Collective Control Part 3: Giant Virtual Controllers

As a sociologist, I have found collective control particularly interesting for two reasons:

First, the theory of collective control provides a way to understand conflict and the escalation of conflicts, and conflict is a pervasive feature of social life. My 2014 article, called “Cycles of Conflict,” offers a PCT analysis of conflict, and my just-published 2019 article comparing mediation with MOL (co-authored by Warren Mansell) discusses the resolution of conflicts. (The articles can be accessed from my ResearchGate page.)

Second, the theory of collective control is scalable upward from an explanation of the collective behavior of pairs of control agents to an analysis of the collective behavior of hundreds or thousands or millions of controllers. My goal as a sociologist has been to use this PCT analysis to explain the widespread uniformities in collective behavior that create the social and cultural patterns we see around us. In my 2004 paper on “The Collective Control of Perceptions,” I began to describe how this PCT analysis can apply on a society-wide scale, and my forthcoming chapter for the Handbook of Perceptual Control Theory elaborates my arguments in considerably more detail.

The most important thing to remember about scaling up the analysis of collective control is that the greater the number of control agents involved in a joint action, the less important the contribution of any single agent to the overall result. In a contest between two evenly matched control agents, a single agent may have quite a bit of leverage to determine the state of environmental variable they both perceive. With only a few controllers involved, the power of a single individual to change the overall outcome may still be substantial. But when thousands or millions of controllers are participating in a joint action, the contribution of any individual control agent to the overall outcome will be so small as to be negligible. Whether an individual even participates or not will have no perceptible effect on the stable result emerging from the collective actions of all the other controllers involved.

Martin Taylor and I have described these collective actions of large numbers of control agents as Giant Virtual Controllers . As I noted in my post on Conflict and Cooperation, the outcome in these cases is as if a more powerful control agent were controlling its perception of the environmental variable in question by using a loop gain equal to the combined loop gain of the participating control agents and a reference value equal to the gain-weighted average of the reference values that the participating agents have for their perceptions of the variable. If all the participating agents were to have exactly the same reference value, the virtual reference value apparently used by the giant virtual controller would precisely equal to that value. But in cases of conflict, the virtual reference value of the giant virtual controller is not necessarily identical to that of any of the participating agents.

My contention has been that giant virtual controllers are the source of the social and cultural stability (such as it is) that we see in social groups, whole societies, and even in social and cultural patterns world-wide.

1 Like

Hi Kent

KM: My contention has been that giant virtual controllers are the source of the social and cultural stability (such as it is) that we see in social groups, whole societies, and even in social and cultural patterns world-wide.

RM: What social phenomena did you observe that led you to this conclusion? What are some examples of social and cultural stability that are explained by the model of a giant virtual controller?



I’ll be coming to that in another post that I’m currently still working on. Thanks for your interest.


A point to emphasise is that any Giant Virtual Controller can be treated and used in more complex situations as though it was a single unitary controller of the same dimensionality. Likewise, unless one can see the entire mechanism, one can never distinguish whether a controller is the unitary controller it may appear to be or is actually a Giant Virtual Controller.

The internal neural structure of the brain is a case in point. It is generally assumed that control is implemented by firings of many individual nerves throughout the internal part a control loops operated by the brain. Powers invented the concepts of “neural bundle” and “neural current” to analyze control loops and systems, to smooth all the firings of relevant neurons.

As Kent says, any one member of a GVC has little influence by itself, and that applies equally to individual neurons just as much as it does to social interactions. Each entire control loop of the kind usually discussed in PCT discussions has linkages that carry only neural currents. Every perceptual controller in our brain is thus a Giant Virtual Controller, which we legitimately treat as a unitary controller for the purposes of experiment and analysis.