Author Combinations, Splits, and Aliasing

Two conversations are happening simultaneously, both in topics that were about one member's choices. That's not only confusing, but one topic page is too long while attempts to fulfill the purpose of the second are being buried.

Let's bring the conversation about best practices for author combinations together, yeah?

Here's an example that hasn't been in an edit battle, to my knowledge. We don't have to have to come to any consensus about this before bringing in other examples. I'm only throwing it to the group as if came to my attention tonight.

From records elsewhere online, David A. Robertson appears to be the same as David Robertson (24).

https://www.librarything.com/author/robertsondavida
https://www.librarything.com/author/robertsondavid-24

What do you think should be done with the two author pages, and why?

>1 aspirit:

First off, I think that there is more than one author on the David A. Robertson page. One writes about First Nations people, one is a mountain climber, there's the author of Linguistic Evidence in Dating Early Hebrew Poetry and someone who wrote The Old Testament and the Literary Critic (these may be the same person).

So that page needs to be split.

Then, I'd alias David Robertson (24) into David A. and assign his works to the appropriate split on the David A. page.

As a rule, I think that splitting and aliasing is vastly preferable to combining and then splitting.

I said in another thread: If the John Smith page is split and aliased, then someone who has entered "John B. Smith"'s book as by "John Smith" will get to that page, and see that their author also uses "John B."

But if the pages are combined and then split, someone who has entered the book as by "John B" will be led to the wrong page.

In addition, in cases such as David A. Robertson, you might have one David A. who also uses just David Robertson, but what about the other David As who don't? They'd wind up on a page with an author name they don't use!

Furthermore, we have a guideline not to combine "surname only" or "first initial/surname" pages with a full name, precisely due to the confusion that would be generated when there are multiple authors sharing a surname or an initial & a surname. It seems to me that the suggestion of combining and then splitting David and David A. contradicts that. It certainly would create confusion.

Edited because I split an infinitive. I'm so ashamed! :-(

I completely agree with lilithcat here. I was actually under the impression that the majority of combiners operated this way (split / alias), and I was surprised to hear Tim express a preference for combine / split in the other thread.

Thanks asprit, for putting this conversation into a separate thread.

I think my first reaction to Tim's preferred method was more or less: "on the one hand this, on the other hand that."

I've thought about it a little more, and I'm inclined to what appears to be the consensus view (with the exception of Tim, of course). I think it works better in practice, and I generally prefer an author name that adds information rather than leaves it out (we don't all go down to look at what, if anything, CK says about the author's legal name or alternate names).

That being said, I think there are probably cases where we need to avoid an absurd result. I gave the hypothetical of George Eliot being aliased to Mary Ann Evans because of a second George Eliot; real examples may pop up from time to time.

So I'd say we should use the split/alias method in general, and use Tim's (combine/split) only when an obvious infelicity would result from split/alias.

Here's an example that's been bothering me that I would like Tim's take on, but I'm also concerned about what that would be.

Author Stacey Abrams has been combined under her Romance byline Selena Montgomery. I saw Collectorator in the edit history a little while back so let it alone, unwilling at the time to get into an argument over an author who publishes political nonfiction (Abrams) and romantic fiction (Montgomery), two genres that get people's blood going merely by existing.

The problem from a record-keeping view, however, is that having Ms. Abrams and her nonfiction works show up under her fiction byline makes no sense.

A split then aliasing would be better in this case wouldn't it?

>5 aspirit:

A split then aliasing would be better in this case wouldn't it?

There's nothing to split, though. There is only one Stacey Abrams and one Selena Montgomery.

The better solution, frankly, is to make Stacey Abrams the canonical name. This is one of those instances where, I think, having a canonical name is appropriate.

>5 aspirit: But there's nothing to split here, it's all a single author. Besides, the one that would serve as alias would again list all works, both romance and NF. BTW, why is her romance alias winning when NF works are more numerous?

Possible solution for authors where standard practices for combining, splitting and aliasing do not work that well would perhaps be to use alt name field to link to another author page the way relationships field does now. The field is not currently coded to do that (it goes to CK history), but I assume it's a minor adjustment since it exists already

>6 lilithcat:

I agree with the canonical name solution here.

>7 SandraArdnas: the canonical name was repeatedly deleted.

Changing the canonical name and having it left alone would correct the problem (on multiple levels) of having

Selena Montgomery
Author of Lead from the Outside: How to Build Your Future and Make Real Change

and a pseudonym likely unknown to political nonfiction readers show up in author touchstones. (I don't remember where I saw a link for "Selena Montgomery" that confused me, but it happened.)

