Simulation of cities to be applied to data on ancient Mesopotamia

This CORONA satellite photograph shows the third millennium B.C. circular-walled town of Tell Beydar in the Khabur basin of northern Syria (lower left). The 3-D effect has been produced by draping the satellite photo over a digital terrain model. The presence of what are thought to be Bronze Age roads, radiating from Tell Beydar, can be seen, in one example, to run up and over the low hills to the right of the center of the image. The evidence of such tracks are being used by Wilkinson’s modeling initiative to show the degree of interaction between ancient cities and also to infer the limits of their agricultural areas.

For decades, people have used computers to model present-day realities and fantasies. Engineers and scientists design cars and predict the weather with them, while video gamers have propelled the game The Sims, which allows the design of simulated human lives to play out on a screen, to become the best-selling computer game of the 21st century.

Now Tony Wilkinson, Research Associate and Associate Professor in the Oriental Institute and Near Eastern Languages & Civilizations, along with colleagues spanning the sciences and humanities, wants to apply this technology to ancient Mesopotamia. If the simulations work as desired, his team will be able to test how and why the first civilizations were born, lived and died.

Wilkinson is a Briton whose soft-spoken manner is belied by the ambition of his project. “It will be a bit like Sim City, but real,” he said. The difference between the Oriental Institute project and a computer game lies not just in the sophistication of the model, but the fact that the database is history itself, and the results will be a new window into its causes. “We’ll run the model to see if we can grow Mesopotamian cities and test the results against archaeological data,” Wilkinson noted.

This new model will be based not on the decisions of a single ruler/player, but on the choices and fortunes of thousands of small communities, a concept known as agent-based modeling. “It’s not a top-down model, where cities or settlements are imposed by the king or administration.” Instead, Wilkinson explained, “we build up cities from the smallest component units and see how these navigate through time.

“It’s a means of modeling that enables you to take entities–an individual, family, village, city–and give them attributes that enable them to make decisions about what to do,” said Wilkinson. “You can integrate this with an empirical crop model, which is a complex mathematical model that lets you project the growth of cereals through different rainfall levels, and you can integrate this model with the agent’s behavior. Then you can take another model, a climate model, and you can integrate that, as well.

“It results in lots of unanticipated behavior that you wouldn’t get with our usual linear thinking. A run of dry years, you think it’s going to collapse, but there are so many coping mechanisms that you might not anticipate.”

Wilkinson’s collaborator, archaeologist McGuire Gibson, Professor in the Oriental Institute and Near Eastern Languages & Civilizations, specifies the new questions they can ask: “We can see exactly what happens if you have less rain, more rain, drought, disease: what does that do to the settlement?”

The Oriental Institute team’s use of agent-based modeling could produce a revolutionary way for archaeology to explain events. “We can test ideas a lot more rigorously–doing repeat runs with different scenarios,” said Wilkinson. “We’ve reached a stage where people are arguing round in circles, using slightly different data sets. Whereas with modeling, we can say under these circumstances, these things happen.”

Gibson argued that, with the new technology, they are reviving some old academic virtues. “What we’re really trying to do is go back to the basics–what (O.I. founder) James Breasted and (former O.I. Director) Robert McCormick Adams were trying to do–bring together texts and archaeology in an integrated study, but make it more analytical using computer capabilities. Fields drift apart as people become more specialized–it’s more than a way of getting us to talk to each other, it’s a way of getting analysis to be done together, as well.”

Wilkinson added, “We’ve done it by stepping outside the Near Eastern field for funding. Our grant comes from a new NSF (National Science Foundation) Program, Biocomplexity in the Environment, for studying the dynamics of coupled natural and human systems. It’s a revival of ecology, a way of returning to old O.I. tradition, going back to the ’30s. Ecology had gone out of fashion in the ’90s, but we’re bringing it back into the equation.” The NSF grant includes $1.2 million to be split between the Oriental Institute and Argonne National Laboratory for a five-year research program.

The project will bring together philologists, who are more accustomed to reading Gilgamesh, with geophysicists, who are more familiar with fluid dynamics and radiative transfer. Wilkinson’s favorite example of a meeting ground is the farming technique of fallowing.

“You’ve got a lot of records on cuneiform tablets of the cropping of a field that say, ‘the land is cropped this year, then rested the next year, then cropped the next.’ This fallow year gathers extra rain. Some of this extra rain evaporates, but about 15 to 20 percent is carried over. By fallowing you get a bonus of soil moisture, which increases your effective rainfall. If a group decides to crop every year they lose their insurance and move to the equivalent of a ‘drier’ climate.

“So when you then run your moisture models and your crop models, you interact a computer simulation of climate with human decision making as preserved in Babylonian texts. You get to examine different effective climates in the same place: one group will be able to ride out drought; another might get more grain in the short term but be very vulnerable when there’s climate stress,” Wilkinson explained.

Gibson added, “That’s in the north. In southern Mesopotamia, fallowing is even more important, absolutely essential to stave off the salt. If you start not fallowing, you see evidence of yield dropping in four or five years. And the texts let you see what that does to a society. People get so far into debt that they can’t stay independent–they sell their children and themselves to big institutions. And then widows and orphans show up on these administrative ration lists. There’s agro-economic data on this happening in Iraq in the’30s and ’50s.”

The project spans history, disciplines and departments, including teams from Argonne National Laboratory, Anthropology and Geophysical Sciences. Under Director John Christiansen, the Argonne team brings to the project a long-term expertise in agent-based modeling.

Wilkinson also described the role of a team under Raymond Pierrehumbert, Professor in Geophysical Sciences and a climate modeler. “They are looking at ways of doing climate models that are useful to the user and the general public. This will help us graft our social and physical models onto the atmosphere that existed then, thousands of years ago, as opposed to now.”

Pierrehumbert noted that Wilkinson’s project and his work “are a match made in heaven. We are trying to develop software to make it easier for researchers to do climate simulations when such simulations are needed to address cutting-edge questions in their disciplines (Wilkinson’s) project will provide an infinite source of climate related questions and an eager pool of users for our climate simulation software.”

Gibson added that scholars in ancient Near Easterner studies might offer geophysicists a chance to pick up data on specific areas at other times, which they can build into their model.

In addition to Wilkinson and Gibson, the O.I. team includes David Schloen, Associate Professor in the Oriental Institute, Near Eastern Languages & Civilizations and the College, whose Patrimonial Household Model forms a key part of the model framework; Christopher Woods, Assistant Professor in Near Eastern Languages & Civilizations and the College; and John Sanders, Senior Research Associate in the Oriental Institute. The project also will incorporate the work of the South West Asia Project at Yale University. That team has been looking at recent environmental change and how it has impacted the landscape, using remote sensing technology.