LCDGT Manual

I know you knew this was coming. How could I post about the LCGFT manual and not post about the Library of Congress Demographic Group Terms manual? I actually heard Janis Young (from the Library of Congress Policy and Standards Division dontchaknow) give her talk on on the LCDGT and the demonym conundrum three times at ALA Midwinter. It got better every time I heard it.

still from the movie Beetlejuice

Pictured: Me listening to Janis

Wait, the what manual?

Okay, so this hasn’t been around as long as LCGFT, and there’s a legitimate chance that you aren’t really sure what I’m even talking about here, so let me do a quick recap on what this vocab is even for. (For the official deets on the LCDGT’s purpose and creation, see the formal document)

Briefly, the LCDGT is a vocabulary consisting of people-characteristics. Where they’re from, what they do, their age, etc. There are 10 categories:

  • Age group
  • Educational level group
  • Ethnic/cultural group
  • Gender group
  • Language group
  • Medical, psychological, and disability group
  • Nationality/regional group
  • Occupational/field of activity group
  • Religious group
  • Sexual orientation group
  • Social group

Okay. That’s a lot of groups, and it is fraught. I know it’s fraught, and I know you know its fraught. Heck, I emailed LC on May 12th 2015 to express some concerns I had.

This is a primer though, so I’m not digging deep, check out the full biz for yourselves at Class Web, or if you don’t have access to that, here are some other options:

  • You can browse it at — though last I checked it’s a bit janky. When you click on a given group…it doesn’t actually display every term in that group. Not sure why, but I’ve seen it posted on a list serv and I know LC knows about it.
  • You can download datasets through this Class Web download — unfortunately though, those aren’t in RDF, just .mrc files
  • You could always use the one I built here *plug plug* — because I had to transform MARCXML rather than RDF, I couldn’t just use the XSLT I’d built for my QueerLCSH project, but I was able to re-use chunks of that code. (and I included some RDFa data (SKOS) in the html, so that’s cool)


So what’s the point of the LCDGT?  Why do we need yet another  vocab?

Sisters movie what omg amy poehler

Those are great questions Amy!


LCSH has a lot of headings which secretly have contributor/creator data or intended audience data embedded in them. That’s not great. So much of what the library-data community has been working on in the last…..aeons is to UN-mix our data. Get your title info out of my non title field! Stop putting your expression biz into my manifestation baz!

So it makes sense that we’d want to keep our subject headings subjects and keep our people info somewhere else.

We want the opposite of this.


  • Fantasy fiction, Korean
  • Children’s films*
  • Revolutionary poetry, Lithuanian

See how these LCSH are smashing together aspects of their subject and creator/audience characteristics? This is what we want to fix.

Another reason is that patrons often select their resources based on these characteristics and we want to make that easier for them. If a patron wants to find a resource on “Finance for Spanish speaking women who are doctors, written by French speaking accountants”, our fancy new ILS-of-the-future will help us help them find it. (I assume it’ll be very good at limiting by facets.

Onto the manual!


MARC Coding of LC Demographic Group Terms Describing Audience in Bibliographic and Authority Records L410

MARC 385 field is where we’re putting intended audience.  Use no indicators, and no ending punctuation (unless term ends with punctuation). Toss a $2 lcdgt on it, and call it a day. It’s repeatable, and LC’s practice is to place each term used in a separate 385, though you can double up (or more!) if you want.

Remember, you can use these in bibliographic records as well as authority records for works.

Bib examples:

245 00 $a Canadian Bates’ guide to health assessment for nurses.
385 ## $a Nurses $2 lcdgt

100 1# $a Blume, Judy.
245 10 $a It’s not the end of the world.
520 ## $a When her parents divorce, a sixth grader struggles to understand
that sometimes people are unable to live together.
385 ## $a Children of divorced parents $2 lcdgt
385 ## $a Preteens $2 lcdgt
385 ## $a Middle school students $2 lcdgt
385 ## $a Junior high school students $2 lcdgt

Authority examples:

100 1# $a Collins, Suzanne. $t Hunger Games (Novel)
385 ## $a Teenagers $2 lcdgt
130 #0 $a Irwin young adult fiction
385 ## $a Teenagers $2 lcdgt

MARC Coding of LC Demographic Group Terms Describing Creators and Contributors in Bibliographic and Authority Records L 412

MARC 386 is where we’re putting creator/contributor information. Just as above in the 385, use no indicators, and no ending punctuation (unless term ends with punctuation). Toss a $2 lcdgt on it, and call it a day. It’s repeatable, and LC’s practice is to place each term used in a separate 386, though you can double up (or more!) if you want.

