Problem bad combinaison and searching to understand

Hi,

I have a problem with a combination / disambiguation and I have tried multiple actions and can not find the solution to properly make things.
This concerns 2 books with several authors (Main and others as contributor oer preface)
who are grouped here :
https://www.libreything.fr/author/leclercdehauteclocqu

In detail:
1 - Philippe Leclerc de Hauteclocque 1902-1947
https://www.librarything.fr/author/leclercdehauteclocqu 1

2 - Thérèse Leclerc de Hauteclocque (2) 1903-1996
https://www.librarything.fr/author/leclercdehauteclocqu-2

For 1: we do not see (1) and this author is a contributor (and not author who is René-Jean Clot.
Is that this poses a problem?

Beyond how to dissociate authors in a clean way?

I remove the books from my library, add them. I tried to associate and then dissociate. I indicated the names of each in the works (in my library) but also on the pages of each book. No change.

Can someone explain to me what to do? Or is it a bug?

Last thing, I completed the CKs of each author and added HTTP links by spending time and during my tests, I saw that these CK could disappear ???? Also, I open several windows to keep the CK and copy / paste ..... but what waste of time and energy

What is the solution ? How to understand? What to do in the future? (PS: It's not the first time I meet this pb when an author is contributor)

Thanks a lot

NB: I gave things in the state to keep the CK (which are in the French language part)

New case for an other author

More or less the same problem

Is it a "é" problem for these

Yves-Marie Péréon
différent with
Yves-Marie Péron

https://www.librarything.fr/author/pronyvesmarie&recalculate=1&check=118...

>2 JMK2020: That's because it's unfortunately coded as pronyvesmarie - so there's no way for the system to make a distinction between Péréon and Péron beyond us splitting the page

>1 JMK2020: The problem is that 'Leclerc de Hauteclocque' is so long that LT can't differentiate between the two names. I took the division out and put a canonical name in so it doesn't have Therese on the disambiguation page, and then redivided.

Please do not remove the canonical title from the disambiguation page.

Tx

So, you have to do with the limits of the system and and do the least worse

>3 Nicole_VanK:

same problem with "¨" in Krüger for example... system transform with Krger... So Kräger is Krüger.....
(I had this pb so do my best with not perfect combinaison/desambiguation... and more, it's a pb between, languages
Txt indicated - for example in french - in the despambiguation zone does not appear in german or english...

https://www.librarything.fr/author/krugerhorst
(I write a txt in . fr and had to copy in .de and .com from .fr. // There were already text in each language)

It's not excellent because it creates confusion

>4 MarthaJeanne:

Idem above (with an other limit)

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

Question :

Is a page or some pages on the site that explain the limits and constraints of the system and explanatory pages to do the better ? Where ?

Tx

>5 JMK2020: All over the place :) Most of the people in the thread had been here for a long time. Some of the information is intuitive once you realize how LT works; some had been explaining through the years...

There are some help pages but they cannot cover everything. Thus all of us always being around to explain and assist.

The two problems you are up against here are

1) The author name only recognizes 20 digits. This only very rarely affects people's names, but is often seen in cases like American Association for ... In these cases, a canonical title before dividing makes things at least look neater.

2) URLs are built from 'standard' letters only. Sometimes letters with extra bits are put in with extra characters, otherwise they are dropped when creating URLs. How they are represented is not LT's decision, but depends on how they were entered. This is why Müller, Möller and, if it exists, Méller, all end up with the same URL. Some of these characters may be usable in URLs today, but they weren't 15 years ago.

Tim has talked about changing from author names in the URL to numbers. Implementing that would be a lot of work, and still wouldn't solve the recurrent problem of "I just went to MY author page, and there are several books there that I didn't write, and I want all the books listed as being by John Doe that aren't mine taken off MY page right now." (Never mind that this John Doe has one book that only he has entered that landed on the page of someone with 20 books each of which has hundreds if not thousand of entries.) The computer has to be told that there are several people with the same name, and which books belong to which person.

Sometime after :

It seems OK now
Is my action is ok (perhaps a updated question from LBTh)
Sorry.... ???????????????????
Is somebody could verify these ? Tx

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

The initial subject was :

QUESTION / SOLUTION ?

An other pb and i can't resolve

It's a confusion between homonym ans Name/Surname ??? !!!

