How to Buy Books (And Not by their Cover) April 4, 2006
Posted by TimTheFoolMan in Books, Computers, Language, Learning, Science & Technology, Technology.trackback
As a tech-oriented person and avid book owner, I have been constantly faced with the dilemma of how to choose between a myriad of selections at the local bookstore, or when shopping at Amazon or Barnes & Noble online. After reading about Donald Knuth’s methodology for grading comp-sci projects (solving long and complex programming problems), I modified my approach, and have been using the following method for the past few years.
A Highly Recommended Approach
Obviously, if you can get first-hand recommendations from someone else, those are a plus. Most of the books that I’ve bought in the past year are a direct result of listening to a topic discussed in a podcast or at a presentation, and catching a recommendation along the way. Depending on the speaker’s knowledge of the subject matter, I may value a recommendation more or less, but that’s not always a definitive guide. After all, sometimes a bozo discovers a good book, and sometimes a leader in the field will recommend something that is far beyond my level of understanding (i.e. most everything except his own books that he recommends).
Use Your Index Fingers
Unfortunately, I don’t always have a good recommendation to go by. That’s when I fall back on the modified-Donald-Knuth-paper-grading approach. Here’s how it goes.
For a given subject you want to buy a book about, let’s say… the history of cartooning, pick a moderately obscure factoid that you think a good book on the subject should cover. For instance, you may have some knowledge of the political cartoons that were created about Abraham Lincoln as President. Now, hit the Index, and find a page where it’s mentioned, and read the entire page (before and after the factoid reference).
If you repeat this process for all the books in front of you, after a couple of books, you should have a good idea of the writing style the author(s) use. This is an important point. If the factoid appears in the middle of the book (and not in the first or last chapter), then it’s probably representative of the bulk of the text. (Some authors are notorious for writing really good first and last chapters, and taking up space in the middle, so choose a factoid that doesn’t appear there. This is analogous to not buying a US-manufactured car in the 1970’s that was built on a Monday or Friday, but for the exact opposite reasons.)
Table of (dis)Contents
Over time, the method I’ve described here has become second-nature. I still check the Table of Contents, as that gives me a clue about the scope of the subject the book intends to cover. However, I always wait to check the TOC until after I’ve been to the Index. That way, I won’t be tempted by the scope (almost always fluffed up) and lose sight of the quality of the writing (or lack thereof).
A Fool and his Words are Soon Parted 
If I’m in a bookshop it’s a good idea to read a random selection of fragments from across the book. You can easily tell from the prose style and the way the arguments are presented if it’s going to hit the spot for you. Otherwise recommendations and reviews are available on pretty much everything on the web.
Part of the problem with recommendations and reviews (at least the ones I’ve seen) is that people can have an axe to grind with an author for reasons that have nothing to do with the book. Spolsky mentions this in a podcast interview with Doug Kaye, where he laments the “one good book that covers the history of unicode,” but found that most of the online reviews were bashing the book because of Usenet flamewars in years past with the author. I’ve seen some of this myself, where a book I knew to be authoritative was inappropriately slammed.
I have no problem with recommendations, but I really take reviews with a grain of salt. – Tim