Reading the Book | Hacker News Reader

Reading the Book

cs.umd.edu

52 pointsmotxilo20 comments

Think about this. Who's the person that learns the most from a class? Is it that brainy person that got an A+ on the exam? That person learned a lot. But the teacher, if they're any good, learns more. When you teach something, you can't learn something half-heartedly. You are trying to explain it to the students, so it needs to make sense to you.

Think about this. Who's the person who learns the most in life? Is it that teacher who had to explain it to the students? That person learned a lot. But the builder, if they're any good, learns more. When you build something, you can't learn something half-heartedly. You are trying to actually make it work in the real world, so it needs to make even more sense to you than it made to your teacher.

Alan Perlis: You think you know when you learn, are more sure when you can write, even more when you can teach, but certain when you can program.

That is pretty absurd. Find the most complex thing you have ever programmed and go teach it to someone. Just because you made it does not mean you can explain it. Further, while explaining there is a pretty good chance you will find some bugs.

Yes, there are some impatient people who unreasonably ask books to painlessly teach them complicated material. But there is also a problem that most technical books really are poorly written, and the reason for this is simple: for the people who are most qualified to teach material, there are usually terrible incentives for them to invest the enormous effort it takes to write books.

For instance, in Academia, there is pretty much a penalty associated with top professors writing books. Writing a good technical book takes of order a year, and one year not publishing papers is easily enough to prevent someone from getting tenure. Tenure committees give negligible weight to producing good texts, and the royalties from upper-level books is insignificant, so why would a professor bother?

This won't change unless the incentives are changed.

I have no evidence, but I suspect most textbooks are written by tenured professors for precisely the reasons you've outlined. I would submit that these people (not untenured assistant professors) are the most qualified to write such books. I attribute the fact that most textbooks are terrible to Sturgeon's law (90% of everything is crap), so I don't think changing the incentives will increase the proportion of good textbooks, though it may increase the absolute number of such books.

This is a reasonable suggestion, but I disagree for three reasons:

1. It's a common misconception that professors work very hard for tenure and then, once they get it, are free to do whatever they want. (In other words, that there is some "threshold".) Instead, the system is set up to keep professors on a treadmill until they are very old through grant "carrots". Yes, it's possible once you secure tenure to just hang out and think deep thoughts, but this leads to salary stagnation or even regression. Instead, most professor continually chase money by cranking out research. Taking time off to write books means significantly less salary, (which is not nearly made up for by book royalties).

(This is my impression as a physics grad student. Other department might be a bit different.)

2. I claim that the hardest part about writing books is distinguishing what's obvious from what's not for the student. Once you've been using the material for years, it's very difficult to remember all the tiny mental leaps that are required to get from point A to point B. Yes, you need a mastery of the material to write a book, but sufficient mastery for textbook writing comes with only a few years of research experience (as opposed to, say, review papers which require extra experience because they are more-cutting-edge/less-canonical).

3. Even if tenured professors didn't have a penalty for writing textbooks and they were the most qualified (in contradiction to my first two claims) there needs to be some serious incentive for them to get off their ass and write a whole book. Those incentives doesn't exist right now. A lack of penalty isn't enough.

I have this thing about learning stuff. If I find something hard to understand, it is normally not me it is my source. I have years of proving I could understand stuff, but these authors generally don't have that level of experience in teaching stuff.

I remember in my A-level Computer Science class when we had to explain stuff, the lecturer would give an automatic zero if it didn't come with an example. I carry this practice with me now of giving examples when I'm trying to explaining stuff and non-technical people have commented they understand the concepts behind some of the technical things I write for them. I find that a lot textbooks fail to provide enough examples and analogies; they present just equation and abstract concepts. It amazes me how people find OOP hard to learn. Then I realize that my lecturer was off the chart in explain stuff. Tons of analogy and the like.

So, yea that is how I judge technical books, how many examples and analogies they provide.

Most textbooks are actually badly written. They are written by busy profs who don't put too much time into making the concepts easily accessible. This is because writing a book takes a lot of time, and the author does not get much monetary compensation back. The author gets some recognition, but by spending that much time on grant writing and paper writing, he would get much more recognition and funds.

