-
IS THERE A D IV IS ION OF L INGUIST IC LABOUR?
CATHERINE J.L. TALMAGE
In 'The Meaning of "Meaning'" Hilary Putnam put forward the view
that there exists in our linguistic community, and perhaps in
linguistic communities generally, a division of linguistic labour.
This view has come to be widely accepted, even, it seems, by
philosophers who are otherwise unsympathetic to the idea that
linguistic meanings are determined socially.' However, this
widespread acceptance of Putnam's view is, I think, largely
unreflective. I suspect that many of those who have endorsed the
view have done so either because they do not understand fully what
the view is or because it fits with one of their most basic ideas
about the nature of linguistic meaning, an idea that they may not
even have consciously acknowledged let alone subjected to critical
examination. In this paper I will discuss both the issue of what
exactly Putnam's view is and the issue of how someone might come to
endorse it. I will conclude, somewhat surprisingly perhaps, that
not only is Putnam wrong in claiming that there is a division of
linguistic labour but that he is wrong for reasons he himself gave
in 'The Meaning of "Meaning'". Indeed, I will argue that the view
that there is a division of linguistic labour arises from a way of
thinking that was itself Putnam's main target in 'The Meaning of
"Meaning'".
I Putnam introduces the notion of a division of linguistic
labour by
way of the following example:
421
-
CATHERINE J.L. TALMAGE
Gold is important for many reasons: it is a precious metal, it
is a monetary metal, it has symbolic value .... etc. Consider our
community as a 'factory': in this 'factory' some people have the
'job' of wearing gold wedding rings, other people have the 'job' of
selling gold wedding rings, still other people have the 'job' of
telling whether or not something is really gold. It is not at all
necessary or efficient that everyone who wears a gold ring...or
discusses the 'gold standard', etc., engage in buying and selling
gold. Nor is it necessary or efficient that everyone who buys and
sells gold be able to tell whether or not something is really gold
in a society where this form of dishonesty is uncommon (selling
fake gold) and in which one can easily consult an expert in case of
doubt. And it is certainly not necessary or efficient that everyone
who has occasion to buy or wear gold be able to tell with any
reliability whether or not something is really gold.
The foregoing facts are just examples of mundane division of
labor (in a wide sense). But they engender a division of linguistic
labour: everyone to whom gold is important for any reason has to
acquire the word 'gold'; but he does not have to acquire the method
of recognizing if something is or is not gold. He can rely on a
special subclass of speakers. The features that are generally
thought to be present in connection with a general name--necessary
and sufficient conditions for membership in the extension, ways of
recognizing if something is in the extension .... etc.--are all
present in the linguistic community considered as a collective
body; but that collective body divides the 'labor' of knowing and
employing these various parts of the 'meaning' of 'gold'. 2
Consider, then, an individual in our linguistic community who
has the 'job' of wearing a gold wedding ring but who lacks the
ability to tell whether or not any given sample is really a sample
of gold.
422
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
Putnam's view, which does seem quite plausible, is that the word
'gold' as uttered by this person means gold. What must be explained
is just how this word as uttered by someone who cannot reliably
identify samples of gold can have this meaning. According to
Putnam, when the person in question acquired the word 'gold' she
did not acquire anything that fixes the extension of this word as
the set of all gold things (TMOM, p. 229). Therefore, he apparently
reasons, this person must, in using the word 'gold' to mean gold,
be relying on certain other members of our linguistic community,
namely, those who have acquired something that fixes the extension
of this word as the set of all gold things. But if individuals rely
on others in this way, then it is clear that our linguistic
community divides not only the labour involving gold but also the
labour associated with the word 'gold'.
1I It is sometimes suggested that Putnam's view that there is
a
division of linguistic labour has absurd consequences. Putnam
holds that an individual can rely on other members of his or her
linguistic community to secure the meanings of his or her words.
But now suppose that there is an individual who, despite being in
other respects a more or less competent speaker of English, not
only lacks the ability to tell whether or not any given sample is
really a sample of gold, but knows simply that the word 'gold' is a
mass noun in the language of her community. This person might well
take to pointing to things and saying, 'Is this gold?'; and she
could also, if she wished, say things like, 'Gold is wet', or 'Gold
is dry'. Surely, however, it would be absurd to hold that the word
'gold' as uttered by this person means just what it does in
English. But if Putnam were right that the labour of knowing the
meaning of this word rests with a small subclass of English
speakers, a subclass that is presumably doing its job, would it not
have to be concluded that the word 'gold' as uttered by this person
does in fact mean gold?
