Library Cataloging System for a Multidimensional Library

[AlHazred]

Here's a poser that's entertained me for weeks now on and off, and I thought I'd share for the other bibliomaniacs on the list.

Say you've got a Library. But not just any Library -- it's the Library of Babel (http://jubal.westnet.com/hyperdiscordia/library_of_babel.html) from Jorge Luis Borges' story (http://en.wikipedia.org/wiki/The_Library_of_Babel), or L-space (http://wiki.lspace.org/wiki/L-space) from Terry Pratchett's Discworld (http://en.wikipedia.org/wiki/Discworld), or whatever; it's the Library in another dimension that has access to all of the universes where writing exists. They've got the books H. P. Lovecraft would have written had he gone to the doctor in time. They've got the books from the Library of Alexandria; they've got all the works of Socrates and Aeschylus and the rest. They've got Plato's Hermocrates, describing both the detailed history and current location of Atlantis. They've got variant stories by your favorite pulp author, including the stuff that wasn't published here, but was a big hit in the Universe Next Door.

The question is, given this hypothetical Meta-Library, how do you catalog the books? If you don't catalog them, then you'll never find anything. You've got, say, sixteen thousand versions of The DaVinci Code, all slightly different; if you just put them in a huge pile in the "Dan Brown Wing," you'll never find the one you're looking for right now, that's got that great plot twist in chapter 16, or the one with the Appendix giving DaVinci's cipher keys.

If you use our world's systems, you've got other problems. Books with the same title and author will get the same number under Dewey Decimal (not sure of the other systems, like the Library of Congress) -- instead of an undifferentiated pile in the "Dan Brown Wing" you'll have them on sixteen miles of shelf between Aisle 44-F and Aisle 10e23.

How do you catalog this mess?

[sinanju]

Until you have a meta-numbering system for clearly identifying the different universes (and thus the books published in each one), you haven't got a prayer. Preferably the universe numbering system would be objective, and clear enough that deciding how to identify any newly-discovered universe would be obvious. In practice, it's almost certainly go be subjective; you start with an arbitrarily determined "base" universe, and define all others by how they vary from it, perhaps with something as simple as the order in which they're catalogued by the library (not necessarily the order in which they're discovered).

And even then, most books won't be identifiable by which universe they come from unless you have the...can't think of the word for it, the "chain of custody" they use to authenticate ownership of works of art, the "pedigree" of the book in question so you can be sure it came from Universe X1743A.

Any way you slice it, it's going to be a nightmare to administer.

[AlHazred]

My current thinking is you need two to three different systems for each book, like a coordinate system for a two- or three-dimensional space.


You need an equivalent to the Dewey Decimal/Library of Congress/whatever system, to describe the book's topic and provenance.
You need a unique identifier for each Universe, to describe where it was written or at least published.
You might need an identifier to identify timeline -- if we assume alternate universes, alternate timelines is also possible. Though, there's a whole lot of crossover for those last two, and you might get by with one multidimensional number/code for that one.

[ghost-angel]

I would say you just need to add on a Universe Specific SKU to either the front or backend of the any other system.

You could get specific with things like Base Dimension or Variant Dimension plus a unique number. I wouldn't imagine it'd actually be complex to implement. Just tough to remember really long numbers.

[AlHazred]

The reason I'm asking this is it might play into a dimension-hopping campaign bit I'm planning. I'm not sure how easy or hard I want to make it for the players to navigate, but I thought it would be interesting to make a sort of prop, where they could scan a list of available titles.

[Markdoc]

I'm with Ghostangel on this one - the 10 digit ISBN identified a book by author/title/edition. The 13-digit ISBN can seperate the same title by publisher/language group/country (and there's a 14 digit GTIN (Global Trade Identification Number) on the way, which will refine regional issues). Amazon has had to deal with this problem already where a single book might exist in a dozen or more different editions, all slightly different. All that happens, as noted above is that you end up with a (say) 20 digit number cataloguing author/title/edition/publisher/language/region/alternate world/timeline.

cheers, Mark

[Lucius]

and a code for "unknown" if no one knows where (when?) the book came from, and multiple editions of unknown provenance are marked a, b, c, etc.

Lucius Alexander

Browsing with a palindromedary

[Nevenall]

I would circumvent the entire notion of a cataloging system because the shear magnitude of changes would be too much for anything but a hyper intelligence or celestial being to keep up with.
What if a user could just direct a question to some "thing" that that thing would respond with pertinent questions of refinement.
"Are you interested in the unpublished works of Xenophon from the first second or third earth dimensional sub group?
Maybe it's a computer, maybe it's a spirit being that embodies books and uses magic to find what you want.

Also, that sounds like an interesting potential campaign.

