Tom Swift and His Electronic Book Machine

Sometime in the next few weeks the Ethnologue 18th edition print volumes should become available through your preferred bookseller.  I won't go into the details of what those volumes contain because I think the process by which they are produced is fascinating and demonstrates the ongoing development of the Ethnologue as a source for useful information.

If you read the history of the Ethnologue you can get a sense of the growth of the Ethnologue from those original mimeographed pages to the large database that sits behind both the website and the varied products (country reports, regional volumes, maps) that we make available. In the early days, the work of turning data into formatted books was done by hand using typewriters, then word processors to produce photoready copy. And that photoready copy pretty much constituted not only the product but the data behind the product. As computer technology developed, the data began to be stored electronically and in a more structured way. As that happened, it became more and more possible to use computers more extensively to do the typesetting. Even so producing a new edition every four years took almost a full year of manual processing involving many people not only in the technical processes but also in the more mundane and tedious work of proofreading and copy editing. Graphic artists, cartographers, language and country experts, typographers, bibliographers, and a host of others all took part in the processinng and preparation of the "book"--a hefty volume that went through various changes in size and format over the years.

With the 17th edition, we made the transition to where we considered the web edition to be the primary focus of our work and began to produce a wider range of print products that represent various views of the data. We also shifted to an annual release cycle, something we could not even contemplate if we didn't have processes in place that automate much of the work. While there is still a lot of manual work to be done by artists, cartographers, bibliographers and editors, the bulk of the formatting and typesetting work is now nearly automatic--automagical, really.

Our database is the starting point. The data we collect (and which many of you contribute) is stored in a large and complex relational database. The fact that it is relational means that we have to do a lot of thinking about how the different bits of data about the languages of the world relate to each other in order to structure the database in the most useful way. If we can identify and account for the multiple possible relationships in how we set up the database to begin with, the possibility of manipulating and analyzing that data in the most useful ways is greatly enhanced.  It also makes it possible for us to produce not only different web views of the data but the larger (and increasing) number of print products that are now available.

For each product, we develop a template that guides how the different fields from the database are to be transformed, combined, and formatted. The website, of course, has a variety of templates that determine the layout of the country and language pages. The country reports have a template that is specific to that presentation of the data.  The 3 regional volumes (coming soon!) have yet another template that takes into account the additional summary tables, for example, that can be produced for each Ethnologue area (Africa, Americas, Asia, Europe, Pacific) and the specifics of page size and layout which differs in those volumes from the format of the country reports.

The Ethnologue Computer Wizards are able to pour the data out of the database into the templates and the result is the basic content of each product. The process doesn't stop there, however.  Back in the day, we would have passed that content to a typographer who would painstakingly turn it into formatted pages--page by page--using computer software to be sure - but still manually (and artistically) designing each page.  I've learned a lot about book design and publishing in the years I've been involved with the Ethnologue and the challenges of creating a visually beautiful two-column, justified, properly-hyphenated set of contrasting odd and even pages with corresponding headers and footers require more skill and patience than most of us would ever imagine. Nowadays, we have Really Smart typesetting software that can do much of that fairly quickly. 

These automated processes don't eliminate all of the manual labor however. We still have to proofread the content, but mostly we do that proofreading in the database now rather than in each resulting product. We still find lots of errors that we need to fix even after we've done as careful a review as we can, but even the proofreading process is being facilitated by computer tools that make it easier for us to spot errors and maintain consistency.

The last step in the process is, perhaps, the most Gee Whiz part of the whole thing.  

Not all that long ago, after we had done the typesetting, we produced a set of photoready pages--printed out on paper--which we had to box up and ship off to a book manufacturer.  That's a company that not only runs printing presses but which also has all the machinery and skills necessary to assemble and bind the pages into the book that eventually gets shipped to the distributors. There aren't many companies around that have the capacity to produce a 1500 page volume with a hard cover binding, so the process of negotiating with the manufacturer about the technical details, getting samples, fixing problems and then actually producing the print run and having it shipped back to us so we could make it available for sale, added months to the production process.

But nowadays we have what is called print-on-demand. No need for photoready paper copies. No need to ship them off to the printer. Smaller products, like the country reports and maps, we distribute electronically as PDF or JPG files. When you add one of those to your shopping cart on the website, you can instantly download the product. For the larger volumes, we produce an electronic version which gets sent off to a print-on-demand book manufacturer. They store the digital product in their computers and whenever a distributor decides to stock and sell the Ethnologue, the book manufacturer is notified. Orders from the distributors are passed on electronically to the printer.  At that point, their computer whirrs and clanks and feeds the book into what I imagine as a gigantic Electronic Book Machine (but which is actually smaller than the gigantic printing presses that were used in the past). That machine cranks out a book, binds it, packs it up, and ships it. This can all be done one book at a time or in large numbers but, always and only on demand, just when they are needed. I always picture some goggled minion dropping a big coin into a slot and pulling a lever and then watching the wheels turn and the motors hum until a book falls out into the bin at the bottom. Kind of like buying a cup of coffee or a candy bar from a vending machine. 

Without these technical marvels, being able to produce a new edition of the Ethnologue each year would be challenging indeed, and being able to produce a set of 3 regional volumes in print a mere 5 months after each web page update would be absolutely impossible. As we gain experience and as we implement new technologies and systems, the time between new editions and book releases will undoubtedly get shorter. We keep looking for more and better ways to make the Ethnologue data available and while it is sometimes challenging to keep up with the technology which keeps getting younger while we keep getting older, we're excited about the possibilities.