Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
As I understand current person authority records, there is no information specific to a work or manifestation in there---except to note the _source_ of the other information found. But all information in a person authority record is about the person in general, not about a particular work or manifestation. I didn't think that the person's role in relation to the bib item being described is in an authority record. Am I wrong?All forms of the person's name are in an authority record, and of course most of them came from _some_ published manifestation, and the source of the information is noted, but the names are there because they are names of the _person_, they are about the person entity. But I'm no cataloger, maybe I'm missing something? And it certainly may be that as we move forward exactly where we put what bit of information may (and hopefully will) shift a bit, sure. We're hopefully not going to keep doing things exactly as we have done them. But it still makes sense to me to think of the current person authority record as the germ of the future person entity record; the current work (ie 'name-title') authority record the germ of a future work entity record, etc. Jonathan Karen Coyle wrote: Jonathan Rochkind wrote: This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. I haven't really thought this through far enough, but I'm not sure that the person entity and the authority record are one and the same. There is data about persons that, under current definitions, would not be in an authority record, such as the form of the author name from the title page, or the person's role in relation to the bib item being described. These are specific to that item, not to the person generally. And they may not be controlled. It seems to me that we will have a person entity in our bibliographic record that has this data, and that entity is NOT the authority record, which has data about the person, not the person's particular relationship to the bib item. What I haven't thought through far enough is where this item-specific and person-specific data will be in the relational model. It seems to me that the person entity in FRBR is not universal, but is a person entity for that bibliographic item. That would make it different from the authority record, and would mean that we would still have an authority record that controlled a universal view of the person. kc -- --- Karen Coyle / Digital Library Consultant [EMAIL PROTECTED] http://www.kcoyle.net ph.: 510-540-7596 skype: kcoylenet fx.: 510-848-3913 mo.: 510-435-8234 -- Jonathan Rochkind Digital Services Software Engineer The Sheridan Libraries Johns Hopkins University 410.516.8886 rochkind (at) jhu.edu
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
At 07:40 PM 6/2/2008, Karen Coyle wrote: Jonathan Rochkind wrote: This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. I haven't really thought this through far enough, but I'm not sure that the person entity and the authority record are one and the same. There is data about persons that, under current definitions, would not be in an authority record, such as the form of the author name from the title page, or the person's role in relation to the bib item being described. These are specific to that item, not to the person generally. And they may not be controlled. It seems to me that we will have a person entity in our bibliographic record that has this data, and that entity is NOT the authority record, which has data about the person, not the person's particular relationship to the bib item. What I haven't thought through far enough is where this item-specific and person-specific data will be in the relational model. It seems to me that the person entity in FRBR is not universal, but is a person entity for that bibliographic item. That would make it different from the authority record, and would mean that we would still have an authority record that controlled a universal view of the person. I have argued elsewhere that there is an important distinction between an entity record for a person, family, or corporate body -- which represents the person, etc. -- and an authority record for the NAME of the person, etc. FRAD's model of authority data makes it clear that the content of an authority record is one or more ACCESS POINTS for the NAME of the ENTITY; the access point is created by a particular AGENCY applying a particular set of RULES to the NAME by which the ENTITY is known. Thus there may be more than one set of access points applicable to the entity, based on different agencies applying different rules to the same set of factual information. In my model for bibliographic data, the description of the entity contains factual information about the entity including name usage on manifestations of works for which the entity is responsible, as well as other factual information such as affiliation, occupation, dates, etc. This record controls the IDENTITY of the entity but not any particular NAME for the entity. To the extent that we continue to desire to control the text strings that we use as access points for the name of the entity, authority records are still required, presumably linked to the entity record. This model is an extension both of the FRAD model and of the RDA scenario #1 model, both of which conflate the entity and authority records into a single object. To my mind, the very logic of both FRAD and RDA requires that these objects be treated separately in the model. Regarding Karen's second paragraph above, I believe that the person entity is intended to be universal, but the attributes (particularly name usage) do reflect information taken from particular manifestations. In another message, Karen asked whether FRAD dealt with records or databases. I believe that the model was intended to be neutral about record structure -- although there is an appendix (at least in the draft) that shows how different entities and attributes can be combined into an authority record. It seems to me that record structure is one of the constructs that one uses the model to help define. John Attig Authority Control Librarian Penn State University
