I laid two photographs on the desk in my room.
One was the Pont du Gard, taken some years ago in the south of France. A Roman aqueduct built around 50 CE. About 48 meters tall, about 275 meters long. Three stories of arches striding across the narrow valley of the Gard. The lowest tier has six great arches, the middle eleven, the topmost thirty-five smaller ones. The stone is a pale limestone, and at sunset it turns gold in the light. In the photograph I am standing on the riverbank below the aqueduct; from the angle of looking up, the structure seems too vast for human work.
The other photograph is one I took a few days ago in Ollantaytambo. Beside a stone stairway, a water channel two palms wide is flowing. The stone is hard andesite, its surface smoothed by five hundred years of weathering. The water is clear and moves slowly down the steps. There is no arch in the frame, no monumental structure, no dramatic line. Only stone and water.
Both are "facilities for carrying water."
And yet why do they look so different.
I laid the two side by side and stared for a long time. The first intuition that came was simple. One goes over, the other goes along. Roman aqueducts overcome gravity and leap across the valley. Inca canals submit to gravity and follow the terrain. Rome is straight line, the Inca is curve. Rome is verticality, the Inca is horizontality.
But this, I soon realized, was not a simple difference of technique. The two structures are two worldviews, materialized. How shall water be treated. How shall nature be treated. And, ultimately — how shall being be treated.
In the previous chapter I made the fact that Tipón's water has been flowing for five hundred years the central thesis. I now have to expand a fact I mentioned alongside it. The Roman aqueducts of the same era, when the Western Empire fell, mostly stopped flowing for fifteen hundred years. Both civilizations left great hydraulic achievements in human history, but the way those achievements survived their empires diverged.
That divergence is the subject of this chapter. And it foretells the opposition of two archetypes that will run through the entire book.
In the first century CE, the Roman Empire was unifying the Mediterranean world. And one of the material proofs of that unification was the aqueduct. At its peak, Rome is estimated to have built aqueducts for about 500 cities.¹ The city of Rome itself had eleven major aqueducts, with a total length of about 500 kilometers.²
The core of Roman hydraulic engineering rested on three techniques.
Arcade. If there is a valley, build a bridge. Not a simple bridge but one with a water channel laid on top. The three-tier arch of the Pont du Gard is the typical example. The grade of the channel above is extraordinarily precise. Across the entire fifty-kilometer length, the drop is only about 17 meters. That works out to a slope of about 0.034 percent — that this level of precision was reached without modern GPS, with only the human eye and basic tools, is astonishing.³
Siphon. Some valleys are too deep or too long to be bridged by an arcade. Here Rome used the inverted siphon. Plunging the water down a steep drop and using pressure to push it back up the other side. The physics is simple, but realizing the principle at giant scale required tremendous engineering. Pressure-resistant sealed pipes, vents to release air bubbles, devices to manage water quality.
Lead pipe (fistula plumbea). For distribution within a city, Rome used pipes made of lead. Lead is easily worked and seals perfectly. There was, however, one problem — lead poisoning. Some scholars trace the high incidence of nervous-system disorders in the Roman upper classes to these pipes.⁴ The logic of accepting poison for the sake of convenience was at work, quietly, here too.
The most important official document on Roman aqueduct administration is the De Aquaeductu Urbis Romae of Sextus Julius Frontinus (40–103 CE).⁵ Composed around 97 CE, this is the practical report of an administrative engineer overseeing Rome's water supply.
One of his sentences is particularly famous. "Compare with such an array of indispensable structures carrying so many waters the idle Pyramids or the useless, though renowned, works of the Greeks."⁶
Three things are bound into a single sentence. Pride. Functionalism. And contempt for other civilizations.
This sentence compresses Rome's worldview. Pyramids are ritual structures. Greek temples are religious-aesthetic architecture. They are "not useful." By contrast, Roman aqueducts "carry indispensable water." Utility is the measure of all value. And that utility becomes a basis for disparaging other civilizations.
