The Intelligibility of Abortive Omniscience



Stig Alstrup Rasmussen

The Philosophical Quarterly, Vol. 37, No. 148 (Jul., 1987),315-319.

Stable URL:

The Philosophical Quarterly is currently published by The Philosophical Quarterly.

Your use of the JSTOR archive indicates your acceptance of JSTOR' s Terms and Conditions of Use, available at JSTOR's Terms and Conditions of Use provides, in part, that unless you have obtained prior permission, you may not download an entire issue of a journal or multiple copies of articles, and you may use content in the JSTOR archive only for your personal, non-commercial use.

Please contact the publisher regarding any further use of this work. Publisher contact information may be obtained at http://www .j stor .org/journals/philquar .html.

Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transmission.

JSTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding JSTOR, please contact

http://www .j Wed Nov 912:31:182005



Despite some reservations about possible-world terminology, Hare has tentatively endorsed what Jaegwon Kim calls "weak supervenience" and rejected what Kim calls "strong supervenience"." Kim's weak supervenience is equivalent to what McFetridge, in his Appendix, labels (xyww), and strong supervenience is equivalent to (XYWW'). Hare is attracted to Kim's weak supervenience for the following reason: Strong supervenience, by governing the connection between the natural and the moral for objects in any possible worlds, makes moral principles necessary, while weak supervenience, by governing the connection for objects only within a given possible world, allows moral principles to be contingent.

But it seems to me to be a mistake for Hare to endorse any notion of supervenience formulated in terms of possible worlds. Kim's weak supervenience is at once too weak and too strong for what Hare wants. When one endorses a moral principle, it applies not only to actual cases, in the given possible world, but also to hypothetical cases, sufficiently similar, in any possible world. II Kim's weak supervenience is too weak to capture this intuition, for it only governs cases in the given possible world. On the other hand, as I have argued above, Kim's weak supervenience is too strong because it constitutes a collective rather than a distributive condition on moral judgement within a single possible world.

The distinction between weak and strong supervenience, focussing on one as opposed to all possible worlds, is not the same as the distinction (that Hare should want) between distributive and collective construals of supervenience. It may be that Hare (and others) thought these distinctions were equivalent, but they are not.

Virginia Polytechnic Institute

and State University



According to a forceful argument repeatedly put forth by Donald Davidson, there can be no communication without a shared world picture, I That is,

II Kim's formulations are found in his "Concepts of Supervenience", Philosophy and Phenomenological Research 45 (1984), pp. 157-67. The same terminology and roughly the same formulations were independently introduced by me in "An Alleged Difficulty Concerning Moral Properties", where I labelled them (WSS) and (SSS). The distinction was suggested to me by some ideas of Rogers Albritton. Hare tentatively endorses weak supervenience on p. 4 of "Supervenience". See also Kim, op. cit., p. 161, note 13.

IZ Recall the passage from Hare's Language of Morals quoted above.

I Several of the essays reprinted in D. Davidson, Inquiries into Truth and Interpretation (Oxford, 1984) are relevant. Cf. in particular "Radical Interpretation", "Belief and the Basis of Meaning", "On the Very Idea of a Conceptual Scheme", and "The Method of Truth in Metaphysics" .



(I) Interpretation requires widespread factual agreement.

Davidson bases this claim on considerations pertaining to what he calls the theory of interpretation, notably the operation of the Principle of Charity as applied by the radical interpreter engaged in devising a theory of meaning for a language, while at the same time striving to make sense of the cognitive and affective attitudes of the speakers of that language. For the purposes of the present discussion, I have no quarrel with (I), or with the inextricability argument Davidson adduces in support of that thesis.' What I wish to question is the rather spectacular "extended claim" that not only does communication require widespread agreement about the nonlinguistic world: a shared language also requires that the presupposed world view must, at least "in its large features", be true? My contention is that Davidson offers no convincing argument in support of the extended claim. In this I agree with Bruce Vermazen, among others; but my argumentative strategy is rather different from his.'

Davidson takes pains to point out that the observation that "agreement, no matter how widespread, does not guarantee truth" in no way affects the acceptability of the extended claim: "this observation misses the point of the argument".' In the context, this remark strongly suggests that Davidson takes the argument from the inextricability of meaning and belief to establish, by itself, that communication presupposes the (near) truth of the shared world view required for communication. But the thesis that false beliefs can be attributed only against a backdrop of attributions of true factual beliefs is an immediate consequence neither of the constraint that factual agreement be maximised as far as possible, nor of Thesis (I). I shall take it, then, that the extended claim rests not on the inextricability argument as such, but on an extension of that argument.

