Sometimes it can be hard to see the Anthropocenic forest through the datafied trees.
Last fall Google launched its Tree Canopy Lab, which uses artificial intelligence to examine aerial imagery and public data on population density, land use, and heat risk to estimate tree canopy coverage across Los Angeles — and eventually other cities — with the expectation that such visualizations will inform tree-planting efforts to reduce carbon emissions and enhance public health. 1 Around the same time, the conservation nonprofit American Forests partnered with Microsoft to introduce the Tree Equity Score, which grafts together data about neighborhood income, demographics, employment, and population density, as well as tree canopy and surface temperature, in order to determine how the presence or absence of trees might map onto other forms of racial and socioeconomic inequity. This summer American Forests revealed that to achieve tree equity, the United States would need to plant 522 million urban trees. 2
Similar technologies have allowed Canopy to deploy proprietary software called “ForestMapper” to help its fashion industry clients plot out more sustainable fiber supply chains, and the U.K. tech startup Dendra to use AI and aerial imagery to identify prime spots for planting, which in this case involves drones firing seed-filled pods into the ground. Meanwhile NCX, a carbon credit exchange built on Microsoft AI, uses its own “precision forest management” tool so that landowners can “quantify the full value” of their wooded properties. As Zack Parisa, the CEO, says, “you can’t manage what you can’t measure.” 3 Microsoft’s Chief Environmental Officer, Lucas Joppa, also frames ecological management as a technological concern: our “wide range of environmental concerns … represent the world’s biggest data challenges, the world’s biggest compute [sic] challenges, and the world’s biggest algorithmic challenges.” And he offers up an even more ambitious ecological-epistemological imaginary: “Imagine if we had a planetary computer that could tell us exactly what we needed to do to protect planet earth — a system that was capable of providing us with information about every tree, every species, all of our natural resources: how could we use all that data to build a better world?” 4
Just imagine! Wouldn’t that be grand? An algorithm that could calculate how many trees would atone for the historical and contemporary inequities of urban planning and environmental injustice, that could undo processes of deforestation wrought through centuries of colonial violence, that could heal a landscape destroyed by clear cutting? A dashboard that grants us datafied dominion over all of creation? A colossal computer that would model all living systems and allow us to turn some knobs and test the impact of design solutions: a windbreak here, a forest preserve there, a pollinator garden over yonder?
As trees become data points, they are all too readily cast as easy fixes for hard problems: sylvan solutions to systemic snafus.
Or maybe not. As trees become data points, they are all too readily cast as easy fixes for profound problems. Trees as tools of carbon capture, tall timber as an instrument for sustainable construction, green barriers as sound buffers along roadways: sylvan solutions to systemic snafus. The media scholar Jennifer Gabrys argues that such approaches are efforts to frame (and tame) hard problems — wicked problems — in computational terms. Forest data sets in particular, she writes, tend to “present the problem of environmental change through … metrics that in turn legitimate specific technological interventions to meet targets for averting environmental catastrophe.” 5 In other words, these technological tools promote techno-solutionist responses to problems that are simultaneously ecological, cultural, social, economic, and political. 6
It’s easier to plant a tree — and to allow a generative design dashboard to tell you precisely where to plant it — than it is to change our individual and collective consumption habits or to muster the political will to eliminate fossil fuels. In 2019, a research team at ETH Zürich mapped the potential global tree canopy and discovered that the world could accommodate an additional 0.9 billion hectares of canopy cover, which could store over 200 gigatons of carbon; as a member of the team told The Guardian, “This new quantitative solution shows [forest] restoration isn’t just one of our climate change solutions, it is overwhelmingly the top one.” 7 Anyone can plant a tree, so the argument goes, and doing so doesn’t require difficult sacrifices or fundamental change. Hence the popularity of initiatives like the One Trillion Tree Campaign — even Donald Trump was a fan — which seem to promise that a billion people each planting a tree will cumulatively constitute a meaningful response to the climate crisis. But at best this is magical thinking, a crowd-sourced form of techno-vegetal solutionism — and as such a distraction from the large-scale, systemic transformations that are required to counter the impacts of global warming. “The notion that tree planting is an elixir for what ails the earth is as popular with polluters as it is with nations, a fact that spawned the ‘carbon offset industry,’” writes science journalist Ted Williams in Slate. 8 For Google and Microsoft, tree projects amount to little more than slickly produced campaigns to greenwash their own extractive, energy-intensive operations. 9
It’s easier to plant a tree than it is to change our individual and collective consumption habits or to muster the political will to eliminate fossil fuels.