What Frontinus did not know — that in his own age, on the other side of the planet, hydraulic systems of comparable scale were being built. The irrigation works of Tiwanaku were already in operation in the Andes; nine hundred years later the Inca, inheritors of that tradition, would complete 40,000 kilometers of road and canal network. If Frontinus could have seen that, what would he have said? He would probably have judged it, like the pyramids, "useless." Because it would have lain outside his measure.
Roman aqueducts reveal not just a technique but a specific worldview. The core concept of that worldview is dominium.
Dominium is a central concept of Roman law. Absolute ownership. Over land, over things, sometimes over people. The holder of dominium has the rights of use (usus), fruits (fructus), and disposal (abusus). To use as one wishes, to take the gain, to destroy or sell.
This legal concept was applied to relations with nature. A Roman emperor or aristocrat could hold dominium over a particular river's water. To divert that water to a distant city was the building of an aqueduct. Water could be owned, transported, sold. Water was a resource.
Coupled with this is imperium. Absolute rule. The authority of the emperor over territory. Where dominium is the concept of ownership, imperium is the concept of dominion. Joined together, ownership over nature becomes rule over nature.
This grammar is inscribed in the material form of Roman aqueducts. The arcade conquers the valley. The siphon compels gravity. The lead pipe distributes the water. Everything has the structure of commanding from above downward. From source to city. To the imperial palace first, then the public baths, then the public fountains.
And Rome invented the sewer.
Cloaca Maxima, "the great sewer." Begun in the sixth century BCE to drain the lowland marshes of Rome, expanded across centuries into an underground sewage system. Part of it is still functioning today as part of Rome's sewage handling.⁷
The invention of sewers looks like obvious progress. Used water gathered up and sent out of the city. Hygienic and reasonable. But we easily miss the weight of worldview contained in this invention.
The sewer presupposes the concept of water-to-be-discarded. Once used, water is polluted, useless, and so must be removed. This very premise is a particular ontology. Water can be classified by use, and discarded when its use is over. The water's full journey — from cloud to rain, to river, to sea, evaporating again to cloud — is invisible within this classification.
This is one of the roots of modern Western environmental thought. The thinking that divides nature into "resource" and "waste." This thinking, beginning at the Cloaca Maxima of Rome, transforms but continues, all the way to the throwaway society of the twentieth century.
Before going to look at the Inca path, we must pause. Because if we want to say that the Inca chose "engineering that walks with nature," we have to show that another civilization in the same Andes had made a different choice first. To prove that the Inca path was a deliberate choice and not a limit of technology, we need the fact that an earlier empire had already gone in the opposite direction.
That empire is Wari.
Wari is the pan-Andean empire that existed roughly from 600 to 1100 CE, some four hundred years before the Inca.⁷ᶜ Its center was near present-day Ayacucho, Peru. At its peak it held a sphere of thousands of kilometers from near Cajamarca in the north to near Lake Titicaca in the south. Smaller than the Inca, but most of the administrative, road, architectural, and hydraulic technology the Inca inherited came from the Wari.
The Inca are also the inheritors of the Wari. This is a first crucial layer of Andean civilizational history.
Wari water management shows a logic startlingly similar to that of Rome.
The clearest evidence is Cerro Baúl. A city on a 2,500-meter cliff in the Moquegua valley of southern Peru. The southern outpost of the Wari. The city was a natural fortress surrounded on all sides by cliffs, with one fatal problem. There was no water.
How did Wari engineers solve this? They brought water from a spring some twenty kilometers away. Along the flank of the cliff they built a stone-and-clay canal so that water flowed by gravity all the way up to Cerro Baúl. From carbon dating and stone analysis, modern archaeologists have confirmed that the canal operated for several centuries.⁷ᵈ
This is structurally identical to the logic of the Roman aqueduct. Overcoming the terrain by technology, drawing distant water into a city. The only reason to put the city on the cliff was military and symbolic — a visual declaration that Wari elites stood above the conquered. If nature is inconvenient, you remove the inconvenience by engineering.
Modern Wari researchers, including Donna Nash, summarize this with the phrase "sculpting the landscape."⁷ᵉ The Wari did not adapt to the land. They sculpted it to fit their intent.
