Introducing the "Melvil Decimal System"

Blog post: http://www.librarything.com/blogs/librarything/2010/08/introducing-the-melvil-de...

Example:

http://www.librarything.com/mds/222.2

Screeenshot:



Help wanted:

I need help entering the divisions from the 1922 edition (here. Feel free to modernize language--for example, I entered "language" for the modern equivalent of "philology" and "Historical books" for "Historic books"). But please, please do not enter them from a newer edition.

Things I know

* This isn't done.
* What you really want to know is where YOUR books stack up. There's a reason I'm holding off on this. (And it's not because I'm pretty tired!)
* Links from catalog missing.
* Links from other pages missing. (Although isn't it cool to see the tags and how well they match up?)

*looks around, wanders off*

Ah, there you are!

And for those who are really confused (I was), read the box on the bottom right:


What is MDS?

Melvil stands for "Melvil Decimal System," named after Melvil Dewey, the famous librarian. Melvil Dewey invented his Dewey Decimal System in 1876, and early versions of his system are in the public domain.

More recent editions of his system are in copyright, and the name "Dewey," "Dewey Decimal," "Dewey Decimal Classification" and "DDC" are registered trademarked by OCLC, who publish periodic revisions.

LibraryThing's MDS system is based on the classification work of libraries around the world, whose assignments are not copyrightable. MDS "scheduldes" (the words that describe the numbers) are user-added, and based on out-of-copyright editions of the system.

The Melvil Decimal System is NOT the Dewey Decimal System of today. Wordings, which are entered by members, can only come from the 1922 edition or from a modernization of that edition's language. It may not come from contemporary editions of the the system.


How irritating (re: OCLC).

Clearly I want to be as non-confusing as possible. But I also want to avoid a lawsuit. OCLC takes very forward positions on this stuff. It sued the "Library Hotel," which put themed books in the hotel rooms as if they were Dewey numbers. It claims a sort of ownership over government-made cataloging records, etc. etc.

I'm still confused! If I go to the google books page I can see a table of contents, mostly truncated. Is this where you think I could get names of divisions? Or are you expecting help from people who own a copy of the 1922 version?

Near as I can tell, Google Books gives me the entire book. And here's a link to it in PDF form (warning: very slow).

>5 justjim:

It's all there. Maybe we should record links to the right page.

Is this supposed to be translated as well? If so, should we do the translation or should attempts be made to find copyright-free books on the DDC in various languages?

This appears to be it at Project Gutenberg. Could you confirm it's the right (or wrong) edition, Tim?

http://www.gutenberg.org/etext/12513


(Not the right version.)

Even signed into google I can't see inside the book. The PDF link 404s on me.

>8 andejons:

I'm so happy you're translating it. OCLC is only studying the idea of a Swedish translation of the (current) DDC.

I don't know if there are out-of-copyright Swedish DDCs. (Don't confuse it with UDC.)

>9 brightcopy:

No, that's the 1876 edition. Use the 1922 edition, which is at
http://books.google.com/books?id=B2FnAAAAIAAJ&dq=Melvil%20Dewey&pg=RA1-P...

12> Right, noted. I was just trying to find an alternate source for justjim, since he's having technical difficulties with google books.

I don't know how different it is. It may be very similar, especially on the top levels, which is all we really need do.

It occurs to me - does google have any copyright claim to the PDF they generate for the book? If not, you could grab it and stick it on LT and provide an alternate link there. If one person is having an issue accessing it at google books, you can bet others will.

I'd like to help out but I'm having the same trouble as justjim - maybe because I'm not from the U.S. Google won't allow me to view the book?

If you're outside the US, you may well have trouble reading it from Google Books, for legal reasons. Which is why a lot of people take the PDF and upload it to the Internet Archive, which is completely legal. http://www.archive.org/details/decimalclassifi02dewegoog is a 1919 edition on the Internet Archive.

The thing about space in the Dewey Decimal System is that the structure does odd things to it. 0 is general, 1-8 are specific examples, and then 9 is the catch-all bin. So of course Catholicism gets 282 and then Unitarianism gets 288. He split Christianity into 8 groups and the misc. Catholicism undoubtedly gets its share of 220-270, whereas Confucianism, Sikhism, Shintoism, N. & S. American and Australian indigenous religions all get stuck in 299.

