As Ernst Mayr wrote, “no other ideology has influenced biology more profoundly than teleological thinking” (Mayr 1992, 117). This approach, which interprets phenomena based on their finality (telos), has permeated the history of biology since Aristotle. It remains prevalent today, particularly in evolutionary biology, both explicitly – but distanced as metaphorical (it is as if nature were acting for the good of the species, optimising organs), and implicitly, through the structure of the still-dominant functionalist reasoning. Indeed, describing the function of an organ requires considering the role it plays in the organism’s survival. Explaining embryonic development involves showing how each stage leads to a viable adult. Understanding the evolution of a trait means identifying the function it served when it appeared (what it was selected for). These biological phenomena are accounted for by their effects while in most other sciences, mechanistic explanations (where causes precede effects) prevail. This paradox is captured by a quip attributed to the geneticist John B. S. Haldane: “Teleology is like a mistress to a biologist: he cannot live without her but he’s unwilling to be seen with her in public”. The biologist “is unwilling to be seen with teleology”, because it implies both a logical and an epistemological fallacy: placing effects before causes, and reintroducing, albeit implicitly, purpose and will into nature, ideas long rejected by science.
Hence my surprise, as a biology student initially trained in philosophy, when I found teleology so ingrained in evolutionary theory. I was even more surprised when, upon reading The Origin of Species, it appeared that Charles Darwin’s theory had provided an explanation of adaptation rendering finality superfluous. I then recalled that the philosopher Henri Bergson shared the same surprise in Creative Evolution. Armed with my newly acquired knowledge of evolutionary biology, ecology and contemporary genetics, I revisited the 1907 book, discovering an in-depth analysis and critique of the teleological (or ‘finalistic’) thinking that underpinned the evolutionary theories of his time – an analysis and critique that felt as relevant to me today as it did then. This is what inspired me to write Du finalisme en biologie. Bergson et la théorie de l’évolution, published in October 2024 by Presses Universitaires de France.
Isn’t Bergson out of date?
The first obstacle I encountered was the view that Bergson was outdated – not just because he belonged to the last century, but also because, while defending evolutionism, he criticised Darwinian theory, was accused of spiritualism, and his fame (considerable at the time) was partly attributed to his popularity with female audiences, discrediting him among ‘serious’ (read male) intellectuals. The following – somewhat tender – verdict summarises Bergson’s place in the intellectual landscape:
Bergsonism was a gateway to fantasy, to daydreaming, it was the truancy that avenged the constraints of school, it was the mockery of intelligence, which vexes and irritates by refusing to mingle with the credulous passions of men, it was the disdain for techniques that demand effort and sustained labour. In those pages, where their adolescence was mirrored, incapable of detaching from itself, unfamiliar with tenderness, clumsy at loving, they found the magic formula for their liberation: intuition and action, and joyfully embraced the promise of life. (Lenoir 1919, 1085 my translation)
Does Bergson still have anything to offer science (and did he ever)? For those undeterred by such presuppositions, a close reading of Bergson reveals a profound epistemological critique of the evolutionary theories debated in his time, articulating (1) an internal critique denouncing the anthropomorphic ‘mechano-finalism’ implicit in these theories, and (2) an external critique highlighting the aspects of evolution these theories fail to explain.
Mechano-finalism
When Bergson wrote Creative Evolution,the Darwinian theory was not yet firmly established. He therefore examined multiple theories, both “mechanistic” (i.e., reductionist, and determinist-probabilist[i]) – Darwinism, but also mutationism – and “finalistic” (i.e., teleological) – orthogenesis and neo-Lamarckism. He rejected all of them for the same reason: they were anthropomorphic, viewing nature as a closed system akin to those studied by engineers, unaffected by time. Finalism sees evolution as oriented toward an end, with each biological event serving as a means to that end: “The doctrine of finality, in its extreme form […] implies that things and beings do nothing other than realize [réaliser] a previously drawn up program.” (Bergson 2023, 42). Mechanism meanwhile is no less anthropomorphic: considering only invariable mechanisms are at play, “[t]he essence of mechanistic explanations is, in effect, to consider the future and the past to be calculable as a function of the present, and to thereby claim that everything is given” (Bergson 2023, 40). The universe is thus reduced to an object that can only exist if one implicitly invokes the existence of a metaphysical entity capable of grasping it all (a “Laplacian demon”). Therefore, finalism and mechanism are two sides of the same coin: finalism likens the direction of evolution to an engineer’s blueprint, while mechanism, though claiming to be blueprint-free, suppresses the engineer only artificially.