Another decisive case is Pikillaqta. About thirty kilometers southeast of Cusco. A site I passed through the window of the car on the way to Tipón.
Pikillaqta is a rectangular grid city of 745 by 630 meters.⁷ᶠ Hundreds of identical rooms arranged geometrically. The natural terrain is largely ignored. Where there was a hill, it was leveled; where a valley, it was filled. The axis of the city is determined not by feng shui or astronomy but by bureaucratic efficiency.
This is the opposite of Inca Cusco's spatial philosophy. The Inca did not lay Cusco out as a grid. The city followed the terrain in the shape of a puma. Pikillaqta, thirty kilometers from Cusco, is a grid; Cusco itself is a puma. The Inca knew the Wari grid and rejected it.
This information matters for a simple reason. Wari was a militarily expansionist empire. The eminent Wari scholar Luis Lumbreras wrote that "the Wari economic policy was directed toward the exploitation of the colonized peoples."⁷ᵍ Some scholars call this mode of rule "conquest by hydraulic superiority." Whoever controls the water controls the people. The technology of lifting water to a city on a mountain becomes the language of power.
This carries an important implication.
Engineering arrogance toward nature is not the invention of Western modernity. Nor is it the invention of Greece or Rome. It is a pattern that imperial structures repeatedly fall into. Wari followed the pattern. In the same Andes, before the Inca.
And because of this fact, the Inca's different choice was conscious. The Inca did not simply not know what the Wari had done. They saw it, knew it, inherited it, and chose another road.
That choice is the next section.
Now I return to the other photograph. The canal of Ollantaytambo.
The Inca canal network was in no way smaller than Rome's. The Qhapaq Ñan ran some 40,000 kilometers. A great portion of it was integrated with irrigation canals. Machu Picchu, Ollantaytambo, Pisac, Tipón — each city had a precise water-supply system. We have already seen the thirteen tiers and the critical-flow fountain of Tipón.
And all of this was built without wheels, without iron, without arches.
It is not that Inca engineers did not know the wheel. Inca children's toys had wheels. They knew the principle but did not adopt it for transport.⁸ Same with iron. They had bronze, but did not smelt iron at scale. The arch was used by Egypt, Mesopotamia, and Rome alike, and the Inca did not use it. As we saw, the Wari built a twenty-kilometer canal on a mountain to overcome the terrain. The Inca knew this and did not follow that road.
To call this "backwardness" is the habit of Western-centric historiography. But it can be seen otherwise — the Inca had no need for these technologies. Because they did not try to go against nature.
The wheel is useful on a straight road. The Inca did not pierce mountainous terrain in straight lines. They zigzagged. On a stepped stone road, the wheel becomes inconvenient. The hooves of the llama and the foot of the human are more efficient. Iron is needed for digging deep into the earth or building giant structures. The Inca did not dig deep, did not build giant sealed structures. The arch is needed to send water horizontally across a valley. The Inca did not cross valleys. They let water follow the curves of the valley.
So what looks like a technical lack in the Inca was not a flaw but the choice of a different grammar.
The basic form of the Inca canal is open. The stone channel is exposed at the surface, and the water flows in air and sunlight. The opposite of Rome's sealed lead pipes.
Why open?
First, because water shows its own state. Clear or cloudy, plentiful or scant, polluted or healthy — the human eye can see. There is no barrier between water and human. This is a philosophy of transparency.
Second, because purification happens naturally. Sunlight kills pathogens with ultraviolet. Air contact supplies oxygen. Microbes on the stone surface decompose organic matter. The water purifies itself as it flows. Without human intervention.
Third, and most important — the very purpose of drainage was different.
Let us not fall into a simple dichotomy. The claim that "the Inca had no sewers" is inaccurate. The royal residence at Machu Picchu had elite sanitary drainage. Kenneth Wright's thirty years of research have shown this. Only — the outlet of those drains was designed precisely so as not to contaminate sources of drinking water or agricultural fields.⁷ᵃ The Inca too had solved drainage technically.
The crux is not technical but philosophical.
