It is becoming increasingly clear that there is no such thing as a biological individual. All organisms are intricately nested collectives: networks of relationships between cells and microbes that make it impossible to say where one “individual” starts and stops. Humans are no exception: We carry around more microbial cells than we do our own.1 Many of our constituent microbes themselves have symbiotic relationships. Moreover, the most dramatic moments in the history of life are the product of symbioses. From this perspective, organisms are ecosystems, and must be understood in the language of ecology.2 Three properties of ecosystems are relevant to this piece. First, ecosystems, like many biological systems, are emergent phenomena, meaning that they are more than the sum of their parts and so cannot be understood when disassembled into their component pieces. Second, ecosystems are dynamic non-linear systems, meaning that effects are in no way proportional to causes. Small changes can bring about large effects which can only be observed in the system as a whole. Third, ecological perspectives are inherently non-reductive, concerning as they do relationships, not things. Indeed, ecosystems are not things at all, but complex networks of processes. As Art Buchwald quipped, “The best things in life aren’t things.”
Despite a growing understanding that ecologies, environments, and cultures can’t be reduced to their “parts” without inflicting severe conceptual damage, the dominant paradigms of today are reductive. There may be a Western human tendency to conceive of, and thus to make, the world in this way. In the words of the psychiatrist and scholar Iain McGilchrist, “An increasingly mechanistic, fragmented, decontextualized world, marked by unwarranted optimism mixed with frantic paranoia and a feeling of emptiness, has come about, reflecting, I believe, the unopposed action of an unopposed left hemisphere.”3 By accepting the premise that the world is singularly knowable, do we exacerbate a human tendency to reduce? Units, time, things, binaries, cause and effect, outcomes, goals...
Here, we offer alternative vantage points from which to perceive and imagine. Writing as a trio, we have used collaboration tools such as email, Box, Slack and Google Docs. Despite these platforms’ cleverness, it would have been more efficient for one person to write. But the process of coming together to talk— and the act of writing together, asynchronously—was important to enact the relationships between our voices, interested as we are in the journey as much as the destination. Or, to borrow from McGilchrist again, the manner matters as much as the means. We draw together examples that illustrate plural, interconnected processes in the hope that they will sprout in non-reductive directions. Using song and relational spaces as our guides, we will explore examples from the human world (design) and the natural world (biology) to nurture questions about non-reductive worlds that may alter the songs we live by.
In the essay The Little Birds Who Won’t Sing, the author G.K. Chesterton finds himself reflecting on the carvings on a building in Flanders. In the style of the early Middle Ages, many of the carvings depict humans doing things. Sailors tugging at ropes, others reaping, others “energetically pouring something into something else.”4 Chesterton puzzles over why they all appear to have their mouths open. Was this an accident? Did it mean anything at all?
Chesterton comes to the conclusion that–of course–they must all be singing. Not just any old songs, but work songs.
In many contexts, work songs are more than songs sung while working. They do more than help pass the time spent in toil. The songs themselves do work. A good example are the sea chanties used by sailors to synchronize their hauling motions. For many jobs, synchronous action is essential. In guiding the work’s motions, the songs themselves are instrumental, acting in the world as much as any hardware. These songs are themselves tools of the trade; relational technologies; essential soundware without which the work cannot be undertaken. They are both work songs and working songs. Songs that both inform, and are informed by, the work. Songs that entrain processes in time. Songs that extend—in time and relationship—our intelligence and ability to think and know.
Chesterton arrives at a puzzle: “If reapers sing while reaping, why should not auditors sing while auditing and bankers while banking? If there are songs for all the separate things that have to be done in a boat, why are there not songs for all the separate things that have to be done in a bank?”5
For much of the rest of the piece, Chesterton imagines the songs that modern bankers and post office clerks might sing; workers not engaged in physical labour nor with obvious need to synchronize their movements. Out of these incongruities, Chesterton is able to tease some considerable humor (“Up my lads and lift the ledgers / sleep and ease are o’er. / Hear the Stars of Morning shouting: / ‘Two and Two are four.’ ”6).
