Doorgaan naar de website
OCLC Support

2024 AskQC office hour member Q&A

Review the most recent member questions from AskQC office hours.

November 2024: Bridging the gap between records and linked data

12 November 2024

What can WorldCat Entities enable/facilitate that VIAF can't?

I would like to describe VIAF as a cluster of authority records. Things that authority files do VIAF does very well, disambiguating this Jeff Mixter from another Jeff Mixter, prescribing an authorized string for Jeff Mixter. With WorldCat Entities data, because of all the work we did into building out a very robust data model, which is much larger in scope than say the MARC Authorities Format, there are things you can say about, in this case, this Jeff Mixter that just don’t fit in the MARC Authority Format. RDA has gone a long way with a variety of 3XX fields to sort of describe authorities in more detail, if you will, but at its root authority records are still really meant for disambiguation and for prescribing an authorized string for a person, place, event, whatever it happens to be. Whereas WorldCat Entities is really meant to describe a person, place, event, etc. as sort of a real-world object in relationship and within the context of other things in your knowledge graph. That is probably the biggest difference between why one would use VIAF versus what one would use WorldCat Entities for.

How would creating an entity interact with other authority systems? Creating an entity in Record Manager wouldn't automatically feed out to any other system, correct?

We’re evaluating now how we could help lower that barrier or more streamline that process. For example, if one creates a new WorldCat entity in Record Manager or directly in Meridian and we know you’re also a NACO member can we just prompt you with creating a robust enough description of that person that we can put together a MARC authority record and ship it off into the NACO process the same way that we do when someone creates a NACO authority record in Record Manager today? We’ve also been evaluating, in talking with ISNI for similar purposes. So we are evaluating those right now and doing, sort of the background work to understand how closely does the WorldCat Ontology, i.e. the things you say about a person, place, event whatever, align with the MARC Authority Format in terms of what is required to create minimally viable, expected, etc. type of NACO record. Then, conversely, that’s the sort of streamlining entities to the traditional authority files. We’re also working on the opposite perspective so, when something is new in LCNAF, or is now in VIAF creating a robust description in WorldCat Entities for that new person, place, event, organization, etc.

Is the function available in WorldShare collection manager only? create WorldCat entity. Thanks!

It is available in Record Manager, which is where one will be doing bibliographic resource descriptions. It will be available in Connexion and then, obviously, in the Meridian application itself. But it’s because Collection Manager is more of a query application, I want data or I want to provide a bunch of data. It’s not really designed for cataloging workflows pe se that that integration is not there.

Could you share examples of discovery layers that are making use of linked data?

Yes, so there are both library systems available and non-library systems, you think of your Googles or whatever. What you see in discovery systems as it relates to integrations of linked data are what we traditionally call knowledge cards. When you do a Google search for Columbus, Ohio it wouldn’t just show you a bunch of web results related to Columbus, Ohio but on the right hand side of the screen it would also say Columbus, Ohio is a place and here it is on a map, and it has a population of whatever and things like that. That same practice can be used with linked data and discovery systems. So, I’ve done a search on Tony Morrison and obviously what the user’s expecting is books by or about Tony Morrison. But what we can also do with linked data is provide an info card, or knowledge card about Tony Morrison to provide more contextual information about her. Sort of taking the user and as opposed to just dumping a bunch of resources in front of them that they might be interested in is providing them context for what their search actually meant. In this case, it’s about Tony Morrison the person, then within that knowledge card, if you will, you can also have Tony Morrisons related to these people or this place or has written about these topics. And now what you’re starting to provide the end user within this case is the ability for them to hop off from a known search term to other searches that might bring back information also of interest to you because they are highly related to the known search that you just did in the system. Most systems do this nowadays, it’s sort of commonplace. You see it a lot on Amazon, for instance, in suggestions like this book is related to this book, or if you like this you might also be interested in this other highly related thing. This is how we can sort of unlock potentially new discovery features that are more challenging to do now based on MARC records as the basis for discovery data.

Can anyone create the WorldCat entities? In the case of LC authority records, NACO members (or authorized institutions) can create them. How do we trust the quality of the entities?

Yes, anyone can create a WorldCat entity. One must be a Meridian subscriber to do so, and we are only offering Meridian to library partners of OCLCs. So, you do not need to be a cataloging partner, but you do need to be a library, archive, or museum of some sort. With quality there are a couple facets there. First is quality from a duplicate perspective, how do we make sure people aren’t creating a bunch of Jeff Mixter person entities. Right now, in place we have duplicate detection, in terms of when an entity is created, we do an initial de-duplication check and then prompt the user if we think they are about to create a duplicate. We do the same for the API, though on the API response if you try to submit an API create for what we think is a duplicate you get back a message saying we are not creating it at all because we think it is a duplicate. If you still want to create this entity go to the web application and create it manually there, because there are a lot of Jeff Mixters in the world so one can understand why I want to create a new one because it’s not the same as the one that’s already there. We’re also working on a generic sort of continuously running duplicate checking like we do across WorldCat today. So that will take care of the duplicate perspective from a quality perspective.

Now, from a quantitative, or qualitative perspective, like how do we know that the description of Jeff Mixter is correct? We’re relying on the community of users the same way for the quality of WorldCat. A large portion of the quality of the content, quality of WorldCat, is based on the careful stewardship and curation by the library community. That’s the same philosophy that we have with WorldCat Entities.

I see a lot of WorldCat Entities, especially for works, that look like they were automatically generated based on the MARC record in OCLC. How are you assuring that automatically generated WorldCat Entities accurately disambiguate the entities they are describing? Does this process also add any descriptive data, alternative forms, or related URIs to automatically generated Entities, or do those need to be manually added?

We did when we launched WorldCat Entities create all of them based on existing data sources. For the works data, yes, that predominantly came out of MARC records. What we are doing now is doing careful evaluation of, in particular, how we clustered together and created the works data. It is important to note that when we say WorldCat Entities what we mean is a WorldCat work is analogous to a Bibframe work. And both of those are different from a WEMI or an RDA work. So a Bibframe work is more at the expression level in the WEMI hierarchy compared to a sort of traditional RDA or LRM work. So, from a mental model perspective, or a conceptual model perspective, I just wanted to make that clear. So, yes, getting back to the question as we continually refine the data, we are looking at not only how we can add more descriptive information but how can we improve descriptions as well as improve the data in general.

With regards to the other part of the question about related identifiers when we create a new WorldCat Entity algorithmically or programmatically, we try to pull together as many external identifiers as we can for that entity. So, there won’t be an expectation that there’s a new entity created today and it’s just a blank identifier with a label and things like that. There will be as much information as we can pull together from around the web across library data will be contributed to the description of that new entity.

I’d like to hear about the pricing structure behind OCLC Meridian. Where can we go to find out more about pricing?

At the bottom of the OCLC Meridian site there is a form to complete that will go to our Sales Team and one of our representatives will reach out to you about use cases, functionality, and pricing. My guess is that I, or my colleague Anne Washington, will be pulled in to do a deep dive demo, tutorial walkthrough, that kind of thing.

Is this tool for cataloging workflow or also discovery workflow? Do we need to be prepared ($$) to become Meridian members if we want to integrate WorldCat entities into a discovery layer?

The answer is both. Meridian is designed for both cataloging workflows. But the WorldCat Entities data, the data that is then created through the Meridian cataloging tool, our authority cataloging tool to be very specific is then freely available to use in your discovery layer. There are terms and conditions, but it is open data so there is no need to pay to use the WorldCat Entities data in your discovery layer.

