Sonifying Derrida’s references

Experimentation and play is valuable and important. It’s good for your brain, and you never know where it will lead, or what fruitful new path you might discover. It’s also incredibly valuable to share our experiments, even if they are incomplete. You never know what connection someone else will make, what unexpected results might arise. With that in mind, here’s a little experiment of mine. It’s not new, and I sort of thought the time had passed to share it, but I’ve been encouraged to post this.

A couple of years ago, I took some R&D time to work through a tutorial on data sonification from The Programming Historian.¹ I’ve been interested in accessibility and different modes of sharing and analyzing data for a while², so I wanted first-hand experience with sonification. I decided to work with data from the Derrida’s Margins project: in particular, the references in Derrida’s Of Grammatology³.

I took the reference data, ordered by page number, and used different pitches to represent the different types of references on a page: high notes are epigraphs, low notes are footnotes; the one you hear repeated the most is quotation. The volume of a note indicates the number of references of that type on a single page, so the louder the notes are, the more references. Silences are pages with no references.

I wanted to experiment with this a little more, so I added a new sound to indicate where the chapters begin - as a way of providing a little more orientation to the listener, like labels on a chart for a data visualization. I also changed the instruments with the goal of finding something that worked with the high amount of repetition in this track and that would sound good together.⁴

Want to play with my experiment to extend it further? Here’s the code, adapted from the script in the Programming Historian tutorial. To generate the MIDI file yourself, download the References data from figshare as CSV and get the python script. Or if you want to try different instruments, download the MIDI file and load it into GarageBand or similar.