Macro-evolution of art: a tree or a network?
If we tried charting the evolution of art forms—styles of painting, genres of literary fiction, or techniques of filmmaking — what would it look like? Would it resemble a branching tree, like the phylogenetic trees commonly associated with biological evolution, or would it resemble a convoluted network? This is a variation of a more general question: what does the evolution of culture look like?
Figure 1: the famous diagram by the anthropologist Alfred Kroeber — “the tree of life and the tree of the knowledge of good and evil” (what a poetic phrasing!). This diagram was meant to foreground the difference between the two: in biology, branching; in culture, branching plus something else — reticulation, or horizontal transmission. Cultural “species” don’t just split, like biological ones, they also merge. We can combine the elements of two separate cultural “species” — say, an image and a text — and voilà: here is a comic book.
However, since Kroeber, the evolution theory has changed. Today, we know two important things that undermine his clear-cut distinction between nature and culture. First, biological evolution, in fact, isn’t exclusively tree-like. Say, in bacteria, the exchange of DNA between different species is a common practice. And so, paradoxically, the evolution of unicellular organisms looks much more like the “tree of the knowledge of good and evil” than the “tree of life”. Second, even though reticulation seem to be frequent in culture, traditional phylogenetic models, branching-only, are used with a huge success for describing culture: languages, artefacts, or folklore.
Thus, it isn’t obvious at all that the shape of art evolution must look like Kroeber’s right tree. We need more than a theoretical assumption, we need to look at some actual attempts to map the genealogy of art: how are those genealogies shaped?
When I started digging into this topic, I didn’t find too much. Few attempts — mostly made decades ago. Apparently, to learn about the patterns of artistic evolution, we need to go into the past. In what follows — three perspectives on art history… Three hypotheses about the shape of art history.
Hypothesis 1: Art history is chaos. Look at Figure 2: a genealogy of abstract art, from Alfred Barr’s Cubism and Abstract Art (1936). This book wasn’t an academic work, but a catalogue of an exhibition in the Museum of Modern Art. Nonetheless, it does reflect the professional view of the history of abstract art.
What do we see? Firstly, no single “root”: no single common ancestor. Cubism, the main hero of this graph, adopted the elements of at least four sources. Another prominent style, Fauvism, was even more omnivorous: six influences. The same is true of virtually every movement here: influenced by many, influencing many. Not a tree — an intertwined web.
Abstract art’s history doesn’t resemble neither of Kroeber’s diagrams — much messier than both of them. If art history is such a mess, do we even have to visualise it? What difference does it really make? However, an exhibition catalogue cannot be our final answer. And, besides, it certainly isn’t the only possible answer. So, from a perfect chaos — to its opposite…
Hypothesis 2: Art history is orderly. Figure 3 is a diagram from Franco Moretti’s Graphs, Maps, Trees (2005). This book isn’t old, like Cubism and Abstract Art, but it does follow an old, half-forgotten tradition of evolutionary thinking in the humanities. The figure shows the history of the “free indirect style” in literature: a technique of storytelling that mixes third- and first-person narration. It is a powerful stylistic tool, but also a complex one — which may be the reason why it was introduced rather late in literary history, by Goethe and Jane Austen, both located at the root of this tree.
Indeed, a tree. Where branching is the only force. Free indirect style, after being first introduced, gets popularised by Flaubert, who becomes the “common ancestor” for all the later branches, including the famous “stream of consciousness” narration of Virginia Woolf and James Joyce.
This diagram certainly is a tree, but was the reality so orderly? It’s hard to imagine that Mario Vargas Llosa hasn’t read Dostoyevsky. If so, should there be an additional branch connecting them? Or should it be added only if it’s proven that Dostoyevsky had a profound influence on Llosa’s writing style? In any case, authors don’t necessarily have just one influence, and it is easy to imagine how this tree could become a complex network, similar to the one with abstract art. After all, Moretti himself notes, in caption: “this figure reflects work in progress”.
Finally, hypothesis 3: Art history is an “orderly chaos”. Figure 4 is a map of Italian schools of painting, published as a standalone chart in 1925. This map is a bit of a mystery — created by some Giovanni Fattorusso according to the principles unknown. The map description vaguely says: “carefully compiled from leading authorities on art: — Vasari, Venturi, Corrado Ricci, Müntz, Crowe & Cavalcaselle, Kugler, Layard, Morelli, Rumohr, Springer, Berenson, Symons, etc.” So, a synthesis, but the exact execution of this synthesis is only known to the “careful compiler” (of whom we know very little).
For a moment, let’s forget about the uncertainties of this map, and let’s consider it, as both diagrams above, not an answer but a question: a hypothesis for how the genealogy of art might be shaped. It is shaped differently from the previous two charts. Neither network, nor tree: a combination of them.
Partly, it still is a tree: we clearly see separate branches and traditions within each school of painting. But, at times, these branches merge — like on Kroeber’s tree of culture. Which exact branches? Apparently, ones close to each other. Siena branch (green, left) had many connections within itself and with its neighbouring branches: Rome, Umbria, and Perugia. Less connections — with Florence. And no direct connections with Padua and Venice. So, closer distance between the branches means higher connectedness.
Interestingly, this pattern of merging the neighbouring branches can also be found in a completely different, non-cultural domain: microbiology. Eugene Koonin, a renown evolutionary biologist, explains one of the hypothesis of the shape of bacterial evolution:
“the tree-like pattern of evolution actually might be a consequence (one might provocatively say, an artifact) of nonuniform, biased HGT [horizontal gene transfer], whereby organisms that appear ‘close’ in phylogenetic trees actually exchange genes frequently, and organisms that seem ‘distant’ in trees are those between which HGT is rare” (Koonin 2011: 164).
This may also be a reason why, looking at some parts of history, we may find a network, and in other cases, a tree. Distance to the object of study is what could matter here: if we look from a close distance, we see a network (like the chart of abstract art, being a rather close look at a short period of 45 years). If we look from a large distance, we see a tree (and that’s what we see on the figures with free indirect style and Italian paintings, presenting a more distant view of 200 and 250 years, respectively).
So: three views of art history, three hypotheses. A rather chaotic network. A perfectly ordered, branching tree. And an intertwined tree, with the vertical and horizontal transmission mixed in some proportion. Three hypotheses — three questions. But what is the answer?
Eventually, the answer must be empirical. After all, someone has to attempt a large-scale mapping of cultural (for example, artistic) evolution: to uncover its actual genealogy. One approximation has already been made: the Music Map, covering roughly a century of Western popular music. It has an interactive webpage, and Figure 5 shows its small portion.
Is Music Map an answer to our question? Probably not. At least, not the final one. After all, it was made by a single individual, while constructing a genealogy aimed at being not a hypothesis but an answer, I believe, should be a collective effort, involving multiple specialists. Music Map does resemble the “orderly chaos” of Italian paintings, but that is all we can say at this point.
For now, the question is open.
