Indeterminacy and Assertion


1. Introduction

This paper will appeal a recent argument for the indeterminacy of translation to show not that meaning is indeterminate, but rather that assertion cannot be explained in terms of an independent grasp of the concept of truth. In particular, it will argue that if we try to explain assertion in terms of truth rather than vice versa, we ultimately will not be able to make sense of the difference between assertion and denial. This problem with such 'semantic' accounts of assertion then illustrates why we need not worry about the purported argument for indeterminacy.

We tend to assert those things we take to be true, and even when we lie, we at least present our utterances as true. It is thus natural see a close conceptual connection between truth and assertion. This close connection can be formulated, for any speaker S and sentence P, as:


S asserts P <=> S presents P as true.


This biconditional presents us with the possibility of understanding either of its sides in terms of the other. [1] We can call these two explanatory possibilities the semantic account of assertion and the pragmatic account of truth. The pragmatic account starts with an account of the speech act of assertion and tries to abstract semantics (content) from it, while the semantic account starts with a theory of truth and tries to connect pragmatics (force) to it. The two approaches can be outlined (roughly) as follows:


The semantic account of assertion.

Step one: Give an independent account of truth.

Step two: Explain assertions as utterances that purport to express the truth.


The pragmatic account of truth.

Step one: Give an independent account of assertion.

Step two: Explain truth in terms of successful assertions.


The pragmatic account has generally been taken to have trouble with the second step: the understanding, or replacement, of the concept of truth with that of (idealized) successful assertion. [2] The first step (coming up with an independent account of assertion) is hardly trivial, but it has generally attracted less criticism than the second. On the other hand, for the semantic account, the first step has usually been the most controversial. Independent accounts of truth, whether in terms of coherence or correspondence, have been surprisingly hard to come up with. [3] Comparatively little attention has been given to the second step, that is, explaining assertion in terms of truth. It has often been assumed that, once we have a notion of truth on board, it will simply be a matter of course to explain concepts like assertion and belief in terms of it. [4]

This paper will argue that this second step of the semantic account is not nearly as unproblematic as it might initially seem. In fact, it may pose more serious problems for the semantic account than the first step. Indeed, the problem that this step in the semantic account of assertion faces is, in many ways, simply a mirror image of that which faces the second step of the pragmatic account of truth. Namely, just as the pragmatic account of truth has trouble explaining how our assertional practices can constitute something as apparently response independent as truth, the semantic account of assertion has a problem explaining just how something as transcendent as non-pragmatic truth can actually get a grip upon our linguistic practices. With the semantic account, whether a certain utterance counts as an assertion becomes, in principle, independent of any fact about the speaker's non-semantically specified behavior. [5] If the force of what we say is determined by our attitude towards independently specifiable semantic values (and cannot simply be equated with specific modes of behavior), it will always be at best a hypothesis that people are making utterances with assertive force (albeit a hypothesis that fits our behavior extremely well). [6]



2. Force and the Massey permutation.

A serious consequence of treating force as conceptually derivative from content is highlighted by Gerry Massey's argument for the indeterminacy of translation. [7] Massey's argument illustrates that our interpretation of an entire language could be drastically permuted provided that one assumes an independent grasp on the concepts of truth and falsity. [8] Massey considers a "first-order quantificational language with identity, alethic modalities, deontic modalities, and prosentences, as well as interrogative, assertive, rejective, imperative, and preventative forces" [9] (which we will call "Massey-English"), and then describes two ways in which it could be translated. The first is simply a homophonic translation (manual h). The second is a "dual manual" (manual d), which Massey describes as follows:

First, in the sentence-mapping dimension, d translates each operator (sentence connective or quantifier) and prosentence by its dual, i.e. conjunction by disjunction, necessity by possibility, prohibition by permission, universal quantification by existential quantification, and 'Yes' by 'No' (and vice versa in each case of course). Negation is self-dual, so d translates '-' by itself. The manual d also translates each predicate by its complement, e.g., 'R' by '#R' (non-R). [10]

As a result, a sentence in Massey-English such as "All rabbits are vegetarians" will be translated by the homophonic manual as "All rabbits are vegetarians" and by the dual manual as "Some rabbits are non-vegetarians." The dual manual will translate each sentence of Massey-English as what is, intuitively, its contradictory. Such a manual would be radically out of touch with the linguistic behavior of the speakers of Massey-English were it not for the fact that the dual manual treats assertion as denial, so when the homophonic manual treats a speaker as asserting that all rabbits are vegetarians, the dual manual will treat a speaker as denying that some rabbits are non-vegetarians. The dual manual thus seems to be able to account for the speakers' utterances just as well as the homophonic one.