[dmjalund]

also - to find out about the universe the book belongs to use the Universe ID code (with Timeline subcode) to find the Encyclopedia that also belongs to the same universe

[Ian Mackinder]

Given just how complicated the entire parallel timeline / universe thing would be, I think the only way to catalogue a Multi-Dimensional Library is to first develop a cataloguing system that is itself multi-dimensional (as in considerably more than 3D). Straight alphanumeric progression simply may not be up to the job.

[dmjalund]

all written information is inherently 1 dimensionl - including indexes, it is how it is organized/categorized that allows it to deal with the 2d and 3d of real life. adding a couple of extra dimensions should be no big deal

[Ian Mackinder]

all written information is inherently 1 dimensionl - including indexes, it is how it is organized/categorized that allows it to deal with the 2d and 3d of real life. adding a couple of extra dimensions should be no big deal

Sure, but when using that to work out and display an organizational chart extending into several additional dimensions (at least), one's command of written expression gets kinda challenged. :)

[dmjalund]

Sure, but when using that to work out and display an organizational chart extending into several additional dimensions (at least), one's command of written expression gets kinda challenged. :) terminology just gets converted into codes we don't understand anyway

[BobGreenwade]

The basic job of determining where a book is shelved could be done just fine using one of our existing systems.

Given that part of the catalog's job will be to find a given edition, a computerized index will be needed. That would help find books from (say) Dimension M-616 or DC-52 or L-41191, so as long as the user knows the designation for the universe of origin one should be able to manage things.

And if you're looking for specific features like an index or glossary, existing computerized catalogs can do that already.

(In short, your initial line of thought should be good enough.)

[Cancer]

Put me in with Ghost-Angel and Markdoc. Stack another pair of catalog numbers (for identification of universe and timeline) onto existing library classification schemes and I think it would work.

The truly interesting part of the library would be the self-documentation of the classification schemes for those other parts of the classification. Where those belong in the library is clear, but it'd be very fun to read....

[Lawnmower Boy]

I'm not seeing the difficulty. Any string of n numbers can be read as an n-dimensional array, giving you as many ways of sorting books as you need. Cross-referencing books on a given subject from different dimensions would be a breeze compared with finding volume 7 of the official history of the German Navy in World War One at your local university library right now. (No shelf scanning. That's cheating!)

[Nevelon]

Put me in with Ghost-Angel and Markdoc. Stack another pair of catalog numbers (for identification of universe and timeline) onto existing library classification schemes and I think it would work.

The truly interesting part of the library would be the self-documentation of the classification schemes for those other parts of the classification. Where those belong in the library is clear, but it'd be very fun to read....

When you are talking about all the information from all the parallel universes the problem is less how it is organized, but more of who knows what's important. I think the librarians are going to be very important. You would need a whole pack of xenolibrarians. The index and a map will tell you where the books are, but who is going to tell you that the Ernest Hemingway from Dimension 432f5a wrote everything in limericks, or that everything published in dimension 863ab4 is porn? You would have specialists in various fields, and when you needed to know something, you'd find the right person, who would then help you find the information.

[Burrito Boy]

They've got variant stories by your favorite pulp author, including the stuff that wasn't published here, but was a big hit in the Universe Next Door.


I assume that includes the new Robert E. Howard stories in "Jeffty Was Five" by Harlan Ellison.

[AlHazred]

Sure. I'm sure they've also got all of Sherlock Holmes' monographs on the subject of detection and crime-fighting, Richard Castle's full run of mystery thrillers, and the complete Books of Bokonon. Me, I want to read all twelve volumes of Life by Unspiek, Baron Bodissey.

"Do you read Sutter Kane?"

[AlHazred]

Here's part of why this is a puzzler for me. Assume you've got a Universe designation - say, Earth-37A. If you assume, as many authors do, that these different parallel Earths are the result of a single choice at a single, discrete point of space-time, then you suddenly have to refer to Earth-37A-1 and Earth-37A-2. Your best bet is probably some sort of array formula that determines the exact point in multi-dimensional space-time where the work was produced; the arrays will quickly become insanely large, but there you go.

Of course, this is exactly why Borges had his Library of Babel in such chaos, and why in his Book of Sand (with infinite pages) it is impossible to find again anything that you've seen once.

[Cancer]

I dunno. This sounds more like a problem from an introductory course in data structures to me. I recognize that the logarithm of infinity is still infinity (and trees reduce search times logarithmically), but before you have an infinite library you can still organize a lot of data.

[austenandrews]

Does the act of cataloging a book create an alternate timeline in which the book wasn't successfully cataloged?

[Markdoc]

Does the act of cataloging a book create an alternate timeline in which the book wasn't successfully cataloged?

Yes, of course - but since it wasn't catalogued, you don't need to worry about cataloguing it :)

cheers, Mark

[Ian Mackinder]

Schrodinger's Library?

[Nevelon]

Schrodinger's Library?

