Abstract: Multimateriality has usually been treated as a theory of making, but made matter also falls out of the coordination for which it was formed. This article supplies the missing half. It places waste in the history of life, where one organism's discarded matter is very often another's food or shelter, and where the oxygen released by cyanobacteria and the carbon buried by ancient forests show that externalised matter can also escape recruitment and become a condition of the world. It defines waste relationally, as recruited or made matter that has fallen out of the coordination for which it was formed, and it distinguishes the trajectories that follow: decay, persistence, dispersal, re-finding and deliberate deposition. Two great human deposits test the definition. The shellmounds of San Francisco Bay were misread as refuse and destroyed by settlers before being returned to Ohlone stewardship, and Monte Testaccio, the hill of discarded olive oil amphorae in Rome, became terrain and the archive of the system that produced it. The article develops three concepts: material afterlife, as a dimension of multimaterial profiles; re-finding, in which discarded forms are recruited again as found matter; and unrecruited poiesis, in which industrial civilisation externalises made matter faster than any living process can take it up. Waste returns multimateriality to its own precondition, so that the history of made things is a loop and not a ladder, and downward answerability arrives as return through air, water, soil and bodies. The article closes with maintenance, repair and the design of afterlives as the labour of holding a shared world together.
I. Introduction: The Missing Half of Multimateriality
The theory of multimateriality has so far been a theory of making. It reconstructs how living beings came to recruit self-produced matter, recruit found matter, select and displace it, transform it, combine it and finally project it before it exists. These are the six layers of a stratigraphy that rests on two preconditions, material dependence on conditions of dwelling and metabolic externalisation, the unrecruited alteration of surroundings that all living produces. The criterion of the framework is recursive uptake: a material transformation becomes multimaterial mediation only when it enters a living being's further coordination. Found and made name relations to aspects of a configuration and not classes of object, later relations accumulate without abolishing earlier ones, and later layers remain answerable to the older conditions on which their viability depends, a principle that can be called downward answerability. Made configurations can also be compared along dimensions such as mobility, mutability, reproducibility, site dependence, required sameness, repairability, temporal trajectory and the burden of symbolic stabilisation, which together form a multimaterial profile. The direction of every analysis in this framework has been upward, from finding toward fabricating.
Each analysis also reaches an edge that it does not examine. Medicines expire and are destroyed. Warehoused metal proves to be stones in sacks. Garments are discarded in bales and shipped to markets where many of them cannot be sold, and synthetic fibres shed into rivers and seas. Bridges stand only as long as they are repaired, and the made configurations that scholars have learned to follow eventually stop being followed by anyone. The framework has no account yet of what happens to made matter after it stops fitting the coordination for which it was made. That is where it is weakest, and it is where this article begins.
The gap matters for more than completeness. The quantity of made matter on Earth has become geological. By one estimate, the total mass of human-made objects, concrete, aggregates, bricks, asphalt, metals and plastics above all, came to exceed the dry mass of all living beings on the planet around 2020 (Elhacham et al. 2020). Of the roughly 8.3 billion tonnes of plastic produced by 2015, most had already become waste, and most of that waste had accumulated in landfills or the wider environment rather than being recycled or burned (Geyer, Jambeck and Law 2017). A theory of multimateriality that stops at the moment of use describes only a thin slice of the material history it claims to explain.
The article makes four moves. It first places waste in the evolutionary history of life, where one organism's discarded matter is very often another's food or shelter. It then defines waste relationally, as matter that has fallen out of coordination, and examines two great human deposits: the shellmounds that the Ohlone and their ancestors built around San Francisco Bay, and Monte Testaccio, the hill of broken olive oil amphorae in Rome. It develops three concepts, material afterlife, re-finding and unrecruited poiesis. And it argues that industrial waste returns multimateriality to its own precondition, metabolic externalisation, so that the history of made matter turns out to be a loop rather than a ladder, and downward answerability arrives as return.
II. Waste Before Humans: Externalisation, Decay and Recruitment
Living beings have always produced matter they do not keep. Metabolism takes in, transforms and releases, and what is released accumulates: faeces, urine, shed skin, moulted feathers, exoskeletons, dead leaves, carcasses, spent webs. In the stratigraphy of multimateriality, such release belongs to the second precondition, metabolic externalisation, the alteration of surroundings that living inevitably produces before any of its products are recruited into further coordination.
