You too can write a book

Thanks to Neeldhara Misra for pushing me to write
this post, based on a thread on Twitter.

This article is primarily directed at academics. Its purpose is to
tell you two things:

1 You Can Write a Book🔗

Let’s say you’re not using someone else’s textbook, or using it only
loosely. That means you’re going to spend a lot of time organizing
your thoughts. You will probably produce some kind of “lecture
notes”. The delta from there to a book is much smaller than you
imagine.

Here’s a pro-tip. Back in about 2003/4, I noticed that the quantity of
reading that students would do before class was at most about six
pages; once it got to about eight pages, they wouldn’t read at
all. (These numbers may be much lower now.) But this automatically
bounds how much you have to write!

In short: let’s say you’re writing up lecture notes. You’re writing
about four to at most six pages per class. Let’s say you have about 30
classes (often many more). You have automatically written about 200
pages without even especially thinking about it. Two hundred pages of
writing is often called a … book. It represents your “take”. So your
take now has a book!

2 You Should Write a Book🔗

What are the incentives to do this? There are many, but they
may not accrue immediately: they may take time. Think of it as a
long-term investment in yourself.

• First, simply: you believe strongly in your view of the world,
  and you’re pursuing it with intensity. Right now nobody else is really
  able to download your brain. Your book becomes how others can download
  it.

• People at other places might use your book, or at least put it on
  reading lists. Even if only one student there reads and internalizes
  that supplemental material, that student now carries your ideas with
  them. Much more concretely, they could be a PhD applicant.

  I have gotten so many great PhD applicants over the years thanks to my
  books! In particular, when they come from a less-well-known university,
  this guarantees for me that they have the preparation I need, and that
  we have a shared mindset.

• What’s out there may not be very good. This is especially an
  issue in programming languages, where some of the widely-used texts
  are basically hot garbage: basically a broken, 1970s view of the
  world. I once wrote a
  position piece on it
  entitled “Teaching Programming Languages in a Post-Linnaean Age”! So
  worldview really matters.

  So that’s another incentive. To drive out the bad with good. Do you
  use the standard text that everyone else uses in your field? If you’ve
  read this far, probably not. Why not? You know (or think you know!)
  what should be taught in your field. Who uses a text? Someone less
  certain. So it’s not their fault. We need to help them along.

• Sometimes we have resources—like software—that many people
  don’t know about, but that are especially well suited to education in
  our field. For instance, the #lang feature of Racket
  is one of the most powerful tools for teaching programming
  languages. But most people don’t (yet) know that.

I speak from experience. I have written several books, some solo and
some with co-authors. I did the very thing you are told to not do as a
tenure-track assistant professor: I wrote a quality
undergraduate-level book. I survived (it didn’t hurt my tenure in the
slightest), and I then benefited from it for a long time.

3 Mechanics🔗

The one big thing I haven’t said, which drives a lot of this,
is the publishing medium. And oh boy, do I have opinions on this!
I wrote up some of them when I published the first formal edition of
my programming languages book, in an essay entitled
Books as Software. Let me summarize/expand.

• Don’t go to commercial publishers. They are either just not
  ready for the modern world or will basically paywall your
  work. Paywalling is the total antithesis of wanting to have impact and
  influence and wanting to drive bad books out of the market.

• Publish it free online. Especially those of us who are
  immigrants from poorer countries know what it’s like to not be able to
  afford high-quality material. The next you is sitting right now in
  Bangalore stuck with a crappy course and crappy book. Be their light.

• Some people like paper. Upload your file to a print-on-demand
  service. Even with a markup, it’ll be a lot cheaper than a book from a
  commercial publisher.

• In a STEM subject, your tenure case is not going to hinge on a
  contract from MIT Press. Getting that contract is actually relatively
  easy. Impact is hard. Go for the hard part, as you do in your
  research. Optimize for it.

• Put out a new release once a year. Don’t fall for the temptation
  of continuous releases. People using your book need to be able to
  depend on a fixed version for the semester. They will have references
  to pages, sections, etc. Don’t break their build.

• Publish permalinks. Not everyone can upgrade their course every
  year. Yes, it means your old mistakes are on permanent display, and
  some people won’t use your latest and greatest. Live with it. Your ego
  is not that fragile. You’ll get over it.

• Make it easy for people to send you corrections. They will (just
  as they did for your software). Sometimes you will even get very
  insightful and creative ideas. Of course you’ll also get various
  dreck. Just as with your software. Because books are software, as my
  essay says.

• Try to provide materials in both PDF and HTML. The reasons
  should be obvious. It’s not always easy. I personally prefer to use
  Scribble for this purpose. But I have also used LaTeX and
  even Google Docs. The latter two are each terrible in their own way,
  but the best tool is the one you use and that lets you get something
  done reasonably quickly. There’s always time to revise. Don’t suffer
  the paralysis of tool indecision and let that become the reason you
  don’t write!

• You won’t make much money this way. It’s okay, you’re probably
  already paid pretty well. And the money you’re not getting is money
  the author of the crappy textbook is also not getting! And you’ll get
  paid in mindshare, which is infinitely more valuable.

• You can also do what I did: I published the free PDF on my Web
  site. On the print-on-demand site, offer a modestly-priced PDF. I let
  readers know where they can get the free version. Therefore, the only
  reason to buy the PDF is the equivalent of a tip-har. Most don’t, but
  a few do. (You can also use one of the tip-jar services, though they
  didn’t provide enough value for me.)

Sure, my revenues have been modest. I view the checks as a little
surprise bonus. Added over time it could probably have bought me a new
bike frame (my unit of measure!), but mostly it’s mostly just
a nice dinner and ice cream. But I’m not doing it for the money. I’m
doing it to spread a worldview and to liberate a field from terrible
books. Both are much more worthwhile to me.

In short: if you’re even slightly tempted to write that textbook—go
for it. I got you, fam.