And could someone please upload the PDF to the Internet Archive then? :-)

Edit: Nevermind - http://www.archive.org/details/decimalclassifi09dewegoog -
That is it, right?

Tim : Looks like it broke somethingin the process. Please fix.

I propose to use a dash (-) to differentiate between separate titles for a category.

For example, in the philosophy department there is 124 "Teleology Final causes" (double spaced). I propose to enter this as "Teleology - Final causes"

Yea or nay?

Furthermore, some boxes do not appear (such as 129.1); does this have to do with missing entries in the CK database for this Dewey Decimal? I can click the empty box and enter the data, though.

What do we do when the modern and the old headings look funny together? Like 491.82 is in modern usage Bosnian/Croatian/Serbian; but 491.83, which I suppose is currently unused, is Croatian. So MDC reads 491.82 Bosnian/Croatian/Serbian 491.83 Croatian. Just let it be, I presume?

Edit: Ah, I suppose we should just leave 491.82 Serbian; just update the language, don't match the usage?

Further edit: #20 I've been using commas and "and".

Further edit: The 510s are pretty much pointless without updating; Dewey's grasp on higher mathematics was pretty negligible, so there's not much left of his 1922 structure.

> justjim, I see from your profile you're in Australia. There's a very limited amount of Google books available in full-text outside the US. I can't access the full-text either from the UK.

>20 Waldheri:, 21 I've been using semi-colons.

Tim, I love this.

One thing: I found myself looking for the Shelf View option and wondering why it had disappeared. You might want to make that box visible all the time, not just when we're below it.

>21 prosfilaes:, 22 - We should probably consolidate formats, or will Tim be able to do some database find-and-replacing? I like the use of semi-colons, though.

Furthermore, do we use the section names given in the first, second or third summary, or from the complete tables? The complete tables often give longer names.

I agree that we should probably consolidate formats; I'll go with whatever other people think is best.

I'd say to go with the longer name, if it exists.

I'm still a little fuzzy on when it's appropriate to modernize the language.

In Zoology (and Paleontology), I updated "Fishes, Batrachia" to "Fishes, Amphibians", even though the term Batrachia is still used in phylogenetics (and is not *exactly* the same thing as Amphibians, although I don't really want to get into that - the details of phylogenetic classification are just as arcane as the those of library classification if not more so).

Should I be doing this across the biological board? Should "Bryophyta" become "Mosses", even though Bryophyta is still a totally valid term? Does "Phanerogamia" become "seed-bearing plants"?

Interesting, so what's the old/new naming policy on categories that have completely changed in contents? 005 jumps to mind, this was apparently "Science" but which now is computer science related stuff.

When you say "and now is" do you mean in the modern Dewey Decimal System? Because my understanding is that this is to get around the copyright limitations of the DDS so you should be using the 1922 values and not the modern values.

28> I was wondering where the numbers came from, and it seems that the numbers are just the current DDNs. Is there a source for MDNs or are you just going to live with the mismatch between DDN and MDN?

Is anyone else filling in the values for entries that have no works? It shows as a blank square, but it can be selected and values entered.

For example: http://www.librarything.com/mds/513.15

This is fascinating and confusing all at the same time. Will this change the Dewey number field in Your Books? I don't see the MDS number in Common Knowledge yet, so I'm not sure where it appears for a particular book.

Does the Dewey Challenge Group need to change names to the Melvil Decimal System challenge? :)

How does wikipedia get around the OCLC to post the Dewey numbers and descriptions?

28> Yes. I find this rather confusing: I can't find computer science books in the 1922 system (the field didn't really exist by then) unless I know that "oh yeah, they replaced complete categories".

30> In my browser (FF 3.6.8 on XP) those squares show up as white with white text. Only by highlighting them can I even see them:

I hate to pile on, but this really seems like one of those "really, you worked on this instead of the huge outstanding list of real shortcoming in LT like not even being able to tell apart 'John O. Dell' from 'John O'Dell' or 'John Odell'?" moments.

>34 brightcopy: seconded.

Also, someone changed the top levels erroneously. This is annoying.

What an interesting way to explore topics! I am inside the 390s which I picked because of an interest in 398 mythology, but am finding quite a variety of subjects. I have not entered very many modern versions though because there is some interpretation involved and I rather like some of the old-fashioned terms, at least for now.

Also, I am using semicolons to separate terms as a dash implies to me a division (i.e. more specific) not necessarily parallel.

Tip: if you have access to the Google book version, opening up Plain Text enables copy and pasting (the transcription/OCR appears to be pretty good).

>34 brightcopy: Personally, I prefer to see small improvements like this coming steadily rather than a complete focus on the big issues that take months or years. The author confusion may be a "real shortcoming", but I actually think this provides more value for time spent.

The 0X0's clobber the X0's and X's, and vice versa.

>36 amarie:: Agreed that this is fun to look at, especially with all the sub-sub-divisions (or whatever they're called) where there are a manageable number of books.

>33 brightcopy:

I've just been changing the URL to get to subcategories that I can't see. If they're never used, oh well. It's the completist in me (at least until I get bored) that wants to add them.

Question: is the name from something like 393 the same as 393.0? I assume the associated titles for 393 are gathered from all decimal divisions within whereas perhaps 393.0 represents only those not .1-.9?

It took a night. I've basically been working 18 hours a day for two weeks. (Tonight I go to Chebeague Island for the weekend.) Much of the time was spent on the most miserable stuff--lots of rewriting code. I needed to put Dewey on a new database footing to move forward on cataloging generally. Once I had that in place, this look from about 11pm to 4am. It was my treat. Don't begrudge me :)

Melvil vs Dewey

I think we should consider using this fellow's system, also composed without reference to in-copyright versions of the DDC, at least for the top levels.

http://everybodyslibraries.com/free-decimal-correspondence/

I was going to propose that, at the least, we start looking at some of the categories and seeing if perhaps new labels would work. That is, it's clear the 0s are about computers. This aboutness is implicit in the numbers libraries have assigned and does not need reference to the modern DDC schedules which, as many of us know (unfortunately), puts computers there.

An interesting time sink, but what use is going to be made of it? It's rather late in the day to set up a rival to the waning principality of Dewey.

>43 TomVeal:

No, it's not really for that. It's for members to see what books they have. Knowing your numbers is less interesting than knowing your words.

PLEASE NOTE THIS THREAD

http://www.librarything.com/topic/97015

I am proposing we make use of the Free Decimal System--another attempt to do something like this.

The 0X0's clobber the X0's and X's, and vice versa.

Fixed. Thanks.

Tim - first of all, it's your site, so screw what I think. So don't take me too seriously and keep in mind that I don't take myself that seriously either.

Okay, having said that - I've seen this "this was only a side project", "this only took a few hours", etc. etc. several times recently. And then that's followed up by days spent in forum threads reporting bugs and suggestions, with a lot of feedback from you. I think you severely underestimate just how much time is spent on little side projects like this.

And yes, I do realize you were working on it at 4am. I was working, too. ;) Couldn't sleep :(

There's just so many serious problems that it gets really frustrating to a lot of us that they're never worked on (that we can see). Things that have been around for years. I could give you a list, but I'm pretty sure you know them by now.

Just my cranky-no-sleep-finally-fell-asleep-at-three-am-my-wife-left-her-alarm-armed-and-woke-me-up-at-seven-thirty-am two cents.

Wait, I don't understand. Do you want to use the 1922 Melvil system, or this guy's FDC? Or do you propose we enter the FDC's section names in the "modern" field?

So, I've changed it on the page in both places.

*We should enter the 1922 wording from the 1922 edition.
*The modern language should come from the FDC or from modernizing the 1922 wording. In many cases the two wordings are the same, in which case you don't need to add the modern working (but you can). And the FDC runs out after about three divisions.

So the FDS (I presume you meant FDS up there) isn't a different system than the DDS, just different words for the same classifications?

Actually, my presumption is that the classification systems are similar but not identical. But in that case isn't using the DDS field but giving it the FDS names wrong in many cases?

I just don't see the point of this exercise.

