“He moved in a direction best characterized as “down”, through the myriad potentialities of earth, and into the clustered improbabilities, and finally into the serried ranges of the impossibilities.” (Robert Sheckley, “Dimension of Miracles”:1968)
If language is only to be equated with a particular lingual (tongue; & anthropic) utilisation, & not difference-in-general (I would say ‘system of differences’, but am wary of overly restrictive metaphysical conceptions of systematicity), then I could accept a convention of increasing scope that goes from language to semiosis to differance, etc.
But my own use is much more flexible. Whether ‘language’ refers to more, or less, specific conceptualisations; whether it references the use conventions of this or that writer or tradition; all these possibilities of discursive inhabitation should be available, but I’m wary of ossifications that might exclude other traditions of consideration, with differing metaphysical distributions of signifiers, as it were, that are not quite so central anymore.
It is not a question of subjective caprice, but rather that of retaining an openness in practice, & not merely settling into the signifying habits of even the most radical discourses, already to varying degrees assimilated, even if not adequately understood, perhaps.
If language is, actually, a ‘system of differences’, I don’t see a problem with its metonymic use for difference-in-general. Surely, context is an adequate guide to which sense is being used? There are languages, & language-in-general. E.g. Zoosemiotics can be seen as the study of animal languages or communication, etc..
On the question of “linguistic idealism”, surely this would depend on what conceptualisations of language are being referred to, & not the signifier “language”, itself. There is no essential link between the signifier,” language”, & those conceptualisations. In order for “linguistic idealism” to obtain, “language” would require characterisation solely according to a metaphysics of Idealism; or, if reduction to the ‘idea of language’ is being referred to, what limiting criteria could such an ‘idea’ or conceptualisation be truly said to have, when language & the sphere of linguistic operation, are both ultimately indeterminable & so ‘universal’?
If that “multiplicity of modes of semiotisation”, concerns those unconscious economies & layers without a tongue, perhaps giving them the signifier, ‘language’, now & then, can give voice to the unspoken silence of their signs, & “linguistic enunciation” need not be condemned, anymore, to its soliloquy of lone import.