Take out the garbage: Delete all unsourced pre-1400 profiles?

Question

It seems to me that hanging on to unsourced medieval profiles does more harm than good.  These profiles often have bogus connections to imaginary parents or children, and this makes WikiTree look bad.  On the other hand, if an unsourced pre-1400 profile of someone who actually existed gets deleted, it's not a big effort to create another profile, and this time around, the profile will have a proper source, and it will by done by someone with a pre-1500 badge, as opposed to the current accretion of gedcom clutter from a bygone era.

Perhaps it would be better to take baby steps toward a purge of pre-1400 unsourced profiles, to see how it works in practice.  For example, start out by deleting all unsourced pre-1200 singletons.  I imagine that it wouldn't be too difficult to create a bot that identifies all pre-1200 singletons, and then a crew (perhaps a weekend challenge) could go through them to double-check for any profiles that might be worth saving, and then merge all the garbage profiles into the profile of Gustave Anjou.

And after the pre-1200 singletons are taken care of, then all pre-1200 profiles that are part of an isolated chain with no more than three or four links.  (for example, son-father-grandfather, or two parents and a child, all un-attached to any other profile).

And then, perhaps getting more ambitious and just deleting all pre-1200 profiles that have the {{unsourced}} tag.  Of course, a message could be generated to warn profile managers beforehand that these profiles will disappear in a week or so unless somebody adds a source.

This post is just brainstorming, and maybe others will have more practical suggestions on how to best "take out the garbage."

Comments

I can understand the concern, but what one sees as "garbage" another can see as needful.  I believe it would be better to develop a Badge or standard statement that stands out at the top of the profile that states this record's validity is "Suspect" and consider establishing a "Caution Rate" that goes with it of 1 to 5, with 5 being the highest caution, with the least amount of support info or ties to back it up.  Perhaps with a roll-over checklist that would indicate why it's suspect too(?)  Or perhaps someone can come up with a similar but better plan that would allow us to keep the records but adequately (yet simply) communicate why the record is suspect.

---

I totally agree with Evelyn.   I was just going to ask about B.C. profiles when I see this comment which would dump a lot of my searches.

I was thinking the same as Evelyn, what is one persons Trash is another person's TREASURE.   I'll keep my treasures.  I don't really have to share them or my sources.

---

I like the Suspect badge idea.  There are some post 1400 profiles that are suspect in my mind.  I like it when people flag the profile - then I know that more research is needed before I connect that person to someone in my tree.

---

A tagging system which registered specific types of concern would be better than one which registered a general concern. This could help create a centralized list for people interested in (and/or concerned about) specific types of problem.

---

We've got Uncertain flags and Uncertain templates.  They're underused because they're like downvoting.  People take them personally.

People don't understand that they're awash in an ocean of junk and nobody's blaming them for it.

---

Maybe this misunderstanding would be less if the tags were more clearly about some specific issue needing checking. At least sometimes.

(I have to say, I don't worry too much about people taking things personally in pre 1500. Harder job is finding anyone willing to talk about most profiles.)

---

RJ - I would be willing to use those on my own profiles I create.  I usually put in the Bio what information I think is unsubstantiated or questionable.  I'll check those out!

Answer 1

As someone who spends a lot of time on pre-1500 profiles I would object to this strategy -- while agreeing completely that the continued existence of these profiles causes problems for WikiTree credibility.  

A number of these profiles are "sourceable"  -- all they require is the research that should be done anyway.  

Some of the efforts to clean them up make sourcing them more difficult.  Some unsourced profiles have been completely stripped of everything there by well meaning people who thought what was there wasn't good enough.  In fact they were right -- it wasn't good enough -- but it contained clues to help source the person.  Fortunately one can look back through changes to get back before the damage was done.

Another problem is people who are too eager to merge.  They identify that C1 is a duplicate of C2 and should be merged.  They are correct, except that C1 has a different father than C2.  If you merge them without researching the parentage, you end up with one of the fathers hanging out there not connected to anything --   which is worse than you started out.

And then there is a complicating factor:  I am assuming that all proifles currently on WikiTree are an invitation to further research and editing.  I don't create new profiles that are weakly sourced, but if I encounter an existing profile for someone who was born in 800 and has no sources, and discover that there is a Geni entry for that person, I may add information from Geni to the profile.  I generally don't remove an Unsourced label that is on that profile because it's not a good source.  But it's added more information and that may help the next person find something that does qualify to remove the unsourced label.  This is, after all, a wiki.  Nobody here is in the business of creating a finished product.

