Why the Indus Civilization Could Adaptbut Its Cities Couldnt

At its height, the Indus civilization built some of the largest and most carefully planned cities of the Bronze Age. Streets followed deliberate layouts. Houses were tied into drainage systems. Craftspeople worked stone, shell, metal, faience, and carnelian. Merchants used standardized weights and seals across an enormous region. At Dalavera, engineers captured scarce seasonal water in a Then, after about 1900 BCE, that urban world began to come apart. The great cities shrank. Some were eventually abandoned. Writing, seals, standardized weights, monumental non-residential construction, and much of the long-distance exchange that had characterized the mature Harapan period faded from the archaeological record. But the people did not simply disappear. Across many parts of the old Harapan world, smaller settlements continued. In some regions, they became more numerous. Farming continued. Craft knowledge survived. Communities changed what they grew, where they lived, and how they organized production. That creates a different mystery from the familiar question of who destroyed the Indus civilization. If the Harapans were incapable of adapting, how did so many communities survive? And if they could adapt, why could their cities not survive with them? To answer that, we need to stop treating a Harapan city as a collection of impressive ruins. A city was a system, and before we can understand why that system unraveled, we need to understand what kept it alive. A city like Harapa, Mohenjo-Daro, or Dalavera was not simply a village that had grown very large. Scale changed the problem. In a small farming community, many households can participate directly in producing food. A large city concentrates people who do not. Craft specialists, traders, builders, and other urban residents can spend more of their time away from farming only if somebody else is producing enough food to support them. That means urban life depends on surplus. But surplus is only the first step. Food has to be produced in the right quantities. It has to move from fields to consumers. Raw materials have to reach workshops. Finished goods need markets. Construction requires labor to be organized beyond a single household. And when exchange extends across hundreds or thousands of kilometers, people need ways to make transactions intelligible to strangers. The archaeological record shows that the Harapans solved many of these problems with remarkable consistency. Their settlements contain standardized weights and measures. Seals appear across cities and smaller sites. Craft debris shows highly developed production of beads, bangles, tools, ornaments, and other goods. Large public or non-residential structures required coordinated labor on a scale far beyond ordinary domestic construction. None of this proves that an all-powerful king directed the system. In fact, that is one of the most unusual things about the Indus world. Unlike Egypt or Mesopotamia, archaeologists have not found obvious royal tombs, enormous palaces, or monuments celebrating individual rulers on the same scale. The political structure remains debated. Some researchers argue that coordination may have depended more on collective institutions and overlapping social groups than on a single ruling dynasty. The problem is that the Harapan script remains undeciphered, so the people who built these cities cannot yet explain their institutions to us in their own words. Archaeology gives us the physical results of coordination without a constitution describing who ordered it, who negotiated it, or how authority was divided. That uncertainty matters. We should not fill the silence by inventing a pharaoh, a merchant oligarchy, or a centralized bureaucracy simply because other Bronze Age civilizations had them. But whatever the political arrangement was, coordination clearly existed. Someone had to agree on what a weight meant. Builders had to follow shared conventions. Traders had to recognize seals and measures. Craftspeople depended on supplies that could come from far outside the city. And large concentrations of people had to be fed year after year. Dalavera makes the dependency especially visible. The city stood on Kadir Island in what is now Gujarat, in an environment where water was scarce and highly seasonal. Instead of simply sitting beside a dependable perennial river, its inhabitants developed an elaborate system for capturing and storing water. Reservoirs were built around the settlement. Drains and channels moved water through an engineered landscape. The city also occupied a position useful for accessing raw materials and participating in trade that connected the Harapan world to regions as distant as Oman and Mesopotamia. The achievement is impressive. But the more important point is what the achievement required. A reservoir is useful only if it is maintained. A standardized weight is useful only if other people accept the same standard. A specialist workshop is viable only if its producers can obtain food and raw materials and find consumers for what they make. A long-distance trade route matters only if enough people continue to use it. Urbanism therefore creates a peculiar kind of strength. By connecting people, resources, skills, and information, it allows a society to do things that small communities cannot easily