Kripke's Puzzle About Belief

Mike Mackus December 9th, 2008 Karen Lewis Philosophy 210 Language and Knowledge within the Individual and the Community The Millian perspective on proper names is immediately appealing for its simplicity and intuitiveness: the meaning of a proper name is completely exhausted by its referent. The name of a person means that given person; the meaning of ‘Wilt Chamberlain’ is Wilt Chamberlain. Frege, however, poses a problem for the theory of direct reference. If the meaning of a proper name is simply its referent then coreferring names would have the same exact meaning and, moreover, the coreferring terms would contribute the same exact thing to a given proposition. For example, the sentence ‘George Orwell is a writer’ and ‘Eric Blair is a writer’ both express the same proposition given that George Orwell and Eric Blair are the same person. Frege points out, though, that the equivalence in meaning of these two sentences is not immediately clear. Frege’s puzzle illustrates that while ‘a = a’ is apparent, there is no obvious reason to know that ‘a = b’. The theory of direct reference seems to support a rule of substitutivity but Frege’s reveals a predicament for such a rule, especially in regards to belief ascriptions. According to Mill, even within a belief context, we should be able to substitute the coreferring terms while maintaining the truth-value; this, however, does not appear to be necessary. Take the sentence ‘Jones believes that George Orwell is a writer’. Using the principle of substitutivity that the theory of direct reference would imply we should then also be able to state that ‘Jones believes that Eric Blair is a writer’. Yet, it is not the case that both of these sentences are true: we can imagine a scenario where Jones has read countless texts by the man he knows as George Orwell and also remembers an old friend he fought with in the Spanish Civil War by the name of Eric Blair- a man that, he believes, could never be a writer; Jones does not believe that Eric Blair is a writer, thus the 1

principle of substitutivity fails to retain truth-value. For Frege’s theory the non-interchangeability of coreferring names is not a problem. Frege would argue that the names simply have different senses or properties associated with them. While Frege offers an easy way out of this puzzle, just as Russell’s theory does, there are many reasons why both Frege’s and Russell’s theories are ones we should reject: Putnam’s argument that meaning cannot exist in the head; Kripke’s counter-examples against descriptivist theories using contingent truths; Kripke’s explanation that associating definite descriptions with names could mean that two people think they are speaking about the same person but are not since each has a separate definite description in mind; etc. Given the rejection of descriptivist theories we seem to be back where we started: with a direct reference theory such as Mill’s. The problem with Mill’s theory, as we saw above, was its inability to conform to a principle of substitutivity but Kripke aims to show us that we are left with a strikingly similar puzzle without resorting to any means of substitution. He constructs one puzzle with the use of, what he believes to be, two self-evident principles: the principle of translation and the principle of disquotation. Kripke goes further to demonstrate that the principle of disquotation itself is sufficient in creating the puzzle. The principle of translation, while rather vague, still remains intuitive. It states that if a sentence in one language expresses a truth then when that sentence is translated into another language it will still express a truth. The second, and probably more important, tool Kripke needs in constructing his puzzle is the principle of disquotation. This principle states that a speaker assents to a sentence ‘p’ if he believes that p and only if he believes that p; the principle also includes a clause that the speaker at hand is normal and not reticent. From here Kripke begins to sketch his puzzle: There is a man named Pierre, a monolingual French-speaker, who lives in

2

France. Many of his friends have come back from a far off place called Londres raving about how beautiful it is. He has even seen pictures himself and agrees with his friends’ assessments. Thus Pierre is inclined to assent to and state himself that ‘Londres est tres jolie’. Using the disquotational principle one can then claim that ‘Pierre croit que Londres est tres jolie’. Next, using the principle of translation one can move the claim into English and state ‘Pierre believes that London is very pretty’. So far we have only taken an utterance that Pierre assented to, attributed the belief in that utterance to him and then translated the attribution of that belief from French to English. A few months later, however, Pierre moves to a run down part of London; he has never ventured into the better looking parts of the city and completely hates the way his neighborhood looks. Over time Pierre acquires English as a second language without using any method of translation; rather he absorbs it directly from difficult conversations with his monolingual English-speaking neighbors. Finally his competence in English reaches a point where he is able to state ‘I hate London! London is one of the ugliest places in the world! London is not pretty.’ Pierre has assented to and uttered the statement ‘London is not pretty’, thus we can use the disquotational principle again to state ‘Pierre believes London is not pretty.’ However, nothing so far has challenged any of Pierre’s prior beliefs or claims. Pierre would still assent to ‘Londres est tres jolie’, allowing us to assert that ‘Pierre believes London is very pretty.’ Kripke even proposes that Pierre could be a logician and that he would never let himself hold contradictory beliefs. Nevertheless, this is not necessarily a problem for Pierre for he is simply ignorant of the fact that ‘Londres’ and ‘London’ both refer to the same city; this is, however, a problem for us as the observers for we are forced to agree with the obviously false P & ~P where P = Pierre believes London is pretty. After dropping such a bomb on us one would believe that Kripke might have a solution in mind; that is not the case. Kripke maintains that this

