A portable method for making invisible system effects visible.
A methodology for studying system dependency, cognitive scaffolding, and non-reversible transformations
Bounded Fictional Analysis is a research method for examining how systems reshape cognition, documentation, and coordination over time. It uses deliberately constructed fictional scenarios to isolate dynamics that are difficult or impossible to study through direct observation of real systems.
This approach is particularly useful when:
- Real systems are too entangled to examine single variables cleanly
- Gradual changes obscure what's due to system effects vs. other factors
- Ethical constraints prevent direct experimentation
- Counterfactual scenarios need exploration before committing to implementation
The methodology systematically examines system integration through five distinct phases:
1. Baseline — Document pre-system state
- How do actors function without the system?
- What are existing capabilities and limitations?
- What cognitive frameworks are already in use?
2. Introduction — Observe initial adoption
- What immediate behavioral changes occur?
- How do habits begin to form?
- What cognitive restructuring happens?
3. Steady-State — Examine full integration
- What new capabilities emerge?
- What old capabilities atrophy?
- How deeply embedded does the system become?
4. Removal — Track discontinuity effects
- What happens when the system suddenly disappears?
- Which behaviors persist despite losing their function?
- How do cognitive frameworks adapt or fail to adapt?
5. Adaptation — Document long-term equilibrium
- What remains after the system is gone?
- Can actors return to pre-system baseline?
- Is the transformation reversible?
The methodology reveals four key patterns across different domains:
Actions that continue after their functional purpose disappears.
Example domains: Deprecated API usage, vestigial workflow steps, phantom limb phenomena
Cognitive frameworks that reference absent infrastructure.
Example domains: Documentation systems, memory architectures, collaborative tools
Artifacts that persist but lose their interpretive framework.
Example domains: Legacy data formats, archived communications, platform migrations
Physical or social structures shaped by systems that no longer exist.
Example domains: Organizational patterns, spatial arrangements, coordination protocols
This methodology applies across multiple fields:
AI Systems
- LLM memory and context windows
- Platform discontinuity and API deprecation
- Tool dependency and migration paths
- Memory system architecture
Documentation & Knowledge Management
- Archival practice and metadata preservation
- Format obsolescence
- Knowledge transfer across system changes
Cognitive Science
- Scaffolded thinking and external cognition
- Tool-mediated perception
- Habit formation and behavioral persistence
Systems Design
- Graceful degradation patterns
- Legacy support requirements
- Migration path planning
Create a bounded world where:
- The system has clear, observable effects
- System presence/absence can be cleanly toggled
- Actors respond in recognizable patterns
- Multiple levels of analysis are accessible
Systematically examine each phase, documenting:
- What changes at individual, social, and structural levels
- Which effects persist across phase transitions
- What proves reversible vs. irreversible
Identify:
- Behaviors that outlast their function
- Cognitive frameworks orphaned from infrastructure
- Artifacts that lose context
- Structures that persist after their cause disappears
Test whether fictional patterns help explain real phenomena:
- Platform shutdowns
- Memory system changes
- Tool discontinuation
- Format obsolescence
✓ Isolates confounded variables
✓ Examines counterfactuals safely
✓ Makes invisible dynamics visible
✓ Generates testable hypotheses
✓ Communicates complex ideas accessibly
✗ Cannot provide quantitative predictions
✗ Does not replace empirical validation
✗ Cannot settle ontological questions
✗ Does not prove causal mechanisms in real systems
✗ Cannot substitute for direct observation
A successful fictional analysis demonstrates:
Cross-domain recognition — People from different fields recognize the patterns
Explanatory power — The framework makes previously confusing phenomena understandable
Predictive utility — It anticipates what happens in similar scenarios
Generative capacity — It reveals new questions or dynamics
Boundary clarity — Limitations are clearly stated
Critical requirements:
- Explicit fictional framing
- Clear consent boundaries
- No hidden probing or manipulation
- Transparent documentation of process
Why this matters: Fictional analysis should explore ideas safely, not conduct covert research.
Important caveats:
- Fictional actors aren't real people
- Patterns suggest hypotheses, not prove facts
- Individual variation is significant
- Cultural context matters
This framework emerged from an earlier experimental thought exercise on system mediation. The analytical approach proved reusable beyond that context and is presented here independently.
Good fit:
- Exploring system dependency dynamics
- Understanding non-reversible transformations
- Examining scaffolding effects
- Studying documentation feedback loops
- Investigating context preservation
Poor fit:
- Measuring specific system performance
- Validating technical implementations
- Making policy recommendations
- Conducting security research
- Reverse-engineering actual systems
At each phase, examine:
- Individual cognition
- Social coordination
- Physical infrastructure
- Temporal persistence
Look for:
- Behavioral lock-in
- Orphaned sophistication
- Context discontinuity
- Residual infrastructure
Ask:
- What changes?
- What persists?
- What's reversible?
- What generalizes?
If referencing this work:
Bounded Fictional Analysis: A methodology for studying system dependency and non-reversible cognitive transformations through deliberately constructed fictional scenarios.
For a complete catalog of related research:
📂 AI Safety & Systems Architecture Research Index