“Whenever I run into a problem I can’t solve, I always make it bigger. I can never solve it by trying to make it smaller, but if I make it big enough, I can begin to see the outlines of a solution.”
– D. Eisenhower
Data and information are the fuel of science. Physics, biology, and other established natural sciences took their most revolutionary leaps in the wake of new data, whether from astronomy, natural history, atomic and nuclear experiments, or molecular biology and genetics.
There is no doubt that the present time is bringing us a volume and wealth of new data, observational methods, and experiments about human societies that will profoundly and irreversibly transform the social sciences.
However, this embarrassment of riches is primarily the result of an explosion in commerce and technology, not scholarship. This raises questions of bias, privacy, reproducibility, and others typical of scientific knowledge creation but alien to business practices. Consequently, while being shaken by a myriad new empirical opportunities, the social sciences, as practiced in universities and research institutions, have arguably not yet started to profit or change substantially.
This situation may be perhaps compared to the beginnings of the Industrial Revolution, when new machines (mostly thermal engines) became common and, despite their dismally low efficiencies, quickly enabled economic production on a massive scale, changing the nature of economics, work, and the structure of human societies.
A science of thermodynamics that described these machines scientifically—and eventually improved them—took another century. But when it did, it ushered in a fundamental understanding of physics (e.g., the principle of conservation of energy) and led to new and previously unimagined concepts, such as the quantification of information, which, of course, brings us right back to where we are now.
The power of social technologies
Today, the power of social concepts, technologies, and data to change the world should be obvious to anyone. The greatest successes of applying the explosive progress in computing and the World Wide Web have been social technologies.
Internet search’s breakthrough (page-rank) is an algorithm invented by sociologists to measure social influence.1Phillip Bonacich, “Power and Centrality: A Family of Measures,” American Journal of Sociology 92, no. 5 (1987): 1170–1182. Google and other companies actively operate predictive ad markets, arguably more sophisticated and stable than our academic knowledge of finance would suggest possible. Wikipedia and OpenStreetMap have shown how collective production of unprecedented quality and scope can be achieved in the absence of structured economic organizations, such as firms, contracts, or price signals.The social sciences must become bigger, more integrated, and more powerful. These organizations and others have shown how collective cognition can operate in large human societies and in the absence of close and continuing social contact. Facebook, Twitter, and many other social networking companies are revealing the evolving structure of immense social networks and pushing the boundaries of our understanding of social mechanisms for the spread of information, opinion formation, and new forms of collective efficacy, for example in response to natural disasters.
These are social science breakthroughs: discoveries so exciting that nothing will ever be the same!
These discoveries are changing our lives in ways that are perhaps more profound than the atomic bomb or the double helix. Their judicious use holds the key to a world of collective creativity, collaboration, justice, and sustainable development. But they also allow surveillance and perhaps social manipulation on a scale and at a level of detail never before possible. To use the power of social technologies for good and to regulate and enforce the boundaries of what is ethical and desirable, we need to learn to better understand social phenomena and their consequences. To this end, the social sciences must be the focus of an unprecedented R&D effort, theoretical and applied. The social sciences must become bigger, more integrated, and more powerful.
The challenge to the social sciences
The challenge to the social sciences as they exist today, however, is profound. To the external observer, say in policy, the social sciences have often appeared irrelevant to social or economic practice. Recent events, from the great recession to the current crisis of inequality and justice to the puzzle of economic growth, only add fuel to this bonfire.
The problem stems from the fact that many different academic disciplines in the social sciences study the same phenomenon but often disagree on even the most basic things, such as on fundamental mechanisms or on the most relevant agents of change.
For example, certainly economic markets are socioeconomic networks2M. Granovetter, “Economic Action and Social Structure: The Problem of Embeddedness,” American Journal of Sociology 91, no. 3 (1985): 481–510. of exchange mediated by prices and other signals, where people and organizations are subjected to incomplete information, have various levels of influence, and make different decisions based on their context, history, and “social horizons.”
Such a lay view of the problem calls for a unified theory of markets that includes current elements from economics, certainly, but also from sociology, social psychology, cognitive sciences, political science, anthropology, and more.Like the story of the steam engine or of genetics, applied work is no substitute to fundamental understanding. The same need for expansion and synthesis applies to the study of other fundamental social phenomena, such as firms, cities, governments and other institutions, neighborhoods, crime, inequality, etc.
But this is currently not the state of the art, of course. The silos that presently characterize the “social sciences of the same phenomenon” seep away their power of explanation by chaining each disciplinary tradition to rigid, and often opposing, worldviews.
By this measure, compared to physics or biology, or, most importantly, to the demonstrable power of social technologies exercised presently by corporations and nonprofit organizations, the social sciences are lagging far behind.
