On Kripke's Use of Hume's Sceptical Doubts and Sceptical Solution: Part I

I. Introduction

Kripke contends that Hume's sceptical anlace in sections IV and V of the Enquiry contains an argument for a conclusion that "might be called: the impossibility of private causation”.  By Kripke's lights, Hume's argument is as follows:

The whole point of [Hume's] sceptical argument is that the common notion of one event 'producing' another . . . is in jeopardy. It appears that there is no such relation as 'production' at all, that the causal relation is fictive.  After the sceptical argument has been seen to be unanswerable on its own terms, a sceptical solution is offered, containing all we can salvage of the notion of causation.  It is just a feature of this analysis that causation makes no sense when applied to two isolated events, with the rest of the universe removed.  Only inasmuch as these events are thought of as instances of event types related by a regularity can they be thought of as causally connected . . . .  When the events a and b are considered by themselves alone, no causal notions are applicable.  This Humean conclusion might be called: the impossibility of private causation.  (K, pp. 67-68).

Though several Kripke commentators have noted various difficulties with Kripke's reading of Hume, these alleged difficulties have focused on the exegetical issue of whether Kripke's Hume (hereafter, KH) does justice to the real Hume.  However, no one has worried about KH's case for the impossibility of private causation nor have they tried to extend KH's difficulties to KW (Kripke's Wittgenstein).  By my lights, KH's argument for the conclusion that private causation is impossible is hopelessly problematic, irrespective of its pedigree.  Furthermore, I believe that the problems with KH's argument can be quite readily extended to KW's case against private language.
Following Wittgenstein's advice to let the proof show us what is proved, I will briefly examine KH's argument in order to determine what he means by "the impossibility of private causation".  For, prima facie, the impossibility of private causation could be understood in quite a number of different ways, some more significant than others.  Ditto, of course, for the impossibility of private language, as anyone familiar with Kripke's book or the literature on this subject will attest.

KH's argument, which is contained in the quoted passage above, appears to go as follows.  First, he uses a sceptical argument (allegedly the sceptical doubts used by Hume in section IV of the Enquiry) to either cast doubt on the existence of productive relations between a cause and its effect or perhaps to establish the nonexistence of such relations.  Second, alleging the "unanswerability" of the sceptical argument (i.e., alleging that the sceptical argument does not admit of a straight solution), KH provides a sceptical solution to it.  Third, the sceptical solution offers a conception of causation that is capable of resisting the sceptical argument, the argument that caused trouble for the productive conception of causation.  Fourth, the new conception is seen to have at least one drawback, viz., it "makes no sense when applied to two isolated events, with the rest of the universe removed".  As such, "[o]nly inasmuch as [two] events are thought of as instances of event types related by a regularity can they be thought of as causally connected . . . .  When the events a and b are considered by themselves alone, no causal notions are applicable."  Now that we have the argument, let's examine it further in order to determine its import.

II. Problems with KH's sceptical argument against the productive conception of causation

To begin with, KH's case against productive relations is problematically vague. In particular, Kripke tell us that Hume's sceptical doubts put our common notion of causation "in jeopardy", that as a result of the sceptical doubts "it appears" there are no productive relations, and "it appears" that the causal relation is fictive.  (As Hume takes 'produce' to be synonymous with 'cause', I will henceforth speak of productive/causal relations, or P/C relations for short).  The question is: Is KH claiming that there are no P/C relations or is he claiming only that the existence of such relations is "problematic" because we have no real access to them? The latter seems the safest interpretation for several reasons.  First, insofar as KH purports to be echoing Hume, he ought not purport to have established the nonexistence of P/C relations. For Hume seems neither to need to show, nor intent on showing, that such relations don't exist.  On the contrary, there seems to be much in the Enquiry to suggest that Hume believes that there are P/C relations between events, although we silly humans have no access to them.  They are, as Hume puts it, secrets that nature has kept from us.  (On this matter, see Galen Strawson's The Secret Connection).

I think it's clear that our inability to find P/C relations is all that Hume needs to establish his main points, viz., whatever our knowledge of causal relations is, it arises entirely from experience, and its companion, reason and the understanding are not the source of our "causal conclusions", i.e., statements like, "Bread will always be conjoined with nourishment in those who eat it", or, "Fire will always be conjoined with heat", etc.  As far as these points are concerned, the question of the existence or nonexistence of P/C relations is by the way.  For our inability to discover such relations is enough.  In a word, Hume's epistemological focus allows him to ignore the ontological status of P/C relations.  That we can find no such relations is enough to get Hume's job done. KH, if he is echoing Hume, ought to be seen as claiming that we humans cannot find P/C relations, not that there are no such relations at all.