Very little of what one organism releases remains unrecruited for long. Dung beetles roll, bury and eat the dung of large mammals and lay their eggs in it. Vultures, hyenas, beetles and blowflies take apart carcasses within days. Fungi and bacteria decompose fallen wood and leaves into soil, and earthworms eat that soil and remake it. Hermit crabs occupy the shells that dead gastropods leave behind. Birds line their nests with shed hair and feathers. Coral reefs are built on the calcium carbonate skeletons of generations of dead coral. In most ecosystems, externalised matter re-enters living coordination through other organisms, and what looks like waste from the standpoint of its producer is resource from the standpoint of its recruiter.
This gives the first principle of the article. Waste is a relation rather than a substance. Faeces are waste for the animal that excretes them and food for the beetle that buries them. A shell is waste for the dead mollusc and a house for the crab. The same matter can fall out of one coordination and enter another, and whether it counts as waste depends entirely on whose coordination is in question.
There are also great exceptions, and they are instructive. Cyanobacteria released oxygen as a by-product of photosynthesis. In photosynthesis itself oxygen has no role: it is what is left behind when the useful parts of the water molecule have been taken away and put to work (Godfrey-Smith 2024). More than two billion years ago the gas began to accumulate in the oceans and atmosphere faster than any process could take it up, transforming the chemistry of the planet and poisoning many of the organisms then living (Lyons, Reinhard and Planavsky 2014). Animals later came to depend on it. Godfrey-Smith (2024) distinguishes transformation of surroundings that is the point of what an organism does, which he calls engineering, from transformation that is a byproduct of doing something else, and cyanobacterial oxygen is his example of the second: it transformed the Earth without re-engineering it. Carbon supplies a second great case. When plant matter is buried before it can break down, its carbon is interred and the oxygen that might have combined with it remains in the air, and in the Carboniferous large quantities of plant matter were buried in this way and became coal (Godfrey-Smith 2024). An older explanation, that fungi had not yet evolved the means to decompose lignin, has been questioned, and the conditions of burial appear to have mattered as much (Nelsen et al. 2016). Unrecruited externalisation on this scale changes the dwelling conditions of every living being that follows. These cases show that waste can escape recruitment, and that when it does, it becomes a condition of the world rather than a resource within it.
III. What Is Waste? Matter That Has Fallen Out of Coordination
Waste can accordingly be defined as recruited or made matter that has fallen out of the coordination for which it was recruited or formed. The definition has three features. It is relational, since matter falls out of a particular coordination and may remain in, or enter, others. It is historical, since waste presupposes prior recruitment or making: a stone that was never picked up cannot be discarded. And it is neutral about what happens next, since matter that has fallen out of coordination may decay, persist, disperse, be re-found or be deliberately deposited.
Mary Douglas (1966) described dirt as matter out of place, the by-product of a system of classification that sorts things into their proper positions and treats what escapes the sorting as polluting. Her formulation captures the symbolic side of waste precisely. Waste is always also a classification, and the classification belongs to multisymbolization: to call something rubbish is to place it outside the order in which things count. Thompson's (1979) rubbish theory added a temporal dimension, showing how objects can pass from the category of the transient, losing value as they age, through a limbo of rubbish in which they have no value at all, into the category of the durable, as antiques, heirlooms and collectibles whose value grows with age.
The multimaterial account keeps both insights and adds a third. Waste is a classification, and it is also a material trajectory that continues regardless of how it is classified. A plastic bottle called rubbish and a plastic bottle called a collectible are made of the same polymer, and both will persist for a very long time. The symbolic reclassification of discarded matter can change its fate, as when scrap becomes a commodity or a ruin becomes a monument, but it cannot change what the matter is or how it decays. Waste therefore sits at a characteristic junction of multisymbolization and multimateriality: the classification is symbolic, and the afterlife is material. The classification is also a binary, and binaries are always L4 operations, compressions of a continuum into two values. The world contains matter more or less in coordination, more or less recoverable, and no line at which the useful becomes the discarded.
Several distinct trajectories follow from falling out of coordination, and the word waste tends to collapse them. Matter can decay, returning quickly to the biological cycles from which it came, as food scraps and wooden tools do. It can persist, remaining materially intact but unused, as glass, ceramics and plastics do. It can disperse, breaking into fragments too small or too scattered to be recruited, as microplastics and airborne particulates do. It can be re-found, recruited again into a new coordination, as scrap metal and second-hand clothing are. And it can be deliberately deposited, placed with care in a particular place for a purpose other than use, as grave goods, votive offerings and, as the next section argues, the shellmounds of San Francisco Bay were. A theory of waste needs to distinguish these trajectories before it can say anything about any of them.