Here the book :
https://www.librarything.fr/work/10773498
With the author
https://www.librarything.fr/author/micheldavid-1

We see : "David Michel (1) has been aliased into Michel David." (AA)

So it's a pb between and with Michel DAVID and David MICHEL

https://www.librarything.fr/author/davidmichel

https://www.librarything.fr/author/micheldavid

Question how separate the combinaison (AA) : micheldavid-1 = David Michel (1) from Michel David
nb :
a) CK for each Michel DAVID is OK (plus HTTP links), separation too, i added desamguiation note on each.

b) I'm working on CK for David MICHEL but i don't want to spend time and time if a simple action (future) remains everything to 0**

I saw it's sometimes the case and historic isn't in memory.... (I already worked on Michel DAVID and David MICHEL).... Lost :-(

What is the solution ?

nb : i send a post to member to change name but.... (and it seems it's another amz confusion)...

David Michel

Michel David

The book by David Michel 1, is seen by its cover to be by David Michel, not by Michel David, so I have deleted the aliasing.

>8 JMK2020:
OK, how do you made to deleted aliased ?

You go to the division page, and click on alias divisions to other authors. If there is something incorrect entered there you delete and save.

>11 MarthaJeanne:
I made it and after I hadd just one name
for
M DAVID
and
D MICHEL ?

On my screens all was des(aliased) ??

is it différent between .com .de or .fr ???

Is it a question of delay vs updates ?

I have no idea what you are saying.

>12 JMK2020: We need some links or a better explanation of what you are seeing.

>13 MarthaJeanne:
>14 AnnieMod:

Thank you for everything.
I 'll come back with a next example.

I'm not sure to understand all the intricacies of the system.
And I would like to know them in depth so that my contributions are useful and made in the rules of the art.

I often encounter problems with bad combinations.
The reasons are multiple: Homonyms for example for the authors or inversion and confusion between name and first name (on this last point OCLC, ISNI, LOC are to be checked and the errors are also quite regular)
There are also the combinations made quickly.
To have and keep data I complete the CKs, put links in the author files and write the disambiguation note. Finally, I assign the work to each author.
In fact, I do multiple searches, I cross and I aggregate part of the information in a personal database. I report part of it in LBThing and in parallel I complete wikidata. Finally, when there are discrepancies I send the information (with the sources) to one or other of the authority sites which will make the modifications. It is very useful - worldwide - and perhaps, one of these daya, ISNI for example will be much cleaner)

For books, I also encounter problems. But that is not the subject here.

In my questions, there are questions like:
How to use and fill the pseudonyms ? How to link them to have only one file authority?
How to find a history in the author files when the CK fields have disappeared (this happens regularly, maybe because of the combinations) ?

But I would come back to all of this with links and / or screenshots - and more explanation - for you to understand the issues and guide me.

Once again thank you

New problem with this homonyms
Gary Sheffield

Why Gary Sheffield (1) have "Works have been aliased into Gary Sheffield."

How separate (1) and " " ?
What are the actions to do to just have clear (1) and (2) ?

Désambigüisation page :
https://www.librarything.com/author/sheffieldgary

1 : https://www.librarything.com/author/sheffieldgary-1
"Works have been aliased into Gary Sheffield"

2 : https://www.librarything.com/author/sheffieldgary-2

As usual : CK, http links and pictures are OK (as desambiguation note)

Tx

>16 JMK2020:

The page to which GAry Sheffield(1) is aiiased needed recalculation. It's actually G. D. Sheffield.

I really need to understand how the authors' combinations and ambiguities work.

Maybe I'm an idiot, but I don't know how to do it right.

I have an example below (from the French version of LBth):

1) I enter a book from the ISBN
https://www.librarything.fr/work/1747107
The author (official authority for France) is William Colby (cf : https://data.bnf.fr/fr/11897239/william_colby/)

2) when I click on William Colby, I have the disambiguation page:
https://www.librarything.fr/author/colbywilliam
With William Colby (1)
But not William Colby (2)

???? Where is 2 ???

3) Finally when I click on William Colby (1), I have the author's page
https://www.librarything.fr/author/colbywilliam-1

4) When I look at the possible combinations, I come up with yet another page:
https://www.librarything.fr/author/colbywilliamegan

I don't understand why there are multiple pages on this author: 3) and 4) with different CKs.

How to understand all this ???

How to have 1 page per author?

Can we not start from the homonym and then distinguish the authors to catalogue the good work to the good author ?

>18 JMK2020: All the books that were assigned to -2 were either deleted OR changed their authors. That's why you do not see -2 anymore.

