From the catalogs of babes

{January 22, 2010}   Starting from scratch: let’s pretend card catalogs never existed

There was a thread not all that long ago on the RADCAT listserv asking people how they got involved in what seems to now be called “radical cataloging,” i.e., basically, anything that questions or deviates from the proscribed traditional standards. Many people cited Sandy Berman as an influence, but I confess I hadn’t even heard of him until I was almost done with graduate school. (I may have even first learned about him on that very listserv.)

Apparently I’ve always been a radical cataloger, because I started deviating from the rules in the very first lecture of my very first cataloging class. It was my second semester in library school, but I had been working at the library where I am now for almost a year at that point, and I had already spent 5 years working for a large retail bookstore chain. The professor was introducing areas of bibliographic description with an exercise where he held up a book and asked students to suggest characteristics that might be beneficial to include in a bibliographic record. Everyone named the obvious components like title, author, etc., right away. The book was green, and I remember him asking the class if we thought that was important enough to be included. I (and several other people) answered yes, and were corrected by the instructor and told that it wasn’t.*

But all I could think about were all the years I spent helping people looking for “that book with the yellow cover” (both in the bookstore and in the arts-oriented library where I work) and how that cover color was information that people wanted to know and wanted to use to find their books, and if that information wasn’t included, we were doing a disservice to a certain percentage of searchers.

So why isn’t cover color included in bibliographic description? I can certainly see obvious reasons why it’s not: covers can vary depending on printing, covers may be multicolored and difficult to describe, books are rebound, the information in the resource and not the resource itself is what’s important, etc. I think these are all certainly valid reasons for excluding color from bibliographic description; the issues and troubles that come from documenting cover color certainly outweigh any benefits derived from including cover description, at least in most libraries.

But in some libraries, like arts-focused libraries, patrons are interested to know what covers look like. This is documented by research as well as my personal observations. So why isn’t color cover included in bibliographic description if it does, in fact, serve patrons?

Because it didn’t fit on a catalog card.
The current cataloging practices we have now evolved directly from the use of cards, specifically card catalogs. I’ve heard Diane Hillman talk about how the semantic web is going to further FRBR and move us away from our archaic self-imposed card-based standards.  I’ve watched Tim Spalding’s talk  about the limitations of standards based on physical cards. We use “main entry” and the “rule of three” because catalog cards did not have space to include every author/contributor. LC prescribes 3 subject headings because any more would tax the available space on a 3×5 card. Modern cataloging has been far too heavily influenced by what kinds of information we could cram into a two-dimensional space a little less than 15 square inches.

Once we were no longer limited to that tiny piece of cardstock, did we start including more information? Has cataloging changed significantly with the new technologies that have manifested between the typewriter and today? It certainly doesn’t seem like it. I know I’ve talked before about discarding these limitations now that we have technology that’s not held bound by these constraints: why not make the title field repeatable, so that multiple versions of a title can be included in a bib record? Why not list all the authors, instead of just the first three? But it leads me to wonder–what else we might include once we’re no longer held back by the tradition of the catalog cards? People claim that RDA will address these issues, but I see RDA as another piece atop the house of catalog cards, teetering precariously, still based on preceding rules and standards and subject to implementation challenges too.

What I would really like to do is sit down and start from scratch. Pretend like card catalogs never existed. If I walked into my library today, with its users and its collection, but without any previous cataloging, how would I organize it? Would I make a card catalog? An online database? An index? A paper list? Piles? Would the height of the book be important? The page count? Would it be enough for my patrons to simply indicate “ill.” or would I describe resources more specifically in terms of maps, sketches, charts, photographs, images, reproductions, etc.? I might include width, rather than (or in addition to) height, so as to be easily able to calculate the linear feet necessary in our increasingly cramped shelf space. I might list all the authors, not just the first three named or the “main” one. I might include categories for artists, illustrators, designers, models, and other contributors that aren’t authors but are certainly creators or co-creators of the work. I might do a lot of things differently if I was given the chance to start fresh and not required to work under the shackles of a system that not only does not serve my niche library, but cripples the evolution of other libraries as well.

Of course, we can’t start fresh—libraries already have large amounts of time, money, and inertia invested in the defunct status quo. Libraries balk at the effort to perform retrospective cataloging and reclassification projects—to throw everything out and develop new cataloging from scratch would be unthinkable. And truth be told, not only is it economically unviable and incredibly taxing to an already overworked personnel, there’s also oodles of valuable data already in catalogs that would be inefficient to simply throw away.

