New Feature: LCC Pages, Improved Classification Pages

We have made some major updates to three classification systems:

* The Library of Congress Classification (LCC)
* The Melvil Decimal System (MDS/DDC)
* Genres

We now have new or updated and "LT2"-ified top-level pages for all three. Examples:

* LCC: https://www.librarything.com/lcc/PE1-3729
- Deeper example: https://www.librarything.com/lcc/DF599.8-649
2. DDC: https://www.librarything.com/mds/391.7
3. Genres: https://www.librarything.com/genre/16/Science-Fiction

The LCC page is entirely new. As with Dewey, LibraryThing exposes the wordings ("schedules"), not just the numbers--something you never see. The full LCC schedules are insanely long--some 400,000 headings--and not easily available. We have used the simplified set of 8,600 headings found on the Library of Congress website and turned into machine-readable format by a number of clever librarians.



All three pages include a line at the top like this:

The "see them here" part is new, and, I think, rather useful.

Other changes:

* Work pages now have LCC and DDC on them, with the genres.
* The right side of work pages has been slightly reordered, to front the covers and move the classifications up.
* The three systems have a "breadcrumb" structure. Click on the headline "LC Classification," "Genres" or "MDS" to get a list of the top-level. Like this:



* The three systems have "relateds" to each other.
* LCCs in your catalog can now be clicked.
* Genres are now broken up into a few categories, such as "Supergenres," "Genres" and "Age-Based Genres."
* The three systems have explainers at the top right of the page.

LCCs in your catalog can now be clicked.

Did you forget to push a file again? They are not clickable either in my catalog or yours. (And because I know you will ask, yes, I know mine are valid in 99.9% of cases. I spent a lot of time on that. The exceptions are a few cookbooks where I was unable to find them in any library source and put in a "stub" of TX.)

"You have 71 books in your library, or see them here."
The "or" is very strange in this sentence.

I understand the idea, given the functionality--a choice of a link to the catalog view or an abstracted bullet list.

Suggested improvement:
"You have 71 of these books in your library; see a list."

I very much like the "relateds" in the right-hand sidebar.

I also appreciate that the landing pages of each system are distinct, at a glance: the DDC set of tiles immediately distinguishable from the LCC's sparse hierarchical tree.

I was a little confused by the navigation within Genres: your example above starts at Science Fiction. When I click on the breadcrumbs-style link back to the parent category, Genres, the filter remains on Science Fiction. I would have expected it to "move back" to the prior level (Fiction?). Instead, it just opened the box of LT's genres. That's useful, just not what I expected from the navigation.

>2 lorax:

Fixed. Should be live in 2-5 mins.

Hopefully, all three of those will have an easily navigable way to access within LT2. Right now, I have no idea how one reaches them or finds them. I keep the links in the notepad module for myself, but new members are unlikely to stumble upon them unless they follow talk

SandraArdnas (#6):

With the exception of LCC (where Tim's promised fix hasn't gone live yet) they're easily reached from the catalog.

And they will eventually have their own "home" in LT2. We just don't have the supporting area for it quite ready yet but we wanted to get the larger feature out.

>7 lorax: Ah, OK thanks. But I assume that requires that you have those columns in your display style. Accessing genres from the drop-down menu in the catalogue that starts with tags leads to something completely different.

>8 conceptDawg: Thanks, that's what I meant, that eventually they should be positioned so that the feature is easier to discover and more prominent than currently

So there's a tension between two uses of classifications:

1. Using it related to your own books
2. Browsing it for fun

I think 1 is well satisfied. I'm not sure how much 2 there will be, but we should encourage it more.

>9 SandraArdnas: Accessing genres from the drop-down menu in the catalogue that starts with tags leads to something completely different.

If you're referring to what I think you are, I complained about this to Tim at least 3 times in the first few weeks after Genre was released: for the tags / authors / series / media / MDS options in that menu, it only shows you the tags / authors / etc. for the Collection you've selected. For Genre, it always shows the total numbers from all your Collections. Please make the behavior for Genre consistent with the behavior of all the other options and just show the totals for the Collection I've already selected, I said, and after the third time, Tim was like, Yes I hear you, and I shut up about it, then went to play with the new Charts and Graphs for a couple weeks.

I hoped it might be a case of "we'll clean this up with the transition to LT2," and didn't bring it up again. Since Tim has specifically included Genre in this classification discussion, I'm hopeful that it'll be cleaned up soon. I just want the drop-down behavior to be consistent.

>12 HeathMochaFrost:

Should be good. If not, give it 5m.

>13 timspalding: Yay, it works! 😊

I was a little surprised to see all the genres still there even if the Collection doesn't have some of the genres included. But since only the genres in that Collection have numbers next to them, it certainly does the job. THANK YOU!

I've added simple links to Genres, MDS and LCC to the Zeitgeist page: https://www.librarything.com/zeitgeist

Eventually we're going to have a better "Explore" tab. We're not there yet.

timspalding (#5):

LCC numbers are still not clickable in the catalog.

>3 paradoxosalpha: I agree with the suggestion to reword that sentence. It makes more sense the way paradoxosalpha has written it. Or you could expand it and say "show list on this page" since that's what it's doing.

>16 lorax:

Ah, thanks. There's a "LC Classification Word" field you can add that is clickable; that what I had been using. I'll be working on the catalog clickability today.