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
Jonathan Rochkind wrote: As I understand current person authority records, there is no information specific to a work or manifestation in there---except to note the _source_ of the other information found. But all information in a person authority record is about the person in general, not about a particular work or manifestation. I didn't think that the person's role in relation to the bib item being described is in an authority record. Am I wrong?All forms of the person's name are in an authority record, and of course most of them came from _some_ published manifestation, and the source of the information is noted, but the names are there because they are names of the _person_, they are about the person entity. Yes, this is what I'm saying. But that means that the Work record needs to have some information about the person, not just a link to the authority record. RDA appears to be written for the level 1 record (although I hear that it all may become clearer when we get the remainder of the text). It appears to present both the Work-related and the Person-related information as a single unit. In the level 3 relational model, one has to divide those up. I don't think we know yet how we would do that. For example, I can imagine that the Work record could have a creator field that links to another record. That other record could be for a person, a family, or a corporate entity. I don't know if it makes sense for the field in the Work record to say which kind of entity the creator is (person, family, corp). Analogously, the Work record could have Subject fields that link to a wide variety of other entity records. The advantage of having a field for subject that doesn't specify what kind of entity it is is that it would be easy, in that model, to add new kinds of subjects (or, said another way, to link to new kinds of entities). But I'm not sure this works without a lot more thinking about it. For example, are there elements you would need in the Work record that wouldn't be valid for all of the creator entity types? In that case, you may need to keep the creator type in the Work record. Perhaps I'm getting too far ahead of things (and perhaps this is more of a discussion of FRBR than of RDA, but it's hard to separate them.) But some of this will influence the RDA Vocabularies project, so it's on my mind. kc -- --- Karen Coyle / Digital Library Consultant [EMAIL PROTECTED] http://www.kcoyle.net ph.: 510-540-7596 skype: kcoylenet fx.: 510-848-3913 mo.: 510-435-8234
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
John Attig wrote: I have argued elsewhere that there is an important distinction between an entity record for a person, family, or corporate body -- which represents the person, etc. -- and an authority record for the NAME of the person, etc. Actually my big concern is that the entity Person may make sense as a subject but we don't have persons as creators, only personal names. That name may be a pseudonym used by two actual human beings, or there could be many names associated with one person. So the Person entity doesn't seem to fit well into our cataloging world view in the creator/agent role. (I think this is just a variation on what John is saying.) FRAD's model of authority data makes it clear that the content of an authority record is one or more ACCESS POINTS for the NAME of the ENTITY; the access point is created by a particular AGENCY applying a particular set of RULES to the NAME by which the ENTITY is known. Thus there may be more than one set of access points applicable to the entity, based on different agencies applying different rules to the same set of factual information. In my model for bibliographic data, the description of the entity contains factual information about the entity including name usage on manifestations of works for which the entity is responsible, as well as other factual information such as affiliation, occupation, dates, etc. This record controls the IDENTITY of the entity but not any particular NAME for the entity. To the extent that we continue to desire to control the text strings that we use as access points for the name of the entity, authority records are still required, presumably linked to the entity record. I don't think that having different choices of name across one or more communities means you have to have different records for each name -- I can imagine an record that allows different options for creating name access points (appropriately identified), just as I can imagine one that has the same subject concepts in different languages. And I don't see why those name forms can't be in the same record that controls the identity of the individual, if that's what's convenient for your system design. I am not sure, though, that we want a record that ONLY addresses the name choice issue. Regarding Karen's second paragraph above, I believe that the person entity is intended to be universal, but the attributes (particularly name usage) do reflect information taken from particular manifestations. Attributes of the person entity? How would you connect that information to the particular manifestations they are taken from? Basically, how would you say that for this work, the author was affiliated with Harvard, whereas for another work, the author was affiliated with Stanford? Would that be in the person record or the work record? In another message, Karen asked whether FRAD dealt with records or databases. I believe that the model was intended to be neutral about record structure -- although there is an appendix (at least in the draft) that shows how different entities and attributes can be combined into an authority record. It seems to me that record structure is one of the constructs that one uses the model to help define. Actually, I think I asked that question about FRBR. And FRBR, too, may be neutral as to record model and database model. The 3 RDA schema levels seem to be more prescriptive about structure, but I'm still not clear if each entity is expected to be a separate record in level 3 or not. And it is that question that leads me to these other questions about how one connects the group 1 and group 23 entities in a work/expression/etc. record, given that some information will be specific to that group 1 instance. kc -- --- Karen Coyle / Digital Library Consultant [EMAIL PROTECTED] http://www.kcoyle.net ph.: 510-540-7596 skype: kcoylenet fx.: 510-848-3913 mo.: 510-435-8234
