segregation & Social Inquiry:
the politics of scientific representation
with Jared Millson and Mark Risjord
Research into residential segregation has implications for our health, our wealth, our children’s education, and our treatment by police officers. Consequently, it broaches upon issues of fairness, autonomy, equality, and justice. This raises a perennial problem in the social sciences: what is the precise role that these values should play in social-scientific research?
Both social scientists and philosophers have grappled with this question. At one extreme is the “value-free” ideal, which holds that values should play no significant role in scientific inquiry. The value-free ideal’s chief attraction has always been its reverence for science’s representational objectivity—the objectivity achieved when scientific representations are accurate. Representational objectivity is not merely a high-minded epistemological scruple; it has practical consequences as well. For example, to address the problems that segregation begets, social scientists must accurately represent its patterns, causes, and effects. Value judgments threaten to turn scientific reasoning into little more than wishful thinking, thereby distorting these aspects of segregation, which in turn leads to ineffective policies for remedying its ills. However, this ideal faces searching criticisms: it ignores the value-laden contexts in which science is produced, it gives poor counsel when evidence is sparse or the stakes are high, and it ignores the ways in which the very objects of social inquiry are saturated with human values. Yet the fate of representational objectivity in the face of these objections has not been addressed. Instead, “objectivity” comes to mean something else, e.g., appropriately formed scientific consensus.
Segregation and Social Inquiry: The Politics of Scientific Representation uses the history of sociological research about residential segregation to show how value-laden science is compatible with representational objectivity. Its core idea is that values should influence the questions that scientists ask, but these values should not influence the answers to those questions. As we show, this preserves the insights of earlier (and bolder) departures from the value-free ideal and suggests a novel account of scientific representation. In the process, we broach on larger issues in the philosophy of science and the methodology of the social sciences, engaging issues such as situated knowledge, underdetermination, inductive risk, thick evaluative concepts in science, modeling, idealization, and measurement.
The social sciences’ methods and subject matters render them especially inhospitable to the value-free ideal. Consequently, Segregation and Social Inquiry shows that objectivity is alive and well among even the most value-laden areas of scientific research. Given people’s tendency to understand segregation and other social issues through highly politicized lenses, this work thereby restores the hope that members of a divided society might still see through the fog of ideology and the blinders of dogma.
Both social scientists and philosophers have grappled with this question. At one extreme is the “value-free” ideal, which holds that values should play no significant role in scientific inquiry. The value-free ideal’s chief attraction has always been its reverence for science’s representational objectivity—the objectivity achieved when scientific representations are accurate. Representational objectivity is not merely a high-minded epistemological scruple; it has practical consequences as well. For example, to address the problems that segregation begets, social scientists must accurately represent its patterns, causes, and effects. Value judgments threaten to turn scientific reasoning into little more than wishful thinking, thereby distorting these aspects of segregation, which in turn leads to ineffective policies for remedying its ills. However, this ideal faces searching criticisms: it ignores the value-laden contexts in which science is produced, it gives poor counsel when evidence is sparse or the stakes are high, and it ignores the ways in which the very objects of social inquiry are saturated with human values. Yet the fate of representational objectivity in the face of these objections has not been addressed. Instead, “objectivity” comes to mean something else, e.g., appropriately formed scientific consensus.
Segregation and Social Inquiry: The Politics of Scientific Representation uses the history of sociological research about residential segregation to show how value-laden science is compatible with representational objectivity. Its core idea is that values should influence the questions that scientists ask, but these values should not influence the answers to those questions. As we show, this preserves the insights of earlier (and bolder) departures from the value-free ideal and suggests a novel account of scientific representation. In the process, we broach on larger issues in the philosophy of science and the methodology of the social sciences, engaging issues such as situated knowledge, underdetermination, inductive risk, thick evaluative concepts in science, modeling, idealization, and measurement.
The social sciences’ methods and subject matters render them especially inhospitable to the value-free ideal. Consequently, Segregation and Social Inquiry shows that objectivity is alive and well among even the most value-laden areas of scientific research. Given people’s tendency to understand segregation and other social issues through highly politicized lenses, this work thereby restores the hope that members of a divided society might still see through the fog of ideology and the blinders of dogma.
relevant publications
The conceptual framework for this project arose from my work on questions in science, but only "Understanding, Truth, and Epistemic Goals" applies these ideas to segregation research. My paper with Emily Sullivan about idealizations and understanding also discusses segregation research in a manner congenial to this project.