I have other concerns, though. One is that the listing of works (actually) by Selena Montgomery is empty because of the combination, for some reason. I'm not confident that will repopulate when a canonical name is entered for both sets of works.

>9 aspirit: I was referring to the fact that her real name should be chosen even without the canonical because her NF works seem significantly more represented in member catalogues. It seems odd that the system chose her pseudonym as default at all. 'Lead from the Outside: How to Build Your Future and Make Real Change' alone outweighs her fiction works

Edit: regarding listing of works under Montgomery, they appear under 'examine and separate author names' as belonging to Selena Montgomery - main author, but somehow don't carry over when you click a link for just works under that name. It's all odd.

|-_-| (#3):

I was actually under the impression that the majority of combiners operated this way (split / alias), and I was surprised to hear Tim express a preference for combine / split in the other thread.


Tim's been wrong before. He knows the code and the data model, but sometimes power users like active Combiners are more familiar with the actual way the site works for people.

>5 aspirit:

I have combined Selena Montgomery and Stacey Abrams, and forced the name to be Stacey Abrams. It's basic to the author system on LibraryThing that real and pen names are combined. If an author is known primarily under their pen name, such as Mark Twain, that can be the winning name. In this case, I have changed her canonical name to Stacey Abrams, because she's published under both and it's weird and confusing to do otherwise.

The David Robinson example is more complicated than it first seems, because David (A.) Robinson isn't one author. So the method everyone wants to use isn't combine and split, it's combine and split one author, combine and split the second author, then alias one split into another split, at least.

Does someone have a simpler example?

Tim's been wrong before. He knows the code and the data model, but sometimes power users like active Combiners are more familiar with the actual way the site works for people.

Maybe. I think a lot of problems people complain arise from not using the simple tool, but aliasing from bucket to bucket in multiple jumps.

In either case, the choice to prefer longer, more specific names that the author rarely or never uses as against the names they use on actual books isn't a data one, it's a matter of taste.

Anyway, we need some simple examples.

>13 timspalding:

The David Robinson example is more complicated than it first seems, because David (A.) Robinson isn't one author.

Not to mention that he has a slew of books listed under David Alexander Robertson.

>14 lilithcat: We could, like, combine them and then split them out.

>15 timspalding:

Funny!

This is what I want to avoid.

Take this guy, the popular author of a slew of relationship books:



The disambiguation page is here: https://www.librarything.com/author/chapmangary

He's then aliased away into "Gary D. Chapman." Here are some of Gary "D." Chapman's books.



As far as I can tell, Gary "D." Chapman exists nowhere in the real world. No book of his is ever credited to that name. His Wikipedia page doesn't use it. His website doesn't use it. His Facebook, his Twitter--nothing uses it.

Sorry, but this is exactly what the LibraryThing system was created to prevent. Libraries "solved" the problem of multiple authors by changing author names until they became unique. They did this because, in an age before databases and unique identifiers, this was the only way to make things unique. Accordingly library technique adds an initial, then a middle name, then dates, then places of birth--anything to make a unique set of letters.

Today, in the 21st century, we don't need to play this ridiculous 19th century game. We can call authors by the names they actually use.

>18 Collectorator:

Strictly speaking, however, the author calculation doesn't go work-by-work, but book-by-book.

>17 timspalding: For what it's worth, US libraries don't choose the primary name on an authority record that way any more, and haven't for quite a while.

I personally prefer having authors split to unique names so they show up as having a gender and being alive or dead in my statistics, and so they can be linked from relatives who are also authors. Could you please address this point of view?

>17 timspalding: But we don't really have unique identifiers for authors, hence the need for splitting and aliasing. With a unique identifier this particular Chapman would have his own page and we would combine all relevant variants into it, end of story.

In the existing system, we combine a more specific name with a less specific one when both are for the same single author - there's no need to split and thus no danger another author ends up with incorrect name. If a split author, we alias to a more particular name. Sometimes, that is the form most often used (so the page already contains most of the works), sometimes rarely, but it still serves as a 'unique' single page for the author.