I would love some more information about how the "Invalid LC Classification" is determined - I've got books with numbers I am literally copying from the LoC catalog, and my LibraryThing catalog is telling me it's invalid.

tidybookshelf (#19):

Examples would help.

>19 tidybookshelf:

Yes, give me an example.

It's possible to have a call number from the LC that's not a LC call number. I don't know any off hand, but there are items in the LC that aren't part of the LC call-number system--archives, for example.

>1 timspalding:

I looked at your catalog and see some: https://www.librarything.com/catalog.php?lcc=invalidlcc&view=tidybookshelf&a...

Some of your own look like errors, but some do not. I will dig into it.

>21 timspalding: >19 tidybookshelf: Library of Congress uses several non-LC call number systems. MLC ("Minimal Level Cataloging") and FLS ("Fixed Location Storage") are assigned as call numbers for low-use books that are stored by size for efficiency's sake. There are also a limited number of call numbers that start with numbers, rather than letters, which are semi-cataloged.

>21 timspalding: >19 tidybookshelf: >23 jonsweitzerlamme: In addition to many Minimal Level Cataloguing examples the following classifications are listed on my LC Classifications page but are not functioning correctly (though all are entered from the LoC catalogue):
UNCLASSIFIED and UNCLASSIFIEDS - link says code UNC invalid
Newspaper - which unfortunately links to LCC - N. Fine Arts
LAW - link says code LAW invalid
CPB Box - which unfortunately links to LCC - B. Philosophy, Psychology and Religion
* - link goes to home page

EDIT: I thought there was some filtering of the results but I'm mistaken - see >33 newcrossbooks:

>24 newcrossbooks: The CPB Box code is not an LCC Code, or so I was informed. It's a Fixed Location Storage code and as such, invalid.

>25 gilroy: CPB Box may not be valid - but I have at least four Cheap PaperBacks that have been catalogued by LoC using it:
Somali Smashout by Peter McCurtin (CPB Box no. 1136 vol. 12)
Light Up the Night by M. L. Buchman (CPB Box no. 3912 vol. 7)
Sons by Robert DeMaria (CPB Box no. 10 vol. 1)
The Medusa Stone by Jack B. Du Brul (CPB Box no. 1901 vol. 16)

Is CPB the code with Box no. a reference to the location?

Yeah, you're most likely to see CPB Box for books that did not originally have a hardcover (vs paperback) edition. I see it primarily for older science fiction.

>26 newcrossbooks: It's a reference to a storage location, yes. But if you erase the CPB code, if someone else has entered the proper LoC code, it should populate in green.

Is the LCC system intended to have all of the available subheads? E.g. https://www.librarything.com/lcc/HD8045-8943 does not have subheads for the various regions or countries. Similarly here https://www.librarything.com/lcc/PQ3800-3999

Edited- OK. I see it gets complicated. But I'm the kind of nut who would want all the details.
https://www.loc.gov/aba/publications/FreeLCC/LCC_H2019TEXT.pdf

And where is PS3601-?
https://www.librarything.com/lcc/PS

I've found a small problem with the listing of History of Africa - LC Classification DT1-3415
https://www.librarything.com/lcc/DT1-3415#

Clicking on the subhead → DT154.1-159.9. Sudan. Anglo-Egyptian Sudan takes you to → DT43-154. Egypt instead.
The link directs to https://www.librarything.com/lcc/DT154.1-159.9# but the text is for DT115-154 Egypt/Local history and description.
When I list the books I have in my collection in this category then it includes a few from Sudan, as well as Egypt, as it pulls in books over the range DT115-154.9 (ie includes DT115-154, which is Egypt and DT154.1-154.9, which is part of the Sudan allocation)

Strangely when I click on the code for one of the recent Sudan entries in my catalogue, DT159.6. D27 P78, Darfur - The Ambiguous Genocide, I'm taken to the correct DT159.6 Sudan/Local history and description page - but am told that I "have no books here", despite the fact that over 20 of the covers displayed are marked with green ticks to show they are in my catalogue...

There are also a couple of semi-official extensions of LCC that I think would be good to treat as valid (even though they won't connect with what LC has):

1) NH classification for photography, so that photography books class with art, instead of with Engineering/Technology (in TR, as the official LCC has it).
This schedule is maintained by the Art Libraries Society of North America:
https://arlisna.org/publications/arlis-na-research-reports/publications-archive/...

2) FC classification for Canadian history. Created & maintained by Library & Archives Canada:
https://epe.lac-bac.gc.ca/100/200/301/lac-bac/class_fc_index-ef/e/n03/FC_Class_2...

>24 newcrossbooks: I thought there was some filtering of the results - but now see that only the last 1000 entries are displayed on the page so the results I thought were filtered out are just older entries:
Microfilm
Comic Book
IN PROCESS
Call number starting with a number (4PQ Fr 4034)

The total number of books with LC classification (Book Level + Invalid Code + Work Level) in Charts and Graphs matches the number on my LC Classification page. Around 10% are invalid (mainly UNCLASSIFIED or MLC).

>33 newcrossbooks: Which link did you follow and what are you trying to figure out?
I go into your charts, click on Invalid and it filters to 561 as it's supposed to on Win 10 and Firefox 92 (64 bit)

So it might be a browser specific bug.

>34 gilroy: Apologies for confusing you.

In >24 newcrossbooks: I was trying to list the invalid codes that had been pulled in from Library of Congress entries (as listed in my catalogue from the Charts and Graphs report) and checking what happened when I tried following the link given to them on the new LC Classification page. Some links are noted as invalid:
UNCLASSIFIED - https://www.librarything.com/lcc/UNC
LAW - https://www.librarything.com/lcc/LAW
But some go to an incorrect page:
Newspaper - https://www.librarything.com/lcc/N
CPB Box - https://www.librarything.com/lcc/B

I couldn't find some of the invalid code types (as listed in Charts and Graphs) so assumed they'd been filtered out and noted this in the post. I now realise only 1000 results are shown on the LC Classification page so the types I'd assumed were filtered out were simply ones that were not in the displayed results. There is no filtering - the invalid types listed in >33 newcrossbooks: are simply ones where I haven't been able to check what happens to links as the 'see in your catalogue link' isn't working yet.

I edited >24 newcrossbooks: and posted >33 newcrossbooks: to try to clear this up.

I use the Library of Congress as my main source and have, from Charts and Graphs, 4330 Book Level + 562 Invalid (one more LAW added today) + 993 Work level codes in my catalogue = 5885 on the LC Classification page. So around 10% of my codes are invalid - but almost all come directly from LoC entries...

And I've found problems with some of the other links on the LC Classification page as noted in >31 newcrossbooks:.

>24 newcrossbooks: CPB stands for (I just checked) "Copyright PaperBack Collection). It's another fixed location; the box references the location in offsite storage.

When call numbers start with NEWSPAPER or LAW that is unfortunately just bad cataloging/bad importing of pre-MaRC data. That is holdings information and should not be in the 050 field of the bibliographic record, which is where this data is being imported from. If there is additional information after that, /that/ should be the actual call number.

>33 newcrossbooks: IN PROCESS mean just that. Microfilm means that you've imported the bib record for the microfilm copy of that item. 4 items are nightmares, let's leave it at that.

>30 eromsted: And where is PS3601-?
https://www.librarything.com/lcc/PS

I have the same question.

Quick example: the work page for The Night Circus shows it's classed in PS3613. Thousands of books have class numbers of PS3601 and higher (but before the PS8001 where Canadian literature begins), so there ought to be a subheading level for them.

Q: Clicking "in your library" on the Genres page gives me, "Genre: Error x No books found." Is it just me, or am I doing something wrong?

>36 jonsweitzerlamme: "When call numbers start with NEWSPAPER or LAW that is unfortunately just bad cataloging" Not really bad cataloging; those records were most likely done before the schedules were created, and just assigned NEWPAPER or LAW pending development of the schedules. The Law schedules (K) are relatively recent creations (began 25 or 30 years ago maybe?). Some parts are VERY recent: KBS, KBT (canon law of Eastern churches) and KI (indigenous law) are just a couple years old, and even KBP (Islamic law) is only about 10 years or so old. Newspapers were originally planned to go in AN, but that schedule has never been developed and probably never will be.

>36 jonsweitzerlamme: I appreciate that the codes are "invalid" - but all are brought in using the Library of Congress as a source and some invalid codes are not being dealt with correctly by LibraryThing at the moment. For instance CPB, "Copyright PaperBack Collection," entries are currently linked by LibraryThing to the Philosophy, Psychology and Religion page - https://www.librarything.com/lcc/B

And, of course, it would be good to correct/update the invalid codes. Sadly most of the Newspapers are what you would call bad cataloging. I've checked some of the entries in the LoC catalogue to find the 050 field usually gives:
datafield tag="050" ind1="0" ind2="0">
subfield code="a">Newspaper
/datafield>

Same with LAW:
datafield tag="050" ind1="0" ind2="0">
subfield code="a">LAW
/datafield>

I have managed to find that the Italian newspaper La Stampa, which was imported as having the classification 'Newspaper 2345', is actually catalogued on the LoC website as having a classification Z2345 .S8 - so that's one fixed, and it gives a clue as to which other newspaper entries might be easily corrected.

>37 jonsweitzerlamme: And I understand what IN PROCESS means, and I have now managed to check that books with this code link to a page noting that the code is invalid: https://www.librarything.com/lcc/PRO

And MICROFILM is fairly obvious, though many of the LoC Newspapers are microfilmed, but it's reported in a different field. Unfortunately MICROFILM entries currently link to the M. Music page - https://www.librarything.com/lcc/M

Comic Book entries link to the B. Philosophy, Psychology and Religion page - https://www.librarything.com/lcc/B

And my example of a call number starting with a number, Chez la Reine de Saba; Chronique Éthiopienne, 4PQ Fr 4034, links to the F. Local History of the United States, Canada and Latin America page - https://www.librarything.com/lcc/F

(Note: While some of the invalid codes send you to an incorrect page, the books with the invalid code are not listed on the page you are sent to)

>40 Proclus: Interesting. And yes, many of my invalid codes look as though they belong to older items that predate the newer code schemes.
Is the Library of Congress going to update all of these placeholder codes eventually?
Was a schedule ever developed for Comic Books?

>39 jasbro: It's not just you; mine is doing the same thing. I'm going to click around a little more, but this is probably a bug.