The distinction between classification and trajectory has an empirical demonstration. From the 1970s, the Garbage Project at the University of Arizona sorted and recorded household rubbish in Tucson and later excavated landfills across North America (Rathje and Murphy 1992). Two of its findings bear directly on the argument. First, what households said they consumed differed systematically from what their rubbish showed they consumed; reported drinking, for instance, fell well below what the discarded bottles and cans recorded. The symbolic account of consumption and its material trace diverged. Second, landfills preserved far more than expected. Deprived of air and moisture, buried food and paper decomposed very slowly, and newspapers decades old could still be read. The landfill, designed as a place where waste would disappear, turned out to be a place where it persisted.
IV. The Shellmound: When Discard Is Not Waste
Around the shores of San Francisco Bay stood hundreds of mounds built over thousands of years by the ancestors of the Ohlone and their neighbours. The archaeologist N. C. Nelson (1909) mapped more than four hundred of them in the first decade of the twentieth century, when many were still visible along the shoreline. The largest, at Emeryville and West Berkeley, rose many metres above the surrounding marshes. They consisted of immense quantities of shells, mussel, oyster and clam above all, together with ash, charcoal, animal and fish bones, stone and bone tools, ornaments and the burials of the dead. Radiocarbon dating has placed the beginnings of the West Berkeley mound some 5,700 years ago, making it among the oldest known village sites on the Bay.
For much of the twentieth century, archaeologists described these mounds as middens, a word that means refuse heaps. On that reading they were the accumulated kitchen waste of people who ate shellfish, a record of diet deposited without intention. The reading made the mounds easy to destroy. Farmers carted their shell away as fertilizer, contractors used it to surface roads, and the Emeryville mound was levelled in the 1920s for industrial development. Later in the century a shopping centre was built on its site, and excavation before construction found the remains of many ancestors who had been buried in the mound. The West Berkeley site lay for decades beneath a restaurant car park. Mid-century excavations removed human remains, funerary objects and thousands of artefacts from it to university collections.
The midden reading was a misclassification, and its error was one of level. An observable feature, the presence of shells and bones, was stabilised at L4 as the category refuse, and the category then licensed treatment that the builders of the mounds would never have accepted. Luby and Gruber (1999) argued that the Bay Area shellmounds were mortuary monuments as much as accumulations, places where the dead were buried among the remains of feasts held in their honour, so that the shells and the burials belonged to one practice of feeding and remembering the dead. Lisjan Ohlone people, whose ancestors lie in the mounds, have insisted that these are sacred sites and burial grounds. The mound was built by discard, in the sense that the shells had been eaten and their contents consumed. Yet the discard was also deposition, the deliberate placing of matter in a place that mattered, and the place grew into a monument through the very act of accumulation. In Godfrey-Smith's (2024) terms, an accumulation that begins as a byproduct of feeding can become, over generations, transformation that is the point of what people do, and the mounds sit exactly at that boundary.
The shellmound thus dissolves the distinction that the word waste presupposes. The shells had fallen out of one coordination, the eating of shellfish, and entered another, the building and tending of a place of the living and the dead. This is the principle of accretion made literal. Multimateriality proceeds by adding new relations to matter without deleting the older ones, and the shellmound is a structure built by exactly such adding, layer on layer over millennia, each generation depositing on what the last had left. It is a made landform that was never made according to a plan, and it was made from matter that had already served its first purpose.
The settler classification of the mounds as refuse repeated a pattern familiar from the history of anthropology. An observable feature was translated into a general category, refuse, and the category licensed removal, levelling and paving. The mound was read through a regime in which discarded matter had no further significance, and that regime was projected onto a people for whom it did not hold.
The history has a recent turn, and it displays the asymmetry of reopening with unusual clarity. After the City of Berkeley blocked a building permit for a condominium development on the West Berkeley site, courts ruled in favour of the developers. In March 2024 the city council unanimously approved the purchase of the site for twenty-seven million dollars, of which the Sogorea Te' Land Trust, an Indigenous-led organisation, raised twenty-five and a half million, and the land was transferred to the trust to hold on behalf of the Confederated Villages of Lisjan (City of Berkeley 2024; Sogorea Te' Land Trust 2024). The mound's reclassification from refuse heap to landmark and then to returned land shows the symbolic side of waste changing its fate. It also shows who bore the cost of reopening a settled classification. The closure that had served developers needed only to stand, whereas the descendants of the mound's builders had to find standing, evidence and a forum, and in the end to raise most of the price of land whose classification as refuse had made its destruction thinkable.
V. Monte Testaccio: A Hill of Amphorae
In the south of Rome, near the east bank of the Tiber, stands a hill about thirty-five metres high and nearly a kilometre around its base, covered in grass and trees. It is made almost entirely of broken pottery. Monte Testaccio consists of the fragments of an estimated fifty million or more amphorae, the great majority of them globular vessels of the type archaeologists call Dressel 20, which carried olive oil from the province of Baetica in southern Spain to the city of Rome (Blázquez Martínez, Remesal Rodríguez and Rodríguez Almeida 1994; Peña 2007). The dumping began in the first century CE and continued into the third, the period when Rome, a city of perhaps a million people, depended on shipments of oil, grain and wine from across the Mediterranean, many of them organised or supervised by the state as part of the food supply known as the annona.