In case of Gary D.Chapman, his most prominent work is catalogued by over 7,000 people and most of them entered it with the middle initial, so clearly this is relevant data by numbers and usage. While I can't comment further on that case because I really don't know who and when uses the middle initial, I can comment on some of mine. Many of my authors are displayed with no middle initial on the cover and title page, but libraries catalogue them with the initial aND the short bios/about the author sections include either the initial or full middle name

>20 NinieB: I personally prefer having authors split to unique names so they show up as having a gender and being alive or dead in my statistics, and so they can be linked from relatives who are also authors. Could you please address this point of view?

If these stats don't work right, we need to fix them. We don't choose a bad overall site principle because a stats doesn't work right.

But we don't really have unique identifiers for authors, hence the need for splitting and aliasing. With a unique identifier this particular Chapman would have his own page and we would combine all relevant variants into it, end of story.

The author has a unique identifier and page when you combine and split. Here, for example, is Steve Martin's. He's Steve Martin 1, and his page is https://www.librarything.com/author/martinsteve-1

In case of Gary D.Chapman, his most prominent work is catalogued by over 7,000 people and most of them entered it with the middle initial, so clearly this is relevant data by numbers and usage.

That is simply not true. Here are the editions. Very few of the editions use his middle initial. The vast majority of his editions do not use it. Rather, someone forced him into Gary D. Chapman by setting the primary author.

>17 timspalding: Today, in the 21st century, we don't need to play this ridiculous 19th century game. We can call authors by the names they actually use.

But a lot of the library sources we draw from still use it -- and as a result the data here has it. Unless this is nicked at the Add/Import level and fixed in the older records, it does not matter what the authors themselves (or the books) use really; what matters is what the sources do - not everyone will update their records to what the books are saying once they add/import.

I'd love not to have unique identifiers and never need to split and/or alias -- but it is what it is.

>23 AnnieMod:

Most LibraryThing data doesn't come from libraries, though. As seen above, most of the data doesn't have these extra bibs and bobs.

>22 timspalding: The author has a unique identifier and page when you combine and split. Here, for example, is Steve Martin's. He's Steve Martin 1, and his page is https://www.librarything.com/author/martinsteve-1

But that is not a fully functional author page like the unsplit ones. It's not just the stats (which apparently don't look beyond the main split page, because they claim gender and living status unknown even when all splits have the info, e. g. alfred adler). You can't link back to them in relationships, you can't touchestone them, you can't combine that page with another one without dragging all the other splits in too, etc.

That is simply not true. Here are the editions. Very few of the editions use his middle initial. The vast majority of his editions do not use it. Rather, someone forced him into Gary D. Chapman by setting the primary author.

Sorry, I didn't check whether primary author was edited or calculated. Still, there's over a 1,000 Gary D.s there. More than enough not to be discounted as irrelevant. So the question is what is the best way to organize and represent that author given the usage.

edited for clarity

>17 timspalding: This could potentially be addressed by just giving "Gary D. Chapman" the canonical name Gary Chapman...

>24 timspalding:

You may want to count the ones with D without the dot while looking for the D ones.

Still the non-initial ones are more but for some authors we are close to it. But that is not the case for some older less reprinted authors from what I had seen.

Michael Francis Gilbert is another case in point.
https://www.librarything.com/author/gilbertmichaelfranci
Every book listed on this page (well, at least all of those which I have) has the name Michael Gilbert on the title page. This author's middle name was indeed Francis, so this isn't "wrong", but it was never used as an authorial identity. But, because libraries and other sources have used this middle name to disambiguate this Michael Gilbert from others, it's stuck. And, now Michael Gilbert (1) is aliased into Michael Francis Gilbert.

What irritates me also, is that when I click on the link which says "You have 103 books and 103 works by Michael Francis Gilbert" it brings up no results, because in my library I have no titles by that author.

Surprised that the same thing hasn't happened to Stephen Edwin King!!

>28 woolly: Your author link shouldn't lead you to Michael Francis Gilbert if you've entered it Michael Gilbert. It should lead to a split page

>22 timspalding: Thank you. And as >25 SandraArdnas: states, I agree that the split pages need to be improved so that they are as functional as unsplit pages.

>29 SandraArdnas: Correct, it leads to the split page which says "Works have been aliased into Michael Francis Gilbert" which I follow in case titles have been given the MFG author. For example, I have the title Death in Captivity where the author has been fixed (algorithmically?) as MFG and doesn't appear in my list of titles by Michael Gilbert.

We're going to have to decide on the best system and THEN make secondary features work as we want them to. We can't work backward.

>31 woolly: I've recalculated the author. Does it show properly now?