We can certainly harvest that data, but we need to add all the other stuff that’s missing—all the stuff that was left off in the past because it didn’t fit on that tiny little card, all the additional authors and contributors and series and width measurements and whatever else proves to be important to us and our patrons. LibraryThing already does this with some of its Common Knowledge data, which is clearly established as important information to the user group the site serves. As an arts librarian, I’d love to see development in the physical description areas, since our patrons seem to be so influenced by the physical characteristics of our resources. I wonder if this could also be crowdsourced/added socially: in the same way that LibraryThing members contribute series and character information, perhaps arts library users could describe their resources in ways that they find important to them? And if each library added the data that was important to them, imagine how fleshed out, detailed, and useful our bibliographic records could be!

Every library is different, and one tiny 3 x 5 card can’t hope to fit all the information needed by all of the different libraries out there. So now it’s my turn to hold up a book and ask which components might be important. Think about your library, its users, and its collection. Pretend catalog cards never existed. Tell me: How would you organize your library’s materials? What information would you record?

* I don’t begrudge the instructor for his answer–it was correct in context in that ‘color’ isn’t included in the traditional 8 areas of bibliographic description, which was, after all, what the lesson was about. He is actually a fantastic instructor who I would recommend to anyone, and I’m totally going to steal that exercise idea someday when I’m teaching cataloging.

Heather says:

I agree with you about ignoring the cataloging rules that only make sense because of the available space on 3×5 cards. How do you feel about the inclusion of cover images in catalog records? I like it, but I do wish that they included thumbnails of the spines as well.

By the way, the New England Law School’s library catalog, Portia, includes notes on the colors of the book covers. They even have a fun search page for it:

Bryan says:

We have been told that the rule of three will be optional in RDA, but I doubt that catalogers will abandon their practice so quickly without a compelling argument to do so.

What I find interesting is that whenever we discuss the rule of three we rarely if ever talk about evidence for justifying it or describing its effects in quantitative terms.

On average, what kind of numbers are we really dealing with here?

It’s interesting to note that not even the CC:DA’s 17 page “Task Force on the Rule of Three” report (, which basically denounced the rule, considered the effects of the rule in quantitative terms. For a profession that is so obsessed with metrics of all kinds, you would think that you would see a something. Not a mention in the report or in other forums. The rule limits the number of name access points because we assume that the numbers will be too great to deal with, but we really don’t know what the numbers are. There is a real disconnect here.

With regard to the rule of three, I think we should be seeking the answers to the following questions:

Of all the materials that we take in annually, what percentage of those materials would be treated under the rule of three? What percentage of the total collection are those materials treated under the rule? 5%? 8%? I think it’s no more than 5%, but that’s only based on informal counting. Let’s check our catalogs! What is the average number of name access points above 4 that appear in the various chief or prescribed sources across various formats? How many of those names typically excluded because of the rule have existing name authority records? For names lacking NARs, might we be willing soften our policies about supporting them with authority records if it meant that we could increase name access? One response might be that we’ll cause a great deal of scatter in the catalog, but how often will that actually happen? Let’s test it out. And if we’re not willing to soften our policies, what kind of automation will help us deal with the names. In Connexion, we have macros that help us create name authorities (i.e. GenerateAuthorityRecord) and convert names in notes to unadorned added entries (i.e. Joel Hahn’s 508-511to700). I can also imagine a variant for 245to700, a combination of 508-511to700 + GenerateAuthorityRecord, a reverse-engineered 700to245or508-511, and other variations on a macro theme. Whatever gets the job done. Alternatively, could we get by without the justifying notes? What’s more important?

I imagine that dimensions data, if we recorded all three of them, could help us better manage shelf space. The subject has come up here and there on AUTOCAT (e.g.

You might be interested to know that LC is contemplating a pilot project to source character names for works of fiction from LibraryThing (see “Appendix E: Pilot to Incorporate Externally-Generated Data from LibraryThing into Library of Congress Catalog Records”:

[…] This post was mentioned on Twitter by Becky Yoose and suzi w., Susan Roby Berdinka. Susan Roby Berdinka said: Starting from scratch: let’s pretend card catalogs never existed #cataloging #librarianship […]

suzi w. says:

As a former bookseller (in the pre-everyone had internet access days) the color was a big deal. “That little red book, I saw it here last year at Christmas, on that table over there” was a joke we told to each other when we were stretched beyond stretch. Who would have known that the Internet would change everything? I catalog a lot of movies, and rely heavily on IMDB for information about actors/actresses/producers/directors/etc.

Also, I am always tempted to add “includes glossary” in the 504, right up there with index and bibliographic notes.


Ivy says:

> I doubt that catalogers will abandon their practice so quickly without a compelling argument to do so.

