|The Enlightenment coffee house|
Social science differs from physical science for two reasons. The first one is a matter of degree. As we go up the scale of complexity in the objects we study, our laws get less exact. Physical laws are very precise. The laws of biology are less so: rabbit breeding is less predictable than the swing of a pendulum. Humans and human societies are even more complex than living bodies, so social scientific laws are even less exact.
The second is a difference of kind. It is a good thing for humans to be able to control and predict the natural world. Not that the power cannot be misused, but it is useful for us to understand animal breeding, global warming and the physics of lasers. On the other hand, it is much more ambiguous for humans to be able to control and predict other humans. To know when this is good, you have to ask with Lenin: who, whom? For instance, predicting crime could be extremely useful, but the idea has furnished material for a large number of cyberpunk dystopias.
These differences mean that in social science, universal scientific laws may be difficult to find, and sometimes not worth finding.
A law narrows down what is possible. Falling bodies on earth accelerate only at 9.81, not 9.79 or 9.93, meters per second per second.
Another goal of social science might be to find and name what Jon Elster calls mechanisms: things that can happen, though they will not necessarily. If mechanisms turn up often enough in social life, then they will enter the public mind and vocabulary. They then expand the ability of the average citizen to understand and predict his or her world. On average this is likely to be a good thing - more likely than the search for deterministic prediction of a specific social situation, like a project to predict outbreaks of conflict.
I know two waves of research which have expanded our vocabulary. The first is game theory. Any educated person knows what a Prisoner's Dilemma is, and many people can identify a Chicken Game or a Battle of the Sexes. You can even nowadays find Guardian articles which complain about the price discrimination practices of train companies. Underneath the hood is contract theory, and perhaps in future this will be a commonplace too.
The second comes from psychology, and it is happening now. Just world theory, loss aversion, and similar concepts are gathering media interest. In a few years, with luck, they will become commonplaces, widely enough known that groups of people can use them to swiftly understand what's going on in a particular situation.
Thinking about social science in this way lets me stay rigorous - a mechanism has to be well-defined and have conditions where it applies - while admitting that social life is, as Dan Kahneman might say, a low-validity environment, and that the people on the ground in a given situation may be better placed to apply the right concept than I am.