This is who -2 was supposed to be: https://www.librarything.fr/author/colbywilliamh : William H. Colby - some editions lacked the H. thus landing them on the William Colby page :) Somewhere along the line, the books shifted.

Now... we have two choices:
- Unsplit the author - we cannot just now because the system is broken and once someone is split, we cannot unsplit.
- Leave it as is in case another book shows up under the name without H. that needs to be split and aliased.

Don't look at the possible combination and please never combine based on them unless you are REALLY sure they are the same person - or if you do, look at the addresses. That last page is colbywilliamegan: William Egan Colby. He should never be combined with William Colby

Why this last guy has a canonical name which makes him more confusing is a different story - I will go investigate.

>19 AnnieMod:

Now... we have two choices:
- Unsplit the author - we cannot just now

I have unsplit the author.

>19 AnnieMod: The only book on the William Egan Colby page was the French translation of Lost Victory which I have now combined.

>18 JMK2020: All we need to not have to sort these things out is for everyone to use the same name when entering the books by the same author. Since that won't happen, we need combiners.

>20 lilithcat: Ha. So it works for some authors. I know we are stuck for some... maybe because of books that fell out of the page but are still assigned?

>21 MarthaJeanne: Now that he is unsplit, we probably should combine the two authors then - they are the same one.

>22 AnnieMod:

I have no clue why it works for some and not for others. This one was smooth as silk, but often I need to move everything to "unknown", save it, and then go back in and switch from"divided" to "single" author. It's very weird.

Oupssss....

Sorry for this long text

For this William Colby case, I haven't touched anything. Sso that you have an example of what I consider to be an anomaly (or something that needs to be improved)

a) I had seen the correspondence between the versions in different languages ​​but did not make the combination to have live factual situation

b) Regarding the William Colby (2), I understood that it could have disappeared if there were no more books present in a member's library (or if name canonique change or was aliased... with sometimes bad consequences for CL fields...)
- There, the disambiguation note is a really good thing when it is clear and explicit. And that's what I'm trying to do

Example :
1: Jean RENAUD (9 authors)
https://www.librarything.fr/author/renaudjean#ck_disambiguation
here, it's easier for me because I know where to get the information and cross it to make the combinations / Distinctions

2: Paul SMITH (26 authors)
https://www.librarything.fr/author/smithpaul#ck_disambiguation
There, it is a little more complicated because I know less well the US / UK resources (and that it requires the same work for me but much more time). For this Paul SMITH, there are still many Ck to find (Date of birth, photos, occupations .... http). To be continued

In any case, for case 1, and after working on it, LBth is more correct than ISNI, BNF, LoC, DNB, VIAF ....

xxxxxxxxxxxxxxxxxx

But back to this William Colby.

As already said, in Fr, .De too, the canonical name in the academic base = William Colby (and this is the name indicated on the books).
Of course the form of the names can change: William E Colby, W.E. Colby ..... Captain Colby, General Colby, Chief Colby ... and this differs according to the countries (without saying Cyrillic alphatet or Corean or others)

My question therefore remains:
How to have one page and only one for an author identified William Colby with different forms of names

A) William Colby according to this book (fr or US) = the head of the CIA 1920-1996 (https://www.librarything.fr/work/1869153/covers/206600971)
It is found here:
https://www.librarything.fr/author/colbywilliam -
CK empty but 2 works (X1 / X1a)

but also here

https://www.librarything.fr/author/colbywilliamegan -
CK OK but 0 work (X2 / X1b)

???????? (for me Egan is not the primary data)

B) William Colby according to this book (US) = 1955- (https://www.librarything.fr/work/1636836)
It also bears the name of William H. Colby (https://www.librarything.fr/work/709123)

It is found here:
/www.librarything.fr/author/colbywilliamh (X2)

??????? (idem)

I'm a little lost, or even completely lost. Do you think i'm idiot ? What is the right medicine ?

Why several pages for 1 single author with different CK and http links on each pages ?

Why are homonyms not primary combined (X) and after separated between X1 and X2?

xxxxxxxxxxxxxxxxx

Personally, this is how I work:

1) I'm looking for an author by his most commonly used name (dependaing on language area) . Let's say "Pierre PETIT" (or a work with the same name of use)
https://www.librarything.fr/author/petitpierre#ck_disambiguation

2) Then I look at occupation, date of birth / death, Country ...

* And it is then that I observe if the works of a contemporary astrophysicist are mixed with that of a musician of the 12th century.

