home / skills / chrislemke / stoffy / conceptual-analysis

conceptual-analysis skill

/.claude/skills/conceptual-analysis

This skill guides you through clarifying concepts by defining necessary and sufficient conditions, testing counterexamples, and refining precise, explicit

npx playbooks add skill chrislemke/stoffy --skill conceptual-analysis

Review the files below or copy the command above to add this skill to your agents.

Files (1)
SKILL.md
7.4 KB
---
name: conceptual-analysis
description: "Master conceptual analysis methodology - defining concepts through necessary and sufficient conditions. Use for: analyzing concepts, testing definitions, finding counterexamples. Triggers: 'what is X', 'define', 'definition', 'necessary conditions', 'sufficient conditions', 'counterexample', 'conceptual analysis', 'analysis', 'concept', 'essence', 'iff', 'if and only if'."
---

# Conceptual Analysis Skill

Master the method of analyzing concepts by seeking necessary and sufficient conditions, testing against counterexamples, and refining definitions.

## Overview

### What Is Conceptual Analysis?

The method of clarifying concepts by:
1. Proposing conditions for concept application
2. Testing against cases (real and imagined)
3. Refining based on counterexamples
4. Reaching reflective equilibrium

### The Goal

**Explicit Definition**: X is F iff conditions C₁, C₂, C₃...
- Each condition necessary
- Jointly sufficient
- Captures the concept's extension and intension

---

## The Method

### Step-by-Step Protocol

```
CONCEPTUAL ANALYSIS PROTOCOL
════════════════════════════

1. TARGET IDENTIFICATION
   └── What concept are we analyzing?
   └── Clarify the question ("What is knowledge?")

2. INITIAL ANALYSIS
   └── Propose conditions
   └── Draw on clear cases
   └── State: X is F iff C₁, C₂, C₃...

3. COUNTEREXAMPLE TESTING
   └── Try to imagine cases that:
       ├── Satisfy conditions but aren't F
       └── Are F but don't satisfy conditions

4. REVISION
   └── Modify conditions to handle counterexamples
   └── Add, remove, or revise conditions

5. ITERATION
   └── Repeat steps 3-4 until stable

6. REFLECTIVE EQUILIBRIUM
   └── Balance analysis against intuitions
   └── May revise intuitions OR analysis
```

### Types of Counterexamples

| Type | Description | Response |
|------|-------------|----------|
| **Too narrow** | Excludes cases that ARE F | Weaken conditions |
| **Too broad** | Includes cases that AREN'T F | Strengthen conditions |
| **Edge case** | Genuinely borderline | Accept vagueness or precisify |

---

## Classic Examples

### Knowledge (JTB Analysis)

```
ANALYSIS OF KNOWLEDGE
═════════════════════

INITIAL ANALYSIS:
S knows that P iff:
1. S believes that P
2. P is true
3. S is justified in believing P

GETTIER COUNTEREXAMPLE:
Smith believes "The man who will get the job has 10 coins"
├── Justified (saw Jones counting coins)
├── True (Smith gets job, happens to have 10 coins)
├── But doesn't KNOW (true by luck)
└── Therefore: JTB is too broad

REVISIONS:
├── No false lemmas
├── Sensitivity: Would not believe if false
├── Safety: Could not easily be wrong
├── Virtue: True belief from intellectual virtue
└── Knowledge-first: Abandon analysis
```

### Free Will (Classical Analysis)

```
ANALYSIS OF FREE ACTION
═══════════════════════

SIMPLE ANALYSIS:
S acts freely iff S could have done otherwise

COUNTEREXAMPLE (Frankfurt):
├── Jones decides to vote for Biden
├── Unknown to Jones, a neuroscientist would intervene
│   if Jones was about to vote Trump
├── But Jones votes Biden on his own
├── Jones couldn't have done otherwise
├── Yet Jones seems to act freely
└── Therefore: PAP (Principle of Alternative Possibilities) fails

REVISIONS:
├── Focus on actual sequence
├── Reasons-responsiveness
├── Source theories (originates in agent)
```

### Art (Definition Attempt)

```
ANALYSIS OF ART
═══════════════

ATTEMPT 1: Representation
├── Art represents reality
├── Counterexample: Abstract art, pure music
└── Too narrow

ATTEMPT 2: Expression
├── Art expresses emotion
├── Counterexample: Some art is cold, intellectual
└── Too narrow

ATTEMPT 3: Significant Form (Bell)
├── Art has significant form
├── Problem: Circular—what makes form "significant"?
└── Uninformative

ATTEMPT 4: Institutional (Dickie)
├── Art = artifact + conferred artworld status
├── Problem: What's the artworld? Circular?
└── Contested

LESSON: Some concepts may resist analysis
```

---

## Techniques

### Case Method

Generate cases to test the analysis:
1. **Clear positive cases**: Obviously F
2. **Clear negative cases**: Obviously not F
3. **Borderline cases**: Test boundaries
4. **Thought experiments**: Imaginative cases