Crucial to Massey's attempted proof of indeterminacy is, of course, the claim that the force of the speaker's utterances is up for grabs in just the same way that their meanings are. Massey gives, as evidence for this claim, a friend of John Searle's who "had gone through life believing that English speakers use the borrowed Greek expression 'hoi polloi' to refer to the few, i.e. to the elite." This mistaken belief did not hamper his ability to communicate because he also "believed that English speakers typically use sentences containing the expression 'hoi polloi' ironically." As a result, neither Searle's friend nor his interlocutors detected the different referents assigned to "hoi polloi" by the speaker and his friends. "Though the one takes the rabble and the other takes the elite to be the referent of the expression, this semantic discrepancy is exactly offset, and so masked, by the different forces -- literal versus ironic -- that they assign the customary use of sentences containing 'hoi polloi.'" [11] Massey takes the moral of this story to be that "difference of force can offset difference of meaning." [12] Massey further argues that while the "hoi polloi" example of content/force compensation is for a single word, one can have entire languages based upon this principle. Indeed, he claims that there is one such language, "Upside-Down Walbiri", which he characterizes as follows:


The male rites of passage of the Walbiri, an indigenous people of central Australia, incorporate the dualization that characterizes the [dual] manual. Boys undergoing initiation are placed in a ceremonial hut with adult males who while in the hut speak only the language they call "Upside-down Walbiri." Upside down Walbiri results from Walbiri by the systematic interchange of expressions with their opposites (antonyms). For some days these linguistic goings-on bewilder the boys. But one by one each boy undergoes a eureka experience after which he understands and speaks Upside-down Walbiri effortlessly. [13]

Massey thus takes the sort of inversion he has in mind thus not to just be a theoretical possibility, but also something that actually happens

That the force of an utterance is also something the interpreter must determine is a point that Massey takes Quine to have missed:

At times Quine seems to believe that force is determinate in a way that sentence-mapping translation is not, but this position is incoherent. What speakers mean by what they say is never given by a sentence alone, but by a sentence together with a force. The candidates for translation -- the arguments of translation functions -- are not forceless sentences but rather sentences imbued with force. [14]

According to Massey, the radical interpreter is simply faced with a set of sentences, and the force with which they are uttered is just as much up for grabs as their content. However, Massey's denial that "force is determinate in a way that sentence-mapping translation is not," presupposes a semantic account of assertion. Consequently, if the semantic account of assertion is correct, Massey's argument may show that translation is indeterminate. On the other hand, those sympathetic with the pragmatic accounts of truth and assertion may be more inclined to take Massey's argument to simply constitute a reductio of the semantic account of assertion and the resulting idea that "that utterance typically has the force of assertion (or rejection) is an analytical hypothesis." [15] This latter position will be the one developed in what follows.



3. Winning.

To see how Massey's argument contributes to a reductio of the semantic account of assertion, it may be worthwhile first to look at the structurally similar (though simpler) case of winning and losing games. Knowing how to, say, play chess involves knowing more than simply how the pieces can be legally moved. Someone who knew all those rules, including that the person who put his opponent in checkmate 'won,' but did not know that the object of the game was to win, would not be able to play chess properly. [16] This parallel suggests that we can see something like a force/content distinction in games as well as in languages, and the connection between truth and assertion could be seen as having the following correlate for any player S:

S is competing <=> S is trying to win.

This biconditional gives one either the option of trying to understand winning in terms of competing (winning just is whatever people compete for -- the pragmatic picture) or understand competing in terms of winning (competing is just trying to win -- the semantic picture).