* I then proceed to the separation by looking if there are other orphan works in LBTh. And finally I complete the note.

In conclusion for this William Colby, I imagine it's easier to combine the William Colbys who have the same name on their books (see above) and after separate each other with additional criteria (2nd, 3rd, 4th first names, age, date ... .)

xxxxxxxxxxxxxxxxx

Sorry for the long note. I have struggled everyday - for a long time - with unique identifiers.

With a clear method and process, I think assigning a unique identifier should be possible. But maybe it is complicated. This is why I use encrypted identifiers: 1 code / number for an author (like Isni or Viaf) then add variations of nominal forms.

Could you give me example (with links and explanation) for a symbolic author.

I'm working on
https://www.librarything.com/author/smithpaul

Question why "Paul Smith (13) : paul smith" ? Where come from lowercase letter p and s ? Why ?

Is this a good start job ?

Tx

>22 AnnieMod:
>23 AnnieMod:
>24 lilithcat:
An other pb here for an Homonym (from one of my books)

David Chandler :
https://www.librarything.com/author/chandlerdavid

I was creating, searching, tidying up when I came up with another problem:
There was - in the middle of all these David Chandler - 2 David Chandler entries for the same author: (1) with many books and (11) with one book,
so i assigned the book 11 in 1

David Chandler (11) is David Chandler (1) ????????????

So 11 is for this title
https://www.librarything.fr/work/16006300

And this Author :
https://business.ucdenver.edu/about/our-people/david-chandler

https://business.ucdenver.edu/sites/default/files/attached-files/vita_david_chan...

but :
David Chandler (11) : David (ed.) Chandler
Works have been aliased into David G. Chandler.
Corporate Social Responsibility: A Strategic Perspective 3 copies

How desaliased 11 from 1 , keep 1 (Historian)... or the others with just 1 page per author from the name David Chandler ?

Please take care of the CK so I can continue without spending a lot of time

nb : note of desambiguation is ok (apart this 11, ... i'll come next time to finish the job... but for this day, i'm desesperate)

>26 JMK2020: Every time I see you write pb I think paperback.

>26 JMK2020:

I am not 100% sure that I understand what you want to do, but if you want to move the book The Art of Warfare in the Age of Marlborough from David Chandler (11) to David Chandler (1) (which looks correct), click on the "edit the division" link in the box on the upper right of the split author page.
In the following page, just locate your book and change the associated number from 11 to 1. Then click the "save page" button.

Does this answer your question?

>28 Felagund: It looks like one issue is the mis-assignment of that work to #11 instead of #1. And another issue is the alias of #11 to David G. Chandler when it's actually #1 that should be aliased there. I have made both of those changes, and I'll spend a little time trying to identify the proper splits for the "unknown" split works.

>26 JMK2020: Felagund addressed how to change which works are assigned to which split authors -- use the "Author Division"/"edit the division" option ("Séparation d'auteurs"/"modifier la séparation"). For how to assign or change an alias (such as to David G. Chandler) you also use "edit the division" but in this case the option at the top of that page to "Alias divisions to other authors" ("Combiner un(e) auteur(e) séparé(e) avec d'autres auteur(e)s"). You put the author code for that other author you want to alias to (in this case "chandlerdavidg") into the box for the correct author split and then save.

As for the broader question of why there are mistaken data, errors, and so on, partly that's a result of this being a public group-sourced database with many many people working it at once -- some people will make errors or have their work conflict with somebody else's work. There may also be bad data from some automated data importing, or data which has been imported from sites which have their own errors (like Amazon). Hopefully people will clean these up when they spot them, and if it's something complicated or confusing, bringing it to this group is appropriate.

>29 scott_beeler: I noticed another problem with the David Chandler page -- somebody had combined David P. Chandler into it, which is the wrong way around. I separated him out, which took most of that split's works with him. There is only one work left by that particular David Chandler, and I have aliased that split into the now separate David P. Chandler page.

Thank you,
how did you manage to separate and clarify (1) and (11)
What actions in relation to the authors?
Should the names on the works also have to be changed?

Complementary question:
What are the consequences - in LBth - of a name in "Canonic name", Legal Name and Other Name, for the author file?

