Abstract art teach us about data visualization

If we look carefully, it can become evident that many of the tools that the digital world offers us are not based on strategies that are foreign or completely new to the humanities. In fact, despite the fact that there are now iterative, massive and automated ways of understanding culture and human societies, the underlying strategies have been with us for a long time.

For example, what could a painting by Picasso and a bar chart have in common? Or, what can modern art teach us about data visualization? While modern art, like Picasso’s, is usually associated with the exploration of individual creativity, data visualizations are associated with the cold description of world events; could they have anything in common? To find that relationship, we have to think about the abstract.

In an article entitled “The Essence of Mathematics”, the philosopher and semiologist Charles Peirce proposed a description and definition of the logical process of abstraction, which is fundamental in mathematical thought. That process, in simple terms, could be explained as follows: let us take a sentence that has a noun and an adjective.

For example, as Peirce himself does, let us think of the phrase “Sweet honey”. As the phrase stands, we know that “sweet” is a quality of honey; hence we put it as an adjective. But what if we modify the phrase in this way: “Honey possesses sweetness.” By doing this, which is no small thing, we are noticing the adjective “sweet”, and we are separating it from its dependence on “honey”.

After that process, we can look at “sweetness” and think about it without relating it to honey. What else has “sweetness”? Sugar, panela, ice cream, even, if we want, people. At this point we are manipulating the abstract quality of sweetness and we are projecting it on many different cases, just as it happens, for example, with an abstract trigonometric formula that applies equally to many cases of triangles.

When we abstract, we take a feature of the experience and separate it from everything else, we give it a special importance that allows us to gain an insight, a novel learning about the world, a noiseless generalization. Peirce understood this with respect to logic and mathematics, but it was also understood by modern artists.

In 1945, the artist Pablo Picasso made a series of lithographs that represented the figure of a bull in many ways, sometimes with a lot of detail and sometimes with very little. When seen separately, each of these works is simply called “The Bull”, but when they are put together they acquire an interesting title: “Eleven States”.

It turns out that Picasso used the same lithographic stone to make each of the versions of the bull, and his process of artistic creation consisted of making small modifications to each previous version of the image. It’s as if we were making a very complex drawing and with a draft we were selectively blurring it.

Thus, little by little, as the states advanced, the graphic representation of the animal became simpler. The eleventh state represented, if you like, the essential bull, the minimal bull. Like many modern artists, Picasso thought that this process of abstraction was useful to find new aesthetic appreciations of the world, and that abstraction, more than a matter of instinct, is a rational process.

The work of art is a sign that manifests the mental process of the artist, and that gains meaning thanks to the process of abstraction. The art critic Guillaume Apollinaire said explicitly in the Cubist Manifesto that “Most of the new painters do mathematics without knowing it, but they have not yet abandoned nature, which they patiently interrogate to be taught the path of life. Picasso studies an object like a surgeon dissects a corpse”.

Data visualizations used to study culture and societies in the digital humanities use strategies similar to those of art and mathematics. Relevant qualities of an event are separated out and we see how they are repeated in different cases, how they can be generalized to obtain information that was not evident before the abstraction.

Many times, these visualizations allow us to bring to a human scale very long extensions of time, space or numbers that cannot be easily counted; in other occasions, the visualizations clean the noise to settle what is there but it is not so evident.

For example, what do you think of my still life/abstract visualization? It is based on the “Still life with quince, cabbage, melon and cucumber” by the Spanish painter Juan Sánchez Cotán. What new information can you give us about the painting?

One last thought. Turning the abstract into signs, into visualizations, involves another interesting process, which is equally humanistic and mathematical: the visualization must be interpreted by someone. Thus, visualization is an interpretation, because it involves structuring data from cultural objects, but it must also be re-interpreted by an audience.

Although computers can perform extremely tedious operations for human beings, and can save us a lot of work, they do not have the interpretative capacity that we have. Visualization synthesizes complex information, but it is our task, just as when we go to an art exhibition, to offer an interpretation, and to evaluate if the abstraction is correct, if it is biased, if it belongs to an avant-garde movement that is already outdated, or if it proposes new points of view on a common subject.

Whoever visualizes the information that is gathered from human complexities, in some sense must have an artist’s eye, and whoever interprets it must have a critical eye.