An American Editor

March 2, 2010

eBooks and the Never-Ending Rewrite

One of the blessings of ebooks is that they are digital files that are easily corrected (note I said easily, not inexpensively), unlike the printed book, which once published becomes a fiscal nightmare if it is error laden. This problem, and what to do about it, came to mind as the result of a recent New York Times article, “Doubts Raised on Book’s Tale of Atom Bomb.”

The Last Train from Hiroshima by Charles Pellegrino was published in January 2010 by Henry Holt to acclaim. Alas, there may be a major problem: The technical details of the mission are based on in-person recollections of someone who was not there. So the question becomes: What is to be done? [Update: According to today’s New York Times, the publisher, Henry Holt, has decided to recall all 18,000 copies of the book. Apparently other issues have arisen, including whether the author truly has a doctorate degree and whether other sources actually exist. Here the publisher is acting as a gatekeeper and warranting the quality of the book; what would be the case if the book had been self-published?]

If this were an ebook the choice would not diminish in either importance or problems. To correct the ebook would lead to versioning and a never-ending attempt to always keep a book accurate and up-to-date — the never-ending rewrite. In one sense, this is good; in another, it is a scholarly nightmare: How will a scholar ever be able to cite or quote an ebook as a source? (Which is another interesting question: Can ebooks be reliably cited?) But failure to correct a major error, one that calls into question the validity and credibility of the book and author, as occurred in The Last Train from Hiroshima, is equally problematic. And what happens when three years from now another history-changing error is found?

Clearly this is not much of a problem in fiction. It doesn’t really matter that a street runs east-west rather than north-south in the real world, nor does it matter all that much that the author uses compliment when complement is meant. But these kinds of errors, as trivial as they are in fiction are volcanoes in nonfiction, especially in the scholarly disciplines. The fiction author will be remembered for having written an intriguing story that held you spellbound or bored you to death; the scholar will be remembered for a work being flawed or flawless in its facts, not in its storytelling.

So what does one do with a book like The Last Train from Hiroshima? The print version is an easy solution: Henry Holt can recall and refund or replace the printed book, destroying the incorrect print version, or it can just do nothing. Perhaps it can issue an errata sheet that buyers can paste into the book acknowledging the error.

But if the book is available as an ebook, the ebook is its own quandary. It is easy to replace the digital file and to even let purchasers redownload the incorrect copy. But at what point does Henry Holt and the author stop making changes? Or should we expect the book to be continuously correct and updated until such time as it is so perfect that no changes can be made? Or should we leave it as is and wait for a “second” edition to be released; that is, should the ebook be considered a mirror image of the released print version or a book in its own right?

Is this the real crux of the matter? To date, no consensus has formed on exactly what an ebook  is. When the only form of a book is the ebook form, then the ebook stands on its own. And in that instance, perhaps one revision of the digital file is warranted to correct an egregious error. But when there is a pbook version as well, the status of the ebook is uncertain and the jury remains out on whether it needs to be a mirror image of the pbook or can stand on its own. (The argument that it does or does not stand on its own is not affected by the presence or absence of “added value enhancements.” The question is whether the core text stands on its own or needs be a mirror image of the pbook.)

Publishing and history are lives riddled with errors. Books with errors have been published before and will be published again. History is knowledge of what occurred in the past and that knowledge is always undergoing change — new insights and facts are regularly being discovered. Consequently, I think there is a limit to the independent life of the ebook. I think scholars and readers of nonfiction need be able to point to a particular book and say that is it.

My solution would be to treat the ebook as the mirror image of the pbook and whatever steps are taken to correct the pbook be taken to correct the ebook. If no steps are taken to correct the pbook (a mistake, I think, that would irreparably harm both the publisher and the author), then no step should be taken to correct the ebook, except, perhaps, to add a 1 paragraph author’s note before the first paragraph of the chapter that is in question indicating that the author is aware that the following material is incorrect.

At minimum, for The Last Train from Hiroshima, the publisher should prepare an errata sheet, one that could be used for all versions of the book. I think this is necessary because the discovery of the error is virtually contemporaneous with the release of the book; had the discovery occurred 12 months from now, my thoughts would be different.

Regardless of how Henry Holt deals with its problem, I am of the opinion that, in the case of nonfiction, the core text of an ebook should mirror the corresponding text of its pbook version — an ebook does not have an independent existence. To ensure scholarly endeavors in the future can point to a specific book and cite it, there must be finality and mirroring will provide it.

Blog at

%d bloggers like this: