Everything Is Just Dandy!

Where Does Meaning Come From?

Alexander R. Galloway

Nobody knows, not really. I’ve said in the past that no one has yet patented a Meaning Machine. And when it comes to the digital and the analog I’m mostly in the dark about the question of meaning. My intuition is that meaning is not exclusively subsumed by either digital or analog phenomena. Meaning seems to be one of those things that spans both the digital and the analog, even if it might need to be thought differently in both contexts. (And here I’m continuing some thoughts that emerged from recent conversations with Beatrice Fazi.)

Understood semiotically the digital is undeniably a site of meaning. Meaning proliferates differentially within sign systems, and is then stored within the signified. But, at the same time, the analog is typically tapped as a source of "deep" and "real" meaning. Recall the romantic aura around the real, or nature; or recall the way analog reality functions as source of truth in empiricism. If pressed I would likely say that meaning comes from a kind of synchronized coordination between digital and analog elements. The analog does (ideological) work by serving as authentic source of value. The digital structures and regularizes that value into storable/archival/communicable signs and symbols, a process that is also deeply ideological.

Yet I remain a vulgar Marxist on this point. Recall how Marx explained the relation between machines and value. For Marx, machines are a bit like batteries: they don’t generate value; yet they can store value deposited in them. Later, during the production process, these machines will tend to leech a bit of their value into the products they produce. We know this because machines run down and have to be repaired, which is to say, replenished with the value they have lost. To repeat, for Marx machines can’t generate value, only store and transmit value. (And, no, I don’t think the Grundrisse changes this basic equation.) I find this an elegant theory because it allows Marx to retain the labor theory of value, while also situating machines as important agents within production and the value chain. Although it *is* a humanism, which many people find distasteful.

Now let’s take a metaphysical leap and replace value with meaning. (That’s assuming a lot I admit, but let’s go with it for the moment.) If Marx were to speak about today’s machines he would say that digital machines might be a site of value but they can never be a source of value. Or after the leap: digital machines are a site of meaning but not a source of meaning. This is what it means to affirm the labor theory of value in the digital age.

Because of this I’m frequently skeptical toward enthusiasts of AI and automation, especially when they claim that robots or algorithms can create meaning out of whole cloth. (Finance guys have their own version of this discourse: markets create meaning.) No one has yet patented a Meaning Machine. Google can’t create meaning, even if it can effectively extract meaning from a network where meaning was previous deposited. It’s a fun game to play: listen for someone saying they’ve found value or meaning, and you will inevitably find occluded evidence of previous labor. It’s an extraction logic not a production logic.

(A more crucial question is whether Marx’s labor theory of value is incompatible with classical semiotics, which claims that meaning is generated differentially within sign systems. In other words if semiotics were true and Marxism false, then digitality could be both a site and source of meaning.)

I mentioned last time Digital I and Digital II as twin processes, Digital I being the introduction of any form of legible difference whatsoever, and Digital II as the regularization of difference under the name of the general equivalent. This naturally leads to another crucial question: is Digital II always social? Does Digital II always presuppose culture or society? Yes, I think that’s the gist of it, or at least we know that symbolic economies are most interesting when they entail culture or society, as they most always do. Sovereignty, sexuation, language — they all seem inherently social. The digital is social as well, even things as seemingly innocuous as the integers or the alphabet. Foucauldians and STS scholars will remind us that even the integers are technologies with their own histories and varying sociality, Platonism and Pythagoreanism be damned. Yet all the while discretization tries to convince us that it’s one of the transcendental categories, that discretization stands somehow prior to or otherwise outside human experience.

And of course if you ask a Deleuzian, culture and society are part of Digital I as well. So it’s culture all the way down.