December 11, 2018

Bruce Clarke

—Drafted just prior to my meeting Lynn Margulis personally in the fall of 2005, this is an early, unpublished synopsis of the conceptual connections that have oriented my research on Gaia theory over the last decade.

From Symbiosis in Cell Evolution to What is Life? and Acquiring Genomes, American biologist Lynn Margulis and her son and co-author, science-journalist Dorion Sagan, have put systems-theoretical concepts to work.[1] As Margulis said in an interview published in 1995, “My work in symbiosis comes out of cytoplasmic genetic systems.”[2] The Gaia theory introduced by British geochemist James Lovelock, co-developed with and co-promoted by Lynn Margulis, is now being mainstreamed in academe and beyond under the label Earth systems science. The aim of this paper is to document how the systems concept that informs Margulis and Sagan’s work is specifically informed by what is termed second-order systems theory, and to explain why that is important.

To begin with, on the scientific stage, Margulis has pointed all of her work toward a major overhaul of classical Darwinism. Her aim in particular has been to surpass neo-Darwinian theory. In the neo-Darwinian "modern synthesis" of Darwinian theory with Mendelian genetics, through a residually social-Darwinist rhetoric of competition, adaptation, and fitness, speciation and biodiversity are largely attributed to the natural selection of random genetic mutations. Margulis counters this evolutionary orthodoxy with the theory of symbiogenesis. On her account, speciation—the emergence and elaboration of viable biological novelty and complexity over evolutionary time—arises not just occasionally but primarily from the permanent mergers of discrete pre-evolved genomes. As new species arise, the bifurcating limbs of the evolutionary tree not only diverge but also converge. Different and previously autonomous living systems discover viable symbioses leading eventually to the merger—or, as in the case of the eukaryotic cell, the permanent cohabitation—of their genomes, and to the natural selection of that no-longer-improvised consortium.

Margulis and Sagan mount a trenchant critique of neo-Darwinism on a conceptual scaffolding itself constructed, I will argue, from a merger of cybernetic systems theories. Their account of symbiogenesis by the merger of genomic structures to create new living systems, exploits descriptions of “structural coupling” and “natural drift” already present in the discourses of cybernetics and autopoiesis. We can hear that biocybernetic note in another passage from the 1995 interview, published in the volume The Third Culture: Beyond the Scientific Revolution. In symbiogenesis, Margulis explains:

The new genetic system—a merger between microbe and animal cell or microbe and plant cell—is really different from the ancestral cell that lacks the microbe. Analogous to improvements in computer technology, instead of starting from scratch to make all new modules again, the symbiosis idea is an interfacing of preexisting modules. Mergers result in the emergence of new and more complex beings. I doubt new species form just from random mutations. (134)

Additionally, Margulis and Sagan stress a major motif running throughout cybernetic theory, the mutual supplementarity of systems and their environments in the production of biotic self-organization.[3]In Margulis and Sagan’s discourse this system/environment holarchy emerges in the close interplay of autopoiesis and thermodynamics. Systems-theoretical idioms allow Margulis and Sagan to extend the emphasis on symbiotic interconnections within living systems to the interrelations or structural couplings amongliving systems and their various environments, which are always compounded of other systems, living and nonliving, within a larger biotic/ecosystemic and material-energetic matrix. In this manner, in what might be called a neocybernetic synthesis, their narrative of life's unfolding over geological and evolutionary time merges physical, biological, and, I will also argue, metabiotic systems theories.

Let me get more specific. Although Margulis is well-known for her investments in and contributions to Gaia theory, she does not adhere to the strong version of the theory, in which the biosphere as a whole is to be considered an organism in its own right. Often in Lovelock's presentations, Gaia come forward as global holism with a vengeance: the living Earth in its totality is presented as a singular living being, of which all other unproblematically-denominated organisms, from bacteria to multicellular metazoans, are the "cells." Margulis has specified that she herself adheres to a modified or "weak" version, which is really anything but weak, but which allows that "Gaia" is not literally an organism, but literally a system, the global ecosystem that emerges from the sum of all local ecosystems.

