• LibraryThing
  • Book discussions
  • Your LibraryThing
  • Join to start using.

splitting & aliasing vs. changing primary & other authors


Join LibraryThing to post.

This topic is currently marked as "dormant"—the last message is more than 90 days old. You can revive it by posting a reply.

Edited: Nov 7, 2011, 12:59pm Top

Since Tim released the new feature allowing us to add other authors and change the primary author, I wanted to open up a discussion about when and how we should use this new feature, rather than or alongside the "aliasing" feature we already have. I know people have sometimes been using splitting and aliasing in order to get books to show up on the right author page (sometimes too for combining purposes), but that's resulted in some strange things: single authors that are split (previous discussions here; example: Henry Wadsworth Longfellow), and author aliases that point to a completely different name/author (see for instance these aliases for "Marks" or the alias of Solomon-15.

With the capability to add other authors and change the primary author, it seems like the need to split and alias has disappeared in some cases. Adding other authors will get the works on all authors' pages. Changing the primary author will also move the work to the right author's page, and take it off an incorrect author's page (so if a work had been listed on a last-name only Smith page, it will disappear from there once the primary author is changed to Bob Smith, and show up on the Bob Smith page).

So when should we use one feature rather than the other? When should we use changing primary & other authors, and when should we use splitting and aliasing? What are the guidelines that should apply? Since changing the primary author is "serious business" according to the instructions, what kinds of cases warrant it? How do we decide who should be the primary author? What if someone sets the primary author to something others don't think is correct?

ETA: Also, how should we deal with previously split & aliased pages (and sometimes, CK on those various split pages, like canonical names) when primary and other authors are fixed? Often what results is an empty page (example) that nonetheless retains the split & alias (and any CK on splits) in the background. Once the authors on the work have been fixed, and the works have moved to the correct author pages, should we remove the split, and remove the alias (it seems you have to do both, for the alias to disappear)? If so, who should be responsible for doing that -- should the person who's cleaning up the authors for works on that page be responsible for undoing the split at the end, or should it just be left for others to find and undo later? Or, should the splits and aliases be left there, even though the split pages are now empty?

Edited to correct some links and typos

Edited: Nov 7, 2011, 12:16pm Top

A very important question, Rsterling. Thanks for opening it!

Nov 7, 2011, 1:43pm Top

I'm still mostly worried about the fact that I changed a primary author and it didn't work, and other people seem to have reported the same problem (http://www.librarything.com/topic/126077#301337).

I think it only makes sense to use aliasing when it's really people who have the same name who are different people. Doesn't make much sense to use it as a bandaid otherwise.

Nov 7, 2011, 1:53pm Top

but that's resulted in some strange things: single authors that are split

That drives me completely berserk. I understand why people do it, but aliasing wasn't designed to accomplish what they are trying to accomplish. It's using a hammer when you need a screwdriver.

Edited: Nov 7, 2011, 2:48pm Top

Ok, well assuming all the bugs get ironed out, how should this work?

3 - It's mostly worked for me, and I've done quite a few now. I just tried recalculating the author page for kleinmaxine, and that got rid of the weird thing where it was showing Zinn's name but linking to Klein's page (with Zinn's name on Klein's page). But now it's showing the name as kleinmaxine on her author page and the work page, rather than Maxine Klein.* But, recalculation might work in some of these cases...

* ETA: and presumably a canonical name will fix the kleinmaxine vs. Maxine Klein issue.

Nov 7, 2011, 2:51pm Top

4: I understand why people do it, but aliasing wasn't designed to accomplish what they are trying to accomplish
I agree, and I think that with this new feature, there's now a better way to accomplish it, and there should no longer be any reason to split single author pages. But maybe I'm missing something...? Hence this thread: what are the cases that require splitting and aliasing? And when should that not be done?

Nov 7, 2011, 3:22pm Top

> 6

In my view, the only cases that require splitting and aliasing are those where more than one author shares the name and there are works by more than one author on the page.

Edited: Nov 7, 2011, 6:29pm Top

So what about last-name-only authors' pages? These generally do refer to more than one person -- more than one "Smith" -- and in the past people have been splitting and aliasing them, and sometimes adding Canonical Names to specify which Smith it is (since we aren't supposed to combine last-name only with full names). Now, because we can change the primary author, that would seem to do away with the need to split and alias. But then, what do we do with the ones that are already split and aliased? If we correct the author on the work page, the work will disappear (after a short lag) from the last-name only author page. But then the split and alias are still there, in the background, and there's still an "also includes" link on the pages the splits are aliased to. Should we just undo all of that, if we correct the primary author?

Nov 7, 2011, 6:42pm Top

So for that question about what to do with already split-and-aliased pages, there's another type I'm finding. Besides the last-name-only authors, which could actually refer to more than one person, there are also author pages where there's a single author with a full name who has nonetheless been split, so that the works would show up elsewhere. So an example would be a translator, who has been split and aliased into various *other* authors, so that the works would show up on the author pages of the original authors of the work. To me, while I understand the motivation, that seems to be an incorrect way to use aliasing, and a case where we should definitely remove the split and remove the alias, once the primary author is corrected.

