|
|
||||
|
Primary:
John Stuart Mill (1867) from A System of Logic.
Alexius Meinong (1904) from ‘The Theory of Objects’. Secondary: Skorupski,
J. (1989) John Stuart Mill, London; Routledge.
|
||||
Mill’s
Theory
|
||||
|
I’ll
start by introducing Mill’s theory so that we know where the
discussion that follows is going to end up. Mill
has a theory of meaning for proper names, that is names such as
‘Stephen’, ‘Archimedes’, ‘Everest’, ‘Fido’, and so on.
He thinks that their meaning is entirely referential, and that the
reference of a proper name is just the thing that it names. Thus the
reference and meaning of ‘Everest’ is the mountain of that name, and
the meaning and reference of ‘Fido’ is the dog of that name. This
sort of theory is called a theory of Direct Reference, because,
as Kaplan says[1], it claims that the terms
to which it applies refer … directly without the mediation of a Fregean Sinn
as meaning. If there are such terms, then the proposition expressed by a
sentence containing such a term would involve individuals directly
rather than by way of … ‘individual concepts’ or ‘manners of
presentation’ … Some
of this definition refers to controversies which we’ll look at in the
next few weeks, but the general intention is clear enough. It’s a
direct theory if there is no other object that the referent (the thing
referred to) has to go through to connect to the referring term – not
meanings, not, ideas, not nothing. We’ll see that there are theories
of direct reference other than Mill’s, but only after a few detours. |
||||
Motivating Mill’s Theory |
||||
|
Anyway,
now that we’ve seen where we’re going, let’s see how Mill gets us
there. Why is Mill interested in names and what problems does he think
need to be addressed by a theory of meaning? There’s a clue in the
fact that his theory is presented in a book called A System of Logic.
He wants to come up with a theory of logic, and of inference in general,
and thereby contribute to a theory about the sorts of things that we can
know. For
Mill, all knowledge is propositional, which is to say that to know
anything is to know that X, where X stands for some sentence that
expresses a positive claim about the world. [Mill’s discussion of
propositions and statements and sentences is confusing in some places
because he never really makes clear whether he’s talking about
propositions as events (like utterances) or as sentences. That needn’t
worry us for now.] Propositions take the general form S is P expressed
in sentences of the form ‘S is P’. For example, ‘Socrates is
mortal.’ (S stands for subject and P stands for predicate.) And the
type of logic that he begins by discussing is the traditional
syllogistic logic that was invented by Aristotle and tells us how we can
make inferences from statements which come in the forms:
S is P
S is not P
Some S are P
Some S are not P
All S are P
No S are P (= All S are not P) In
medieval times the terms that could fit into the positions marked by S
and P might have been called categoremata. Happily for us, Mill
just calls them names. We
can determine the truth or otherwise of syllogistic inferences that are
made on such propositions by treating the names as labels for sets of
objects in the world. There’s a fairly well-known method by which this
becomes a trivial matter (so long as we keep the number of names down to
about three!) This is the method of Venn (or Euler) diagrams. Consider
the famous inference:
p1. Socrates
is a man
p2. All men
are mortal
---------------------------------- c.
Socrates
is mortal We
can check whether that’s true or not by drawing the diagram:
Men
Mortals
The
shading of the lune of the set ‘Men’ indicates that there are no
elements of the set of Men that are not also members of the set of
mortals. And the X in the lens stands for Socrates himself who is an
element in the set of Men and is therefore necessarily an element in the
set of mortals – just as the conclusion to the inference claimed. Now
the important thing here is that the only role that the names play in
the inference is to describe the elements that belong to the
corresponding sets. In set theory we say that the elements of a set are
its extension, and by a natural extension (ha) we also apply the
word to the things that are named by a name; thus the extension of the
name ‘mortal’ is everything that will die, the extension of the name
‘man’ is everything that is human, and the extension of the name
‘Socrates’ is just Socrates (everything that is him.) For
syllogistic reasoning therefore the function of names is exhausted by
their extensions. Mill
concluded, for reasons like this, that the ‘meaning’ of names –
the role that they played in propositions – was entirely referential.
The term which he used for this was denotation. The denotation of
‘Socrates’ is Socrates, and the denotation of ‘Man’ is every
individual thing that is a man. Not the class of things that are human
note, but the things themselves. Perhaps a good way to think of
denotation for Mill is to treat it as equivalent to ‘is truly
predicable of’: thus to say that X is denoted by ‘Man’ is the same
as saying that ‘Man’ is truly predicable of X. Mill’s
denotation was paired with another term, connotation, which Mill
took to be something like the sense that ‘meaning’ generally has
when it is something more than mere denotation. This is a distinction
that Mill was very proud of and saw as the means to solving many
philosophical puzzles. Since we will see something like this
denotation/connotation distinction proposed by others for the same
purposes it will be as well to have an understanding of this earliest
version of the distinction.[1] i.
