Musings on Digital History

Ethics, Bias, and Diversity Issues in DH

Author and Professor Safiya Umoja Noble writes in her shocking book Algorithms of Oppression: How Search Engines Reinforce Racism that while we may think that search engines are equal opportunity providers, search engines actually have led to biased algorithms which discriminate against people of color (especially women of color), and reinforce the privilege of white users. Bias is seen in search engine results, photo searches (in quite offensive ways), and in how things are prioritized when searching. I think most people don’t consider the level of human involvement when it comes to the internet, but the biases of developers can certainly be seen in these ways. To be fair, it seems as though sometimes the sin is one of omission: people not considering that their choices impact people of color in these ways (this is true outside of the internet as well). At the end of her study, Noble states that

“I am particularly mindful of the push for digital technology adoption by Black / African Americans, divorced from the context of how digital technologies are implicated in global racial power relations.”

Safiya Umoja Noble, Algorithms of Oppression: How Search Engines Reinforce Racism (NYU Press, 2018) 171.

In the age of COVID, this statement is haunting. Access to internet has been granted to thousands of families for virtual schooling. Considering this push, and how it could hurt black children, who are already marginalized on the internet, oof. It is a bleak thought.

Bias an ethics extend to the archive-sphere. As we have discussed previously, it is important to acknowledge what is not in archives, and what gaps exist due to racism and collecting priorities. But once you access archival material, there are a host of ethical considerations to be aware of. Just because you can easily publish interesting bits from archival research, doesn’t necessarily mean you should. Historian Michelle Moravec asks:

How are we thinking about our responsibilities as the users of these digitized archival material, when what we write is online, and when our reuse of digitized materials may at the least violate copyright and the worst cause harm to individuals?

Michelle Moravec, “What Would You Do? Historians’ Ethics and Digitized Archives,” Medium, April 14, 2016.

I have been thinking about this quote all week, and how it pertains to my Instrumental Women project. Part of my DH project will be an interactive database of women who make musical instruments from around the world. I have compiled a list of over 200 instrument makers, and I am in touch with another organization which has a list of 100 makers working in South America, alone. For months now, even though the database is ready to go live, I have hesitated on whether I should post the database. On one hand, all of the information I have compiled has been from readily accessible online sources, thus if I can find it, can’t anyone? The reason I have been cautious is that women in this industry have experienced terrible trolling, sexism, and even threats. By creating this public database in order to celebrate and help support women instrument makers, I do not want to create a easy access platform for trolls and those who wish to do harm. Because I believe in consent, I have now decided that I will be asking for written permission of these makers in order to make sure everyone is not only aware of the project, but has the right to opt out. The added bonus will be that I will be able to be in contact with all the women in the database, which will no doubt help me as I embark on my dissertation, wherein I plan to conduct oral history interviews with many of these women.

PS– I encourage you all to check out the work of ethnomusicologist Kyra Gaunt.  I attended a presentation Dr. Gaunt gave last year on just this thing.  She is writing a book entitled “PLAYED: Twerking at the Intersection of Music, YouTube, and Violence Against Black Girls,.”  In the following article, she states that”My book is about the intersection of twerking, the way the music is being monetized through streaming and the way artists will exploit girls’ games in order to gain attention in the attention economy. There’s this halo-effect for artists and girls because they know when Cardi B and Megan Thee Stallion release a song, there’s a whole slew of algorithms that are going to pull their videos in. That’s the monetization of music. YouTube is profiting off of the content of very young girls. And before last year when the fine came from the FTC, they were practicing plausible deniability.”

4 replies on “Ethics, Bias, and Diversity Issues in DH”

Hey Jayme – wonderful blog post. I had a lot of similar thoughts related to Noble’s work. Your connection between COVID and further tech and digital issues for POC is something I haven’t seen anyone else discuss yet this week.
I had no idea your thoughts about Instrumental Women! I knew you were working on the project and we had discussed it, but I assumed it just had not come quite to fruition yet. Your query is an important one, and one I often do not think about as a early American historian. I respect your decision to communicate with these women and am excited to see developments.

I have definitely reflected on how our online world this year has disproportionately affected families of color, but I didn’t draw the connection between that and this weeks readings. Especially in high school, where students are asked to do some basic research, this could have far-reaching effects. Normally such issues could be rectified in person fairly easily, where you can tweak a search or explain certain concepts as issues arise, but online it would be much more difficult. A student of color may receive different search results based on their personal data, or perhaps they would search a term different, and it would be unpredictable to know what they would get as an output. I imagine the reliance on algorithms as objective tools has to make online teaching more complicated.

Sidenote: my high school librarian also worked at the library at Carnegie Mellon, so she ensured our district had access to the same databases that the universities do and taught us how to use them. So I learned EBSCO and JSTOR in high school, and I am learning most people did not. I can’t imagine trying to navigate this years teaching with just Google searches. Although as we learned this week, database research can also be skewed by algorithms, so who’s to say?

Jayme, your database sounds like it will be such an important reference, and yet as you rightly note, one that could be exploited to nefarious ends — but there are some really excellent databases out there that do the same, and have identified some methods to dissuade misuse and empower the persons represented in the database. And your opt-out/consent-driven approach sounds like an excellent way to do that!

One recent similar example comes to mind: the Queer Cartoonists and Cartoonists of Color databases have a submission form that allows creators-only to submit and update their own entries, and the database admins have called on comics librarians and other information professionals to help with the management/cleanup *after* data is received, so there can still be some best-practices in place alongside ethical info-gathering. Maybe this could be an option if/when your database grows (which it totally could and should, I’ll bet musicologists and music librarians will use the heck out of this!)

…and by “databases that do the same,” I mean databases that have considered what it means to aggregate and open up this information — not databases that harass creators! 😬 (oh, for an edit button!)

Leave a Reply to Robert Carlock Cancel reply

Your email address will not be published. Required fields are marked *