So, does this mean that this extra information about an entity or person will be accessed through the linking will be in the $$1 and this linking field would be made viewable in a library catalogue?

Yes, the description of the WorldCat Entity would reside – when resolves or follows that link in subfield $1. There are a variety of reasons for that. One of the most important ones is what I was just talking through from the cataloging perspective which is Jeff Mixter, my name, is probably the same in any Latinized script, but in terms of discovery of me, let’s say resources by me, if I’m not a Latin script language speaker, if I speak Russian or Hebrew, etc. stuffing all that information about Jeff Mixter, even just different forms of one name, into a MARC record is particularly scalable. The idea would be all these WorldCat entity links have now been embedded, will continue to be embedded, into subfield $1s across MARC records and that the discovery system would use those links and then the data delivered from those links to make enhancements to the discovery system or to enable new features in the discovery system. Whether those links are viewable in the catalog or not, that is really a discovery system preference.

If the 245 field is updated (e.g., due to a typo), is the 758 text string automatically updated at some point?

If the 245 have a typo in it, for your example, and you do update it and it’s now out of sync with the title that is in the 758 you are welcome to report that to bibchange@oclc.org and we’ll add that to our workflow to get the 758 updated.

Will the BIBFRAME web application be available to Record Manager users? Or will it be part of a Meridian subscription?

That is yet to be determined. What we can say is that the Bibframe editor will include all of Meridian. So, from within the Bibframe editor, everything you can do in Meridian you’ll be able to do in the Bibframe editor.

21 November 2024

Can I just confirm, on Record Manager we can add entities during record creation, but we are not able to add entities through the WorldCat Entities website without a subscription - is that correct?

That is correct. Any cataloging user can, within Record Manager search and find entities in WorldCat Entities and automatically add those identifiers to a MARC record. But you need to be a Meridian subscriber to be able to create new WorldCat Entities or edit existing WorldCat Entities. But finding them and adding them to a WorldCat record can be done by any cataloging user.

How do you decide what data goes into entity descriptions? E.g. for persons that might include things like alternative names and date of birth, but it could also include demographic information etc. I'm curious how you decide what data is of value in your entities.

We seeded WorldCat Entities with an initial set of data, and they included a lot of basic information, for want of a better term, e.g., labels, alternative labels, date, birthdates, death dates things like that. Since then, we’ve been enriching the data sort of via at-scale additions, or even one by one, with additional claims. So, what data gets added is really up to the users of the data. It’s similar to the WorldCat model; what is of use to a cataloging user gets added to a MARC record. The only caveat with the WorldCat Entities data is there is a data model there. And when it comes to, maybe demographic information, things like that, gender for example. When we did our Diversity, Equity, and Inclusion review there were some concerns from a DEI perspective about certain things being said about certain classes of things, a person, a place, etc. So, what you can say is definitely constrained by the data model. For either DEI purposes, or just common sense purposes, e.g. places don’t have a dates of birth, but within the context of what you can, or are allowed to, if you will, say about an entity based on the data model, that is really up to the users of the data. Then OCLC through our at-scale enhancement of the WorldCat Entities data is pulling from WorldCat data, from authority data, from VIAF, from FAST to add new claims to the WorldCat Entities data.

Can you do linked data in your local catalog (after you export and download a record from OCLC)? For example, if you catalog in LC, and you have books about "Gettysburg" in multiple locations can you create linked data to show where these books are? In general, can you provide an example of how linked data would be done?

That would be outside the traditional metadata perspective. That would be, sort of, adding information at the holdings level. Which you can certainly do.

As a cataloging educator I am always striving to keep up with overall cataloging, metadata, and classification developments and how that changes what I teach. What basic knowledge and skills do library and information science MS students--who want to work in InfoOrg in libraries of all kinds--need to have in hand upon graduation?

With incoming library students, it’s good to get an idea about Bibframe. A really good tool is RDF.

It’s important that students today have a good understanding of models, of the Bibframe model, of the IFLA LRM (Library Reference Model). And that students still have an understanding of MARC records. You can always learn a particular product or programming language but first must have a really good understanding of the fundamentals to be able to use those productively in whatever system you’re going to work in.

Is there a process for institutions who want these identifiers in their local catalog to get new versions of the WorldCat records after you've added them?

Yes, there is. If you have Collection Manager, it is very easy to get those updates. There are a number of ways you can select what kinds of updates you want. See WorldCat updates for more information.

What does quality control of these automated processes look like at scale? For instance, if an external endpoint is providing bad mapping or private/harmful information, how does OCLC fix that on its end?

There are 3 distinct categories of mitigating harm or inaccuracies that we’ve implemented so far. The first 2 relate to WorldCat Entities. The one aspect of quality control is duplicates. The solution we have right now is an at-the-edge prevention. So, if a user in Meridian tries to create a new Jeff Mixter they’ll be prompted. There’s a quick search that’s done for duplicates and if potential duplicates are found the user is prompted saying We think we may have found this person, or place, or whatever it happens to be, in the system already then, it’ll provide a list of those potential duplicates. And you trust the due diligence of the cataloger to check those. But, because there probably are a lot of Jeff Mixters around the world, then the human can always override that and say No this is definitely a new, or different, Jeff Mixter. From the API perspective, if our API detects a potential duplicate it’ll reject the request and send back to the client a message saying This entity looks like a duplicate if you want to actually create this entity please have a human log into Meridian and manually create this entity. So that’s one aspect of quality.

Also, within WorldCat Entities, for sort of inappropriate, harmful, private, inaccurate information being added we have the same policy that we do with VIAF or with WorldCat, or other data sets which is a couple of things. You can have a data request take down. So, if someone adds a birthdate to someone, and that person whose birthdate added doesn’t want their birthdate there they can contact OCLC have a data request take down which we already have in our general workflows and data management processes here and we can do 2 things. If the request is just “get me out of the system” we have a way to delete an entity and then delete any references across WorldCat Entities back to that entity. So, we have basically scrubbed the whole existence of that entity from the Entities knowledge graph. More regularly the request is “I don’t mind that my description is there, I just don’t want my birthdate there.” In that case we have a way of removing the birthdate then locking that property at that entity level so that that statement, that birthdate, can never be added to that person’s description again.

Then with regard to the other aspect of quality is the links we put into WorldCat. How do we know that we’re adding the right identifier to, for example, Jeff Mixter – to that 100 field that has Jeff Mixter in there. And what we’ve done is, thus far, all the automatic and at-scale additions and identifiers that we’ve been doing to WorldCat with WorldCat Entities has been very conservative. So, we are only adding these identifiers to headings that are controlled and in which that control identifier or URI has a same as relationship to something in WorldCat Entities. It’s a very, very, very high degree of accuracy for adding those identifiers to people, places, events, organizations, because they are authority control based. And then for catalogers adding them in Record Manager we had the same sort of trust, belief that the cataloger knows best. It is a community curated data set, and they are the experts. But again, like with other problems if there’s an incorrect link catalogers can remove that link, or they can request that link be taken away or whatever.

So, the quality falls into not only where the identifiers are being put, that accuracy, but also the quality of the entities themselves, be that either duplicate problems or sort of private, harmful, or inaccurate information problems.

Since name authority records which have already been used to "seed" WC Entities are enhanced by NACO participants, will those additional linked data elements eventually make their way into WC Entities?

Yes.

What's the future of BIBFRAME editing tools from OCLC? Is that going to be Meridian, or something new?