>33 SandraArdnas: Sadly not. See: https://www.librarything.com/author/gilbertmichael-1. There should be 129 copies (see the author page for Michael Francis Gilbert).

>34 woolly: Death in Captivity should show properly under Michael Gilbert 1, now, it does for me. The rest I haven't checked, but for that one someone changed the primary author from the calculated one so it ended up ONLY on aliased page. If you have other books than need such recalculating, feel free to open the thread here in combiners and people will chime in and help resolve possible messes

>22 timspalding: "If these stats don't work right, we need to fix them. We don't choose a bad overall site principle because a stats doesn't work right."

We've been asking you to fix this for years! See, for example: http://www.librarything.com/topic/257394#unread

>35 SandraArdnas: Thanks SandraArdnas. It does for me too.

Here's a case that resembles the Gary Chapman example discussed above.
There are several authors named Brian Ellis. One of them is a philosopher who published as Brian Ellis, but in LT, many of his works are attributed to Brian David Ellis (his full name) or B. D. Ellis. To my knowledge, he never published under (or otherwise used) either of those names.

A split / alias strategy makes sense. But right now, the alias goes to B. D. Ellis, which, again, is not a name he was known by. An imperfect but straightforward solution would be to enter "Brian Ellis" as the canonical name on the B. D. Ellis page.

I'm not in a rush to address this; I just thought it was a good case study.

Just noticed that an Australian library named after him has an LT account. It's the Brian Ellis Library, not the B. D. Ellis Library :-)
https://www.librarything.com/profile/BrianEllisLibrary

>22 timspalding:

I think this debate involves a few more-or-less separate factors:

1. Aesthetics/preferences.

Some prefer a longer form name, especially over a generic name. It adds clarity and specificity. And it's not purely aesthetics: it's easier to recall that the author of books on Arab culture is James Alexander Campbell than James Campbell (3). I understand the force of Tim's argument for authors whose names I've heard so often they're set in my mind - I wouldn't want names like Agatha P. Christie, Charles W. Dickens or Jane Myra Austen to be the end points for familiar authors. But I'm not sure if anyone has explained why a canonical name would not be an adequate solution for this problem - it seems like it's a pretty complete answer to me.

And about that "(3)" after the split names. I don't like it generally, and I wouldn't be surprised if it's part of the reason so many editors prefer aliasing when there's an opportunity. It's obnoxious. The "D." in Gary D. Chapman has a good deal more right to be there than a (1) after his name.

2. Work flow

The split/alias model seems to be a better fit with how I work. I generally only do author combination/splitting/aliasing when I run into someone that looks wrong, and I'm often not trying to do the complete job. Combining the variants of a particular name is an easy first step - and I might not even be aware that there are works under a shared name, particularly if it's an obscure author. If, at some later point, someone decides to combine "Charles Alexander Hamilton" into "Charles Hamilton", all the links, CK and so forth that have built up have to be transferred.

Another difficulty with the combine/split model is that it requires editors to have a clear understanding of the balance of the particular versus the shared name at least with respect to how they're assigned to the works. But this requires (1) combining all the shared name variants; (2) combining all the particular name variants; (3) figuring out which works under the shared name really were authored by the particular author; and (4) assessing whether the shared name or the particular name is predominant. And (5) - if I understand Tim correctly - if the particular name is favoured by the data, we should verify that in real life (and on title pages) the author actually used the particular name.

This strikes me as a complex and comparatively lengthy process. On the other hand, split/alias allows editors to proceed more piecemeal.

3. Problems

The aesthetic problem of a Charles W. Dickens could also be regarded simply as wrong. I think we should be flexible to avoid such outcomes, using Tim's methodology to avoid such pitfalls with famous authors. But where it's unavoidable, is the canonical name solution inadequate?

I believe r.orrison raised a problem with the combine-and-split approach that I don't think has been addressed: when a particular name is the most common name, you get all "James Campbell" under a "James W. Campbell" page. I suspect this won't be too common a problem if people are careful to alias when a particular name is the dominant name - and I am not sure if that's Tim's preferred policy. If you take "combine-and-split" as a general approach, this becomes a much more common problem.

So, that's about where I am. I think we ought to avoid the absurdities that Tim warns about, although I don't think Gary D. Chapman rises to that level. But in most cases, I think the split-and-alias approach makes more sense to work with, risks fewer problems and creates a better final structure.

In the end, for me, I am not willing to combine John W Smith with John R Smith and John G. Smith.

