Some semi-random thoughts, as headlines continue to focus on “post-factual” news:

There is a lot in common between what we’re seeing now in Britain and the US, and  researchable “news” in fragile and conflict affected states. The recent Brexit referendum and US Presidential campaigns are just the two, latest and possibly most prominent cases that come to mind wherein the tone and content of open media forces a radical rethink of how researchers can use it.

Fact-free politicians (and their allegedly fact-free constituencies) are just part of an emerging landscape. Analysts are barely able to make sense of the political framing of reality in their attempts to decipher what parties and politicians actually stand for, what they actually think and believe, or how much damage or disruption they actually intend to cause in future. It won’t be any easier for future historians of US and British politics who will have to contend with the morass of today’s media output as they attempt to reconstruct our present reality.

Guerrilla archiving efforts –  undertaken by climate scientists who fear political interence in the US in the form of mass deletions of decades of scientific research –  suggest that the stakes are substantially higher and that the problems associated with it are likely to become much more acute. Wilfully created gaps in current understanding are consequential, with more or less immediate resource, budgetary and policy  implications. By comparison, gaps in the future historical record are a problem for historians, and deciphering and validating the details of past events and developments that can be found in archived yellow press doesn’t seem like such a big deal.

There are some basic reasons why researchers should pay close attention to this. They revolve around the fundamental importance of context, precision and analytical judgement.  The lessons here are a bit more subtle, but they’re worth noting, especially among knowledge workers who rely heavily on “open source intelligence” and among the legions of people whose understanding of research process doesn’t get much more sophisticated than Google or Wikipedia.

It’s a challenge I’ve dealt with in the research and consulting projects I’ve done on guerrilla vernacular communications in West Africa, Afghanistan and elsewhere. I had the pleasure and privilege of working for a short time with the late and eminent Africanist, Stephen Ellis, before he left us in 2015. He expressed it best, in his books and in his work. In West Africa, oral culture prevailed and information was more often than not broadcast through cultural filters and images. Stephen always held that sense could still be made of documentary evidence (like local newsprint). These were sources that on the surface of it only transmitted utterly sensational and fantastic details. To use them effectively required a particular sensitivity to meaning – to the truth of a matter – that could only really develop when well supported by deep knowledge and additional, supplementary forms of inquiry.

Media outlets come in all shapes and sizes.  Some are loud and boisterous, while others are more stoic. “Newspapers of record” are a recognized form of the latter.  Some try to report what happened, while others try to convince readers why and how they happened. Media output, in other words, can serve more than one purpose, and only one of them is to provide researchers and analysts with a source of evidence needed to  determine the factual basis of past events: what happened, when it happened, who was involved, what they said about what happened and so on.  Reconstructing past events is a tricky business, and some media environments are so highly politicized – the rhetoric so overheated and contentious – that verifiable facts are almost impossible to discern from the collection of color and misdirection in which they’re embedded. Propaganda in wartime and in crisis environments is especially difficult to parse for verifiable details – here, the “facts” have less to do with manifest content  –  explicitly used words and text – than with the latent meaning that can be teased out of the wider contextual landscape and from more technical meta-data.