The theories’ limitations
Beyond this epistemological critique, Bergson pointed out several limitations of the evolutionary theories of his time:
- They cannot explain the unpredictability of evolutionary trajectories.
- They fail to account for the regularities proper to life, which articulate directionality and creativity, the appearance of similar traits and processes across different lineages, and the increasing divergence of forms.
- In the end, they neglect what is fundamentally alive in organisms, and the distinctive causality of biological phenomena.
According to Bergson, these limitations all come from what I termed the ‘mechano-finalism’ of the evolutionary explanations that fall short when it comes to integrating the efficacy of duration. To address these shortcomings, Bergson introduced the concept of élan vital. This concept, which he sometimes qualified as an “image”, was not meant to replace scientific explanations but to open new avenues of thought. The élan vital which Bergson compares to consciousness, does not represent a spiritual principle but rather a capacity for invention, intrinsic to both life and consciousness, which stems from the recording of time. Because evolution is a history and consciousness a memory, both have the ability to draw more from themselves than they contain, to produce something new. The recording of time means that in evolution, as in consciousness, events do not repeat. Processes are not governed by deterministic causality but resemble maturation. With the élan vital, Bergson invited science to consider a model of causality distinct from that of the physical sciences, capable of accounting for both historicity and creativity.
Contrary to what dominant historiography suggests, Bergson’s critique was well received by several theorists of the Modern Synthesis, who appreciated his emphasis on creativity and a sui generis form of causality in organisms, and evolution. This is because the core of Bergson’s critique addressed key epistemological issues in Darwinian theory, issues that persisted throughout the 20th century and remain unresolved today – largely due to the ongoing (now dual-faced) finalism within the theory.
Finalism today: identifying and explaining it (without justifying it)
Finalism manifests in evolutionary theory today in two forms.
Mechano-finalism today
The first is what Bergson already criticised as implicit in all mechanistic (i.e., deterministic or probabilistic[ii]) theories, which I have referred to as mechano-finalism. In contemporary evolutionary theory, this is present in the adaptationist assumption still dominating orthodox views that all traits have a function and are perfectly adapted to it (Dennett 2014, e.g. 197). Thus, a final cause, naturalised through natural selection, is seen as explaining both the emergence and adaptation of organs. This teleological approach assumes natural selection acts in a goal-oriented manner, optimising species. Although biologists explicitly distance themselves from this view, it nonetheless influences their reasoning, especially when modelling evolutionary dynamics using optimisation algorithms. Indeed, the shift from Darwin’s natural language to the mathematical language of population genetics has paradoxically reinforced implicit teleology, leading us to forget that nature, unlike our algorithms, does not strive for an optimum. This is why, despite their precision, mathematical tools cannot predict the forms evolution will take. These forms, unlike algorithmic optima, are not pre-determined in an eternal space of possibilities that groups together all theoretically optimal biological forms. Instead, ‘good biological forms’ evolve over time, shaped by the histories of species and environments; they are not biologically possible from all eternity. For instance, molars as teeth for crushing food were not possible before the historically situated emergence of articulated jaws: while hypothetically conceivable, teeth without jaws would make no biological sense, as they would not be functional (Montévil 2019).
The finalism of intentionality
Finalism also appears in a second form: the finalism of intentionality, where biological entities are viewed as rational agents with a consciousness resembling our own, arranging the best possible means to achieve certain ends (survival and reproduction) – always striving to optimise energy expenditure. For instance, genes are described as agents seeking to maximise reproductive success through their ‘survival machines’ – organisms. In this view (and less surprisingly), organisms are also considered rational agents. For example, lions are said to kill cubs that are not their own because they ‘know’ the mother will not ovulate while lactating and want to increase their reproductive chances (infanticide interrupts lactation, allowing ovulation to resume). This teleological view seems rooted in our tendency to consider (recognise?) living beings as agents. But this finalism, while often explicit, is claimed as a metaphor, whose referent (the organisms’ agency) would not need to be elucidated. As a result, agency is reduced to a human-like intentionality without deeper analysis.
