Re: [RDA-L] Uniqueness of titles proper
Thomas said snip No, there is no equation of 'preferred title for the work' and the authorized access point for the work. The preferred title for the work is one element only. Mapping it in MARC would mean mapping it to 240 $a,$n,$p,$k -- but not to the rest of the 240 subfields. RDA 5.3 says to record additional elements to differentiate identical titles of works. To differentiate the title one could add tags in bibliographic or authority records corresponding to the additional elements in RDA 5.3: Form of Work - 380 $a Date of Work - 046 $k Place of Origin of the Work - 370 $g (authority record only) Other Distinguishing Characteristic of the Work - 381 $a snip If we can add all these distinguishing characteristics to make the preferred title unique, why can't we add the most salient distinguishing characteristic of all to make the preferred title unique, the author? You can say that the relationship between the creator and the work is a different kind of relationship from the relationship between the date of composition and the work or the form of the work and the work, or the place of origin of the work and the work, or other distinguishing characteristics (yipes, who knows what kinds of relationship to the work these may have; why do we have to dance around so much to avoid the creator here?). On Sat, Oct 5, 2013 at 6:14 PM, Brenndorfer, Thomas tbrenndor...@library.guelph.on.ca wrote: No, there is no equation of 'preferred title for the work' and the authorized access point for the work. The preferred title for the work is one element only. Mapping it in MARC would mean mapping it to 240 $a,$n,$p,$k -- but not to the rest of the 240 subfields. RDA 5.3 says to record additional elements to differentiate identical titles of works. To differentiate the title one could add tags in bibliographic or authority records corresponding to the additional elements in RDA 5.3: Form of Work - 380 $a Date of Work - 046 $k Place of Origin of the Work - 370 $g (authority record only) Other Distinguishing Characteristic of the Work - 381 $a In the MARC environment, the burden for differentiating lies mostly with the use of qualified authorized access points (and RDA anticipates this use as well-- it's just that RDA doesn't assume authorized access points are the only way ever to do this). These same qualifying elements are strung along the access point until the condition of uniqueness from the LC-PCC Policy Statement is met. In the current environment, authorized access points for works (130 or 1XX+240) aren't created for every record. But when they are created, the policy is to follow RDA 5.3 to differentiate works by creating unique authorized access points for works. The goal is not to have a unique title for every work. The goal is to supply all the elements necessary to differentiate the work from other works so that when users are looking at the bibliographic data they can know which work is involved. Because only the Preferred Title for the Work is initially a core element, other elements should be brought in. In RDA, any element becomes a core element if the resource or entity is not differentiated from another entity. With RDA we can meet this requirement by: 1. have a stack of discrete work elements starting with Preferred Title for the Work (in some future scenarios, this may be the only method) 2. qualify the authorized access point for the work with those same elements 3. both approaches (for example, copying and normalizing Date of Work in 046 $k is a useful idea -- even if it's not also needed in an authorized access point for a work) To compare these approaches, have a look at the MARC-RDA examples of authority records: http://www.rdatoolkit.org/sites/default/files/examples_of_rda_authority_records_041113.pdf The RDA records are much simpler, much cleaner, and far easier to understand than the MARC records. When trying to understand bibliographic data, I now routinely start with the RDA approach, and then work backwards to understand the complexities and shortcomings of the MARC/AACR2 approach. Thomas Brenndorfer Guelph Public Library From: Resource Description and Access / Resource Description and Access [ RDA-L@LISTSERV.LAC-BAC.GC.CA] On Behalf Of J. McRee Elrod [m...@slc.bc.ca] Sent: October-05-13 11:36 AM To: RDA-L@LISTSERV.LAC-BAC.GC.CA Subject: Re: [RDA-L] Uniqueness of titles proper Steven quoted RDA: If the preferred title for a work is the same as or similar to a title for a different work ... differentiate them ... To resort of pre FRBR/RDA language we all understand, I think this mist be understood as saying: If the preferred title [main entry] for a work is the same as or similar to a title for a different work ... differentiate them ... Conversely, one may as does the PS, understand preferred title too mean authorized access point. It is clearly
Re: [RDA-L] Uniqueness of titles proper