do alone. But connection also creates dependency. The more specialized the system becomes, the more each part depends on other parts continuing to function. A farmer depends on predictable water and demand. A craft specialist depends on agricultural surplus. A trader depends on producers and consumers. A city depends on all of them at once. This does not make cities inherently fragile. For centuries, the mature Harapan system worked well enough to support large settlements across a vast and environmentally diverse region. And that is precisely why the next part of the story matters. The Indus civilization had never lived in a perfectly stable environment. Different parts of its world received different rainfall, relied on different water sources, and cultivated different combinations of crops. Harapan communities had already developed ways to manage uncertainty long before the great urban centers began to decline. So climate change alone cannot be the explanation. The real question is what happened when the strategies that helped individual communities survive no longer lined up neatly with the requirements of the cities they had built. The Harapan world was too large and too environmentally diverse to have depended on a single way of producing food. Its settlements stretched from the plains of Punjab and sinned to the semi-arid landscapes of Haryana and Rajasthan and south into Kachh and Gujarat. Rain did not fall at the same time or in the same quantities across this enormous territory. Rivers behaved differently from one region to another. Some communities lived close to major waterways. Others depended far more heavily on seasonal rainfall. So when we talk about Harapan agriculture, we are really talking about several overlapping agricultural strategies. And those strategies reveal something important. The Harapans already knew how to live with uncertainty. At Harappa itself, the archaeobotanical evidence published so far is dominated by winter crops, especially wheat and barley. These were productive cereals capable of generating large quantities of food. That made sense for a major urban center. Feeding a concentrated population rewards crops that can produce substantial and relatively predictable surpluses. But Harappa was not Gujarat. In the far drier southwest, farmers faced a different problem. Rainfall was concentrated into the summer monsoon, drought was recurrent, and the margin for error could be much smaller. Archaeobotanical evidence from sites such as Rajdhi shows farmers using combinations of crops rather than relying on a single agricultural calendar. Over time, the mix changed. Summer crops such as millets, rice, pulses, and sorghum became increasingly important at some sites, while winter cereals continued in smaller rolls. This was not simply agricultural backwardness replacing sophisticated farming. It was flexibility. Different crops respond differently to heat, rainfall, soil, and the timing of the monsoon. Planting across more than one season can spread risk. If one set of conditions fails, another crop may still produce something. In the northeastern part of the Harappan world, village communities followed still other strategies. Some combined winter and summer crops. Others used forms of sequential cultivation and intercropping, growing different plants in the same agricultural system rather than betting everything on one harvest. There was no single Harappan solution because there was no single Harappan problem. This matters because resilience is often misunderstood. We tend to imagine a resilient civilization as one that preserves its existing institutions despite changing conditions. But resilience can also mean changing the institution itself. A household might shift crops. A pastoral community might alter where animals graze. A village might change planting seasons, rely more heavily on local resources, or reduce its dependence on distant markets. Population can move. An entire city cannot do any of those things as easily. Dalavera shows just how far Harappan communities could go to make a difficult environment habitable. The city stood on the arid island of Kadir, where permanent access to water was a fundamental constraint. Its inhabitants captured water from seasonal streams and runoff through an extraordinary system of reservoirs, channels, and drains. They did not passively occupy a favorable environment. They engineered around an unfavorable one. And they maintained this settlement for many centuries. That alone should make us suspicious of any explanation in which Harappan urban society simply encounters dry conditions and immediately collapses. These were people accustomed to variation. They understood seasonal scarcity. They developed different farming systems in different landscapes. They moved materials across enormous distances. They built infrastructure to capture water where nature did not conveniently provide enough of it. So environmental stress matters, but only if we explain what kind of stress it created and why existing adaptations eventually stopped being sufficient for the urban system. There is another reason the distinction matters. Not every agricultural strategy solves the same problem. Imagine two farmers. The first concentrates labor on a small number of highly productive crops. Under suitable conditions, that strategy can create large surpluses. The