name: mip-cluster description: Phase 3 of MIP pipeline - Cluster all metaphor annotations into 4 major CMT systems, with quantitative analysis. allowed-tools: Read, Write

MIP Clustering Phase

Analyze all chapter annotations and cluster metaphors into 4 major Conceptual Metaphor Theory (CMT) systems.

Your Task

Read all 9 annotation files and perform systematic clustering to identify the 4 most significant conceptual metaphor systems in the novel.

Process

1. Load All Annotations

Read each file in mip-analysis/annotations/:

• chapter-01.json through chapter-09.json
• Extract all metaphor annotations into a unified dataset

2. Cluster by Domain Mapping

Group metaphors by their source→target domain mappings:

• Identify recurring patterns
• Note variations within patterns
• Track frequency across chapters

3. Rank CMT Systems

Evaluate each potential system by:

• Frequency: How many instances across all chapters?
• Distribution: How evenly spread across the novel?
• Thematic significance: How central to the novel's themes?
• Variety: How many different lexical manifestations?

4. Select Top 4 Systems

Choose the 4 most significant systems. Expected candidates:

1. DESIRE IS LIGHT - green light, Daisy's luminosity, brightness imagery
2. MORAL DECAY IS PHYSICAL WASTE - ashes, dust, grey, rot
3. THE PAST IS A PLACE - spatial metaphors for time, "back" imagery
4. WEALTH IS DISPLAY/PERFORMANCE - shirts, parties, possessions as presentation

Output Files

mip-analysis/cmt-systems.json

{
  "total_metaphors_analyzed": 67,
  "systems": [
    {
      "id": "system-1",
      "name": "DESIRE IS LIGHT",
      "mapping": "DESIRE IS LIGHT",
      "frequency": 18,
      "distribution": [2, 1, 3, 2, 4, 1, 2, 1, 2],
      "key_instances": ["ch1-003", "ch5-007", "ch9-012"],
      "lexical_manifestations": ["green light", "gleaming", "luminous", "bright"],
      "thematic_function": "Gatsby's romantic aspiration and its ultimate unattainability"
    }
  ],
  "unclustered": [
    {
      "id": "ch3-005",
      "reason": "Unique instance, no recurring pattern"
    }
  ]
}

mip-analysis/cmt-summary.md

# CMT Systems Analysis Summary

## Quantitative Overview
- Total metaphors analyzed: X
- Metaphors clustered into systems: Y (Z%)
- Unclustered/unique metaphors: N

## System Rankings

### 1. [SYSTEM NAME] (N instances, X% of total)
**Mapping**: TARGET IS SOURCE
**Distribution**: Chapters where it appears
**Key examples**: 3-4 representative quotes
**Thematic significance**: Why this matters to the novel

[Repeat for all 4 systems]

## Distribution Analysis
[Chapter-by-chapter breakdown]

## Clustering Notes
[Any decisions made, edge cases, overlapping metaphors]

Success Criteria

• All 9 annotation files read and processed
• 4 distinct CMT systems identified with clear mappings
• Quantitative data accurate (frequencies, distributions)
• Each system has 3+ key examples identified
• Summary provides clear rationale for system selection
• Ready for /mip-outline phase