I will continue to complete this CK fields... and am happy to have already been able to complete some data on (11)

Could you see this, before i complete, is-il good ?
https://www.librarything.fr/author/smithpaul#ck_disambiguation

Tx

xxxxxxxxxxxxxxxxxxxxxxxxxxx

>28 Felagund:
For the rest, I think I understand most of the attachments / detachments

I also understood that multiple actions of members or bad data (external sites) could generate confusion
I am trying to clarify all of this with now automatically a clear and precise note in the disambiguation space for the namesakes. I can't always find ... but it is generally more precise than the data in the official authority files

xxxxxxxxxxxxxxxxxxxxxxxxxxx

>27 jjwilson61:
"pb" is/for problem ;-)

>31 JMK2020: "how did you manage to separate and clarify (1) and (11)
What actions in relation to the authors?
Should the names on the works also have to be changed?"

You go to the author page. You look at the right menu and locate the "Vous pouvez modifier la séparation". You switch the work from 1 to 11.

On a separate note: You may want to move to the .com version of the site so the explanations we are all trying to give you make sense. Once you know the UI, move back to the French side - everything is in the same place, the labels are different. I think that this is at least part of the problem - you cannot find what we are pointing to because you are looking at the interface in French.

>31 JMK2020: "What are the consequences - in LBth - of a name in "Canonic name", Legal Name and Other Name, for the author file?"

Legal name and Other name are harmless and can be set to any NON-split author at any time (for split authors, you set them in the split portion of them).

Please never set Canonical name UNLESS the name in the DB is really wrong. When you set these, they hide issues and lead to bad combinations unless people pay attention (and most people just look at a name and do not bother with the URL - and that causes issues). So leave the field empty - half of the complicated author issues are because someone set a canonical name somewhere at some time.

>32 AnnieMod:

1)
Yesssssssssssssssssssss ;-)
Well, it was enough to say it: .com> .fr
I could always fight with myself and all the test on .fr

2)
Tx for "Please never set Canonical name UNLESS the name in the DB is really wrong"
it is clearer. Indeed, there's sometime strange name

xxxxxxxxxxxxxxxxxxxxxxxx

What about the names of authors (or contributor, publisher, photographer ....) for the books themselves?

Example of names for the same author and the same book:

.com (US) version:
Paul Read Smith
Paul R. Smith

.de version :
P.R. Smith

.Fr version
Paul smith

What will be the canonical (and unique) page of the author? (see below**)
Under what name?
Will there be only one page for this paul read smith ?
Will there be any differences between the .com or .it or .fr ?

Is it possible to have different pages depending from language of this Paul Smith (with a different number of homonymous authors) ?

xxxxxxxxxxxxxxxxxxxxxxxx

From :
https://www.librarything.com/author/chandlerdavid

we have
David Chandler (1)
so page A : https://www.librarything.com/author/chandlerdavid-2

but David P. Chandler too
so page B :
https://www.librarything.com/author/chandlerdavidp

So, 2 different page with different CK (data, http links, pictures) between A and B.

maybe this is normal and I must understand that on this David Chandler (1), you have to go to the alias David P. Chandler. And that there was not only one page ?

>33 JMK2020: "Well, it was enough to say it: .com> .fr"

Not really more (if that is what you mean) but different enough to trip you if you do not understand. So look at .com to see what we are talking about, then look at the .fr site to see what corresponds to it. The translations may not be the same - which can make it harder. Thus the "use .com to learn things, then move over" :) And you cannot expect people to post the texts from the French side - we may remember sometimes but... :)

On the second one. The data is the same across all sites. There are elements which can have different values (Common Knowledge) but the data is the same. So Paul Smith on the .com site is exactly the same record as on the .fr side or the Bulgarian site.

Think of the authors as a single entry with a list of 49 sets of common knowledge fields. Most of those sets stay empty and LT will helpfully bleed between them (so if you fill something on the French side and the English one is emtpy, it will tell someone looking on the English side that there is something on the French side - do they want to confirm it or add something different on the English side - but besides that it is the same author name.

Which name bubbles at the top depends on the number of books being added.

The Chandler example is because of aliasing - not because of different languages. When an author is aliased, they indeed will have two pages:
- The split portions (author-x format one)
- Their actual page
Some may even have more than that if more than one part from different splits are aliased.

In such cases the main page is the actual page (so https://www.librarything.com/author/chandlerdavidp in your case) but the https://www.librarything.com/author/chandlerdavid-2 still allows CK (because noone decided to disallow it). It is just one of the quirks of the system.

If I see "Works have been aliased into" note, I usually will NOT add CK - I will add it on the main page.

