All posts by Jill Walker Rettberg

#dayofDH2014 3: Technological bias and cultural filters

I kept writing and reading instead of doing the more DH-specific data visualization I was intending to do. But it’s so interesting! I’m writing about filters, you know, Instagram style filters, but I’m extending the notion of filter by seeing technological and cultural filters as basically the same thing. I talk a bit about that in the presentation I did a couple of weeks ago, and am working on writing that out properly so it makes sense.

But one example of how cultural and technological filters mix is the skin-tone bias in photography, especially 20th century photography, although today’s digital cameras aren’t perfect either. Early photographic colour film was designed to show white people’s skin in detail, but did a terrible job of representing a person with darker skin – especially when people with very different skin tones were in the same image. There were complaints from parents – for instance at poorly lit children with darker skin in diverse school groups – from the 1950s on, but it wasn’t until the 70s that Kodak actually did anything about the problem, and then only because companies wanting to sell dark woods and chocolates wanted film that would show the detail better.

An older "Shirley card" for skin tone and colour calibration of cameras.
An older “Shirley card” for skin tone and colour calibration of cameras.
Recent, international calibration image.
Recent, international calibration image.

Lorna Roth, who wrote an interesting article about this history (“Looking at Shirley, the Ultimate Norm: Colour Balance, Image Technologies, and Cognitive Equity“, 2009), notes that an important reason there were no real campaigns for film producers to create film that did a better job of representing darker skin tones was our general assumption that technology and science are objective – we thought photographs just were that way. But “[h]ad NASA, the U.S. intelligence service, or meteorological scientists already completed their research on photography of low-light areas at the time of the popular development of still photography, the evolution of film chemistry might have unfolded quite differently,” and it could certainly have changed once those developments were in place.

Sylveeta McFadden’s article in Buzzfeed, “Teaching the Camera to See My Skin“, is a very interesting personal take on this. And the comments on the Jezebel article about the same topic are wonderful explanations of how disgust at how badly conventional photography represents darker skin is a pretty common motivation for taking selfies:

Her piece is beautiful and I struggle with this thought ALL the time. Growing up all of my girlfriends (and immediate female relatives) were white. I would watch them effortlessly take a photo or get their photo taken and in return get an image that looked just like them. I never really felt that way. I still don’t – unless I take my own photo. And people call it vanity but really I just want to be able to see myself in a picture. I don’t see myself in other people’s photos, I just don’t.

And another:

Ive alwaaays felt this way too. Some people laugh at me for wanting to take selfies rather than have someone take the photo but I’ve always felt kind of shitty pre smartphone era when the photos would come back developed and I just woudnt look like me. I think there is a features proportions issue as well as skin colour issue. – Ive felt like its not just me but other black people I see in photos too .

McFadden writes the same thing at Buzzfeed:

I couldn’t help but feel that what that photographer saw was so wildly different from how I saw myself. Is that how you see me? Could you not see blackness? Its varying tones and textures? And do you see all of us that way? (..) I started taking pictures to self protect. I just couldn’t bear seeing anymore shitty pictures of me. I didn’t want know what I wanted these images to say, but I knew I could make something beautiful.

I’m going to be using this case in talking with students about how technological determinism, cultural co-construction and how technology encodes cultural biases, that’s for sure.


First item of the day: picking up my J1 scholar visa form which needed a travel validation signature so I can leave the US to go to Canada on Thursday to give a talk with Scott on our visualizations of the ELMCIP Electronic Literature Knowledge Base. I suppose I don’t need the signature to leave the country, but I certainly will for them to let me back in.

Getting signatures on forms isn’t digital humanities, just one of the not very exciting admin tasks that all scholars have to attend it.

The day before Day of DH 2014

I’m currently on sabbatical at the University of Illinois at Chicago, where I’m writing a book about digital self-representations with the working title Seeing Ourselves Through Technology: How We Use Selfies, Blogs and Wearable Devices to See and Shape Ourselves. 

For the Day of DH tomorrow I’m planning to focus on the most digital methods and tools-ish things I need to do this week.

First I have to put in at least 20 minutes of writing on my book project. Which is a humanities approach to the digital and to digital tools, but doesn’t really involve digital tools.

Then I’m going to finalize my sample and finish adding data on dissertations on electronic literature to the ELMCIP Knowledge Base of Electronic Literature so that I can update my network visualizations in Gephi and make the final changes to my paper analyzing genres of electronic literature based on this network analysis. I presented the paper at the ELO conference in Paris last year, but need to add a few more dissertations and make quite sure that the sample makes sense – all known dissertations between two specific dates would be good, right now there are a few outliers I think. The paper is to be published in the May issue of EBR: the Electronic Book Review, so it needs to be finished soon. If you’re interested in this kind of visualization, I wrote a tutorial last year  which includes a ready-made Gephi file with 32 dissertations about electronic literature and the creative works they reference.

Here’s the genre map from the earlier version of the paper, which I hope to update beautifully tomorrow:

If I can fit it in, I’ll also get to work preparing a talk I’m doing with my husband and collaborator Scott Rettberg at Wilfred Laurier University near Toronto about the ELMCIP project, how we’ve built and used the knowledge base, about the visualizations we’ve done, and about how we’ve designed practical research courses for undergrads on digital humanities. We were invited by Jeremy Hunsinger, who teaches in  Communications Studies at WLU and is the co-editor of several really useful anthologies, and are excited to get to meet up with him and his colleagues.