Any Port in a Storm

tdm-pilot.org, the Constellate Beta, and almost 10,000 articles from the Public Historian.

What do all of these have in common? They were all plan B, and they all seem good enough.

Originally, the plan was to use JSTOR's Data For Research tool. It would enable me to download wholesale collections of articles and journals for analysis using R and topic modelling methods. Unfortunately, the best laid plans fell through. The DFR tool simply could not easily fulfill my needs, so we had to look elsewhere.

Luckily, a new pilot project had been launched, that would use Jupityr Notebooks and the digital toolbench to analyze data and produce research. It could download the data, do some basic analysis of its own, and easily export it to a Jupityr Notebook for topic modelling analysis. If you can't design your own code, store bought is fine.

As is often the case in digital work, I failed upwards, and one of the core lessons I've been slow to internalize in my time as a digital humanist rose quickly to meet me. Where will your research live? What is its lifespan?

It's not something that needs to be tackled in normal historical work. Journals are published and archived, books will always exist, even as they are occasionally launched digitally as well. But digital work, especially when designed with digital tools, has a ticking clock. It might exist for a long while, as the creators of the code that you may be using for research have made it freely available online for anyone to use, and the research is published on websites, the code in plain sight. Some research, like mine is now, is being designed using digital tools made online. The digital toolbench pilot project creates and stores my dataset by stripping words from articles and books from a variety of databases who have an agreement with this service. The word salad is then diligently saved to my computer by me, but from there, the Constellate Beta has handily coded a way to export the data straight to a Jupityr Notebook for my analysis. As long as Jupityr Notebooks are around, I can continue to poke and prod at my data. But while my notebooks will comprise the core of this research, it's imperative that as I complete I record my findings in other ways. Screenshots of the results, careful observations and formal notes, even a paper analyzing my findings.

As I discovered when I had to switch the service to complete my work, digital tools have a timeline that is often invisible to me, and it needs to be taken into account when planning. Luckily, when Plan A did not work out, Plan B was ready and waiting for me to seize onto it, and hopefully it will have legs and enable me to take my research where ever it needs to go. Any port in a storm is good enough.