His musings touch on bigger questions, however. What kind of world would it be where bank clerks had to work in time and in tune with each other? Where their actions were organized by a sound that they collectively made, rather than a clock? It is hard to imagine a bank clerk needing to synchronize the strokes of her pen; the dotting of ‘i’s or crossing of ’t’s. A clerk on her own should accomplish these tasks just fine, without acting together in time with other clerks. Certainly today, to synchronize may be hard, given that tasks of clerks are increasingly automated and the role of bank employees is to serve the unique conversational needs of individuals. Compare this with a team of sailors hauling up an anchor, which no one sailor could accomplish alone. The task can’t be broken down to any smaller unit of activity. Tens of sailors must work at once, in the same place, and moreover in time with each other. The chantey is what makes this possible. The chantey, emerging from the voices of the sailors, causes their actions to cohere. The actions make no sense without the song.
This example provides a metaphor for some of the themes we want to explore in this piece. Songs, like all music, are processes in time, not things. Work songs are often sung in groups, the song is the sound of the relationships between the working members of the group. The accomplishment of the work emerges from the collective actions and voices; it cannot be completed by any one of the parts alone. It is more than the sum of its parts. The song is a technology of relation. The work, guided by song, resists reduction.
As Chesterton observes, to actually sing songs while being a great financier makes no sense, “Because the essence of being a great financier is that you keep quiet.”7 Nonetheless, financiers, like scientists, medics, programmers, and almost all other professions have tacit songs, unsung and unsingable songs that guide their work and shape their actions, as a sailors’ chantey might. Just as a sailors’ chantey allows their actions and work to cohere in time and place, so do the unsung songs that guide the actions of workers in silent professions. These tacit songs might be thought of as the expression of paradigms, uncodified norms, guiding mythologies or taboos that direct and coordinate work and enquiry.
In our view, it is frequently these tacit songs that underwrite our tendency to reduce. In resisting reduction, we are resisting songs that guide us towards reduction. Our proposition is that resisting reduction might be as simple (!) as rewriting the songs that underpin a field of enquiry or sphere of thought in such a way that they guides us away from reduction.
As an example, we want to discuss the field of “generative design.” GD is a field of design pioneered by a company called The Living, an organization that sits at the “intersection of biology, culture and synthetic biology.” GD is based on the premise that human preconceptions fetter the processes of design. To this end, GD seeks to avoid the preconceptions implicit in any human design by using genetic algorithms to evolve new solutions to design briefs. For Lorenzo Villaggi of The Living, the black box within which his generative algorithms work is free from the limits of human “preconception… and simple intuition.” These ever changing states of preconception – and worse – of intuition, are uncomputable engineering nightmares. Villaggi posits that without human preconceptions, genuinely original solutions are free to arise: left-field notions too strange for a human to dream up. Humans, the argument goes, are too bound up in webs of habit and meaning to handle true novelty on a day to day basis. Villaggi hopes to design using “goals” and “constraints rather than form.” This involves developing a viable co-design practice between human and computer.
GD is evolving rapidly. Many big players like Autodesk (parent company of The Living) are starting to invest in its future, and it may become a prominent feature in the “design landscape of tomorrow.” Given that its corporate upbringing will affect the manner of its maturation, we think it worthwhile to speculate on its principles and possible futures.
