Main text
1. The Illusion of the Copy
Two genetically identical newborns are born in the same maternity ward. They share the entirety of their DNA, three billion base pairs, each replicated identically from a single fertilised cell. Their genomes are, in the most technical sense, indistinguishable. But their fingerprints differ radically. Each line of the epidermis, each papillary bifurcation, each point of friction created by movements in utero during development, produces a unique pattern. So too does their susceptibility to disease diverge: one may develop schizophrenia, the other not; one may suffer from asthma, the other remain healthy; one will have a cardiac predisposition, the other robust circulation. Their temperaments, despite undeniable resemblance, depart in divergences that parents notice immediately after birth. One sleeps deeply; the other wakes at every sound. One accepts a change of position; the other resists. That which genetic resemblance promised, indistinguishability, is not fulfilled by nature. Monozygotic twins present themselves as the figure of impossibility: if the perfect genetic copy does not produce a perfect copy of the organism, then reproduction does not copy. The genome is not a blueprint that determines construction. Reproduction produces a variant.
The divergence between genetically identical twins calls into question the fundamental premise of genetic determinism: that the genetic code unequivocally specifies form. But the mechanism of this divergence is precisely quantifiable and allows us to understand how reproduction, far from being repetition, is a systematic operation of differentiation. During intrauterine life, even before any post-birth environmental experience, the two fetuses develop in slightly distinct contexts. Mechanical pressures in the uterus, nutrient circulation, maternal hormonal fluctuations, none of these is uniform or identical for both. One fetus, positioned slightly differently, undergoes variable compression on its hands; this friction determines the unique pattern of epidermal ridges. Signalling molecules crossing the placenta arrive in marginally diverse concentrations; this variation alters the intensity of development of certain neural circuits, producing distinct susceptibilities to affective disorders or differential resistance to pulmonary inflammation. None of these variations is coded in the genome; none is replicable; none can be predicted from the DNA sequence. The genome provides constraints, boundaries of possibility, not specification. Within these boundaries, contingency unfolds inevitably. Two organisms with identical genomes inhabit not only different environments, but also intrauterine contexts that are never fully replicable. Reproduction, even in its most mechanically uniform form (the fertilisation of a single egg that subsequently divides), does not produce duplicates.
The phenomenon of twins reveals an architectonic truth about life: form is the remainder of material operations, not an essence. Where there was hope of identity through absolute genetic identity, there emerges the realisation that biological form is inextricably dependent on context. The process that produces twins is precisely the process that produces divergence: the same genetic origin, released into distinct contexts, reorganises itself differently. This is the pattern that will generalise to all reproduction. The perfect genetic copy does not produce a copy of the organism because the organism is not a linear reading of the genome. The organism is the continuous integration of the genome with the field of variations in which it forms. What contemporary biology has discovered is that no replication produces identity because no replication occurs in a vacuum. All replication occurs within a specific material context, and that context is part of the reproductive operation.
The biological mechanism of this variation is precisely quantifiable. Every DNA replication introduces opportunities for error. The human mutation rate stands at approximately ten per ten billion base pairs per generation, that is, for every cell division, in approximately one out of a billion nucleotide positions, a spontaneous substitution, insertion, or deletion occurs. The human genome contains approximately three billion base pairs. By direct calculation: each human individual is the carrier of between one hundred and two hundred de novo mutations, genetic variations present in them that did not exist in any ancestor. Neither of these individuals is, therefore, genetically identical to either progenitor. Each child is a new generation of alterations. But variation does not end with point mutations. Meiotic recombination, that process through which a germ cell duplicates its genome and then divides it to form two gametes, each with half of the original material, produces, through the crossing-over of homologous chromosomes, a redistribution of genetic content. Each gamete receives a unique combination of chromosomal segments that has never existed before, fragments of paternal and maternal DNA recombined in a new pattern. Independent segregation of chromosomes adds to this: during meiosis, twenty-three chromosomal pairs separate randomly; each gamete receives one of the two versions of each pair. This produces two raised to the twenty-third power, more than eight million, different chromosomal combinations that a single individual can generate. Given that the reproductive partner contributes an equally unforeseen combination, the probability of two distinct individuals (excepting monozygotic twins) inheriting exactly the same genetic combination is astronomically miniscule. Each descendant is numerically distinct. This is not an accident of reproduction; it is its fundamental operation.
The variation that emerges in monozygotic twins during development is merely the clearest form of a general principle: no reproduction is a copy. There is an even more fundamental aspect. The error rate of DNA replication, one in every ten billion nucleotides, means that the genome is not a stable structure upon which life rests. It is a structure continuously perturbed by spontaneous error. Every dividing cell introduces some mutations. Every growing tissue accumulates genetic variants that neighbouring cells do not share. A developing brain is not genomic; it is a mosaic of genetic variants. This is not a weakness of replication; it is an essential property of it. The possibility of error is what permits variation, and variation is what permits response to contingency. If replication were perfect, no variation would emerge, no adaptation would be possible. What appears to be a defect from the standpoint of engineering (copy error) is, from the standpoint of life, a mechanism of innovation. Evolution does not operate despite the imperfection of the copy; it operates through it.
The conceptual error that dissolves here is precisely that of confusing the transmission of material with the reproduction of identity. Reproduction as a biological phenomenon is not the conservation of form through time; it is a response to irretrievable cessation. The preceding analysis established that every organism is a "remainder", a form that emerges from the dynamism of material differences, maintains itself while conditions hold, and transmutes when they reconfigure. No living form persists indefinitely; death is integrated into the structure of the living. Reproduction is not a denial of this finitude, it is its only possible logic. How can a type of continuity persist when each individual ceases? The answer is not through recourse to identity: seeking the perpetuation of the individual organism is asking the impossible. The answer nature implements is the transfer of possibilities. Hereditary material, genome, but also epigenetic structures, regulatory molecules, maternal influences derived from membranes and fluids, is not a program that a new organism must replicate; it is a set of constraints and resources through which a new individual reconstructs itself. The analysis of the origin of the universe established that form is an event, not a substance. Each organism is an irreversible event. But events, under a regime of universal death, can link together. The prior organism confers no guarantee of identity upon the posterior organism. It offers only the possibility of being new; it offers the constraints through which that novelty will take form. This is the architecture of life in contingency: not the repetition of one, but the systematic institution of another.
2. Genealogy of Transfer
The modern understanding of reproduction does not arise through simple observation of offspring. It emerges through a radical inversion of what seemed self-evident: that heredity operated by the mechanism of visual resemblance. Two pivotal philosophical figures structure the contemporary genealogy of reproduction as transfer. Both share a fundamental diagnosis: the material that is transmitted is not visible form, but something categorically distinct. Both refuse the metaphor of "inheritance as identity legacy" and replace it with an analysis of what actually persists through time.