Now Davidson in fact goes on to offer such an extension. The argument for the

extended claim is embodied in the following passage:'

We do not need to be omniscient to interpret, but there is nothing absurd in the idea of an omniscient interpreter; he attributes beliefs to others, and interprets their speech on the basis of his own beliefs, just as the rest of us do. Since he does this as the rest of us do, he perforce finds as much agreement as is needed to make sense of his attributions and, interpretations; and in this case, of course, what is agreed is by hypothesis true.

This argument would seem to have just two premisses, viz. (I) and (II):

(II) The notion of there being an omniscient interpreter is intelligible;

2 I borrow the term 'the inextricability argument' from M. Dummett who, in "The Significance of Quine's Indeterminacy Thesis", Truth and Other Enigmas (London, 1978) pp. 387-8, discusses what he calls Quine's "inextricability thesis".

3 D. Davidson, op. cit., pp. 199-200.

4 B. Vermazen, "The Intelligibility of Massive Error", The Philosophical Quarterly, 33 (1983) pp.69-74.

5 D. Davidson, op. cit., p. 200. 6 Ibid., p. 20l.


and Davidson has it that (I) and (II) entail the extended claim. But this will not work. Firstly, (I) and (II) cannot possibly support a conclusion stronger than:

(III) It is intelligible that an omniscient being should interpret us in a

manner such as to attribute to us a largely true picture of the world.

Surely, the intelligibility of our not being massively wrong in no way entails either that we are, as a matter of fact, by and large right; or that it is unintelligible that we should be massively wrong.

But, secondly, the real flaw in Davidson's argument is that (1) and (11) do not even entail (III), unless we can take it for granted that

(IV) The omniscient being's attempt at interpreting us would meet with success.

(IV) is however far from an innocent assumption. This may be gleaned from the fact that, if we are in fact massively wrong about the features of the world, (I) and (II) combine to rule out (IV). Disregarding here the first worry pointed out above, (IV) and the extended claim are, in the presence of (I) and (II), roughly equivalent. To get an independent and non-question-begging argument going along the suggested lines, Davidson therefore needs an underpinning of (IV) which does not pass through the extended claim. For unless such an argument is adduced, his line of reasoning will support only the rather lame disjunction of the extended claim and the negation of (IV): either the fact that we communicate ensures that we share a largely true world picture, or an omniscient being would be debarred from interpreting us.

There might be a temptation to think (IV) a more or less trivial consequence of the omniscience enjoyed by the being introduced in the course,bf Davidson's argument. However, omniscience must not, in the present context, be taken to imply privileged access to what the target language speakers mean by their words, or to what they believe. What is in question is omniscience exclusively about the features of the world about which the target language speakers communicate. As Vermazen points out, Davidson's'

. . . omniscient interpreter doesn't know what the speaker's beliefs are:

"He attributes beliefs to others ... just as the rest of us do". He is omniscient about everything but the speaker's mental economy. If his omniscience weren't thus limited, Davidson's argument couldn't get going, for the interpreter would know what the speaker's words meant, and so wouldn't have to employ the Principle of Charity.

One might raise doubts as to the coherence of the kind of limited omniscience required of the Davidsonian being: granted the presence of human agents in the environment, could there really be knowledge of all facts save the ones about the beliefs of those agents, and the ones about their having associated particular meanings with their linguistic expressions? That is, in the light of the above, there is a serious threat that (II) may fail. I shall however not pursue this line of attack. The

; B. Vermazen, op. cit., p. 71.



salient point is that, given Davidson's intended construal of 'omniscience', the thought that (IV) follows trivially from the omniscience possessed by the Davidsonian being will seem plausible, only if we invoke a caveat which usually attends Davidson's presentation of the inextricability argument for (I). The caveat is suppressed by Davidson in his argument for the extended claim, and is to the effect that the agreement required for communication must not be taken to rule out what the interpreter can discount as explicable error, on the part of his interlocutors. Bearing this caveat in mind, one may plausibly hold on to (IV): an omniscient being, even in the restricted sense, would be in a singularly favourable position to locate explicable errors on the part of others. It would seem, then, that the omniscient being would be unlikely to fail in a sustained attempt to interpret us.

