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

How Dewey works

Build the Open Shelves Classification

Join LibraryThing to post.

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

Sep 14, 2008, 7:50am Top

As a number of you are having trouble mapping the new classification categories to Dewey I thought I'd give you some background on why it's such a pain.

Dewey is not a subject classification scheme. It is based on discipline. It was devised at the end of the period where people could go to a library to self-educate themselves (read the law, for instance) It assumes that you are trained (or are self-training) in a particular discipline and that you want to read books related to that discipline, not all the books on a particular subject.

So, let's take the subject transportation. If you are trained in economics, you think of transportation systems as a way to move goods from one place to another, so there are transportation books in the 380s. If you are a city planner, you think of transportation systems as a way to move people around a city, so books that deal with that aspect of transportation are in the 711s. If you are an engineer, you think of transportation as something you can build, so there are books on engineering various means of transportation in the 620s and on manufacturing them in the 680s (? - Don't have the schedule open at the moment). If you are running a business, including one that involves transportation, the books on the internal workings of business are in the 650s. If you are worried about the environmental effects of buses, it's with social problems in the 360s. If you're an engineer trying to find out about the composition of bus exhaust, it's in the 660s or the 621s, since a lot of exhaust is nano-particles and the last revision was rather stupid about the way it dealt with nanotech.

You get the general idea. So there's no good way to get your top categories mapped onto Dewey in any complete way. The map that was discussed in the top-level categories thread will probably do what the coordinators want it to (with the corrections offered by several people).

Sep 14, 2008, 3:04pm Top

I think the mapping of Dewey to Top-level is to get some approximation of hoany books might be under any given category.

You're right about 'discipline' being the guiding term, rather than subject. When I was at the university of Pennsylvania in the 1960's, its library system still used Dewey. One of the departmental libaries was the "Fin Arts Library", and it included 4 departments, fine arts, architecture, landscape architcture, and city planning. Obviously, they didn't all get along with one another. So, all their books were divided up into two camps, those under the purview of fine arts and architecture, and those under landscape architecture and city planning. The former had shorter loan periods and many books that could not circulate. The 380's had been given to Fine Arts, but City Planning wanted people to read transportation books. So, many transportation books that would normally be classified as 385 (railroads) or 388 (public transit +) were given a 711 call number, so that City Planning students would have no restrictions on how long they could be taken out.

In the 1970's, the University of Pennsylvania converted to Library of Congress, so I suspect these rules do not exist today.

Sep 20, 2008, 10:56am Top

vpfluke, the U of PA situation you describe sounds like a case where a better solution might have been to simply put a prefix for the particular school's library rather than to change the call # that was actually the most accurate.

However i think that it would be better if a classification system was based on the actual entity/topic/subject rather than a particular disciplinary approach, altho it makes sense to use disciplinary approach as one of the ways a particular subject could be arranged.

btw I believe LC basically tends to make discipline the basis of its system also. Altho in the case of LC, the various subject areas were constructed by various subject depts of the L of Cong with sometimes not too much consultation with librarians from other subject areas.

I have to confess that of the 2 imperfect systems, I definitely prefer DDC because with all its faults it is more consistent, and its notation is more user-friendly. LC was made for one particular library. If I was the person to make the decision as to what system an academic library should choose and I had to choose between LC and DC i'd definitely choose the latter. Hey, it worked for the British Library, which is about as big, if not bigger, than the L of Congress

Sep 21, 2008, 8:17am Top

The problem w/ Dewey is the labeling. The labels only hold 9 characters across in a readable font and in an academic setting the Dewey numbers often go out to 15 or 20.

I thought LC was subject not discipline based, but I haven't done any extensive classification in LC.

The U of Pa thing was a peculiarity of large academic libraries in the 60s. They thought if they developed specialist libraries it would help their graduate students. They forgot that it would have the undergrads running all over creation. It kept student workers busy for about a decade, first moving the books to the separate collections and then re-integrating then.

Edited: Sep 21, 2008, 8:42am Top

>4 aulsmith:, The problem w/ Dewey is the labeling. The labels only hold 9 characters across in a readable font and in an academic setting the Dewey numbers often go out to 15 or 20.

Why are you limited to 9 characters in Dewey? Use a two-inch label and run it along the spine just like LC. You can easily fit 12-15 (or more) digits in that space. The labels wouldn't be upright, but neither is LC. In fact, neither are some DDC numbers on thinner books in most libraries.

If labeling is The problem with Dewey, that can easily be worked around. I think the problems with Dewey are far more complex than labeling :)

That said, I think that DDC is easier for the average user to understand at a glance than LC; and as an employee of a public library, ease of use for the patron is important to me.