A major problem is profile managers who are not pre-1500 certified and don't intend to become pre-1500 certified.  I recently adopted ALL the orphaned pre-500 profiles -- including a lot of junk ones that might be subject for deletion -- to keep them from being adopted by someone who is not qualified to do anything with them.

At the moment, when people have added the "unsourced" designation to a profile, they have usually intended that as a flag to encourage sourcing the profile.  If they understood that placing that designation on a profile was to flag it for deletion, they would probably choose differently which profiles to tag and which not to.  So if there was an intent to delete a specific group of profiles,it really should require placing some new designation like "Delete this profile" on it.  Otherwise people would actually feel betrayed that a designation they employed for one purpose got used for a purpose they did not intend.

Comments

The profiles entered by many folks prove that A LOT of work went into A LOT of people over A LOT of years and just dumping them would be a true slap in the face of all those who dedicated their lives to working to help others.

I've been working for 10 years, my husband has worked for 40 years on branches of our trees that have "research worthy" names and to think of dumping them would be an absolutely travesty. I'm sure we all started out with such excitement that we used the first available source then found more reliable ones and have so many that getting them all fixed at once is ludicrous.

I can personally say I started by using Ancestry but learned FAST that it was the LEAST reliable and bailed out of the site after only one year of reading wrong profile after wrong after multiple duplicates of the wrong profiles.  But rather than fixing them, they just perpetuate them and then let themselves be hacked. 

I've learned and continue to learn. I feel that suggesting edits and not getting ignored or receiving condescending comments on questions or suggestions would be a much better pathway to go. 

I've also never heard any negative comments on the reputation of wikitree from any source other than within wikitree... So who are the ones saying wikitree has a bad reputation.   I can't say enough GOOD things to anyone who asks.. but will never recommend Ancestry or My Heritage to anyone.   Ancestry because of unreliable/duplicate data and My Heritage for their bait and switch billing methods.   So far I've been cheated out of over $100 of fees paid because they change billing mid year and cheat me out of the remainder of the year.   I'll never sign up with them again.

 

---

If people have put years of work into knowing about a historical person, then they will have sources?

---

@Rebecca
I don't really understand how an unsourced profile can be said to represent a lot of work Have a look at the unsourced list for the 13th and 14th C.There are many examples of profiles added in the early days of wikitree which will be unsourceable .Heres one example
https://www.wikitree.com/wiki/Wells-41
 A 13th century woman with the improbable christian names of Terry Lynn . (her descendants have equally improbable names).
Even if she was a real person called Maud Wells without any location  she would impossible to source.( where to start looking?)
 I personally agree that if we could find a way of  getting rid of this garbage , it would help those that have the skills to work on improving whats left .This is in itself a gagantuan task with very few able and willing to take it on.
( I am pre 1500 'certified', but I don' t count myself as one  of them for the earlier periods since for the most part I would be reliant on secondary sources.For 13th and 14th C English profiles, I don't really have the skill set to  read/translate original records written in latin and medieval French.Indeed, even interpreting  transcriptions when they are availabe can be difficult.

---

Helen, a very very excellent example of a bogus profile.  One issue is that she exists all over the internet.  Her husband had a slightly more probable name, so I googled him.  The husband lives on Find a Grave as a real person, and he lives on Geni.  Geni, however, has already identified him and the whole family as fictional.

I did a quick fix on the husband's profile.  

https://www.wikitree.com/wiki/Wielher-12

As a fictional person he doesn't get a biography, only research notes.  I cleaned up the original  Gedcom sourcing and added Geni as a source.  (And I'm using "source" here strictly to mean "where the stuff came from.")  

What he and his family now need is to be marked as "Disproven Existence," and de-linked from parents, wife, and child.   That way anyone who comes looking for more information about the name will see what we know about it.  Geni links back to WikiTree profiles, so anyone who looks there will link back to our current information.

What I'm curious about is where the fictional genealogy came from.  Is it a recent creation or did it come from a well-meaning book in the 1866?  The use of 20th century name combinations like  "Terry Lynn" marks it as a recent creation.  If we find out who created it, that needs a write up under frauds and fictions.  Creating fake individuals like these is a crime, and crimes need publicity rather than being hidden.  Did I just say I favor public hangings?

---

I think you did, Jack :D

I have come across a lot of dubious profiles from the same Gedcom, or at least the same provenance. I mean they were imported by the same person.

The first thing to do in such cases IMO is go down the generations until a real person is found. Then entirely disconnect the real from the fake, so that the set of fake profiles is quarantined and prepared to be Disproven and dismantled.

---

@Helen, 

It certainly DOES represent a lot of work, because those names aren't just pulled out of thin air, they are linked but, as I do, I'll add names to my tree with approx dates, or even just names, and then go back to do research when I have time to dedicate to just one person at a time and add the sources.  

I use many sources, and set a specific number of people I will post into my tree then I stop adding and start researching but I never ever just pick a name and then see if I can link it to myself.   That would be useless and a waste of time. 

Giving me examples is just fine, but is not relevant to me at this time.  I don't sit at my computer 24/7 typing in data aimlessly.   This is my hobby as well as for many other folks.  It's NOT a career, or a job.   It's a hobby and something I do for pleasure.  I get thrilled with history and have become a real history junkie but having someone beat me up because my source doesn't "fit the definition" is like telling an excited child their accomplishment for the day was worthless.    My work is important to me.  It's not worthless nor is the time I put into finding these names and then researching. 

I have a reasonably large tree and it was built from good sites built by hardworking people.   Just because I didn't stop and fill in the other blanks doesn't mean they didn't do years of work, or that I still HAVE years of work ahead of me.

Saying I didn't take time to build my tree is just plain wrong and rather arrogant, to be nice about it.   I take the time (10 years so far) and my husband has done (30 years) enormous amounts of research and has original family books that I haven't even touched yet, so we ARE doing work, even with one name or a complete profile...If wikitree removed future sources for the me and others to find would be the worse move wikitree could ever do for the average person.  

Wikitree can do as it wants... but removing profiles from an entire era would be an even worst move.   Removing Charlemagne, Ceasar, and Ptolemy and other historic people (whether they have wikitree profiles or not) would be more than just wrong.   They are history and can't be erased. 

My choice would be easy if wikitree removed every name before 1400 without a source.  

---

I beg to differ. "Teri Lynn" as given name for a medieval profile DOES look like it was pulled out of thin air and is just plain wrong. Either someone plain invented it and it got passed along; or someone duplicated a recent person to fabricate a medieval ancestor... and it too got passed along.

Either way, the profile cannot remain connected. It must be labelled as an error and detached from the rest.

---

@Rebecca, I definitely don't think anyone is proposing deleting Charlemagne. Does he seriously have no sources?

Secondly I think that any genealogist who does serious work without keeping track of sourcing rationales is inevitably going to come into some sort of conflict with other genealogists IF they decide to work on a "single tree" project. This is because:

1. There will inevitably be differences between the work of any two genealogists who have worked on their own, and this is even more true in cases where there has not been some sort of "strictness" about making sure there is always a sourcing rationale you can find for every person.

2. What we see before 1500 is that everybody's family trees overlap, so if everyone has their own individual vision and can not explain why it is better than everyone else's then the single tree project will fail. This is more or less what happened with other websites in the past.

So does Wikitree really have an option?

---

@Rebecca -- please DON'T create several profiles and then plan to do the research later.  If you are like me in even the tiniest way, distractions occur and "later" may be two years away.  Do profiles with an adequate amount of research one at a time.  Then "later" will never bite you (as it has me).

---

@isabelle -- I agree completely on Terri Lynn.  But profiles like hers are cancers.  You don't solve anything by snipping out part of the cancer.  You need to investigate to find out how far the cancer has spread, what kind of cancer it is, and what's the best treatment for the whole cancer.

---

Please understand,,, I do not create ANY profiles on wikitree other than my own and my immediate family.  I am talking about building my own private files with my own program which is not shared anywhere.  

And the point was missed completely.   Deleting history... ANY history is wrong.  I only used Charlemagne as an example and YES I have plenty of sources but my sources rarely have satisfied anyone. 

The original post was a broad "delete the garbage post".   While my words are obviously getting twisted to suit the "crime" rather than looking at is a true informative opinion as a hobby researcher, I will end my part in this. 

