Patrick F. Spear » Posts

My friend sent me this incredible video the other day. Though you really should watch it for yourself, I’ll give you the quick play by play: a research team decides to map the extents of an underground ant colony by pumping cement down the entrance, waiting a month for the cement to cure and then carefully excavating the structure created. The result is a sprawling 10-ton labyrinth of chambers and tunnels.

The structure is something that intrigues me on pretty much every level.

First of all, the colony looks like something that has been intricately design by a central intelligence- some kind of ant architect, squeaking out instructions to to the ant engineers- despite the fact that we know this not to be the case. Instead, the colony is a perfect example of how form can arise directly from an algorithm or a set of generative rules. Each ant is born with a set of instincts, which are instructions that will direct the ant’s behaviour when presented with a set of circumstances. The colony is a result of those instructions being applied in parallel- with no central authority- all over the site, wherever an ant is found. If this is interpreted as as a kind of computational architecture, this process is pretty incredible.

It is pretty incredible from a cognitive design and wayfinding point of view as well. The form of this colony of can be seen as the mapping of a thought process directly onto space. As such, for an ant traversing the labyrinth, every turn and and every space would make inherent sense to the ant’s mind- the space and the thought process would compliment each other perfectly, such that the two working in tandem make both more efficient. Such a scenario is the ultimate goal of wayfinding- “encoding” knowledge into form and signage to reduce the amount of brainpower needed to traverse a space.

It’s also interesting to realize as well this structure is as much a product of the environment as it is of the ants. The colony may be a mapping of a thought algorithm onto space, but the end resultant colony will be a reflection of how that algorithm deals with a particular space. Any site would contain roots, rocks, loose soil, predators, and unpredictable sources of food, all of which flex the ant’s mental model and thus change the shape of the resultant colony.

What would happen if the ants were instead presented with an “ideal” site? Could the structure that arises be considered a “pure form”- the absolutely direct mapping of an algorithm into space? Or would the “ideal site” be another construct that is based on the algorithm in question?

Obviously, ants aren’t the only insects to create large, complex nests. Other notables are termites, whose nests take on cathedral like forms with high-reaching spires…

…and wasps, who create orbs with rippling landing bays…

…and each of these show just how different thought processes would give rise epiphenomenally to different physical forms.

If we wanted to try and do the same for the human thought process, the results would be hopelessly marred by higher reasoning capabilities, the advantage of complex memory, individual experiences and culture. But, what if we could factor those things out somehow? At a base enough level, the maner in which spatial information is processed must be standardized enough to be representative of a fundamental human condition.

If we wanted to create an inherently natural and navigable building- one that human minds seamlessly integrate into, a structure that takes advantage of the nature low-level spatial processing so as to free our complex thinking abilities from the simple task of wayfinding- what would it look like?

As for the original video- I’ve pieced together from the video comments that excerpt is taken from the BBC Planet Earth Series, but I couldn’t track down which part it was from. If anyone happens to know, do share with the rest of us in the comments here.