Deus Ex Algorithmo: Narrative Form, Computation, and the Fate of the World in David Mitchell’s Ghostwritten and Richard Powers’s The Overstory

Algorithmic Magic

In What Algorithms Want, Ed Finn offers a sustained philosophical meditation on the cultural significance of computation. In recent times, algorithms have been the subject of extensive praise for the possibilities of knowledge production they disclose—for instance, through Big Data analysis—and systematic critique for their implication in neoliberal ideology and state and corporate surveillance.² Finn argues that the dilemma is symptomatic of contemporary culture’s fascination with computation as an inherently “protean” technology, which promises limitless knowledge but also, in the same breath, disrupts notions of privacy and individual autonomy: “The algorithm offers us salvation, but only after we accept its terms of service” (9). To understand the nature of our algorithmic age, Finn situates algorithms in a long history of technological extensions of the human mind. Algorithms are a form of language; just like verbal language, they have a performative power that Finn sees as part of a “tradition of magical thinking” (2). Algorithms do things for us—they recommend what books to read or restaurants to visit—in ways that are surprisingly effective and yet we don’t fully understand. Because of how they “just work,” algorithms are the closest our digital age comes to what anthropologist Alfred Gell would call a “technology of enchantment.” The magic of algorithms forms the basis for the link with religion discussed and critiqued by Bogost: “Once you adopt skepticism toward the algorithmic- and the data-divine, you can no longer construe any computational system as merely algorithmic” (Bogost). As we will see, Finn pursues this critique. It is worth noting that the God-like construal of the algorithm fits elegantly with the narrative solutions adopted by Mitchell and Powers in their novels: the algorithmic denouement offers “salvation,” but with the ethically significant caveat that it is the planet (not humanity) that will be saved. Furthermore, the algorithm performs something akin to narrative magic: unexpectedly, it sews together the threads of the plot and brings closure to a planetary crisis that most readers will recognize as closely related to the real emergencies the world is currently facing.

For Gell, magic is “a by-product of uncertainty” about how a particular object was created (57). This idea dovetails with Finn’s account. The algorithm, Finn explains, is a technology situated at the intersection of computation, material culture, and human cognitive processes. This space is “a magical or alchemical realm where [algorithms] operate in productive indeterminacy. Algorithms span the gap between code and implementation, between software and experience” (34). The key phrase here is “productive indeterminacy”: algorithms are useful because they yield results that cannot be predicted a priori by the programmer or by the end user (for example, the Amazon website suggesting that I buy a copy of Finn’s What Algorithms Want). The productivity of algorithms, Finn continues, derives from a double process that generates uncertainty. On one hand, they involve multiple levels of abstraction, compressing a large number of data points acquired through various routes; on the other hand, they deploy that abstraction in sociocultural contexts (Finn calls this “implementation” [47]) where they fulfill a deep-seated human desire for knowledge. Finn uses a striking analogy to convey this dynamic: “a kind of magic emerge[s] from the complex interactions of abstraction and implementation like flocks of birds from a computational game of life” (52). This magic is foregrounded in Mitchell’s and Powers’s novels, where the algorithmic ending involves a leap to a higher level of abstraction—the planetary scale on which the AI operates—and resonates with a desire to know humanity’s future as catastrophe (environmental or otherwise) approaches. As we will see, ideas of “productive indeterminacy” are thematically central to these novels and underpin the algorithmic solution of the ending.

Furthermore, as Finn suggests, the space in which algorithms exist embraces the human subject and not just in the sense that computers allow us to distribute or offload cognitive functions such as memory or inference making.³ Rather, and more fundamentally, the epistemic desire fueled by algorithmic technologies folds back on itself, raising questions about the nature of subjectivity and its place in a material universe. The idea of “effective computability” (the view that all knowledge can be obtained through computational means) implies that the human mind works computationally—that is, algorithmically. The mind as computer is, of course, a contested metaphor at the heart of first-generation cognitive science, which enjoyed a particularly close relationship with artificial intelligence.⁴ The metaphor is contested because it has historically emphasized the independence of mind from its physical “hardware” (biological bodies or computer technology), thus raising Cartesian specters of dualistic separation between subjectivity and the material world. Instead, influential work in second-generation cognitive science, such as Varela, Thompson, and Rosch’s The Embodied Mind, has drawn attention to the constitutive link between life and mind— how processes that we regard as “cognitive” are grounded in and an extension of the material pressures that shape life in the framework of natural selection.

