I have been unduly cruel to Michel Houellebecq for cheap lulz. Appropriating dry technical texts is an entirely valid and often highly entertaining literary technique. And the publicity has actually made me want to read the book.
It was picked up on immediately because he picked a dry technical text that people actually read. I’m presuming here that words in French Wikipedia are subject to the same horrors they’re put through on English Wikipedia and the pattern of traumatised textual flesh is distinctive and obvious.
Spotting and marking for death anything interesting, well-written or showing signs of coherent authorship is, in practice, a reliable heuristic for eliminating puff pieces. English Wikipedia has a house style, and it’s really obvious when someone’s quoting a chunk of it. It’s what happens to text when too many people edit it and all nuance is iteratively wrung out. Also, there’s lots of dangling subclauses as successive writers argue in the article and try to get their favourite nuance or contingency covered. It’s most visible on articles that were made featured a few years ago and have since sunk into dilapidation.
If someone hasn’t studied and written about the Wikipedia house style academically, they damn well should. It would be an interesting exercise for an individual to try to write as badly as an overedited Wikipedia article, so as to make their own fake Wikipedia text for fiction. This may even be amenable to computerisation.