Tag Archives: research

Version control: In which Facebook makes improvements and my work suffers

In a paper I presented last year I discussed how hiding your sex/gender from your public Facebook profile didn’t actually remove the gendered language the system uses to refer to you. So, when leaving the box unselected, like so

'Hiding' sex status on Facebook

your public profile would still use sentences such as, “If you know Anne, add her as a friend or send her a message” (emphasis mine).

Facebook's gendered language

I see this as a concern for various reasons I’ve discussed elsewhere and I won’t go into them now. (Google+ also had a similar practice during its testing phase, which it changed fairly quickly after receiving a lot of user feedback. Frances Haugen from Google talks about the situation here, for anyone wanting a refresher.)

I mention this again this week because I’m in the process of finalising a paper which reviews gender and sex usage in social media systems and discusses potential social effects leading from these system design choices. It was going well and I had a great structure that appeared to work, but yesterday I went to check out Facebook’s system to confirm my statements and noticed that they appear to have addressed the above issue. Now, when looking at a person’s public profile, regardless of whether they’ve chosen to hide their declared gender/sex status, you see shorter sentences that avoid pronoun usage altogether.

Facebook system avoiding gendered pronouns

I think this is a great move! Eventually I’ll make time to look into other instances of this and evaluate how easily other systems can introduce similar changes (it’s difficult because I’m not that familiar with other languages/cultures). But for now I have to finish writing my paper. And this improvement throws a spanner into the works. (Or, rather, removes one that I was hoping to talk about a bit.)

I can work around it. I just need to do some work on changing the way I’ve structured everything. But it does raise one major difficulty I’ve been having throughout my research; that of being unable to easily record how the systems I discuss change over time when they are ‘closed’ systems.

Last year when Google Profiles allowed users to hide their gender status I kicked myself because I didn’t think to take a screenshot (a documented first-hand account that may constitute a better academic reference than a random person’s blog post) before it was changed – honestly, I didn’t think they’d fix it! I could keep taking regular screenshots of various interfaces, but that’s time consuming when I don’t actually know what I may want to focus on later. It’s what changes that becomes interesting, and I don’t have forewarning. In this case with Facebook removing gendered pronouns I actually have screenshots; what I don’t have is an idea about when, exactly, within the twelve months between my two screenshots these changes were actually made.

Couple this with the fact that some users see different iterations of a system interface, depending on their location or the server they’re using, documenting something accurately for later use can be quite difficult.

And this is one important (though small) reason I love open source software: I can find out which version is running on a particular server and then look through the code personally to document it accurately. This allows one to take a historical look at these services.

With all their talk about ‘transparency’, I wish Facebook and Google+ would take this direction. Just for me.

It would make my work a little easier.