Visual thinking gone public: for the better or worse?


Using vision to think with is not something new; people have acknowledged the expressive potential of visualization techniques since the 19th century. The upcoming popularity of the visualization of data and the spill of this technique from the domain of experts to the domain of amateurs is something we have never seen before. Creating knowledge and thoughtful insights by means of effective visual metaphors as a solution for the current overload of abstract data that is present within our society is something we should embrace. But is this really the case?

Visualization techniques not only attract scientists, analysts and statisticians but also art designers and other artists. However, datavisualization is not only made available for these professionals but also any other curious mind with a computer as a result of the many online user-friendly interfaces that make possible the mapping of large amounts of data into visuals and the sharing of those visualizations with a large online audience. Many Eyes is one of these online visualization tools, created by Fernanda Viégas and Martin Wattenberg who believe that datavisualizations contribute to a more informed society and will ultimately lead to the further democratization of our society when more and more people are able and encouraged to create their own datavisualization (Viégas et al. 2007). Cybercritics like Robert Kosara warns for a visualization Cargo Cult where visualization techniques are used without understanding their application or usefulness within the field. While this is true, I argue that we should also be aware of how aesthetics can work within certain fields.

First of all, the field of datavisualization is made up out of two extreme domains being art and science and it’s obviously hard to find a balance when working with extremes. While the latter is predominantly concerned with the function and effectiveness of the visualization and thereby neglecting the possible contribution of aesthetics to this function, the former often has the tendency to neglect functionality in its visualization design which leaves the visualization without an argument to convey or a story to tell. Instead of blurring the boundaries, Kosara believes that we need to make a clear distinction between art and sciences within the field of datavisualization or, put in his words ‘it will be too late’ (Kosara, 2010).

This critical distinction between visualizations as art and visualizations as information can be considered helpful while it aims at identifying the broad field of datavisualization and exploring it’s unique characteristics. Secondly it helps create an understanding of what  ‘good’ and ‘bad’ visualizations are in relation to the context, audience and purpose for which the visualization is created. On the other hand, focusing on clear-cut boundaries can ignore the possibilities that arise from the diversity of the field and deny the existence of datavisualizations that incorporate both ends of the spectrum i.e. that are both beautiful and informative at the same time. Furthermore there needs to be room to explore the boundary between art and science within the field of datavisualization in order for us to learn how these two domains can compliment each other. This means understanding aesthetics as a technical design choice but also understanding how graphic design can be enriched by science.

When the aim is to attract a broader audience beyond the domain of the experts, the power of aesthetics to awaken the senses and arouse curiousness can be used to visually stimulate the reader to discover the meaning and knowledge behind data. Moreover it seems that aesthetics can help create insight into data pattern, which according to Colon Ware is fundamental in the extraction of meaning through visualization (Ware 2008, 4). This is however not an easy task; conveying an argument with words is difficult and using visual design does not make it easier. On the contrary, designing requires certain complex skills. Colin Ware argues that during the process of designing, the designer must be able to critically analyze which patterns will lead to the right cognitive actions and visual queries provided by the visualization technique. Ware calls this skill ‘critical seeing’, a skill that is mastered through years of experience honing critical perception’ (Ware 2008, 64).

This is where the words of cyber critic Robert Kosara fall into place, claiming that a datavisualization makes more sense in either being artistic or insightful than fail at being both (Kosara 2010 b). When talking about finding a balance between form and function, one has to keep in mind that the right balance does not exist: it is context dependent. This consideration is also useful when applied to the visualization tool itself. Taking in account the context, audience and goals of Many Eyes, a pragmatic approach is more appropriate than an artistic one. The average user simply does not posses the skills and expertise that is required to create an effective visual design. Finding the right graph to fit the data in order to create a meaningful result is already a difficult task for users without any experience with visual design. From the research I have done, I can conclude that Many Eyes is encouraging (instant) usability at the expense of the visualization design and that this is a strategically good choice.

Moreover, by experimenting with online user-friendly visualization tools such as Many Eyes and thereby exploring the balance between form and function, we can learn about the existing trade-offs and become better judges on how to compromise in different contexts. Furthermore online visualization tools will have to consider certain trade-offs in their design in supporting this goal. In this way the democratization of datavisualization can lead to new insights concerning a healthy balance by providing a learning environment. 





Kosara, Robert. 2010, april 7. The Visualization Cargo Cult. (Blog). Eagereyes.



Kosara, Robert. 2010b, may 3. Visualization Can Never Be Art. (Blog). Eagereyes.



Viégas, Fernanda, Martin Wattenberg, Frank van Ham, Jesse Kriss, Matt McKeon. 2007.

Many Eyes: A Site for Visualization at Internet Scale.EEE Transactions on Visualization & Computer Graphics 13 (6 ):1121-1128.


Ware, Colin. 2008.Visual Thinking for Design. Burlington: Morgan Kaufmann Publishers.


Nice article

Nice article Michelle, it probably is your get real paper summarized? I believe the artistic vs science (or pragmatic may be à better word as not all pragmatic visualisations are also scientific) distinction is really useful. Maybe it would, with emergence and democratization of dataviz even be à good idea to adress only one of both as research object. As you say, there are so many different visualization... And pragmatic ones surely cant be judged and critizised by the same criteria as more artistic ones.

Other tools

Hi Michelle,

Nice article! The problem with Many Eyes seems to be that it is no longer developed, since it was an IBM experiment that ran until 2010 (Also: everything you post on there belongs to IBM, just like the way Facebook owns you...).

Did you also check out some other datavisualisation tools? Tableau public seems very interesting, D3 already a bit less user friendly but more tweakable/programmable. Any other ideas for good datavisualisation tools?

More tools

Here is a great list, not all of them web-based though.