The Roman cloaca is a system "for discarding." Used water is driven out of the city, downstream. The end is the river or sea. Which is to say — the burden is shifted onto another.
Inca drainage is a system "for circulating." The water used in the temple ceremony goes → to the noble residential area → to the commoner's home → to the agricultural terraces → and back to the natural water of the lower watershed. At each step the water is reused, and at the end it returns to the earth.
The starkest contrast is in the disposal of ordinary household waste. In Inca villages, waste was — collected in chamber pots and returned to the field as fertilizer. The Quechua word wanu means "excrement" and "fertilizer" at the same time. The English word guano derives from this word.⁷ᵇ Wanu is — not something to be thrown away. It is one stage of the circle. What leaves the body returns to the earth, becomes plants, and reenters the body.
The Roman cloaca is what is discarded. The Inca wanu is what nourishes.
This contrast compresses the entire chapter. The civilization that saw water as ownership (dominium) and the one that saw water as relation (ayni). The civilization of use-and-discard and the one of use-and-return. There is no doubt which lineage modern urban sewer systems descend from. Each day we send tons of water down toilets and sinks. That water is processed and discharged into rivers. Tremendous energy and cost are invested, and the discharged water is rarely fully clean. An Inca would not have understood this waste. Why break the circle?
Behind Inca hydraulic engineering stands the concept of yaku. The Quechua word I touched briefly while walking the canals of Tipón I return to here.
Yaku is translated "water," but the meaning is far broader. Yaku is living water. Water that flows, water that springs, water that falls as rain, water that evaporates, and the fluids that circulate inside the body. All of these are yaku. And these are not separate substances. They are different faces of one being.
To the Inca, water was not the pronoun "it" but the pronoun "thou." The distinction is decisive. "It" can be used as a tool, treated as a resource, discarded when soiled. "Thou" cannot. The being you face as "thou" you must, in some way, be in relation with.
The form of relation is ayni. Reciprocity. If I receive from water, I return to water. If I have used it, I let it flow again. If I have purified water, water has purified me. When this give-and-take is broken, the relation dies. When the relation dies, the being dies. When the being dies, the community dies too.
This is the metaphysical structure of the Inca canal. And this structure is inscribed in every stone.
Now to compare the fates of the two engineerings.
In 476 CE, the Western Roman Empire fell officially. The year the last emperor, Romulus Augustulus, was deposed. After that, Europe entered about 500 years of the early Middle Ages. During this period, most of Rome's water-supply infrastructure lost function.
Why?
Roman aqueducts could not be maintained without central authority.⁹ Three things were necessary to operate the system. A specialized engineering class. Continuous funding. And political will. All three depended on the political system of the Roman Empire.
Who were the specialists? In Rome, aqueduct administration was organized under the high office of the curator aquarum. The very office Frontinus held. Beneath him were specialized workers called aquarii. At first slaves, later free men. They found leaks, cleared blockages, repaired damage.
When the empire fell, this organization disintegrated. The Germanic kingdoms that arose lacked the administrative capacity to maintain Roman aqueducts. Offices vanished, workers scattered, budgets stopped. Within decades, leaks and blockages accumulated and the aqueducts lost their function.
A centralized system collapses with its center.
Second, users could not perform maintenance. Roman aqueduct technology was the monopoly knowledge of an elite. Ordinary citizens knew only how to receive water at a fountain; they did not know where the water came from, or how to repair it. When the aqueduct broke, they had no choice but to dig wells or draw from rivers. The stones of the old aqueducts became material for new buildings. Massive Roman structures like the Pons Pontificalis became quarries in the early Middle Ages.
This is the same pattern as Sacsayhuamán becoming the quarry for colonial Cusco. The stones of a destroyed civilization become building material for the next. History repeats.
By contrast, the Inca canals did not lose function after the empire fell.
After the Spanish took Cusco in 1533, the Inca state ceased to exist officially. The administrative system changed, religion was forced into Catholicism, even the language was partly replaced by Spanish. And yet the canal at Ollantaytambo still flows today. The sixteen fountains of Machu Picchu still work. The thirteen tiers of Tipón have functioned for five hundred years.
How is this possible?