These teleological approaches are problematic for several reasons. First, they treat natural processes as if they followed human logic, without questioning this assumption. This introduces a double pride, as anthropomorphism is reinforced by anthropocentrism, assuming our intelligence is the perfect way to access reality and that our theoretical frameworks align perfectly with nature’s structure. Moreover, these finalistic views fail to explain both evolutionary oddities, such as traits that are poorly adapted or non-functional, and the unpredictability of evolutionary trajectories – despite the precision of our models, we still cannot predict evolution.
How then can we account for this unpredictability and develop new scientific models specific to evolutionary biology? Bergson’s philosophy, beyond its critical analysis, offers positive insights, potentially helping us overcome the pitfalls of finalism.
Understanding evolution without finalism: a reflection on natural history and its agents (with Bergson and beyond)
To pave the way for understanding the unpredictability and petulance of evolution, largely overlooked in the dominant theories of his time, Bergson introduced the image of an élan vital emphasising the creativity of history: in evolution, the continuity of recorded time allows for the emergence of novelties. Bergson also uses the image of consciousness: life’s evolution appears as a stream of consciousness, transformed by the passage and recording of time. This comparison is not just heuristic but holds an ontological meaning: the élan vital derives its creativity not only from history but from individual memories – the consciousness of individual living beings (Tahar 2024a). Bergson does not conceive of this consciousness present in all living beings on the human model: it is not necessarily representational or teleological (goal-directed). Instead, it is characterised by the ability to create something new from the past, to “bring something new into the world” (Bergson 2023, 211) by inserting memory into the present. “Consciousness is synonymous with invention” (Bergson 2023, 231). All organisms, through the creativity of their memory, contribute to the overall creativity of evolution.
What does this Bergsonian proposition offer contemporary biology? Bergson’s philosophy encourages us to conceive of evolution by linking its historicity to the creativity – and thus the agency – of living beings. According to Bergson, an organism’s behaviour is not merely an effect but an action: “the organism is a contingency machine, a mechanism put together to produce anti-mechanical, extra-mechanical actions. It’s a machine for producing the unforeseeable” (Bergson 2024, 82). Through living beings, the élan vital “inserts indetermination into matter” (Bergson 2023, 117). However, to fully grasp the interplay between history and the agents shaping it, we must develop a new conception of causality, specific to evolution, that addresses key challenges in contemporary biology:
- The complexity of biological interactions.
- The inability to predict evolutionary trajectories.
- The existence of regularities, despite unpredictability, which require explanation.
- The difficulty of elucidating the agency of living beings, which we nevertheless instinctively view as agents.
Rethinking causality
The complexity of biological interactions and the unpredictability of evolutionary trajectories confront us the limitations of the dominant mechano-finalist approach. They invite us to consider a non-deterministic (and even non-probabilistic) form of causality accounting for the constant changes in the space of possibilities, i.e. to the historicity of evolutionary processes. This has led several researchers to shift away from traditional physicalist and reductionist models of causality and propose one specific to biology, capable of integrating the plurality of processes and their historicity. Instead of causes inducing their effect, researchers like Stephen Jay Gould (2002, Chapter 10), or more recently Maël Montévil and Matteo Mossio (2015) among others have suggested talking about constraints – plural and historically situated – steering, rather than determining evolutionary trajectories. These constraints arise from selective pressures, morphological laws, and genealogical history (Fig. 1)[iii]. They exert a stabilising influence by channelling variability, meaning that not all evolutionary outcomes are possible (for example, given their morphology, pigs will probably never evolve wings, Fodor 2007). The concept of constraints has therefore become crucial for studying stabilisation in biological organisation. However, constraints are not only stabilising on an organisational level; they are also creative on an evolutionary scale, opening up new evolutionary possibilities. Jaws and molars are one example, but the case of the black heron’s wings is perhaps more striking (Gould 2002, 1225-1226).
The black heron’s wings
Today, the black heron uses its wings primarily as umbrellas, to shade water and improve prey tracking. However, this is unlikely to be the function for which wings were originally selected; they most likely evolved for flight and later acquired this additional function. Even before that, the initial function of feathers in the heron’s ancestors was certainly thermoregulation. How can we explain this functional overload of the heron’s wings? The answer lies in the heron’s behaviour: it shifted from using its wings mainly for flight to using them as parasols. And this new behaviour may eventually alter selective pressures – larger wings, which would not be useful for flying, might prove advantageous for catching more fish and thus be selected. Creativity here stems from the animal’s interaction with its environment, leading to a functional excess (Fig. 2), and perhaps a morphological change. This is why the causal power of constraints in evolution must be considered in relation to the activities of organisms. In fact, constraints are better understood as norms when considering their role in evolutionary trajectories: they channel evolution while opening up new possibilities because living organisms internalise them through practices specific to their unique situations (Tahar 2022).