The system concept promotes the observation of emergent phenomena in coherent ensembles exhibiting behaviors unseen in any of their parts taken alone. The system concept implies a bounded ensemble, at once distinct from an external environment that surpasses it and internally "autonomous" and "self-regulatory," that is, having its own systematicity. Living beings, then, areproperly referred to as systems, as in this statement fromWhat is Life?— "Minimal life on Earth today is a system, a minute membrane-bounded sphere, a bacterial cell" (58). The difference is that the system concept is not limited to living beings per se, but is properly applied as well to coherent and bounded nonliving ensembles of material-energetic components—for instance, physicochemical or meteorological systems—and to metabiotic or higher-level ensembles of living systems and their environments, aptly called naturalsystems—on the one hand, populations, biological communities, and their ecosystems; and on the other, neural, psychic, and social systems arising from the systematic integration of multiple levels of cognitive—or, in their idiom, sentient—components. Margulis is surely right, then, to maintain that, while Lovelock is welcome to his metaphors, properly speaking Gaia is not an organism per se but a metabiotic natural system: 

Planetary physiology—Gaia—is the result of the interaction of innumerable beings.
. . . Gaia is symbiosis seen from space . . . . the global autopoietic system. (What is Life? 189-90)  

In this last formulation, we see in Margulis and Sagan another, less-noted but equally maverick and crucial commitment to yet another controversial theory, Humberto Maturana and Francisco Varela's concept of autopoiesis, the key term in their theory for a formal criterion of living organization. Under her own name, Margulis began advancing the claims of autopoiesis at least as early as her wonderfully audacious 1990 essay “Big Trouble in Biology: Physiological Autopoiesis versus Mechanistic Neo-Darwinism,” republished in Margulis and Sagan’s 1997 volume Slanted Truths: Essays on Gaia, Symbiosis, and Evolution. In fact, Gaia and autopoiesis fit together as interlocking, macro- and micro- modes of biological systems theory. As we noted, Gaia captures the global totality of the biosphere as the ecosystem of all ecosystems, and biological autopoiesis defines the minimal formal requirements for living systems, beginning with the cell. The bacterium is "the smallest autopoietic structure . . . the minimal unit that is capable of incessant self-organizing metabolism" (What is Life? 78).[4]

Now, the work of Maturana and Varela is deeply informed by cybernetic systems theory and its development since the 1950s of theories of self-organization. The concept of self-organization provided an initial framework within which to explore classical cybernetic analogies between artificial and natural systems.[5] But, for Maturana and Varela the crucial difference is that both artificial and natural systems cannot be said to be self-repairing and self-maintaining. They presented the concept of autopoiesis precisely as a means to distinguishbiological from technological organization: “Our proposition is that living beings are characterized in that, literally, they are continually self-producing. We indicate this process when we call the organization that defines them an autopoietic organization” (Tree of Knowledge 43). In its conceptual genesis, autopoiesis is set forward to emphasize the autonomyof living systems from their environments (including all other systems): “an autopoietic system,” Maturana and Varela write, “pulls itself up by its own bootstraps and becomes distinct from its environment through its own dynamics” (Tree 46-47). This description is motivated by the issue of control central to classical cybernetic concerns, and responds to prior evolutionary notions of adaptation as implying a form of environmental control over living systems. For Maturana and Varela, the notion of autopoiesis represents an effort to balance the conceptual scales by stressing the priority of life’s operational autonomy over its dependency on environmental contingencies. In Varela’s interview in The Third Culture, he succinctly summarizes the main lines of the concept:

Autopoiesis attempts to define the uniqueness of the emergence that produces life in its fundamental cellular form. It’s specific to cellular level. There’s a circular or network process that engenders a paradox: a self-organizing network of biochemical reactions produces molecules, which do something specific and unique: they create a boundary, a membrane, which constrains the network that has produced the constituents of the membrane. This is a logical bootstrap, a loop: a network produces entities that create a boundary, which constrains the network that produced the boundary. . . . It doesn’t require an external agent. . . . It is, by itself, a self-distinction. It bootstraps itself out of a soup of chemistry and physics.[6]

If we now turn back to the work of Margulis and Sagan and look at the way that they introduce the concept of autopoiesis into their volume What is Life?, we note that their stress is equally on the environmentalside of the autopoietic-systemic coin. Their emphasis on this side of the system/environment dyad does several important things.  It promotes the integration of autopoiesis into the Gaia paradigm, and by doing so, it lifts the concept of autopoiesis away from its cellular foundations and toward its metacellular, and ultimately, metabiotic natural extensions. I would argue that Margulis and Sagan’s autopoietic spin on the Gaia concept is precisely metabioticin a fashion analogous to Niklas Luhmann’s extension of biological autopoiesis to both psychic systems—the autopoiesis of consciousness—and social systems—the autopoiesis of communication.  Both psychic and social systems are natural yet metabiological: these higher-order systems self-organize their forms within and only within the medium of living systems, but at the same time, psychic and social “selves” are not organic but systemic—emergent, co-evolving forms of virtualautopoiesis spun off from the literal metabolic looping of living systems.[7]

