Some time ago, I discovered an online essay by Lev Manovich, called “Database as a Symbolic Form“. It is a condensed version of a chapter in his book, The Language of New Media, which has been sitting on my shelf for almost a year, one of the many books that I am always intending but never quite managing to read. The essay’s central argument assumes that, where the age of the novel and film privileged narrative as “the key form of cultural expression,” the computer age privileges the database in its stead. It argues that new media objects often lack stories as such, being comprised of many equally significant elements that have no essential beginning or ending and no form or development of any kind. In my opinion, however, the assumption that a database does not function narratively is highly suspect for several reasons.
First, from a technical perspective, the elements in a database are never actually equal in significance. They are always entered in a sequence, and they are assigned their position in a sequence. Some element will always occupy the first position (1,1) and will function as a beginning. Some element will always occupy the final position (x,x) and will function as an ending. Other elements will always occupy the positions between them and will function as a development. This beginning, this ending, and this development will always combine to form a narrative, even if this narrative is only of the simplest kind, even if it only says, “Look, though there are only seemingly random numbers, here is the highest number and here is the lowest, and here is the one that is repeated most,” even if it only says, “Look, though there are only unrepeated and seemingly meaningless symbols, this one looks something like this one that came before it, and there seem to be many symbols that have curves, while only a few have angles.”
No matter how random and meaningless the elements of the database might seem to be, these narrative functions are always operative, because of the conventions that govern reading and writing, whether these are the coded conventions of a machine reader or the social conventions of a human reader. Even if the writers or the readers do not in fact follow the established conventions of the code or of the culture, they must nevertheless follow some convention, must produce some sort of narrative, and must always do so in the context of what the established convention is, even if only through opposition to these conventions. It will never be possible for them to write or read without a narrativity, and it will never be possible for this narrativity to be entirely dissociated from the established narrative conventions.
Second, every element in the database is itself the function of one or more narratives. It is always artificially isolated from a story that is ongoing in the world beyond the database, even and especially if the elements are random numbers chosen for their randomness, even and especially if they are only meaningless symbols created for the purpose of meaninglessness, even and especially if they are only natural elements chosen for their naturalness. In every case they will be the products, the signs, the representations of at least one and probably many narratives.
To ignore the role of these source narratives in determining the data in the database is to ignore their physicality, their historicity, their locality. These sources are not always visible through the data that they produce, but they are nevertheless essential to the production of the data as such. In this sense, the database might even be said to be more narrative even than a traditional narrative, because it combines all of its source narratives into a single master narrative while still maintaining these sources as separate narrative elements in ways that are difficult for traditional narratives to accomplish.
Third, it is obvious, particularly in light of the kind of work that Jacques Derrida and others have done on the function of the archive, that it is impossible to understand the database apart from the narrative of its own production. In every case, the database is constructed by a particular producer for a particular purpose, even if that purpose chooses to take a form that appears random or purposeless. The database is therefore always and entirely implicated in the narrative of its own production and creation, in the narrative of its own purpose, whether political or aesthetic or functional or whatever, and in the narrative of what it may in fact produce in those who read it.
There is no escaping these narrative aspects of the database, and there is no separating them from the social, political, cultural, and economic implications that such narratives entail. To pass over the narrative function of the database is to impose on narrativity an artificial limit and an illusory exteriority. The only database that could actually occupy this position would be one that was neither written nor read, one that was neither populated nor empty, one that was neither ordered nor random, one that could be defined only by a language so paradoxical as to have become a theology.