Admittedly, many cases are only about one or two different authors, but whatever we come up with has to also work for the cases where there are several different authors with similar names.

>41 MarthaJeanne: In the end, for me, I am not willing to combine John W Smith with John R Smith and John G. Smith.

This. The moment we end up in this situation, something in the system had gone very very wrong.

Could someone check my work and tell me if I did things right? Devorah Blachor wrote one nonfiction book under her own name and several mysteries under the name Jasmine Schwartz. I split Jasmine Schwartz but assigned all the works to one author so that I could alias Schwartz to Blachor. Or something. I went back & changed my mind & got confused a few times. Anyway, does this seem right & if not, could someone explain what I should have done & how? Many thanks to you all. And, yeah, I wholeheartedly agree with >41 MarthaJeanne: & >42 AnnieMod:

https://www.librarything.com/author/blachordevorah

>43 susanbooks:

No, please do not split a single author.

If there are no other authors using either name, you can combine the two pages.

>44 lilithcat: Thanks! When I combined the 2 pages it made Schwartz the main page (that name has more books) but Blachor is her real name and it's the name under which she published the book with the most copies in the system. So when a user looked for the author of Blachor's book, they'd come to the Schwartz page (even though the latter is a relatively obscure mystery writer while Blachor is a NY Times columnist, McSweeney's writer, etc). Is there a way to make her real name the system default (without doing the ridiculous thing I did)?

>38 I-_-I: I suspect that canonical would harm rather than help. It shows everywhere as such and would sooner of later be combined with the split Brian Ellis. Besides, none of those members entered it as such. Those who entered it Brian Ellis end up on a different page

>38 I-_-I:
>46 SandraArdnas:

There is a lot of B. D. Ellis in the edition data, maybe not a majority, but enough that it certainly doesn't look like someone forced Brian Ellis to B.D. Ellis. Might be a case where Ellis uses a different name on the cover than on the colophon (and maybe the title page).

I would think a disambiguation notice would suffice if there's a danger of recombination.

>47 gabriel: Disambiguation is not seen at all unless you open the author page. Unfortunately, there's no warning when trying to combine as with work combinations. If you just do a search for combinable authors from Brian Ellis page and the results show you another Brian Ellis not yet combined, clicking combine will immediately do so without an intermediate page where DN would show. OTOH DN on Brian Ellis page saying 'do not combine with Brian Ellis' isn't very meaningful. Once a canonical is set the only place you can see anything else site-wide is the page itself among other names.

All of the issues we're discussing here would vanish if split authors actually got a unique page of their own, not tied to the other splits. If the main split page were to function as merely disambiguation page, where non-assigned works end up and we do the splitting of various authors, but once split they get an independent page that can be combined with their other aliases. I have no idea how feasible it is code-wise, but it would make combiners work much easier, while securing that all members go to the correct author page with all thew works no matter whether their entered author name is a split or aliased one.

This is even more direly needed for institutions and editorial offices as authors.

>47 gabriel:

At least some academic journals (esp. in the sciences?) have a blanket style rule of reducing author names to XY Surname. At least I know that my father, who went by J. William Lloyd (or if forced to, by James William Lloyd), was always transformed into JW Lloyd when he published in epidemiological or statistical journals. Same for his co-authors. Possibly some of Ellis's journal articles came out this way..

>50 Crypto-Willobie:

Possibly some of Ellis's journal articles came out this way..

I know absolutely nothing about Ellis beyond what I can see on LT, but it's not just a couple obscure articles. His most common books have plenty of "B.D. Ellis" or "Brian David Ellis" in them - around half of the editions on the works I checked.

It might be originally from library data that decided to distinguish Brian Ellis from the other Brian Ellises, but wouldn't then the normal procedure be for him to be "Brian D. Ellis", from what others have said?

I guess this Brian Ellis discussion just goes to show that there are no simple solutions. I'm fine with not going the canonical name route for the reasons given above... but the fact remains that it's absurd for Brian Ellis' LT "homepage" to be under B. D. Ellis, a name not generally (and certainly not primarily) associated with him.

>52 I-_-I: Ah, but there isn't a single homepage. For those who entered him as simply Brian Ellis, that split is the home page. It's where clicking on author name leads you. Only those who entered him as B. D.Ellis (or Brian D.Ellis) end up on the other one.

>53 SandraArdnas: right... good point :)

>40 gabriel:

Thank you for this long and interesting response. I have just left a lot of comments on the other topic, so I shall be brief here.