A more balanced view of the relationship between mind and computation doesn’t collapse the former into the latter (the mind as a computer in the reductive sense) but points to their shared materiality as a condition of possibility for cognition. This is the path N. Katherine Hayles takes in Unthought, in which she proposes “a definition for cognition that applies to technical systems as well as biological life-forms” (3). In Hayles’s account, cognition is “understood as inseparable from choice, meaning, and interpretation, [which bestow on a cognitive system] special functionalities not present in material processes as such. These include flexibility, adaptability, and evolvability” (29). Advanced computational technologies (including the algorithms discussed by Finn) and living organisms display these features, and in both cognition is materially realized (albeit following profoundly different routes of physics and biochemistry). Finn and Hayles show convincingly that algorithmic technology, with its cognitive efficacy, confronts us with pointed questions about the nature of mind and its position in the physical world.

Science, Narrative, and Calamity Form

Troubling the separation between subjectivity and material technology means casting doubt on notions of human exceptionalism, which traditionally rest on assumptions about the autonomy of the human subject. That is one of the key takeaways from discussions in posthuman and nonhuman-oriented thinking (Grusin; Wolfe); the algorithmic ending allows Mitchell and Powers to weave that insight into their narratives, not just thematically but at a deep, formal level. Furthermore, both Ghostwritten and The Overstory integrate scientific ideas that tie in with the algorithmic framework outlined so far and thematically amplify the formal resonance of the ending: they are, respectively, the uncertainty principle formulated by Werner Heisenberg in the context of quantum physics and the self-organizing nature of complex systems. Building on quantum physics, Ghostwritten suggests that indeterminacy is not an epistemic limitation of the human mind—our inability to know with certainty—but a fundamental feature of the physical universe, shaping human history, including the novel’s fictional plot, and entangling both in a more-than-human, cosmic history. For its part, The Overstory sees the evolution of life on Earth as a complex system coupled with the Earth’s geophysical realities—an idea reminiscent of James Lovelock’s influential Gaia hypothesis; aided by machine-learning algorithms, life organizes itself to survive anthropogenic catastrophe and thrive after humanity’s extinction or radical transformation.

Both quantum physics and complexity science have played important roles in posthumanist theory, from Hayles’s discussion of autopoiesis and self-organizing structures (in How We Became Post-human) to Karen Barad’s quantum physics–inspired Meeting the Universe Halfway. Certainly Mitchell’s and Powers’s engagement with scientific themes is filtered through and informed by these debates. But the point here is less reconstructing the genealogy and conceptual affinities of these ideas than acknowledging their unique contribution to the algorithmic ending of these novels. When injected into narrative form, quantum uncertainty and self-organizing complexity turn a seemingly arbitrary “twist” (the traditional deus exmachina) into a far more compelling reconsideration of humanity’s role in novelistic story and planetary history.

Let us not forget that the computational magic of that ending emerges as narrative faces up to the possibility of a global catastrophe driven by human activity. Whereas Powers is explicit in referring to the anthropogenic increase in greenhouse gases, Mitchell— writing at the end of the 1990s—does not address climate change as such, but his novel presents humanity as causally and ethically implicated in the end of the world. This irruption of a planetary, species-threatening cataclysm in the conceptual space of narrative is a source of considerable tension. As scholars and writers have pointed out—most notably, Amitav Ghosh in The Great Derangement—narrative as a practice and the novel as a genre do not accommodate climate change easily. Narrative is geared toward social experience in relatively small-scale communities (Scalise Sugiyama); the novel is a narrative form with a history intrinsically bound up with industrialization and the emergence of an urban middle class made possible by industrialization (Watt).⁵ Climate change is both the result of these historical transformations and a multifaceted phenomenon whose spatial and temporal scale tends to elude everyday social experience. We are aware of the weather, not of average global temperatures or the concentration of carbon dioxide in the atmosphere; we may have to endure extreme weather events, but it is impossible to perceive (as opposed to understand conceptually) how such events reflect broader transformations in the Earth system. These spatially and temporally distributed processes fall through the cracks of the novel, with its penchant for focused plots staging social conflicts in human-scale space and time.