ETA - it's doing the same thing with LC Classification. I'll try to see if there's already a bug report, and if not, I'll submit one. Good catch, jasbro.

>41 newcrossbooks: OK, I'm understanding you, sorry. The issue is that these invalid call numbers are being treated as valid, that makes sense.

>30 eromsted: >38 HeathMochaFrost: The subheading currently stops at PS3576, so only covers American Literature up to the year 2000. Books published after 2001 are in the series PS3600-3626 and so do not currently get listed under the American Literature subheading.

The post 2001 books are all included in the books displayed under the general PS American Literature heading. The link from any PS3600-3626 book only currently takes you to the main PS list, but it should really link to a list of all books in the post 1961 sequence (or will LibraryThing add a new PS3600-3626. 2001- split?).

P. Language and Literature
⮑ PS. American literature
⮑ PS1-3576. American literature
⮑ PS700-3576. Individual authors
→ PS700-893. Colonial period (17th and 18th centuries)
→ PS991-3390. 19th century
→ PS3500-3549. 1900-1960
→ PS3550-3576. 1961-2000
→ PS3600-3626. 2001-

>42 newcrossbooks: No separate schedule for comic books. They go in PN6700-6790, I think

>46 Proclus: I've managed to update some of the entries coded 'comic book' using entries in the Michigan State University Library catalogue. They do indeed fall in the range you suggest.

The coding system doesn't seem to be as fixed as I was assuming though, with different libraries using completely different codes for the same books. I found that my Turkish Akbaba magazines are listed with comic books, though they're really more of a humerous magazine similar to the British Punch.

This is because the library I had used as a source has coded them as PN6790.T9 A43
P. Language and Literature > PN. Literature (General) > PN1-6790. Literature (General) > PN6010-6790. Collections of general literature > PN6700-6790. Comic books, strips, etc.

I checked some other US library holdings of the magazine and found a wide range of codes have been chosen:
PN6222 .T9 A333 (P. Language and Literature > PN. Literature (General) > PN1-6790. Literature (General) > PN6010-6790. Collections of general literature > PN6146.5-6231. Wit and humor > PN6157-6222. By region or country)
PL21 .A115 (P. Language and Literature > PL. Languages and literatures of Eastern Asia, Africa, Oceania > PL1-8844. Languages of Eastern Asia, Africa, Oceania > PL1-481. Ural-Altaic languages)
NC1660.T9 A313 (N. Fine Arts > NC. Drawing. Design. Illustration > NC1-1940. Drawing. Design. Illustration > NC1300-1766. Pictorial humor, caricature, etc.)
AP95.T8 K34 (A. General Works > AP. Periodicals > AP1-271. Periodicals > AP95.A-Z Periodicals. By Other language, A-Z > AP95.T8 - Turkish Language)

In fact I would have thought Akbaba would be better coded in the Humerous periodicals range (AP101-115), which is where Punch magazine is (AP101.P8). It's Turkish language so it would code as AP115.T8

And I'm a little confused by the fact that T9 has been used by some libraries for the language, rather than T8. Is T9 another option for Turkish?
And where can I find a comprehensive list of language codes? I can find that T3 = Tamil, T4 = Telugu, T8 = Turkish, T27 = Tagalog, T28 = Tajik, T34 = Tatar and T83 = Turkmen but what, for instance, is the code for Tigrinya?

>47 newcrossbooks: T* are (most likely) the Cutter numbers for the languages. There are multiple kinds of Cutters (and therefore multiple tables for those Cutter numbers): Cutter 2-figure, Cutter 3-figure, Cutter-Sanborn, LC Cutter...there may be more; but those are the most obvious ones in my memory.

If you go here: https://calculate.alptown.com/, choose LC Cutter, the type in the language, you will get the full Cutter number for that language. That doesn't mean that's the set in stone Cutter number. Most institutions will Cutter to fit an item on the shelf. So, even though Tagalog is T34 at the above site, the library might have needed to put it at T27 to fit on the shelf.

One of the things that drives me insane about this update to LT is the "canonical LC number." There's no such thing. There's the number Library of Congress uses. There's the number(s) that not Library of Congress uses. Every library (as in public, academic, etc) is given discretion to use the call number that works for their institution. That's why there is supposed to be knowledge management for cataloging practices at a library. Cataloging follows a proscribed set of rules for punctuation, data placement, level of cataloging, etc; but there are no proscribed rules for classification (yes, there are probably some exceptions) beyond the classification scheme fitting your institutional needs.

If you think Akbaba fits better in AP101-115, then classify it as such in your library. So long as you remember why it's classified that way, that's the beauty of the LC classification system. You make it fit your needs. It's the only part of cataloging I ever liked.

For what it's worth, if you look up Tigrinya using the above site and LC Class Numbers option, you get...

BS325.T49 The Bible--General--Texts and versions--Modern texts and versions--Non-European languages--African languages, A-Z--Tigrinya Table BS5
DT380.4.T54 History of Africa--Eastern Africa--Ethiopia (Abyssinia)--Ethnography--Individual elements in the population, A-Z--Tigrinya
KQ9000.T54 Africa: Law of indigenous peoples--Individual peoples, chiefdoms, kingdoms and empires--Individual ethnic groups, A-Z--Tigrinya Table KL-KWX14
PE1130.3.T54 English philology and language--Modern English--Language--Grammar--Readers--Textbooks: Grammars. Readers, etc., for foreign students--By language--African languages, A-Z--Tigrinya
PJ9111-PJ9111.95 Oriental philology and literature--Ethiopian languages--Special--Tigrinya Table P-PZ15a
PN6519.T56 General literature--Proverbs--Modern--Other special, A-Z--Tigrinya

>48 lesmel: Thanks very much for the detailed reply - very useful.