>34 AnnieMod:
Thank you for all your answers. They allow me to better understand how the system works.

When I wrote ".com> .fr" I just wanted to say that to understand each other, it was better that I refer to LBTh.com (rather than LBTh.fr.

For the different CKs between versions and languages, I had already observed it.
On this point, when I modify or correct a CK from .fr, I also look at the CKs in the other languages ​​and modify them if necessary (if it's just a question of date and place of birth or death + occupation and studies. And I do that in English or German for example). I created several profile (fr, com, de, it, ...) to have access to CK as date of death)
However, I do not change CK which are strictly local : some awards (but I change if it is a question of booker prize, nobel and other international grand prizes) ...

xxxxxxxxxxxxxxxxxxxxxxxxxxx

About multiple pages like Chandler, I'm starting to understand aliases better.
However, I still don't quite understand how to make sense of the alias.
For example how to aliased David P. Chandler to David Chandler
rather than
David Chandler to David P. Chandler.

xxxxxxxxxxxxxxxxxxxxxxxx

Beyond this last point, being able to have several pages could be the subject of further development.
But for the moment, it is so and it is enough to understand that to find the information, David P. Chandler could be more complete than David Chandler (x)

However, I complete the CK David Chandler (x) because - beyond the note of disambiguation - by opening the different David Chandler (x), in other windows, if the CK and the http or still photos are present, we observe the different authors quickly and we can then attribute the works with care.
Maybe there will be an IT development in the future (already wrote) ?

Once again, thank you for your answers. When I have a next case on the meanings of aliases to make, I will ask you to do it perfectly

>35 JMK2020:
1. Open the Chandler split page https://www.librarything.com/author/chandlerdavid
2. Go to "edit the division"
3. Look at the page. It opens on "Assign works to authors" but there is a second tab in there called "Alias divisions to other authors". This is where you alias.

You always alias a split part (aka an author-x) into the author page but you do it from the split author page. So we say that Chandler-2 is aliased into P. Chandler BUT you do it by going to Chandler and telling it that its "-2" part belongs to P. Chandler.

Does this make more sense?

>36 AnnieMod:
OK. GIANT...
Perfect, Super ;-)
Thanks

>33 JMK2020: Regarding the different versions of "Paul Smith", for example, being attached to the same work, sometimes that is how the records are entered. It can be deliberate (the author chooses to not use a middle initial in an early printing, but then adds one in a later printing to distinguish from another "Paul Smith"). Or it can be accidental, that a particular data source or user enters the author as Paul Read Smith, Paul R. Smith, P. R. Smith, Paul Smith. Within that work, given these conflicting options, this site will automatically select an author name. I'm not totally sure how it chooses but I think it has something to do with which name is most commonly present in the records (most different entries, or most different copies entered by users).

Most of the time that's fine and can be left to handle itself. However, in certain cases it may be best to override this, if the automatic selection chooses a clearly wrong name (like the cover artist instead of the main author), or a garbling of two names together, or just nothing. You can do this on the work page in the "Other Authors" section by clicking "Add Other Authors" ("Ajouter d'autres auteur(e)s"). Once in there, you can edit the current main author by clicking on the pencil icon, entering and saving the name. This is also where additional authors can be added where necessary (co-authors, various contributors to anthologies, and so on) using the "Add" option there.

I had thought these were the same across different language versions of the site. Common Knowledge is language-specific but I think the author names are shared. I may be wrong on that though.

As for the relationships between authors, there has been some discussion of splitting and aliasing here, but for a little broader information... Where an author has published under two different names that are unique, those two author pages can just be combined. For example: Mark Twain, Samuel Langhorne Clemens, and Samuel L. Clemens have been merged, because they all refer only to the same person and not to any other. However, "Samuel Clemens" is not unique; aside from Mark Twain there are 2 others by that name in the LibraryThing database. So it cannot be merged into Mark Twain. Instead the Samuel Clemens page has been split into those three authors, and then the correct one has been aliased into Mark Twain. (Split-author components are not merged, they are aliased only.) The other two splits have not been aliased anywhere because they don't have another name to alias into -- they publish only as "Samuel Clemens".

To try to make clear which split is which, the Disambiguation Notice down at the bottom of the page can be filled in with brief information on the different authors (such as full name if longer than the split name, or birth/death dates, or nationality or occupation).
https://www.librarything.com/author/clemenssamuel

More detailed information can be entered on the individual author splits...
https://www.librarything.com/author/clemenssamuel-2

Or on the page which they are aliased into...
https://www.librarything.com/author/twainmark

Where there have been merged author names or aliased author names you can see them at the top of the author page. For example that Mark Twain page lists several variants on his name such as the Samuel Langhorne Clemens, and then also notes that it includes the aliased split-author "Samuel Clemens (1)".