An example: GD was applied to generate the stand layout for the Autodesk University Conference 2017. The goal was to radically alter the layout for the conference to discover “new and unexpected design solutions.” First, the AI was given specific constraints for the size and proximity of booths, along with fixed geometric constraints like columns. Next, avenues were plotted through the space (as vectors from edge to edge) to subdivide it into macro regions. Streets emanate from these avenues to form meso-regions, and then these are subdivided into micro-regions to create zones for the booths. These regional divisions are able to change as the AI goes to work. In the first phase, the 30,000 design solutions were then ranked according to two metrics: “buzz” and “exposure.” In the second phase, design solutions were assessed against further metrics. The final space resembles a funneling system of grids, arranged around wider access avenues with a large central core. A sub-network of streets allows for a cyclical journey through the network of booths and pavilions aside from the straight-line approach to the core pavilion via the avenues.
In our opinion, GD presents two interesting problems. First, preconceptions are important. They create an architecture of the specific, based on a vast body of amassed collective human knowledge—cultural muscle memory—that is intrinsic to design; a landscape of interconnected ideas that act, or silently sing, through time and space. This web of ideas includes history (precedent and temporal context), tested methodologies (symmetry, balance, proportion perspective, scenography), context (cultural, political, urban, spatial), humor (juxtaposition, quotation), and so on. Second, preconceptions are unavoidable and are inevitably baked into AIs. As we will argue, GD is founded on a limited set of evolutionary models—the reductive work songs of 20th century neo-Darwinism—that don’t reflect the full scope of evolutionary processes in the world.
GD is underpinned by genetic algorithms. These are based on evolutionary processes or, as we want to argue, a particular model of evolutionary process. Solutions to the design problem are iteratively generated. Each solution within each of these generations is evaluated according to a fitness function, which stands in for natural selection, performing an editing role. Novelty is introduced at each generation by random mutational changes and by recombination; the latter a process simulating sex, in which features of each parent are randomly shuffled. Given the various starting constraints, and fitness functions, solutions are free to evolve.
Within the frameworks of GD, a particular understanding of evolution, or theoretical song, is mobilized to work in the world. This model of evolutionary process, a particular paradigm, or school of evolutionary thought, informs algorithms that generate design solutions. These solutions result in the development of new structures, housing and directing human life in new ways. In the hands (or algorithms) of GD, a non-material set of concepts, a body of evolutionary thought, is made to act, just as a work song acts, in the material world.
The problem is that genetic algorithms are based on one particular model of evolution. It is the dominant one for sure, but nonetheless one among several. This dominant paradigm is known broadly as the neo-Darwinian synthesis. Within this framework, novelty arises only through gradual mutation, and the shuffling of preexisting traits through sexual recombination. It is a model that has dominated evolutionary thinking in the US and Western Europe from the mid-twentieth century. It has never been the only model, however: In the early years of the twentieth century, in Russia, alternative forms of Darwinism emerged that emphasized cooperation rather than competition as the primary force driving evolutionary change. These perspectives were largely ignored by Western theorists.8
Increasingly, driven partly by developments in the microbial sciences, cooperation has reentered mainstream evolutionary discourse as a prominent driver of evolutionary change. One of the most vocal doubters of the neo-Darwinian model was the biologist Lynn Margulis. Margulis argued that many of the most dramatic moments in the history of life had not arisen because of gradual random mutation, but rather through the coming together of different organisms; symbiotic mergers; breathtaking leaps across species boundaries; or in her words, “the long-lasting intimacy of strangers.” Novelty, in other words, could arise quickly. Preexisting life forms could collide into something new. Collaborative symbiosis was not a biological curiosity, a quirky exception to the evolutionary rule, but fundamental to the evolution of new life forms. Margulis revived the word symbiogenesis to describe this process.9
Margulis did not question the importance of natural selection, the process by which disadvantageous evolutionary outcomes are weeded out. Natural selection remains the editor no matter how the novelty and variation that it is editing arises. Furthermore, in her view, the steady trickle of random mutations still resulted in novelty, but it was not the only source of novelty; it was one of several routes by which new evolutionary riffs could arise. Competition, so beloved by the neo-Darwinians, was given similar treatment. Of course competition is important, it’s just that collaboration is too. Neither is one inherently better than the other, it’s just that the books should be balanced. A view of evolution based only on competition between genes distorts our view of reality because it privileges reductive competitive interactions over anti-reductive collaborative interactions.
