Analyzing the Structure of Mondrian's 1920-1940 Compositions

TL;DR

This study analyzes Mondrian's 1920-1940 compositions to test if their structure results from recursive splitting, finding that this hypothesis is generally not valid and that Mondrian's design involves more complex decisions.

Contribution

The paper introduces a novel interactive analysis tool and provides a comprehensive statistical analysis of 147 Mondrian paintings to evaluate the recursive splitting hypothesis.

Findings

Recursive splitting hypothesis is generally invalid for Mondrian's compositions.

Mondrian's design involves complex decisions beyond simple splitting.

Splitting alone cannot fully capture the aesthetic and structural essence of Mondrian's art.

Abstract

Mondrian was one of the most significant painters of the 20th century. He was a prominent member of DeStijl, the movement which revolutionized art by setting it free of the obligation to make images of existing objects, persons, or situations. DeStijl was one of the interrelated movements in early 20th century Europe including Cubism, Constructivism, the Futurists, and Dada. Their disruptive ideas changed the meaning of Western art. It was Mondrian, more than anyone else, who worked restlessly to find expression for the purest possible kind of beauty and truth, based on a theory called Neoplasticism. He was already famous during his lifetime and still now, his name is almost synonym for modern art. We analyze the structure of the system of black lines in his paintings and put the hypothesis to the test that the paintings could be obtained by recursive (binary) splitting. We used a novel tailor-made interactive analysis tool and apply it to as many Mondrian paintings as possible (in total 147). The results will be explained in a visual manner, but we also present statistical findings from the analysis of the 147 paintings. Our main conclusion is that the hypothesis of splitting decomposition is in general not true. It is possible to make Mondrian-like compositions by splitting, yet one misses out on a great deal of Neoplastic beauty if one would work by splitting only. It is possible to consider all crossings as pairs of Tees, but that is clumsy, and it leaves out essential information. Moreover there are other important design decisions of Mondrian, such as the keeping-distance to the canvas-edge which are not well-described by splitting.