In defence of the hybrid view – meaning of name versus content of belief – response to Richard G. Heck, Jr., Mind, vol. 104, p. 79, 1995
Alex Byrne
In “The Sense of Communication”(1) Richard Heck argues that a popular view in the philosophy of mind and language is untenable. Although we think there is much in Heck’s paper that is highly suggestive, his central argument fails, and the purpose of this note is to bring out that failure. The view in question–which Heck calls the Hybrid View–distinguishes between the meanings of names and the contents of beliefs which are expressible using names. According to the Hybrid View the meaning of a name is its referent: names do not have senses. Thus (a) “George Orwell wrote 1984” means the same as (b) “Eric Blair wrote 1984”. However, the Hybrid View tells a different story about the beliefs one expresses when one utters (a) or (b). The content of a belief expressed using a sentence that contains a name has more to it than just the referent of the name and whatever properties the sentence ascribes to the name’s referent. There is something like a Fregean sense in the content of the belief. Thus though (a) and (b) mean the same thing, the belief that a speaker expresses by uttering (a) need not have the same content as the belief he would express by uttering (b). In short, the Hybrid View says that names are Millian, while beliefs are Fregean.
1. The premises
Heck’s argument against the Hybrid View is complicated and not easy to follow. Early on in his paper he raises the question: “how must the senses which different speakers associate with a given name be related if transmission of knowledge by means of language is to be possible?” (p. 81). And he says he will argue “that the answer is not that speakers may diverge as much as they like regarding the senses they attach to a given name, so long as they attach the same reference to it” (p. 81). A further argument is supposed to show that the previous conclusion–that attaching the same reference is insufficient for linguistic transmission of knowledge–is “inconsistent with the Hybrid View” (p. 81). This further argument, Heck says, “rests upon two crucial auxiliary premises” (p. 81). So Heck’s argument for the conclusion that the Hybrid View is false has three premises: that attaching the same reference is not sufficient for knowledge transmission, and the two crucial auxiliary premises. The first of these three premises may be stated thus:
(R [Mathematical Expression Omitted] K)
“Preservation of reference and means of expression [i.e. words
individuated orthographically] … does not suffice to enable the
transmission of knowledge” (p. 95) (even if “defeating conditions”
do not obtain).(2)
We shall see exactly what this means when we come to Heck’s argument for it. As to the two crucial auxiliary premises, Heck says that he will not argue for them directly, but instead argue that “they ought to be acceptable to a proponent of the Hybrid View” (p. 81). The conditional status of these two premises will be important later on.
Heck never explicitly formulates the two crucial auxiliary premises as such, beyond saying that the first (which we will call “AP1”) concerns knowledge and understanding, and the second (AP2) concerns understanding and meaning. A little reconstruction is needed to pin down AP1, but it is clear enough what AP2 is, or at any rate ought to be.(3) We now turn to identifying these premises, beginning with AP2.
Heck defines the notion of the cognitive value of a sentence as follows:
(CV=CB)
“[T]he cognitive value of [a] sentence … is what is, or ought to
be, common to the beliefs different speakers who understand it
would form were they to accept it as true.” (p. 87)
(The cognitive value of an expression–most relevantly a name–is the contribution the expression makes to determining the cognitive value of sentences containing it.(4))
Heck argues in [sections]4 that a proponent of the Hybrid View ought to accept:
(M [right arrow] CV)
“[T]he meaning of a name [or a sentence] … determine[s] its cognitive
value” (p. 89). (So, if two names differ in cognitive value,
they must differ in meaning.)
(M [right arrow] CV) is AP2.(5)
Heck also argues in [sections]4 that a proponent of the Hybrid View ought to accept the converse of (M [right arrow] CV):
(CV [right arrow] M)
“[T]he cognitive value of a name [or a sentence] … determine[s]
its meaning” (p. 89). (So, if two names differ in meaning, they
must differ in cognitive value.)(6)
We shall now draw a conclusion–which will turn out to be AP1–from (CV [Mathematical Expression Omitted] M) and the following claim that Heck endorses in [sections]5:
(U [Mathematical Expression Omitted] K)
“If one has communicated successfully with (understood an assertion
made by) one who knows what she said to be true, then all
that is required, if one is to come to know something, is that one
accept the conclusion as true” (p. 92), “[m]odulo defeating conditions”
(p. 92, fn. 19).
