It is . . . subsumed and amassed cultural capital that mapmaking societies bring to the task of making maps; not the patiently acquired mastery of this or that individual more or less carefully passed on—often in secret—through speech or gesture or inculcated habit. It is the endlessly reproduced and everywhere disseminated wisdom of thousands of such individuals, caught up, stored, annotated, corrected, indexed, epitomized, reduced to formulae, taught by rote, so that what once was an epochal discovery or invention is reduced to common knowledge, grounded into a taken-for-granted fact of life. (Wood 1992, 48)
In attempting to rethink database technology (and dataveillance) within the context of computer networks, solicitation techniques, and information mapping software, information is consistently updated in a cybernetic or cyclical fashion—a process that is heightened by its constant production of graphic representations of space (maps). Although one could rightly argue on purely technological grounds that computerized maps can condense data more efficiently than database outputs or that the power of maps stems from their representational status as iconic, easily recognizable signs, such approaches nevertheless fail to account for the reflexive element of reproduction, spurred on by the need to accumulate increasingly more precise data profiles. Geocoding psychographics from a database, for example, is essential for successful mail-order campaigns (catalogues), yet at the same time there is no absolute requirement to view (in the graphical sense of the term) the addresses and geographical entity in question (that is, the topographical market). As such, this indexical approach could be interpreted as making a clear distinction between human features and topological characteristics: ‘‘It is important to understand that geocoding software applications do not require a mapping system [such as GIS]. A geocoder simply reads data, analyzes the address specific portion of the data, and then attempts to match each address with a location, a latitude and longitude’’ (MapInfo 1995,3).
The introduction of computer maps by GIS, however, calls into question this ‘‘split’’ between human and geographic data—a bifurcation often characterized by empirical and scientific claims. In other words, traditionally (Martin 1991, 4–5),
data relating to dynamic human populations are very different in their geographic properties to those relating to the physical world: the location of any individual is almost always referenced via some other spatial object, such as a household address or a census data collection unit. Unlike a road intersection or a mountain summit, we are rarely able to define the location of an individual simply by giving their map reference.
Thus in addition to the obvious claims of efficiency and ‘‘clarity’’ —to the extent that GIS produces pictures of relationships— such computer maps also enter into a search for empirical or ‘‘true’’ representations of both the geographic and human world. Given the inherent problems in predicting human behavior and the ever-changing environment (natural and human), we might ask ourselves how an accurate picture is ever achieved. One answer (to return to the cybernetic qualities of diagrammatic thinking) is by cleaning. According to the Group 1 Software corporation, ‘‘Dirty data is the number one cause behind inaccurate mapping analysis’’ (MapWorld 1996, 5). In the case of the database, then, the process of capturing the world ‘‘out there’’ must be up to date. It cannot be missing a zip code, or consumer promotions, political campaign material, and state documents would never reach their desired locations. Computer ‘‘cleansing’’ programs such as those marketed by Group 1 locate the correct addresses while also attempting to reconstruct an accurate and ‘‘true’’ representation of the world’s topography. In this respect, one could view ‘‘the dirt’’ (or incorrect data) as what escapes initial attempts at ‘‘cleaning’’ (because it is miscoded or misformatted), hence driving the never-ending cycle toward pure, white, clean data—the accurate representation of human topography.