From: Resource Description and Access / Resource Description and Access [mailto:RDA-L@LISTSERV.LAC-BAC.GC.CA] On Behalf Of Adger Williams Sent: October-08-13 10:01 AM To: RDA-L@LISTSERV.LAC-BAC.GC.CA Subject: Re: [RDA-L] Uniqueness of titles proper Thomas said snip No, there is no equation of 'preferred title for the work' and the authorized access point for the work. The preferred title for the work is one element only. Mapping it in MARC would mean mapping it to 240 $a,$n,$p,$k -- but not to the rest of the 240 subfields. RDA 5.3 says to record additional elements to differentiate identical titles of works. To differentiate the title one could add tags in bibliographic or authority records corresponding to the additional elements in RDA 5.3: Form of Work - 380 $a Date of Work - 046 $k Place of Origin of the Work - 370 $g (authority record only) Other Distinguishing Characteristic of the Work - 381 $a snip If we can add all these distinguishing characteristics to make the preferred title unique, why can't we add the most salient distinguishing characteristic of all to make the preferred title unique, the author? You can say that the relationship between the creator and the work is a different kind of relationship from the relationship between the date of composition and the work or the form of the work and the work, or the place of origin of the work and the work, or other distinguishing characteristics (yipes, who knows what kinds of relationship to the work these may have; why do we have to dance around so much to avoid the creator here?). RDA 5.5 indicates that if the authorized access point for the work (i.e., in name-title form) is used to represent the work, then the differentiating elements are applied when the authorized access point is not unique. RDA 5.3, in making these extra attribute elements 'core' for differentiating reasons, covers the situation when authorized access points are not used (i.e. not in the current MARC scenario). I do think that specifying a relationship to a responsible agent does the job of differentiating as well-- for example, an easy way to differentiate translations is to specify the translator relationship to the expression. In fact, the original FRBR report indicates the range of user tasks for that kind of relationship covers exactly that premise. While RDA focuses on the 'Find' user task of the relationship between a Creator and a Work (RDA 18.2), the original FRBR report specifies that the relationship also serves the 'Identify' and 'Select' user tasks. But that handful of elements in RDA 5.3 has value beyond their role for disambiguating entities. Date of Work is useful generally for identifying the work when people are looking for the work that came out a particular year (such as with motion pictures). Form of Work is a nascent element at this point, as the whole form/genre infrastructure in cataloging needs more attention. There is a difference between attribute elements and relationship elements. As an attribute, Form of Work, is limited to the data present in the field. As a relationship element, Form of Work would become like a subject access point (like a 655). As an entity, Form of Work would have its own attributes and its own separate relationships (such as hierarchical relationships as seen in subject headings). But the reality today, in the current MARC environment, is that 5.5 governs, in that authorized access points for works (in name-title form) are used, and the only time differentiating elements are required is when there is a conflict in access points. That being said, I do see RDA 5.3 as opening the door for the changes necessary to move past this limited use of this data as derived from AACR2 for constructing headings. In RDA, 'core' elements are defined primarily in their role for disambiguating entities. But the elements should be added on their own as separate elements anyways, even if not needed for differentiation, mostly because they assist users in other ways such as 'Identify' (as in confirm resource described is resource sought) and 'Select.' That applies to both these attribute elements, and the relationship elements or subelements such as Creator or Translator. Thomas Brenndorfer Guelph Public Library
Re: [RDA-L] Uniqueness of titles proper
Steven Arakawa said: Both will have 245 00 $a John Rawls. One difference between RDA and AACR2/LCRI was that monograph title main entry conflicts in the same way that serial title main entry conflicts, so one of the collections will need a 130 John Rawls with a qualifier. How much simpler to insert a word in other title information ala Margaret Mann. 100 Ashbery, John. 240 Poems. Selections 245 The tennis court oath : a book of poems / John Ashbery. With the exception of music and Shakespeare, our clients dislike uniform titles (130 or 240) not on the item. They particularly don't like them if the title is distinctive. I agree with them, particularly for ILS which display the 240 rather than the 245. In most collections, there would be too few hits under authors such as Asbery, John, to make the uniform title helpful. A little pragmatism and awareness of patron preferences please. See also Kevin Randall's perceptive comments in this thread. His description of our practice as bizarre and inconsistent is spot on. I can only assume lack of communication between theorists and end users. __ __ J. McRee (Mac) Elrod (m...@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__
