The Database Avant-Garde

Talk about flattering, Steven Johnson responds to my review of Interface Culture, taking minor issue with my characterization that there hasn't been a real emergence of an "interface avant-garde". And using his broad definition of the term "interface" he's clearly right.

If one is ooking at interface mainly from the perspective of information, as Johnson does, the broad definition is probably the best approach, if a bit confusing. If however we look at it from the perspective of the computer as a medium, I think a further articulation is warranted. Here we can use interface in its more traditional sense, as the inputs and outputs. In addition to interface we also have the database and the algorithm.

Now if we look on this level we can see that Johnson's examples of new interfaces for information (Google, Technorati, del.icio.us) actually are actually all actions on the database. The big exception is Google which has radically innovated on both the database and also on the algorithm. What we are seeing is not as much the formation of an interface avant-garde but more a database avant-garde. It all adds up to better access to information, improvements to Johnson's broad interface, but all done with out anything but minor improvements to the more traditional interface (which it should be noted is far larger then just the GUI). We may be interfacing better with our information, thanks mainly to our better databases, but the interfaces to our machines continue to limp along..

Interface Culture

Interface Culture might just be Steven Johnson's most important book, but at the same time its probably his least read, and deservedly so. This is a less a book that was written before its time and more of a book that was written too soon. If one could extract an algorithm that Johnson used to generate the book in 1997 and run it today the result would be markedly better.

Johnson's core argument is simple, interface is important and needs to be understood. As simple as it sounds, its never quite been articulated properly, we still lack a language to properly address interface. Johnson comes the closest of anyone in this book, but ultimately he's betrayed by his examples, which are more rotted then dated. The materials to properly write this book just where not around in 1997 and it suffers. Johnson also has a problem I wish a lot more of us had, he writes too well. The prose is so efficiently polished it leaves on craving a clunky paragraph or two simply to break the pace and create a space to actually think about what he's written.

Wrapping up his little breakneck ride, Johnson calls out for the creation of an interface avant garde, a subculture of radical interface designers. Since the books publication there have been numerous moves in this direction, but I'm unsure if any have actually gelled into real form. Flash designers circa 2000, skinners, game modders, they are more like microcultures of interface, never quite reaching the mass and velocity necessary to self-replicate into full fledged subcultures.

The one group that has emerged is the information architect/interaction/experience designer, a set that seemingly seeks obscurity through a constant renaming process. There is no question though this is a subculture, and they tend to focus on a space Johnson quite accurately brought to the for, text as interface. But as an "interface subculture" I find them rather lacking. In Language of New Media Lev Manovich postulates that a core task of new media is the creation of interfaces for databases. And it seems to me much of what the information architects are doing is prepping the database for future interfaces. There is of course a degree of interface innovation going on, but its yet to reach any level fitting of the "avant garde". But much like the shortcomings of Interface Culture in many ways this comes from a lacking in the contemporary database, not from a lack of concepts.

Manovich's book is in many ways a better starting point then Johnson's although they are quite complementary. Manovich devotes quite a lot of space to interface, but essentially stops right where Johnson starts, focusing on the screen as interface through history. Neither break very far from the screen either, interface clearly needs a dose of physical computing. And the tools for analysis of computing culture are still being built, beyond the interface and database lies algorithm (anyone know a good book here?), protocol and beyond..