“…errors cause planes to crash, buildings to collapse, and knowledge to regress. The beneficial properties need to reside in the type of exposure, that is, the payoff function and not in the ‘luck’ part: there needs to be a significant asymmetry between the gains (as they need to be large) and the errors (small or harmless), and it is from such asymmetry that luck and trial and error can produce results.”
Recently I was clicking around an audio file I had loaded into a granular sampler software. I was changing the Grain Location within the audio, positioning the cursor at different points of a graphic that reminded of the up-down mountain topography of a Tour de France stage, listening for something good. Nothing was sounding good, so I kept moving the cursor along the mountainous waveform, now adjusting the Grain Size knob as well, trying to get something going. But there was no magic–no sound that inspired me to hear beyond itself, to hear through the sound towards a feeling evoked.
The reason I was clicking around an audio file in a granular sampler in the first place was that I had made the decision to use the software to make a series of pieces. (A series is always better than a single piece because you can leverage what worked in one piece to work in multiple pieces, and failing that, it takes the guesswork out of What to do next? for a while.) My plan was vague: load your audio files, see what you get, and build on that. But so far I wasn’t getting much to build on. Was this idea simply a dead end, an error?
In his writings about his concept of antifragility, Nassim Taleb notes that our errors are valuable because they reveal information. There’s also an asymmetrical relationship between small errors (i.e. those with little downside) and their potentially big payoffs. Making errors is a good practice to practice because errors have “beneficial properties” built into them in that they tell us what isn’t working, or what we don’t know. In short, it’s through making errors that we learn. In the context of an uncertainty-strewed craft such as music production, when we make errors we strengthen the systems that guide our work.
Despite there being no magic at the moment, I pushed on with tinkering with my audio. I opened the Modulation page in the software, which shows a representation of the sample’s ADSR (Attack Decay Sustain Release) Envelope. I played with the sustain length, making it shorter which caused the sample to skip and stutter, then made it super long which created a drone.
Wait, what’s that sound?
The drone kept going…and going. I stopped, took my fingers off the keyboard, and listened. It sounded gossamer yet full of little pulsations. It was also super quiet, so I boosted it 10 dB. I clicked on the Effects page, noticed that there was reverb on the sound, turned it off, then turned it on again to see what exactly the difference was. (Not much.) Through tinkering, what seemed to be an error a few minutes ago was becoming an opportunity.
The sustain on the drone was so long that I could play a chord in the lower register with it and, while the chord resonated, play higher notes. I went back and forth from one end of the keyboard to the other for a while until I remembered to record a take. Then I tried separating the low and high register parts into different takes. This way the first (lower) part wouldn’t be listening to the second (higher) one, and the length of each part could be different. Recording the lower and higher parts as separate takes of differing lengths added unpredictability into the process. I liked the resulting sound because it wasn’t something I could have played all at once I couldn’t predict how it would go.
One of the most common experiences I have when working on music is encountering my own errors. (“Cataloguing of negative results” Taleb advises in the essay’s “Heuristics to Maximize Antifragility” section.) Among the many types of errors are those of process, judgement, performance, and patience. Process errors include doing things that I know from experience rarely lead me to interesting results, such as using four-bar loops. Errors of judgment include not listening closely enough to what it is in a sound that isn’t quite right. Errors of performance happen when I don’t take enough risks while recording. All of these errors are connected to my errors of patience. I find that when an error tells me what isn’t working, I can use this information to slow myself down and alter the course of my practice. The example of my clicking around an audio clip in search of something good illustrates this. My workflow felt like an error until trying this and that began generating something with promise.