It will be a new application that will include all the Meridian functionality. Meridian is really for describing authoritative things, the people, places, events, and organizations that are important in describing bibliographic resources. And the use case for that is: I already have my Bibframe editor or I’m still using MARC so all I want to do here is describe authoritative things, really manage that type of data. Our Bibframe editor will be primarily focused on describing bibliographic resources. But, again, going back to one of those challenges at the beginning, it will also include all Meridian functionality. So, while you are describing a bibliographic resource in Bibframe you can also describe, if you needed to, all the people, places, events, and organizations that are important to describing that resource.

It sounds like OCLC has done some in depth work on DEI issues in metadata management. Do you have any links you can share about this work?

There is a brief write-up on the DEI audit of the WorldCat Ontology here. But of course there is also a lot other DEI related work going on in our metadata work including the Reimagine Descriptive Workflows report and follow-on activities.

October 2024: Using MARC for new RDA elements

8 October 2024

Can a work have more than one representative expression?

Yes.

How does the information in the 387 field help the library user? How will the average user benefit from having eng in the 387 and the 041 versus just the 041?

That is going to be very dependent on how different systems are implementing these fields and index and display them.

A lot of these elements are available in the authority format and the bibliographic format. And, again, that’s a system issue on how that would be implemented.

So, are statements of responsibility no longer required in the 245 or 250 if we record it in the 881 field? (Per the Oxford Classical Dictionary example)

In Official RDA there are really no core elements. Depending on which entities you describe there are some minimum requirements, and those are very slim. So, the statement of responsibility relating to title proper is not required by Official RDA, however, it does not mean that it wouldn’t be required by your cataloging community or your agency. Also, just because something is not required doesn’t mean that it’s not a good idea to record it sometimes.

This example doesn't include the 245 info in the 881. You can pick and choose?

Is it possible to use punctuation to indicate line breaks in the source in the 881?

Yes, you can pick and choose in 881.

Yes, in Official RDA there’s an option to use some form of punctuation to indicate line breaks – use whatever your agency thinks is appropriate is one option. Another option is to use 2 slashes.

Can you provide an example of how the 881 field looks in a public catalog display?

It will look different for every library.

How important is it to use $0 in RDA fields like 340?

This will be really dependent on different communities and different systems. RDA has different implementation scenarios. Subfield $0 speaks of a more sophisticated implementation scenario. It allows for better dealing of controlled terms, more flexibility in display, it depends on the system.

Does 335 essentially duplicate or replace the LDR/008 positions that currently record this information?

There is some overlap.

Extension plan: What happens when the ‘plan’ changes? For example, a volume is published and catalogued as a monograph in 2010, no sign that it is v.1 of anything. In 2024, v.2 is published. Now what?

RDA doesn’t talk about this. It would be up to the library. Keep in mind that the Extension plan is not a required element in RDA.

Do public libraries cataloging in OCLC need to include these elements for records to be considered full level?

No. These are for your information to use if you want to but are not considered part of a full level record.

There is a lot of overlap and redundancies. Why use 881?

Field 881 is not something I would use for every resource every time. There are use cases for it, especially with rare materials.

Also, is the 881 subfield $f only for what is found on the title page, or would you also add things found on the verso? Like full addresses of publishers?

The manifestation statement can be taken from anywhere.

It looks like the new 34x are elements drawn from MARC 300. In that case, do we still use 300 as is, then add 34x or do we forgo 300 altogether? I hope we can still use 300 to summarize the characteristics of the manifestation as a whole.

That’s going to depend on your cataloging community decision. OCLC allows you to use both 340 and 300.

17 October 2024

Do we know why the MARC language code list isn't a valid choice in 387?

@Lisa Hatt, I think what Kate is trying to say is that the Language Code List we all know and love doesn't have a source code that you would add in the $2 subfield. That being said, I'm looking at the source code list for language sources at the moment, and according to that, iso639-2b is equivalent to the "regular" MARC language codes. So that would be the code you put in $2 if you want to use the MARC language code list (unless there is another restriction of which I am not aware).

Right. And you can take the codes from various places and specify the source in $2.

Do you have to put in capitals in the 881 because this older title page has all capital letters? Could you have put this same info. not in all caps? So, do we need to add this every time a title page has the title in all caps? Even though it may contain exactly the same info. that is in the 245? And there's no punctuation in the 881?

The manifestation statement element in RDA has a couple of different options for transcription. One of them is basic transcription. There’s also an option to use normalized transcription in manifestation statement. There are different valid options in RDA. None of these elements are required by RDA or OCLC.

Examples of Official RDA bib records from NLNZ include OCN 1459746030, 1458466211, and 1458836890.

For files of MARC records created since NLNZ implemented August 1: https://natlib.govt.nz/librarians/pu...ns-new-zealand For further information email us: nznb@dia.govt.nz. Some more OCLC records 1457435504 and 1457252538

How are these fields going to be used? Are they indexed in OCLC?

Keep in mind these are fairly new and we haven’t seen a lot of libraries use them, that could change in the future. In Bibliographic Formats & Standards we have information about indexing at the bottom of each page/MARC Element.

For the Oxford dictionary example, I don't understand why can't we put the editor's names in 245 $c and treat each edition as a separate work. Why do we need to use both 250 and 881 to record the editor's names?

We are conflating some different things here. The 245 is about information that appears on the manifestation and has nothing to do with the work. So, a statement of responsibility relating to the title proper is what appears on the manifestation relating to the title proper. That last part is key. If it is only related to the designation of the edition, it’s not related to the title proper. In MARC that gets recorded in Field 250. But it can be super confusing.

September 2024: Cataloging audiobooks

10 September 2024

Is there a good way to know when the MLA/OLAC best practices has been published outside of just checking the website?

It depends – the website is useful, but you might also be interested in joining the OLAC Listserv, which will most likely have an announcement about it.

re: slide 21 = why is the term "sound discs" used rather than "audio discs"?

RDA instruction 3.4.1.3 Recording Extent has a basic instruction to use a term from 3.3.1.3, which would be "audio disc" in this case, and an alternative to use another term, which would be "sound disc" in this case.

Why in the 240 10 is the narrator's full name not used?

It’s really more of a common practice not to include the full name unless there is a chance of a conflict. It’s important to remember that the surname is not being used to refer to the narrator but as another distinguishing characteristic of expression and there are other terms you can use for that, like the surname of someone involved with the expression such as a publisher. There are no RDA instructions for it.

So $h is not phased out in the 240 like it was in the 245?

Yes.

Shouldn't there be a date in the LC call number?

That was just a classification for the author, it’s not an example of a call number which would need a cutter and a date.

When do you add ǂh Spoken word to an 830 when there is a series statement?

If the series is representing the sound recording as opposed to the series representing the published version of the book.

Considering that an audio recording of an interview (nonmusical sound recordings and not an audiobook) is displayed as “audiobook” in WorldCat Discovery, could you explain why there are no requirements to have display any of the sound icons (sound recording (other), sound recording (cassette), sound recording (CD), sound recording (LP))? Why do these icons exist if we can’t assign them? Particularly important in case where the nonmusical sound recording is not an audiobook. See: Format display in search results - OCLC Support Thank you!

That is something we are looking into for the future.

Why ǂh Spoken word and not ǂh Sound recording?

We want to remember what the words in Field 240 are doing. So, this is giving us part of an authorized access point for an expression of a work, so we want to use expression elements to distinguish this expression from others. One of the expression elements we can use is the content type, like used in Field 336. You can use one of the content type terms in subfield $h to distinguish the expression. This would, for instance, distinguish this from the text version of the same work.

For the last example of the book accompanied by audio recording, why is the 337 'unmediated'?

It is referring to the book element.

You mentioned that 240s can be put into bib records without having an associated authority record. Is that mentioned in the best practices documentation?

Not every person who catalogs is a participant in NACO, but everyone who catalogs is able to create a 240 Field. So, if there is not one in the authority file and if you feel comfortable creating one then it’s good to include it on the record.

Did the publisher name on the disc have audio as part of the name? Should it match exactly?

Part of the job of cataloging a sound recording is to determine the publisher. The copyright holder is not necessarily the same as the publisher and we chose to use the name of the publisher found on the item and not on the disc which we determined to be the copyright holder. It’s partly cataloger’s judgment.

Would 240 be preferred or a 700 entry in my previous message or are both correct? 700 1_ |aHenry, Emily.|tFunny story.|hSpoken word|s(Whelan)

People tend to prefer the format of the 700 but as things stand today the 240 is the correct place for the authorized access element.

For the audio enabled books, would you use two 700$t for both the print expression and the audio expression, rather than one 240 for the audio expression?

For this case, for a 240 for an audio-enabled book we have to go back to the principle that the community basically decided that we’re not going to add a separate 700 for the print element but instead use an 006 Field and use the 33x fields. We are going for consistency in that the elements are where we expect them to be.

I'm seeing more records with 700 fields that repeat the info from the 100 and 245. Is this expected?

It is possible that these are due to limitations of some ILS systems and such things are fine on a local level, but we ask that users follow best practices in WorldCat.

19 September 2024

I have seen records for audiobooks with the "Spoken word" qualification in 8xx in either subfield h or s. Which is correct?

I believe you are addressing the qualification within a series then it would be subfield $h.

Regarding the 385 field, what is the reason to use the field twice with different vocabularies?

Sometimes vocabularies don’t match. What I was doing by including both was following the best practices put out by OLAC. Field 385 is optional, but they are indexed under keyword searching.

In the 8XX, should we remove links to the print edition of the series?

It depends on what that series is referencing. If there is a series for the sound recordings then you would use the 8xx for that sound recording series. But if the series is for publication, e.g. the 3rd novel in the “x” series there may not be a separate series for the audio recordings. It’s really just referring to the series within the books in which they’re published. In which case I would leave it. Unless of course it was very specific, then that would be a different matter.

Thank you for the 8xx answer. To follow up, the series title may be "Doe, Jane. X Series." but have the print publisher in the 643 Field of the authority record... I can keep this in the audiobook record if no audiobook series exists?

Yes.

For Playaways can 007$b 's' (standalone device) be used instead of z?

We are following what was, at least, in the Draft of the OLAC Best practices document.

Recently I've seen book records merged into sound recording records for Wonderbooks. Is it okay to create a book record as opposed to a sound recording record?

What you do locally is entirely up to you. However, in WorldCat we require a single record for a single resource and the best practice is that Vox and Wonderbooks (all audio-enabled books) be catalogued on the sound recordings format.

So, the leader for Vox? i for non-musical?

Type: i is correct

Should bibliographic records reflect a difference between CDs and CD-Rs? I often see popular audiobooks are manufactured on CD-R but this is not reflected in available OCLC records. Should I assume my CD-R copy is different from the available records?

The publishing information would most likely be different between the two so I would assume they are coming from different sources. But more information is needed - when contacting us via email attachments can be included.

August 2024: Name that Identifier! : Standard Numbers, National Bibliography Numbers, & Others

13 August 2024

Often times there is a barcode on the last page of the book what exactly is that barcode?

User response: It sounds like you might be referring to print on demand information. That numeric string includes information about when that specific copy was printed (often the last few numbers in the string represent a date).

Is there anything recommended for resources that have the same ISBN number but are completely different? Example: OCLC # 1411852771 and OCLC # 1450623348.

The situations can be different. It’s quite possible that the publisher has inappropriately published the same number on different resources. A quick internet search can be useful for finding which resource it is valid for.

Lately I've encountered multiple records with duplicate 016 fields from the same institution, where one field ends in period. Is it possible on the OCLC side to find these and remove them?

Yes, I think we can do a search for those and take a look into that.

I frequently see 020 ISBN information repeated in the 024 field. Can we delete the 024 field information?

Sometimes the same number can be in 2 different schemes. This is not always the case. It can be an ISBN, and that same string of digits can also be an EAN number. So, the first indicator in the 024 Field is important to look at because it is possible it is both a valid ISBN number and a valid EAN number. So, no, do not routinely delete them from the 024 Field.

22 August 2024

Is it true that 979 ISBNs don’t' have a 10-digit version?

It is true that some of the newer ISBNs that started being assigned in the past few years do not have a 10-digit version. Also, ISMNs have changed, they used to be 10-digits and began with an M but now are using 13-digits that start with 979.

How are 024 numbers indexed?

Slide 10 shows that the standard number index includes the 024 Field.

Are there cases where numbers that may be found in the 037 field could also be copied into the 024 field?

No, the 037 Field tends to be a stock number or the like and would not be appropriate in Field 024. There is a list of source codes for 024 for the ones recorded in Subfield $2 and that’s the first place to check. If the source of acquisition organization is not on that list then it shouldn’t be recorded in Field 024.

1 is represent US LCCN center. How about other numbers?

For field 022 which is for the ISSN, I mentioned there is a Subfield $2 that specifies the source and that is used for a national ISSN agency. So, there is no LCCN center but there is certainly a US ISSN Center that is at the Library of Congress. We do have a partial list on our BFAS (BibFormats) of Subfield $2 source code numbers. There’s a link for the full list of codes.

June 2024: Customizing MARC record delivery with Collection Manager

11 June 2024

Is there an ETA on Cambridge, Oxford, and Sage for automated collection loading?

Yes, Sage will be available later this month. Cambridge will be later this year, so later 2024. Oxford will be in 2025.

I would like to update a query collection with additional criteria, but discovered I could not and have to make a new collection instead. Are there any plans to make these editable?

We don't have that on the roadmap at the moment. The method that we typically suggest is what you're looking at in terms of making new collections.

What are some reasons why you would disable MARC record delivery for a collection? Most of our collections have it disabled.

It's possible you might not want records for e-resources, that's the biggest one with some libraries. They don't actually put their e-resources in their local catalog, they just have those available in WorldCat. That's probably specific use cases by library. Libraries have extremely varied workflows.

Is there any way to be notified when a provider updates its collection in Collection Manager?

If you're looking for something like an automated response where you get a notification when they're updated, that doesn't exist currently. We do post regular updates about new and updated collections in the Knowledge Base and Central Index, and on the Community Center so you can definitely check those out as well.

(Alma library) We’re getting automated MARC updates from OCLC WorldCat Partners Service, matching on vendor order number (035) from YBP. Are the settings for this controlled in Collection Manager? Some of the records contain bad diacritics.

You may want to open a Support@oclc.org request and they will get it reported to the right place. It would be hard to say without looking at the specific records. But yes, WorldCat Catalog and Partner Collections are in Collection Manager, so the configurations are there. Also, if you want to send those directly to Bibchange@oclc.org, we can fix those records for you.

How are the record files delivered?

In World Share Collection Manager, there is a tab called My Files. It's right under the Collection Manager section and files are delivered there. You can also get them through FTP if you prefer.

Are diacritics problems usually present in records batch loaded to WorldCat?

In WorldCat, we pretty much allow all the Unicode characters in the bibliographic record format. The thing is, it might not be the correct diacritic for whatever that text is, but it's a valid diacritic in Unicode.

I am a vendor who creates MARC records for customers so I'm just trying to understand their experience. Do most users customize their records?