3

is, in the fullest sense, a puzzle and that every theory of meaning must some how come to grips with it. As noted above, the principle of translation appears to be somewhat vague as well as weak. The only condition for successful translation is the retention of truth-value which does not necessarily entail an accurate translation. Furthermore, one may propose a sentence such as ‘This sentence is in English’ that cannot be translated successfully according to the principle of translation. With the question of this principle’s validity in the back of his mind Kripke goes a step further in order to construct a puzzle with the same result without the use of any translation: One day Peter, being an excellent pianist himself, is recommended to the pianist Paderewski by a friend. Peter enjoyed Paderewski’s entire a catalog of music and thus assented to and asserted the claim ‘Paderewski had tons of musical talent’. Through use of the disquotational principle one can then state ‘Peter believes that Paderewski had tons of musical talent.’ Not long after Peter hears friends speaking of a man named ‘Paderewski’ during a discussion of politics. His friends tell him that Paderewski was a nationalist leader and prime minister of Poland. Peter draws the conclusion that these two people named ‘Paderewski’ could not be the same (even though the two people are one in the same) for politicians are not renowned musicians. Given this, Peter would also assent to the claim that ‘Paderewski has no musical talent’ and one could rightly claim ‘Peter believes Paderewski has no musical talent’. Once again, it does not appear that Peter is the one with logical inconsistencies; he simply lacks knowledge of the fact that Paderewski the musician is also Paderewski the politician. But for us, the observers, we are left agreeing with the logical inconsistency of Q & ~Q where Q = Peter believes Paderewski has musical talent. Kripke displays that there is no need to use the principle of translation in order to generate his puzzle. In doing so he eliminates the possibility that the trouble only lie in confusion between languages.

4

Should we consider that there might be a problem with the disquotational principle and the use of it? In a general sense, the answer must be no: there is no way to refute the principle of disquotation for it is an operation we use in everyday reasoning and the move occurs in only one step. Therefore if there was a problem with the principle it would be between an agent’s assent to an utterance ‘p’ and our attribution of belief that p to that given agent. While this step may be questioned, say, by an argument along the lines that ascribing belief is a much more difficult process than simply hearing a speaker claim to have a belief, I believe it doesn’t get to the core of Kripke’s puzzle; that is, Kripke’s puzzle does not only concern an individual’s belief but it also bears on the knowledge of an individual and the knowledge of a linguistic community. After taking a closer look at what is implied in Kripke’s argument through the lens of David Sosa I will draw an argument about what it means for a language community to have knowledge and, furthermore, how Kripke’s puzzle sheds light on the system of language and the system of knowledge; moreover, Kripke’s puzzle demonstrates that these two systems are indeed intertwined and perhaps composed of one another. After properly addressing these points I believe Kripke’s puzzle disappears. David Sosa proposes a puzzle analogous to Kripke’s in order to highlight an implicit premise that motivates Kripke’s original puzzle. Sosa’s analog is an unsuccessful construction of the puzzle which enables him to delineate what procedure takes place in the Kripke examples as opposed to his own. Sosa’s (non-)puzzle is as follows: There is a man named Rock who lives in Paris, Texas. He is not particularly fond of his community so he assents to and asserts ‘Paris is not pretty’. The disquotational principle allows one to then hold that ‘Rock believes that Paris is not pretty’. Rock has also seen many photographs that his friends have brought back from France, particularly of the city Paris. Rock loves the sights in his friends’ pictures and assents to

5

the statement ‘Paris is pretty’. This then allows one to claim ‘Rock believes that Paris is pretty’. Thus one can say ‘Rock believes that Paris is pretty and Rock believes that Paris is not pretty’. It is apparent that this puzzle does produce the same results as Kripke’s but why not? Obviously it is because Rock is speaking of two different places, Paris, Texas and another city by the same name in France. The crucial question, however, is, when juxtaposed with Kripke’s puzzle, where does the divergence come that makes it fail and, moreover, what procedure gets Kripke’s puzzle past the point where Sosa’s fails? Sosa argues that the successful puzzle must assume a conditional of the nature if S believes p and S believes ~p then there is a contradiction. In the case of the unsuccessful analog the name ‘Paris’ has multiple referents and is thus ambiguous. However, as Sosa proposes, this means that along side the above conditional there needs to be an additional principle, that being the hermeneutic principle, which states if a name has a single referent then it can be properly represented logically by a single constant. One might then say that in the unsuccessful analog we do not end in a contradiction because Paris is ambiguous. But by using ambiguity in this sense, where a term is ambiguous only if it has multiple references, there is a presupposition that ambiguity cannot be of the kind where a term has only one referent and multiple senses associated with it; by presupposing this Kripke already eliminates the possibility for a Fregean theory. Furthermore, Sosa notes that the only justification for the hermeneutic principle is a Millian approach. Thus even though Kripke avoids use of the principle of substitutivity he still resorts to hermeneutic which is also implied by Millianism. Nonetheless, the hermeneutic principle seems to perform the same basic operation as substitutivity insofar that both rules require two co-referring terms to have the same meaning. The puzzle now appears identical with Frege’s. Now, are we to resort back to a Fregean notion of sense in order to better capture these ambiguities; is the problem at hand simply different modes of presentation? Sosa

