Semiotics

Beyond Signification: The Co-Evolution of Subject and Semiosis 43

with associations of pleasurable satisfaction of hunger. One has experienced them several
times, and from these several experiences one has abstracted a category – a signified is in
place – and now a signifier is needed in order to ask someone to provide an instantiation of
the category. It does not matter too much whether the signifier is going to be chips, french
fries, fritos or skyfies, as long as a serving can be ordered.
A way to understand these onomasiological processes is by means of the distinction
between type and token. Once a cognitive architecture has developed that can hold these two
separate and distinct, then the processes of symbolisation can proceed. Let us consider the
possibility that the animal mind, even at its highest level of development, cannot hold this
distinction, so that for the pre-human creature every instance of its prey, for example, is
identical with the category of prey. This is entirely consistent with the principles of salience
and pregnance. Thom (1983, pp. 273-274) has an original way of explaining this, by saying
that the hungry predator is its prey, in that it is entirely possessed by the internalised image of
its prey (what to the human would rather be an ideal type), and when an instance of this type
is perceived, externally, then the type vanishes, as it were, to be entirely replaced by the
particular token. Since the one merges completely with, or substitutes completely for, the
other, the two are not held distinct at any stage. The animal will not produce a symbol, like
chips, that stands for the object in general, i.e. for the type rather than the various tokens that
instantiate it.
As Thom says, ―man is freed from the enthrallment of things by giving them names‖
(1983, p. 274). The essential Saussurean insight is that it is a matter of relative indifference as
to what these various names are, just as long as there is a sufficient number of different names
for different ideal types. Man can at any and every stage keep type and token distinct due to
the symbols that have been formed out of each idealised type, the precondition for all other
processes of abstraction. Stjernfelt (2002, p. 341) has this way of putting the matter:

Higher animals may not only recognize tokens as instantiations of types, they may make

use of these types to symbolize, to reason, argue, use diagrams. Probably, the special human

privilege is abstraction, making it possible for us to make explicit and contemplate such types,

reasonings, diagrams with any particular token placed in brackets and thus facilitating control,

experiment, and quick development of these signs.

There is a consequence of this abstraction that is highly characteristic of the symbolic
order, which is the power for various kinds of self-reflexivity. Among the things that can be
described by a universal category is any and every aspect of language itself; language is in
this sense self-referential and self-reflexive.^4 One can refer in one‘s discourse, for example, to
a past instance of discourse, what someone said or wrote, their choice of words, and what
they might have meant. One can identify saliences and pregnances in this discourse and
nominalise them, just as one does with other sorts of objects, thereby ̳capturing‘ the
indexicality of their interrelationships, but now far removed from the concrete immediacy of
animal indexicality. It is part of our challenge to understand this more fully.
The symbolic has so to speak ̳reflected into itself‘ its own natural history in the ascent of
man. The very difference of the symbolic from the iconic-indexical is in fact subsumed within
the symbolic order as part of itself. This can perhaps best be understood in terms of the

(^4) ―Semiosis explains itself by itself; this continual circularity is the normal condition of signification‖ (Eco 1976, p.
71). See also Jakobson (1960) on the ̳metalingual function‘.