The tech giants’ hubristic projects, with their satellite maps and interactive apps, not only serve to minimize the scale of the problem; they also risk impoverishing our understanding of trees. As Gabrys argues, such tools can render forests as mere timber stores or carbon sinks, rather than as “sites that sustain cultural narratives or indigenous cosmologies,” as systems and repositories of knowledge that resist algorithmization. 10 This is, in fact, a new version of an old problem. The rise of scientific forestry in Prussia and Saxony in the late 18th century brought about an abstracted and operationalized perspective of wooded landscapes. As the political scientist James C. Scott argues in his influential book Seeing Like a State, the forest became “an economic resource to be managed efficiently and profitably,” while “the actual tree with its vast number of possible uses was replaced by an abstract tree representing a volume of lumber or firewood.” 11
Yet there are other, more capacious ways to think through trees. By recognizing the deep roots and copious branches of arboreal thinking, we can not only appreciate trees as foundational models in intellectual and social history; we can perhaps also graft those branches onto the evolving methodological tools that we are using to repair our damaged planet.
Decision Trees, Old and New
Decision trees are the artificial dendrological descendants of interactions between ecology and cybernetics in the 1960s, 12 and today these biotechnical models are being used to scaffold decision-making in fields ranging from global policy to urban administration to environmental design. A decision tree is an algorithm that functions like a flowchart: it subjects a data set to a series of questions or if-then propositions, addressing one variable at a time in order to iteratively split the data into branches and, ultimately, into “leaf” nodes that represent a final classification or decision. If, for instance, we were using an algorithm to determine where to plant trees in a particular city, we might start at the root node, with questions about a neighborhood’s existing tree canopy. At each subsequent stage in what data scientists call “recursive partitioning,” we would ask further questions that would then allow us to successively split the data into branches with maximum “purity” or homogeneity, so that all data in a branch will belong to the same class.
Algorithmic decision trees are being used in fields ranging from global policy to urban administration to environmental design.
Modeling machine learning through decision trees can help make computational processes observable and comprehensible; yet as the sociologist Adrian Mackenzie observes, that intelligibility can also produce a “highly restricted framing of differences,” as well as an over-emphasis on the purity of its classifications. 13 (The drive for purity can lead to “overfitting,” which compromises the algorithm’s generalizable utility for other data sets.) When we’re dealing with complex phenomena informed by intersecting variables — qualities or events shaped by the entangled impacts of race, class, and gender, and by their historical and cultural legacies — it’s reductive, to say the least, to sort entities into “A” and “B” categories, e.g., sufficient tree canopy or not, environmentally just or not.
Closely related to the decision tree is the “random forest.” This is an ensemble — or forest — of decision trees in which each node works on a random subset of variables; the trees are run in parallel, and the results then aggregated. The utility of a random forest — much like the hardiness of a biological forest — depends upon diversity. Random forests can alleviate overfitting and readily lend themselves to generalizable uses; but precisely because of the “randomness” of their operations, they take longer to train, and are more difficult to interpret. 14 And as with a biological forest, random forests constitute an ecology in which many factors affect the system’s operation; it’s hard to isolate the impact of any single variable. If a random forest tells us that a particular (physical) forest grove is prime for harvesting, it’s hard to know exactly what qualities make it so — and what unaccounted-for variables or contexts might shift our decision. Clearly these operations embody specific and widely accepted models of reasoning, which can potentially obfuscate their own logics from the people who deploy them.
The transformation of trees into data might seem a modern phenomenon; but the use of trees in technologies that record data has a long history.
The transformation of trees into data might seem a phenomenon of the modern era; but the use of trees in the technologies that record data has a very long history. In the Bronze Age, the Sumerians fashioned reeds into the styli they used to make wedge-shaped marks, or cuneiform, on clay tablets. Around the same time, the ancient Egyptians were making scrolls from pressed strips of the papyrus that was abundant in the Nile delta; the red and black inks with which they wrote were composed of gum from acacia trees, mixed with charcoal and lead or copper. Later, in the second century BCE, the Chinese developed the first known paper-making technology by mashing together tree bark with hemp and rags.