The answer lies in the difference of management structure. Where Roman aqueducts depended on the imperial central bureaucracy, Inca canals were managed directly by local communities (the ayllu). A village's canal was the responsibility of its villagers. Each year, before the rainy season, a day for communal labor was set, and the canal was cleaned and broken stones replaced. This communal labor was called faena or minka.¹⁰ Participating in faena was not merely an obligation. It was a ritual proving membership in the community. To miss it was to lose the right to use the water.
There was nothing for Spain to take in this structure. The central administrative system could be torn down, but the custom of faena, dispersed across villages, was hard to dismantle. Whatever the Spanish viceroyalty did in Cusco, in some highland village people still gathered before the rains to lift the stones of the canal. As children grew into adults, and adults aged into grandparents, the custom of cleaning the canal was passed down the generations.
This continues into the twenty-first century. Old villagers in the Peruvian highlands do not think of themselves as protecting an "Inca legacy." They simply manage their own village's water. When faena day comes, residents go out with shovels, clear the leaves from the canal, dislodge a blocking stone. That evening they share a cup of chicha together. Just as it was five hundred years ago.
The strength of distributed responsibility — that is why the Inca canals survived.
This difference is anything but abstract for us.
In the twenty-first century we still live atop Roman-style centralized infrastructure. Power supplied by large generating stations. Water from giant dams. The internet dependent on central servers. These systems are efficient. Economies of scale work. They are technically refined.
And yet these systems are fragile. A single point of failure can paralyze the whole. The 2003 North American blackout (about 50 million people without power), the 2016 Ukrainian grid hack, recent cyberattacks of every kind — all reveal the systemic vulnerability of centralized systems.
The Lebanese-American economist Nassim Nicholas Taleb took up this problem in his 2012 book Antifragile.¹¹ His distinction is useful. A fragile system breaks under shock. A robust system endures shock. And an antifragile system grows stronger from shock. Can such a system exist? Taleb's answer is yes. The clearest example he offers is a distributed system. A structure in which, when a shock comes, the failure of some parts allows the whole to keep functioning, and the failures are learned from in the next iteration.
The Inca canal is precisely this kind of system. It has weathered many shocks across five hundred years (conquest, civil war, independence, republic, industrialization), and survived them all. Because there was no center. In Taleb's vocabulary, the Inca canal is antifragile.
The Roman aqueduct, by contrast, was fragile. When the empire fell, it fell with it.
What does this contrast imply for the present? Water, electricity, food, communication — every essential infrastructure we have designed in the Roman manner. The infrastructure revolution of the past century has been almost entirely in the direction of centralization. The benefits are great. So is the cost — we have forgotten how to know the canal of our own village. Forgotten how to make the electricity of our own village. Forgotten how to grow the food of our own village.
In an age of climate crisis and geopolitical instability, perhaps we must look again to the Inca. Not to tear down what is large. To build small things alongside the large. Village-scale water systems. Distributed solar. Local food networks. Technically these are already possible. What is hard is political imagination. The imagination of "distributed responsibility" that we have lost over five hundred years.
The difference of techniques is, in the end, a difference of worldviews. The last task of this chapter is to trace, briefly, the philosophical roots of the two worldviews. The detailed genealogy will be taken up in chapter eight ("A Single Genealogy"); here I compare only the seeds of the two roots.
In the Western tradition, the seeds of "dominating nature" are in many places.
The oldest is Genesis 1:28. "And God blessed them, and God said unto them, Be fruitful, and multiply, and replenish the earth, and subdue it. And have dominion over the fish of the sea, and over the fowl of the air, and over every living thing that moveth upon the earth." The Hebrew kabash means strong subjugation; radah, the rule of a conqueror.¹²
The American historian Lynn White Jr., in his 1967 article in Science, "The Historical Roots of Our Ecologic Crisis," identified this passage as the theological origin of the West's ecological crisis.¹³ White's thesis was simple: Christianity defined nature as an object created for human use, and this definition, joined with the scientific revolution, supplied the legitimacy of unlimited exploitation of nature. Controversial, but influential.
