Abstractions might benefit deeply from language, but I don't see a tautology here. I suspect that anyone who does a lot of abstract thinking (math, software des

Abstractions might benefit deeply from language, but I don't see a tautology here. I suspect that anyone who does a lot of abstract thinking (math, software design, visual arts, etc.) spends time engaging with abstractions that have no names or linguistic references, and which may never manifest in named constructs. It can be a relief to have such a "mental thing" become concrete enough to attract a name -- now we have more tools to manipulate it with, and we can communicate it more effectively -- but that doesn't always happen. And when it does, sometimes you have to discard some of the interesting parts in order to wedge the thing into a nameable compartment: it's not really the same abstraction any more. In software design, sometimes I know that I am working within a certain project-specific "space" in my head that informs the decisions I'm making. It has structure, influenced by the structure of the project and by my problem-solving approach; but there aren't any words in there, and I'd be hard pressed to give it a better name than "a mental model." And yet it's a palpable "thing," it's often unique to the specific project, and it is useful in getting the work done -- I definitely notice when I get distracted and lose track of it, and recognize it when it I find it again. I'm trying hard not to sound all fancy and mystical about this -- it's not magical stuff. It's just tricky to talk about kinds of abstractions that don't lend themselves well to being named. :)