Crowdsourcing DH

It should be obvious that as technology improves for Digital Humanities, it also finds a way to make DH more difficult. Some of these ways include Wikipedia, crowd-sourced information, fabricated sources and deepfakes.

Wikipedia is sometimes very useful. I find it especially useful to jog my memory about things I’ve already learned but have started to forget. There are some problems with Wikipedia, however, namely that it is crowd-sourced. Crowd-sourcing is when the public is invited to contribute to the project. I also specifically use the word “contribute” rather than collaborate, because in many cases there is very little collaboration going on. The contributors often do not know each other, they often don’t directly add to what another contributor wrote, sometimes they specifically contradict one another or remove sections written by the other contributors. Collaboration implies they are knowingly working together to achieve the same results.

There is also the question of whether the people producing the crowd-source product are even qualified to write on it. In the case of the Digital Humanities Wikipedia page, the original creator was a DH professional at Stanford, which is promising, but it’s nearly impossible to vet every contributor for every crowd-sourced project. Wikipedia is getting better at this by requiring citations and installing bots to determine whether a contribution is in good faith or “potential graffiti.”

Then we get into fabricated sources. Everyone has seen a TV episode, movie, or book that features a “this artifact was a fake all along” trope. It’s pretty common in media, but it’s a real fear among history professionals, and while there are real scientific methods for determining the authenticity of artifacts using carbon dating, chemical analysis, handwriting analysis, and more, it’s much harder for digital sources to do the same thing. Enter deep-fakes. Deep-fakes are digitally altered media (typically video or audio recordings, but can include digital pictures or other sources) that is meant to trick the audience into believing something that did not actually happen. For instance having a video of a president giving a speech, but the president never actually said the words in that speech, they’re all computer generated and nearly impossible to tell from a true recording of their voice. This causes problems because despite not tricking most professionals, if it goes public, many members of the public may find it realistic enough to believe.

There’s a fairly common saying in history about how to fix historical conundrums: “With more history.” Right now, that seems like the only way to go forward, as professionals we need to continue to be vigilant about the provenance and authenticity of sources, and make sure that when we contribute to public history forums, we do it in good faith, with accuracy, and sources for our information.