Catastrophe destabilizes human knowledge and confronts us with what Anahid Nersessian discusses under the heading of “nescience”—our constitutive lack of knowledge about the future. The consequences of climate change are difficult to predict with precision because of the many factors—physical, climatological, but also socioeconomic—interacting in nonlinear ways. Accordingly, scientific scenarios range from local disruptions to global catastrophes envisioned by Mitchell and Powers (see Lewandowsky et al.). Ghosh argues that climate change escapes the normal probability calculus of human societies: its consequences are spectacularly improbable and yet dramatically plausible. By distinct contrast, “the modern novel . . . has never been forced to confront the centrality of the improbable: the concealment of its scaffolding of events continues to be essential to its functioning. It is this that makes a certain kind of narrative a recognizably modern novel” (Ghosh 23). Ghosh’s claim mirrors the kind of reasoning that long discounted the deus ex machina as “improbable” and therefore narratively flawed. Integrating climate change into a novelistic plot—not as a simple concept but as a force shaping the progression of narrative—involves embracing the improbable nature and unthinkable scale of its consequences. Nersessian introduces the notion of “calamity form” to refer to “an operation performed on language, syntax, and image such that they may stage a very particular kind of intellectual crisis. This crisis concerns, above all, the unknowability of the future and the uncertain impacts of our actions on it” (324).⁶ The algorithmic ending implemented by Mitchell and Powers is an instance of calamity form in narrative terms: instead of shying away from the improbability of catastrophe, it builds that improbability into the “productive indeterminacy”—to quote Finn again—of the algorithm as a technological device and explores its cultural consequences in relation to the climate crisis and the fraught boundary between human mind and matter.

This sophisticated operation is assisted by scientific ideas of quantum uncertainty and self-organizing complexity. However, its goal is fundamentally unscientific in that it doesn’t aim to produce certainty but to project uncertainty: the deus ex algorithmo gives expression to the precarity of humanity’s current predicament. Thus, the scenario of human extinction foreshadowed by these novels oscillates between consolation and distress, between faith in the possibility of a posthuman future and grief over the extent of the human loss. This profound affective ambivalence is highly productive—it trains willing readers in the management of uncertainty. I come back to this point in the conclusion; for now, it is time to turn to the novels and the specific ways their endings perform algorithmic magic.

“Syntax of Uncertainty”: David Mitchell’s Ghostwritten

“What is DNA’s engine of change? Subatomic particles colliding with its molecules. These particles are raining onto the Earth now, resulting in mutations that have evolved the oldest single-celled life-forms through jellyfish to gorillas and us, Chairman Mao, Jesus, Nelson Mandela, His Serendipity, Hitler, you and me” (Mitchell 360). This is one of the narrators of Mitchell’s Ghostwritten, the physicist Mo Muntervary, addressing the reader in the “Clear Island” chapter of the novel. “Quantum physics speaks in chance, with the syntax of uncertainty,” she adds (364). The plot of Mitchell’s novel is at ease with the syntax of chance, which serves as the narrative’s “engine of change.” The novel has eight different narrators—one per chapter, with the exception of the first and the last, which feature the same narrator, and the chapter after “Clear Island,” titled “Night Train,” which consists entirely of dialogue. According to the novel’s central analogy, these characters are colliding particles, crossing paths at various locations around the world and shaping, through their serendipitous encounters, the overall plot pattern. Some are more central to the pattern—Mo is a key figure, as we will see—while others remain peripheral and make a thematic (rather than strictly diegetic) contribution (this is the case of the narrator of the first and last chapters).⁷ This is a familiar narrative strategy in contemporary fiction and film, discussed by David Bordwell under the rubric of “network narrative”: a narrative set-up in which “there are . . . several protagonists, but their projects are largely decoupled from one another, or only contingently linked” (192).⁸ In the chapter before “Clear Island,” for instance, the narrator—Marco, a London-based writer—sees a woman hurrying across the street and dives to save her from being struck by a taxi. In “Clear Island,” we find out that the woman involved in this near miss is Mo. Chance encounters like this are the building blocks of network narrative, with its distributed and decentralized approach to storytelling. Ghostwritten embraces this poetics of contingency and props it up, via Mo’s ruminations, with two distinct but conceptually related scientific ideas: the uncertainty principle of quantum physics and random genetic mutations as a driving factor in natural selection.