The semantic picture, where winning (and thus losing) are taken to be independently intelligible, allows the concept of anticompeting (trying to lose) to be on just as good conceptual footing as competing (just as the semantic account of assertion leaves it on equal footing with denial). If one accepts the semantic account of competing, a case analogous to Massey's dual manual could be constructed for the interpretation of a pair of chess players. We could either interpret them as playing a game in which one won if one placed one's opponent in checkmate (the 'homophonic' game), or as playing "anti-chess" in which one won if one was placed in checkmate by one's opponent. Both manuals match the players' behavior equally well because while the first treats the players as trying to win (competing), the second treats the players as trying to lose (anti-competing). A difference in games played is compensated for by a difference in the object of playing the game. [17] One could then, presumably, come to have such doubts about how to interpret one's opponent, and, ultimately, oneself. How, then, do we really know that we are trying to win rather than trying to loose?

However, it seems doubtful that the 'dual manual' for chess players represents an intelligibly distinct alternative to the homophonic one: the indeterminacy results make us lose our grip on the idea that we ever really had an independent understanding of what winning and losing were. Winning isn't of any value if isolated from succeeding at what you are trying to do in a particular gaming situation (winning if you are competing, losing if you are anticompeting). When detached from the pragmatics of competition then, there seems to be no intelligible difference between winning and losing. [18] Competing at Chess and anti-competing at Anti-Chess are supposed to be different practices, but for them to be so, one needs to give a substantive account of the difference between winning and losing, but once winning and loosing are divorced from any connection to human behavior, it is hard to see where such an account could be found.

It is worth remembering that we can teach people how to play games without making any reference to the concept of winning. One plays chess by trying to put one's opponent in checkmate, one plays baseball by trying to score more runs than one's opponent, one plays hearts by trying to 'score' fewer points than any of one's opponents, etc. One could successfully play any of these games without knowing that what one was trying to do was "winning" and one could presumably have a culture that played all sorts of games, but had no word in their language corresponding to "winning." Still, each game has a preferred outcome, and one could introduce into the language a term that picked out all and only those preferred outcomes ("let's call what we try to do in each game 'winning'"). Such a term, once introduced, would be quite useful (for instance, it can make talking about games simpler ("I think the Pirates will win" vs. "I think that the Pirates will score more runs than their opponent"; "Whatever game he plays, he always wins" vs. "when he plays chess he puts his opponent in checkmate, when he plays hearts he scores the fewest points, ")). With such a term in play, games can be specified independently of any explicit reference to what any of the players actually try to do (since such a reference is implicit in the concept of "winning" itself). However, once games are specified in this way, it can be tempting to think that "winning" picks out an independently intelligible property. Consequently, one may come to think that certain outcomes are preferred because they have this property, and that one could coherently describe certain practices where this property was never preferred (indeed, where it was dispreferred). These imagined practices would correspond to inverted games, and the intelligibility of such games rests on the assumption that "winning" picks out an independently intelligible property. Such an assumption, of course, sits badly with the genealogy suggested above.

While the concept of winning allows one to specify games with no explicit reference to what players try to do, the implicit reference remains. As a result, there can be no purely 'semantic' description of games that have no reference to what players try to do, and no coherent description in which players always try to lose. [19]

Consider again the person we are trying to teach how to play chess. As stated before, teaching him the rules about how the pieces can be moved is certainly not enough, and telling him that the person who puts their opponent in checkmate is the 'winner' need not help either. If the player sees no reason to want to be called the "winner", he may try to win, lose, make the game continue as long as possible, see to it that there are as few as possible pieces left on the board, or whatever. Yet the rules for moving the pieces and the stipulation that the person who places his opponent in checkmate wins seem to be all that is involved in the 'semantic' characterization of chess. It might then seem that, in order to learn how to play the game, the player needs to be taught an additional rule, namely, that each player should try to be the "winner." However, this last 'rule' is not plausibly considered a rule of chess. Rather, it is at best a clarification of what it is to be a winner in a game -- if one doesn't know that one is supposed to try to win, then one doesn't know what winning is. [20] It is because we have this general understanding of what winning is that lists of rules need only state what is required to win, not that one should try to win as well. [21] It seems, then, that an interpretation in which everyone openly tried to lose would not be a coherent one (it would involve everyone trying for the allegedly dispreferred outcome). One could imagine a society like ours in which everyone tried to do what was called "losing", but that would only give us reason to treat their word "losing" as meaning the same as our word "winning."

