Every week or so, I overhear – may even participate in – a conversation about “class”, and it rapidly becomes clear that nobody involved – least of all me – has any very clear idea about who’s working-class, who’s petit-bourgeois, who’s “middle class” whatever the hell that means, who’s haut-bourgeois, who’s ruling class.

Perhaps the problem is that we conceive of classes as categories into which people fall, rather than aspects or “moments” of the relation individuals stand in with respect to the processes of exploitation. For example, a bus driver when he’s driving his bus stands in the relation of “worker” to the Transit Authority, but if he’s invested in a few houses in Queens and rents them out, he stands to his tenants in the relation of small-time landlord, which is surely echt petit-bourgeois.

Similarly, a “knowledge worker” – a computer programmer, say – toiling away in a cubicle maze for Bank of America is distinctly a wage slave as regards his conditions of work and the sadistic brutality with which his boss is apt to treat him. But he’s also making more money than the bus driver, probably, because he knows Tensorflow or what have you. Is this intellectual capital? Is he a capitalist, or at least a rentier? Is he a worker in the cubicle, and a petit-bourgeois on his commute, or at his gym?

Then suppose you got into the carpenter’s union because your cousin was a shop steward, and now you make a union wage at what most people would call a “working-class” job, rather than the exiguous pittance that your non-union counterpart gets? Is your cousinhood a kind of capital?

There’s a certain tendency to treat such cases as if they were liminal – as if these individuals dwelt in some ill-defined no man’s land between the well-mapped trenches of “class”. And there’s another solution, which is the vulgar reliance on income: The bus driver is working-class because his hourly wage is comparatively low, and the computer programmer is petit-bourgeois because he’s making four times that.

The former approach reifies “class”; the latter collapses it into a position on the income curve.

I’d suggest that the best way to view “class” is that it’s an abstraction – a valuable abstraction, like mass or momentum or potential energy or impedance – which reflects the various ways individuals interact with the underlying machinery of exploitation in the society. If you take the view (as I do) that “class” affects “consciousness”, then it follows that each such relation an individual has is apt to have some reflection in his consciousness.

But whatever a “class” may be, it’s not a collection of people.