Something must change.
Falling back on old models of privacy, ownership, or regulation, while natural and prudent, will likely fail to capitalize on the potential for good brought about by these technologies, and for the expansion in our knowledge of social phenomena they enable.
How then can we expect the social sciences to change in the era of applied social technologies? Like the story of the steam engine or of genetics, applied work is no substitute to fundamental understanding. A more fundamental understanding of social phenomena requires several elements that presently remain challenging across the practice of the social sciences. I suggest that these can be summarized as (i) synthesis, (ii) relevance, and (iii) the use of the scientific method as an instrument to accumulate predictive knowledge.
These epistemological mechanisms have been used successfully in the natural sciences, of course, but they need to take their own life and form in the context of the multidisciplinary communities in the social sciences.
Synthesis, relevance, and the scientific method
The problem of synthesis speaks to the issues already brought up of creating a single “best theory” of social phenomena such as markets, firms, cities, etc., that includes all the insights from traditional disciplines, makes observable predictions, and is falsifiable, so that it can be improved.
The problem of relevance is perhaps harsher, and speaks to the relative importance of certain ingredients to explain observable outcomes, qualitative or quantitative. For example, how important are social networks for the functioning of markets? Or for generating economic growth and human development? Is learning necessary, or essential? Or are representative simple agents with assumed preferences enough for most purposes?
I picked controversial examples on purpose, issues debated across the social sciences for many decades. Such debates must reach more conclusive results faster: irrelevant considerations—wherever they come from—must be dropped and Occam’s razor must always be at work. We must strive collectively to build epistemological structures—interacting with big and small data—to answer these questions as simply and clearly as possible, so that our understanding of critical social phenomena can advance.The use of data is becoming a necessary condition for academic impact.Methodologically, the use of social technologies by corporations, governments, and other institutions—their data collection and “experiments”—are in many ways breaking taboos and pushing the envelope on what is possible and acceptable. This will always be disputed ground, but I think it is fair to conclude that such practices are revealing that people are willing to share their information—under certain general conditions—in situations where they experience benefit (personal or collective). This casts, in my opinion, issues of privacy and incomplete and potentially biased data in a new light and asks for a practical yet principled approach to the uses of information in social scientific research.
If anything, researchers and scientific institutions can provide individuals and organizations with higher standards of data use than corporations already do. Researchers are also in the unique position to help set new standards that allow discovery while enlarging the sphere of the common good. But only plunging in and learning to swim will do.
The social sciences have produced some of the most extraordinary insights and concepts in any scientific discipline. Phenomena such as the division of labor and knowledge, the emergence of cooperation between strangers, cultural evolution, and economic growth are as fascinating and powerful as anything to be found in the natural sciences. These problems—and more—continue to be at the vanguard of research and policy. They are at the base of the greatest challenges in human societies and are keys to our continued tenure of the planet.
To solve them the social sciences must become bigger.
This means that they are more capable of synthesis across their marvelously multifaceted approaches to the complex problems of changing societies. It also requires that social scientists learn to better use the vast empirical possibilities of digital technologies to discover what matters most to describe each given social phenomenon in a way that exercises the virtuous cycle embodied in the scientific method faster.
All this is already happening, of course, though perhaps not always in the ways one might have planned. Interdisciplinary collaborations within the social sciences and beyond are emerging, often with the goal of capitalizing on big data, dealing with issues of sustainability and resilience or under the interdisciplinary banner of complex systems. Such efforts are certainly growing faster than traditional departments and opening new spaces for syntheses.
The use of data is becoming a necessary condition for academic impact, even in disciplines traditionally relying on axiomatic approaches to theory. And above all, vast and varied “applied social science experiments” are underway by a growing number of technology companies and governments. Digital economy companies do primarily applied social sciences research, much like Bell Labs did applied physics or Celera Genomics and its biotech heirs do applied biology. Embracing the connections between applied work and theory—by companies and universities alike—will shake and revitalize the social sciences, especially as the boundaries between the digital and physical worlds blur and disappear.
This is not your parents’ social science. It should not be: it is something new and exciting, shaken and stirred by technology and data, challenged by philistines from other sciences, discovering the power of its concepts in new places and their deeper meanings and reveling in its potential to change the world.
References [ + ]
|1.||↑||Phillip Bonacich, “Power and Centrality: A Family of Measures,” American Journal of Sociology 92, no. 5 (1987): 1170–1182.|
|2.||↑||M. Granovetter, “Economic Action and Social Structure: The Problem of Embeddedness,” American Journal of Sociology 91, no. 3 (1985): 481–510.|