August Weismann, at the end of the nineteenth century, proposed a demarcation that would transfigure all of biology: the distinction between germline and somatic body. The body, that which we see, which ages, accumulates damage, and dies, is merely a temporary structure. The germline, those cells that give rise to gametes, is not simply part of the body; it is a material continuum that persists through many successive bodies. Each body is a temporary vehicle of a lineage that precedes it and will succeed it. Transmission across generations is not the transmission of bodily form; it is the passage of biological material capable of instantiating new bodily organisation. Weismann concluded, and experimental observation confirmed, that characteristics acquired during an organism's life, modifications to the somatic body caused by use or disuse, are not inherited. The body that modifies itself leaves no mark on the germline. This means that heredity does not operate by simple copying of observable form. That which it transmits is, necessarily, something different from the form we see. Weismann dissolved the illusion that "resemblance" explained transmission. What is transmitted is material, but material under a different regime from the body that ages. The ontological consequence of this separation is radical: the body is a temporary effect; what persists is the genetic lineage, passing through multiple bodies like a thread through beads. No body recovers the prior form; each is a new reconstruction from hereditary material that carries no memory of the prior form.
Richard Dawkins reinterpreted this insight through the concept of the "replicator". His analysis was twofold. First, he identified the gene as the fundamental unit of replication: a sequence of DNA capable of making copies of itself, persevering through generations of organisms that host it. Second, and radically: the organism is not the unit of natural selection; it is a "vehicle" for replicators. This perspective inverts the common causal order: organisms do not reproduce to ensure their own survival; organisms are machines through which genes replicate. What we have called the "reproductive success" of an individual is, in this reading, a measure of the efficacy of its hosting of genes in producing copies of those genes in the next generation. The gain of this perspective was immediate: it managed to explain phenomena that seemed contradictory to evolution by natural selection, altruistic behaviours among relatives, sterility of certain individuals in social colonies, asymmetries between paternal and maternal investment. If natural selection acts upon replicators, then an individual that foregoes its own reproduction to increase the reproductive capacity of close relatives (carriers of copies of its genes) acts in a way that, in terms of replicator propagation, is highly efficient. But the concept of the replicator carries the risk of reification. The gene as replicator can be understood as possessing intention, "the gene wants to make copies of itself", "the gene struggles for survival". This anthropomorphisation is a categorical error. A gene is a sequence of DNA; it possesses no intention, executes no plans, "wants" nothing. What actually happens is that DNA molecules with a certain sequential structure possess chemical stability in an intracellular context; this stability allows them to be copied; copying allows them to persist in new generations; persistence is a consequence of physical properties, not intentionality. Natural selection is not a cosmic force that chooses; it is a differential process of elimination. Genes that produce non-viable or sterile organisms do not propagate; genes that produce reproductively successful organisms continue to appear. Asymmetry emerges not from nature's favour, but from material constraint: only those genes capable of supporting replication in viable cellular contexts persist as lineages. The correction is essential: the replicator is a material constraint, not an agent. Selection is differential elimination, not intelligent choice.
James Griesemer subsequently proposed that reproduction was not the transmission of abstract information, a genetic code passed from generation to generation as if it were an intact blueprint. Reproduction was material overlap. The daughter cell does not receive a "copy" of the mother cell's genome; it receives the genome through a direct material mechanism: the chromosome is duplicated, and the two copies are physically separated between two daughter cells. What persists is material continuity, chemical bonds linking nucleotides in sequence, which are never completely broken, existing in both daughter cells as parts of a continuous material structure. This means that heredity is not the transmission of a message (which could, in principle, be expressed in any medium, ink, paper, radio waves). Heredity is participation in a material lineage. The genome does not exist as a blueprint; it exists as a chemical molecule reacting with other chemical molecules in the specific context of the cell hosting it. The implication is that no copy is ever perfect in the informatics sense: each "copy" of the genome is the same physicochemical process, the same molecular structure in a new cellular context. Reproduction maintains continuity of material relationship, not identity of form.
Griesemer's critical contribution is to demonstrate that the concept of "genetic information" as something independent of its materialisation is a conceptual fiction. Information is always embodied in a substrate. The DNA that exists in a progenitor's germ cell is not the exact same DNA that will come to exist in the daughter cell, in the sense that every chemical molecule has a history, location, and context. But material continuity is real: it is the same polymerisation of nucleotides, the same helical structure, the same base of replication. Reproduction does not copy in the sense of "making a photocopy": reproduction is a bifurcation of material continuity. Two daughter cells do not possess genomes that are "copies" in the sense of having been described and then redrawn. They possess genomes that are a continuation of the progenitor genome through a material act of duplication and separation. This dissolves the paradox of the perfect copy: there is no copy because there is no act of copying. There is only the material continuation of one structure into two related structures. The concept of replication that emerges from this analysis is entirely materialist: genes replicate not because they "decide" to make copies, but because the chemical structure of DNA, in a dividing cell, with enzymes present, with nucleotides available, in an appropriate environment, inevitably produces a new molecule structurally related to the first. Replication is an emergent property of physical dynamics, not the intention of an agent.
Susan Oyama, through "Developmental Systems Theory" (DST), deepened this critique of "genetic inheritance" as a concept. Oyama argued that the word "inheritance" creates the illusion that something is transmitted intact across generations, genetic information, as if it were a parcel sent by post. But during development, genetic material does not execute a "program"; it interacts dynamically with its cellular, hormonal, maternal, and environmental context. The same genome, in different contexts, produces radically different phenotypes. A caterpillar transforms into a butterfly using the same genome, mediated by hormonal signalling intrinsically dependent on the developmental context. A plant grows to different heights depending on light availability; a fruit fly develops wings of different sizes depending on incubation temperature. The same genome is not executed uniformly; it is read in diverse ways by the developing cell, a reading that depends integrally on external signals. That which is "inherited" is not, therefore, a coded program; it is a set of resources that the developing organism uses to reconstruct itself. Heredity does not guarantee identity; it offers materials through which identity will be constructed anew, differently.
Eva Jablonka and Marion Lamb expanded this understanding through a diagnosis of "four dimensions of inheritance". First: genetic inheritance, the transfer of DNA sequences, as previously described, through mutation and recombination. Second: epigenetic inheritance, chemical modifications to DNA (cytosine methylation) and the histone proteins surrounding it (acetylation, phosphorylation), which alter the accessibility of the DNA sequence for transcription without altering the sequence itself. These epigenetic modifications are, to a degree, inheritable through meiosis and development. An experience of deprivation in a progenitor can leave epigenetic marks that persist in offspring, the genetic sequence remains identical, but the way it is read changes. This is the register where the somatic history of the prior organism transfers consequences to the posterior organism without passing through the genetic code. Epigenetic plasticity reveals that inheritance includes the space where the genome is silenced or amplified, where reading is determined not by sequence but by chemical markers accumulated through experience. Third: behavioural inheritance, learning, transmission of knowledge, imitation of techniques among individuals within and across generations. A bird learns a song by listening to parents; that learning is not coded in any gene; it is behaviour transmitted culturally. A primate learns a food acquisition technique by observing its mother; that technique becomes a group pattern, transmitted through demonstration and imitation. Behavioural inheritance is, in a sense, more flexible and innovative than genetic inheritance: it allows the accumulation of knowledge that is not fixed in code, which can vary rapidly as the environment changes. Fourth: in humans only, symbolic inheritance, language, conventions, narratives, cultural meanings. Through symbols, humans transmit not merely techniques or patterns, but interpretations of the world, value systems, conceptual structures that organise experience before any direct contact with phenomena. Symbolic inheritance operates in a completely distinct register from the biological: it is not limited by somatic compatibility, can be transmitted between non-related organisms, and can be revised rapidly without the need for genetic evolution. The instability that transfers across generations is not, therefore, reducible to the genome. It is multidimensional. An individual inherits not only genes, but epigenetic patterns, learned behavioural capacities, and symbolic structures that define world perception. Each of these dimensions introduces variation that is not predetermined by the prior generation. Each of them is a site where difference breaks out systematically. What is transferred is, therefore, a multifaceted capacity to be different, not a blueprint of identity, but a heterogeneous set of constraints, resources, and potentialities that each new organism receives and singularises. The transfer of instability is the transfer of this multiplicity of registers in which the new organism will have to negotiate identity not in genetic isolation, but in continuous dialogue with surrounding epigenetic, behavioural, and symbolic contexts.
