Databasing Life Patterns edit
Over the last two decades, many of us have felt the gradual and expanding involvement of technological information and the internet in our lives. However, more often than not, we fail to appreciate the subtle and pervasive implications these developments may have for the ways we think and behave. The accumulation of information, from the growing expansion of the trivial to the serious aspects of life that are recorded in databases (e.g. financial, medical or legal records, online habits) and the increasing sophistication of computer technology converge to confer to data and information a new and interesting role in the lives of people and the functioning of institutions. Information is not any longer confined within the world of computer-based experts. It increasingly infiltrates social life, constructs the perception of social events, defines priorities and relevancies, and frames the ways we approach and deal with them.
In a recent and provocative book that may bring tears to many humanists, the American legal scholar Ian Ayres (1) describes how modern technologies of communication and computing are involved in constructing relationships and profiles of people by virtue of manipulating available data. Such relationships or profiles are impossible to conceive and construct, unless one has access to huge and constantly updated databases that are possible to run and analyze through powerful computers. They derive from the comparison of our habits or choices over time and across different life activities that usually evade our perception and understanding. Who we are or how we act is assumed to lie hidden in the data that records habits, transactional patterns and other characteristics of individuals and must be brought forward through database analysis and data permutations. The profiles constructed range from the analysis of people’s online consumption and navigation habits to more complex activities that in traditional settings require consultation with human experts, such as medical, legal of financial expertise, sports coaching, sexual or partner preference mapping and others. Very indicative of these trends is the concept of “digital shadow” that projects the amount and diversity of data that can be tied to an individual but it is not precisely of her or his own making. These are data others produce of us (surveillance cameras, airlines, hospitals) including the data (traces) our web and online habits leave.
In being able to extract patterns, relationships and causalities that often elude human perception, inspection and understanding, database analysis seems to be able to crowd out human expertise and other traditional modes of human conduct from a variety of fields, activities or life patterns. The argument sounds undeniably old. It has been heard several times over the last few decades, arousing high expectations only to lead to gradual disillusionment. But there are reasons to believe that the argument regains actuality and relevance these days. This is due to the conditions established by an entire new range of technological, organizational and cultural arrangements that capture, store, process and circulate data of an immense variety. These conditions confer to the argument a credibility that was not possible to obtain during the time Dreyfus and others deconstructed the technological illusions of artificial intelligence and software engineering.
Ayres opens his book with a symbolic event that ushers us in the new age he seeks to describe, that is, the uproar created in the circles of wine expertise in US by the Princeton economist Ashenfelter. Bordeaux wine quality is known to depend on ripe grapes with high juice concentration. Both of these key characteristics are heavily influenced by the rainfall and temperature distribution over the year. Combining data over weather conditions, Ashenfelter was able to come around the high uncertainty of predicting wine quality from tasting young wines in the making and predict the quality of wine for the years 1989, 1990 in the region of Bordeaux with astonishing accuracy. His mathematical vision of wine caused a variety of angry reactions from the establishment of commercial interests, rites and activities that have centred around wine quality prediction but, if we are to believe Ayres, it managed to gain recognition internationally.
Ayres’s key claim is about the significance of database analytics that offer an inspection of life states and opportunities that transcends human expertise deriving from intuition, observation, and acquaintance with reality. Simplifying a little, we may say that the issue is no longer whether machines can map human intelligence but rather the variety of things than can be accomplished by drawing on the wide availability of standardized data, organized in huge and often interoperable databases that are possible to crunch by powerful processors. He amasses persuasive examples from a large variety of fields (e.g. baseball and sport coaching, chess, car stealing, e-dating, finance) that demonstrate the superiority of “database analysis” over “observational expertise”. We find the type of claims people like Ayers make challenging and, crucially, timely. Avoid confronting such arguments amounts to turning one’s back to reality. On the other hand, we do not share with Ayres and his likes the optimism that “super crunchers”, as he calls number crunching supported by huge databases and ample computer processing capacity, will invariably lead to better decisions and even less so to a better society. Better is an ethical not a cognitive term and while in some cases ethics and cognition may go hand in hand, in many others may not. Let’s have a quick look over some of the issues the databasing of life patterns is bound to give rise.
The issue whether machines are nowadays (given the construction of huge databases and powerful processing capacity) better able than humans in analyzing and predicting reality may be a misplaced one. In a world dominated by technological information, which is interoperable and granular, human agents qua cognitive decision makers are already in a disadvantaged position, in the same way that a pedestrian or cyclist cannot compete with automobiles in highways. But cognition, no matter how important, is only part of what defines human agency and humanity. Most importantly, the mediation of reality through databases follows principles that are predicated on just one (important but slim) form of cognition that gives premium to classification and standardization of data and events. Data are not recorded in databases haphazardly. Rather, they need to conform to the categories of the database and be in forms that are compatible with the underlying mechanics of computerized data processing. Classification and standardization thus presuppose the direct or indirect operation of a conceptual, logically constructed scaffold on which categories are crafted and make sense. The diffusion of databases implies that such a logically constructed scaffold gains significance at the expense of other implicit and associative ways of perceiving and framing life events. What is recorded in databases must pass through the bottleneck of the conceptual scaffold on which the database is crafted and the standardized forms of data or information that the technological system admits. Information that does not fit the categories of the database and the prevailing data standardization will most probably fail to be perceived or deliberately ignored or distorted.
The artefact of the database descends from millennia old information recording techniques such as list and tables, and other non-verbal forms of writing. Geoffrey Bowker has recently claimed that “databases are not a product of the computer revolution”, as most people may think; “if anything the computer revolution is a product of the drive to database”. The non-verbal cognitive organization of the database contrasts with traditional strategies of narration and the importance narrative has assumed in making sense of reality, including its contribution to constructing life trajectories and personal identities. Or, as Manovich has suggested, the database reverses the order of the classical elements of narrative (i.e. the plot and the description), squeezing narration and storytelling and hugely prioritizing description. Logical connections of database elements take command and constitute modern forms of life as derivatives of database associations. Structuralists and post-structuralists will of course claim that this has always been the case and that the database just makes evident the logical (cognitive polarities, differences) operations that underlie human thinking and cognition. Be this as it may, the distinction does have a merit and it is important to uphold. Databases formalize these differences, standardize their inscription forms and vastly increase the permutations (as Levi-Strauss may have said) of the recorded elements.
Brief as this commentary is, it suggests a series of complex trade-offs between the positive and negative attributes of recent technological developments. The issue is not to question the achievements or prospects of database analytics but rather to mark its territory and dissect the hidden assumptions on which its superiority is predicated. In considering what falls systematically outside the principles of database construction and algorithmic reasoning, we may at least obtain a better view of what is gained and lost when technological information becomes the key vehicle for understanding and acting upon, and, ultimately, constructing reality. On the other hand, it is important to recognize, as we have pointed out in a previous Telos article, that the salience the database and its cognitive derivatives currently assume are supported by a huge and powerful institutional machinery that by design or implication gives priority to logic over other forms of conducting one’s life. Perhaps, next generation Paris or London restaurants and wine bistros will score and price wines according to database analytics. We need Louis to sing for us “what a wonderful world”…
1. Ian Ayres, Super Crunchers: How Anything Can Be Predicted, London: John Murray, 2007.