Subscribe to Salvage and receive two print issues per year, plus digital access including audio and back issue PDFs
How Life Doesn’t Work
It is quite certain that in terms of merely mechanical principles of nature we cannot even adequately become familiar with, much less explain, organised beings and how they are internally possible.
– Kant, Critique of Judgment
The relation of end has proved to be the truth of mechanism.
– Hegel, Science of Logic
I.
‘Biology,’ Murray Rothbard once wrote, ‘stands like a rock in the face of egalitarian fantasies.’ He wrote this, as Quinn Slobodian documents in Hayek’s Bastards, while promoting a new fusionism, a crossover of right-wing libertarian thought with ‘paleoconservative’ racism and patriarchy that formed a substrate of Trumpism.
Fascism has often marketed itself as, in the words of Bavarian minister Hans Schemm, ‘applied biology’. Interwar fascism hymned, per Hitler, the ‘eternal laws of selection’. It espoused a mythopoetically aggrandised version of the Social Darwinism that was, as André Pichot has shown, thoroughly mainstream in the biology of the day. Today, the new reaction rests on a new bedrock of ancient biological myths – scientific racism, neo-eugenics, IQ racism, neurogenetic determinism – refurbished in explicitly capitalist terms.
In the ‘ethno-economy’ foreshadowed by figures like Peter Brimelow and Charles Murray and boosted by tech capitalists today, everything from the chromosome to supposed phenotypes like ‘whiteness’ or ‘neurocastes’ is recast as economic assets. IQ becomes a ‘reliable benchmark of value comparable to a stock market valuation’. Border fascism and pronatalism become essential ‘meta-market’ conditions for thriving accumulation, curbing (dixit Elon Musk) ‘dysgenic fertility’. This, again, appropriates and radicalises claims that, however incoherent and ill-evidenced, are still widespread among the proponents of evolutionary psychology and sociogenomics. It is also congruent with the spontaneous ideology of tech and big data. Life is understood as a manipulable codebase: weed out the bad alleles, avert the bad outcomes. Indeed, as Anita Say Chan points out in Predatory Data, the very idea of eugenics, being predicated on the keeping of extensive breeding and demographic records, is an application of ‘big data’.
Whenever the goal is to naturalise historically contingent inequalities, biology usually turns out to be the most expedient weapon. But why? In one sense it is obvious. Biology is a ‘rock’ against egalitarian freedom insofar as it supports the three major claims that Albert O Hirschman finds in reactionary rhetoric: social radicalism is perverse because it is against nature, is futile for the same reason, and as a result jeopardises valued social ends. But why should biology be so susceptible to such reactionary meanings? Why is the common-sense understanding of biology equivalent to that which is unalterably given? Why, for example, should it be that when trans people engage in a form of top-down morphogenesis, changing their bodies according to their rational desires, this is struck down as contrary to biology, as though the freedom to do so was not also biology? And why is popular biological writing filled with transparently ideological accounts of universal domination, deceit, competition and exploitation across biota? Why is it that, as Ambika Kamath and Melina Packer write in Feminism in the Wild, mainstream biological research is still inflected with the racist, sexist, colonial and ruling-class perceptual schemata of past epochs? Why isn’t it sufficient simply to deflate the scientific pretensions of the new far-right – its nonsensical arguments about vaccines, or the long discredited phrenology and IQ racism championed by Elon Musk – by appealing to more mainstream findings?
The problem, I submit, goes deeper than partisan research funded by tech millionaires or networks of racist think-tanks like the Human Diversity Foundation. It goes deeper, even, than the genetic reductionism of neo-Darwinian thought from R A Fisher to Richard Dawkins – the traditional foil of left-wing biologists like Steven Rose, Richard Lewontin, Stephen Jay Gould and Richard Levins. It even goes beyond the question posed by Jonathan Marks as to whether science is racist. The ontological question at the heart of biology – what is life? what is the meaning of life? – is urgently of the essence.
According to the neo-Darwinian synthesis of selection and genetics, life is a thoroughly meaningless, mechanistic process, the organism a passive vehicle for the gene pool, behaviour and functions controlled by the genes in a one-way causal process. Following August Weissmann, opponent of Peter Kropotkin’s mutualist biology, neo-Darwinians hold that changes in function acquired in one lifetime can’t be passed through the germ line. Evolution proceeds through minute, random genetic mutations selected passively by environmental pressures, as well as genetic drift. This paradigm extends to ‘origin of life’ studies, in which mechanistically-minded Marxists like Alexander Oparin, JBS Haldane and JD Bernal were formative. Applying Darwinian evolution to prebiotic chemistry, they argued that life was built, in Bernal’s aphorism, ‘from atom to molecule; from molecule to polymer; from polymer to organism’. A mechanical process of aggregation. Today Noam Lahav formulates the orthodoxy: ‘The origin of life was a Darwinian evolutionary process in which assemblies of inanimate molecules gradually became living entities.’
If we reduce life to complex chemistry, however, in what sense are we still in the domain of biology? Obviously, the molecular priors of life had to be assembled, whether in ancient clay beds, or in the heat of black smokers. But, as the biologist J. Scott Turner puts it in his defence of vitalism, biology ‘could only be a distinct science’ if ‘life were a distinctive phenomenon’. The effect of mechanical reduction is to close the ontological gap between life and death. And although vitalism is an insufficient response to that closure – I will argue for a refurbished Aristotelian ontology – it does at least correctly pose the question of how living things can be agentive, conscious and immanently free if they differ in no essential way from mindless matter. In short, there does seem to be something immaterial about life – its rational structure, its self-organisation and goal-directedness – which cannot ultimately be grounded in the properties of its material constituents. As Hans Jonas writes, the ‘parts of which the organism consists’ are ‘temporary, transient contents whose joint material identity does not coincide with the identity of the whole which they enter and leave’. The organism ‘is never the same materially yet it persists as its same self – by not remaining the same matter’.
