Do we need to correct William the Conqueror's siblings?

+19 votes
EuroAristo Members,

In a separate discussion about accuracy on WikiTree, RJ Horace wrote: "OK, start with William the Conqueror.  3 of his 6 siblings aren't his, and his Conteville step-family is [absurd] with little resemblance to anything known to history."

RJ, can you cite your source for this?

John, Darlene, and EuroAristo members, is there something that needs to be fixed? I'll readily admit that I don't the first thing about any of this. I leave it to the experts.


WikiTree profile: William de Normandie
in Genealogy Help by Chris Whitten G2G Astronaut (1.5m points)
retagged by Darlene Athey-Hill

Thanks Chris, I just brought this to the attention of the new Medieval Project

I added a couple of more tags to bring this to the attention of more project members.

7 Answers

+19 votes
Best answer

I've started to investigate the profile for Felicia, and hopefully can write that up by the end of the day.

I suspect that Rohesia de Conteville, given she has the other last name of FitzHenry, may be a mistaken version of Rohese FitzHerbert who was at one time thought to be an illegitimate daughter of Henry I.

If anyone is working on any of the other profiles or wants to write up Rohese, please post here, so we are not all trying to edit the same profiles.

by John Atkinson G2G6 Pilot (641k points)
selected by Chris Whitten
I've finally finished writing up the profile of Felicia and I've detached her from Robert of Normandy as her father.
+20 votes
Absolutely. His unwed parents also had a daughter, Adeliza, with an interesting history of her own. By her husband Herleve was mother of Robert and Odo. The others are myth, difficult to prove that, but I've yet to see a reliable source for any of them. As a quick thought Weir does not name them. If we are to be taken seriously they should be dealt with ASAP.
by C. Mackinnon G2G6 Pilot (346k points)
+20 votes
I'd be interested to see the rest of the discussion, but at first sight RJ has as usual made a good point which is very much thinking about what is really important to improve in Wikitree.

I think we could benefit greatly if we could recruit more people with the skills needed in pre-1500. I also know by trying, that many such people really are interested enough to have an occasional look. I believe in the potential.

But at the moment it is hard to argue that Wikitree has much potential, because any quick browse will find obvious problems. In some ways pre-1500 Wikitree is still worse than some of the old gedcoms it merged into itself.

One of the problems is the bias against deletion. The old gedcom merges created all kinds of things which need deletion. Profiles grew and got duplicated, and they stay that way a long time.

Anyone who works in pre-1500 will know that it is approximately 1 million times more pleasant and easy to build a family up who is not yet in Wikitree. Any family who was popular in all the old gedcoms is typically going to be stuck in limbo, and very difficult and unpleasant to work on.

What can we do about that?
by Andrew Lancaster G2G6 Pilot (146k points)
Hi Andrew. Maybe you could spin this off into a new post?

Yes, perhaps I went off topic and perhaps you wanted specific remarks about that family? I thought I'd make a comment about the general type of problem. Actually I mention such concerns quite often. But my apologies if I've gone off topic.

Concerning the example case, to me it is obvious that even the most famous monarchs in wikitree normally have a circus full of extras in their family. Semi-fictional siblings of royalty were very popular in the old gedcoms. ...But to me the bigger problem is the question of why it stays that way, despite so much effort.

I guess my challenge to all of us in this great big project of Wikitree is how we can make our pre-1700 parts NOT look like the result of a massive gedcom merge.

I am using pre-1700 as shorthand for the part of Wikitree where the most gedcoms originally overlapped. This part of Wikitree faces qualitatively different challenges than other parts - whether they are medieval or not.

After a frustrating session today though, I'm honestly not sure I feel like trying to start a new thread on that. It would not be the first time, and on G2G such posts generally get hit with a round of responses from people who don't work on the same parts of Wikitree, and think I am being a heretic 

Hi Andrew. Yes, I would consider this a tangent. See Discussion Rule #1. I can understand not wanting to rehash it if you have have discussed it quite often in the past, but in that case also see rule #5. :-)  If you do want to consider a proposal, restricting it to something pre-1700 or pre-1500 could be quite sensible. You don't need to invite responses from people who work on later profiles.