Suppose that A utters a sentence s to B, and B realizes this fact. Suppose that the meaning of s in A’s language is the same as the meaning of s in B’s language. Surely this is sufficient for B to have understood A’s assertion (more exactly, for B to have understood what A literally said).(7) That is, preservation of meaning (and means of expression) is sufficient for understanding. But given (CV [Mathematical Expression Omitted] M), since cognitive value determines meaning, preservation of cognitive value (and means of expression) is sufficient for understanding. Let’s just stipulate that we are only concerned with cases where the speaker knows, and the hearer accepts the speaker’s asserted sentence as true. Then given (U [right arrow] K), that understanding is sufficient for knowledge transmission, it follows that:
(CV [right arrow] K)
Preservation of cognitive value (and means of expression) is sufficient
(modulo defeating conditions) for transmission of knowledge.
Heck clearly believes (CV [right arrow] K) (8) and its addition to the two premises we have so far identified is the most obvious way to make Heck’s argument valid. We therefore claim that it is the fugitive AP1.(9)
2. The argument against the Hybrid View
Heck’s three premises are, or ought to be, these:
(R [Mathematical Expression Omitted] K)
“Preservation of reference and means of expression … does not
suffice to enable the transmission of knowledge” (p. 95) (even if
“defeating conditions” do not obtain).
(CV [right arrow] K)
[AP1] Preservation of cognitive value (and means of expression)is sufficient (modulo defeating conditions) for transmission of
knowledge.
(M [right arrow] CV)
[AP2] “[T]he meaning of a name [or a sentence] … determine[s]its cognitive value” (p. 89). (So, if two names differ in cognitive
value, they must differ in meaning.)
To show that these premises imply the falsity of the Hybrid View, it will be useful to give the example Heck uses to argue for (R [Mathematical Expression Omitted] K).(10) Consider, Heck says, two speakers, Tony and Alex, and:
[s]uppose Eric Blair were to become amnesiac and check himselfinto a hospital. The doctor, Tony, deciding that she needs to have
some name by which to call him, dubs him “George Orwell”. And
suppose further that Alex says–not intending to refer to Tony’s
patient–“George Orwell wrote 1984” and that Tony forms, in reaction
to Alex’s assertion, the belief she would express to other
members of her staff as “George Orwell wrote 1984”. This belief
is true: Tony’s new patient happens to be Eric Blair, that is, “the
other” George Orwell. But surely it would not count as knowledge,
even if Alex knows that George Orwell wrote 1984: it
would not even count as justified. Thus, preservation of reference
and means of expression is not sufficient for successful communication,
since it does not suffice to enable the transmission of
knowledge. (p. 95, fn. omitted)
So here is a case of attempted knowledge transmission using “George Orwell wrote 1984”, with no defeating conditions, where the reference of “George Orwell” is preserved, but in which knowledge fails to be transmitted. (And the case is not one where the trouble is traceable to the predicate “wrote 1984”.) So the cognitive value of “George Orwell”–the contribution “George Orwell” makes to the cognitive value of the sentence “George Orwell wrote 1984”–cannot be simply its reference (or be determined by its reference). For by the second premise, (CV [right arrow] K), if the cognitive value of “George Orwell” were simply its reference, the cognitive value of “George Orwell wrote 1984” would be preserved, and knowledge would be transmitted after all. As the cognitive value of “George Orwell” is not simply George Orwell, it follows that there could be another name for George Orwell (“Eric Blair”, perhaps) with a different cognitive value. So “George Orwell” does not necessarily have the same cognitive value as any co-referential name. Now the third premise, (M [right arrow] CV), says that meaning determines cognitive value: if two names have the same meaning, they must have the same cognitive value. So, if the meaning of “George Orwell” is simply its reference, it must have the same cognitive value as any co-referential name. But we just saw that the consequent of this conditional is false. So the meaning of “George Orwell” is not simply its reference–contra the Hybrid View. Hence the Hybrid View is false.(11,12)
(It is worth noting that the detour through cognitive value is entirely unnecessary. Take Heck’s first premise, (R [Mathematical Expression Omitted] K), and his claim that understanding is sufficient for knowledge transmission, (U [right arrow] K). In arguing for Heck’s second premise, (CV [right arrow]K), we appealed to the fact that preservation of meaning is sufficient for understanding. So, by (U [right arrow] K), preservation of meaning is sufficient for knowledge transmission. And (R [Mathematical Expression Omitted] K) gives us the desired conclusion, that meaning is not reference. However, as this simple version of the argument is certainly not the one Heck has in mind, we will stick to criticizing the argument as originally given. It will be easy enough to see how to adapt our criticisms to the simple version.