Delete if you must... I personally (and others also) don't feel it's proper.  I'll take care of my few profiles and future research elsewhere.   

---

I can't see how especially the pre-1500 part of wikitree can accommodate thousands of different hobby researchers who will all have different trees, unless it gives up the aim of being "the" single tree site, or even any kind of single tree site.

There are of course heaps of places on the web which do accommodate thousands of people's unsourced trees. Wikitree kind of started from such material, and from that experience, still not fixed, we know it is totally incompatible with the aim of having one single tree.

It is never possible to merge them, so especially in pre-1500 you end up with dozens and dozens of profiles for the same person, but unmergable, because with all kinds of different details. The only logical solution, surely, is to use the standard of trying to work out which options were true, or most likely true. Once we take that road, I find the amount of dispute is surprisingly small.

---

Thank you Jack, once again decisive ! (  do you want them hoist from a branch of the tree ?)

Rebecca, none of the answer was personal or relates to your particular tree, I have no knowledge of it. If you look at that particular link I pointed to I'm sure you will agree that characters such as Terry and her descendant Patty Sue et al cannot have been researched in any way. No-one is suggesting removing documented historical figures (as long as they are not BCE)

When any profile is created,  the inclusion of  some sort of evidence  is important (that's why there is now a tick box about sources when a new profile is created)  The honour code for wiki-tree says that we cite sources and care about accuracy.  I would suggest that we can't evaluate accuracy without those sources .Unfortunately, there a number of  profiles and lines such as the one I cited which can only be called garbage.They don't help enhance wiki-trees credibility (and any claims to 'relationships' that pass through them will unfortunately be false)

Answer 2

John, Your idea would be much harder to implement than you think. First lets deal with the unsourced part of it. I recently adopted several profiles that were added in 2011 and not touched since. They are unsourced but at first glance they look like they have sources.They are not marked unsourced so they are not easy to find.

Comments

Still, Wikitree needs to find a way if it is going to keep momentum in pre-1500.

---

I think the pre-1500 momentum will continue for these reasons:

(1) We are not creating new junk pre-1500 profiles.  So there is a finite number of junk profiles to attend to.  We may not know the number, but it is finite and shrinking.

(2)  Even without an organized clean up program, the day to day activity of pre-1500 qualified WikiTreers is improving pre-1500 profiles.  So they're getting better.  

(3) There's an essential difference between WikiTree and soc.gen.medieval.  While WikiTree is developing a culture of excellence, it is still oriented to a large number of people, not just a few experts.  soc.gen.medieval is oriented to the true experts.   I think it will simply always be the case that WikiTree looks to soc.gen.medieval to find the latest expert thinking, not the other way around.  I don't ever expect to see someone at soc.gen.medieval say, "Well, it says such and such on Wikitree, and that should settle the matter!"

---

No one knows everything on all of the English medieval Genealogy,so they may have and still might check in and follow some of you wikitree experts. I know you mean as far as your point goes in terms of being progressive. But honestly, even someone does not know vasts amounts on a topic could find the answer and even well document and sourced. The problem is that the word tends not too go far and wide unless the professional experts making a living off they're literature can bring said claim too relevant mention. These experts still want too make a living though so usually if they look into an concise from someone amateur and find it correct they will put it into they're own words without simply referencing said person so then you never know if they got the answer from you. XD

Answer 3

I would change that to delete (or "reconsider"?) profiles that cannot be sourced... A lot of unsourced profiles (tagged or not) represent identifiable people and could be sourced. 

I agree there are a lot of fanciful profiles and the problem should be addressed.

Answer 4

I agree that something should be done, and your suggestions are all good ones, especially cleaning up all the unconnected orphan profiles.

From this G2G:

1. Everyone who has pre-1500 rights should be able to edit project protected profiles.

2. Everyone who has pre-1500 rights should be able to modify an LNAB. 

And I still think we should remove all profile managers on Pre-1500 profiles who are not certified. I always marvel at the profiles with multiple managers, and none can edit the profile. See Godith de Furnesio, where this was sort of fixed (but still has surname problems and should not have a father.)

All of these suggestions (mass deletion, expanded edit/LNAB/merge rights, and removing PMs) have been rejected in the past. 

