62 music, philosophy, and modernity
was not immediately what it became and is’ (Herder 1985 : 445 ).^6 In
order to understand the very different constitution of different natu-
ral languages Herder, like his friend Hamann, considers languages in
terms of their bearing the traces of their historical origins in concrete
encounters between a people and their specific world. Responding to
the argument for the divine origin, Herder contends that the fact that
language transcends what any individual could invent is not a result
of its being created by God, but is rather a result of language being a
product of ‘whole millennia’ (ibid.).
This temporalised perspective moves him away from the idea that the
primary function of language is what is expressed in Leibniz’s idea of
a ‘universal language’ (ibid.: 499 ), towards his expressive conception:
‘Are there not a thousand characteristics in one language and millions
of traces in the difference of languages precisely of the fact that peoples
gradually learned to think through language and gradually learned to
speak through thought?’ (ibid.: 448 ). In this formulation Herder, who
elsewhere has trouble with the representationalist problem, that he
identifies in S ̈ussmilch and Condillac, of how humankind can have rea-
son without already having language and have language without already
having reason, gets close to the kind of circumvention of the problem
proposed by Rorty. Instead of there being a point of wholesale tran-
sition to language or to thought, the expressive means of the species
develop in a manner in which those means and intentional states can-
not be separated because each is part of the other, there being no
scheme/content relationship between them. How, then, does Herder
see the development of language?
One crux of Herder’s approach to language – his notion of ‘Beson-
nenheit’, ‘reflection’ – is echoed by Rorty’s remarks on metalanguage
as what allows one to talk of somethingasxorasy. Reflection, for
Herder, enables one to foreground a characteristic of something, to
‘see something as something’, which means that it can also be seen as
something else. The notion underlies the fundamental claim of inferen-
tialist semantics, namely that terms gain their sense by their inferential
relations to other terms. Herder’s sheep in theEssaygains its identity
by having the characteristic of bleating, which means it is not some
other animal with its own characteristic noise. As Taylor argues ( 1985 :
230 ), Herder’s approach therefore presupposes a holistic view of lan-
guage, which leaves open the question of how, if we need the whole of a
6 For a general outline of Herder and Hamann on the origin of language, see Bowie ( 2003 ).