Reading list, 2 March 2019

Some interlinked reading and listening ...

A new academic study showing that the collections of 18 sampled US art museums' collections are hopelessly white and hopelessly male has been written up in many outlets - such as on Artnet. The study itself is also available online. Here's the method: 18 museums with their collections fully available online were selected by the researchers as credible and representative, then ...

We scraped the public online collections of 18 major U.S. art museums, retaining the
museum name, artist name, and a web link pointing to the artist’s entry in the
museum’s collection. Then, we deployed a large random subset of scraped records to the
crowdsourcing platform Amazon Mechanical Turk (MTurk) and asked crowdworkers to
research the demographics of each sampled artist, using the web link as a starting point.
Crowdworkers reported their inferences of gender, ethnicity, national origin, and birth
year, along with a numerical rating of confidence in each inference. We put in place
multiple safeguards and checkpoints to ensure the quality of our data. Starting with the
data obtained from crowdworkers, we first eliminated records that did not correspond to
individual, identifiable artists. For each remaining record, we aggregated the inferences
of multiple crowdworkers and if their responses were sufficiently self-consistent, we made a final inference of each artist’s demographic characteristics.

In line with that - a piece of research from last year, co-commissioned by Art Agency Partners and Artnet, asked African American Artists Are More Visible Than Ever. So Why Are Museums Giving Them Short Shrift?

And the Art Agency Partners podcast this week was a recording of a panel discussion they held recently at Frieze art fair: Californian Museum Leaders on Expanding the Canon.

And some randoms ...

Adrian Ellis writes about the increased public scrutiny of museum boards in the social media age, the first in a new series for The Art Newspaper.