The amphorae were discarded rather than reused because of what they were. Once emptied into smaller containers at the warehouses nearby, the Dressel 20s could not easily be put to other uses. Their clay had absorbed oil, which would have turned rancid in any new contents. Their thick, curved walls broke into large fragments that were hard to reduce to the small pieces used as aggregate in Roman concrete, and the residual oil may have made the fragments unsuitable for mixing with lime. Most other amphora types were reused, repurposed or ground into building material. The Dressel 20 had a material afterlife determined by its fabric, its shape and its contents, and that afterlife made it waste.
The discard was organised. Excavations have shown that the hill was built in terraces. Workers laid intact or half-broken amphorae in rows to form retaining walls, filled the space behind them with sherds, and scattered lime over the fragments, probably to neutralize the smell of the decaying oil. The heap was thus a made structure, built by the systematic placement of matter that had fallen out of its first coordination. Like the shellmound, it grew by accretion, and like the shellmound, it became a landform.
Monte Testaccio shows made matter taking the place of dwelling. Dwelling names the non-human-made conditions of a place, its terrain, water, climate and ground. The hill was made, yet within a few centuries of the last dumping it had become part of the terrain of Rome, a landform like the seven hills among which it stands. For those who came later it was found: ground that no living person had made, answering as ground does to weather, slope and drainage, and so functioning in their coordination as dwelling while remaining a deposit of earlier making. Medieval Romans held games and carnival processions on its slopes, and later its interior, which stays cool, was dug into for wine cellars, whose successors now house restaurants and clubs. People who walk on the hill today walk on the discarded packaging of the imperial oil supply without needing to know it. The area around the hill kept an association with what the city discarded. In the late nineteenth century Rome built its municipal slaughterhouse at Testaccio, and the district's cuisine, according to local tradition, grew around the quinto quarto, the fifth quarter of the animal, the offal left after the four prime quarters had been sold, which slaughterhouse workers took home or received in part payment. A hill of discarded jars stands beside a cuisine of discarded cuts. Made matter, once it has fallen out of coordination and persisted long enough, can become found matter for later generations, part of the given world into which they are born.
The hill is also an archive. Many amphorae carry painted inscriptions, tituli picti, recording the weight of the vessel and its contents, the name of the merchant or shipper, the place of origin and the officials who checked the load. Read in their millions, these inscriptions allow historians to reconstruct the organisation of the oil trade, the families and firms that ran it, and the changing role of the state in supplying the capital (Blázquez Martínez, Remesal Rodríguez and Rodríguez Almeida 1994). The oil itself was eaten, burned in lamps and rubbed on bodies long ago, and it left no trace. Its packaging, thrown away because it could not be used again, preserved the record of the system that moved it.
This reverses a principle that holds elsewhere in the study of made things. Successful coordination tends to leave little trace, because it prevents the disruption that would call attention to itself, while misfit is recorded in accounts of failure and repair. Discard supplies a third kind of record. The successful consumption of oil erased the oil, and the discard of its containers preserved an archive of the whole supply system. Archaeology is largely the study of such afterlives, and it knows the past disproportionately through what the past threw away.
VI. Material Afterlife as a Dimension of Multimaterial Profiles
Made configurations differ in their afterlives as much as they differ in mobility, mutability or reproducibility. Material afterlife can be defined as the trajectory that a made or recruited form follows after it has fallen out of the coordination for which it was formed. It deserves a place among the dimensions of a multimaterial profile, because it varies enormously and because its variation has consequences as far-reaching as those of any other dimension.
Some afterlives are short and closed. A wooden bowl, a woven basket or a garment of undyed wool will rot within years or decades and return to soil. A pharmaceutical tablet is designed to be destroyed after its expiry date, and regulation governs the manner of its destruction. Some afterlives are long and open. A ruined bridge becomes part of a landscape, its arches standing in a river centuries after its road has gone. A ceramic sherd will persist for millennia. Some afterlives are dispersive. A polyester garment sheds fibres every time it is washed, and those fibres, too small to see, travel into waterways, soils and the bodies of animals. Some afterlives are toxic. Lead paint, mercury from felt-making and gold-mining, and radioactive waste persist in forms that harm the living beings they encounter long after their makers are gone.
