There is now a better way. Petabytes allow us to say: “Correlation is enough.” We can stop looking for models. We can analyze the data without hypotheses about what it might show. We can throw the numbers into the biggest computing clusters the world has ever seen and let statistical algorithms find patterns where science cannot.
While the poor guy is getting shellacked on the comment boards, he’s on to something. He probably overstates his case for the natural sciences, but his argument is more telling for the social sciences. If theory, even universal theory, about human behavior is time bound and context dependent, and society is innovating and changing at an exponentially rapid pace, then what good is universal theory?
Bent Flyvbjerg’s wonderful book Making Social Science Matter makes a related but different argument about the shortcomings of applying scientific principles to social science. he argues for an emphasis in social science on phronesis, or knowledge on the “art of living,” rather than episteme, or knowledge for its own sake. Here’s a telling passage from an essay derived in part from his book.
Regradless of how much we let mathematical and staistical modeling dominate the social sciences, they are unlikely to become scientific in the natural sciences sense. This is so because the phenomena modelled are social, and thus “answer back” in ways natural phenomena do not.
This is the guiding principle behind my own thinking about race scholarship. it is much more instructive for use to be guiding our scholarship towards knowledge that enhances the art of living in a multicultural democracy over the quixotic search for some universal law of race relations.
Comments 1
james braselton — February 17, 2009
HI THERE WOW A PERABYTE OF DATA THEN BLACK HOLE MEMORY AND BLACK HOLE RAM BLACK HOLE SSD AND LHDD LASER HARD DRIVES AND SOLID STATE FLASH DRIVE.