“Raw Data" Is an Oxymoron
by Lisa Gitelman, Editor
The MIT Press, Cambridge, MA, 2013
208 pp., illus. 16 b/w, 10 color photos. Paper, $30.00
Reviewed by Jan Baetens
“Data”, literally: “what is given”, is neither an issue of our digital age, although our storage and transmission capacities are now stretching far beyond the utterly thinkable, nor a characteristic of hard or social sciences, which seem to have the privilege (or the burden) the work with this kind of “given”. Data is as strongly present in humanist disciplines as in any other ones, and the term itself as well as its frequent use are much older than the communication and information society in which it has become so ubiquitous.
The ambition of this fascinating collection of essays gathered by Lisa Gitelman (who has a strong publication record in the field of media and data studies, first as the editor of New Media, 1710–1915, second as the author of Always Already New: Media, History, and the Data of Culture, two remarkable books published by MIT in 2003 and 2006, respectively), is threefold: definitional, historical, and critical.
In the very first place, the contributors of the book try to define the notion of “data” as precisely as possible. On the one hand, they do so by comparing the term to the key set of notions that often appear as synonyms or quasi-synonyms, such as facts, documents, information, evidence, figures, etc.: what is being highlighted in this respect, is the apparent “neutrality” of data, which appears as the “raw” material of further interpretation and explanation. On the other hand, they complete this comparative approach with an attempt to formulate also a number of intrinsic properties, such as the fact that data is always networked, that it never exists in small samples, and that it is highly context- and discipline-sensitive.
Second, the chronological architecture of the book draws immediately our attention to the necessity of taking into account the historicity of data, both as a theoretical and methodological concept and as a material reality. Leading us from an analysis of “data before the fact” (as goes the title of the essay by Daniel Rosenberg, who reconstructs the history of the word “data” through the OED and Google Books Ngram Viewer, the former still being much more reliable than the latter) to the ironical and self-critical “thick” description of data-collection in today’s geography and hydrography, “Raw Data” Is an Oxymoron demonstrates that data is something that changes dramatically over time, not only because we do not have given the same name (data or something else) to objects, items, figures, statistics, events that we tend to consider data today, but also because the things or objects we identify as data are always anything but stable. New methods, new theories, new words, new politics, new people relentlessly force us to rethink the data we have gathered and to modify them in retrospect (if possible, of course, but we all know what scientists mean by more or less radical “data-massaging”).
Third, and here is where the English culinary metaphor comes to the fore, all authors of the book show themselves highly critical of the validity of the adjective “raw” we routinely associate with the “data”, as if data were something neutral, non-constructed, unprepared, in a word natural. What Gitelman and her colleagues demonstrate with great profundity and wit is that there is nothing natural in data, but that, on the contrary, data can only exist if, regardless of any eventual interpretation and social use or abuse, it is first wanted for, then gathered via special techniques, stored, protected against all forms of corruption, permanently updated and, more often than not, restored, reconstructed, if not remade with the help of new (but rapidly outdated) insights and procedures. Of the essential frailty of data, the book gives numerous often exhilarating, sometimes painful and shocking, but always thought-provoking examples which show also how directly any of us is involved in the daily making, unmaking, and remaking of data.
However, the strongest claim of the book is that none of these three approaches can exist independently from the others. All definitions should entail a historical dimension and be open to critical evaluation, and vice versa, in all possible combinations. The aim of “Raw Data” Is an Oxymoron is therefore not, in a kind of naïve deconstruction, to warn us against the dangerous belief in the absolute value of the objectivity of data, which rapidly appears to be a myth. The message of the book is much subtler: data is not something that is true or false in itself; the value of data is the value of the way is being used. Neither is it something that has to be opposed to reading or interpretation: data itself is always already a matter of reading and interpretation. And it is our common responsibility, not to long for “pure” (or “raw”) data, but to contribute as consciously and as dutifully as possible to the ongoing co-creation of data and interpretations.
The essays in this book are all wonderful appeals in that direction. What makes them even the more inspiring and challenging is the shared disposition to expose the data problem in sometimes very difficult and complex fields in layman’s terms (and the well-chosen iconography enhances that objective). Given the disciplinary width of the collection, which brings together very abstract forms of mathematical speculation with hands-on political commitment and agitation through the use of paper clippings or the elaboration of new forms of sociology with the help of an old-fashioned index card system, this is not a minor achievement. “Raw Data” is an Oxymoron is a must-read for all of us, including for those for whom data work seems boring, frightening or ideologically suspect. It is great reading, great fun, great thinking, and it should make clear that critical data work is an all scholar’s duty that blatantly displays the urgency of a dialogue between the two cultures.