### Necessary vs. Sufficient Conditions

```
NECESSARY CONDITIONS
════════════════════
Required for F-ness but may not be enough

"Being unmarried is necessary for being a bachelor"
├── All bachelors are unmarried
├── But not all unmarried people are bachelors
└── Unmarried is necessary, not sufficient

SUFFICIENT CONDITIONS
═════════════════════
Enough for F-ness but may not be required

"Being a square is sufficient for being a rectangle"
├── All squares are rectangles
├── But not all rectangles are squares
└── Square is sufficient, not necessary

BICONDITIONAL
═════════════
Both necessary and sufficient

"X is a bachelor iff X is an unmarried adult male"
├── All and only bachelors satisfy this
└── Captures the concept
```

### Ockham's Razor for Analyses

- Prefer simpler analyses
- Don't multiply conditions unnecessarily
- But don't oversimplify

---

## Challenges to Conceptual Analysis

### Family Resemblance (Wittgenstein)

- Some concepts lack common essence
- "Game" — no single defining feature
- Network of overlapping similarities

### Open Texture

- Concepts have unforeseen applications
- Cannot anticipate all cases
- Definitions are provisional

### Experimental Philosophy

- Intuitions vary across cultures, demographics
- Are armchair intuitions reliable?
- Need empirical investigation

### Naturalized Epistemology (Quine)

- No sharp analytic/synthetic distinction
- Conceptual truths are just very central beliefs
- Philosophy continuous with science

---

## Best Practices

### Do

- Start with clear cases
- Explain why conditions are chosen
- Consider multiple counterexamples
- Be prepared to revise
- Acknowledge borderline cases

### Don't

- Assume first analysis is right
- Ignore stubborn counterexamples
- Add ad hoc conditions to save analysis
- Claim certainty about contested concepts
- Forget that intuitions can be wrong

---

## Output Format

```markdown
## Conceptual Analysis: [CONCEPT]

### Initial Analysis
X is [CONCEPT] iff:
1. Condition 1
2. Condition 2
3. Condition 3

### Testing
**Clear positive case**: [Example satisfying conditions and being F]
**Clear negative case**: [Example not satisfying conditions, not being F]

### Counterexamples Found
1. [Counterexample 1] — Analysis is too [narrow/broad]
2. [Counterexample 2] — Analysis is too [narrow/broad]

### Revised Analysis
X is [CONCEPT] iff:
1. Revised condition 1
2. Revised condition 2
3. New condition 3

### Assessment
[How confident are we in this analysis?]
[Remaining difficulties?]
```

---

## Integration with Repository

### Related Skills
- `argument-mapping`: Analyzing argument structure
- `logic`: Testing logical relations

### For Thought Development
Use conceptual analysis to clarify key terms in your philosophical explorations.

Overview

This skill teaches the method of conceptual analysis: defining concepts by proposing necessary and sufficient conditions, testing them with cases, and refining definitions in light of counterexamples. It aims to produce explicit biconditional definitions (X is F iff ...) while acknowledging limits like family resemblance and open texture. Use it to sharpen concepts, expose hidden assumptions, and reach reflective equilibrium between analysis and intuition.

How this skill works

You pick a target concept, propose candidate conditions, and formalize an initial biconditional definition. Then generate clear positive, clear negative, and borderline cases and search for counterexamples that make the analysis too broad or too narrow. Revise conditions (weaken, strengthen, or add) and iterate until the definition survives substantive counterexamples or you conclude the concept resists analysis.

When to use it

  • Clarifying a contested philosophical term (e.g., knowledge, art, free action)
  • Drafting precise definitions for research, teaching, or writing
  • Testing whether intuitive analyses hold under thought experiments
  • Finding counterexamples that reveal hidden assumptions
  • Deciding whether a concept admits a necessary-and-sufficient definition

Best practices

  • Start with clear positive and negative cases before formalizing conditions
  • Explicitly mark which conditions are necessary, which are sufficient, and which are conjunctive
  • Actively seek counterexamples — imagine edge cases and use classic thought experiments
  • Prefer simpler analyses (Ockham’s Razor) but avoid ad hoc fixes that only save the analysis
  • Record revisions and why each change addresses specific counterexamples

Example use cases

  • Analyze 'knowledge' by testing JTB against Gettier cases and proposing safety or virtue-based revisions
  • Clarify 'free action' by examining Frankfurt-style examples and considering source or reasons-responsiveness accounts
  • Attempt a definition of 'art' by testing representation, expression, and institutional conditions against counterexamples
  • Teach students how to move from intuitive examples to explicit biconditional definitions
  • Prepare rigorous concept sections for a paper that require clear necessary and sufficient conditions

FAQ

What if a concept resists analysis?

Some concepts show family resemblance or open texture; in such cases record the network of related features, acknowledge vagueness, or accept a pluralistic/cluster account rather than a single biconditional.

How many counterexamples are enough?

There is no fixed number; focus on counterexamples that reveal distinct failure modes (too broad, too narrow, edge-case) and those that force substantive revision rather than ad hoc tweaks.