|The Devil’s Advocate™|
Our fascination with algorithms, big data, artificial intelligence and exponentially accelerating technologisation leads us to believe we can reduce the world’s organisation, and therefore its problems, down to predictable, rationalisable, atomised units.
There is some irony that this blind faith in the model mistakes the map for the territory. Over-reliance on the model causes over-reliance on a model.
But a model only models what it can model.
In Seeing Like A State, James C. Scott calls this the problem of legibility — because a simplistic model can’t adequately react to the nuances of an autonomous organic network, political administrations oblige their populations to organise themselves to best fit the model rather than having the model fit them. The model itself, by its existence, queers the pitch, skews incentives. People optimise for the model, often undermining the model’s original goals — tax planning, right? They fill in windows, leave their chimneys unfinished to avoid paying taxes calculated on the number of windows, or payable upon completion of structures.
Thus, a model is not just an inadequate representation of how a system behaves; it is a politically-enforced model that corrupts the behaviour of the system in itself.
Also pitted against the reductionists and the high modernists are the systems theorists and complexity people, two of whom are featured in the video in the panel. Joe Norman (to the right) makes an interesting assertion that, in any system, informality — arrangements outside the model or that the model cannot see and therefore treats as non-existent — are fundamental to its operation. Indeed, the “formal” parts of a system are just small islands in a sea of informal relations.
- Jane Jacobs makes the same observation about the modernist city planners of the 1940s and 1950s.