Catalog Tree

Via Greg Smith at serial consign, an interview with the Arnheim [N.B.: corrected from errant Amsterdam] graphic information design firm catalogtree.

Catalogtree creates compelling graphics and navigations through complex data spaces. In some ways, it made me think of a less visually bound or encapsulated version of Sea Dragon, but I think that’s an unfair comparison, as there is a flexibility in this work, permitting a rapid adoption of interface style, and there is no pretense at constructing a larger semantically unified content space, a la Photosynth.


From the interview:

In examining your portfolio I notice your very distinct style of not only graphic design, but information visualization. Broadly speaking, what is your philosophy for graphically communicating data?

Our approach to design – not just in designing info-graphics – is to devise a set of rules by which the content should ‘behave’. A design is a visual outcome generated by these rules. We feel this way of working might be an answer to the flow of dynamic content in newer media and might give a fresh look on the processing of content in older media.

Data sets – large once especially – are perfect for this approach: The content shapes itself into an intriguing picture.

We try not to be dogmatic however, if we don’t like the generated outcome, we change either the rules or the outcome. One needs to breathe, right?


Many of your visualizations are a series of infographics that break down larger issues into numerous data snapshots. Does this have to do with the way that your studio processes information? Is this an aesthetic decision?

If the data allows it, we like to break down information into a graphical hierarchy similar to poster designs: a larger motive or trend is visible at first glance, and more detailed information becomes clear on closer inspection. Furthermore, the design should reflect some of its content. The data does not always allow for this: many times a simple bar graph is best. That said, a design cannot exceed its content: bad data sets lead to bad graphics, however simple or conventional the design is.