025.431: The Dewey blog

Genrifying with DDC

A colleague shared a blog post with me: Genrify your Catalog, not your Collection! Written by Leanne Ellis, a School Library Coordinator for the New York City Department of Education’s Department of Library Services, it’s a great quick read on some of the pitfalls you can encounter if you ditch classification altogether—but it also has some helpful tips for getting the best of both worlds.

Many of the benefits touted by libraries who have genrified their collections are not incompatible with an organized classification system. I’m thinking of more eye-catching signage, more colloquial labels, or shelving attractive covers facing out. All of those are ways you can promote usage of your collection without sacrificing the organization you get with a classification system.

Reorganizing a collection is no trivial task, so there’s a lot to be said for drawing on some of the advantages of a genrified system without losing out on the organization and interoperability you get with the DDC and other classification systems.

I’d be interested to hear from anyone who has tried methods like this and what your results were. And of course, we can still improve things on the editorial side! If you have ideas for changes you’d like to see in the DDC, I hope you’ll get in touch.

New changes to DDC MARC fields

At last year’s virtual ALA Annual, I presented a discussion paper to the MARC Advisory Committee (MAC) around the issue of modernizing citation of DDC numbers in the MARC formats. I summarized the issue in a blog post then. In short, the problem was to better express the recency of a DDC number when you could only cite the edition number—and we’ve decided not to designate new editions in the future.

MAC response to the discussion paper was positive. Reinhold Heuvelmann of the German National Library (DNB) suggested a different approach that would expand the existing subfield $2 in DDC MARC fields rather than designating new ones. Then, at this year’s virtual ALA Midwinter (the last one!), I presented a formal proposal to implement these changes, using the same basic approach as the discussion paper, plus the DNB’s feedback. That proposal was approved, and its recommendations are now live with the June 2021 MARC update.

To review these changes, see the documentation for the MARC Bibliographic 082 field. I think of this as the “bread-and-butter” DDC field; you’ll see the same sort of changes in other DDC fields, such as Authority 083. As of this writing, the changes are in red. The main change is the redefined subfield $2:

$2 - Edition information

Information on the edition from which the DDC number was assigned. If DDC numbers are assigned from more than one edition, each DDC number or range of numbers is contained in a separate 082 field. If the DDC number is assigned from a printed version, record edition number (use “23” for print-on-demand), a slash (/), the language of the edition, a slash (/), and the year of publication or printing. If a DDC number is assigned from an electronic version, record edition number (use “23” or “22” for WebDewey, as appropriate), a slash (/), the language of the edition, a slash (/), and the date on which the DDC number was assigned by the cataloger. Languages are recorded according to the three-character codes in the MARC Code List for Languages. Dates are recorded according to Representation of Dates and Times (ISO 8601) in the pattern yyyymmdd.

082	00$a792.8/2$223/eng/20190402

What does this mean for your practice? That depends on if you’re using WebDewey or a printed version (either the traditional, published editions or a more recent print-on-demand version).

WebDewey: Keep using 23 as the edition number (unless you’re using DDC 22!). After a slash, add a language code. Note that this is the language code of the DDC you’re using, not the language of the resource or the language of the cataloging. If you’re reading this, it will probably be “eng” for English. Then add a further slash and whatever the date is as you assign the number. A macro or other shortcuts like the constant data in OCLC Connexion could help with this. Because the date is recorded in ISO 8601 format, it’s not wrong to give only the year, or only year and month. But it’s good to be precise.

Print: Keep using 23 as the edition number (unless you’re using an older edition). The print-on-demand versions, which were introduced in 2018, are all considered edition 23, like WebDewey. After a slash, add a language code. If the cover of your print DDC says “Dewey Decimal Classification”, it’s probably “eng” for English! Then add a further slash and the publication or printing year of your print DDC. That’s easy if you’re using print-on-demand, since they’re named for their year of printing. Otherwise, just check the title page. (Hint: the print DDC 23 is 2011.) Don’t record month or day in this case.

Like any change, this means some adjustment. A priority for this change was that it be backwards compatible: any MARC record out there with just an edition number is still valid, even though it provides less information than the current standard. I’m excited that with these changes, catalogers everywhere will have more detailed information about the “provenance” of assigned DDC numbers.