If user needs and access isn’t a compelling arguement, then this is a profession that I don’t want to be a part of.

Your metrics are interesting, especially, as you note, for a profession so seemingly obsessed with them. However, while I used “the rule of three” as an example in my post, I think this metric is a little too narrow. When two authors co-write a text equally, one is the main entry and one is added–I’m arguing that that idea is also moot. They both wrote the text, we have the technology to indicate they both wrote the text, and if we were designing cataloging rules today based on literary warrant and available technology, we’d be able to easily do so. To make such strange rules as added entry co-authors and the rule of three not only alienates users but makes more work for catalogers, which increases the chances of error, which in turn alientates more users…

Macros in Conexxion and crosswalking fields in MARC is great an all, but it’s a band-aid on a bleeding artery, a finger in the dam. The rest of the world has already surpassed library technology and we’re still trying to jumpstart MARC? We need to step back and start from scratch, at least for brainstorming a new and better way to catalog to better serve library user’s needs. Like I said, we needn’t throw away the data we already have, but imo we do need to figure out how to translate or repurpose it into new structures, instead of trying to hold our old structures together with duct tape and string.

Ivy says:


I’m curious–why *don’t* you add glossary notes? Simply because it’s not included in AACR2 or some other reason?

Ivy says:

Heather, that link to Portia is awesome! I can’t decide which fascinates me more–the actual ability to search the catalog by color, or the fact that it’s a *law* library that instigated it! :)

I love the inclusion of covers in/along with records, and I also agree with you about the spines–that is often how people see and visually identify the books, after all. I don’t know if you saw it, but LibraryThing just announced a “Shelf Browse” vizualization, and the very first comment was about spine images!

[…] OCLC and others in the cataloging world have been on a kick concerning making our bibliographic data work harder. We’ll cover this more systematically in class, but for now Ivy over at FtCoB sets the stage: Starting from scratch: let’s pretend card catalogs never existed […]

[…] right! I’m barely one paragraph into the article Dr. MacCall recommended at this URL and I’m already loving […]

Kathleen says:

What I find interesting about this post is that it presupposes that many useful pieces of information are currently NOT included in cataloging records. I don’t find that to be the case at the institution where I am Technical Services Librarian. We have illustrators, added authors, genre/form headings in 655s, very distinctive icons for patrons, multiple subject headings, and lots of freedom to make our records as user friendly as possible, all the while using AACR2 which works just fine for us.

I think “LibraryWorld” as a whole, and catalogers in particular, are running scared from the new web that is out there. We will be making wholesale changes, based on impressions that patrons find our current practices unusable. I’m not so certain that that is the case.

And anyone who has read anything at all about FRBR and RDA, and I’ve read and read and read about both, and attended workshops, should be able to see that this computer data-base model is an ill-fitting and monstrously complex schema which does not mesh at all well with patrons’ needs.

I am all for new outlooks, new ideas, and new tools, but we also need to remember that “new” is not always good, better or best.

I’d 2nd what Bryan wrote, where are the metrics for the rule of three? How many records have 4 authors, or 5 or over 100? I’ve cataloged a few offprints of scientific papers with dozens of authors. What percentage of our catalogs do they represent?

We have started moving away from the restrictions of the card. I regularly use 246 to records additional titles. About 1/4 of the records in our catalog contain both a summary and table of contents. But moving beyond the card and redesigning as if it never existed are different.

Ivy says:

Hi Kathleen,

I’m not “presupposing” that useful information isn’t included in records–I’m basing it off my personal experience at my library (which is what this blog has always been based on). I’m glad you find your current records and standards to work for you at your library–which is not the “LibraryWorld” as a whole, either, and it’s certainly not my library. I’ll go out on a limb and guess that your library is probably a mid-to large-sized either public or academic library that serves a generalized audience with an overall variety of subjects and demographics, as opposed to say, a small, private library at a fashion and design school serving a narrow demographic of mainly 18 to 24-year-old women with little to no previous library education.I think that the larger and broader a library is, the more likely it is that current standards might still be adequate. But there’s plenty of data out there proving that these standards don’t work for specialty populations like the arts.

What I’d like to know is: how do you know your users are being served? What are the precision and recall rates from your catalog? Do you have a user survey that shows data supporting that the info in your bib records is working for your patrons? Or are you saying that it works for you internally and assuming it works for your patrons based on that (see Hoffman’s “Meeting Users’ Needs in Cataloging,” 2009)? I’m not trying to undercut your points–I’m genuinely curious, and I hope you’ll reply, because I think that kind of analysis is necessary so we don’t lose any parts of cataloging that are working. I’m not into “new” just for the sake of new, either–but I am into “better for the sake of better.”

