language has been privileged as "the system of all systems"; hypertext may function to interrogate this (as though hypertext were outside language)

But "Media for human receptors are always mediated by language" (Jonathan Turner, October 06)

We think without language; and we can also identify non-linguistic systems: e.g., pre-linguistic babies or animals. But at the human level, the privileging of language makes sense if you take "language (metaphorically) to be any human coding" [1]

To use an image rather than text that describes the image challenges the primacy of text. But we still use language to understand what the image means. We still require language for communication. [2]

Language "proper" is "constructed of phonemes or graphemes syntactically"; other experiences, such as listening to music, don't invoke this medium.

What is "natural" language? Isn't all language artificial, dependent on agreed, arbitrary sets of meanings?

"Natural" in this context means languages that have emerged in human cultures, opposed to artificially produced languages such as Esperanto or computer languages.

Some thought is still prelinguistic: e.g., premonitory, or awareness of a mood.

[1] This debate depends on "language" being used metaphorically: e.g., the "language" of music. Structuralist attempts (e.g., early Barthes) to see language everwhere as the universal code have not been very helpful. Most processes are not structured like a language: they don't have syntax, as that is understood by linguists.

[2] Cf. the "dual coding" model of Allan Paivio, who argues for the independent existence of both imagery and language: e.g., Mental Representations: a Dual Coding Approach (1986).