## For Lack of a Word

January 7, 2009

This from an encyclopedia entry:

*** is a notyetcoinedologism which refers to the practice of imposing structure on or between sets of data, in a more or less arbitrary fashion. Generally certain semantic cues are followed, resulting in a theory which is partially natural, while many connections will be found only after the choice of structure has been chosen. In its most natural degree ***ing is the process of observing structure that would be similarly observed by any from some large pool (e.g., humankind). As such, it is the bedrock for scientific theory of any kind. In its less natural degrees ***ing can be used to author mnemonic devices. It can also be used to mine new relationships between disparate sets of objects for the purpose of creating artwork, literature, music etc. An instance of *** is called a *** map. Pseudosciences may qualify as bodies of ***, although usage of the term *** presupposes a motive of playfulness and creativity. One does not *** to produce an irrefutable theory.

The origins of the word date to the year 2009 from a blog posted by Andrew Marshall, although the actual term was suggested by one of the blog’s readers. It may be a portmanteau of, or take inspiration from, the following words: algorithm, supervenience, append, arbitrary, map making, giving birth, theory weaving, superlogic, artificial, imposing, forcing, analogy, metaphor.

Examples where the modeling is mathematical include the imposition of a total or partial ordering on a set of real world items (or categories of items); the assignment of numbers to the members of a set, where qualities of the numbers are taken into consideration; the use of directed graphs to account for and suggest adjacency or local partial ordering. However, the product of *** is not usually entirely mathematical, as the natural language value of the objects involved is not forgotten. Frequently two sets with some similar interrelationships will be identified, whereby the modeling is almost entirely at the natural language level.

Concrete example: a piece of music with $n$ movements will be identified with a region $R$ containing $n$ counties. The (standard) duration of each movement gives a natural order to the movements, and the land mass of the counties gives a natural order to the counties. The two are identified accordingly. It is then noted that duration and note count are roughly correlated, as land mass and population are. We may then ask to what accuracy is population correlated with note count. As we scour the data relevant to the music and the land we might find that the number of sections in each movement corresponds precisely to one greater than the number of large rivers passing through the interior of each county. Upon such an observation, we will find a way to identify the land masses between rivers with sections of movements, perhaps still according to our land mass:duration correspondence, perhaps according to some other distinction. Having made such an identification we will observe moods of the movements and find ways of seeing each county as embodying that mood. We might, finally, write a short story involving $n$ people, each from a different county of $R$. These characters’ personalities will share perceived moods of their corresponding musical movements. We will give a clue or two in the story, possibly referencing the piece of music in some telling context, but the *** map is regarded, here, as scaffolding around a building or the wax prototype of a bronze sculpture. Some of it will not survive.