OK. Sounds like your suggestion to post something wasn't quite what it seems. Misunderstandings. Just to be clear, I made no proposal, and it does not sound like it would be wise to make any.

I made some very short and casual observations, and based upon the responses, I do hope you will read those again. I think there is a serious under-estimation of how terribly poor the quality of medieval profiles are on Wikitree, and how this is something hard to change.

There are some excellent people working hard on medieval profiles, but the system is working against the good ones, and it works harder against them the harder they work to try to make Wikitree better. They are working on people who are ancestors to most Wikitreers, and because of their work, more Wikitreers will find those connections. This is what most genealogists really want.

No one cares though, because there is definitely an over-confidence around, and misunderstandings about what the situation really is. We have one of those "Mission Accomplished" problems.

I say all this with absolutely positive and hopeful intentions. 

Let's be clear. Obviously I am trying to get your attention. I hope you will see that my intentions are good.

+19 votes

What is common with early pre-1500 genealogy is that there are what I think of as different levels.

There are some relationships we are sure about, there is definite confirming primary evidence - with William the Conqueror's siblings, we are sure about a sister, Adelaide on his father's side, and two half-brothers on his mother's side - Robert, Count of Mortain, and Odo, Bishop of Bayeux

Then the next level down there are relationships that seem likely, there is perhaps some primary evidence, but perhaps not as much detail - so William on his mother's side may have had one or two half-sisters.  There is also possibly a Conteville step-brother.

Finally there is a third level of relationships that are highly unlikely.  Evidence, if there is any, often comes from much later sources and these are often the relationships we unfortunately find proliferated on online family trees.  I'd put Felicia and Rohesia in this group, plus some of the Conteville siblings.

The issue is that documenting all of this takes time and as RJ and Andrew have stated William the Conqueror isn't an isolated case.

by John Atkinson G2G6 Pilot (641k points)
Yes I'd say there is a relatively small number of us who have the knowledge and enough energy (in bursts at least) to really carefully review such cases, and then actually do something about. In the case of royal families, the work can be particularly unrewarding. I say this as one of the people who put in a big effort on the next generations of William's tree for example.

I would love to be able to attract more people with the necessary skills, but the more systemic problem here is that actually Wikitree as a community is not exactly warm and welcoming to this very necessary type of work.

In fact, it is also not really welcoming to anyone who wants to discuss such things, or discuss whether we can discuss such things. You are not supposed to say things can be done better, because Wikitree does not like deleting or criticizing the work of anyone.

So it is MUCH easier to add to Wikitree than to prune it.

The further back you go, the bigger the impact of this problem is.
John and Andrew, first, let me say thank you for your invaluable work on Medieval profiles. Very few of us can do it.

Are those half-sisters marked as uncertain?

I see that Normandy-215 does have the {{Uncertain Existence}} Research Note Box. Does Conteville-61 need it? And do they both need a few words of explanation in a Research Notes section?

Andrew, I hope you'll consider proposing changes to or other policies if you think they are warranted.

Chris, I don't know about posting proposals, those rules keep changing, and clearly they are not welcoming to heretics. Anyway, short version:

Surely, "uncertain" tags are for cases which are likely to be possible, and not for every proposal ever made. 

Background. One practical reason to follow this approach is that the second population of proposals is many times greater than the first. We do really look at many many proposals, and it takes a LOT of time to filter them, but we simply can't be making it our main job to list them all as if they were equally serious. They just aren't. Please, please, please believe this.

Philosophical. If you would really want all proposals in, then we'd need more categories of uncertainty. Basically, we'd need a "junk" category. And that would raise the "what on earth are we doing?" questions. 

Executive summary: It would mean we can't work on real genealogy anymore, and Wikitree would die because all resources are focused on things which are in conflict with real genealogy. 

My opinion. Such approaches accelerates the problems which threaten this community's future as a real genealogical project, and can move us down the list from Geni B to Geni C, instead of letting us achieve our real potential as something new and better. We have to remember what we are doing here.

I'll second this -  "uncertain" tags are for cases which are likely to be possible, and not for every proposal ever made. 

Quite frankly, some profiles are imported by what I call "wishful thinkers", connecting them to important historical persons, gods, or anything in between. 

For these "internet lore" profiles, you do not need the skills of genealogists to sort them out, you need the skills of historians for that specific time period and location to sort out if they even existed.