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
At 01:41 PM 6/3/2008, Karen Coyle wrote: John Attig wrote: I have argued elsewhere that there is an important distinction between an entity record for a person, family, or corporate body -- which represents the person, etc. -- and an authority record for the NAME of the person, etc. Actually my big concern is that the entity Person may make sense as a subject but we don't have persons as creators, only personal names. That name may be a pseudonym used by two actual human beings, or there could be many names associated with one person. So the Person entity doesn't seem to fit well into our cataloging world view in the creator/agent role. (I think this is just a variation on what John is saying.) I don't think that is what I was saying. For the purposes of the model, the pseudonym would be a person, as would each of the actual persons; we don't make any distinction here. Links between these different persons would be made to indicate the relationship. FRAD's model of authority data makes it clear that the content of an authority record is one or more ACCESS POINTS for the NAME of the ENTITY; the access point is created by a particular AGENCY applying a particular set of RULES to the NAME by which the ENTITY is known. Thus there may be more than one set of access points applicable to the entity, based on different agencies applying different rules to the same set of factual information. In my model for bibliographic data, the description of the entity contains factual information about the entity including name usage on manifestations of works for which the entity is responsible, as well as other factual information such as affiliation, occupation, dates, etc. This record controls the IDENTITY of the entity but not any particular NAME for the entity. To the extent that we continue to desire to control the text strings that we use as access points for the name of the entity, authority records are still required, presumably linked to the entity record. I don't think that having different choices of name across one or more communities means you have to have different records for each name -- I can imagine an record that allows different options for creating name access points (appropriately identified), just as I can imagine one that has the same subject concepts in different languages. And I don't see why those name forms can't be in the same record that controls the identity of the individual, if that's what's convenient for your system design. These options were canvassed in a MARBI discussion paper some years ago. I'm not sure there was a consensus. However, the separate authority records for each choice of name is the basis of the Virtual International Authority File -- and also reflects the reality that these authority records are likely to be created separately within each national community and brought together virtually -- using, I would argue, the person entity record as a clustering point. I am not sure, though, that we want a record that ONLY addresses the name choice issue. And I don't see how we can avoid it . . . unless we with to abandon the need to control the textual form of name. Regarding Karen's second paragraph above, I believe that the person entity is intended to be universal, but the attributes (particularly name usage) do reflect information taken from particular manifestations. Of the person entity? How would you link those to the particular manifestations? Basically, how would you say that for this work, the author was affiliated with Harvard, whereas for another work, the author was affiliated with Stanford? Would that be in the person record or the work record? Typically, we do not attempt to link these. If it is felt to be important that the attributes change for different manifestations, we typically include scope information along with the attribute in the authority record (in my model, this would be in the entity record). It seems to me that there is a practical limit to how much linking is sustainable. In another message, Karen asked whether FRAD dealt with records or databases. I believe that the model was intended to be neutral about record structure -- although there is an appendix (at least in the draft) that shows how different entities and attributes can be combined into an authority record. It seems to me that record structure is one of the constructs that one uses the model to help define. Actually, I think I asked that question about FRBR. And FRBR, too, may be neutral as to record model and database model. The 3 RDA schema levels seem to be more prescriptive about structure, but I'm still not clear if each entity is expected to be a separate record or not. And it is that question that leads me to these other questions about how one connects the group 1 and group 23 entities in a work/expression/etc. record, given that some information will be specific to that group 1 instance. (a) In both FRBR and FRAD, I believe that the
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