Of course, given our general practice of trying to win games, we can certainly make sense of someone trying to lose a game. For instance, we can imagine a skilled chess player deliberately playing badly so that he will not be put in the embarrassing position of having beaten his boss. However, even in such cases, the player is still at least pretending to try to win. If he were too obvious in his attempts to lose, the boss would probably be even more upset with him. A particular game of chess in which both players were openly trying to lose would be almost unrecognizable, [22] and, indeed, would almost inevitably result in a draw. [23] This suggests that our idea of trying to lose makes sense only against a general practice of trying to win, and thus the fact that there are particular cases of trying to lose does not imply that there could be a practice in which everyone always tried to lose without even making a pretense to try to win. [24]


4. Assertion and the Dual Manual

Just as the resultant possibility of anti-competing at anti-chess suggests that we don't have a grasp of winning and loosing independently of our grasp of competing. Massey's manual should suggest that we don't have a grasp of truth and falsity independently of our grasp of successful assertions. Just as the actual practice of competition is what provides the traction that allows us to distinguish winning from loosing, the actual practice of assertion is what provides the traction that allows us to distinguish truth from falsity. What Massey shows is that if we take truth and falsity as primitive, and thus treat it merely as an "analytical hypothesis" that our linguistic practice is one of assertion rather than denial, assertion and denial, truth and falsity, each become impossible to distinguish from one another. Quine is precisely right to assume that "force is determinate in a way that sentence-mapping translation is not." There are certain sorts of behavior which we can simply characterize as asserting, [25] and truth is understood (though perhaps not defined) in terms of that. [26] What Massey shows is that if we divorce assertion from any such direct behavioral connections, then it will ultimately be indistinguishable from denial.

It has been argued here that the fact that it makes room for the type of dual manual that Massey describes can serve as reductio of the semantic account of assertion. [27] The claim that such a manual provides a reductio presupposes that we cannot, ultimately, make sense of the possibility of such dual manuals. However, Massey argues that permutations of the sort he describes are not only possible, but they actually happen from time to time. The two most prominent examples he gives are that of the ironist and "Upside Down Walbiri", but upon closer investigation neither of these turn out to be real examples of such dual manuals.

For instance, the ironic use of "hoi polloi" that Massey appeals to is clearly parasitic upon general practices of non-ironic assertion, [28] and it is far from clear whether it could be generalized. For instance, while Searle's friend could use sentences containing "hoi polloi" ironically, and while for certain utterances the question of whether the speaker was being ironic might come up, this could not generally be the case. The literal-meaner and the confused ironist might apply the phrase "hoi polloi" to the same groups of people, but there is plenty of other evidence that will separate the two. If one sincerely asserts a sentence using the term 'hoi polloi' one may defend it if challenged, while the ironist will usually claim to be misunderstood if his utterance is challenged in the same way. [29] The difference in force will correspond to a difference in how the speakers react to contradiction, requests for clarification, etc. Because of this, the case of irony points to, at best, a need to consider a wider range of behavioral evidence when determining force, not a type of indeterminacy. A better example might then be lying, since, unlike the ironist, the liar will be inclined to defend his utterance if challenged etc. Unfortunately, lying is a type of assertion, and while it makes sense against a background of sincere assertion, a practice in which everyone lied all the time and everyone knew it would be unworkable. [30] We can 'throw' a game while pretending to win, and we can occasionally lie while pretending to tell the truth, but just as a gaming practice cannot explicitly revolve exclusively around losing, one could not have a non-parasitic linguistic practice explicitly based upon lying.

