III. THE ARBITRARINESS OF
LINGUISTIC SIGN
The epistemological premiss of linguistic conven-
tionalism, for all the authors thus far recalled, is the
explanation of the semantic relation given by Aristotle
at the beginning of De interpretatione: names are the
conventional signs of ideas, but ideas are the natural
signs of things. The semantic relation is accordingly
validated by the natural sign's mediation between
name and thing. This is the premiss for numerous
projects of an artificial language, the best known being
the Ars signorum of George Dalgarno (1661) and the
Essay towards a Real Character and a Philosophical
Language of John Wilkins. Analyzing the mind's con-
tents, drawing up tables of categories of all simple and
complex ideas, then assigning a symbol to each of these,
one could, it was thought, obtain a language which,
eliminating the mediation of words, would be free of
the ambiguity and uncertainty of human languages.
Contemporary epistemological inquiries, however,
were working in the direction of a criticism of the
Aristotelian view of the relation between name and
thing, and in general of a criticism of the idea of
language as the mirroring, the phonetic translation, of
a scheme of natural signs. The semantic investigations
of Hobbes (Human Nature [1650], Chs. 5, 13; Levia-
than [1651], Chs. 4-7; Logic [1655]) concern rather
the theory of truth than the nature and function of
language. Nevertheless, at least two aspects of Hobbes's
doctrine contributed to the subversion of the Aris-
totelian conception of the semantic relation. In the first
place, the idea of reasoning as “computation,” in which
signs are the essential things: in this perspective it is
no longer thought that conditions speech, but on the
contrary it is the use of signs that conditions thought.
In the second place, the idea that universality pertains
not to things or to ideas but only to names: with the
consequence that the universalizing function is a func-
tion not of thought but of language.
The development of this second theme is one of the
fundamental motives of Lockean semiotic (Essay on
Human Understanding [1690], III). The meaning of
words is the ideas they stand for. But the collections
of ideas are the product of an abstractive function
which is itself arbitrary. It is arbitrary because it forms
collections of simple ideas which have no real pattern;
this is the case with the names of mixed modes. Or
else it is arbitrary because the collection of simple
ideas, though it have a real pattern, yet never expresses
this pattern but only the nominal essence, that is to
say a pattern, determined by the choice of the speakers,
which never reaches to the real essence of the thing
and does not even exhaust its properties. The semantic
relation, therefore, is never stable and exhaustive: the
choice made in the linguistic act never rests on the
real essence of the thing; and the determination and
the range of significance vary from time to time, ac-
cording to the needs of communication, and the state
of knowledge, and current linguistic usage. Through
this dynamic conception of meaning, Lockean semiotic
constitutes the crisis of Aristotelian conventionalism,
anchored in the theory of the idea as natural sign of
the thing; and becomes the first radical affirmation of
the arbitrariness of linguistic sign.