Comments

Yeah I see what you mean Kirk. I do know of a pre-1500 That I wanted too work on in the future if I can get certified. But yeah the pre-1500 Is pre-1500 man I wish I could work on some of those and learn even more. XD ill wait until the time comes and the iron is hot. Thx pre-1500 people love ya for all the hard work. XD

---

I see some work has just been done on Godith last year after some requested changes by me. I would say that having Eldred as father is probably now at least the best guess, and having Workington as his "surname" is probably the best choice given Wikitree policy. I think my messages on that profile refer to earlier versions.

Answer 5

I'm not sure what the answer is John.  The fear is that you throw out a lot of good work and profiles along with the garbage.  I am guessing that this question arises out of these two threads on SGM ("wikitree" and "gold standard") which show just how disrespected wikitree is held by serious medieval genealogists.  Their complaints are valid and it is hard to defend the amount of garbage on wikitree.

Wikitree should be the perfect format to grow a peer-reviewed, medieval database.  But the tree is so corrupted with junk from early gedcom uploads that almost no one on SGM thinks it is worth the effort to fix.  My personal choice was to continue with wikitree and try to make my little corner of it as good as possible.  I can't see wikitree deleting thousands of perfectly good profiles, but I do think there needs to be changes if wikitree is ever going to earn respect as a reliable medieval database.

I would make lots of changes.  I am not sure any of them would make a real difference in fixing the underlying problem of a flawed database.

Comments

I agree Isabelle and Kirk.  The way wikitree handles names is absolutely terrible and one of the big problems with the medieval period on wikitree.  The concept of "their conventions, not ours" does not work for this time period, and is not used anywhere else in genealogy but on wikitree.  It is often an embarrassment frankly. Experienced genealogists look at it and say what the hell are you people doing. 

During much of this time people did not have  a LNAB the way we think of it.  They spoke in one language and wrote in another.  Spelling was almost random and certainly unimportant.  It is wikitree which is imposing a ridiculous standard which increases confusion rather than decreasing it.

1. There needs to be standardization of names to match how they are commonly found in sources.  This will make the names understandable and recognizable to our readers.
2. The concept of "their conventions, not ours" needs to be specifically rejected for this time period.
3. We need a title field,or the suffix field extended so we can have normal looking names.  All these names with the title inserted into the middle of the name look just terrible
4. We need a searchable name field where alternative names and spellings can be placed without it being in the display name.

---

Joe:

ad 1. What standard would you propose to use for names? English? French? German? Latin? Most medieval rulers are known by different names depending on the language the secondary literature we depend upon is written in (can't figure out how to avoid the dangling participle ...).

ad 2. That leads directly to a partial refutation of a categorical rejection of "their conventions, not ours": If we truly want to be an international site and not an exclusively English speaking one we need to name them in the language closest to what they spoke.

ad 3. and 4. Wholeheartedly agree, have been asking for something like that for a long time.

---

I knew this would come up Helmut.  And I really hope you don't think I am being disrespectful.  Wikitree's attempt to be everything to everybody is a major part of the problem.  But wikitree is an English language site. I know we have many international users, but all of them seem to be doing just fine with all of wikitree's pages written entirely in English.  You and Isabelle, and Ales, and Magnus (where is Magnus?) and Eva and Bea all do fantastic work without English being your primary language.  Would you really be offended if we said William the Conqueror of Normandy, King of England instead of Guillaume le Conquérant de Normandie, Roi d'Angleterra? (I hope I got that right).  The one that drive me the most nuts is the profile of Анна Ярославна Киевская  - that's Anna of Kiev if you are having trouble recognizing her name or finding her.

Wikipedia's response to this problem is to create separate encyclopedias for every language.  This is impractical for wikitree.  There needs to be some standardization, including in language, so that we can write good, meaningful profiles that are searchable and findable by the users of wikitree.

---

Well, I have no problem (comparatively) with her being Анна Ярославна Киевская. More precisely it's a minor problem with the ton of dross currently encumbering her profile.

I have no problem with William the Conqueror, oddly enough, but I have one with "Eleanor" of Aquitaine - a name she never used to be sure, she never spoke English. But I'm in the minority and I accept that.

The important thing is to try and construct something as we can. I'm not sure we can change the database fields. I'm afraid we can't. But we can try to clean up the garbage. Getting rid of what is really garbage in order to focus on what can be saved would IMO be a step in the right direction.