I must admit that I originally assumed that every book would have a fixed classification code assigned to it - but it has rapidly become clear that things aren't that simple...

>47 newcrossbooks: Often there is one, and only one, correct classification for a book— but sometimes there are more than one that's acceptable.
Your Akbaba magazine, as you recognized, should strictly speaking be in AP115 (Humorous periodicals in languages not specifically provided for in earlier numbers); but, it could just as well be in PN6222.T9 (general collections of wit & humor in Turkish; you can use this number if you want the periodical to sit with other humor, and not have the periodicals so widely separated from the books of the same sort).
The other numbers you found are just wrong. ;)

>49 newcrossbooks: The first rule of library school is that cataloging is specific to your library. For example, a book about soccer in Indonesia can be put under soccer--Indonesia if your library is strong in sports, or Indonesia--sports if your library is strong in southeast Asia. I have a book (The Greatest Book in the World and Other Papers) that is ex-library and has an 814 Dewey number (American essays) printed on the spine. I've put it under 002.04 (History of the Book--Essays) because that makes more sense for my library.

>50 Proclus: I strongly disagree with the idea that "Often there is one, and only one, correct classification for a book". During the last three years of working as a librarian, I re-classified every book in our library and it was agonizing to decide where to place them. Most fiction books can be fairly easily placed (but genres can overlap greatly and one person's thriller is another person's horror is a third person's general fiction).
The vast majority of the books in our library could have easily been placed in either of at least two, if not three or more different places.

>52 PawsforThought: I know Dewey is different, but for LC classification, yes, there is often One and Only One correct classification. If you have an edition of Shakespeare's Romeo & Juliet in English, for example, then PR2831 is the only place it should go. This is the case for just about all literary works by individual authors, since genre is irrelevant in LC in such situations.

>45 newcrossbooks:
(or will LibraryThing add a new PS3600-3626. 2001- split?)

That was basically my question: why isn't it already there?

I checked several other "literature in X language" and found several that do include the "2001 and following" class numbers--for example:

Italian
https://www.librarything.com/lcc/PQ4900-4926

Danish
https://www.librarything.com/lcc/PT8177-8177.36

Portuguese
https://www.librarything.com/lcc/PQ9300-9326

It's true that there are *many* languages in there that don't have anywhere near the specificity of these (mostly) European languages, but why would PS *not* have the "2001- " group when these other languages/countries do?

On a related note ... I found that PR, English Literature, needs a lot of work, because right now it's just one big bucket.
https://www.librarything.com/lcc/PR0-10000#

>53 Proclus: We didn’t use Dewey at my library. And I wasn’t primarily talking about fiction, where things are usually fairly easy to place. Non-fiction can be a beast, because there is so much interdisciplinary work.

>1 timspalding: How can genres be added? I would like to see something like an overall "Language: Vocabulary and Grammar" and "General Education."

Thank you for your consideration.

Carla Girtman

>56 ariatracker: Genrething changes should be made in the Genrething threads:
https://www.librarything.com/topic/333142

>56 ariatracker: You can't add custom genres. You can only add and remove from the existing ones

I'll try asking this here, but am happy to be redirected: This book https://www.librarything.com/work/25775371/editions/206573300 has acquired an incorrect Dewey. Is there any way of forcing a recalculate?

>59 rodneyvc: Someone mentioned the "Canonical DDC/MDS" Common Knowledge field in another thread, so I wondered if adding the Dewey/MDS number for cooking would change that display -- and happily, it worked. :-) I changed the Genre to match, too. If someone else decides there's a more precise class number for the work than plain old 641.5, they can go into Common Knowledge and change it again.

We can only speak to the changes to DDC but thank you! And thanks to LibraryThing designer conceptDawg.

>60 HeathMochaFrost: You're a genius - I hadn't spotted the introduction of the Canonical DDC/MDS field!

>59 rodneyvc: >60 HeathMochaFrost: I had a similar problem with LCC which I tried to fix via the Combiners! group.
https://www.librarything.com/topic/335655

I eventually managed to fix the problem by adding the LCC to the CK field - but then found some interesting behaviour when I tried removing it to see if it reverted back to its previous LCC...

I really like the Genre and Classification pages. However (no one is every totally satisfied!), I would still like to see the Genres marked as Fiction or Nonfiction by being columnized or colorcoded or something else so that they are more easily visually grasped as a group-that-belongs-together.

>32 Proclus:

In the process of adding NH and FC. It's a little tricky to do.

Note for self: MA, ZZ also formerly acceptable. Unsure where they come from. See get_lcc_classessAA
NH now found on IA at https://web.archive.org/web/20210518070619/https://www.arlisna.org/images/resear...

What's going on with "missing" sections of LCC? For example, I'm looking today for books in QH325-327, but clicking on that takes me to the "Biology (general)" page, QH301-705.5, and the subheads that are available don't cover the whole range. Attempting to hack the URL to get me books in the early QH 300s just always kick me right back to the general page.

