As a credentialed conflict avoider, I've been loathe to stick my head into this fray. Indeed, this battle about meeting the needs of academia appears to be waged at times with an ideological fervor to rival that of the recent US election. It seems to me that the fervency with which people approach this issue has made it difficult in some cases for the arguments to follow a path towards resolution. It is perhaps also complicated by the wide assortment of changes being proposed to remedy the perceived problems. Some arguments for change suggest that PG should direct its energies towards making its library suitable for scholars by including more information in the files, particularly pagination and provenance, presumably packaged with XML. I have no problem with including such information. However, I don't think it should be required of all texts, nor do I believe that it really solves the scholarship issue. Including page scans _would_, to the degree that a solution is possible, and requires approximately 0% extra work for most of our valiant volunteers. And, PG has made it clear that this is acceptable, and has already done so for some projects. I feel that Marcello gave the most persuasive and concise summary of the situation, and I didn't notice any overt disagreement. Marcello Perathoner wrote:
The best value for Academia (and the least work for us) would be just to include the page scans. Any transcription you make will fall short of the requirements of some scholar. I think we should use our time for producing more books for a general audience instead than producing Academia-certified editions of them.
HSH's comments justify such an approach. Her Serene Highness wrote:
I need to know EXACTLY when the original was published, who published it, and where, since there are variant texts out there. Even a single word change that might have occurred in the copying process could change the meaning of a vital sentence.
Of course, there is a simple, if unsatisfactory, answer to all these questions for PG texts: they were published by PG, on the PG website, and each file states when it was published. Each work we publish is the "PG variant" of that text. As an academic, I find it dishonest and unhelpful for a scholar to cite a physical volume when the volume they consulted is an electronic edition. It is virtually impossible to guarantee that "even a single word change" was not introduced in the transcription process. Even with DP's careful processes, I would not wager that most of our books enter PG completely error free (or correction free, for that matter.) Page scans allow for an additional layer of safety for any scholar concerned about the adherence to a given print edition, though a certain level of trust in the provider is still required. Thus, while I hope that PG's holdings are as accurate as possible, it would also be my hope that scholars using PG would cite PG. Evidently this is not always the case. Michael Hart wrote:
I've also heard that many of those who complain, actually use our eBooks in secret, and ONLY want the provenance so they can steal them without giving credit where credit is due.
This suggests to me two things. 1) We can include page scans and information about provenance, _when available_, with the files so that academics can feel confident in the reliability of those PG holdings. Not so that the original sources can be dishonestly cited, but to provide the necessary data for certain scholars to confidently cite PG's edition. We can point to this in our documentation to enhance our scholarly credibility. 2) We can prominently suggest an appropriate style of citation of works in PG's holdings. (I've seen this done with other digital collections.) Perhaps if the citation style also takes into account the original source, some otherwise reluctant scholars would be appeased. Is this something we can all agree on? -- David Newman www.davidnewman.info