I decided to base my digital bibliography exercise on Edith Wharton’s The House of Mirth. Of the three databases I chose for my exercise (Google Books, HATHITrust, and Project Gutenberg), I’m most familiar with Google Books, so I decided to go there first. I entered in my search terms and got two actual results (i.e. Wharton’s text, and not texts about Wharton’s text). The first one listed was the full text of Wharton’s The House of Mirth (with illustrations by A. B. Wenzell), published by Charles Scribner’s Sons in 1905. Google offered two versions of this edition of Mirth for download, EPUB and PDF. The second search result was a 2007 Digireads.com version that cost $2.99 to download. While the Google Books PDF was free and a fast download, I was pretty annoyed to discover that I couldn’t search the text – I tried on my work computer (which uses Windows) with Adobe Reader and my MacBook with both Preview and Adobe Reader.

Although the online Google version was searchable, since there were no ways to highlight or annotate, it didn’t seem very useful beyond yanking quotes out of the depths of the novel for use in other projects (which is actually how I tend to use Google Books). Indeed, Google even seems somewhat prepared for this – their primary source of textual manipulation (when viewing the book on my Mac – this feature disappeared on my work computer) is the ability to ‘clip’ a line into plain text format, a link to an image of the selected text, or a link to embed the text. While it might be neat to generate a digital image of the text, it actually limits the user to ‘clipping’ in rectangular forms only, meaning you can’t carry over onto the next line unless you want additional words from surrounding sentences caught in the rectangular clipping field. I’m not sure what the point of this clipping is – I really don’t think I’ve ever seen someone use it (or so rarely that I can’t recall). Google also allows you to generate a link for the specific page of text that you are currently reading, almost as a digital bookmark for later citations. There didn’t seem to be any ways to report errors for Google beyond writing a review for the text, but that leaves me questioning: what is a book review supposed to review? The actual content of the novel penned by Wharton? Or the scanning quality of the book? I’ve seen this happen on Amazon for Kindle versions a few times – people give a book low reviews based on the amount of grammatical and/or digital formatting errors, which confuses/frustrates those who are interested in the quality of the story.

Next up was HATHITrust, which I’ve encountered briefly before. I got a little lost the last time I was searching around for quick text downloads (actually, for Woodchipper, a data-mining tool we used in Technoromanticism), which turned me off to the site initially. However, when I searched for Wharton’s text on HATHI, I got four full-text hits for four different editions of Mirth: C. Schribner’s Sons (1905), C. Scribner’s Sons (1922), C. Scribner’s Sons (1933), and First Scribner/Macmillan Hudson River Edition (1989). When I clicked on the 1905 edition, I discovered that it was the same digital text that I encountered on Google Books (except for a badly digitized front cover scan). It even had the same pink thumbtip of a careless scanner in the bottom corner of a page! However, HATHITrust includes a watermark next to the “Digitized by Google” that reads “Original from UNIVERSITY OF CALIFORNIA.” I re-checked the Google Books version, and there is no such notation made for the edition’s provenance, which is odd, since it appears to be the same exact book and scans. HATHI attributed all of the universities that held the physical copies of Mirth contained in their database (two from UC, one from University of Virginia, and one from University of Michigan). It also revealed that all four digital texts were “Digitized by Google.” So… why weren’t they all available on Google Books?

Also, since the one version I was most interested in obtaining in PDF form (the 1905 one) was also offered on Google Books, I found it a bit silly that I had to log-in via UMD partnership in order to download it. It was a long process of “Building” the PDF, then downloading it, all to obtain pretty much the same text as Google. I was able to search the HATHI PDF on my work computer using Adobe Reader in a hit-or-miss fashion (I was sent to the correct page with a box appearing roughly around the portion of text that contained my search term), but I was unable to search it at home using my MacBook with either Adobe Reader or Preview. In HATHI’s site version I thought it was interesting that I could toggle between views (Classic and Plain Text), which might have made searching easier (otherwise the site just directs you to the right page with no highlights or line indicators), but the very first time I tried toggling over to Plain Text, I caught a number of typos on the page I happened to have open, the most glaring being the running head, which read: THE HOUSE OF M1KTH. HATHI does have a Feedback link at the bottom of the page that allows for error reporting, though I’m not sure I would have the will to submit a new one for each Plain Text page.

Like Clifford, I found Project Gutenberg to offer the most variety in file formats, and like her, found the image-lacking disclaimer pointless, as the HTML and plain text versions did not contain images either. Project Gutenberg offered HTML, EPUB (no images), Kindle (no images), Plucker, QiOO Mobile, Plain Text UTF-8, and MP3 files of The House of Mirth; for my purposes I converted the HTML version to a PDF file, one which (finally!) is fully searchable. Unlike either Google Books or HATHI, there seems to be no printed referent for Project Gutenberg’s text. The only noted provenance is a release date of the digital text (June 1, 1995) and a few notes at the end of the text:

Notes:
1. I have modernized this text by modernizing the contractions: do n’t becomes don’t, etc.
2. I have retained the British spelling of words like favour and colour.
3. I found and corrected one instance of the name “Gertie,” which I changed to “Gerty” to be consistent with rest of the book.
-Linda Ruoff

There is also a notice at the end of the text that “Updated editions will replace the previous one–the old editions will be renamed.” It almost seems as if Project Gutenberg is leaving little to no room for discussion on authoritative editions, variants, and the like (though you are free to email them with errors you may discover). There also appears to be no interest in preserving a digital transmission history of their edition of House of Mirth, as any discrepancies will be obliterated with no discernible trace (unless you leave a note, as Linda Ruoff did).

All in all, in order to accomplish the two things I want most in a digital text (searchability – a digital affordance, and writeability – a print affordance), I had to save a PDF file from an HTML version of The House of Mirth – one that had no perceivable basis in print. Project Gutenberg’s version is pure text, no book, which leaves me wondering: how would I cite these quotes that I am able to find at a moment’s notice? Would I have to turn around and utilize Google Books’ scans to pin specific quotes to page numbers? Makes one wonder, are Post-It Flags really so terrible?