Some might respond to this objection by insisting that there
simply is no distinction between what a word as uttered by a
particular individual means and what the word conventionally means
and, hence, that it is perfectly correct to hold that the word
'gold' as uttered by the person in question means gold. Putnam
himself, however, presumably
423
-
CATHERINE J.L. TALMAGE
would not respond in this way. In 'The Meaning of "Meaning'" he
says explicitly that 'we don't assign the standard extension to the
tokens of a word uttered by Jones no matter how Jones uses the
word' (THOM, p. 246), and this would seem to be a straightforward
denial of the claim that the notions of conventional meaning and
word meaning are to be equated. But if Putnam allows (as I think he
should ~) that these notions are distinct, then he is obliged to
take this objection seriously.
Although Putnam cannot dismiss the objection outright, he can,
it seems, argue that it is founded on a superficial reading of the
passage quoted in the preceding section. In that passage, Putnam is
concerned not with everyone who is, in a broad sense, a member of
our linguistic community, but only with those who have a 'job' in
the 'factory' concerned with gold. 4 But what is involved in having
such a 'job'? As Putnam himself suggests, to have a 'job' in the
'factory' concerned with gold is to consider gold to be important
for some reason. Moreover, everyone in our linguistic community who
does consider gold to be important must, Putnam claims, have
acquired the word 'gold'.' But what is involved in acquiring this
word? In a later passage, Putnam identifies two ways in which
speakers inform others of what the words they utter mean (TMOM, p.
229). First, a speaker may give another person an ostensive
definition. So, for example, someone who means gold by the word
'gold' may point to a sample of gold in the presence of someone
else and say, 'This is a sample of gold'. Second, a speaker may
give another person a description. So, for example, someone who
means gold by the word 'gold' may say to someone else, 'Gold is a
bright yellow metal'. Now the individuals who are addressed in
these two examples clearly cannot be counted as having acquired the
word 'gold' unless they understand fully what has been said to
them. In particular, they must understand that the word 'gold' as
uttered by the individuals who addressed them is a natural kind
term. But in what does such an understanding consist?
The following answer seems to be in keeping with Putnam's own
analysis of natural kind terms (see TMOM, pp. 230-33, 243 &
250). In the first example, the person must understand that what is
really salient to whether a particular thing can correctly have the
word 'gold' applied to it is not, say, whether it has the same
superficial features as
424
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
the sample that was identified ostensively, but whether it has
the same hidden structure as this sample. Similarly, the person in
the second example must understand that what is really salient to
whether a particular thing can correctly have the word 'gold'
applied to it is not, say, whether it fits the description 'bright
yellow metal', but whether it has the same hidden structure as the
things to which this description is causally connected. Thus it
would seem that acquiring the word 'gold' involves acquiring an
appropriate notion of what is salient to the correct application of
the word, either by coming to associate the word in an appropriate
way with an ostensively identified sample of gold or by coming to
associate the word in an appropriate way with a description that is
causally connected to samples of gold. ~
Through a consideration of certain views expressed by Putnam, we
have, then, arrived at a basic understanding of what is involved in
acquiring the word 'gold'. But given this understanding, it would
seem that Putnam can easily answer the charge that his view that
there is a division of linguistic labour has absurd consequences:
If a person in our linguistic community knows only that the word
'gold' is a mass noun in English, then she cannot have had this
word defined for her either ostensively or through the use of a
description; but if she is not acquainted with either an
ostensively identified sample of gold or a description that is
causally connected to samples of gold, then this person cannot have
acquired an appropriate notion of salience. Without an appropriate
notion of salience, however, this person cannot be counted as
having acquired the word 'gold'; but if she has not acquired the
word 'gold', then gold cannot be important to her, which is just to
say that it cannot be the case that she has a 'job' in the
'factory' concerned with gold. But then given that he holds that it
is only those who have such a 'job' who can reap the benefits of
dividing the labour of knowing the meaning of the word 'gold',
Putnam can, it seems, consistently deny that this word as uttered
by the person in question means what it does in English.
III This answer to the original objection seems perfectly
satisfactory.