I want to highlight here how much the words ‘self-identifies’ appears in this memo (and throughout the manual. I think that’s a great decision on LCs part to make sure it’s clear that we are to listen to the people whose work we are describing. They are the final (and best) arbiters of their own lives.

Interestingly, unlike the increased use of $3 in the LCGFT manual to indicate different genres for different pieces, there is no use of $3 to distinguish which LCDGT terms are to apply to which creator/contributor (in the common case of multiple). We just code all the LCDGT terms for all the creators/contributors. I’m curious to see how that shakes out in practice.

Just as with the 385, you can use these in bibliographic records as well as authority records for works.

Bib examples:

100 1# $a Russell, Rachel Renée.
245 10 $a Dork diaries : $b tales from a not-so-fabulous life / $c
Rachel Renée Russell.
386 ## $a Virginians $2 lcdgt
386 ## $a Lawyers $2 lcdgt
[“Rachel Renée Russell is an attorney. … Rachel lives in Chantilly, Virginia”–Author blurb.]

100 1# $a Sadler, Matthew.
245 $a Tips for young players / $c Matthew Sadler.
386 ## $a Chess players $2 lcdgt
386 ## $a Britons $2 lcdgt
[“Britain’s No. 3 ranked player Grandmaster Sadler answers key questions…”–Page 4 of cover.]

Authority examples:

100 1# $a Frank, Anne, $d 1929-1945. $t Achterhuis
386 ## $a Girls $2 lcdgt
386 ## $a Teenagers $2 lcdgt
386 ## $a Jews $2 lcdgt
386 ## $a Holocaust victims $2 lcdgt
400 1# $a Frank, Anne, $d 1929-1945. $t Anne Frank’s diary

130 #0 $a City anthology series of American poetry
386 ##$a Americans $2 lcdgt

Broader Term References L 468

This isn’t the hill I want to die on, but this memo happened to alert me to the fact that Marines is an NT of Soldiers. The marines hate that. My understanding is that they see the U.S. Army members = Soldiers, not just any member of the U.S. Armed Forces. Like I said, it’s not my hill nor battle, but there it is.

An interesting example given is:

Children of gay men

not BT Children

[Although Children of gay men appears to refer to people under thirteen years of age, the term refers to any child of gay men, including those who have reached adulthood. The BT Children is therefore not appropriate.]


It’s a fine line and perhaps one that should have been avoided, because essentially this means that the vocabulary is using the same word ‘Children’ in two different ways. One of those ways is an absolute technical definition, “someone who is the child of someone else.” The other is dependent on the subject’s age. So for example, while for now any books authored by li’l Harper Harris-Burtka could be given 386s of

386 _ _Children $2 lcdgt

386 _ _Children of gay men $2 lcdgt

As soon as he hits adulthood, his future works can only be given the second term.

There’s a general rule about assigning multiple BTs

when the term is intrinsically part of two or more groups. The BTs may be from different categories.

BT Information scientists
BT Library employees

I quibble here again that I know many librarians who are not library employees, thus seeming to violate the ‘intrinsic’ aspect of the definition. To some, a librarian is anyone with an ML[I]S, to others its someone who works in a library in a professional capacity.

Another interesting thing I thought was that no terms are made for language groups. That is, German speakers and English speakers share no BT with some term like, ‘Germanic language speakers’.

“Used For” References L 473

8. Pejorative or outdated terminology. Generally avoid making UFs for pejorative or long disused terminology.

Words or phrases that were formerly pejorative but are not any longer may be provided as UFs. For example, research indicates that use of the word Cheeseheads to refer to people from Wisconsin was pejorative but is now considered acceptable, so it may be a UF to Wisconsinites.

This was the rule I cited way back when to request that LC remove a term from their UFs for the LCDGT. I also just like really like the idea of people doing intense research to figure out if ‘Cheeseheads’ was offensive or not.

Assignment of Terms: Audience Characteristics L 480

This memo and the next are your two biggies, the ol’ “How to assign these dang terms” memos.

There’s a strong emphasis on explicit statements from the resources. Phrases that indicate what the intended audience is are strongly preferred over cat judge and inferences. It goes on to say:

In case of doubt, do not include terms for audiences that are only inferred.

So don’t just run around assigning these terms willy-nilly.

Assignment of Terms: Creator and Contributor Characteristics L 485

Again, much weight is placed on the self-identification of the creators/contributors. I especially like the phrase:

Avoid assigning terms based on a photograph or picture of the creator, or based on the creator’s name, because they can be misleading regarding age, ethnicity, gender, etc.

That’s a great caveat to include. Another warning is under the edition section

Demographic terms assigned to earlier or later editions of the same resource may be reused with caution in the new cataloging record. Creators may self-identify with different demographic groups over the course of their lifetimes.

I appreciate that LC is aware of these realities and provides guidance in the memos.


One thing that I want to highlight is the ‘Overlapping terms’ instruction in both this and the previous memo. There are many terms which may seem to overlap with each other and thus adding both would be redundant. An example given is:

  • African Americans
  • Americans

Here’s the thing though, terms from different categories are not overlapping terms, within the LCDGT. Remember the 11 groups above — African Americans is an ethnic/cultural group, whereas Americans is a national/regional group. Though outside LCDGT-land, all African Americans are obviously also American, within this sphere, it isn’t redundant because they’re part of different groups.


Each group has its own memo, most of which have special provisions in addition to the general rules given in the broader memos. (The Gender category, Medical, Psychological and Disability category, and Sexual Orientation categories have no additional provisions so I didn’t note anything about them)

Age Category L 510

If assigning an Age Group term, consider also assigning an Educational Level group term, because of how often patrons will be seeking/associating the two.

Do not routinely assign the term Adults, unless explicitly stated by the resource.

Do not assign any Age term if the resource is ‘general audience’

Educational Level Category L 515

Reciprocal instruction of the above instruction re: Age/Education.

If the resource is American use terms reflecting the American educational system, but if it isn’t then you’re welcome to propose (or use, if already accepted) new terms for the non-American educational system.

The memo includes special instructions on proposing such terms.

Ethnic/Cultural Category L 520

Similar to the crosses between Age/Education, Ethnic-Cultural has a post-coordination connection to National-Regional.

There’s an option (which I think y’all should opt to use) — if creating an new term make a UF in every official language of that ethnic/cultural group.

Finnish Canadians
UF Canadiens d’origine finlandaise
[Canada has two official languages, French and English.]

Language Category L 530

Do not just assign a term based on the language of the resource, there should be a better justification than that for assigning a language group term.

National/Regional Category L 540

There’s a lot of instructions here on creating and assigning demonyms. It’s all well and good if we stay at or above the first-order administrative subdivision divisions. Unfortunately SOMEONE suggested to the Policy and Standards Division that they create demonyms for below that level.

This created so many issues and problems that there’s an entire paper devoted to discussing it.

They would love your input, so read that paper (good luck…) and answer those questions at the end!

Occupation/Field of Activity Category L 545

There’s a note in the background section that members of religious orders (e.g. Bendectines) are in the Social category, not this one — why aren’t members of religious orders in the religion category?

Common sense note not to apply terms to creators when its redundant with the work being cataloged. I.e. don’t add “Authors” to a creator characteristic when cataloging a book.

No BTs may be constructed from an Occupation/Field of activity term to a Gender, Religion, or Sexual orientation category.  This is a good rule, as it allows that any given occupation is not limited to a single gender, religion or sexual orientation — acknowledgement of the great variance in humanity is a good thing.

Religion Category L 550

Terms are established for religions and sects, but not individual congregations

Social Category L 560

The social group is the catch-all for things that don’t really fit in the other groups.

No BT references are made to indicate age/gender/sexual orientation (good job, again)

In terms of “members of a specific organization” the only allowed terms are world wide scouting organizations and political parties. Otherwise they prefer you to construct a term based on the identity of that organization.

i.e. rather than adding a term, “American Humanist Association members”, add “Humanists”

That’s all! Go forth and start applying those terms. Write to LC, write to Janis, post stuff on your blog, yell at me on twitter. Do what yer gonna do.

We finished!


One last thing….

Here’s something cool — unwashed randos, just like you and me (i.e. non SACO members) can submit LCDGT proposals through this survey monkey for a limited time! Before you click through though! Let me save you some time — they’re going to want you to have AT THE READY several memos from the manual. But for some reason, they didn’t actually link them. I am a kinder, gentler person and will provide those links for your LCDGT-creation convenience.

Categories of Terms: L 465

Broader Term References: L 468

Used For References: L 473



*Yes! It must be noted that Children’s X is actually one of the stated exceptions in the LCGFT manual (See memo J 270), but the fact that they had to make an exception for it, demonstrates that they’re aware it breaks the rules. I provided it here to see if you were paying attention.