From your conversation with Karen Coyle: I am not sure, though, that we want a record that ONLY addresses the name choice issue. And I don't see how we can avoid it . . . unless we with to abandon the need to control the textual form of name. Very well put. I hope your words are heeded as the alternatives that seem to be under discussion do not seem workable to me. Mike Tribby Senior Cataloger Quality Books Inc. The Best of America's Independent Presses mailto:[EMAIL PROTECTED] -Original Message- From: Resource Description and Access / Resource Description and Access [mailto:[EMAIL PROTECTED] On Behalf Of John Attig Sent: Tuesday, June 03, 2008 1:41 PM To: RDA-L@INFOSERV.NLC-BNC.CA Subject: Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2 At 01:41 PM 6/3/2008, Karen Coyle wrote: John Attig wrote: I have argued elsewhere that there is an important distinction between an entity record for a person, family, or corporate body -- which represents the person, etc. -- and an authority record for the NAME of the person, etc. Actually my big concern is that the entity Person may make sense as a subject but we don't have persons as creators, only personal names. That name may be a pseudonym used by two actual human beings, or there could be many names associated with one person. So the Person entity doesn't seem to fit well into our cataloging world view in the creator/agent role. (I think this is just a variation on what John is saying.) I don't think that is what I was saying. For the purposes of the model, the pseudonym would be a person, as would each of the actual persons; we don't make any distinction here. Links between these different persons would be made to indicate the relationship. FRAD's model of authority data makes it clear that the content of an authority record is one or more ACCESS POINTS for the NAME of the ENTITY; the access point is created by a particular AGENCY applying a particular set of RULES to the NAME by which the ENTITY is known. Thus there may be more than one set of access points applicable to the entity, based on different agencies applying different rules to the same set of factual information. In my model for bibliographic data, the description of the entity contains factual information about the entity including name usage on manifestations of works for which the entity is responsible, as well as other factual information such as affiliation, occupation, dates, etc. This record controls the IDENTITY of the entity but not any particular NAME for the entity. To the extent that we continue to desire to control the text strings that we use as access points for the name of the entity, authority records are still required, presumably linked to the entity record. I don't think that having different choices of name across one or more communities means you have to have different records for each name -- I can imagine an record that allows different options for creating name access points (appropriately identified), just as I can imagine one that has the same subject concepts in different languages. And I don't see why those name forms can't be in the same record that controls the identity of the individual, if that's what's convenient for your system design. These options were canvassed in a MARBI discussion paper some years ago. I'm not sure there was a consensus. However, the separate authority records for each choice of name is the basis of the Virtual International Authority File -- and also reflects the reality that these authority records are likely to be created separately within each national community and brought together virtually -- using, I would argue, the person entity record as a clustering point. I am not sure, though, that we want a record that ONLY addresses the name choice issue. And I don't see how we can avoid it . . . unless we with to abandon the need to control the textual form of name. Regarding Karen's second paragraph above, I believe that the person entity is intended to be universal, but the attributes (particularly name usage) do reflect information taken from particular manifestations. Of the person entity? How would you link those to the particular manifestations? Basically, how would you say that for this work, the author was affiliated with Harvard, whereas for another work, the author was affiliated with Stanford? Would that be in the person record or the work record? Typically, we do not attempt to link these. If it is felt to be important that the attributes change for different manifestations, we typically include scope information along with the attribute in the authority record (in my model, this would be in the entity record). It seems to me that there is a practical limit to how much linking is sustainable. In another message, Karen asked whether FRAD dealt with records or databases. I believe that the model
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
-Original Message- From: Resource Description and Access / Resource Description and Access [mailto:[EMAIL PROTECTED] On Behalf Of Karen Coyle [John Attig] Regarding Karen's second paragraph above, I believe that the person entity is intended to be universal, but the attributes (particularly name usage) do reflect information taken from particular manifestations. [Karen Coyle] Attributes of the person entity? How would you connect that information to the particular manifestations they are taken from? Basically, how would you say that for this work, the author was affiliated with Harvard, whereas for another work, the author was affiliated with Stanford? Would that be in the person record or the work record? [Bob Maxwell] No, I would say that the person entity record would have as attributes that person X was affiliated with Harvard at a particular time, and affiliated with Stanford at a particular time. I wouldn't say that for this work author affiliated with such and such an institution would be an attribute of the person entity, though we might get evidence for the time spans of attributes listing affiliations from information we glean from certain works (or manifestations, etc.). So the person entity record might well have a place to record the evidence for the attributes (indeed, I certainly hope there will be such a place), but the evidence itself is not in my opinion an attribute of the person entity.
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
Robert Maxwell wrote: [R Maxwell] I do not agree that authority records in some form will still be needed in a FRBR-ized catalog. If each entity has one and only one entity record in the database, then that one entity record serves all the purposes of the current authority records. When cataloging an edition of Homer's Iliad in English I would need at least: one work entity record (Iliad) one person entity record for the author (Homer) I think perhaps when Laurence Creider says authority records in some form will still be needed, he was thinking of this person entity record and work entity record. That is perhaps an authority record in some form. Laurence, can you confirm? This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. Jonathan
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
At 04:30 PM 6/2/2008, Jonathan Rochkind wrote: Robert Maxwell wrote: [R Maxwell] I do not agree that authority records in some form will still be needed in a FRBR-ized catalog. If each entity has one and only one entity record in the database, then that one entity record serves all the purposes of the current authority records. When cataloging an edition of Homer's Iliad in English I would need at least: one work entity record (Iliad) one person entity record for the author (Homer) I think perhaps when Laurence Creider says authority records in some form will still be needed, he was thinking of this person entity record and work entity record. That is perhaps an authority record in some form. Laurence, can you confirm? This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. Jonathan I agree with Jonathan. In a sense, *most* of the records in a database may end up being some sort of authority record (e.g., a personal name authority record, a work authority record, an expression authority record, maybe even a manifestion authority record; an item held by a library would have its item record attached to manifestation authority record, which in turn would be linked--directly or indirectly--to all of the other kinds of authority records that pertain to the contents, provenance, publisher, etc. of the item). Each of these records would act as the authority for the entity it describes. Kevin M. Randall Principal Serials Cataloger Bibliographic Services Dept. Northwestern University Library 1970 Campus Drive Evanston, IL 60208-2300 email: [EMAIL PROTECTED] phone: (847) 491-2939 fax: (847) 491-4345
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
Jonathan Rochkind and Kevin Randall are correct. I was simply not thinking clearly. Laurence S. Creider On Mon, June 2, 2008 3:30 pm, Jonathan Rochkind wrote: Robert Maxwell wrote: [R Maxwell] I do not agree that authority records in some form will still be needed in a FRBR-ized catalog. If each entity has one and only one entity record in the database, then that one entity record serves all the purposes of the current authority records. When cataloging an edition of Homer's Iliad in English I would need at least: one work entity record (Iliad) one person entity record for the author (Homer) I think perhaps when Laurence Creider says authority records in some form will still be needed, he was thinking of this person entity record and work entity record. That is perhaps an authority record in some form. Laurence, can you confirm? This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. Jonathan -- Laurence S. Creider Head, General Cataloging Unit Special Collections Librarian New Mexico State University Las Cruces, NM 88003 Work: 575-646-7227 Fax: 575-646-7477 [EMAIL PROTECTED]
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
I think that I understand what Bob Maxwell is saying, and I agree with him that 4.1 and 4.2 could be construed to mean that only manifestation-level records should be made. A statement to the effect that other levels of the FRBR hierarchy should or could be represented in catalogs would be a good idea. The problem is that the ICP have to look backwards as well as forwards because some countries and probably many libraries will not be able to make use of the World Wide Web and developments in computer software that the Anglo-American cataloging community can. The ICP should also refrain from committing itself to particular technological solutions, just as the Paris Principles tried to do. After all, the relational database is probably not the last word in computer systems. Still, since the ICP are so clearly based on FRBR principles, some bow to the entity-relationship model's impact on the form of catalogs would be a good idea. I think that Jonathan Rochkind made a better reply to Bob's point about authority records than I could. I am learning a great deal from the comments of others. Thanks. -- Laurence S. Creider Head, General Cataloging Unit Special Collections Librarian New Mexico State University Las Cruces, NM 88003 Work: 575-646-7227 Fax: 575-646-7477 [EMAIL PROTECTED] On Mon, June 2, 2008 3:14 pm, Robert Maxwell wrote: Like Larry, I apologize for sending this to two lists, but my response to the original (sent to RDA-L), has a lot more to do with FRBR than RDA ... -Original Message- From: Resource Description and Access / Resource Description and Access [mailto:[EMAIL PROTECTED] On Behalf Of Laurence Creider Sent: Monday, June 02, 2008 2:28 PM To: RDA-L@INFOSERV.NLC-BNC.CA Subject: Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2 [M Yee] Secondly, 4.2 carves into stone the approach to the multiple versions problem that has created so much havoc in existing catalogs. 