Unfortunately, however, it gives rise to a second, more serious
objection. Given what has been said, it would seem that the notion
of
425
-
CATHERINE J.L. TALMAGE
salience acquired by any individual who has acquired the word
'gold' will in fact fix the extension of this word as the set of
all gold things. For instance, in the first of the two examples
given in the preceding section, the person who is addressed
acquires the word 'gold' by acquiring a notion of salience that
implies that something is in the extension of this word if and only
if it has the same hidden structure as the sample that was
identified ostensively (i.e., a particular sample of gold). This,
however, directly refutes Putnam's claim that members of our
linguistic community commonly acquire the word 'gold' without
acquiring anything that fixes its extension as the set of all gold
things. But if this claim is refuted, then Putnam has no grounds
whatsoever for holding that some individuals rely on others to
ensure that the word 'gold' as uttered by them means gold. And if
there is no division of linguistic labour in this case--the very
case Putnam uses to explain his view--then there is no reason to
believe that there is a division of linguistic labour in any
case.
IV One may be reluctant to accept the foregoing conclusion, for
one
may think that by rejecting the view that there is a division of
linguistic labour, one renders oneself unable to account for the
fact that it is possible for an individual to mean by a word just
what the experts mean by it even though he or she lacks the
experts' knowledge. Consider, for example, the person in our
linguistic community who has the 'job' of wearing a gold wedding
ring but who cannot reliably identify samples of gold. Now suppose
that this person has knowledge of the hidden structure of a certain
bracelet (i.e., suppose that she knows such things as the atomic
weight, relative density and melting point of the stuff out of
which the bracelet is made). Despite having this knowledge, this
person would be unable to tell whether or not the bracelet is
really gold. 7 But how can it be that someone who has expert
knowledge about the bracelet nonetheless lacks the experts' ability
to tell whether or not it can correctly have the word 'gold'
applied to it? Putnam's answer would of course be that this person
lacks the experts' ability because, unlike the experts, she has not
acquired anything that fixes the extension of the word 'gold' as
the set of all gold things; and he would use this answer as the
basis for his claim that this person is
426
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
able to use this word to mean gold solely by relying on the
experts. But how can one account for this case if one refuses to
invoke a division of linguistic labour?
In fact, one can quite easily do this. For suppose, for example,
that the person in this case holds that what is really salient to
whether a particular thing can correctly have the word 'gold'
applied to it is whether it has the same hidden structure as the
things to which the description 'bright yellow metal' is causally
connected. Given this, there is an obvious answer to the question
posed in the preceding paragraph: the person is unable to tell
whether or not the bracelet can correctly have the word 'gold'
applied to it because, although she has knowledge of the hidden
structure of the bracelet, she lacks knowledge of the hidden
structure of the things to which the description 'bright yellow
metal' is causally connected. Now presumably the knowledge that the
person lacks is possessed by the experts. However, it clearly does
not follow from this that the person is linguistically dependent on
the experts, for it is apparent that what she is relying on them
for is information about certain things in the world and not
information about the meaning of a word. It may be concluded,
therefore, that this case involves only a mundane division of
labour.
V I have argued that if one reflects on what is involved in
acquiring
the word 'gold', one comes to see that there is no basis for the
claim that some members of our linguistic community rely on others
to ensure that they mean by this word what is conventionally meant
by it. I will now consider more closely why the view that our
linguistic community divides the labour associated with this and
other words has come to be so widely accepted.
As we have noted, it is possible for an individual to mean by a
word what the experts mean by it even though he or she lacks the
experts' knowledge. Now I suspect that some of those who have
endorsed Putnam's view that there is a division of linguistic
labour have taken themselves simply to be acknowledging this
possibility. But since one can, as we have seen, acknowledge this
possibility while accepting the existence of only a mundane
division of labour, it is clearly a mistake for these people to
claim that they agree with
427
-
CATHERINE J.L. TALMAGE
Putnam. Most of those who have endorsed Putnam's view, however,
have understood that it consists in more than the acknowledgement
of this possibility. They believe, with Putnam, that the mundane
division of labour that exists in our linguistic community
engenders a division of linguistic labour. But why exactly do they
believe this?
