Crowd Resourcing 

The Library of Virginia is looking for a few good transcribers.

click to enlarge Kathy Jordan and Sonya J. Coleman are using open source code and lots of volunteer help to transcribe manuscripts from the Library of Virginia’s treasure troves.

Scott Elmquist

Kathy Jordan and Sonya J. Coleman are using open source code and lots of volunteer help to transcribe manuscripts from the Library of Virginia’s treasure troves.

Nearly two centuries ago, Isaac Reed died of strangulation in a Henrico County cellar. He was handcuffed and tied in such a way that three men could hoist him up and down or leave him suspended, his toes barely touching the ground.

His supposed offense? Reed, a free black man, had been accused of stealing $100, a crime to which he never confessed. One witness testified that he saw Reed before his death, and that the man had begged him to pay the money, pledging five years' service in return.

The details of the Henrico coroner's report on the death, which happened July 17, 1827, provide a chilling glimpse of the realities of the racial system that structured American society for so long. Because of the Library of Virginia's new crowd-sourced Transcribe program, Reed's story is preserved for posterity and easily available to the public.

"This is an opportunity to attract users that might not have otherwise come to us," says Kathy Jordan, a digital initiatives and Web services manager at the library.

Transcribe is built on open-source code developed by the University of Iowa and George Mason University's Center for History and New Media, and supported partly through funding from the Institute of Museum and Library Services and the Dominion Foundation. It opens up selected Library of Virginia collections of scanned manuscripts for members of the public to try their hands at transcription. The goal is to make the digitized documents more searchable and readable.

The tool neatly kills two birds with one stone: It chips away at the massive digitization job confronting the library — which possesses some 119 million archival items — and increases public engagement with the institution. "It's amazing, the number of people who enjoy doing this kind of work," Jordan says.

With most manuscripts between three and 10 pages long, and none exceeding about 20, each project offers a manageable chunk to transcribe. Volunteers can even work on a single page of a manuscript, leaving the rest to someone else, although most get drawn into the stories and are reluctant to step away before finding out their conclusions.

Part of the appeal of the manuscripts lies in the collections the library has selected for inclusion. Although the project may expand to include new collections, it's now focused on five: broadsides, African-American narratives, the Civil War 150 Legacy Project, manuscripts related to Gabriel's conspiracy, and letters written by Patrick Henry during his three years as governor of Virginia.

Digital collections specialist Sonya Coleman notes that the Civil War 150 and African-American narrative manuscripts in particular have attracted much attention because of how personal the stories are. "I think it's very important for the public to become invested in that personal story," she says.

To dip into the documents on display at the Transcribe site is to fall down a rabbit hole of historical questions, whether about an individual who wrote a particular diary, a former slave who gained his or her freedom papers, or the mechanics of early-20th-century communication as described in the charmingly forthright "Directions for Using Phone" published around 1900 in Blackstone. Each document brings what could have been a dry historical footnote into sharp relief.

Although the library initially faced concerns that crowd sourcing transcriptions would produce poor or inaccurate work, Coleman says the finished products "have generally been of very good quality."

"We've found that the crowd is really very self-policing," Jordan notes, perhaps not a surprise when considering the high level of accuracy that other crowd-sourced resources such as Wikipedia have achieved.

Nevertheless, the library isn't leaving the job entirely up to amateurs. Every transcription must be reviewed by at least one staffer at the library before it can be approved, adding a key level of quality control that will satisfy the rigorous standards of scholars interested in using the collections for their work.

So far, Transcribe has been a success, gaining momentum since its launch Aug. 26. As of late October, almost 2,000 pages had been transcribed by more than 1,500 people, and Jordan reports that people from as far away as Argentina and China are diving into the manuscripts.

Down the road, the staff hopes to expand the project into public transcribe-a-thons, or perhaps work with college classes to transcribe lengthy documents such as diaries.

"In many ways," Jordan says, "we don't know what's next." S


Latest in Arts and Culture

Comments (5)

Showing 1-5 of 5

Add a comment

Subscribe to this thread:
Showing 1-5 of 5

Add a comment

More by Sarah Vogelsong

Connect with Style Weekly

Most Popular Stories

Copyright © 2021 Style Weekly
Richmond's alternative for news, arts, culture and opinion
All rights reserved
Powered by Foundation