Since I launched this web-based book, the top question I’ve gotten from readers is “How can I get an e-book or PDF version?” Frequently this is attached to a comment like “Now that’s something I would pay for.”

As an author, of course I’m happy that readers want to pay for books at all. Speaking of which—this is an apt moment to thank the readers who have taken the advice in How to pay for this book seriously, and who have supported this work with payments and purchases. You are successfully sustaining this project.

As for the rest of you: I do understand why you’d want an offline version of a web book. Because you don’t always have a network connection. Because it’s convenient. Because it works better with your Kindle. Really, I get it.

But I’m not going to do it. And here’s why.

I was wondering whether I should spell the word ebook or e-book. I settled on the latter—unlike email, common usage suggests that e-book is not quite ready to lose its hyphen. Beyond hyphenation, it’s interesting to notice another key difference between e-books and email.

Today’s email is based on an open standard that dates back over 30 years (namely RFC 822). Like any network, it was apparent back then that the value of an internet-based email network would depend on the number of people connected to it (a proposition known as Metcalfe’s Law). Conversely, if internet users ended up with thousands of smaller, incompatible email systems, email would be less valuable to all. So an open standard made it possible for all these systems to exchange messages.

Today, we take it for granted that we should be able to send & receive email using whatever hardware & software combination we like. For instance, if I told you I’d invented a new email system that required you to buy a certain hardware device to access your inbox, you’d probably say that’s a terrible idea.

Yet that’s exactly the situation we have with e-books. Though open standards for e-books exist, those formats aren’t the basis of our e-book economy. Today’s reading devices—like the Kindle, Nook, and iPad—ostensibly let you use e-books from any source. But in practice you’re strongly encouraged to have a monogamous book-buying relationship with the company behind the device.

You could imagine a world of e-books more like that of email—or, for that matter, digital music—where you can get your content from any vendor, in an open format, and then load it onto any device. But that’s not how it works. These companies tout convenience as a big benefit, but it’s been defined narrowly: you can read with any device you want, as long as it’s ours.

Furthermore, if you still think that companies like Amazon and Apple have benign motives, it’s time to wake up. For instance, Apple violated federal antitrust law by improperly influencing the price of e-books. Amazon is currently applying coercive pressure to Hachette, a major publisher, for refusing to capitulate to Amazon’s pricing demands.

It’s important for us to criticize technology companies. And it’s not hard, because they’re full of shit. Primarily in the way they drape themselves in lofty principles that are regularly contradicted by their actions. Still, this behavior isn’t surprising, because they’re economic entities. They’ll always do what’s in their financial interest. They do not have moral agency.

But as readers and writers—as human beings—we do. Thus, we have to consider our own role in these transactions. As a writer, can I participate in these marketplaces while overlooking the bad behavior? I don’t think so. Whatever benefit I’d be getting is based in part on their market manipulation. Moreover, if I accept the benefits of these marketplaces now, I’m not going to have any basis to complain later. If Amazon or Apple turns out to be a monster, I will have helped build that monster.

“But wait—you could still release your book in EPUB [or some other open e-book format].” If you can show me an e-book format that gives me the same control over typography and layout that I can get in a web browser, I’ll consider it. As far as I know, it doesn’t exist. (If I’m wrong, please contact me.)

Thus my option here is to use EPUB to make a typographically ugly book about typography. Do I want to make that? No. Do you really want to read it? I doubt it. A key point in why does typography matter is that typography influences reader attention. If you like this book, in part it’s because the typography is helping you enjoy it. If I remove the typography, I can already foresee the results.

The PDF market doesn’t suffer from the same anticompetitive problems as the e-book market. PDF started as a proprietary Adobe file format, but it eventually became an open standard. Today, writers can make PDFs with many tools; readers can view PDFs in many ways. That’s all good.

What’s bad are the limitations of the format itself. PDF is fundamentally a digital simulation of paper. So it’s great for making paper documents available in the digital realm. But for natively digital documents—like this one—it removes functionality and imposes design constraints.

Thus, as a format for digital books, I have to vote against PDF. As a typographer, that’s painful, because PDF preserves layout and typography better than the typical e-book formats. In all other respects, however, it’s an example of the Shirky Principle—a backward-looking format that wants to impose yesterday’s constraints on today’s projects. As a reader and writer, that’s not good enough.

I can imagine a parallel universe where we have an open e-book format that’s widely supported, that has all the typographic and layout precision of PDF, and that’s easy to charge money for.

In this universe, however, it doesn’t exist. So we must compromise. Though the web has a number of problems—lack of a payment model for content is at the top of the list—I consider it our best opportunity for digital books. It’s based on open standards, with no vendor lock-ins. It has finally—especially because of the arrival of webfonts—developed a useful level of design sophistication.