A more serious case is "Upside-Down Walbiri", since Massey claims that it shows how an entire language (not just a small subset of a standard one) could be characterized by a dual manual. However, the claims Massey makes about Upside-Down Walbiri turn out to be almost entirely unsupported by the source he cites for them. [31] In particular, there seems no reason to say that Upside-down Walbiri "incorporates the dualization that characterizes the [dual] manual." Not only is there no evidence for any dualization of logical connectives, quantifiers, or force (indeed, imperative and assertive force are clearly preserved), but the 'antonyms' in Walbiri are nothing like Massey's predicate complements. [32] That "Upside-Down Walbiri" is not characterized by anything like the dual manual for Walbiri should not be surprising. Indeed, if such dualization were to occur, the results should be indistinguishable from Walbiri, which would defeat the purpose of having a special language.


5. Conclusion

If assertions were understood in terms of truth, the sorts of indeterminacy that Massey has in mind would be a coherent possibility. They are not, however, a coherent possibility, since their existence would undermine our grasp on the concept of truth itself. Consequently, the type of cases that Massey focuses on show not that meaning is indeterminate, but rather that the semantic account of assertion (and thus Massey's argument for indeterminacy) is dubiously coherent. This alone is hardly enough to establish the alternative, pragmatic, approach to truth and assertion, but it provides further motivation for trying to develop the pragmatic line.




1. One could also, of course, take the two to be interdefined.


2. This can be done either through some sort of identification, as in the theories of Peirce, James, and more recently Putnam (Peirce 1878, James 1907, Putnam 1981), or in terms of the pragmatics of truth talk, as in various redundancy or 'prosentential' theories which can be understood as denying that "true" picks out a property (Grover, Camp and Maxwell 19??, Brandom 1994).


3. Failures at this first step in the semantic account have often been what drove people towards the pragmatic account.


4. Davidson, for instance, helps himself to the concept of truth and simply defines belief and assertion in terms of 'holding true' (Davidson, 1984). Lewis characterizes languages in purely semantic terms and provides the following answer to the problem of "what the members of a population P must do in order to make it the case that a certain possible language £ is their actual language"( Lewis 1969, p.177.)


The convention whereby a population P uses language £ is a convention of truthfulness and trust in £. To be truthful in £ is to act in a certain way: to try to never utter any sentences of £ which are not true in £. To be trusting in £ is to form beliefs in a certain way: to impute truthfulness in £ to others. (Lewis 1972, p 167.)


That we can characterize speakers as being truthful and trusting is taken to be fairly unproblematic. Indeed, Lewis characterizes the interpretee's utterances as assertions through the "principle of truthfulness" that states that speakers tend only to utter sentences in their language that they take to be true.


5. I.e., other than its being understood as aiming at the truth.


6. Lewis' principle of truthfulness actually seems like more than a well confirmed hypothesis. He claims that "the fundamental principles of our common-sense theory of persons implicitly define such concepts as belief, desire and meaning," and that if such fundamental principles (which include the principle of truthfulness) are made explicit, they would "have a status akin to analyticity." (Lewis 1974, p. 112.) This might suggest that Lewis' account is, in some sense, closer to the pragmatic. If the principle of truthfulness is taken to be analytic, then it is analytic that our assertion-like behavior really is assertion, and not denial. Identifying assertion with assertion-like behavior (i.e.: assertion just is a certain type of verbal behavior), which is essential to the pragmatic account, amounts to making the principle of truthfulness 'analytic,' so Lewis's methodology could be understood as involving a type of backhanded commitment to the pragmatic account.


7. Most recently presented in Massey 1992.


8. Though the language in question is still an artificial language and represents only a fragment of any natural language such as English.