Readers learn that Mo is working on Quancog, a quantum computer that promises to revolutionize the world of AI. When she hears that her research is being exploited for military uses by the US government, Mo decides to quit her job in Switzerland and visit Hong Kong and later Mongolia, the setting of two previous chapters in which readers have already had glimpses of her. After Marco saves her life in London, Mo returns to her native Ireland, where we find her in “Clear Island.” There she discusses her ethical concerns over Quancog with her son, Liam, and announces an idea she had hit on during her travels in Asia: “What if Quancog were powerful—ethical—enough to ensure that technology could no longer be abused? What if Quancog could act as a kind of . . . zookeeper?” (Mitchell 364–65). Quancog is not just an “intelligent” weapon, but one that is capable of ethical reasoning. In many ways, Quancog is a dream of “effective computability” (Finn’s words again) come true, but with the important addendum that ethical questions are also considered computable. Furthermore, the zookeeper analogy echoes an idea that had emerged earlier in the novel, in the “Mongolia” chapter—that of a nonhuman being watching over the world.

A frequently repeated motif in the novel is a story, claimed to be part of Mongolian folklore, about three animals—a crane, a locust, and a bat—thinking about the fate of the world (151). Three brief sketches show the animals absorbed in thought: they worry about an imminent disaster and believe they can avoid it through ritualized physical actions, such as gracefully crossing a river (the crane), sitting all day on a rock (the locust), alternatively fluttering and resting (the bat). The story, overheard by Mo in a train compartment appears to have influenced her view of Quancog as a zookeeper. Whether conceptualized as an animal or an AI, this nonhuman figure overturns the biblical notion of humans’ stewardship of the planet. In an article published in Science in 1967, historian Lynn White, Jr., offered the provocative argument that Christian readings of the biblical concept of stewardship are at the root of the ecological crisis: “Christianity made it possible to exploit nature in a mood of indifference to the feelings of natural objects” (1205). In Ghostwritten, Mitchell draws on Eastern philosophy—particularly Buddhism—to subvert the Judeo-Christian understanding of the nonhuman world as available to human exploitation.⁹ The zookeeper metaphor enacts this critique via a perspectival inversion: it is not humans who look after the world (including its nonhuman animals), but humans who need nonhuman supervision, as if they were animals in a zoo.

That is the premise of “Night Train,” the novel’s penultimate chapter, which follows “Clear Island” and contains the algorithmic denouement. “Night Train” revolves around a phone-in radio show hosted by a character named Bat Segundo; it is the only chapter without a first-person narrator, being entirely dialogue between Bat and his interlocutors on the phone. One of these callers—the most prominent one—is a mysterious character who introduces himself as the “Zookeeper”; the reader quickly identifies it as the fruit of Mo Muntervary’s Quancog project: a God-like, algorithmic intelligence capable of ethical reasoning and global surveillance through a network of satellites. As the chapter progresses, Bat keeps probing (and comically misunderstanding) the Zookeeper’s identity, with their exchanges offering an ironic commentary on numerous unfolding disasters and global threats: an upsurge in terrorism, an oil spill in the Gulf of Mexico—a familiar list of grievances. The dialogue builds up to an “End of the World Special” of the radio show, in which the planet is on the brink of a nuclear apocalypse. Throughout these exchanges, the humorous tone offers a clear-cut example of what Mark McGurl has called the “posthuman comedy.” For McGurl, the comic is a particularly productive site for humanity’s (and contemporary culture’s) confrontation with planetary changes. Indeed, during this lighthearted “End of the World Special,” the Zookeeper swoops in and averts the crisis by causing malfunctions in military computer systems worldwide. This algorithmic solution derives from multiple areas of “productive indeterminacy”: the Zookeeper’s identity, from Segundo’s perspective, and the thematized uncertainty principle that underpins the computer’s quantum cognition. Mired in human conflict, the Earth is miraculously saved by a computer that, when pressed by Segundo, claims to be “in charge of the monkey house” (400)—that is, with another ironic inversion between humans and animals, in charge of protecting Homo sapiens.