Connotation
connects a name to attributes: ‘Man’ denotes things which are
human, but it connotes the attribute ‘humanity’. ii.
Connotation,
where it exists, determines denotation:
‘Virtuous’ is a name applied … in consequence of an attribute
which they are supposed to possess in common, the attribute which has
received the name of virtue. It is applied to all beings which are
considered to possess this attribute; and to none which are not so
considered. (VII, 31) iii.
Proper
names do not connote: They denote the individuals who are called by them;
bu they do not indicate or imply any attributes as belonging to those
individuals. When we name a child by the name Paul, or a dog by the name
of Caesar, these names are simply marks used to enable those individuals
to be subjects of discourse. It may be said, indeed, that we must have
had some reason for giving them those names rather than any others; and
this is true; but the name, once give, is independent of the reason.
(VII, 33) And
there are just a couple of points further to be made which will be of
some interest when we come to the criticism of the Millian View. iv.
It’s
possible for a name to have a connotation without a denotation: Mill
seems to admit this possibility but does not have any comment to make on
how to interpret propositions which involve such non-denoting terms.
What would he say about ‘Pegasus inhabits Mount Helicon,’ for
example. v.
All
general names are connotative. vi.
The
meaning of a connotative name is its connotation.
|
||||
|
Problems |
||||
|
The
Millian View has been pretty universally condemned. There are any number
of criticisms of it, but the following are the most often cited. Note
that many criticisms of the Millian view rely upon an assumption that
the meaning of a sentence is somehow derived from the meaning of the
parts of the sentence. (a)
Talking about identity (Frege’s Puzzle) Probably
the most obvious difficulty is seen in identity statements. When we
compare a pair of statements like: a.
Venus
is a planet b.
Mars
is a planet we
know that the difference in meaning between the two is due to the fact
that ‘Venus’ and ‘Mars’ have different meanings. The MV explain
this by pointing out that V and M meanMV different things
because V and M are proper names and their meaningsMV are
just their references and they refer to different objects – to whit,
Venus and Mars. So far, so good. But we also think that the following
statements have different meanings: c.
Hesperus
is Venus d.
Phosphorus
is Venus and
in this case the difference in meaning can’t be explained by the MV
because Hesperus and Phosphorus (the evening and the morning stars) both
refer to the same thing (i.e. Venus). Therefore both H and Ph meanMV
the same thing, and so the sentences must meanMV the same
thing. That means that ‘meaning’MV means something
different from ‘meaning.’ The
difference is even clearer when we compare the sentences: e.
Venus
is Venus f.
Phosphorus
is Venus Our
knowledge of the first statement is purely logical, and we don’t need
to know anything further about the actual object Venus. Our knowledge of
the second statement, on the other hand, is of a different kind. It is
something we need to find out, and it might be false; cf: g.
Mars
is Mars h.
Phosphorus
is Mars Since
the statements express different sorts of knowledge, it is very clear
that their meanings are different. (b)
Talking opaquely There’s another whole class of types of statement
using proper names in which the role of the name in the statement
can’t properly be reduced to its reference. Suppose we have the
following: a.
Cicero
was an orator. b.
Cicero
is Tully. Then
it follows that c.
Tully
was an orator. This
sort of thing we can do because the role that ‘Tully’ and
‘Cicero’ play in the sentences is pretty much exhausted by their
referential role. It is their reference that determines whether those
sentences are going to be true or false. This is a case where the MV
gets it right, and this is hardly surprising because this sort of
semi-syllogistic inference was just what Mill’s theory was intended to
make sense of. In such cases we say that the names T and C can be
substituted salva veritate (ie. that substituting one for the
other does not affect the truth value of the sentences in which they
occur.) We may also say that the names occur in a transparent context,
or an extensiona one. We
don’t hear much more about ‘transparency’ in the literature, but
we do hear a fair bit about the property of opacity with which it is
contrasted. Consider these sentences: d.
John
believes Cicero was an orator. e.
Cicero
is Tully. Here
it is pretty clear that we can’t conclude that f.
John
believes Tully was an orator. Now
Tully and Cicero are the same person, so T and C have the same reference
and the same meaningMV, but it’s pretty clear that the
two sentences d. and f. don’t have the same meaning because it’s
possible for one to be true and the other false – and if there’s
anything that we’re sure of it’s that if two sentences can be true
and false in different situations then they can’t have the same
meaning. Here, if John doesn’t know that Cicero is Tully, then it’d
be quite possible for him to believe that Cicero was an orator and yet not
believe that Tully was an orator. The
problem here is said to arise because the role that ‘Tully’ and
‘Cicero’ play in the sentences is not exhausted by their
referential role, and this fact seems somehow to be marked by the
context in which the names occur. We call this an opaque context.