9. Massey 1992, p.333.


10. Massey 1992, p.333.


11. Massey, p. 331. The example can be found in Searle 1987.


12. Massey, p. 331.


13 Massey p. 338.


14 Massey, p. 332.


15 Massey p.341. Furthermore, it should be noted that Massey's indeterminacy would be a reductio of the semantic account of assertion in a much stronger way than Quine's "Gavagai" could be understood as a reductio of the radical interpretation process. Quine's indeterminacy thesis was not taken as a simple reductio of his methodology because the methodology (all there is to meaning must be available to an interpreter) seemed well motivated (considerations of publicity, the tie between meaning and communication, etc.) and did not have an appealing alternative (the inability to provide any plausible alternative being the main problem with many criticisms of Quine's methodology). On the other hand, the semantic account of assertion is less well motivated, and the alternative pragmatic account has been motivated and developed quite independently of any concerns about avoiding indeterminacy. Because of this, the temptation to take indeterminacy simply as a consequence rather than a counterexample is much less tempting in Massey's case than it is in Quine's.


16 For a discussion of this, see Dummett 1973, p. 296. See also Brandom 1994.


17 The second interpretation would, of course, also insist that the meanings of "win" and "lose" had been switched, so that players who claimed that they wanted to win 'really' meant that they wanted to lose, etc.


18 It seems, of course, otherwise when working within our gaming practice: we can specify the rules of a game and what constitutes winning it without making any reference to the fact that we try to win, we can play by the rules and deliberately lose, etc. (Just as we can specify what constitutes winning an individual game, we can specify the truth conditions for individual sentences.) But all of these practices, like those of lying, stealing, etc. can only work parasitically as exceptions and cannot exist independently of the more integrated practices they seem to flout or rise above.


19. The only way in which it might be possible is a game in which all the players tried to loose but were each under the mistaken impression that the others were trying to win (compare Lewis's discussion of a community of deluded liars in Lewis 1972). The potential parallel with language should be obvious. Just as the concept of truth allows us to specify languages with no reference to how speakers actually try to use the language, the reference is always there implicitly.


20. In this respect I think that Searle, at best, expresses himself unhappily when he claims that:

I think that there are rules crucial to competitive games which are not peculiar to this or that game. For example I think that it is a matter of rule of competitive games that each side is committed to trying to win. Notice in this connection that our attitude to the team or player who deliberately throws the game is the same as that towards a team or player who cheats. Both violate rules, though the rules are of quite different sorts. (Speech Acts, 34.)

While we could understand someone as having violated a rule if they threw a game, the rules need make no reference to winning; someone who throws a chess game violated the rule of trying to put your opponent in checkmate, etc. The idea that "trying to win" is an independent rule suggests that there could be freestanding non-competitive games (as opposed to competitive games played non-competitively) which still made reference to winning.


21. Just as our stating when a sentence is true is enough for us to know how to use it (we need not add, that, for this sentence, like most others, we use it when it is true).


22. Though we can, of course imagine cases where both participants tried to lose while putting on the pretense of trying to win. Instances of this purportedly could be seen at a number of track meets where it became known that the runner who came second would be tested for drug use. In such races, once it became clear that a certain runner would win, many of the other runners made a noticeable effort not to come second. While it was quite obvious that the runners were trying to let others pass them, they could not do this overtly (by, say, stopping or running backwards) and so had to put on a pretense of trying for the second spot (the results were, supposedly, quite comical).


23. I am speaking here, of course, about a single playing of chess which takes place against a general background of trying to win.


24. This phenomenon is, of course, structurally similar to stealing, lying and, at least according to some, having false beliefs. Each represents something which could, and does, occur against a background of owning, truth-telling and true belief, but none of these could turn out to be the rule rather than the exception.


25. For a characterization of this speech act, see Dummett 1973, Stalnaker 1978.


26. The claim that assertion and denial are on the same conceptual footing may stem from the focus on 'assent' and 'dessent' within the Quinean tradition. This later pair do seem to be on the same footing conceptually.


27. The resulting defense of the claim that we can know that our interlocutors are generally making assertions rather than denials is thus stronger than Ramberg's defense of a similar assumption on Davidson's part (Ramberg 1989, p. 68). According to Ramberg, the assumption that our interlocutors are making assertions is, like the principle of Charity, a presupposition of our engaging in the interpretive process. This comparatively add hoc way of dealing with the Massey cases leaves it coherent to assume that there could be such inverted practices, its just that we can't interpret others if we do.