Re: [RDA-L] Uniqueness of titles proper
Mac said: steven Arakawa posted: I understand that work titles can conflict and we would need to break the conflict in such cases ... Only if neither has an author main entry (or author as part of AAP as Kevin would say). Of course two different works should not have the same preferred title if they are by the same person, but that is a rare problem. It certainly makes a lot of sense to say that there cannot be a conflict if the authors are different. But still I'd like to point out that this is not what RDA says. Look at 5.3: If the preferred title for a work is the same as or similar to a title for a different work, or to a name for a person, family, or corporate body, differentiate them by recording as many of the additional identifying elements in the following list as necessary. There is no exception for works with different creators. The LC-PCC PS for 6.27.1.9 shows a marked discrepancy to RDA 5.3, because it exchanges the title of the work for the AAP: If the authorized access point is the same as the authorized access point of another work represented by a bibliographic record or name/series authority record, add a parenthetical qualifier to the access point. Mind, I'm absolutely in favour of this practice, and the German-speaking community will do likewise. I think it would be a nightmare if we'd have to verify there is nothing else with the same work title in virtually every case, or add an identifying element if there was. But we should still bear in mind that this is not really RDA. Heidrun -- - Prof. Heidrun Wiesenmueller M.A. Stuttgart Media University Wolframstr. 32, 70191 Stuttgart, Germany www.hdm-stuttgart.de/bi
Re: [RDA-L] Uniqueness of titles proper
RDA 5.3 says those identifying elements should be recorded as separate elements, as parts of the access point, or both. So I don't think the PS is going against RDA by limiting the situations when the elements are added to the access point for the work. A lot of our difficulties seem to stem from the fact that we are creating MARC records that essentially describe manifestations, but then we try to add a lot of information about the works and expressions therein. -- John Hostage Senior Continuing Resources Cataloger Harvard Library--Information and Technical Services Langdell Hall 194 Harvard Law School Library Cambridge, MA 02138 host...@law.harvard.edu +(1)(617) 495-3974 (voice) +(1)(617) 496-4409 (fax) From: Resource Description and Access / Resource Description and Access [RDA-L@LISTSERV.LAC-BAC.GC.CA] on behalf of Heidrun Wiesenmüller [wiesenmuel...@hdm-stuttgart.de] Sent: Saturday, October 05, 2013 06:49 To: RDA-L@LISTSERV.LAC-BAC.GC.CA Subject: Re: [RDA-L] Uniqueness of titles proper Mac said: steven Arakawa posted: I understand that work titles can conflict and we would need to break the conflict in such cases ... Only if neither has an author main entry (or author as part of AAP as Kevin would say). Of course two different works should not have the same preferred title if they are by the same person, but that is a rare problem. It certainly makes a lot of sense to say that there cannot be a conflict if the authors are different. But still I'd like to point out that this is not what RDA says. Look at 5.3: If the preferred title for a work is the same as or similar to a title for a different work, or to a name for a person, family, or corporate body, differentiate them by recording as many of the additional identifying elements in the following list as necessary. There is no exception for works with different creators. The LC-PCC PS for 6.27.1.9 shows a marked discrepancy to RDA 5.3, because it exchanges the title of the work for the AAP: If the authorized access point is the same as the authorized access point of another work represented by a bibliographic record or name/series authority record, add a parenthetical qualifier to the access point. Mind, I'm absolutely in favour of this practice, and the German-speaking community will do likewise. I think it would be a nightmare if we'd have to verify there is nothing else with the same work title in virtually every case, or add an identifying element if there was. But we should still bear in mind that this is not really RDA. Heidrun -- - Prof. Heidrun Wiesenmueller M.A. Stuttgart Media University Wolframstr. 32, 70191 Stuttgart, Germany www.hdm-stuttgart.de/bi
