Hello Reader,

The London Underground map is a lie.

Stations that are miles apart look like they are inches apart. Curves become straight lines. The Thames, a winding river cutting across the city, is reduced to a gentle kink. Every geographic fact has been distorted or discarded.

And it's one of the most useful designs ever made.

That's abstraction: reducing complexity by focusing on structure instead of details. Not oversimplification: you don't lose what matters. Not generalization: you don't broaden for the sake of it. You strip away what doesn't serve the goal and keep what does. The Underground map warps distances so you can find your station, trace your route, and count your stops in seconds.

The model does not need to be accurate. It needs to be useful and get you from point A to point B.

Why Use It

Your brain can't hold every detail of a complex system. You think it can. It can't.

When you focus on surface specifics, you drown in noise. When you focus on core structure, you see how things connect. Research on construal level theory shows that abstract thinkers outperform concrete thinkers in pattern recognition and information synthesis because they've organized their information.

Abstraction also enables transfer learning. Once you understand a pattern in one domain, you can apply it elsewhere. Velocity is speed plus direction. Once you hold that, you see it everywhere: marketing is reach plus targeting. Negotiation is terms plus timing. Every abstract structure you build becomes a lens you can carry into new territory.

The thinkers who move fastest across domains aren't the ones who know the most details. They're the ones who've built the most portable patterns.

When to Use It

Deploy abstraction when:

• Complexity clouds understanding. You're drowning in details and can't see how the system works. Strip to essential components first.
• Teaching or explaining. Someone needs to grasp a concept quickly. Build the simple model before adding nuance.
• Designing systems. You're architecting products, processes, or organizations. Abstract models reveal structural patterns before implementation buries them.
• Solving novel problems. You're in unfamiliar territory. Map it to abstract patterns you've seen before.
• Transferring learning. You solved something in one domain and need to apply it elsewhere. Extract the core pattern and test it.

Don't use it when details determine outcomes. Engineering tolerances, legal language, and medical protocols require precision, not simplified models. A subway map won't help you walk between stations. Abstraction is a tool, know when to zoom back in.

How to Use It

In 1931, Harry Beck was a 29-year-old engineering draftsman at London Transport. He'd been let go in budget cuts and was doing occasional freelance work for the organization. The Underground map at the time showed geographic accuracy (distances, real curves, true positions). It was technically correct. It was also practically useless. The central London lines were crammed together in an unreadable tangle while outer lines stretched across empty space.

Beck's insight was simple: underground travelers don't care about geography. They care about sequence and connections. Which line am I on? Where do I transfer? How many stops to my destination? He stripped everything else away. Distances became uniform. Curves became clean 45- and 90-degree angles. The Thames became a thin blue line, present only as a landmark.

London Transport was skeptical. They printed a small trial run. It sold out immediately and had to be reprinted.

Beck made the map less accurate and infinitely more useful. The process he used applies to any system:

1. Identify core components. List everything in the system. Ask: what's essential to how it works vs. what's contextual detail?
2. Strip to the minimum structure. Remove one element at a time. If the system still makes sense without it, that element is a detail, not a structure. Keep removing until the pattern breaks – that's your core.
3. Name the pattern. Give the abstraction a clear label. "Routes and connections" beats "simplified geographic representation." Names make patterns portable and shareable.
4. Map relationships. Show how components connect. Beck used colored lines and station dots. You can use diagrams, frameworks, or simple sentences.
5. Test in new contexts. Apply your abstraction somewhere else. If it clarifies that domain too, you've found a transferable pattern. If not, refine.
6. Document what you lost. Beck's map lost walking distances, neighborhood context, and surface geography. Every abstraction sacrifices something. Know what you gave up – and when you need it back.

Next Step

Map a system you work with daily:

• What are the 3–5 core components?
• How do they connect?
• What details did you strip away?
• What's the thing people actually need that current complexity obscures?

This week: Take one complex system (a product feature, a team workflow, a decision process) and draw it as a simple diagram. Three to five nodes. Lines showing connections. No details.

This month: Find something in your work that's "geographically accurate but practically useless." Redesign it to show only what people need to navigate.

This quarter: Build an abstraction library. When you encounter useful patterns, capture them in stripped-down form. Reference them when new challenges appear.

Where It Came From

Abstraction originated in mathematics and computer science. Mathematicians abstract numbers into algebra, algebra into abstract algebra, each level revealing deeper structural patterns. Computer scientists use abstraction layers to build complex systems: hardware abstracts physics, operating systems abstract hardware, and applications abstract the OS.

Construal level theory formalized how abstraction affects cognition: people thinking abstractly make better aggregation judgments and transfer learning more effectively than those stuck in concrete details.

Harry Beck's 1931 Tube map became one of the most influential design artifacts of the 20th century. It has been adopted by transit systems worldwide. Beck was paid five guineas for it (~£400 in today's money). He never received royalties. But the insight outlived the oversight: strategic reduction beats comprehensive accuracy, every time.

Share or Join 👉

• Discover hundreds of mental models in the Re:Mind App.
• Share this newsletter with a friend.
• Get insights from me on Instagram or Linkedin.