[tei-council] Sex, gender, and http://purl.org/tei/fr/3601624

[James Cummings]

Hi TEI Council,

A number of people on twitter pointed out that our use of ISO 
5218 for the data.sex datatype.  Melissa Terras has since 
submitted a feature request asking us to re-examine this at 
http://purl.org/tei/fr/3601624.

I won't repeat the whole discussion on the ticket, please go and 
read it there, but simply summarise the basic problem and 
encourage you to read and comment on the ticket itself.

As you may know @sex and tei:sex/@value are a TEI datatype 
'data.sex' which implements ISO 5218 which is a recommendation 
for single-digit representations of sex. This has 0, 1, 2, and 9. 
With male coded as '1' and female coded as '2'.  Some believe 
that this represents a deeply-rooted patriarchal system that 
implies that women are somehow secondary to men. The standard 
itself explicitly states that no such implication is intended and 
it was simply codifying existing practice amongst member states. 
(Countries seemed to use odd/even numbers or 1 / 2.)  The TEI 
moved away from a system of 'm', 'f', 'u, and 'x' used in P4 when 
developing P5 partly because this standard existed (and where ISO 
standards exist for things we are modelling we try to use them if 
appropriate), and partly because values like 'm' and 'f' are 
generally anglo-centric. However, it has been pointed out that 
this might be interpreted as inherent support for this 
patriarchal worldview.  I'm quite confident that the TEI does not 
want to cause offence to any of its users but needs to balance 
that against providing a robust and rigorous scheme.

One solution is to retain our use of ISO 5218 but note on the 
data.sex reference page the offence that may be caused and 
possibly propose alternatives.

One such alternative is the use of @ana to point to one or more 
categories in a local or remote taxonomy, with the benefits of 
being able to categorise multiple vectors (e.g. biological sex vs 
gender vs sexual identity vs allSortsOfOtherThings), but with the 
drawback that it is not a simple international standard. I don't 
believe the TEI itself should propose this taxonomy because I 
feel it may be too culturally-specific.


If you have (appropriate) thoughts on this issue please post them 
on the ticket http://purl.org/tei/fr/3601624.

Many thanks,

-James

-- 
Dr James Cummings, James.Cummings at it.ox.ac.uk
Academic IT Services, University of Oxford