========================================================================= Date: Fri, 1 Aug 1997 16:33:54 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Jean L. Hirons" Subject: Corp.main entry & uniform titles MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Colleauges: In a previous message to this list I said that I would separately comment on corporate main entry and uniform titles. In this message I want to address three things: 1) corporate authorship and main entry for serials; 2) uniform titles for serials created according to 25.5B to distinguish different serials with the same title; and 3) uniform titles used to collocate serials for translations and language editions. One of the hallmarks of AACR2 is that it provides "an integrated and standardized framework for the systematic description of all library materials" (AACR2, preface). This goal of standardization is carried from Part I into Part II where rule 21.1B2 governs the entry for all types of library materials, granted with a few specific cases separately accommodated. I have come to doubt whether this goal is realistic and wonder whether trying too hard for conformity has meant that some publications are forced into shoes that just don't fit. Reading Martha Yee's paper and speaking to other AV experts convinces me that title main entry is a good thing for this category of publications. But surely not for classical music, most books, and many serials. Are there general principles that we can develop that will more rationally deal with the types of publications that should and should not be entered under corporate body? Being a serialist, I have to admit that I've never understood the problem with the concept of corporate "authorship." Perhaps I think of "authorship" in a broad sense, but certainly an organization is the "author" of its annual report. Or at least it's the only author that anyone really cares about or would think to search under! So I was relieved to see that no one so far is calling for doing away with the main entry concept, as has been suggested so many times in the past. For serials, however, this is not enough. I was pleased with Hagler's recognition that "the high incidence of non-distinguishing titling" is a serious problem for the identification of a work and that the addition of the corporate body is often the only way in which the work can be identified. Unlike the earlier rules, AACR2 does not accommodate generic titles in its rule for corporate body main entry. AACR2's limitation on corporate body main entry has resulted in a need to use uniform titles to distinguish different serials with the same titles. (Some might argue that these aren't really "uniform titles"; however, we use the same fields and have yet to come up with a better term.) The code says nothing about what to do when the qualifier changes, but LCRI 25.5B says to create a successive entry when a corporate body qualifier changes but not when the place changes. Thus, we have tended to favor place as a qualifier, even though we often would prefer the more identifying corporate body. Our rationale has always been that the uniform title is only needed to distinguish one serial from another, but the desire on the part of many is for the uniform title to identify, not just distinguish. How can a patron pick out of from a long display of similar titles the one he is seeking when the qualifier is the place of publication rather than the more meaningful name of the issuing body? I've often thought it would just make a lot more sense in such situations to enter the serial under corporate body to begin with and forget about the artificial construct of the uniform title in such cases. Jim Cole recently reminded me of the former rule 6 in the British text of AACR (1st ed.) that required a corporate main entry for a serial whose title was generic or included the name or abbreviation of the name of the issuing body. In an article he contributed to Serials Librarian, Cole suggests that this is a viable alternative to the current 21.1B2 and that it be applied to all publications. Or, as Hagler discusses, perhaps we should adopt the concept of "key title" as defined in ISBD(S) and used in conjunction with the ISSN. The key title is a constructed title meant to identify the serial. We currently construct and give both key and uniform titles in our records, which many times are identical to one another. This is one redundancy that we need to eliminate. The problem is that AACR2's emphasis on the nature of the contents, while ideal perhaps, is just not practical. We need to be able to identify the serial from an "entry" for purposes of citations, online search displays, links, and added entries without having to look at the entire bibliographic record. We now have six ways of doing this, depending on the situation: title (245); corporate body and title (110/245); uniform title (130) (per 25.5B); uniform title and title (130/245) for translations, language editions, and legal works; corporate body and uniform title (110/240); and key title (222). Surely we can simplify this! An unrelated but long time problem for serials catalogers has been the required use of uniform titles for language editions and translations according to (25.3C, 25.5C1). While we understand the usefulness of such collocation for literary works, most serials are a different breed altogether. Does a reader seeking an article in English from a translated Russian journal really want to look up the journal under the Russian title? The problem for serials catalogers is that the original title can change (as happened recently with many of the former Soviet journals), while the translated title the same. A new entry must be made based on the change in uniform title, even though the title under which it is probably checked in and most likely to be sought has not changed. Is the work and its collocation really so important in such situations? Are there better ways of bringing them together, such as linking fields (765 and 767) which can hot link to the related record if one should have the original in one's library? In FRBR terms, when is the work foremost and when is the expression more important? Is it always in the best interests of our users to keep all aspects of the work together? Jean Hirons Acting CONSER Coordinator LC ========================================================================= Date: Fri, 1 Aug 1997 19:55:18 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Bibliographic relationships Comments: cc: autocat@ubvm.cc.buffalo.edu The newly posted paper for the future of AACR conference, Bibliographic relationships by Sherry L. Vellucci, clocks in at 35 pages. Vellucci rarely uses a two syllable word where a four syllable one is available with the same meaning, so these 35 pages require careful reading. She lays a firm historical foundation, beginning with Panizzi's and Cutter's purposes of the catalogue, and continuing with Tillett's and later studies of bibliographic relationships. Occasionally one feels like faulting her, as she did one IFLA study, for being too theoretical. But then she pulls one back down to earth with some actual examples of authority records. Reading this paper is made easier for me in that she uses "main and added entries" (p.17) to mean access points as opposed to the whole record so entered, which is the way those terms are used by my clients. The distinction already made in other papers between work and item is here extended to three levels: work, item or publication, and copy. She reports studies which describe as many as four levels. The idea of work authority records for all works is supported (also called access control records (p.24)). My major concern with the universal application of uniform titles is our clients' rejection of titles in a language different from the text. They will accept uniform titles in 730, but not 130 or 240. I don't recall ever having the same text in two languages in the collection of any of our 50 special library clients. In collocating for the global bibliographic database, the requirements of specific libraries should not be neglected. (Law firms whose working language is English even want English parallel title transcribed first for bi and multi lingual works, regardless of the order of languages on the title page.) Describing present methods of expressing relationships, she mentions dash entries (p.15). We abandoned dash entries, even for cards, when we adopted MARC. Is anyone still using them? On the other hand, the listing of titles under an entry in the book catalogues we print does look a bit like dash entries used to look. (Those titles are, by the way, a mixture of titles for which that entry is the main entry or an added entry. The same is true of author search OPAC displays in the PC based systems we support.) She proposes consistent treatment of parts, as opposed to direct entry of parts with distinctive titles. Our clients would not accept this. They very much dislike 245$a$p titles. They often have us change the 245$a to a 440, and the 245$p to 245$a. Aren't all titles issued in series, parts of a larger whole? Where would the line be drawn, if not at distinctive title as at present? This suggestion suggests a too long absence from dealing with the consumers of catalogue records one creates. The series entry can collocate the whole work, not the item or copy main entries. She wisely gives attention to the interrelationship of rules and MARC, the impact of the globalization of the bibliographic data base, the possible impact of hypertext, the fact that catalogues may describe and give access to remote resources, and the design and structure of catalogues as related to the code. Her four concluding principles are that the bibliographic record should identify bibliographic relationships (independent and dependent); these relationships should be reflected by linkages; multi level description (work, item, copy) should be provided; like bibliographical relationships should be treated consistently. Both identification and linkages should be bidrectional. The way 780/785 are now used would seem to me to come closest to her ideal of bidirectional linkages. Yee's point that more than bidirectional linkages are needed to unite serials which have gone through several title changes was not picked up. Yee's statement that change in authorship may not signal a new work might lead to the collocating of the successive editions of standard legal texts with differing authorship statements in the edition area, but she does not specifically mention that problem, as she does the successive title change problem for serials. Of the papers so for, Yee's seems to me to be the most anchored in actual experience creating records for patrons. __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sat, 2 Aug 1997 09:27:16 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Component parts and successive parts Comments: To: G.Cornelius@NHM.AC.UK Comments: cc: autocat@ubvm.cc.buffalo.edu In-Reply-To: <199708020358.UAA11263@bmd2.baremetal.com> Gillian Corenlius remarked: >Currently our library uses UKMARC and each level of a multi-level work (with >it's corresponding statement of responsibility) is usually entered in field >248. I could get excited about 248 as opposed to 245$p. It would give more flexibility in mapping, e.g., 248 alone on circulation card and pocket labels, since currently one usually doesn't get as far 245$p before running out of room. (Most code discussions seem to proceed without reference to our very mundane daily concerns.) For those of us a bit awed at the concept of work authority records (doubling our work load we fear), how about collocating successive serial records either by making 780/785 repeating for all past and future titles, or by introducing 244 for later titles and reactivating 247 for former titles? Or would it be considered wrong to use a 2XX field for a related work, since all related works are still only in 7XX? If we are to have a work authority record for the whole span of a serial through multiple title changes, it would seem to me that record would have to have something in the way of multiple title fields for the successive titles. It seems to me that the same serial under a different title, while not a constituent part (except in the work record?), is a bit more closely related than other 7XX "related works". Mac J. McRee (Mac) Elrod Vancouver Island, British Columbia, Canada Homepage: http://www.islandnet.com/~jelrod/mac.html ========================================================================= Date: Sat, 2 Aug 1997 22:03:09 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Some ideas for the ideal catalog Having mused over some of the conference papers, there are some thoughts that I felt I should tie together and see if they make sense to others. The discussion of authority records for all works would seem to mean that three types of authority records would result: works, authors, and subject headings. A key question would be how to tie them all together using rules that are equally applicable in a card catalog format and in an on-line environment. The list of works would in some ways resemble a name-title section of a card catalog. These would be some principles: - The concept of main entry as the primary collocating device would remain here, as one must always account for a single entry arrangement, such as physical arrangement on the shelf. - The list of works would really be a list of "uniform headings for works," in either main entry title or name-title form. - Cross-references would have to replace the concept of simple title added entries right across the board. This would be the sharpest departure from the card catalog design. All variant titles would point to a uniform title or name-title, which in a card catalog would require a trip to a different section of the catalog to see the full record. In an on-line environment, the resulting flexibility for display designs could be enormously beneficial. For example, common titles (such as "Annual report") would be split into mutliple SEE references pointing to the respective name-title headings. - Expand the use of qualifiers. For example, all works which are series should have the qualifier (Series), even in the event of no conflict. The same can be said for serials, television programs, and works originally in a different language. In MARC, the qualifiers should be in a separate subfield, so they can removed from display as appropriate. - Subsets of this list of works could be created in an on-line environment, such a a list of series, serials, videos, etc. Having the qualifier sheared off for these subsets would be beneficial. While the display may seem unusual at first, in that name-titles will file alongside titles, the benefits seem to be quite substantial. The actual headings for works could be highlighted, put in bold, etc. Under authors and subjects, the card catalog immediately points to the main entry form of the name of the work, or the name + 245 form if the on-line form is flexible enough. For authors, this is not sufficient, because all references to variant titles for which the author may have been responsible, even in a secondary capacity, would be lost. Perhaps in an on-line display, under the author, we could construct a complete subset of all the titles found in the list of works, with appropriate cross-references. A patron could scan the list, and if he or she only knows a variant form of a title, then that should appear alphabetically in the list pointing to an established heading for the work. This would make the most sense when scanning for titles translated from another language. Once a work is selected, then what? This is where Martha Yee's table of relationships comes into play. Under the selected work would be a set of options: Edtions of ... Works containing ... Performances of ... Works about ... Other works related to ... Added to this could be a complete breakdown of the title changes of a serial, with volume and date information somehow incorporated. For example, someone searching a serial title would look in the list of works. There might be a cross-reference to an established heading for the serial (which may be an early title bearing no resemblance to the searched title). Once the work is selected, the table of relationships would reproduce a list of all known titles that the serial may have been catalogued under, including the one for which the patron searched. All other relationships would also be mapped into this table. Holdings records could be attached to the authority record for the work, and be mirrored from each different title under which the serial has been catalogued. Reproducing such a system in a card catalog may be possible (although it would seem to me that a proliferation of cards would result). For an on-line environment, I can only see an enormous increase in the usability of the catalog. This is how I see things at this point: 1. List of authority-controlled names of works (really all titles cross-referenced to established forms, with name-titles to fulfill the persistent need for single-entry collocation) ------------> Table of relationships under each work (The list of works could be broken into different subsets in an on-line environment.) 2. List of authority-controlled names of authors, corporate bodies, etc. Cross-references from variant forms of the names would appear here. ------------> list of all titles associated with the author, displayed in list of works form ---------------------------> table of relationships under each work 3. List of authority-controlled subject headings (including works and authors, but with a single point of control to ensure consistency). Cross-references, etc., for subject headings would appear here. ------------> list of all titles associated with the subject, displayed in list of works form ----------------------------> table of relationships under each work The simplest arrangement would in fact be the list of works. The author and subject lists would introduce another layer in the hierarchy, as authors and subjects would have their own authority records and associated cross-references. Perhaps a superset of all authority records could be created for technical services staff. This would be useful for conflict elimination, and to have a centralized place for names and titles also appearing as subject headings. The subject heading organization seems a little haphazard, but I think it may be appropriate to be directed to the actual texts of works held in the library when searching the title as a subject. Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Sat, 2 Aug 1997 16:24:03 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Some ideas for the ideal catalog Comments: To: ThomasB@MSN.COM In-Reply-To: Thomas, Thank you for helping us visualize what a catalogue using work authorities might be. I'm having some difficulty translating this into how a printed book catalogue would look - the form of catalogue used by about one third of our customers. Multiple indentations eat up paper. If *each* item catalogued in the library is also to have a work record, there must be, I think, a way of automatically abstracting it from the record for the item if that item represents the first time the work is catalogued. Or conversely, a way of automatically carrying the information from the work record forward into the record for the item. Creating *two* records for each item would overburden already overburdened cataloguers. OPAC software must be flexible enough to allow the suppression elements not wanted in displays. For example, monolingual special collections would *not* accept the display of titles in languages other than the language of the item text. The collocating function is not required if the library only owns texts in one language. >The discussion of authority records for all works would seem to mean that >three types of authority records would result: works, authors, and subject Perhaps works, persons, and subjects. Not all persons are authors. And not all personal added entries are for authors. I assume "persons" includes both individuals and organizations. >- Expand the use of qualifiers. For example, all works which are series should >have the qualifier (Series), even in the event of no conflict. Hmmm. Some libraries catalogue a "series" as a serial, while others catalogue it as monographs in a series. Just yesterday I did what quacked like an annual serial, but then added individual distinctive titles and authors after the first several numbers had been published. >Holdings records could be attached to the authority >record for the work, and be mirrored from each different title under which the >serial has been catalogued. I would seem better to me to have the holdings statements associated with the item records. How would you distinguish between the print, electronic, and microform versions in a unified work record? Somehow holdings attached to a work authority record just doesn't fly for me. The work authority record acting as a device to pull together the holdings of the serial under different titles and in different formats would seem a better way to go. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sun, 3 Aug 1997 11:15:13 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: JSC conference homepages Comments: cc: autocat@ubvm.cc.buffalo.edu An increasing number of people are asking me off list how to access the future of AACR conference papers. Is the general information page not linked to the papers page and vice versa? I would gather not from the questions I am getting. (Perhaps here is an instance where Vellucci's bidirectional linkages are needed.) Perhaps the urls could be reposted occasionally to autocat and other lists for people who were on holiday when they were originally given. Broad discussion seems vital to me. If the implications of some of the papers are carried through, the changes created by AACR2 will seem quite minor. They are talking about the first major reformation of the way catalogues are created since Panizzi. If we are caught unaware, we will only have ourselves to blame. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sun, 3 Aug 1997 11:26:51 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Impact of work authority records The more I think about the implications of work authority records, the more worried I get. We have about 300,000 bibliographic records in our working file at the moment. Allowing for successive entry serial records, and successive editions of titles, I would expect we would need about 200,000 additional records if we were to have work authority records. (We simply don't have translations.) How this would impact on what we would deliver our customers I can't imagine. At present only a tiny minority of our customers have authority records as part of their OPACs. We have name and subject authorities here, and use them to create cross references in book catalogues, and of course to control the headings in our own records. But by and large they do not migrate to customer catalogues. I suspect the same is true for most libraries with PC based OPACs. Much more helpful to us would be a way of *reducing* rather than *increasing* records. Customers complain about what they consider redundant records. If work records could *replace* item records, perhaps with repeating collations and (in MARC records) 008 and 006 broken out into separate fixed fields (as is the case in Catss), we could combine many of our item records. For example, multiple 300s would allow the records for the sound cassette and sound disc of the same music recording to be combined; the records for the film reel, VHS video cassette, Beta video cassette, and disc, of the same feature film to be combined; and the records for the print and microform of the same serial to be combined. In all of the above cases, the records are identical except for the microform gmd, collation (not seen at every entry point), and fixed fields (never seen). If this avenue is to be explored, MARC revision *must* be discussed in tandem with code revision. We simply can't continue to have these two basic standards develop independently of each other. In sum, I think we need to explore reducing the number of records in our files, not expanding them by yet another layer. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 4 Aug 1997 09:32:35 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Impact of work authority records In-Reply-To: <7MK5zEJ3B0MD092yn@slc.bc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII "Work authority records" would only be useful if a particular catalogue contained more than one record for a work. I would guess that would me that in a catalogue of 300,000 records, you might only need 50,000 work authority records at the most. A cataloguing system which had such records would need to do a couple of things: (1) Warn the cataloguer when it look as if two different bib. records involved the same work -- e.g., if the had an author and a title in common -- so that the cataloguer could exxamine the records. (2) If the catalogue was obtaining bib. records from a system like OCLC, only download the work authority records for which the catalogue had two or more corresponding bibliographic records. This should (at least in concept) be easy to implement as an automatic process in the centralised cataloguing source, if that source had kept a record of what bib. records had been sent to the catalogue. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 On Sun, 3 Aug 1997, J. McRee Elrod wrote: > The more I think about the implications of work authority records, the > more worried I get. We have about 300,000 bibliographic records in our > working file at the moment. Allowing for successive entry serial > records, and successive editions of titles, I would expect we would need > about 200,000 additional records if we were to have work authority > records. (We simply don't have translations.) ========================================================================= Date: Sun, 3 Aug 1997 15:55:33 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Bibliographic relationships MIME-Version: 1.0 I found Sherry Vellucci's paper on bibliographic relationships rather alarming. I would be amongst the first to insist that catalogues should rigorously and consistently reflect the different kinds of relationship between works, but the following points seem equally important: 1. In my experience the majority of monographs are one-off manifestations of single works, never to be repeated. Are we in danger of creating a cataloguing sledgehammer to crack a bibliographic nut? To change the metaphor, are we sure that the big "work" in multiple editions is not going to become a dinosaur? 2. Most OPACs cannot cope properly with the complexity of AACR2. This is particulary true of Web-based OPACs: take a look at how the LC experimental OPAC displays uniform titles. Isn't it a bit soon to be suggesting that we increase the burden on OPACs? 3. MARC is one way of implementing AACR. It is not synonymous with AACR and there are many smaller libraries which do not use it. 4. Library users are frequently baffled by catalogue entries and the more we depart from established practice in bibliographic citation the worse this will be. 5. AACR is difficult to apply and even national bibliographies are riddled with errors. We are not going to be getting more staff with higher qualifications. Shouldn't we be thinking of making it *easier* to apply? -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Mon, 4 Aug 1997 14:17:48 +1000 Reply-To: Giles S Martin Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Bibliographic relationships In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII On Sun, 3 Aug 1997, Robert Cunnew wrote: > I found Sherry Vellucci's paper on bibliographic relationships rather > alarming. I would be amongst the first to insist that catalogues should > rigorously and consistently reflect the different kinds of relationship > between works, but the following points seem equally important: > > 1. In my experience the majority of monographs are one-off > manifestations of single works, never to be repeated. Are we in danger > of creating a cataloguing sledgehammer to crack a bibliographic nut? To > change the metaphor, are we sure that the big "work" in multiple > editions is not going to become a dinosaur? It may be true that most of the works in a library are one-off manifestations; but I don't think that the same is true as a proportion of what users look for. There is likely to be a strong correlation between demand for a work and the number of different manifestations of a work. This is likely to remain true of mayerial published in physical manifestations. It may not be true of Internet resources, which generally appear in only one copy -- very few sites are mirrored somewhere else. > 2. Most OPACs cannot cope properly with the complexity of AACR2. This is > particulary true of Web-based OPACs: take a look at how the LC > experimental OPAC displays uniform titles. Isn't it a bit soon to be > suggesting that we increase the burden on OPACs? The Web OPACs that I have seen are just Web front-ends to a normal OPAC database, and sometimes don't even have the same functionality as their text-based counterparts. It should be possible to move a lot further away from the older model. You could have a system with one Web page for each bibliographic and authority record in the tradition MARC-based OPAC, all linked to each other with hyperlinks corresponding with access points, references and tracings. On top of this could be a key-word based search engine, which would give priority to authority records over bibliographic records (so that when you looked for "Mozart", you would find the authority records for W.A. Mozart before all the hundreds of bib. records). If the system had "work authority records", these would be intermediate in position between the author pages and the manifestation pages, and would have an intermediate priority (after authors, and before manifestations) in key-word searches. Constructing these Web pages from MARC records would be the sort of task that could be carried out by a computer program, and then tidied up manually. > 3. MARC is one way of implementing AACR. It is not synonymous with AACR > and there are many smaller libraries which do not use it. Agreed. In addition, there are several dialects of MARC that each implement AACR differently. > 4. Library users are frequently baffled by catalogue entries and the > more we depart from established practice in bibliographic citation the > worse this will be. The catalogue display needs to explain a lot of thinngs to the catalogue user. The traditional 3x5 card display includes a lot of information is a very limited space, so that experienced users can identify what the various parts mean; but I am not convinced that most users unerstand all of the information there or how they can use it. > 5. AACR is difficult to apply and even national bibliographies are > riddled with errors. We are not going to be getting more staff with > higher qualifications. Shouldn't we be thinking of making it *easier* > to apply? The first priority should be to make life easier for the catalogue user -- to "save the time of the reader", in the words of Ranganathan. The second priority should be to do this in a cost-effective way. These priorities should mean that AACR should be as easy as possible to apply, given that it has to describe a complex bibliographic universe. And these should also mean that errors should be minimised, since they waste the time of both catalogue users and cataloguers. However, are the errors caused by AACR being hard to use? I suspecct there are other causes, including some cost-cutting by library managers. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 ========================================================================= Date: Mon, 4 Aug 1997 08:20:14 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Component parts and successive parts MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT "Mac" wrote: > > I could get excited about 248 as opposed to 245$p. It would give > more flexibility in mapping, e.g., 248 alone on circulation card and > pocket labels, since currently one usually doesn't get as far 245$p > before running out of room. > But where is this better than (his earlier suggestion) turning 245$p into 245$a and 100/245$a into a 700$a$t? B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 4 Aug 1997 00:16:18 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Component parts and successive parts Comments: To: EV@BUCH.BIBLIO.ETC.TU-BS.DE In-Reply-To: <65FAB4412D0@buch.biblio.etc.tu-bs.de> Bernhard Eversberg wrote: >> I could get excited about 248 as opposed to 245$p. >But where is this better than (his earlier suggestion) turning 245$p >into 245$a and 100/245$a into a 700$a$t? It avoids a duplicate entry under the 100. It would also be better when the 245$p/248 title was not distinctive enough to function well as a 245$a. (I am *very* opposed to Vellucci's suggestion that all parts, whether with distinctive titles or not, be entered subordinately.) Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sun, 3 Aug 1997 23:50:52 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Impact of work authority records Comments: To: ulgsm@DEWEY.NEWCASTLE.EDU.AU In-Reply-To: Giles said: >"Work authority records" would only be useful if a particular catalogue >contained more than one record for a work. I would guess that would me >that in a catalogue of 300,000 records, you might only need 50,000 work >authority records at the most. Hagler advocates making a work authority record for *every* work, whether there are multiple items of the work or not. What you propose would be like only making authority records for authors or subjects if cross references are needed. Some libraries do this, but most libraries have authorities for all headings, cross references needed or not, if they have authorities at all. Certainly creating work authority records only when there were two or more instances of the work would greatly reduce the impact on work load. About 50% of our cataloguing is original. Neither OCLC, RLIN nor Catss have records for much of the specialized material acquired by our libraries at the time the material is acquired. I am sure there are other special libraries for which this is the case. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 4 Aug 1997 10:17:04 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Component parts and successive parts MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT J. McRee Elrod wrote: > Bernhard Eversberg wrote: > > >> I could get excited about 248 as opposed to 245$p. > > >But where is this better than (his earlier suggestion) turning 245$p > >into 245$a and 100/245$a into a 700$a$t? > > It avoids a duplicate entry under the 100. > But that, surely, is only a card argument? > It would also be better when the 245$p/248 title was not distinctive > enough to function well as a 245$a. (I am *very* opposed to Vellucci's > suggestion that all parts, whether with distinctive titles or not, be > entered subordinately.) > Whereas we (in Germany) would welcome this very much since it would bring our separate galaxies closer. Here, the utilities' databases have more functionality, and from the view of cooperative cataloging and resource sharing, USMARC databases look like quite a bit of a mess when it comes to series and multiparts. We want all holdings for all items attached to the right level, and as little duplication as possible. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 4 Aug 1997 12:05:44 METDST Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Relationships: what can be done now Relationships : What can we expect, what can we do? Ground-breaking work has been done, by Vellucci, Tillett, Smiraglia, Leazer and others, to shed light on the nature and the manifold aspects of bibliographic relationships, and the requirements of catalog users and of consumers of bibliographic records have been analyzed. Whatever the type of relationship, the important practical consideration is, how can we provide devices like access points, index displays, hot links, or collocation lists to enable catalog users to navigate cobwebs of related records. Chances are poor for relationships expressed only in notes - nobody on this list needs to be told about the inadequacies of keyword access. Of course, it depends largely on the database software, but no software can be intelligent enough to turn contents notes into name-title access points of the same quality as from a 100/240 combination for example. So the rules have to tell us in what way to express relationships in the record so as to ensure proper access points can be generated. "Relational" databases are not the answer for a proper handling of bibliographic relations, although the very name seems to imply that. Sherry L. Vellucci draws attention to studies (Green,1996) that dispel this popular misconception (on p.29 of her paper). In this posting, I'm not trying to theorize about what AACR3 and a revamped USMARC should be like. I'm asking, instead, what can be done with present-day data and existing software. For these will be around for a while before a kinder and gentler bibliographic universe opens up. As of now, all relationships expressed through the presence of a 240 or a 700$a$t or 600$a$t can be turned into access points to support collocation of related publications ("manifestations of expressions of works"). In Tillett's terminology, the relationship types of "equivalence", "derivative", "descriptive" (as long as reflected in a 6XX) and "sequential" can thereby be collocated in a browse list. And here, it may well be asked if existing software is already making the most out of these fields as they are. And here, more than anywhere, keyword access is not the answer. Allow me a little aside here: From what I read, the AACR Future Development papers say precious little about what links should be presented in what ways at the user interface of OPACs. Is it all to be trusted to the vendors? That's the tradition, as we all know, but will the Toronto Conference indeed not concern itself with questions like what indexes a catalog should have, how names, titles, keywords, dates, place names etc. should be treated when indexed, how a brief entry should be constructed for the presentation of result sets, in what ways collocation might be achieved in the online environment, whether to keep up ISBD formatting as a standard or to favor labeled displays, and so on and so further. I mean, the details of cards had been nailed down in the code to the last dot, but what about OPAC displays? Is it not about time to think about guidelines or minimal standards here, at least? But now, it gets practical. Having been involved with questions of improving international exchange, I have done some experimenting, mostly with music records, to see and to demonstrate what can or might be done with USMARC data as they are. Look at this, it is a screenful from an alphabetical uniform title index, which can be browsed up and down, line by line: ------------------------------------------------------------------------------ 1 mozart, wolfgang amadeus: quintets, clarinet, k anh 90;score 1 mozart, wolfgang amadeus: quintets, horn, k 407 ;score arr 6==>mozart, wolfgang amadeus: quintets, horn, k 407;rec 5 mozart, wolfgang amadeus: quintets, horn, k 407;score 1 mozart, wolfgang amadeus: quintets, horn, k 407;score arr 23 mozart, wolfgang amadeus: quintets, piano, k 452;rec 4 mozart, wolfgang amadeus: quintets, piano, k 452;score 1 mozart, wolfgang amadeus: quintets, piano, k 452;score arr 1 mozart, wolfgang amadeus: quintets, piano, k 478;rec 2 mozart, wolfgang amadeus: quintets, piano, oboe, clarinet, horn, bassoon 1 mozart, wolfgang amadeus: quintets, strings, select;score 1 mozart, wolfgang amadeus: quintets, violins, k 406;score arr 3 mozart, wolfgang amadeus: quintets, violins, k 515;rec 5 mozart, wolfgang amadeus: quintets, violins, k 516;rec 3 mozart, wolfgang amadeus: quintets, violins, k 593;rec 1 mozart, wolfgang amadeus: quintets, violins, k 593;score 2 mozart, wolfgang amadeus: quintets, violins, k 614;rec 1 mozart, wolfgang amadeus: quintets, violins, k 614;score 4 mozart, wolfgang amadeus: quintets, violins, no 4;rec arr 3 mozart, wolfgang amadeus: quintets, violins;rec 4 mozart, wolfgang amadeus: quintets, violins;score New search term? Just type it! [Sh+F8] = Extended index" [Enter] = Title display [F1] = HELP ------------------------------------------------------------------------------- This index was composed by appropriate software out of unadulterated USMARC data, using the contents of 100/240, 600$a$t and 700$a$t, plus several other subfields. (Nothing new involved here, no "work authority records" of any kind!) To be more precise (since not everything is apparent here), the following subfields have been indexed, in this syntax: $m, $n, $s, $k, <$p>;medium ($l) with "medium" = "score", "rec" or "book". If you now press [Enter], it brings up the first of the records belonging to the line the arrow points to: ------------------------------------------------------------------------- Mozart, Wolfgang Amadeus: [Quintets, horn, strings, K. 407, E flat major.] [This appears as part of:] (press '0': main record '2': related records) The Art of the French horn sound recording. [S.l.] : Everest, [197-]. - 1 sound disc : 33 1/3rpm, stereo. Dennis Brain, horn; Max Salpeter, violin; Cyril Preedy, piano (in the... [Subjects:] 1. Trios (Piano, horn, violin). 