Thursday, February 21, 2008

Picture or Graph? A Conceptual Tussle

My recent post on informational images addressed the unambiguous world of pictures, as opposed to other forms of informational graphics. Occasionally these forms cross-breed and produce unhappy results. A case in point: this image, from a set of graphs designed to supplement a LIFE magazine feature on the economic outlook in 1953.

If you read the text, the graph has been created to compare economic performance data from the period 1929-1935 with data from 1950-51 and projections for 52-55. But the thing we're looking at takes us somewhere else. Indeed here's a case study of how pictorial logic can gum up an altogether different visual form.

The cityscape at the top of each visual field creates an implied space across which various characters (inluding Paper Sack Man, Milk Bottle Man, and Mr. Telephone Services) travel. Alas, this same visual field is asked to function as a Cartesian graph tracking time on the x axis and dollars on the y (but without numerical information of any kind). We are left with a weird perception that perhaps the black non-durable goods line occupies a position more distant from us than does the gray-green durable goods line. I'm so wrapped up in trying to interpret an image that ought not be an image that I can't get to the data, let alone an analysis of it. In short, three-dimensional inferences goof up the two-dimensional information.

Images: economic data graphs, LIFE Magazine, January 1953.


Ivan said...
This comment has been removed by a blog administrator.
Rob Dunlavey said...

DB, I think there are definitely loose ends in these figures that makes them not work very well (when examined closely). However, I think they could be redesigned and made to be much clearer and still contain the 2-d/3-d space.

-The cityscapes at the top of each figure contain information that should be more obvious. The first image should clearly read as a city in economic decline while the second shows increasing economic activity. This should accentuate the information in the graphed items.

- the cartoon figures and lines need to be pushed down so they are clearly in the 2-d coordinate space. Since there is not a significant vertical axis with units marked, the lines have a lot of latitude to be placed in relation to each other rather in strict accordance with the scale.

This pair of illustration suffer from poor execution more from any inherent ambiguity (which is very real!) of mixing 2-d and 3-d spaces.