Meme Fitness Correlates with Compressibility

Claim

[!abstract] Thesis Statement The “fitness” of a meme (probability of successful transmission and replication) is positively correlated with its compressibility: memes that can be encoded in fewer bits while maintaining reconstruction fidelity spread more successfully than informationally dense alternatives.

Formal Statement

Let $M$ be a meme with information content $I(M)$ bits. Define:

• Compression ratio: $\rho(M) = I(M) / L(M)$ where $L(M)$ is the compressed representation length
• Reconstruction fidelity: $F(M, \hat{M})$ measuring similarity between original and reconstructed meme
• Transmission probability: $P_T(M)$, probability of successful human-to-human transmission
• Fitness: $W(M) = P_T(M) \cdot F(M, \hat{M}) \cdot R(M)$ where $R$ is replication motivation

Claim: For memes in competition, $W(M) \propto \rho(M)^{-1}$ when controlling for content utility.

$$\frac{\partial W}{\partial \rho^{-1}} > 0$$

High compressibility (low $\rho$, meaning $L(M) \ll I(M)$) predicts high fitness.

Why This Matters

This claim is central because it:

1. Connects information theory to cultural evolution: Provides a quantitative framework for memetics
2. Explains empirical regularities: Why proverbs, jokes, and slogans dominate cultural transmission
3. Predicts anti-memes: High-$\rho$ ideas (incompressible) should systematically fail to spread
4. Enables analysis: We can actually measure compression ratios and test the prediction

Supporting Evidence

Theoretical Support

1. [[Data Processing Inequality]] — Information is lost in transmission; only compressible memes survive lossy channels
2. [[Rate-Distortion Theory]] — There’s a minimum transmission rate for acceptable fidelity; compressible memes fit under this bound

Empirical Support

1. [[Proverb Compression Analysis]] — Cross-cultural analysis shows proverbs have high compression ratios (4:1 to 10:1 meaning-to-words)
2. [[Viral Tweet Entropy Study]] — Viral tweets show lower per-character entropy than non-viral ones

Historical Precedent

• Oral traditions preferentially preserve metrically structured content (compression via prosodic redundancy)
• Religious texts develop compressed formulations (creeds, mantras) for reliable transmission

Counterarguments

Objection 1: Complex ideas do spread

Some complex, incompressible ideas (e.g., calculus, quantum mechanics) have spread widely.

Response: These spread through institutional infrastructure (schools, textbooks, formal training) that provides error-correction and extended transmission time. Without institutions, they’d be anti-memetic. The claim is about unaided human-to-human transmission.

Objection 2: Emotional salience matters more than compression

Many viral memes succeed through emotional punch, not information efficiency.

Response: Emotional salience is orthogonal to compressibility. A meme can be both emotionally salient AND highly compressible (most successful memes are). The claim is that among emotionally equivalent memes, compressibility predicts fitness. Additionally, emotional content may enable compression by hooking into existing cognitive structures.

Objection 3: Network effects dominate

Meme spread depends on network structure, not inherent meme properties.

Response: Network structure affects which memes get exposed, but compressibility affects whether exposure leads to adoption and retransmission. Both factors matter; the claim is about the latter.

Confidence Assessment

Factor	Rating	Notes
Theoretical grounding	4/5	Solid info-theoretic foundation via DPI and rate-distortion
Empirical support	3/5	Suggestive but not rigorous; needs more data
Logical coherence	5/5	The argument follows directly from Shannon
Novelty	4/5	Formalizes intuitions; few have done this rigorously

Overall Confidence: 0.75

What Would Change My Mind

• If high-compression memes failed in controlled transmission experiments while low-compression succeeded
• If viral content analysis showed no correlation between entropy measures and spread
• If someone demonstrated that receiver reconstruction happens via different mechanisms than compression/decompression

Implications If True

1. Pedagogy: Compress ideas before teaching (the “curse of knowledge” is partially a compression failure)
2. AI alignment: LLM outputs should be calibrated for human compressibility
3. Epistemics: We should expect compressed (and thus simplified) versions of truths to dominate discourse
4. Engineering: Meme design could be optimized via rate-distortion analysis

Related Claims

• [[Proverbs Are Error-Corrected Codes]]
• [[Anti-Memes Have High Conditional Entropy]]
• [[Tokenization Boundaries Affect Meme Spread]]

Sources

• Dawkins, R. (1976). The Selfish Gene, Chapter 11
• Sperber, D. (1996). Explaining Culture, Chapter 5
• Cover & Thomas (2006). Elements of Information Theory