3. Why the argument fails
The trouble lies with Heck’s first auxiliary premise, (CV [right arrow] K). Remember that Heck says he will not argue for either auxiliary premise directly, but will instead argue that both ought to be acceptable to a proponent of the Hybrid View. So we shall first argue that Heck has supplied no good reason why the Hybrid View theorist in particular should accept (CV [right arrow] K).
But we can hardly leave the matter there. For on the face of it (CV [right arrow] K) may well be true, in which case Heck could withdraw the conditional status of (CV [right arrow] K) and simply assert it outright. In fact, as we have noted, Heck does appear to believe (CV [right arrow] K) himself. But we shall finally argue that this escape route is blocked, for (CV [right arrow] K) is false.
First, why does Heck think that a proponent of the Hybrid View ought to accept (CV [right arrow] K)? This premise follows, we argued, from:
(CV [right arrow] M)
“[T]he cognitive value of a name [or a sentence] … determine[s]
its meaning” (p. 89).
And:
(U [right arrow] K)
“If one has communicated successfully with (understood an assertion
made by) one who knows what she said to be true, then all
that is required, if one is to come to know something, is that one
accept the conclusion as true” (p. 92), “[m]odulo defeating conditions”
(p. 92, fn. 19).
Heck clearly thinks that the Hybrid View theorist, in particular, ought to accept (CV [right arrow] M) (see Heck’s [sections]4). We are not going to dispute this, nor are we going to dispute that (CV [right arrow] M) is true.
What about (U [right arrow] K)? As it, in conjunction with (CV [right arrow] M), implies the false (CV [right arrow] K), and since we are not denying (CV [right arrow] M), we must deny (U [right arrow] K). Heck, as we have observed, evidently thinks that (U [right arrow] K) is true. So perhaps he thinks that the Hybrid Theorist has a proprietary reason to accept (CV [right arrow] K) because he has a proprietary reason to accept (CV [right arrow] M) and a perfectly general reason, acceptable to theorists of all stripes, to accept (U [right arrow] K). But Heck does not argue for (U [right arrow] K) and, as we shall see shortly, there is every reason for taking it to be false.(13) So if this interpretation is correct, there is no reason why the Hybrid View theorist need accept (U [right arrow] K), and so no (obvious) reason why he need accept (CV [right arrow] K).
However, some material early in Heck’s paper strongly suggests that he does think that the Hybrid View theorist, in particular, should accept (U [right arrow] K). So we now need to show that (what we take to be) Heck’s reason is mistaken.
In [sections]2 of his paper, Heck argues that the Hybrid View implies that “understanding an utterance depends, in respect of names, only upon preservation of their reference … reference is all that matters” (p. 83). That is, understanding an utterance, in respect of names, simply is getting the reference right: getting the reference right is both necessary and sufficient for understanding. And in [sections]3 he suggests that the proponent of the Hybrid View needs to answer the question: “[w]hat sort of argument can be given for the claim that even so much [as preservation of reference] is necessary for understanding?” (p. 85). He answers that the only available argument is that “in communication, we use other speakers as a source of information … It is thus important that we not get the references of names occurring in sentences uttered by others wrong, lest we acquire false beliefs” (p. 85). And much later, in [sections]5, Heck says that [w]e wish not to misunderstand because understanding is a precondition of the transfer of information. But the transfer of information we desire is not merely the transfer of true belief. it is the transfer of knowledge” (p. 92). This connection between understanding and knowledge Heck encapsulates in (U [right arrow] K). So all this suggests that “the only available argument” (p. 86), acceptable to the Hybrid View theorist, for why so much as reference needs to be preserved, will appeal to (U [right arrow] K). That is why–Heck might argue–a proponent of the Hybrid View theorist has a proprietary reason for accepting (U [right arrow] K).(14)
Let us just grant that an argument, in the style Heck may have in mind, for why preservation of reference is necessary for understanding, does require (U [right arrow] K).(15) The obvious difficulty is this. If Heck is correct that the Hybrid View implies the claim that preservation of reference is necessary and sufficient for understanding, then any argument for the Hybrid View is an argument for this claim. Since, as Heck notes, there are good reasons to think that the Hybrid View is correct,(16) there are also good reasons to think that preservation of reference is necessary for understanding. It is hard to see why a proponent of the Hybrid View needs any additional argument for this. Hence we have no reason to believe that the Hybrid View theorist, in particular, is committed to (U [right arrow] K), and so no reason to think him committed to the second of Heck’s premises, (CV [right arrow] K).