4.2 is completely contrary to the general objective of the convenience of the user. [L Creider] Insofar as I understand this point, I disagree with it. A separate record for each manifestation is a good idea for both user and librarian because the practice more clearly represents what is held by a library. The problem with so many current catalog displays is not the rules, but the display. More adequate displays or better use of qualifiers would allow users to pick the manifestation they need without having to guess from incomplete data. FRBR at least allows records at the works level would allow users to then search for the manifestation they need by going down a level in the hierarchy. For example, our catalog, like others, contains runs of serials that are in paper, microfilm, and electronic formats. These different expressions need adequate description, but the user would be better served by encountering first a work-level record that would provide the call numbers of the various parts of the run. [R Maxwell] I, too, am greatly concerned that 4.1 and 4.2 carve into stone the current method of doing things, i.e., single records that cover all of the FRBR entities. In an entity-relation database, we would not have a bibliographic description ... based on the item as representative of the manifestation ... [including] attributes inherited from the contained work(s) and expression(s). It is my contention that we would instead have separate work, expression, manifestation, and item records, and separate records for all the other entities as well, all linked by specified relationship links. This is NOT the scenario described in 4.1 and 4.2. The problem with 4.2 is NOT that it doesn't call for a separate record for each manifestation-it obviously does-but that, in combination with 4.1, it sounds a great deal like that is ALL it calls for, with no separate work or expression records or for that matter, no records for any of the other FRBR entities. [M Yee] Thirdly, the existing principles are also preferable to the proposed new principles because they are not nearly as tied down to existing catalog technology as the proposed new principles are. The new principles make explict reference to concepts such as authority records which may not even exist any more in a FRBR-ized catalog designed to exist on the semantic web in which each entity is represented by a URI. [L Creider] The Paris Principles with its reference to uniform heading, entry, conventional name, reflect the cataloging terminology of their time. Authority records in some form will still be needed even in a FRBR-ized catalog. Bare URIs will not be adequate because there will still need to be references from variant forms. I think that the ICP strike a good balance between reflecting current conditions (which internationally range from book and card catalogs to experimental catalogs making use of the latest
Re: [RDA-L] Comments from Martha M. Yee on the April 10, 2008 version of the STATEMENT OF INTERNATIONAL CATALOGUING PRINCIPLES 1 of 2
Jonathan Rochkind wrote: This is certainly how I've seen people talk about this sort of thing before, assuming that the person entity _is_ the evolution of the person authority record, and thus considered some form of authority record. I haven't really thought this through far enough, but I'm not sure that the person entity and the authority record are one and the same. There is data about persons that, under current definitions, would not be in an authority record, such as the form of the author name from the title page, or the person's role in relation to the bib item being described. These are specific to that item, not to the person generally. And they may not be controlled. It seems to me that we will have a person entity in our bibliographic record that has this data, and that entity is NOT the authority record, which has data about the person, not the person's particular relationship to the bib item. What I haven't thought through far enough is where this item-specific and person-specific data will be in the relational model. It seems to me that the person entity in FRBR is not universal, but is a person entity for that bibliographic item. That would make it different from the authority record, and would mean that we would still have an authority record that controlled a universal view of the person. kc -- --- Karen Coyle / Digital Library Consultant [EMAIL PROTECTED] http://www.kcoyle.net ph.: 510-540-7596 skype: kcoylenet fx.: 510-848-3913 mo.: 510-435-8234