Bourdieu's work does not always distinguish between written and spoken language, but it is clear from his writing that the kinds of controls he talks about (above and elsewhere) are primarily of the written kind, though control of spoken language, i.e. correct pronunciation of official French is also involved. That the French state is much better at controlling a linguistic marketplace of written language, rather than spoken language, however, is clear even from the above presentation. Given what else he says about the evaluation of linguistic products' it is also clear that written tokens of language will be more harshly evaluated than will spoken ones, and because they are also scarcer, they have higher symbolic value. But attention to subtle differences of pronunciation will also result in evaluations that can only benefit those speaking the legitimate' language, in its standard (Parisian, or at least northern) pronunciation.
``Though a large part of language is invariable, there exist, whether in pronunciation, lexicon, or in grammar, a whole set of differences associated meaningfully with social differences which, though negligible to the linguistBourdieu faults most linguists, from Saussure on down (though he probably exempts sociolinguists) for accepting standard language as the domain of their enquiry, and ignoring the fact that they legitimate it by focussing exclusively on la langue and ignoring la parole. , are pertinent from the sociological point of view, because they enter into a system of linguistic oppositions which retranslates a system of social differences." (Bourdieu op. cit. p. 41.)