Now, at the outset of their exposition of autopoiesis, and in the spirit of Maturana and Varela’s original presentation, Margulis and Sagan use that concept to counter the popular overemphasis on DNA, what Donna Haraway calls the “gene fetish,” the spurious pseudo-religious presentation of DNA as comprising the “essence of life.”[8] Biological autopoiesis provides a crucial criterion of distinction between living systems and the non-living macromolecular structures that carry the information for the assembly of those systems. So we read in What is Life?:

DNA molecules replicate but they don’t metabolize and they are not autopoietic. . . . Viruses . . . are not autopoietic . . . they do not metabolize. . . .  Viruses do nothing until they enter an autopoietic entity. . . . The smallest cells, those of the tiniest bacteria . . . are the minimal autopoietic units known today. . . . Viruses mutate and evolve; but . . . the cell is the smallest unit of life. (18)

That is to say, blueprints may render carbon copies of themselves, but in themselves they do not make or maintain the building for which they state the specs. DNA molecules are as dependent upon living cellular systems for the fact of their replication leading to actual living consequences as those systems are dependent upon DNA for the form and transmission of their operational dynamics.  In other words, Margulis and Sagan stress the interlocking of genetic information and metabolism in living processes, in a manner that literally fleshes out the cybernetic formalism of the “circular organization” recalled by Varela and set forth in his and Maturana’s original description of cellular autopoiesis.

However, Margulis and Sagan immediately go on to consider that

metabolism . . . requires energy. In accordance with the second law of thermodynamics, autopoietic self-maintenance preserves or increases internal order only by adding to the “disorder” of the external world. . .  All living beings must metabolize and therefore all must create local disorder: useless heat, noise, and uncertainty. This is autopoietic behavior, reflecting the autopoietic imperative required for any organic being that lives, that continues to function. (19)

It is the case that considerations of thermodynamic entropy, of living systems as bound up with material-energetic environments ruled by the second law, are absent from Maturana and Varela’s discussions. And in general, Margulis and Sagan put more emphasis on autopoietic behaviorthan on autopoietic form. They do not elaborate on the circular or self-referential aspects of the concept, what Varela termed in the 1995 interview “the entire loopiness of the thing” (213).  Put another way, they do not stress the constructivist side of the autopoietic idea, the specifically second-order cybernetic emphasis on the self-referential nature of all observing systems.  As Maturana and Varela point out in The Tree of Knowledge, “Behavior is not something that the living being doesin itself (for in it there are only internal structural changes) but something that we [as observers] point to” (138). But that is just to say that Margulis and Sagan focus their attention on, so to speak, the external politics of autopoietic relations, taking as their concern the flip-side of the autopoietic loop—in that the boundary formed and maintained by any autopoietic entity both cleaves it from and cleaves it tothe environment from which it sets itself apart.

Or again, Margulis and Sagan put the micro-matter of cellular autopoietic autonomy into a global frame, and in that process overcome Maturana and Varela’s tendency to stress internal autonomy over external conditions.  In the following remark, the “autopoietic view” given is one that has already been expanded by Gaian considerations: “The autopoietic view of life differs from standard teachings in biology. Most writers of biology texts imply that an organism exists apart from its environment, and that the environment is mostly a static, non-living backdrop.  Organic beings and environment, however, interweave” (19). And from this Gaian perspective of total biospheric interconnectedness, a level to which Maturana and Varela never take the concept of autopoiesis, the paradox of autopoietic borders that are at once both thermodynamically open and organizationally closed returns in the form of a global environmentthat is also and in itself, not precisely a livingsystem but a higher-order autopoietic consortium of living autopoietic systems.

In the section of What is Life? called “The Autopoietic Planet,” we read

The biosphere as a whole is autopoietic in the sense that it maintains itself. . . . The combined activities of autopoietic surface life have led to an atmosphere in which oxygen has been maintained at levels of about 20 percent for at least 700 million years. . . . As an autopoietic system, Gaia therefore shares an essential quality with individual living systems. (20)

The recognition that there can be natural yet nonvital—precisely, metabiotic—modes of autopoiesis—that autopoiesis describes a general mode of self-referential systems functioning, oneform of which is living systems—therefore, solves the central problem with the strong form of the Gaia hypothesis.  Margulis and Sagan give voice to the standard objection to Lovelock’s version: “Some evolutionary biologists have suggested that Earth life in its totality cannot constitute a living body, cannot be a living being, because such a body could only have evolved in competition with other bodies of the same sort—presumably, other biospheres” (23). But through a nuanced revision of the autopoiesis concept, one can grant that objection while at the same time overcoming the limits of its neo-Darwinist logic. Through the more generous systems-theoretical reading of the autopoiesis concept, one retains the really important claim underlying Gaia, that the biosphere in its entirety is systematically co-evolved from, integrated into, and co-dependent upon all the living systems that comprise it, as they are from, into, and upon it.