Is there just no way to get narrowed LCC listings that aren't from a certain restricted list? It would be really helpful if we could bring up an arbitrary range, like with Melvil!

>66 melannen:

Yes, we only have about 8.6k ranges in, with their schedules. So you can choose others outside those points.

In theory it could allow arbitrary ranges, so long as it didn't show the schedules. (The schedules corresponding to that list could be inside a schedule, or crossing many, or have none.) I'm not really convinced that's a common use-case, however. Can you describe more about what you're trying to do and why?

I'm working on the "invalid" ones that it's pulling in from the global level.

>68 timspalding: Will someone be working on PR? Because right now it looks like one big range with no subheads:
https://www.librarything.com/lcc/PR0-10000#

>69 HeathMochaFrost:

Thanks. I'll take a look soon. I think some level must have come disconnected, because there's a whole missing structure there.

Oh God, none of the LCC structure under PR got entered. Misery. Okay, looking at how to get it in.

>67 timspalding: Well, I'm looking for books in that LCC range! :P

Basically: I wanted to find some more books about Precambrian life and abiogenesis, as research for a project; so I did what I would do in my library or with Melvil - find a book on the topic that I own, look up its call number, find other books near it on the shelf. The books I own on that topic are all in QH325 or QH327.

With that particular topic, Melvil/Dewey doesn't work great, because it splits it up between a bunch of different numbers in entirely different subtopics. And the tagging is inconsistent enough that I couldn't be sure I wasn't missing stuff. So I thought this would be a perfect use for the new LCC system! QH325-327 seems to have just what I want, but if there's a way to get LT to show me only books in its database in that range, or even a slightly broader version of that range, I can't find it? Since site search doesn't have a way to search other than the entered headings either - even just having it on site search would work for my purpose (most library catalogs let you search any call number range, or a specific call number and get books near it on the shelf.)

Right now my only option using LCC is to get all of biology jumbled together, unsorted, which is unhelpful. Even an ability to get just the books in that range that aren't under any of the entered subject headings would be vastly better, or to get only the QH300s, or get the results sorted by call number.

>71 timspalding: I'm sorry to bring misery. :-( I wish I could help you fix it.

But I am kinda glad to have identified an actual problem and I don't know how no one else mentioned it earlier. (Well, except for my note at the end of this comment: https://www.librarything.com/topic/335290#7613502.)

Okay, valid LCCs should no longer be labelled as invalid.

Invalid LCCs should no longer be coming in from the work level. (About 850 remain; they will be removed within 24h.)

Also, new bug:

When I did try to look through everything in Biology (general), I discovered that the "Show All 500" button only seems to work in cover mode - it does nothing in title mode (which makes trying to ctrl-f search in the results difficult.) And "show all 500" doesn't really make sense? Surely it's "show first 500"? (I did finally find a subhead with less than 500 results; "Show All 465" makes a lot more sense, but there aren't many subheads that small.)

I love the new look of this feature!
I do have a question though - as I look at the "see a list" feature some weird things show up. For example four of my pre-school picture books show up in "Information" - and they are not, as far as I can grasp, about "information". ... two of my ten or so Henry and Mudge Books, Gooseberry Park by Cynthia Rylant, and one or two others. Do you know why this kind of thing is happening?

>76 dutchgirldtd: The classification is set by the Melvil Decimal System (MDS) information that LibraryThing has pulled in for the book. Unfortunately the MDS number seems to vary from library to library, depending on how they choose to classify the book.

For your Henry and Mudge books the MDS information 'calculated' by Library Thing varies from title to title:
Henry And Mudge And The Bedtime Thumps - 306.4 - Social Sciences > Social Sciences > Culture and Institutions > Specific aspects of culture
Henry and Mudge in the Sparkle Days - 028 - Information > Library and Information Sciences > Books and Reading
Henry and Mudge Get the Cold Shivers - 813 - Literature > English (North America) > American fiction
Henry and Mudge and the Long Weekend - 813.54 - Literature > English (North America) > American fiction > 20th Century > 1945-1999
HENRY AND MUDGE AND THE FOREVER SEA - 028 - Information > Library and Information Sciences > Books and Reading
Henry & Mudge and the Happy Cat - 420 - Language > English > English and Old English (Anglo-Saxon)
Henry and Mudge and the Careful Cousin - 813 - Literature > English (North America) > American fiction

Checking one of your two "Information" books in more detail:
Henry and Mudge in the Sparkle Days is catalogued by 1022 members but, looking at the Work details, the MDS information is taken from only 12 catalogue entries:
Work details: https://www.librarything.com/work/441756/workdetails/103995086
DDC/MDS
028 - Information > Library and Information Sciences > Books and Reading - Primary; 4 copies
514 - Natural sciences and mathematics > Mathematics > Topology - 2 copies
428 - Language > English > Standard English usage (Prescriptive linguistics) - 1 copy
441 - Language > French > Orthography - 1 copy
531 - Natural sciences and mathematics > Physics > Mechanics - 1 copy
554 - Natural sciences and mathematics > Earth sciences & geology > Europe - 1 copy
563 - Natural sciences and mathematics > Fossils & prehistoric life > Protozoa; Radiates - 1 copy
720 - Arts and Recreation > Architecture > Architecture - 1 copy

It is clear that many of these MDS entries are suspect, probably left over information from books titles that were incorrectly merged at one time and later separated, but 028 is used most so is the one that is 'calculated' by LibraryThing to be the primary.