Re: [RDA-L] Uniqueness of titles proper
John, Maybe so, but 5.3 definitely says that the element(s) used to differentiate two works with the same title must be recorded *somewhere*. My understanding of current LC practice is that two works with the same title only need to be differentiated by additional element(s) if they would otherwise have the same AAP. If this isn't the case, the additional elements do not have to be recorded *anywhere* (although a cataloger might still choose to do it, cf. LC-PCC PS for 0.6.3). My impression also was that LC is quite aware of the discrepancy I mentioned. Have a look at slide 36 in this LC presentation: http://www.loc.gov/aba/rda/source/refresher_module_B_oct2011.ppt Here, LC explicitly points to the difference between strict RDA and their current practice. The reason given is that this applies to current implementation in MARC still relying on authorized access points. Anyway, I think we agree that the current practice has a lot to do with today's MARC environment. Heidrun John Hostage wrote: RDA 5.3 says those identifying elements should be recorded as separate elements, as parts of the access point, or both. So I don't think the PS is going against RDA by limiting the situations when the elements are added to the access point for the work. A lot of our difficulties seem to stem from the fact that we are creating MARC records that essentially describe manifestations, but then we try to add a lot of information about the works and expressions therein. -- John Hostage Senior Continuing Resources Cataloger Harvard Library--Information and Technical Services Langdell Hall 194 Harvard Law School Library Cambridge, MA 02138 host...@law.harvard.edu +(1)(617) 495-3974 (voice) +(1)(617) 496-4409 (fax) From: Resource Description and Access / Resource Description and Access [RDA-L@LISTSERV.LAC-BAC.GC.CA] on behalf of Heidrun Wiesenmüller [wiesenmuel...@hdm-stuttgart.de] Sent: Saturday, October 05, 2013 06:49 To: RDA-L@LISTSERV.LAC-BAC.GC.CA Subject: Re: [RDA-L] Uniqueness of titles proper Mac said: steven Arakawa posted: I understand that work titles can conflict and we would need to break the conflict in such cases ... Only if neither has an author main entry (or author as part of AAP as Kevin would say). Of course two different works should not have the same preferred title if they are by the same person, but that is a rare problem. It certainly makes a lot of sense to say that there cannot be a conflict if the authors are different. But still I'd like to point out that this is not what RDA says. Look at 5.3: If the preferred title for a work is the same as or similar to a title for a different work, or to a name for a person, family, or corporate body, differentiate them by recording as many of the additional identifying elements in the following list as necessary. There is no exception for works with different creators. The LC-PCC PS for 6.27.1.9 shows a marked discrepancy to RDA 5.3, because it exchanges the title of the work for the AAP: If the authorized access point is the same as the authorized access point of another work represented by a bibliographic record or name/series authority record, add a parenthetical qualifier to the access point. Mind, I'm absolutely in favour of this practice, and the German-speaking community will do likewise. I think it would be a nightmare if we'd have to verify there is nothing else with the same work title in virtually every case, or add an identifying element if there was. But we should still bear in mind that this is not really RDA. Heidrun -- - Prof. Heidrun Wiesenmueller M.A. Stuttgart Media University Wolframstr. 32, 70191 Stuttgart, Germany www.hdm-stuttgart.de/bi -- - Prof. Heidrun Wiesenmueller M.A. Stuttgart Media University Wolframstr. 32, 70191 Stuttgart, Germany www.hdm-stuttgart.de/bi
Re: [RDA-L] Uniqueness of titles proper
Steven quoted RDA: If the preferred title for a work is the same as or similar to a title for a different work ... differentiate them ... To resort of pre FRBR/RDA language we all understand, I think this mist be understood as saying: If the preferred title [main entry] for a work is the same as or similar to a title for a different work ... differentiate them ... Conversely, one may as does the PS, understand preferred title too mean authorized access point. It is clearly impossible to have a unique title for every work. This demonstrates why we we the MRIs as opposed to the Toolkit. __ __ J. McRee (Mac) Elrod (m...@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__
Re: [RDA-L] Uniqueness of titles proper
No, there is no equation of 'preferred title for the work' and the authorized access point for the work. The preferred title for the work is one element only. Mapping it in MARC would mean mapping it to 240 $a,$n,$p,$k -- but not to the rest of the 240 subfields. RDA 5.3 says to record additional elements to differentiate identical titles of works. To differentiate the title one could add tags in bibliographic or authority records corresponding to the additional elements in RDA 5.3: Form of Work - 380 $a Date of Work - 046 $k Place of Origin of the Work - 370 $g (authority record only) Other Distinguishing Characteristic of the Work - 381 $a In the MARC environment, the burden for differentiating lies mostly with the use of qualified authorized access points (and RDA anticipates this use as well-- it's just that RDA doesn't assume authorized access points are the only way ever to do this). These same qualifying elements are strung along the access point until the condition of uniqueness from the LC-PCC Policy Statement is met. In the current environment, authorized access points for works (130 or 1XX+240) aren't created for every record. But when they are created, the policy is to follow RDA 5.3 to differentiate works by creating unique authorized access points for works. The goal is not to have a unique title for every work. The goal is to supply all the elements necessary to differentiate the work from other works so that when users are looking at the bibliographic data they can know which work is involved. Because only the Preferred Title for the Work is initially a core element, other elements should be brought in. In RDA, any element becomes a core element if the resource or entity is not differentiated from another entity. With RDA we can meet this requirement by: 1. have a stack of discrete work elements starting with Preferred Title for the Work (in some future scenarios, this may be the only method) 2. qualify the authorized access point for the work with those same elements 3. both approaches (for example, copying and normalizing Date of Work in 046 $k is a useful idea -- even if it's not also needed in an authorized access point for a work) To compare these approaches, have a look at the MARC-RDA examples of authority records: http://www.rdatoolkit.org/sites/default/files/examples_of_rda_authority_records_041113.pdf The RDA records are much simpler, much cleaner, and far easier to understand than the MARC records. When trying to understand bibliographic data, I now routinely start with the RDA approach, and then work backwards to understand the complexities and shortcomings of the MARC/AACR2 approach. Thomas Brenndorfer Guelph Public Library From: Resource Description and Access / Resource Description and Access [RDA-L@LISTSERV.LAC-BAC.GC.CA] On Behalf Of J. McRee Elrod [m...@slc.bc.ca] Sent: October-05-13 11:36 AM To: RDA-L@LISTSERV.LAC-BAC.GC.CA Subject: Re: [RDA-L] Uniqueness of titles proper Steven quoted RDA: If the preferred title for a work is the same as or similar to a title for a different work ... differentiate them ... To resort of pre FRBR/RDA language we all understand, I think this mist be understood as saying: If the preferred title [main entry] for a work is the same as or similar to a title for a different work ... differentiate them ... Conversely, one may as does the PS, understand preferred title too mean authorized access point. It is clearly impossible to have a unique title for every work. This demonstrates why we we the MRIs as opposed to the Toolkit. __ __ J. McRee (Mac) Elrod (m...@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__
Re: [RDA-L] Uniqueness of titles proper
Thomas said: RDA 5.3 says to record additional elements to differentiate identical ?titles of works. In real life, I suspect most will add data to make titles unique only for title main entries, and for author/title subject and added entries for works in the *very* rare case of two works by the same author having the same title proper. Otherwise, I suspect most will ignore identical titles proper; they are very common. The access point with an author can be unique, even if the title alone is not. The PS seems to me to support that supposition, referring as it does to access point rather than preferred title. But when they are created, the policy is to follow RDA 5.3 to differentiate works by creating unique authorized access points for works. I assume the reason you want an 046 date is that the 264$c applies to the manifestation rather than the work? The MARC21 definitions of 046 subfields seem also to apply to manifestations. Since all we now have are manifestation records, our clients are quite happy with differences in the description making that differentiation. While some use is made of 008 dates one and two, I am aware of none of our clients making use of 046 in their ILS. How would the data you suggest adding to fields other than 245 become part of an access point? 046 for example? The RDA records are much simpler, much cleaner, and far easier to understand than the MARC records. For you perhaps. __ __ J. McRee (Mac) Elrod (m...@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__
Re: [RDA-L] Uniqueness of titles proper
From: J. McRee Elrod [m...@slc.bc.ca] Sent: October-05-13 6:40 PM To: Brenndorfer, Thomas Cc: RDA-L@listserv.lac-bac.gc.ca Subject: Re: [RDA-L] Uniqueness of titles proper RDA 5.3 says to record additional elements to differentiate identical ?titles of works. In real life, I suspect most will add data to make titles unique only for title main entries, and for author/title subject and added entries for works in the *very* rare case of two works by the same author having the same title proper. Otherwise, I suspect most will ignore identical titles proper; they are very common. The access point with an author can be unique, even if the title alone is not. The PS seems to me to support that supposition, referring as it does to access point rather than preferred title. RDA 5.3 can also be read as not requiring those differentiating elements be added to the authorized access point when the overall access point (1XX+240) is unique. Rather RDA 5.3 provides the option that the data could be recorded elsewhere, as is now possible with the new RDA-influenced MARC fields, if even in anticipation of possible conflicts. Regardless of potential conflicts, there is great value in that data for users, especially Date of Work. But when they are created, the policy is to follow RDA 5.3 to differentiate works by creating unique authorized access points for works. I assume the reason you want an 046 date is that the 264$c applies to the manifestation rather than the work? The MARC21 definitions of 046 subfields seem also to apply to manifestations. Since all we now have are manifestation records, our clients are quite happy with differences in the description making that differentiation. While some use is made of 008 dates one and two, I am aware of none of our clients making use of 046 in their ILS. Date of Work maps specifically to 046 $k. That's a great field to add to bibliographic records, such as for motion pictures, where users want the original date of the movie, not the date the DVD was published. The 500 note for the details of the original movie is not machine-actionable and often buried in the record. We've added a single element line for Date of Work based on the 046 values to our brief displays. In the brief display, that Date of Work is of great significance because the 130 uniform title is not mapped for display there. Adding these qualifiers to the access point for the work does the job of creating a unique identifier for the work that can stand on its own. But in the end the user task these individual elements serve is identify. What matters is that the identifying data be presented somewhere to the user. The RDA approach of first emphasizing discrete data elements, rather than authorized access points, opens up so many doors for great functionality, such re-shaping web-based displays and entering data to be more amenable to machine action. The RDA records are much simpler, much cleaner, and far easier to understand than the MARC records. For you perhaps. One only has to spend a short time with them to come to conclusion. No tag numbers to memorize. No subfield codes. Little to no punctuation requirements. Logical consistency in element name labels for the same function, such as relationships between entities, and identifying variants, and co-ordinated notes on main elements instead of generic MARC 500. It's better to start with that, and then go ahead and teach the special implementation rules, such as constructing entries for collocation in card catalogs. One can start from the same base RDA instructions and then create an entirely different kind of catalog based on reciprocal relationships and hyperlinks, with more machine-actionable data. If that didn't exist today I would imagine people would be clamoring for a cataloging code that wouldn't be out-of-step with all the new technology and methods of organizing and displaying data. A clean break wouldn't cut it either, as others would be clamoring for a bridge from legacy catalogs to newer systems. Thomas Brenndorfer Guelph Public Library
Re: [RDA-L] Uniqueness of titles proper
steven Arakawa posted: I understand that work titles can conflict and we would need to break the conflict in such cases ... Only if neither has an author main entry (or author as part of AAP as Kevin would say). Of course two different works should not have the same preferred title if they are by the same person, but that is a rare problem. The best way to avoid the title proper of the first work in a collection being the title proper of the collection, is to supply a collective title, and record the individual work titles in 505. If at least one of two identical titles proper has a GMD, our clients consider that different GMD as distinction enough, and do not want a 130 or 240. (That's one of the reasons some want GMDs inserted in RDA records.) __ __ J. McRee (Mac) Elrod (m...@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__