Yes indeed, a lot of us are running scared. There is a lot of change in the air, from technology to funding to marketing that’s all going to have an effect on our positions and professions. it think it’s only logical to be scared–I’d worry more about anyone who wasn’t. And I think a lot of people in the library world are doubly scared of technology, database design, and the semantic web, because, hey, we were never taught that stuff in school! Now we’re out there, many of us on our own, trying to keep up with and teach ourselves complex and complicated subjects–of course we’re frustrated and scared. On the other hand, many of us are not running from the “new web”–we’re trying to make out catalogs more like it, and that’s not going to happen without some major changes.

Kathleen says:

Hi Ivy,

You have raised some excellent questions, and I hope that sometime soon I can provide a fuller answer than the one I have time for here. In the main, I am basing my impressions on my 3 years of work in our Circulation Dept., in which I was on the front lines of both helping users use our OPAC, and using it myself without the added experience of creating records, maintaining authorities, etc., which I now have as Technical Services Librarian.

As you have correctly surmised, I work in a medium-sized public library. We have, in my opinion, an exceptionally well-informed TS staff here. We are all interested in and keeping up with new developments in our field. Our supervisor is exceptional in both her knowledge and her encouragement of of staff development.

I do not have any experience in a library such as yours, and, as you say, without actual surveys, all our “evidence” is anecdotal and experiential.

It has simply become my impression, from the immense amount of reading I’ve done, and from interactions with colleagues at conferences, etc., that a great many librarians have accepted the idea that we are rapidly becoming irrelevant, and that we must therefore throw out both the baby and the bathwater lest the net-savvy pass us by.

I just think we should be a little more appreciative of what we have which already works, for our patrons most importantly, but also for us. The wholesale condemnation of current cataloging practices, and the rabid insistence that our patrons are the ones demanding something different, is not the whole picture.

Kathleen says:

Oh, and I should have said, that I did not take your initial post as “wholesale condemnation,” nor as “rabid insistence.” I think your post was thoughtful,thought-provoking and very even-handed.

Ivy says:

I think anecdotal and experiential evidence can definitely be valid (goodness knows I use it plenty myself!), but I confess I was curious if you had something more just because it seems like such a hard thing to quantify and I was curious and hopeful for an example of such a thing in practice! :)

If you’re from the library I suspect, then I think you guys are the ones who fleshed out those Videofashion records I was admiring a while back. I’d say you guys do indeed have an exceptionally well-informed TS staff, and I respect and admire all the effort you guys are putting into your work!

oneblankspace says:

In my freshman comp course nearly 20 years ago, one of the textbooks was available in either a blue-and-red cover or a white-and-blue cover, with different (but consecutive) ISBNs, but otherwise identical–both ISBNs appeared on tp verso.

In a similar course at a different college the next year, my Little, Brown Handbook was white.

Heather says:

@Ivy Thanks for the info on LibraryThing.

As for law students, the major series of casebooks have distinctive covers, sort of like the old orange Penguin covers. When I worked in a law library, students would regularly ask for the “black torts” or the “red con law.” The colors made it easier to pull from the reserves stacks.

Booker says:

> we needn’t throw away the data we already have, but imo we do need to
> figure out how to translate or repurpose it into new structures

So how much resistance to enhancement of records is driven by fear of the size of the job to get that enhanced data populated for X,000 existing records? Your post implies (or am I projecting?) that the huge job of retrofitting keeps people from adding tools for new records. I hope that’s not a large concern, because unevenly applied improvements still are improvements.

LG says:

I add either table of contents or summary notes to every single book I catalog – if the table of contents says the book has a glossary, that gets listed there. In my summary notes for the children’s books I catalog for my library, I say whether the books include glossaries, timelines, etc. (our users are not children, btw, but rather students who will eventually become teachers, so I don’t have to concern myself with what children might think when coming across my records in our catalog).

If you know the cataloging rules well, you should also have a good idea of where you can break them (making sure to break them in a consistent manner, and recording what you do for the catalogers that will come after you).

Kathy says:

I think Open Library is also a step in the right direction. Getting the records out of catalogs and more accessible can only help in changing both the way we catalogue our data and the systems we use to encode that data.

I’m one of the many unemployed recent MLIS grads, and in the last year have had plenty of time to research and think about what I want to contribute to this profession. Going forward, I plan to obtain a computer science degree and begin closing some of the gaps between our old systems and some of the new ones out there. The world desperately needs librarians who are also programmers so we can get out of the rut we’ve been in for the last 50 years.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

et cetera
%d bloggers like this: