2018.07.03. Self-organization vs. Centralized Administration

(Right now: In McDo.) 

One of the easily recognizable manifestations of the emergence of complex behavior is through the complex pattern formation of groups of animals in motion. From flocks of starlings moving about in the sky, to the schools of fishes under predatory attack, to ants colonies and wildebeests herds, we can see the occurrence of a globally coordinated behavior that is driven by the simple local interactions of agents with limited information. This phenomenon is referred to in literature as self-organization, a driver of complexity wherein complex, unpredictable, and organized behavior emerges with little to no influence from a top-level, centralized administration, but, instead, from the bottom, through the multitudes of simple nearest-neighbor interactions.

A school of tuna forming a tornado-like vortex. Image from here.

Starlings creating wonderful self-organized patterns in the sky. The image originally appeared here.

Self-organization is an important concept with ubiquitous examples in nature and human society. Apart from the actual physical patterns observed in animal groups discussed above, one can also observe this in non-living systems, such as in earthquakes: despite the very slow rate of driving through tectonic motion, the distributions of earthquake moment magnitudes are statistically very similar across various periods and geographical regions, and follows a power-law scaling across a very broad range of values. In human interactions, despite the very personal reasons for choosing friends, humans also exhibit power-law distributions of the number of actual friends. In economics, as early as the 1700’s, the idea of an “invisible hand” that self-regulates the market, a central tenet of the laissez-faire market philosophy, may be viewed as a manifestation of self-organization. 

Oftentimes, self-organization is given a positive connotation, no doubt due to the layman’s definition of the word “organization.” The result of the self-organized process is always assumed to be positive, beautiful, or, well, organized. But the thing is, it need not be. 

Part of my research interests is on the signatures of self-organization in the urban setting, particularly in the street network architecture. Interestingly, self-organization, in this particular complex system, indicates unregulated growth that is not governed by a centralized plan. Barthelemy et al. [1] studied these two extremes in the case of the city of Paris, France. The road network in the French capital has been growing organically, i.e. in a self-organized manner, for centuries; but during the second empire, Baron Hausmann introduced drastic, top-level changes in the city road network layout. The researchers found that the planned changes actually improved the navigability of the road network, as it changed the centrality profile of the network structure. In that sense, planning actually had a “better” result in the sense of efficiency or overall aesthetics. But the self-organization is still present in the sense that most of the development are not imposed from the top level but are based on local conditions. Other works have also looked at the differences between the planned and self-organized cities, particularly in terms of robustness to attacks [2], which have important implications in the main function of road networks as the veins of transportation within the urban setting.

A portion of Metro Manila, as viewed from hilly Antipolo. 

One of my students, Michelle, has been studying road network properties of mostly self-organized cities, with a particular emphasis on the cities in the major metropolitan areas in the Philippines. Our work provided an empirical morphological description of the cities, that is, a description of the shapes and geometries created in space by the road network. Our findings have revealed the similarities and differences between the Philippine cities and representative cities that were drawn from strict master plans, namely, Canberra in Australia and Brasilia in Brazil [3]. Some of these findings, along with additional analyses of the additional data sets, are being prepared for publication and/or conference presentation.  These findings add to a growing literature on urban systems that view cities as complex systems capable of self-organization. Moreover, it introduces an analysis of the Metro Manila conurbation, which, while being one of the most populous and economically vibrant in the southeast Asian region, is not being featured so much in the complex systems literature. In a sense, there is a developmental aspect to these kinds of studies, as these works may form a quantitative basis for formulating policies for the better regulation of the metropolis.

Needless to say, there are other aspects of human societal interactions that may be viewed from the lens of self-organization. Another example that I am particularly interested in is in the evolution of a centralized police force for maintaining order in the society [I discussed the origin of the idea in a post in my private research blog, but I intend to also discuss some of the details in this blog soon]. My student, Merimae, has worked on in-silico experiments that simulate the assumed dynamics of the society in the form of modified prisoner’s dilemma (PD) games in a grid. One of our interesting conclusions so far is that a generally cooperative society emerges when there are punishers for defection and when the agents change strategy (i.e. whether to cooperate or defect) based on local conditions [4]. Again, a manifestation of self-organized behavior!

But then again, several interesting questions remain. Would we be better off if the form of punishment itself were self-organized, i.e. anyone can punish anyone else for behaviors deemed to be disruptive or bad. The fear with a dedicated force for punishment, of course, is that, because it is paid for by the general population and it is armed, if it goes rogue, then the society may be paying for its own dangerous undoing. But with a decentralized, one-on-one punishment, society may go out of control, and lead to similar catastrophic consequences. Or will it?

Galileo facing the Roman Inquisition by Cristiano Banti (1857).
The image is in the public domain and downloaded from here.

And the same goes for the idea of organization and authority in general. If order can be achieved from the bottom-up, is the current modern setup — that of surrendering our individual freedoms to a centralized government of a few, selected people — an optimal strategy? After all, the starlings that we discussed in the intro have no president or prime minister, and they can organize. Can the same work for humans? Just copy and adjust based on your neighbors, and everything will be just fine? The first thing to be eliminated in this kind of self-organized “governance” is the abusive figure of authority, who is oftentimes portrayed as (or, is actually) incompetent, insecure, and just plain wrong. Is it worth it?

These questions, I know, invoke post-apocalyptic scenes a la Revolution (yeah, well, my wife and I watched it, and we thought the first two seasons were good). But these are just some of the questions that came to my mind upon exploring this idea of self-organization. I hope to have more discussions with my (possibly) new students to find some other frontiers to explore under this paradigm. 

References: 

[1] M. Barthelemy, P. Bordin, H. Berestycki, and M. Gribaudi, Self-organization versus top-down planning in the evolution of a city, Sci Rep 315(3), 2153 (2013). 

[2] A.P. Masucci and C. Molinero, Robustness and closeness centrality for self-organized and planned cities, Eur Phys J B 89, 53 (2016). 

[3] M.T. Cirunay and R.C. Batac, Spatial metrics for characterizing the urban road networks of cities with different levels of global planning, Proceedings of the Samahang Pisika ng Pilipinas (2016) [link to article]. 

[4] M.S. Villamayor and R.C. Batac, Emergence of cooperator dominance in an iterated prisoner's dilemma games with punishing agents and dynamic strategies, Proceedings of the Samahang Pisika ng Pilipinas (2018) [link to article].