second grows a broader combination of crops across different seasons. The second farmer may produce less of any single crop, but if rainfall arrives late, one harvest fails, or economic demand changes, there are more alternatives. Neither strategy is inherently superior. One can maximize output, the other can spread risk. And the difference becomes much more important once cities enter the equation. A village needs enough food to survive. A metropolis needs enough surplus to support thousands of people who may not be producing their own food at all. That creates pressure towards scale. The same concentrated production that helps make urban specialization possible may also reduce the number of fallback options available when conditions become unpredictable. The archaeological evidence is not complete enough to say every Harapan city followed exactly this pattern. Even at Harappa, only part of the plant assemblage has been fully published. But the contrast between regions is difficult to ignore. Some agricultural systems appear strongly shaped by the demands of large-scale production. Others retained much greater flexibility at the local level. And as the mature Harapan period drew toward its end, that flexibility would become increasingly valuable. Because the next challenge was not simply that the climate became worse. The more dangerous change was that water became less reliable across different parts of the Harapan world in different ways, and for periods long enough that communities could no longer treat the problem as a temporary bad season. The question was no longer how to survive one failed harvest. It was how to organize life when the assumptions behind the old system could no longer be trusted. The problem was not that the Indus world suddenly became a desert. It was that the old patterns of water became less dependable for long enough to change the choices people had to make. For a farming society, rainfall is only one part of that problem. What matters is whether rain arrives in the right season, whether rivers carry enough water, whether floods reach fields at useful times, and whether those patterns remain predictable enough for households to plan the next harvest. Across the Harapan world, those conditions were never identical. The Indus and its tributaries drew water from different combinations of monsoon rainfall, winter precipitation, and snowmelt. Farther south, communities in Saurashtra depended on a different hydrological regime. A drought that seriously reduced water in one region could leave another comparatively less affected. That regional variation is crucial. Recent paleoclimate and hydrological research does not point to one single drought that struck every Harapan settlement with the same force. Instead, it suggests repeated episodes of severe water stress stretching across centuries. A 2025 study combining high-resolution paleoclimate archives with river flow modeling identified four major drought intervals between roughly 4,450 and 3,400 years before present. Each lasted more than 85 years, and their intensity and geography differed. The final major drought identified in that study lasted more than a century. Think about what that means in human terms. A five-year drought can be remembered as a crisis. A drought pattern extending across decades changes expectations. A century-scale shift can outlive the people who first experienced it. Children grow up under conditions their grandparents considered exceptional. Farming practices change. Settlement decisions change. Infrastructure designed around older patterns becomes harder to justify. What begins as emergency behavior can become the new normal. Other paleoclimate evidence points in the same direction, including repeated weakening of summer and winter rainfall around the broader period of Harapan transformation. That matters because Harapan agriculture could draw on more than one rainy season. If summer rainfall weakened, winter moisture might soften the impact in some areas. If both weakened, the number of reliable fallback options narrowed. But even here, we have to resist the temptation to tell a single civilization-wide story. Hydrological modeling suggests that central and lower parts of the Indus system could suffer substantial reductions in river flow while regions farther south or east experienced different conditions. During some drought intervals, Saurashtra remained comparatively favorable. Other regions near the Himalayan foothills retained access to more persistent moisture. So the map of environmental opportunity was changing. That can produce migration without an invading army. It can produce abandonment without a command from a king. And it can produce Imagine a farming household supplying part of its surplus to an urban market. As rainfall and river flow become less predictable, the household faces a choice. It can continue maximizing the crops and schedules that produce the largest marketable surplus when conditions cooperate. Or it can reduce risk by changing crops, planting across different seasons, relying more heavily on animals, moving closer to more dependable water, or producing primarily for local needs. None of those choices are dramatic in isolation. But a city does not experience them in isolation. A large urban population depends on thousands of households continuing to produce more food than they consume