So the problem could be the book, don't blame yourself for everything. Search for the best book. Recently, I found that often the best book is an online tutorial, resource, blog. My 2 cents.

For many years, I wished for a boss who would hand me a book and say, "here's a book, read up on it." Little did I know that I had to marry him first.

</twee>

Aww, that's so cute I just barfed in my mouth a little.

As a Java programmer, I once was welcomed at a new job: "Here you are. Effective Java. Read it all, cover to cover." Having programmed in Java for a couple of years, it was like a punch direct to my face. And it was awesome.

This whole post is just giving excuses for bad writing. Even though some subjects can be very difficult, a knowledgeable person, who knows how to teach and how to write can produce something useful. Personally, I tend to keep trying different books until I find something that makes sense to me. But I admit this is not always possible with more esoteric subjects. A good text book, like a good teacher, makes a big difference.

As Pliny the Younger proclaimed: "There never was a book so bad that it was not good in some passage or another."

Some well-read fellows called Mortimer Adler and Charles van Doren wrote a book called How to Read a Book.

It serves two useful purposes.

1. It provides a framework for thoroughly digesting important books.

2. It is a rich vein of humour for visiting friends to mine when they spot it on your book shelf.

http://www.amazon.com/How-Read-Book-Touchstone-book/dp/06712...

My advice: if you don't understand it at first, read it again. And again. Follow the references and read them if you need to. Or google what you don't understand, read up on it on Wikipedia, or another source. People give up way too easily.

Some of you might remember the invisible wall in your mind between you and college-level (and higher) math/science books/papers that you had before you went to college. Breaking that barrier has been one of the most important things I got from college.

Nothing is too difficult to understand (someone had to understand it to write a book about it), so don't give up too early.

Some books are bad. However, some of the best books shouldn't be easy to read. I think there is a lot of value in struggling through hard books. At times I think the internet has made things too easy.

Look at any great math book. Large leaps of logic are declared 'obvious', and left as exercises for the reader. While such exercises often aren't easy, struggling with such exercises helps you learn. Before the internet provided tutorials for everything, you had to really struggle when you couldn't figure something out. Yet, I think I really learned to program from that struggle.

You learn to find information that isn't spoonfed to you. When you do, you will better be able to dig beneath hype and propoganda, and make up your own mind. Look for clues, and try thinking for yourself instead of getting all the answers from someone else.

It depends on why the book is hard. Sometimes it's just sloppy thinking and writing; sometimes it is an inherently hard subject. It isn't always clear when starting which it is so it is useful, if you start having trouble to compare the one you are reading with another on the same subject.

For reading that is inherently difficult, Mortimer Adler and Charles Van Doren give some very helpful information in How to Read a Book. Unfortunately it is not an easy read itself.

Another book I found helpful was A Time to Learn by Phillip Bandt, Naomi Meara, and Lyle Schmidt. It has some really good ideas on diagramming complex work to help you understand them, especially how their ideas fit together.

TL;DR

Sorry, I had to, especially since it's kinda true (just glad there was a summary at the end). I guess I'm lucky this wasn't required reading for my job. ;)

As a physics major I never actually read the books; I know this is why I wasn't a straight A student (close enough though) and probably why I took away less than I know I could have.

There were other majors in our department that would beat themselves up over the material, taking hours on end to really understand and apply it (this was impressive). They were the ones with straight A's, not because they were smarter than everyone but because they took the time to rip into the material. I'm always in awe of these kind of people.

TL;DR Don't read the book, devour it.

"In the US, we like to think reporters have no bias, but of course, they do. In Europe, the biases are much clearer."

Not related to the main point of the article, but I don't agree with that. It was very subtle, but in my opinion that is a big lie.

I think it is strongly related to the main point of the article, insofar there are types of books that allow you a broader margin for thinking and constructive critiques, and on the other hand, for technical books this is not so obvious. What don't you agree with? What and why is a big lie?

Reply on news.ycombinator.com