Researching with a second brain
Or, why the research community should be writing with tools like Roam Research
For the past few months, I’ve been using Roam Research to produce my forthcoming book—Screens, Research and Hypertext—as well as all the other bits of supporting materials that go along with a new book.
It’s admittedly a bit of an unconventional choice, as Roam bills itself as a note-taking application and not a publishing tool. It’s part of the leading wave of so-called personal knowledge management tools.
The idea behind PKM is to develop a “second brain,” one that both aids in remembering important ideas and helps surface connections between those ideas. That second one is arguably the more important task. Indeed, if all you’re after is remembering things, then there’s quite a lot of evidence that you’ll be better served writing notes by hand.)
But for all their virtues as a memory aid, handwritten notes aren’t especially good at surfacing new connections.
Roam’s secret sauce is its out-of-the-box use of bidirectional links. At this point, we’re all familiar with the HTML hyperlink. Those are single directional. The text linking out “knows” that it is linking to something. But the target text has no idea what things link in to it.
Bidirectional links do just what they say on the tin — they create links that run from the source text to the target text and also from the target text back to the source.
Roam then surfaces those inbound links at the end of each note.
(If you’re old like me, you might remember the days when comment threads on blogs incorporated trackbacks. A bidirectional link is a sort of automated trackback.)
The beauty of bidirectional links is that they surface connections you’d forgotten. For example, if you were to review an old note, you’d see both the things you linked out to when you wrote the note and also the later things that you linked back to that note.
Roam is also smart enough to surface unlinked references.
Imagine that you wrote a note called “bidirectional links.” A few months later, you’ve forgotten that original note, but you write a new note that includes the phrase “bidirectional links.” Roam will automatically add a new section called unlinked references in both your current note and your “bidirectional links” note.
Later still, were you to reopen your newer note, you’d see that link back to the original “bidirectional links” note. In a very real sense, Roam has “remembered” connections that you didn’t even realize you’d made.
Hence the “second brain” moniker.
There are obviously a number of substantial advantages to a tool like Roam for researchers. Research is all about finding the links between ideas. Conventional writing tools limit those links to the ones you can hold in your own head.
To be sure, Roam is not a publishing tool, for all that I’ve used it as one. My book is deliberately nonlinear. Readers are meant to follow threads of ideas across different essays. So Roam matched the ethos of my project in a way that traditional publishing tools—which are mostly geared toward producing linear documents—did not. If you’re writing a more traditional publication, other tools may better serve your needs.
That said, linear documents are a poor choice for communicating research.
As Joel Chan points out, in the research world, document-centric tools like EndNote and general purpose software like Word and Excel still dominate the research process. These are certainly tried-and-true—they’re what I learned to use as a PhD student in the late 1990s.
They were also badly suited to knowledge synthesis way back then, and they haven’t improved with age on that front. Tools like Roam (along with competitors like Obsidian and Notion) may still be a bit rough, but they are already a step-change over what most researchers are doing now.
