Computational evidence of first extensive usage of violet in the 1860s

Abstract Tager showed that works of art rarely contained the color violet until the Impressionists in the mid‐19th century. This article further investigates the prevalence of different colors in paintings. We analyzed 4117 Western and Eastern paintings, extending Tager's previous research. Combining color science, image analysis algorithms, and new research on color categorization, we were able to calculate the frequencies of occurrence of several different color categories in digital images of paintings from several major art museums. Our analysis shows how the frequency of use of the various color categories vary between different cultural periods and regions. The color violet was shown to have a substantial presence on only 2% to 4% of the artworks created before the early 1860s. It quickly became very popular soon after that date, rising to a substantial presence of 36.5% to 47.9%. Textual analyses indicate a similar trend in the popularity of the color term violet. We discuss several explanations for the steep increase in the use of violet in paintings in the early 1860s. Neither cultural factors with roots in ancient Greece, nor the introduction of cheap purple and so‐called violet pigments in the 1860s, can sufficiently explain the trends that we found. An increased sensitivity to violet colors due to continuous evolution of the human visual system offers an alternative explanation. We finally show that contemporary developments in color theory and their adoption by Impressionist painters may almost naturally have led to an increase in the use of violet from 1863 onwards.