The compounding effect of reading well

The name of this project -- 1.01 raised to the power of n -- is an attempt to capture something about how understanding actually develops. Small improvements, sustained over time, compound into something qualitatively different than what you started with.

This applies to reading in a specific way.

Two ways to read

Most people read for information. They read to find out what something is, how it works, what to do. When they find the answer, they stop reading. The book is a container; once you have extracted the relevant content, its work is done.

The other way to read is for understanding. This means engaging with the argument being made -- not just the conclusions, but the reasoning. Why does the author believe this? What assumptions is the argument resting on? Where might it be wrong? What would I believe if I disagreed?

Reading for information produces knowledge. Reading for understanding produces the capacity to think.

These are different compounds. Knowledge decays when you stop using it. The capacity to think, once developed, persists and grows.

Why rereading matters

Reading for understanding is also why rereading is so valuable.

When you read something for the first time, you have a limited base of experience to connect it to. You understand the words but may not have the context to see what is actually being claimed. Some passages land. Others go past.

When you return to the same text with more experience -- with more problems encountered, more decisions made, more patterns seen -- the same words mean something different. Not because the text has changed, but because you have. You can now connect the argument to things you have seen that you had not seen before.

This is why the best books about foundational things reward return. "The Mythical Man-Month" at the start of your career is interesting. At ten years, it is haunting. You recognize the situations. You remember the projects. The abstractions land with the weight of things you have personally experienced.

The slow and fast return

The problem with reading for understanding is that the return is slow and invisible in the short term.

When you read a tutorial, you can deploy what you learned in an afternoon. The feedback loop is hours. This produces a real and satisfying sense of progress.

When you read Naur on programming as theory building, the return is not visible for months or years. You carry the idea, it shapes how you reason about a problem, and one day you make a decision differently because of it. The causal chain from reading to outcome is long and hard to trace.

This is why most people default to reading for information. The feedback is immediate. The learning curve is clear. The sense of progress is real.

The slow compounding of reading for understanding produces something more valuable, but it requires accepting that you will not see the return for a long time.

What "reading well" actually involves

A few things help:

Read the argument, not just the conclusions. The conclusions are often available in a summary. The argument is where the value is. Why does the author believe this? What evidence are they drawing on? What is the structure of the reasoning?

Write as you read. Not summary -- response. What do you agree with? Where are you uncertain? What does this connect to that you already know? Writing forces you to form and hold specific thoughts, rather than letting the words flow past.

Return to things. Plan to read the most important material at least twice, with time in between. The second read will be a different experience.

Read primary sources. The closer to the original thinking, the better. Summaries and explainers are useful for deciding what to read, not for doing the reading. The ideas worth internalizing are worth getting directly.

Go slowly. Dense writing is dense for a reason. If you are reading faster than you can think, you are extracting information, not building understanding.

None of this is efficient in the short run. In the long run, it is the most efficient thing you can do.

1.01 to the 365th power is 37.8. The same base raised to zero is 1. The math is unambiguous about what accumulation does over time.