map gender to sex, not sex role

VladimirAlexiev commented 7 years ago

Gender is mapped to 300055147 sex role. However, that concept has no children and is a bit unclear. In https://jira.getty.edu/browse/ITSLOD-497 I suggest to delete it, since there's already:

300055146 sex, with children male, female
300400503 sexual orientation, with children: LGBT, bisexual, heterosexual etc

I think AAC should map to 300055146 sex

workergnome commented 7 years ago

There's a long discussion with @cbutcosk about this at https://github.com/american-art/aac_mappings/issues/47.

Basically, "sex" is about biology, and unless the museum is determining gender using DNA tests, it's not really correct to use it.

"Sexual orientation" is about whom they wanted to sleep with, and that's also not what museums describe.

"sex role" is about the gender that they appeared/performed, which is the most relevant characteristic for museums.

azaroth42 commented 7 years ago

Either way it would be a discussion for the future. Can we tag as 1.1?

workergnome commented 7 years ago

Yup. Done.

VladimirAlexiev commented 7 years ago

300055147 doesn't have any child concepts. (Patricia answered: "sex role" does not require children – it is a concept useful on its own.) So how are you sure it would be correct to map eg the value "male" to a property typed with 300055147?

VladimirAlexiev commented 7 years ago

@workergnome Please reopen for "version 2'.

unless the museum is determining gender using DNA tests

This comment is disingenuous. I know you're a man without testing your DNA. Since 300055147 doesn't have any children, why do you think it's correct to use it for male/female or man/woman?

workergnome commented 7 years ago

I disagree with you, but it's a long conversation, so let's table it for now. I also agree that we should keep open for discussion at v.2.

VladimirAlexiev commented 7 years ago

According to the mapping, it should be like this (eg from https://github.com/american-art/GM/issues/18):

<constituent/whoever> crm:P2_has_type <constituent/whoever/gender>.
<constituent/whoever/gender> a crm:E55_Type;
  crm:P2_has_type <thesaurus/gender>.
<thesaurus/gender> a crm:E55_Type;
  skos:broadMatch aat:300055147; # sex role
  skos:prefLabel "gender".

Allowing museums to have their own notions of what is Gender strikes me as a crazy idea.

workergnome commented 7 years ago

At least here in the US, prescriptive requirements for gender are a major political issue—see #47, as well as https://medium.com/@malpinder/falsehoods-programmers-believe-about-gender-cf1a55085ab2 for an example of the sort of critique strictly defined gender roles will engender.

VladimirAlexiev commented 7 years ago

Allowing museums to use any set of Gender values is fine: they could use one of the 15ish in Wikidata if they want, see http://vladimiralexiev.github.io/pres/20150212-sex-or-gender/#sec-2-5.

Allowing them to have their own notions just doesn't match any current business requirement. Pray tell, what do they have in their data? You mean, like, fa'afafine (Q1399232), māhū (Q3277905), kathoey (Q746411), fakaleiti (Q350374), hijra (Q660882) etc etc is just not enough?

designforcontext / aac_review_tool

map gender to sex, not sex role #50