Afterlife interacts with the other dimensions of a profile. High reproducibility combined with a long afterlife generates accumulation, since billions of persistent objects are made and each outlasts its use; this is the problem of plastic packaging. High mobility combined with a dispersive afterlife generates distant harm, since the object travels before it falls apart and its fragments travel further. Site dependence combined with a long afterlife generates ruins, landforms and the obligations of heritage. Designers increasingly try to specify afterlives projectively, through biodegradable materials, design for disassembly and deposit schemes for containers, which is an attempt to extend projective multimateriality past the end of use.
The dimension also clarifies a historical asymmetry. For almost the whole of human history, most made things had short and closed afterlives, because they were made from materials that living processes had produced and that other living processes could decompose. Stone, ceramic and metal were the great exceptions, and they are the materials through which the deep human past is now known. The twentieth century inverted this proportion. Synthetic polymers, made from hydrocarbons and designed for durability, are materials that almost no living process can take apart, and they are now produced in quantities that dwarf all earlier persistent materials. The long, open and dispersive afterlife has become the normal case.
The most extreme case of an afterlife that must be projected is radioactive waste. Some of the material produced by nuclear reactors and weapons programmes will remain dangerous for tens of thousands of years, far longer than any human institution or language has lasted. Finland has built a repository at Onkalo, in tunnels deep in the bedrock, designed to isolate spent fuel for about a hundred thousand years. In the United States, the planners of a repository for defence waste in New Mexico commissioned experts to design markers that could warn people ten thousand years in the future, who might share none of our languages or symbols, not to dig there (Trauth, Hora and Guzowski 1993). The exercise was an attempt to extend multisymbolization across an afterlife longer than all of recorded history, and it exposed the limits of symbolic stabilisation when the matter it concerns outlasts every community that could read it. The material afterlife is certain, while its symbolic accompaniment cannot be made to last as long.
VII. Re-Finding: The Found and the Made in Reverse
The distinction between found and made is a relation rather than a classification of objects: the same configuration can be found in some respects and made in others. Waste shows the relation running in both directions. Matter that one population made and discarded can become found matter for another, recruited again into a new coordination. This is re-finding, and it is one of the oldest and most pervasive relations to matter in human history.
The monuments of Rome were quarried by later generations. The Arch of Constantine incorporates reliefs taken from earlier monuments of Trajan, Hadrian and Marcus Aurelius, and for centuries the travertine of the Colosseum and other ancient buildings was stripped and burned for lime or reused in palaces and churches. Medieval builders across Europe set carved fragments of Roman buildings, spolia, into new walls, sometimes as mere stone and sometimes as displays of inherited authority. In each case the made form of an earlier age was found by a later one, and recruited at the second layer of the stratigraphy, as found matter, before being transformed and combined again.
Re-finding sustains vast economies today. In India, networks of kabadiwalas buy newspaper, bottles, metal and plastic from households door to door and sell them on through chains of dealers to recyclers, so that a great deal of what wealthier households discard never reaches a dump. In Rio de Janeiro, Millar (2018) described the catadores who reclaimed recyclable materials from the Jardim Gramacho landfill, for whom the dump was a workplace and the discard of the city a livelihood. In Accra, traders at the Kantamanto market buy bales of used clothing from Europe and North America unseen, sorting, repairing and restyling what can be sold. On the beaches of Alang in Gujarat, ocean-going ships are driven ashore at high tide and cut apart by hand, and their steel is re-rolled into building materials, at great cost to the health and sometimes the lives of the workers who dismantle them.
Kolkata supplies a case in which the re-finding extends to a whole landscape. East of the city, the sewage of millions of people flows into a system of shallow ponds and fields, the East Kolkata Wetlands, where for about a century it has been used to raise fish and irrigate vegetables. The wastewater, which would otherwise have to be treated at great expense, is taken up by algae, fish and plants, and the city is fed in part by what it discards. The wetlands were recognised as a Ramsar site of international importance in 2002. They show the principle of section II, that one organism's waste is another's food, reorganised as a human institution at the scale of a metropolis.
Re-finding reveals that the stratigraphy of multimateriality is recursive in time. The layers of recruitment, displacement, transformation, combination and projection do not end with the finished object. The finished object, once discarded, can be recruited again at a lower layer, as found matter, and passed back up through the stratigraphy into a new form. What falls out of one coordination is available to be found by another, and much of human material life consists of such second and third lives.
VIII. Unrecruited Poiesis: Industrial Waste and the Return to Externalisation
In the stratigraphy of multimateriality, wild bearded capuchins at Serra da Capivara provide a threshold case. They strike quartzite cobbles together and detach sharp flakes that resemble early hominin tools, but they do not use the flakes (Proffitt et al. 2016). They transform stone without recruiting the transformation, and their flakes stand to lithic technology as dispersing metabolic waste stands to a recruited structure. The general principle is that material transformation becomes multimaterial mediation only when the transformation enters subsequent coordination.
