Bad diagramming

Advertisements

Reconstructing the mathematical paper

img_1892.jpg

Group 1 divided their paper up into sections that were outline and sections that were filling-in, a simple job for the diagrams, but a little more subtle when it came to the text

img_1887.jpg

Group 2 re-presented their paper as a comic strip, prioritising diagrams and adding a visualisation the theorem, and bringing annotations into the diagram to remove the need to switch between diagram and text. They suggested extending this by stepwise adding annotations, repeating images and adding annotations as they became relevant. Another participant noted that some outlets ask for graphic abstracts rather like this. The sequential math website (https://sequentialmath.com/) was also noted as a reference point.

IMG_1823.jpg

Group 3 worked on a colour-coding presentation of an argument, and attempting to present associated ideas on one page.

IMG_1780.jpg

Group 4 laid out the pages of the paper according to the way the argument progresses around the diagram

Deconstructing the mathematical paper

As part of the Reimagining the Mathematical Paper workshop at Bridges 2018, participants were asked to find different methods for taking apart and classifying a few pages of a published paper. Some of the methods they came up with are as follows:

  • Ordering sections according to symbolic content (other participants asked whether there tended to be more symbols in certain sections, and though the group hadn’t paid attention to this as they went, they agreed this might be of interest)
  • Noticing which sentences were in the imperative mood
  • Finding alternative ways to understand particular sentences
  • Noticing the agency of the sentences: often there was an agentless sentence followed by a ‘we’ sentence
  • Finding five different ways to read a paper: backward, first sentence of each paragraph only, figure captions only, etc. This technique is used by a participant in the refereeing process!
  • Considering the exclusive we, the inclusive we, the royal we, passive voice, and so on
  • Reducing the amount of text and instead integrating labels into the image, making it interactive etc.