Would it be possible to fill databases with whatever Dewey-based system? So a field for the 1922 Melvil, and one for any other 'open source' system?

How are we getting along with the LT's Open Non-Dewey, anyway?

FDS is clearly Dewey in one sense. It was built by looking at the earlier system and what was assigned in the real world. That is, he reverse-engineered Dewey.

How are we getting along with the LT's Open Non-Dewey, anyway?

You will get words for meaningless numbers. So you can browse books by terms. Also, when I extend it to your books, you will be able to see how your books are categorized, which works a lot better when there are words, not just numbers, which really only librarians know (and most don't know more than a level or two deep).

It appears that FDS is looking at Dewey and finding Library of Congress Subject Headings that say it the same or better. Though sometimes actually not as well since LCSH is an evolving system also with its own peculiar ways.

>49 timspalding:
An interesting view would be to compare the two in a list format. I am curious about "old names" of things but now one needs to click through to view both versions of individual divisions.

Also, I have been using the modern field for divisions left blank in 1922, like 21st Century literature (which is too specific to be covered by FDS), so I hope that is the right idea. I went ahead and changed eras to add parenthesis since this appears to help the CK format (Melvil just has a space between, for example, Colonial and 1607-1776).

I think you're doing it right. Indeed, wonderfully.

Incidentally, I see this is as being mostly useful to us--to members looking at LT and their books. Libraries aren't going to use it and nobody else cares.

Still, we're going to release this as public domain, without conditions.

I noticed there are no books in LT with a MDS of 174.7. I wonder what it says about us all?

;)

Halfway done with the 900s. It's only taken 6 hours. Think I need to take a break to eat.

Indeed.

How does the system know what books go into what categories? Is that based on the current classifications (i.e., if the category changed, the book shows up in that same number category even if it doesn't fit the description)?

>59 carlym:
I believe they are based on the classification numbers in the book data which are probably from the current DDC. I know they have changed some of the numbers over the years, but are you seeing any that are wildly off the 1922 names/topics?

#56
Are there any book in that category?

>56 readafew:

We're slags and wasters, even as we deny ever having been on the dole. But we know precisely where to shelve our books.

>61 TLCrawford:

It's digital invisible ink. If you highlight the cells (which appear blank), you'll see there is white-font text on a white background.

59: Nothing is changing with the numbers.

All we're doing is adding word labels for numbers.

64: So this is based on the numbers in the existing Dewey field, just with added descriptions? That makes sense.

60: I haven't investigated it enough to know what oddities might be out there, but I think the computer science one (003?) is an example that has been mentioned already. I'm not complaining, just curious.

Add me to the just don't see it crowd. I know more or less what books I have. I don't need to see them listed in some order other than my own.

I don't begrudge you your treats, but I really wish that some of the list of issues could become the treat du jour occasionally and get worked on.

So I was looking at some of the subgroup names under Biology (570-579), and they don't seem to match up well with the books that are listed in each category. I imagine Dewey has been reinterpreted since 1922 for these types of categories?

My library is mostly fiction so I've tended to sort by LCC rather than Dewey, but anyway could somebody please take a look at the new MDS wording for 'American' and 'English' Literature fiction?

813 = Literature > American Literature > Fiction

823 = Literature > British Literature > English fiction (includes: The Secret River by Kate Grenville an Australian author)

843 = Literature > French literature > French fiction (includes: The Tin Flute by Gabrielle Roy, French-Canadian author)

ETA: The MDS original wording seems much clearer to me

Literature > American > Fiction

Literature > English > Fiction

Literature > French > Fiction

>67 jlelliott: Yeah, most the biology books are totally off between 1922 names and current usage. Although I love that a) lampreys have their own category, and b) that it seems to be fairly conserved in the modern scheme.

>68 kathrynnd: It appears that the form is from FDS in following Library of Congress Subjects Headings, at least the part of adding "Literature" to each nationality. I assume the intent of using British instead of English is to include the at least Scotland and Wales, though technically not Ireland. It's CK so anyone can change them, though consensus somehow is always good.
http://www.librarything.com/mds/823

> 68 - There's a note in the 1922 DDS stating that French Canadian authors should be grouped with the 840's. Canadians writing in English may be assigned to either the 810's (American Literature) or 820's (English Literature).