That is why it is so overwhelming to work on such subjects as this question. Who is a real person and who is not? Are they related at all? What should be considered reliable sources in this case? How much time am I willing to spend on sorting it out? 

So I'll "third" it:  "uncertain" tags are for cases which are likely to be possible, and not for every proposal ever made.

I have done preparatory work for disconnecting a fabricated line of descent from King Gustav Vasa. The connection was originally created by forgery in the first decades of the 1800s, by a nobleman who desired Vasa blood he didn't have. Disbelief was published a few decades after his death. The work was thoroughly discredited before 1900 - this nobleman was also revealed as a forger of many other historical documents in his collection.

Yet this line has been entered in Wikitree, back in 2014. The interesting thing is that the invented wife/mistress of a real Vasa descendant had her profile created with the suffix "Fictional", so the creator was clearly aware of making a profile for a non-existant person. Rules may have been different in 2014.

Anyway, "Uncertain" does not fit the situation. Fortunately I don't do this type of work very often, so I will be able to keep watch over this case until it has been satisfactorily resolved through all stages of the process.

I apologize for not being able to contribute on William the Conqueror.

Thanks Maggie. I sometimes feel very alone on such points. I think most of us who work on medieval profiles have similar concerns, but there seems to be a lot of reluctance to say it clearly in the public discussions. 

The enormous amount of very low quality material available on the internet for these generations presents a critical challenge to Wikitree's open arms approach to gedcom-swapping genealogy. We just can not be "nice" to every theory out there. We can't even mention them all.

The reluctance to talk about such problems seems to lead to an overconfidence about our medieval section? That seems to influence Wikitree's decision-makers to push things in the wrong directions. They apparently assume the the old gedcom merges went great and are just being tidied up a bit. 

So, everything is set-up to make it difficult for us to fix things. We still have powerful profile owners, not only pre-1700 but even pre-1200. If any American family thinks they are descended from them, it can take months of haggling to fix up a small family.

Our rules don't keep changing. They are slowly evolved through careful deliberation. If you want to propose a change to see

Hi Chris, some rules change then, but in this particular case what I am actually saying is that the medieval area needs more evolution, and needs to be more different than it already is from the way we handle recent generations. 

I guess the typical concerns of people working in the medieval area are pretty well-known to people in those policy discussions, wherever they are, and have been often discussed. The medieval world is relatively small and we know we tend to agree on some things. 

An obvious one is that profile managers can't really do anything positive in this area, but they can do negative things, and they do. Why do we still have them? (Many of my colleagues will more cautiously propose that at least they should have pre-1500 certification.)

Yes, I know there are various ways to propose this, and I'm sure there'll be more discussions in other places over coming weeks, months and years. But as I understand it, this is not a new topic, and it is known to be one where the people working in that area tend to have similar thinking that Wikitree's higher echelons do not agree with. 

It is very hard to put a positive spin on what the reason must be, because the only possible thing they can be thinking is that someone might question the weak points in a gedcom that connects to them, and they want to maintain the power to make our lives miserable if we try to do that.

Anyway, that is the main effect in practice. There is no positive effect of the current system of having unqualified profile managers who can "own" medieval people and make the lives miserable of editors working properly on medieval profiles. 

...and that leads to the problems being discussed in this thread. The problem is that we can not have Chris Whitten and John Atkinson spring personally into every other such case, because there are thousands upon thousands.

+14 votes
Why not take the easy way first like the Early Scandinavia Project has done. All uncertain connections are removed and said connections discussed in the bio and/or resarch notes on both the master profile and the disconnected profile.

That way we would get a more accurate genealogy with the possibility of reconnecting relationships that proof are found for.
by Juha Soini G2G6 Pilot (122k points)
If you really remove ALL uncertain connections, this would mean breaking a very large part of our medieval tree into individuals with no connections, or groups of twos and threes.

But perhaps you are already unconsciously defining "uncertain" in a way which excludes minor uncertainty.

