The Manifold Mismatch: Topological Constraints and Non-Human Agency in Complex Adaptive Urban Systems

Cities have always been more than just collections of buildings and roads; they are complex systems where human plans constantly collide with the physical world. For centuries, architects and planners have tried to impose neat, straight lines and perfect squares onto the landscape, assuming that humans are the only ones who matter in the design process. They treated the ground, the water, the wind, and the animals as passive backdrops to be managed. However, the ground often refuses to stay flat, water does not always flow where pipes are laid, and animals move according to their own needs. This friction between the rigid, human-made order and the messy, living reality of the planet is a fundamental tension that has shaped every city in history, long before computers were invented. Today, as we rely more on artificial intelligence and digital models to design our future, a critical question arises: are these new tools helping us understand this complex relationship, or are they simply making the old mistakes faster?

A new study by Arda Arın at MEF University tackles this question by looking at the "Manifold Mismatch." This is a term the author uses to describe the structural gap between the straight, logical grids humans draw on paper and the curved, shifting, and unpredictable forces of nature and technology that actually shape a city. The research argues that for too long, urban design has been anthropocentric, meaning it focuses almost exclusively on human comfort and human logic. The study suggests that this approach is failing because it ignores the "agency" of non-human things. In this context, agency means the ability to act and influence outcomes. The paper proposes that rocks, rivers, fungi, and even computer algorithms are not just objects to be used, but active participants that co-create the city alongside humans.

To understand how this works, the author traces the history of how we use computers to design cities. The study identifies four main stages of this evolution. First, there were rule-based systems, where designers used simple geometric rules to generate shapes, much like following a recipe. Next came topological analyses, which focused on how things connect and move, mapping out where people walk or drive. Then, the field moved to complexity-based simulations, which tried to model the city as a living, changing system rather than a static picture. Finally, we have arrived at the current era of artificial intelligence, where machines learn from massive amounts of data to predict and generate new designs. The author finds that despite these technological leaps, all four stages share a common flaw: they still treat the city as a problem for humans to solve, ignoring the active role of non-human forces.

The core of the study is a massive review of twenty-nine real-world examples from around the globe, spanning thousands of years. These cases prove that non-human forces have always been shaping cities, often in ways that forced humans to adapt. For instance, in Cappadocia, Turkey, ancient people did not build cities on top of the rock; they carved their homes directly into the volcanic tuff, letting the wind and water erosion dictate the shape of the landscape. In the Aztec capital of Tenochtitlan, the city was built on floating islands held together by the roots of willow trees, which acted as living anchors against the water. In Venice, the city's survival depends on millions of wooden piles driven into the mud that, over centuries, have turned into stone-like columns through a chemical process that humans cannot control but must rely on. These are not just historical curiosities; they are examples of "ecological actants"—living and non-living things that actively shape the built environment.

The study also highlights how modern technology is creating new forms of non-human agency. In Miami, rising sea levels are pushing water up through the porous limestone ground from below, bypassing seawalls entirely. The city is now forced to install sensor-driven pumps that react to the water in real time, effectively letting the ocean dictate the city's daily operations. In Rotterdam, a massive port terminal operates entirely without human workers, using automated cranes and vehicles that communicate with each other to move cargo. This space is designed not for human eyes but for the efficiency of machines, creating a city layer that exists independently of human perception. The research shows that when we ignore these forces, the city breaks. Mexico City, built on a drained lake bed, is sinking by up to fifty centimeters a year, cracking its subway lines and buildings because the ground is reasserting its own physical reality against human plans.

To fix this, the author proposes a new way of using computers, called the "Manifold Mismatch simulation framework." Instead of just asking a computer to optimize a building for human use, this method would simulate the city as a battlefield of competing forces. It would model the pressure of the wind, the flow of groundwater, the growth of tree roots, and the movement of animals as active agents that push and pull against human structures. The computer would then show where these forces clash, predicting where a bridge might crack, where a neighborhood might flood, or where a road might become impassable. This allows designers to see the tension before it becomes a disaster. The goal is not to eliminate these conflicts, which is impossible, but to negotiate with them. It shifts the role of the architect from a commander who imposes a perfect order to a mediator who helps humans and non-humans coexist.

The findings suggest that the future of city-making depends on accepting that we are not the only designers. The study validates a three-part classification of these non-human forces: ecological actants like plants and animals, sensing infrastructures like sensors and pumps, and algorithmic subjects like the software that runs our smart cities. By recognizing these entities as active partners rather than passive obstacles, we can build cities that are more resilient and better suited to the complex reality of the planet. The paper concludes that the most advanced cities of the future will not be those with the smartest algorithms alone, but those that learn to listen to the manifold of forces—geological, biological, and digital—that are already shaping them. The city is not a machine to be perfected, but a continuous, multi-species negotiation that we are only just beginning to understand.