The knowledge produced through and on those botanical apparatae was sometimes represented in the form of a tree — a graphic schema that facilitated cognitive work. Tree thinking is rooted deep in the history of classification, which dates back to classical philosophy — perhaps most explicitly in the Arbor Porphyriana, described by the philosopher Porphyry in the 3rd century BCE as a means of dividing genera into species; a millennium later, his widely translated writings influenced the medieval logicians who created the tree diagram — an ancestor of the decision tree.
Such diagrams served functions both practical and philosophical. Etymologies, an encyclopedia produced by Isidore of Seville in the 7th century CE, references stemma, or “twigs that lawyers draw so they can then measure the degrees of kinship”; the book then offers variations on the arbor juris, proto-family trees that served to guide moral, ethical, and legal decisions. According to scholar Jean-Baptiste Piggin, the arbor juris “has been used in Roman law systems for two millennia to determine the legal boundaries of kinship. It helps compute, for example, which kinfolk may marry without committing incest, which are exempt from testifying against an accused, or which may inherit if a person dies without making a testament.” 15
Descartes famously used the metaphor of a tree to describe the entire field of philosophy.
Centuries later, Arbor Scientiae, or Tree of Science, written in 1295 by the Spanish philosopher Ramon Llull, proposed a classification of human knowledge into sixteen categories, each represented by a tree. In contrast, many epistemological trees of the medieval period, as the librarian-historian Scott Weingart writes, “featured single roots from which all knowledge branched, based either in an ancient knowledge category or, perhaps more frequently, in God or some angelic form.” 16 Descartes’s Principles of Philosophy, published in the mid-17th century, famously used the metaphor of a tree to describe the field: “the whole of philosophy is like a tree whose roots are metaphysics, whose trunk is physics, and whose branches, emerging from the trunk, are all the other sciences.” A century later, in their celebrated Encyclopédie, Denis Diderot and Jean le Rond d’Alembert sought to reject genealogical orders altogether, arguing that maps of human knowledge are inevitably arbitrary, infinite in possibility. “The universe,” proclaimed d’Alembert, “is but a vast ocean, on the surface of which we perceive a few islands, whose connection with the continent is hidden from us.”
Still, the encyclopedists had to contain that vast ocean and its islands within their book; and so they visualized their conceptual map as a “figurative system of human knowledge,” which was then illustrated with an engraving by Jean-Michel Papillon in the unmistakable form of a tree. 17 Yet another century later, Charles Darwin reified tree logic by manifesting his theory of universal descent through the diagram of a tree of life. 18 “The affinities of all the beings of the same class have sometimes been represented by a great tree,” he wrote. “I believe this simile largely speaks the truth.” 19
All the while, as the field of scientific forestry was emerging in the same period, naturalists were amassing specimen collections that sought to capture trees as holistic material entities. Many of these new collections were contained in book-shaped repositories called xylotheques. Each “volume” in a xylotheque was a box made of a particular species of wood; its spine was covered with the corresponding bark, and its sides were composed of various tree cuts: horizontal slices of branches, cross-sections of truck, samples of sapwood, mature wood, heartwood. Within a box there would be an assortment of anatomical parts: seed capsules, roots, leaves, buds, flowers, and wax models of its fruit.
While individual trees might embody ideals of enlightenment, their dense aggregation can provoke disorientation and confusion.
