Theme 3 of 13 · 21 ideas
The same rulebook idea, moved inside your own head. Languages, notations, metaphors, forms and learned habits all tilt what you find easy to think. A dozen fields discovered this separately and gave it a dozen names. The evidence says the tilt is real and measurable, and that it is a tilt and no cage.
The same ground, told as a story: Thinking Inside Frameworks.
The ideas
Each one is a short page. Walk them in this order, or start with whichever catches your eye.
- Framework (a medium for thought): Any structure people build to think with, such as a language, a form, a metric or a model, quietly shapes what they find easy to think.
- Framework pricing (mechanism): A framework makes some thoughts cheap and others expensive, and across a whole population people mostly think the cheap ones.
- Cost asymmetry: Inside any tool or system of notation, some things take one step and others take fifty, and people mostly do the one-step things.
- Defaults: Whatever a system gives you when you make no choice is what most people end up with, so whoever sets the default sets the outcome.
- The Columbia slide (example): Engineers briefed NASA on possible shuttle damage with bullet-point slides, and the layout buried the caveats that mattered most.
- Notation as a tool of thought: The symbols you write a problem in decide which steps are easy, so a better notation puts harder thoughts within reach.
- Leibniz’s dy/dx and the British lag (example): British mathematicians stayed loyal to Newton’s way of writing calculus, while the rest of Europe used Leibniz’s handier symbols and pulled ahead.
- Linguistic relativity: The language you speak nudges what you notice and remember, but it does not decide what you are able to think.
- Russian blues (example): Russian has separate words for light blue and dark blue, and its speakers tell those shades apart faster, unless a word task interferes.
- Pirahã and exact number (example): A people whose language has no number words can still match exact quantities they can see, but slip once the task needs memory.
- Influence, not imprisonment: The tools you think with tilt your thinking in measurable ways, but careful testing shows they do not trap it.
- Framing effect: Describe the same facts in two different ways and people will make two different choices.
- The Asian disease problem (example): Describe the same outcomes as lives saved or as lives lost, and most people switch which plan they choose.
- Conceptual metaphor: The comparison you use to describe a problem quietly suggests which solutions make sense.
- Crime as beast, crime as virus (example): Readers told crime was a beast preying on a city favored tougher enforcement than readers told it was a virus, with the same statistics.
- Theory-ladenness and paradigms: What an expert sees when looking at evidence depends on the theory they already hold.
- Einstellung effect: Once you know one way to solve a problem, that way keeps pulling your attention back and hides better ones.
- Chess experts and the Einstellung effect (example): Chess experts who had found a familiar solution said they were hunting for a better one, while their eyes kept returning to the old one.
- Law of the instrument: People reach for the tool they already have, and then see every problem as one that tool can fix.
- Inductive bias: A system that learns from examples has to make assumptions that go beyond those examples, and the assumptions shape everything it concludes.
- Prompt formatting and the 76-point swing (example): Changing only spacing, separators or capital letters in the wording given to an AI model moved its test accuracy by as much as 76 points.
← The medium sets the rules · All themes · When the map moves the territory →