In conclusion, the MDS numbers in your catalogue are all green 'calculated' numbers - you can edit them to make them more consistent and this should then move all the books into the same category.

>77 newcrossbooks: Not quite right. The numbers from the library records are Dewey Decimal Numbers but the DDS classification tables are copyrighted so LT can't use them. The MDS tables are the closest to the actual DDN tables that LT can legally use.

>77 newcrossbooks: Picture books will sometimes end up in the "books and reading" classifications if they've been cataloged either as a) examples of the craft of book-making or b) books about how to read, rather than as fiction.

>78 jjwilson61: The numbers pulled in from library records are indeed Dewey® Decimal Numbers but LibraryThing stores the information under their Melvil Decimal System. It's the same decimal classification system, as originally copyrighted by Melvil Dewey in 1876, but is based on the last copyright free issue, the eleventh edition from 1922, so doesn't necessarily conform to the latest updates.

The Dewey Decimal Classification® (DDC) is now maintained by the Online Computer Library Center (OCLC) who, I understand, used to be quite litigious in protecting their name and the use of their system so I didn't want to risk mentioning them.

>79 melannen: I'm sure every library has a reason for classifying their books in the way they do but the whole point of the decimal classification system is to ensure that all books on the same subject are shelved together. It doesn't really make any sense to me for LibraryThing to have the different books in the 'Henry and Mudge' series of pre-school readers classified in so many different ways.

The Library of Congress Classification system is designed to handle juvenile fiction but it I'm not sure that the Dewey® system is. Under the Children’s and Young Adults’ Cataloging Program (CYAC) these books would not be given a Dewey® number but would instead get the classification E "for easy fiction intended for children through grade 3 or age 8." There is a similar Fic category that "applies to fiction for children beyond grade 3 or age 8, as well as young adults through high school."
https://www.loc.gov/aba/cyac/classification.html

E and Fic designations are currently treated as invalid Dewey® Classifications by LibraryThing.

I note from the OCLC support page for field 082 in their Bibliographic Formats and Standards
"For works intended for a juvenile audience, a lowercase j is added as a prefix."
https://www.oclc.org/bibformats/en/0xx/082.html

A quick test suggests that LibraryThing does not handle Dewey® numbers in this format correctly - the first part of the number in front of j is ignored and the number is read from the first decimal point (eg j962.7043 is read as 7043).

>81 newcrossbooks: That's not an official Dewey format though - it's just how LC, specifically, puts together call numbers for juvenile nonfiction. Every library more or less works out its own format for separating juvenile nonfiction, if they do at all, but the actual Dewey is still just the bare number, as far as I know; the J is an added prefix to the actual Dewey. (I do wish LT would be smarter about stripping out the js though.)

Most libraries don't class their Juvenile Fiction under Dewey or LCC at all - they have their own systems (which, again, vary a lot, but are usually based mostly on author's surname.) A library that *does* have picture books classed under Dewey or LCC very often has them for other reasons than just a browseable picture book collection - an art library that has them for the illustrator, for example, or a community college library that has just a few as resources for early childhood education or adult ESL classes. They're going to file them where it makes sense for how their patrons are going to use them - that's built-in to the Dewey system, that libraries have flexibility in assigning numbers based on how their collections are used.

Unfortunately, because of the way LT calculates the green data, the vast majority of copies that don't have a valid Dewey/LCC number at all get ignored, and those few libraries that assign them a number for a specialist reason float to the top. That doesn't make it a wrong number, just an odd way LT handles the data. (I'm pretty sure the ones about prehistoric life are just plain wrong numbers though.)

There's plenty of bad Deweys out there for children's fiction. Whether it's idiosyncratic libraries or Amazon errors I don't know, but it's not LT's error in any case.

>83 lorax: So what's a good Dewey for children's fiction? There doesn't seem to be any way of differentiating adult and children's fiction under the Dewey system - yet most public libraries have a section for children's books.

I don't know how 'official' putting a 'j' in front of a Dewey number to signify that it's a book for a juvenile audience is. I guess it probably is, and is in response to the LoC replacing Dewey numbers with 'E' and 'Fic' for children's book entries, but LT certainly doesn't handle 'j' prefixed entries properly (see >81 newcrossbooks:).

And, as I've found to my cost, LibraryThing has a habit of attaching similar titled books to entries that they don't belong to - and when you separate them out the unwanted LCC and DDC are left behind and can overide the correct LCC and DDC entry for your book.

I think you're exactly right, that there isn't a good way. So most libraries don't use it, so it's less likely that the values that are there are good.

I'm not sure what you mean about the LoC replacing Dewey numbers with "E" or "Fic" for children's books; E is for history in the LoC, and "Fic" isn't an LoC classification at all. You posted upthread about the "Children's and Young Adult's Cataloging Program" originating those classifications - is that operated by the LoC? I hadn't heard about it. PZ in the LoC is the closest I can think of. But even then it's a parallel system to Dewey, not slotting something else in for a particular subset of books.

>85 lorax: When I check my invalid DD Code entries I have a lot of LCC coded PZ entries that have 'E' or 'Fic' (in square brackets) as DDC entries.