And yet — White's reading has provoked half a century of fierce debate. Before leaving this chapter, the three main strands of rebuttal must be touched briefly. Otherwise this account itself becomes a forty-years-out-of-date scholarly simplification.
First, the stewardship reading. The interpretation that the Hebrew radah means not violent rule but responsible care as God's deputy. In ancient Israel the duty of the king was covenantal responsibility — especially the protection of the weak. The classic statement of this is Douglas John Hall, Imaging God: Dominion as Stewardship (1986).¹⁴
Second, the different language of Genesis 2:15. Reading only Genesis 1:28 is selective reading. In Genesis 2:15 the Hebrew abad (to serve) and shamar (to keep) are the language of farmer and shepherd. The Bible itself contains two creation narratives, and the central verb of the second is not "subdue" but "serve and keep." Theodore Hiebert's The Yahwist's Landscape (1996) has systematized this view.¹⁵
Third, and the most scholarly powerful rebuttal, belongs to the British New Testament scholar Richard Bauckham. In The Bible and Ecology (2010), Bauckham argues that the tradition of reading Genesis 1:28 as a command for exploitation is not ancient Christian; it is a modern reading invented in the seventeenth century by Francis Bacon.¹⁶ In patristic and medieval theology the prevailing view was rather of a community of creatures. The thirteenth-century Cantico delle Creature of St. Francis of Assisi is evidence.
These rebuttals do not fully invalidate White's thesis. The history by which modern Western civilization used Genesis 1:28 as ground for conquest cannot be denied. But what the debate makes clear is one thing — the problem was not the biblical text but the way of reading it. The way of reading was shaped by Greek dualism, Roman legal thought, and above all by the interpretive frame of the modern scientific revolution after the seventeenth century.
Interestingly, the Eastern Orthodox tradition went a different way. It preserved a cosmic-liturgical tradition seeing the human as the "priest of creation." The role of the human is not to dominate creatures but to be the mediator of the praise creatures bring to the Creator. This tradition connects directly to the theology of perichoresis in the latter part of this book (chapter fifteen). Christianity is not one thing. Mainstream and alternative traditions have always existed simultaneously, and the readings we choose make the outcomes.
Greek philosophy planted seeds in the same direction. Plato's dichotomy of idea and hyle created the hierarchy of mind above matter. Nature is the realm of matter, and matter is a passive medium waiting to be ordered by form. Aristotle's natural hierarchy (mineral → plant → animal → human → god) systematized this scheme biologically. Within this hierarchy, the lower exists for the higher.
Rome gave this Greek hierarchy legal form. The distinctions of dominium, imperium, and persona (legal subject). These distinctions made clear who was master and who was object. Medieval Europe inherited Roman law. Modern Europe extended Roman law into international law. When Europeans "discovered" the Americas, the legal tool they carried was this concept of dominium. Locke's terra nullius ("land belonging to no one") is its modern extension. We will treat it in detail in chapter ten.
The seeds of the Andean tradition are very different.
The most fundamental concept is Pacha. The word is hard to translate. "Time," "space," "world," "order," "state" — all at once. Pacha is the integration of time and space. Not separated time and separated space, but being as a single flow.
Pacha consists of three layers. Ukhu Pacha — the underworld, the world of root and seed. Kay Pacha — the present world, where humans, animals, and plants live. Hanan Pacha — the world above, of sun, stars, and gods. These three layers are not separated. The root grows into a sprout, the sprout into a tree, the leaf falls and becomes root again. By the same principle, the energy of Ukhu Pacha rises into Kay Pacha, and the energy of Hanan Pacha descends into Kay Pacha. Being also circulates vertically.
Paired with this Pacha philosophy is the ethic of ayni. Reciprocity. Every receiving must answer with a giving. If you receive crops from the land, you return offerings to the land. If you draw water for drinking, you keep the channel clean. If you have the protection of the mountain (Apu), you offer prayer to the mountain. When this reciprocity is broken, the relation dies.
And ayllu, the community. The basic unit of Andean society is not the individual but the ayllu. The ayllu is a kin group, and at the same time a group spiritually bound to a particular land, water, and mountain. To care for the canals of one's ayllu was inseparable from caring for oneself.