It is very difficult to find cases of complete certainty. We can't do DNA tests on everyone.
I count totally unsourced in medieval genealogy as uncertain. Most of the rest would be plausible until you get an occasional document naming multiple generations.
We should be aiming for one consistent policy on this, and currently it is
Chris, perhaps we might have need of something more besides "uncertain" for those pre-1500 which are often gedcom imported profiles from WikiTrees early days. Some of us who work these profiles might not be able to post a proposal because we do not know what could be a workable solution.

Chris that particular policy page does not appear to be immediately applicable to this case where we are talking about relationships themselves being marked as uncertain (or alternatively being simple broken off).

I offer my words in bold, above, as a potential wording for a future guideline. 

1. "Uncertain" tags are for cases which are likely to be possible, and not for every proposal ever made.

As a secondary part to the guideline I would suggest:

2. If there are several relationships which are likely to be possible, then an uncertain connection is appropriate in cases where one of those alternatives is a "clearly leading hypothesis" compared to the others. Otherwise, no link should normally be made, and all the likely links should be discussed within the profile body.

3. A possible reason for allowing an uncertain link when there is no leading hypothesis would be when the removal all links will completely disconnect two families who are known to be linked. Such cases should normally be discussed on G2G or some larger forum.

Maggie, I think it is very easy to post a proposal for any Wikitreer? I wish more people would do it. I know some people make a big contribution to pre-1500 without even having certification.

Only remark I would make is that we've been seeing comments on profiles getting deleted more quickly and thus sourcing and other types of advice are being removed before they are acted upon.

Juha, "I count totally unsourced in medieval genealogy as uncertain"

I might be seen as a purist by some people but I would be careful about this. For better or worse, on Wikitree, "best practice" in these periods is to first at least try to understand where an idea came from before we start making decisions that no one is likely to look at again for a long time.

...Believe me, I know very well that this task is MUCH HARDER than actually just making a fresh good profile in the cases where you have good sources at hand. At least I can say that if you know your field well, there are a lot of typical places you can look to check if unsourced information comes from there. Google helps a lot too, because often these unsourced ideas come from websites, and the websites themselves do often explain their sources and reasoning better than we do, sadly.

Very occasionally I DO find that unsourced information actually came from a new article or book or SGM discussion by a well-respected medieval genealogist. Those will normally come up on google if you are tenacious enough. 

Of course at some point you are going to need to make a decision. But it is best to do this when you have already got yourself a good perspective on what kinds of ideas are out there, including the silly ones. 

...By doing this, you can also write up the profile in such a way that you warn future editors about why certain other ideas are not correct or likely. This can save you problems in the future.

Andrew, naturally you need to look for eventual sources first, but if the only thing you find is an occasional unsourced family tree floating around the internet, then the profile is unsourced, though you cannot be certain the family ties in question did not come from some little known publication not (yet) digitalized or available online. That is why IMO the unsourced medieval family connections should be counted as uncertain and not discarded totally.
Yes, at a certain point we have to decide whether we can see what's going on. And I'm sure everyone draws that line differently, but we should make some kind of effort first. You'd be surprised how often the results of that effort are interesting and fast.

On the other hand, back to the topic, I think if you discover there is no source marking a link as uncertain is one of the options I would NOT normally consider.

Uncertain means "uncertain". As discussed above it often means "quite likely". Generally I would say this designation is used when you've looked at the sourcing and decided how conclusive it is, not when you did not yet find it or get around to finding it.

By using it to mean that it is currently unsourced, I fear you are innovating in a way that others are not going to understand. I think there are other better ways to mark something as unsourced.

We can't have our own private ways of working on this wiki.
I'm sorry Andrew, but I can't understand what you are talking about, and it seems to me that you do not get my point. I don't know if it is my language skills or your academic mind but we do not seem to meet on even terms here. Let's agree not to discuss this further between us.
Juha, no I don't think so, because this is a basic rule of the community. Uncertain actually means uncertain. We signal that something is waiting for sources in OTHER ways.

Please don't make up your own systems.

Andrew, I should perhaps add that Early Scandinavia works on profiles before 1200. Main sources for most profiles before 900 is the Icelandic sagas written in the early 1200's.

The profiles vary from historical proven persons such as Harald Bluetooth to mythological ones, Fornjot for example was an ancient being, a giant/jotun, whose three sons represented nature phenomena and ruled the sea, wind and fire.