New Dewey MARC discussion paper available for comment

By now, you’re probably used to citing Edition 23 in MARC records when cataloging with Dewey. Edition 23 was introduced in 2011, and what you see today in WebDewey is still considered Edition 23. Since the DDC is now maintained via continuous updates rather than discrete editions, we have no plans to have an Edition 24. That means just citing a number as being from Edition 23 becomes less and less useful over time, since numbers can change within an edition (that was true in the past too, but is more acute now).

To address this, the editorial team has put forth a discussion paper to the MARC Advisory Committee (MAC) outlining the problem. We’ve suggested a way forward that largely preserves current practice if you’re using a print version (whether print on demand or older print editions) and uses dates in a new subfield if you’re using an electronic version (like WebDewey).

The full paper is available on the Library of Congress’s MARC Standards site. There are some discussion questions at the end—we’d love your feedback on these or any other aspect of the paper. Note that since this is a discussion paper rather than a formal proposal, MAC will not be deciding whether or not to implement the outlined changes right away, but rather how to proceed. This is a great chance for you to be involved in that discussion.

What would this change mean for your workflow? Are there other approaches we should consider? You’re welcome to leave comments here or send them to [email protected]. Please have feedback in by Monday, June 22 in time for MAC to consider.

How do the editors decide which concepts get a Dewey number? An explainer

Welcome to the first of a new series of explainers about Dewey Decimal Classification basics. I love chatting with folks about the work I do, especially answering questions about how the classification grows and changes. I'm looking forward to sharing what I know in this series, and I'll be glad for any suggestions you have for potential topics (you can email us at [email protected] or leave them in the comments below). In this post, I'll talk about literary warrant, the basis for judging which concepts get their own Dewey number.

Melvil Dewey originally published the DDC in 1876 as a method of organizing knowledge in print. He had a very practical focus on standardizing library classification and cataloging with the primary goal of efficiency. Though this terminology hadn't been coined yet, he was working with an early conceptual idea of literary warrant—that is, the principle of basing a classification on the amount of published literature. To help understand literary warrant, it helps to contrast it with other methods of determining warrant: for example, a classification could be based on a specific religion's understanding of how the world is organized, which would be a type of cultural warrant. Literary warrant uses the body of published literature as a way of representing the structure of knowledge.

In the modern history of the DDC (since the publication of the 16^th edition in 1958), the editors have committed even more strongly to the idea of literary warrant—instead of editing the classification in line with abstract ideas about the way things are, a new classification number is added only when a specific number of resources have been published, then collected and cataloged by libraries. That number has been arbitrarily set within our editorial rules at twenty works. When librarians ask about creating a new number for a specific topic, one of the first things we do is determine whether at least twenty works have been published primarily about that topic, demonstrating a need for a new number. (This differs from the Library of Congress Classification, where just one new book on a topic justifies the inclusion of a new provision in the schedules.)

In one way, relying on literary warrant removes some of the personal bias of the editors. Instead of making decisions based on our own decisions of what is valid (which would be heavily shaped by our cultural background and personal opinions), we have a clear test of whether a concept should be included in the DDC or not. On the other hand, literary warrant perpetuates a different cultural bias. We know that not all knowledge is reflected in published literature. The only way we have to measure literary warrant is based on library catalogs (like, for example, WorldCat), but we know that not all published works are collected by libraries or fully cataloged.

It's important to recognize that literary warrant is just one warrant that could be chosen to organize information, which has benefits and drawbacks. For an interesting review of how literary warrant works as well as information about other types of warrants, check out the Literary warrant article by Mario Barité on ISKO's Encyclopedia of Knowledge Organization.

Follow the Rules

"If all else fails, read the instructions."  Who wants to take the time to read the instructions when what needs to be done is "obvious"?   But sometimes what’s "obvious" turns out not to be correct.  For instance, last week I wanted to move a friend's car a few feet.  I found the car key, inserted it into the ignition, and turned.  Obvious way to start a car, right?  But there was no response.  It turns out that my friend's car is designed to start only when the brake pedal is depressed.  (Safety feature, I guess.)

What really was obvious about my futile car-starting attempt was that my efforts didn't work.  Sometimes failure to seek out and follow instructions will result in a noticeably different outcome from the one anticipated and desired:  The cake doesn't rise.  Nothing happens when you flip the switch.  Your computer freezes up.  But sometimes the incorrectness of an outcome that results from failing to follow directions isn't readily apparent. 

Since the assignment of an inappropriate Dewey number doesn't announce itself, starting with a knowledge of the basic rules that govern the DDC is important.  A knowledge of the basic rules is all the more important, because, as a general principle, the Dewey tables and schedules do not include instructions/reminders that are considered "standard operating procedure."

A common variant on "if all else fails, read the instructions" is "is all else fails, read the manual."  In the case of the DDC, the Manual is not where the basic rules of the DDC are recorded.  (The Manual contains notes that are too lengthy to be included in-place in the schedules and tables.)  It is in the Introduction that the basic rules of the Dewey Decimal Classification are found.  The Introduction is part of volume I of the print DDC 23 and is accessible from every WebDewey screen (indeed, you don’t even have to have a WebDewey subscription in order to access the Introduction from WebDewey or elsewhere).

Over the next little while, we are planning to use the blog to review basic principles that govern use of the DDC.  We will consider, for example, the rule of application and the rule of zero; we’ll review the difference between an interdisciplinary number and a comprehensive number; we’ll consider rules that govern the addition of standard subdivisions.  If there’s a specific rule or principle you would like to see addressed, please let us know by commenting on this post.

What is Knowledge Organization?

A special double issue of Knowledge Organization devoted to the question “What is Knowledge Organization?” has just been published as v. 35, nos. 2/3, of the journal—I co-edited it with Ia McIlwaine (former editor in chief of the Universal Decimal Classification [UDC]).  Six internationally known researchers contributed to the special issue, including our own Rebecca Green.  Topics addressed include the nature of the field of KO; epistemology, theory, and methodology in KO; relationships in KO; short- and long-term KO research questions; and KO systems.  The issue also features an interview with Ingetraut Dahlberg, founder of the International Society for Knowledge Organization (ISKO) and the journal Knowledge Organization.  The table of contents and preface are available here.

ISKO will hold its 10th International Conference, “Culture and Identity in Knowledge Organization,” 5-8 August 2008, at the Université de Montréal.  Immediately preceding the conference in the same venue on 5 August, ISKO NA (the North American chapter of ISKO) and OCLC are co-sponsoring a workshop on terminologies, “Everything Need Not Be Miscellaneous: Controlled Vocabularies and Classification in a Web World.”  The detailed agenda for both meetings is available here.

Postscript: Janus views must be the rage this week—it has come to our attention that the Journal of Information Science has just published a special issue on the history of information science in the UK (v. 34, no.4).

Literary Authors: AutoDewey and LC Name Authority File

At the recent Dewey Breakfast/Update at ALA Midwinter, we described the Library of Congress AutoDewey software for literature, which takes advantage of those places where Library of Congress Classification (LCC) and DDC are similar enough to facilitate mapping from LCC to DDC.  Before AutoDewey can assign a DDC number, the cataloger must often choose form (poetry, drama, fiction); when the literary periods of LCC and DDC do not match, the cataloger must also choose a literary period.

We spoke of the possibility of adding information about authors’ literary periods to Library of Congress name authority records and asked for advice about how best to do that.  At the breakfast there was general agreement that the more information given, the better.  Specific suggestions were that there should be links between the authority records and the relevant WebDewey records; that in addition to the DDC numbers we should give captions indicating what the numbers mean; that we should indicate which numbers are comprehensive or interdisciplinary numbers; and that we should begin with the latest authors. 

One librarian said she needed to think more about whether it would be best to give information about an author’s literary period without specifying form, or just to give a full built number showing form as well as literary period.  This is an issue especially for new authors, where we know in what form a particular work has been written, but we do not know whether that will be the only form in which the author will write.  If we give a poetry number for an author, we don’t want that number used for the author’s later drama or fiction!

The presentation from the Dewey Breakfast/Update has been posted on the OCLC Dewey Services discussion page.  We welcome additional advice.  Please reply directly to this blog entry.  If you prefer, you may also send comments and suggestions directly to [email protected].

Dewey Numbers in Authority Files + EPC Meeting

As promised in late August, the discussion paper on Dewey Numbers in Authority Files is now available.  This is one of the topics on the agenda for next week’s meeting of the Decimal Classification Editorial Policy Committee (EPC) at the Library of Congress, October 11-13.  Among the other topics on the agenda are updates to research methods, Italian provinces, historical periods for Italian regions, and crimes. The committee will also review topics related to groups of people in Table 1, psychology, and the social sciences.  Ia McIlwaine, chair of the Universal Decimal Classification Consortium and former editor-in-chief of UDC, will join EPC to discuss UDC-DDC cooperation in improving the presentation of religion (see here for an earlier discussion of the topic), and Diane Vizine-Goetz (OCLC Office of Research) will brief the committee on current DDC research initiatives.  We’ll post daily reports from the meeting next week.

You use "Regular" Google? Dude, that's so 2003

Back to PLA, then, for one last note of the goings-on at this excellent meeting. The Dewey-blog contingent decided that, since we'd spent much of the week learning about what the big names are up to, what we really needed was a sense of how the smaller, independent, unsung heroes of Libraryland are coping with everything the 21st century is throwing at them. So we thought we'd check out what Google, Inc., a lesser-known Mountain View, CA, company is doing in the name of organizing the world's information and making it universally accessible and useful. Which is quite a lot, as it happens. Actually, that bit about "organizing the world's information [etc.]"? Turns out that's Google's mission, and we know that for sure because Ben Bunnell told us 142 times in his talk on Thursday. Ben captured all our hearts with his bright-'n'-breezy overview of Google Book Search (whereby Google aims to index "every word in every book") and Google Scholar (ditto for scholarly journal articles); and then, on Friday, David Ferriero (director of research libraries at New York Public Library) brought us up-to-date with a chronological account of the challenges Google and its five partner libraries continue to face in providing access, through Book Search, to the copyrighted works held by those libraries. Both speakers gave shout-outs to OCLC's collection analysis of the "Google Five" (NYPL and the university libraries of Harvard, Michigan, Oxford, and Stanford), which found that 60% of all the works that the libraries collectively hold are represented in only one of the five. Many of the records that you'll retrieve when you use Book Search come complete with a "Find in a Library" link to OCLC's Open WorldCat, which lets you determine whether books are available in your local library; Google's intention is to extend this feature to more records in the near future. Now: Wouldn't it be cool if Book Search grouped subsets of search results into subject categories, so that users could browse classes of related material? Oh, but I guess that would require the books already to have been assigned class numbers or something. Hmm. Hang on a minute ...

I never metadata I didn't like

The Dewey roadshow reached San Antonio this week, remember, as teachers, learners, and practitioners gathered for their annual review of the state of education for library and information science. The development of the Metadata Education and Research Information Center (MERIC) -- a web-based repository or "clearinghouse" for metadata education resources -- was the subject of yesterday's session organized by ALISE's special interest group on Technical Services. Sherry Vellucci (Rutgers University) summarized the events leading to the implementation of the MERIC prototype that is accessible here (with the title Metadata and Cataloging Education—Web Clearinghouse): the identification, at the Library of Congress's Conference on Bibliographic Control for the New Millennium (in 2000), of the need for training and support programs that would help new LIS professionals participate in the development of metadata standards for the bibliographic control of digital resources; the constitution in 2003 of the Task Force for Preparing Cataloging and Metadata Educators and Trainers, jointly run by ALISE and the ALA's Association for Library Collections and Technical Services (ALCTS); and the publication in 2002-2003 of the report in which Ingrid Hsieh-Yee (Catholic University of America) reviewed the then-current state of metadata education and recommended (among other proposals) that a central clearinghouse should be established to facilitate access to standards and tools in the area. Ingrid herself talked about the different sets of core competencies required of all LIS students, of metadata catalogers, and of leaders of metadata projects; it's intended that the clearinghouse will provide support for learners' acquisition of core competencies at all levels of expertise. Anita Coleman (University of Arizona) then discussed the technical issues arising during the process of designing the MERIC prototype; Pat Lawton (University of Wisconsin–Madison) presented details of the collection policies and short- and long-term goals of the project; and Shawne Miksa (University of North Texas) spoke about plans for evaluation. A beta version of the clearinghouse will be available in summer 2006, with a full version scheduled for the fall.