Where all the books are both checked out and on the shelf until you go to get them.

[Old Man]

How do you catalog this mess?

I'm late to the thread, but I read this and immediately thought of just a really, really big card catalog.

[Cancer]

Does the act of cataloging a book create an alternate timeline in which the book wasn't successfully cataloged?


Yes, of course - but since it wasn't catalogued, you don't need to worry about cataloguing it :)

... which brings up a variant Russell's paradox: the library cataloging all libraries which do not catalog the those libraries which have themselves in their catalogs ...

[Beast]

how about this

since info is what is wanted,I figure a contact is in order
pretty much had to buy it to the highest levels but making it a PDA and takes a full phase to send off the info
the uncontrolled is a GM choice on how long it takes the library assistant to find the information
I made it usable as an attack so you can find out info on your target(figure it has a built in camera or data input)

you might want to add Eidetic Memory to it so you do not have to call the Library assistant back if you for get the info(might start getting a bit pissy)


Trans Dewey decimal library card/PDA: Contact (Contact has access to major institutions, Contact has extremely useful Skills or resources, Contact has significant Contacts of his own, Contact is slavishly loyal to character, Contact limited by identity), Usable As Attack (+1), Transdimensional (Any Dimension; +1), Spirit Contact (x2), Organization Contact (x3) (66 Active Points); OAF (-1), No Conscious Control (Only Effects cannot be controlled; -1), Concentration (0 DCV; -1/2), Extra Time (Full Phase, Only to Activate, -1/4) 13-

[IndianaJoe3]

I just thought of another wrinkle to the problem. This cataloging system has to contain information as to whether a particular book is from a universe that is relevant to the searcher. Is Sherlock Holmes a noted amateur criminologist, fictional character, or police detective that dabbled in writing about the fictional adventures of a London ophthalmologist?

[AlHazred]

Beast may have hit upon the least expensive solution. My current thinking is a really, really good-roll Detect, with Dimensional, MegaScale, and a host of other Adders/Advantages.

But of course, the real bone of contention isn't necessarily the power build (after all, you can build anything in Hero) -- it's how to justify it. :)

[Ian Mackinder]

I just thought of another wrinkle to the problem. This cataloging system has to contain information as to whether a particular book is from a universe that is relevant to the searcher. Is Sherlock Holmes a noted amateur criminologist, fictional character, or police detective that dabbled in writing about the fictional adventures of a London ophthalmologist?

... Or, to use the premise of 'Without A Clue' (one of my favourite Holmesian movies), was it Dr Watson who was the real brains (select one of the above), and Holmes just the front man?

[Curufea]

Shame I didn't see this thread earlier. I've an acronym for you - FRBR (http://en.wikipedia.org/wiki/Functional_Requirements_for_Bibliographic_Records)
It's the latest hot thing we use in library land :)
You are basically breaking down an item's relationship to similar items in a hierarchical manner. It's great for expressing different editions, and it gives a reasonable basis for relevance ranking in search results.

As a schema, it can also be expanded upon to include subdivisions or even higher divisions (such as the dimension the item is from).

It's not a classification system per se - but most of the current existant systems are very inadequate for even today's books, let alone tomorrow's media, or fictional dimensional media. But it can be used as a basis for creating a classification system.

[AlHazred]

FTFA:


FRBR comprises groups of entities:


Group 1 entities are Work, Expression, Manifestation, and Item (WEMI). They represent the products of intellectual or artistic endeavour.
Group 2 entities are person and corporate body, responsible for the custodianship of Group 1’s intellectual or artistic endeavour.
Group 3 entities are subjects of Group 1 or Group 2’s intellectual endeavour, and include concepts, objects, events, places.

Group 1 entities are the foundation of the FRBR model:


Work is a "distinct intellectual or artistic creation."
Expression is "the specific intellectual or artistic form that a work takes each time it is 'realized.'"
Manifestation is "the physical embodiment of an expression of a work. As an entity, manifestation represents all the physical objects that bear the same characteristics, in respect to both intellectual content and physical form."
Item is "a single exemplar of a manifestation. The entity defined as item is a concrete entity."

So, let me get this straight. Each discrete book in a library (a Group 1 entity), would have four codes of reference: one indicating this specific work (e.g. ALHAZ001 for the Necronomicon), one indicating what kind of book it is (e.g. PYSCROLL for a papyrus scroll), one indicating something to do with this specific form of item (e.g. ALAZIF for the Arabic original text which was called Al Azif), and one indicating this instance of publication (e.g. 738DMALD for a first edition published in Damascus in 738 by Al-Din the bookbinder). Do I have that substantially correct?

[Curufea]

Not really, no.

The work would be the Necronomicon.
The Expressions are the various ways it occurs - translations, audio books, editions.
Manifestation is how that expression is produced - e.g. hardcover published by Oxford University Press in 1999, Penguin paperback, audio book read by Neil Gaiman, papyrus scroll.
An Item is the physical thing.