Industrial civilisation has become, on a planetary scale, the capuchin at the cobble. It produces material transformations of enormous extent that no one takes up into further coordination. Most plastic ever made has become waste, and most of that waste remains in landfills or the environment (Geyer, Jambeck and Law 2017). Carbon dioxide from burning coal, oil and gas accumulates in the atmosphere faster than oceans, soils and plants can absorb it. Mine tailings, fly ash, electronic waste and persistent chemicals accumulate in quantities without precedent. These are products of the most elaborate projective multimateriality ever achieved, and they end as matter that no living process recruits. The concept of unrecruited poiesis names this condition: made matter that has fallen out of coordination and enters no other, persisting or dispersing as a condition of the world.
The carbon in question forms a loop within a single element. It was taken up by ancient plants and plankton, buried before it could decay, recruited by industrial societies as fuel, and returned to the air as carbon dioxide within a few centuries. Godfrey-Smith (2024) notes that burning all the accessible fossil fuels on Earth would barely change the vast reservoir of atmospheric oxygen, whereas carbon dioxide is so scarce in the atmosphere that the same burning alters it proportionally a great deal, with correspondingly large effects on temperature. Waste in this case is small in proportion to the medium that receives it and large in consequence.
Unrecruited poiesis returns multimateriality to its own precondition. The stratigraphy begins with metabolic externalisation, the unrecruited alteration of surroundings that all living produces, and it rises through recruitment and transformation to projective fabrication. Industrial waste brings the top of the stratigraphy back to the bottom. The comparison with cyanobacteria is exact in form, whatever its differences in scale and consequence. Their oxygen was a by-product of the metabolic process that sustained them, released faster than anything could take it up, and it changed the dwelling conditions of every later organism. Industrial carbon dioxide and plastics are by-products of the multimaterial process that sustains contemporary human life, released faster than anything can take them up, and they are changing the dwelling conditions of every living being now alive and to come. The history of multimateriality therefore turns out to be a loop rather than a ladder. Its highest achievements return to its lowest relation.
The comparison invites a misreading that Godfrey-Smith's (2024) discussion of nature and the natural helps to avoid. If human action is part of the history through which living beings have transformed the Earth, then its being natural in that sense cannot guide choice, since whatever people do is natural in that sense and the description cannot tell one action from another. Continuity of kind therefore excuses nothing. What marks the present is speed and scale. The trouble with a warming climate lies less in temperatures unprecedented in the history of the planet than in the pace of the change, which few species and few institutions can absorb.
Geologists have begun to register the loop in their own terms. Zalasiewicz and colleagues (2014) proposed that the durable made objects of the present, from ballpoint pens to concrete, will form a distinctive class of technofossils in the sediments now accumulating. On a beach in Hawaii, researchers described a new kind of rock, plastiglomerate, in which melted plastic had fused with sand, shell fragments and volcanic rock into a durable stone likely to enter the geological record (Corcoran, Moore and Jazvac 2014). The title of this article records the return. Multimateriality began with secretion and rose to stone. Its waste is now becoming sediment and, in places, stone again.
IX. Downward Answerability as Return
Downward answerability is the principle that later layers, however elaborate, remain answerable to the older conditions on which their viability depends. The bridge must fit gravity; the medicine must fit a metabolising body; the financial claim must eventually meet the copper, the harvest or the borrower. Waste shows downward answerability in a further form. It arrives as return: matter that falls out of coordination re-enters dwelling, the non-human-made conditions of air, water and soil, and through dwelling it comes back to bodies.
The return is often literal. Microplastic particles have been found in human placentas (Ragusa et al. 2021) and in human blood (Leslie et al. 2022). Mercury released by industry and by small-scale gold mining settles into waterways, is transformed by bacteria into methylmercury and accumulates up food chains, as the poisoning of fishing communities at Minamata in Japan showed in the 1950s. Persistent chemicals used to make textiles and packaging resistant to water and grease now occur in the blood of most people tested for them. The sea absorbs part of the carbon dioxide that is released and becomes more acidic, which is a problem for many shell-building invertebrates (Godfrey-Smith 2024), so that the same gas threatens the kind of animal whose shells built the mounds of San Francisco Bay. The multimaterial hinterland of made things, which extends backward to mines, fields and oil wells, turns out to extend forward as well, into the bodies of people who never used the products whose waste they carry.