Thank you very much for these very clear explanations with a significant example (M Twain)

Regarding the names to add to the works themselves (translator, photographer, ...., other authors, ...), I understood how to do it and already practice (for example for all collaborative works). I just wanted to understand the form to give to the name when there is ambiguity and the consequences vs LBTh.

xxxxxxxxxxxxxxxx

Example for this author (today) : Serge Andolenko
Let's call this textual form (A)
https://www.librarything.fr/author/andolenkoserge

The form of the name in the authority base fr (France) fom BNF (Bibliothèque Nationale de France, Nat. Library of Fr.) = Sergej Pavlovič Andolenko (B)
https://data.bnf.fr/fr/12788521/sergej_pavlovic_andolenko/

When I imported my book (from the BNF site), it was (B) that was imported into my library but also into LBTh ("other authors" frame)
https://www.librarything.fr/work/26986762

(B) not corresponding to what is indicated on the book (and not being very expressive), so I changed the name in my library and also in the book sheet to indicate (A)

In addition, in the CK (fr) author file, I added 2 other forms of names (B but also the Cyrillic / Russian version) AND put (A) in canonical name for the .fr and .com version.

Finally, I looked to see if it could be associated with works by the same author with the multiple textual variants. For this rather local author, the case is simple.

Did I do well?

xxxxxxxxxxxxxxxxxxxxxxxxx

Beyond that, there are several language-specific CKs (not modifiable from another language): Gender, Date of birth, death.
Tim listed them in another post. It seems to me that there are 5.

About this point, when I complete CKs, I always look at what is already present in other languages ​​to possibly supplement or correct (with discernment) **.
I have multiple LBth accounts for each language to get it right.

** especially the dates and place of birth / death. For the other CKs, ... I would have to live several lives .... but I put aside these authors to come back to them maybe one day (and match the CKs)
Example of differences between CK and languages ​​(fr vs com / Education, Relationship, Award ...)
https://www.librarything.fr/author/zolaemile

xxxxxxxxxxxxxxxxxxxxxxxxx

If I ask all these questions and try to understand how LBTh works, it's because sometimes I take a long time to search, find, and share CKs (or make the right connections between works). Zola is not the most complicated case - when you know it - but it takes a lot of resources and time.
So I spend a lot of time but have already observed that pages of authors or CK disappear ... and we do not always find the histories. It may be due to combinations or other actions of members ... It sometimes discourages me

Is the Canonical Name you entered any different from what LibraryThing would display if you didn't enter it? If so, it's unnecessary.

And will probably make trouble in the future. Please do not add canonical names. There are very, very few times when they are justified.

>39 JMK2020: "AND put (A) in canonical name for the .fr and .com version."

WHY? Which part of my explanation up in >32 AnnieMod: "Please never set Canonical name UNLESS the name in the DB is really wrong. When you set these, they hide issues and lead to bad combinations unless people pay attention (and most people just look at a name and do not bother with the URL - and that causes issues)." was unclear? :)

>40 r.orrison:
Yes

>41 MarthaJeanne:
>42 AnnieMod:
For this author
https://www.librarything.fr/author/andolenkoserge

Here is the form of the name that was displayed as author in the book and in the author file in LBth

Sergej Pavlovič Andolenko (A)
(this form is from BNF !!!!!!!!!!!!)

Loc : Andolenko, Serge (B)
https://id.loc.gov/authorities/names/nr93044551.html

Nat Bib.nl : Andolenko, Serge (B)
http://data.bibliotheken.nl/doc/thes/p130634778

....
And perfect official from gov : ANDOLENKO, Serge (B)
https://www.leonore.archives-nationales.culture.gouv.fr/ui/notice/5107

I don't understand why keep (A),
If I look for the author page in TBth or try to find some works, I 'll use "Andolenko, Serge" and not "Sergej Pavlovič Andolenko" (I repeat here : (A) was the name in my library (import), the name as author in LBth for the book and the name of the author page... that's not the common or canonic form...