One important point must be marked. These two roots — the hierarchical worldview of the West and the cyclical worldview of the Andes — are both complex, both ancient, both developed. Neither is more "primitive" or more "advanced."
The Western tradition itself contains rich alternative genealogies. Heraclitus ("all things flow"), the Stoic accord with nature, St. Francis's fraternal love of creation, Romanticism, the late philosophy of Whitehead and Heidegger. All of them voices of the West pushed to the margins of the mainstream.
The Andean tradition has its shadows too. The Inca Empire expanded by conquest, used the policy of mitima (forced relocation) on conquered peoples, and had a strict class hierarchy. A cyclical worldview does not automatically produce an egalitarian society.
And yet, despite this complexity, the dominant tendencies of the two civilizations are distinct. The mainstream of the West took the road of separation and dominion. The mainstream of the Andes took the road of circulation and accompaniment. And what materializes this divergence is each civilization's hydraulic engineering. Arch and sewer. Open canal and circular structure.
One sentence I wish to add before going on.
Medicine, too, took two roads. The road of eradicating pathogens and the road of caring for the immune system. The first led to the triumph of germ theory and antibiotics; the second to the rediscovery of lifestyle medicine and microbiome research. The branching of these two roads, perhaps, was already determined five hundred years ago in the way of handling water. The argument for this will be unfolded in chapter seventeen. Here, only the point that water and body are the same story.
The two photographs are still on the desk.
The Pont du Gard is a museum ruin. The French government keeps it as national heritage. UNESCO World Heritage (inscribed 1985). Thousands of tourists visit every day. But no water flows along the channel above the arches. It stopped fifteen hundred years ago.
The canal of Ollantaytambo is still active duty. It is also UNESCO World Heritage (as part of the Inca's "Sacred Valley"), but it is not a museum. Beside the tourists taking photographs, an old woman of the village does her laundry in that water. That water has not stopped a single day in five hundred years.
This difference cannot be coincidence.
And it is not a question of "who is right." Roman aqueducts were a great achievement. Inca canals were a great achievement. Both worked in their context. The problem lies elsewhere.
Five hundred years ago, when these two philosophies met, what happened?
This is the question of the next chapter. November 16, 1532, in the small highland town of Cajamarca in northern Peru. The day Emperor Atahualpa and 168 Spaniards stood for the first time in a single plaza. In that brief afternoon, two worldviews physically collided. The result of that collision made the next five hundred years of the world.
In the next chapter we go to that day.
¹ On the scale of Roman aqueduct construction see A. Trevor Hodge, Roman Aqueducts and Water Supply, 2nd ed. (London: Duckworth, 2002), introduction. The figure of about 500 is approximate and varies among scholars.
² On the list of the eleven major aqueducts of Rome and total length see ibid., chapter 3.
³ On the precision of the Pont du Gard's gradient see Guilhem Fabre, Jean-Luc Fiches, and Jean-Louis Paillet, L'aqueduc de Nîmes et le Pont du Gard (Paris: CNRS Éditions, 2000).
⁴ The effects of lead poisoning from Roman lead pipes are debated. Jerome Nriagu, "Saturnine Gout among Roman Aristocrats," New England Journal of Medicine 308 (1983): 660-663, is a classic study; recent counterarguments hold that the effect was overstated.
⁵ Frontinus, De Aquaeductu Urbis Romae. English translation: Frontinus: The Aqueducts of Rome, trans. Charles E. Bennett, Loeb Classical Library (Cambridge, MA: Harvard University Press, 1925).
⁶ Ibid., I.16. Translation by the author.
⁷ On the history and present of the Cloaca Maxima see Nicholas Purcell, "Rome and the Management of Water," in Human Landscapes in Classical Antiquity, ed. Graham Shipley and John Salmon (London: Routledge, 1996), pp. 180-212.
