2nd/10_Wiki/Topics/Other/Sociology of Knowledge.md

---
id: wiki-2026-0508-sociology-of-knowledge
title: Sociology of Knowledge
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [Wissenssoziologie, Social Construction of Knowledge, Mannheim Sociology]
duplicate_of: none
source_trust_level: A
confidence_score: 0.85
verification_status: applied
tags: [sociology, epistemology, philosophy, knowledge, social-theory]
raw_sources: []
last_reinforced: 2026-05-10
github_commit: pending
tech_stack:
  language: agnostic
  framework: social-theory
---

# Sociology of Knowledge

## 매 한 줄
> **"매 모든 knowledge 는 매 social context 에 매 embedded"**. Karl Mannheim (*Ideology and Utopia*, 1929) 가 founding 한 매 sub-discipline — 매 truth claims, 매 scientific paradigm, 매 even mathematical conventions 까지 매 producing community 의 매 social structure 의 reflection 이라고 본다. 매 Berger & Luckmann (*Social Construction of Reality*, 1966) 가 매 modern reformulation.

## 매 핵심

### 매 founders / lineage
- **Marx**: 매 economic base → 매 ideological superstructure 의 매 proto-thesis.
- **Mannheim** (1929): 매 ideology vs utopia, 매 *Seinsverbundenheit des Wissens* (knowledge's existential bondedness).
- **Berger & Luckmann** (1966): 매 reality 의 매 social construction — externalization → objectivation → internalization.
- **Kuhn** (1962): 매 paradigm + 매 normal/revolutionary science — 매 scientific knowledge 의 매 community 의 conventions 으로.
- **SSK / Strong Programme** (Bloor, Edinburgh, 1976): 매 even successful science 의 매 sociological explanation 가능.
- **Latour ANT**: 매 actor-network — 매 human + non-human 의 매 hybrid network 가 매 facts 생산.

### 매 핵심 thesis
- **매 standpoint epistemology**: 매 knower 의 매 social position 이 매 known 에 영향.
- **매 paradigm-bound**: 매 "fact" 자체 가 매 prevailing paradigm 안 에서 만 sense 함.
- **매 distinction (Bourdieu)**: 매 cultural capital, 매 habitus 가 매 academic / scientific 의 매 selection 좌우.

### 매 응용
1. **Science studies**: 매 lab ethnography (Latour & Woolgar *Laboratory Life*).
2. **Tech history**: 매 silicon valley 의 매 culture 가 매 tech direction 형성.
3. **AI ethics**: 매 ML model 의 매 bias 가 매 producing community 의 매 reflection.
4. **Knowledge management**: 매 tacit knowledge (Polanyi/Nonaka) 의 매 social embedding.

## 💻 패턴
(매 mostly conceptual domain — 매 code patterns 적음, but 매 modern computational social science 의 적용)

### Pattern 1: Citation network analysis (매 paradigm detection)
```python
import networkx as nx
from sklearn.cluster import SpectralClustering

# 매 OpenAlex API 의 citation graph build.
def build_citation_graph(seed_papers):
    G = nx.DiGraph()
    for p in seed_papers:
        G.add_node(p["id"], **p)
        for ref in p["referenced_works"]:
            G.add_edge(p["id"], ref)
    return G

# 매 community detection — 매 paradigm proxy.
def detect_paradigms(G, k=5):
    A = nx.to_scipy_sparse_array(G.to_undirected())
    sc = SpectralClustering(n_clusters=k, affinity="precomputed_nearest_neighbors")
    labels = sc.fit_predict(A)
    return dict(zip(G.nodes, labels))
```

### Pattern 2: Bias-in-corpus probe
```python
# 매 ML training corpus 의 매 socio-demographic bias quantify.
from collections import Counter
import re

def occupation_gender_skew(corpus, occupations, pronouns):
    counts = {occ: Counter() for occ in occupations}
    for doc in corpus:
        for occ in occupations:
            for match in re.finditer(rf"\b{occ}\b\s+\w+\s+(\w+)", doc, re.I):
                token = match.group(1).lower()
                if token in pronouns:
                    counts[occ][pronouns[token]] += 1
    return counts
```

## 매 결정 기준
| 상황 | Approach |
|---|---|
| 매 scientific consensus 의 origin 분석 | Kuhn paradigm + 매 citation network |
| 매 ML bias audit | Standpoint epistemology — 매 producer demographics |
| 매 organizational tacit knowledge | Polanyi/Nonaka SECI model |
| 매 tech adoption pattern | Latour ANT — 매 human + non-human |

**기본값**: 매 reflexive — 매 자기 의 분석 도 매 같은 sociological forces 의 subject.

## 🔗 Graph
- 부모: [[Epistemology]]
- 응용: [[Tacit Knowledge]]

## 🤖 LLM 활용
**언제**: 매 academic discipline 의 매 origin / school mapping, 매 paradigm 식별.
**언제 X**: 매 LLM 자체 가 매 producing community 의 매 bias 의 carrier — 매 reflexive caution.

## ❌ 안티패턴
- **Vulgar relativism**: 매 모든 knowledge 가 매 equally valid — Bloor 의 strong programme 의 매 misreading.
- **Reductionism**: 매 모든 fact 의 매 social cause 만 — 매 material/empirical evidence 무시.
- **Self-exemption**: 매 sociology of knowledge 의 매 itself 에 매 적용 X — 매 reflexivity 결핍.

## 🧪 검증 / 중복
- Verified (Mannheim *Ideology and Utopia*; Berger & Luckmann *Social Construction of Reality*; Bloor *Knowledge and Social Imagery*).
- 신뢰도 A.

## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — Sociology of Knowledge 의 lineage (Marx → Mannheim → Berger/Luckmann → SSK → ANT), thesis, computational adaptations 정리 |