KNOWLEDGE by the conventions which

KNOWLEDGE
by the conventions which specify that 'cat' means cat, and so on. An additional
set of conventions must determine what 'cat' and the other words in the sentence
mean. But now consider that 'The cat is on thc mat' can be included in any
number of literary works. Presumably, in each of these works it can have a
different non-literal meaning. In one work it may contribute to the fact that the
work as a whole states that first impressions are a poor guide to character. In
another, it makes it possible for the work to state, say, that the best laid plans often
go awry. For each of these works, a different set of conventions would be
required. The number of literary works in which a sentence can be used is
potentially infinite. If sentences in works of literature can have non-literal
meanings, a sentence can have a different meaning in each work. An infinite
number of conventions are needed to account for the infinite number of non-
literal meanings of a sentence such as 'The cat is on the mat-' This ig impossible,
and I conclude that sentences in literary works have only literal meanings.
One could object that readers of literature grasp what a work means (and
states) without employing conventions. Perhaps readers grasp what authors intend
and recognise what state without a knowledge of semantic conventions.
Since readers must rely on their knowledge of semantic conventions to discover
what authors intend to say, this is an implausible suggestion. Still, a couple of
linguistic phenomena suggest that grasping what a sentence states is a matter of
grasping the author's intentions. The first of these phenomena is irony, while the
other is the existence of malapropisms. We can understand what people are saying
when they speak ironically or employ malapropisms. In doing so, it seems, we do
not employ semantic conventions. Malapropisms and ironical statements
apparently mean something besides what they literally mean. Instead.
understanding these statements seems to be a matter of grasping a speaker's or
author's intentions. Perhaps something similar happens when we read literature.
Let us consider whether the existence of malapropisms and their meanings
supports the view that sentences in works of literature can have additional. non-
literal meanings. An example of a malapropism will be useful. A student once
said to one of my friends. a punctilious, demanding teacher of English, and a
woman of unimpeachable character, 'You are nothing but a pederast!' It took a
sæond. but my friend soon recognised that the student meant that she is a pedant.
The student's sentence literally means that my friend is a pederast. It seems to
mean, in addition, that my friend is a pedant.
One might think that the existence of malapropisms supports the view that
non-literal meanings exist in literature. Grant that malapropisms have additional,
non-literal meanings. Even if this is so. grasping the additional meaning is not a
matter of divining a speaker's intentions. Rather, when we understand what the
producer of a malapropism means, we rely on our knowledge of the same
character. On this view, literature can employ semantic representation even if it
employs mainly literally false sentences.
If a sentence has two truth-values, it must have two meanings: a literal
meaning and a non-literal meaning. This is the case since the truth-value of a
sentence is a function of two factors. ne u-uth-value of a sentence depends on
how the world is, but also on what the sentence means. In order to see that this is
the case, consider again the sentence, ' ne cat is on the mat-' (Assume that
cat, and nothing else, is on the mat.) This sentence would be false if the world
were, in certain respects, otherwise than how it is. In particular. it would be false
if the cat were not on the mat. The sentence would also be false if it meant
something other than what it does. If, for example, it meant that the bat is on the
mat, it would be false. In general, the truth-value of any sentence depends on what
it means.