(Opacity is a term for this phenomenon introduced by Quine. It’s not
much more confusing than other names that have been suggested.) We can
also call these contexts intensional using a coining that
contrasts with extensional seen previously. [Note that this is
intensional with an ‘s’. Intentional with a ‘t’ is quite
different.] Opaque contexts are a great difficulty for the MV. (c)
Talking about existence When somebody says to you ‘Atlantis does not exist’, or ‘there is such a place as Brazil’ you know exactly what the sentence means and you know what it would be like for the sentence to be true or false. But if Atlantis does not exist, then there is nothing for ‘Atlantis’ to refer to, so ‘Atlantis’ has no referent. According to the MV this means that the proper name ‘Atlantis’ has no meaning, and it follows that it doesn’t contribute any meaning to any sentence in which it occurs. This could have a couple of consequences. i.
It
could be that the sentence ‘Atlantis does not exist’ has no meaning
at all, because the fact that some of its parts have no meaning means
that whatever process is responsible for building sentence meanings from
sentence component meanings is short-circuited. This is surely not
right, because we know just what the sentence means, and so it does
have a meaning. ii.
It
could be that ‘Atlantis does not exist’ does have a meaning, but it
has just the same meaning as the sentence ‘God does not exist’,
because the proper names ‘Atlantis’ and ‘God’ make exactly the
same contributions to the sentences in which they occur (ie. None at
all) and those sentences are otherwise identical. Again, I think we can
agree that the two sentences certainly seem to be saying
different thing. We can certainly imagine that one of those sentences
could be true and the other false, and how could this be if they both
meant the same thing? Usually
it is claimed that the first of these options is what happens. But this
has the unfortunate effect that any true statement of non-existence is
meaningless (and can a meaningless statement be true? Pair o’ ducks!)
On the other hand, any true statement attributing existence (to Brazil,
say) must be tautological, since the fact that it is meaningful at all
is the guarantee that the thing it says exists does exist. This, if it
were the case, would be a boon to science. Do electrons exist? Simply
ask if that is a meaningful question. If it is, then they do. Does
phlogiston exist? Do the celestial spheres exist? Do the four humours
exist? (d)
Talking about non-existents There
is a related problem with respect to statements about things that are
known not to exist. When somebody tells you that Pegasus lives on Mt
Helicon, or that Sherlock Holmes lives at 22B Baker Street you neither
assume that Pegasus or Holmes exists nor collapse in puzzlement trying
to understand what is meant by those sentences. But according to the MV,
as we’ve seen, at least one of those reactions would be
justified. There
are a couple of typical reactions to this sort of problem: i.
Some
people may say that, sure, there’s no Sherlock Holmes, and no Pegasus,
but there are undoubtedly the ideas of SH and P. And it is those
ideas that are the referents of the names ‘SH’ and ‘P’. But this
doesn’t seem to be right because when I say something about myself, or
Prime Minister Howard, or the Sydney Harbour Bridge, what makes those
statements true or false are facts about me/Howard/the bridge, and not
at all facts about the ideas of me/Howard/the bridge. Similarly,
what’s needful for the truth or falsity of statements about P/SH are
facts about P/SH and not facts about the ideas of P or SH. If we want
to talk about the idea of Pegasus, we have to mention ‘the idea of
Pegasus’, not ‘Pegasus’ Note
that I’m not saying here that the notion that language gets its
meaning from some sort of relation to ideas completely ridiculous.
We’ll see later that there are good reasons for many theories taking
such a tack. I am claiming that the notion of ideas as referents
in a direct reference theory is not workable. ii.
Another
way of approaching this problem is to claim that Sherlock Holmes does
exist as a fictional man, and that this fictional person is the
referent of the name. Since there is a referent, there is a meaning, and
so the sentence itself is meaningful. But this doesn’t look right,
because a fictional man is not a man, and we really want this referent
to be a man of some sort. When we say that ‘Pegasus is a winged
horse’ we mean this to be a true statement, or at least a meaningful
statement. But if Pegasus is a mythological winged horse then he isn’t
a winged horse. So all statements about Pegasus are false, even the
true ones. Anyway,
one of the advantages of a direct theory like the MV is that it relates
real things like objects with the real things like the names that name
them, and thus it is able to treat meaning as a thing that is explicable
in terms of the sorts of objects that belong to our
scientific/materialist view of the world. (Some people like that.) But
if we have to talk about fictional objects in the world then meaning
becomes something rather sui generis. (Most people don’t like
that.) What
sort of loony would want to defend a claim about the existence of
nonexistent objects? Let’s talk about Meinong.
|
||||
