I think a lot of people are missing the point here, or just read the first two points of the article and didn't go further.
TLDR:
- reading many books doesn't mean that you don't read some of them in depth, or for enjoyment
- reading many books quickly is often richer reading a single one end to end
- reading many books is something you can learn and practice, it is not "shallow"
- over time, what remains of a book read end to end and a book skimmed is often the same
---
The first couple of entries focus on reading for the purpose of gaining information, and he expands upon what that means in more detail further down. The whole "5 books a day" is for his work. He brings up that he reads in order to answer a specific question or solving a specific problem, and reading books in cluster, which is in fact more "studying" or "working" than reading.
He explicitly brings up the fun he gets from reading, in particular fiction, and I don't think he skims through 5 novels a day.
In fact, he advocates reading the best book on every subject you don't care about to gain insight and future pleasure from knowing about that domain (or that the first book about a subject that you read, probably more carefully than others, has to make you "travel").
I am in a bit of a similar spot in that I churn through books (mostly programming/systems/science-related, mostly textbooks), and when I work on something, I can "read" through 10 books a day. Not only will those 10 books be about the same subject, and I can pick the one book that is clearly better, but by reading fast and in group, I can focus on concepts, how they are differently examined by different authors, and see relationships between these different approaches. It is a very rich way of reading, and you get better at it over time.
Reading 5 books at once in a day can also be: read one book much more carefully, and skimp around and discard the 4 others as have not much more to say than that first one. Or realizing they are "factoid" books that will be useful when I need something very specific. Or reading the first book, and only using the other 4 in order to get a different explanation for a certain concept.
Sometimes, I will also pick one of those books and study it in great depth, annotating and dissecting pretty much every page. In those cases, I can maybe go through 10 pages a day. The insights into reading and how to extract information from books gained by these hyperfocused readings translates to being able to skim through a table of contents and index and know that that pretty much counts as "having read" the book.
---
Even book I read in detail end to end, 5 years later, I will remember at most one big idea. More importantly, I will remember the overall quality of the book and the table of contents. That's great value, because when I will have to study X, I can zoom in on that book and get my answer.
Say I want to read up on prolog implementations, I'll remember the implementations I wrote from SICP and PAIP, I'll remember bratko and the odd NLP books I had, I'll remember that odd Partial Evaluation book, and that will allow me to get a good start for my renewed interest. As I go back to those books, I'll have further things to look up, say the alice book for some storage/query engines material, potentially transaction processing and ddia if I get into more datalog-y stuff. That's because I spent many days skimming 10+ books about each subject, and discarding most of them as just repeats or as "discoverable" resources (as in, the interesting info in them can just be googled ad hoc).
Building that kind of organic feel for where good information can be found and what "shape" it has can go very quickly after a while. Extracting the valuable information out of them can also be practiced and become extremely quick. Valuable information is the information I will be able to act upon. For technical information, that can often be completely skipping the content and just reading/doing the exercises, that can be taking very "academic" notes (say, math definitions and proofs or algorithms), or very "philosophical" notes (concepts, relationships between concepts, forming or refining personal concepts).
I used to filter a lot by typography and quality of writing as well, but that signal is not as good these days. Plenty of good typography around, but mediocre (even if well edited) content, and on the other hand, great content but shitty direct to print self-publishing.
TLDR:
- reading many books doesn't mean that you don't read some of them in depth, or for enjoyment
- reading many books quickly is often richer reading a single one end to end
- reading many books is something you can learn and practice, it is not "shallow"
- over time, what remains of a book read end to end and a book skimmed is often the same
---
The first couple of entries focus on reading for the purpose of gaining information, and he expands upon what that means in more detail further down. The whole "5 books a day" is for his work. He brings up that he reads in order to answer a specific question or solving a specific problem, and reading books in cluster, which is in fact more "studying" or "working" than reading.
He explicitly brings up the fun he gets from reading, in particular fiction, and I don't think he skims through 5 novels a day. In fact, he advocates reading the best book on every subject you don't care about to gain insight and future pleasure from knowing about that domain (or that the first book about a subject that you read, probably more carefully than others, has to make you "travel").
I am in a bit of a similar spot in that I churn through books (mostly programming/systems/science-related, mostly textbooks), and when I work on something, I can "read" through 10 books a day. Not only will those 10 books be about the same subject, and I can pick the one book that is clearly better, but by reading fast and in group, I can focus on concepts, how they are differently examined by different authors, and see relationships between these different approaches. It is a very rich way of reading, and you get better at it over time.
Reading 5 books at once in a day can also be: read one book much more carefully, and skimp around and discard the 4 others as have not much more to say than that first one. Or realizing they are "factoid" books that will be useful when I need something very specific. Or reading the first book, and only using the other 4 in order to get a different explanation for a certain concept.
Sometimes, I will also pick one of those books and study it in great depth, annotating and dissecting pretty much every page. In those cases, I can maybe go through 10 pages a day. The insights into reading and how to extract information from books gained by these hyperfocused readings translates to being able to skim through a table of contents and index and know that that pretty much counts as "having read" the book.
---
Even book I read in detail end to end, 5 years later, I will remember at most one big idea. More importantly, I will remember the overall quality of the book and the table of contents. That's great value, because when I will have to study X, I can zoom in on that book and get my answer.
Say I want to read up on prolog implementations, I'll remember the implementations I wrote from SICP and PAIP, I'll remember bratko and the odd NLP books I had, I'll remember that odd Partial Evaluation book, and that will allow me to get a good start for my renewed interest. As I go back to those books, I'll have further things to look up, say the alice book for some storage/query engines material, potentially transaction processing and ddia if I get into more datalog-y stuff. That's because I spent many days skimming 10+ books about each subject, and discarding most of them as just repeats or as "discoverable" resources (as in, the interesting info in them can just be googled ad hoc).
Building that kind of organic feel for where good information can be found and what "shape" it has can go very quickly after a while. Extracting the valuable information out of them can also be practiced and become extremely quick. Valuable information is the information I will be able to act upon. For technical information, that can often be completely skipping the content and just reading/doing the exercises, that can be taking very "academic" notes (say, math definitions and proofs or algorithms), or very "philosophical" notes (concepts, relationships between concepts, forming or refining personal concepts).
I used to filter a lot by typography and quality of writing as well, but that signal is not as good these days. Plenty of good typography around, but mediocre (even if well edited) content, and on the other hand, great content but shitty direct to print self-publishing.
(edit: formatting)