The reason they believe this is, I think, that they are
committed (consciously or not) to a traditional idea about the
nature of linguistic meaning, namely, the idea that the meaning of
a word is to be equated with a conception that is associated with
the word. To see how a commitment to this idea gives rise to the
belief that the mundane division of labour engenders a division of
linguistic labour, consider once again the person who has the 'job'
of wearing a gold wedding ring but who cannot reliably identify
samples of gold. Now this person does not in fact associate with
the word "gold' a conception that fixes its extension as the set of
all gold things--she, we may assume, conceives of the stuff to
which she holds it correct to apply this word as a bright yellow
metal. On the other hand, the experts presumably do associate with
this word a conception that properly fixes its extension. Thus if
one accepts not only the claim that the person in question means
gold by the word 'gold' but also the traditional view that for a
word to mean gold is for it to be associated with a conception that
fixes its extension as the set of all gold things, then one will
quite naturally conclude that the meaning of this person's word is
secured by something that the experts possess.
It may seem that we have at last found a way to support Putnam's
claim that there is a division of linguistic labour. But,
unfortunately, the crucial idea is one that Putnam himself rejects.
Recall the first stage of his famous Twin Earth thought experiment.
In 1750 Oscarl, who has acquired the word 'water' by being shown
samples of water but who lacks knowledge of the hidden structure of
these samples, conceives of the stuff to which he holds it correct
to apply this word as a colourless, tasteless liquid. Moreover,
since there is in Oscarl's linguistic community no one who is
expert about this stuff, the other members of this linguistic
community who have acquired the word 'water' presumably conceive of
the stuff to which they hold it correct to apply this word as a
colourless, tasteless liquid. Accordingly, if one holds that this
word cannot mean water unless there is associated with
428
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
it a conception that fixes its extension as the set of all
samples of water, then one will have to claim that no one in
Oscarl's linguistic community means water by the word 'water'.
But suppose that Oscarl has a Doppelganger in the present.
Oscar3, as we shall call him, likewise conceives of the stuff to
which he holds it correct to apply the word 'water' as a
colourless, tasteless liquid. However, in Oscar3's linguistic
community there presumably are individuals who have knowledge of
the hidden structure of this stuff. Hence if one believes both that
it is possible to mean what the experts mean without knowing what
the experts know and that meaning is to be equated with associated
conception, then one will quite naturally hold that Oscar3 does
mean water by the word 'water' and that this is possible just
because the conception associated with this word by other members
of his linguistic community fixes its extension as the set of all
samples of water. And in doing so, one will, of course, be claiming
that there is a division of linguistic labour in this case, one
that is indeed engendered by a mundane division of labour.
How would Putnam respond to this way of treating these two
cases? There is no question that he would object to the assessment
of the first case, for he holds explicitly that even though the
word 'water' was not in 1750 associated with a conception that
fixes its extension as the set of all samples of water, Oscarl does
mean water by this word. But in holding this, Putnam is clearly
rejecting the idea that meaning is to be equated with associated
conception; indeed that this traditional idea must be rejected is
the central thesis of 'The Meaning of "Meaning'". How, then, is it
that Oscarl means water by the word 'water'? The fact that it is
water to which he applies this word does not on its own ensure that
this is the case,.for if he applied~this word to water while
holding that what is really salient to whether something can
correctly have this word applied to it is whether it fits the
description 'colourless, tasteless liquid', he would not mean by
the word 'water' what is conventionally meant by it. The only
plausible explanation is that Oscarl means water by the word
'water' because in acquiring this word through the use of ostensive
definition, he acquired a notion of salience that fixes its
extension as the set of all samples of water.'
429
-
CATHERINE J.L. TALMAGE
This brings us to Oscar3. Would Putnam likewise object to the
assessment of this second case? In fact, given what he says about
the word 'gold', it seems likely that instead of objecting to this
assessment Putnam would wholeheartedly endorse the conclusion that
Oscar3 is able to mean water by the word 'water' solely by relying
on other members of his linguistic community. But are there any
grounds for distinguishing between these two cases in this way?
After all, Oscarl and Oscar3 have both acquired the word 'water' by
being shown samples of water. The only real difference is that
Oscar3 can find out from others the hidden structure of these
samples, whereas Oscarl is without this resource. This, however, is
just to say that there exists in Oscar3's linguistic community a
mundane division of labour that does not exist in Oscarl's
linguistic community. But why suppose that this mundane division of
labour engenders a division of linguistic labour? The only reason
for supposing this is, it seems, a commitment to the idea that it
is only because some members of Oscar3's linguistic community
conceive of the stuff to which the word 'water' is correctly
applied as H20 that this word as uttered by Oscar3 means water. But
such a commitment is clearly inconsistent with a rejection of the
idea that meaning is to be equated with associated conception,
something that is, as we have seen, implicit in Putnam's discussion
of Oscarl. It must be concluded, therefore, that Putnam's view that
there is a division of linguistic labour rests on a traditional
idea that he himself repudiates.