Intimacy has its roots in the Latin word intimus, meaning inmost, and indeed, many of these symbioses involve the literal in-habitation of another organism. Chloroplasts (the structures in which photosynthesis takes place) and mitochondria (the structures in which respiration, or energy generation, takes place) did not evolve inside plant and animal cells. Rather, the ancestors of plant and animal cells engulfed photosynthesizing and respiring bacteria that they then failed to digest. This merger provided the ancestors of plants with the ability to eat light, and the ancestors of plants and animals with the ability to generate energy from oxygen and carbohydrates. The subsequent impact of these metabolic collaborations is hard to overstate.
One of the poster organisms for the kind of intimate relationship is a single-celled creature, Mixotricha paradoxa, a type of protist. Inside each cell are bacteria, which play the energy-producing role that mitochondria do in animals, plants and fungi. Instead of fine hairs, or cilia, that occur on the surface of many types of cell, there are thousands of hairlike bacteria that move synchronously, allowing Mixotricha to propel itself. What’s more, Mixotricha itself lives in the gut of an Australian termite, allowing it to digest wood. The termite is a prolific pest for humans: Not only has it almost single-handedly made vegetable farming in Northern Australia untenable, but it bores into wooden structures, often dwellings, causing destruction. Mixotricha is a nested riot of interspecies interactions that are intensely collaborative, but not harmonious: The constituent organisms compete with each other as well.10
Margulis considered Mixotricha to have five genomes: its own along with the four bacteria that it partners with. Its intimacies can thus be construed on several levels. Its component organisms are physically intimate: They touch and enclose each other. They are also behaviorally intimate: They work together in a coordinated way to bring about movement, etc. They are also genetically intimate: Their constituent genomes work with one another. The organism cannot be reduced to any of these constituent parts. It is an embodied relationship, an irreducible system. A chorus of interspecies kin-making, to borrow a concept from Donna Haraway.11
Our point here is that given a different evolutionary song, or even introducing a new harmony to an existing evolutionary song, the influence of genetic algorithms—the work that they do—might change. In this case the reductive logics of the neo-Darwinian perspective could be replaced by symbiotic, collaborative, anti-reductive logics of intimacy and co-creatureliness.
How would the structures authored by GD differ when evolved using genetic algorithms based on symbiogenesis rather than mutation and sexual recombination?
What would happen if we replaced genetic algorithms with symbiogenetic algorithms?
We can only imagine. But it feels appropriate when designing structures that we ourselves will live within, to take inspiration from the evolutionary process by which organisms grow to live within each other, intimately. To take inspiration from the evolutionary process by which a termite might de-design a wooden human structure, based on the abilities of the Mixotricha collective, its intimately housed symbionts, to break down wood.
In the process of symbiogenesis, new organisms are in a sense negotiated into existence, arising out of an interaction or dialogue between organisms that already exist. Are there ways to think about this in a human context? What is the exchange at play between different forms of materials in a building, for example? What is the power dynamic between the natural and man-made within the context of, say, a reservoir?
A good example can be found in Land/Architecture (our term), a design practice12 that attempts to build dialogic, reciprocal relationships between human and natural worlds. Land/Architecture attempts to build with landscapes rather than on top of them. It is an architectural practice that pays “scrupulous attention to the genius of place.”13 Land/Architectures are structures that embody a conversation with a place, a dialogue with the landscape. To explore this concept, we choose as a case study a rural house set into the volcanic landscape of Lanzarote, Canary Islands, by the artist Cesar Manrique.
