My book of choice for any bibliographic project will usually be Moby-Dick. Katie and Susie can both attest to this after having to sit through a semester of me geeking out over the textual history of the novel. Of course, by posting later than some of the others, I can only echo what they have said: Project Gutenberg provides the most formats for a given text, including an audio option, which neither HATHITrust nor Google Books gives you (as they only allowed for pdf downloads, and with HATHITrust permission was required, and Google payment), and it was the certainly the easiest to download, because it came with virtually no strings attached. But while I have traditionally always turned to it first for my canonical etext needs, I found it the least transparent of the three versions of Moby-Dick I collected.

For those unfamiliar with Melville scholarship in general one name pretty much reigns as the foremost editor of Melville’s novels, especially Moby-Dick: Hershel Parker. He has edited since the 60s three ‘authoritative’ versions of MD that have formed the foundation of most of Melville scholarship and editing practices since. As someone heavily invested in Melville, Parker’s imprint is typical in any edition I come across, and the lack of it is suspicious. It is not a bad thing, of course, but it raises questions. Project Gutenberg does not note an editor or recognize their copy-text in either of the two full-text editions of MD, but instead does include the note:

Produced by Daniel Lazarus, Jonesey, and David Widger

I do not recognize any of the names personally, and these people are not specifically named as editors, so it is difficult to determine what sort of mark they may have left on the text, and without providing information about the copy-text, the text’s specific origins are unknowable to an outsider. Of course, Project Gutenberg provides a (somewhat reasonable) defense for this:

Creating the works from public domain print editions means that no one owns a United States copyright in these works, so the Foundation (and you!) can copy and distribute it in the United States without permission and without paying copyright royalties.

This is what made Project Gutenberg’s text of MD so easy to acquire, versus HATHITrust and Google, who expressed copyright claims to their digital versions and locked the downloads behind certain obstacles, and while I can appreciate the reverence paid to access, the unclear provenance of the text, other than its recognition as a “public domain text” does not point me to the copy-text being reliable. This perhaps is fine for a general reader, but unsettling for a scholar.

On the other hand, HATHITrust and Google Books both provide some more concrete information because the book is viewed through images of a scanned hard copy. What is unfortunate is that the two public domain editions available on each platform were also very dated. HATHITrust’s edition of MD is from a 1929 Macmillan edition (which is about the time Melville was rediscovered but well before academics began critically editing his work) and Google Books full text edition is from the 1851- the year the book was published. Google’s edition wins, for me at least, because the 1851 edition at least is more reputable than whatever edition served as the copy-text of Project Gutenberg’s edition, and it stands to reason may have served as the copy-text for HATHITrust’s version. Easily accessing the first edition of the book leaves little questions to scholars as to what they are working with, and can actually be very useful not only as a text itself, but as an artifact of the novel’s original form (before critical editing).

Of course, I can’t spend all my time musing on editions and validity. The formatting of the texts is also interesting for one major reason: in the Gutenberg edition, since it does not mimic the page scrolling format Google Books and HATHITrust adhere to, we find awkward moments in the text where the body of the text is interrupted by Melville’s footnotes (which he typically wrote in to clarify any esoteric nautical information). In the page scans from the other two databases, this does not occur, because they reproduce the pages and so the text remains in a more traditional form (with footnotes at the bottom, clearly demarcated as outside of the body).

In response to the Duguid article, where one of the primary critiques of Google Books is the poor scanning of pages and distorted words, Google’s edition of MD looks to be pretty polished. In my sampling of the scanned pages, I did not find cut edges, distortions at the spine, or anything of that sort. That problem, however, was prevalent in the HATHITrust version, where the illustrations of the cover page were cut off near the spine, and some marginalia went over the edge of page (someone made a note on the Table of Contents that spanned the margin between Chapters XIII and XVIII that I think might have said ‘BORING!’ , but I cannot be sure).

Finally, in terms of feedback, HATHITrust made the process the easiest by providing, on the same page as the book was read on, a little button that opened a survey asking about the quality of the book, where any errors could be reported including missing, distorted, curved, and blurry text. Google unfortunately, only allowed users to review the book, which could be more concerned with plot and enjoyment, instead of textual quality. Project Gutenberg did not provide any easily accessed method of evaluation, but does include links on the home page to get in contact with them, and to submit missing pages for texts (which I suppose counts as one form of correction).

I was surprised, especially after reading Duguid, of what I found in Google Books. Their images of the Moby-Dick text looked more professional and refined than the HATHITrust edition, was an 1851 first edition, and posed no issues in the formatting of the text. The same could not be said of the HATHITrust and Project Gutenberg versions, whose scans were less sophisticated, contained marginalia (incomplete and cutoff at that) or posed formatting issues by presenting a text with footnotes incorporated into the body without separating them in any way. As I said, the Duguid article made me fearful of what I would find on Google, and their issues with Tristram Shandy are of course valid concerns, but perhaps it’s possible Google has learned or has improved their process since that article was published in 2007, since while Google Books’ major downside was the lack of a reporting feature, of the three editions I have looked at, it was surprisingly the one that needed it the least.