Elucidating and integrating agency
This conception of biological causality as normativity requires us to consider the organisms’ agency and offers insights into understanding it outside teleological assumptions. It allows us to recognise the active role organisms play in evolutionary processes, without attributing representational consciousness or human-like intentionality. Agency refers to organisms’ ability to introduce indeterminacy into physical causality by producing new behaviours and assigning new meanings to the constraints – or rather norms – to which they are subject. This biological agency is neither mysterious nor abstract: it is visible through the organisms’ behavioural flexibility and inventions, explainable by their phylogenetic history and their individual memory (Longo 2019; Tahar 2024b), i.e., by the original ways in which they respond to biological situations that are always “one of [their] kind” (Bergson 2023, 32). Agency can have both ecological and evolutionary consequences, transforming the selective pressures organisms face (for more on how inventive agency impacts evolution, see Tahar 2023). Evolution’s creativity, therefore, comes, at least in part, from the creativity of living beings.
Reading Bergson today invites us to take the historicity of evolution seriously, and to connect the unpredictability of processes with the creativity of living beings. Following this path in light of contemporary biological discoveries could lead to concepts that compensate for the shortcomings of the current theory of evolution. In doing so, we can understand the processes it underdetermines, while avoiding the pitfalls of finalism.
It is time to revisit Creative Evolution and pursue the collaboration between philosophy and science that Bergson envisioned, to resolve the challenges that evolutionary theory has long ignored.
[i] Determinism and probabilism correspond to the same approach to the world, according to which nothing new can happen, since every event pre-exists at least as a probability.
[ii] For Bergson, mechanism refers to a physicalist, reductionist approach to the world, according to which, if not the necessary effect, then at least the probable ones can be determined from the cause.
[iii] All figures come from Tahar, M. Biological constraints as norms in evolution. HPLS 44, 9 (2022). https://doi.org/10.1007/s40656-022-00483-1
References
Bergson H. (2023). Creative Evolution. Trans. D. Landes. Abingdon, Oxon; New York, NY: Routledge.
Bergson H. (2024). Freedom: Lectures at the Collège de France, 1904 – 1905. Ed. N.F. Schott & A. Lefebvre. Trans. L. Lawlor. London New York Oxford New Delhi Sydney: Bloomsbury Academic.
Dennett D.C. (2014). Darwin’s Dangerous Idea: Evolution and the Meanings of Life. New York: Simon and Schuster.
Fodor J. (2007). Why pigs don’t have wings. London Review of Books, 29 (20): 19-22.
Gould S.J. (2002). The Structure of Evolutionary Theory. Cambridge: Harvard University Press.
Lenoir R. (1919). Réflexions sur le bergsonisme. Nouvelle revue française, 75: 1077-1089.
Longo G. (2019). Confusing biological rhythms and physical clocks. Today’s ecological relevance of Bergson-Einstein debate on time. https://hal.archives-ouvertes.fr/hal-02903712 [accessed on 21 April 2024]
Mayr E. (1992). The Idea of Teleology. Journal of the History of Ideas, 53 (1): 117-135.
Montévil M. (2019). Possibility spaces and the notion of novelty: from music to biology. Synthese, 196: 4555–4581.
Montévil M. & Mossio M. (2015). Biological organisation as closure of constraints. Journal of Theoretical Biology, 372: 179-191.
Tahar M. (2022). Biological constraints as norms in evolution. History and Philosophy the Life Sciences, 44 (1): 9.
Tahar M. (2023). Agency, inventiveness, and animal play. Novel insights into the active role of organisms in evolution. Special Issue on Levels of Biological Agency, Spontaneous Generations,11 (1).
Tahar M. (2024a). La philosophie animale de Bergson. Conscience du vivant, créativité instinctive et biologie contemporaine. Thaumazein Rivista di Filosofia, 12 (1): 83-107.
Tahar M. (2024b). Time and evolutionary history. In R. G. Delisle, M. Esposito & D. Ceccarelli (eds.), Unity and disunity in evolutionary biology. (pp. 551-573).Heidelberg/New York/Dordrecht: Springer, 2024.
Comments