Extended Brain (2026) — The Genetic Text: How David Haig Reimagines Evolution as Interpretation

Bibliographic Reference

Extended Brain. (2026, March 5). The genetic text: How David Haig reimagines evolution as interpretation. Extended Brain (Substack). https://extendedbrain.substack.com/p/the-genetic-text-how-david-haig-reimagines

Core Argument

David Haig’s framework reconceptualizes the genome not as a program that is executed but as a text that is interpreted. This shift — grounded in the empirical discovery of genomic imprinting and intragenomic conflict — has consequences for how we understand mutation, selection, meaning, and purpose. Haig identifies five conditions for genuine biological meaning: (1) a persistent physical substrate (DNA), (2) causal effects in the world, (3) differential replication based on those effects, (4) accumulated modifications through iterated selection, and (5) context-dependent reading (interpretation). The article traces the implications from molecular biology (the flaw as source code) through developmental biology (interpretation, not execution) to biosemiotics (life as sign-reading at every scale) and naturalized teleology (purpose without a designer).

Key Concepts (relevant to clonal evolution)

1. The genome as contested text

Haig’s kinship theory of genomic imprinting shows that the maternal and paternal genomes within a single organism have partially conflicting evolutionary interests. Paternally expressed genes (e.g., IGF2) promote growth to extract maternal resources; maternally expressed genes (e.g., H19, IGF2R) restrain growth to conserve resources across offspring. The genome is not a unified program but a negotiated text with multiple authors whose disagreements are written into gene regulation. This is directly relevant to cancer: the same imprinted growth-regulatory loci where evolutionary conflict plays out are the loci where loss of imprinting (LOI) drives tumorigenesis.

2. The flaw as source code

Imperfect copying is not a defect in inheritance — it is the condition for evolution. The first mutation was meaningless; it became meaningful only when selected for what it did. Meaning is conferred retroactively by selection. For clonal evolution, this frames somatic mutation not as degradation but as the same productive-error mechanism operating within a single organism’s lifetime — with the crucial difference that somatic selection optimizes for cell-level fitness, not organism-level fitness.

3. Interpretation, not execution

The same gene produces different effects in different contexts (Pax6 produces a camera eye in mice, a compound eye in flies; HOX genes produce legs, wings, or flippers depending on organismal context). Context-dependent reading is not noise — it is the evolved mechanism by which texts achieve adaptive outcomes across varied situations. For cancer, this means a mutation’s functional effect depends on cell type, tissue context, and the existing mutational background — a single driver mutation may mean different things in different clones.

4. Five conditions for biological meaning

Haig’s conditions map cleanly onto somatic evolution: (1) DNA as persistent substrate, (2) mutations have causal effects (driver vs passenger distinction), (3) differential replication = clonal expansion, (4) accumulated modifications = sequential driver accumulation in the Bozic-Nowak framework, (5) context-dependent reading = the same mutation in different tissue contexts or clonal backgrounds produces different phenotypic outcomes.

5. Naturalized teleology — purpose without a designer

Haig argues that a gene’s function is the reason it exists — not “as if” purpose (Dennett’s intentional stance) but genuine purpose grounded in the causal history of selection. Applied to somatic evolution: a driver mutation’s functional effect (increased proliferation, reduced apoptosis) is the reason that clone expands. The “for” is not metaphorical — it is the most economical description of the selective history.

Concepts Referenced

Genomic imprinting, kinship theory, intragenomic conflict, loss of imprinting (LOI), biosemiotics, semiotic niche, semiotic freedom, epistemic cut (Pattee), naturalized teleology, interpretation vs execution, text metaphor for genome

Entities Referenced

• David Haig — Harvard evolutionary biologist; kinship theory of genomic imprinting; author of From Darwin to Derrida (2020)
• Howard Pattee — Physicist; epistemic cut concept (matter-symbol boundary)
• Jesper Hoffmeyer — Danish biosemiotician; semiotic freedom, life as semiosis
• Michael Levin — Developmental biologist; “Picasso tadpoles” demonstrating interpretative developmental competency
• Daniel Dennett — Philosopher; intentional stance (contrasted with Haig’s genuine teleology)
• Jakob von Uexküll — Early biosemiotician; tick’s Umwelt as three-sign universe
• Genes: IGF2, H19, IGF2R, Pax6, HOX, GNAS, syncytin (viral co-option), antifreeze proteins (broken-enzyme repurposing)
• CRISPR — Bacterial immune system as captured viral code repurposed

Limitations

• This is a popular science essay (Substack), not a peer-reviewed source. It interprets Haig’s work for a general audience. Claims should be verified against primary sources before being cited as evidence.
• The article synthesizes Haig, Pattee, and Hoffmeyer into a unified narrative; the synthesis itself (biosemiotics as extension of Haig’s framework) is the author’s interpretation.
• Cancer connections are not made in the article; they are inferred from the framework.

Relevance to Clonal Evolution

This article provides the philosophical scaffolding for understanding why genomic imprinting matters to somatic evolution. Key bridging concepts:

1. Functional haploidy of imprinted genes: Because only one allele is active, imprinted tumor suppressors require only one hit to inactivate. This accelerates the effective driver mutation rate — a parameterization concern for the branching-process-model.

2. LOI as epigenetic driver: Loss of imprinting can produce growth-factor overexpression (biallelic IGF2) or growth-restraint silencing (H19) without any DNA sequence change, at rates faster than genetic mutation. This provides a mechanism for epigenetically-driven clonal-expansion.

3. Interpretation framework for clonal heterogeneity: Haig’s insight that “the same text means different things in different contexts” applies to somatic mutations — a BRAF V600E mutation means different things in melanocytes (driver), colon (passenger in some contexts), and thyroid (driver). Context-dependent interpretation of mutations is the rule, not the exception.

4. The flaw as source code — somatic edition: The same logic that makes germline copying errors the source of evolutionary novelty applies within the soma: somatic mutations are “textual corruption” that, when selected, becomes functional innovation at the cellular level. The cancer genome is not degraded — it is reinterpreted under a different selective regime.