and remaining connected to the system that moves that surplus toward the city. If enough households change strategy, move, or reduce their exposure to distant exchange, the problem compounds. The city still needs food. The specialist still needs customers. The trader still needs volume. The water system still needs labor. But the countryside supporting all of them is reorganizing around a different objective, reducing risk. This is why calling the Harapan Transformation a climate collapse explains too little. Climate did not issue an order for Mohenjo-Daro to be abandoned. A drought does not automatically destroy a trade network. Lower rainfall does not directly erase standardized weights. Environmental pressure matters because it changes incentives, costs, and the reliability of the systems connecting one part of society to another. And the Harapans did respond. Settlement patterns shifted. In some areas, populations dispersed toward smaller communities. Agricultural strategies changed. Drought-tolerant crops became more important in parts of the Harapan world. Regions with more favorable water conditions could attract people while stressed areas lost them. That is adaptation. But now the central paradox of this story becomes sharper. The Harapans could respond to declining water reliability precisely because they had alternatives. They could change crops. They could change location. They could change the scale at which they lived and produced. The city had fewer options. A reservoir can store water, but it cannot make a century of regional rainfall decline disappear. A grain surplus can buffer a poor harvest, but urban life requires surpluses year after year. Trade can move shortages between regions, but only while other regions have enough to exchange and the network remains worth maintaining. The strength of the Harapan system had been its ability to connect many different landscapes. Under increasing uncertainty, that same system had to keep coordinating landscapes that were now changing in different directions. And that brings us to the most important mechanism in the entire story. Because when water becomes less predictable, the obvious solution is to diversify. Grow different crops. Spread risk. Depend less on one season, one market, or one source of supply. For a household, that can be an excellent survival strategy. For a metropolis, it creates a much harder problem. The crops and production systems best suited to surviving uncertainty are not always the ones best suited to generating the large, predictable surpluses that cities require. The system was approaching a contradiction. Resilience at one scale could undermine resilience at another. This is where the difference between surviving and sustaining a city becomes critical. A farming household does not need to maximize production under every condition. It needs to make it through the next season. If rainfall becomes unreliable, the safest response may be to spread risk. Plant crops with different water requirements, divide production between summer and winter seasons, keep more livestock, or reduce dependence on a distant market. A large city faces a different equation. Cannot simply survive on average. It needs enough food arriving in sufficient quantities with enough regularity to support a concentrated population year after year. That includes thousands of people whose labor may be invested in activities other than farming. Potters, metal workers, bead makers, builders, traders, people maintaining wells, drains, walls, streets, reservoirs, and whatever institutions coordinated life at urban scale. Every specialist represents a wager that agriculture somewhere else will continue producing enough surplus to release that person from full-time food production. That is one of the foundations of urbanization, and it creates a trade-off. At Harappa, the plant remains studied so far show a strong emphasis on large winter cereals, especially wheat and barley. These crops can produce substantial yields, and researchers have argued that the demand created by a growing urban population may have encouraged farmers supplying Harappa to emphasize this kind of production. Under favorable conditions, that makes sense. Large yields support large populations. Large populations support more specialists. More specialists increase craft production and exchange. Urban growth reinforces the agricultural system that feeds it, but a system optimized for output can become harder to adjust when the conditions behind that output become less reliable. This does not mean wheat and barley somehow caused Harappa to decline, nor does it mean Harapan farmers were trapped in a rigid agricultural system they did not understand. The evidence actually shows attempts to diversify. Millets became more important over time at Harappa, apparently adding another option under changing ecological conditions. The deeper problem may have been scale. A few households can change strategy quickly. The provisioning system for a city cannot necessarily change at the same speed without consequences elsewhere. Suppose farmers shift away from crops that produce the largest predictable marketable surplus and toward a wider range of crops selected for household resilience. That may be an intelligent response to uncertainty. But if the result is less surplus moving toward the urban center, the city