6

would have us believe so. But we have also noted reasons why a Frege-Russellian descriptivist theory is also shortcoming. While this might feel like a terrible position to be in, not having an appealing theory without devastating objections being weighed against it, I believe this leaves us at an extremely valuable crossroads where we can synthesize all our results thus far, successes and shortcomings alike, in order to form a new hypothesis. Kripke’s puzzle not only tells us a lot about language but it also tells us about how beliefs interact with language. While Putnam assured us meanings do not exist inside a language users head we obviously cannot take this to mean beliefs do not exist inside one’s head. The beliefs throughout Kripke’s puzzles are extracted from the speaker’s head and put in terms of a proposition. The proposition then has its existence outside of any specific speaker; the proposition is simply there for anyone to utter. As a group we take this proposition and examine it; we fill it with meaning and the knowledge of the linguistic community. The original speaker (or thinker) does not necessarily have the knowledge of the rest of the community and, in the case of all Kripke’s examples, must not have the same knowledge. It seems then there must be a distinction between the use of the individual and the understanding of the community. The original speaker’s system of knowledge plays a crucial role in the content of his utterances (the product of his language). The linguistic interface (specifically, the semantic-linguistic interface for our purposes) is in a direct relation with the speaker’s knowledge of the outside world. If this is the case, that knowledge has a direct influence on what one can assert, then language must have an individualistic side to it. Knowledge is shaped and colored by each person’s unique experience. The unique knowledge of an individual has implications on his semantics and the way he uses linguistic expressions. This is obvious in all of Kripke’s examples: Pierre has no knowledge of the fact that ‘Londres’ and ‘London’ refer to the same city and thus uses the names

7

as if they were different; more interestingly, Peter is not faced with the troubles of translation and must resort to an additional step of thought in order to believe ‘Paderewski’ has two different referents: he assumes if Paderewski is a musician then he cannot be a politician and if Paderewski is a politician he cannot be a musician. While Kripke’s example of Pierre appears to be one of simple ignorance, Peter poses a problem because he resorts to his own knowledge, his own rules of inference, in order to determine the way he deploys a linguistic expression. Kripke’s example without the principle of translation requires more than an individual to be ignorant of the actual state of affairs; when constructing the puzzle without any translation Peter is forced to import his own knowledge about the nature of politicians and musicians. Knowledge dictates the way in which one can use a name or any other term. On the other hand, however, when the linguistic community hears an assertion that resembles the results of the puzzle, P & ~P, there is an obvious tension because the linguistic community as a whole possesses the necessary knowledge required in order to know that there is a contradiction. We see then that the nature of language and the nature of knowledge have two components: language exists within the individual and his knowledge effects the way in which he uses this language; and language exists throughout the entire linguistic community directly connected to the knowledge of the community which helps maintain stability in the semantics of terms. The stability I refer to can be regarded as such: there is a necessary agreement between the language community in terms of the relationship between semantic knowledge and knowledge of the world that is needed for successful communication. We see that language and knowledge exist in two elements: within the individual and throughout the community. Thus the relationship necessary for successful communication must be that the individual user’s semantic knowledge cannot be changed by his knowledge (of the world) insofar as it inhibits his ability to

8

communicate successfully. (Likewise, his semantic knowledge cannot inhibit his knowledge of the world to the point of inhibiting his ability to operate within it.) More broadly, this principle states that the individual must have some sort of necessary knowledge that cannot diverge too far from the knowledge of the linguistic community. As Lepore and Stone note, the notion that the language community shares the same meaning of different linguistic terms is only an assumption that we make in order to explain linguistic competence and use and thus “it is not a logical consequence of our knowledge of meaning that we share the meanings of our terms with others in our community” (Lepore and Stone 147). Kripke’s examples make shared meaning a logical necessity but as we can see such is not the case. By viewing language and knowledge as having an existence within the individual and throughout the group we see that Kripke’s puzzle disappears. Language, in its broadest sense, exists throughout the community, while knowledge, although shared among the community, exists mainly within the individual. I believe Kripke’s puzzle confuses this element of the interaction between the system of language and the system of knowledge and with the principle proposed above I believe that we can make the puzzle disappear by simply stating that the agents within Kripke’s examples violate the requirements for successful communication.

9

Works Cited Lepore, Ernest, and Matthew Stone. "Logic and Semantic Analysis." Philosophy of Logic. Ed. Dale Jacquette. Chapel Hill, MA: North Holland, 2006. 145-76. Sosa, David. "The Import of the Puzzle About Belief." The Philosophical Review 105 (1996): 373-402.

10