The return is also unequal. Waste travels, and it tends to travel away from those who produced it toward those with less power to refuse it: to informal dumps on the edges of cities, to ship-breaking beaches, to rivers downstream of factories, to countries that import discarded clothing and electronic scrap. Liboiron (2021) has argued that pollution is itself a form of colonialism, since it presupposes access to land and water that can be treated as a sink. In multimaterial terms, the afterlife of made matter is distributed very differently from its benefits, and those who bear the afterlife are frequently those who never enjoyed the use.
Downward answerability as return also completes the loop described in the previous section. The made world does not escape the found world by discarding its products into it. The discarded products become part of the found world, altering its conditions, and every later act of making must then fit a world that earlier waste has changed. Seen this way, climate change is downward answerability arriving through the atmosphere: the found conditions of dwelling, reshaped by unrecruited industrial externalisation, return to set new conditions of fit for everything that living beings make.
X. Maintenance, Repair and Lokasaṃgraha
If waste is matter that has fallen out of coordination, then maintenance and repair are the labour of keeping matter within it. Graham and Thrift (2007) and Jackson (2014) have shown that repair is as constitutive of technological order as invention, and that most of the made world persists only because people continually clean, patch, replace and adjust it. Every repaired object is an object whose afterlife has been postponed. A bridge restored after war, a garment darned at the elbow and a machine kept running with improvised parts all extend the period during which made matter remains in fit.
The Bhagavad Gītā offers a term for the orientation this labour expresses. Kṛṣṇa urges Arjuna to act for lokasaṃgraha, the holding together of the world, performing the actions that sustain the shared order without attachment to their personal fruits (Bhagavad Gītā 3.20, 3.25). The term fits maintenance well, because maintenance is effort directed at keeping a shared world from coming apart, and because its success is largely invisible. A well-maintained drain, road or roof attracts no attention. Its failure does.
Waste is thus the point at which maintenance either ends or fails. It ends when a thing has completed its use and is deliberately released from coordination. It fails when a thing that could have been kept in fit is allowed to fall out. The difference matters, because it separates the designed afterlife from the neglected one. Some traditions design afterlives with care. During Durga Puja in Kolkata, images of the goddess modelled from river clay on frames of straw and bamboo are carried to the Hooghly at the end of the festival and immersed, returning to the river the matter from which they were made. The practice is a deliberate completion of a material trajectory, in which the image falls out of ritual coordination by a ritual act. In recent decades, as images came to be decorated with synthetic paints and non-biodegradable ornaments, the same act of immersion began to leave lead and plastic in the river, and authorities and organisers have had to regulate how it is done. The designed afterlife of clay was disrupted by materials whose afterlives no one had designed.
Lokasaṃgraha in the age of unrecruited poiesis accordingly requires more than repair. It requires the projective design of afterlives, so that what is made can fall out of coordination without falling out of the world's capacity to take it up. Much of what is now called the circular economy is an attempt at exactly this. Its success will depend on whether made matter can again be given afterlives that living processes, or human institutions of re-finding, can recruit.
XI. Conclusion: From Stone to Sediment
The theory of multimateriality began with secretion and rose to stone. It traced how living beings learned to recruit, displace, transform, combine and project matter, and it showed that every later relation remained answerable to the older ones. This article has added the half that the theory lacked. Made matter does not only rise. It falls out of coordination, and what happens next is as various and as consequential as anything that happened before.
Waste is a relation rather than a substance. The shellmounds of San Francisco Bay, built from the shells of mussels, oysters and clams eaten over thousands of years, show that discard and deposition can be the same act, and that a colonial regime that read such places as refuse misclassified them at the cost of their destruction. Monte Testaccio, built from tens of millions of olive oil amphorae that could not be used again, shows made matter becoming terrain and becoming the archive of the system that discarded it. Re-finding shows the stratigraphy running in reverse, as discarded forms are recruited again as found matter. Unrecruited poiesis shows the top of the stratigraphy returning to its bottom, as industrial civilisation externalises made matter faster than anything can take it up.
Life did not only learn to make worlds. It also learned, repeatedly, to leave behind what it made, and much of what it left became the ground on which later life was built. Shells became hills, broken jars became a hill, cities were built from the stones of older cities, and one city's sewage feeds its people through the fish of its wetlands. The difference in the present lies in the scale and persistence of what is being left, and in the fact that so little of it can be taken up again. From secretion to stone and from stone to sediment, the history of multimateriality is a loop in which everything made returns, sooner or later, to the found world, and every afterlife becomes a condition that later making must fit.
Selected References
The Bhagavad Gītā. See van Buitenen, J. A. B., trans. 1981. The Bhagavadgītā in the Mahābhārata. Chicago: University of Chicago Press.