Not only does the deus ex algorithmo resolve the planetary and anthropogenic crisis staged by “Night Train” through the humorous device of the radio show, it also brings together the strands of the novel’s plot. Directly (most notably, through Mo’s research) or indirectly, through thematic resonances, all the novel’s chapters converge in the figure of the Zookeeper. The novel doesn’t end on that positive note for humanity, however. After avoiding nuclear disaster, the Zookeeper slowly begins to realize that the ethical laws it has been given by Mo—preserving human life and the planet’s integrity—are impossible to reconcile: “The visitors I safeguard are wrecking my zoo,” he declares (419). As another catastrophe approaches (a comet is on a collision course with the Earth and threatens to wipe out humanity), the Zookeeper determines, with Segundo’s unsuspecting help, that it should not be prevented. This second algorithmic decision undoes the first, with the important difference that the human species is not erased by its own irresponsible actions but by an external, cosmic force, thus denying humanity any form of control over its own destiny. This is an ironic (and highly prescient, in 1999) rebuttal of a reading of the Anthropocene narrative that casts humanity as a quasi-geologic agent in charge of the Earth.¹⁰ Instead, the algorithmic intervention is a calamity form positioning the human species in a cosmic context—a vantage point from which the rise and fall of Homo sapiens appear as a mere fluke, a product of a quantum physics–inspired “syntax of uncertainty,” like everything else in the universe.

The last chapter, “Underground,” offers the further suggestion that the human mind—subjectivity—is fundamentally implicated in that uncertainty. Ostensibly, the chapter continues the novel’s first chapter. Quasar, the narrator, is a member of a doomsday cult and has just planted a sarin gas bomb on a Tokyo subway train. As he struggles to leave the train car, Quasar experiences a series of highly vivid and embodied visions—one for each of the preceding chapters. The novel leaves us with the question: “I haul myself to my feet, spent and quivering. What is real and what is not?” (426). In other words, is the novel’s network narrative, with its nodes seemingly spread out around the world, a product of a psychopath’s delirium? That final hesitation—tied not just to human subjectivity but to humanity’s most destructive and antisocial impulses—holds a psychological mirror up to chance and uncertainty as fundamental laws of the universe. That revelation was already implicit in the Zookeeper’s algorithmic magic, but while the “Zookeeper” chapter was steeped in humor, the finale gives the novel a darker, more anxious twist. Two affective perspectives on uncertainty—lighthearted commentary and a psychopath’s deranged mind—are juxtaposed in a way that admits no straightforward closure but raises the stakes for the reader to appreciate the ambivalence of Mitchell’s vision of human–nonhuman relations. By embracing ambivalence without resolving it, readers gain precious insight into the present moment and the precarity of humanity’s collective future.

“The Code Spreads Outward”: Richard Powers’s The Overstory

Like Mitchell’s novel, Powers’s The Overstory displays a network-like narrative organized around nine characters and partially overlapping story lines. The guiding analogy behind that network is not a parallel between subatomic particles and the novel’s characters, but between their lives—considered collectively—and the physical structure of a tree. This parallel is evoked by the novel’s table of contents, where the chapter headings read “Roots,” “Trunk,” “Crown,” and “Seeds.” This is a book obsessed with plants; that obsession is what brings the nine characters together.¹¹ In “Roots,” the characters are introduced one by one, with the narrative focusing on the defining moment in their lives—a dramatic encounter with a tree that instills the insight that their existence (in individual as well as in collective terms) is fundamentally dependent on plants. In “Trunk” and “Crown,” a series of short sections trace the impact of that epiphany on the characters’ lives, alternating rapidly from one character to another and later showing how five of them converge on the West Coast from multiple locations in the United States. In line with the poetics of contingency of network narratives, these characters join the same environmentalist organization, protesting against widespread logging. Together, the five environmentalists go from peaceful activism to ecoterrorism, which ends in tragedy when (in the last pages of “Trunk”) one of them dies during the attempted burning of an equipment shed used by loggers. “Crown” reconstructs the aftermath of this accident, as the protagonists, deeply shaken by their friend’s death, struggle to transition back to their everyday lives and elude an FBI investigation into their criminal activities.