feels the loss even when the farmers themselves are more secure. The same choice can therefore produce two different outcomes depending on where we measure success. At the household level, greater flexibility. At the urban level, less predictable supply. And once that distinction appears, the apparent contradiction at the heart of the Indus decline becomes much easier to understand. Adaptation was not failing. It was changing its objective. The clearest contrasts appear when we compare regions. In Gujarat, farmers at several sites used flexible combinations of crops and seasons. At places such as Kanmer and Kursara, the archaeological record suggests changing crop choices over time, with later production becoming more strongly shaped by local ecological conditions. At Rajdhi, farmers followed another combination. In villages farther northeast, communities also used diverse summer and winter crops. These systems offered multiple ways to respond when rainfall, demand, or local conditions changed. But the relationship between diversity and urbanism was not simple. Evidence from settlements in the hinterland of Rakhigarhi shows that farmers were already using rice, millet, pulses, and multi-season cultivation while Indus cities were developing. So it would be wrong to argue that cities always required monoculture while villages alone possessed diversity. Harappan urbanism could coexist with flexible agriculture. The real issue was whether the surrounding production system could continue generating enough surplus, reliably enough, while households were simultaneously adapting to new risks. That balance would have differed from city to city, and it could change over time. This is why Harappa is useful as a case study but dangerous as a template for the entire civilization. Its agricultural record suggests a tension between the demand for high yield supply and the need for greater flexibility. Other Harappan regions solved the same problem differently. That regional diversity may have been one reason communities survived so well. But it also meant there was no single agricultural adjustment capable of preserving the entire urban network. When stress increased, farmers did not have to make decisions for the Indus civilization. They made decisions for households, villages, fields, herds, and local exchange partners. A farmer deciding to plant more drought-tolerant millet was not voting to abandon Harappa. A family moving closer to more dependable water was not trying to destroy an urban economy. A community reducing production for distant exchange was not choosing civilizational collapse. Each decision could be rational at the local level. But complex systems can fail without anyone choosing the final outcome. If enough producers reduce risk in ways that also reduce surplus, urban provisioning becomes harder. If enough people relocate, the geography of supply changes. If exchange becomes less predictable, specialists lose access to markets and raw materials. If the specialist economy contracts, living inside a large city provides fewer advantages. Then population dispersal becomes even more attractive. That creates a feedback loop. Environmental uncertainty encourages local adaptation. Local adaptation reduces some community's dependence on the urban system. Reduced dependence weakens the scale of that system. A smaller urban system creates less demand for specialized production and long-distance exchange. And as those networks contract, communities have even more reason to organize around local conditions rather than civilization-wide standards. There does not have to be a single moment when the breaks. It can become less valuable one connection at a time. That helps explain something that a simple drought theory cannot. Why did the Harapan world not merely rebuild the same cities when conditions improved? Because by the time individual environmental pressures eased, the social and economic geography around those cities could already have changed. People had developed other strategies. Production had moved. Settlement patterns had shifted. Networks had contracted. The old urban arrangement was no longer simply waiting to be switched back on. A temporary pressure can produce a permanent transformation if the responses to that pressure change the system itself. And across the Indus world, those responses were not moving everyone in the same direction. They were pulling different regions apart. Once communities began solving environmental and economic problems locally, the Harapan world no longer had to change in one direction. And it did not. This is one of the most important features of the period after roughly 1900 BCE. The mature Harapan world had been extraordinarily diverse, but many of its settlements still participated in a recognizable shared material system. Seals, weights, styles of pottery, craft traditions, exchange networks, and urban forms connected communities separated by enormous distances. In the centuries that followed, that shared package became less uniform. Regional differences became more visible. Archaeologists often describe this as localization. But localization should not be imagined as the sudden replacement of one civilization by several completely new peoples. In many places, the more striking pattern is continuity combined with change. Communities remained in the wider landscape. They continued