Sep 21, 2008, 9:05am Top

I meant the problem with Dewey for large research libraries is the labeling. Dewey has a bunch of problems, most notably the religion section.

Why we don't use those other labels? I'm not sure. You also have a add a location, Cutter and, depending on how you cutter, an edition distinguisher. So the total label would run in the 30 to 40 character range. There's also the problem of standard label printers. I work at a place big enough that the labels are chosen by an entirely separate unit from the Cataloging Dept., so I'm only guessing why we do things the way we do. The bottom line is we cut the Dewey at 5 decimals, rendering some sections completely whacked. Fortunately we have closed stacks, so it doesn't matter.

If you work in a public library, you're cutting the Dewey's a 3 decimal places, aren't you? That's what most public libraries do.

I do think that Dewey's progression from the general to the particular is very useful in browsing, but the way Open Shelves has been developing, I don't think I'd find a lot of agreement.

Sep 21, 2008, 9:35am Top

As someone who just spent the past 4 months filling all the 'holes' in Dewey column of my spreadsheet version of our library inventory, I can attest to how woefully inadequate the DDS really is. We're in the process of reshelving all of our books now (~have~ to do it!), but we'll be reconsidering an alternate method shortly thereafter.

From my point of view, I find Ethics being a subcategory of Religion nonsensical. They are 2 entirely unrelated subjects. Also, the whole way Literature is subclassified is tedious at best, but I am at a loss to try to redefine it.

Sep 21, 2008, 5:15pm Top

General ethics is in philosophy (170). Christian ethics is in religion (240?); Buddhist ethics with Buddhism; Jewish ethics with Judaism ...

The miscellaneous category in literature is weird, but it's actually pretty easy to browse the shelves. And it's much better than LC where they skipped every other number in a chronological system. When they run out of numbers they're either going to have to reclass or double back and start shelving the newest stuff between two sections of older material.

Sep 21, 2008, 9:13pm Top

I took a look at the University of Pennsylvania website. They still have 15 departmental libraries. However, the cleavage at the Fine Arts Library described in Message 2 no longer seems to exist. Penn has gone to Library of Congress, but there are lots of books still in Dewey, including its aberrant economics classification, where railroads were classified by the Wharton School of Business Library (Dietrich) at 380 rather than 385 (a specific book about the problem of railroads in New York State in 1940 is classified as 380.837S N7A7). The main library (Van Pelt) used 385 for railroads.

I wonder if there are any academic libraries in the U.S. which are still cataloging new books in Dewey? I'm sure Penn is not the only academic library that has old books in Dewey.

Nov 11, 2008, 5:03pm Top

Purdue is on Dewey. Many smaller academic libraries use Dewey (I could give a handful of examples off the top of my head).

Another point about Dewey: the philosophy of DDS is very different from what Open Shelves intends to do. Open Shelves is all about organizing the books we have in (public) libraries now. Dewey intends to organize all of the world's knowledge into a logical system.

Dewey's system is rather presumptuous, and that is (I think) its downfall. It was set up in such a way that new categories cannot be added in a way that is useful. For a great example, look up the books on computers in your public library. They're shoehorned into the 000s. Dewey had no idea there would be computers.

Nov 11, 2008, 5:20pm Top

>10 karl.stutzman:

Sometimes it's a saving grace. He'd probably have interjected some weird racial or linguistic bias into the category, or balanced it such that 9/10 of the computer category was for computers employed by Christians.

Nov 11, 2008, 5:43pm Top

I work at Pratt Institute, which still uses Dewey. We were a public library back in the early part of the past century and never bothered converting our collection when we had the opportunity. We will probably never have the staff or funds to convert the collection to LC. As an art library, we therefore have multiple floors of 700s.

Edited: Nov 25, 2008, 11:03am Top

Which means in LCC you'd have several floors of Ns. I don't think many specialized collections do well when shoehorned into general systems.

Edited: Dec 5, 2008, 4:13pm Top

In response to #12

I personally wouldn't see any really good reason for Pratt to switch from Dewey to LC. They are both general classifications (as post #13 notes) and frankly i think (my opinion only of course) of the two, Dewey is more browser-friendly than LC. I understand the British Library, surelly comparable in size to the L of C, uses Dewey.

PS: my mom was a Pratt grad!

Dec 6, 2008, 12:43am Top

The Mid-Manhattan branch of the New York Public Library, basically the circulating arm of the main research library across the street, uses Dewey rather than LOC.

Big research libraries aren't usually browsable by common patrons. I think some libraries went to LOC because it was easier to buy LOC catalog cards for their own library's card catalog.

Dec 6, 2008, 9:24am Top

It's also much easier to class in LC than in Dewey.

Edited: Jan 4, 2009, 2:51pm Top

I am new to this group, but I have something to say about a couple of things if it's not to late.

1) I do not like, as an Architect, to categorise my architecturebooks under Art & Architecture, It feels quite wrong. You can of cause find some art in architecturebooks, but as a whole it's problematic.

For now I have used Amazonbookstore's catagorizing and they have been tagged under; "Professional & Tecnical, Arts & Photography"
This does not feel right and Architecure are as much an own Universityfield as Psysics and Medicine and are so much more than some books about beautiful houses.
Therefore I suggest Architecture as an own category.

Under is some examples of different ways I have tried to categorise some own my books:

architecture, history,

Art, history, architecture, folklore,

Architecture, landscape, garden

Architecture, Technology, Building, Craft

Architecture, History & Periods, International, Norway, Oslo, food, cooking

Architecture, History & Periods, International, Sweden

Architecture, Interior Design, Asia

Architecture, History & Periods, International, Norway

Architecture, Technology, Project Planning & Management, energy

Architecture, Landscape, Garden

Architecture, Landscape, History & Periods, Garden

Architecture, Landscape, Garden

Architecture, Interior Design, Achitects A-Z, Capability Brown

Architecture, Interior Design, Achitects A-Z, William Morris

Architecture, Technology, Project Planning &
Management, energy, fireplace

Architecture, Technology, Craft, Woodwork

Architecture, Interior Design, Building Types & Styles

Architecture, Urban & Land Use planning

Architecture, Building Types & Styles, Theory, debate

About Cooking. This is not something that just happens at home. I have a lot of cookbooks, but my way of categorise this is to start with FOOD.
Food can be so much. Professional cooking, nutrision, healtfood, industrial foodmaking, cooking, etc. So I think "Food" should be the upper category, and everything about food down from there, altso cooking (not cokery, ugly word)

Jan 4, 2009, 7:17pm Top


Some of us are less hierarchical than others in classifying. My own tagging is not hierarchical, and neither is my bookshelf arranging.

Food comes from plants and animals, but I assume that biology and agriculture should be separate categories.

Maybe 'cookery' was used to broaden the concept of 'cuisine'. Anyway, 'food' seems a better choice.

May 10, 2009, 11:31am Top

I worked for many years in the Gallaudet University Library, which has the world's largest collection of books and other materials related to deaf people and deafness--over 92,000 books--and which still uses Dewey plus Cutter numbers.

Standard Dewey crams virtually all deaf-related books into 305, 362, and the 370s, but this is totally inadequate for the size and scope of the Gallaudet collection. After a two-decade fling with an in-house-created classification system that ultimately proved inadequate in itself, the deaf collection was recataloged by recasting the entire Dewey system within a "deaf" context and adding a collection prefix to distinguish it from the main, non-deaf library collection.

So, for example, literary works by deaf persons are in the DEAF 800s, subdivided similarly to regular Dewey; a biography of a deaf astronomer goes into DEAF 520.9; religious ministry to the deaf is somewhere in the DEAF 200s; a work on American deaf history is in DEAF 973; and a book on the Nazi attempt to eliminate deaf people from the population in the name of "eugenics" goes into DEAF 943.

Even with putting the Dewey system within this "alternate world", it still has to be modified in some places to give adequate classification detail. One good example of this: Standard Dewey puts both American English and British English languages into the 420s. But, for historical and geographical reasons, American Sign Language and British Sign Language are totally different from each other, not the same language at all or even related to each other, so they cannot be classed together. The Gallaudet Library had to modify the 400s (Language and linguistics) area to put American Sign Language into the DEAF 410s, and British Sign Language finds itself displaced to the otherwise-underused DEAF 480s. The DEAF 400s had to be rewritten and expanded to not only accommodate the hundreds of natural sign languages that exist around the world, but also to accept other deaf communication systems such as pidgin sign languages, artificial sign languages, speechreading, cued speech, and others.

Similarly, the entire DEAF 610s (Medicine) class had to be rewritten to provide adequate classification detail for the thousands of audiology-related books in the collection, and some other classes are also modified or expanded.

May 10, 2009, 11:27pm Top

# 19

What you have written is useful to know. At some point, people have talked about using facets to have a different dimension of classifying books. I think people want the OSC to be flexible and not restrictive.


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


No touchstones

You are using the new servers! | About | Privacy/Terms | Help/FAQs | Blog | Store | APIs | TinyCat | Legacy Libraries | Early Reviewers | Common Knowledge | 114,464,346 books! | Top bar: Always visible