Arboreal models and metaphors are widespread, yet it is crucial to consider their geographic specificity. In a recent conversation, the Danish anthropologist Janne Flora told me that in Greenland, much of which lies above the Arctic Circle — above the “tree line” — arboreal referents were rare. The Inuit of Greenland encounter trees only in the form of driftwood; trees are thus associated with wood tools, and colonial intrusion. Rather than imagining their world in arboreal terms, Flora said, the Inuit think through animals. 20 Arboreal epistemology is also affected by scale. In Western cultures, the forest has traditionally symbolized the unconscious, the Kantian sublime, the chaotic — a primordial place, the opposite of civilization. And while individual trees might embody ideals of knowledge and enlightenment, their dense aggregation can provoke disorientation and confusion. 21
As Robert Macfarlane writes, “The word ‘panic’ comes from the ancient Greek panikos, in reference to the goat-god Pan, whose presence caused sudden, irrational fear in those who entered his disorienting woods and forests.” 22 But this unsettling mixture of knowing and not-knowing is resonant with the cultural history of the forest. In a recent interview, the literary scholar Robert Pogue Harrison describes the forests depicted in Judeo-Christian mythology and literature. “Forests are never just evil,” he says, “they are evil and enchanted, they’re places where you go and get lost, but they’re also places where you go and find yourself; they’re places of sin, but they’re also places of redemption.” 23
Trees of Ethical Deliberation
If it is abundantly clear that trees have profoundly influenced ways of thinking, techniques of decision-making, modes of orientation, and methods of classification, what’s less evident are the ethics that are implied or entangled in the morphologies of a decision tree or a tree of knowledge. What values are embodied in its root node? How might the sequencing of its branches be informed by politics? What order, what ontology, is being expressed? 24
Momentous decisions have been made under the shelter of trees.
Such questions arise as well when we ponder the long history of trees (I am referring now to actual, physical trees) that have served as sites for passionate and partisan deliberation. Many momentous decisions have been made under the shelter of trees; trees have witnessed and even seeded the germination, hybridization, invasion, and, on occasion, destruction of peoples and nations. 25 As told in the first pages of the Bible, it was Eve’s choice to consume forbidden fruit from the Tree of Knowledge of Good and Evil that condemned humanity to sin, suffering, judgment, and death. Its counterpart in the Garden of Eden was the Tree of Life, a variation of which is central to many religions and mythologies. It was beneath the Bodhi Tree, the “tree of awakening,” where Siddhartha Gautama achieved the spiritual enlightenment that transformed him into “the Buddha,” and where he made the decision to devote his life to teaching. 26
Centuries ago, in what is now North America, the Great Peacemaker of the Iroquois and his disciple, Hiawatha, brought together leaders from the Mohawk, Onondaga, Oneida, Cayuga, and Seneca nations to form the Haudenosaunee confederacy. The Peacemaker directed those gathered to uproot a great white pine and thus make “a cavity in which to bury all weapons of war”; the Great Tree of Peace (Skaęhetsiˀkona), with its needles in clusters of five, has come to symbolize the union. 27 In 1681, the French trader René-Robert Cavelier, Sieur de La Salle, summoned representatives from the tribes of the Great Lakes region, including the Potawatomi, Miami, and Illinois, to propose joining forces against the Iroquois. Allegedly they gathered beneath an immense oak and there signed a treaty that would give rise to new alliances and divisions among Native tribes, the French, and the English; for centuries afterward the tree, located in what is now South Bend, Indiana, would be memorialized as the Council Oak. 28 The hardy trees that stood on battlefields of the Civil War and have survived — the Burnside Sycamore in Sharpsburg; the Sickles Oak at Gettysburg; the Manassas White Oak, among others — are today revered and protected as “witness trees.” 29 Many of their sylvan siblings south of the Mason-Dixon line served as scaffoldings for the lynching of Black people well into the 20th century — witnesses for a perversion of justice in Jim Crow America.
The surviving trees that stood on battlefields of the Civil War are today revered and protected as ‘witness trees.’
Trees have grounded decisions that led to the founding of cities and institutions and the cultivation of civic identity. Legend has it that in the early 16th century Spanish conquistadors gathered under a ceiba tree on a large island in the Caribbean and founded San Cristóbal de la Habana; five centuries later that site continues to be honored in present-day Havana. 30 Another legend recounts the spring day in 1792 when two dozen stockbrokers met under a buttonwood tree on Wall Street and signed the Buttonwood Agreement, which established the market that would become the New York Stock Exchange. In many New England towns, the elm has been accorded special status as a civic symbol; as the historian Thomas Campanella writes, charismatic specimens near town centers “served as community totems” and “secular counterpoints to the meetinghouse.” The decision, in the early 19th century, to give pride of place to indigenous elms, rather than to imported species like Lombardy poplars and ailanthus, signaled the rise of a new “arbor-nationalism.” 31
In Somalia, elders have held court for centuries under acacia trees to resolve disputes in accordance with Xeer, the customary legal system. 32 In Slovenia, the linden is widely used as a site of communal governance. And in Africa and Australia, baobab trees, which can live thousands of years, have sheltered civic meetings, open-air markets, and religious ceremonies. “Senegal has few rivers and no mountains, so baobabs sprout from the scrubby landscapes as majestic way points,” journalist Dionne Searcey writes. “Throughout history, entire communities were constructed around these trees. Baobabs serve as town halls — gathering places where municipal decisions are made, babies named and scores settled.” 33 And in the late 19th century, in Queensland, Australia, workers massed under a ghost gum tree in front of the Barcaldine railway station, an action that led to the organization of the Central Queensland Carriers Union. This Tree of Knowledge later hosted a reading of the Labour Party Manifesto, which led to the founding of the party. In 2006 the tree was poisoned and killed; the crime remains unsolved, and today there is a memorial where the tree once stood.
