One of Graham Harman”s comments on a recent post of mine was that he felt the use of a partly symbolic notation - all those Ps and Qs - was a rhetorical fail: the gains in precision, such as they were, were outweighed by the loss in vividness. Vividness is a virtue in writing that seeks to make sense of the world through metaphysics, partly because it allows the unruly variety of the world itself to impinge on the conceptual harmony of the proposed system, and partly because it enables the system itself to enlist unexpected allies from among the range of referents it calls as witnesses. If desiccated symbolic notation shrinks from trials of strength in which it must stand up to concrete referents and persuade them to support it, vivid writing attracts and compels support through a mixture of glamorous allure and ingenious conceit.

Part of me wants to object that there are certain kinds of arguments that can only be made with the assistance of abstract language and formal notation. I’m not sure how one might explain the concept of a generic subset without it, for example. You can present someone with the image of a bowl full of very heterogeneous things - bacteria, lightbulbs, raisins, cat whiskers, little cubes made of ivory - and that will give them a fairly good starting point, but it’s still necessary to explain how it is that a generic subset isn’t constructible (and why this matters), why it has to be infinite and so on. When you have firmly established all of the salient properties of a generic subset, it’s then fully tooled-up as a battering-ram for knocking down people’s assumptions about counted-as-one multiplicity being the correlate of some act or other of human cognition, since a generic subset is a very determinedly irregular sort of mathematical object which it’s difficult to imagine being the correlate of anything in particular. It’s actually a good example of something that can be demonstrated to exist but never fully represented as such: the image of a bowl full of bus tickets, fork handles, anchovies, ammonites, dark matter, dandruff and maple syrup can only ever represent a finite, constructible part of an infinite and non-constructible multiple.

The value of this is that it equips you with one very effective way of saying “yah boo sucks” to the argument that anything you can think is just the thought of something thinkable by you: clearly, “thinking up” the generic subset (establishing the procedure whereby the possibility of its existence is asserted) is not the same as “thinking of” it (possessing a representation of the thing you’ve shown to be able to exist). The “speculative” aspect of “speculative realism” seems to me to have to do with just this sort of “thinking up” (epitomized by Meillassoux’s style of reasoning in After Finitude), and mathematical or very technically abstract argumentation is evidently a very powerful ally in that venture. It is equally “realist” inasmuch as the very purpose of such argumentation is to drive a wedge between “thinking up” and “thinking of”, thereby reasserting the possibility of speaking of a reality independent of our powers of conceptualization without falling into philosophical naivety or incoherence*.

How many times does this point need to be made, though (a question every Derridean must have asked themselves at least once)? And what in particular can you go on and do once you’ve made it? As Graham suggests, it may be better to take it as a fait accompli - to downgrade one’s assessment of the philosophical threat posed by correlationism - and just get on with talking about the world and the things in it, in as vivid and engaging a fashion as one can. I don’t mean to suggest that the style of argumentation pursued in Prince of Networks is somehow merely chatty or blas?: it entails a supremely careful balancing and evaluation of stances, and argumentative moves of startling audacity and strangeness. But because it is not a Houdini-like attempt to unpick the locks of a logical straitjacket, it can afford itself world enough and time to unfold a rich panoply of conceits and figures.

I haven’t altogether lost hope that a copy of Logics of Worlds will arrive on my doorstep some time in the next couple of months, and it will be fascinating to see how Badiou’s determinedly eccentric mixing of poetic description (of French opera, slave revolts and ivy-clad walls at eventide) and technical formalisation plays out at length in that book. He is still a model writer for me, both as a polemicist (The Meaning of Sarkozy is utterly tremendous) and as a maths teacher, but I suspect that “Being and Event II” will be the culmination of that style.

* It’s possible that the universe is not only inconceivably weird, but also weird in a way that no rational procedure can even demonstrate as a possibility. This I think is where being and thinking must finally take their leave of each other. Meillassoux and Badiou seem equally unwilling to take that step, preferring to elect their respective axioms and principles as the guardians of rational consistency and sense. The madness of mathematical reason is still a reasonable madness, when all’s said and done.