That really varies from library to library. Many libraries do make some customizations, some don't. It varies depending on their comfort level with Collection Manager and making those customizations.

Is customization something that is set up initially for all of their records, or is this the way that it is done here?

There are default settings for customization that everyone has unless they make changes.

If a library was adding like a proxy to a URL is that something that would be done via this customization pane, or is that part of the default customization?

If you're adding a proxy to Knowledge Base links, there's actually a different section of the configuration for that. There's a section called proxy and authentication, and that's normally where libraries would add that proxy information, and then that would be added to their URLs from that section.

For the process that adds brief MARC records when there are none in WorldCat, would it be possible to run it on local collections?

It’s not currently creating records for local collections.

In export preferences is there a way to exclude French 655s but retain LCGFT 655s when both have a second indicator of 7

We have an enhancement open for this: https://community.oclc.org/t5/worlds...-managerideas/
exclude-the-content-of-a-subject-heading-field/idi-p/38056

20 June 2024

How would I be able to add the e-collections if they are not in the collection list?

You can create a local collection if you want to just add a specific set of titles that your library owns.

I see many open access KB collections in WorldShare, but do those records only include URLs for open access versions, excluding URLs for any other versions?

It can include a mix. Generally speaking, if it says open access they should be open access, but it is possible that we have a collection that has a mix of records where some may require authentication. An example of that might be a directory of open access journals, which could include a mix of open access and proprietary URLs. I think it would be the exception to the rule that there would be URLs on paywall platforms.

Sometimes in KB, it will say we have 98/100, does that mean our system is not set up for an autoload?

We might need to look at the specific use case, so the answer is it doesn't necessarily mean one way or the other because those 98 out of a 100 could have been manually selected or that it could have been via a feed too.

May 2024: Who's Your Audience?: Ways to Record Audience in Bibliographic Records

14 May 2024

How do things like content warnings (sometimes including in the front or back matter) fit in with the Library Bill of Rights (LBR)?

The LBR is very broad – so I was quoting from the ALA Interpretation of the LBR. They're both put out by the ALA organization. It's a little ambiguous but they do point out that it's important to attribute where that information came from without really going further. And it does say that that can be discouraging to users that that could be prejudicial. You have to keep in mind where ALA is coming from and what might be useful and appropriate. There's no ultimate right answer. This is really a question of user needs and cataloguing ethics.

Is there any rating for dyslexic materials?

LCDGT does have "Dyslexics" as a term. Broader Term is People with developmental disabilities. There is also "Dyslexics, Parents of" which says USE Parents of dyslexics. Not the same as a rating system, but it is there for possible use for audience.

Is it that the information in 387 is repeated from other fields? What the 387 for? What is the difference between 385 and 387?

This is an issue of MARC aligning with RDA. You could record the same information in 387 and 385 when you're talking about intended audience. 387 is a lot broader than just that 1 element and it's the subfields that let you know whether it's intended audience or something else. If you’re concerned about which field to use it’s going to be 385 for intended audience. And, of course, you also have Field 521 available if you prefer to do a note.

What do we mean by official RDA?

Access.rdatoolkit.org = has no instruction numbers

Original.rdatoolkit.org = has instruction numbers

23 May 2024

Isn't "Canadian home video rating" an attribution of the rating being used? (slide 15)

Yes.

If people want to see a system that provides faceting based on audience, you can try searching at https://orbiscascade-network.primo.exlibrisgroup.com/discovery/search?vid=01ALLIANCE_NETWORK:NRSG. For example, search for "novels" and then look in the right-side column for the audience facet, which is coming from both the fixed field for Audience as well as the 385 field.

Does OCLC do any advocacy with public catalog/discovery layer/ILS companies to help them determine how they display or interpret audience from MARC records? i.e. advocate against use of 008 and towards use of an alternate field?

No.

April 2024: PCC and OCLC: Overview and what's new

9 April 2024

Could you speak a bit more about the production expectations of funnels, please?

User response: As coordinator of Art NACO, I ask that participants do enough records to keep their training in their brain.

The idea of a Funnel is the expectation is that users contribute to the Funnel as a whole which takes the burden off the individual. According to the Governance document p. 4, production requirements for funnel is "none." Go to https://www.loc.gov/aba/pcc/about/PCC-Gov-Doc.pdf

Can non-North American users of OCLC products join PCC?

Yes.

As a new cataloger, how do I jump into NACO or request that someone be given a NACO record if we're not a NACO institution?

Email authfile@oclc.org if you are not a NACO member and need a name authority record created.

I had kept Connexion Client 2.63 even after the 3.0 version was up. Now that it is gone is there a way to be able to keep two sessions open? or is that completely gone?

Connexion Client 2.63 is still available.

Could you please give some examples of funnels, i.e., topics that are covered?

Arabic, Armenia, CJK, Comics and Fiction Hebraica, Military, etc. are many of these and are all listed on the PCC website.

I'd be interested to hear about any interactions between WorldCat Entities URIs and PCC-authenticated records. E.g., field 758 in WorldCat records. For now, I understand, PCC catalogers are treating 758s in WorldCat copy as pass through elements (ignoring them) when authenticating. Will there be a point where PCC will interact with these fields in any way?

What you are seeing in 758 fields, currently is mostly a project of OCLC, however, the 758 field writ large is intended to allow you to link to entity descriptions in any number of places, including wiki data, or other places that describes things like, works. Similarly, the subfield $0, and $1 in various access point, fields are valid in OCLC and could be added, and many of them are added, programmatically or have been, but many have also been added manually by OCLC members. More broadly speaking, the ability to add linked data URIs in MARC records has been expanding and so the PCC will be and is looking at ways of refining best practices and recommendations for folks to add certain things or maybe not add certain things because it's better to add them programmatically and things of that nature.

Field 758, Resource Identifier, has a lot of flexibility, it is not just limited to recording WorldCat URIs. The field has been around since 2018 in MARC, you’re just hearing a lot about it now because of OCLC’s project.

When you say language-based funnels, do you mean that they're creating subject headings in those languages (e.g., bidex for Spanish) or that they're making English-language SHs for works in other languages?

There is particular example of language based/subject area that has Funnels across all 3 programs, there is a Hebraica NACO Funnel, a Hebraica BIBCO Funnel and an Isael/Judaica SACO Funnel. The NACO Funnel focuses on the funnel would focus on names of people who use Hebrew as their main language that they publish in. The BIBCO (bibliographic) Funnel would look at things like transliteration of transcription fields and making sure that the records follow PCC standards. Then the SACO (subject) Funnel would then look at topics of interest or things that are related to the language community and the culture and the geographic area of Israel. So, here's a case where you have the same kind of expertise, but based on the program that the Funnel is in, that expertise is applied in a slightly different way.

I am interested in developing the same features in NACO file between Record Manager and Connexion Client.

Both Record Manager and Connexion allow users to contribute to NACO and have the same authority work capabilities.

Do you need to have specialized training for each funnel: Records reviewed and eventually become independent?

Yes, each new member will go through training then review until eventually becoming independent. Every Funnel operates a little bit differently.

User response: FYI: it's okay for a Funnel member institution to stay in a Funnel forever. You can keep contributing modest amounts over a long period of time. And there isn't pressure or a requirement to become independent.

Is it still true that when you join NACO through a funnel, the library becomes the member, not the cataloger who was trained?

Yes.

I have been running across French language books with series names that aren't necessarily authorized yet. The information for these series is included in publishers’ catalogs, publishers like Gallimard. My question is, what PCC funnel or group would I join to add the series so they can be authorized in Connexion, for example.