Heck might reply that the demand he intended to press on the Hybrid View theorist was not, after all, an argument for why reference must be preserved, if understanding is to be possible. Rather, he might say, the Hybrid View theorist must explain why reference must be preserved (see p. 91). But, first, one would like to know why the argument that can be given by the Hybrid View theorist is not also an explanation (some arguments are explanations). And, second, it seems that once one has argued that meaning is reference, then given the close conceptual connection between meaning and understanding, one is in an excellent position to explain why preservation of reference is necessary for understanding.
Finally, then, let us see why (CV [right arrow] K) is false. As Heck argues, preservation of reference is not sufficient for knowledge transmission. But an exactly parallel argument shows that preservation of cognitive value is not sufficient for knowledge transmission either. Claims about cognitive value in the abstract are hard (for us, anyway) to understand, so let us begin with a specific proposal. Suppose (as a Fregean about names might) that the cognitive value of a name is given by a description, and consider the following case. A patient checks into the hospital and is assigned room 101. Tony dubs him “Winston” and the cognitive value she attaches to the name is: the amnesiac in room 101. Alex is thoroughly unaware that Tony has seen the patient, but by sheer chance she also dubs him “Winston” and attaches the same cognitive value to the name. Alex utters “Winston will never recover” in Tony’s presence, and Tony forms the belief that she would express by saying “Winston will never recover”. Even if Alex knows that Winston will never recover, does Tony’s belief count as knowledge?
It is clear that Tony’s belief isn’t knowledge in this case any more than it was in Heck’s example, even though here both Tony and Alex attach the same cognitive value to “Winston”. Heck argues that if it is only by chance that Tony and Alex attach the same referent to the name “Orwell” then “even if Tony’s newly formed belief were true, she should merely have been lucky” (p. 9 1), and hence her belief isn’t knowledge. What the preceding example brings out is that identical cognitive values can also occur as a matter of luck, and hence requiring that different speakers attach the same cognitive value to names won’t guarantee knowledge transmission.
Admittedly, the example concerns a particular Fregean proposal about the cognitive value of names. But it is very hard to see why any proposal about cognitive value should not be susceptible to a similar example. Whatever cognitive value is, surely speakers may by chance attach identical cognitive values to identical expressions, in which case knowledge transmission will not occur. So (CV [right arrow] K) is false.
It seems clear, to us at least, that Tony/Alex-style examples show that understanding can occur as a matter of luck, and so understanding does not guarantee knowledge transmission. Therefore (U [right arrow] K) is also false.(17)
Of course, one might wonder why Heck’s George Orwell example and our Winston example do not count as cases where “defeating conditions” obtain, in which case the falsity of (U [right arrow] K) would not follow.(18) (And a weakened version of (CV [right arrow] K), with a broader range of defeating conditions, would survive.) But if they do count, then Heck’s argument for his first premise,(R [Mathematical Expression Omitted] K), collapses. For (R [Mathematical Expression Omitted] K) says that reference preservation is not sufficient for knowledge transmission, even if defeating conditions do not obtain. And if the George Orwell example has a defeating condition then obviously (R [Mathematical Expression Omitted] K) cannot be drawn as a conclusion.(19,20),
Department of Linguistics and Philosophy ALEX BYRNE Massachusetts Institute of Technology Cambridge MA 02139 USA abyrne@mit.edu
Department of Philosophy MICHAEL THAU Princeton University Princeton NJ 08544 USA mikethau@pucc.princeton.edu
(1) Heck (1995). All page references are to this paper. (2) The parenthetical addition is clearly implicit. An example of a defeating condition is absence of trust (see p. 92, fn. 19). More on defeating conditions later. (3) There is an exegetical problem here. Heck’s explanation of the two auxiliary premises is this: AP1 is that “conclusions about what is necessary for the transfer of knowledge yield conclusions about what is necessary for understanding; and [AP2 is] that conclusions about what is necessary for understanding yield conclusions about meaning” (p. 81). It looks, from the way they are linked together in the quoted sentence, as if AP1 and AP2 together give us that “conclusions about what is necessary for the transfer of knowledge” “yield conclusions