I agree that many books are in the wrong category. That isn't a fault of MDS or DDS. The numbers are assigned by librarians, and their judgments about where particular works belong seem to vary widely.

Thanks Tom, my concern was the modern category wordings, not placement, esp. the change of English to British Literature with English fiction as a sub category.

>70 amarie: I agree the modern MDS may have been following Library of Congress Subject Headings in the use of British, but the FDS form uses English literature (820) and English fiction (823) and Tim has suggested we follow the FDS wording.

>72 kathrynnd: Good point, I changed it to English Literature.

The problem is that some areas have been completely melted down and recast since 1922, such as 510 (Mathematics), 540 (Chemistry), and I think 780 (Music), so in many cases what you're doing is meaningless, even if you do modernize the language.

As others have pointed out, there's been a lot of changes since the 1922 edition. But I'd also thought I'd point out that there's been changes in the editions even in the past twenty or thirty years. And if the source of lot of these numbers are library records from different institutions and different ages, those discrepancies already exist even if we had the current version of Dewey mapped.

Ideally, the library record indicates the version of Dewey in one of the subfield, but I've seen many records where the institution doesn't use that subfield.

I think this means for cases where certain categories have had large shifts or there are only a few volumes that correspond to the call number over a large period of times the oddities will become far more noticeable.

Still, not an end-all, be-all. The same problem has been a headache for other folks with really, really large collections who try to implement some version of a "shelf browse". Still ends up being useful. (Maybe if you went by publication date or something as well you could do some clustering even in a call number range which would help things "line up"). Maybe allow some sort of indicator/tags associated with a particular version of the call number scheme or multiple words.

All the same, I'm probably not going to use the call number browse much when there's tag/keyword/subject type searching on librarything.

Once MDS is fully fleshed out, someone who really liked it could manually change his books' "Dewey" fields to correspond to MDS numbers.

I would like to start translating this into Spanish, but I couldn't find out if there's a Spanish Public Domain edition of Dewey's... should I just translate the titles from the 1922 edition, from the Free Decimal Correspondence, or just wait?

Hmmm. Not too sure this is going to be very helpful. http://www.librarything.com/mds/669.723

#78: What's not going to be very helpful? I suspect most of its value is going to be in the upper levels; anything more than one place after the decimal point is not going to be of much use in many cases.

There is a lot of bad numbers in the system, though. I was running through my read books sorted by MDS and found 813 (American fiction) had a lot of British authors, and more annoyingly books translated from Dutch and Japanese (895.63).

It would be nice if we could mark blatantly incorrect books so that they (optionally) wouldn't show up on the list.

Many, many sources appear to be outputting their local decimal classification, which LT then trusts to be exactly DDC numbers. There are a lot of really bizarre categorizations, though, which must be due to something entirely different.

In some cases two and even three decimals are a definite necessity. Take a look at things like 499 or 899. The class 894.51 lumps together Finnish, Hungarian, Estonian, and probably even some other nationalities. Darn little room for style and genre categorizations there, in contrast to the class 81. For the time being I guess we should be glad that there are not too many works from those regions.

78>

Very interesting pair of books there, Noisy! I would have expected Peter Pan to come under 629...