Margulis and Sagan continue: “in our view, autopoiesis of the planet is the aggregate, emergent property of the many gas-trading, gene-exchanging, growing, and evolving organisms in it” (23).  And from this form of the Gaian perspective, one overcomes the autonomy fetish that represents the conservative residuum in Maturana and Varela’s original exposition of autopoiesis. As Margulis and Sagan stress, the upshot of the merger of symbiogenesis and Gaian autopoiesis is the recognition that “Organisms are less self-enclosed, autonomous individuals than communities of bodies exchanging matter, energy, and information with others” (23).  That is to say, from the eukaryotic cell up, there is really no such thing as pure individuality.  And the sym- in symbiosis stands not just for the merger of different living species, but for the merger of life altogether with extrabiotic thermodynamic system/environments and metabiotic elaborations of psychic and social observing systems, which are themselves built up from life’s own self-referential sentience. I will conclude for now with a passage that, in drawing these various conceptual lines together, lays out the implicitly second-order systems-theoretical framework of Margulis and Sagan’s biological discourse:

Mind and body, perceiving and living, are equally self-referring, self-reflexive processes already present in the earliest bacteria. Mind, as well as body, stems from autopoiesis. . . . Changing to stay the same is the essence of autopoiesis.  It applies to the biosphere as well as the cell.  Applied to species, it leads to evolution. (31)    



[1] Lynn Margulis, Symbiosis in Cell Evolution: Microbial Communities in the Archean and Proterozoic Eons, 2nd edition (New York: W. H. Freeman, 1993); Lynn Margulis and Dorion Sagan, What is Life? (Berkeley: University of California Press, 2000); Margulis and Sagan, Acquiring Genomes: A Theory of the Origins of Species(New York: Basic Books, 2003).

[2] John Brockman, The Third Culture: Beyond the Scientific Revolution (New York: Touchstone, 1995), 135.  The interview with Margulis is available elsewhere on this site.

[3] See the classical presentation of self-organization in Heinz von Foerster, “On Self-Organizing Systems and Their Environments,” in Self-Organizing Systems, ed. Marshall C. Yovits and Scott Cameron (New York: Pergamon Press, 1960), 31-50; rp. Heinz Von Foerster, Understanding Understanding: Essays on Cybernetics and Cognition(New York: Springer, 2003), 1-19.

[4] By integrating the discourse of autopoiesis into her development of the theory of symbiogenesis, Margulis both underscores the systems-theoretical paradigm of her work and participates in Maturana and Varela’s own polemical resistance to the neo-Darwinian synthesis.

[5] That is, informatic and computational systems on the one hand, and biological and neurological systems on the other—as in John von Neumann’s famous comparisons of the neuron to the vacuum tube: both can be said to operate as switching mechanisms within information-processing networks.

[6] Third Culture 212. Available on-line at:

[7] Luhmann’s central statement is Social Systems, trans. John Bednarz, Jr. with Dirk Baecker (Stanford: Stanford University Press, 1995). For commentary on Luhmann and other sociological systems-theoretical perspectives, see Felix Geyer and Johannes van der Zouwen, ed., Sociocybernetics: Complexity, Autopoiesis, and Observation of Social Systems (Westport, CN: Greenwood Press, 2001).

[8] In this passage Haraway ventriloquizes “the structure of denial in technoscientific fetishism”: “’Scientific mapscould not be fetishes; fetishes are only for perverts and primitives. Scientific people are committed to clarity; they are no fetishists mired in error. My gene map is a non-tropic representation of reality, that is, of genes themselves.’” Donna Haraway, Modest-Witness@Second-Millennium: FemaleMan©Meets Oncomouse™: Feminism and Technoscience(New York: Routledge, 1997), 137.

Bruce Clarke

Bruce Clarke is the 2018-19 Library of Congress Chair of Astrobiology and the Paul Whitfield Horn Professor of Literature and Science in the Department of English at Texas Tech University.

Related Posts