Thus, it is not so much the name which is false but the strange form given by the BNF which is seen in LBth because of the importation of a book.
What is certain is that by importing from AMZ the name would have been (B). But i don't use amazon as primary sources (and more generally no merchant site) and just if it is necessary to find some data (that i verify)

xxxxxxxxxxxxxxx

NB : I saw and understood 32 and 42 ;-)

xxxxxxxxxxxxxxx

For a next ambiguous or strange case, I will ask you what to do

xxxxxxxxxxxxxxx

>41 MarthaJeanne:
>42 AnnieMod:
A few mn after
OK OK,
it's just a question of combinaison rather canonical name
https://www.librarything.fr/author/andolenkoserge/names

https://www.librarything.fr/commonknowledge/changelog.php?item=110586&type=2...

Sorry

>43 JMK2020: Except for very small authors coming from really bad sources, the authors’ names tend to sort themselves once all combinations are done. Think of canonical names as last resort, not as an easy way to fix everything. It looks fixed. Until it comes to bite us :)

>43 JMK2020:
Russian names in particular don't lend themselves to being standardized in one "official" form the way you would like.

"Andolenko" is actually simpler than most because it doesn't include any of the Cyrillic letters that don't map 1:1 to the Latin alphabet. Most other Russian names will inevitably appear in a variety of spellings because the conventions for rendering Cyrillic differ in different languages. And just in English there are at least three different transliteration systems in use!
Take Dostoevsky for example: possible spellings include Dostoevsky, Dostoyevsky, Dostoevskii, Dostojewskij, etc. All of these are correct and legitimate forms of the name -- depending on the particular context. Conversely, any form you chose as "correct" will inevitably look "wrong" to someone else.

In addition, the Russian name system (given name, patronymic, surname) may be adapted somewhat to local conventions when published. Russians use either the full name or, frequently, two initials + last name (F.M. Dostoevskii). English or German or French publishers are more likely to use the more familiar form of given name + surname (Fyodor Dostoyevsky etc.).

Which is "right"? I suppose one could argue that the "native" form is the most correct, which would give us Достоевский, Федор Михайлов. But if the Russian form is therefore listed as the canonical name, I think a lot of users would be understandably annoyed to see the name presented in an alphabet they can't read.

By the way, you don't need multiple accounts to use the different language versions of LT -- you have to log in separately, but you can do so using your regular account; accounts aren't inherently linked to a particular language site.

>43 JMK2020: The author name on LT is supposed to be the one most recognizable by most people which LT determines by the name most commonly entered in LT. You shouldn't override that unless it's completely wrong, not just that it's in a form that you don't like.

>45 spiphany:
I understand and pratice with vietnamian or other transliteration. And sure, the canonical form here or there, isn't canonical name, there ou here... Babel is and stay a myth ;-)

Just one account ? Is'nt it a limit to change 4 or 5 CK as date of birth/death (if false in another language) ?

xxxxxxxxxxxxxxxxx

>44 AnnieMod: : OK
So Now that I have understood a certain number of actions (above all with organisation process) .... I think I will have to go over a certain number of author files that I have completed ..... !!!!
it's like that ;-)

xxxxxxxxxxxxxxxxx

>46 jjwilson61:
OK, I saw 32 and after

New problem with Louis Carré / Louis Carr (Combinaison)

https://www.librarything.fr/author/carrlouis (.fr)
https://www.librarything.com/author/carrlouis (.com)

Louis Carré is for 2 Authors

Louis Carr.... ???????????????

I separated and test with this or that but i can't exclude Carr from Carré ?

Help Help

I'll wait before complete CK and http links, pictures...

This is not a problem. Author page URLs are formed by the valid letters in their names. é does not count as a valid letter, so the URL is the same for Carré as for Carr. The author page has been divided, giving separate pages for three authors.

Similar with Robert A Smith and Roberta Smith, Karl Möller and Karl Müller. This just how it works. Tim would like to find a different way of dealing with it, but it's complicated, and this works.

>49 MarthaJeanne:
Tx.

I was not sure there's no way to change this Big probleme "é" + "Similar with Robert A Smith and Roberta Smith, Karl Möller and Karl Müller."

I can't help for software but it's sure there is smth to do.
There is here an example from a french site
Seach = "Louis Carré" for decease from 1970
an (1) after others crossed metafdatas is
qkGmpDmuwbHf
https://deces.matchid.io/id/qkGmpDmuwbHf

But Sudoc, ISNI, ......... have an idRef clear (When metadata are OK)

Resume : LBth had to change the id Ref (and take care with wikidata Ref...) or more complete wikidata