Tinderbox as Visual Tool

I was initially attracted to Tinderbox in v.5 for the way it allowed notes and metadata to be treated as objects in a variety of graphical/visual views.

The software’s commitment to the visual was, on the one hand, accessible and appealing. There were maps, charts and something called treemaps. More significantly, notes were “deep” (they contained text, text links, and all the attributes and values I could imagine), but I could interact with them as flat objects arranged on a plane. All of this felt intuitive, accessible and revolutionary. I’d never used my computer this way, and it felt liberating enough that once, in a moment of euphoria, I compared it to swimming like a dolphin.

On the other hand, this commitment to the visual often proved to be intellectually challenging. When I was staring at an outline of notes, each with its text and long list of key attributes and values, I could ignore my confusion about what I was doing or working on because at least, it seemed, I was working. That confusion was harder to ignore in a map that made glaringly obvious just how much of a jumble all my thoughts were in. In this way, the visual continually brought me up short and forced me to take account of where my thinking stood.

But if Tinderbox offered me a graphic presentation of my mental jumble, the app also made it possible for me to create order in the chaos by arranging my thought–objects sensibly, meaningfully, and then capturing the insights that emerged visibly in badges, colors, shapes, links, adornments, and more. As strange as it sounds, this commitment to visual tools and graphical/spatial sense-making felt polemical, something I’d never imagined software could be. There was a point-of-view embodied in the Views, and I felt pushed by it to find clarity and empowered as I made the attempt.


So fast forward from my initial impressions of Tinderbox to last Summer and Fall. I was course planning in Tinderbox and made a series of videos about how I was going about it. I was also trying to make headway with my research for a writing project, but I was struggling to get anything done, even if it took me awhile to realize it. When I did, I assumed it was mostly because of the move to online teaching and the inevitable difficulties of finding your feet in a still new project. Also, there was COVID. Eventually though I began to suspect that the trouble arose because I was trying to do my work in a tool I’d lost sight of how to use. I was having fun making magical little machines that did all kinds of idiosyncratic things, but in many ways they amounted to a collection of distracting TBX toys.

How could this have happened?

Strange as it may sound, I think the problem arose in part from being too deeply immersed in the forums. The forums are incredibly friendly and supportive, and they’re a space on the internet that I value. But as I spent more and more time there, I slowly internalized a variety of implicit (imagined?) notions that together built up into a revised conception of how to use Tinderbox that was emphatically textual and profoundly non-visual. Here are a few of those implicit notions:

  • The outline is the base view and is rounded out by the attribute browser.
  • The map view is a limited and probably introductory view.
  • The other views are experiments or curiosities of limited use.
  • Use and manipulation of information generally happens (and in advanced work probably will happen) through action code or AppleScript.
  • Information should be organized and regularized in anticipation of this coded/scripted manipulation. (Also probably for export. Because, maybe.)
  • TBX is a tool for producing and interacting with a linear text file in a non-linear way. (This is a factual statement that carried outsized metaphorical weight for me.)

I’m not sure whether most forum participants would agree or disagree with these statements, and it doesn’t much matter one way or the other: Tinderbox is a varied and complex tool that invites each person to figure out how best to use it for what they want to do. What does matter is that these and other text-centered notions had begun to operate as the fixed coordinates I navigated by, and this wasn’t working for me at all.

The concreteness that came from being able to treat information and ideas as objects (without flattening them out to mere objects) had evaporated. Everything in my files had become abstract and indistinct, and I felt like I was spending all my time building or servicing vague structures rather than thinking about my courses or my research.

I was lost.


So I pulled back from the forums last Fall and set aside all the elaborately beautiful little machine–files I’d built up and was dazzled by. I wanted distance from what had become a dead-end and to find again the messy provocation and the polemic that had first attracted me to the app: the idea that arranging complex textual data graphically in space could both clarify thought and be made an integral component of thinking and writing.

To give myself a chance to see things with fresh eyes, I decided to tackle a different aspect of my writing project, one that involved the difficult problem of how to annotate film stills directly but in a way that would allow me to use the annotations as notes for a developing, multi-film commentary.

So far, this is going well. I love how the annotations look, and I’m rediscovering brilliantly useful visual tools I haven’t touched in a year or more. And when I want to get a look at how things are shaping up, I can dive (like a dolphin) into the notes’ texts for a bit before coming up for visual air. Importantly, I’ve avoided building machines. Instead, I create simple prototypes and simple agents only when some task has become a pain and I see that using them can make it easier. This feels a lot like painting with only red, blue, yellow and white, an enabling restriction. It also feels like a rediscovery of Tinderbox’s visual heart.