Mallarmé would argue that this phenomenon is natural: that the language is there, ready to serve, and that ‘things exist, we do not need to create them; we only need to seize the relationships between them’ (‘Les choses existent, nous n’avons pas à les créer, nous n’avons qu’à saisir les rapports’). This is perhaps the most marked Mallarmean claim, uttered in an interview with Jules Huret at the ‘mature’ stages of his career in 1891. By this point, Mallarmé had, tortuously, achieved his dream. He was living and writing in Paris, surrounded by a horde of young disciples who would gather in his home at 89 rue de Rome every Tuesday: avid ‘mardistes’ hungry for their master’s parole. And what he did tell them was literary sacrilege: ‘To name an object is to remove three-quarters of a poem’s joy… to suggest it, that is the dream’ (‘Nommer un objet, c’est supprimer les trois quarts de la jouissance du poème… le suggérer, voilà le rêve’). Indeed, what Mallarmé is suggesting here is that we use the relationships between things to work out the things themselves. To do this, we need to avoid talking about the thing itself. We must not name it, define it nor describe it; instead, we must allude to it by its absence– only the evocative absence of the word can make up for the inadequacy, even redundancy, of language. This complex image could be likened to a spider’s web, from which the spider itself is absent. Even though the spider is not there, the web implies some anterior and inherent presence of a spider and therefore automatically conjures up the image of the spider in our minds. In fact, the spider metaphor is one of Mallarmé’s favourites, and appears throughout his work and correspondence.
"use the relationship between things to work out the things themselves" reminds me of Benjamin’s literary montage and "__ by ___" [what’s the phrase?] method of commenting
Alos: not naming objects and, essentially meaning, no self-reference; spider’s web; implicit