about meaning”. What is the “conclusion[s] about meaning”? Surely it is that “the meaning of a name exceeds its reference” (p. 81) (contra the Hybrid View). (The assumption that belief is Fregean–as Heck says, “intensional”–will apparently be assumed in deriving this conclusion (see pp. 81-2). So for the overall argument to work, the first premise, (R [Mathematical Expression Omitted] K), must give us “conclusions about what is necessary for the transfer of knowledge”. But it doesn’t: (R [Mathematical Expression Omitted] K) only gives us conclusions about what is not sufficient for the transfer of knowledge. So as (R [Mathematical Expression Omitted] K) is unquestionably Heck’s first premise, Heck’s official explanation of either AP1 or AP2 can’t be quite right, else the argument will fail. (4) In fact Heck never defines the notion of the cognitive value of an expression (instead, just the cognitive value of an expression for a speaker–see p. 89, fn. 15). But it’s clear the above definition is what he has in mind. (And, following Heck, we are ignoring indexical expressions.) (5) As we noted in fn. 3 above, Heck says the thrust of AP2 is “that conclusions about what is necessary for understanding yield conclusions about meaning” (p. 81). By the definition of cognitive value ((CV=CB)), if a speaker understands a sentence, then the belief she would form were she to accept it as true contains the sentence’s cognitive value. So claims about cognitive value might well be put as “conclusions about what is necessary for one to draw conclusions about meaning from claims about cognitive value”. And (M [right arrow]CV) does indeed straightforwardly allow one to draw conclusions about meaning from claims about cognitive value. (6) Putting (M[right arrow]CV) and (CV[right arrow]M) together, Heck says that the “meanings of expressions … might as well be identified with, their cognitive values” (p. 89). (7) Following Heck, we are taking “understanding a speaker” to mean understanding what the speaker literally said, not understanding what the speaker meant. Alternatively, we can simply stipulate that we are only considering cases of attempted communication where what the speaker literally said = what the speaker meant. (8) At p. 9 1, Heck argues as follows. “What [according to the Hybrid View] must be common to different speakers who understand a name, if communication among them is to be successful, is thus just the reference of the name: hence, given what was argued in the last section, the meaning of a name [according to the Hybrid View] is just its reference.” Now, what was argued in Heck’s “last section” ([sections]4) was that the meaning of a name is its cognitive value (see fn. 6 above). So Heck’s argument at p. 91 is only valid if “what must be common to different speakers who understand a name, if communication among them is to be successful” is the cognitive value of the name. So the cognitive value of a name is whatever to be preserved to ensure communication. And since Heck clearly believes that communication is sufficient for knowledge transmission, he appears to be committed to (CV [right arrow] K). (9) As we noted in fn. 3 above, Heck says the thrust of AP1 is “that conclusions about what is necessary for the transfer of knowledge yield conclusions about what is necessary for understanding” (p. 81). We noted in fn. 5 above that claims about cognitive value might well be put as “conclusions about what is necessary for understanding”. But trying to express (CV [right arrow] K) in this kind of terminology seems only to give something like: “conclusions about what is necessary for understanding yield conclusions about what is sufficient for the transfer of knowledge”. And this is not what Heck says. But as we also noted in fn. 3, Heck’s official elucidation of AP1 and AP2 can’t be right if his argument is to succeed. So charity enjoins us to say that Heck mischaracterises AP1. (10) Heck gives a similar example immediately before the example about to be quoted. But Heck does not use it to argue for (R [Mathematical Expression Omitted] K). Rather, it is employed to argue for a weaker thesis: that preservation of reference does not suffice to enable the transmission of knowledge. (I.e. the thesis obtained by deleting “and means of expression” from (R [Mathematical Expression Omitted] K). (11) Clearly a Tony/Alex-style example could be given for any name. Therefore we may take the conclusion of the argument to be that no name’s meaning is simply its reference. (12) Heck takes the Hybrid View’s claim that beliefs are Fregean to be important to the argument: “the meaning of a proper name exceeds its reference if belief is intensional [i.e. Fregean]” (p. 82). But it is unclear why. And if the claim is not doing any work, then the conclusion of the argument is that names are not Millian, regardless of the status of