The above chart indicates that, although they are small, some differences in the media used by male and female artists exist, with the notable observation that men remain the dominant producers of conventional media, such as painting and sculpture.
Drawing this inquiry further, basic log ratio analyses indicate that there are also differences in the 1,000-plus characteristics applied to the artworks by The Art Genome Project. That is, in some respects, women and men produce different characteristics of art according to this taxonomy. Although the large majority of features show no substantive difference by gender, there are those that do. Analyses in ongoing research further suggest that, provided with these features of The Art Genome Project, a machine learning algorithm can classify art by the gender of the artist with a relatively high degree of precision. How this finding maps onto the claim that women and men make art with different characteristics is a central focus of the research.
If the preliminary finding that men and women, in some respects, produce art with different characteristics is upheld by further scientific scrutiny, then the next step is to investigate whether the characteristics more commonly appearing in women’s work are also associated with lower prices or other outcomes of artistic value. If so, then the undervaluation of “female” characteristics may contribute to our broader understanding of gender inequality in the art world. Evidence that this hypothesis could be upheld is provided by a small but negative correlation found between the odds of an Art Genome Project feature appearing in women’s, as opposed to men’s, artwork and the listing price of the artworks displaying that feature.
Sociologist Pierre Bourdieu once quipped that “sociology and art do not make good bedfellows.” His reasoning was grounded in the tension between the art world’s desire to focus on individual creative genius, and sociology’s insistent aim to explain phenomena in terms of social forces. Who and what defines art and quality, which institutions matter and how they are accessed, who knows whom, whether advantage is accumulated from a prejudiced past, and where conscious and subconscious biases of culture interrupt economic valuation—these are the questions that sociologists ask to explain greatness. It is not a denial of quality, talent, innovation, or genius, but a way to contextualize them.
As shown here in a brief analysis, efforts to analyze this context in the domain of the contemporary art world can be aided by online art platforms. Data from these platforms help answer not only questions about gender, but also race and ethnicity, age, career stage, and socioeconomic status. And not only about inequality, but also networks and curation, the shifting influence of roles and institutions, the position of art in social life, the tastes and demographics of collectors, and the emerging sway of algorithmic recommendation systems and social media. In other words, the marriage of these platforms with scholarly research will teach us more than ever before about the roles, patterns, and processes that together (re)produce the art world as we know it.