Here's a commonplace of philosophy of language: Contra Humpty Dumpty, you can't make words mean anything you like just by privately intending them to be understood differently. If I want my words "please open the window" to mean "please close the door" I will need to say or do some suitable thing change the context, and prepare my listeners (e.g. say "let's now speak a modified version of English in which `open' means 'close' and `window' means door"). What an utterance of mine can mean thus has something to do with how I can reasonably be hope to be understood.*
If you apply the same idea to fictions, so that the properties of a fiction have some systematic relationship to what reactions the author could reasonably hope to inspire in their community this suggests a weird possibility. The existence of scholarly bodies studying modern languages (a fairly recent phenomena), might have very literally expanded the realm of literary techniques which it is (metaphysically) possible for writers of those languages to use.
Here's what I have in mind: there might be patterns which one could not reasonably expect to be found by a dedicated single reader, but could reasonably expect to be found by a dedicated community of readers, building on each others work, using word-frequency searches, and being financially motivated (by the drive to publish) to explore even the most remote conjectures. If so, then the existence of English departments might make it the case that a text can do something by displaying one of these patterns (e.g. allude to Joyce, make it true in the fiction that the series of words displaying the pattern was not produced by chance) while a text with the very same pattern could not have had this function before.
*of course there are many books worth of caveats and subtleties to be dealt with in figuring out exactly how to phrase this