All around the world trees and forests are losing ground to the insidious violence of the climate crisis.
The elm and the baobab are being subjected to slower and more insidious violence. American elms have been succumbing to the elm leaf beetle since the mid-19th century; European elms have been vulnerable to the elm bark beetle since the end of the First World War. Now the dangers posed by insects and pathogens are greatly exacerbated by increasing global trade and intensifying global warming. The baobab is threatened by urbanization and population growth as well as by drought, which is driving elephants to dig into the trees to access their interior water stores. 34 All around the world trees and forests are losing ground to metropolitan development and resource extraction, and to the ramifying impacts of the climate crisis. 35
As trees fall, so do ways of life, habits of sociability, histories of political and cultural discourse; and so also do the multiple material forms of knowledge that trees embody and possess, and that make manifest the entanglement of our own lives with those of other species and the planet. “Trees,” as the science journalist Jim Robbins writes, “are giant organic recording devices that contain information about past climate, civilizations, ecosystems and even galactic events, much of it many thousands of years old.” 36 To count and scan tree rings is to learn about not only the age of individual specimens but also the evolving worlds in which they grew. That information can then be digitized and shared through the International Tree-Ring Data Bank, a public archive maintained by the U.S. National Oceanic and Atmospheric Administration. Arboreal encodings regarding historical precipitation, water sources, temperature, and light can contribute to climate modeling and to our understanding of historic methods of forest management.
Indigenous forest stewardship is informed by ethical concerns, spiritual commitments, and political ontologies.
One historic method that is drawing increasing attention in this year of record-setting wildfire is the practice of controlled burning. In an essay in this journal, Timothy A. Schuler describes how ranchers in Kansas came to understand that the indigenous practice of prescribed burns was crucial to maintaining the health of the prairie. As he writes, “They observed that annual or semi-annual burns were effective at preventing trees from encroaching and that the fire also fueled the growth of native grasses.” 37 Such recognition of indigenous stewardship underscores not only that many of our “pristine” forests were effectively managed by their early inhabitants but also that traditional knowledge is in many ways incommensurable with data-driven intellection.
Indigenous forest stewardship is often informed by ethical concerns, spiritual commitments, and political ontologies. 38 A Runa hunter in Ecuador, or a matsutake forager in the Pacific Northwest, for instance, will discover sylvan logics that are unavailable to, say, a state forestry service or a logging company. Anthropologist Eduardo Kohn, who has spent significant time with the Runa in the Ecuadorian Amazon, describes the experience as formative, leading to the realization “that anthropology can never be just anthropology. It must also include that which lies beyond the human.” Kohn ascribes agency and communicative capacity to all manner of non-human beings, including trees. “Forests are good to think with,” he writes, “because they themselves think.” 39 For anthropologist Anna Lowenhaupt Tsing, it is the mushroom that has inspired her extended exploration of “interspecies entanglement.” “If we open ourselves to their fungal attractions,” she writes, “matsutake can catapult us into the curiosity that seems to me the first requirement of collaborative survival in precarious times.” 40
You’ve likely heard of the Wood Wide Web, an underground network of mycorrhizal fungi that connect the roots of plants and trees.
There has indeed been much discussion in recent years about arboreal, fungal, and floral intelligences. Scientists have learned that trees respond to numerous stimuli: energy sources, mechanical impulses, soil structure, humidity, temperature, atmospheric gas composition, and various biotic signals, including the presence or absence of nearby plants, predators, or diseases. 41 Plants can remember particular stresses, such as drought, or extreme heat and cold, or exposure to radiation. And they have evolved methods of interspecies communication. You’ve likely heard of the Wood Wide Web, an underground network of mycorrhizal fungi that connect the roots of plants and trees, transferring water, carbon, and nutrients, sending chemical warnings of attack, and either nurturing desirable neighbors or sabotaging invasive ones. The ecologist Suzanne Simard — who was the inspiration for a character in Richard Powers’s novel The Overstory — is widely credited for her germinal research in the study of forest communication networks. Almost a quarter century ago, Simard published a paper in Nature that, as a recent commentator wrote, “upended a dominant scientific narrative — that competition is the primary force shaping forests.” 42
In this light it is clear that recognition of interspecies networks has epistemological and political implications: it suggests the potential for cooperative thinking, for the co-production of knowledge, as alternatives to prevailing models of dominance, competition, and ownership, as reflected in everything from intellectual property to seed patents. 43 Yet even today there is continuing resistance; as Simard lamented, in a recent interview, “The people who are devising forest policies — and then, ultimately, practices — are still stuck in the old model. … We’re still trying to grow trees that are tall and dominant because they’re going to provide commodities in the future.” 44
Traditional healing, historical memory, indigenous stewardship, subterranean communication networks: these are forms of arboreal agency and sylvan intelligence that can lead us to ways of “thinking with trees” that are more complex and nuanced, more grounded, more open-ended, messier and humbler, than the data-driven processing of decision trees and random forests. 45 And recently some extraordinary art projects have been exploring the integration of arboreal epistemologies with Anthropocenic ecologies.
Last year the Dutch design studio Formafantasma launched a gorgeous show, Cambio, at the Serpentine Galleries in London, on the “governance of the timber industry … and its tentacular expansion across the globe.” 46 I regret that the pandemic kept me from visiting, but the experience of the exhibition online, with its many artifacts and videos, was moving. In one video, “Seeing the Wood for the Trees,” the artists have staged a kind of trade show in the middle of the forest, juxtaposing machine vision and green screens and performance dashboards with archival documents that record colonial legacies; the result is a powerful argument, blending datafied and embodied ways of knowing, for community forest rights.
For the past couple of years, artist Julia Christensen has been collaborating with the Innovation Foundry at NASA’s Jet Propulsion Lab on a project called The Tree of Life. Christensen and her scientist-colleagues are using sound to manifest the myriad ways of knowing and being that trees represent. Together they’ve constructed a network wherein a CubeSat satellite and sensor-studded trees will “sing to each other” about their “operational conditions” for 200 years — a period well beyond the traditional timeframe for technological decay — with the trees functioning as “living terrestrial antennae.” 47 The challenge of stewarding a satellite for two centuries, and of imagining how to translate data sets such that the tree and satellite have meaningful things to say to one another, evokes the complexity of interspecies communication recorded by Simard and other ecologists. You might say Christensen has built a biotechnical version of a mycorrhizal network.
What might a forest say to a satellite — or to us?
What might a forest say to a satellite — or to us? Perhaps its communication is more ambient and affective than semantic. Earlier this year, amidst the long, dark days of the pandemic winter, I happened upon a sound archive, tree.fm, that combines photographs and audio recordings to allow you to view and listen to forests around the world; the site’s tagline is “listen to a random forest.” I tweeted about it, and other cabin-feverish folks seemed appreciative. “I’m so starved for other places, this almost made me cry,” was one response. Another responder did cry, at least in emoji. Yet my caption, quoting the tagline, seemed to mislead and even frustrate commenters from the tech world. Apparently, they were expecting to encounter an algorithm and were disappointed to be immersed instead in the rustling leaves of Atatürk Arboretum, in Istanbul, or the birdsong of Bitza Nature Park, in Moscow, or the rushing water of a coastal forest, in Ibiza. Here the randomness wasn’t about parallel computational processing; rather, it was about a kind of grounded sublimity, a sensation of poetic disorientation as one forest after another materialized onscreen.
The popularity of hiking surged during the pandemic. For centuries forests have redeemed us humans. Now we must reciprocate that redemption.