3. Mechanisms of Reproductive Instability
The transfer of instability operates through multiple mechanisms, each capable of introducing variation in an offspring generation without that variation being reversible or correctable by natural selection, because any correction would itself be eliminated by selection in the next generation. De novo mutation is the most fundamental. DNA replication is not a perfect process; each time the double helix is unwound and recopied by the enzyme DNA polymerase, small errors occur. The error rate of polymerase is extraordinarily low, approximately one in every ten billion nucleotides, but opportunity is vast. A dividing human cell copies three billion nucleotides; statistically, that division will introduce a mutation. Across many divisions, producing an entire organism, mutations accumulate. The result: every living human individual carries between one hundred and two hundred mutations that did not exist in any progenitor. These uninherited, de novo mutations are introduced during gametogenesis, the production of spermatozoa or ova, or during the earliest stages of embryonic development. Once present, they become hereditary; any offspring will carry that mutation if the progenitor transmits it. Most of these mutations are neutral, altering no protein function because they occur in a non-coding region or a redundant codon position. Some mutations are deleterious, impairing necessary protein function, reducing viability or reproductive capacity of the bearer. Rarely, a mutation is beneficial, adaptively improving protein function in a specific environment. The frequency of these categories is uncontrollable; variation is introduced blindly, without prior knowledge of future reproductive context.
Meiotic recombination operates through an equally non-teleological mechanism. During meiosis, each diploid cell (with two chromosome sets) undergoes replication and two divisions, producing four haploid cells (with one chromosome set). The first meiotic division is where crossing-over occurs: homologous chromosomes, one inherited from the maternal progenitor, the other from the paternal progenitor, pair up closely and exchange segments. A fragment of the paternal chromosome recombines with a fragment of the maternal chromosome, creating a hybrid chromosome that has never existed before. Statistically, each chromosome pair undergoes between one and three recombination events, depending on its size. The exact position of each event is random. Then, in the second meiotic division, the already replicated chromosomes separate again. The order of this separation is also random: there is no known mechanism that determines which of the two copies of each chromosome proceeds to each daughter cell. The result is that each gamete receives a unique combination of chromosomes. An individual with twenty-three chromosome pairs can generate approximately eight million different combinations. When two individuals reproduce, each contributes one of these combinations. The probability of two distinct gametes being genetically identical approaches zero. Each descendant is, therefore, a genetically new combination.
Phenotypic plasticity introduces an additional source of variation that is not genetic in origin, but is phenotypically manifest. The same genotype, in different environmental contexts, can produce radically different phenotypes. A fruit fly developed at twenty degrees Celsius produces short wings; the same species, same genotype, developed at thirty degrees produces long wings. A plant of the same genetic variety cultivated in nutrient-scarce soil remains dwarf; replanted in rich soil, it grows to normal height. A population of bacteria, genetically uniform, divided between a sugar-rich environment and a poor environment, differentiates phenotypically, bacteria in the rich environment store reserves and reproduce rapidly; those in the poor environment enter reduced-energy metabolism and grow slowly. The norm of reaction, the range of phenotypes that a genotype can produce across a range of environments, is a constitutive property of development. No genotype specifies a single phenotype; each genotype specifies a contingent relationship between environment and developmental outcome. A genotype may be coded identically in a hundred different organisms; if those hundred organisms experience different environments during development, they will exhibit a hundred different phenotypes. Phenotypic plasticity is the mechanism through which the environment channels development; it is not a developmental error, but a universal feature.
Epigenesis, covalent modifications of DNA and surrounding proteins, represents a fourth source of variation. The DNA molecule wraps around the protein called histone; eight histones form an octamer around which the DNA helix wraps approximately one hundred and forty-seven times. This wrapping conditions which DNA segment is accessible for transcription (conversion into RNA) and which is packaged and protected from reading. Enzymes can add chemical groups to histones, acetyl, methyl, phosphoryl, altering the surface charge of the protein, loosening the wrapping, allowing access to genes. Other enzymes can add a methyl group directly to cytosines in the DNA sequence; this does not change the sequence, but marks it as "silenced". The same gene, with identical sequence, can be acetylated in one tissue (and thus active) and methylated in another tissue (and thus silenced). Modification is responsive to the environment, nutritional stress induces epigenetic patterns distinct from a rich alimentary environment; thermal variations induce epigenetic remodelling; social signals (in bees, the quality of larval food) induce epigenetic differentiation between queens and workers from identical genomes. These epigenetic modifications are, to a certain extent, inheritable. An organism undergoing epigenetic modification during development can transmit some of these epigenetic marks to its offspring, not through an alteration in DNA sequence, but through the maintenance of acetylation or methylation patterns during replication. Epigenetic inheritability is imperviously complex and incompletely understood, because many epigenetic marks are erased during meiosis and gametogenesis, while others are reestablished de novo during offspring development. But the essential point is: even where epigenetic inheritability exists, it is not identical transfer of the parental pattern. It is the transfer of material that will undergo remodelling in the new organism.
4. Implication: Continuity of Difference
Life persists not despite systematic difference, but by means of it. Every organism that is born is genetically new, phenotypically contingent, and epigenetically distinct from its progenitors. No generation repeats the previous one, even if we were to imagine identical environmental conditions. Each individual is unrepeatably particular. The error lies in confusing continuity with identity. Heraclitus offered the decisive image: the river remains, but the waters change. Life realises this logic on a generational scale. There is no identity of form, because each organism ceases; there is continuity of operation, because the death of one organism does not extinguish the process it instantiated, but transfers it to another. The descendant does not inherit life as a property; it instantiates anew the same operation of adaptation, reconstruction, and response to constraints under numerically distinct conditions. What is transferred is not immortality, but renewed finitude.
Cosmological analysis established that form is an event. Form is not an enduring substance; it is a transient configuration emerging from the dynamism of material differences. When the conditions sustaining a form alter, the form transmutes or ceases. No form is indefinitely stable; stability is always provisional, metastable, doomed. The individual organism is a form of this type, an irreversible event. The analysis of the organisation of matter established that this contingency is not a defect of material organisation; it is its constitutive regime. The living is a series of non-repeatable events maintaining an organisational pattern. Each heartbeat occurs once; future beats reiterate the operation of circulation, but never reiterate the unique event of the prior beat. Each cellular reproduction follows a conserved mechanism (mitosis), but each result is the generation of new cells. Contingency transforms into an irreversible series: there is no return to a prior state, no complete determination of a future state, no essence that would remain if all form ceased. But there is continuity of pattern. Organisms remain alive because they continuously reconstruct themselves. When an organism ceases to reconstruct itself, when the fundamental operation paralyses, death occurs. Reproduction perpetuates this operation: not through the repetition of the prior operator, but through the institution of a new operator. The new operator is not a copy of the prior one; it is a new realiser of the same operation in numerically distinct materiality. This is no guarantee of continuity, it is a fragile structure, permanently threatened by failure. A mutation can deactivate an essential gene; a developmental accident can eliminate an organism before viability; a sudden environmental change can exceed the phenotypic plasticity of a population. But while the operation continues, while each organism reconstructs itself, transfers material to the descendant, and the descendant reconstructs itself differently, continuity subsists.
This structure exposes a question that cannot yet be resolved in this ontological register. If the prior organism does not determine the posterior, but merely transmits to it a field of possibilities, what is the basis of responsibility for the effect produced in the descendant? The question is no longer strictly biological; it shifts to the plane of obligation. What this chapter establishes is merely the structure: the transfer of instability offers the possibility of relationship, but does not determine it. It is in that gap between possibility and determination that the ethical question erupts.
Life does not copy itself, it transfers itself as difference; each descendant is new, not a repetition.
5. The Myth of the Genetic Program
When the Human Genome Project was announced as complete in 2003, the public narrative was one of determinist triumph. It seemed that the “book of life” had finally been deciphered and that the genetic sequence would come to explain, almost by itself, the form of bodies, the origin of diseases, and the architecture of behaviour. The dominant vocabulary was computational: the genome appeared as software, the cell as an execution machine, development as a program.
The actual outcome was far less comfortable for this metaphor. The human genome revealed roughly twenty thousand genes, a modest number relative to the expectations of the time and comparable to that of far less complex organisms, such as Caenorhabditis elegans. Furthermore, the simple relationship between gene and trait proved unsustainable for the vast majority of relevant phenotypes. Height, psychiatric susceptibility, cognition, or immune response depend on thousands of variants, non-linear interactions, and environmental contexts without which nothing is updated. What biology found was not a closed script, but a network of constraints whose outcome always depends on context.
This is why the metaphor of the genetic program fails at the decisive point. A program, in the strong sense, specifies an outcome unequivocally when executed under prescribed conditions. The genome does not do this. Monozygotic twins, despite sharing the same genetic sequence, diverge in fingerprints, pathological vulnerabilities, and measurable phenotypic traits. If the same “instruction” does not produce the same outcome, then we are not dealing with a program, but with biological material whose efficacy depends on regulation, development, environment, and chance.
Persisting in talking about “genes for” intelligence, depression, or aggressiveness is not merely a linguistic shortcut; it is a conceptual error with public consequences. The program metaphor obscures phenotypic plasticity, environmental co-causality, and developmental openness. For the core question of this chapter, this is decisive: inheritance does not eliminate difference through a sovereign code. It transfers resources, sensitivities, and limits within which difference emerges once again.
6. Genealogy of Contingent Inheritance
The historical genealogy of the "genetic program" metaphor is not the narrative of an objective error that the scientific community committed collectively in the past and subsequently corrected through the accumulation of knowledge. Rather, it is the story of a radical historical contingency, the statistically improbable confluence of three heterogenous disciplinary lines that casually met and intersected in the nineteen-fifties and sixties, producing a synthesis that seemed inevitable, almost natural, as if deducible from first principles. Evelyn Fox Keller, historian of biology and epistemologist, provides a precise, demystifying, and deeply political genealogy of this contingent confluence. The metaphor of the "genetic program" results from a specific encounter among four distinct intellectual traditions: first, early twentieth-century Mendelian genetics (inheritance in discrete, measurable proportions, genes as units of hereditary transmission); second, Claude Shannon's mathematical theory of communication (the central idea that a message is transmissible across different material media and reducible to abstract symbols in an invariant manner); third, nascent molecular biology (Watson, Crick, the structure of DNA, the chemical explanation of replication); fourth, theoretical computer science (Turing and abstract machines with discrete states, von Neumann and cellular automata with logical instructions, the formal concept of a program as a sequence of executable logical operations). In the 1950s, the genetic code was deciphered by the international molecular community, discovering that each triplet of nucleotide bases (each codon) specifies a particular amino acid in the final protein. From that discovery, DNA was universally recognised as an "information repository", the informational metaphor completely conquered scientific discourse. Digital computers of first, second, and third generations were constructed during this period and began to be recognised not merely as calculators, but as logical machines capable of executing arbitrary programs inscribed in symbolic language. The conceptual synthesis that followed was almost automatic, almost inevitable in the cultural circumstances, almost deducible as a logical conclusion: if DNA stores "information" coded in chemical language (based on four symbols, the four bases), if that information is transmissible across different material media without loss (as Shannon demonstrated for abstract information), if computers execute programs inscribed in symbolic language and produce outputs completely specified by the programs, then why not think of the cell as a biological computer, DNA as a program inscribed in biochemistry, protein as computational output, and the whole organism as a sophisticated biological machine executing inherited genetic software? The elegance of the synthesis, perfect logical coherence, apparent explanatory power, mathematical elegance, completely obscured its deeply contingent and constructed nature. It did not have to be so. It was not the result of logical deduction from necessary and universal axioms. There could have been an alternative intellectual tradition in which heredity was conceptualised not as abstract, transmissible information, but as dynamic, responsive material, not as statically inscribed code, but as a living process in continuous development, not as a rigid program to execute, but as an open resource for contextual reconstruction. But there was not, the dominant metaphors that captured the scientific imagination, those that established productive research domains, were the computational metaphors of information and program. Keller's interpretative and political gain is not to reject the real underlying biological mechanisms (DNA replication, transcription of DNA into RNA, translation of RNA into protein, all of which function empirically, are universally documented in all living cells, are incorporated into all contemporary functional molecular biology, and are absolutely indispensable for understanding). It is to recognise with epistemological precision that the metaphor of the "genetic program" is exactly that, a metaphor, not a transparent, direct description of reality, not an unmediated capture of the being of things as they appear in themselves, and that metaphors reveal while simultaneously obscuring, guiding scientific investigation while misleading researchers when reified. Keller's work opens the possibility of decoupling real mechanisms (which continue to be studied, remain true, and remain indispensable) from the interpretative metaphor (which is contingent, historically produced, and revisable). We can retain the mechanisms and reject the metaphor. In fact, it is imperative to do so if we wish to understand reproduction and inheritance with philosophical clarity.
Richard Lewontin provides a crucial and empirically grounded deconstructive complement: his systematic thesis of the "triple helix" of gene, organism, and environment in continuous co-construction. The central and deeply radical thesis is that biological causality is never unidirectional, never a one-way flow from inherited gene to developing organism, but always multidirectional and reciprocal, forming circular causal webs where each element affects and is affected by all others. The gene influences the organism because it encodes structural, enzymatic, and regulatory proteins; but the organism simultaneously influences the gene because genetic expression depends integrally on the specific cellular context, the epigenetic state in which the chromosome resides, the molecular signals arriving from the surrounding cytoplasm, and the local density of transcription factors. The environment influences the organism because nutrition, temperature, radiation, and environmental stress systematically alter the developing phenotype; but the environment also influences gene expression because chemical mutagens cause mutations in DNA, because certain environments activate or silence genes via epigenetic mechanisms, and because chromatin structure modifies in response to environmental signals. The organism influences the environment by growing within it, transforming it through its presence and active metabolism, selectively extracting resources from it, and modifying microhabitats. None of the three, gene, organism, environment, unilaterally determines the final outcome in isolation. All participate simultaneously in a complex causal web where the direction of effect is not prescribed a priori, not determined by a logic prior to real operation. Fundamental and inescapable consequence: there is no clean or discrete separation between "the genetic" and "the environmental", this seemingly natural dichotomy is a historically produced conceptual artefact, not a reflection of real causal structure found in nature. Heredity is not the pure transmission of isolated genetic material, it is always co-constructed dynamically between genetic material inherited from the progenitor and the epigenetic, cellular, and environmental context of expression and development of that genetic material. Genetic material is a necessary, ineliminable condition; context is an equally necessary condition. Neither is sufficient in isolation to produce an outcome.
Paul Griffiths and Russell Stotz contribute an even more radical and philosophically deconstructive perspective: the gene is not a natural ontological entity with a fixed essence and discernible boundaries in nature as molecules and cells are. It is a functional concept that is radically contextual in every particular application. What counts as a "gene", what demarcates the conceptual boundaries of the term, how it is particularised in a specific organism, depends entirely on the conceptual, experimental, and interrogative context in which one is working. In a classic twentieth-century Drosophila genetics experiment, the white allele is operationally defined as the "gene for eye colour" because mutations at the white locus cause a discrete, measurable change in phenotype: white eyes instead of wild-type red. In contemporary genetic engineering with CRISPR editing tools, the same DNA sequence is the "gene to edit", a physical, discrete unit, a sequence of nucleotides to be cleaved and modified. In genome-wide association studies (GWAS), a "gene" becomes a statistical locus correlated with phenotypic variance across a population. The concept of the gene is not a static object discovered in nature; it is an operational instrument constructed by biological inquiry to isolate specific relations within a complex matrix.
7. Epigenetics and the Dissolution of the Code
Inheritable epigenetics offers an additional complication and fundamental sophistication: alterations in gene expression, chemical methylation of DNA, modifications of histone proteins around which DNA wraps, three-dimensional chromatin conformations, can be partially inherited for one or two generations, sometimes more. This is categorically not "inheritance of acquired traits" in the classic Lamarckian sense that twentieth-century biology rightly rejected. The sequence of DNA is not altered, there is no change in the genetic information encoded in the bases. But the epigenetic state, the chemical information superimposed upon DNA that determines how genes are accessible to transcription, how they are transcribed into RNA, and how they are expressed as protein, can be maintained through mitotic cell division and across a generation or two (meiosis, fertilisation, development). A paradigmatic example: exposure to a specific pattern of nutrition during intrauterine development can alter methylation patterns in critical genomic regions regulating metabolism and stress response; in the offspring of that individual, when exposed to a similar nutritional pattern or when facing stress, some methylation patterns can be partially maintained, producing a phenotype similar to the progenitor's despite the genetic sequence being absolutely identical atom for atom. But these epigenetic states are contingent, reversible, and responsive to change, they can revert if environmental pressure changes completely. If nutrition shifts to a drastically different pattern, if the environment becomes completely different, the methylation pattern that had been maintained can disappear entirely and be replaced by a new pattern. Inheritable epigenetics is the inheritance of a contingent and reversible functional state, not the inheritance of a permanent and chemically irreversible genetic change inscribed in DNA. It offers yet another empirically well-documented dimension of how inheritance is an open construction, permanently dependent on present context, not the rigid, deterministic execution of a pre-coded program.
8. Inheritance as Resource
If we abandon the program metaphor, inheritance changes status. It is no longer a closed instruction that the descendant executes; it is a set of resources with which the descendant constructs itself. The genome is neither an integral blueprint nor a deterministic script. It is a material constraint: it opens certain possibilities and closes others, but it does not unequivocally fix the final outcome.
The distinction between constraint and determination is central. To determine means to impose a single result for the same input conditions. To constrain means to limit the space of possible outcomes. The same genome, exposed to different environments, distinct developmental times, and different cellular accidents, produces different phenotypes. This suffices to show that the genome, in isolation, does not specify the organism. The outcome emerges from the interaction among genetic sequence, inherited cytoplasm, epigenetic regulation, prenatal environment, postnatal environment, and the contingent history of development.
This is why expressions like “gene for height” or “gene for depression” are merely crude shorthand. The most rigorous formulation is: there are genetic variants that contribute to certain traits in certain population and environmental contexts. Inheritance transmits heterogeneous material with which one works in the present. The descendant does not execute a code; it reconstructs itself with partially inherited resources under conditions that never exactly repeat those of the progenitor.
In this perspective, to inherit is not to receive a destiny, but to enter a conditioned field of possibilities. The living constructs itself in interaction with what it receives. And it is precisely for this reason that sex becomes philosophically decisive: if inheritance no longer determines the outcome, sexual reproduction will show, to an extreme degree, how continuity itself depends on the dissolution of genetic identity.
To inherit is not to receive a destiny, it is to receive material with which to work; the result is always a construction.
9. The Dissolution of Genetic Identity
Sexual reproduction constitutes a mechanical operation of dissolving individual genetic identity. The phenomenon is radically distinct from any merely somatic or hereditary continuity: it is not the transmission of a pattern that maintains itself across generations, but the systematic dismantling of that pattern in every reproductive act. Meiosis, the first phase of this process, reduces the chromosomal content of the progenitor cell from diploid ($2n$) to haploid ($n$). Each chromosome of the parental individual is replicated, pairs with its homologue, and thereafter enters a regime of physical exchange. Crossing-over, the exchange of chromosomal segments between homologous chromosomes, is not a metaphor for transformation: it is a material act of dissociation. DNA fragments that operated in the progenitor as an integrated totality are physically separated, regrouped into unique combinations, and distributed to distinct gametes. No germ cell resulting from this process carries the genetic identity of the progenitor as a totality. Each gamete is already an unprecedented singularity, a combination that the progenitor's body never experienced in its present form.
The mechanism of meiosis constitutes a two-stage descent, each responsible for an irreversible type of fragmentation. In meiosis I, two homologous chromosomes, one of paternal origin, the other of maternal origin in the progenitor, pair up in a structure called a bivalent. Crossing-over engages precisely here: segments of DNA are literally severed in both chromosomes, and the exchanged segments are reinscribed into the opposite sequence. The result is that each chromosome of the bivalent now carries a hybrid configuration of paternal and maternal material. When chromosomes segregate into daughter cells, each cell receives chromosomes that are mosaics, neither pure paternal nor pure maternal in origin, but heterogenous compositions that the progenitor individual never contained as a cohesive form. Independent chromosomal segregation executes in this context: each pair of homologues segregates randomly. In human beings with 23 pairs of chromosomes, this means $2^{23}$ possible combinations, approximately 8 million variations. No progenitor body is capable of generating, through sexual reproduction, two germ cells that are identical in chromosomal combination. During meiosis II, the operation replicates on a reduced scale: each duplicated chromosome, consisting of two sister chromatids originally identical, will segregate randomly into distinct daughter cells. Sister chromatids may reside in separate gametes, or both in the same gamete. The combinatorics of this second phase further multiply real possibilities. The result is that each gamete represents a singular point in the potential combinatorial space of the progenitor, a point that no other gamete of the progenitor will ever repeat, whether by statistical chance or biological intention.
Fertilisation does not resolve this combinatorial indeterminacy; it amplifies it exponentially. When a paternal gamete (one of 8 million paternal possibilities) fuses with a maternal gamete (one of 8 million maternal possibilities), the product is a singularity on a combinatorial scale that transcends not merely parental form, but the domain of the biologically repeatable or predictable. The mathematical probability of two genetically identical offspring from the same parents is approximately 1 in 70 trillion ($70 \times 10^{12}$). Genetic identity is, therefore, not merely rare, for practical purposes of biological continuity, it is unrepeatable. Each fertilising act produces a singularity that cannot be regenerated from the same parental pair. The contribution of each progenitor to the descendant is formally 50% of genetic mass per progenitor. But this proportion masks what is truly at stake: it is not a quantity that persists, but the fragmentation of totality. The progenitor contributes isolated segments, not identity as an integral form or gestalt. In the $\text{F}_2$ generation (grandchildren), each grandmother contributes, on average, 25% of genetic material. In the $\text{F}_3$ generation, each great-grandmother contributes 12.5%. In the $\text{F}_4$ generation, each great-great-grandmother contributes 6.25%. In the $\text{F}_5$ generation, each ancestor from five generations back contributes, on average, 3.125%. In the $\text{F}_6$ generation, each ancestor contributes 1.5625%. This geometric dilution is a process of irreversible dispersion: no individual ancestor is biologically recognisable as a totality beyond two or three generations. After five generations, any ancestor contributes, on average, less than 3% of genetic material to descendants. After ten generations, the contribution is less than 0.1%. After twenty generations, less than two centuries in organisms with short reproductive cycles, the contribution is negligible. The genetic sequence may circulate through many organisms across many generations; the genetic identity of the singular individual, as the cohesive totality it was, does not persist in any of them. What "persists" is statistical, not form: the probability that a specific allele of a progenitor circulates through descendants, impossible to trace as the continuity of a person.
Somatic death is the death of the body as a function, the cessation of metabolism, circulation, and energy processing. The death of genetic identity, however, occurs in every sexual reproductive act and constitutes a distinct regime of finitude: it is not the cessation of cellular life, but the dissolution of the reproducible unit as a biological totality. The individual that reproduces sexually cannot, biologically, generate a copy of itself as a cohesive identity. It can only dismantle its own singularity and participate, as a disintegrated fragment, in the construction of another radically new singularity. "Deferred death", a term occasionally used in evolutionary biology, inadequately describes this complex phenomenon. It is not a matter of postponed somatic death, of bodily death to come in the future. It is a matter of identity death incorporated into the act of transmission, a death that occurs while the body lives. The progenitor is alive while the descendant is generated; but its identity as a reproducible totality is biologically dead at the moment generation occurs. It is death-in-life: identity dies; the body continues alive as a physical system. This is the fundamental structure of finitude incorporated into the mechanism of continuity: not a death that follows life, but a death-in-life that is a necessary condition for another life to exist as a radically new and singular form. Sexual reproduction institutionalises self-forgetting in the act of transmission. What persists is not form; it is the fragment circulating, irretrievable as a totality. Otherness is not a contingency that befalls reproduction, it is the internal structure of the reproductive act, mechanically guaranteed in every operation of chromosomal segregation and gametic fusion. Sexual reproduction is not, therefore, a mechanism for perpetuating identity; it is a mechanism for the systematic production of difference, an operation that dissolves the same to instantiate the other.
10. Genealogy of Sex as a Problem
The problem of sex emerges in the mid-twentieth century as a central question of evolutionary biology: sex is costly, and natural selection should favour asexual reproduction consistently and universally. John Maynard Smith's argument, formulated in the 1970s in systematic and rigorous work, structures the question with unparalleled precision. An asexual female transmits 100% of her genes to each offspring produced. A sexual female transmits only 50%, since half of the offspring's genome comes from the genetically distinct male. If an asexual population competes with a sexual population in the same ecosystem for resources, the asexual population doubles its numbers in every generation (each asexual female produces clones that are equally fertile). The sexual population, cut in half, reproduces only at the rate of its females, that is, at half the reproductive capacity of an equivalent asexual population. The mathematical advantage of asexuality is, therefore, twofold: (1) each progenitor transmits a complete genome, unfragmented by meiosis; (2) there is no investment in males, considered biologically "costly" from a purely reproductive perspective, since half of the reproductive population does not produce direct offspring. This is the so-called "two-fold cost of sex". Empirically, based on simple mathematical calculation, asexuality should have swept sexual populations out of evolutionary history tens or hundreds of millions of years ago. But it did not. Sexual organisms dominate complex life forms. The question Maynard Smith raises is tautologically simple and yet empirically refractory: why has natural selection not eliminated sex? The cost is measurable in clear numbers; the advantage is invisible to simple mathematical calculation.
George C. Williams, contemporaneously, proposed an answer that repositioned sex as an evolutionary response to variable environments, not as a solution to a present, immediate problem, but as an investment in possible futures that reduces long-term risk. In a stable, predictable environment, asexuality (which guarantees constant genetic identity) is superiorly adapted in the short term. Genetic consistency allows each clone to inherit not only genes but also the maternal body's fit to the existing niche. In a variable environment, where conditions oscillate unpredictably between different states, sexuality (which produces continuous variation) offers biological insurance: some offspring will possess genotypes adapted to future conditions that the environment may assume. Sexuality is, in this light, a "genetic lottery", a betting strategy that distributes genes across multiple combinations, sacrificing short-term efficiency to protect long-term survival. When the environment changes, climate fluctuates, nutritional availability varies, predatory pressures rise or fall, some of these genetic combinations are likely to succeed where others fail. Asexuality, by contrast, is a strategy of "certainty": each offspring is identical to the mother, hence equally well or ill-adapted to the same conditions. Williams does not claim that environmental variability "explains" sex deterministically, but that it offers an ecological context in which sex ceases to be merely costly: it becomes an investment with a visible rate of return in an uncertain environment. The logical implication: organisms in stable environments should be asexual; organisms in variable environments should be sexual. But environmental variation alone does not explain the observed universality of sex in complex eukaryotes. Stable environments exist and persist; few eukaryotes in them are asexual.
W. D. Hamilton introduced a radically distinct perspective that aligned the problem with deep parasitological research, repositioning sex as a response not to internal environmental variability, but to continuous and cyclical external coevolutionary pressure. The Red Queen hypothesis, referencing Lewis Carroll and the idea that one must run as fast as possible just to stay in the same ecological place, structures sex as a mechanism of constant escape from parasitic adaptation that never ends. Parasites adapt to the genotypes of frequent hosts through mutation and directed selection. When a host population is asexual (all copies genetically identical), parasites rapidly evolve infectious strategies that conquer the entire population irreversibly. A single parasitic strategy can neutralise an entire clonal population. Sexuality, on the contrary, produces rare genotypes continuously. Each offspring is a genetic combination that does not exist at high frequency in the population. Parasites adapted to genotype A encounter rare genotypes B, C, D, and E just as genotype A becomes frequent enough to make coevolutionary parasitic investment worthwhile. Once parasites successfully evolve against A, that genotype becomes disadvantageous; rare genotypes gain a selective advantage simply by being unadapted to the existing parasitic arsenal. Coevolution is, thus, an endless arms race without final resolution or possible equilibration. What was once a genetic advantage (high frequency) becomes a disadvantage (a preferred target for the parasite). Every transition in frequency alters the selective landscape. Potamopyrgus antipodarum, a small freshwater snail that exists in both sexual and asexual forms within the same geographic region, offers a paradigmatic and measurable example: sexual populations display greater resistance to specialised local parasites; asexual populations (clones generated within the same rivers) suffer a demonstrably higher parasitic load. The divergence is reproducible across multiple populations and geographies. The Red Queen does not win; she runs to avoid losing her place in evolution.
Richard Michod returned to the question of cumulative genetic damage and offered a mechanism that none of the previous hypotheses adequately explained: the irreversible accumulation of error on a demographic scale across many generations. Muller's ratchet hypothesis, named after Hermann Joseph Muller, who proposed it in 1964, describes an inevitable dynamic in large asexual populations. In asexual populations, the class of individuals with the fewest deleterious mutations can be lost through genetic drift. Once lost, that class cannot be restored, because there is no recombination to recover it. The result is that, from generation to generation, the average mutational load increases. The process is irreversible: the ratchet turns in one direction only, accumulating damage with no internal possibility of return. This is the mathematical reason why an asexual genome cannot fully recover. Sex does not “repair” mutations, but allows recombination and redistribution of the load: an individual with mutations at loci A, B, and C can reproduce with another carrying mutations at loci D, E, and F, generating offspring with a lower concentration of damage in each genome. Sexuality, in this perspective, is less repair than redistribution. The genome imposes the limit; sex responds.
An adequate synthesis recognises that multiple mechanisms sustain the universal persistence of sex in eukaryotes and in most complex organisms. No single cause suffices to explain the observed universality. Environmental variability (Williams) offers a context in which sex ceases to be strictly costly. Parasitic coevolution (Hamilton) offers continuous, irresolvable selective pressure favouring rare genetic variation. Mutational accumulation (Michod) offers long-term demographic weight making recombination indispensable to prevent irreversible genetic degradation. Population dynamics, in which asexual numbers should grow exponentially but fail to do so in real ecosystems, suggest that in any concrete ecology, the combination of these factors renders sex adaptive despite its persistent cost. Sex is not an answer to a single, well-defined problem. It is a multifactorial solution to an ecosystem presenting multiple simultaneous pressures: variability, parasitism, accumulated genetic error, and competitive population dynamics. The universality of sex in complex life forms is due to this redundancy of causes, none can be eliminated without another taking its place in importance. No real ecology is simple enough for asexuality to be advantageously sustained over time. Sex persists because no ecosystem exists without combinations of all these pressures operating simultaneously.
11. Cost, Red Queen, Muller
Empirical documentation of the cost of sexual reproduction is unequivocal in organisms where experimentation is possible and verifiable across multiple controlled generations. In Drosophila melanogaster, the fruit fly that has served as a laboratory model organism for over a century, the asexual reproductive rate (when artificially induced or through genetic mutation) doubles the sexual rate in each generation under controlled conditions. An asexual female Drosophila that clones herself produces two genetically identical daughters in 10 days; in 30 days, four (daughters and grand-daughters exponentially); in 40 days, eight. A sexual female, by the mechanical nature of meiosis, produces only daughters that inherit 50% of the genome (and sons inheriting the remaining 50%, which reduces per-female reproduction). Mathematically, the asexual population should reach $2^n$ individuals while the sexual reaches only $n$, duplication versus linear growth cut by half. But when Drosophila is maintained in the laboratory for 100 generations in a controlled environment, without external parasitic pressure and without substantial environmental change, asexual lines (which could be artificially maintained through selection or genetic manipulation) consistently and documentedly collapse. Individuals accumulate deleterious mutations that cannot be removed or compensated. Vigour decreases, fertility declines, and death rates rise in response to minor perturbations. The asexual population enters an imminent extinction spiral. The sexual population, with half the reproductive rate in the ideal scenario, persists across hundreds of additional generations. Sexuality operates, thus, as long-term insurance: in a protected, stable environment, the cost is purely a cost, with no apparent benefit. But the "investment" in insurance becomes inescapably adaptive as soon as any real pressure occurs, environmental shift, a new parasite, or merely the pattern of genetic error that inexorably emerges in small, isolated lineages. Sex is not advantageous in the short term; it is indispensable in the long term because it constitutes the sole mechanism preventing accumulated genetic degradation.
Empirical documentation of the Red Queen hypothesis in Potamopyrgus antipodarum offers direct evidence of host-parasite coevolution in real time, observable in living populations and measurable. The freshwater snail exists in two distinct reproductive forms: sexual and asexual, within the same rivers. In populations of the St. Bathans River in New Zealand, where the parasitic nematode Microphallus occurs at high frequency (prevalence between 30% and 80% of snails), sexual populations have a significantly lower infection rate than asexual populations in the same geographic region. Data show that asexual genotypes (genetically homogenous clones), when frequent, suffer a parasitism rate of up to 80%, virtually the entire population affected. Sexual genotypes, by virtue of their continuous diversity, maintain an infection rate of 30-40% even under high parasitic pressure. The pattern repeats across different rivers and freshwater systems: where the parasite is prevalent, sexuality is maintained as the dominant reproductive form; where the parasite is rare or absent, asexuality becomes common. The mechanistic interpretation is clear and reproducible: the parasite evolves against frequent asexual genotypes, becoming specialised in infecting that specific genotype; as the parasite evolves this specialisation through directed mutation, asexual genotypes become diseased and suffer measurable population decline; sexual females produce rare genotypes that have no evolutionary history of exposure to the specially adapted parasite; the parasite takes time to evolve against these new, unknown genotypes; meanwhile, generations of asexual hosts suffer increased infection and mortality. Sexuality offers a continuous renewal of genetic "secrets" that the parasite has no time to fully decipher. Coevolution does not converge toward equilibrium: it is a process of perpetual pursuit where every innovation of the parasite (the Red Queen running) is met by an immediate innovation of the host (the Red Queen running too, to stay in the same ecological place). Both run eternally; neither catches up to the other definitively. This is the true structure of coevolutionary competition: never rest, never victory, only continuous movement of escape and pursuit without truce.
Empirical documentation of Muller's ratchet emerges from long-term studies of asexual bacterial lines maintained in laboratory settings for hundreds or thousands of generations without any possible recombination. Lines of Escherichia coli, replicated asexually for hundreds of generations in controlled environments, display a measurable accumulation of deleterious mutations following Muller's theoretical pattern with extraordinary precision. Lines decrease in cell size, reduce metabolic efficiency, and show increased mortality rates under environmental stress. Genetic sequencing reveals that deleterious mutations fix in populations in proportions and patterns predicted by Muller's theory. The process is, as expected, irreversible: no natural selection can "reverse" the accumulation because the class of individuals with the fewest mutations was lost through genetic drift and cannot be recreated in a strictly asexual population without a recombination mechanism to reshuffle segments. Sex, including simple forms such as bacterial conjugation, where two organisms exchange plasmid DNA, offers a mechanism that prevents this cumulative damage. A bacterium with mutations in genes A, B, and C can conjugate with another bacterium carrying mutations in genes D, E, and F; progeny can receive non-mutated alleles from both "parents", resulting in an individual with a relatively reduced mutational load. Recombination does not restore the ideal ancestral state; but it redistributes damage so that no single individual carries the worst possible scenario, the class that would have accumulated all deleterious mutations. The frequency of sex in living organisms correlates measurably with genome size: organisms with larger genomes, typically complex eukaryotes, display universally high sex frequencies; organisms with small genomes, prokaryotes, have sex rarely or not at all. The mechanistic speculation is direct and testable: larger genomes have a higher absolute mutation rate (more sites where mutations can occur); therefore, the accumulated error load is graver per generation; therefore, selective pressure for a mechanism to relieve it is stronger and biologically detectable. The pattern is no evolutionary coincidence: it is a direct response to the constraint of genetic size. The ratchet does not stop while the genome remains large; it is a permanent mechanism of generational risk.
12. The Descendant Is Other
The synthesis of the third section returns us to the fundamental mechanism governing sexual reproduction: the incorporation of alterity into continuity as an internal structure, not an external contingency or chance. The descendant is irreducible to either of its progenitors, not because biology selected "desirable variation" or because "the environment modulates gene expression", but because the mechanism of genetic transmission is, structurally, the dissolution of identity as a reproducible totality. The child is not a modified version of the parent. It is a combination that neither father nor mother ever was at any point in their life as an integral form. The geometric reduction of genetic contribution, 50% per progenitor, 25% per grandmother, 12.5% per great-grandmother, is not a number describing "how much" of the progenitor persists as a cohesive form; it is a number describing with mathematical precision "how much" of the progenitor is forgotten in each subsequent generation. What persists is not parental identity. It is the fragment: an allele circulating through multiple bodies without a totality to enclose it, a chromosomal segment that operated in a parental context and now operates in a completely new and singular context, a statistical probability distributed across generations of individuals, none of whom contains the original totality. Genetic continuity is continuity without identity, the circulation of components through forms that do not contain them and cannot reconstitute the totality that generated them. The descendant carries genes; it does not carry the progenitor as a totality. In this sense, alterity is not a category that the conscious subject discovers later upon encountering the other displaying difference. Alterity is a mechanical fact inscribed in the code of sexual reproduction, prior to any form of ethical or moral recognition that will come later. It is a condition of generation before it is a philosophical category.
When contemporary phenomenology speaks of the "irreducibility of the other" as the foundation of ethics, it begins already too late: sexual biology has already executed that revolution before any philosophy constituted itself. The other is the product of the biological operation that generated it. No one is "recognised as other" merely because they present themselves differently; one is other because the mechanism that produced them produced alterity as its own mechanical content. Embodied finitude is not a philosophical abstraction; it is the biological structure governing every reproductive act without exception. The individual body is finite in time: self-consumption, ageing, and somatic death are facts of metabolism. Genetic identity is finite through recombination and segregation: reproductive death in which identity dissolves so that another, different identity may exist. Both deaths, somatic and genetic, are constitutive conditions of sexual life, not accidents or pathologies that could be avoided. No biology could operate otherwise without ceasing to be sexual. The vulnerability resulting from this double finitude, a body that decays and ages, an identity that dissolves in each generation, a descendant that is other and operationally autonomous, renders the relationship with the other not optional, but structural; not a choice, but a fundamentally biological situation.
The return to the three sections offers a complete synthesis that resituates the problem of reproduction inside embodied finitude as a permanent and inescapable structure. First section: reproduction transfers instability, each organism receives not a stable genome as a fixed program, but a platform of instability, sensitivities, and capacities that update only in a concrete ecological context. Second section: inheritance does not program, genetic transmission contains no instruction manual determining development; it contains potential sensitivities and capacities that act only in the context of environmental interaction. Third section: sex dissolves identity, genetic continuity passes through the death of parental identity as a reproducible totality, incorporating alterity into the very act of transmission. Triple synthesis: finitude is incorporated into the operation of continuity such that no individual can perpetuate itself integrally and no descendant is a copy of what generated it. What each progenitor transmits is not the certainty of continuation; it is the inscription of finitude into the descendant. The descendant is not an extension of the progenitor requiring instrumental care; it is a vulnerable other claiming response because the relationship that produced it produced it radically as biologically other. Here, in the biological impossibility of repeating oneself in genetic continuity, space opens for the vulnerability of the other to summon response, not response as a moral obligation already constituted from outside, but as a structure of situation that emerged from the biological mechanism. An ethics that does not recognise this origin in biology is an ethics that fails to understand its own condition of possibility. Responsibility is not added afterward; it is engendered in the act that made the other exist as an other distinct from any parental totality.
The aphorism concluding this section, Sex does not overcome death, it incorporates it; whoever reproduces sexually dies in part so that another, different, may exist, crystallises the integrated logic of the three sections into a single radical conceptual operation. Death here is the death of genetic identity as a reproducible totality, not the somatic death of the body inhabiting space. The progenitor does not die bodily in the act of sexual reproduction; life continues in the body that generates. But its genetic identity as a reproducible unit dies in the very act of transmission. What persists is partial death: the totality that was an individual now circulates as a fragment across multiple others, impossible to reunite into a single cohesive form. No victory in this death. No defeat. No teleological purpose justifying or transcending it. It is pure mechanism: the incorporation of finitude into the heart of continuity. The sexual individual transfers not merely genes, it transfers its own biological impossibility of extending itself integrally. This impossibility is the structure making the descendant exist as other. It makes alterity exist not as a posterior category, but as a fact prior to all moral recognition. In this death-in-part lies the possibility of ethics: because nothing guarantees that the other brought into existence by my reproductive desire will be my image or extension. It will be biologically other. Biology guarantees this mechanically through crossing-over, segregation, and exponential recombination. The resting place of the question "how can the finitude of one give rise to responsibility for the other?" finds its answer not in moral philosophy, but in the mechanism that brought both into being. Responsibility is prior to morality; it is engendered by the biological structure that incorporates alterity into the very act of generating continuity. The ethical knot is not untied by an abstract concept; it is tied by the biology that made the other appear as unrepeatable. Finitude is not a problem to be solved; it is the structure of existence opening the possibility of the other.
Sex does not overcome death, it incorporates it; whoever reproduces sexually dies in part so that another, different, may exist.