Over the course of the last several years, the Institute has taken an ever-increasing interest in the concept, application, and structure of databases. Undoubtedly, the history of databases could reach back as far as language itself, if the definition-in-use were broad enough; but the modern digital database (standardized, software-based, and machine-independent) is a recent and fairly well-documented invention.
The relational model developed in response to a variety of concerns: from rapid access to information, to data independence from the storage implementation, to the desire for a conceptual structure of information that would be easily accessible to human users. This simple fact is important because it reminds us that absolute flexibility with regard to faithfully representing "real world" data, although a priority, was often compromised for economic, technological, or other competitive concerns.
These tensions sometimes present themselves as limits to the database architect at the scale of the schema. Most often, however, the architect operates well within the bounds that three decades of innovation have left her: objects, reduced to products, become things with attributes (weight, name, price, availability), the preferred species of data. The cynic might invert the process, suggesting that objects are now already database subjects without need for market-driven reduction.
Certainly, the relational model of the digital database was developed as a kind of approximation of thought and memory. A "schema," for centuries, has been located between understanding and experience, between brain and stimuli, between Man and her Environment. The relational model was a particular specification for how information about the world is ordered and understood (whether in the single human memory or more expansive cultural forms of memory). What can be said now that these databases (and all of their mutations) have exploded into ubiquity? Does it still make sense to say that the database is like us? Or has it become that we are like the database?
We didn't ask this question to simply wish databases away; nor did we want to engage in the inverse-identical task of developing database models that more accurately describe some a priori "Memory." Instead, we sought out the boundaries of the relational database and tried to develop schemas that wavered between counter-intuitive applications of the model and functional implementations of open-ended problems.