The OCLC support page for field 082 in their Bibliographic Formats and Standards notes the use of these entries for "some juvenile works" under "Use the following guidelines for entering Dewey number from LC copy."
It looks as though the 'j' prefix for juvenile works also comes via the Dewey Program at the Library of Congress.
https://www.oclc.org/bibformats/en/0xx/082.html

I agree with >83 lorax:. I was just looking at the classifications and found that I own several religious works, which surprised me.

A couple of examples: Curious George Goes to the Aquarium shows up as both 279 Religions > History of Christianity > Pacific and 406 Language > Language > Organizations and management.

A Turn for Lucas, which is a book about a Brazilian/American brother and sister who play soccer together on a youth team, shows up as 259 Religions > Christian pastoral theology, homiletics and religious orders > Other ministries and church work and 952 History and Geography > Asia > Japan.

These two books are just a couple of examples. There are several other children's books showing up as religious works which clearly are not. I have no idea how these books get classified, but these are pretty far off IMO!

>87 rretzler: I'm not sure I agree so much with >83 lorax: that the problem is nothing to do with LT.

As I mention in >84 newcrossbooks: - one problem I've noticed is that LT "has a habit of attaching similar titled books to entries that they don't belong to - and when you separate them out the unwanted LCC and DDC are left behind and can overide the correct LCC and DDC entry for your book.

I would guess this is what has happened in your two examples:
417 members have catalogued Curious George Goes to the Aquarium, but only 2 of the entries contain DDC values that LibraryThing accepts as valid (and both are clearly incorrect).
The Library of Congress gives the book a Callnumber PZ7.C9217 1984 and a DDC of E (in square brackets).

14 members have catalogued A Turn for Lucas, with 2 of the entries contain DDC values that LibraryThing accepts as valid (and both are clearly incorrect).
The Library of Congress gives the book a Callnumber PZ7.A93382Tur 2006 and a DDC of E (in square brackets).

LibraryThing is currently treating DDC 'E', and 'Fic', entries as invalid - yet the Dewey system, which is supposed to make it easy to keep all books on one topic together on the shelves, appears to have no way of differentiating children's books from other books. You can force a book to accept a particular Dewey Number by adding one to the Common Knowledge - but how can you then keep all children's books together?

Should LibraryThing start accepting 'E' and 'Fic' as 'valid' Dewey Classifications and add a children/young adults classification?

Doesn't setting canonical DDC/LCC override the calculated one? If so, it can be used for cases like >87 rretzler:

>89 SandraArdnas: The LCC for the books in >87 rretzler: are both correct (PZ7 is a general number for “Juvenile belles lettres” or juvenile fiction, for authors whose first works have been published between 1870 and 2014).

You can force a more apt DDC on the books by adding one to the Common Knowledge but if you give these a 'normal' DDC then they'll be mixed up with all the other books. I think most libraries prefer to keep all their children's books grouped together, which doesn't really fit in with the current Dewey system.

Under the Children’s and Young Adults’ Cataloging Program (CYAC) these books would get the classification 'E' "for easy fiction intended for children through grade 3 or age 8" in place of the Dewey number. There is a similar 'Fic' category that "applies to fiction for children beyond grade 3 or age 8, as well as young adults through high school."
https://www.loc.gov/aba/cyac/classification.html

I find that lots of PZ7 books in my catalogue have 'E' or 'Fic' instead of a Dewey number, but that LT treats the records as invalid.

>90 newcrossbooks: But if a library has E or Fic on their book record then when searching their library on DDN those books will be in their separate category. The data in the work record doesn't really matter when searching or browsing within an individual catalog.

>91 jjwilson61: If a collection on LT has books with E or Fic for their book record then they will be listed as having an invalid DDN so, unless you count this as their separate category, you can't search for them in your catalogue by DDN.

And the data in the work record does really matter when you try to display all books in LT that belong to a particular category. It should be an easy way of getting an idea of books in any category of interest that you might like to add to your own library. Unfortunately many categories seem to have books that shouldn't really be there, and lots of these spurious entries seem to be books for children or young adults. Worse, you find books that you have in listed as E or Fic in your catalogue appearing in odd categories because an incorrect DDN has been set in the work record.

You just have to look at the books listed under something like Religions/History of Christianity/Pacific to see the sort of problem that there currently is with the way LT works out the DDC of children's books.
https://www.librarything.com/mds/279

Look at a book like Liar, Liar, Pants on Fire. A children's book that has been catalogued by 172 members has 8 different DDCs listed against it, all from single catalogued entries:
279 - Religions > History of Christianity > Pacific
359 - Social Sciences > Public Administration, Military Science > Navy; Naval Science
427 - Language > English > Historical and geographic variations, modern nongeographic variations of English
431 - Language > German and Germanic > Orthography
464 - Language > Spanish > {Formerly "Synonyms"; No longer used}
469 - Language > Spanish > Portuguese
511 - Natural sciences and mathematics > Mathematics > General Principles
813 - Literature > English (North America) > American fiction
LT appears to have chosen the lowest number listed, 279, to be the default for the book.

>92 newcrossbooks: I think you're conflating two different things with DDC, the site-wide category and the search of your catalogue for that particular entry. It is certainly possible to search for particular DDC in your catalogue (ddcnum:Fic if it's Fic). The default search 'all fields' would also find them, but might pull in some other stuff with that string