That’s a good question. We do not, as yet, have a Funnel that focuses solely on series. Funnels focus on name authorities then series authority work is another level of training.

18 April 2024

Are there plans at OCLC to reassign subject headings that have undergone splits? (This was done recently for Primitive societies.). Are the reassignments made by catalogers with access to the resources involved, or are automated processes used? Does OCLC post information anywhere about these & other special cataloging projects?

OCLC has done, in the past, when subject headings have gone under splits. The biggest one that comes to mind is when Illegal aliens was split and became 2 different headings, and changed to better terminology. We did that with a combination of automation and manual evaluation. Each project would depend on the circumstances. We haven’t posted information about that anywhere, but we do publicize when we do such a project.

Is there a formal training program for SACO like there is for NACO?

Yes, though it is different from NACO.

If you have questions about applying for SACO, contact coop@loc.gov.

I'm trying to find the app for the OCLC BSAF for my iPhone. Can you put it in the chat?

It’s not an app. but how the information is presented. In your phone’s browser use oc.lc/bfas to get to the page.

A non-CONSER library asking for an update to a CONSER bib record. What is the best way to go about this?

OCLC staff will make updates to CONSER records if it is appropriate. Email bibchange@oclc.org for requests to update bibliographic records, including CONSER records.

Are there any series specific live NACO training workshops these days, or do we just continue to use the written resources?

The NACO series course can be done without a trainer, but then you could ask a trainer to give it as an alternative. Either way, there would need to be a reviewer of series authorities created after the training.

Just to clarify, a NACO member can use their NACO authorization to edit/replace a BIBCO record?

Yes.

Is there a way to view lists of newly approved or changed LCSH? (I'm aware of the Tentative Monthly Lists which are not yet approved.)

Yes, approved lists are at https://classweb.org/approved-subjects/

How/if/when do changed headings filter through to OCLC records?

OCLC holds a copy of the LC/NACO Authority File and records that are contributed using OCLC Connexion and Record Manager are sent to LC overnight. LC loads the records then distributes them to all the NACO Nodes. That process typically takes 2-3 days.

Is the PCC Wiki linked from the PCC website?

Yes: from the home page: https://www.loc.gov/aba/pcc/

March 2024: Music Score Cataloging for the Uninitiated

12 March 2024

How would you define piano score?

A score for piano, which is the 1st example that we used, the music was originally written for piano so you would use the subject heading Piano music. A piano score refers to an arrangement. So, a piano score is typically a score that was originally written for some sort of ensemble, it could even include voices, for example with an opera. However, if you have a symphony that is arranged for a piano, i.e. you have an entire symphony but it's been arranged so that it could be played on the piano, those are called piano scores, and in the subject heading in the 650 you would name the original format, which could be symphonies, string quartets, etc. whatever the original was.

Could you please explain when FMus code "l" ("el") is used vs, for example, code "a"?

Before RDA catalogers were asked to make a distinction between a score that was for multiple instruments and a score that was for a single instrument. Consequently, you would code a score for multiple instruments as “a” and a score for a single instrument as “z.” When we came up with RDA it was decided rather than attempting to replace all of the codes that we had been using we added a code to redefine “a”. It was decided to just add code “l” (el) which would encompass both scores for a single instrument and scores for multiple instruments. So, you can use “l” in an AACR2 record if you are going to update it, it's still valid. But that's why people have been using “a”, because that was the past practice. Current practice for the last decade or so has been to use “l” instead.

Is Subfield $v scores allowed or required?

If you are referring to anything other than a score for a solo instrument it would be required if that is appropriate.

…or is it better to use a 655 rather than a subfield $v? or do you use both?

You would use both. Sometimes the 655 would not be as useful, it depends on what the main subject heading is.

Is it right to consider that that the genre/form term « Scores » (655) can be use more freely than the subdivision « $v scores » for subject in 650 (not only for musical compositions, but for all resources with notated music, for example Methods) ?

Method books can be cataloged as either a book or a score, it depends on the main component. If you catalog it as a score, then you should code it as a genre term (655).

If you have introductory material in your score in a particular language such as English, would you enter eng in the Lang fixed field or zxx for the predominant score part? In other words, do you record a spoken language in the Lang field if it includes any of that language?

If it is an instrumental work, you record zxx in the Fixed Field for language. Then in 041 you use subfield $g for any supplementary material. If you have vocal music then you would record the language of the song in the Fixed Field, then record it again in an 041. If you have supplementary text in a language different from the song, then use an 041 to record the language(s) of that supplementary material. If the text of the song and any supplementary material are in the same language then you do not need an 041, a note will suffice.

655 Notated music vs 655 Score?

Score would be appropriate. Notated music would go in Field 336.

Has LC stopped using conventional collective titles for music or just for literature (ex. poetry)?

LC has not stopped using conventional titles for music.

I've been using a 546 rather than a 348 for staff notation - should I change this?

That was changed recently, within the last 2 or 3 years, the preference now is to record that information in the 348 field rather than a 546.

Is 382 $s for number of players or number of instruments? Is there ever a time when one player would play two instruments?

Yes, there are times when players play two instruments. Those instruments have separate codings for subfield $d, that’s called doubling, but the subfield $s would still be the same, it would be for the number of players.

I understand that sheet music can be reissued frequently. If the date is the same on the reissued work, but other things have changed, like place of publication, at what point do you make a new record?

We ask that you look at Bibliographic Formats and Standards Ch 4 When to input a new record for guidance: https://www.oclc.org/bibformats/en/about/input.html

So, if I feel that a method should be catalogued as a score, is it ok to create a duplicate record in OCLC (if someone else has catalogued the same method as a printed monograph)? Thanks!

We ask that you never add duplicates to WorldCat. If you feel strongly about it, we suggest you add a record to your local catalog in the other format (but add your holdings to the WorldCat record).

Are we still permitted to use the conventional collective title "Selections" as the preferred title for compilations?

Yes, as appropriate, but depending on what it is you are looking at you might also want something more specific than “works” for instance “chamber music selections,” etc. depending on what the compilation is.

Vocal scores are sometimes published in different keys for different voices. Do you put these in the 384 field or in the 250 field, or in both?

Certainly, put it in 250 field, if you are creating a separate record for that edition.

21 March 2024

For the 048 instrumentation, if you have something like a contrabassoon, can you just code it for bassoon wd since there's no code for the basso instrument? Or do you use other woodwind wz?

Field 048 was created for a past era and its usefulness nowadays is waning, especially in comparison to field 382. One could simply say that now, the medium of performance is better handled in the preferred field 382 than in field 048. It is important to remember that there are other more detailed and useful sources that can be coded in field 048 (and/or in field 382, for that matter) than the MARC Instruments and Voice Code List. The IAML UNIMARC Codes for Medium of Performance (https://www.iaml.info/unimarc-field-...um-performance), in particular, has a specific code for contrabassoon ("wdb") and individual codes for many "ethnic" instruments (you'll excuse the expression). BFAS field 048 has a link directly to the IAML/UNIMARC list and both fields 048 and 382 are linked indirectly to that list via the "Musical Instrumentation and Voice Code Source Codes" list.

Confirming that no notation of "music" goes into a 300 b subfield for a score.

That is correct. If you were cataloging a book that included music, you would record that in a subfield $b, but if it is a score then you would not record music in subfield $b.

Did I see you record the language of notes, prefaces, bio info etc. as a 500 and not a 546? Would the language of libretto be a 500 or 546 in that case?