---

I tend to be with Isabelle and I think Wikitree can and should aim to be international. There are no real borders in genealogy.

Furthermore I think the solution to the name problem is simply not to have such inflexible "pseudo important" fields. The standard wiki search is like a google search, and just searches any words you put in. I don't think LNAB being important wins anything.

Concerning cyrilic alphabets, I think a more convention solution would be a view setting which can convert them to one of the standard transcription styles. Анна Ярославна Киевская just says Anna Yaposplavna Kievskaya, so you can see Anna and Kiev.

EDIT: sorry, Yaroslavna.

---

I would have a tiny problem with Frederic the Great, I think. But ultimately "Namen sind Schall und Rauch" (Goethe, Faust), the bigger problem bearing on how WikiTree is perceived with respect to it's medieval profiles is that we never get around fixing problems. We PPP profiles and tell people to discuss with the project any significant changes, it gets put up on G2G, a very few people nod gravely, and years later the error is still not corrected, maybe a maintenance category has been added. Maybe we need a process similar to default approval of a merge: I'm proposing a significant change to a pre-1500 profile and if nobody objects explicitly within a certain time frame the change is approved. Example: There can really be no dispute about the centrality of a profile like Charlemagne's, yet we seem to be unable to correct something as simple as his birth place - it is not known and certainly not "Aachen, Rheinland-Westfalen, Deutschland".

---

Since I was mentioned by Joe, I'd just like to say that I agree about WikiTree being at root an English language site. It's built into the structure - and by using WikiTree I have accepted this.

As for the pre-1500 Augean stable, of which Sweden has it's fair share, my personal attitude to this is that I have avoided getting the badge. I love researching from primary sources = the Swedish church records. They go back to the early 1600s in the best cases, and this is where I prefer to improve the tree and hone my skills.

---

If we look at say pre 1200, most of the lines of ancestry people with British roots can trace any further are not English. Furthermore British ancestry no longer dominates the family trees even of people in English speaking countries.

Anyway, if there was some kind of reason to make a trade off then we'd need to consider. But I see no real reason to not let Wikitree be increasingly flexible about language. There seems to be an idea that its a technical issue, but I do not think it is.

---

@Helmut - just saw you're moving forward on Charlemagne.

I was going to say that, when we see something that needs changing and we think we can change it ourselves, we should just post a message on the profile and/or on G2G (depending on how important the profile is) and if there is no reply after two or three weeks - especially with profiles with a lot of managers - just do it. It will still be slow, but some things will get better. For instance the 3 or 4 spurious children who were attached to the infamous Anna Iaroslavna profile one year ago are gone now.

---

If we want a truly international site, the main language has to be English, just as in medieval Europe it had to be Latin.

What matters is to get the tree structure correct. If that is done, it is a vastly easier job to translate the words that desceibe each entry into another language. The structure itself is coded in some computer database language.

Answer 6

I think sourcing on its own is not going to be a good enough reason to delete profiles, at least not yet. I guess many real people and families have unsourced skeletal information and therefore should be improved rather than deleted. But it would be good to develop some way of making bigger deletion decisions to clear out the enormous backlog.

Unfortunately though, I can not think of a simple method. Many ugly looking articles which are clearly uploads from poor quality gedcoms of long ago, do contain a real person within.

Answer 7

You may wish to re-read my recent proposal: not pre-1500 certified? then don't adopt pre-1500 profiles! - which was unanimously approved (but there was a small caveat) - but turned down.

http://www.wikitree.com/g2g/564762/proposal-not-pre-1500-certified-dont-adopt-pre-1500-profiles

Comments

Oh yes... *sigh*

Answer 8

Have to say some of the criticism seems invalid, for one specific reason:

The only way to achieve a quality standard is not to put stuff out until it's ready.

WikiTree tends to forget the difference between contributors and users.  Many members are hardly aware that non-members pay the bills.

There's little difference between what we see and what goes out to the general public.  All the unchecked half-baked work in progress is broadcast to the world.

And there's nothing added to say "work in progress" or "yes we know it's not right" or "assembled by smart-matcher".

Obviously, it isn't going to be good.  It isn't supposed to be good.  The only way it could be good is if all the work were being done somewhere else.

Comments

Doesn't that mean the criticism is valid? I think I don't follow completely.