A huge part of being a historian is digging through the evidence, teasing out connections in order to build an argument or narrative. For some areas of resarch this is easier than others; one thing I like about working in the twentieth century is that there are a lot of preserved records, and not nearly as much temporal distance between the lived experience of today and people of that time. In comparison, scholars who work on much earlier time periods have a lot less to go off of–which is not to say that they can’t draw really sharp and eye-opening conclusions from limited documents (see, for instance, the deservedly famous cultural history work of Robert Darnton, Carlo Ginzburg, or Natalie Zemon Davis). But an abundance of documentation does not necessarily aid the historian. In the modern era, integration of technology increases personal engagement with data, yet it also significantly complicates making sense out of the chaos.
Once something is on the web, it’s there forever, so they say. There’s truth to that: one remarkable aspect of the Internet is its robustness and resistance to censorship. Paradoxically, the more a person, or group, or government seeks to control and eliminate information, the more it is preserved and spread (sometimes referred to as the Streisand Effect). Yet the Internet is also fundamentally transient–and not merely in the sense that a website here today might be gone tomorrow. Instead, transience is baked into the very experience, and compounded by the scale of data available. I’ve lost track of how many times I’ve tried in vain to look up a blog post or web page that I can hazily remember enjoying previously.
But in the case of a medium like Twitter, that difficulty might be even more irresolvable. The simplicity and limited space of a tweet leads it to be wonderfully direct, but in so many cases the context is absolutely crucial in understanding meaning: who/what is this tweet responding to, and how are others reacting in turn? In fact, the individual experience of someone reading that tweet is wholly unique. The context in which it exists has been created through the curated set of follows that the user has chosen for their feed. Thus a message that might appear to be a non-sequitur to one viewer could be a crucial part of an ongoing debate for another, or a voice in a chorus of affirmation to a third.
In previous centuries, gatekeepers limited the amount and diversity of voices preserved in historical record, even if informal debate and discussion certainly existed in both living room and neighborhood bar. Notably, many historians of today have recently sought to recover previously unheard and unrecorded voices, particularly of marginalized groups. But the historians of the future will face a somewhat different task in dealing with a nearly overwhelming multiplicity of recorded perspectives. With so many contrasting and conflicting records, how do you get interpret what’s really going on? Do you privilege certain voices, trying to prioritize the ‘important’ contributors? How do you know your research is comprehensive? Historians have always struggled with these questions, but now that we do have access to so much that they take on new relevance.
As a brief example of the challenges involved, this week I worked a bit with Storify, trying to gauge its potential usefulness in curating a narrative out of individual tweets. But it wound up being far harder than I expected. I chose as a topic the discussion surrounding a recent New York Times editorial about the supposed decline of political history–a moderately controversial piece that provoked a good deal of reaction from other historians. But practical issues caused problems from the very beginning. I began thinking about curation about two weeks after the initial publication, and the volume of activity on Twitter made identifying all the relevant conversations difficult, particularly since there was no centralized hashtag easily gathering tweets together. Many other tweets were simply retweets of the initial article (and Storify’s search tools were even worse than Twitter’s). The result simply felt incomplete and fragmentary, despite my own perspective of the past few weeks making me certain that there had been interesting reactions and discussion existing within the medium.
The task wasn’t fruitless: I identified a distinct dichotomy in responses, with non-historians often passively accepting the article’s premise and remarking about the tragedy, while many historians forcefully rejected the assertions that political history was dead. But as a whole, trying to splice Twitter into a neat narrative that was both democratic and reflective of experienced contributing voices felt like a much harder task than anticipated. In doing so, however, I found one more-successful implementation of my goals. Tom Sugrue storified a set of his own tweets discussing the controversy, calling out some of the most relevant reactions and blogs about the subject. In part, his concentration on more thought-out longform responses rather than individual quickly-tweeted reactions led to a more comprehensive-feeling result.
And indeed, one of the most valuable aspects of Twitter is its capability to direct me to more weighty articles or thinkpieces that I might have otherwise missed. This all is not to say that I don’t think that individual tweets themselves are valuable… Only that Twitter is more innately transient than even the larger Internet as a whole, and thus should be understood and treated differently than other media. The Library of Congress’s desire to preserve Twitter is a revolutionary attempt to record the kind of daily interactions that previous documentary evidence leaves out (even if the project is suffering some serious setbacks). But it also may demand that future historians reconsider how they engage with their research, ensuring they have a clear sense of their goals and the limitations of the data.