Yes, I did use a 500 Field because the particular notation of the score (this was for a trio). Since it was an instrumental work the work itself has no language and that’s why we coded it as zxx in the Fixed Field for language. So, information about the supplementary texts would go in the 500. If there was a libretto the language of the libretto would go in 546 because the libretto has text. And therefore, if it were a libretto in for example Italian with English translation, that's the kind of note I would put in a 546 $a.

Is there a particularly effective way to search OCLC for existing music score records? Monographic records for example, are most easily searchable by ISBN, title, etc.

Searching WorldCat Indexes: https://help-nl.oclc.org/Librarian_Tool...rldCat_Indexes

For band music, in the 048 field, do you include the number as well, i.e., od01 for 1 band?

I believe that ensembles do not use numbers, generally speaking.

Could you please repeat guidelines for recording distributor numbers? Which MARC field, indicators?

The distributor would be recorded in Field 028 1st indicator 6.

In your example of 505 notes you record the title and the parenthetical associated with the title in separate subfield t’s. Can you explain the reasoning behind this?

The reason I chose this setting is in part because the titles that are in parentheses are somewhat descriptive and not necessarily the title alone. This is optional, if it makes more sense to have them as a string that would be fine as well.

I often encounter score records where a very recent date is recorded as the date of publication. When I compare this record with the item I have in hand, everything matches but the only date in the item is the copyright date. On occasion that same copyright date is recorded in the OCLC record. Most often than not, these records are coded as "M". Should I be using these records as is, should I "correct" the fixed field date, or should I create a new record (assuming no other matching OCLC record).

If the record is Encoding Level M and it was loaded by a vendor, it is most likely that it was the date the vendor received the item. The preference is to work with the vendor record and if it is missing information like the copyright date to adjust it based on the item you are cataloging. It’s best to upgrade vendor records.

Are there any resources you recommend for non-Western musical scores (one of our professors is a composer of a lot of traditional Asian instruments). Lots of spelling variations etc.

There are non-Western musical instruments included in the medium of performance thesaurus so that would be a place you could look to see what is the standardized spelling of musical instruments.

For Incomplete compilations of musical works do you use 700 fields OR a 240 field but not both?

Best practice is to not do both.

For a work such as a concerto for two pianos, there could be two identical scores. Should the record describe just one score (300 1 score) or indicate identical scores (300 2 scores) because two are needed for performance? Or is this handled in a note?

2 identical scores, if that is how it was issued.

It seems like preferred titles are almost always present in score records, but sometimes not - even when we might expect to see them. Can you give any further guidance about supplying these, especially when there is no authority record and no current conflict. Is it best practice to supply one anyway?

It depends because there are so many formats of scores. Often you will see preferred titles being used so that, for example, if you have an opera like Don Giovanni, Don Giovanni is a distinctive title so if you have a full score for Don Giovanni and the title page says Don Giovanni then it is not necessary for you to have a preferred title in a 240 field. On the other hand, if you have a vocal score for Don Giovanni and the title page simply said Don Giovanni you will still need a preferred title of Don Giovanni in 240 with subfield $s vocal score.

February 2024: The Lifecycle of a CIP Record

13 February 2024

When is the CIP request made by the publisher? How long prior to publication?

We do ask publishers to submit their CIP application request as far in advance as possible. Prepub book link does limit the application to at least 8 weeks from the print date on the application.

Is there a way for a library to know if a book is in the CIP process? Sometimes we get a book order request for Pre pub books that do not have records in yet.

I don't think there is a way aside from having a bibliographic record in the library on that elaborate would know that a. Unless the publisher also says, otherwise it must go through the process of getting that bibliographic record created. I think that's the only way at this point that I can think of that would let libraries know that the book is going through the process.

Would self-publications be eligible for the CIP program? And if the self-publishers try to send in the application, does the system reject them automatically?

To request CIP data, one must be a CIP Publisher. As I showed in those slides at the beginning, it said that when they go into their Prepub Book Link account, they can choose requests of data. If you are an author, a self-publisher, or, you're a small publishing company you do not have the choice to request CIP data. You can only request an LCCN as part of the Prepub program. So, once a publisher has established themselves and believes that they now meet the criteria we need to be a CIP publisher, they have an option within their Prepub Book Link account to apply for CIP. And they're asked to show 3 books by 3 different authors that have been acquired by at least 1,000 libraries in the US. And if they meet that criteria, then they are approved for CIP. And then when they go into their account, they'll be able to see that request for data. If they do not reach it, their approval goes back, and the apply to CIP button reappears and then, in future should they get to the point where they do meet the criteria, they may apply again to become part of the CIP program.

Is the CIP request for English titles only? How about Spanish, French, or Chinese titles?

We can accept CIP requests in the Romance Languages – English, Spanish, French, Italian - unfortunately, we cannot accept full CIPs in Chinese at this point. Maybe in the future, but right now we cannot. And some of that is based on our current cataloging client, and some of the additional cataloging rules that will go into play. So, unfortunately, non-Latin scripts are not in scope.

Why are CIP records coded as authenticated in PCC? meaning why is field 042 coded pcc.

They are Full cataloging records and there are authority records for all access points.

Can we edit the 050 00 of a PCC CIP record to correct the date?

Yes.

If LC has a CIP request for a title in non-Latin script, do you transliterate or transcribe the title (or do both)? If you provide both the transliterated and transcribed title, would both be included in the LC CIP data block?

Again, we can't accept non-Latin scripts as a part of the CIP application. But if the item has a transliterated title, then we would accept those parallel titles for CIP.

Do you ever have to remind the publisher to send the book to you?

Yes, we do. Some publishers are better at sending the required copy upon publication, and sometimes we do have to give others a bibliographic of a nudge.

How do records make their way from LC's catalog to OCLC? GU is a member of the ECIP program and there have been occasional records that never seem to make their way into OCLC.

It tends to come down to the cataloging teams and their workloads, for instance titles that come through the literary group may take longer. While I would love to give a definitive answer it just really depends on the team.

How long on average should it take for it to get through LC's end of things (shelflisting and Dewy number) and out to OCLC? There have been a number of times when I have found a record in WorldCat that is coded as CIP (it still has a 263 field and ELvl =8) but it has been upgraded (completed 300 field, etc.) to a full level record. The book was published several years prior so there has been time for the Library of Congress to receive it. How and when are these records updated to ELvl = [blank] by LC?

It can take up to 6 months from when the publisher sends it, and the bibliographic records are updated. It can take longer depending on how long the publisher takes to send the required information. Sometimes CIP staff must remind the publisher that a work is due.

Again, this kind of depends on our workload. It does take a while for the book, when the publisher sends it, to go through the security processes at the library. It must go through scanning to make sure the package is safe, this is a result from many years ago when there was the anthrax situation, so it goes through additional security measures. Then it must go through the mail room and then be routed to the teams, then checked in from what we call the technical unit. That unit then makes sure that the book is checked in and needs to be routed to different teams. It can take up to 6 months from the time the publisher sends it to when we get it. Sometimes the publisher actually sends the book fairly quickly, so sometimes we can get those books out and bibliographic records updated pretty fast, otherwise, it can take up to 6 months, maybe longer, it just depends on when the publisher sends the book.

All the records should come through. I think there is a download process weekly or nightly that the bibliographic records come from, but as with anything with technology, there are glitches and sometimes records always make it over to OCLC. When we're notified, we do try to try to get those records sent but, unfortunately, glitches happen, and not all the records may make it despite our best efforts.

OCLC does get daily feeds of records from the Library of Congress and loads those every night. And when there are glitches, we contact each other and try to figure it out.

