This is a potentially huge topic and one that will not be answered simply or completely on a blog. The goal is to gather material for a series of lectures for my class and basically see where it takes us.
The first thing I'd like to do is explore map types and try to establish a common frame of reference for discussing them. After that, I'm just going to let the Train of Thought roll along at a leisurely pace, occasionally changing tracks as new ideas present themselves (either from my own rumination or your comments) until the whole thing is derailed by the Cow of Disruptive Thought.
From a top down perspective, there are three general data types, which leads to the following map types...
SPATIAL MAPS
The oldest and most basic type of map, a spatial map translates physical location data into a visual reference to allow you to physically go from point A to point B. The most interesting point about spatial maps is the fact that they do not have to be pin-point accurate to get you where you want to go. The first maps were rough approximations, at best, and even some modern maps (like a map of a carnival or fairground) are less interested in spatial distances than they are spatial relationships. A map of the Six Flags Amusement Park, for instance is not drawn to actual scale, but it clearly lays out that the Superman ride is after the Shockwave ride and before the Titan. And all of these rides are laid out on the map as huge drawings on a tiny area that, if they actually existed in that ratio, could be seen over the curve of the horizon.
TEMPORAL MAPS
A temporal map defines not physical space, but temporal progression, in that if event A happens, you can expect event B to happen. A sheet of music, for example, shows the progression over time as does a storyboard or workflow. Events A and B can take place anywhere, and need not have any spatial relationship to each other. Shot 1 on a storyboard might be followed by Shot 2 in a film, but they might have been recorded half a world apart, for instance.
In a relational map, we are looking for abstract ties between data points, how they relate to each other. Data Point A and Data Point B may not exist in the same physical relationship, in other words, you can't find B just by knowing the location of A. They may not even exist in the same temporal location. Data Point A might only apply in the 1800's while Data Point B applies to today. They simply have a variety of non-causal similarities that allow them to be mapped relative to each other. For example, Human Male 1 in Texas has never met Human Male 2 in China, but they can be mapped together because they are both Human Males. Vin diagrams are excellent examples of relational maps.
MIXED MAPS
The combination of the three types above can lead to a wide variety of mapped data representation. In gaming, for instance, we adventure maps that not only represent a spatial layout of a dungeon, castle, spaceship, etc., but also have a meta-function as temporal maps, arranging hallways and rooms in a manner that ensures the players will complete the dungeon in a way that builds up the excitement slowly and keeps the challenge level steady even as it allows them room to explore. You never fight the big boss monster in the first room, for example, but only after traversing the whole dungeon and facing a succession of increasingly more challenging monsters.
Another example would be the board game 'Black Death' which represents an abstract spatial map of the lines of disease communicability in medieval Europe, but also a relational map, where certain squares (marked with a + or - and a number) are more or less likely to spread the disease based upon the virtue of their being slums or hard to traverse terrain. It is also a temporal map, as the movement of disease using its virility rating doesn't represent the disease literally walking across Europe, but slowly being passed, over time, from person to person in one area and then breaking out in another, nearby area. The loss of a disease chit from the board through its Lethality also reflects the temporal nature of the disease running its course and killing off the population.
Yet a third example would be a farmer in a rural area giving directions by using not only spatial references ('turn left at the farmhouse') but also relational ('that was completely flattened by a meteor') and temporal ('in 2001').
THAT'S A START...
Now that we have a set of definitions, I think there are a few questions that raise their heads, the first and most important being, why do humans feel a need to spatially map abstract data? I'll look into those and anything else that comes down the tracks as the series continues...
