Exploring with images

I decided it would be useful to gather together a big collection of images of exemplary projects and installations which have been swimming inside my head. I realized that having all these pictures stuck inside the pixels of the computer was holding me back from drawing relationships between them in a more natural way. Specifically, I realized I wanted to push them around on a table and explore how all these projects relate to each other along different dimensions.

Axes

There’s a tradition in design of drawing two orthogonal axes and arranging items on a Cartesian plane. But what two axes am I interested in rating all these installations along? One axis running from “static” to “interactive,” and the other axis running from “solo” to “cooperative?” Or one axis running from “path-directed” to “open-ended,” and the other axis from “discouraging collaboration” to “encouraging collaboration?”

Dimensions

Then I thought, well, it looks like perhaps I want to rank all of these projects along maybe five or six axes all at the same time—in other words, I want to describe them in five or six dimensional space. It’s hard for us Earth creatures to interact with six dimensional objects, so we tend to use tricks like t-SNE to reduce dimensionality down to the one or two or three we’re more familiar with.

Flirted with the idea for a minute, but I realized I don’t think that’s an appropriate tool to use for my set of 20 or 30 data points. Instead, I’m just drawing some two-dimensional comparisons I think may bear fruit and arranging the pictures as feels appropriate.

Here’s an arrangement I explored for starters: one axis runs from “one user” to “cooperative” while the other axis runs from “static” to “interactive.”

picture of approximately 25 photographs arranged on a table

No eureka moment yet, though I think that pursuing this sort of exploration may start to yield some larger-scale findings about the sorts of exhibits/installations that already exist, and perhaps suggest where there may be holes out there. Or maybe I’ll decide that the field is well-covered and that there aren’t any noteworthy holes for me to address. In either case I’ll keep exploring and post further findings.

ps: The cards I’m using are a smattering of projects of interest in my research. Here is a PDF I made, printed, and cut up into pieces so you can see exactly what range of projects I’m playing cards with.