Post-Publication Name Change Policies,
Why they Matter,
and Whether they Work

Danica J. Sutherland (she)
or from 2019ish – 2020: D. J. Sutherland (intentionally vague)
or before 2019ish: D           J. Sutherland (he)

Robotics DEI Seminar, University of Michigan – March 23, 2023

Slides available at djsutherland.ml/slides/umich-name-change

(Swipe/arrows to move through slides; m for a menu to jump; ? for more.)
PDF version

HTML version

Name changes

To steal an insight from Robyn Speer (Queer in AI talk @ ACL 2020)

There are only two hard things in computer science:
cache invalidation and naming things. – Phil Karlton


For trans people, naming things (yourself) can be hard

…and getting other people to update the way they refer to / think of you can be harder

How do you change your name?

Ask people to call you something new

How do you legally change your name?

Generally easy: (woman's) last name when you get married

Everything else varies dramatically by jurisdiction

  • In the UK: write on a piece of paper, have a witness sign it
  • In many places (esp. on temporary visas): literally impossible
  • Most US states: go to court, announce in the paper
  • Costs up to thousands of dollars

What happens after you change your name?

I flew here on a passport with the wrong name on it,
but a driver's license with right name in my pocket

Who changes their name?

  • Most trans people: ~80% of trans Americans change or want to change their name [US Trans Survey 2015]
  • ~75% of American women getting married [NYT Upshot 2015]
  • People reclaiming cultural identity, religious conversions, etc
  • People estranged from their father's family
  • People who just don't like their given name
  • …
How I started wanting a name change

(Not actually this tweet)

  Mae

A few hours later, on academia.stackexchange.com

Answer from Rebecca J. Stones (a mathematician):

Next steps

  • (Reading, therapy, talking to people, trying new things, …)
  • Iterated through a bunch of possible names
    • Sorry to about seven different possible future Sutherlands who might want a gmail address!
  • Eventually settled(ish) on one that keeps my initials
    • Explicitly because I wanted citations
      to “D. J. Sutherland” to stay correct
  • Started publishing + giving talks as "D.J." in the interim

Now, what to do about the old papers?

How it used to be (1960s)

Lynn Conway:

How it used to be (2010s)

[Rebecca J. Stones on academia.stackexchange, 2015]

How it used to be (2013)

publicationethics.org/case/change-authors-name-after-publication

How it used to be (2020)

[Article in Science by Katie Langin, Feb 2021]

A non-solution

“Just start using the new name + note on your website/ORCID/…”

  • Have to disclose the name change constantly, forever
  • Readers won't realize you and “old you” are same person
  • Will still get cited under the old name, forever

“But the sanctity of the published record!”

A non-solution

“As an aside, I believe you should not rebrand your older papers with your new first name. This confuses all the database searches. I am afraid that your best bet is to stick to your initials only.

“Similar things happen to divorced women. For instance, [someone] was essentially forced to keep the name of her ex-husband […] decades after her divorce.

“I am sure many people told you so, so…”

What happened

What happened

Before me:

  • ACM adopts an early policy, after much resistance
  • Name Change Policy Working Group
  • Committee on Publication Ethics (COPE) forms a working group in 2019
    • Publishes an update in January 2021
    • "Anticipate sharing the initial guidance document…within the next few months"🦗🦗🦗

What happened

Lots of arguing with lots of different publishers

Some formal policies, some making changes without one

What happened: Formal policies

An example: ACM

  • One of the first actual policies
    • Took lots of convincing, by Tess Tanenbaum + others
  • Slow and messy implementation
  • My request (one measly paper) took five months
  • Others have waited a year or more
“Apparently a few months ago the folks in publishing made a snazzy presentation to the board indicating that waiting times were going down and the list had started shrinking. It appears to have achieved this by noticing that there were a series of single-paper requests in a row, and selecting exactly that range to base the presentation on.”

Policies aren't enough

“As it turns out, when no one monitors the mailboxes you send your name change requests to, the policy is as good as nonexistent.”

“I have written to plosone at plos dot org and privacy at plos dot org in May 2022, August 2022 and October 2022, requesting to have my name updated on a publication I have in PLOS One. I've received no response aside from one automated acknowledgement email, and my name has not been updated.”

Bureaucratic violence – Vagrant Gautam, Feb 2023

An example: IEEE

…but then they actually instituted a policy! (Sep 2021)

and it's a pretty good one, implemented reasonably well

An example: OpenReview (ICLR)

  • For me: after lots of back and forth, got everything fixed
  • Wanted approval from program chairs
    • For me, got it before even telling me they were asking
    • What if no PC replies? (Maybe all have retired…)
    • What if a PC is just transphobic?
    • Is the publisher okay with PCs deciding to admit a paper with [a cis-presenting author] but not an identical paper with [a trans-presenting author]?
  • OpenReview no longer requires PC permission

An example: arXiv

  • I made new versions of all my papers with the name fixed
  • You don't have to do this anymore:
    • Invisible retroactive name changes for the most important places
    • As of last year, editing citations/etc was officially against policy (but the person I was talking to let me do it)
    • Submit new versions if you need to

An example: AAAI

  • No public policy
  • They will change names, but implementation is bad
  • Months and months of ghosting me before I complained elsewhere in the org and my one paper (where I sent them a fixed PDF) finally got fixed

Generally

Dozens of publishers (for thousands of venues) tracked here

The big missing things in general:

  • updating citations
  • prompt implementation

What happens after?

…how many authors actually get their citation info from the publisher page?

Semantic Scholar

  • Send an email: all papers fixed regardless of publisher

DBLP

  • Send an email (or in my case no email?), can get all papers immediately fixed regardless of publisher

Google Scholar scholar.hasfailed.us

  • There's no one to email
  • You can change your own profile page, but it's not obvious this doesn't affect results from searching for the paper
  • After name is fixed at publisher, remains wrong at Google Scholar until they invalidate their cache (months later)
    • Recently added fast-tracked (5-10 days) updates for name changes, for some publishers

People's .bib files they reuse between projects

  • Often remain wrong
  • I've corrected bibs from multiple coauthors in paper drafts
  • rebiber pulls DBLP entries for conferences it knows about
    • but design means rebiber itself can easily be out of date
  • ACL camera-ready checks now warn for outdated names

Do people actually cite me by the right name?

Semantic Scholar API => arXiv downloads => pdfgrep:
some errors in processing, 1/3 of citing papers not on arXiv

Do people fix their bibliographies?

Seven papers corrected my name in later arXiv versions
(two of them because I emailed and asked)

Compare: >100 papers deadnaming me after arXiv was fixed,
~100 after corresponding official versions were fixed

Do people look at their own bibliographies?

Ten (different) papers cite me by both correct and deadname in the same version

Takeaway, if you're changing your name:

  • You absolutely can!
  • Once Google Scholar gets the updates,
    most people will start citing you correctly
  • Old citations won't change unless you really work at it
  • It's still a lot more work than it should be
  • Ask for help dealing with publishers if you need it
    • (Me, on the Queer in AI Slack, the NCPWG, …)

Takeaway, if you're writing papers: