Science vs. Design in The New Urban Science

Cheng Hou
Civic Analytics & Urban Intelligence
2 min readOct 23, 2016
Timeline of New Urban Science Institutions

We are living in the age of cities. Yet understanding what city is and how urban world works is still one of the biggest challenge in the human life. Since 1995, many labs, departments and institutions were set up and focused on the new urban science. The definition of what urban science actually is changes rapidly since more and more characteristics of cities were discovered by deeper research. An interesting question is the tension between science and design in the new urban science.

Traditional methods of urban science relies on field work, site visits, and surveys to understand cities. In contrast, the scientific approach aims to explain the cities through universal laws applying on the urban data. There is no certain answer to evaluate the two different urban science study methods. But the several examples indicate that the scientific approach would be more essential.

Distribution of racial ethnic groups based on samples from 2005 to 2009

The above image shows the distribution of racial ethnic groups in New York City which is an example in the urban science study. Using the traditional method, we prefer to conclude the result by saying some sentences like ‘People living in Manhattan are mainly white men’. While in the scientific method, we just save the entire distribution data and use it in further analysis. The cities are becoming so complex that simple answers may miss a lot of indispensable information. So using the big data generated by the city itself to describe characteristics and analysis is a better idea. We probably cannot display all the data in limited numbers of images or tables. But we have universal laws to handle the data and analysis in order to give predicted answers to other complex problems. It is how data-driven policies work. And we believe that it is closer to the real rules and it excludes the certain impact of some political factors.

--

--