October 1, 2024

the chaos archive of old mailing lists

Thinking

I didn't have much training in methodology as part of my PhD, so I have always been sort of butterfly-meme about methodology generally, if you know what I mean. My dissertation (and current book project) focuses on the rise of civilian computer cryptography, which starts in the late 70s but goes through the late 90s, so it encompasses a wide variety of sources, from very formal to very chaotic. I've done a good amount of research at the excellent Charles Babbage Institute archives, using institutional records and some personal collections they have, and a ton of research using the historical newspaper search on ProQuest. I also use a good number of government sources, some of which I access through the national archives, but some of which I randomly just found online?¹ So even my more traditional archival research has sometimes been a bit wacky.

But at least half of my research, starting in the 90s, is based on old listserv records and other weird born-online digital ephemera. Lots of Usenet and mailing lists, lots of other informal writing that has been floating around the web for thirty years, and some documents that other people found or FOIA-requested and then hosted on other websites.

From a logistical standpoint, this shit is incredibly hard to deal with. I've had real trouble figuring out how to keep track of it all and how to cite everything. Nobody's ever given me a good answer on how to cite a born-digital document that just sort of floats around, with no definitive origin or canonical location.² Citation managers like Zotero don't have a good solution either, especially for things that are technically websites but really aren't, like forum posts. The mailing list archives in particular can disappear at any moment if the host decides they don't want to host them (e.g., Google).

From a more epistemological standpoint, I've done a good amount of thinking about what role this type of material can play in my arguments, since it's sort of like trying to make cultural claims based on what's going on in a weird corner of reddit. But since my work focuses on the idea that people's political ideas about technology inspire them to use and advocate for technology in particular ways, it is important to go looking for how people talk about it.

Take as an example the 1990s debates about the Clinton administration's "Clipper Chip"--essentially, a voluntary encryption standard for phones that could allow the government to decrypt and wiretap encrypted phone lines. Many different groups of people advocated both for and against the proposal in the press, in Congressional hearings, in formal correspondence and published documents, and in person. One of those places was on the Usenet cryptography newslists.

While old Usenet posts and other emails are part of a necessarily incomplete collection, are as unreliable as any other social corner of cyberspace, and represent the communications of a small community, their contemporary importance is worth the uncertainty. As cryptographer Matt Blaze put it in a contemporaneous summary of a Clipper chip meeting with NSA and FBI representatives: the feds "seemed to base a large part of their understanding of public opinion on Usenet postings. Postings to RISKS [an academic mailing list], sci.crypt and talk.politics.crypto seem to actually have an influence on our government."³ Usenet was a niche platform in the 1980s and early 1990s, but the topic was also still a pretty niche topic, and those involved in the debate went where the experts and most influential voices were—and so should historians today. In 2011, Blaze recalled that the NSA representatives were "enthusiastic" about Blaze and Bellovin circulating their notes to the newsgroups.⁴

With that said, because the Clipper debate was a much broader political movement than just those who communicated on early 1990s mailing lists, those sources should not be taken in isolation. For me, since I am concerned with the question of how did ideas about cryptography change over time, and why and how did those ideas spread within and change American society, I needed to combine those corners of online history with other sources to get a more complete picture of how ideas moved and changed and shifted. In this case, that meant looking at sources like institutional records of lobbying organizations, government records, and press coverage.

The other problem with born-digital sources is that there's also just so much of it, and it can be so incredibly wild. Both the scale and the absurdity has been hard to convey in my writing.

Thanks to Gareth Millward for inspiring this edition of the newsletter with a discussion about how hard modern digital sources can be. Are we missing obvious points? Do you have a good resource to recommend? Can you teach me the mysteries of methodology? Let me know!

Reading

I read an excellent longform ProPublica article about the difficulties of shifting clinical practice in medicine last week. It's a wide-ranging piece, discussing how entrenched recommendations can become, even when there is quite strong evidence that a specific treatment is useless or even actively harmful, from the reversal on peanut allergies in babies to hormone replacement therapy for menopausal women. The end of the article--which discusses the differences between public-health and individual outcomes--is particularly interesting for those of us with an interest in numerical epistemology but zero intuition for statistics.

Doing

It's Rosh Hashanah this week, which means I will be baking copious amounts of challah and cake and trying to fit five pounds of beef into my small dutch oven. Shana tova!