Voyageur Technologique

How to Read a Book

I started read “How to Read a Book” by Adler Mortimer (HRB) after one of the engineers I look up to, Star Simpson, professed the book enabled her to understand many technical papers. As someone who’s desperately trying to hack their way to understanding the field of Cognitive Science via intense reading, I figured trying to polish the art of what I was doing every day would be time well spent.


HRB starts out the way most self-help books do, describing the writer’s own ascent for ignorance to enlightenment. Basically, Adler graduates from university thinking he knows everything. He doesn’t and is distressed. He realises he’s never really “read”, he’s only let information flow through him without connecting it.

From this relatable intro, Adler descends into a rant about the current education system’s failures in the context of “reading”. If you are actually interested in educational reform, read “The Educated Mind” by Kieran Egan, which is way better written and actually provides many practical solutions.

Overall, this book is quite lengthy and repetitive. Maybe that’s just how books were written those days? Or maybe the author was actually afraid of the readers leaving his grasp without any progress? As a result, I gave up after finishing Part II, since he addresses specific works of fiction for the book’s remainder.

Regardless, below is a summary of what Adler considers “reading” and the various techniques required to be a “reader”, instead of just a “consumer”. I’ve tried to make more hacker-friendly than the Wikipedia summary and the book’s summary table.

The techniques can be divided into three “readings” or as prefer to call them “perspectives”, which are each in turn divided into four “rules” or more understandably “approaches”. Each “approach” is to be done continuously while reading the book to aid understanding.

The Analytical Perspective (top-down)

  • Classify the book. What is it’s focus?
  • Summarise the book. What is the most important content of this book?
  • Summarise all the parts (this division might be chapters or more arbitrary divisions). How do each of the parts contribute to the whole?
  • What problems is the author trying to solve?

The Interpretive Perspective (bottom-up)

  • What are the important words that the author is defining?
  • How does the author define these concepts in phrases? What about paragraphs?
  • How do the paragraphs support the author’s arguments?
  • What are the author’s arguments? Can you imagine instances where they hold true?

The Critical Perspective (ways you can disagree with the author)

  • Show where the author is uninformed. Where are they lacking information?
  • Show where the author is misinformed. Where have they poorly interpreted the information given?
  • Show where the author is illogical. Where are the gaps in their reasoning?
  • Find when to withhold judgement due to insufficient evidence, but please define what evidence you would need to be convinced and evaluate if that is reasonable.

I’m particularily fond of criticising the work as the final step. You must understand what the author is doing and how they can hold this world-view. Only then can you truly understand whether it needs to be corrected and in what ways.

When trying to apply this procedure, the usual rules for learning a skill still apply. Don’t try to do them all at once. Try each technique individually and see how you can make it work. Eventually, you’ll be able to do them simultaneously with less effort.

If these instructions aren’t clear enough, then by all means download the PDF and plough through it. However, since life is short, skip Part B where he rambles on about education, which are of no benefit to someone who’s trying to figure out how to “read”. Additionally, be prepared to be beaten over the head repeatedly with basic concepts.


Aside from these rambles on education and these helpful “reading” techniques, Adler also spends a considerable amount of time expounding the benefits of Great Books. He means “reading” classics and associating them together to gain a complete understanding of humanity’s knowledge and it’s epistemology. He claims that to understand any field, one must go to the primary discoverers and read their publications.

If, in the same college, two men were lecturing, one a man who had discovered some truth, the other a man who was repeating secondhand what he had heard reported of the first man’s work, which would you rather go to hear? Yes, even supposing that the repeater promised to make it a little simpler by talking down to your level, would you not suspect that the secondhand stuff lacked something in quality or quantity? If you paid the greater price in effort, you would be rewarded by better goods.

And that the quality of materials, not the quality of mentorship matters not for teaching:

The great books can be read in or out of school. If they are read in school, in classes under the supervision of live teachers, the latter must properly subordinate themselves to the dead ones. We can learn only from our intellectual betters. The great books are better than most living teachers as well as their students.

I’m sympathetic to this point of view. Many failures of Computer Science as a field resulted from not acknowledging the lessons taught to us by those who came before. For examples of this, check out this video by Alan Kay and this other video by Brett Viktor.

However, his claims lack nuance. Surely the distillation of knowledge isn’t completely worthless. Compared to a primary discoverer, a regular user of the knowledge must contribute something meaningful.

I also can’t seem to resolve the idea that one should totally ground themselves philosophically before proceeding experimentally. Especially when I’ve wasted countless hours researching when I really should have just built something. For example, when I was first learning to program, I got stuck in the minutea of different programming languages instead of following literally any tutorial to completion. Surely there is some middle ground between researching and building?

Regardless of this lack of intellectual humility, HRB still useful if you need to read a bunch of things and you get the feeling you’re not quite doing it right. Personally, it’s reassurance alone was worth the cost of entry.