farming, keeping animals, producing pottery, and using technologies with roots in the Harapan period. What changed was the scale at which those practices were organized and the degree to which distant regions continued to resemble one another. Gujarat provides a particularly useful example. It had always been one of the more environmentally demanding parts of the Harapan world. Rainfall was limited and unpredictable, yet the region contained major settlements, craft centers, agricultural communities, and routes connecting inland production with maritime exchange. That meant local communities had experienced balancing two different pressures. They had to survive in an uncertain environment. But during the height of Harapan urbanism, many of them also operated inside a larger economic system. At Kursara, the agricultural record illustrates how those priorities could change. During an earlier phase, barley strongly dominated the crop assemblage. Researchers have suggested that its importance may partly reflect wider economic demand during the urbanizing period. Over time, however, the system diversified. Barley became less dominant. Summer pulses appeared. Rice entered the crop package. And by the late Harapan phase, summer crops including pearl millet, rice, and sorghum formed much more of the assemblage. Kanmer shows another significant change over time. Rajdhi followed yet another strategy, with a strong reliance on millet-based agriculture suited to arid conditions. These communities were not all converging on one new solution. They were becoming more responsive to local conditions. That distinction is critical. If an integrated economic network rewards farmers for producing a crop demanded elsewhere, local agriculture can partially reflect the needs of that network. If that network weakens, the calculation changes. A crop that produces less value in distant exchange may still be more dependable under local rainfall. A production system designed partly around external demand can shift toward one designed more around local survival. At Kirsara and Kanmer, researchers have proposed exactly this possibility. The movement from relatively specialized cereal production toward more diverse, multi-season systems may reflect both changing ecology and declining economic pressures associated with the urban network. The evidence does not allow us to prove that trade decline caused every crop change, but the correlation reveals the larger point. Economic integration and environmental adaptation were interacting. When one changed, communities could reorganize the other. And that means the decline of urbanism did not have to spread like a wave from one ruined capital. It could emerge independently from thousands of local adjustments. A farmer changes what is planted. A settlement depends more on nearby resources. A craft producer loses access to a distant raw material. A household relocates toward more reliable water. A community trades less frequently with a city that is itself shrinking. None of these decisions destroys a civilization. But together they can change the geography on which that civilization depends. This is a classic complex system problem. The individual decision and the collective outcome are not the same thing. Imagine a network of settlements in which every community is connected because connection provides benefits. One region supplies grain. Another provides stone. Another produces shell objects. Another sits along a route carrying goods toward the coast. A city concentrates consumers, specialists, and exchange. As long as enough parts of the network participate, participation itself is valuable. But now introduce prolonged uncertainty. One community reduces production for export. Another changes crops. Another becomes more pastoral. Another loses population. Another turns increasingly toward local exchange. Each decision can improve resilience for the people making it. But every local retreat slightly reduces the scale of the larger network. And as the larger network becomes smaller, the incentive for the next community to remain fully connected can weaken as well. This is how adaptation can produce fragmentation without coordination. No Harapan ruler needed to announce the end of the urban age. No council had to decide that civilization would become regional. People only had to keep responding rationally to the conditions directly in front of them. The result could still be civilization-wide. This also helps explain why the transformation looked different in different places. Some settlements were abandoned. Others shrank. Others continued. Some communities placed greater emphasis on drought-resistant crops. Some relied more heavily on livestock. Some material traditions persisted while others became increasingly regional. There was no single moment at which every Harapan community crossed the same threshold. The mature urban system did not disappear like a light being switched off. It lost coherence. That distinction matters because a civilization-wide system depends on more than population. It depends on repeated interaction. Shared standards matter because people keep using them. Specialized production matters because consumers remain connected to producers. Long-distance trade matters because enough goods continue moving to justify the route. Urban centers matter because concentrating people