In recent decades, biological research has forced the question through discoveries pointing to the innate agency, plasticity and adaptability of life. Far from biology being a synonym for the mutely unalterable, it might well be another word for freedom.
![]()
II.
Biology suffers, as Ernst Mayr famously said, from ‘physics envy’: the urge to reduce life to quantitative, mathematically defined mechanical entailments. This reduction, a legacy of the seventeenth-century mechanistic revolution, climaxed with what Julian Huxley called the ‘modern synthesis’ of Darwinian selection and Mendelian genetics.
Physics envy thrives in the strand of neo-Darwinism spawned by mathematician and eugenicist Ronald Fisher, with its themes of mathematised selection and genetic reduction. Mayr’s impatience with this was rooted in a more capacious variant of the modern synthesis originating with the liberal Darwinian Theodosius Dobzhansky, which also influenced more radical biologists like Stephen Jay Gould and Richard Lewontin. Dobzhansky resisted reducing organisms to sub-organismal elements, stressing variable genetic expression in different environments (‘norms of reaction’). Yet all wings of neo-Darwinism agree that the gene is the common element linking otherwise separate dimensions of development, inheritance and adaptation, and that adaptations are products of random molecular noise passively selected by environment.
Above all, they renounce the Aristotelianism which, dethroned by the mechanical philosophy, and seemingly interred by Darwin’s proofs that life evolved by unguided means, was blamed for two millennia of stasis. There is no small irony in this, since Aristotle effectively founded the field of biology with his later writings, from History of Animals, to Generation of Animals, Movement of Animals, Progression of Animals, Parts of Animals, and On the Soul. What neo-Darwinians abhor in Aristotle, though, is the imputation of mind-like properties such as reasons, forms and purposes to nature.
Whereas mechanical materialism treats matter as a self-sufficient cause, for Aristotle it couldn’t be the cause of anything by itself because, without being organised by rational form, matter was nothing but pure potentiality. In his hylomorphic ontology, all true substances are compounds of form and matter. And whereas mechanical materialism reduced nature to a bare syntax of bodies with extension and movement, capable of transmitting force to one another in a clockwork-like system, Aristotle famously identified a logical structure of four causes for anything that exists: the material cause (what it’s made of), the efficient cause (the force that moves it), the formal cause (its rational organisation), and the final cause (its purpose). It is the formal and final causes that neo-Darwinians reject, because talk of rational organisation and purpose smacks of design. That this is a misprision is particularly evident in biology: to say that the heart exists, and is organised in the way that it does, for the sake of circulating blood does not imply external design. It does imply a relation between parts and wholes in which the former serve the latter. And it does imply that the whole is striving toward a final end which, far from being some imposed objective, is simply the fullest realisation of its vital possibilities. David Bentley Hart suggests that Aristotelianism describes ‘the inherent logical structure of anything that exists insofar as it exists’, and I will argue that this is nowhere more obvious than in biology.
In its origins, Boris Hessen and Henryk Grossman argue, mechanical materialism expressed a revolt of the practical bourgeoisie not against Aristotelianism proper, but rather against a desiccated scholastic Aristotelianism. The mechanical philosophers were mostly not exposed to or interested in Aristotle’s biological writings, but they rejected the ontology of natures and powers, wholes and parts, and formal and final causes, admitting only material and efficient (motor) causes as suitable for explanation in a mechanical universe, allowing final causes (rational ends) to work only in the human soul, which was a subject for priests or psychologists rather than scientists. The resulting mechanistic materialism was highly successful as method, contributing to a drastic expansion of the bourgeoisie’s powers. But as its radical possibilities were exhausted, and as the method mutated into metaphysics, it could and did turn reactionary – particularly in biology.
As Marc Shell warns, to ‘argue or assume that matter is ontologically prior to thought’ is to surrender to ideology: think of phrenology, physiognomy, gene determinism and the collapse of mind and social life into biochemistry. To which we could add the reduction of life to ‘naturally occurring machinery’ (AG Cairns-Smith), exhaustively explicable in ‘physico-chemical terms’ (Jacques Loeb), which flattens the ontological abyss between life and death. Metaphysical materialism, applied to organisms, renders life engineerable, passive material – and thus, through breeding, rearing and interventions in the genome, as material to be controlled and shaped for capitalist ends. Or what Melinda Cooper calls ‘life as surplus’.
The old teleology, though, in Engels’ words, ‘gone to the devil’, keeps returning. Biology cannot avoid intentional, goal-directed language. Talk of the adaptive ‘functions’ of the phenotype, or the ‘decisions’ of a cell, invokes teleology. Darwinian natural selection is implicitly teleological – or at least what David Depew calls a ‘consequence etiology’ insofar as a trait only exists because it enables the survival and reproduction of the organism. But neo-Darwinism insists that natural selection is only a ‘pseudo-teleology’. The language of goals is either a methodological fiction or a façon de parler. For Dawkins, organisms only give ‘the appearance of having been designed for a purpose’. But ‘design’ is a red herring, giving the false impression that the only alternative to the mechanistic paradigm is an externally imposed order. Either way, the organism would be a machine: a contrivance, either of accidental causes or of demiurgic design. Organisms are ‘no longer like machines,’ Richard Lewontin wryly observed, ‘they are machines’. Yet, living systems exhibit causal powers of self-organisation and self-reproduction that no mechanism can replicate. As Robert Rosen writes, ‘a material system is an organism if, and only if, it is closed to efficient causation’. And if, as Annie Crawford persuasively argues, it is impossible to rephrase teleological statements in metaphysically neutral language without some loss of information – try, for example, to rephrase the statement, ‘to accomplish metabolic process X, enzyme Y evolved a specificity for Z’, without any purposive implications whatsoever and without losing some of the sense of the statement – this may be because life is teleological.
The causal poverty of the mechanistic framework periodically provokes short-lived, inadequate rebellions of ‘vitalism’, where life is explained by a superadded vital principle that still leaves the underlying mechanical framework intact. It is often assumed, partly thanks to the intellectual heft given the idea by such accounts as Daniel Gasman’s The Scientific Origins of National Socialism, that vitalism is tendentially reactionary or fascistic. That is not so. Modern vitalists like Hans Driesch were as firmly anti-fascist as mechanistic biologists like Jacques Loeb. Fascism was not committed to a single ontology of life. In fact although, as Ze’ev Sternhell describes, it often defined itself against ‘materialism’, in its ‘applied biology’ it inherited the unresolved contradictions of bourgeois materialism, oscillating between mechanicism and vitalism.
Popular biological writing tends to assume that scientific explanation can only choose between these post-seventeenth-century options. In recent decades, however, biology has returned to agentive, goal-directed conceptions of life that could point beyond this. Discoveries in genetics, epigenetics, cell biology, cognition, immunology and evolutionary theory have eroded the mechanistic paradigm. A major casualty has been the sovereign gene, and associated doctrines of hard heredity. The Human Genome Project, intended to decode life’s ‘blueprint’, found fewer protein-coding genes in the human genome (20,000) than in the water flea genome (31,000). Genes, far from being fundamental determinants of life, are raw materials moulded by higher-order processes like regulatory networks and agencies like cells: on which, more later. The genome, as biologist James Shapiro suggests, seems to be more of a read-write system – a library of functions that is constantly revisable during cell division, the growth of the organism, and the evolution of a population – than a rigid code. Further, as developmental biologist Mary Jane West-Eberhard notes, research guided by genetic reductionism produced ‘massive documentation of a heretofore widely underestimated capacity for adaptive condition-sensitive behaviour and development’: phenotypic plasticity. Insofar as these changes are sustained and then genetically assimilated, genes are ‘followers, not leaders’ in evolution.
Conrad Waddington’s argument for ‘genetic assimilation’ of acquired characters is one of a range of recently revived ideas. Richard Goldschmidt’s once execrated argument for evolutionary macromutations is supported by evidence of large-effect genomic restructuring, while his doubt about the ‘particulate’ gene is partially vindicated. There are even some qualified grounds for a partial revival of Lamarck, as argued by Snait B Gissis and Eva Jablonka, in Transformations of Lamarckism. Lamarck was once derided by Mayr as a ‘dirty word’ in biology for his claim that characteristics acquired in one’s lifetime could be inherited. As Richard Burkhardt has shown, Lamarck’s argument was hardly unique to him at the time and was subsequently caricatured. Critics assumed that a minor throwaway line, suggesting that the giraffe’s neck evolved as a shorter-necked ancestor stretched to reach leaves, was the strongest exposition of the idea. Nevertheless, the heritability of epigenetic changes, laboratory experiments showing heritable cellular adaptations arising without random mutations, and of course the non-genetic transmissibility of cultural and habitual behaviours with major biological consequences all suggest it’s time to abandon the anathema. It may even be, as suggested in immunological research by Edward J Steele and others, outlined in Lamarck’s Signature, that Weismann’s barrier between soma and germ cells is selectively permeable.
‘Niche construction’, a process originally described by Lewontin and Levins in which animals shape environmental selection pressures, combined with extensive evidence of intergenerational learning, implies that organisms co-direct evolution, both its objects and its subjects. Even sexual selection, stripped of what Evelleen Richards shows are its sexist, colonial and heteronormative trappings, and of folksy reductions of it to the operations of selfish replicators (alleles), describes how bodies are formed by the erotic and aesthetic desires of sexual partners: the astonishing Paisley design on the tail of the male Great Argus pheasant, the vivid stripes and patterns of reef fish, or the extravagant displays and shimmering electric colours of the bower bird are all famous examples of traits arising from sexual choice. In many cases, like the Argus tail, or the chucks made by male túngara frogs, the traits are either costly or dangerous for the animal in pure survival terms, but persist because of their beauty. Desire is a deep-historical force that, through thousands of generations of attraction and creation shaping the organism’s bauplan, crafts the living world into miscellaneous wonders. Its glamorous productions are achieved against the grain of selection, which alone would relentlessly shear off the useless or burdensome.
Finally, there is the greatly expanded world of animal cognition, imagination, and tool use among animals: spiders and bees turn out to be ingenious creative architects, birds are tremendous navigators, and tool use is ubiquitous among mammals and birds. There are, too, highly developed and informationally rich syntaxes found across different animal languages, from the waggle dance of bees to the use of name calls by elephants, dolphins and parrots, to recursion in the languages of whales and songbirds, to the syntactic regularities discovered in the ultrasonic vocalisations of mice. In addition, there is a growing body of work on ubiquitous ‘prosocial’, moral behaviour and attitudes among animals as diverse as rodents, birds, monkeys and dogs. As Marc Bekoff and Jessica Pierce argue in Wild Justice, ubiquitous morality among animals is exactly what evolutionary continuity should lead us to expect, and what the evidence points to. One might even suggest — as a deliberate provocation, but also completely in earnest — that reason, morality and mind are, not freakish evolutionary byproducts of more originally selfish ‘replicator’ strategies, but bootstrapped realisations over evolutionary time of life’s innate propensities.
This freshet of ideas and discoveries leads Eva Jablonka and Marion Lamb to advocate an ‘extended evolutionary synthesis’ accommodating these agentive and symbolic dimensions. In the philosophy of biology, renewed emphasis on organisation, functional integration and emergent levels of regulation has stimulated curiosity about Aristotelian biology with its stress on natures and powers, form, vital activity, goals, and the downward causal power of the whole. Philosophers like George Molnar, Nancy Cartwright, James G Lennox, Ruth Groff and John Greco have gone further in mapping this terrain than modern Aristotelians like Philippa Foot and Michael Thompson. None of this is uncontested, but the goalposts have moved. The Dialectics of Biology Group, which anticipated some of these trends, would appear moderate by today’s standards.
Above all, what the new biology stresses – scandalously, astoundingly, against the grain of all prior popular understandings, and certainly against the capitalist view of life as a manipulable series of inputs and outputs – is the innate freedom of living things.
![]()
III.
The most judicious, and theoretically conservative, account of the emerging biological paradigm is Philip Ball’s How Life Works. Ball, a former editor of Nature, has distilled decades of research contradicting the dominant, genocentric paradigm. As a neo-Darwinian physicalist, he has no interest in reviving Lamarck, let alone Aristotle. Yet the portrait he supplies is more radical than his own aims.
At the heart of Ball’s analysis, ascending from genes, proteins and regulatory networks to cells, tissues and organisms, is the concept of ‘causal spreading’. As organisms become more complex, multiple levels and directions of causality emerge – it isn’t all one-way traffic from the gene upward. This idea is akin to Denis Noble’s theory of ‘biological relativity’ or Steven Rose’s ‘epistemological pluralism’. The critical move, though, is his insistence that life is a truly agentive, goal-directed process. This begins with the cell, whose ‘rediscovery’ is, according to the molecular biologist Michel Morange, ‘probably the most important change that has taken place in molecular biology since its rise in the 1950s’. Cells are not molecular factories or building blocks, Ball says, but truly ‘cognitive’ entities that cognise both their environment and their inner state and establish meaningful relations between them. To understand how and why, it is necessary to revise the old story of the sovereign, all-determining gene.
From Darwin’s ‘gemmules’ to Rosalind Franklin’s discovery of DNA structure, the gene was conceived of as a particulate unit of inheritance. According to genetic fundamentalism, DNA code was read linearly and from a fixed starting point, transcribed into RNA, which in turn was translated into a single protein type with a definite function in the cell’s development. DNA contained, in James Watson’s words, ‘instructions’ for development. The organism thus became an epiphenomenon of sub-organismic particles and the cell what Dawkins called a ‘convenient working unit for the chemical industry of genes’. This, as Ball acidly puts it, is like explaining novels by saying that words give rise to sentences and books.
But most genes in the human genome have no assigned function. Many are pleiotropic, influencing multiple effects, while most traits are polygenic. Genome-wide association studies (GWAS) have revealed that the relationship between genes and traits is probabilistic, not deterministic. Knocking out supposedly crucial genes often has either no effect or an unexpected one. Just 2 per cent of our genome codes proteins while the rest, once written off as ‘junk DNA’, often codes regulatory RNAs. Even the gene’s supposed discreteness is under revision: the ENCODE project has shown that genes are modular entities formed from non-contiguous sequences and promoter regions, not discrete DNA stretches. Barbara McClintock’s transposons – genes capable of inserting themselves into remote parts of the genome – are central to genome function, with around 65 per cent of the genome capable of ‘jumping’ behaviour.
As Francesca Bellazzi argues, this revised gene can still be considered a ‘natural kind’ – although, I would add, only insofar as it has a purpose within a wider living system. But there is no ‘gene for’ any trait and no unity between the molecular and the Mendelian gene. The older gene concept was a reification, objectifying life’s intrinsically relational and purposive qualities. Genes are not the prime movers of development but, as Fred Nijhout puts it, ‘passive sources of material on which a cell can draw’.
Another challenge to the sovereign gene comes from RNA, once thought to be a go-between from DNA to the protein it coded. In this picture, non-coding RNAs (ncRNAs) were just transcriptional noise. It’s now known that ncRNAs regulate gene expression. They can either enhance the transcription of a gene and thereby amplify its effects, or suppress transcription so that it has no effect at all. They do this by methylating different parts of the DNA: methylating coding regions will enhance transcription, while methylating promoter regions will suppress transcription. An example of this phenomenon is the LacI repressor discovered by Jacques Monod and François Jacob, which blocks the transcription of genes for lactose-metabolizing enzymes in E. coli. These ncRNAs also respond to environmental conditions and leave ‘epigenetic marks’ that persist across generations, hinting at quasi-Lamarckian inheritance. Tellingly, the role of ncRNAs rises as organisms become more complex. While 90 per cent of bacterial genomes code for proteins, only 2 per cent of the human genome does. Metazoans rely on flexible regulation rather than sheer genetic quantity.
The proteins produced by DNA were once seen as the ‘workhorses of the cell’. The ribosome, on the ‘workhorse’ view, reads the RNA strand and translates it into a polypeptide, which folds into a unique 3D structure depending on patterns of attraction and repulsion created by hydrophilic and hydrophobic regions. Each shape permits only one kind of ‘molecular recognition’, and thus one kind of function. In reality, protein folding is far more dynamic. A single gene can produce on average six functionally distinct proteins through alternative splicing of exons: a process regulated by a spliceosome consisting of RNA and RNA-binding proteins (RNABPs). Many proteins aren’t structured but assume multiple conformations depending on their interactions, a flexibility that allows cells to respond to diverse circumstances without precise genetic instructions – again, this flexibility is far more important in metazoans than in bacteria. And since crystallographic studies focus on biomolecules that can be crystallised, about half of human proteins are unknown: what Ball calls a ‘dark proteome’. The modular structure of proteins also permits an evolutionary shuffling of functional molecules to create new multidomain molecules with novel potentials for development.