The trouble with this line of defence is that it is not open to Davidson, since (IV) is now saved at the cost of undermining the otherwise plausible (I), as applied to the case under consideration. The reason why there is, on the present suggestion, a virtual guarantee that the omniscient being would succeed in his interpretative endeavours is that, no matter how wrong we might explicably be, the being would locate and explain our errors. In consequence, (IV) is ensured by omniscience at the cost of jeopardising (I). By the same token, what remains of the original argument no longer supports (III). The picture now emerging is that of an omniscient being who successfully interprets us as being, by and large, wrong, albeit explicably so. I submit that this picture would seem to be perfectly intelligible. If it is, Davidson must be wrong in concluding as follows:'

But now it is plain why massive error about the world is simply unintelligible, for to suppose it intelligible is to suppose there could be an interpreter (the omniscient one) who correctly interpreted someone else as b~g massively mistaken, and this we have shown to be impossible.

Nothing of the kind has been shown. Davidson's argument will work only if (IV) is supported; but, given the intended reading of 'omniscience', (IV) seems capable of support only by appeal to considerations to do with "explicable error", and that appeal forces the admission that an omniscient interpreter could interpret us as being massively mistaken. Furthermore, it has transpired that there are more ways than the one pointed out by Davidson in which to give content to the supposition that massive error is intelligible. For instance, there might be a dedicated omniscient interpreter who failed to interpret us. This might be because we are inexplicably, as well as massively, mistaken; but the intelligibility of there being such an interpreter would have to be taken seriously, too, if we found reason to join Davidson in suppressing, in the case of the omniscient interpreter, the caveats usually qualifying his argument for (I). Finally, since Davidson's argument misfires, we are left with no ground for thinking unintelligible the more pedestrian possibility that there might be a massively mistaken interpreter who wrongly interpreted speakers of the target language as being massively right about factual matters.

Finally, I wish to suggest that invoking an omniscient being is the wrong sort of move, if the object is to get from (I) to the extended claim. For support we go along

x D. Davidson, op. cit., p. 201.


with the inextricability argument and thesis (I). There is then no denying that, in so far as we manage successfully to communicate with others, there is no such thing as an interpretaton of our fellow-speakers involving the imputation to them of massive error. The obvious way to get from here to the extended claim is to invoke a "verificationist" premiss to the effect that to the extent that we are, as a matter of principle, debarred from establishing that someone else is mistaken, there can be no content to the suggestion that he might be. It may deserve note in passing that this kind of verificationism ought in any case to strike the adherent to the inextricability argument as congenial: that argument relies heavily on disallowing facts about mental content and linguistic meaning not capable of some form of .public manifestation. However that may be, Davidson's introduction of the

omniscient being is a move in the opposite direction; for the effect of appealing to omniscience is exactly to set up as intelligible a standard of truth such that, even where we have no purchase on the answer, there is a genuine question as to whether or not we go wrong in interpreting others as being massively mistaken or, by and large, right about the world:

University of Copenhagen



One of the fundamental aspects of Donald Davidson's account of the everyday activity of communication by language is the fact that he describes speakers of a language as theorizing interpreters. Indeed, his belief that the concept of interpretation can be used to explain linguistic understanding leads him on occasion to use the terms 'interpretation' and 'understanding' as if they were interchangeable; he can say, for example, that " ... what is essential to my argument is the idea of an interpreter, someone who understands the utterances of another".' (ITI, p. 157) In addition to this, he refers to anyone who successfully interprets utterances as the possessor of a "theory" of interpretation. The philosophical theory of meaning Davidson offers is intended to model the precise form this everyday theorizing must take; we must see the interpreter as the possessor of just such a theory if we are to make sense of what he does.

To see how the notion of 'interpreting' paves the way for a correspondingly wide use of the notion of 'theorizing', we need only look at certain features of our standard employment of the term 'interpretation'. We speak of interpreting a dream, or a word in a foreign tongue, or a piece of technical jargon; in all these examples, the process of interpretation is linked with a transition between languages or between sectors of the same language: the point is to clarify meaning by giving it

, All references to Davidson's writings will be to the collection Inquiries Into Troth And Interpretation (Oxford, 1983), henceforth to be known as ITI.

Sign up to vote on this title
UsefulNot useful