mcld.co.uk

We had an interesting conversation here yesterday about designing new musical instruments. We're interested in new instruments and interfaces, and there's quite a vogue for "user-centred design", "experience design" and the like. But Andrew McPherson pointed out this paper by Johan Redstrom with an interesting critique of this move, essentially describing it as "over-specifying" the user. If we focus too much on design for a particular modelled user experience, we run the risk of creating tools that are tailored for one use but aren't repurposable or don't lend themselves to whole "new" forms of musical expression.

The twentieth century alone is littered wth examples of how it's only by repurposing existing technologies that new music technology practices come about. Here's a quick list:

• The Hammond organ was meant to be used in churches as a cheap pipe-organ alternative, but it really took off when used in R&B, rock and so on.
• The mixing desk is widely used as intended, of course, but it unexpectedly became a musical instrument in the hands of dub reggae people like King Tubby and Lee Scratch Perry.
• The saxophone (I didn't know this) was apparently intended to have a consistent timbre over a wide pitch range - it wasn't intended for the throaty sounds we often recognise it for these days, and which earned it a firm position in jazz. (OOPS the sax was pre-20th century, my mistake - it doesn't strictly belong on this list.)
• The vinyl turntable famously wasn't designed to be scratched, and we all know what happened with that in hip-hop and beyond.
• The development of the electric guitar was clearly driven by the desire simply to make a normal guitar, but amplified. Hendrix and others of course took that as a starting point and went a long way from the acoustic sound.
• The TB-303 was supposed to be a synth that sounded like a bass guitar. Turn its knobs to high-gain and you get those tearing filter sounds that made acid house. (Indeed it was discontinued before it got really famous, showing just how unexpected that was...)
• The microphone led to a number of changes in vocal performance style (for example, it allowed vocalists to sing quietly to large audiences rather than belting). The most obvious repurposing is the sophisticated set of mic techniques that beatboxers use to recreate drum/bass/etc sounds.
• 1980s home computers had simple sound-chips only capable of single sounds. But pioneers like Rob Hubbard broke through these constraints by inventing tricks like the "wobbly-chord", and created a rich genre of 8-bit (and 16-bit) music whose influence keeps spreading.
• AutoTune was supposed to subtly make your voice sound more in-tune. But ever since the Cher effect, T-Pain et al, many vocalists push it to its limits for a deliberately noticeable effect.

The only successful twentieth-century musical instrument I can think of, that was successful through being used as the designer intended, is the Theremin! (Any others? Don't bother with recent things like the ReacTable or the Tenori-On, they're not widespread and might well be forgotten in a few years.)

So, given this rich history of unexpected repurposing (kinda reminiscent of the fact that you can't predict the impact of science) - if we are designing some new music interface/instrument, what can we do? Do we go back to designing intuitively and for ourselves, since all this user-centred stuff is likely to miss the point? Do we just try building and selling things, and seeing what takes off?

==

One important factor is hackability. There's quite a telling contrast (mentioned in the Redstrom paper) between the "consumer" record player and the "consumer" CD player - in the latter, the mechanisms are quite deliberately hidden away and all you have is a few buttons. The nature and size of vinyl makes that a bit difficult, so most record players have the mechanism exposed, and it's this exposed mechanism that got repurposed by scratch DJs.

(There are people doing weird things with CD players, and hacked CD players are relevant to the glitch aesthetic in digital music. But maybe if the mechanism was more exposed, more people would have come up with more and crazier things to do with them? Who can say.)

But it's not neccessarily a good thing to expose all the mechanism. In digital technology this could end up leading to too-many-sliders and just poor usability.

(Another relevant paper on this topic: Thor Magnusson's "Affordances and constraints" paper, considering how users approach music technologies and their constraints.)

In a paper I wrote with Alex McLean (extended version coming soon, as a book chapter), we argue that the rich composability of grammatical interfaces (such as programming languages) is one way to enable this kind of unbounded hackability without killing usability. Programming languages might not seem like the best example of an approachable musical environment that musicians can fiddle around with, but the basic principle is there, and recent work is making engaging interfaces out of things that we might secretly call programming (e.g. Scratch or the ReacTable).

==

Another factor which is perhaps more subtle is ownership - people need to take ownership of a technology before they invest creative effort in taking it to new places. There was some interesting discussion around this but I personally haven't quite pinned this idea down, though it's obvious that it's important.

For inventors of instruments/interfaces this is quite a tricky factor. Often new interfaces are associated with their inventor, and the inventor generally likes this... Also it's rare that the instrument gets turned into a form (e.g. a simple commercial product) that people can easily take home, live with, take to gigs, etc etc, all without reference to the original inventor or the process of refining original designs etc.

I don't even think I've really pinpointed the ownership issue in this little description... but I think there is something to it.