It seems like a good time to make an all-in-one-place list of things that aren't accounted for in the first wave of modeling I did:
- In those models, individuals' preferences are fixed and can't change in the process of deliberating with one another;
- the focus on proposals, rather than a broader range of approaches to deliberation, is limiting;
- there's a need to think about when strategy is important, that is, when conflict between people's individual interests and the good of the group may color the timing or content of their proposals, or what they're willing to accept at various times in the process, or otherwise complicate the deliberation;
- there are interesting and relevant questions about what to do when a group is too big for a simple all-to-all process, though there's plenty to do without adding that complication;
- the individuals need a language for describing classes of proposals that are good or bad, so they can talk in general terms before narrowing down to absolutely specific proposals;
- a group's process of deliberating about a particular issue is always embedded in the larger context of the group's business and the communities it overlaps with. People may choose to give way on one issue in order to maintain goodwill for the group's long-term health, or so they'll have influence on another issue they care more about; or they may have "external" needs that come into play in the deliberation, such as a strong need to feel heard whether it helps the group reach a decision efficiently or not.
- I am interested in making connections to Scott Page's research on the value of groups with diverse problem solving strategies, as in strategies for exploring the search space, which is a different thing from having different preferences, as we're considering here.
- How does a formal process interact with existing structural inequalities? Many groups have strategies to counterbalance a dynamic in which some people, for instance people of certain races and genders, tend to have relatively more or less sense of entitlement or personal safety speaking for their preferences in a group than others.
Some of these list items overlap with each other.
There are some folks I'd like to have conversations with about the structural inequalities question. I'm thinking of the language question and the issues vs. proposals question as sort of methodological, whereas the changing preferences question is more philosophical and/or empirical, as is the issue of structural inequality. Those four questions are the ones I'm most inclined to pursue right away, though I'm keeping an eye on the strategy question, and half-expecting it to rear up on its own in the models, whether I invite it or not. I think in the short term I can pursue the methodological ones independently of the philosophical ones, and vice versa.
I think that if I leave out changing preferences and strategy, the project is a constraint satisfaction problem, in which the constraints are not all known to any one person at the outset because each individual has access to one piece of the full set of constraints, and the problem is one of designing a good process for discovering a satisfactory solution. I can pursue the questions of language and issues within that frame, and review existing techniques for constraint satisfaction problems.
Simultaneously I want to try to understand the more philosophical/empirical questions better and see how they might change the framework.
Just how do people's preferences change when they deliberate together? One approach to this might be the perspective of Marshall Rosenberg's work on what's sometimes called Nonviolent Communication, other times Compassionate Communication: the fundamental idea there is that people often come into conflict over strategies they want to pursue to address their needs, and by shifting focus to the basic needs themselves, they can find more strategies available that avoid conflict. This may or may not overlap with political scientists' research in the subject of deliberation, which includes studies on the exact question of how preferences change as a result of deliberative communication.
On structural inequality and process, I'm aware of specific strategies that groups adopt, such as "step up, step back," and progressive stacks. I know a little about the idea of microaggressions, and probably more than that, but it would be good to take a direct look into existing work on race, gender and other factors in deliberative processes.