What concerns me is how society privileges and emphasizes the abstract over the contextual.

Some examples of this:

STEM vs the Humanities: Employers and society in general tend to favor and admire people with STEM backgrounds over humanities backgrounds. Even within STEM, the "hard sciences" are seen as more valuable than the "soft sciences", where hardness and softness are largely judged by how much abstraction they require (with, for example, physics and math being seen as very 'hard').

Securities Fraud: As Matt Levine likes to say, "Everything Is Securities Fraud". Why?

"When a company harms some class of people in some bad but essentially non-monetary way, it is hard to quantify the damages as money; when a company harms its shareholders, you can just see how much the stock went down."

The more measurable - that is, the most abstractable - the harm, the more likely you are to get justice for it.

Even within the financial industry, a highly abstract domain, securities stand out as especially abstract and thus valuable:

Patrick Mackenzie writes:

"I think our level of awareness is less than where it should be at the moment. Partially this is for Seeing Like A State style reasons. Securities are so legible . They have conveniently surveillable mark-to-market prices and trade continuously. [...] Loans, on the other hand, remain quite illegible."

bitsaboutmoney.com/archive/dep…

Deposit franchises as natural hedges

Many observe that banks seem to be blowing up due to predictable consequences of rising interest rates. How did we get here?

^{Patrick McKenzie (patio11) (Bits about Money)}

Speaking of Seeing Like a State, that book is full of examples of overvaluing abstraction and legibility - and it also shows that this preference is nothing new.

For example, Scott documents the rise of fiscal forestry:

"The early modern European state, even before the development of scientific forestry, viewed its forests primarily through the fiscal lens of revenue needs. [...] Missing, of course, were all those trees, bushes, and plants holding little or no potential for state revenue."

In Germany, this "fiscal lens" was imposed on the forest itself. "To this end, the underbrush was cleared, the number of species was reduced (often to monoculture), and plantings were done simultaneously and in straight rows on large tracts. The German forest became the archetype for imposing on disorderly nature the neatly arranged constructs of science."

The neat and orderly nature of the German forests mean unskilled (and thus cheaper) labor crews could carry out maintenance and logging.

Yields were more uniform and easy to turn into standardized products. The forests were easier to study scientifically: "Now that the more complex old-growth forest had been replaced by a forest in which many variables were held constant, it was a far simpler matter to examine the effects of such variables."

These abstracted German forests were better for many human purposes, but worse in one crucial way—they could not survive. Germans had to invent a new term: waldsterben, or "forest death".

Given the long growth cycles of trees, it took nearly a century for the fact that the forests were dying to become clear. Or perhaps I should say, "clear to the people in charge". Who knows what the plants and the animals and the people who relied on the forest knew?

This is a recurrent problem with abstraction. Because it is inherently more visible than the contextual, and because it is more likely to be wielded by those in power, it can be hard to alter.

Ideally, the abstract is valued in equal measure to the contextual. Just as abstractions can help make the contexts of our individual lives better, so can feedback from individuals and contexts about issues with the abstractions make the abstraction better.

But we do not value abstraction and context equally. And so the forests die.

As someone in tech, I am particularly concerned with how tech perpetuates this dynamic.

The tech industry constantly tries to commoditize human needs, replacing real relationships in all their complexity and interdependency with context-less, ephemeral "relationships" mediated by apps. Instead of asking a neighbor or a new friend for help fixing a broken screen door, we hire someone on TaskRabbit. Instead of asking someone we know to cook us something when we're sick, we order a meal online.

The Amish are remarkable for their care in determining what technologies they'll adopt. This applies to engineering innovations like cars and televisions, but also to new social technologies. For example, starting in 1955, the Amish refused to pay Social Security taxes.

"The Amish objected to participating, in part on the basis that they believed they were religiously obligated to take care of each other and should not be transferring that obligation to the state." - David Friedman

Of course, money itself is itself an abstraction, one with tremendous impact on social relationships over the millennia.

I'm going to quote extensively from Aaron Bady's review of David Graeber's Debt (thenewinquiry.com/blog/david-g…)

"Is in the very nature of a question like “What do I owe my parents?” that there is not and can never be a final, numerically answer. It is a question that we re-visit and re-negotiate every minute we are with them; obligation and love form an endless Möbius strip"

If money is an abstraction that tends to replace human relationships, capitalism is a way of structuring a society that rewards and accelerates that replacement—and modern, highly financialized capitalism the most extreme version of that.

See again Matt Levine's contention that everything is securities fraud—at the rate we're going, maybe someday soon, it will be.

I'm going to pause here - there's a dozen different directions I could go with this and I'm not sure which one to take. 😂

Wait let me try a poll, let's see if anyone fills it out.

Where should I take this thread?

• Abstraction v Context on the left (38%, 5 votes)
• Psychological drives behind preferring abstraction (30%, 4 votes)
• How power influences abstraction vs context (61%, 8 votes)
• Ideas for centering context when building tech (61%, 8 votes)

That's an interesting take. I'll admit my default is towards abstraction and while I can see that it has its issues, I can see the issues in context more glaringly. Context has high variance. If you're well-embedded in a good community with plenty of people willing to help you, outcomes will probably be better than depending on an average bureaucracy. But a lot of people aren't. Vulnerable people who are unlikable or friendless or lonely, without systems to take care of them, simply don't get the help they need.

This was most starkly brought to my mind by a post I read recently, remarking that some disabled people aren't allowed to be unlikable, because if they are, it may mean death to them (less dramatically, not eating that day, not being helped to the toilet, etc).

I think my society tends to prefer contextual, individual solutions. Until recently, nursing homes were seen as family failure, for example. Having seen what this can do to people (abandonment, care burden) it's very hard for me not to come on the side of abstraction.

But it's totally fair to remind us that it also has its costs.

Apologies if this is too wordy or takes things in a different direction from which you were thinking.

@modulux

No need to apologize - I enjoy when people engage with my ideas, even if they have differing opinions.

And I'm not sure we disagree here. This thread is focused on the failure modes of abstraction but there are failure modes in over-valuing context too.

@modulux

That said, some people (including myself!) have a tendency to see the failure modes of context and think "we can fix this with abstraction!" And sometimes this is true, but sometimes it's not, and what you actually need is to improve things at the context level.

To use your nursing home example: yes, institutional and societal support is so important in reducing unequal care burdens but also professional nurses cannot actually replace the experience of being cared for by loved ones.

I wish I could find the HN comment by the Honeywell engineer that talks about like 6 different safety features none of the open source DIY thermostats have but I cannot seem to find it anymore. Stuff like

- it needs to lock out instead of continuing to try to start the gas heat so you don't flood the house with gas if the ignition is faulty

- it needs to automatically turn on the heat no matter what if if gets too cold

- it needs a mechanical safety feature to shut itself off if the heat gets stuck on

etc etc

@feld My Netatmo doesn't do all that, and neither do the other thermostats connected to the other boilers I have access to.

Fundamentally, those are things the boiler handles (if it's less than 20 years old), not the thermostat. My boiler, even though it's 10 years old, has all these internal features; it also controls the water temperature, etc.

The thermostat is just for deciding if the house is cold or not - I expect all the other controls to be handled by the burner itself.

In other words, I'm not replacing the internal thermostat (the one that regulates the flow temperature, etc.), just the room thermostat. Besides, my sensor is located very far from the boiler, which is in a separate, ventilated room.