beliefs. Heck’s argument that the Hybrid View theorist ought to accept the two auxiliary premises does not seem to depend, as far as we can make out, on the fact that this theorist holds that beliefs are Fregean. But there is a case to be made that Heck’s argument for the first premise, (R [Mathematical Expression Omitted] K), depends on the assumption that beliefs are Fregean. For assume that beliefs are Millian. Suppose that Tony hasn’t dubbed her patient “George Orwell”, and that knowledgeable Alex asserts “George Orwell wrote 1984” as before. Suppose Tony comes to have a belief she would express by asserting “George Orwell wrote 1984”. Does it amount to knowledge? Yes, presumably. Now restore Tony’s dubbing of her new patient as “George Orwell”, keeping the rest of the example unchanged. By Millianism about belief, the belief that Tony would now express by asserting “George Orwell wrote 1984” is the very same belief that she had in the previous case where the dubbing was absent. It certainly is not obvious why adding the dubbing should remove knowledge–Tony would have believed the same fact whether or not the dubbing occurred. So the Millian about belief has some reason to resist the claim that Tony’s belief does not amount to knowledge, and so need not accept (R [Mathematical Expression Omitted] K). (13) More exactly: Heck does not give any reasons for (U [right arrow] K), with “defeating conditions” understood as they must be for the argument against the Hybrid View to succeed; that is, understood as not applying to Heck’s George Orwell example. We grant that Heck has motivated the view that understanding is sufficient, in most ordinary contexts, for the transmission of knowledge. (14) An argument in precisely this style appears in Heck’s [sections]3, but here the principle that the Hybrid View theorist ought to accept is “that communication [is] essentially a means for the transmission of information, of true beliefs from one speaker to another” (p. 86). Somewhat confusingly, it seems to turn out in [sections]5 (pp. 90-1) that the Hybrid View theorist ought not to accept this principle after all, because it’s false. Or, on an alternative interpretation, it turns out that this principle, although true, does not give the “correct explanation” (p. 91) of why reference must be preserved. In any event, it is natural at this point to read Heck as suggesting that the form of the argument in [sections] 3 was perfectly correct, but the wrong principle was selected: it ought to have been (U [right arrow] K) instead.. (15) This is a major concession, for it seems obviously wrong that the Hybrid View theorist must assume (U [right arrow] K) in the precise form required for Heck’s argument against the Hybrid View to work. But we need not develop this objection here. (16) A powerful argument for the Hybrid View can be extracted from familiar considerations concerning the explanation of behaviour and Kripke’s arguments against descriptivism. (17) It might be objected that understanding a speaker who says that p is equivalent to, or at least requires, knowing that he said that p (cf p. 82). If this is right, then since Tony/Alex-style examples are cases where the hearer does not know that the speaker said that p, they are also cases where the hearer does not understand. So they do not show, after all, that (U [right arrow] K) is false.
We think this is wrong. Granted, understanding someone implies that one knows what he said. Perhaps it is tempting to infer from this that understanding requires propositional knowledge. But this is to take a large and unobvious step away from the original platitude: understanding, in our view, is more a case of knowing-how than knowing-that.
More importantly, even if we are wrong about this, Heck’s argument is still unsound. The second premise, (CV [right arrow] K), remains false. But what of our earlier argument in [sections]1 for (CV [right arrow] K), which appealed to (U [right arrow] K)? Since (we are now assuming) (U [right arrow] K) is true after all, the argument must falter elsewhere. And indeed it does: the argument also required that “preservation of meaning (and means of expression) is sufficient for understanding”. And if understanding requires propositional knowledge then Tony/Alex-style examples show that this is false. (18) At p. 92, fn. 19 Heck indicates there are defeating conditions additional to those he gives as examples. (19) Heck cannot claim that only the Winston example is a case where some defeating condition has come into play, since the example has the same structure as his own. (20) Thanks to John Barker, Richard Heck, Jim Pryor, Gideon Rosen, Robert Stalnaker, Jason Stanley, and a referee for Mind.
REFERENCE
Heck, Jnr., Richard G. 1995: “The Sense of Communication”. Mind, 104, pp. 79-106.
COPYRIGHT 1996 Oxford University Press
COPYRIGHT 2004 Gale Group