Blázquez Martínez, José María, José Remesal Rodríguez, and Emilio Rodríguez Almeida. 1994. Excavaciones arqueológicas en el Monte Testaccio (Roma): Memoria campaña 1989. Madrid: Ministerio de Cultura.
City of Berkeley. 2024. Ordinance No. 7,899-N.S., authorizing the acquisition and transfer of real property at 1900 Fourth Street (the West Berkeley Shellmound). Adopted 12 March 2024.
Corcoran, Patricia L., Charles J. Moore, and Kelly Jazvac. 2014. "An Anthropogenic Marker Horizon in the Future Rock Record." GSA Today 24 (6): 4–8.
Douglas, Mary. 1966. Purity and Danger: An Analysis of Concepts of Pollution and Taboo. London: Routledge and Kegan Paul.
Elhacham, Emily, Liad Ben-Uri, Jonathan Grozovski, Yinon M. Bar-On, and Ron Milo. 2020. "Global Human-Made Mass Exceeds All Living Biomass." Nature 588: 442–444.
Geyer, Roland, Jenna R. Jambeck, and Kara Lavender Law. 2017. "Production, Use, and Fate of All Plastics Ever Made." Science Advances 3 (7): e1700782.
Godfrey-Smith, Peter. 2024. Living on Earth: Life, Consciousness and the Making of the Natural World. London: William Collins.
Graham, Stephen, and Nigel Thrift. 2007. "Out of Order: Understanding Repair and Maintenance." Theory, Culture and Society 24 (3): 1–25.
Jackson, Steven J. 2014. "Rethinking Repair." In Media Technologies: Essays on Communication, Materiality, and Society, edited by Tarleton Gillespie, Pablo J. Boczkowski, and Kirsten A. Foot, 221–239. Cambridge, MA: MIT Press.
Leslie, Heather A., Martin J. M. van Velzen, Sicco H. Brandsma, et al. 2022. "Discovery and Quantification of Plastic Particle Pollution in Human Blood." Environment International 163: 107199.
Liboiron, Max. 2021. Pollution Is Colonialism. Durham: Duke University Press.
Luby, Edward M., and Mark F. Gruber. 1999. "The Dead Must Be Fed: Symbolic Meanings of the Shellmounds of the San Francisco Bay Area." Cambridge Archaeological Journal 9 (1): 95–108.
Lyons, Timothy W., Christopher T. Reinhard, and Noah J. Planavsky. 2014. "The Rise of Oxygen in Earth's Early Ocean and Atmosphere." Nature 506: 307–315.
Millar, Kathleen M. 2018. Reclaiming the Discarded: Life and Labor on Rio's Garbage Dump. Durham: Duke University Press.
Nelsen, Matthew P., William A. DiMichele, Shanan E. Peters, and C. Kevin Boyce. 2016. "Delayed Fungal Evolution Did Not Cause the Paleozoic Peak in Coal Production." Proceedings of the National Academy of Sciences 113 (9): 2442–2447.
Nelson, N. C. 1909. "Shellmounds of the San Francisco Bay Region." University of California Publications in American Archaeology and Ethnology 7 (4): 309–356.
Peña, J. Theodore. 2007. Roman Pottery in the Archaeological Record. Cambridge: Cambridge University Press.
Proffitt, Tomos, Lydia V. Luncz, Tiago Falótico, et al. 2016. "Wild Monkeys Flake Stone Tools." Nature 539: 85–88.
Ragusa, Antonio, Alessandro Svelato, Criselda Santacroce, et al. 2021. "Plasticenta: First Evidence of Microplastics in Human Placenta." Environment International 146: 106274.
Rathje, William, and Cullen Murphy. 1992. Rubbish! The Archaeology of Garbage. New York: HarperCollins.
Sogorea Te' Land Trust. 2024. "Let the Land Breathe: Breaking Ground at the West Berkeley Shellmound." https://sogoreate-landtrust.org/let-the-land-breathe/.
Thompson, Michael. 1979. Rubbish Theory: The Creation and Destruction of Value. Oxford: Oxford University Press.
Trauth, Kathleen M., Stephen C. Hora, and Robert V. Guzowski. 1993. Expert Judgment on Markers to Deter Inadvertent Human Intrusion into the Waste Isolation Pilot Plant. SAND92-1382. Albuquerque: Sandia National Laboratories.
van Buitenen, J. A. B., trans. 1981. The Bhagavadgītā in the Mahābhārata. Chicago: University of Chicago Press.
Zalasiewicz, Jan, Mark Williams, Colin N. Waters, et al. 2014. "The Technofossil Record of Humans." The Anthropocene Review 1 (1): 34–43.