With the last-name-only author pages, I'd also favor removing the split and aliases after authors are corrected. But since the page could still conceivably refer to more than one person, maybe there's some argument that the split structure should be kept even if the works disappear, in case the split is needed later. But I just can't see the need, and it's confusing to have splits that don't have any works attached to them, not least because they get renumbered, then any remaining CK ends up incorrect, etc.

Nov 7, 2011, 8:19pm Top

To me, while I understand the motivation, that seems to be an incorrect way to use aliasing, and a case where we should definitely remove the split and remove the alias, once the primary author is corrected.


Nov 7, 2011, 8:39pm Top

#10 However, when a single name author wins out and there are 60 or more of that name to change to primary author, it becomes more onerous than a simple alias. (Yes, I know, an exaggeration but I have seen 25 or so!)

As an aside, how do you change primary author? I know you can add but I thought the primary could not be deleted.

Nov 7, 2011, 8:40pm Top

You can edit it. It's under the "other authors" area on work pages. (Maybe it should be called "authors," or whatever, but changing the primary author is decidedly the minority case.)

Nov 7, 2011, 8:51pm Top

#13 Ah! I tried it and it works. Thanks.

This is an example of a lengthy list of aliased books: http://www.librarything.com/author/agatha

Easier to alias than to change each one. And I know - many can be combined and need to be.

Edited: Nov 7, 2011, 9:12pm Top

13 - In that case, I'd just see splitting as at best a temporary measure for grouping things that need their primary author changed. But I don't really think it's a good solution.

Tim, one issue I'm finding is that it's not obvious that you can undo the split and undo the alias after the primary author has been changed. The split and alias pages have no works on them once the primary author has been changed for all the works. You can still change to "single author," and it will un-split the author -- but that doesn't get rid of the aliases. You then also have to go to the alias page, and click "save page" again to un-alias it. That works -- it gets rid of the aliases -- but you wouldn't really know it at the time, because you get an error message. (It doesn't *look* like you can do any of this stuff once there's no works on the page, but you can.)

Nov 7, 2011, 9:20pm Top

#12 Another quick question - I have this book with no author and no other copies: http://www.librarything.com/work/11618284

The author is listed in WorldCat so I added as primary author. THe author is not in the system. The work page does not show the author. When clicking on the other author link (where I made Hollaus as primary author), the author page has no name. What did I do incorrectly?

Nov 7, 2011, 9:29pm Top

15 - I think this may be connected to a known bug about adding authors who aren't already in the system. I'd recommend taking this over to the "bugs" thread for the feature, and looking for other similar reports there:

For the author page not showing the name, a temporary solution would be to add a canonical name.

Nov 7, 2011, 9:38pm Top

#16 Thanks - will add name

Nov 17, 2011, 4:29pm Top

For me, I'm now limiting aliasing "single name" author pages, when the single name author is an LT misinterpretation that it is a single name, or when it should be a single name that is distinguishable from other names.

A recent example of a non-single name interpreted as a single name:
H. Smit on the Smith page.

A recent example of a distinguishable single name:
Lucas (musical group)

These are rare occurances.

Otherwise, rsterling has talked me into using the "other authors" instead of aliasing.

Nov 17, 2011, 5:06pm Top

18 - That makes sense: sometimes the single-name page makes sense, and does need to be aliased. Your H. Smit example is a great one. Other examples would be: a publishing house called just "Smith" that put out a lot of works without other indications of authors; or, an author who wrote simply under the name Smith. Those cases would require aliases.

For single-name authors where the full author name can be identified, I think the proper solution now is to change the primary author to reflect the correct full name.

(The next question is then what to do with any previous splits.)

Nov 17, 2011, 6:41pm Top

#19 Your comment:
The next question is then what to do with any previous splits.

Think of it a "job" security. It will take months and months and give us all something to do.

Nov 17, 2011, 6:54pm Top

That it will.

I'm just wondering what the consensus is on what to do with all the unused/empty splits this creates, and all the CK attached to them (including disambig notices enumerating the splits, and any canonical names or other CK attached to individual splits). My preference, as I mentioned, is to get rid of empty splits on pages like "Smith" and "unknown," once the works are moved off. In other words, get rid of the splits for works where primary author was corrected, but leave any splits where correcting the author isn't a solution.

So for a hypothetical Smith page, where
Smith (1) = John Smith
Smith (2) = Jane Smith
Smith (3) = H. Smit
Smith (4) = artist who goes only by the single name Smith
we would want to get rid of the splits for John Smith and Jane Smith but keep the ones for H. Smit and the artist known as Smith. So, we would change the primary author for all works listed under splits 1 and 2, which would empty out those splits, then re-save the alias page to get rid of the aliases for John Smith and Jane Smith, then (in my view) manually renumber the splits so that H. Smit and the artist Smith get 1 and 2 (and get rid of or move any CK on those splits, accordingly).

Nov 17, 2011, 7:55pm Top

#21 Whew! Definitely job security.

