A Twitter thread by Venkatesh Rao.

A good hack to find interesting ideas to work on is to start with a word that seems to pick out an important concept, but has been rendered annoyingly vacuous by abusive over extension. Examples: strategy, meaning, irony. Other examples?

Back around 2002, I had a friend in grad school who was studying law and philosophy of language. He was also a Navy vet. We proposed a little project (didn't really get off the ground) to analyzing the concept creep of "strategy" over the decades and reconstruct it more cleanly

The trick to the clean-up exercise is to avoid the temptation to rigorously analyze the history of the concept and propose a cleaned up definition. Thar be legalism and bureaucracy.

Instead, just impose a *private* definition on yourself and stick to it with discipline.

The private definition should be chosen not for Grand Unified Theory potential of the term's history, but pick out and exaggerate an elegant new core in the connotation cloud. It's like caricature mixed with rebasing (in the sense of picking a new origin for a coordinate system)

The aim is to almost achieve a semantic disruption of sorts. If the "market" for a term is all the ways it is used, arranged in some sort of tag cloud, with a bad/bureaucratic vacuous center, you disrupt by picking a more meaningful thing on the margins.

It turned into a shtick for me at some point, and is probably the conceptual basis of how I do a lot of my meme-inception type writing. Eg. Gervais Principle is based on trying to disrupt the word "sociopath"

My book Tempo (which grew out of my postdoc project, which in turn kinda grew out of my false-start project with my buddy to reconstruct "strategy") attempts something similar. It is built around a somewhat idiosyncratic, but not too heterodox core sense of "strategy".

But the BIG trick is to not reveal what you're doing until/unless it actually catches on, and people start saying things like "sociopath in the gervais principle sense". It's like a test of success in the form of taking a word-stock into private equity, then re-IPOing it.

A common pattern I've noticed in terms that are vulnerable to this pattern of concept creep and semantic decay is that they pick out attempts to create agency out of ramification, and there is some actual agency there. Just hard to pick out, acquire, and exercise well.

Take for example, the word "strategy". It attempts to create agency at higher levels of abstraction in thinking about conflict. The agency involved in pulling a trigger is obvious. The agency involved in moving tokens around on a map in a war-room is not.

A good heuristic for picking out this condition is to apply John Wannamaker's observation "I know half my advertising works, I just don't know which half."

This is why "marketing" is hard. The semantic noise induced by the illegibility of the agency sought.

One I'm working on right now is "temporality". Why not just say "time"? Because "time" offers no agency unless you know how to travel at relativistic speeds or shrink yourself down to sub-Planck quantum levels. "Temporality" is a way to pick out locus of agency just above "time"

There are banal things you can do that alter time experience. I'm not even talking drugs or base jumping. I'm talking stuff like drinking coffee or going to sleep. Those alterations are not to bare-metal time proper, but to the lowest level reification that offers some agency.

Most of the terms people are suggesting in the replies: community, agile, data-driven, fit this pattern. They are reifications in search of a locus of agency that may or may not exist, and you won't know till you try to "land" the concept on a pattern of agency that seems to work

How do you know your attempt at semantic disruption is working? Two signs:

Social proof: people resonate wit your sense of usage. They train on, and add to, your use instances.

Material proof: navigating with that concept yields better-than-random results, ie it finds agency

I'd like to try and teach semantic disruption skills at some point, but I don't yet have a clear sense of what kind of preparedness you'd need to pick up on this. It's a bit like being a meme VC, knowing when a concept is right for disruption and that an angle on it is working

Implicit in this model is that conceptual meaning is not something that evolves from fluid usages and eventually "arrives" by landing in a dictionary or other canonical reference work, like an IPO. Conceptual meanings are like companies with stocks on a meaning stock market.

The value of the stock goes up and down as the clarity and agency go up and down through use and abuse. Successful concepts ossify and attract growing abuse, creating disruption opportunities. As a memester, you have to watch the meme markets for "startup" opps basically.