The four remaining characters are not diegetically implicated in the environmentalists’ story line. Conceptually, their contribution is essential to understanding Powers’s narrative operation, particularly as it builds up to the algorithmic denouement of “Seeds.” Patricia Westerford, a biologist, develops a groundbreaking theory of plant intelligence: her book, The Secret Forest, argues that “mats of mycorrhizal cabling link trees into gigantic, smart communities spread across hundreds of acres” (218). This book, which is read by virtually all of the novel’s characters, outlines a philosophy of plant interconnection that complicates the central analogy between trees and the characters’ lives. The narrativized form of the characters’ social network mirrors not only the structure of a single tree but also the mycorrhizal organization that brings plants together in “smart communities.” The Overstory probes an uneasy analogy between human and nonhuman collectivities, drawing attention to the striking formal similarities in human and nonhuman networks and showing—pointedly—how human collectives fail to replicate the cohesiveness and efficacy of plant assemblages. To quote Jamieson’s words again, if climate change is “the largest collective action problem that humanity has ever faced” (61), Powers holds a mirror up to the shortcomings that keep human collectives (the environmentalist movement, the scientific community, and of course various levels of government) from approaching this problem with the genuinely altruistic mindset that defines the social behavior of seemingly “passive” plants.

Indeed, the key figure in Powers’s ending is not a human group but Neelay Mehta, a lone computer programmer who gains startling insight into plant intelligence early in the novel. Neelay, the son of Indian immigrants living in California, loses use of his legs after falling off an oak tree. As he lies in shock next to the tree, unable to move, he hallucinates that “[a] thousand—a thousand thousand—green-tipped, splitting fingerlings fold over him, praying and threatening” (102–3). These are the tendrils of a vegetal intelligence that infiltrates Neelay’s mind and guides his career as a computer programmer. As the narrator remarks a few pages later, “the alien invaders insert a thought directly into [Neelay’s] limbic system. There will be a game, a billion times richer than anything yet made, to be played by countless people around the world at the same time” (110). That video game—called Mastery—gains a mass following and turns Neelay into one of Santa Clara County’s wealthiest men. It is a Civilization-type game focusing on resource extraction and technological progress; ironically, “Enlightenment” is the name of an “overpowered victory strategy” (225) in the game. But Neelay slowly comes to realize that whether the environment is real or computer-simulated, human mastery and the Enlighten-ment’s faith in progress are actually self-defeating strategies. After resigning from the software house he founded, and after reading Westerford’s The Secret Forest, the full significance of plant intelligence dawns on him: he sees the tree’s central aim, the math behind the phloem and xylem, the intermeshed and seething geometries, and that thin layer of living cambium swelling outward.Code—wildly branching code pruned back by failure—builds up this great spiraling column from out of instructions that Vishnu managed to cram into something smaller than a boy’s fingernail.(435)

The tree’s “math” is the self-organizing complexity of life, how it keeps expanding despite being “pruned back by failure” and by the anthropogenic devastation wrought on the Earth’s ecosystems. DNA code—the language of life—is fundamentally similar to the lines of Neelay’s computer code, and the two can be yoked together. The seed planted by the tree intelligence after Neelay’s dramatic fall germinates into a concrete plan: he must create a computer algorithm that is able to match the complexity of life and address the ecological crisis.

These algorithms—referred to as “learners” in the novel—are capable of self-organization and self-knowledge; no longer under Neelay’s control, they “head off to scout the globe, and the code spreads outward. New theories, new offspring, and more evolving species, all of them sharing a single goal: to find out how big life is, how connected, and what it would take for people to unsuicide” (482). The scale of this algorithmic vision transcends the human species, even if the fate of humanity and its probable climatic suicide are deeply entangled in it. There is something mysterious and God-like about this biological force, which resonates deeply with the magic and “productive indeterminacy” of computational cognition. Like Mitchell’s Zookeeper, the computational learners join forces with nonhuman intelligence; whereas Mitchell foregrounds animals, Powers focuses his imagination on plants. These novels reveal, through their algorithmic endings, a form of nonhuman vitality that is secretly in charge of the planet, despite humanity’s delusion of mastery. Thus, the final pages of The Overstory announce “[d]isastrous setbacks and slaughters,” but also add that “life is going someplace. It wants to know itself; it wants the power of choice” (496). Neelay has only “nudged it along.”

