Diagrammatic representations are omnipresent and are used in various application domains. One of their major goal, in particular for information visualization, is to make data visual in a way that a spectator can easily understand the graphical encoding to finally derive insights from the data. As we see, there are various different ways to visually depict data by using visual features in various combinations. In this paper we come up with some thoughts about existing diagram styles, for which we first discuss the benefits and drawbacks of each of them focusing on aesthetics based on readability. Additionally, we describe some initial results on the aesthetics of diagrams which we recorded in a web-based experiment. In this, we asked participants to vote for one of two given diagrams of a given repertoire of 70 of them covering all examined aspects which focuses more on aesthetics in the sense of beauty, not readability. The major result of this experiment unhides a trend towards colored, 3D, and radial diagrams which stands somewhat in contrast to readability user studies in information visualization oftentimes tending towards 2D and Cartesian diagrams for data exploration.
There should be much more evaluation in this field, in particular for simpler diagrams and also more visual features.