continues to produce advantages greater than the costs of feeding and maintaining them. Once communities could survive more effectively by reorganizing around regional conditions, some of those advantages began to weaken. And that created a second feedback loop. More localization reduced the scale of inter-regional exchange. Less exchange reduced the advantages of standardization and specialization. Fewer advantages made large urban concentrations less attractive. Smaller urban populations created still less demand for distant goods. The system did not merely respond to pressure. Its responses changed the system itself. By the time this process had advanced far enough, the map still contained people descended from the Harapan world. It still contained farmers, craftspeople, settlements, and technologies. But it no longer contained the same network. And when a network loses enough of its scale, even the institutions that once made it powerful can begin to lose their purpose. By this point, the Harapan transformation was no longer only a problem of water, farming, or settlement. It was becoming a problem of scale. During the mature Harapan period, people separated by enormous distances participated in systems that looked remarkably standardized. Weights followed shared conventions. Seals and ceilings circulated through settlements large and small. Craft specialists produced objects using techniques and raw materials that connected distant regions. Writing appeared on seals and other portable objects within this shared material world. None of these things necessarily required a single central government. But they did require something else. Participation. A standard only works while enough people continue to recognize it. Imagine a merchant carrying goods between two Harapan settlements. A standardized weight allows both sides to measure according to a shared system. That reduces uncertainty. The merchant does not need to invent a new measurement every time he reaches another city. A seal can carry information within a shared cultural and economic system. A specialist workshop can produce highly refined objects because customers exist beyond the immediate household or village. These tools become more useful as the network grows. But network effects work in both directions. When participation falls, the value of the network can fall with it. And that appears to be one of the most important features of the post-urban Harapan world. As durbanization proceeded, several hallmarks of mature Harapan integration declined together. Writing disappeared from ordinary archaeological use. The standardized weight system faded. Steatite seals ceased to play the same role. Long distance exchange contracted. Large non-residential architecture became far less characteristic. This does not mean every craft tradition vanished. Far from it. Some forms of specialized knowledge continued. Fiance ornaments, for example, remained important and in some regions were produced in substantial quantities. That distinction tells us something critical. What was disappearing was not skill itself. It was the particular scale and organization of the mature Harapan system. A craft that could use local materials and sell to local consumers might continue perfectly well. A craft dependent on distant raw materials, specialized exchange routes, or large concentrated markets faced a more difficult future. The important question is not whether trade collapsed before the cities or whether the cities collapsed before trade. The evidence is too complicated for such a clean sequence. More likely, these processes reinforced one another. A shrinking city creates fewer customers for specialist producers. Fewer specialists reduce demand for distant raw materials. Lower trade volume makes long routes less profitable or less necessary. Less exchange reduces the practical advantage of maintaining shared standards. Greater localization makes communities increasingly responsive to regional needs. And as the benefits of integration decline, maintaining a large city becomes still harder to justify. That is a feedback loop. No single link has to destroy the system. Each weakening connection makes the next connection slightly less valuable. This helps explain why looking for one final catastrophe can be misleading. There may never have been a moment when the Harapan network simply stopped. Instead, its density decreased. Some routes mattered less. Some standards were used less often. Some specialists found smaller markets. Some settlements became more self-reliant. And eventually, the network that had made urban concentration productive no longer operated at its previous scale. This also explains why the disappearance of writing should not automatically be read as the disappearance of people. The Harapan script was strongly associated with the urban material world in which seals and standardized practices mattered. If those institutions became less useful, the objects carrying the script could disappear even while communities continued speaking, farming, worshiping, remembering, and transmitting knowledge. The same logic applies to weights. The disappearance of a civilization-wide measuring system does not mean people stopped measuring things. It means they no longer needed to participate in that particular shared system to the same extent. That is what makes the Harapan case so revealing. Civilizations are often recognized archaeologically through the things that connect people across space. Standardized objects, common technologies, shared symbols, large settlements, trade routes. When those connections weaken, the archaeological civilization can seem to vanish even when much of the population remains. By the late Harapan period, the people were still there. But the scale at which they organized their lives was changing. The integrated urban network had once rewarded participation. Now localization increasingly rewarded independence from it. And once that happened, the old Harapan system did not need an army to destroy it. It only needed fewer people to keep choosing it. Once we see the Harapan decline as a transformation of systems rather than the disappearance of a people, the most famous explanations begin to look incomplete. Climate mattered. But there was no single drought that struck every Harapan settlement equally and destroyed the civilization in one blow. Environmental stress varied from region to region, and communities responded differently depending on their water sources, crops, economic connections, and available alternatives. Nor does the evidence support the old image of an invading army sweeping across the Indus cities and bringing the civilization to an abrupt end. Later movements of people from Central Asia did reshape the population history of South Asia. Ancient DNA shows that steppe-related ancestry entered the region after the mature Harapan period. But the successfully sequenced genome from a Harapan period individual at Rakigari contained no detectable steppe pastoralist ancestry. That does not tell us the ancestry of every Harapan individual across an enormous civilization. It does tell us why migration and conquest cannot simply be collapsed into the same explanation. People moved, populations mixed, violence existed. None of those facts demonstrates that a single invading population destroyed Mohenjo-Daro, Harappa, Dalavera, and the wider urban system around 1900 BCE. And there is another simple explanation we should resist. We cannot confidently blame the collapse of a Harapan king or central government, because we do not actually know that such a political system existed in the form familiar from Egypt or Mesopotamia. The script is undeciphered. The institutions remain debated. So the safest explanation begins with what archaeology allows us to see directly. The cities became smaller. Population redistributed. Regional agricultural strategies changed. Long-distance exchange contracted. Writing, seals, standardized weights, and much of the old urban material system disappeared. Yet smaller settlements persisted. Craft knowledge persisted. Agriculture persisted. People persisted. That is not the pattern of a society that simply failed to survive. It is the pattern of a society changing the scale at which survival was organized. Hydroclimatic stress changed the constraints. Communities responded by altering crops, production, settlement, and economic connections. Those adaptations could make sense locally while weakening the dependable surpluses and large networks on which urban concentration depended. As urban populations declined, the markets supporting specialists became smaller. As exchange contracted, participation in civilization-wide standards became less useful. As local production became more important, regional differences became stronger. And each of those changes reinforced the next. That is why the disappearance of the cities should not be confused with the disappearance of the Harapans. When we say the cities couldn't adapt, we do not mean that every Harapan city instantly became physically uninhabitable. Different cities followed different trajectories, and some continued to be occupied after the mature period. What failed to survive at its former scale was the larger mature Harapan urban system. The dense combination of great cities, specialized production, long-distance connections, common standards, and the agricultural surpluses that helped sustain them. For centuries, that system had been a solution. It allowed populations to concentrate, crafts to specialize, goods to travel immense distances, and communities across very different environments to participate in a recognizable shared world. Then the conditions changed, and eventually, the strategies that made sense for survival no longer necessarily made sense for maintaining that system. The ruins therefore record two stories at once. One is a story of loss. Great urban centers declined. Networks fragmented. Technologies of coordination disappeared from use. A form of urban life that had flourished for centuries was never reconstructed in the same way. But the second story is adaptation. People moved. They changed what they produced. They reorganized around different landscapes and different scales. They built new regional cultures from material that had not simply vanished with the cities. And that may be the most important lesson of the Indus Transformation. Civilizational resilience does not always mean preserving the civilization in the form we recognize. Sometimes survival requires abandoning the very system that once made success possible. The Indus civilization did not simply disappear because its people failed to adapt. In part, its old urban world disappeared because they did.