But pronouns and gender identities are completely different things. We don’t need to store preferred pronouns in the database because there is no automatically generated text about the artist anywhere like: “[Artist] was born in [location]. [Pronoun] is/are active since [year].”
Would “gender non-conforming” be a better catch-all?
I understood @hiccup’s position like this:
As long as gender is a fixed list, MB is making decisions about which identities are significant (enough to be on the list), which ones are synonymous, which are meaningful etc.
I agree that MB shouldn’t be involved in that sort of decision (which, of course, is currently the case), which is why I voted for “Free-form writing (e.g. tags)”.
To me gender non-conforming means you identify with one gender (usually the one the doctors assigned to you at birth), but don’t stick to many of the “rules” society puts on you regarding how someone with that gender should look and behave.
So it’s not necessarily related to gender identity.
I still don’t understand why MB is aiming to make up a new list. MB should be driving to a standard so that other users of the data can work within a standard. Any list made up by a single person here at MB will have bias in it that will come back to bite at a later time.
What happened to the suggestions about Wikidata’s gender list? I know I have seen a few suggestions here pointing to officially agreed lists. If MB pushes to that standard then won’t that also help the standard to be taken up by other websites and databases too?
If every database makes up their own lists that just leads to more of a mess \ confusion \ argument.
Yes, we should have a list of choices. But importantly MB should not be setting the contents of that list.
Rename gender to sex and use tags for transgender, etc.
This suggestion would help biologist to get less confused I can tell you.
After reading both threads, this proposal seems effective at representing multiple gender identities in MB; I agree the best approach is a pre-defined list of gender identities with an optional disambiguation free-form text box.
This discussion is important. I agree by offering a gender field in MusicBrainz, it creates space for this topic. When thinking of gender, it should be thought of as fluid and changing like culture. So, building in ways to capture the fluidity of that data is useful. Since MusicBrainz captures data about music, people, and art, consider what these represent of culture. Music is a component of human culture. Labels for artists already exist as their birthplace or current area of residence. Nationality and citizenship are a part of human culture. In the same vein, gender is part of human culture. Over time, culture and perspectives change. To consider the nationality example, the break-up of Yugoslavia into several nations is a change captured in MB already. Gender can also be this way. It is challenging because a gender binary is deeply-rooted in many societies and thus our information systems. However, multiple gender identities won’t go away in 1, 5, or 15 years. If discussion does not happen today, then it will in the future. Capturing accurate data and avoiding a bias that leaves some people out of the documented music history are important considerations.
Thus, I believe the proposal of a pre-defined list with disambiguation box is effective at standardizing gender data to a limited extent, and offering more flexibility where needed. An externally-curated list for gender options could be used initially. I believe this is a good compromise to create a standardization mechanism while also permitting infinite options.
To go a step further, the gender disambiguation could work similarly to genres created from tags. If a disambiguation is used frequently, it could be included in the list. Or be distinguished in another way.
However, this approach does not address the open-ended questions of what happens if an artist changes gender identity after their career starts.
If questioning whether this data is valuable to collect at all, I believe there is value. Gender identity is an interesting social characteristic of music. Interesting areas could be explored by MB gender data in relation to tags, artists, and more. To erase gender is to erase this possible social exploration. Additionally, this metric can be a useful metric of diversity and ways to improve data coverage. If Wikipedia leads a Wiki Women in Red group to improve coverage of women on Wikipedia, a similar campaign could be sustained by knowing this data:
All else aside, I think gender is interesting to capture because for some artists, gender is a defining component of their recording identity and themes in their music.
Adopting a better standard for capturing gender is important, but it should allow for basic standardization of data with flexibility to grow and adapt in the future. I think this proposal captures that.
@jflory This is an excellent comment, thank you. I especially like how you emphasize how both gender and music are important parts of our culture and both are connected. This shows how the discussion about gender in MB is actually relevant and has value in itself. I fully agree with you on what you have written.
hey! i’ve submitted a proposal for a pronoun attribute, if you’re interested: