The biggest puzzle for me coming out of this week’s reading lies somewhere between Bodenhamer’s epistemological concerns and Monmonier’s many representations of the same data.
These images are probably familiar to people:
The first is a map of the world “upside down.” My primary reaction to this map is how well it illustrates the primacy afforded to North or ‘up,’ at least in Western culture. Look how superior Australia looks! What I don’t realize when I look at that map is the extent to which I appreciate the altered perspective, but still more or less consider it to be a distortion of physical reality. For that I need these next two photos, taken from Apollo 17 as it orbited Earth in 1972. The first is the actual photo as it was taken. The one beneath is the photo as it was made public (known as the “Blue Marble”); it has been flipped. This may be a slightly embarrassing thing to admit, but it actually takes me a minute to understand how it is possible to take an upright photo with the South Pole on top. The extent to which this spinning object with no real point of reference in the universe exists in my imagination as having an actual ‘up’ side is a fairly hard idea to break.
I bring up these examples because I think they illustrate the problem with how all things in mapping are relative–literally, in the sense that space is only measured by relationships, but everything about how that relationship is oriented is subject to interpretation. I think this raises difficulties but also really interesting possibilities for mapping.
Monmonier creates several maps, number line plots, and scale labels for the same data. The differences are rather dramatic. His most prominent example in the statistical data chapter has to do with infant mortality in New Jersey. He is trying to figure out how to represent Essex county, where the majority of deaths occur, but which is also heavily populated. The many versions create really disparate perceptions of how focused in Essex county the problem is. It seems fair to modify the visualization of the data to something closer to a per capita representation: without it, it seems like there is a unique problem in that county that other counties do not have; with it, you can’t see as well the concentration of total deaths and may have a skewed view of the actual numbers. And this seems a far more straight forward mapping project than a lot of others.
The question I am getting to is one of process: How does one ascertain the scale and the set up that best demonstrates a desired narrative, without warping the data too far to support that narrative? To what extent is our freedom to reconfigure these relationships limited to the graphicacy of our audience–their ability to recognize the world as the world from a unique but no less true point of reference?