All of these objects (descriptive records) are connected to each other, and to the author/s involved. When a user looks up a work, they would be able to see all of the different editions/versions nested together, including the availability of the physical items attached.

Example record for At the Mountain of Madness: http://trove.nla.gov.au/work/9640772

This site shows relationships for a Harry Potter book-
http://www.frbr.org/eg/hp-goblet-1.html

[shadowcat1313]

are audio recordings cataloged by the Dewey Decibel System?

[Curufea]

are audio recordings cataloged by the Dewey Decibel System?

No. The Dewey Decimal Classification System does not record media format. However MARC records do. Modern systems use MARC (http://en.wikipedia.org/wiki/MARC_standards) (MAchine Readable Cataloguing) or DC (http://en.wikipedia.org/wiki/Dublin_core) (Dublin Core) for their records. DC is primarily used for non-book items as it has more relevant fields though it isn't as detailed as MARC. DC does have the advantage of being more easily transformed into XML however. And XML = easier machine-to-machine interfaces. Hence why I use it a lot these days (and am even translating a very bad HTML reference page on Doctor Who to XML (http://www.curufea.com/rassilon/) at the moment)
Anyhow, the holdings area of the MARC or DC for the record is where libraries put the call number (or shelf number) for the item. It can be Dewey, Library of Congress, or any other classification system they like.

Call numbers in MARC for the record are in tags 050 to 099 - http://www.nla.gov.au/librariesaustralia/cathelp/bib/ecbdclas.html
Which library holds the item and what that library's call number is (in case it differs from the above, usually with prefixes and suffixes) is stored in the 850 tag - http://www.nla.gov.au/librariesaustralia/cathelp/bib/ecbdhold.html
Physical description of the item is in the 300 tag - http://www.nla.gov.au/librariesaustralia/cathelp/bib/ecbdphys.html

[Curufea]

Here's an example MARC record for an electronic version of "At the Mountains of Madness".


000 00780cam a22002293a 4500
001 000044735737
003 AuCNLKIN
005 20091009135052.0
007 cr cn|nnn||n|n
008 090419s2009 at es 000 f eng d
035 __ $a(OCoLC)450681233
040 __ $aSUA$beng$cSUA
100 1_ $aLovecraft, H. P.$q(Howard Phillips),$d1890-1937.
245 10 $aAt the Mountains of Madness$h[electronic resource] /$cH.P. Lovecraft.
260 __ $aAdelaide :$bThe University of Adelaide Library,$c2009.
500 __ $aThe University of Adelaide Library eBooks @ Adelaide.
506 __ $aFreely available.
516 __ $aElectronic text.
653 __ $aLiterature
830 _0 $aeBooks @ Adelaide.
850 __ $aSUA$b1353267$cOnline access
856 40 $uhttp://ebooks.adelaide.edu.au/l/lovecraft/hp/mountains


The record that users see is here-
http://trove.nla.gov.au/work/9640772?selectedversion=NBD44735737

Sorry, but I am a librarian and this is my work - so apologies for too much information.

[shadowcat1313]

ok, never try to outpun a librarian

[ghost-angel]

Shoot, for a build - unless there's some burning need to make it a major part of the campaign

Access Perk; 5pts - information is based on GM needs and Player requests.

I wouldn't take it any further than that mechanically.

[Curufea]

For something as large as a multidimensional library, the simplest answer is to use no classification system at all.
Use accession numbers. Give each item a unique number.
All the "finding of stuff" happens with the records about those numbers.

[The Weapon]

One thing you need to watch out for is subjects that are classified differently in different universes. For example the "eaters of hashish" are classified as a historical cult in this universe, in others they're classified as a revival movement that brought Islam back to it's proper roots against the brutal resistance of heretics.

[Lucius]

One thing you need to watch out for is subjects that are classified differently in different universes. For example the "eaters of hashish" are classified as a historical cult in this universe, in others they're classified as a revival movement that brought Islam back to it's proper roots against the brutal resistance of heretics.

Astrology and Astronomy would have completely different classifications in one world, and in another be basically the same subject.

Lucius Alexander

Classify a palindromedary

[TSandman]

Ooook*




* But how do you catalog an infinite number of documents?

[Curufea]

One thing you need to watch out for is subjects that are classified differently in different universes. For example the "eaters of hashish" are classified as a historical cult in this universe, in others they're classified as a revival movement that brought Islam back to it's proper roots against the brutal resistance of heretics.

This is basically what authority files are for, subject or title authority - in this case subject. We're currently working on people authorities at my work as well (ie how do we know this John Smith the actor is also this John Smith that wrote this book or this John smith the CEO)

http://en.wikipedia.org/wiki/Authority_file