Dear MDS blog,
I have a few thoughts about MDS category 281 — A fairly obscure section of the classification system, but one about which I happen to know quite a lot.
The first observation: There is no uniform scheme for allocating the subcategories, although the existing ones pretty well cover the field. I suggest there might be at least three methods for doing this job in a better way: Theological, liturgical, or geographical. This last is problematic, because most of the Christians in the Middle East are under such pressure (from war, economics, or religious or ethnic persecution or discrimination) that large numbers are emigrating. And as a result, in Western Europe, Australia, and the Americas, all these Churches are represented.
The liturgical choice would group Eastern Churches according to the liturgical family to which their worship belongs; this would group them into three or possibly four groups: Syriac (Antiochene), Coptic/Ethiopic (Alexandrian), Byzantine, and possibly Armenian (or else the Armenians could be grouped with the Syriac Churches , or else under some other banner (Cappadocian, perhaps?) with the Byzantines). One problem: Some of the Byzantine Orthodox Churches also have parishes or dioceses which use Western liturgies. If this categorization is used, the Catholic Eastern Churches need to find a place here, too, since they use similar liturgies.
Probably the best choice would be theological, at least as an historical marker (since ecumenical dialog is gradually reducing theological disputes among these Churches, and between these Churches and the Roman Catholic Church). This would yield categories such as: Assyrian Church of the East (vice Nestorian); Non-Chalcedonian Churches (with subcategories for Armenian, Syriac, Coptic, and Ethiopic Ethiopian/Eritrean; Chalcedonian Orthodox Churches; and then Heretical Churches (e.g., Arians). This last has its problems (since various Churches including the Non-Chalcedonians and Church of the East have been accused of heresy by others), but needs to be fitted in under some name or other. BTW, the Chalcedonian Orthodox are those we usually call Eastern Orthodox (Greeks, Russians, etc.)
The series 281.1 through 281.4 is confusing and overlapping; if it is meant to include the 'undivided' Church before 1054, then all the Chalcedonian Orthodox and Roman Catholic (282) books covering that period need to be included here and not elsewhere. Arguably, however , some of the Church Fathers belonging to what became the Assyrian Church of the East and/or the Non-Chalcedonian Orthodox which split off before 1054 also ought to be included here (St. Ephrem the Syrian comes to mind). It is probably best to eliminate the Ante-Nicene/Nicene dichotomy entirely, or else eliminate 281.1 (which actually includes both 281.3 and 281.4; so if they are kept, perhaps they should become 281.12 and 281.13). This section also seems to overlap with category 27, Christian History.
Not to beat a dead horse, but which are the "Oriental Churches" of 281.5? True, "Oriental Orthodox" has been used as a synonym for "Non-Chalcedonian", but the term Oriental Orthodox includes both Armenians and Copts and Ethiopians/Eritreans as well as most of the Syriac Churches. Apparently 281.5 means the Syriac Non-Chalcedonian churches and any others they didn't know where to put. It's not a good category as it stands.

Finally, a casual perusal of the books actually listed in MDS under most of the categories 281.1-281.8 are miscatalogued. Most ought to be in 281.9 since they concern the Eastern (Chalcedonian) Orthodox. Perhaps this situation represents the general unfamiliarity/confusion of most Americans about the Eastern Churches. If it's any consolation, the same thing seems to happen in theological libraries, where they ought to know better.

I hope this helps, and leads to better, more consistent cataloguing. However, I'm not sure if these suggestions follow the proposed rules for MDS, and avoid all copyright problems.


Mike Cassidy

> 42
I was going to propose that, at the least, we start looking at some of the categories and seeing if perhaps new labels would work.

Yeah, I was wondering about that, because I was looking through 9XX and its subcategories and some of the descriptions were driving me batty.

Like "South America and Latin America." I see there are books about Latin America in there, so I guess its apt, but it's still weird because half of Latin America is in North America (over in 972) and it's more of a cultural subdivision than a political one, the latter being what most of the categories seem to be using. I changed the "and" to a semicolon just to separate the concepts a bit.

>70 amarie:
I assume the intent of using British instead of English is to include the at least Scotland and Wales, though technically not Ireland. It's CK so anyone can change them, though consensus somehow is always good.

The FDC is wrong on this, because he lists 941 as "Scotland and Ireland" and 942 as "England and Great Britain" which doesn't reflect how the books are actually being sorted--books about Great Britain as a whole are under 941. I've changed it for now.

Isn't it hard to change the labels without changing the categorization of the books themselves? For example, I get the point about South America and Latin America being grouped together, but if the existing categories contain books about both, isn't changing the label just more confusing? (I'm not talking about changing the semicolon to "and"--I mean a substantive change.) I agree that a lot of the categories could be better organized, but it seems too late for that.

The numbers are from libraries and are probably Dewey Decimal in most cases. It seems to me that this exercise isn't about classifying works so much, but is instead about labelling the existing classification in a way that doesn't violate copyright. As such, if the MDS is way off from what the modern Dewey Decimal System is doing, then we should try to find a label that fits the books that are actually under that number. The MDS is just a convenient starting point but it should not be the end point.