The web also offers some unique benefits. Unlike PDF or other static e-book formats, a web-based book can be part of the searchable, hyperlinkable flow of web information at large. It’s easier to update, so it maintains a better connection to the author. It can have software-based features provided by the server or the web browser. It is, in short, a living thing.

And yes, the payment model is lacking. But if we want a fix, we don’t have to wait for some standard-making authority to draft a technical recommendation that browser makers may or may not implement in the coming years.

Rather, we can fix it right now, ourselves. We can do this by starting to practice a habit of readers voluntarily paying authors. If we do that, we can create a rich future for digital books. If we don’t—if we settle for the severely limited view propounded by companies like Apple and Amazon—we’ll have no one to blame but ourselves.

Reader, I’ve made my choice. I’ve made this book available under a voluntary-payment scheme (see How to pay for this book). More recently, I’ve made its underlying web-publishing system, called Pollen, available to other authors for free.

The rest is up to you.

—Matthew Butterick
27 May 2014

My original objection to EPUB, an open e-book format, was that it didn’t support the typographic luxuries that I, a finicky typographer, couldn’t live without. A reader recently suggested I take a look at EPUB 3, a revision of the EPUB format which amounts to an HTML/CSS website packaged into a standalone file, with all the same support for styling (including webfonts) that one would find in a web browser.

Looking back over this piece, however, I made a mistaken assumption. I called EPUB an “open e-book format”. This is true in the sense that no single company controls it. But my mistaken assumption was that EPUB was then (or eventually would be) supported by Amazon Kindle devices. That is not true: Kindles have never supported EPUB. There’s no reason to think they ever will. So EPUB fails to deliver the intended benefit of an open format, which is to let authors like me publish to one format and have the format readable everywhere, as one can with email or digital music.

Why won’t Kindles ever support EPUB? Because Amazon has always dominated the e-reader industry, and has its own e-book format called KF8. So there’s no incentive for it to support EPUB. Furthermore, as an author, there would be no point releasing an e-book in a format that isn’t readable on most e-readers (namely, Kindles). So we’re back to the original problem.

Still, in the nearly seven years since I wrote this piece, it feels like the underlying considerations are going obsolete.

First, various litigation and free-market pressures have meant that e-books have increased in price, and printed books have decreased, to the point where the e-book is usually only $2-4 cheaper. I wonder if a lot of the perceived popularity of e-books in the years 2010–14 was mostly enthusiasm for artificially cheap pricing—$9.99 in most cases—that Amazon was imposing on the market. As evidence, consider that based on the sales trends in the early 2010s, e-books were predicted to eventually displace printed books. But that never happened. E-book sales flattened, while printed books have continued to sell well.

Second, since 2014, computer screens themselves have gotten a lot better for reading (see screen-reading considerations), thereby reducing the comparative advantage of hardware e-readers. My only evidence here is anecdotal: nobody has asked me for an e-book in years.

To be fair, whatever the virtues of browser-based reading, there’s still no easy way of using a website without an active network connection. This is especially perplexing since the main job of a web browser is to copy everything from a remote web server onto your machine. It seems like offline access should be an easy upgrade. At various points there have been efforts to make these “offline web applications” possible, but the idea never took off and now seems to be left for dead. I imagine it has something to do with the fact that so much of the web is supported by user tracking, and offline access defeats that entire industry. Or maybe it’s just that ever-increasing network access means that the offline slice of human existence gets ever smaller, and the need for offline reading likewise shrinks.

Roughly, my impression is that e-readers have carved a nice niche for themselves for serving mass-market books that don’t have specialized design or presentation needs, which includes many or even most titles. I don’t see any evidence that layout-intensive books (like this one) have succeeded on e-reader platforms. Whatever the possibilites that EPUB 3 or KF8 offer in terms of text styling & layout, it’s a lot of new work to gain access to a small set of customers, many of whom would be willing to get a printed book. In other words, it’s a waste of time.

To that end, I remain interested in finding a path to convert this online book into a paperback, probably by some print-on-demand path. I don’t have the solution yet, though I have made progress. But I still have no interest in selling an e-book or PDF. This time, the market seems to have caught up with me.

As of last year, the Kindle supports EPUB. Well, sort of—Amazon is not supporting the format natively, but rather converting EPUB books to its proprietary KF8 format. Worse, apparently Amazon won’t convert EPUB books purchased from Google Books or Kobo, which defeats one of the key virtues of an open format. Thus, we still haven’t made much progress since the bad old days of 2014—as I wrote then, “in practice you’re strongly encouraged to have a monogamous book-buying relationship with the company behind the device.” I still don’t buy e-books.

—Matthew Butterick