Some concepts return cyclically, ideas of multiple decades that purport to change everything with renewed urgency each time they appear, usually under different names.  One such concept rose to importance in the 1830s in Ghent, and returned in the 1900s in Berlin, to reappear in the 1930s at the University of Vienna, and again in the 1940s in Brooklyn, then in the 1970s in Tokyo.[1] In the 2000s it reappeared in the digital corporate world—this time coded as “big data.”  Each time the idea appeared it was rebranded as new; each time it purported to make predictions about the future based on the analysis of present data; each time it spoke more about the recent past than the future; and each time the buzz lasted about a decade.  So I understand this as an idea of the decade.  But which decade do I choose? And which version of the concept?  Two books along this nodal timeline of may help give us some indication of where our current fetishes for the digital humanities and big data are (and not where they are going).

The first is a book of what I will call speculative fiction, though it is more usually classified as social science.  Published in Japanese as Jōhōkezaigaku (Econo-Information Studies) in 1976 by Masuda Yoneji, the book is neither simple economic analysis of a current social phenomenon nor a specific proposal for near-future policies reacting to the contemporary socioeconomic distribution of information.  Rather, it ventures into the world of long-term sociological prediction and, therefore, into the realm of science fiction.  Despite all the book will get right, to truly understand this work of near future fiction pitched as social science, I think it helps to consider it along with some truths from a work about social science which was sold as science fiction—and here, I refer to my second book, Isaac Asimov’s Foundation. 

Beginning in May 1942, Asimov’s Foundation series of stories are based on the premise that future history is predictable through a new science called “psychohistory” developed by the Shaman-trickster, mathematical genius Hari Seldon.  Though the depiction of Seldon’s successful predictions based on massive data sets might tell us much about big data today, we also might learn a lesson about psychohistory and our future “deep time” from Seldon’s inversion—in the character of Salvor Hardin—the shrewd, manipulating, neo-con who, though knowing quite well the power and truth of Seldon’s science, also understands how part of its power lies in glossing it as religion for the masses who will never understand it.  Hardin teaches the reader that mysticism, fictionality, and the minor malleability of history by individuals in unpredictable ways are integral to the proper functioning of the predictable grand system. 

We might read both these books as warnings about our contemporary interest in big data; our fetish (coinciding almost too easily with the rising interest in discourse history) too often denies the importance of individual, stochastic, or random possibilities in favor of big trends.  When we consider recent big data processing mistakes, the risks of this approach (denying the individual, the outlier, or the random) appear today everywhere from 1990s banking decisions to bundle risk to the Japanese government and industry’s decisions, in the 1960s, to build nuclear plants in historical tsunami paths.  From healthcare to border security, risk management structures continually gamble on probalistic logic without regard for the disasters that await in the random or “statistically insignificant” probabilities. These are the probabilities for which the humanities are made; let the unpredictable, immeasurable, sublime human individual continue to be the place on which we focus critical acumen and insight.

Notes