Sunday, March 15, 2009

Emergence 5: Emergence Facilitated by Social Structure

What social structures facilitate emergence (particularly strong emergence), and how can we best modify our current social structures to better facilitate beneficial emergence?  This is a challenging set of questions, partially because it has been little studied, and partially because it is difficult to directly experiment with social structure when we have but one interconnected society to observe, and where we can't really "experiment" on society as a whole directly.  None-the-less, this post will attempt to explore the circumstances and attributes of society that lead to--or repress--emergent phenomena.

One method to gain insight into the circumstances that facilitate emergence in society is through the use of agent based modeling.  Early efforts to model societies did not use autonomous agents, but rather used hierarchically organized units controlled by a centralized controller.  Sawyer, Social Emergence p. 147, citing Connah and Wavish 1990.  However, when researchers began to experiment with decentralization, and designing systems of autonomous agents not subservient to any centralized control, they were better able to reproduce emergent phenomena.  Id.  Additionally, agent based modeling has shown that social norms must be known and understood by the individual agents in order for cooperation (one emergent phenomena) to emerge--it is not sufficient for these norms to be imposed from above.  This observation that "social emergence cannot occur unless the participating agents have explicity representations of mutual beliefs, team plans, and team goals and are capable of communicating to negotiate and coordinate their beliefs and plans," (Sawyer p. 179, quoting Tambe, 1997) suggests the importance of pushing both the decision making and understanding about the overall social system down to the lowest possible level as a means to facilitate the kind of behavior that can result in emergence.

Another method to gain insight into the circumstances that facilitate emergence is to look to historical examples.  Some sociologists have observed that emergence is a function of both the number of units and the complexity of the rules of interaction between members of a society.  Baas (1994); Darley (1994) (paraphrasing Sawyer, Social Emergence, p. 97).  While it is purely anecdotal, history also makes clear that powerful cultural, scientific, and spiritual innovation takes place in environments of free-flowing interaction between peers far more than in rigid and hierarchal structures:  the Renaissance arising out of the merchant class of Italian city-states; classical philosophy arising out of the quasi-democratic gentry of Athenia and other Greek city states; innovation springing from the university environment modeled after the generally peer-society of monastic orders rather than from the hierarchal structures of state administration or militaries; the cultural advances of relatively more free and freely communicating societies such as 1960's/1970's USA and Western Europe vs. 1960's/1970's USSR and Eastern Europe; etc.  This kind of historical analysis could go on indefinitely, but I think this brief introduction is sufficient to propose--though not definitively prove--the importance of peer-to-peer communication, peer equality, leisure time, and routine interaction between diverse groups as important to the facilitation of social emergence.

What about "strong" emergence specifically?  Thus far the limited examples available to us on how and where emergent properties come to exist in society does little to explain how strong emergence (emergence that is fundamentally non-reducible and capable of downward causation, as opposed to merely practically non-reducible) arises.  This is especially troubling because it is strong emergence that (as I will explain in the concluding post in this series, next) holds the greatest promise to re-shape society.  I think there are two lessons from this.  First, we need to study those systems that do exhibit strong emergence to see what can be learned (see below).  Second, we should realize that strong emergence may not always be clearly identifiable (e.g. the noosphere, the "global brain," synchronicity, etc.), and that by facilitating emergence generally we may in fact be facilitating strong emergence.  At a minimum, I am aware of no evidence to suggest that the structures that facilitate emergence generally can inhibit strong emergence specifically.

Frustrating, but as the next post will discuss, there may be further insight to be gained by analogy from analysis of another system that does (apparently) exhibit strongly emergent properties:  the human brain.

Readers may also find my litigation checklist of interest.


Anonymous said...

If something is strongly emergent why would it need facilitation?

Jeff Vail said...

Strong emergence is a phenomena that arises out of a specific structure (at least that's the best of my current understanding). For example, consciousness is a strongly emergent phenomenon that arises out of the physical structure and operation of the human brain. By making a structure that was not sufficiently like the human brain (or that did not sufficiently exhibit the structural prerequisites for strong emergence) into something that does cross that threshold, we can faciltiate strong emergence. So, for example, while consciousness may be strongly emergent, it doesn't emerge from moder day computer systems--even those that intentionally try to achieve artificial intelligence. That's not to say that we couldn't facilitate it by making that computer's structure and operation more like that of the human brain. Similarly, while modern economic systems may not exhibit strong emergence, that attribute could be facilitated by modifying the structure.

Of course, there's the side issue of "degrees" of strongly emergent phenomena: rat brains and human brains are, in many fundamental structural ways, fairly similar, and clearly they result in the emergence of some level of intelligence, but it also seems clear that this is not qualitatively the same as what emerges from the human brain. This suggests (at least to me) that there are structural attributes that can be enhanced to facilitate an increased order or a greater degree of strongly emergent phenomena...