A characteristic of emergent phenomena in a system of many interacting parts is that they exhibit collective behaviour where it looks like the many parts are "dancing to the same tune". But who is playing the music, who chose it, and who conducts the orchestra?
Consider the following examples.
1. A large group of starlings move together in what appears to be a coherent fashion. Yet, no lead starling is telling all the starlings how and where to move, according to some clever flight plan to avoid a predator. Studies of flocking [murmuration] have shown that each of the starlings just moves according to the motion of a few of their nearest neighbours. Nevertheless, the flock does move in a coherent fashion "as if" there is a lead starling or air traffic controller making sure all the planes stick to their flight plan.
2. You can buy a freshly baked loaf of bread at a local bakery every day. Why? Thousands of economic agents, from farmers to truck drivers to accountants to the baker, make choices and act based on limited local information. Their interactions are largely determined by the mechanism of prices and commercial contracts. In a market economy, no director of national bread supplies who co-ordinates the actions of all of these agents. Nevertheless, you can be confident that each morning you will be able to buy the loaf you want. The whole system acts in a co-ordinated manner "as if" it has a purpose: to reliably supply affordable high-quality bread.
3. A slime mould spreads over a surface containing food supplies with spatial locations and sizes similar to that of the cities surrounding Tokyo. After a few hours, the spread of the mould has reorganised so that it is focussed on paths that are similar to the routes of the Tokyo rail network. Moulds have no brain or computer chip but they can solve optimisation problems, such as finding the shortest path through a complex maze. In nature, this problem-solving ability has the advantage that it allows them to efficiently locate sources of food and nutrients. Slime moulds act "as if" they have a brain.
A biologist Michael Levin discusses the issue of intelligence in very small and primitive biological systems in a recent article, Collective Intelligence of Morphogenesis as a Teleonomic Process
[I first became aware of Levin's work through a podcast episode brought to my attention by Gerard Milburn. The relevant discussion starts around 36 minutes].
The emphasis on "as if" I have taken from Thomas Schelling in the opening chapter of his beautiful book, Micromotives and Macrobehaviour.
He also mentions the example of Fermat's principle in optics: the path light takes as it travels between two spatially separated points is the path for which the travel time is an extremum [usually a minimum]. The light travels "as if" it has the purpose of finding this extremum.
[Aside: according to Wikipedia,
"Fermat's principle was initially controversial because it seemed to ascribe knowledge and intent to nature. Not until the 19th century was it understood that nature's ability to test alternative paths is merely a fundamental property of waves."
Similar issues of knowledge/intent/purpose arise when considering the motion of a classical particle moving between two spatial points. It takes the path for which the value of the action [time integral of the Lagrangian along a path] has an extremal value relative to all possible paths. I suspect that the path integral formulation of quantum theory is required to solve the "as if" problem. Any alternative suggestions?
No comments:
Post a Comment