Protein complexes, rather than catalysing specific reactions as was once assumed, form intricate regulatory networks – interactomes – that exert top-down control over genetic activity. At the heart of the interactome are transcription factors (TFs). Once thought to toggle specific genes on/off, they bind on multiple sites, and are sometimes substitutable with one another, as is the case with the factors which regulate the genes encoding bone morphogenetic protein. All of this, before we even get to the cell, undermines the standard model of bottom-up causation. The multiple potencies of genes, proteins and regulatory networks frustrate the search for individuated mechanisms. Rather than genes dictating development in linear fashion, the combinatorial nature of interactomes allows organisms to achieve the same outcome by different pathways: for example, just as different transcription factors can substitute for one another in the production of bone morphogenetic protein, so skeletal muscle in mice can be built with either or both of two transcription factors MyoD and Myf5. The redundancy in the system gives it robustness. This, Ball argues, takes ‘causation largely out of the hands of the molecules themselves’, as higher-order organisation channels the ‘frenzied macromolecular party’ within each of our 37 trillion cells into structured outcomes. Complex life evolved, not by acquiring more genes for more complex programming, but by regulating them better from above. And even to say as much – to speak of ‘regulation’, let alone ‘better’ regulation – already implies purposive, goal-directed action at the level of the organism.
![]()
IV.
What happens to a cell when freed from its organismic constraints? Research led by the biologist Michael Levin, assembling ‘xenobots’ from frog embryo cells, shows that cells will adapt to their new environment and construct a coherent body plan to survive. Similarly, living human lung cells scraped from deceased tissues can be nudged to form self-repairing multicellular structures.
Cells, then, are cognitive, context-sensitive units, making autonomous decisions. According to Ball, citing the biologist Dennis Bray, they are the ‘touchstones of human mentation’. This, of course, opens a can of worms. In what way are cells cognitive? Do they truly make decisions in ways at least analogous to, or prefigurative of, those made by humans? Certainly, some evidence would point in this direction. For example, single-celled protists like certain zooplankton exhibit memory, timing and learning in their daily motions up and down the sea. From a mechanistic perspective, what look like decisions must surely be the outcomes of stochastic chemical fluctuations, genetic programming or mechanical transmissions of force. Yet that seems, prima facie, inadequate: the simplest cell appears to have the goal of persisting, and in service of that goal a form-determined ability to cognise inner and outer states and act on those cognitions. It thus exhibits a basic intentional relationship to the reality in which it persists. And as cells form communities, becoming multicellular organisms, they appear to cooperate for the survival and reproduction of the organism. The question, which will come up later, is whether such (non-metaphorical) claims for cellular cognition and decision-making are even intelligible from the materialist, neo-Darwinian perspective that Ball defends.
Cells within an organism start making decisions during differentiation, as they commit to specific fates. This process is conditioned but not determined by genes or regulatory systems, and physical constraints also define the range of stable cell states, as Conrad Waddington described with his famous ‘epigenetic landscape’. In Waddington’s model of ‘canalization’, cells travel down a branching series of pathways, representing progressively narrower developmental channels until they reach their basins. Research on mouse embryos suggests a murkier reality. Cells gradually commit to ‘blood-like’ or ‘bone-like’ identities but retain the ability to switch fates via ‘tunnels’ connecting the pathways. To avoid getting stuck in the wrong fate, cells leverage genetic ‘noise’ – random fluctuations of gene expression – to explore alternative routes. They actively interpret and respond to their environment. Ball approvingly cites Michael Levin and Daniel Dennett’s argument, echoing Humberto Maturana and Francisco Varela in their Autopoiesis and Cognition, that it is ‘cognition all the way down’.
From the collective decision-making of cells, tissue is formed. Some organisms exhibit remarkable regenerative abilities, such as planarian worms, which can precisely reconstruct lost body parts, or deer, which regenerate antlers in a way that preserves past injury patterns. This indicates remarkable body plan memory. Levin’s work suggests that bioelectrical signalling – voltage gradients across cell membranes – guides development by encoding positional information, influencing gene expression and tissue architecture. Cells also exploit mechanical forces, such as the folding of a cluster of cells as it expands in a confined space, to produce organisation. Ball describes how early chicken embryos use positional information and mechanical forces to break symmetry and create the structure necessary for organs to develop: a cluster of hypoblast cells migrates to one edge of the embryonic disk, defining a structural axis. Opposing cells then express Wnt and Nodal genes to initiate gastrulation, setting up the basic body plan.
The formation of body parts similarly exploits physical constraints. In 1901 Theodor Boveri proposed that chemical concentration gradients (morphogens) establish polarity in developing embryos, a thesis supported in John Saunders and Mary Gasseling’s 1968 work on limb patterning in chicken. The latter showed that a molecule, later identified as Sonic Hedgehog (SHH), helps coordinate positional information through an activator-inhibitor mechanism similar to those proposed by mathematician Alan Turing. This mechanism also assists feather patterning, lung branching, and blood vessel formation. Such constraints do not, and could not, determine form alone – they can only provide what Ball calls ‘structure for free’. Rather, what emerges from his account is that development depends on all of these aforementioned mechanical and biochemical causes – Hox gene expression, transcription factors, molecular signals, concentration gradients, environmental feedback – being functionally integrated and organised by higher agencies, beginning with the cell. Which raises the tricky question of what agency is, and how a cell could possess such.
![]()
V.
What sorts of decision-making agencies are involved, and how are they explicable? Ball, quite reasonably, defines biological agency as ‘the ability of a living entity to manipulate and control itself and its environment in order to realise a goal’. This means that agency is intrinsically teleological. And, as Matthew Tugby argues in Teleology, all living systems are teleological to the extent that they pursue goals of persistence, self-organisation and self-reproduction.
Over decades of exaggerated anathema, teleology has often been conflated with external design, inverting its original sense of internal striving toward an end. This is no doubt in part because the internal goal-directedness of an organism resembles no machine or mechanism. Machines exist near thermodynamic equilibrium unless in use by a human, do not resist entropy, and passively transmit mechanical or electrical force. Organisms exist far from equilibrium, resist entropy, and possess intrinsic causal power.
The physicist Erwin Schrödinger hoped to ground life’s ability to maintain dynamic order, by consuming ‘negative entropy’ from its environment, in the patterns of sustained, ‘well-ordered atomic association’ found in nature, such as aperiodic crystals. Ball, drawing on Jeremy England’s research into how complex physical systems often settle into well-adapted states, adopts a similar approach. Such dissipative systems, it is claimed, could give rise through successive iterations to the self-sustaining dynamic order typical of life. From the functional integration of physical systems, so the argument goes, would emerge the capacity for goals, which – like everything else in biology – must have evolved. From this elementary goal-directedness would evolve the basic cognitive capacities of cells, and from their complex organisation would arise perception and mind. This is a non-reductive physicalism in which the agentive, cognitive and mental capacities of living systems supervene on and are realised in physical processes with no non-physical remainder.
This is a common approach in mainstream biological writing. For example, Nick Lane claims, in 2022’s Transformer, that the dynamic order of living systems can ultimately be grounded in chemical cycles, and ‘the power’ of chemistry to ‘self-organise, to grow’. But it falls apart given a moment’s thought. Physical states, to begin with, are not remotely ‘adapted’ in the sense that organisms are: adaptation already implies goals. The ‘self-organisation’ of dissipative systems is not even analogous to self-directed action: there is no ‘self’ to organise, just a passive transmission of mechanical force. To treat goals as an emergent property of functional integration is to invert cause and effect: for anything to be integrated by functions is for it already to be a goal-directed process. The ‘capacity to develop goals’ cannot really have evolved into existence: selection can only act on existing variations, and Ball is clear that ‘all living systems exhibit’ goal-directedness, so there is no variation for selection to act on. Goal-directedness isn’t just a phenotype; it is what distinguishes living from non-living systems.
It would be far better to ground talk of goals in the immaterial natures and powers expressed in organic processes that obviously exceed mere law-like physical regularities. The argument, in short, presupposes the very mind-like, intentional strivings that it is supposed to explain. To escape this deadlock, Ball falls back on ‘integrated information theory’, which holds that consciousness is already present in quantum reality, as particles of information embedded in physical systems, awaiting integration. But, as John Searle points out, this amounts to treating an effect of mind as its cause: information is only information to a knowing mind.
The concept of ‘causal emergence’ also does much heavy lifting here. It is a truism in biology that entities at higher levels of organisation display properties that are not found in their constituent parts. That is what is called emergence. For example, the folded shape of a protein is not simply a function of its lower-level biochemical properties. The pumping of a heart is, in the same way, not reducible to the properties of cardiomyocytes. What, though, does it mean to speak of ‘higher levels of organisation’? In both cases, we are talking about the formal organisation of matter toward a final, normative end: the thriving and reproduction of the organism. The language of emergence, as soon as it aspires to be more than surface description, illicitly relies on the kind of formal and final causes that organise matter and actualise its potentials. Having thus covertly smuggled forms and goals into the picture, where they are needed to explain higher-level organisation, it then tries to explain them as an emergent property of higher-level organisation.
The effort to retain the physicalist, neo-Darwinian paradigm eventuates in a series of logical contortions. And in so doing it frequently leads to the error, common to Ball, Lane and Morange, of assuming that non-physicalist explanation in biology must be ‘vitalist’. And ‘vitalism’ has supposedly been ‘disproved’. This claim generally refers to the contention of a short-lived variety of ‘physical vitalism’ in the eighteenth century, that there was some sort of vital substance, an argument that contemporary vitalism rejects. But if vitalism is out of the picture, that only leaves more or less sophisticated physicalisms. This provides little coherent basis for the concepts of agency, goal-directedness and the immanent power of self-organisation that current biological research points to. We can’t ground agency in information, teleology in a world exhaustively consisting of lawlike physical regularities, or self-organisation in dissipative physical systems. It would be more coherent to talk of biological agency in terms of the Aristotelian ‘soul’ (psyche), as the form or actualisation of material potential in living bodies – but that, too, is often conflated with a ‘vital principle’, as though Aristotle was adding something to an otherwise mechanical picture. Likewise, teleological processes are difficult to explain as emergent properties of an antecedent material order, since it is unclear how a non-intentional, non-goal-directed process could possibly give rise to one with goals. However, they are much more explicable in terms of a hylomorphic unity of form and matter where form actualises material potentiality, because matter organised by form possesses certain dispositional powers (dunamis) that are inherently directed toward the ultimate realisation of the possibilities contained in those powers. The capacity for an initially homogeneous embryo to polarise and gastrulate is such a power, and the natural striving of the embryo toward a developmental peak is such a goal. Indeed, form (eidos) and finality (telos) must be an essential part of the explanation for why the same molecular stuff can perform such varied organic roles, while the same goals can be met by various molecular pathways.
Indeed, the breakthroughs of recent decades become abruptly more intelligible in Aristotelian terms. Genes and proteins, for example, can be counted among a suite of dispositional powers whose actualisation depends on formal constraints (regulatory architecture, interactomes, signalling) that enable organic goals to be reached in various ways, and on the downward selection of the formal whole. RNAs and transcription factors are both formal constraints and efficient instruments for the actualisation of the body’s material potentials, modulating how organismal form is achieved, but subject to the formal and final oversight of cellular agencies. The substantial form of the cell, where Ball first identifies real biological agency, confers immanent powers of self-organisation. The pathways of developmental plasticity and canalization represent the formal orchestration of the space of potentials into outcomes normed by what is good for the organism. Bioelectrical and molecular signalling are formal constraints that encode final ends. Ultimately all of this activity is orchestrated, not by matter but by the strivings of the ‘soul’.
![]()
VI.
Decisions ‘about biological truths’, writes the historian Gabriel Motzkin, are ‘political decisions’. Of course, no particular ontology of life maps neatly to any particular politics. Soft heredity coexists with eugenics, teleology with facile right-libertarianism, mechanism with radical politics, and there are forms of right-Aristotelianism that support authoritarian conservatism. Nonetheless, the portrayal of life as plastic, self-organised and goal-directed is basically subversive. For it is in no way commensurable with the fatalistic view of life as, in Kriti Sharma’s satirical words, a competitive struggle over ‘energetic costs and profits’ where ‘the genetic cream rises to the social top’ and ‘nothing is ever given away’.
In fact, from an Aristotelian perspective, such capacities are exactly what would ground freedom. For Aristotle, freedom is rational self-direction: an action is only really free when its principle (archē) lies in the agent, not in external compulsion or in accidental, privative conditions like ignorance. In living systems endowed with what Aristotle would call a ‘rational soul’ – and despite his own restriction of the rational soul to humans, it is hardly self-evident that any organism fails to participate at least to some degree in rationality, to the extent its form permits – that principle would be its self-movement guided by logos toward rational ends.
Strangely, although much of the research described above is decades old, it hasn’t yet fully been assimilated by mainstream biological writing. James Shapiro and Denis Noble reveal how dozens of discoveries challenging the modern synthesis have been excluded from mainstream popularisations of evolutionary theory. Similarly, West-Eberhard notes that plasticity’s role in evolution, though well-documented, is still treated as ‘controversial’. Scientific impasses like this may be symptomatic of the deeper ‘epistemological cultures’ described by Evelyn Fox Keller, shaped by the intellectual and institutional ‘needs’ of scientists. These are themselves constrained by the social and political conditions of intelligibility. As Steven Shapin argues, these are not ‘contaminants’ of scientific work but ‘constitutive of the very idea of scientific knowledge’.
Biology is at its most ideological when disavowing and thus naturalising its social determinants – whether it is the relationship between mechanical philosophy and the rising bourgeoisie, the industrial revolution and Lavoisier’s chemistry, Victorian capitalism and Darwinian selection, the mechanisation of labour and the thermodynamic revolution, or the rise of hard heredity and fantasies of industrially rationalising biology (eugenics) that so pervaded the modern synthesis. Beyond intelligibility constraints, there are the institutional matrices in which scientific research takes place: the Cold War fusion of science with state, capital and military research produced the information theory that now informs gene sequencing and disease research. Molecular biology and genomics have been complicit in the commodification of life itself, with the 1980 Supreme Court ruling in Diamond v Chakrabarty setting a precedent by allowing the patenting of genetically modified organisms. Mechanistic materialism, of which neo-Darwinism is the most pervasive expression, has been the language of the appropriation, control and commodification of life’s powers, and that has surely skewed the range of what is intelligible.
For those seeking ‘a liberatory biology’, as the Dialectics of Biology Group put it, there is hope here. Although it hasn’t been a theme of Marxist biologists, historical materialism is inflected with a basically Aristotelian ontology. This is casually acknowledged by Terry Eagleton, in his short pamphlet on Marx, who describes him as ‘a sort of Aristotelian essentialist, who holds that there is a human nature or essence’. In fact, as documented in the work of David Depew, Scott Meikle and George McCarthy, Marx’s Aristotelianism went much further, drawing on a hylomorphic conception of matter, a dialectic of forms, and an understanding of essences, law and necessity, with ancient Attic roots. Marxism is, as Ernst Bloch argued, a species of ‘left-wing Aristotelianism’. This is sometimes given short-shrift, as when John Bellamy Foster, in Marx’s Ecology, dismisses Marx’s entire relationship to Aristotelian naturalism in a single statement claiming that Marx eschewed ‘Aristotelian final-causes’ in favour of an Epicurean materialism of the perverse wilfulness and freedom of the atom: a strongly hylozoic materialism bearing little relation to the mechanical philosophy. If, however, we now find that final causes are indispensable in biological explanation, this legacy is worth revisiting.
Historically, radicals affiliated with materialism and its disenchantments as a strike against traditional authority. But they have always been ambivalent about ‘metaphysical materialism’, which today would be called physicalism. As China Miéville explains in his evisceration of the myths of ‘folk Marxism’, in ‘Beyond Folk Marxism: Mind, Metaphysics and Spooky Materialism’, that equivocation extends to the Marxist tradition. This is nowhere clearer than in the traditions of radical and Marxist biology which emphasise what Marx and Engels called, in a back-handed compliment to idealist philosophy, the ‘active side’: agency. From the left-Lamarckian tradition that Adrian Desmond and PJ Hale have shown spans Victorian radicalism and Lunacharsky, to more orthodox left-Darwinian biologists like Steven Rose, they have insisted on the innate history-making power of the organism deriving from life’s radical indeterminacy and goal-directedness.
The recourse to dialectical materialism, as in Lewontin and Levins’ work, clarifies what is at stake here. For dialectical materialists, contradiction is at the foundation of nature. And since contradiction is an act of rational unfolding, this illicitly – but, I claim, correctly – imparts to nature a mind-like and autopoietic capacity of a sort that could ground life and freedom. In other words, Marxist biology, far from being committed to ontological materialism could playfully be lauded as a cunning idealism that appropriates and inverts the resources of materialism for a project that consistently historicises and de-materialises the supposed physical foundations of behaviour and aggrandizes the rational and social powers of mind to overcome the material obstacles to its collective purposes: ‘materialism without matter’, as Alberto Toscano puts it.
The tacit Marxist argument for the reality of downwardly causal rational relations in life is nowhere clearer than in its understanding of humanity as an historical species which brings itself into being through rational, collective labour. World history is ‘nothing more than the creation of man through human labour’, the ‘self-mediated birth’ of the species as Marx puts it in The Economic and Philosophic Manuscripts of 1844. The work of Peter Corning, Lesley Newsom and Peter Richerson, and Sarah Blaffer Hrdy even suggests there may be some basis in evolutionary biology for the claim that social cooperation drove the evolution from Australopithecines to Homo sapiens, and that a specific form of cooperation – alloparenting – laid the basis for advanced moral capacities like empathy and theory of mind. This suggests that new, distinctly human phenotypes were produced not in the first instance by changes in allele frequencies but by a Lamarckian change of habits enabling the fuller self-actualisation of a latent capacity for social freedom.
A similar case can be made for Marx and Engels’ claim, in The German Ideology, that language and consciousness are, from the beginning, ‘a social product’. Of course they are wrong in their assumptions about animal minds, and about the rich pragmatics and syntactical complexities of animal communications. Moreover, for language to have evolved by means of social habit and intergenerational learning, humanity must already have been nascently in possession of it. Even a very minimal act of reference like ostension (pointing) would be utterly opaque – no one could be sure what exactly was being pointed at, or why – unless it was already part of a complete semiotic system understood by everyone involved. That form-determined material capacity for linguistic rationality had to have been elaborated over the course of evolutionary time, otherwise its cultural evolution in the human species could not take place. Nonetheless, there are growing bodies of research, developed by Michael Tomasello, Daniel Dor and others, supporting a case for cultural origins to human language. This suggests that a change of habit, in this case a striving to communicate rationally with conspecifics, to share attentional frames and intentional states with them, elicited a new phenotype with astonishing global consequences. Which, again, strongly supports the role of downwardly causal social and rational relations in the making of the species.
We can, indeed, think of the whole pattern of natural history as a striving toward the realisation of dispositional potentials, forms and goals already implicit in life, with progress achieved not in a linear Whiggish fashion, but in the course of what Stephen Jay Gould and Niles Eldredge called ‘punctuated equilibrium’ – or, as Lenin put it in his conspectus on Hegel’s Science of Logic: ‘Leaps! Leaps! Leaps!’
It is well known that relatively minor physiological pivots can result in new phenotypes with enormous consequences. For instance, a small gene-regulatory change in a fish can turn myocytes into electrocytes, producing an electric organ that alongside neural timing and sensory systems produces a radically new biological form capable of stunning prey and predators, perceiving objects in the environment as an electric image and developing novel forms of social behaviour and mate choice: a whole new cognitive and social niche. Even more compelling are those leaps made possible by incremental changes of habit. One example of this would be the cognitive liberations that, according to Carol and James Gould, are frequently achieved by new habitual breakthroughs, as when birds learned to nest on platforms resulting in the explosion of new mental capacities, and eventuating in a self-reinforcing spiral of cognitive gains. Another would be the moral liberation enabled by the aforementioned practice of alloparenting, resulting in the development of a sympathetic imagination that on current evidence seems to exceed that available to other animals.
In the twentieth century, and long before the Anthropocene was heard of, the Soviet geochemist Vladimir Vernadsky hoped that a new leap might be made possible by the rational, socialist governance of the earth. Having coined the concept of ‘the biosphere’, he looked forward to a new ‘psychozoic era’ in which the earth was enveloped in the ‘noösphere’ – consciousness becoming a planetary force. If Vernadsky was premature, it was perhaps only because capitalism (state capitalism included) has appropriated, alienated and blackboxed life’s immanent rationality and freedom as immanent externalities to be exploited. A first step toward breaking capital’s grip over life would be to reject the hideous, necrotic image of life bequeathed by metaphysical materialism and put to excellent use by the accountants and administrators of capitalism so that life can be converted into flows of energy, labour processes, factories.
For life, insofar as it is consubstantial with agency, purpose, meaning-making and plastic capacities for self-realisation, from cells to souls, is indeed the deep-historical process of that liberation. From which purview, a ‘liberatory biology’ is not a paradoxical or accidental byproduct of an otherwise wholly deterministic chain of molecular antecedents and consequents, but a tautology: biology is nothing other than the actualisation, through and against the material confinements and privations that obstruct it, of freedom.
![]()
Richard Seymour is a founding editor of Salvage, and author of Disaster Nationalism (Verso, 2024). He writes at Patreon.
This essay appears in Salvage #16: The Construction Site of Hell (Autumn/Winter 2025).