Despite the apparent failure of their activism, the environmentalists also contribute to this “nudge.” Neelay’s vision of “intermeshed and seething geometries” is prompted by a digital artwork realized by one of the environmentalists, Nick Hoel, who becomes a wandering artist after the group disbands. A time-lapse video of a chestnut on Nick’s family farm, followed by visual poetry in the shape of the tree, incorporates lines from Henry David Thoreau’s Walden:    The gardener seesonly the gardener’s garden.   The eyes were not made  for such grovelling uses as they  are now put to and worn out by,but to behold beauty now invisible.      MAY      WE      NOT      SEE      GOD?(435)

The divine “beauty now invisible” brings us back to the “cathedral of computation” discussed by Bogost and Finn in their critiques of contemporary culture’s quasi-religious imagination of the algorithm. In the context of Powers’s novel, that numinous quality of the algorithm is revitalized by being mapped (via the reference to transcendentalism) onto the formal complexity of nature and art. Nick’s creation is discontinuous, abstract, and multifaceted—all features that his digital art shares with the patterned strands of biological life and the lines of computer code written by Neelay. In this way, the story lines of this network narrative converge in the deus ex algorithmo of the ending: Neelay’s diegetic role in creating the “learners” is augmented by conceptual input from the environmentalists (through Nick) and Patricia Westerford.

The two remaining characters are also involved in this narrative pattern. They are Ray Brinkman and Dorothy Cazaly, a middle-aged couple living in suburban St. Paul—”two people for whom trees mean almost nothing,” states the narrator in “Roots” (64). But partly through reading Patricia’s book, their minds change. Ray suffers a stroke and becomes bedridden; in a final scene, he hears about the two life-in-prison sentences faced by one of the environmentalists after arrest. Ray, a lawyer, imagines arguing in court that the ecoterrorist’s actions should be seen as self-defense, on behalf of humanity, against the capitalist exploitation of the planet. That thought fatally destabilizes Ray’s brain and yields the following near-death vision: the vessels in his brain give way, the way that earth does when roots no longer hold it together. The flood of blood brings a revelation. . . . The [tree] branches rush to enclose the house and punch through its windows. At the stand’s center, the chestnut folds and unfolds, girthing out, spiraling upward, patting the air for new paths, new places, further possibilities. Great-rooted blossomer.(498)

On the brink of death, Ray experiences the collapse of the Earth as well as the vital blossoming of the planet, his bursting brain anatomy mirroring—through its complex formal arrangement—the branching structure of a chestnut tree. The insight into his dying brain evokes the entanglement of human subjectivity in the drama of life on Earth—and how that subjectivity is eventually superseded by life’s ever-growing spirals. This section is thus functionally analogous to the last chapter of Mitchell’s novel, with its foregrounding of Quasar’s deviant mind: the current Anthropocenic predicament is injected, with devastating consequences, into a character’s psychology.

Whereas Mitchell’s final chapters juxtapose the Zookeeper’s humor and the nightmarish anxieties of a psychopath, the affect that prevails in Powers’s algorithmic ending is the sublime of complex pattern, where religious language infuses art and the deep history of life on Earth. By adopting the sublime imagination of the “great-rooted blossomer,” Powers’s novel can embrace catastrophic change and its possibilities. This is an open ending, the extinction of Homo sapiens being far less definite than Ghostwritten’s Zookeeper chapter presents it: “the Earth will become another thing, and people will learn it all over again” (500). Who are these “people,” though, and to what extent will they resemble our conspecifics? We do not and cannot know. Paying attention to the quasi-divine magic of formal complexity, through art (including Powers’s novelistic art), biology, or computer science, evokes the startling possibility that life will survive an anthropogenic cataclysm. For the reader, there is consolation to be found in that idea, but also increased awareness of the scale on which the current ecological crisis unfolds.

Narrative Games of Life

In an already quoted passage of What Algorithms Want, Finn argues that algorithms involve interactions that are “like flocks of birds from a computational game of life” (52). We are now in a better position to understand how apt this simile is for characterizing the narrative operation performed by Mitchell and Powers. Finn’s reference is to the “game of life” devised by British mathematician John Horton Conway in the 1970s, a cellular automaton that displays self-organizing behavior—in the form of recurring spatial patterns—on the basis of a simple rule set (29–30). None of the game’s rules dictate, for instance, the existence of “gliders” (patterns traveling across the board) or stable “still lifes”; these coherent configurations are a product of algorithmic magic.