I don't really understand why Wikitree has profiles for mythical beings. I fear that is way off-mission (to the extent we have a clear mission), and things like that seem to be effecting the rest of Wikitree, and not only because of the negative impression it must create of the quality of work we do overall.

If more and more people use Wikitree for writing non-genealogical articles, doesn't this risk making us a Wikipedia "b" for the people who want to post stuff that would be deleted from Wikipedia? I guess I now understand Juha's sarcastic remarks about Wikipedia editors.

In genealogical terms, obviously fully mythical people tend to have multiple family trees and so I can imagine you can't make really solid trees, and such profiles will basically just be stand alone articles?

In any case, here Juha is clearly advising people about how to handle uncertain genealogical relationships in non-mythical parts of Wikitree. But from what you say, Juha is not working on normal genealogy? Discussions like the ones the rest of us are concerned with, about which evidence is strongest, would be irrelevant to his work?

So overall I am uncertain about what you are saying. I guess you are saying there is a part of Wikitree which works differently, or something like that. I think that would require another discussion about what Wikitrees mission is, and where the limits should be. I thought the Wiki was for genealogy. I work on Wikipedia itself for non-genealogy, and don't need a "b" version.

On the other hand, I think the point I made to Juha is very simple and straightforward, and was clearly not understood by Juha or me to be about any special isolated part of Wikitree. It was about how to handle the genealogical relationships of real people.

Marking relationships as uncertain is not normally the way that we signal that someone should work on finding sources. We would normally signal that in different ways.

Was my description wrong, do you think?
+13 votes

I did a ton of research and posted it in a long email to the Medieval Project email group.   Because I included materials for discussion that can't be included in G2G for a variety of reasons.  

Multiple vetted sources for this time period agreed with Adelaide, Robert and Odo. also seen as Eudes.  Robert and Odo are half brothers.   

Vetted sources do not always agree and there are some guesses made by even acclaimed historians that can muddy the waters.  Many of the hints come from single lines in old charters written in Latin that can have more than one meaning when translated.  So rather than clog up G2G with all of the screen prints and foreign language straight copy of text and then translations I instead sent it to our smaller group for discussion / editing down to a workable level for G2G.

Sources I used included:,%20Kings%201066-1603.htm#_Toc359663361   and his siblings listed under his father  at


 Henry Project by Steven Baldwin.   

Transferred to website of the American Society of Genealogists, 1 July 2020

Canonical URL:  he shows family as:  He shows family as

So where are the other 3 Felicia, Emma and Rohesia coming from?  There is another Herleve or Herleva and perhaps they have been conflated?  
Rohesia is shown as daughter of  with no father shown  source on this profile is only an ancestry tree.   I would vote to disconnect...  
Emma is shown as Daughter of  [uncertain] and  [uncertain]
And they disconnected her from her conjectured son which is what connected her to William as a sibling
I had more on each of those 3 that cast doubt on them in the larger email.   This is just a synopsis of where I looked and what I found...
by Laura Bozzay G2G6 Pilot (855k points)
Just wanted to add that all of those sources cited their sources which I included in the larger email.   And I also compared what Wikipedia has because that is where a lot of the internet trees are getting their info so I wanted to see what it said compared to what the other sources had and what sources it was using.  

Felicia is a huge can of worms when you look to see all the variations out on the internet for her.   Shudder....
Thank you for your hard work on this, Laura!
+7 votes
Adelaide of Normandy was his sister

Robert of Mauritania

Odo of Bayeux.
by Lisa Taylor's G2G3 (3.1k points)

Related questions

+7 votes
1 answer
501 views asked Jan 31, 2017 in Genealogy Help by Living Horace G2G6 Pilot (649k points)
+9 votes
3 answers
+100 votes
104 answers
+3 votes
2 answers
508 views asked Oct 1, 2013 in Genealogy Help by Roger Travis G2G6 Mach 2 (27.8k points)
+15 votes
4 answers
1.1k views asked Jan 15, 2016 in Genealogy Help by Jack Day G2G6 Pilot (478k points)
+11 votes
4 answers
5.4k views asked Apr 10, 2013 in Policy and Style by Vic Watt G2G6 Pilot (366k points)

WikiTree  ~  About  ~  Help Help  ~  Search Person Search  ~  Surname:

disclaimer - terms - copyright