Another reason to see KH as claiming only that the existence of P/C relations is problematic rather than saying that they are nonexistent is simply that the first quoted passage nowhere explicitly claims that P/C relations don't exist.  The principle of charity counsels us to accept KH's claims "as is", and "as is", they do not explicitly claim that P/C relations don't exist.

However, there is a fly in the ointment here.  The fly is that the significance of KH's impossibility result seems to require him to show much more than simply that talk of P/C relations is problematic or that we humans cannot find any P/C relations.  The problem is that if there are P/C relations between events then causation is private, whatever humans may be able to say or think about the matter.  But since it's obvious that our inability to find P/C relations does not mean that there are no such relations, it's also obvious that KH, as we've interpreted him above, leaves open the possibility that there are P/C relations between events.  Thus, on the above interpretation of KH, he has offered us no reason to think that private productive causation is impossible, for he has not established that P/C relations don't exist.

Actually, things are even worse for KH.  For in order to establish the impossibility of private causation, KH must establish not only the nonexistence of P/C relations -- he must also show the impossibility of such relations.  So even if we see KH as claiming that P/C relations don't exist (and as the above makes clear, there is nothing to suggest that KH is making any such claim), he nowhere bothers to show that such relations are impossible.  So whatever else KH's impossibility of private causation claim says, it does not show that it's impossible for one event to produce another event without the help of any other events.  Needless to say, I find this to be a rather crushing blow to the import of KH's impossibility result. For it shows that KH has not threatened the possibility of what we all very naturally regard as private causation, viz., private productive causation.

Similar remarks can be applied to KW's case against private language. Whatever else KW has shown, it's difficult to see him as having shown the impossibility of private language in any significant sense.

More particularly, and contrary to a popular misconception of Kripke's book, KW does not show that it's impossible for a "pure" Robinson Crusoe, i.e., an individual who has lived an entirely solitary life, to speak a meaningful language or follow "substantive rules". As such, it's difficult to see what import there is to KW's case against private language.  The problem is as follows.  Since KW allows the community to "act at a distance" in providing substance to someone's rule-following (i.e., since Kripke allows that someone who was merely "physically isolated" from others could come, despite his/her physical isolation, to exhibit behaviors that we would or could judge to be rule following and using language, K, p. 110), it's clear that a solitary figure who was recognized to be a rule follower by some community was a rule-follower long before the community declared him to be a rule-follower.  In short, we shouldn't hold that a pure Crusoe whose behavior, once we are exposed to it, warrants our calling him a rule-follower, only becomes a rule follower upon our declaration.  This is like saying there weren't any natives in America until Europeans discovered them.  (See here my paper on Martinich on my "Assorted Problems in the Secondary Literature" page.).

Two further points of similarity between KH and KW.  Neither can be said to establish the "unanswerability" of their respective sceptical doubts (i.e., neither establishes that there can be no straight solution to their sceptical doubts).  As I noted in the Overview section of this site, it is no part of Hume's own sceptical solution to his sceptical doubts that those doubts be established to be "unanswerable".  (Cf. K, p. 66).  All Hume claims of his sceptical solution is that it accounts for our causal conclusions without contravening any part of his sceptical doubts.  I suggest that Kripke, in requiring that the sceptical solutions of KH and KW begin by granting the "unanswerability" of their sceptical doubts, is requiring too much.

Relatedly, and more importantly, even if we grant that there are no straight solutions to the sceptical problems of KH and KW, each draws an illegitimate conclusion from their respective sceptical solutions, that private causation and private language are impossible respectively.  The problem is that KH and KW have shown, at best, that their respective sceptical solutions make causation and language, "non-private".  But in order to justify the claim that private causation and private language are impossible tout court, KH and KW must show not only that there can be no straight solutions to their original sceptical doubts. KH and KW must also show that any sceptical solution to their respective sceptical doubts will echo their own solutions in requiring community aid for causation and language.  Needless to say, they no where bother to show any such thing, nor show any awareness of the need to do so.  Surprisingly, among the major commentators on Kripke's book, Donald Davidson, in his "The Second Person", is the only one to have noted this gap in KW's case against private language.

 

On Kripke's Use of Hume's Sceptical Doubts and Sceptical Solution: Part II

---

Ultimately, the problem with Kripke's appeal to Hume as providing an analogy with KW is that KW makes a variety of claims for which no analogues can be found in Hume.  In particular, KW's sceptic's sceptical doubts allegedly lead to bizarre and intolerable results, viz., "the concepts of meaning and of intending one function rather than another will make no sense" (K, p. 13), that "my answer [to 68 + 57] was an unjustified leap in the dark" (K, p. 15), that "nothing justifies . . . answer[ing a mathematical problem] one way rather than another" (K, p. 15; K, p. 21), that "I apply the rule blindly", (K, p. 17), that in my answers to computations I am "simply following an unjustifiable impulse" (K, p. 18), there can be no fact as to what I mean by 'plus', or any other word at any time" (K, p. 21), there is no fact about me that distinguishes between my meaning a definite function by 'plus' . . . and my meaning nothing at all" (K, p. 21), ". . . the entire idea of meaning vanishes into thin air" (K, p. 22),,"there can be no such thing as meaning anything by any word" (K, p. 55), "Wittgenstein's main problem is that it appears that he has shown all language, all concept formation, to be impossible, indeed, unintelligible" (K, p. 62).  Now, if we try to hold on to the analogy between KW and Hume, it seems we must find Hume saying similar things about causation as KW says about rule following and meaning.  In short, we ought to find Hume claiming that his sceptical doubts, or a sceptic's sceptical doubts, purport to show that the idea of causation vanishes into thin air or that causation is impossible.  However, Hume never says any such thing, not even, no how.

Just here, I think, is a crucial difference between the respective scepticisms of Hume and KW, viz., Hume never claims that his sceptical doubts threaten the idea that bread causes nourishment, or threaten our assurance that the next time we eat a piece of bread we will be nourished by it (or render that assurance impossible), or leave our causal conclusions without any foundation whatsoever.  On the contrary, Hume is challenging a particular view of causation, viz., the view that holds that our causal conclusions are, or can be, "founded on reasoning, or any process of the understanding."  Hume no doubt thinks it too obvious a point for him to stress that denying that our causal conclusions are founded on reasoning does not, does not mean that causation is impossible, that causal conclusions are without foundation, and so illegitimately drawn.

 In contrast, KW's sceptic makes an illegitimate move from his self-proclaimed successful sceptical attack on meaning facts to the bogus conclusion that the notion of meaning is illegitimate, unintelligible, and impossible.  At best, KW's sceptic shows that insofar as one's conception of meaning is such as to require the existence of meaning facts", there can be no meaning.  But surely every beginning student of philosophy is aware that undermining one conception of meaning is a far cry from undermining any and every conception of meaning.  Hume was certainly aware that bringing out problems with one concept of causation did not warrant any sweeping claims about the impossibility of causation or about the lack of any possible ground or foundation for our causal talk, our causal “language-game”.

This, I contend, is the reason KW's sceptical anlace seems so wrong and so tortured, viz., it contains an impossibility claim that is both illegitimate and unnecessary to his ultimate position.  By dropping the impossibility of meaning claim, we get a more sane, albeit less exciting or controversial, KW.  At best, all KW's "sceptical doubts" really show is that our language game of meaning, appearances to the contrary notwithstanding, makes no use of, indeed, has no need for, "meaning facts".  The proper conclusion to draw from this is not that language or meaning is impossible but simply that a concept of meaning which relies on "meaning facts" cannot be our concept.

The job of KW's sceptical solution then would not be that of solving the impossibility of meaning, since no such conclusion has really been established.  (As I remark elsewhere, the impossibility of meaning would be impossible to solve for anyone who, like KW, accepts the nonexistence of meaning facts).  Clearly, we need to appreciate that KW's sceptic fails to appreciate that he has, at best, undermined only one conception of meaning.  Most importantly, KW's sceptic has at best undermined a conception of meaning different from the one we actually use!  Thus, the job of the sceptical solution would become the manageable one of showing how our notion of meaning gets by without appeal to "meaning facts".

Nicely enough, if KW would drop his sceptic's bogus claim that meaning is impossible, or change it to the claim that one conception of meaning has been shown to be impossible, the analogy with Hume is rendered more plausible.  For both KW and Hume would then be seen as asking how we manage to do such-and-such, without ever claiming that their sceptical doubts threaten the possibility, the intelligibility, or the legitimacy, of undeniable features of our life, viz., causation and meaning respectively.  To be sure, there is a price to pay in making KW and Hume genuinely analogous.  For doing so requires appreciating that KW's sceptic has no case at all for claiming that meaning is impossible, unintelligible or unjustifiable, which in turn allows us to appreciate that there is really no threat to rule following or meaning to be found in Kripke's book.

Top of Page  |  Problems and Projects Page |  WRPL Ultimate HomePage

Last modified October 5, 2012
JAH, Professor
Dept. of Philosophy