Pharmakon Miner

Cross-corpus semantic search system for mapping ancient Greek pharmacological terms to modern biochemistry and quantum mechanisms.

---

🎯 Purpose

Extract untranslated Greek pharmakon (φάρμακον) terms from ancient texts and triangulate them against modern scientific corpora to identify:

• Modern compound equivalents
• Biochemical mechanisms
• Quantum biological effects

---

🏗️ Architecture

┌─────────────────────────────────────────────────────────────────┐
│                        PHARMAKON MINER                          │
├─────────────────────────────────────────────────────────────────┤
│                                                                 │
│  Greek Text: "κυκεών mixed with βάλσαμον heals fever"          │
│                            ↓                                    │
│  ┌─────────────────────────────────────────────────────────┐   │
│  │              ENTITY EXTRACTION                           │   │
│  │  • Greek terms: κυκεών, βάλσαμον                        │   │
│  │  • Symptoms: fever                                       │   │
│  │  • Context: preparation, dosage, ritual                  │   │
│  └─────────────────────────────────────────────────────────┘   │
│                            ↓                                    │
│  ┌─────────────────────────────────────────────────────────┐   │
│  │           MULTI-CORPUS SEMANTIC SEARCH                   │   │
│  │                                                          │   │
│  │  [greek_corpus]      → ancient text context             │   │
│  │  [science_corpus]    → biochem/pharmacology             │   │
│  │  [physicists_corpus] → quantum mechanisms               │   │
│  └─────────────────────────────────────────────────────────┘   │
│                            ↓                                    │
│  ┌─────────────────────────────────────────────────────────┐   │
│  │              CROSS-REFERENCE & MAPPING                   │   │
│  │                                                          │   │
│  │  βάλσαμον ≈ Commiphora/Balsamodendron (myrrh family)   │   │
│  │  κυκεών ≈ barley drink + ergot alkaloids               │   │
│  └─────────────────────────────────────────────────────────┘   │
│                            ↓                                    │
│  ┌─────────────────────────────────────────────────────────┐   │
│  │              CONFIDENCE SCORING                          │   │
│  │                                                          │   │
│  │  🟢 High: 3-corpus hit                                  │   │
│  │  🟡 Medium: 2-corpus hit + strong semantic score        │   │
│  │  🔴 Low: single corpus, speculative                     │   │
│  └─────────────────────────────────────────────────────────┘   │
│                                                                 │
└─────────────────────────────────────────────────────────────────┘

---

📊 Corpora

Corpus	Contents	Use Case
greek_corpus	Ancient Greek texts, medical papyri, Hippocratic corpus	Source terms, historical context
science_corpus	Modern biochemistry, pharmacology, botany	Compound identification
physicists_corpus	Quantum biology, photobiomodulation	Mechanism mapping

---

🔄 Pipeline

1. Extract Greek Pharmakon Terms

• Identify untranslated Greek terms (φάρμακον, βοτάνη, etc.)
• Handle variant spellings (βάλσαμον vs βαλσάμινον)
• Normalize via lemmatization before embedding

2. Pull Surrounding Context

• Wide context window (±3 sentences)
• Extract: symptoms, preparation methods, effects, dosage
• Medical papyri bury details in preparation instructions

3. Embed and Search Science Corpus

• Semantic search for modern equivalents
• Match on: compound names, plant families, effects

4. Map Ancient → Modern Compounds

• Entity co-occurrence analysis
• Botanical family mapping
• Effect/symptom correlation

5. Confidence Scoring

• Weight by number of corpus hits
• Penalize single-source matches
• Boost multi-corpus convergence

---

📦 Output Schema

{
  "greek_term": "κυκεών",
  "transliteration": "kykeōn",
  "context": "ritual drink, grain-based, ergot contamination",
  "modern_matches": [
    {"compound": "ergotamine", "corpus": "science", "score": 0.89},
    {"compound": "LSA (ergine)", "corpus": "science", "score": 0.84}
  ],
  "quantum_links": ["serotonin receptor binding", "neural coherence"],
  "confidence": "high",
  "sources": ["Hymn to Demeter", "Wasson et al."]
}

---

🧪 Demo Cases

κυκεών (Kykeon)

Aspect	Ancient	Modern
Description	Ritual drink at Eleusis	Barley beverage
Active compound	Unknown “mystery”	Ergot alkaloids (LSA, ergine)
Effects	Visionary states	Serotonin receptor agonism
Sources	Hymn to Demeter	Wasson, Hofmann, Ruck (1978)

βάλσαμον (Balsamon)

Aspect	Ancient	Modern
Description	Healing resin	Myrrh family (Commiphora)
Uses	Wounds, fever	Antimicrobial, anti-inflammatory
Mechanisms	—	Terpenoid compounds
Quantum angle	—	Photobiomodulation synergy?

---

🛠️ Implementation

Function: translate_pharmakon()

def translate_pharmakon(greek_term: str, context: str = "") -> dict:
    """
    Translate ancient Greek pharmakon term to modern equivalents.
    
    Args:
        greek_term: Greek term (Unicode or transliterated)
        context: Optional surrounding text for better matching
    
    Returns:
        PharmakonResult with modern matches and confidence
    """
    # 1. Normalize/lemmatize Greek term
    normalized = lemmatize_greek(greek_term)
    
    # 2. Search greek_corpus for usage context
    greek_hits = search_corpus("greek_corpus", normalized, context)
    
    # 3. Extract symptoms/effects/preparation from context
    entities = extract_medical_entities(greek_hits)
    
    # 4. Semantic search science_corpus
    science_hits = search_corpus("science_corpus", 
                                  f"{normalized} {entities}")
    
    # 5. Search physicists_corpus for mechanisms
    quantum_hits = search_corpus("physicists_corpus",
                                  f"{entities} mechanism")
    
    # 6. Cross-reference and score
    return build_result(
        greek_term=greek_term,
        greek_hits=greek_hits,
        science_hits=science_hits,
        quantum_hits=quantum_hits
    )

Dependencies

• Knowledge RAG system (ChromaDB + embeddings)
• Greek lemmatizer (CLTK or custom)
• Entity extraction (spaCy or LLM-based)

---

📋 TODO

• Build translate_pharmakon() function
• Add Greek lemmatization (handle variants)
• Create entity extraction for medical terms
• Build confidence scoring algorithm
• Test with κυκεών and βάλσαμον cases
• Expand to full Hippocratic corpus terms

---

🔗 Related

• Knowledge RAG Setup
• Greek Studies
• Quantum Biology MOC
• Ancient Greek MOC

---

📚 References

• Wasson, Hofmann, Ruck (1978) — The Road to Eleusis
• CLTK (Classical Language Toolkit) — Greek NLP
• Dioscorides — De Materia Medica

---

Last updated: 2025-02-02