I think this is a great idea and will be very useful when fully fleshed out (ie applied to user libraries.) However, I don't understand how the books are being categorized. For example why is Marley & Me the first book listed under technology?

I love this Dewey browser for finding new books in a given category, but when it comes to user libraries, I'd much rather see the Tag Mirror restored and fully-functional.

87> Because the Dewey Decimal Number of the majority of the Marley & Me editions cataloged is 636.7527092 and 636 in the MDS (which partially maps to the Dewey Decimal System) is Technology>Agriculture>Domestic Animals.

The Dewey 600 division has recently been called "Technology" but it is actually a much broader category than that. When I was learning Dewey, we often called it "Useful Arts" (as opposed to "Fine Arts" which were the 700s).

Useful Arts included works on all skills that were considered useful to everyday life. Building, Cooking, Sewing, Child care, Construction, Medical arts, Welding, Leatherwork, and Agriculture. Agriculture includes all domestic animals including pets. (Wild animals are in the 500s. 590s to be precise - Animals.)

Looked at in that light, putting Marley in the 600s makes sense.

>86 jjwilson61:: I understand that no one is cataloging books here; that's why I'm asking about changing labels. If it's the difference between MDS and modern practice, I get it, but I thought some folks were thinking about trying to make the numbering and associated labels more sensible, which is hard to do when the books have already been assigned to categories. To use the previously cited example, we can't move books about North American-Latin American countries out of the South America and Latin America number, so even though that category isn't the best way to sort things, it's too late to do anything about it unless no books about North American-Latin American countries have been assigned that number. (I assume that's why Lampbane didn't change the name.)

Aside from the obvious (seeing the different Dewey categories for my own books), I'd like two things.

1: The ability to add related Dewey numbers
2. The ability to flag (or preferably move) books in the wrong Dewey category.

>92 infiniteletters:
I think both of those would be useful. At the moment, though, I think we should have the flagging option fairly soon (like, right after we get done with the labeling stuff).

Once we get an LC browser, too, I'd like to have an option of adding related LC numbers to Deweys and related Deweys to LCs. Presumably, Dewey-to-Dewey (and LC-to-LC) relationships could be added at the same time.

92.2>

Moving will never happen. It's catalog-data based, and catalog data, no matter how wrong, is completely and utterly sacred. Sure, a lot of the wrong numbers are based on bad source data, but some of them are probably people re-using Dewey for their own purposes, and even if not, erroneous user data is still user data. Just like we couldn't "move" a copy of Moby Dick attributed to Mark Twain in a user's catalog, we can't "move" a copy of Peter Pan that someone has filed under Metallurgy.

You could flag a Dewey number as wrong at the work level and let the next most popular float to the top. It might be nice to have a page where you can see all the Dewey numbers assigned to a work and the number of books each represents and choose the Canonical Dewey or Melvil or whatever number for that work.

And perhaps there could be a section on the Dewey Browser (and LC browser, when we get one) that specifically shows your books that you've given the Dewey number you're viewing.

Great Idea, It woul be nice that the themes definition would reach the sixth and seventh level

example. 150.195.2

Excellent job

library of congress

94, see 95.

Either that or combining. *steeples fingers*

A very minor change:

Could librarything.com/MDS redirect to librarything.com/mds ?

I keep typing in MDS and get a 404 before I remember to change to lowercase.

Here's a weird bug; in the list of Common Knowledge changes, http://www.librarything.com/commonknowledge/changelog.php?user=164934&uid=4c... , I've got a line that says

Aug 22nd, 2010: 5:50:31 pm Die graphische…ünste series eng Sammlung Göschen (75) profile catalog prosfilaes

for http://www.librarything.com/work/7993056

Below that I've got

modern wording eng Chinese literature
modern wording eng Japanese fiction
modern wording eng Japanese literature

all for Die graphische…ünste which links to http://www.librarything.com/mds/7993056

>100 infiniteletters:

Done.