“As soon as I stood on Lanzarote again, I noticed an obsession to go back and relive the adventures of my youth in the lava; but now—comparatively—with much experience; I felt great interest in finding, or if possible, repeating the sensations, of living anew the magic of the volcanic fissures and thereby experience this aesthetic feeling.”14
Cesar Manrique, 1988
The house of Cesar Manrique does not stand out. When approached from the road it resembles the cubic vernacular common to the island of Lanzarote. A series of pure white interlocking oblongs rise out of a black volcanic geology. The traditional collection of low slung forms sit above a network of subterranean bubbles formed during the cooling of a lava torrent in the eruptions of 1730. The house appears to nestle within a prehistoric site, yet its geological context is no older than any London Georgian terraced house.
Commonly, these volcanic bubbles were filled in or used as underground water stores. They were of little value to the people of the island. When Manrique returned to Lanzarote, the place of his birth, having lived for a number of years among the avant-garde in New York City, he purchased the land for a song. He began to work with the natural contours of the landscape, connecting the underground chambers using small charges of dynamite.
The words of the architect Louis Kahn spring to mind when considering the manner in which Manrique worked with the site in cultivating the dwelling: “Listen to the house... It will say what it wants to be.”
In the essay Towards a Critical Regionalism, Kenneth Frampton terms this practice of working with the landscape and its natural features, “cultivating the site.”15 This feels apt and inherently dialogic: Manrique never used plans and worked incrementally to create his home. He responded to what the landscape presented to him, reacting instinctively (“Cesar loved to dance”16), and the layout of the house and the manner of its construction are emblematic of this. The resultant plan is an arational response to the conditions of the site, in that its logic is generated from fostering a close relationship with the landscape and the culture of the island: Setting creates its sense.
The forms above ground look to the traditions of Lanzarote and thus welcome its people. The cultivation and careful integration of the network of volcanic chambers works with the contours of the volcano, which is both destroyer and creator of this place:
“He will open his mouth—bomb and fire—and will bomb and set fire to the whole island. And all the strands—silver, aluminum, wax—of Lanzarote will be his lips.”17
It is the dialogues that Manrique has made with the petrified lava and the traditions of the island that are compelling. The house arises out of these interactions. When reading the house it is hard to discern where the landscape ends and the house begins. Manrique, the man, is the living link between his house, the volcano, and the culture of Lanzarote. The house is a metaphor for his practice, the act of its construction substantiates a multitude of artistic and living experiences: painter, partygoer, sculptor, builder. The house is neither a pastiche of the local style nor an assertion of any avant-garde modes that are “deaf to other melodies.”18 There is no distinction between the artist, the island, and his home.
There are many precedents for this approach. One example we want to explore is the relationship between plants and certain groups of fungi.
All plants form intimate relationships with fungi. One of the most ancient of these relationships is the mycorrhizal relationship, from the Greek word for fungus (mycos), and root (rhiza). These fungi make fine, threadlike filaments that grow into plant roots, and outward into the soil. Much of the phosphorus (and in some cases nitrogen) that plants require is supplied by the fungi, which are deft scavengers, and more efficient than plants’ own roots. In return for these nutrients, the plant provides the fungus with sugars that it makes in its leaves in the process of photosynthesis. Fungi can connect multiple plants together, forming large networks affectionately referred to as the Wood Wide Web. These networks can grow to be large and extremely complex.19
The relationship is generally thought of as a mutualism; a symbiosis in which both plants and fungi benefit. However, the relationship is dynamic. It is constantly negotiated. At any one time, plants or fungi may be giving more than they receive, or vice versa. Young plants connected to older plants via fungal networks may spend much of their heavily-shaded young lives taking from the network, only paying back when they are older and are able to find more light.20 Overall, both plants and fungi may benefit, but only when the books are balanced over longer time scales. In this sense, the benefits provided to and gleaned from all partners are not predetermined, but emerge out of an ongoing negotiation, a continuously unfolding process of exchange.
Importantly, these processes of exchange encompass multiple currencies besides phosphorus and carbon (in the form of sugar). The fungus supplies water to plants, defends their roots from infection, and can prime plant immune systems so they are less vulnerable to attack. Besides carbon, plants provide fungi with physical protection, shelter from the hectic bustle of the soil.21 The array of currencies—some, like phosphorus being stuff, and some like protection being services—form a dynamic, self-braiding line of exchange between plant and fungal lives.
By partnering, plants gain a prosthetic fungus along with the benefits and possibilities that fungi experience. In turn, fungi gain a prosthetic plant, along with the possibilities and abilities that plants have developed. Their reach and abilities are extended by the association, just as ours might be extended by the domestication of a plant, an animal, or the invention of mobile telephones. Just as a Land/Architecture might design a building in conversation with the environment they find themselves working within, plants and fungi build their own bodies and physiological possibilities in conversation with each other, on and around each other, entangled with each other.
The mycorrhizal relationship is a relational innovation, a form of extended intelligence,22 that underpins the evolution of life on land. In the earliest days of terrestrial life (around 450 mya), the ancestors of plants partnered with the ancestors of fungi, forming the first mycorrhizal relationships.23 The ancestors of plants were sea-dwelling algae-like organisms, with the ability to eat light (through the process of photosynthesis), but without the ability to digest solid materials to obtain key nutrients. Fungi are unable to photosynthesise, but are metabolically ingenious and able to digest solid substrates.
The plant–mycorrhizal fungal merger enacts a fundamental polarity visible today: Plants eat light with photosynthesis and eat the earth using fungi. Life on land was made possible by this relationship. The process by which ecosystems are able to assemble themselves on, within, and in relationship with their terrestrial context is founded on a dynamic relationship between plants and fungi. In turn, this relationship is one in which plants and fungi build their own bodies on, within, and in relationship with each other.
Land/Architecture can be understood as a refraction of this type of dialogic process, one that resists reduction by prioritizing the conversation between a life, or a form, and its context—whether another life, or a landscape, or both. Architectural traditions like modernism, by contrast, impose notions conceived a priori. The notion is prioritized over the conversation by which the notion arises.
At a larger architectural scale, Land/Architecture approaches can be seen in the increasing role that conservation, restoration, expansion, and refurbishment have in the contemporary architectural process. The work of French practice Lacaton & Vassal comes to mind in their series of projects to expand, evolve, and create dialogues with enormous housing blocks through economic additive expansions to their facades—here, the cultivated landscape is the existing architectural stock, the perception of the culture to it, and by proxy, the inhabitants too.
For both plants and mycorrhizal fungi, like Land/Architecture, the processes by which they live and grow cannot be understood in the absence of these relationships. Their lives enact an entangled flourishing in the sense that they are not reducible to a fungus-only or plant-only existence. The limits and possibilities that they face are shared.
In The Master’s Tools Will Never Dismantle the Master’s House, the poet Audre Lorde wrote: “Difference must be not merely tolerated, but seen as a fund of necessary polarities between which our creativity can spark like a dialectic.”24 The history of life is a history of entangled sparks and flourishings enabled by relational innovations that thrive amid difference. Can we think about the relationships between the members of Mixotricha, between buildings and the landscapes they find themselves within, between fungi and their plants as models for human innovation, moving forward? Of course this is an old idea. The history of domestication is the history of extending intelligence by refiguring the relationships between humans and nonhumans, whether animal vegetable or mineral. Work songs, like many other ancient and immaterial technologies of relation, have grown from an understanding of the world based on relation and exchange rather than severance and reduction.
What would the world look like if we spent as much time and money developing new ways to build relationships with other organisms and places as we did with machines? What would the world look like if we diverted one percent of the money being spent on developing artificial intelligence to developing work songs that resist reduction? If we defined intelligence inclusively, as the ability of organisms to respond flexibly to the problems that their lives present them with? These relational innovations can be considered ways to develop an extended, rather than an artificial, intelligence; ways of thinking with, rather than thinking about; ways of talking with the world rather than about it.