Distributed agency
How many actors can materially redirect the process through action, refusal, interpretation, or defection?
T Working theory
Unprogrammability is a degree, not a switch.
A collective process becomes more unprogrammable as control disperses, feedback intensifies, information stays incomplete, institutions become permeable, and the number of plausible trajectories expands.
Framework12analytical dimensions
The model
Structure still matters. It simply stops being destiny.
Unprogrammability does not deny institutions, resources, leadership, class structure, legal constraints, technology, or coercive power. It asks how those structures interact with actors who can learn, conceal preferences, improvise, coordinate, defect, and react to one another.
The model is therefore relational and dynamic. It focuses less on isolated variables and more on how changing relationships create new possibilities, close old ones, and make previous plans obsolete.
Analytical profile
Twelve dimensions of unprogrammability.
Organizational plurality
How many overlapping organizations, factions, networks, and informal centers of initiative coexist?
Information opacity
How much remains unknown about preferences, loyalties, thresholds, capacities, and intentions?
Feedback intensity
How quickly does one actor’s move change the incentives and expectations of others?
Coalition heterogeneity
How internally diverse are the actors temporarily aligned around a shared objective?
Institutional permeability
How possible are defections, refusals, splits, cross-boundary alliances, or selective enforcement?
Legitimacy volatility
How rapidly can beliefs about rightful authority, acceptable conduct, and institutional credibility change?
Coercive uncertainty
How certain is it that police, military, intelligence, or paramilitary institutions will act cohesively and as ordered?
External coupling
How strongly do foreign actors, diasporas, markets, wars, sanctions, aid, or transnational media alter the process?
Temporal compression
How much faster are events moving than organizations can deliberate, verify, coordinate, and adapt?
Path dependence
How strongly do early contingent choices and accidents reshape later options?
Outcome openness
How many qualitatively different endpoints remain viable at the current phase?
Low ↔ high
Think in profiles, not labels.
DimensionLower unprogrammabilityHigher unprogrammability
AgencyFew actors control consequential decisionsMany actors can redirect the process
OrganizationTight hierarchy and enforced disciplineMultiple overlapping organizations and factions
InformationPreferences and capacities are visiblePreferences, loyalties, and thresholds remain hidden
FeedbackResponses are slow and weakly coupledEach move rapidly changes others’ behavior
InstitutionsOrganizations remain cohesiveSplits, refusals, and cross-boundary alliances emerge
Outcome spaceEnd states are tightly constrainedSeveral qualitatively different outcomes remain live
This is a conceptual diagnostic, not a validated quantitative index. A numerical score should not be published until a defensible measurement model exists.
Process model
The arrows run both ways.
Structural conditions→Trigger→Preference revelation→Institutional response→Adaptation→Reconfiguration→Outcome→Institutionalization
Recursive, not linear: later responses can reactivate earlier mechanisms, change coalition composition, alter legitimacy, and reopen branches that previously looked closed.
What would weaken the theory?
A serious concept needs failure conditions.
The framework becomes less useful if “unprogrammable” merely means “complicated,” if its dimensions cannot be distinguished empirically, if it adds no explanatory value beyond existing concepts, or if apparently open processes are routinely governed by hidden central control that the model systematically misses.
The site should therefore seek negative cases: episodes with apparent decentralization but tight scripting, apparent spontaneity but concealed coordination, or intense contention that nevertheless remains highly predictable because institutional constraints dominate.
How the project tests claims