Edited: Apr 20, 2012, 8:53am Top

I have some questions about the change primary author function. When changing the primary author we see:

Warning: Changing a work's primary author on the "work" (ie., global) level is serious business, and should only be done when truly necessary. All edits are tracked and checked by other members. See WikiThing for more information.

Is there a place where we can see/track/check other people's changes?
Is there a thread where we justify/discuss changes?

Also, is it a convention / preference to leave the role of the primary author blank when that role is author?

Edited: Apr 20, 2012, 11:58am Top

Is there a place where we can see/track/check other people's changes?
These are recorded in the Helpers Log.

Is there a thread where we justify/discuss changes?
There's no specific thread that I know of, but discussions of individual cases might come up in the combiners group.

Also, is it a convention / preference to leave the role of the primary author blank when that role is author?
That is certainly my preference. I've seen one person suggest that we should add author for everything, but that doesn't seem to be common practice.

We probably should have more discussion and guidelines on these. My take is that in 99% of the cases, the primary author should reflect the way the author's name is written on/in the book.

Cases where changing primary author is very useful:
1) Where someone has entered 2 author names in the author field, such as "Amy Black and Charles Smith." The primary author can be changed to "Black, Amy" and "Smith, Charles" added as another "main" author.
2) Where someone has entered only a last name (or only a first name) instead of a full author name: so the book only has "Black" on it, but it's actually by "Amy Black."
3) Cases where the LT record says "Amy Black" but the book cover actually says "Amy L. Black" -- especially if there's more than one Amy Black in LT.
4) Cases where the primary author is misspelled: Amy Blcak instead of Amy Black.
5) Cases where the primary author is just wrong: e.g. there were a bunch of books that got imported with Enid Blyton as the author when they were totally unrelated to her (bad Amazon data I think).
6) Movies: I think there's less of a standard here, but I might change the primary author on a movie record to the director, especially if it's a movie based on a book, and is showing up on a book author's page.
7) Cases without an author: when the author can be determined, we can add it.

There are no doubt more.

In all of these, though, it's often useful to do any work combining first, since that saves some work.

Apr 20, 2012, 12:22pm Top

+1 to 24) rsterling

My order is:
A) combine works
B) if needed, recalculate title/author

C) then the rsterling cases with the addition:
8) replacing translators and editors with the author edited or translated, adding editor/translator as secondary entry, for instance,
Selected poems of Amy Black by Charles Smith. Replace "Charles Smith" with "Amy Black", add "Charles Smith" as editor.

D) then go to Amy Black page and combine if appropriate.

Edited: Apr 20, 2012, 8:36pm Top

> 24

8) Cases where an Editor is listed as the author of a work. This is not uncommon for pre-20th century works. But Mark Schorer is not the author of Pride and Prejudice-- Jane Austen is; David Bevington did not write Othello, William Shakespeare did.

There are thousands of such examples to be found on LT: I've fixed many though not all of the Shakespeares. But when I'm deposing an editor from the position of Primary Author, I take care to add him or her back in as Editor.

Apr 20, 2012, 2:49pm Top

25, if you do A) right the first time, you shouldn't have to do D)

You probably wouldn't consider going back to fix all your aliased-then-combined changes, but they'll get fixed eventually I suppose.

Apr 21, 2012, 12:08am Top

#24: My take is that in 99% of the cases, the primary author should reflect the way the author's name is written on/in the book. ... Cases where the LT record says "Amy Black" but the book cover actually says "Amy L. Black"

One problem is that the way the author's name is written on the book depends on the book. Harry Potter und der Stein der Weisen was written by Joanne K. Rowling, for example. The original editions of the Lucky Starr stories were written by Paul French; the new editions remove the pseudonym and call the author Isaac Asimov.

Apr 21, 2012, 12:09am Top

Ok, so 97%? ;)

Edited: Apr 21, 2012, 8:41am Top

Thanks for clarifying rsterling.

Edited: I meant for the answers and guidelines in #24, not just the 97% clarification :)

Edited: Apr 21, 2012, 8:39am Top

Name authority is a very good thing. That's why an LoC imported author name trumps the title page. And the title page trumps the book cover. (And also why the title page blows away an Amazon imported author name.)
So make that 96%.

Apr 24, 2012, 2:36pm Top

Except when the LoC is wrong. I don't have a specific example from them but WorldCat has a dime novel by "Anthony Comstock" which would have him spinning in his grave. The "Roger Starbuck" story in question was really written by Augustus Comstock no matter what a USC cataloger says.

The Nowadays Girls in the Adirondacks is sometimes attributed in library catalogs as by a real person with a name similar to the Stratemeyer Syndicate pseudonym used.

Probably better than Amazon data, I agree.


Group: Combiners!

788 members

46,407 messages


This topic is not marked as primarily about any work, author or other topic.

You are using the new servers! | About | Privacy/Terms | Help/FAQs | Blog | Store | APIs | TinyCat | Legacy Libraries | Early Reviewers | Common Knowledge | 116,146,468 books! | Top bar: Always visible