My 2016 paper in Sociological Theory Cycles of Conflict
reports a simulation model with four agents.
Kent
Thanks! Got it.
Thanks Kent, really excellent and useful article. Could you, accroding to the footnote 6, share the Excel file?
Eetu
Happy to do so. Here you are: Simulation Model of Four-Agent Collective Control as reported in McClelland 2014 Cycles of Conflict
Kent
For some reason your 4 person spreadsheet made me realize that your “collective control” model (as well as my implemation of it) doesn’t really apply to a tug of war. In your model, all individuals involved in the conflict can set their references for the state of the commonly controlled variable to anything they like and have an appropriate effect on those variables. But in a real tug of war, the individuals on each side of the 0 point – the line that defines the “sides” of the tug of war – must set their references for the position of the flag – the controlled variable – on their own side. If they set their references for the flag being on the other team’s side they would have to push on the rope to move teh flag toward that reference position and you can’t really push very well with a rope.
So it seems to me that your model of “collective control” (and my implementation of it) really only applies to situations where the feedback connection between individuals and their commonly controlled variable is rigid, so that all individuals can exert force on that variable in both a negative (pull) and a positive (push) direction.
The model can be correctly applied to the tug of war situation if it takes into account the fact that, in reality, the individuals involved in the tug of war can only apply force to the controlled variable in a negative (pull) direction.
I think this realization shows (to me, anyway) how important it is to test PCT models against data. Without the constraint of reality, modeling can be quite misleading.
If they set their references for the flag being on the other team’s side they would no longer be engaged in tug of war. But accepting that premise … pulling in a tug of war team is subordinate to controlling a higher-level perception, typically demonstrating who is stronger. Team B has a secret agenda of controlling a perception of seeing members of Team A fall on their backsides in the dirt. To do this, they first control a perception of the flag on their side, and then they control a perception of the flag being on Team A’s side. They do this not by pushing the rope, but by allowing Team A to pull it.
Very nice paper. Especially thought-provoking is the converse relationship between aggressiveness and solidarity (agreement on the collectively controlled variable and its reference value). Intuitively, with less agreement as to goals each member is in conflict with other members as well as with the collective adversary group and its members.
“Perceptual control theory implies that failure to control, which can happen when stalemated conflicts prevent combatants from reaching their goals, is inevitably
frustrating, and that such emotional reactions set in motion reorganization processes in the brain, as individuals cast about for alternative ways to get back in control.”
In subjective experience, emotion may seem causative. Is it? Or is emotion a perception derived from corporeal sensation associated with preparations for action that cannot be executed (frustration)?
According to Bill’s model, emotion is higher-level cortical perception derived from a combination of perceptions of the interior environment in the somatic branch of the hierarchy and perceptions of the external environmental in the behavioral branch. An emotion is a narrative explanation or rationalization of those perceptions. Two people, or the same person on two occasions, may perceive very different emotions on the basis of similar perceptual inputs from these two sources. A well-known example is ‘stage fright’ being perceived by a more experienced actor as ‘eagerness to perform’.
I’ve frequently remarked that emotion is like water: it fills the available container. If frustration results from disturbances or conflicts that can’t be effectively changed or perhaps aren’t even perceived, the narrative of the emotion may give the villain role to some other person or circumstance.
Conflicts of the sort studied in the research that you quote certainly involve collective agreements about narratives and assignments of roles that explain the distress resulting from frustration of control, and which help participants to identify CVs and reference values in common.
When there is less ‘solidarity’ in the group, the frustrations of intra-group conflict plausibly lead to stronger efforts against the perceived adversaries. Aggression against group members is restrained by control of being a group member, and so they are less available as targets of aggression. Emotion fills the available container.
I put a revised version of my spreadsheet model of virtual control up at mu Dropbox site. The only change is that this spreadsheet lets you specify the location of the systems involved in the conflict. You enter their locations in the yellow cells labeled “CS1 Location” and “CS2 Location”. The locations are specified by numbers on the real number line.
The graph of the results of a Run of the model now shows the locations of the control systems (in solid red and blue lines) as well as the locations of the references of both systens for the state of the controlled varable (in dashed red and blue lines). The references still bound the Dead Zone of the conflict and the graph shows variations in the sinusoidally disturbed “virtual” controlled variable relative to that zone.
When you open the spreadsheet you will see that the references for CS1 and CS2 are on opposite sides of each other. If this were a tug of war CS1 would be trying to push the controlled variable into CS2’s territory and CS2 would be trying to push it into CS2’s territory. Of course, this couldn;t be done if CS1 and CS2 were connected to the controlled variable by rope. But it could be done if CS1 and CS2 were connected to the controlled variable by rigid poles. The spreadsheet shows that, in this case, there would still be a conflict with a Dead Zone.
The difference between a “tug of war” type conflict using rope versus one using rigid poles is a difference in the nature of the feedback connection from system outputs to the controlled variable. Since Kent’s model handles conflicts with both of these types of feedback functions, I’m wondering if he has used it to explain any real world examples of virtual control with non rigid feedback functions.