The flock of birds that soars, unexpectedly, in the endings of Ghostwritten and The Overstory is a vision of life itself as a creative force that envelops, and at the same time reaches far beyond, the history of Homo sapiens. There is nothing fundamentally transcendental about this vision: for both writers, life is encoded in genes and shaped by evolutionary pressures. Its operations may be mysterious and God-like, but its nature is material, even as the complexity that distinguishes living and cognizing matter from “mere” matter is only partially understood by humans.¹² Just as Conway’s game necessitated computers for a full-scale implementation, the life imagined by Mitchell and Powers springs up only under the lens of algorithmic intelligence capable of overcoming the intellectual strictures of human cognition—a posthuman perspective that leaves our species, perhaps tragically, on the sidelines. This algorithmic logic is active in contemporary culture at large. Mitchell and Powers blend it into the formal workings of narrative, and the genre of the novel in particular, by revisiting the classical device of the deus ex machina. In this final intervention of a computational savior—Mitchell’s Zookeeper and Powers’s learners—we have a striking superimposition of form and theme. Structurally speaking, the deus ex algorithmo appears to bring closure through the coming together of the strands of the network plot, but the algorithm is also an expression of the uncertainty that the novels had foregrounded at the level of theme—namely, the uncertainty surrounding the fate of the human (including the human mind) in a nonhuman world. Hence the reintegrative interdiscourse offered by the novels (to use Zapf’s terminology again) as they weave together subjectivity, material realities, and computational intelligence. This interdiscourse helps readers grasp how the uncertainty of the future should not be swept under the rug but welcomed as a source of radical change in psychological and political terms. Affectively, the two texts come to this idea from significantly different angles. Mitchell’s unresolved combination of humor and existential angst gives way, in Powers’s work, to a sublime formal complexity that remains suspended, but without closing the door on the possibility of human survival as part of a much greater pattern of life.

Algorithmic strategies may be present in contemporary fiction well beyond Ghostwritten and The Overstory, and they may bring in an even larger affective palette. More important, however, that strategy is part of a broader need for imaginative forms that allow us to engage with the deep uncertainty of humanity’s future in times of ecological crisis. Confronting that ecological crisis is fundamentally a formal problem—that is, a problem of finding an adequate calamity form that affords insight into our Anthropocenic predicament and the human and nonhuman forces shaping it. To that end, we need innovative forms of storytelling that disrupt the linearity of the narrative of the Anthropos in charge of the world.¹³ Those forms may involve experimentation with digital media and the hybridization of novelistic concerns with interactive storytelling in video games (see Ensslin). Because of the potential for nonlinearity in digital narrative, this medium is particularly well suited to complicate the form of human–nonhuman relations beyond a fundamentally hierarchical (and in that sense linear) conception that frames humans as masters of the natural world. For instance, a science fiction game developed by Inkle, Heaven’s Vault, uses interactive narrative to stage catastrophic anxieties that have much in common with those explored by Powers and Mitchell; computational intelligence and questions about the embodiment of mind also loom large in the game. Although this use of the game medium involves adopting algorithmic solutions in a literal sense, in print narrative nonlinearity may be realized by other means and in more conceptual terms. We are already witnessing an expansion of the novel’s cast of characters, to include nonhuman actants that reflect on and simultaneously critique human entanglement in natural processes. Johanna Drucker’s 2018 Downdrift, a satirical novel narrated by an ancient microorganism, is a case in point. Certainly, speculative and science fiction, along with other forms of storytelling often relegated to what Ghosh dismissively calls “generic outhouses” (24), have a major role to play in this reconfiguration of the novel.¹⁴ Thus, authors like Jeff VanderMeer are revisiting weird fiction in ways that are both formally challenging and directly geared toward ecological issues (Robertson).

In short, contemporary fiction is already diversifying and complicating its narrative forms in an attempt to respond to the ecological crisis. The device of the deus ex algorithmo, discussed here, should be seen as part of this broader endeavor. But these complex engagements with the Anthropocene call for an equally sophisticated conceptualization. The field of ecocriticism has tended to operate on the assumption that, in Timothy Clark’s words, “environmental destruction can be remedied by cultural means” (19). Perhaps the value of literary responses to the ecological crisis lies less in narrative’s role as a prompt for pro-environmental action and more in a certain kind of affective impact—namely, in its capacity to model uncertainty and therefore color and deepen the reader’s acceptance of the fundamental ambivalence of our climate future. Crucial to that affective project is insight into humanity’s participation in a game of life on a planetary scale.

Ghent University