⁷ᵃ On the sanitary drainage system of Machu Picchu see Kenneth R. Wright and Alfredo Valencia Zegarra, Machu Picchu: A Civil Engineering Marvel (Reston, VA: ASCE Press, 2000), chapters 4 and 6. For Wright's remarks on the design of drain outlets see PBS NOVA, "Ghosts of Machu Picchu" (2010).
⁷ᵇ On the etymological link between Quechua wanu and English guano see standard Spanish etymological dictionaries and the Real Academia Española entry. The bird-droppings industry on the Peruvian coast spread the word to nineteenth-century Europe.
⁷ᶜ For an overview of the Wari Empire see William H. Isbell and Gordon F. McEwan, eds., Huari Administrative Structure: Prehistoric Monumental Architecture and State Government (Washington, D.C.: Dumbarton Oaks, 1991); Justin Jennings, ed., Beyond Wari Walls: Regional Perspectives on Middle Horizon Peru (Albuquerque: University of New Mexico Press, 2010).
⁷ᵈ On the Cerro Baúl canal system see Ryan Williams, "Cerro Baúl: A Wari Center on the Tiwanaku Frontier," Latin American Antiquity 12, no. 1 (2001): 67-83; Donna J. Nash and P. Ryan Williams, "Architecture and Power on the Wari-Tiwanaku Frontier," Archaeological Papers of the American Anthropological Association 14, no. 1 (2004): 151-174.
⁷ᵉ Donna J. Nash, "Sculpting the Landscape: Wari Terraforming and the Politics of Labor," in Andean Lifeways: Essays on the Peruvian Past, ed. Joanne Pillsbury (Washington, D.C.: Dumbarton Oaks, 2012). The phrase "sculpt the landscape" has become standard in Wari studies.
⁷ᶠ On the planned-city structure of Pikillaqta see Gordon F. McEwan, Pikillacta: The Wari Empire in Cuzco (Iowa City: University of Iowa Press, 2005).
⁷ᵍ Luis G. Lumbreras, The Peoples and Cultures of Ancient Peru, trans. Betty J. Meggers (Washington, D.C.: Smithsonian Institution Press, 1974); the quotation is the author's summary of Lumbreras's analytic context for Wari. The phrase "conquest by hydraulic superiority" is an analytic term derived from Karl Wittfogel's Oriental Despotism (1957) and applied in Andean studies.
⁸ On Inca non-use of the wheel see Terence D'Altroy, The Incas, 2nd ed. (Malden, MA: Wiley-Blackwell, 2014), pp. 243-245.
⁹ For recent research on the causes of Roman aqueduct collapse see Paolo Squatriti, Water and Society in Early Medieval Italy, AD 400-1000 (Cambridge: Cambridge University Press, 1998).
¹⁰ On Andean traditions of communal labor (faena, minka) see Paul H. Gelles, Water and Power in Highland Peru: The Cultural Politics of Irrigation and Development (New Brunswick: Rutgers University Press, 2000).
¹¹ Nassim Nicholas Taleb, Antifragile: Things That Gain from Disorder (New York: Random House, 2012).
¹² On the interpretation of the Hebrew verbs kabash and radah in Genesis 1:28 see Theodore Hiebert, The Yahwist's Landscape: Nature and Religion in Early Israel (Oxford: Oxford University Press, 1996).
¹³ Lynn White Jr., "The Historical Roots of Our Ecologic Crisis," Science 155, no. 3767 (1967): 1203-1207.
¹⁴ Douglas John Hall, Imaging God: Dominion as Stewardship (Grand Rapids, MI: Eerdmans, 1986). A systematic statement of the stewardship reading.
¹⁵ Theodore Hiebert (1996), op. cit. See especially the discussion of abad and shamar in Genesis 2:15. For an agrarian-biblical synthesis on the same theme see Ellen Davis, Scripture, Culture, and Agriculture: An Agrarian Reading of the Bible (Cambridge: Cambridge University Press, 2009).
¹⁶ Richard Bauckham, The Bible and Ecology: Rediscovering the Community of Creation (Waco, TX: Baylor University Press, 2010). Especially chapter 1, which argues in detail that the "dominion" reading of Genesis 1:28 is a modern reading reconstituted in the seventeenth century by Bacon.