A few thoughts on college textbooks

11/24/2013 01:44:00 PM
A good textbook is never ever revised.
A couple of things have me thinking about textbooks lately. The first is that I'm gearing up to teach Principles of Macroeconomics soon, and the second is a more long-run project of mine to learn various programming languages, which has me buying a few textbooks on programming. Over the years, I've learned a few principles about how to select and use textbooks:
  1. Use a first-edition textbook published this year
  2. Overlapping textbooks are good.
  3. Don't tie your class to one particular textbook
The first point gets at one of the worst scams a college professor can ever commit: assigning a Second Edition (or third, fourth etc) to their students. The reality is that publishers coerce their authors into "revising" their textbooks every year, for no other reason than to destroy the textbook resale market. This would, to some extent, be acceptable if each revision really meant that the textbook was whole-heartedly keeping up with cutting-edge developments in the field. But the unfortunate reality is that they don't: authors put minimal work into revisions meaning that as the years go by the "revised editions" will become almost as hopelessly outdated as the first edition. If you are in a rapidly changing field like programming, this is totally unacceptable: something from even just three years ago is basically useless, even with the "revisions." If you are in a field like neoclassical microeconomics that has been well-established for decades, then there is no reason at all to have revisions--I'd point to the famous Mas-Colell, Whinston, and Green (MWG), the first year microeconomics graduate textbook used by all English-speaking economics grad schools, as an example. MWG is the bible of microeconomics, and has never ever been revised, because when you do it right revisions can only hurt your book.

The second point is something I learned as an undergraduate. Miami University had a little math library in the math building that contained all the major textbooks for basically all of the math courses that they teach. You couldn't check books out, but they had huge tables and a big chalkboard right in the room, which was even better, in my opinion--you could just show up, grab any book (they were all always there!) and start doing some more practice problems. For many of these courses, like linear algebra, calculus I through III, differential equations etc, the curriculum is pretty standard no matter which textbook you use or who's teaching. What I discovered is that each textbook explains things in a slightly different way, and that it was good not to read a particular textbook, but instead to read the relevant chapters of all the textbooks. Usually, when I didn't grasp something that one textbook was saying, there was a different textbook that explained it perfectly. I think a big fallacy professors often make is trying to judge a particular book's strengths and weaknesses, without recognizing that these depend entirely on who is reading it--what you think is a particularly good treatment on a topic won't get through to some of your students, while other books you may think aren't as good will actually succeed with those students. That is, every student is different, and they should be encouraged to read different textbooks until they find an explanation that makes sense to them. Besides, there were times when I ran out of practice problems for a particular topic long before I really understood the topic--in these cases, I could just find another textbook and keep practicing until I really mastered it.

The third point is closely related to the second. Because every student is different, they should have considerably more say in which textbook they use. I understand that some courses are so specialized that you have no choice but to require a specific text, to make sure everyone is learning the same topics. But courses like Calculus aren't like that. For the most part, every calculus textbook covers the exact same topics in roughly the same order, and students should be encouraged to explore different authors' treatments on the same topics. Moreover, there's nothing new whatsoever being taught in any Calculus course, so forcing students to pay top dollar for a brand new $n$-teenth edition of a textbook is totally insane and a total scam--let them buy whatever used $(n-1)$-teenth edition they find on amazon. Yes, the page numbers may differ slightly, and a few of the practice problems will be in a different order. If you are teaching your class in a way where these cosmetic differences matter, you are doing your students a huge disservice. That kind of teaching deserves it's own special ring in Dante's Inferno.