By Tess |
February 01, 2024, edited February 14, 2024
abstractions make complex thought possible. multiple abstractions can often achieve similar but not identical results.
everyone has a tendency to push abstractions that benefit themselves, and powerful entities can be quite effective at this. since our relationship to some of these entities is adversarial, many of the abstractions we use are working against us.
let's call abstractions we've accepted without too much thought "borrowed abstractions".
note that abstractions are distinct from specific ideas about the topic (aka object level beliefs). an example may clarify:
suppose someone says "your career isn't as important as your family, you should have a baby in your twenties"
if your takeaway is you should have a baby in your twenties, you have accepted the object level idea.
if your takeaway is you disagree. you think you shouldn't have a baby in your twenties because your career is more important, you accepted the abstraction.
in the latter case you rejected the object-level argument, but implicitly agreed with the frame that there is necessarily a trade-off between career and family. you may have also adopted that abstraction into your long-term worldview without realizing it.
being good or bad at chess is a popular abstraction, but it's also a fairly obvious one. if we agree on how the game works, it's pretty clear some people consistently win against others. even without being told some people have more skill, you would probably come to that conclusion.
on the other hand, good or bad social skills is a much more questionable abstraction. if you've been told negative social interactions are a signal of poor social skills and you don't question this frame, you might think the only solution to social problems is to change how you interact. however, there might exist alternative abstractions that are more useful. for example you could instead believe each pair of people have a level of compatibility between them. this would indicate the thing to do is find friends you get along with better.
(note I'm not advocating for latter example or object-level conclusion, I'm simply demonstrating how an alternative abstraction can lead to different results)
borrowing abstractions is essential to efficiently building an understanding of the world, but with enough data and thought we can take control of the abstractions we use. we can more deeply understand why we use the ones we do, build new ones and modify or scrap the ones that get in our way.
I call this owning our abstraction stack.
even if a borrowed abstraction is generally useful, there will be times when there's a more optimal way to understand the situation.
consider the first example. maybe after thinking really hard you decide there is a coherent family-career tradeoff for you and many people. with this work done, you're equipped to identify people in unique situations where this isn't the case, and adapt your advice appropriately. if you had borrowed the abstraction without consideration you wouldn't be able to make this call.
there's plenty of cases where more optimal abstractions for everyone are out there, waiting to be discovered or popularized. this is especially the case for topics specific to a single demographic, company, friend group, etc.
other people and organizations tend to gravitate towards abstractions that benefit them. this means in places where the game is zero-sum, abstractions from others tend to be stacked against us. if we put in the work to build our own, we get to choose abstractions that give advantage to us.
adapted from this twitter thread