This conclusion forces a reassessment of the significance of
'The Meaning of "Meaning'". Putnam himself says that its
significance lies in its having drawn attention to the fact that
philosophers of language have traditionally overlooked two
contributions to the determination of extension--the contribution
of the real world and the contribution of society (TMOM, pp. 245
& 271). Our investigation supports the claim that the real
world contributes to the determination of extension; but it also
makes it clear that this contribution is mediated by the beliefs of
individual speakers. Consider, for example, Oscarl and his
Doppelg/inger on Twin Earth, Oscar2. Oscarl, let us suppose, holds
that for something correctly to have his word 'water' applied to it
it must have the same hidden structure as the things that were
identified ostensively when he acquired this word (i.e., samples of
H20).
430
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
Similarly, Oscar2 holds that for something correctly to have his
word 'water' applied to it it must have the same hidden structure
as the things that were identified ostensively when he acquired
this word (i.e., samples of XYZ). Neither Oscarl nor Oscar2 has
knowledge of the hidden structures of the relevant samples, and so
each one conceives of the stuff to which he holds it correct to
apply his word 'water' as a colourless, tasteless liquid.
Nonetheless, Oscarl and Oscar2 have acquired different words:
Oscarl's word 'water' is correctly applied to samples of H20,
whereas Oscar2's word 'water' is correctly applied to samples of
XYZ. This shows that the conception associated with a word is not
necessarily the word's meaning, for it may be the case that what
the word means depends crucially on how the world is. However, the
fact that the real world can contribute to the determination of the
extension of a speaker's word does not make the speaker
linguistically at the mercy of the real world, since the real world
can make no contribution at all unless it figures in the speaker's
own notion of salience.
What, then, of the alleged contribution of society to the
determination of extension? What Putnam's Twin Earth thought
experiment reveals, I have argued, is that there is no such
contribution. The experts conceive of the things to which words are
applied in ways that are more sophisticated (and sometimes more
accurate) than are the ways in which others conceive of them. But
if we appreciate the force of the idea that the real world
contributes to the determination of extension--if, that is, we come
to acknowledge the fact that meaning cannot be equated with
associated conception--we will not be tempted to see the experts as
thereby knowing more about meanings than others do, but simply as
thereby knowing more about the things of the world. Hence, in my
view, Putnam has failed to see the full implications of the central
argument of 'The Meaning of "Meaning'". Once the full implications
of this argument are appreciated, Putnam's claim that there is a
division of linguistic labour is perhaps best seen as a telling
reminder of just how difficult it is 'to stop thinking of the
meanings of words as conceptions that are associated with them.
431
-
CATHERINE J.L, TALMAGE
VI In 'The Meaning of "Meaning'" Putnam sums up his discussion
of
the notion of a division of linguistic labour with the following
suggestive remark:
...there are tools like a hammer or a screwdriver which can be
used by one person; and there are tools like a steamship which
require the cooperative activity of a number of persons to use.
Words have been thought of too much on the model of the first sort
of tool. (TMOM, p. 229)
If, as I have argued, there is no division of linguistic labour,
then the analogy Putnam suggests must be rejected. But once we deny
that words are like steamships, should we then hold that they are
like screwdrivers? This is, I think, a somewhat better analogy; but
it too is misleading. There is, it seems, just one basic way to use
a screwdriver to drive in a screw. So, it seems, each person who
uses a screwdriver successfully has acquired the same basic skill.
Thus this analogy suggests that each person who uses a word to mean
what is conventionally meant by it must have acquired the same
thing. This suggestion is not, however, in keeping with the
findings of this paper, for the discussion in section II makes it
clear that although each person who uses the word 'gold' to mean
gold must have acquired a notion of salience that fixes the
extension of the word as the set of all gold things, there is not
one particular notion of salience that each one must have acquired.
In light of this, it seems that it would be more appropriate to say
that words are like ropes--just as one can learn to tie any number
of different knots and still succeed in securing something with a
rope, so one can acquire any of a number of different notions of
salience and still succeed in meaning by a word what is
conventionally meant by it.
