---
id: wiki-2026-0508-procedural-rhetoric
title: Procedural Rhetoric
category: 10_Wiki/Topics
status: verified
canonical_id: self
aliases: [Persuasive Games, Computational Rhetoric]
duplicate_of: none
source_trust_level: A
confidence_score: 0.9
verification_status: applied
tags: [rhetoric, game-design, persuasion]
raw_sources: []
last_reinforced: 2026-05-10
github_commit: pending
tech_stack:
  language: design
  framework: Bogost framework
---

# Procedural Rhetoric

## 매 한 줄
> **"매 argument 의 made not by words, but by rules"**. Ian Bogost 의 *Persuasive Games* (2007) 의 coined — 매 systems/processes/simulations 의 medium 의 argumentation. 2026 의 LLM-driven dynamic narrative + simulation games (Civ VII, FrostPunk 2) 의 procedural rhetoric 의 mainstream.

## 매 핵심

### 매 Bogost's framework
- **Verbal rhetoric**: words, written/spoken.
- **Visual rhetoric**: images, layout, color.
- **Procedural rhetoric**: rules, systems, feedback loops.
- 매 game 의 unique medium — 매 "what you do" expresses argument.

### 매 Mechanisms
- **Constraint**: what you can't do says as much as what you can.
- **Feedback loops**: reward shapes belief about what matters.
- **Simulation gap**: simplifications reveal designer assumptions.
- **Failure states**: what counts as losing encodes values.
- **Resource economy**: scarcity → priorities.
- **Agency vs. determinism**: how much player matters.

### 매 Examples
- **September 12th** (Frasca): bombing terrorists creates more terrorists.
- **PeaceMaker**: Israeli-Palestinian dual-perspective.
- **Papers, Please**: bureaucratic complicity, moral fatigue.
- **This War of Mine**: civilian war experience.
- **FrostPunk**: authoritarianism as survival logic.
- **Civilization**: progress narrative encoded in tech tree.
- **Cookie Clicker**: critique of incremental design.

### 매 응용
1. Serious games (training, education).
2. News games (Bloomberg's Build the wall vs Don't, NYT graphics).
3. Activist games (climate, refugees).
4. Marketing simulations.
5. AI-driven dynamic narrative (2026 Inworld, Convai).

## 💻 패턴

### Encoding argument as mechanic
```
Argument: "Bureaucracy dehumanizes"
Mechanic: Time pressure + paperwork + small reward for thoroughness, big penalty for missing detail
Result: Player feels the dehumanization rather than reads about it
→ Papers, Please
```

### Resource scarcity as ideology
```
Argument: "Survival justifies authoritarianism"
Mechanic: Heat ↓ over time, citizens demand law book, only authoritarian laws prevent extinction
Result: Player chooses oppression "rationally"
→ FrostPunk
```

### Simplification as critique
```typescript
// Climate simulator: emissions only knob
// By omitting "carbon offsets", "green tech", argues these are insufficient distractions
const tempAtYear = (year: number, emissions: number) =>
  baseline + emissions * year * sensitivity;
// What's missing IS the rhetoric
```

### LLM-driven NPC dialogue (2026)
```typescript
// Inworld-style: NPC values encoded as system prompt + memory
// The game's argument now adapts to player — emergent procedural rhetoric
const npcPrompt = `
You are a refugee in an unnamed conflict.
Your values: family safety > nation > ideology.
You distrust both sides equally.
Respond based on the player's prior actions: ${recentActions.join(', ')}.
`;
// The NPC's reactive logic IS the rhetorical move
```

### Failure state as moral
```
Lose condition: Empire collapses if happiness < 50%
Encoded argument: "Subject welfare is instrumentally necessary, not intrinsically valued"
vs.
Lose condition: Game ends when one citizen dies
Encoded argument: "Each person has infinite worth"
```

### Reward loop as ethics
```
Reward: +XP per enemy killed
Argument (unintended): violence is the path to progress
Reward: +XP per peaceful resolution
Argument: dialogue valued
→ Designers encode ethics whether they intend to or not
```

### Counter-rhetoric (anti-game)
```
Players expect: shooter rewards aggression
Anti-game: every kill ages your character +1 year, game ends at 100
→ Mechanic refutes the genre's implicit argument
```

## 매 결정 기준
| 상황 | Approach |
|---|---|
| Educational claim | Simulate causal model, let player explore |
| Empathy goal | Constrain player to NPC perspective |
| Critique of system | Make player IS the system, feel its pressures |
| Information delivery | Verbal/visual still better for facts |
| Complex policy | Procedural model + verbal scaffolding |

**기본값**: identify the felt experience the player should have, design mechanics that produce it, then verify via playtest.

## 🔗 Graph
- 부모: [[Game-Design]]
- 변형: [[Persuasive-Games]]
- 응용: [[Beat Saber 엑서게임 연구(Beat Saber Exergaming Study)]] · [[Edtech-Industry-Trends]]
- Adjacent: [[Media-Literacy]] · [[Information-Society]]

## 🤖 LLM 활용
**언제**: analyze a game's procedural argument, brainstorm mechanics that embody a thesis, generate playtest probes.
**언제 X**: never collapse procedural rhetoric to "narrative" — the rules ARE the argument.

## ❌ 안티패턴
- **Skin over substance**: stick a "climate change" theme on standard mechanics — argument absent.
- **Sermon mechanic**: the only "right" choice — no agency, weak rhetoric.
- **Unintended argument**: reward loop says A while writing says B — players believe the loop.
- **Realism worship**: simulation accuracy ≠ rhetorical clarity.

## 🧪 검증 / 중복
- Verified (Bogost *Persuasive Games* 2007, *How to Do Things with Videogames* 2011, Frasca *Videogames of the Oppressed*).
- 신뢰도 A.

## 🕓 Changelog
| 날짜 | 변경 |
|---|---|
| 2026-05-08 | Phase 1 |
| 2026-05-10 | Manual cleanup — Bogost framework + 2026 LLM-NPC angle |