To your factors I want to add a few:

1. Allowing the system to work vs. forcing it forever. The point of the system is to get better as more data is entered. I want that to happen. So I view forcing primary authors and canonical names as heavy-handed solutions, to be employed only when truly necessary.

2. Canonical names and name variants. You write "The aesthetic problem of a Charles W. Dickens could also be regarded simply as wrong. I think we should be flexible to avoid such outcomes, using Tim's methodology to avoid such pitfalls with famous authors. But where it's unavoidable, is the canonical name solution inadequate?" I guess my problem here is that if there's a page for Charles Dickens, that should be the page for Charles Dickens. I don't think the canonical names should be used to have one page for Charles Dickens that's split and another for Charles W. Dickens that's been canonical-named into Charles Dickens.

3. Features vs. data. I do want members to tell me things that aren't working. If they are causing members to make data choices--if people are twisting the data to make the link work here or there, that needs to be fixed so we can do the data correctly.

4. Martha Jeane writes "In the end, for me, I am not willing to combine John W Smith with John R Smith and John G. Smith." I understand that and in most cases would agree. But If the John Smiths are mostly using "John Smith," I want John W. Smith aliased into John Smith (2), or whatever. I don't want John Smith (2) aliased into John W. Smith. As I've said before, I want the system to use the names people really use.

But the system doesn't let us alias into a split. If that's what you want to happen, you have to make it possible.

>55 timspalding: I want John W. Smith aliased into John Smith (2),

We would all love that, and it would solve many of the issues raised in the recent threads. Currently, you can't alias (or combine) into a specific split. You're the developer, make it possible.

I was going to do a little research on disambiguation pages so I searched for John Smith but the link on the search page goes to a Doctor Who page, specifically https://www.librarything.com/author/doctorwho-316092. What's going on?

I found a John Smith disambiguation page at https://www.librarything.com/author/smithjohn

We would all love that, and it would solve many of the issues raised in the recent threads. Currently, you can't alias (or combine) into a specific split. You're the developer, make it possible.

If members want John W. Smith to be part of John Smith, they can do it. They refuse to do so because they perceive a problem with John W. Smith being combined into John Smith at the same time as John C. Smith is combined into John Smith. Although on the system level, it's merely aggregating names, which are then separated into people by the split process, members feel that it's aggregating people.

Perhaps we just need the ability to alias something into something in a way that is visually and terminologically separate from combination.

So, John Smith and Smith John would be "combined," but John W. Smith and John E. Smith would be "aliased" into John Smith.

I don't see the point in aliasing into splits directly. As the aliasing code has it:

"Be sure to find the exact author code you wish to alias the division into. You can only alias to the "top level" of an author (eg., "adamsjohn" not "adamsjohn-1").

This is to allow the splitting operation to be free, and allow restructuring of author divisions as necessary. If John W. Smith could go into John Smith (2) directly, the UI would have to lock certain works on John Smith (2). I think this would make split assignment a lot more annoying to do.

>60 timspalding: By that rationale we could also combine John, Joseph and James Smith because all are J. Smith. So, yeah, I don't like that idea.

>60 timspalding: If members want John W. Smith to be part of John Smith, they can do it. They refuse to do so because they perceive a problem with John W. Smith being combined into John Smith at the same time as John C. Smith is combined into John Smith. Although on the system level, it's merely aggregating names, which are then separated into people by the split process, members feel that it's aggregating people.

It's really not about mere perception. It's a practical problem. With disparate names all combined into the split author page, usually one of those will surface as 'majority' name and be the default for all splits. This results in listing along the lines:
John W. Smith (1): John C. Smith
John W. Smith (2): John Pinkleton Smith
John W. Smith (3): John Smith etc.

Everina Maxwell has been combined into the Avoliot author page. The problem is that with the upcoming release of Winter's Orbit, the Everina Maxwell byline is more frequently entered (currently three times as often on LT) than the name for the author page. I'm confused from the discussion so far about how to handle multiple names. What might be the best way to handle this case? Separate then recombine, split then alias, or some other way?

Is this the same person? If so, I would try separating and then recombining. If not, just separate.

>64 MarthaJeanne: Yes, same author. Her novels might also be the same work. Separating then recombining fixed her name. (Thank you!)

I'll watch for reviews of the new release to see if it's different enough from what she had published on Archive of Our Own for the works to remain separate.

>65 aspirit: You should probably add the other name to CK.