member reviews the submission, pastes it into the metadata record in CONTENTdm, and indexes the collection, at which point the transcription is live and available for searching. This workflow has obvious drawbacks. Such a mediated system is more costly in staff time than is ideal and the asynchronous nature of the submission process means that multiple users could work on the same page simultaneously, resulting in duplication of effort. Nevertheless, the pilot moved into the testing phase, where another issue emerged: some staff members voiced concerns about the quality of submissions, questioning whether the public was qualified to do the work. While these concerns do have some validity, the overall consensus was that some imperfect access was better than none, and any staff-side inefficiencies would be worth the trade-off in public outreach benefits. The Civil War Diaries Transcription Project site was launched in early spring of 2011. Standard promotional tools of press releases and blog posts drew a little attention, but the crowds needed to drive the project proved to be elusive, so the next two months were spent focusing on promotion to historians, Civil War enthusiasts, and genealogists. In early June the project was featured on the American Historical Association blog 8 from there, it was picked up by reddit.com, a social media news site.9 The response to the reddit post was enormous the day it went up, web statistics jumped about 7,000 percent (from typically 1,000 users to about 70,000), and the Digital Library server crashed, remaining out of commission for the next several days until the traffic became more manageable. Since then, response has calmed down quite a bit, but the project still retains a core stable of loyal transcriptionists. One of these core transcriptionists, Dave Hesketh, is a 69-year-old retiree in the north of England who has currently transcribed 140 pages and is still going strong. Regarding the family whose papers he has spent the most time working with, he says that they “have become almost an extended part of my own:” …[T]hese diaries & letters reveal the lives of “ordinary” people…, rather than those of politicians, generals, and the like (whose actions, words, and deeds are generally sanitised for public dissemination). These people come alive you come to share their hopes, their fears, their everyday concerns—the price of food for themselves, and for their animals, the cost of RLI 277 12 Experimenting with Strategies for Crowdsourcing Manuscript Transcription ( C O N T I N U E D ) DECEMBER 2011 RESEARCH LIBRARY ISSUES: A QUARTERLY REPORT FROM ARL, CNI, AND SPARC