These remarks reveal just how little the view developed in this
paper has in common with the traditional account of linguistic
meaning. Traditionally it is held, first, that whatever determines
the extension of a word is its meaning and, second, that what does
determine the extension of a word is a conception associated with
it. I
432
-
IS THERE A DIVISION OF LINGUISTIC LABOUR?
have argued that Putnam's Twin Earth thought experiment shows
that this second idea is mistaken, that associated conception does
not always determine extension. In addition, I have identified
something that does determine the extension of a word, namely, the
speaker's own notion of what is salient to the correct application
of the word. But since, as we have seen, two individuals can mean
the same thing by a word without having acquired the same notion of
salience, it would, in my view, be wrong to conclude that the
meaning of a word is to be equated with the speaker's notion of
salience. Thus the traditional idea that whatever determines the
extension of a word is its meaning must, I believe, be rejected.
But if this traditional idea is mistaken, then how should we
conceive of meanings? The question itself is ill-founded, I think,
for the simple reason that there are no such things as meanings.
Thus, in my view, there simply is no point in trying, as Putnam
did, to determine the meaning of 'meaning'.
UNIVERSITY OF BRITISH COLUMBIA VANCOUVER, BRITISH COLUMBIA
CANADA V6T 1Z1
NOTES See for example Tyler Burge, 'Individualism and the
Mental', in Midwest Studies in Philosophy IV: Studies in
Metaphysics, P.A. French, T.E. Uehling, Jr. and H.K. Wettstein
(eds.) (Minneapolis: University of Minnesota Press, 1979), p. 117;
Akeel Bilgrami, 'An Externalist Account of Psychological Content',
Philosophical Topics Vol. XV, No. l (Spring 1987), p. 200; and
Donald Davidson, 'The Social Aspect of Language', unpublished
manuscript, p. 3. While Burge is sympathetic to the idea that
linguistic meanings are determined socially, Bilgrami and Davidson
are not. Hilary Putnam, 'The Meaning of "Meaning'", in Mind,
Language and Reality (Cambridge: Cambridge University Press, 1975),
pp. 227-28. Hereafter this article is cited in the text as TMOM.
That the notions of conventional meaning and word meaning are
distinct has been argued persuasively by Donald Davidson. See
especially 'A Nice Derangement of Epitaphs', in Truth and
Interpretation: Perspectives on the Philosophy of Donald Davidson,
E.
433
-
CATHERINE J.L. TALMAGE
LePore (ed.)(Oxford: Basil Blackwell, 1986), pp. 433-46.
Davidson's arguments make it clear that it is not necessary to
equate word meaning with speaker's meaning in order to draw a
distinction between it and conventional meaning. Although Putnam is
surely right not to be concerned with every member of our
linguistic community, he should perhaps extend his discussion to
cover those who, despite not having a 'job' in the relevant
'factory', are nonetheless qualified for such a 'job'. (Extending
the discussion in this way would have no effect on the conclusions
reached below.) Of course it is in principle possible for someone
in our linguistic community to consider gold to be important
without having acquired the word 'gold', for he or she might
somehow have acquired another word (presumably peculiar to his or
her own idiolect) that has the same meaning. Putnam, however, is
clearly excluding such exceptional cases. One can, of course, have
an appropriate notion of what is salient to the correct application
of the word 'gold' without being able to give an explicit statement
of that notion. What is crucial is that one use the word as a
natural kind term, and not that one be able to explain what a
natural kind term is. I am assuming, as Putnam surely is, that
someone who has the 'job' of wearing a gold wedding ring but who
cannot reliably identify samples of gold does not have knowledge of
the hidden structure of gold. My reason for assuming this is that I
take it that if someone did have this knowledge then he or she
would have, or at least be qualified for, some 'job' in the
'factory' concerned with gold other than that of wearing a gold
wedding ring. It is not entirely clear that this is the answer that
Putnam would give, however. Although Putnam notes that this case
does not involve a division of linguistic labour in any ordinary
sense, he claims that it can be construed as involving a division
of linguistic labour across t ime (TMOM, p. 229). Thus Putnam might
hold that Oscarl is able to mean water by the word 'water' solely
by relying on certain members of a future linguistic community.
This view, however, is highly implausible, for surely Oscarl would
have been able to mean water by the word 'water' even if no one in
the future had taken an interest in discovering the hidden
structure of water. In any case, that Putnam makes this suggestion
is, I think, evidence that he has not completely abandoned the idea
that meaning is to be equated with associated conception.
434