Dangerous editing gender on base first name.

Question

DBE 501 (1500-1699)
I started chancing the gender of some females named William, in my opinion a clear male name, but then I found several females with that name that were married to someone with a real male name. Concluded that William can also be a female name and changed them all back to their previous state. And will stop editing gender based on first name.

Comments

I once addressed a widow as Mrs Bertha Smith and was told in no uncertain terms that since her marriage she'd been Mrs Arthur Smith.

If she appeared here as Mrs Arthur Smith, the gender would be fine, but the 1st name would be wrong.  And the LNAB.

Were they born William?

---

I don't know, there are no (on line) sources mentioned.
And, indeed, I also find some females who bare the name of their husband.
But as I do not know their name at birth I can do nothing about them.
Maybe put a note in the Bio.

---

If it is quite clear that William is her husband's name that is entered as her first name, I would be inclined to change her name to Unknown as she was not born by that name.

Pierre as you say it would be good idea to add a note in the bio and I will also contact the PM before changing it. The PM might be able to find her correct name.

---

I recently did a profile for a man that included many local newspaper articles and a picture of his gravestone included in Find a Grave.  Even his census and son's marriage application listed him as such.  They all referred to him as June.  I finally found an article that made me think that his name was a nickname variant of the suffix Junior, named after his father.  Quite a surprise.  I wondered to myself how often he used his legal name.  Not often.

---

My wife's brother is named Joyce William. Their dad was named Joyce William. His dad was named Joyce William.

---

I believe that Joyce is one of those names that was formerly a man's name but has morphed into a unisex name or a woman's name. Consider Joyce Kilmer, for example, and see http://www.behindthename.com/name/joyce

---

How about Johnnie Cash's song "A Boy Named Sue"?

---

This is one of the error types that I am concerned will produce more errors than it will resolve. It is useful heuristically for sussing out the possibility of something being amiss, but there will be a sizable number of correct names which are identified as possibly erroneous based on gendering. Especially with a database of names as small as WikiTree has. The biasing of the distribution of the names due to the over-representation of US citizens in the database will significantly skew the results.

---

Fixing this error doesn’t generate new errors – it reveals new errors not previously detected.  Many of these Single Sex Marriage errors were caused by file to gedcom conversions years ago which somehow reversed the husband and wife.  By fixing the gender it often creates a new 303 Mother is male or 203 father is female.

Look at this profile of male William Henry Spangenberg (1887 - 1953) married to male Olive Estella (Farr).  When you check their son William Henry Spangenberg, Olive is the father and William is the mother.  By correcting the gender of Olive to female, you create a new error of George’s father being an female.

As far as the name being correct and incorrectly being flagged as an error, it certainly happens.  But it is not going to create more errors than are fixed.  I wouldn’t be surprised if this false positive error rate is less than 1%.  It is up to the “Correction Officers” to be a little careful on any profile where the error isn’t obvious.  If we are overly cautious about the rare case of a false positive error, WikiTree will never get cleaned up.

Answer 1

First Name	Total	Gender Empty	Gender Female	Female (%)	Gender Male	Male (%)
William	398069	6247	1158	0,31	362140	98
Williams	1195	18	350	31,56	741	66,82
Williamson	507	5	88	18,18	391	80,79
Williamina	214	3	199	98,51	0	0
(William)	100	0	4	4,6	83	95,4
Fitzwilliam	40	2	1	2,56	36	92,31
William?	28	1	0	0	24	96
McWilliams	19	1	7	41,18	9	52,94
(William	17	0	3	27,27	8	72,73
William)	16	0	5	35,71	9	64,29
Williamine	13	1	12	92,31	0	0
(William?)	10	0	0	0	9	100
*William	10	0	0	0	9	100

 

This is a distribution of name William and all its variations. So indeed 1158 are female. But I never heard for a woman to use name William. 

There could be also errors with marriages. We have a few thousand single sex marriages in era's when they were not common.

Comments

I prepared a search to analyse any first name.

http://www.sdms.si:92/wikitree/ShowFirstNames.htm

---

You missed ones that named their kids with a variant, like Willie or Will. I have a few of those in my family (Willia Lou is a female, for example).

---

Check link above. It is all there

---

The below Error help pages has been updated with a link to the page where you can search a name and see the distribution mentioned by Aleš above

• Space:DBE 203 
• Space:DBE 501 
• Space:DBE 502 
• Space:DBE 503 
• Space:DBE 504 
• Space:DBE 505 
• Space:DBE 506 
• Space:DBE 507 
• Space:DBE 508 
• Space:DBE 509 
• Space:DBE 510 
• Space:DBE 511 
• Space:DBE 512 

 

Answer 2

Almost invariably in this situation the problem is not that the Gender is wrong, it’s that the Marriage is wrong.  Often it is a woman’s first husband somehow “married” to her second husband.  Or a father “married” to his own son. Or sometimes it is just a completely random “marriage” of two men who didn’t live within a century of each other.  The fix is not to change the Gender of William, but to remove the Marriage.  Usually, a brief google search or search on ancestry.com is enough to reveal the correct relationship between the two men and you can help untangle yet another knot on WikiTree.