102: Slightly better. :)

Now instead of a 404, I get "Not a valid MDS". http://www.librarything.com/MDS/003.3

To beg the help from the people who owns a 1922 version would be the wisest idea. try to find another decimal class definition is just running in circles.

In that case use the OCLC technique

> Not a valid MDS

I'm not seeing that on the very page you link to. Am I misunderstanding.

I saw it when infiniteletters posted, over a month ago; maybe it took a while for the upper-case version to percolate? It looks fine to me now, too.

Uppercase is working fine for me today; thanks for adding URL case-insensitivity. :)

Did anything come of this in the end? Is the data being employed on the site somewhere and I just haven't noticed it (entirely possible)?

It's used in the MDS browser.

I'd love an option to show ratings beside the titles, and to sort by rating....

I hadn't looked at this in a while, so when I followed Zoe's link I noticed the green checkmarks for (I think) the first time. I was checking out the top levels, thinking "Cool, I have popular books in all the headings." Then I hit history and had to hit next before there was a book in my library.

I can't quite tell if this means I'm not as interested in history as other topics or my interest is just more focused and thus I'm not likely to buy popular titles. Both could be true, I suppose.

It's an open-source variant on the Dewey Decimal System. The two lines with pencils are for people to enter modernized and corrected labels for the numbers selected - because the last public-domain DDC listing was 1922 and it was...somewhat archaic.

Beyond that - you don't put your books into it; you could put it into your books, in the DDC field, if you wanted to (actually, as far as I know, the numbers are the same for the same subjects). Or in the new classification fields that are supposed to be coming up. You can see what books you own that fall into each classification by looking at the list below the 'pretty boxes that open' and looking for green (or other color) checkmarks.

Beyond that...I suppose it still isn't finished.

I think it's sorted by number of copies.

That I don't know; I've never looked beyond the first couple of pages.

I don't believe Amazon returns the dewey number so it must come from library sources. Not that that can't be ratty too.

>115 Collectorator:

I believe "Not set" refers to numbers that the 1922 edition did not have a decimal classification for. Books seem to be placed under the modern decimal classification that is imported from library sources.

You can always follow Tim's link in message #1 to verify that the 1922 edition didn't have an entry—it may have been missed. Otherwise I don't think we are making up classifications to cover those entries yet.

Is there a way to access the MDS programmatically? REST would be fine; indeed ideal. What I am looking for is a way to import the MDC categories as genres in Delicious Library 2 via Applescript. I can (probably) manage the Applescript part, if there is some way to access the classification names by DDC or (preferably!) ISBN using REST.

So apparently there is some disagreement as to the capitalization of the word "and" in the MDS. I prefer lowercase, but apparently some people disagree (http://www.librarything.com/commonknowledge/changelog.php?item=1&type=201&field=39). Is there a standard rule on the subject?

>124 Calion: I prefer lowercase too, but I've pretty much given up on standardizing the style, even in my own usage. And, and, & & are used fairly interchangeably; in my view, the incompleteness issue is much more pressing. Style can be standardized later.

For the last week or so I've been putting in an effort to flesh out descriptions for the lower rows, and most of the 000s to 200s are now looking much less skeletal. I figure it's the least I can do before I start squeaking my wheel for LCC.

(Which, in turn -- at the risk of tipping my hand -- is the least I can do before I try to broker some sort of cooperation whereby LT becomes a catalyst for the completion and popularization of BC2, thereby saving the world. Or something like that. Hey, a guy can dream, right?)

Finally, although it's low on my personal wishlist, I think replicating the existing UDC outline would be pretty low-handing fruit, and it's CC-BY-SA... (http://www.udcsummary.info/php/index.php?id=13358&lang=en)
And unless/until a critical mass of the existing user base decides to jump on board (Continental Europe, I'm looking at you...), the outline should more than suffice, thank you very much.
Given the fields, I'd gladly do all the data entry required; should take a weekend at most, and far less with many hands.

ETA:
To put it in more appropriate terms: after MDS 025.431 comes MDS 025.432, then MDS 025.433, then MDS 025.434. The road forward is clear. :-D
(And hell, if we ever get a working Z39.50 for an Indian library, I suppose MDS 025.435 still exists, too!)