28. I.e., Searle's friend takes most of his and others' utterances to be non-ironic.


29. Which is part of the reason why irony should not be equated with denial. If there is an 'ironic force' it plays a role significantly different than merely being the opposite of assertion.


30. For a discussion of this, see Lewis 1969, 1972.


31. Hale 1971.


32. Indeed, pace Massey, there is little reason to think that Walbiri has a set of 'antonyms' any richer than what is implicit in most languages. For instance, Walbiri antonyms for, say, most biological terms, are constructed along principles that would seem to be available to the speakers of any language. Hale describes this process as follows:

It is apparent that the principle employed here is that of opposing entities which, according to some taxonomic arrangement or other, are most similar, i.e., in formal terms, entities which are immediately dominated by the same node in a taxonomic tree. Thus, members of a given class of objects are opposed to other members of the same immediate class -- a large macropod is opposed to another large macropod (kangaroo/euro), a eucalypt is opposed to another eucalypt (red-gum/ghost-gum), and so on. (Hale 1971, p. 476.)

The Walbiri antonym structure is of interest then because the presence of Upside-down Walbiri makes explicit the type of abstract semantic structures of the sort that usually remain more implicit in other languages, and thus "provides us with a surprisingly uncluttered view of certain aspects of this semantic theory and is certainly not irrelevant to the much discussed, though occasionally incoherent, question of whether the semantic structures we, as students of language and culture, imagine to exist do, in fact, have any 'reality' for the speakers who use the system."( Hale 1971, p. 478.) Furthermore, while Massey claims that the adults in the hut speak "only" Upside-down Walbiri, what Hale actually says is that "one guardian speaks [Upside-down Walbiri] while the other answers, or rather interprets the message, in Walbiri." (Hale1971, p. 474.) The dialogs the initiates would hear would thus be something like the following:


A: Withhold fire from him.

B: I should give you water.

A The other is quenched.

B: You are thirsty.


Needless to say, this sort of running translation will make Upside-down Walbiri much easier to learn than simple immersion. It should be noted that, even if what Massey said were true about it, the learnability of an 'Upside-down' language parasitic on one's own does not show that a dual-language could be learnt as one's first language. That is to say, Upside-Down Walbiri may still be parasitic on Walbiri in a way that would prevent it from showing that denial is conceptually on par with assertion.


Works Cited


Brandom, R. 1994. Making it Explicit, Cambridge: Harvard University Press.

Davidson, D. 1984. Inquiries into Truth and Interpretation, Oxford: OUP.

Dummett, M. 1973. Frege, Philosophy of Language, Cambridge: Harvard University Press.

Hale, K. 1971. "A note on the Walbiri tradition of antonymy", in Steinberg and Jakobovits, ed., Semantics: an Interdisciplinary Reader in Philosophy Linguistics and Psychology. New York: Cambridge University Press, 1971.

James, W. 1907, 1979, Pragmatism, Cambridge: Harvard University Press.

Lewis, D. 1969.Convention, Cambridge: Harvard University Press.

Lewis, D. 1972. "Languages and Language". In Lewis 1983.

Lewis, D. 1974. "Radical Interpretation", in Lewis 1983.

Lewis, D. 1983. Philosophical Papers, Volume 1. New York: OUP.

Massey, G. 1992. "The Indeterminacy of Translation: A Study in Philosophical Exegesis," Philosophical Topics, vol. 20, n.1.

Peirce, C. S. 1878. "How to Make our Ideas Clear", in Writings of Charles S. Peirce, Volume 3: 1872 - 1878, Bloomington: Indiana University Press, 1986.

Putnam, H. 1981. Reason, Truth and History, New York: CUP.

Ramberg, B. 1989. Donald Davidson's Philosophy of Language. New York: Blackwell.

Stalnaker, R. 1978. "Assertion" in Davis, S. (ed). Pragmatics, a Reader. New York: OUP, 1991.

Searle, J. 1987. "Indeterminacy, Empiricism, and the First Person", The Journal of Philosophy Vol. LXXXIV, no.3, 123-146.