me of not following the rule correctly. Needless to say, this does not deserve the title of substantive rule following talk. The second problem is more fundamental, viz., if I can arrive at justification conditions for attributing correct and incorrect rule following to Jones, and furthermore can tell when Jones is going right or wrong (by my lights, anyway) then I can surely do the same for myself. And I can then occasionally come to recognize that one or more of my responses are incorrect. Not everything I give as an answer to computation problem am I going to count as correct. So, confined to my own justification conditions, it is not true that I am left to call everything that comes out of my mouth, correct. So much then for the idea that confined to an individual's justification conditions, all we can say of an ICI is that s/he is licensed to follow the rule as it strikes him/her.
- If
someone whom I judge to have been computing a normal addition function
(that is, someone whom I judge to give, when he adds, the same answer I
would give), suddenly gives answers according to procedures that differ
bizarrely from my own, then I will judge that something must have happened
to him, and that he is no longer following the rule he previously
followed.
. . . From this we can discern rough assertability conditions for such a sentence as "Jones means addition by 'plus'. " Jones is entitled, subject to correction by others, provisionally to say, "I mean addition by 'plus'," whenever he has the feeling of confidence -- "now I can go on!" -- that he can give 'correct' responses in new cases; and he is entitled, again provisionally and subject to correction by others, to judge a new response to be 'correct' simply because it is the response he is inclined to give. These inclinations . . . are not to be justified in terms of Jones's ability to interpret his own intentions or anything else. But Smith need not accept Jones's authority on these matters: Smith will judge Jones to mean addition by 'plus' only if he judges that Jones's answers to particular addition problems agree with those he is inclined to give, or, if they occasionally disagree, he can interpret Jones as at least following the proper procedure. (K, pp. 90-1).
COMMENTARY: I'm not sure this passage needs my critique. It is pretty obvious that what Kripke says here is either illegitimate or otherwise fails to show how or why "the situation is very different if we widen our gaze from considerations of the rule follower alone and allow ourselves to consider him interacting with a wider community". However, a few comments anyway. First, how does KW justify saying of someone who "suddenly gives answers . . . ", that he is no longer following the rule he previously followed? That we would say this is clear enough but given the sceptical challenge, it's not clear that such a claim would or could be justified. For the person in question might tell us that his answer accords with the rule he was using all along, viz., quus. So for all we can tell, the person who "suddenly gives bizarre answers" is following the rule he previously followed. So much for the legitimacy of accusing him of not following the rule he was following previously.
Of course, there's also no justification for me to judge of another that s/he has been computing "a normal addition function" either; for all I can tell, s/he might be computing a non-normal addition function. For one's answers to addition problems are obviously compatible with computing a wide range of functions; surely we haven't forgotten the sceptical challenge already! Let no one under the influence of too much Kripke prose suggest that the reason it's legitimate for me now to claim that another person is computing addition or legitimate for me now to claim that another person is no longer following the rule s/he previously followed, whereas it wasn't legitimate before, is because we've already moved into the sceptical solution phase of the discussion, which means we've given up the search for facts about meaning in favor of assertion conditions.
While I know many who have actually given such a reply to my complaints, I know of none who have made the reply stick. The problem is that the sceptical challenge is based on considerations that are immune to the shift from truth conditions to assertion conditions. It is not, in fact, the desire or felt need for facts that leads to there being alternative interpretations of any past behavior involving, e.g., '+'. It is also not the desire or felt need for facts that lead us to hold that one cannot legitimately assert of someone who has yet to compute equations involving numbers greater than 56 that s/he means plus rather than quus. To be sure, we can, if we like, turn our backs on the infinitude of competing meanings compatible with any finite use of '+' we may have made in the past and simply decree that any circumstance in which someone has given answers in accord with addition constitutes a circumstance which legitimates asserting that s/he means plus by '+'. But we can and should appreciate both that such a maneuver not only has all of the advantages of theft over honest toil but also that a similar maneuver could also be made by fans of meaning facts as well.
As for the "rough assertability conditions" for meaning attributions (note: One will look in vain in Kripke's book for any refinement of these "rough" ACs), the only thing that keeps these ACs from being exactly like the ACs for an individual considered in isolation, and so the only thing that allegedly allows them to give substance to our rule following talk, is the phrase, "subject to the correction of others". This is ALL that the community allegedly has going for it that is unavailable to the ICI, that is, other people to "correct" him/her. However, as many critics have pointed out (in particular, see Ayer's and Gellner's comments), talk of correction by others here is bankrupt, since KW allegedly accepts the sceptic's case against meaning facts. So long as we buy the sceptical argument we can at most speak of agreement or disagreement of responses among members of a community. But clearly, that your answer is different from mine to some particular computation problem does not show that I am wrong and you are right, or vice-versa. Nor, clearly, does it show that I am not following the rule I followed previously.
As such, talk of one of us "correcting" the other is illegitimate. Since it is, so too is the idea that KW has, by "widening our gaze" to include community members, provided a way of giving substance to rule following talk which the ICI cannot have. In the ICI case, as well as the community case, we can challenge proposed answers to computation problems by pointing out their incompatibility with meaning particular functions (i.e., if someone says, 68 + 57 = 5, we deny him/her the status of a plusser, while allowing that s/he may be a quusser). But we never can rule out, in either case, that someone who seems to be a plusser, is really computing some other function whose difference from plus awaits a computation problem that we have yet to consider. (NOTE: Recall from my previous commentary that an individual is not forced to regard anything that comes out of his/her mouth as correct or as in accord with his/her rule. This certainly seems to be something KW believes, but he is wrong about it. It is, in some sense, KW's key mistake).
Finally, note that Kripke explicitly claims that Jones' meaning attributions to himself are not justified by appeal to his intentions. Recall however that on p. 88, Kripke told us that our inability to get at the ICI's intentions prevented us from being able to make sense of an ICI following rules. Clearly, intentions are a red herring in the discussion. That intentions do not and cannot justify meaning talk is obvious. As such, KW's move from, (i) we can't use an ICI's intentions to make meaning attributions about him/her, to (ii) we can't make ANY legitimate meaning attributions about an ICI, is bogus. If it were not bogus, a similar argument would allow us to conclude that all we can about anyone is that s/he is licensed to follow rules as s/he sees fit. Clearly, KW doesn't believe this.
- If
Jones consistently fails to give responses in agreement . . . with
Smith's, Smith will judge that he does not mean addition by 'plus'. Even
if Jones did mean it in the past, the present deviation will justify Smith
in judging that he has lapsed. (K, p. 91).
COMMENTARY: Of course, the mere fact that Jones' answers are at odds with Smith's does not warrant the claim that Jones does not mean addition by 'plus'. For it may be that Smith is not adding but quadding. The real problem here though is that Smith is obviously doing more than saying that Jones' answers differ from Smith's own answers. He is also saying that his (i.e., Smith's) own answers are the answers according with meaning addition by 'plus'. And Kripke never tries to explain how Smith can come to judge legitimately that he himself means addition by 'plus', nor, obviously, does he show that Smith needs other people in order to determine that he means addition by 'plus'.
Last
modified September 20, 2011
JAH, Professor
Dept. of Philosophy