Can you talk more about how one becomes a CIP partner? What are the requirements?

We are generally contacted by the group that does the cataloging, e.g., a university press, then we have a program they have to go through where they have to get certain things clarified, and they have to sign an agreement with the Library Congress saying they will catalog for us. They used to only allow BIBCO libraries to contribute but they can be solely NACO libraries now because the Library Congress still does look at their records as well. Once we get all the paperwork and everything and we get them officially part of the program we show them how to use the PBBL system. They use the system as Connie demonstrated in her portion of the presentation. Then they send us the records which are uploaded, finished by LC staff then added to the catalog and once LC finishes the record it will be distributed to OCLC. Right now, we have approximately 34 different cataloging partners.

Is there a rule for the publisher to please choose only one national library to do a CIP record? For example, Crabtree Publishing asks both LC and Library and Archives Canada (LAC) to do CIP records.

From the LC side there is no requirement that the publisher only submits to one national library. I don't know if Crabtree also publishes widely in Canada, it may be that Crabtree Publishing is large in Canada. That may be a reason why they participate in the Library Congress and the Canadian CIP program, but we don't require publishers just to work with just one national library.

22 February 2024

Has the Library of Congress ever hired remote catalogers to create CIP records or only in-house employees?

The CIP Program does have a contractor to create some of the CIP records. They do about 10% of records. Then an LC staff member will finish out with a shelflist number and add the Dewey number then send the data block to the publisher.

With self-publishing on the rise, will CIP be available to authors who are publishing that way?

Unfortunately, self-published authors are still out of scope for the CIP Program. We do have our sister program, the Pre-assigned number (PCN), that will give a brief bibliographic record that’s created, and an LCCN back to the author. At the end when authors send in their books there is a chance that our selection librarians could select it to be a part of our permanent collections.

How does a library become a CIP cataloging partner?

The CIP Cataloging Partner Program is voluntary. We are usually approached by a library or a publishing arm of a library, like for a university. We go through a process. They must sign official documents saying that they are a partner of the library, we teach them how to use the Prepub link. We set them up with their own group within Prepub Link and then they can catalog for us until such a time that they don’t have the staff to do so and then they may ask to leave the program. And it’s open to all libraries that are NACO or SACO participants.

When information in the CIP block differs from the item, I catalog in hand is that because the information was in the galley but doesn't make it to the print book? For example, I'll see an edition statement in the CIP block but not on the item.

That is correct. We are given the gallies and the information on the form and we base everything on those, and any change requests we receive. There are often many changes that happen in between which is why we have the verification process. So, if you do see something in the CIP data block that doesn’t match the copy in hand that would be why. 

Do you keep the downloaded galleys indefinitely?

No, we do not keep them indefinitely, they are in the system and once the requests are archived the gallies go away and are no longer available to us.

So, if we use a CIP record for a book-in-hand, we can delete the 263 field in our catalogue's record?

Certainly, you can do whatever you want with your own catalog record.

But any changes made by LC during the verification process do not happen in OCLC, correct?

That’s not quite correct. When LC is making changes in their system it doesn’t automatically come to OCLC instantly. The records get redistributed to OCLC and we upload those into WorldCat on a daily basis. When a completed, or upgraded CIP is upgraded to Full level by the Library of Congress we do receive that record and upload it into WorldCat. Now if someone else has upgraded it to Full level, which could have happened, then the changes from the Library of Congress are not applied. But if the record is still at Encoding level: 8, when the full record is received from LC then the changes from LC are not applied to the WorldCat record. But if the WorldCat record is still Encoding Level 8 when the Full record is received from the Library of Congress the full record from LC then overlays the record in WorldCat.

Why does LC not always apply the translation cutter table to CIP titles with an 041 $h and a 240?

There are places according to the classification table where the translation table is not applied. So that may be the situation or it may simply be an error. We now have, for compilations we’re not using all of the 240 translation information. That may also be part of the situation.

So, if a PCC CIP cataloging partner were to do a CIP record that had a traced series and an 830 field would LC keep that 830 in the record?

Yes, LC would keep the 830 in the record.

What happens with surplus/duplicate CIP records? Are they deleted or suppressed?

By surplus I assume you mean the books that were never published. Every now and then we do look at old CIP records that are in WorldCat, that don’t have any library holdings and have been there for say 5 years or so but haven’t been upgraded yet, we may delete them. We look at them on a case by case basis. I’m sure the LC staff have that issue too because there may be some that they have in their system that were never updated because the book never got published.

By surplus, meaning as in when a paperback cover is held when a hard cover is then received, what does LC do with the extra hard copies?

The Library does have a surplus books program. It is available to schools and things of that nature where they can come and look and take the surplus books.

Is there a possibility for OCLC to not treat the eBook links with that "Available onsite" as a gov doc/open access link? It causes confusion for our patrons.

When the LC provides CIP to OCLC there is not an 856 link in the eBook records, any 856 links get added by a library who has that link. If there are local links feel free to delete them. But we do like to have the links for the providers that are in the 856 fields and the eBook records that are for multiple of that someone can purchase access to.

Does OCLC receive all the CIP records LC creates?

We are supposed to, at OCLC, receive all the records that LC creates. Once in a while there is a glitch, and something may not come through, but the policy is, the intent is, that we do receive all the records and load them into WorldCat.

My institution submitted a CIP record through Voyager that was never uploaded to WorldCat. What should we do?

If you know you uploaded the record, contact a CIP program staff member to make sure that it’s completed. We do have some workload disadvantages and sometimes even though the partner has finished it, it may take a while for staff to complete it. So, check with us to make sure that it is completed, and then if we see that it is completed, we'll either check with Cynthia or talk to another Library Congress employee in a separate system to see if we can then get that record redistributed to OCLC.

January 2024: Reading LC Authority Records

9 January 2024

In Connexion, what is best search to use to locate the authority record for "Prince" (the artist formerly known as ...)? Since I know his last name is Nelson, I can easily find the authority record. But what if I didn't know his last name? It seems a Browse search on "Prince" is better than a "Search" search. I hope I'm making sense.

Do a browse search on personal names. You get the exact record for Prince. Browsing is left-anchored and is simply a list of terms presented alphabetically. Search is keyword (so any element in the 1xx, 4xx, 5xx) will be presented in a list containing only items that include that term.

So, what makes an 053 different from a 050? Is it specifically for literary authors?

053s are LC class numbers in Authority records, Call Numbers recommended for use. 050s are LC call numbers in WorldCat records in actual use.

Can you explain when CYAC would be useful?

Those are subject headings that are specific for children and young adult resources. So, if in your library, a lot of your users are children or young adults that is who those headings were designed for. They can be more simplified in their terminology than LCSH, or, sometimes, when there's a need felt to have an English language version of a title established. So, it's just dependent on what your primary user community is.

As a rule, is it preferable to use one or the other of 372 and 374 ... not both? for a name authority record.

This is strictly a case of cataloger’s judgment and it's perfectly valid to have both 372 and 374.

18 January 2024

What does $w nnnc mean? Or where can I find a key to those codes?

Subfield $w is a Control subfield and those are explained in the MARC 21 Authorities guide for tracings, https://www.loc.gov/marc/authority/adtracing.html $w nnnc means to suppress the reference in displays and display the 663 note instead.

Do series authority records use a specific prefix? nb, no, nr, ns?

The series authority records are in the Names file so it could be any 1 of those prefixes. As well as the end prefix, you can tell it's a series authority record from the fixed field coding. It'll also have those 64X fields, like the 645 field. That's another indication it's a series authority record.