2. Quintets (Horn, violin, violas (2), violoncello). 3. Horn and piano music ------------------------------------------------------------------------- (BTW, seeing a record display, you can press F7 to see all index entries belonging to it.) This gets us to the whole/part relationship: in fact, what we did was to extract all 700$a$t fields and make 100/240 combinations out of them, embed these into a basic record frame and then link these part records with the main (container) record using its control number. This way, more or less, every work listed in the contents note gets its own little analytic record. If you press '2', the index is displayed at the related 100/240 combination. If you press '0' as the display suggests, the programm collocates the related main record with all subordinate records and you see this: --------------------------------------------------------------------------- Brain, Dennis: The Art of the French horn sound recording. [S.l.] : Everest, [197-]. - 1 sound disc : 33 1/3rpm, stereo. Dennis Brain, horn; Max Salpeter, violin; Cyril Preedy, piano (in the... The 3d work originally for viola da gamba and continuo CONTENTS: Brahms. Horn trio in E flat, op. 40.--Mozart. Horn quintet in E flat, K. 407.--Marais, M. Le Basque [Subjects:] 1. Trios (Piano, horn, violin). 2. Quintets (Horn, violin, violas (2), violoncello). 3. Horn and piano music [Added entries:] I. Salpeter, Max. -- prf. II. Preedy, Cyril. -- prf. III. Grainger, Eileen. -- prf. IV. Parry, Wilfrid. -- prf. V. Carter String Trio. -- prf. VI. Brahms, Johannes -- Trio -- piano, violin, horn -- op. 40 -- E flat major. VII. Mozart, Wolfgang Amadeus -- Quintets -- horn, strings -- K. 407 E flat major. VIII. Marais, Marin -- Pieces de violes -- 4e livre. -- 1er partie. -- 39- 40; -- arr. ----------------------------------------------------------------------------- Here, the Brahms, Mozart and Marais pieces are not really part of the main record any more but added to this display by way of linkage. It looks like the complete record before it was cut up. (This cutting up is not really necessary for the indexing, but we did it for other reasons: to demonstrate whole-part linking.) The real displays are more instructive (on a PC screen) because of color. If you want to see more, you can telnet into this database: telnet 134.169.20.3 login: opac password: opac mm beethov (or any other name, like mm copland ) (that's the command to access the MarcMusic database) the index will be displayed at "beethoven" Then, follow instructions (and use '?' for help). To get out, press 'x' on any title display, then 'y' to confirm. There are more indexes. The instructions tell you how to get into them. At least look at the anyword index (6) if you are a keyword addict. There are no authority records yet in this database, hence no references in the index. A name authority record would generate index entries like cajkowskij, petr iljic -> tchaikosky, peter ilich (This single one is present, you can check out how it works.) It appears that for OPAC users, well-designed browseable indexes are one way to make relationships visible and navigable. Using existing uniform title data, one can already do a decent job. Some formal improvement of the 240 and 700 subfields would get us further, and provision of more such fields for cases where presently none are prescribed by the code. The biggest issue may well be (as J. McRee Elrod indicated) the rules for the uniform title as such (240$a and 700$t), esp. with international exchange in mind and monolingual audiences to consider. A structured, browseable index has a welcome side-effect: it makes all those little inaccuracies of spelling and punctuation both very apparent but also less harmful (the example shown above has very few flaws though). MARC data tend to be less than perfect where indexing had not originally been intended or did not appear feasible. This way of indexing can still provide useful relationship collocation. The term "browseable index" is used here as opposed to internal indexes which cannot be looked at but which are used implicitly by "find" commands. Well, just compare the above with mere keyword access where you don't even ever get to see an alphabetical index display. (There is, however, an "anyword" index as well in this database, and you can browse it too.) The database has some 40.000 records, so there are examples for just about everything, at least for the music cataloger. But look at "shakespeare" and "macbeth" too. On the subject of whole/part relationships there's a bit more to say. But that's to be another posting. Regards, B.E. P.S. J. McRee Elrod wrote: [about relationships and work authorities] > If this avenue is to be explored, MARC revision *must* be discussed in > tandem with code revision. We simply can't continue to have these two > basic standards develop independently of each other. If this were indeed still the case, as I'm convinced it isn't, it would constitute a decent scandal. Conceptually independent of each other though these standards may have been conceived and may still be viewed by some theoretically inclined people, they are in actual fact inextricably intertwined. And who is using MARC for anything else but AACR records? Bernhard Eversberg Head of Library Computing Universitaetsbibliothek Braunschweig Germany B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 4 Aug 1997 10:05:23 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Redundancy in bibliographic records Comments: To: EV@buch.biblio.etc.tu-bs.de In-Reply-To: <66198EE0566@buch.biblio.etc.tu-bs.de> Concerning have a 7XX which duplicates the 1XX, you asked: >But that, surely, is only a card argument? It is even worse for COM and printed book catalogues than for a card catalogue. The library world is not universally OPAC. Generally speaking, I think we should attempt to move away from redundancy in bibliographic records. One way to do that is to stop deconstructing the record, display the ISBD as a unit, and without labels, as we did in the card catalogue. Just because we can do something, doesn't mean it should be done, e.g., deconstruct a bibliographic record and label its parts. Your comments on the need to have code standards for OPAC displays were excellent, I thought. Not only should the code and MARC not be discussed in isolation from each other, but bibliographic record standards should not be be created in isolation from standards for the catalogues in which they are to appear. This sort of practicality has so far not characterized most of the papers posted for the Toronto conference. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 4 Aug 1997 16:11:03 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Re: Some ideas for the ideal catalog -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of J. McRee Elrod Sent: Saturday, August 02, 1997 4:24 PM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Re: Some ideas for the ideal catalog Thomas, Thank you for helping us visualize what a catalogue using work authorities might be. I'm having some difficulty translating this into how a printed book catalogue would look - the form of catalogue used by about one third of our customers. Multiple indentations eat up paper. If *each* item catalogued in the library is also to have a work record, there must be, I think, a way of automatically abstracting it from the record for the item if that item represents the first time the work is catalogued. Or conversely, a way of automatically carrying the information from the work record forward into the record for the item. Creating *two* records for each item would overburden already overburdened cataloguers. [] **** I have always viewed bibliographic records as quasi-authority records. At this level, for works with only one published item, few changes would have to be made to AACR bibliographic records, it seems to me. For MARC, it would make more sense to have indicator values set up to create a true cross-referencing pattern for titles that mimic the author and subject authority indexes. Perhaps what is needed are indepedent values for different output devices. Currently, some indicator values in MARC have meanings geared for card catalogs (the 1st indicator for 245 for example). Why not leave these, but have different values geared for online displays? Today, MARC seems little more than a printer command language. Online advantages such as keyword or control number searches seem to be constructed in an ad hoc and haphazard fashion. Some collocation functions have completely disappeared. It would behoove the powers that be to make AACR an "output neutral" code. Or, at the very least, make provisions for the different logistical and logical requirements of card catalogs vs. online catalogs. ****** OPAC software must be flexible enough to allow the suppression elements not wanted in displays. For example, monolingual special collections would *not* accept the display of titles in languages other than the language of the item text. The collocating function is not required if the library only owns texts in one language. [] ****** Once work records are required, why not establish layers of authorized forms? For example, have the original language form followed immediately by a local language form. The second form could display as a qualifier, or it could replace the original form altogether in displays (although not in the electronic work authority record). The ideal form would marry universal bibliographic control with local bibliographic control, with no need to obliterate one for the other. ****** >The discussion of authority records for all works would seem to mean that >three types of authority records would result: works, authors, and subject Perhaps works, persons, and subjects. Not all persons are authors. And not all personal added entries are for authors. I assume "persons" includes both individuals and organizations. [] ****** I prefer "names" to "persons." Unfortunately, "names" could also mean "names of works" as in the LC name authority file, which include persons and works. ****** >- Expand the use of qualifiers. For example, all works which are series should >have the qualifier (Series), even in the event of no conflict. Hmmm. Some libraries catalogue a "series" as a serial, while others catalogue it as monographs in a series. Just yesterday I did what quacked like an annual serial, but then added individual distinctive titles and authors after the first several numbers had been published. [] ******* Series headings would appear as work records, but I suppose work records should be similar enough to bibliographic records to allow for quick conversion. By conversion, I mean changing descriptive elements such as collocation and how holdings records are attached. Or how a "table of relationships" is set up after the work is selected. In an online environment, one should, through a table of relationships, find all separately catalogued items if the series work record is selected, and also find all parts of the series if the work record for a part is selected. There is no compelling reason why a card catalog arrangement should significantly change from current standards with regard to series-- this should be a coding issue in MARC standards. On a related note, there is already a rudimentary hierarchy of established work headings in a bibliographic record: 240 --> 245 --> 246. The heading for a physical item would always be constructed out of the title proper (field 245), and a list of these headings should always be the final result of searching. What I would like to see is the simple concept of a "heading" enacted in full. Why not have the heading (i.e., title or name-title) be a distinctive, self-contained unit at the top of the record, instead of hopelessly mangled with the ISBD descriptive record? If truncated forms are required to make the data fit on a catalog card, then that should be the exception, not the rule. ********* >Holdings records could be attached to the authority >record for the work, and be mirrored from each different title under which the >serial has been catalogued. I would seem better to me to have the holdings statements associated with the item records. How would you distinguish between the print, electronic, and microform versions in a unified work record? Somehow holdings attached to a work authority record just doesn't fly for me. The work authority record acting as a device to pull together the holdings of the serial under different titles and in different formats would seem a better way to go. [] ****** Holdings records should still be attached to bibliographic records. However, in an on-line environment, there is no reason why a grouped function could not be allowed. Under the work record, one should be able to tap into all associated holdings, with the list arranged by the heading of each bibliographic item (i.e., name + title proper). Currently, I find checking holdings under each successive entry for a serial to be a real bother. Why not a master grouping >>option<< under a work authority record? Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Tue, 5 Aug 1997 21:02:02 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: Bibliographic relationships In-Reply-To: MIME-Version: 1.0 In article , Giles S Martin writes > >These priorities should mean that AACR should be as easy as possible to >apply, given that it has to describe a complex bibliographic universe. >And these should also mean that errors should be minimised, since they >waste the time of both catalogue users and cataloguers. However, are the >errors caused by AACR being hard to use? I suspecct there are other >causes, including some cost-cutting by library managers. > Giles, I think many errors *are* caused by the complexity of AACR and I'm not convinced we need all that complexity. We certainly don't need more. As for cost cutting: if cataloguers are less qualified or have less time to attend to detail, isn't this something that code revisers will need to take into account? Thanks for your comments. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Wed, 6 Aug 1997 08:53:56 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Re: Bibliographic relationships Robert Cunnew wrote (in part): "As for cost cutting: if cataloguers are less qualified or have less time to attend to detail, isn't this something that code revisers will need to take into account?" As the conversations on code revision proceed, this is the kind of question that's important to consider. I would ask the conference to consider to what degree specific sociocultural trends ought to be embedded in the code. In other words, cost-cutting (and, to put it bluntly, the dumbing-down of technical services) is a response to particular social conditions, ranging from tax revolts to the devolution of literacy (and elevation of the "quick-fix fix"). A revised AACR should acknowledge the actual world we live in, but ought it to reify these kinds of trends? I would rather see a code which actually declares a standard that takes a bit of work to achieve, and let individual managers make their excuses :-). I don't mean to misstate Mr. Cunnew's point -- it just provided me with a jumping-off point for some of my concerns. David Miller Levin Library, Curry College Milton, MA dmiller@curry.edu ========================================================================= Date: Wed, 6 Aug 1997 15:52:01 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Bibliographic relationships MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT David Miller wrote: > A revised AACR should acknowledge the actual world we live in, but > ought it to reify these kinds of trends? I would rather see a code > which actually declares a standard that takes a bit of work to achieve, > and let individual managers make their excuses :-). > This brings me to another suggestion: Should a revised code (or any cataloging code for that matter) have an introduction for the layperson? For that's what "individual managers" often are when it comes to cataloging. Were there this kind of introduction (5 pages maximum, better less), carefully outlining the philosophy and structure of the code (not an easy task, I knoe), it would become easier to point out to those managers why their excuses are weak ones or even irresponsible. But in any case, I think the "layperson's introduction" should be a good thing to have, for what we are doing must appear very esoteric to every taxpayer who didn't grow up in our profession. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Wed, 6 Aug 1997 10:21:35 EDT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Jenifer K. Marquardt" Subject: Layperson's introduction MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Content-Transfer-Encoding: 7BIT I agree with Bernhard Eversberg's suggestion of including a layperson's introduction in a new AACR. In addition to the reasons he mentioned, an introduction would also be valuable as a teaching tool, providing a consistent foundation for student's taking cataloging classes. Jenifer Marquardt Monographs Original Cataloger University of Georgia Libraries ========================================================================= Date: Wed, 6 Aug 1997 08:10:35 -0700 Reply-To: Daniel CannCasciato Sender: "International Conference on the Principles and Future Dev. of AACR" From: Daniel CannCasciato Subject: Re: Bibliographic relationships In-Reply-To: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hi All, to jump in here, too: > As for cost cutting: if cataloguers are less qualified or have > less time to attend to detail, isn't this something that code revisers > will need to take into account? In a time when there is more need for the ability to distinguish different materials from one another, I don't see this as a direction I'd like the profession take. We could, instead, accept that part of our task is difficult and commit ourselves to improving our workflows and the rules themselves. The code can certainly stand to be simplified, but there are a number of approaches to that. One is to simplify the rules themselves. Another, however, is to make explicit that which is ambiguous. This doesn't necessarily mean that the length of a bibliographic or authority record would be any shorter; it could very well be a longer, more detailed record. However, application of the rules would be a streamlined intellectual process if the rules themselves are clearer. Another might be to have continued education a mandatory aspect of professional participation. Those most familiar with the code and it's intent can (?) apply it more quickly and appropriately than those not as familiar. While this does not simplify the code, it does streamline the intellectual process of using the code. The next year or so should be pretty interesting, following this conference. Daniel ------------------------------------------- Daniel CannCasciato Head of Cataloging and Interlibrary Loans Central Washington University Library 400 East 8th Ave Ellensburg WA 98926-7548 509 963-2120 509 963-3684 (FAX) dcc@cwu.edu ========================================================================= Date: Wed, 6 Aug 1997 14:01:56 CDT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Jim E Cole Subject: Issues Related to Seriality I greatly enjoyed reading the Hirons/Graham paper "Issues Related to Seriality," which discusses matters that have long needed discussing and discusses them well. The possibility of considering ongoing unnumbered works to be serials brings to mind the ISBD(S) definition of a serial, which begins, "A publication in printed form or not, issued in successive parts usually having numerical or chronological designations and intended to be continued indefinitely." (ISBD(S), rev. ed., p. 6) It also brings to mind example 50 (p. 71), which contains a note "Collection non numerotee." I do wish that for all serials, both print and nonprint, we could return to cataloging from the latest issue. This does, to be sure, cause (or better, permit) certain changes to Areas 1-4 of the description that hitherto have been forbidden under AACR2, but it allows for the creation of records of use by patrons, check-in staff, and others who may recognize the latest form of a subtitle, for instance, and overlook a record having an earlier form in Area 1--even if the change is as simple as Appalachia : a journal of the Appalachian Regional Commission which has now become Appalachia : journal of the Appalachian Regional Commission. The basic problem that needs to be considered is the fact that many online catalogs--NOTIS, for instance, but even OCLC--alphabetize through subfield "b" (the subtitle), and this greatly affects the placement of the records in any index list. Also, if an online version of a serial is cataloged using the latest chief source and the print version is left behind with the earliest, then the two stand a good chance of being separated in the index of an online catalog; it would be wise to bring them together whenever possible. For check-in purposes, too, the current place of publication and publisher are much more of interest than, say, who published the serial in the 1870s when it began. Giving the earliest place of publication in Area 4--and then in a uniform title, if needed--can be a disservice to the patron, who associates the title with the current place of publication, which it has had throughout the bulk of its many years of existence, and does not even realize it ever was published elsewhere. Jim Cole --- Jim Cole, Editor THE SERIALS LIBRARIAN SL home page: http://www.ames.net/serialslibrarian ========================================================================= Date: Wed, 6 Aug 1997 22:12:06 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: Issues Related to Seriality In-Reply-To: <9708061901.AA25195@isua2.iastate.edu> MIME-Version: 1.0 In article <9708061901.AA25195@isua2.iastate.edu>, Jim E Cole writes > >I do wish that for all serials, both print and nonprint, we could return to >cataloging from the latest issue. I've always had a lot of respect for the present earliest-issue rule in AACR2. To me it's very pragmatic: it recognises that cataloguers haven't the time to check every issue for minor changes. >This does, to be sure, cause (or better, >permit) certain changes to Areas 1-4 of the description that hitherto have >been forbidden under AACR2, but it allows for the creation of records of use >by patrons, check-in staff, and others who may recognize the latest form of >a subtitle, for instance, and overlook a record having an earlier form in >Area 1--even if the change is as simple as > > Appalachia : a journal of the Appalachian Regional Commission > >which has now become > > Appalachia : journal of the Appalachian Regional Commission. For check-in, shelf-filing, holdings lists and many other purposes I would strongly advocate using key title or uniform title: ie title proper plus qualifier if necessary. That is what these are for. If your software doesn't permit this isn't the fault with the software rather than the rules? I know the qualifier might be the original place of publication. Perhaps we should move away from using place as a qualifier. It is becoming less important as communications improve. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Thu, 7 Aug 1997 10:03:53 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Issues Related to Seriality In-Reply-To: <9708061901.AA25195@isua2.iastate.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Wed, 6 Aug 1997, Jim E Cole wrote: > I do wish that for all serials, both print and nonprint, we could return to > cataloging from the latest issue. This does, to be sure, cause (or better, > permit) certain changes to Areas 1-4 of the description that hitherto have > been forbidden under AACR2, but it allows for the creation of records of use > by patrons, check-in staff, and others who may recognize the latest form of > a subtitle, for instance, and overlook a record having an earlier form in > Area 1--even if the change is as simple as > > Appalachia : a journal of the Appalachian Regional Commission > > which has now become > > Appalachia : journal of the Appalachian Regional Commission. > > The basic problem that needs to be considered is the fact that many online > catalogs--NOTIS, for instance, but even OCLC--alphabetize through subfield > "b" (the subtitle), and this greatly affects the placement of the records in > any index list. Our Innopac catalogue behaves the same way as NOTIS, so including the subtitle when filing is not at all unusual. The problem here is that the user could approach this title in three different ways: Appalachia Appalachia a journal ... Appalachia journal ... In each case they ought to be sent directly to the record, if there is no other title starting with those words. However, in most catalogues the one-word search on "Appalachia" will bring up multiple hits. In ours (which is a long way from Applachia), it brings up 22 hits starting with: Appalachia 1903 Appalachia In The Sixties Decade Of Reawakening Appalachia Preschool Test Of Cognitive Skills Apt Appalachia Variations On An Old Slave Song Brigg Fair A Appalachia Variations On An Old Slave Song With Final C Appalachian Folk Songs For Recorder Appalachian Geomorphology An Annotated Bibliography Appalachian Ouachita Orogen In The United States So even in our catalogue the three alternate forms of the title would file differently. > Appalachia Appalachia 1903 > Appalachia a journal ... Appalachia In The Sixties Decade Of Reawakening > Appalachia journal ... Appalachia Preschool Test Of Cognitive Skills Apt Appalachia Variations On An Old Slave Song Brigg Fair A Appalachia Variations On An Old Slave Song With Final C Appalachian Folk Songs For Recorder Appalachian Geomorphology An Annotated Bibliography Appalachian Ouachita Orogen In The United States The solution to the problem has two parts. The first is that OPACs should create two index entries for title-subtitle combinations -- particularly for very short titles, which will file a long way away from the title combined with the subtitle. The information is already present in the bibliographic record, and it is unreasonable for cataloguers to have to add an extra field when the operation can be carried out mindlessly by a computer. The second part of the solution is to allow, or even expect, cataloguers to make added entries which the sub-title varies and this may affect access to the record. This can be done regardless of whether you base the description on the first issue or the latest issue: it just requires someone to watch for subtitle changes; and you would have to watch for subtitle changes anyway if you catalogued from the latest issue. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 ========================================================================= Date: Thu, 7 Aug 1997 00:01:50 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Issues Related to Seriality In-Reply-To: from "Giles S Martin" at Aug 7, 97 10:03:53 am Content-Type: text Once again we have examples from particular OPAC systems that do not even follow current filing rule standards. What if Giles' example followed some more systematic set of filing rules (say filing by title proper (subfiled by other title) then the example he provides would look like: > Appalachia > Appalachia : a journal ... > Appalachia : journal ... > Appalachia 1903 > Appalachia In The Sixties Decade Of Reawakening > Appalachia Preschool Test Of Cognitive Skills Apt > Appalachia Variations On An Old Slave Song Brigg Fair A > Appalachia Variations On An Old Slave Song With Final C > Appalachian Folk Songs For Recorder > Appalachian Geomorphology An Annotated Bibliography > Appalachian Ouachita Orogen In The United States that's not so bad? the comments coming across in aacrconf once again seem to be using the logic: please make aacr conform to my opac system's particularities. i don't get it. is the function of aacr really to make notis or dra or iii work better? the recent comments about serials also seem to be considering serials processing needs over all other needs (check-in, etc.). of course, it may be that serials processing needs are more significant, but successive entry (I'm guessing) does meet the needs of potential catalog readers working from citations (assuming that writers of papers cite serial titles from the article/volume cited). i don't know if current citation practice of interest to anyone, but at least some catalog users are going to be coming to a catalog to follow a citation. it seems to me that the catalog should at least meat that known function (as well as all of the other hunches about how people may or may not use catalogs). --ralph papakhian Giles S Martin said > > On Wed, 6 Aug 1997, Jim E Cole wrote: > > > I do wish that for all serials, both print and nonprint, we could return to > > cataloging from the latest issue. This does, to be sure, cause (or better, > > permit) certain changes to Areas 1-4 of the description that hitherto have > > been forbidden under AACR2, but it allows for the creation of records of use > > by patrons, check-in staff, and others who may recognize the latest form of > > a subtitle, for instance, and overlook a record having an earlier form in > > Area 1--even if the change is as simple as > > > > Appalachia : a journal of the Appalachian Regional Commission > > > > which has now become > > > > Appalachia : journal of the Appalachian Regional Commission. > > > > The basic problem that needs to be considered is the fact that many online > > catalogs--NOTIS, for instance, but even OCLC--alphabetize through subfield > > "b" (the subtitle), and this greatly affects the placement of the records in > > any index list. > > Our Innopac catalogue behaves the same way as NOTIS, so including the > subtitle when filing is not at all unusual. The problem here is that the > user could approach this title in three different ways: > > Appalachia > Appalachia a journal ... > Appalachia journal ... > > In each case they ought to be sent directly to the record, if there is > no other title starting with those words. However, in most catalogues > the one-word search on "Appalachia" will bring up multiple hits. In ours > (which is a long way from Applachia), it brings up 22 hits starting with: > > Appalachia 1903 > Appalachia In The Sixties Decade Of Reawakening > Appalachia Preschool Test Of Cognitive Skills Apt > Appalachia Variations On An Old Slave Song Brigg Fair A > Appalachia Variations On An Old Slave Song With Final C > Appalachian Folk Songs For Recorder > Appalachian Geomorphology An Annotated Bibliography > Appalachian Ouachita Orogen In The United States > > So even in our catalogue the three alternate forms of the title would > file differently. > > > Appalachia > Appalachia 1903 > > Appalachia a journal ... > Appalachia In The Sixties Decade Of Reawakening > > Appalachia journal ... > Appalachia Preschool Test Of Cognitive Skills Apt > Appalachia Variations On An Old Slave Song Brigg Fair A > Appalachia Variations On An Old Slave Song With Final C > Appalachian Folk Songs For Recorder > Appalachian Geomorphology An Annotated Bibliography > Appalachian Ouachita Orogen In The United States > > The solution to the problem has two parts. The first is that OPACs > should create two index entries for title-subtitle combinations -- > particularly for very short titles, which will file a long way away from > the title combined with the subtitle. The information is already present > in the bibliographic record, and it is unreasonable for cataloguers to > have to add an extra field when the operation can be carried out > mindlessly by a computer. > > The second part of the solution is to allow, or even expect, cataloguers > to make added entries which the sub-title varies and this may affect > access to the record. This can be done regardless of whether you base > the description on the first issue or the latest issue: it just requires > someone to watch for subtitle changes; and you would have to watch for > subtitle changes anyway if you catalogued from the latest issue. > > Giles > > #### ## Giles Martin > ####### #### Quality Control Section > ################# University of Newcastle Libraries > #################### New South Wales, Australia > ###################* E-mail: ulgsm@dewey.newcastle.edu.au > ##### ## ### Phone: +61 49 215 828 (International) > Fax: +61 49 215 833 (International) > ## > The web of our life is of a mingled yarn, good and ill together > -- All's Well That Ends Well, IV.iii.98-99 > -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Thu, 7 Aug 1997 15:26:30 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Issues Related to Seriality In-Reply-To: <9708070501.AA16980@browndwarf.ucs.indiana.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I certainly don't want AACR to conform with everything that Innopac does, since I find some of its practices annoying. What I do want is for a library system to lead users from their OPAC searches as directly as possible to the record they want. The original posting talked about a journal with a varying subtitle. If users put in the subtitle as part of their search, then I think that they would expect to find both "Appalachia : journal ..." and "Appalachia journal ..." in the same place. In other words, from their point of view, they would expect the system to ignore colons in the same way as it ignores other punctuation in titles. They don't know that there is a subfield code there, they don't know the difference between "title proper" and "other title information", and they don't want to have to learn about these things. So I was saying that a title like "Appalachia : journal ..." should file in two places, because users could look for it in either way, and the two forms would be separated in most catalogues. We didn't do this in card catalogues or in book catalogues, because we wanted to keep the size of the catalogue down, but this is no longer such an important consideration in an OPAC. The logical corollary of that is that if the subtitle changes, then you need at new access point under the new title-subtitle combination. Incidentally, in the listed examples that I gave, you would need an extra access point under "Appalachia" for each of the two entries starting "Appalachia Variations On An Old Slave Song", since the title proper is just "Appalachia" (it's the musical work by Delius). Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 On Thu, 7 Aug 1997, Ralph Papakhian wrote: > Once again we have examples from particular OPAC systems that > do not even follow current filing rule standards. > What if Giles' example followed some more systematic set of > filing rules (say filing by title proper (subfiled by other > title) > then the example he provides would look like: > > > Appalachia > > Appalachia : a journal ... > > Appalachia : journal ... > > Appalachia 1903 > > Appalachia In The Sixties Decade Of Reawakening > > Appalachia Preschool Test Of Cognitive Skills Apt > > Appalachia Variations On An Old Slave Song Brigg Fair A > > Appalachia Variations On An Old Slave Song With Final C > > Appalachian Folk Songs For Recorder > > Appalachian Geomorphology An Annotated Bibliography > > Appalachian Ouachita Orogen In The United States > > that's not so bad? ========================================================================= Date: Thu, 7 Aug 1997 01:47:28 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Issues Related to Seriality In-Reply-To: from "Giles S Martin" at Aug 7, 97 03:26:30 pm Content-Type: text Greetings, I am not able to follow Giles' line of reasoning. Giles presents the hypothetical case: "If users put in the subtitle as part of their search, then I think that they would expect to find both "Appalachia : journal ..." and "Appalachia journal ..." in the same place." "If" users do such and so, then Giles "thinks" such and so. Such and so is surely going to happen now and then. If it is often enough (does anyone have any evidence?) then that would certainly be significant argument for AACR to do away with the distinction between title proper (the main title) and other title information. Somehow I don't imagine that this is a problem unique to serials. But if one eliminates that distinction, the whole "theory" of uniform titles (if there is such a theory) is thrown into disarray. What, exactly, is being proposed? Since we are engaging in hypothetical scenarios, what would happen if user X searched Appalachia : journal Appalachia journal Appalachia a journal and had no retrieval results? Do you suppose the user would try?: Appalachia Is there any monographic corolloary? Must the user searching either of these titles: Hamlet, a tragic drama or Hamlet, une drame tragique have all of there searching needs met? We can postulate many hypotheticals. Say, as an Armenian immigrant, I wander into the IU Music Library and, interested in hearing a recent CD, search for the title: Mazhig flud I find nothing! Should this need also be met by AACR? I'm sorry. I have not yet had the opportunity to read all of the papers posted for the AACR Conference. But I would hope that any discussion on this listserv could focus very carefully on the purposes of AACR and cooperative cataloging without digressing to peculiarities of various OPAC systems (which usually don't even follow existing standards) or to hypothetical unsuccessful searches, which all of us can create for any set of catalog rules. Cataloging/bibliographic rules have not, at least for about 400 years, attempted to account for any/all potential, possible, hypothetical inquiries (I think). They have tried to account for certain very basic searches common to the western European academic tradition, which now seem to be on the wane. --ralph papakhian Giles S Martin said > > I certainly don't want AACR to conform with everything that Innopac > does, since I find some of its practices annoying. What I do want is for > a library system to lead users from their OPAC searches as directly as > possible to the record they want. > > The original posting talked about a journal with a varying subtitle. If > users put in the subtitle as part of their search, then I think that they > would expect to find both "Appalachia : journal ..." and "Appalachia > journal ..." in the same place. In other words, from their point of > view, they would expect the system to ignore colons in the same way as it > ignores other punctuation in titles. They don't know that there is a > subfield code there, they don't know the difference between "title > proper" and "other title information", and they don't want to have to > learn about these things. > > So I was saying that a title like "Appalachia : journal ..." should file > in two places, because users could look for it in either way, and the two > forms would be separated in most catalogues. We didn't do this in card > catalogues or in book catalogues, because we wanted to keep the size of > the catalogue down, but this is no longer such an important consideration > in an OPAC. > > The logical corollary of that is that if the subtitle changes, then you > need at new access point under the new title-subtitle combination. > > Incidentally, in the listed examples that I gave, you would need an extra > access point under "Appalachia" for each of the two entries starting > "Appalachia Variations On An Old Slave Song", since the title proper > is just "Appalachia" (it's the musical work by Delius). > > Giles > > #### ## Giles Martin > ####### #### Quality Control Section > ################# University of Newcastle Libraries > #################### New South Wales, Australia > ###################* E-mail: ulgsm@dewey.newcastle.edu.au > ##### ## ### Phone: +61 49 215 828 (International) > Fax: +61 49 215 833 (International) > ## > The web of our life is of a mingled yarn, good and ill together > -- All's Well That Ends Well, IV.iii.98-99 > > On Thu, 7 Aug 1997, Ralph Papakhian wrote: > > > Once again we have examples from particular OPAC systems that > > do not even follow current filing rule standards. > > What if Giles' example followed some more systematic set of > > filing rules (say filing by title proper (subfiled by other > > title) > > then the example he provides would look like: > > > > > Appalachia > > > Appalachia : a journal ... > > > Appalachia : journal ... > > > Appalachia 1903 > > > Appalachia In The Sixties Decade Of Reawakening > > > Appalachia Preschool Test Of Cognitive Skills Apt > > > Appalachia Variations On An Old Slave Song Brigg Fair A > > > Appalachia Variations On An Old Slave Song With Final C > > > Appalachian Folk Songs For Recorder > > > Appalachian Geomorphology An Annotated Bibliography > > > Appalachian Ouachita Orogen In The United States > > > > that's not so bad? > -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Thu, 7 Aug 1997 00:49:05 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Filing (was Issues Related to Seriality) Comments: To: papakhi@INDIANA.EDU In-Reply-To: <9708070647.AA18532@browndwarf.ucs.indiana.edu> This statement by Giles was questioned: >"If users put in the subtitle as part of their search, then I think that they >would expect to find both "Appalachia : journal ..." and "Appalachia >journal ..." in the same place." I don't question it. Based on a catalogue use study, and based on where student card filers filed the cards before professionals revised the filing, that is *exactly* what a user would expect. John Rather at the Library of Congress in an internal memo proposed that filing rules be simplified to nothing before something (e.g., ground water before groundwater), filing as spelled (e.g., Mc and Mac would not interfile, umlauts would be ignored). We refiled the catalogue that way, and filing revision became *much* less work. Now we have abandoned filing to the computer, but this is pretty much what happens if ":$b" and ".$p" are ignored in 245. We also found that people expected numbers to file numerically before letters, rather than as spelled. So that "Appalachia 1903" would have filed before "Appalachia : a ...", rather than as shown by another poster, and so it does in every PC system we support. Perhaps filing should be considered part of catalogue construction, and be discussed along with code revision and MARC. All are interdependent. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Thu, 7 Aug 1997 10:23:46 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Filing (was Issues Related to Seriality) MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Mac writes: > > Perhaps filing should be considered part of catalogue construction, and > be discussed along with code revision and MARC. All are interdependent. > Perhaps indeed. Well, I'm really not saying German rules were so much better, but we've had this integration of filing rules into the code for a long time, and German catalogers are always mildly puzzled when learning that this is not so with AACR. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Thu, 7 Aug 1997 21:51:58 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Issues related to seriality MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=iso-8859-1 The paper by Hirons and Graham is very well presented and makes some sensible suggestions. However, these are wrapped up in a radicalism which no-one with an existing catalogue to worry about needs. What we do need are gradual adjustments to the existing rules. There are also one or two omissions in the paper. "Recommendation 1: Adopt a three-dimensional approach to the cataloging rules." If AACR is going to give some decent definitions re publication status, could we please have one for "published"? This causes more problems with monographs than serials but it's still a serious omission. "Recommendation 2: Incorporate the concept of the ongoing publication into the cataloging code." I don't think this is necessary: all we need to do is extend the definition of serial a little and add a new chapter on singlepart updating items. "Recommendation 3: Adopt Model B as a short term solution. Work towards the adoption of Model C as a longer term approach." Model B differs from the present definition of serials in including unnumbered as well as numbered items. Fair enough; but don't lets forget that many items AACR2 recognises as serials are already unnumbered: they have chronological designations. Model C is worrying because it dispenses with the notion of serials altogether. "If there were more of an emphasis on identifying ongoing publications rather than transcribing them down to the exact punctuation, we wouldnt end up with meaningless designations, such as: "1992-1993- 1994-1995," but could give "1992/1993-1994/1995." " AACR2 *doesn't* insist on exact transcription of punctuation and the latter example (which we follow) is already quite legitimate. "The placement of the bodys name on the chief source of the first issue determines whether it is recorded as a statement of responsibility or as the publisher. The placement often changes on subsequent issues, while in reality, the body is continuing to serve the same function." No. This is a CLA rule interpretation, nothing to do with AACR2. We follow a different interpretation: a body goes into the statement of responsibility if it is different from the publisher *or* the title is generic. This doesn't contradict AACR2. "Recommendation 4: Retain description based on the earliest issue for publications issued successively; for publications issued in a single updating part, base the description on the latest iteration." Yes, this is the only practical option. "Recommendation 5: For ongoing publications, replace the concept of chief source with that of source of title. Allow greater flexibility in the selection of title within the parameters of the prescribed sources. Define new terminology for sources within online publications." The phrase "greater flexibility" worries me: we look to AACR to answer our questions in a consistent way, not to turn them back on us as some kind of bibliographic psychoanalyst ("Well, where would you *like* to take the publisher from?"). "Recommendation 6: ... apply successive entry to changes of entry for successively-issued publications and a latest entry approach to changes of entry for updating publications." I was relieved to read this. It may cause problems but I think it's the only viable approach. An additional justification: cataloguing serials under titles they don't carry is a particular problem in host item citations in analytic records. The second part is quite reasonable too and doesn't require major surgery. "Might we not give more credence to the title of a journal (which is entered under title main entry) than to that of an annual report or conference publication (which is entered under corporate body)?" So our space-age serials cataloguing machine is to be drawn by a main- entry donkey? If we're going to be radical, surely that must be first concept into the bin? "Recommendation 7: Modify the rules for title changes to reduce the number of meaningless changes." Yes, but we must retain the link between what it says on the cover and the title proper in the record. One step forward would be to regard titles of the form "Post magazine & insurance week" as splitting (into title proper and other title information) *before* the & if layout and typography justify. But this is a situation where ISSNs always seem to change. "Recommendation 8: For ongoing publications, create rules that focus on identification rather than transcription. Consider new displays and arrangements of data that might result from and facilitate such a focus and that would keep similar information together in the record." On the face of it the second point has good possibilities, although it could get rather complicated: London : Smith & Brown, 1900-1937 ; Green & Yellow, 1938-1970 ; Birmingham : Pink & White, 1970- Also, spare a thought for those not using MARC, for which the above could create problems. "Recommendation 10: Explore new ways of arranging the rules that will facilitate a three-dimensional approach to cataloging ... " The present arrangement of AACR2 is not helpful: it is easy to miss things because of the split between ch 1 and ch 2-12. We mustn't allow too much theory to create an arrangement which is even more complicated. Would it be better to adopt a simple arrangement by area of description, with subrules where necessary for special materials? This would help iron out the existing discrepancies between different treatment of the same situation for different kinds of material. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Thu, 7 Aug 1997 17:03:56 -0400 Reply-To: Arlene Taylor Sender: "International Conference on the Principles and Future Dev. of AACR" From: Arlene Taylor Subject: Interdependence of AACR, MARC, filing rules Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Mac wrote: > Perhaps filing should be considered part of catalogue construction, and > be discussed along with code revision and MARC. All are interdependent. Charles Cutter's rules, at the last turn of the century, contained rules for entry, rules for description, rules for subject heading, rules for form heading, and rules for filing. These rules were not possible to maintain. They were broken into smaller pieces, probably for good reason. Mac did not suggest here that the rules be published together, only "discussed" together. Some previous posts, however, have suggested that MARC be included in AACR. This would have made sense a number of years ago, but now, with the advent of other encoding standards, I wonder if we would really want to tie the standard wording of the content of bibliographic records to the coding of those records. There are already places that have begun coding records with HTML rather than with MARC. In addition the Dublin Core may become a standard for cataloging internet resources. The "standard" will be the elements of the Core and the container in which it is housed. They have no plan now to write a standard for the wording of the *content* of the elements. That content could just as well be done according to AACR (along with other standards for subject headings, classification numbers, etc., which is also true with MARC). I think we should be very careful about including other standards with AACR. It would make things quite complicated because the other standards have overseeing bodies of their own, and because the other standards have lives and uses of their own. **************************************************** Arlene G. Taylor ** Associate Professor Department of Library and Information Science School of Information Sciences University of Pittsburgh ** Pittsburgh, PA 15260 e-mail: taylor@lis.pitt.edu ** voice: 412-624-9452 fax: 412-648-7001 ** http://www.pitt.edu/~agtaylor **************************************************** "What the caterpillar calls the end of the world, the master calls a butterfly." --Richard Bach ========================================================================= Date: Thu, 7 Aug 1997 22:46:56 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Re: Filing (was Issues Related to Seriality) -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of J. McRee Elrod Sent: Thursday, August 07, 1997 12:49 AM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Filing (was Issues Related to Seriality) John Rather at the Library of Congress in an internal memo proposed that filing rules be simplified to nothing before something (e.g., ground water before groundwater), filing as spelled (e.g., Mc and Mac would not interfile, umlauts would be ignored). We refiled the catalogue that way, and filing revision became *much* less work. Now we have abandoned filing to the computer, but this is pretty much what happens if ":$b" and ".$p" are ignored in 245. We also found that people expected numbers to file numerically before letters, rather than as spelled. So that "Appalachia 1903" would have filed before "Appalachia : a ...", rather than as shown by another poster, and so it does in every PC system we support. Perhaps filing should be considered part of catalogue construction, and be discussed along with code revision and MARC. All are interdependent. Mac [] I agree with this 100 percent. The computer filing that Mac describes is based entirely on the ASCII values of the respective characters, with space before numeral before letter. A universal approach to filing rules should be a >>mandated<< requirement in AACR revisions. As an example, I am thinking about the year 2000 problem, where filing of our CODOC organized government document collection is done with two-digit dates. I want to move to 4-digit dates, but how does one get -2000 to follow -99? One solution is replace the ASCII character "-" with a character that files higher than numbers or letters. A good solution would be: -99 followed by \2000 (a downward slash before the year). This new number files exactly where it should in the online catalog. People will still have to be trained to use the character, but at least this is based on near universal conventions (ASCII values), and it is far better than a mysteriously intermixed collection as two centuries slowly collide in the filing order. Perhaps still problematic is the automatic sorting of other numerical values in our index terms. For example, how does one get a good, consistent ordering of series numbering using the numbering from the series added entry? I would think we would have to rely on artificial intelligence and expert systems for computer programs to apply the myriad understandings humans have about ordering alphanumeric elements. Computers seem to find these complications (such as filing 100 before 20) extremely intractable, whereas filing clerks don't seem to have as much trouble. The same problem seems to exist with chronological subdivisions in subject headings. How about a thorough review of this whole concept of sorting and filing? Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Fri, 8 Aug 1997 10:47:30 +1000 Reply-To: Giles S Martin Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Fundamentals Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Can I first step back a bit from this discussion, and try to list a few of the propositions that I regard as axiomatic (but which others might think more dubious)? Part of the reason why I want to do this is my mathematical background, which makes me want to do things in a logical order. And part of the reaso is my feeling that there have been some disagreements about fundamental concepts in AACR. (1) AACR is about the *description* of the *bibliographic universe* so that *users* can find the *bibliographic items* they want. The description should describe as precisely as possible the entities and relationships in the bibliographic universe, so that users can identify bibliographic items quickly and accurately. (2) *Description* includes: (a) Transcription of *cataloguing information* about bibliographic items from the items themselves and from other sources. (b) Providing *access points* linked to the cataloguing information. (3) The *bibliographic universe* includes everything in library collections, everything available from publishers, and everything available on the Internet. (It may include other things as well, but this will do for the time being). (4) *Bibliographic items* include: (a) Physical items in library collections. (b) Computer files available on the Internet. (c) Works contained in or manifested in bibliographic items. (d) Sets (in the mathematical sense) of bibliographic items, such as *manifestations*, *editions*, *serials* and *series*. (Note that this definition is recursive. Note also that the sets in part (d) must make some sort of bibliographic sense -- e.g. the set of all books with red bindings in my library is not included, nor is the set of all books that are out on loan to me.) (5) *Manifestations* are the sets of all physical items which share a common title, publisher, format and content, and are therefore treated as being identical. (6) *Cataloguing information* includes: (a) The names (*titles*) of bibliographic items. (b) The names of people and organisations who are responsible for producing and distributing bibliographic items, which for the time being we will call *authors* (although of course in real life they are called lots of other things -- such as editors, performers, producers, publishers). (c) The physical description of bibliographic items. (d) The *access points* which are linked to the descriptions of bibliographic items. Cataloguing information does not include information about the topical content of bibliographic items, although that may often be deduced from the cataloguing information. (7) *Access points* are formally constructed names for *authors* and *works*, based on the names found in bibliographic items. Access points are constructed so that, as far as possible, there is a one-to-one relationship between access points and authors or works. (8) Although AACR does not specify it, access points are designed to be filed in a particular order (which may in practice differ according to the kind of library catalogue, partcularly with OPACs) (9) OPACs may include access points which are not specified in AACR but which are included in the cataloguing information prescribed by AACR, such as key words in contents and ISBNs. (10) *Users* approach the catalogue with names for *authors* and *works* based on how they have seen or heard them named, and perhaps with an imperfect memory of that. As far as practicable, they should be guided to the formally constructed *acccess points* by *references* from the names chosen by users. (There are a few terms above indicated by asterisks that need formal definition, but which I haven't formally defined. In addition, of course, the definitions given above need to be further refined -- but they can never, by the nature of things, reach the rigour of a formal definition in mathematics.) I think I've said enough in this message. There are some more points that I want to make arising out of the "Filing" debate that I will put into a further message. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 ========================================================================= Date: Fri, 8 Aug 1997 11:20:48 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Filing In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII In the message I just posted, one of my proposed axioms was: (10) *Users* approach the catalogue with names for *authors* and *works* based on how they have seen or heard them named, and perhaps with an imperfect memory of that. As far as practicable, they should be guided to the formally constructed *acccess points* by *references* from the names chosen by users. I want to go on from that point to look at the "Filing" debate. Ralph Papakhian, not unfairly, criticised me for putting up some "hypotheticals". He is right: it often worries me, too, that cataloguers makes decisions based on what we think users will look for, rather tha seeking some empirical evidence. The empirical evidence that we use is the information we find in the bibliographic universe -- names used on title pages, etc. -- and not what users come to the catalogue with. Of course, in a sense we have to depend on what we find in the bibliographic universe. We need to describe it accurately, as a service to our users, who often depend on an accurate description. Further, the words used in the bibliographic universe are much easier to identify that the words that catalogue users might bring to the catalogue -- partly because the relevant words will be brought to it in the future! However, with the invention of OPACs, it is now possible to obtain precise empirical evidence of what searches are being brought to the catalogue; and this may be relevant to what gets put into the catalogue. For example, over a four-month period last year, in this library, the most popular title searches (used more than 100 times) were: abnormal psychology 151 articulate mammal 106 australian industrial relations 120 australian law journal 117 australian psychologist 618 biochemistry 110 biology 165 bulletin 119 business review weekly 112 chemistry the central science 110 child development 192 companion to ethics 230 concepts and schemata 140 death of nature 139 essay writing for students 204 ethics 101 film art an introduction 164 first global revolution 198 for and against feminism 105 fundamentals of nursing 490 fundamentals of physical geography 147 in the underworld 135 journal of advanced nursing 207 lancet 127 learning to labour 115 microeconomics 196 much ado about nothing 147 nature 160 new sociology for australians 341 new woman new earth 121 norton anthology of poetry 108 philosophical ethics 135 process of learning 140 psychology at work 118 science 153 scientfic american 161 second sex 106 social psychology 123 social psychology a practical manual 286 social science and medicine 140 society and change 101 sociology 228 sociology a brief but critical introduction 101 sociology one 212 sociology themes and perspectives 381 studying law 108 survive exams 186 textbook of anatomy 101 turning point 110 Of course, most of these titles are the names of books prescribed for reading by undergraduate students: that's what you would expect in an academic library. A few will give you a lot of hits, so that the user might need some way of limiting the search to find the right book or journal -- such as "biology", "bulletin", "ethics", "nature", "science" and "sociology". So catalogues will need to have mechanisms here (which are not prescribed by AACR, except for additions to common uniform titles like "bulletin"). By my estimate, five of the list include what might be subtitles: chemistry the central science 110 film art an introduction 164 social psychology a practical manual 286 sociology a brief but critical introduction 101 sociology themes and perspectives 381 In fact, the relevant bibliographic records do have a $b subfield: Chemistry :$bthe central science /$cTheodore L. Brown, H. Eugene LeMay, Jr., Bruce E. Bursten Film art :$ban introduction /$cDavid Bordwell, Kristin Thompson Social psychology :$ba practical manual /$cedited by Glynis M. Breakwell, Hugh Foot and Robin Gilmour Sociology :$ba brief but critical introduction /$cAnthony Giddens Sociology :$bthemes and perspectives /$cMichael Haralambos ... [et al.] So with most of them, if you filed the title under just the "title proper", you would be a considerable distance from the title sought by the user. A title search on just "Sociology" gives 34 entries with the one word title, and a further 611 titles starting with "Sociology"! This means, that at least in these cases, the choice of an inappropriate filing order will give our users real problems -- not hypothetical problems. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 ========================================================================= Date: Thu, 7 Aug 1997 20:56:21 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Fundamentals & axioms Comments: To: ulgsm@dewey.newcastle.edu.au In-Reply-To: As the gourmand ungrammatically said to the waiter handing back the menu, "I see nothing here to object to". You made some sound points. Just some addenda. > (b) Providing *access points* linked to the cataloguing information. "Linked" surprised me. I am more accustomed to having this relationship called "justified". True, in an authorities environment the access points are linked to fields in the MARC record, but in other environments the access points themselves are embedded in the record, often in forms differing from the transcribed ones which justify them. I am more comfortable thinking in terms of an ISBD based record, plus access points, being incorporated into a catalogue, than thinking of a deconstructed record with information dispersed, although the latter will happen in some environments. It seems to me the code must be written in terms of a discrete record per manifestation. One poster's idea of integrating the rules by area (field) across formats, with exceptions for particular formats noted in each area, sounded good to me. >(6) *Cataloguing information* includes: > (a) The names (*titles*) of bibliographic items. > (b) The names of people and organizations who are responsible for >producing and distributing bibliographic items, which for the time being >we will call *authors* (although of course in real life they are called >lots of other things -- such as editors, performers, producers, publishers). I am a bit uncomfortable with this lumping together of 245$c and 260 (to speak in MARKese). Perhaps you mean to include place, publisher, date as physical description below, but for me physical description begins with collation. Also, there is no mention of notes or entries bringing out history and relationships (440, 580, 780, 785, etc.). > (c) The physical description of bibliographic items. >Cataloguing information does not include information about the >topical content of bibliographic items, although that may often be >deduced from the cataloguing information. True, subject analysis is another field, but there is overlap, e.g., autobiographies and annual reports. >(10) *Users* approach the catalogue with names for *authors* and >*works* Every catalogue use study I have done found that through at least the junior year of university, the most frequent search is subject. Only seniors, graduate students, faculty, and staff would search more for known items than for subjects. I would put "When searching for a known item" in front of (10) above. I also wonder if some of those "title" searches you quoted from your OPAC use record were not intended as subject searches. Your remarks on catalogue arrangement in your filing post are, I think, spot on. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 8 Aug 1997 14:55:44 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Fundamentals & axioms Comments: To: "J. McRee Elrod" In-Reply-To: <16m6zEJ3BI8Z092yn@slc.bc.ca> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Some quick responses to some of Mac's comments are below my signature block. #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 On Thu, 7 Aug 1997, J. McRee Elrod wrote: > > (b) Providing *access points* linked to the cataloguing information. > > "Linked" surprised me. I am more accustomed to having this relationship > called "justified". True, in an authorities environment the access > points are linked to fields in the MARC record, but in other > environments the access points themselves are embedded in the record, > often in forms differing from the transcribed ones which justify them. I used the word "linked" fairly deliberately, as I was trying to cover all forms of catalogues. In the card catalogue, the access point is linked to the description by being written above it (as a heading for filing purposes) or below it (as a tracing on a main entry). A book catalogue is similar. In an OPAC, the links are provided by a connection from an index to a bibliographic record, and by the inclusion of fields containing headings in the bibliographic record (which can take you back to the index). In fact, the technical details are often more complex than this. > >(10) *Users* approach the catalogue with names for *authors* and > >*works* > > Every catalogue use study I have done found that through at least > the junior year of university, the most frequent search is subject. > Only seniors, graduate students, faculty, and staff would search > more for known items than for subjects. I would put "When searching for > a known item" in front of (10) above. I also wonder if some of those > "title" searches you quoted from your OPAC use record were not intended > as subject searches. I think that in an academic library the majority of searches are for known items: undergraduates looking for set texts, postgraduates looking for books and articles that they have found cited somewhere, or that they have found in a journal index. The situation is probably different in public libraries and school libraries. However, I just said "authors and titles" because the universe I am talking about is the universe of AACR. While there are some links between AACR and the universe of subject cataloguing, there are some important differences. One is that that in the AACR universe entities are more well-defined, and are more precisely labeled. The concepts "William Shakespeare" or "Macbeth" are much better defined than "chemistry" or "sociology". Cataloguers will find much easier to agree on whether a work is by "William Shakespeare" or if a item is a manifestation of "Macbeth" than they will agree on whether a book is about "chemistry" or about "sociology". They will find it easier to agree on who is the author or what is the title than on what is the subject or what the name of the subject is. Giles ========================================================================= Date: Fri, 8 Aug 1997 01:06:50 +0530 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: John Attig Subject: Fundamentals and other things MIME-version: 1.0 Content-type: text/plain; charset="us-ascii" It being after midnight where I am, I am going to throw caution to the winds and enter this discussion, hoping that I will not be totally incoherent. I'd like to return to a different type of fundamental than the ones Giles has set forth [although I do appreciate his list and even agree with almost all of it -- at least until someone starts drawing implications]. I'd like to recall the purpose of the exercise we are engaged in: a discussion of the principles and future development of AACR2. To many of the comments I have seen recently, I have been tempted to ask: "But what do you expect AACR2 to do about it?" I think we might want to begin to direct our attention to the question of how the cataloguing rules can help (or hinder) the problems we are describing. As much of the discussion has touched heavily on the terrible things our online systems do to our lovely data, I think that it might be time to return to an issue that I raised some time ago: what is the cataloguing code attempting to codify? Should be look to AACR2 to codify the display of cataloguing information (such things as layout of records, arrangement of access points, etc.)? How far do we want to take this? We could decide that the organization of information prescribed in Part I of AACR2 is to be our standard of display and that (for instance) a labelled, as opposed to a paragraphed, display of cataloguing information does not comply with the rules. We could add a specification of filing order for access points to the code and declare that any display that insists on presenting lists of data arranged in the ASCII collating sequence are in violation of the rules. We could do that, but I don't think that the people who pay the bills at our institutions would be willing to pay for a system that fully complies with that rigorous a standard -- or that our users will be willing to wait around while a system sorts data and formats displays. Beyond this somewhat cynical point about costs and benefits, I think that any system designer would react negatively to such a specific display standard. The response would be (1) sure, we can do that if that's what you really want and you are willing to pay the costs, but (2) you should realize that you are stifling the creativity of some very talented people and turning your back on a lot of past and future lessons learned from our experience with human-machine-information interactions. Speaking in my own voice again, I think that system designers have added significant value to our information systems beyond what the rules currently specify and that their creativity can be stimulated, rather than stifled, by our better articulating goals and principles rather than rules and specifications. So, I guess I'm arguing against making the rules too specific about such things as displays. If not a display specification, what is the code? For one thing, it is an organized list of data elements, along with specific rules for recording appropriate information for each element. We have some trouble with the details of the rules for recording information, but we also have trouble with the concept of an organized list of elements: the fact that there is an explicit sequence of elements and that there are a wealth of implied relationships among the elements. There is also another role for the code, one that seems to me to be a weakness of AACR2. The current rules are strong on creating single bibliographic descriptions, but are not equally strong on creating a catalogue. The relationships between records are implicit in Part II of AACR2, but the approach there is still from the individual record (what entries to make for that item and then what form to use for that name or title). The tools for articulating relationships among records are principally entries (Ch.21) and references (Ch.26). These two methods can be in conflict (anyone want to talk about names of treaties?) and the rules do not resolve these conflicts. It is not even clear that these two methods exhaust the possibilities for articulating relationships among bibliographic entities. [Digression: The IFLA Functional Requirements for Bibliographic Records document has made a good start in *defining* what entities and relationships are important. One of the next steps is to study the ways in which these relationships should be made functional within an information system.] Going back to my original point, I think that most of the topics proposed for the Conference and the papers that have been posted are addressing this type of issue. Most deal with defining or re-defining fundamental bibliographic entities or with relationships among entities. I think we need to bridge the gap between such general issues and their specific implications and applications. One specific example: I completely agree with Giles that our users cannot be depended upon to agree with our decisions about where the title proper of a publications ends and where the other title information begins. So what can the rules do? The rule for title added entries (21.30J) calls for making an added entry for the title proper and permits making additional title added entries under "any version of the title." Should the rules be more prescriptive and require added entries under title proper plus subtitle in all or some circumstances? Does anybody doubt that "any version of the title" permits such entries? If not, then shouldn't that point be made explicit in the rule? Beyond that, should the rules be more explicit about how title access (or any other kind of access) is provided? Is it time for the rules to recognize that a significant component of access in most information systems is through keywords rather than complete headings? Should the rules include specifications for what elements should be so indexed and how the results should be presented? In sum: What is the appropriate way for the rules to address Giles' problem? Or is this a problem best solved outside the rules? Apologies for a somewhat rambling message. I look forward to the continuing discussion. John Attig Penn State University Libraries University Park, PA 16802 jca@psulias.psu.edu ========================================================================= Date: Fri, 8 Aug 1997 08:37:51 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Issues related to seriality MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT At the end of his recent posting, Robert Cunnew wrote: > Would it be better to adopt a simple arrangement by area of description, > with subrules where necessary for special materials? This would help > iron out the existing discrepancies between different treatment of the > same situation for different kinds of material. > This should indeed be given serious consideration, also with regard to reasoning put forth in other postings toward integration of code and format. It would also eliminate a lot of tedious duplication and almost-duplication in chapters 2 to 12, and the frequent necessity of flipping back and forth between these and chapter 1. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Fri, 8 Aug 1997 11:10:17 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: AACR + filing rules MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT > Mac wrote: > > > Perhaps filing should be considered part of catalogue construction, and > > be discussed along with code revision and MARC. All are interdependent. > and Arlene Taylor replied: > Charles Cutter's rules, at the last turn of the century, contained rules > for entry, rules for description, rules for subject heading, rules for > form heading, and rules for filing. These rules were not possible to > maintain. They were broken into smaller pieces, probably for good reason. > What was that reason? Mac's suggestion is, I think, a very important one and not to be dismissed because there may probably have been good reasons at some point in the distant (pre-MARC) past. Not only is the tight integration of a cataloging code with filing rules a sound and logical proposition, but also, if one cares to look beyond the AACR world, large-scale examples can be found where this approach WAS found possible to maintain and where nobody ever found a reason to split filing rules from the code. Many colleagues in this country really hate the idea that Germany should go on indefinitely using a code not compatible with AACR, but up until now there is still much reluctance here to adopt the latter. Probably for good reason. P.S. To my knowledge, the real reason was just a practical one, not a good one: American libraries had those extra employees called "filers", so that true and real "catalogers" didn't have to waste their time filing their cards themselves. I'm sure this is largely a thing of the past. Somehow this division of labor never caught on here, and hence no division of filing rules from cataloging rules. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Fri, 8 Aug 1997 07:45:35 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Topics outside AACR such as subject cataloguing Comments: To: ulgsm@dewey.newcastle.edu.au In-Reply-To: Giles said: >I think that in an academic library the majority of searches are for >known items: undergraduates looking for set texts, postgraduates looking >for books and articles that they have found cited somewhere, or that they >have found in a journal index. The situation is probably different in >public libraries and school libraries. Perhaps this is a Canadian/Australian difference. When at the circulation desk we asked "how did you find this?" about the item being checked out, the majority answered "by subject" through the first three years of their academic career. (The "set texts" were purchased at the book store.) They were looking for materials on subjects assigned for term papers. In training new cataloguers, the absence of generally accepted guidelines for assigning subject headings was a major handicap. Whether AACR should contain such guidelines is another question. Perhaps AACR can't be everything (integrated with MARC, include catalogue construction guidelines, include subject cataloguing guidelines), but we do have clearer guidance in descriptive cataloguing than in subject cataloguing or catalogue construction. We seem to be engaged in improving what is already the best. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ > >However, I just said "authors and titles" because the universe I am >talking about is the universe of AACR. While there are some links >between AACR and the universe of subject cataloguing, there are some >important differences. One is that that in the AACR universe entities >are more well-defined, and are more precisely labeled. The concepts >"William Shakespeare" or "Macbeth" are much better defined than >"chemistry" or "sociology". Cataloguers will find much easier to agree >on whether a work is by "William Shakespeare" or if a item is a >manifestation of "Macbeth" than they will agree on whether a book is >about "chemistry" or about "sociology". They will find it easier to >agree on who is the author or what is the title than on what is the >subject or what the name of the subject is. > >Giles > ========================================================================= Date: Fri, 8 Aug 1997 11:36:25 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: AACRCONF Coordinator Organization: NLC-BNC Subject: Papers received from Sherry L. Vellucci and Rahmatollah Fattahi MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit Please note that the following conference papers are now available on the JSC Web site: Bibliographic Relationships by Sherry L. Vellucci. AACR2 and Catalogue Production Technology: The Relevance of Cataloguing Principles to the Online Environment by Rahmatollah Fattahi. ========================================================================= Date: Fri, 8 Aug 1997 12:25:57 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: William Gerlach Subject: AACR and the display of the bibliographic record Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" First, I would like to say that it is wonderful that this list was set up to allow discussion of the papers that have been made available. I find that I get so busy with the HOW of cataloging and getting the materials (be they here physically or on the Internet) available/accessible to the user that I rarely take the time to think about the WHY, but this list has given me a reason to do so. This message is partly in response to John Attig's post, which caused me to re-read Rahmatollah Fattahi's paper in which he states that "AACR does not provide ANY guidelines for display" (emphasis mine). While I agree with this point, the divisions of the rules for description into 8 areas at least IMPLIES that certain elements of the description are related to each other and as such should (or maybe not?) be displayed together. It may be argued that these areas simply reflect the card-based bias of AACR, since these areas are given in the order in which they appear in the traditional catalog card format. While this may be true, there are varying levels of relationships between the elements in these areas which can provide a jumping-off point for guidelines for display. The most obvious relationship which occurs to me is between the title portions of the series and the numbering within the series in the series area. To my mind these elements must necessarily display in proximity to one another, since the number alone has little or no meaning without the title. Another area which it makes sense to me to display in proximity to one another, but for the reason of separating it from other elements, is the physical description area. My reasoning is that the elements in this area are perhaps the only PURELY descriptive elements in the bibliographic record: I can think of no reason to use them as access points, although I could see that they could be used as limiting elements (such as someone who wants books with accompanying slides). While some bibliographic records (including those that are MARC-based) may have such information coded elsewhere for limiting, not all users of the rules will. On the other hand, an area in which the elements have little relationship to one another in the note area. The discussion is complicated by the fact that those of us who use MARC are now using fields outside of the 5XX to generate notes and access points from the same field. While it may be an accident of MARC that variations in title (B4 of the note area) and other title information (B5) are displayed in many OPACs next to the title and statement or responsibility area, rather than next to other notes, I would argue that this may be a more reasonable placement of such notes/access points. Similarly, notes relating to edition and history (B7) as well as the other notes which correspond directly to one of the first 6 areas may make more sense in proximity to those areas. There are certainly arguments which can be made one way or the other for the other areas as well, but any future changes to AACR which deal with display should take the relationship between elements in one area and between elements in two different areas into account. Bill Gerlach Cataloguer/NDNACO Funnel Project Administrator NDSU Libraries ========================================================================= Date: Fri, 8 Aug 1997 12:29:40 CDT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: David Prochazka Subject: Re: Filing MIME-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII In our library, we've grappled with the issue of filing by complete title vs. title proper ever since we went online in 1985; some of our manual files (e.g. our media booking system) are filed by title, and since we collect in a narrow scope (health sciences), different works with the same title-proper are commonplace. For online catalogs, rather than choose only title proper or complete title as the title access point, why not index both? That way, we won't have to second-guess or spend lots of time on empirical studies to determine user preferences; we might also spend less time determining where we think users would think the title proper ended. I see many chief sources that are (intentionally?) ambiguous as to title proper vs. other title information. I suspect most of us do. For the work "Medicine PreTest self-assessment and review" which graphically sets off the word "Medicine" on the chief source, we chose "Medicine" as the title proper; that's fine, but grammatically, "Medicine PreTest self-assessment and review" is also a very logical title to look for. I'm suggesting that this record have title indexing under "Medicine" and under "Medicine : PreTest self-assessment and review," perhaps not so much for the sake of where it lands in alphabetic arrays, but for the sake of the user who could easily type in "Medicine PreTest self-assessment and review" as a title search and not realize that we own it, because we only indexed it under the title "Medicine." I can't imagine that this should be a problem from the programming end, and it would certainly increase the potential for retrieval. --------------------------------------------------> David Prochazka, Cataloging Librarian Library of Rush University 600 S. Paulina--Rm. 581//Chicago, IL 60612-3874 phone: 312-942-2731//fax: 312-942-3143 e-mail: davidp@lib.rush.edu --------------------------------------------------> ========================================================================= Date: Fri, 8 Aug 1997 11:33:10 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Scott Piepenburg Subject: Re: AACR and the display of the bibliographic record Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" As a trainer, I work with many librarians around the United States (and a few in Canada) and am constantly hit with the question of "which automated system more accurately displays the record from AACR." This then takes us on the discussion that AACR is a means to select and format the information, MARC is a means for storing the information, and it us up to the designer of the automated system to determine how that information is being displayed. Many librarians equate AACR with display. What we need to do is look at the function of each. USMARC is NOT a cataloging code, although it seems as if much of what we do as catalogers is determined by how it will display. As a former system designer and current librarian and trainer, I've ALWAYS encouraged librarians to catalog according to AACR, code and organize according to the appropriate MARC format, and create displays and indexes by having the automation system organize and extract from the MARC record what is important. Someone earlier accurately stated that the code should NOT be linked to a particular format or system. If the information is coded correctly, and then formatted correctly, automation systems can be designed to extract what each library deems important in the cataloging record. That gives each library the flexibility it needs for its users. - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - Scott Piepenburg District Cataloger/System Administrator Dallas Public Schools spiepenb@dallas.isd.tenet.edu http://www.dallas.isd.tenet.edu "I will not be pushed, filed, stamped, indexed, briefed, debriefed or numbered. My life is my own."-- No. 6 ========================================================================= Date: Fri, 8 Aug 1997 14:12:19 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Filing In-Reply-To: from "Giles S Martin" at Aug 8, 97 11:20:48 am Content-Type: text Giles S Martin said .... Greetings, The evidence of these searches could also suggest that they were cataloged incorrectly. That is, based on the way people are searching for the titles, it appears that what has been included as other title information really ought to be considered as part of the title proper (the chief name of an item). 286 searches on "social psychology a practical manual" may be a pretty good indication of what the chief name is? No one has ever claimed (I think) that catalogers' decisions are always the best, nor that descriptive cataloging shouldn't be revised as necessary. --ralph papakhian. > > By my estimate, five of the list include what might be subtitles: > > chemistry the central science 110 > film art an introduction 164 > social psychology a practical manual 286 > sociology a brief but critical introduction 101 > sociology themes and perspectives 381 > > In fact, the relevant bibliographic records do have a $b subfield: > > Chemistry :$bthe central science /$cTheodore L. Brown, H. Eugene > LeMay, Jr., Bruce E. Bursten > Film art :$ban introduction /$cDavid Bordwell, Kristin Thompson > Social psychology :$ba practical manual /$cedited by Glynis M. > Breakwell, Hugh Foot and Robin Gilmour > Sociology :$ba brief but critical introduction /$cAnthony Giddens > Sociology :$bthemes and perspectives /$cMichael Haralambos ... > [et al.] > > So with most of them, if you filed the title under just the "title > proper", you would be a considerable distance from the title sought by > the user. A title search on just "Sociology" gives 34 entries with the > one word title, and a further 611 titles starting with "Sociology"! > > This means, that at least in these cases, the choice of an inappropriate > filing order will give our users real problems -- not hypothetical > problems. > > Giles > > > #### ## Giles Martin > ####### #### Quality Control Section > ################# University of Newcastle Libraries > #################### New South Wales, Australia > ###################* E-mail: ulgsm@dewey.newcastle.edu.au > ##### ## ### Phone: +61 49 215 828 (International) > Fax: +61 49 215 833 (International) > ## > The web of our life is of a mingled yarn, good and ill together > -- All's Well That Ends Well, IV.iii.98-99 > -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Fri, 8 Aug 1997 15:19:48 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Adam Schiff Subject: Re: Topics outside AACR such as subject cataloguing Comments: To: "J. McRee Elrod" In-Reply-To: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Isn't the Subject Cataloging Manual a generally accepted guideline for creating and assigning subject headings? ************************************** * Adam L. Schiff * * Principal Cataloger * * University of Washington Libraries * * Box 352900 * * Seattle, WA 98195-2900 * * (206) 543-8409 * * (206) 685-8782 fax * * aschiff@u.washington.edu * ************************************** On Fri, 8 Aug 1997, J. McRee Elrod wrote: > > In training new cataloguers, the absence of generally accepted > guidelines for assigning subject headings was a major handicap. Whether > AACR should contain such guidelines is another question. Perhaps AACR > can't be everything (integrated with MARC, include catalogue > construction guidelines, include subject cataloguing guidelines), but we > do have clearer guidance in descriptive cataloguing than in subject > cataloguing or catalogue construction. We seem to be engaged in > improving what is already the best. > > Mac > > __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) > {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ > ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 8 Aug 1997 17:33:54 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Topics outside AACR such as subject cataloguing Comments: To: aschiff@u.washington.edu In-Reply-To: Isn't the Subject Cataloging Manual a generally accepted guideline for >creating and assigning subject headings? It is the practice of one library. It is too complex for most smaller libraries. It does nothing to help classed catalogues. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 8 Aug 1997 17:06:08 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Limiting subtitles (was Filing) Comments: To: davidpro@LIB.RUSH.EDU In-Reply-To: >For online catalogs, rather than choose only title proper or >complete title as the title access point, why not index both? All the systems we use and support have a truncating feature, where automatically (or by a character such as "?"), you get titles which match your title search key, plus titles which begin that way. In terms of screen display, and COM and book catalogue arrangement, all fifty of my customers are united in wanting arrangement to observe subtitle. The one exception is when information they don't consider title information is included as other title information, such as when national cataloguing agencies include the date and place of a symposia or conference in the subtitle, or the statement of scope of a serial. My clients want these things in a note, and don't see them as other title information. Perhaps we should restrict subtitle to what can be actually considered part of the title. This would both simplify filing, and reduce problems with changing statements found on serials defining their subject or audience. I agree with my customers that these statements, like where and when a symposia or conference was held, belong in notes. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ That >way, we won't have to second-guess or spend lots of time on >empirical studies to determine user preferences; we might also >spend less time determining where we think users would think the >title proper ended. >I see many chief sources that are (intentionally?) ambiguous as to >title proper vs. other title information. I suspect most of us do. >For the work "Medicine PreTest self-assessment and review" which >graphically sets off the word "Medicine" on the chief source, we >chose "Medicine" as the title proper; that's fine, but grammatically, >"Medicine PreTest self-assessment and review" is also a very >logical title to look for. >I'm suggesting that this record have title indexing under "Medicine" >and under "Medicine : PreTest self-assessment and review," >perhaps not so much for the sake of where it lands in alphabetic >arrays, but for the sake of the user who could easily type in >"Medicine PreTest self-assessment and review" as a title search >and not realize that we own it, because we only indexed it under the >title "Medicine." I can't imagine that this should be a problem from >the programming end, and it would certainly increase the potential >for retrieval. >--------------------------------------------------> > David Prochazka, Cataloging Librarian > Library of Rush University > 600 S. Paulina--Rm. 581//Chicago, IL 60612-3874 > phone: 312-942-2731//fax: 312-942-3143 > e-mail: davidp@lib.rush.edu >--------------------------------------------------> > ========================================================================= Date: Fri, 8 Aug 1997 17:21:17 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: AACR and the display of the bibliographic record Comments: To: spiepenb@FALCON.DALLAS.ISD.TENET.EDU In-Reply-To: <199708081633.AA11912@falcon.dallas.isd.tenet.edu> >Someone earlier accurately stated that the code should NOT be linked to a >particular format or system. If the information is coded correctly, and >then formatted correctly, automation systems can be designed to extract what >each library deems important in the cataloging record. That gives each >library the flexibility it needs for its users. The unit card developed over generations. Why we think a non librarian systems person can do better, I don't know. Just because we *can* do something (deconstruct and label parts of the MARC record) does not mean it is a good idea. For me, the best standard for display is the ISBD. It takes the old unit card information and arrangement, and introduces punctuation which makes clear what each part is. I've never seen an OPAC display which is so clearly and easily understood as the unadorned ISBD. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sat, 9 Aug 1997 16:06:02 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: AACR2 and technology AACR2 and catalogue production technology by Rahmatollah Fattahi (33 p.) posted to the Future Development of AACR papers site is, I think, the most helpful and hopeful so far. Since I feel the question should be how to get systems developers to make better and more coherent use of the excellent form for bibliographic records which has evolved over generations, I was suspicious of a paper which "aimed at re-examining the principles of AACR2 in light of the different characteristics of the catalogue production technology" (p.24). I need not have been concerned. While Fattahi makes some superb suggestions for expanding the scope of AACR (standards for levels of display for example), RF gives new underpinnings for the validity of such concepts as main entry (used in the sense of person responsible as opposed to record so entered in contrast to Hagler and the AACR2 glossary), corporate entry, and uniform title. Fattahi points out new functions of the catalogue (in additional to the traditional collocation and identification) made possible by anyword searching and Boolean searching, and the increase in remote usage. RF suggests the possible use of hypertext links (e.g, 100 author to author's webpage, 260$b publisher to publisher's webpage). Like other papers RF makes the distinction between works and manifestations. But rather than work authority records, RF speaks of "super records" for works. While it is not explicitly stated, I would assume a super record would only be created where more than one manifestation existed, which makes much more sense to me than having each manifestation or item backed up by an authority record, in the way each name entry is backed by an authority record. I can see a super record uniting the manifestation records for serials which have changed titles, standard works which have gone through multiple editions, works which have been replicated in microform or talking book, sound recordings which have been issued on cassette and disc, films which have been issued on film reels, video cassette, and video disk, etc. RF's paper is peppered with practical suggestions, for example, that there should be more direct and less subordinate entry for corporate bodies, in view of the limitations of one line brief display. (Linkages could provide the collocation function for subbodies related to the larger body.) He mentions the redundancy of such headings as "Canada. Canadian High Commission". To address the problem of mislabeling of fields (e.g., non authors labeled as authors), RF suggests that AACR "prescribe the terms to be used in labels" (p.23). In an increasingly multilingual world, it would seem better to me to eschew labels, and depend upon the integrity of the ISBD bibliographic description, at least in the third and forth proposed levels of display. Also, since relationship (author, editor, translator, illustrator, Festschrift honoree, defendant, composer, etc.) is not contained in the entry (1XX,7XX), this does not seem programmable to me. If you haven't time to read all the papers, I would suggest reading at least this one, and strongly supporting Fattahi's conclusions and recommendations. RF describes very well (or delineates as Vellucci would say) the reality of OPAC use of bibliographic records, and makes practical, down to earth, suggestions for amelioration. It is difficult to believe this was written by a professor rather than a practicing cataloguer :-{)}. __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Sun, 10 Aug 1997 16:54:25 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: AACR2 and technology In-Reply-To: MIME-Version: 1.0 In article , "J. McRee Elrod" writes >AACR2 and catalogue production technology by Rahmatollah Fattahi (33 p.) >posted to the Future Development of AACR papers site is, I think, the >most helpful and hopeful so far. ... >It is difficult >to believe this was written by a professor rather than a practicing >cataloguer :-{)}. I agree that some good points are made here, but regarding your last comment above I regret I cannot agree! Only a professor could take so long to make a few sensible points. For me the important areas were: MAIN ENTRY I accept that for many libraries this is a useful concept but some of the arguments put forward in favour (and where are the arguments against?) don't stand up. Main entry isn't necessary to collocate different manifestations of a work. A uniform title may be, but not main entry. Clearly AACR needs to retain a chapter on the assignment of main entry points, but given that many libraries have no use for the concept this chapter needs to be optional, in the same way that the chapter on uniform titles is optional. This means ensuring that no other rules *assume* main entry, which creates difficulties for those not using the concept. For example: 1. If the body changes under which a serial receives main entry, a new record for the subsequent issues is required. 2. Added entries for series are to be made under the appropriate heading, ie the main entry. 3. The rules for added entries cannot be used on their own, ie a library assigning author access points which are main entry neutral needs to grub around in two sections for guidance. A single unified chapter on access points is required, backed up by an optional chapter on main entry which refers back to the main chapter. UNIFORM TITLES The problem here is that uniform titles *and* transcribed titles seem to be indexed by most OPACs as "title", so a catalogue user will not know which he has found and may therefore miss the required collocation of manifestations. The only way round this seems to be to include within a record a link that can be followed up by the catalogue user to find other manifestations (including serial continuations). Couldn't this function be served equally well by a work number? The link would enable you to search for other records with the same number. NAME HEADINGS I applaud Fattahi's suggestion to introduce more direct headings for corporate bodies. I have never been able to satisfactorily explain to library assistants headings like France. Direction des Assurances I don't think it matters if the country name appears in the name or not. We need a single rule for corporate bodies that hinges on whether or not the name is dependent on the name of another body, ie no special rules for government bodies. If necessary qualifiers can be added: Direction des Assurances (France) The point is that "France" is not regarded as a corporate body. For that matter, neither are most conference names. Perhaps those rules also require amendment, but I'm not sure what could be done there. DISPLAY Yes, recommendations would be useful here. That's all they could ever be, of course - recommendations. Level 1: surely this should give title and statement of responsibility, not main entry heading and title. This would make the recommendations main entry neutral (desirable) and in any case the statement of responsibility usually gives more information. Qualifiers added to main entry headings? That sounds a backward step. Level 2: to give main entry plus statement of responsibility is hardly appropriate at this level, given the space required. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Sun, 10 Aug 1997 13:14:59 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Elisabeth D Spanhoff Subject: Re: AACR2 and technology On Sun, 10 Aug 1997 16:54:25 +0100 Robert Cunnew writes: MAIN ENTRY >I accept that for many libraries this is a useful concept but some of >the arguments put forward in favour (and where are the arguments >against?) don't stand up. Main entry isn't necessary to collocate >different manifestations of a work. A uniform title may be, but not >main entry. The best argument against main entry I've seen lately is Martha Yee's own conference paper. How many of her difficulties do you suppose would remain if she were not bound by the requirements of main entry? Would there even be a need to agonize over the definition of 'work?' >UNIFORM TITLES >The problem here is that uniform titles *and* transcribed titles seem >to be indexed by most OPACs as "title", so a catalogue user will not know >which he has found and may therefore miss the required collocation of >manifestations. The only way round this seems to be to include within a >record a link that can be followed up by the catalogue user to find >other manifestations (including serial continuations). Couldn't this >function be served equally well by a work number? The link would enable >you to search for other records with the same number. - In addition to the work number (a wonderful idea) what about some field in the record to record the kind of relationship the given manifestation bears to the original work; so that displays such as "Adaptation of Shakespeare's Hamlet," "Translation of Shakespeare's Hamlet" , or strings to that effect, can be automatically generated? Elisabeth Spanhoff espanhoff@juno.com ========================================================================= Date: Sun, 10 Aug 1997 12:53:49 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: AACR2 and technology Comments: To: robert@CUNNEW.DEMON.CO.UK In-Reply-To: >MAIN ENTRY > >I accept that for many libraries this is a useful concept but some of >the arguments put forward in favour (and where are the arguments >against?) don't stand up. Main entry isn't necessary to collocate >different manifestations of a work. A uniform title may be, but not >main entry. Main entry + uniform title makes the best collocation of the manifestations of a work; main entry makes the best collation of the works of an author (more important in literature than elsewhere I will admit.) Even if you were willing to depend upon an added entry to collocate the works of an author, how in the world would you create unique uniform titles for novels like Slate, Canary, Cobalt, and Vermilion (all novels by Nathan Dyne); not to mention Chemistry, Physics, etc., etc. I can't think of anything other than author a catalogue user is likely to know. I don't want to see any more manufactured entries like main entries for treaties! I hope these don't make it past the next rule reivision. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 11 Aug 1997 09:53:30 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: AACR2 and technology In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I don't understand the objection to countries as corporate bodies. The French Republic (which is given the formal name "France" can be responsible for publication in its own right, just as other governmental and non governmental bodies can. In our catalogue, we give an author entry to "France" for: (1) its Constitution; (2) its treaties; (3) its laws; (4) a legal case where the French Republic was defendant ("Nuclear tests case: Australia v. France") I don't think that it would be appropriate to give the entry to a particular department, minister or agency of the French Republic in any of these cases. Further, all except the laws have analogous works for non-governmental organisations, which may have a fundamental document called a "charter", "constitution", "memorandum and articles of association" or "rules and by-laws", which enter into agreements with other organisations, and which take part in legal cases. Indeed, even individual human beings can enter into agreements, or be parties to law cases, in just the same was as the French Republic can and does. (Incidentally, I just checked the LCNA record for "France" (n79-6404). In my opinion, it ought to have 410 field for "French Republic" and "Rpublique franaise", since those are the official names of the government in French and English). Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 49 215 828 (International) Fax: +61 49 215 833 (International) ## The web of our life is of a mingled yarn, good and ill together -- All's Well That Ends Well, IV.iii.98-99 On Sun, 10 Aug 1997, Robert Cunnew wrote: > NAME HEADINGS > > I applaud Fattahi's suggestion to introduce more direct headings for > corporate bodies. I have never been able to satisfactorily explain to > library assistants headings like > > France. Direction des Assurances > > I don't think it matters if the country name appears in the name or not. > We need a single rule for corporate bodies that hinges on whether or > not the name is dependent on the name of another body, ie no special > rules for government bodies. If necessary qualifiers can be added: > > Direction des Assurances (France) > > The point is that "France" is not regarded as a corporate body. For > that matter, neither are most conference names. Perhaps those rules > also require amendment, but I'm not sure what could be done there. ========================================================================= Date: Mon, 11 Aug 1997 13:25:07 +0100 Reply-To: mh@bodley.ox.ac.uk Sender: "International Conference on the Principles and Future Dev. of AACR" From: Mike Heaney Subject: Time is of the essence In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII Functional Requirements of Bibliographic Records largely ignores the time aspect,but I suggest there is much to be gained from an analysis, at each level (Work, Expression, Manifestation, Item) of: (a) how objects/entities of library interest exist over time; (b) how or whether they change over time (c) how or whether their existence is reflected in some sort of physical reality, tangible or not (d) whether the physical reality is continuous or intermittent I believe that this would give us a better insight into what are users are really looking for, and what will satisfy their needs, when they conduct bibliographic searches. Cataloguing revolves around objects/entities which can have a physical existence over time. We can record that existence continuously (and often do at the Item-specific level, and for serials), or we can take what are essentially snapshots of the state of an object at a given moment in time (for example the editions of a Work). For some objects (e.g. performances) the actual time-span of their existence is brief. The extent to which this is important depends on how that existence impinges on the real world. If a performance is recorded, we can view the performance on different occasions and act as if the original performance still exists. For books, the "performance" is the metal or lithographic plates (or computer file) used to produce the Items: the plates/files may be destroyed/deleted, but the Items remain. When entities (at whatever level, from Work down to Item) change, they can do so by augmentation (adding more to what is there) or metamorphosis (where something new replaces what was there). The issues relating to time become more prominent for serials and for online resources, but they affect every aspect of cataloguing. We have survived by fudging the issue while book Mainfestations were the focus of cataloguing, because their production technology allowed us to ignore these considerations; but this has not endeared us to, for example, our colleagues cataloguing manuscripts or rare books with provenance histories. I shan't take up more bandwidth/diskspace on the list with this, but I've written in more detail (if somewhat ramblingly -- this is by no means a fully thought through position) and put it on the Web at http://www.bodley.ox.ac.uk/users/mh/time978a.htm Mike Heaney Bodleian Library michael.heaney@bodley.ox.ac.uk ========================================================================= Date: Mon, 11 Aug 1997 08:53:14 EST Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Neil R. Hughes" Organization: University of Georgia Libraries Subject: Re: Definition of work We may well be able to do without the idea of main entry in the bibliographic record (and of course as a result, in our online catalogues that have made this feasible), but we cannot do without a definition of "work" because of the extent to which the idea of authorship and principal responsibility for intellectual content is bound up in our users' approaches to the catalogue. I am quite willing to enter a videorecording of Verdi's _Aida_ under title, but I am not willing to forego an analytical added entry for: Verdi, Giuseppe, $d 1813-1901. $t Aida. ... because there is nothing that librarians can or should do about the fact that this is the approach that the majority of users will take to this work or any manifestation of it that we might arbitrarily decide is a "different work." No matter how significant the contributions of the video's producer & director, no matter how much effort was put into the performance by the cast, orchestra and conductor, nor how much money was given to support the production by the NEA or other agency, to our users, this is still Verdi's Aida ... completely present in one of its countless manifestation in the container we have in hand. It simply ignores what music librarians know is the primary approach by the typical user to the catalogue to say that the video is "a new work" and that Verdi's opera--which actually cannot be reduced to _any_ of its physical manifestations (score, vocal score, chorus score, audio recording, video recording, etc)--is merely "another work" that gave rise to or inspired the "new work" we have in hand. It's certainly no waste of time to theorize about such things, for as responsible professionals we should certainly examine the definition of a work as it relates to our catalogues (and Martha Yee's paper certainly does an outstanding job of analyzing and describing the theoretical problem we face as designers of a cataloguing code and of bibliographic files). But it is to me all vanity and the most irresponsible sort of political correctness to try to overturn 1,000 years of cultural conditioning of our users and force them to see that they are "wrong" to approach this item as if it too were principally Verdi's creation. Our job is to provide access, and to remove the authority record for Verdi ... $t Aida from our catalogue structures and strip it from our bibliographic files would certainly provide me with lots of ammunition in the argument for shutting down libraries and cutting their funding. I contend, and I challenge anyone to disprove it, that nothing about having an entry on the bibliographic record for the videocassette for Aida under Verdi's name hinders access by those very few people who think that a video of Verdi's _Aida_ isn't as much "his" as a copy of a vocal score printed in the 1920s. To leave off such an entry, however, would hinder access by many (all those who wish to see all manifestations of Verdi's _Aida_ in the catalogue, to name but a few). But in order to justify such an entry on the bibliographic record, we need an authority record for it, and to have an authority record, we must acknowledge that _Aida_ is a work and the person most responsible, above all others, for its creation (BECAUSE OF HOW OUR USERS APPROACH IT!) is Verdi--not the librettist, the camera operators, the cinematographer, or any other person or body responsible for a given manifestation of it. This requires at least a clear conceptualization, if not a definition of work, whether or not it requires the perpetuation of the concept of main entry. I believe it would create chaos in our authority files to assert that each new capture of a new performance of a musical or musico-dramatic work is itself a new work requiring an authority record. Think of the qualifiers one would need! Imagine trying to sort out which version of Aida (for example) began life as a film and was then transferred to videotape vs. those which began life as video productions (this is _not_ always evident on the items!). We must take great care that in our zeal to make the code uniform for all types of materials that we do not obscure access for large groups of users and supporters of our libraries. As a music cataloguer, I am perfectly willing to have different rules for "work entry" (if we may use such a term in place of main entry on the momentary assumption that main entry will go away) for different styles and types of music, Western and non-Western, as long as I am not forced into dropping the all-important name/uniform title access that is required by users of Western classical music just to be shown that I'm wrong (or if I'm not, that I should be) about what the principal means of access to the aforementioned video is by the majority of its library users. Unfortunately much of the debate surrounding main entry and the definition of a work has been reduced to a popularity contest among contributors to the item in hand and stiff-lipped assertions of "They're all equal under God's eyes!," instead of a careful analysis of how we arrived where we are today and whether or not a lot of that isn't absolutely on the money where access is concerned. The cultural conditioning that has made librarians assume that music is more important than words when ascertaining authorship of a song cycle has worked even more profoundly upon our users than it has upon us, because few of them have ever even paused to consider that it might be otherwise. Among those music librarians who are willing to eschew main entry, many (if I may dare to speak for more than myself) would be willing to have two, co-equal analytical added entries on the bib. record for a song cycle: an author/title entry for the musical work under the composer's name, and another under the poet's name for the text(s)--and if I may, I would remind everyone that these two titles may be utterly different, depending on the language variant that might have been chosen by the composer and many other factors. There must simply be the recognition that the musical work is a new (or separate, in the case of poetry that was composed expressly for the song cycle) work and must have its own, concomitant authority records associated. When the users stop looking for different manifestations of Verdi's Aida and start looking only for known items (each of which is arguably a "different work" just because it has different users' fingerprints on it), then I'll stop harping on this point. > The best argument against main entry I've seen lately is Martha Yee's own > conference paper. How many of her difficulties do you suppose would > remain if she were not bound by the requirements of main entry? Would > there even be a need to agonize over the definition of 'work?' <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> Neil R. Hughes Music Cataloger University of Georgia Libraries Internet: nhughes@libris.libs.uga.edu Telephone: (706) 542-1554 Fax: (706) 542-4144 ========================================================================= Date: Mon, 11 Aug 1997 15:00:35 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: franzmeier Organization: Staatsbibliothek zu Berlin / IIE Subject: Uniform titles for serials Mime-Version: 1.0 Content-Type: text/plain; charset=iso-8859-1 Content-Transfer-Encoding: 8bit In her posting of 1 August on this subject Jean Hirons wrote: "We need to be able to identify the serial from an 'entry' for purposes of citations, online search displays, links, and added entries without having to look at the entire bibliographic record." Apart from deploring the existence of six different ways to create such a citation under present AACR/LCRI conditions, Jean's main concern seems to be that in the case of uniform titles with their emphasis on distinction rather than on identification this purpose is not served well in many instances. And she goes even further and questions the usefulness of the whole concept of uniform titles for serials. As to the many ways to create a title short enough for citation purposes and long enough to identify the serial, at least one of the solutions used at present under AACR/LCRI conditions could be dispensed with, in my opinion. It is the solution described in LCRI 25.5B(6) to qualify a generic title (proper) by adding the responsible corporate body in the form it has in the respective name authority record. The resulting uniform title is nothing else than the traditional author/title-type of citation with the sequence of the constituent elements simply reversed (and parentheses around the corporate body instead of a colon between it and the title proper, okay). But it conveys exactly the same information. So why not use the traditional author/title sequence instead which has at least two advantages: It is extremely familiar, is the way in which works by personal authors are cited (and certain publications of corporate bodies as well); and it is at the same time easily derivable automatically from the existing (stored) data without the necessity to input data into an extra field (MARC 130). I know of course that this question is intrinsically related to the question of "main entry". Without going into the details of this longtime controversy, I would like to point to the fact that there are situations, that there are titles (proper) where nobody has ever denied that they need to be cited together with the responsible corporate body, titles like the famous "Bulletin". Now, fortunately, there are no more than two choices in which form the issuing body can be brought into contact with the title (proper) in such cases: In the form of the name in which it appears on the publication; or in its name authority form, i. e. formalized, normalized. If the latter is used, why let it precede the title in one case, and make it follow the title in another, as the present practice wants it? In my opinion there is absolutely no necessity to proceed in such a divided manner. The "normal" position of a normalized name is that of a heading, i. e. it precedes the descriptive parts of a catalogue entry, which begin with the title proper. On the other hand, persons and corporate bodies in their descriptive form (as on the piece) normally only follow the title proper. This is their proper place, certainly not only in my opinion, and not only in ISBD-oriented descriptions: it is the solution adopted for ISSN key titles, as we know (with a clear emphasis on identification, by the way). Therefore, If the whole problem of main entry for serials were (re)considered, just for a change of perspective, not so much from the (content-oriented) angle whether the work e. g. records "the collective thought of the body" (21.1B2), but from the more formal angle whether the corporate body is indispensible to name, to cite, to identify the given serial publication, the solution could be quite easy. Easy in so far as the decision process is inevitable anyway, and has to happen also when the issuing body is to be added in parentheses under current LCRI practice. There is nothing new about it, nothing additional, nothing complicated, on the contrary: the suggested procedure would allow for a consistent treatment of all cases where a heading, a constructed form of name is used together with a title: and it would save us at least one unnessesary (hybrid) variant for the construction of citable titles. One important difference between the solutions described above should also be remembered in this context: Using the normalized name authority form will always imply that national,at least not really international forms of names come into the play. This is legitimate, of course, is practical and may even be inevitable under certain circumstances. The more, in my opinion, normalized headings should be kept separate, should not mingle with the descriptive elements, should appear in a position where everbody expects normalized elements, i. e. before or on top of the title proper, but not following it; and even less between the Main title and the Section title, where they have to be put in many cases according to the present LCRI. Günter Franzmeier German Union Catalogue of Serials Berlin, Germany ========================================================================= Date: Mon, 11 Aug 1997 11:23:36 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "James E. Agenbroad" Subject: AACR and Subject Headings MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Monday, August 11, 1997 While AACR is not used to create topical subject headings, at least in the US isn't it widely used to create subject headings for persons, geographic names, corporate bodies and uniform titles that are the subject of works? If this use of AACR is valid but insufficiently acknowledged (I didn't find it in the index) perhaps it should be stated near the beginning of chapters 22-25. Regards, Jim Agenbroad ( jage@LOC.gov ) The above are purely personal opinions, not necessarily the official views of any government or any agency of any. ========================================================================= Date: Mon, 11 Aug 1997 11:43:21 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Uniform titles for serials Comments: To: franzmeier@SBB.SPK-BERLIN.DE In-Reply-To: <33EF6152.44D5@sbb.spk-berlin.de> >So why not use the traditional author/title sequence >instead which has at least two advantages: It is extremely familiar, is >the way in which works by personal authors are cited (and certain >publications of corporate bodies as well); and it is at the same time >easily derivable automatically from the existing (stored) data without >the necessity to input data into an extra field (MARC 130). The difficulty I would have with this is that there would be no clear and consistent way to determine main entry for a serial. A title which is distinctive and unique at time of cataloguing may cease the to be so at any time. Would you have the first remain under title, while the second identical title is under corporate body? Responsibility for serials is so mixed, that I am quite happy with the shift to title main entry, with the exception of those which are *about* the issuing body such as annual reports. But the contrast between main entry under title and *no* entry under title is too much I think. We title trace annual reports, using " - " as opposed to "/$c" in 245 if the title and body are not not grammatically linked. e.g., "of the"; in which case we do a 246 with the " - ". We find it a handy search key when the hits on the corporte body exceed the maximum our system will report. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 11 Aug 1997 23:25:47 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: Uniform titles for serials In-Reply-To: <33EF6152.44D5@sbb.spk-berlin.de> MIME-Version: 1.0 In article <33EF6152.44D5@sbb.spk-berlin.de>, franzmeier writes >As to the many ways to create a title short enough for citation purposes >and long enough to identify the serial, at least one of the solutions >used at present under AACR/LCRI conditions could be dispensed with, in >my opinion. It is the solution described in LCRI 25.5B(6) to qualify a >generic title (proper) by adding the responsible corporate body in the >form it has in the respective name authority record. The resulting >uniform title is nothing else than the traditional author/title-type of >citation with the sequence of the constituent elements simply reversed >(and parentheses around the corporate body instead of a colon between it >and the title proper, okay). ... > On the other hand, persons and corporate bodies in their >descriptive form (as on the piece) normally only follow the title >proper. This is their proper place, certainly not only in my opinion, >and not only in ISBD-oriented descriptions: it is the solution adopted >for ISSN key titles, as we know (with a clear emphasis on >identification, by the way). I share your distaste for putting what are effectively name headings *after* titles. However, we have adopted the opposite solution from the one you recommend. For brief citation - check-in, box labels, issue, labels, "in" analytics - we use a compromise between AACR and key-titles (which many of our serials don't have and which we don't necessarily know anyway): * Title proper if distinctive and unique, otherwise: * Title proper plus statement of responsibility if present, otherwise: * Title proper plus qualifier in square brackets We take the qualifier from the descriptive part of the record, following the preferences established for key-titles. We also omit initial articles from the title proper and the statement of responsibility (as per key-titles), with the exception of "in" analytics, where we keep them in as as per AACR. The advantage of this approach is that all serials are cited by title and that qualifying text is more likely to be found on the actual item. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Tue, 12 Aug 1997 03:25:49 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Enrique Gildemeister Subject: Re: Uniform titles for serials So much of the confusion and lack of agreement about serial uniform titles comes from the fact that they were not provided for in AACR2. It took a lot of rule interpretations from LC to make "AACR2-style cataloging" for serials. Rex Bross and Mitch Turitz wrote interesting articles for _The Serials Librarian_ special issue v.22, no.1/2 on serials cataloging, outlining the major problems in dealing with unique seial identifiers. I think we still need uniform titles to create unique titles that can accurately identify a serial in all the linking fields and other areas in which there must be consistency and clarity. Serials are constantly evolving; they are fluid. I hope that the Joint Steering Committee will consider making uniform titles a part of the future canon of AACR and provide for them in the rules themselves. The biggest bone of contention has been the choice of qualifier for the uniform title. In an older version of LCRI25.5B, place was given preference over corporate body as qualifier. The main problem with the corporate body qualifier is that the RI's have ruled that if the body changes, a new entry has to be created, as if the title proper had changed. The new RI leaves choice of qualifier to the discretion of the cataloger, but as things stand now, a change in corporate body still necessitates a "title change". Jean Hirons said in a recent posting to SERIALST that the catalogers in the Serial Record Division at LC like the idea of a non-varying uniform title which has cross-references in the 246 fields. I suggested something similar to that myself on SERIALST, and I think it goes a long way to resolving a lot of the problems of using uniform titles for serials. Another problem is the fact that serial uniform titles (except series) are not under authority control. To my knowledge, AACR2 says nothing about authority records at all and just asks that references be made. Coming back to the model of 130/246, the serial bibliographic record itself then behaves like an authority record. Even if JSC fleshes out a workable policy, I predict that there will still be controversy and change, but at least the debate will be on the rules themselves and not just on LC policy. Rick Gildemeister Serials Cataloger and Head of Cataloging Lehman College of the City University of New York ========================================================================= Date: Tue, 12 Aug 1997 10:32:58 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Uniform titles for serials MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Rick Gildemeister points out: > So much of the confusion and lack of agreement about serial uniform titles > comes from the fact that they were not provided for in AACR2... and > > I think we still need uniform titles to create unique titles that can > accurately identify a serial in all the linking fields and other areas in > which there must be consistency and clarity. but then he comes round to: > Another problem is the fact that serial uniform titles (except series) are > not under authority control. But this is not strictly true. For what reason had the ISDS been set up, one is beginning to wonder, and what are their Key Titles for? If ISDS does what AACR2 fails to do, then why not use those titles wherever possible? They are even printed out as such in many serials. Not in many of those problematic cases, right, but what else is wrong with Key Titles that AACR2 don't even make added entries (access points) for them? After all, a key title is "The unique name assigned to a serial by the International Serials Data System" (AACR2 glossary). Just what we need - or not? But why not? B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Tue, 12 Aug 1997 09:07:01 -0400 Reply-To: Regina Reynolds Sender: "International Conference on the Principles and Future Dev. of AACR" From: Regina Reynolds Subject: Key titles & uniform title MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Thank you, Bernhard Eversberg for making a point that I have been meaning to write about. Indeed, there is no reason at all to have two fields in serial records which are serving exactly the same function: unique identification of a serial title. I have been long trying to have the key title substituted for the uniform title in LC cataloging but the uniform title seems quite entrenched. Besides the elimination of redundancy in the record, the key title offers the advantages of being an international standard and of being assigned by agency in the country of publication, To answer the obvious question about what to do if a key title is not available, I have a simple answer: cataloging agencies should construct a key title according to the rules set forth in the _ISDS Guidelines_. These rules could be incorporated into cataloging codes. In CONSER records the presence or absence of an authentication code (nsdp or iscs/c) in field 042 shows whether the key title has been assigned or confirmed by an ISSN center or not. ISSN centers would be encouraged to accept key titles constructed by cataloging agencies as long as they did not conflict with another title in the ISSN database. I absolutely agree that this could work. Why not? If key titles were to become the accepted means of identifying serial titles in catalog records, I can see progress on another source of conflict and expenditure of cataloging time: arguing about title changes in shared databases. The ISSN and key title are part of a very successful international standard. When the ISDS Network (formerly ISDS, the International Serials Data System) began many ISSN centers operated independently of the cataloging departments in the national libraries where most ISSN centers are located. Over the years, more and more centers are combining their ISSN operations with their cataloging operations. Also over the years the criteria for determining title changes under AACR and under ISSN rules have come closer and closer together. An area for great potential would be working towards complete compatibility between ISSN rules for title changes and those in AACR and other cataloging codes. Then the full potential of using the key title as the unique title identifier could come into play. Then the possibility of using the ISSN record as the *authority record* for serial titles could be developed. There is much that can be gained from making use of the ISSN database of over 800,000 records, authoritative records with the work of unique title identification already completed. I look forward to sharing further thoughts on this topic. Regina R. Reynolds email: rrey@loc.gov Head, National Serials Data Program voice: (202) 707-6379 Library of Congress fax (202) 707-6333 101 Independence Avenue, S.E. ISSN Web page: lcweb.loc.gov/issn/ Washington, D.C. 20540-4160 ========================================================================= Date: Tue, 12 Aug 1997 10:02:32 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "E. Spanhoff" Subject: Re: Definition of work Getting rid of main entry does not mean we now start entering everything under title or that we must forego access points, such as your example for the video of Verdi's Aida. By all means, let's keep the authority record for Mr. Verdi's work and continue to accommodate the bibliographic relationships the library community has found useful to distinguish over the years. What concept of 'work' is required to support these relationships? Patrick Wilson answered that question as early as 1989 (no new concept) in his splendid paper, "Interpreting the Second Objective of the Catalog," Library Quarterly 59:4, 339-353 (see p. 349). For clarity, soundness, and good sense, his discussion of 'work' has never IMHO been surpassed.. I invite you to reread it. E. de Rijk Spanhoff State Library of Louisiana On Mon, 11 Aug 1997 08:53:14 EST "Neil R. Hughes" writes: >We may well be able to do without the idea of main entry in the >bibliographic record (and of course as a result, in our online >catalogues that have made this feasible), but we cannot do without a >definition of "work" because of the extent to which the idea of >authorship and principal responsibility for intellectual content is >bound up in our users' approaches to the catalogue. I am quite >willing to enter a videorecording of Verdi's _Aida_ under title, but >I am not willing to forego an analytical added entry for: > >Verdi, Giuseppe, $d 1813-1901. $t Aida. ========================================================================= Date: Tue, 12 Aug 1997 11:48:49 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Regina Reynolds Subject: ISDS Guidelines MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I have been asked to further identify the _ISDS Guidelines_. I apoligize for citing a work that may not be familiar to all and furthermore for citing an earlier title. (This is really an inexcusable error for any serialist, let alone head of an ISSN center. Mea culpa!) The correct title is _ISDS Manual_ published in 1983 in Paris by the International Serials Data System, now the ISSN Network. The Manual is currently undergoing revision but the basic rules for assigning key titles will remain the same. Provisions for assigning key titles to electronic works are being added and other provisions are being expanded and clarified. The rules for creating key titles and those for creating uniform titles of the distinguishing kind are quite similar. In the majority of cases handled by the U.S. ISSN center, the key title and the uniform title end up exactly the same or are adjusted to be the same so as to reduce the possibility of future problems if one of the titles were to change and not the other. A significant difference is that the key title is supposed to reflect the title at the time of ISSN registration while the uniform title is supposed to be taken from the earliest issue. In practice, this distinction is not always followed. However, recent discussions on this list have pointed out the desirability of providing current identifying information for serials, which is an ISSN practice. I believe that harmonizing the two sets of rules is very possible and would result in mutual benefit to both the ISSN and AACR communities. Regina R. Reynolds email: rrey@loc.gov Head, National Serials Data Program voice: (202) 707-6379 Library of Congress fax (202) 707-6333 101 Independence Avenue, S.E. ISSN Web page: lcweb.loc.gov/issn/ Washington, D.C. 20540-4160 ========================================================================= Date: Tue, 12 Aug 1997 08:28:41 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Collocating works (was Uniform titles for serials) Comments: To: Riquili@AOL.COM In-Reply-To: <970812032548_806083260@emout13.mail.aol.com> The erudite Rick said: >Jean Hirons said in a recent posting to SERIALST that the catalogers in the >Serial Record Division at LC like the idea of a non-varying uniform title >which has cross-references in the 246 fields. I suggested something similar >to that myself on SERIALST, and I think it goes a long way to resolving a lot >of the problems of using uniform titles for serials. Since most of the second indicators in 246 are taken, I would prefer a reactivation of 247 for earlier titles and a new 24X for later titles. A set of 247 and 24X second indicators could save a lot of 246$i or 580 keying. Whether these should be limited to small changes in title, or replace all successive entry, could be a matter of considerable debate. Repeating 780/785 for all former and later titles could also get us away from the one missing link in the chain and multiple search problem created by successive entry. But in the absence of the super record solution, I would prefer the 24X one. Many of the problems the super record is proposed to solve (successive entry for serials, the same work in multiple formats, successive editions) could be solved by other means (247/24X, repeating 300s, sticking with the original author, respectively). __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Tue, 12 Aug 1997 08:28:41 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Collocating works (was Uniform titles for serials) Comments: To: Riquili@AOL.COM In-Reply-To: <970812032548_806083260@emout13.mail.aol.com> The erudite Rick said: >Jean Hirons said in a recent posting to SERIALST that the catalogers in the >Serial Record Division at LC like the idea of a non-varying uniform title >which has cross-references in the 246 fields. I suggested something similar >to that myself on SERIALST, and I think it goes a long way to resolving a lot >of the problems of using uniform titles for serials. Why a non-varying uniform title for serials? Isn't this what the 222 Key title is supposed to be? Wouldn't a non-varying 022 ISSN help as well? In terms of other forms of the titles, since most of the second indicators in 246 are taken, I would prefer a reactivation of 247 for earlier titles and a new 24X for later titles (unless latest title is always in 245). A set of 247 and 24X second indicators could save a lot of 246$i or 580 keying. Whether these should be limited to small changes in title, or replace all successive entry, could be a matter of considerable debate. Repeating 780/785 for all former and later titles could also get us away from the one missing link in the chain and multiple search problem created by successive entry. But in the absence of the super record solution, I would prefer the 247/24X one. Many of the problems the super record is proposed to solve (successive entry for serials, the same work in multiple formats, successive editions) could be solved by other means (222/247/24X, repeating 300s, sticking with the original author, respectively). __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Tue, 12 Aug 1997 11:36:36 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Definition of work In-Reply-To: <19970812.100233.16486.0.espanhoff@juno.com> from "E. Spanhoff" at Aug 12, 97 10:02:32 am Content-Type: text Will the conference be avoiding post-modernist theories of the "work?" --ralph p. -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Tue, 12 Aug 1997 21:17:04 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: Key titles & uniform title In-Reply-To: MIME-Version: 1.0 In article , Regina Reynolds writes >Thank you, Bernhard Eversberg for making a point that I have been meaning >to write about. Indeed, there is no reason at all to have two fields in >serial records which are serving exactly the same function: unique >identification of a serial title. I have been long trying to have the key >title substituted for the uniform title in LC cataloging but the uniform >title seems quite entrenched. > This is an attractive suggestion but I can see problems that would need to be addressed: 1. I have never seen a key-title printed in a serial. 2. ISSNs *are* printed in serials but they're not always the current ones, so making it harder to pick up changes in key-titles. 3. There seems to be no easy way to find out key-titles. They don't seem to be available on the Web. 4. Key-titles don't fit into AACR and this can be confusing. Another thing I don't like about the ISDS manual is (I believe) shared with the LC rules for uniform titles: when a title is common to more than one serial the first occurrence of that title registered doesn't carry a qualifier. This is messy and doesn't help identification of the first registered title. Is it beyond our present automated systems to retrospectively add a qualifier? -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Wed, 13 Aug 1997 10:22:00 +1000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Key titles & uniform title In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Both AACR and the MARC format find it difficult to link bibliographic records because of their heritage from the catalogue card era. In a card catalogue, to link to another bibliographic record, you need to specify the exact heading under which the user will find the record in the catalogue. In an computerised catalogue, other means are available of linking records, which need not be visible to either the user or the cataloguer. For example, pages in the WWW are linked by URLs, which need not be typed in by the user (except perhaps for the first entry point). You move from page to page by pointing and clicking at references. Just the same could happen in an OPAC. In such an environment, the cataloguer would need to tell the computerised system what links to create, in the case of new titles. (For old titles, there would be a default assumption that the existing unique identifiers, such as key titles, ISSNs and uniform titles, were accurately specified). The cataloguer would point at two different bibliographic records, and tell the system something like "This is an earlier/later title relationship), and there could be a system generated link between the two. For the user looking at one title, the minimum they need be told is that links exist, e.g. "This journal has now changed its title." They could follow up that link if they wished. The same can be done for monographic relationships, e.g. "This work has been translated into other languages" or "This work has been peformed on film or video". But indicating these links through uniform titles is a remnant of the card catalogue. In fact, even book catalogues often didn't need to use uniform titles for this purpose, since relationships could be show by records being printed next to each other on the same page. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 2 4921 5828 (International) Fax: +61 2 4921 5833 (International) ## [Note that the telephone numbers have just changed. From inside Australia, the area code is now (02).] ========================================================================= Date: Wed, 13 Aug 1997 08:06:00 +0200 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Knut Hegna Subject: Re: ISDS Guidelines In-Reply-To: Your message of "Tue, 12 Aug 97 11:48:49 EDT." Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii The ISSN Network has a www-site at: http://www.issn.org/ -- Knut Hegna - knut@dtv.dk DTV, Udviklingslaboratoriet, Box 777, 2800 Lyngby, Danmark Tlf + 45 45 25 73 51 / Faks + 45 45 88 89 84 ========================================================================= Date: Wed, 13 Aug 1997 08:37:07 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Key titles & uniform title MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT In reply to the suggestion to consider the ISDS Key Title as a unique identifier for a serial, Robert Cunnew points to a few problems: > This is an attractive suggestion but I can see problems that would need > to be addressed: > > 1. I have never seen a key-title printed in a serial. > > 2. ISSNs *are* printed in serials but they're not always the current > ones, so making it harder to pick up changes in key-titles. > > 3. There seems to be no easy way to find out key-titles. They don't > seem to be available on the Web. > > 4. Key-titles don't fit into AACR and this can be confusing. > Of course these problems need to be addressed. The ISDS Key Title is, however, the closest thing currently in existence to a unique identifier, and at 800.000 records the available database is not exactly small. Thus, NOT to build on this but instead start all over constructing unique identifiers where presently there are none would be the closest thing to re-inventing the wheel. It is to be hoped that this matter gets high priority at the meeting, and from what Regina Reynolds writes one can be confident that the ISDN center (located inside LC!) will be competent to handle what problems remain to be addressed. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Wed, 13 Aug 1997 09:23:06 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Kevin M. Randall" Subject: Re: Key titles & uniform title Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" At 09:07 AM 8/12/97 -0400, Regina Reynolds wrote: >An area for great potential would be working towards complete >compatibility between ISSN rules for title changes and those in AACR and >other cataloging codes. Then the full potential of using the key title as >the unique title identifier could come into play. Then the possibility of >using the ISSN record as the *authority record* for serial titles could be >developed. There is much that can be gained from making use of the ISSN >database of over 800,000 records, authoritative records with the work of >unique title identification already completed. I have always been a firm believer in corporate body main entry for serials, and believe it would eliminate much of the need for uniform titles. It is encouraging to see that at least the concept of main entry is still being advocated in some of these papers, whether or not its use will be expanded. Perhaps a workable rule would be something to the effect of: "If the title of the publication is nondistinct (such as "Journal"), or if the name of the corporate body is in the title, or if *in the cataloger's judgment* the corporate body has a relationship with the publication such that it would be unlikely the publication would exist apart from that body, enter the title under the name of the corporate body." Yes, that last "or" clause leaves lots of room for interpretation, but the rule would be infinitely easier to work with than the current unwieldy rule and its even more unwieldy LCRI. However, that would not completely eliminate the need for uniform titles. And after reading Regina's excellent argument for use of key title, I am leaning in favor of that approach, at least for the cases where corporate body main entry is not an option, if not for all cases. (I would still argue for extended use of corporate body main entry.) The idea of the ISSN record as the authority record is intriguing, and would make a lot of sense, IF these records were readily available! I have wondered for a very long time WHY the ISSN records are not in an international database such as OCLC. They seem to be available only by purchase on CD-ROM or tape from ISSN. Such an important resource should be just as easily available as the CONSER records are. Until that happens, I would want to hold off on making key title the "authoritative" form of the title in AACR. Kevin M. Randall Head, Serials Cataloging Section Northwestern University Library Evanston, IL 60208-2300 email: kmr@nwu.edu phone: (847) 491-2939 fax: (847) 491-7637 ========================================================================= Date: Wed, 13 Aug 1997 20:51:13 +0100 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Cunnew Subject: Re: Key titles & uniform title In-Reply-To: <2.2.16.19970813092307.198f9c5e@hecky.acns.nwu.edu> MIME-Version: 1.0 In article <2.2.16.19970813092307.198f9c5e@hecky.acns.nwu.edu>, "Kevin M. Randall" writes >Perhaps a workable rule would be something to the effect of: "If the title >of the publication is nondistinct (such as "Journal"), or if the name of the >corporate body is in the title, or if *in the cataloger's judgment* the >corporate body has a relationship with the publication such that it would be >unlikely the publication would exist apart from that body, enter the title >under the name of the corporate body." That sounds rather familiar - AACR 1967, chapter 6 if I recall rightly? Do we really want to go back to that? I remember trying to explain to library assistants why some of the check-in cards in our Roneodex index were under title, some were under corporate body and some were under the name of the country. In our online check-in system everything is under title. There may be several screens of "Annual report" but it's easy to find the one you want by following the qualifiers and the principle is easy to grasp. -- Robert Cunnew Librarian, Chartered Insurance Institute, London ========================================================================= Date: Wed, 13 Aug 1997 16:56:19 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "James E. Agenbroad" Subject: AACR and Subject Headings (2) MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Wednesday, August 13, 1997 After sending my earlier message on this topic I came across the kind of statement I had in mind in the 1949 "ALA cataloging rules for author and title entries" on page 82 in the first rule for personal authors: "The form adopted for a given person is used without variation whenever it occurs as a heading, whether as author, added entry, or subject." I did not find any similar statement in the first rule for corporate bodies (page 126). Regards, Jim Agenbroad ( jage@LOC.gov ) The above are purely personal opinions, not necessarily the official views of any government or any agency of any. ========================================================================= Date: Thu, 14 Aug 1997 08:22:47 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Key titles & uniform title MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Kevin M. Randall writes > >Perhaps a workable rule would be something to the effect of: "If the title >of the publication is nondistinct (such as "Journal"), or if the name of the >corporate body is in the title, or if *in the cataloger's judgment* the >corporate body has a relationship with the publication such that it would be >unlikely the publication would exist apart from that body, enter the title >under the name of the corporate body." > BUT this would dictate American Chemical Society: Journal and Association of Computing Machinery: Communications instead of Journal of the American Chemical Society and Communications of the ACM That's how they are always cited and looked for. And these are the most important and most-cited periodicals in chemistry and computer science. Does anybody go looking for them under the corporte names, I wonder? B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Thu, 14 Aug 1997 10:20:17 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Linking. Part 1 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT To link or not to link, and how ------------------------------- Part 1: Concepts. Authority linking In this posting, the first of two, I try to clarify some concepts of record linking. Not everybody reading this list will need that. I don't mean to lecture, but in the course of my activities I have come across too many misunderstandings and misconceptions to assume that everybody talking about linking, or listening and asking no questions, has a firm grasp of the subject. If you do, just skip this - many statements will appear banal to you. Or correct me where I'm wrong. The second part won't come before next week. It will deal with links between bibliographic records, and more specifically, records for the parts of series or multipart publications. Which, in fact, is my primary motivation for all this. It is the biggest remaining stumbling block for improving exchange between Germany and the AACR/MARC world. What is a link? --------------- Every cataloger establishes links between records many times a day, but not all are aware of this fact. If you look up "Tchaikovsky, Peter Ilich (1840-1893)" in the name authorities and faithfully transcribe or copy this into a 100, 600 or 700, you establish a link to the authority record. Likewise, if you look up "Afro-American musicians" in the Red Books and copy it into a 650, this constitutes a logical link from your bib record to the subject authority record - whether you have it in your local system or not. Programmers call this type of links "textual links" because it is a text string that connects both records. As soon as you change one character in the text string, you break the link - as far as the computer is concerned, who (unlike humans) has great difficulty or finds it impossible to match strings that are not 100% identical. If the change is made in a bib record, it is just this one record that is affected and loses its logical connection. If the change is made in the 1XX field of the authority record, ALL bib records linked to it get disconnected at once. Sophisticated software should prevent the cataloger from doing either and thus protect the data from disruption. But is a "true" link not something better than this, something fundamentally different from a text string identical to another text string? Devices like a "next generation hypertext link", or "system generated links", sound like they must be of an altogether different character! But no, they are not. On the data level, all links are basically alike. There is no such thing as an immaterial, direct, ethereal, intrinsic, magic rubber band connection between any two records. (Don't laugh, some people seem to have mental images like these.) At the root of it there is always a character string (not always displayed by the software) in the one record that matches a character string in the other record, nothing else. IDnrs are character strings as well, and so are URLs. It all depends on the software whether or not this link does anything, whether it is purely logical and abstract or has any function, and whether and how well it is protected against change (meaning disruption). And what of exchange? CAN links be exchanged? Once a record goes on exchange, all it consists of is characters or bytes. In order to exchange a link, one needs to exchange both ends of the link. Nothing more, but maybe less: if the receiving system already has one of the two records, only the new one has to be exchanged (downloaded). But beware: the other record may be a new, better version, and the one locally owned may have its own local additions. However: The receiving software must be able to "see" there is a link and then use its own ways and means to put it in effect - or there will be no link in the recipient's local system. Assuming for the moment that all systems can handle links, the important distinction to be made is between DATA links and INFORMAL links, like in contents notes or other notes fields. The latter unfold their beauty only in the eye of the (human) beholder, reading something like "Previous ed. under the title 'Twilight of the demigods'", and the reader will have to go ahead and search for that title, the software cannot present this sort of link as a hypertext button to click on, for example. This can be done only for DATA links, consisting of controlled character strings as described. The use of INFORMAL links is largely a passive one, from the POV of the catalog software: human readers have to realize it and take action to follow it. This certainly holds for everything in contents notes, despite keyword indexing. DATA links can be categorized into TEXTUAL and IDENTIFIER links: ** USMARC data currently use TEXTUAL links exclusively; Names, subject headings, uniform titles are all recorded as text strings in the bib records. One has to take great care to transcribe those entry forms faithfully, or use cut-and-paste facilities. Advantages: ** The links are immediately humanly readable; ** There is no trouble with exchange: bib records are usable even when the authority records are absent. Disadvantages: ** These links are easily broken (catalogers have to be instructed not to make even minor "corrections", or the software must prevent it) ** esp., the 1XX in the authority records must never be changed unless all occurrences in the bib records are changed as well. On the large (national) scale, changes in established headings are therefore extremely problematic and costly. ** IDENTIFIER links consist of IDnrs instead of the text strings (though IDnrs are text strings too). One might have something like 100 10 $3n98765 instead of 100 10 $aTchaikovsky, Peter Ilich,$d1840-1893 assuming n98765 is the IDnr of Peter Ilich's authority record. Advantages: ** Changes in established headings are not problematic, e.g., Tchaikovsky's form of name may be changed in the authority record's 100 without having to change so many bib records as long as the IDnr remains unchanged (there's rarely a need to change an IDnr). ** Savings in storage space (though nobody cares anymore). Disadvantages: ** Link is not humanly readable, software has to be in place to insert the name for the number in all displays and indexes. That means: ** Exchange is more problematic: authority record has to be present or bib record is not usable (not correctly indexable or displayable). A problem with authority records, for both kinds of links, is this: Changes made later in the central authority file don't automatically get transferred into all local systems affected. Therefore, the number of outdated headings and references, and lacking references, will steadily increase on the local level. This calls, at least, for periodic revisions of local authority files. Links don't take care of themselves! Software has to be provided to see after them. Whereas presently there are no linkage IDnrs in USMARC exchange records, it should be easy to enhance these records with an extra subfield $3 (or whatever) in every authority controlled field, carrying the IDnr. Then, local systems could either use or discard these number subfields, or retain them for later use if they cannot utilize them right away. Since the $a would be redundant (and would require maintenance!) a local system would ideally not have the $a but only the $3 with the number. Technically, this is perfectly possible, but as of now, certainly not all local systems operate in this way. If they do, they have to generate their own $3 (or whatever) using their own procedures since USMARC exchange data don't provide it. In the German format MAB2, as in UNIMARC ($3, incidentally), there already is provision for IDnrs of related records, authority or bibliographic. It is all a matter of definition (and then implementation into software) what links can be accomodated in what fields. It is not a built-in property of any field that it can or cannot support a link. Theoretically, there can be many links in one field (e.g., several subfields $3 might occur in a note field). It is also possible that one particular field has a $3 in certain cases and a $a in others. As long as USMARC says nothing about this, everything is possible in local implementations. They can differ widely in their provision and functionality of linkage while utilizing exactly the same data. Up until now, I have been referring largely to name and subject authorities as candidates for links with bib records. The second posting will deal with links between bib records, and more specifically, links between parts of series or multipart publications. Part 2 will begin with the concept of "bidirectional" links, and whether or not they are useful or necessary to have. Bernhard Eversberg Universitaetsbibliothek Braunschweig Germany B.Eversberg@tu-bs.de Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Thu, 14 Aug 1997 09:03:17 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Kevin M. Randall" Subject: Re: Key titles & uniform title Comments: To: B.Eversberg@tu-bs.de Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" At 08:22 AM 8/14/97 METDST, Bernhard Eversberg wrote: >Kevin M. Randall writes >> >>Perhaps a workable rule would be something to the effect of: "If the title >>of the publication is nondistinct (such as "Journal"), or if the name of the >>corporate body is in the title, or if *in the cataloger's judgment* the >>corporate body has a relationship with the publication such that it would be >>unlikely the publication would exist apart from that body, enter the title >>under the name of the corporate body." >> > >BUT this would dictate > >American Chemical Society: Journal >and >Association of Computing Machinery: Communications > >instead of > >Journal of the American Chemical Society >and >Communications of the ACM > >That's how they are always cited and looked for. And these are the most >important and most-cited periodicals in chemistry and computer science. >Does anybody go looking for them under the corporte names, I wonder? Bernhard seems to think I was advocating a return to pre-AACR2 form of entry for serials. I was talking simply about choice of main entry, and said nothing at all about changing the way the title is transcribed. The titles mentioned above would be entered as: American Chemical Society. Journal of the American Chemical Society. Association of Computing Machinery. Communications of the ACM. This does nothing at all to prevent people finding them under their titles proper. Kevin M. Randall Head, Serials Cataloging Section Northwestern University Library Evanston, IL 60208-2300 email: kmr@nwu.edu phone: (847) 491-2939 fax: (847) 491-7637 ========================================================================= Date: Thu, 14 Aug 1997 07:25:04 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Linking. Part 1 Comments: To: B.Eversberg@tu-bs.de In-Reply-To: <751B5072FEB@buch.biblio.etc.tu-bs.de> >A problem with authority records, for both kinds of links, is this: >Changes made later in the central authority file don't automatically >get transferred into all local systems affected. As the bibliographic utility Catss works, the ASN (authority sequence number) is substituted for the text in the entries of the bibliographic record. Our records are in our own files on their machine. When LC changes the text of an authority, that change displays automatically when our records containing that ASN are viewed. We download our records the 15th of each month for our most active file, annually for our less active ones. It astounds me that this arrangement seems unique to Catss. Perhaps before we go too far attempting to improve AACR and MARC, we should work on creating systems which make better use of what we already have. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Thu, 14 Aug 1997 11:21:26 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Regina Reynolds Subject: Re: Key titles & uniform title In-Reply-To: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Tue, 12 Aug 1997, Robert Cunnew wrote: > In article , > Regina Reynolds writes > >Thank you, Bernhard Eversberg for making a point that I have been meaning > >to write about. Indeed, there is no reason at all to have two fields in > >serial records which are serving exactly the same function: unique > >identification of a serial title. I have been long trying to have the key > >title substituted for the uniform title in LC cataloging but the uniform > >title seems quite entrenched. > > > This is an attractive suggestion but I can see problems that would need > to be addressed: > > 1. I have never seen a key-title printed in a serial. > > 2. ISSNs *are* printed in serials but they're not always the current > ones, so making it harder to pick up changes in key-titles. > > 3. There seems to be no easy way to find out key-titles. They don't > seem to be available on the Web. > > 4. Key-titles don't fit into AACR and this can be confusing. > > Another thing I don't like about the ISDS manual is (I believe) shared > with the LC rules for uniform titles: when a title is common to more > than one serial the first occurrence of that title registered doesn't > carry a qualifier. This is messy and doesn't help identification of the > first registered title. Is it beyond our present automated systems to > retrospectively add a qualifier? > > -- > Robert Cunnew > Librarian, Chartered Insurance Institute, London Key titles are available in several ways: the CONSER database (available on the major U.S. shared cataloging databases, in the Library of Congress online catalog, on U.S. MARC tapes) contains key titles for all U.S. and Canadian serials which have been assigned ISSN, and for many core serials from around the world. The ISSN database is available as a CD-ROM and as a tape service from the ISSN International Centre (20 rue Bachaumont, Paris 75002 France, http://www.issnic.com). Availability over the Web is currently being tested by ISSN centers. Terms of access to the Web are currently under discussion. If key titles were to be adopted by AACR as the unique identifier for serials, the rules for creating key titles would have to be included in AACR so that catalogers could create provisional key titles when authenticated key titles were not available. Finally, creation of key titles differs from LC rules for uniform titles by requiring retrospective qualification of key titles originally assigned without qualifiers. Thus, when the second occurrence of a title necessitates that title be qualified, the first title is amended by adding a qualifier. To circumvent this need somewhat, all one-word titles are qualified when the ISSN is first assigned, even if there are no conflicts. Regina R. Reynolds email: rrey@loc.gov Head, National Serials Data Program voice: (202) 707-6379 Library of Congress fax (202) 707-6333 101 Independence Avenue, S.E. ISSN Web page: lcweb.loc.gov/issn/ Washington, D.C. 20540-4160 ========================================================================= Date: Thu, 14 Aug 1997 11:18:53 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Cynthia Watters Subject: Re: Linking. Part 1 Comments: To: B.Eversberg@tu-bs.de I appreciated the posting about linkage, and might use this opportunity to digress into my concern about linkages which was occasioned by my vendor's (DRA) problems. That is, the linkage of authority records with each other. I think we really need to establish hierarchies of authority records so that the authority record for Huis clos, for example, is linked to and "hung off" the authority record for Sartre. Same thing for subject authorities with subfields. Instead of independent authorities for the $a heading and that for the $a same heading with $x, y, or z attached. Our vendor says they can't link a heading with multiple authority records, so if a name-title heading is linked to the name-title authority record it can't also be linked to the name authority record. Likewise, of course, a change to the name in the name authority record does not automatically cause a change in the same name in name-title authority records. I think this is a fundamental problem for references display and functioning in OPACs, for ordering of hits on hit lists, and for maintenance. Does anyone have any ideas about how to convert so, for example, a name title authority record is a "sub-record" linked to a name authority record? Or any disagreements with the need? Thanks, Cynthia Watters Catalog Librarian Middlebury College watters@myriad.middlebury.edu ========================================================================= Date: Thu, 14 Aug 1997 08:49:15 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Impact on bibliographic utilities What would happen in local systems which don't even have access point authorities, if work authorities or super work records were introduced? What would be the impact on bibliographic utilities of either work authority records (one for each work even if represented by a single manifestation?) or super work records (representing only works which have more than one manifestation?)? Is anyone from the bibliographic utilities reading the papers and list? If so, why aren't you contributing? If this raises utility costs, we would be priced out of our market niche. Wouldn't there be an impact on library budgets in these lean years? Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Thu, 14 Aug 1997 10:47:13 CDT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Jim E Cole Subject: Key title There has been some discussion on this list regarding the use of the key title as a uniform title for serials. To be sure, arguments can be advanced in favor of this proposal. I do hope, however, that the inherent problems will also be fully discussed. One of the problems is that there are two standards for the creation of key title for serials with generic titles proper--one for serials published within the territorial confines of the CONSER Program, allowing the use of data from authority files following the space-hyphen-space, and the other for serials published in the rest of the world, requiring the transcription of the "name of the issuing body in the sequence and form given on the title source" (ISDS Manual, p. 42) The CONSER provision, actually an exception to the general rule, brings the AACR2 uniform title and the key title into close alignment. One often sees very little difference between key titles constructed according to the general rule and those constructed according to the CONSER exception. Sometimes, however, this is not the case. To use a Canadian example, one that does *not* have a key title, the difference in the two is rather striking: (Key title according to CONSER exception) Working paper - Ontario Agricultural College. Dept. of Agricultural Economics and Business (Key title according to the general ISDS standard) Working paper - Department of Agricultural Economics and Business, University of Guelph Do we want such a mixture of headings in our catalogs, depending upon the place of publication of the serial? If we believe that our libraries acquire so few serials from outside the territorial confines of CONSER that this would have little effect if any upon our catalogs, we seem to have a rather provincial outlook on things. Another problem is the fact that not all serials have key titles at the time a library catalogs them; some are so ephemeral in nature or local in coverage that they may never have one. Do we want to set these aside, ask the authorized agency to assign a key title and ISSN, and then catalog them later, when the key title has been assigned? Does this delay serve our users well? The incorporation of key titles into our catalogs--and the use of the ISSN records as authority records--would require their storage and maintenance in our catalogs, and this involves cost. Do we want to pay for this? We must also consider the fact that there is no monographic equivalent of key title. The use of the key titles as uniform titles for serials would create a separate rule for serials, or a special one at least. Do we really want to do this? AACR2 uniform titles for "generic title" serials, and their CONSER key title equivalents, may be properly described as "posterior corporate entry." All that has been accomplished is the transposition of the corporate heading to the end of the title. We have spent a great deal of time in the past seventeen years creating posterior corporate entries; they now take up a fair amount of computer storage space for our OPACs. We could almost completely avoid the use of uniform titles for "generic title" serials merely by adopting a rule applicable to *all* publications, one that would require the entry of a publication issued by or under the authority of a corporate body under the name of the body if the title proper of the publication consisted solely of a generic term or phrase that did not adequately identify the publication except when taken in conjunction with the name of the body. This is rule simplifcation; and is this not what we are to be about? Jim Cole, Editor The Serials Librarian ========================================================================= Date: Thu, 14 Aug 1997 10:12:54 -0700 Reply-To: Daniel CannCasciato Sender: "International Conference on the Principles and Future Dev. of AACR" From: Daniel CannCasciato Subject: AACR and systems In-Reply-To: <970814111853.9515@myriad.middlebury.edu> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Hello All, After reading Fattahi's paper and after having read many of the postings on this list, I'm wondering which is the cart and which is the horse: the code or the catalog display? While I like the idea of the super record (and will be trying to test it locally to see if I can make it work) many of the other suggestions for changes to AACR seem to be system dependent. Or, more precisely, the suggested changes are spurred by system inadequacies and system strengths. (Inadequacies appear to be winning the battle at this point.) To me, this is a flaw. While the two are teamed, to be sure, I feel the code should be driving the system vendors to take advantage of what we do. For example, the fact that many corporate headings are too long to display on one line is not, in my opinion, a cataloging problem. It's a display problem. We should complain to our vendors about it. If we did so collectively, I bet we could get some real results. The fact that corp. names are not conventionalized WHO Commission ... vs. World Health Organization. $b Advisory ... " is not really much of an issue if you have a system that makes use of authority records. [There are other issues in this area that deserve atttention, I agree. Fattahi makes some good suggestions.] But many of us don't have systems that utilize authority records. Many of us have systems that don't display author-title references very intelligibly. But should we adapt the code to fit a vendors interest (or lack of interest)? I'd say no. Especially as the code is being used by many many libraries that don't have a vendor. Writing a code to take advantage of hypertext links only makes sense if you have hypertext links available. Daniel ------------------------------------------- Daniel CannCasciato Head of Cataloging and Interlibrary Loans Central Washington University Library 400 East 8th Ave Ellensburg WA 98926-7548 509 963-2120 509 963-3684 (FAX) dcc@cwu.edu ========================================================================= Date: Thu, 14 Aug 1997 13:17:49 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Regina Reynolds Subject: Correction: ISSN URL In-Reply-To: <9708141547.AA11045@isua2.iastate.edu> MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII The correct URL for the ISSN International centre is as was given earlier on the list: http://www.issn.org Regina R. Reynolds email: rrey@loc.gov Head, National Serials Data Program voice: (202) 707-6379 Library of Congress fax (202) 707-6333 101 Independence Avenue, S.E. ISSN Web page: lcweb.loc.gov/issn/ Washington, D.C. 20540-4160 ========================================================================= Date: Thu, 14 Aug 1997 14:02:07 +0000 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Pamela Simpson Subject: Re: Key Title MIME-version: 1.0 Content-type: text/plain; charset="us-ascii" Jim E. Cole wrote: >> Another problem is the fact that not all serials have key titles at >>the time a library catalogs them; some are so ephemeral in nature or local >>in coverage that they may never have one. Do we want to set these aside, >>ask the authorized agency to assign a key title and ISSN, and then catalog >>them later, when the key title has been assigned? Does this delay serve our >>users well? I think the idea would be for all serials catalogers to assign a key title at the time of cataloging, using the same standard that ISSN centers currently use. Although these key titles would not yet be authenticated by an ISSN center, (and some never would be, if the title never received an ISSN), they could be created by anyone who possesses the level of expertise required to correctly formulate the uniform titles currently assigned. Although the discrepancy in the form of name that Jim pointed out would certainly have to be resolved, the rules are really quite similar. >> >> The incorporation of key titles into our catalogs--and the use of >>the ISSN records as authority records--would require their storage and >>maintenance in our catalogs, and this involves cost. Do we want to pay for >>this? Using ISSN records as "authority" records for serials does not necessarily mean that they would be used in the same way name headings are. Clearly, we would all need access to the ISSN database, but we would not necessarily have to store them in our catalogs to be used to flip headings or refer users to different forms. The work of identifying a serial in the process of assigning an ISSN involves a kind of "authority" work that has not been formally recognized. ISSN catalogers search the ISSN database and spend a great deal of time researching conflicts and resolving questions regarding the history of titles, determining whether or not a serial is "the same one" that was formerly published in another country, etc. Steve Shadle and I presented a workshop at the 1993 NASIG conference on the use of the ISSN database as a cataloging resource; a report of this workshop appears in the proceedings of that conference (Serials Librarian 24 (1994) no. 3/4). One of the great advantages of using ISSN records as an authoritative source is that ISSN centers are often in touch with the publishers, and so are in a position to determine relationships and histories which are often not described explicitly on the pieces themselves. It is unfortunate that this rich resource has not been easily available to catalogers. The CD has been available since 1993, but it is quite expensive. I hope that the International Center will find a way to offer affordable web access. This standard could offer a great deal more to the library community if it were more readily available. Unfortunately, funding for the ISSN program, both in the U.S. and internationally, is a perpetual problem, and the International Center may not be able to afford to offer free access to the records. >> We must also consider the fact that there is no monographic >>equivalent of key title. The use of the key titles as uniform titles for >>serials would create a separate rule for serials, or a special one at least. >>Do we really want to do this? Yes, I think we do. Or at least, I don't think we should rule something out just because it involves a separate rule for serials. Trying to accomodate items that are published serially with the same rules we use for things that are published only once reminds me of my favorite quote from this year's NASIG, by Carroll Davis, "Cataloging a monograph is like performing an autopsy; cataloging a serial is like studying a gazelle in the wild." Pamela Simpson *********************************** Pamela Simpson Serials and Electronic Resources Cataloging Librarian E506 Pattee Library The Pennsylvania State University University Park, PA 16802-1805 (814) 865-1755 Fax: (814) 863-7293 p2s@psulias.psu.edu *********************************** ========================================================================= Date: Thu, 14 Aug 1997 16:23:00 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Disipio Mary F." Subject: Beyond MARC Content-transfer-encoding: 7BIT Please note that the following conference paper is now available on the JSC Web as a HTML document Beyond MARC by Mick Ridley. ========================================================================= Date: Fri, 15 Aug 1997 09:54:17 +1000 Reply-To: Giles S Martin Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Linking. Part 1 In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII The Australian Bibliographic Network (ABN) works in a somewhat similar way to Catss as Mac just described it. Since ABN was based on WLN software, the same would be true of WLN. Internally in ABN, headings in the 1XX, 440, 6XX, 7XX and 8XX fields (in bib. records) are represented by the system number of the vocabulary record. When the vocabulary record is changed, all the linked bibliographic records are channged as well (except for a few bugs in the system!) When you change a heading in a bib. record, the system looks for a match in the vocabulary file, and links to the match if there is one; otherwise, it creeates a new vocabulary record. Similarly for new bib. records coming into the system: headings either link to existing bib. records or create new vocabulary records. (They are called vocabulary records until someone checks to make sume they are valid. Only after they are validated are they called authority records.) Unfortunately, the only link with LC's authority files is through the kind of link that Bernhard Eversberg described in his "Linking. Part 1" message: that is, through the text matching the correct LC heading. LC authority records are not loaded into ABN. However, most people using ABN are often unaware of the system that I have described, unless the system breaks down. When you display a bibliographic record, or output it to a local system, you see the headings from the vocabulary records, and not the system numbers that are really present in the bibliographic records. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 2 4921 5828 (International) Fax: +61 2 4921 5833 (International) ## [Note that the telephone numbers have just changed. From inside Australia, the area code is now (02).] On Thu, 14 Aug 1997, J. McRee Elrod wrote: > As the bibliographic utility Catss works, the ASN (authority sequence > number) is substituted for the text in the entries of the bibliographic > record. Our records are in our own files on their machine. When LC > changes the text of an authority, that change displays automatically > when our records containing that ASN are viewed. We download our > records the 15th of each month for our most active file, annually for > our less active ones. > > It astounds me that this arrangement seems unique to Catss. ========================================================================= Date: Fri, 15 Aug 1997 13:39:04 +1200 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ruth Lewis Subject: Re: Linking. Part 1 Comments: To: Giles S Martin Thanks for that Giles. The New Zealand Bibliographic Network (NZBN) uses much the same software, and does the same things with vocabulary headings as ABN. The name or subject heading is not actually part of the bib. record, the bib record just contains a link to the vocabulary heading. But to the user it looks seamless (most of the time). NZBN manages to exchange data with ABN, LC and others, as well as provide data to local systems. I'm not sure how this is done, but the problem of data exchange using this type of linkage is obviously not insurmountable. Ruth Lewis Ruth Lewis Music Room National Library of New Zealand ruth.lewis@natlib.govt.nz Telephone (64 4) 474 3000 ext. 8888 Toll free 0800 736 561 Fax (64 4) 474 3042 These opinions are my own and are not necessarily National Library of New Zealand policy. ========================================================================= Date: Thu, 14 Aug 1997 19:48:15 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Linking. Part 1 Comments: To: ulgsm@dewey.newcastle.edu.au In-Reply-To: Giles S Martin wrote: >The Australian Bibliographic Network (ABN) works in a somewhat similar >way to Catss as Mac just described it. Since ABN was based on WLN >software, the same would be true of WLN. When we investigated WLN, we found it did not have cascade, that is, if you had a standard heading with a free floating subdivision, that string had to be established if it was to verify. But then, when the main heading changed, this new authority remained unchanged. Has this problem been solved on ABN? Why is ABN abondoning its WLN clone? Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 15 Aug 1997 14:05:32 +1000 Reply-To: Giles S Martin Sender: "International Conference on the Principles and Future Dev. of AACR" From: Giles S Martin Subject: Re: Linking. Part 1 In-Reply-To: Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII As far as ABN is concerned, each combination of a main heading with a subheading is a different string of characters, with no relationship except that they file near each other. So you can establish a heading with a subheading without also having the heading by itself. For example, you can have the LCSH string "Education--Research--Australia" without also having "Education" and "Education--Research". Furthermore, if you wanted to change "Education--Research" to "Educational research", you would have to change all of the hundreds of combinations starting with those words, one by one. There are a lot of small reasons why ABN is changing, and one big one. The big reason is that the system design makes it difficult to makes small changes! So there a lot of known bugs in ABN, and a lot of improvements that people would like to make, but it would be easier to fix these by getting a new system that by fixing the problems one by one. Giles #### ## Giles Martin ####### #### Quality Control Section ################# University of Newcastle Libraries #################### New South Wales, Australia ###################* E-mail: ulgsm@dewey.newcastle.edu.au ##### ## ### Phone: +61 2 4921 5828 (International) Fax: +61 2 4921 5833 (International) ## [Note that the telephone numbers have just changed. From inside Australia, the area code is now (02).] On Thu, 14 Aug 1997, J. McRee Elrod wrote: > When we investigated WLN, we found it did not have cascade, that is, if > you had a standard heading with a free floating subdivision, that string > had to be established if it was to verify. But then, when the main > heading changed, this new authority remained unchanged. Has this > problem been solved on ABN? Why is ABN abondoning its WLN clone? ========================================================================= Date: Fri, 15 Aug 1997 01:07:56 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Linking. Part 1 In-Reply-To: <970814111853.9515@myriad.middlebury.edu> from "Cynthia Watters" at Aug 14, 97 11:18:53 am Content-Type: text Greetings, I agree that what Cynthia describes is a fundmental problem and am baffled by the inability of current systems to account for the heirarchies required. As an explanation, all I can guess is that the heirarchy assumed by AACR2 is a function of the flat/card file and the technology of the card catalog. The end product of that technology has not been dpulicated by online systems. Imagine the large card catalog room with thousands of card trays. One enters the room views guides of various kinds: subject, authors, etc. One walks to authors and sees guides: A, B, C, etc. One walks to S and checks the guides .... Sellers... Shakespeare! One checks the guides ..... Tragedy of Hamlet see Hamlet One checks the guides ..... Hamlet, etc. Within a few moments, the antique technology of the card catalog can direct the interested reader to the file of cards displaying all entries for Shakespeare's Hamlet in one place (assuming the library followed certain cataloging rules, and that catalogers and filers did not make mistakes). This very simple process, assumed by most every librarian and library user, to be available in any catalog, actually does not exist (or at least does not exist elegantly) in any system I have seen. It assumes a heirarchical reference/index structure as suggested by Cynthia. Once again, the problem is not with the rules, but with library systems. --ralph p. Cynthia Watters said ...> > I appreciated the posting about linkage, and might use this opportunity to > digress into my concern about linkages which was occasioned by my vendor's > (DRA) problems. > > That is, the linkage of authority records with each other. I think we > really need to establish hierarchies of authority records so that > the authority record for Huis clos, for example, is linked to and "hung > off" the authority record for Sartre. Same thing for subject authorities > with subfields. Instead of independent authorities for the $a heading > and that for the $a same heading with $x, y, or z attached. > > Our vendor says they can't link a heading with multiple authority records, > so if a name-title heading is linked to the name-title authority record > it can't also be linked to the name authority record. Likewise, of course, > a change to the name in the name authority record does not automatically > cause a change in the same name in name-title authority records. > > I think this is a fundamental problem for references display and functioning > in OPACs, for ordering of hits on hit lists, and for maintenance. > > Does anyone have any ideas about how to convert so, for example, a > name title authority record is a "sub-record" linked to a name authority > record? Or any disagreements with the need? > > Thanks, > Cynthia Watters > Catalog Librarian > Middlebury College > watters@myriad.middlebury.edu > -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Thu, 14 Aug 1997 23:11:37 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: "Nesting" in authority control Comments: To: ulgsm@dewey.newcastle.edu.au In-Reply-To: Giles S Martin wrote (re ABN): >Furthermore, if you wanted to change "Education--Research" to >"Educational research", you would have to change all of the hundreds of >combinations starting with those words, one by one. On Catss you don't. It is automatic. Some call this "nesting". It is accomplished by the "casc"ade command. That, along with the ability of Catss to compute labels (090$c1-2$d1-100 gets you 300 labels), is why we did not leave Catss for WLN. Is anyone aware of any inhouse system which will do these two things? We've found none. I agree whole hearted with those who have said we should improve systems, including being able to adequately handle authorities, and parse holdings statements, before making changes in AACR to make it "better" for systems. It's time for the tail to stop wagging the dog. Nac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 15 Aug 1997 09:14:14 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: "Nesting" in authority control MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Mac, time and again, you relate all those marvelous facts about CATSS. One is beginning to wonder if not the only two problems with CATSS are that it is on the left side of the border, and that its name is not OCLC. (just for the weekend) B. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Fri, 15 Aug 1997 14:34:37 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Beyond MARC MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT http://www.nlc-bnc.ca/jsc/r-beyond.htm Mick Ridley's paper is a welcome addition to the conference papers, in that it offers views of a non-librarian, computer expert. We get precious few such views, sometimes they could do us a lot of good. Much of what he says is additional confirmation of what was already stated or suggested in this list, like on uniform titles, language bias, or his suggestion of a 3-tier structure for records (work, manifestation, copy). For the latter, however, the "Functional Requirements..." study and Vellucci's paper go a lot further, and Ridley seems to have missed those. Mick Ridley points out, as one would expect from a computer person, that too many things are optional in the rules (resulting in inhomogeneous data), and there's still a lack of standardisation with certain fields, like publishers' names (producing in imprecise results of queries). This would mean that traditional bibliographic standards (title page transcription) may have to be re-evaluated over and again. Ridley still adheres to the superficial and now outdated (or so I think) attitude toward main entry as being a concept of the past. He obviously did not follow this list (or AUTOCAT, for that matter). He admits, in his introduction, that his views are probably prejudiced from the programmer's side. It is one particular prejudice I want to get at here: on page 3, he writes: "... Indexes are just a device for making queries go faster, and are hidden from just about everyone." Wrong. Very wrong. Indexes, if well-designed, are an invaluable medium for browsing and collocating, and should as such never be hidden from anyone. Everyone should be encouraged to look into them. This is clearly a computer person's prejudice. They work from the assumption that everybody approaching a catalog has a clear concept of what they are looking for, and know the correct spellings and everything. If this is true for every other database application context, then library catalogs are the odd counterexample. The very concept of a "query" (as in "SQL queries") is the wrong paradigm for ibrary catalogs, but computer people are indoctrinated with this. To enable browsing (on shelves as well as in files) has always been an important feature of library service. Not every index is good for browsing. Sometimes indeed, the software has just not been designed to be able to present an index for browsing. Then, Ridley's statement is a handy excuse. But sometimes, a software is explicitly designed to be based on browsing. For example, look at an index like this, with title keywords and subject headings interfiled (an * marks the subj headings) 108 double 2 double bass * 5 double bass -- orchestra studies * 1 double bass and harpsichord music * 1 double bass and harpsichord music, arranged * 4 double bass and piano music * 4 double bass and piano music, arranged * 1 double bass and piano music, arranged -- scores and part * 1 double bass and piano music, arranged -- scores and parts * 5 double bass music * 1 double bass music double basses 2, arranged -- scores * 6 double-bass 1 double-bass and harpsichord music, arranged * 1 double-bass and percussion music * 3 double-bass and piano music * 1 double-bass and piano music -- scores and parts * 5 double-bass and piano music, arranged * 1 double-bass and piano music, arrangedn * 7 double-bass music * 1 double-bass music double basses 2 -- scores * 1 double-bass music double basses 4 -- scores and parts * or another section: 211 pianoforte 1 pianoforte--concertos 1 pianoforte-album 1 pianoforte-literature 1 pianoforte-playing 1 pianoforte-sonate 1 pianoforte-spiel 2 pianoforte-werke 1 pianoforte/clavicembalo 2 pianofortebegleitung 1 pianofortes 2 pianoforti 88 pianos 1 pianos 2 with orchestra * 7 pianos 2 with percussion * 1 pianos 2 with percussion -- scores * 1 pianos 2 with percussion -- scores and parts * 1 pianos 3 with orchestra * 1 pianos concerto * 1 pianos music * 1 pianotorte 1 pianotrio 4 pianto 1 piatero 7 piatigorsky 1 piatnitsy 8 piave 1 picassos 2 piccadilly Everybody reading this list will be able to figure out what the advantages of this kind of index are over the hit-and-miss approach of a query language. Rule revision, as was stated many times now, should now concern itself more with indexing and collocation in OPACs. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Fri, 15 Aug 1997 11:40:36 -0230 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Charley Pennell Subject: Re: Linking. Part 1 Comments: To: mac@slc.bc.ca MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit Mac- As has been discussed many times before on these fibre pathways, WLN uses the same model of authority linkages to bibliographic records *within the utility*. Where the model breaks down is when this data is transported out of the control of the utility, namely to the local site or to another site using the bib record. If memory serves me, our beloved Canadian utility would not output authority linkages per se to the local library, preferring to replace the link with the actual text from the authority file at the point in time the record was exported. This prevented many of us from utilizing the linkages in our own system and shackled us needlessly to a database halfway across the country. Our present Sirsi Unicorn system uses proprietary subfields (much like Bernhard's |3 model) to accomplish hard linkages to authority file records. The text itself remains in the field and is flipped when the linked authority record is replaced. Records reported out to USMARC files will not have the proprietary subfield, since it is not defined in that standard. Just as well, as the authority control number is an internal one (just like Mac's ASNs) and not a standard control number like the authority 001. On the other hand, what does this issue, and some of the other discussions vis-a-vis the various MARC formats, have to do with revision of AACR? _______________________________________________________________________ """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" Charley Pennell cpennell@morgan.ucs.mun.ca Head, Acquisitions/Cataloguing Division voice: (709)737-7625 Queen Elizabeth II Library fax: (709)737-3118 Memorial University of Newfoundland St. John's, NF Canada A1B 3Y1 World Wide Web: http://sicbuddy.library.mun.ca/~charl8P9/chuckhome.html Cataloguer's Toolbox: http://www.mun.ca/library/cat/ _______________________________________________________________________ """"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""" -----Original Message----- From: J. McRee Elrod To: AACRCONF@INFOSERV.NLC-BNC.CA Date: Thursday, August 14, 1997 1:15 PM Subject: Re: Linking. Part 1 >>A problem with authority records, for both kinds of links, is this: >>Changes made later in the central authority file don't automatically >>get transferred into all local systems affected. > >As the bibliographic utility Catss works, the ASN (authority sequence >number) is substituted for the text in the entries of the bibliographic >record. Our records are in our own files on their machine. When LC >changes the text of an authority, that change displays automatically >when our records containing that ASN are viewed. We download our >records the 15th of each month for our most active file, annually for >our less active ones. > >It astounds me that this arrangement seems unique to Catss. > >Perhaps before we go too far attempting to improve AACR and MARC, we >should work on creating systems which make better use of what we already >have. > >Mac > > __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) > {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ > ___} |__ \__________________________________________________________ > > ========================================================================= Date: Fri, 15 Aug 1997 10:36:38 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Cynthia Watters Subject: Re: Linking. Part 1 Ralph says the problem is not with the rules but with library systems. I agree it is not at all with the rules, but my vendor asserts (and has me convinced), the problem is with the MARC format which treats each string with additional subfields as totally unrelated to the string without subfields. Cynthia Watters Catalog Librarian Middlebury College watters@myriad.middlebury.edu ========================================================================= Date: Fri, 15 Aug 1997 09:43:23 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Relation of linkages and revision of AACR Comments: To: cpennell@morgan.ucs.mun.ca In-Reply-To: <199708151410.LAA13027@piva.ucs.mun.ca> "Charley Pennell" wrote: > As has been discussed many times before on these fibre pathways, WLN uses >the same model of authority linkages to bibliographic records *within the >utility*. Where the model breaks down is when this data is transported out >of the control of the utility ... We (and I assume others) order the authority records to which our own records are linked. Our headings which match our authorities link up and produce cross references in printed catalogues and on cards for our card catalogue customers. (So far as I know we are the only ones automatically printing - as opposed to typing - card catalogue cross references.) But our private file on Catss retains the numbers as opposed to text, and when re down loaded, reflect the changes made in the authority at LC. > On the other hand, what does this issue, and some of the other discussions >vis-a-vis the various MARC formats, have to do with revision of AACR? Too much of the discussion has to do with revision of AACR for automated bibliographic systems. It's not AACR which needs fixing so much as it is the automated bibliographic systems. We cataloguers must return to our traditional role of designing and constructing catalogues, and stop focusing on the creation of ever more complex bibliographic records. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 15 Aug 1997 10:10:36 EDT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Zhishan Xu Subject: Re: Linking. Part 1 In-Reply-To: Your message of "Thu, 14 Aug 1997 11:18:53 EDT." <970814111853.9515@myriad.middlebury.edu> I have watched with great interest regarding linkages of authority record. For me, the linkages to a bibliographic record should go beyond authority record. If we regard a bibliographic record as an information entity representation, and a starting point to navigate all the information about the same information entity, then the linkages to a bibliographic record should be more multi-dimentional, and highly dynamic than what our current online systems have provided us. I think it would be very important to study what should be linked to bibliographic record, and its impact on cataloging rules, bibliographic data encoding system such as MARC, and online systems. This study may change our mindset on what shoulde be future cataloging principles. Thanks! Amanda Xu Serials Cataloger MIT Libraries Massachusetts Institute of Technology Cambridge, MA 02139 Email: zxu@mit.edu ========================================================================= Date: Fri, 15 Aug 1997 15:29:11 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Michael Gorman Subject: Posting MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit It seems to me that many of the communications to this listserv are interesting (sometimes not very) but have little to do with AACR2 or any possible revision thereof. Displays, MARC, the niceties or grotesqueries of individual online systems, whether AACR2 is based on the card catalogue (I don't think it is), subject headings (other than those based on names (personal, corporate, geographic) and titles), etc. etc. Subjects that do have something to do with the matter on hand are links between records, various questions revolving around serial publications, descriptive cataloguing of electronic documents and resources, main entry (though a lot of that discussion is shadow-boxing), etc. Perhaps it would be more useful if the discussion confined itself to the latter topics. Michael -- ________________________________________________________ Michael Gorman michael_gorman@csufresno.edu Dean of Library Services telephone: (209) 278-2403 CSU-Fresno fax: (209) 278-6952 5200 N. Barton Fresno, CA 93740-8014 *It is a luxury to learn; but the luxury of learning is not to be compared with the luxury of teaching* RDH ________________________________________________________ ========================================================================= Date: Fri, 15 Aug 1997 15:25:51 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Michael Gorman Subject: Posting MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Dear Friends I tried on Wednesday, for the third time, to send a disc with the paper that Pat Oddy and I wrote for the Toronto conference to the NLC = BNC. I just wanted you all to know that the paper was completed lo these many weeks ago and it is some problem with North American mails, not our forgetfulness or indolence, that has prevented its "publication" on the Web. Michael -- ________________________________________________________ Michael Gorman michael_gorman@csufresno.edu Dean of Library Services telephone: (209) 278-2403 CSU-Fresno fax: (209) 278-6952 5200 N. Barton Fresno, CA 93740-8014 *It is a luxury to learn; but the luxury of learning is not to be compared with the luxury of teaching* RDH ________________________________________________________ ========================================================================= Date: Sat, 16 Aug 1997 23:39:50 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Ralph Papakhian Organization: Indiana University W&G Cook Music Library Subject: Re: Linking. Part 1 In-Reply-To: <970815103638.8700@myriad.middlebury.edu> from "Cynthia Watters" at Aug 15, 97 10:36:38 am Content-Type: text Greetings. I'm not quite sure what Cynthia means about "each string with additional subfields as totally unrelated to the string without subfields." But, here is a result of searching in NOTIS staff mode: ********************** BIBLIOGRAPHIC INDEX -- 1 ENTRIES FOUND 1 - 1 DISPLAYED MOZART WOLFGANG AMADEUS 1756 1791 .GIRL IN GARDENERS DISGUISE *SEARCH UNDER 1 MOZART WOLFGANG AMADEUS 1756 1791 .FINTA GIARDINIERA (AF 400 ccb) ********************** If one actually conducts this search or redirects it, the result will be all instances of Mozart Wolfgang Amadeus 1756 1791 .finta giardiniera ... including strings with additional subfields. Here is the result: ********************* BIBLIOGRAPHIC INDEX -- 32 ENTRIES FOUND 1 - 12 DISPLAYED MOZART WOLFGANG AMADEUS 1756 1791 .FINTA GIARDINIERA 1 *ESTABLISHED HEADING (AF 100 ccb) 2 .DRAMMA GIOCOSO LA FINTA GIARDINIERA EIN VERGLEICH DER VERTONUNGEN VON PASQUALE ANFOS <1989> (BM 600) 3 .FINTA GIARDINIERA <1973> sound (BM 100) 4 .FINTA GIARDINIERA <1989> sound (BM 100) 5 .FINTA GIARDINIERA K 196 <1954> sound (BM 100) 6 .FINTA GIARDINIERA KV 196 DRAMMA GIOCOSO IN TRE ATTI <1981> sound (SB 100) 7 .FINTA GIARDINIERA KV 196 DRAMMA GIOCOSO IN TRE ATTI <1981> sound (SB 100) 8 .FINTA GIARDINIERA KV 196 DRAMMA GIOCOSO IN TRE ATTI DI GIUSEPPE PETROSELLINI <1981> sound (BM 100) 9 .FINTA GIARDINIERA MOZARTS MUENCHENER AUFENTHALT 1774 75 BAYERISCHE STAATSBIBLIOTHEK <1975> (BM 600) 10 .FINTA GIARDINIERA OPERA BUFFA IN 3 ACTEN KOCH VERZ NO 196 <1800> music (BM 100) 11 .GIRL IN GARDENERS DISGUISE K 196 <1967> sound (FW 100) .FINTA GIARDINIERA AH DAL PIANTO 12 *ESTABLISHED HEADING (AF 100 cbb) BIBLIOGRAPHIC INDEX -- 32 ENTRIES FOUND 13 - 22 DISPLAYED MOZART WOLFGANG AMADEUS 1756 1791 .FINTA GIARDINIERA AH DAL PIANTO 13 .OPERA ARIAS OPERNARIEN <1983> sound (BM 700) .FINTA GIARDINIERA CON UN VEZZO ALLITALIANA GERMAN 14 .ARIEN LIEDER UND ROMANZEN AUS DEUTSCHEN OPERN <1973> sound (BM 700) .FINTA GIARDINIERA CRUDELI FERMATE 15 *ESTABLISHED HEADING (AF 100 cbb) 16 .OPERA ARIAS OPERNARIEN <1983> sound (BM 700) .FINTA GIARDINIERA ENGLISH 17 .FINTA GIARDINIERA K 196 DRAMMA GIOCOSO IN THREE ACTS <1991> visual (BM 700) 18 .FINTA GIARDINIERA K 196 DRAMMA GIOCOSO IN THREE ACTS <1991> visual (BM 700) 19 .FINTA GIARDINIERA K 196 DRAMMA GIOCOSO IN THREE ACTS <1991> visual (BM 700) .FINTA GIARDINIERA GERMAN 20 .FINTA GIARDINIERA K 196 OPERA BUFFA IN THREE ACTS <1950> sound (SE 100) 21 .GARTNERIN AUS LIEBE LA FINTA GIARDINIERA OPERA BUFFA IN 3 ACTS KV 196 <1972> sound (BM 100) 22 .GIRL IN GARDENERS DISGUISE K 196 <1967> sound (BM 100) BIBLIOGRAPHIC INDEX -- 32 ENTRIES FOUND 23 - 32 DISPLAYED MOZART WOLFGANG AMADEUS 1756 1791 .FINTA GIARDINIERA GERMAN 23 .OPERA BUFFA <1977> sound (BM 700) .FINTA GIARDINIERA GERMAN ITALIAN 24 .FINTA GIARDINIERA A COMIC OPERA IN THREE ACTS K 196 <1900> music (BM 100) .FINTA GIARDINIERA GERMAN SELECTIONS 25 .FINTA GIARDINIERA ABRIDGED <----> sound (BM 100) .FINTA GIARDINIERA OUVERTURE 26 *ESTABLISHED HEADING (AF 100 cbb) 27 .BRUNO WALTER CONDUCTS MOZART <1960> sound (BM 700) 28 .FINTA GIARDINIERA OVERTURE <1940> music (BM 100) 29 .OVERTURES <1967> sound (BM 700) 30 .SYMPHONY NO 39 IN E FLAT MAJOR K 543 TITUS OVERTURE K 621 LA FINTA GIARDINIERA OVERT <1970> sound (BM 700) .FINTA GIARDINIERA SELECTIONS 31 .FINTA GIARDINIERA K 196 HIGHLIGHTS OPERA BUFFA IN 3 ACTS <1959> sound (BM 100) .FINTA GIARDINIERA VOCAL SCORE GERMAN 32 .GAERTNERIN AUS LIEBE OPER IN DREI AUFZUGEN IN VOLLSTANDIGEM CLAVIERAUSZUG MIT DEUTSC <1829> music (BM 100) ****************************************** There are problems with this display (and NOTIS has plenty of other problems indexing in both public and staff modes). But I believe that the software, in this instance, is implementing the AACR and USMARC correctly. So it may be that a particular system cannot index MARC records correctly. It is not logical to therefore conclude that all systems cannot index MARC records correctly because of the structure of MARC. The reference above, should lead to an index point where all instances of the correct title can be found (including additions, etc. to the title). I haven't seen a WEB based catalog that can do that (admitting that I have only looked at a few). You know, if you can put a man on the moon ... --ralph p. Cynthia Watters said > > Ralph says the problem is not with the rules but with library systems. > I agree it is not at all with the rules, but my vendor asserts (and has > me convinced), the problem is with the MARC format which treats each > string with additional subfields as totally unrelated to the string > without subfields. > > Cynthia Watters > Catalog Librarian > Middlebury College > watters@myriad.middlebury.edu > -- A. Ralph Papakhian, Indiana University Music Library Bloomington, IN 47405 812/855-2970 papakhi@indiana.edu co-owner: MLA-L@listserv.indiana.edu ========================================================================= Date: Sun, 17 Aug 1997 14:57:03 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Posting Comments: To: michael_gorman@CSUFRESNO.EDU In-Reply-To: <199708152229.PAA09045@zimmer.csufresno.edu> >It seems to me that many of the communications to this listserv are >interesting (sometimes not very) but have little to do with AACR2 or any >possible revision thereof. The greatest handicap AACR suffers, I feel, is its isolation (except in practice) from MARC and OPAC systems. I would not like to see us make changes in AACR2 (based as it is on the long evolution of bibliographic description) because of problems which really belong to MARC and/or automated catalogue systems, and which should be rectified *there*. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 18 Aug 1997 08:16:36 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Linking. Part 1 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Zhishan Xu wrote: > I have watched with great interest regarding linkages of authority > record. For me, the linkages to a bibliographic record should go > beyond authority record. > Wait for Part 2, to be posted later this week. > > I think it would be very important to study what should be linked to > bibliographic record, and its impact on cataloging rules, > bibliographic data encoding system such as MARC, and online systems. > This study may change our mindset on what shoulde be future cataloging > principles. > There are several studies into this matter already (Tillett, Smiraglia, Leazer, Vellucci...) and Part 2 will be based on these. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 18 Aug 1997 13:16:27 +0100 Reply-To: mh@bodley.ox.ac.uk Sender: "International Conference on the Principles and Future Dev. of AACR" From: Mike Heaney Subject: Re: Michael Gorman's posting In-Reply-To: <199708152229.PAA09045@zimmer.csufresno.edu> Mime-Version: 1.0 Content-Type: TEXT/PLAIN; CHARSET=US-ASCII On Fri, 15 Aug 1997 15:29:11 -0700 Michael Gorman wrote: > It seems to me that many of the communications to this istserv are > interesting (sometimes not very) but have little to do with AACR2 or any > possible revision thereof. Displays, MARC, the niceties or grotesqueries > of individual online systems, whether AACR2 is based on the card catalogue > (I don't think it is), subject headings (other than those based on names > (personal, corporate, geographic) and titles), etc. etc. I fully agree that if AACR were an independent, self-contained construct then we could all be happy with (most of) it. But surely we should have some regard to the uses to which it is put? AACR is used primarily in library catalogues, where it co-exists with classification schemes, subject headings, coded information (in MARC), keyword indexes (in computer systems generally), call numbers &c.; and all these elements have a role to play in bringing together the users and the items they want. The library system in use until recently at the University of Oxford actually incorporated (and had done for a decade or more) many of the things we're now talking about -- entity-relationship structure, record-to-record links &c -- and used on its own would have been a wonderful system. Unfortunately it was very bad at communicating with the outside world -- imported and exported records had to lose most of the advanced elements. AACR2 may be OK in itself but -- like our old system -- can fail to address the issues in the real world. I agree with Mac (J. McRee Elrod) that most of the problems derive from the MARC structure and the use made by automated systems of that structure rather than in AACR itself. Do we, therefore, just turn away and say "not our fault"? Do we try to "take over" MARC? (And which MARC? -- of all the MARC formats I know, USMARC is the most closely tied to the catalogue card era, and is the least tractable) Or should AACR, as one of only several elements in a MARC record, recognise its subsidiary place in this structure and try to accommodate itself better to it? What then of the library world (which still exists) which doesn't use MARC (being either still a hard-copy catalogue or a small standalone system)? Though Michael avers that AACR2 is not based on the card catalogue, it does inhabit a world in which the instruction "make a reference" is enough: so at 26.2A3 "Lewis, C. Day- . see Day-Lewis, C." is intended to guide the user to the entries for C. Day-Lewis, and then use intelligence to work out that name-title entries uder Day-Lewis also fall within the scope of the reference. Similarly, publisher phrases such as "The Society" require the user to make inferences about their meaning. In that sense, AACR2 does belong to a pre-computer age. If systems designers (being guided or misguided by AACR2 as encoded in MARC) have not made the inferences and interpretations implicit in AACR2, then shouldn't we be making them explicit instead? Or should we be redesigning AACR in the knowledge that its primary implementation will be in computer systems of one kind or another, with an awareness of how computers can hold and process data? Mike Heaney Bodleian Library michael.heaney@bodley.ox.ac.uk ========================================================================= Date: Mon, 18 Aug 1997 14:51:42 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Michael Gorman's posting MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Mike Heaney writes: > ... > If systems designers (being guided or misguided by AACR2 as > encoded in MARC) have not made the inferences and > interpretations implicit in AACR2, then shouldn't we be making > them explicit instead? Or should we be redesigning AACR in the > knowledge that its primary implementation will be in computer > systems of one kind or another, with an awareness of how > computers can hold and process data? > > The issue of filing rules must be mentioned here again. Sorry if I'm getting on everybody's nerves. Human filers of catalog cards can make inferences a computer can impossibly make when it has to arrange browsable indexes or result sets. Therefore, the elements to be used for sorting by computer have to be controlled more explicitly and tightly than headings for cards. Make filing rules part of AACR and you automatically raise the awareness of everybody involved, including systems designers. And there are more such elements than there had ever been for cards, for we want indexes of publishers' names, place names, but above all, titles, and ALL titles in all fields and subfields. Which brings me back to the initial article problem. If the legendary fairy were to grant me my three wishes (for this conference, that is), this one would be among them. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 18 Aug 1997 10:23:53 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Main entry for performances I read Martha Yee's very interesting paper, "What is a Work?", shortly after reading the "Draft Document of the Task Force on the Cataloging of Works Intended for Performance," produced by CC:DA. My comments on main entry for recorded performances refer primarily to Yee's recent paper, but I will refer to the Task Force document from time to time. I'll begin by saying that I advocate title main entry for recorded performances, in any medium (theater, music, dance, mixed-media, etc.) This view stems in part from my background: formal study of theater (two academic degrees), an essentially lifelong involvement with performance in different media, continual experience as a reflective audience member, and ongoing informal study. I am not arguing against the concept of main entry, nor do I believe that proper respect for the work of playwrights, composers, etc. depends on it. I am also not advocating any "death-of-the-author" philosophy. I recognize that library users in most situations conventionally regard music and theater performances as being works by composers and playwrights, and there is no reason to disappoint them in this. There are several points that I take as axioms: 1) that a performance is different in its very nature from a written or printed document; 2) that audience members do not attend performances in order to read scripts or scores; 3) that code revision should not reify culture-specific notions about the nature of art forms; 4) that assigning main entry according to our standard hierarchies of value results in patent absurdities in too many instances; and 5) that we cannot ask catalogers, even those with academic backgrounds in the arts, to act as critics in order to determine main entry. 1) A performance is different in its nature from a script or score. I hope that, on the experiential level at least, it will take little argument to establish this. Performances are evanescent, scripts/scores are tangible (even electronic documents are tangible in a way performances are not). The acts of reading as compared with watching/listening involve the person in very different kinds of experience, including different sensory emphases, very different senses of social occasion, different opportunities for reflection, etc. Another way of saying this is that "intended for performance" is not "performed." (Main entry for works "intended for performance" should, of course, be under author! Recorded performances are something else.) Yee: "In AACR2R we have considered the following changes to be substantial enough to cause the creation of a new (but related) work (signalled by a change in main entry) ... adaptation of an art work from one medium to another (e.g. an engraving of a painting) ... " (p.7). If two 2-dimensional media can be recognized as sufficiently different to cause a change in main entry, how can a physical document and a live performance be considered the same work? The primary reason, apparently, is that the labors of composers and playwrights are regarded as having all performed realizations already latent within them from the moment of their creation. Other artists are only needed to serve as intermediaries for these already-existing performances. (This is not unrelated, perhaps, to the medieval view of sexual reproduction, in which the male seed contained the complete person in miniature, with the womb needed only as a convenient receptacle.) I'll mention some specific examples of the absurd consequences of this view later on. For now, let's just say that it ignores the evidence. One further point can be made regarding our discomfort with live performances as items in a library's collection. As Yee notes (p. 24), there are "three 'layers' of creative activity" involved in a recording of a performed script or score: a) the original writing, b) a performance, and c) a recording. The first and third are collectible, while the second is not. The fact that we can't get our hands on the performance itself may lead us to assimilate it to something we can touch, an "item in hand," a script or score. Instead, we should look at a recording as a surrogate for the vanished event, and put our attention on providing access to the event, as well as the surrogate will allow. Regarding the use of film (or audiotape, etc.) as "a mere recording medium": the lack of creative camerawork does not retroactively transform a performance back into a score. 2) Audience members do not attend performances in order to read scripts or scores, meaning that we do not attend expecting only to witness the work of the composer/author. This is true notwithstanding the occasional audience member who does indeed show up with a musical score. Of course, it's true that most people will come to see a performance of _Macbeth_ with the idea that it's a work by Shakespeare, and may very well believe that Peter Sellers's stagings of opera are indeed works by the respective composers. However, wouldn't you want your money back if all you were shown in a theater were holographic projections of scripts and scores? Less facetiously, what audience would there be for Wagnerian opera, if Wagner's original stagings were the only ones used -- or for Brecht, if all productions only followed his modelbooks? Audience members indeed come to witness the collaborative work of all involved, even if they don't realize the full extent of what that means, and even if they sometimes dislike the results. I'm not dissing audience members here. I'm one myself very often, and I frequently go because of the playwright or composer, knowing nothing about the performers. Different user group perceptions of what a performed work represents can easily be reconciled using title main entry, with added entries for the composer/authors, works performed, performers themselves, and so on. There is no need for mutually exclusive choices where access is concerned. I'd go further and say that the persistence of different user group perceptions indicates that a performance is at least two things at once: a work in itself, as well as one possible realization of another, already existing work. Taking Losey's film of _Don Giovanni_ as an example, we should not need to decide which group of users is right: those who believe it's Losey's work vs. Mozart's. Why should a cataloging code adjudicate this disagreement? Instead, realizing that the film represents both, do not privilege either, enter under title, and provide the added entries wanted by the user groups. 3) Code revision should not reify culture-specific notions about the nature of art forms. (I suppose somebody will mutter about "political correctness" at this point.) We have the opportunity to take a broader global/historical view of the varied relations between composing and performing, recording and enacting, improvisation and memory, tradition and innovation, than the current code allows. On p. 21, Yee notes that "even within the music field, there is general acknowledgement that the primacy of composition over performance is culture- specific, and functions best when applied to western classical music. Users are not as consistent in considering works of western popular or folk music, or non-western music of all kinds to be primarily the works of composers." This is more satisfying to me than the quote from a Task Force member which suggests that music consists of "a very detailed set of instructions for a performer," which defines out most of the music that has ever been made. (A house does not consist of a blueprint, nor are all habitations built from printed plans.) The principle of the primacy of writing over performance also causes difficulties for non-Western theater and dance forms, where it is often a nonsequitur. Would an hours-long dance drama based on a brief section of the _Mahabharata_ need to receive a uniform title for that section of the epic? Well, probably not, if no equivalent of a script with "detailed instructions for performance ... closely followed" could be located (per Task Force report, see Yee p. 23). But are we then creating a situation in which "author main entry" signals "works in the Western high-culture tradition of authorship?" Is this the kind of function for which main entry is intended? 4) Assigning main entry according to our standard hierarchies of value results in patent absurdities in too many instances. A few examples will have to do. Yee mentions Losey's film of _Don Giovanni_, but consider also Bergman's film of _The Magic Flute_. We cannot regard these films, "with frame composition, camera angles, cutting, etc." -- the "etc." including visual imagery, casting, placement of text -- as being inherent in the musical scores or even the composers' imaginations, without slipping into a muzzy sort of mysticism. The glories of these works exist to a great degree in that they have latent within them, not the performances themselves, but the power to inspire the creation of so many realizations, new works, "derivative" if you like (but not in a pejorative sense). Generations come back to these works over and over, and in collaboration they create new works, the value of the new works varying with the abilities of the collaborators. This, it seems to me, is an appropriate tribute to genius, that it can serve inspiration in this fashion. By contrast, not a single camera angle can be justified as being foreseen by any 18th- century composer. In AACR2R, "changes not [considered] to be substantial enough to cause the creation of a new work" include "providing a choreography for an existing musical work, such as a ballet (p. 7-8)." This gives one pause. How can a piece of music be considered a ballet? Is ballet not a dance form, and is music composed for it not generally called "ballet music", meaning written to accompany the dance? How are specific movements, overall patterning, alternation of solo and ensemble passages inherently contained in the music as such? How can anything like "detailed instructions" for movement be contained in a musical score proper? Rhythm alone is a weak indication of specific choreography, and new choreographies, such as a new _Sacre du Printemps_ are not bound by current practice to use composed rhythms at all. In general, the concept of musical composition taking precedence over choreography is untenable. Most Western choreography involves either commissioned music, or existing music not originally composed for dance. In the first instance, the dance is in development while the music is being composed, independently or in an active collaboration. (Merce Cunningham's company does not hear the music until the first performance.) If we must invoke hierarchies of precedence, we'd have to say that the music is the "servant" of the dance in most instances. But the question itself is obfuscating. Martha Graham's _Appalachian Spring_ was not composed by Aaron Copland, nor was Copland's _Appalachian Spring_ notated by Graham. In the case of new choreography to existing music, it makes no sense to regard the dance as the work of the composer. Balanchine choreographed his _Serenade_ using a work of Tchaikovsky's composed decades earlier. Must this dance receive a main entry for Tchaikovsky? We cannot pretend that this represents an informed understanding. 5) We cannot ask catalogers, even those with academic backgrounds in the arts, to act as critics in order to determine main entry. As the AACRCONF discussions progress, there are repeated calls to take into account the impacts of code revision on existing staff expertise and workflow. I'm generally leery of these cautions (wanting to distinguish the cart from the horse), but here such a warning is appropriate. If catalogers are required to determine main entry based on a) whether a printed document contains sufficiently detailed instructions for performance, and b) whether those instructions were faithfully carried out, there are two probable consequences. The first is that, if catalogers are actually true to these guidelines, they will need to become conversant with a whole host of specific practices of both authorship and performance, and make informed judgements on a case-by-case basis. This is in fact the business of criticism, not that of bibliographic description. The second probable consequence is that catalogers, having neither the time nor the inclination to do this, will simply apply main entry under author in most cases, assuming that there were in fact detailed instructions faithfully carried out. The latter consequence effectively ignores the guidelines and negates the discussion. Let's take two examples from dramatic literature. Shakespeare's plays, as is well known, do not include detailed instructions for performance, and therefore immediately fail the first test. And yet, productions of the plays are often hotly debated vis-a-vis their faithfulness to S's "intentions," which are objectively undiscoverable. The most diverse productions have been mounted over the centuries, by artists who deeply believed that they were being faithful to Shakespeare. What shall we do here? In the absence of detailed instructions in the scripts, we might try to salvage main entry under author by determining that, nevertheless, a given production carried out instructions which -might- have existed, making the production Shakespeare's work indeed. This will mean getting involved with textual criticism, social history, etc., for each and every videotape! Nearer our own time, we have the case of Gertrude Stein's plays. This isn't the place for the argument that these are indeed works intended for performance, but the case is not difficult to make. Few of these texts have stage directions, descriptions of settings, conventional act/scene divisions, or even character indications. There are no instructions whatever in most cases, just the text to be spoken. Here we can refer to the statement by a Task Force member, to the effect that some performances are "realizations of 'texts' that contain an inadequate set of instructions (and therefore a new work MUST be created with each performance) ... ". This statement misses an essential point, that many writers and composers have deliberately left out types of information that we conventionally expect. The term "inadequate" is inherently pejorative, and assumes that the writer/composer simply did a bad job of it. This needs to be proven -in each case-, not just asserted. Is it the descriptive cataloger's job to analyze a writer's failure? In short, asking the cataloger to determine how much improvisation is enough to create a new work, or what constitutes detailed instructions and exactness in execution, is inappropriate. It can be a fascinating study for one's leisure hours, or a full-time job for the scholar. On the job, it will result in a lot of questionable judgement calls and hair-splitting, if followed conscientiously. It's time to bring this message to an end, and my thanks if you've followed it to this point! There's more that can be said, and probably will be. There is one sensible exception to main entry of performances under title: the case where the composer/author(s) and the performer(s) are the same, meaning exact congruence. I have a recording of Bartok playing his piano works, which I'd enter under Bartok, not the title of the disc. He created the performances as well as the scores. I appreciate the opportunity that the JSC has provided for widespread input into an important discussion, and look forward to further participation. David Miller Levin Library, Curry College Milton, Mass. dmiller@curry.edu ========================================================================= Date: Mon, 18 Aug 1997 17:09:26 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Main entry for performances MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT David Miller promulgates these axioms: > > 1) that a performance is different in its very nature from a written or > printed document; > 2) that audience members do not attend performances in order to read scripts > or scores; > 3) that code revision should not reify culture-specific notions about the > nature of art forms; > 4) that assigning main entry according to our standard hierarchies of value > results in patent absurdities in too many instances; and > 5) that we cannot ask catalogers, even those with academic backgrounds in the > arts, to act as critics in order to determine main entry. > Much though there is to be said in favor of these statements, one should certainly not regard main entry rules as anything in the way of art criticism or reflecting a Western hierarchy of values. Catalogs are not, IMHO, vehicles to give credit for achievements, in the arts or elsewhere. Catalogs are finding tools for items and collocation tools for manifestations of expressions of works. It was expressed several times that OPACs mostly fall short of the latter, and that we therefore need the "work authority record". What other way is there to define this record but by using those elements that are the most stable and the most universally known and easy to look up items of data? These happen to be the name of the person who created the first manifestation of the abstract work, and the title it has come to be known by? What else are we going to use as identifiers for collocation? The title alone? It does not mean to give a composer undue credit when listing performances under his name, but it is a way to collocate performances which one might wish to compare but which, when recorded under performer, would be widely scattered. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 18 Aug 1997 08:36:42 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: 740 for single magazine issue Comments: To: lgallagh@bna.com In-Reply-To: <9707188719.AA871919165@smtpgate.bna.com> > Mac-- Are you sure 730 has $v (volume)? It's not in the > current ed. of OCLC Bibliographic Formats and Standards. > I was looking for $v but had to use $n instead. The Catss LHF has it. This is not the first difference between OCLC and Catss standards my postings have uncovered. I suppose one would have to appeal to USMARC as the final authority, which I don't have. But it doesn't matter much what USMARC has does it? One is dependent on what one's system will accept. I do have the OCLC manuals, but rarely consult them. If they don't like a field or subfield in a Catss record we are uploading into an OCLC account, the software lets us know. We find they object to many things which we consider quite legitimate (such as an 020 for a collection level record recently mentioned on this list). Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 18 Aug 1997 12:37:07 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Re: Main entry for performances Comments: To: B.Eversberg@tu-bs.de The need for collocation between a script/score and derivative works, including performances, is most important. I am concerned about the concept of "work" as "the abstract work, and the title it has come to be known by," though. If I read Vellucci's paper correctly, this is how the IFLA document understands "works" -- as something which exists prior to any incarnation. The problem here is that such an abstract "work" is impossible to delimit. Again, this is the job of criticism, not of cataloging. What is the abstract _King Lear_? Critics have been working on that for centuries, and will continue to do so. That's why _King Lear_ is important. Look at it another way: works which are easily circumscribed ideationally are works with no generative potential -- i.e. are likely to spawn little in the way of bibliographic families (from Smiraglia). If you can understand an abstract "work" without experiencing it in some concrete manifestation, there's not that much there to begin with. Any nursery rhyme has more generative potential than that. I haven't yet got a solid handle on the "super works" concept, but it seems to me that this is where a script/score/notation and its derivative works may be collocated -- at the authorities level. On the bib. record level, author/title added entries can be indeed tied to work records -- records that pertain to -what the writer actually created-. David Miller Levin Library, Curry College Milton, MA dmiller@curry.edu ========================================================================= Date: Tue, 19 Aug 1997 14:17:28 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Main entry for performances MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT David Miller writes: > The need for collocation between a script/score and derivative works, > including performances, is most important. I am concerned about the > concept of "work" as "the abstract work, and the title it has come to > be known by," though. If I read Vellucci's paper correctly, this is > how the IFLA document understands "works" -- as something which exists > prior to any incarnation. The problem here is that such an abstract "work" > is impossible to delimit. Again, this is the job of criticism, not of > cataloging. Exactly. The job of cataloging can only be to determine the "title it has come to be known by", nothing more and nothing less. This can and must be based on bibliographic evidence alone, not on critical evaluation, not on philosophical reasoning, not on literary analysis. The result is not one that is supposed to do anybody "justice" in giving them credit, and neither is it supposed to accurately reflect results of scholarly research, the result is only supposed to facilitate locating and collocating publications. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Tue, 19 Aug 1997 08:28:49 EST Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Neil R. Hughes" Organization: University of Georgia Libraries Subject: Re: Main entry for performances I find no new arguments in David Miller's original posting re: "Main entry for performances," though it is undoubtedly one of the most cogent explications of the interesting assumption that anything other than title main-entry somehow requires too much subjectivity and arcane knowledge on the part of the cataloguer. My further observations follow: > There are several points that I take as axioms: > 1) that a performance is different in its very nature from a written or > printed document; No argument here. > 2) that audience members do not attend performances in order to read scripts > or scores; Nor here. > 3) that code revision should not reify culture-specific notions about the > nature of art forms; And how would one avoid that? It is "reifying" culture-specific notions (questionably-applied egalitarianism; one-size-fits-all = optimal service; etc.) to advocate the elimination of the hierarchy that most users demonstrably bring to their library searches for Western classical music. > 4) that assigning main entry according to our standard hierarchies of value > results in patent absurdities in too many instances; and "Too many instances" is too vague to support the argument; be specific, please. I too can show this--applying the principle of composer main-entry to Karnatic music, to name one instance--but I believe the arguments below fail to demonstrate any absurdity in the application of composer main entry to Losey's film of Mozart's Don Giovanni or Bergman's of Mozart's Zauberflote. They merely support a statement that it is arbitrary to do so, and perhaps that it is also harsh from a librarian's standpoint to do so. But they do not show that access is adversely affected by composer main-entry in such cases, or that a better citation form or list-sorting in a catalogue can be achieved by adopting title main-entry. > 5) that we cannot ask catalogers, even those with academic backgrounds in the > arts, to act as critics in order to determine main entry. Criticism is not necessary; only knowledge of the code that tells one what to do in such-and-such an instance. I believe that expanding the current code to allow for more specific guidance for non-Western, non-"art music" genres is all that is necessary. One is still left with questions of the structure of the authority file and the representation of "the work," though. While the claim that Losey's film of Don Giovanni is a new work is something with which I might even agree, I question the need for a name/uniform-title NAR to represent it unless someone reissues it with animated segments, or a dubbed soundtrack sung in English, etc. etc. Name/uniform-title NARs and the need for them for given works are fundamental to the question of what should constitute main entry, at least where music is concerned. As a person who has lived & breathed "classical music" since the age of five, I apprehend one salient fact clearly: Mozart's work, Don Giovanni, is present, discrete and complete, in Losey's film and constitutes the majority of the intellectual content of Losey's film, whatever other riches may orbit there resulting in a second, new work. (Having seen the lengthy trailer issued prior to the release of that wonderful film and heard Losey himself speak reverently of the Mozart work, I have little doubt that he'd agree with me, though sadly he's no longer around to respond.) > 1) A performance is different in its nature from a script or > score. I hope that, on the experiential level at least, it will > take little argument to establish this. Performances are > evanescent, scripts/scores are tangible (even electronic documents > are tangible in a way performances are not). The acts of reading as > compared with watching/listening involve the person in very > different kinds of experience, including different sensory > emphases, very different senses of social occasion, different > opportunities for reflection, etc. Another way of saying this is > that "intended for performance" is not "performed." (Main entry for > works "intended for performance" should, of course, be under > author! Recorded performances are something else.) I return again & again (_ad nauseam_, no doubt) to my point that experts, dilettantes, aficionados, lovers and all cognoscenti of Western classical music would simply state that the preceding statement is baseless and that a performance of a work is _exactly_ the same thing as a work intended for performance; in fact, it is a more complete and accurate representation of that work. So I say let us build our catalogues to reflect arbitrary tradition, rather than adopting an untested and iconoclastic view (which Mr. Miller supports only by insistence on a personal approach to these works, rather than extended observation of users' approach to the catalogue--either card or online--in the music library, which is what I claim as support for my position). Can we really assume that title main-entry would be better in all cases merely because it is better for some? That is the classic Procrustean bed. > > Yee: "In AACR2R we have considered the following changes to be substantial > enough to cause the creation of a new (but related) work (signalled by a > change in main entry) ... adaptation of an art work from one medium to another > (e.g. an engraving of a painting) ... " (p.7). If two 2-dimensional media can > be recognized as sufficiently different to cause a change in main entry, how > can a physical document and a live performance be considered the same work? > The primary reason, apparently, is that the labors of composers and > playwrights are regarded as having all performed realizations already latent > within them from the moment of their creation. Other artists are only needed > to serve as intermediaries for these already-existing performances. (This is > not unrelated, perhaps, to the medieval view of sexual reproduction, in which > the male seed contained the complete person in miniature, with the womb needed > only as a convenient receptacle.) I'll mention some specific examples of the > absurd consequences of this view later on. For now, let's just say that it > ignores the evidence. I contend that it is not up to catalogue librarians to change users' views, medieval or otherwise. Only to construct catalogues that serve their needs. > Regarding the use of film (or audiotape, etc.) as "a mere recording > medium": the lack of creative camerawork does not retroactively transform a > performance back into a score. ... nor does the presence of creative camera work make a given performance any less the complete realization of the composer's work, as I said re: Losey's film above--or make it more the "work" of someone else, unless we are prepared to create authority records for the work of each of the camera operators, the gaffer, the executive producer, and the caterer, with 510s all linking them to all other records related to the item in hand. > > 2) Audience members do not attend performances in order to read scripts or > scores, meaning that we do not attend expecting only to witness the work of > the composer/author. No one has ever said that they did. This is why the code allows for performer main entry in certain instances, but subordinates it to cases where the recorded music is (for example) all by one composer. This is arbitrary and convenient, and reflects no judgments, moral or otherwise, on the importance of the performer's contribution, which I believe almost everyone is prepared to acknowledge is extremely important, if not equal, in either case. > This is true notwithstanding the occasional audience > member who does indeed show up with a musical score. Of course, it's true that > most people will come to see a performance of _Macbeth_ with the idea that > it's a work by Shakespeare, and may very well believe that Peter Sellers's > stagings of opera are indeed works by the respective composers. However, > wouldn't you want your money back if all you were shown in a theater were > holographic projections of scripts and scores? Less facetiously, what audience > would there be for Wagnerian opera, if Wagner's original stagings were the > only ones used -- or for Brecht, if all productions only followed his > modelbooks? Audience members indeed come to witness the collaborative work of > all involved, even if they don't realize the full extent of what that means, > and even if they sometimes dislike the results. And what has that to do with how they approach searching for an edition of the play in question when they go to the library? Do we have _any_ evidence that they look for Brecht's plays on videocassette differently than they do the same entities in print? > > I'm not dissing audience members here. I'm one myself very often, and I > frequently go because of the playwright or composer, knowing nothing about the > performers. Different user group perceptions of what a performed work > represents can easily be reconciled using title main entry, with added entries > for the composer/authors, works performed, performers themselves, and so on. > There is no need for mutually exclusive choices where access is concerned. Again I must insist that this statement is fundamentally inaccurate, at least where Western classical music is concerned, and I wish I had some extensive transaction logs from academic or conservatory music libraries' online catalogues handy to prove my point. As is so often the case, we are wrestling with a very important question for which there is little but anecdotal evidence and supposition on either side of the argument, but my observations & experience are not imaginary. There is a demonstrable need for mutually exclusive choices, and the more specific they are to the type of work in question, the better-served the user is. The extant rules work well for Western classical music, most of the time. Let's offer some choices for other musics, and start reflecting the realities of our more diverse curricula and expanding world views in the arts, but let's not do that by offering a macaroni-&-cheese non-choice of title main-entry for everything. If we do abdicate responsibility for deciding what constitutes the work and apply title main-entry to all such items, the only viable solution to the resulting confusion for the users may be what Martha Yee proposes on p. 25-26 of her paper, where she proposes that one be allowed to choose among list displays matching one's personal approach to the work: 1. Editions of [work] 2. Works containing [work] 3. Performances of [work] 4. Works about [work] 5. Other works related to [work] Where she and I differ re: this proposal is where the entry for Mozart's Don Giovanni as it pertains to the Losey film would be found in choices 1-5. She claims it would necessarily be found under 5, "Other works related to ... ," whereas I claim it could only properly reside under 2, "Works containing ... ," because of my intractable view that a performance of a musical work equates to the work and that Mozart's opera is discrete and present within the Losey film. This is an important distinction, and we would have to decide arbitrarily that it is one or the other to make such choices viable in our online catalogues. (Does one tag the added entry 700 1[blank] for a related work, or 700 12, for an analytic entry?) But I stray far afield from discussions of AACR again.... > > I'd go further and say that the persistence of different user group > perceptions indicates that a performance is at least two things at once: a > work in itself, as well as one possible realization of another, already > existing work. Taking Losey's film of _Don Giovanni_ as an example, we should > not need to decide which group of users is right: those who believe it's > Losey's work vs. Mozart's. Why should a cataloging code adjudicate this > disagreement? The code does not now adjudicate any such disagreement; it merely reflects a cultural tradition in such a way that logical relationships among records may be built up in a hierarchical mode, rather than a flat, Mozart's-o.k., Losey's o.k., we're-all-equals-here amorphous loop of authority and bib. records. If the code happens to reflect a disagreement too, I'd say let's work on that--if it affects access. But again, is it necessary to eliminate helpful choices? > Instead, realizing that the film represents both, do not > privilege either, enter under title, and provide the added entries wanted by > the user groups. This statement illustrates what I mentioned in my earlier posting about the arguments re: definition of a work and the concept of main entry for performances being reduced to a popularity contest. We are not currently "privileging" anyone or anything by making an arbitrary choice for main entry. I claim we are assisting more users by retaining composer (or when appropriate, performer) main entry for "performances of works intended for performance" than we would be by eliminating these choices. We are retaining more manageable authority and bibliographic files, and we are harming no one. We do not deny legitimate changes to The Canon; we do not halt the progress of a society that has long recognized Bergman, Losey, or the author of every song text who ever drew breath as an artist in his/her own right in so doing. We are merely bringing order to bibliographic chaos from a logical basis that will have to change as our users' conceptualizations of works change, but I do not think it is up to us to effect those changes through the design of our catalogues. > > Yee mentions Losey's film of _Don Giovanni_, but consider also Bergman's film > of _The Magic Flute_. We cannot regard these films, "with frame composition, > camera angles, cutting, etc." -- the "etc." including visual imagery, casting, > placement of text -- as being inherent in the musical scores or even the > composers' imaginations, without slipping into a muzzy sort of mysticism. The > glories of these works exist to a great degree in that they have latent within > them, not the performances themselves, but the power to inspire the creation > of so many realizations, new works, "derivative" if you like (but not in a > pejorative sense). Generations come back to these works over and over, and in > collaboration they create new works, the value of the new works varying with > the abilities of the collaborators. This, it seems to me, is an appropriate > tribute to genius, that it can serve inspiration in this fashion. By contrast, > not a single camera angle can be justified as being foreseen by any 18th- > century composer. Nor could Mozart have imagined the _actual_ first performance of either of these operas, nor any subsequent production or performance. This may be support for the notion that composer main-entry is politically incorrect, but not for the idea that it's bad access in a catalogue, nor for the arbitrary definition of the work one has in hand. > > In AACR2R, "changes not [considered] to be substantial enough to cause the > creation of a new work" include "providing a choreography for an existing > musical work, such as a ballet (p. 7-8)." This gives one pause. How can a > piece of music be considered a ballet? Is ballet not a dance form, and is > music composed for it not generally called "ballet music", meaning written to > accompany the dance? How are specific movements, overall patterning, > alternation of solo and ensemble passages inherently contained in the music as > such? How can anything like "detailed instructions" for movement be contained > in a musical score proper? Rhythm alone is a weak indication of specific > choreography, and new choreographies, such as a new _Sacre du Printemps_ are > not bound by current practice to use composed rhythms at all. There are now uniform titles for choreographic works parallelling those for musical works (which may or may not be known in conjunction with a given choreographic work in a given instance). While related-work links among these records have not, in my opinion, been adequately explored, recognition that there are in all cases at least two works here (and not one new one, representing a synthesis) is an important step in the right direction. Dance music illustrates vividly a sub-area (again, within the Western traditions) where one may see--even in a music library!--two distinct approaches to works by users, and I am in complete agreement with the present solution to this, whether I personally think the music is more important than the dance or not. It reflects our users choices. > In short, asking the cataloger to determine how much improvisation is enough > to create a new work, or what constitutes detailed instructions and exactness > in execution, is inappropriate. It can be a fascinating study for one's > leisure hours, or a full-time job for the scholar. On the job, it will result > in a lot of questionable judgement calls and hair-splitting, if followed > conscientiously. A good description of the art of good descriptive cataloging over the centuries, if I've ever heard one, but I think a _non-sequitur_ in the present argument. <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> Neil R. Hughes Music Cataloger University of Georgia Libraries Internet: nhughes@libris.libs.uga.edu Telephone: (706) 542-1554 Fax: (706) 542-4144 ========================================================================= Date: Tue, 19 Aug 1997 08:47:17 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Main entry for performances Comments: To: EV@BUCH.BIBLIO.ETC.TU-BS.DE In-Reply-To: <7CDB06839CD@buch.biblio.etc.tu-bs.de> >Exactly. The job of cataloging can only be to determine the "title it has >come to be known by", nothing more and nothing less. This can and must >be based on bibliographic evidence alone, not on critical evaluation, >not on philosophical reasoning, not on literary analysis. The result is >not one that is supposed to do anybody "justice" in giving them credit, >and neither is it supposed to accurately reflect results of scholarly >research, the result is only supposed to facilitate locating and collocating >publications. This is the best argument for series title entry in 440 I have ever seen, as opposed to duplicating the 100 in an 800, with the title (which is what my patrons are seeking) in often poorly indexed 800$t. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Tue, 19 Aug 1997 14:02:36 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Stephen Hearn Subject: composer main entry Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" I've found the discussion of David Miller's argument for entering performances under title only fascinating, but parts of it and the responses to it have been a bit too abstract. So I offer a few examples which are helping me focus my attention on the choices involved and how they would affect a hierarchically ordered catalog. The advantage of a "main entry" is that it offers a specially identified data element for use in constructing more complex index entries. Considered by itself, its status as the "main" entry is pretty meaningless; it only becomes meaningful when it is used to sort, or more importantly subsort a list of headings. And bear in mind that we are not talking about the terms someone would search; we're talking about the complex, sorted entries which their simple search will allow them to browse, and which AACR makes possible. Taking music-in-performance as an example, let's consider how someone would like to find the entries under a performer's name, say Leonard Bernstein. If one consistently assigns a composer name main entry to Bernstein's recordings as a conductor, one can use that main entry to subsort the entries under Bernstein by composer name, and then title; e.g., Bernstein... (a.e.) Beethoven, Ludwig... (m.e.) [Concertos, orchestra, piano] Bernstein, Leonard... (m.e.) Candide... Copland, Aaron... (m.e.) [Symphonies, no. 3] Wagner, Richard... (m.e.) Tristan und Isolde This follows the filing order called for by Library of Congress Filing Rules (1980, p. 86). If the three composer names above, with or without their uniform titles, become added entries and indistinguishable from other name entries on the record, it's hard to imagine how they could be used for subsorting. Hence, a rule calling for title main entries for all performances would result in something simpler, but less explicit about what the titles (title proper, not uniform titles this time) represent: Bernstein... (a.e.) 5 piano concertos... (m.e.) Candide... (m.e.) Third symphony... (m.e.) Tristan und Isolde... (m.e.) The problem of sort order in real files/indexes is of course much more complex than this, and usually not well handled in the case of uniform titles. However, I think these examples illustrate one of the options that would be lost in going to title main entry only for performances in AACR. Stephen Hearn E-mail: s-hear@tc.umn.edu Authority Control Coordinator Phone: 612-625-2328 University of Minnesota Fax: 612-625-3428 ========================================================================= Date: Tue, 19 Aug 1997 15:55:08 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Jennifer Kolmes Subject: Re: Main entry for performances In-Reply-To: <539A5AA60A4@libris.libs.uga.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In response to David Miller's remarks about main entry for performance, it seems to me that the best argument against applying such a practice to classical music is practical rather than theoretic. Most classical works do not have nice specifc titles such as _Don Giovanni_, but instead have hopelessly unspecific titles such as "Symphony number 1 in G major." Such a title is of course meaningless without knowledge of the composer. Aside from following tradition established with books, this is why the tradition of main entry by composer is so firmly established with Western classical music. You'd need a lot better arguments than any presented by Miller to convince me of the wisdom of changing this practice, at least as far as classical music is concerned. Jennifer Kolmes Head of Cataloging University of Arkansas Fayetteville, AR 72701 ========================================================================= Date: Tue, 19 Aug 1997 16:13:18 EST Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Neil R. Hughes" Organization: University of Georgia Libraries Subject: Re: composer main entry To Stephen Hearn's first and preferred list-sort I would add the following important element, i.e. a further subsort by titles-proper under each of the composers+uniform titles listed there. This is a primary example of a portion of what Ralph Papakhian was pointing out on the list last week, i.e. that our current catalogues are not based on the current code and its intentions re: access points. I have inserted some made-up (but quite possible!) examples below, under the entries for Beethoven, Copland, and Wagner to show how the sort could be made even more useful. (I've included only a portion of each title-proper; inclusion of other-title information in most cases would make the list truly useful, as others have pointed out. I also took the liberty of correcting the order of the elements in the Beethoven uniform title, which are based on the arbitrary relationship of solo instrument-accompaniment rather than the alphabet.) A list such as I advocate would necessarily include all added entries for composers+uniform titles as well, but let's say for now that in this mock online catalogue, added entries are listed separately after main entries and we just haven't chosen to show any here. Be warned though--my imaginary titles-proper in all cases except the Wagner example represent scores, recordings, videocassettes, and study kits! > Bernstein... (a.e.) > Beethoven, Ludwig... (m.e.) > [Concertos, piano, orchestra] 5 concerti per pianoforte e orchestra The complete piano concertos Concerts de piano integrale Five masterworks of the keyboard : ... Die funf Klavierkonzerte Piano concertos 1-5 > Bernstein, Leonard... (m.e.) > Candide... > Copland, Aaron... (m.e.) > [Symphonies, no. 3] Copland's third symphony in facsimile Dritte Sinfonie Symphonies no. 3 & no. 2 ("Short Symphony") Symphony no. 3 > Wagner, Richard... (m.e.) > Tristan und Isolde Wagner, Richard... (m.e.) [Tristan und Isolde. Vocal score. English & German] The romance of Tristan and Isolde Tristan and Isolde Tristan und Isolde = Tristan and Isolde ... etc. etc. This is something that libraries with music collections have largely given up with the advent of online catalogues and the demise of cards, and there is nothing like it for bringing together all manifestations of a work (as we now define "work") held by a given library. I personally think it was too much to give up, even in exchange for keyword access, which can be a real time-waster if your collection is large enough! I realize that Mr. Hearn's list was such that one presumably could have selected the Beethoven uniform title and then gone to a list of all titles-proper connected to it; I prefer to have it all up front, as I did here. My scrolling tolerance is high (witness my e-mail messages). But in either case, those who say we can't have list displays of more than two lines per item because our users are too dim to take it all in have managed to make their cries be heard more loudly than those who value Cutter's objects or the Paris principles! <><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><><> Neil R. Hughes Music Cataloger University of Georgia Libraries Internet: nhughes@libris.libs.uga.edu Telephone: (706) 542-1554 Fax: (706) 542-4144 ========================================================================= Date: Tue, 19 Aug 1997 16:14:46 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Re: composer main entry Stephen Hearn has helped with a point that was puzzling me: why did there seem to be such resistance to name/title added entries for composers and playwrights (etc.), if main entry for recorded performances was given to the title? Added entries should certainly serve to collocate the original works, as realized in performances -- but the subsorting issue is important and makes the discussion more complex. I'm still working on Neil Hughes' posting from this morning, but at the moment I'll throw out this: what technical reasons necessarily prevent adequate subsorting via added entries? We could talk about present limitations of MARC and specific automated systems -- but looking further ahead, must subsorting be confined to the main entry? I've urged title main entry for performances pretty vigorously, but I don't assume that this is a good idea given the status quo. Part of what we're doing here is to imagine what we want, without sacrificing the good that we have. Intelligible subsorting is important: how can we facilitate this without misrepresenting intellectual/creative responsibility for performances? Thanks, all -- David Miller Levin Library, Curry College dmiller@curry.edu ========================================================================= Date: Tue, 19 Aug 1997 20:56:29 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: M.E. for performances: reply to Neil Hughes Thanks to Neil Hughes for giving me a good tough bone to chew on! Before replying to some of his points, I'll make a couple of preliminary comments. The first is that my assertions have nothing to do with frustrating any existing users' expectations where searching and display of headings are concerned, nor with changing their minds about music. ("Descriptive Cataloger Holds Music Library Patrons Hostage in Re-Education Session") I am concerned with what main entry designates, apart from its importance for collocation. Given the implications of the conference papers I've read so far, and the discussions that have gone on here, we certainly have room to imagine how adequate collocation of headings can be brought about without relying on a chain of main entries. The second prelim. is that my original posting may have been somewhat skewed vis-a-vis Martha Yee's conference paper, in that I was also responding to the recommendations of the Task Force on the Cataloging of Works Intended for Performance. That draft document does receive mention in Martha's paper, but naturally isn't her primary subject. So, though my posting was directed at the present paper, my comments were colored by the draft document, and in particular by the following recommendation (quoted by Yee, p. 23): "Realizations of pre-existing texts which consist of instructions for performance should be entered as follows: If the instructions are detailed, and if they are closely followed in the performance, the performance should be considered the same work as the pre- existing text, and creation of the text should be considered the primary function in the creation of the work, with performance being secondary. If, however, the original instructions are not detailed and/or are not closely followed in the performance, the performance should be considered a new work, but one related to the pre-existing text; i.e. if improvisation and/or adaptation and/or creative or intellectual work beyond mere performance occur, the performance should be considered a new work. This new work would be entered under title, unless there are only two authorship functions involved, and a more specific rule assigns primacy to one of the functions." My advocacy of title main entry is not based on this recommendation, by the way. I still believe that conscientious application of this recommendation will indeed require catalogers to serve as art critics in a sufficient number of instances to bring most departments to a dead halt. My specific comments: "(NH) I find no new arguments in David Miller's original posting re: 'Main entry for performances,' though it is undoubtedly one of the most cogent explications of the interesting assumption that anything other than title main-entry somehow requires too much subjectivity and arcane knowledge on the part of the cataloguer." This overstates my case. I have a disc titled _Rudolf Serkin plays Beethoven_. If I need to assign personal-name main entry, I really only have two choices - - neither of which involve arcane knowledge. The question is, does either of them represent the work accurately? As Fattahi puts it (p. 13): "Main entry is a uniform construct for the naming and identifying of works and also for the useful collocation of arrangement/display of the different expressions and manifestations of a work." It's not simply a matter of facilitating file structure according to an amiable, established convention -- but of delimiting where one work ends and another begins. " (DM) > 5) that we cannot ask catalogers, even those with academic backgrounds in the > arts, to act as critics in order to determine main entry. (NH) Criticism is not necessary; only knowledge of the code that tells one what to do in such-and-such an instance. [ ... ] As a person who has lived & breathed "classical music" since the age of five, I apprehend one salient fact clearly: Mozart's work, Don Giovanni, is present, discrete and complete, in Losey's film and constitutes the majority of the intellectual content of Losey's film, whatever other riches may orbit there resulting in a second, new work. (Having seen the lengthy trailer issued prior to the release of that wonderful film and heard Losey himself speak reverently of the Mozart work, I have little doubt that he'd agree with me, though sadly he's no longer around to respond.) [ ... ] > (DM) Taking Losey's film of _Don Giovanni_ as an example, we should > not need to decide which group of users is right: those who believe it's > Losey's work vs. Mozart's. Why should a cataloging code adjudicate this > disagreement? (NH) The code does not now adjudicate any such disagreement; it merely reflects a cultural tradition in such a way that logical relationships among records may be built up in a hierarchical mode, rather than a flat, Mozart's-o.k., Losey's o.k., we're-all-equals-here amorphous loop of authority and bib. records." Let's accept, for the sake of argument, that Losey's film is indeed a sufficient fulfillment of Mozart's composition, that it should be given composer main entry. Let's imagine that other opera cognoscenti feel differently, and deny the film's status as an adequate representation. Let's also imagine, please, that some film connoisseurs take a converse view: that Mozart's composition has only served as the vehicle for a consummate Losey work. (None of these hypothetical people may actually be found, but analogous situations will arise.) If we provide main entry under Mozart, we have indeed declared whose work it is primarily. More, we have declared that the film does in fact fulfill Mozart's work. And, though this wasn't the intent of Neil's comment, he does display personal critical expertise and relevant specific experience (viewing the trailer, hearing the filmmaker speak) to back this judgement. Again, composer main entry is not simply a convenient element for structuring a file. In arguably ambiguous cases, assignments of primary intellectual/artistic responsibility do involve either the exercise of a critical function, or the abdication of it under the guise of following a convenient rule. " (DM) > Another way of saying this is > that "intended for performance" is not "performed." (Main entry for > works "intended for performance" should, of course, be under > author! Recorded performances are something else.) (NH) I return again & again (_ad nauseam_, no doubt) to my point that experts, dilettantes, aficionados, lovers and all cognoscenti of Western classical music would simply state that the preceding statement is baseless and that a performance of a work is _exactly_ the same thing as a work intended for performance; in fact, it is a more complete and accurate representation of that work." I am sure I understand what Neil means by this -- I'm not such a materialist as all that. My life with Western classical music also goes back to preschool, for what that's worth. Still, I will now exasperate everyone by asking: what, then, is a "thing?" For the purpose of constructing a library catalog, I must regard a physical artifact (the composer's manuscript, a CD-ROM compilation of comparative editions) as a different thing from an event which took place once, has now passed, but has left traces on a recording medium. If I'm going to provide the kinds of enhanced links between works that many are calling for, I need first to separate thing from thing. In an undifferentiated universe, we need no links, but neither can we act or make decisions. Polymorphous perversity is paradisiacal indeed! " (DM) > Regarding the use of film (or audiotape, etc.) as 'a mere recording > medium': the lack of creative camerawork does not retroactively transform a > performance back into a score. (NH) ... nor does the presence of creative camera work make a given performance any less the complete realization of the composer's work, as I said re: Losey's film above--or make it more the 'work' of someone else, unless we are prepared to create authority records for the work of each of the camera operators, the gaffer, the executive producer, and the caterer, with 510s all linking them to all other records related to the item in hand." The first part of Neil's reply, again, assumes sufficient critical knowledge to make an accurate judgement (assuming that we agree on what a "complete realization" is.) The second part is an overstatement. We do not now create added entries for the persons mentioned, nor would title main entry mandate that. " (DM) > Audience members indeed come to witness the collaborative work of > all involved, even if they don't realize the full extent of what that means, > and even if they sometimes dislike the results. (NH) And what has that to do with how they approach searching for an edition of the play in question when they go to the library? Do we have _any_ evidence that they look for Brecht's plays on videocassette differently than they do the same entities in print?" No, I don't, but I mention this aspect of audience psychology in terms of what the "work" is that people come to witness, and what that means for naming. Of course, they should be able to find the plays via author-title entries. "(NH) The extant rules work well for Western classical music, most of the time. Let's offer some choices for other musics, and start reflecting the realities of our more diverse curricula and expanding world views in the arts, but let's not do that by offering a macaroni-&-cheese non-choice of title main-entry for everything. [ and earlier ... ] Can we really assume that title main-entry would be better in all cases merely because it is better for some? That is the classic Procrustean bed." For what, besides Western classical music, do the extant rules work? Not, apparently, for the most of the rest of the world's music. The rules work about as well for theater -- that is, for work in the Western mainstream and little else. Composer main entry makes no sense for dance, and title main entry is already the norm for film. And then there's the entire, sprawling area of intermedia and multimedia ... It begins to seem as though playwright/composer main entry may be Procrustes' mattress of choice. Dare I suggest that Western classical music and mainstream theater are the exceptional situations? "(NH) If we do abdicate responsibility for deciding what constitutes the work and apply title main-entry to all such items, the only viable solution to the resulting confusion for the users may be what Martha Yee proposes on p. 25-26 of her paper, where she proposes that one be allowed to choose among list displays matching one's personal approach to the work: 1. Editions of [work] 2. Works containing [work] 3. Performances of [work] 4. Works about [work] 5. Other works related to [work]" I won't quibble with the phrase "abdicate responsibility" here, but want to say that I think Yee's proposed structure of entries is excellent, and not simply as a way out of Bedlam! Much current thinking is converging toward this kind of structured display, roughly analogous to the collapsed display of subject heading subdivisions according to type, by the way. To what degree will coding at the field level accomplish this, as compared with links at the "super record" level? I find this fascinating. "(NH) There are now uniform titles for choreographic works parallelling those for musical works (which may or may not be known in conjunction with a given choreographic work in a given instance)." Yes, and I wonder what these are supposed to designate. Will a uniform title for _Giselle_ stand for all possible choreographies of _Giselle_ -- and if so, what are they assumed to have in common, besides, presumably, the same piece of music? Again, what is being named here? Over to you! David Miller Levin Library, Curry College dmiller@curry.edu ========================================================================= Date: Wed, 20 Aug 1997 01:09:39 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Re: composer main entry -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of David P. Miller Sent: Tuesday, August 19, 1997 4:15 PM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Re: composer main entry I'm still working on Neil Hughes' posting from this morning, but at the moment I'll throw out this: what technical reasons necessarily prevent adequate subsorting via added entries? We could talk about present limitations of MARC and specific automated systems -- but looking further ahead, must subsorting be confined to the main entry? [] Would not the term "main" entry imply that this is only one of many ways of creating a heading for an item already? A first among equals, if you like, with no restriction on subsorting, ultimately, in an automated environment. As a case in point, are not alternate headings in a MARC authority record coded exactly as if they were the main heading. My take on this issue is that main entry headings must always be constructed because we must always assume a single entry limitation at some point or another. Book catalogs, cutters for shelf arrangement, brief summary displays, citation forms -- all come to mind. In other words, main entry to me means that if there was one, and absolutely only one way to create a heading for a work, fulfilling the needs of unique identification and efficient collocation, what would that form be? The right answer, surely cannot be, oh, let the computer take care of it. The right answer, surely cannot be that this is too much of an intellectual exercise, better suited for literary criticism. I see uniform titles, main entry headings, and work records as all intricately intertwined. Perhaps these specific functions are >>too<< intertwined, with overlapping fields, partly because of the slight card-catalog bias in AACR, but mostly because of the overt card printing focus of MARC. The unique work heading consisting of the 100 name plus 240 uniform title or 245 title proper are poorly integrated and often do not meet the needs of computerized systems, which understand only hard logic and clear separations of all data into discrete, simple units (at least without resorting to complicated spaghetti code). This is especially so when these name and title elements are joined in name-title form when a reference to a work is required, creating a mold of inconsistency that replicates problems throughout our automated environments. The discussion on musical works also extends to series, where I would strongly disagree with J. McRee Elrod. The only good main entry heading is a consistent clear one, independent of as many crutches and extra baggage as possible. If a series heading had to stand alone, independent of all other data, what would that form be? Creating inconsistency in the treatment of main entry, which applies to works, independent of considerations of physical manifestations (the prime directive of Part 2 of AACR) will certainly create as much labor as one might think would be eliminated. A name-title series heading generates an authority record. The heading on this authority record is the same one that can be used as a related work or as a subject heading, where the 100 author would certainly not be the author responsible for the series. It is precisely for this reason that a mutilated title-only series heading would not suffice, as collocation in a "lowest common denominator" catalog would be obliterated, with the link severed from the original series author to all related works. (Alternatively, we could create subject added entries for every item in the series in name-title form, but I think it is certainly in the public understanding that a series title is a fundamentally different concept than just a slight variation on the title of a part of the series. As an example, I recently created a related added entry for Isaac Asimov's Foundation Trilogy, a work, certainly, in its own right, despite its physical manifestations into separate parts. And it sure is nice to have this kind of collocation under Asimov's name, which certainly would not be the case in a traditional card catalog with main entry titles for all series. Why have two or three extra degrees of separation when there is no compelling discomfort in name-title series headings, bringing all relevant works together?) Again, we can assume complicated programmed arrangements of mixing and matching heading elements to remedy choices for headings, but certainly one of the hallmark achievements of AACR has been its ability, at the >>very least<< to bridge a simple, single entry world with the complex multi-index, multi-keyword world of automation. Whatever changes are made, this absolutely must not be lost. Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Wed, 20 Aug 1997 11:00:04 +0900 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Mervyn Islip Organization: University of South Australia Subject: Re: composer main entry MIME-version: 1.0 Content-type: text/plain; charset="US-ASCII" Content-transfer-encoding: 7BIT > The problem of sort order in real files/indexes is of course much more > complex than this, and usually not well handled in the case of uniform > titles. However, I think these examples illustrate one of the options that > would be lost in going to title main entry only for performances in AACR. > > > > > Stephen Hearn E-mail: s-hear@tc.umn.edu > Authority Control Coordinator Phone: 612-625-2328 > University of Minnesota Fax: 612-625-3428 > While I appreciate the special problems of music and other specialist libraries, AACR2 was designed for "use in the construction of catalogues and other lists in general libraries of all sizes. They are not specifically intended for specialist and archival libraries..." AARC2R 0.1 (p. 1). In revising the code I believe this should be kept in mind. Therefore, I would like to advocate abolishing main entry, not as a concept, but in the construction of the bibliographic record. What I am suggesting is that the headings for the author(s) associated with a work be listed together in the record, rather than separated into main and added entries. In MARC this would probably mean authors, etc. would only appear in 7xx tags (assuming 7xx and not 1xx was chosen as the tag for this purpose.) However, I recognize the problem that this would create for uniform titles, particularly for music and literature. One way around this would be to provide a name-uniform title added entry where required, although I realize this may not allow for the complex sort order displays outlined in discussion so far. In MARC the 100/240 for uniform titles would be replaced with a 7xx name-title added entry. This may seem a radical idea, but I am sure I have heard it before somewhere, and I think we should be considering all options in revision of the code. Mervyn Islip Senior Cataloguing Librarian University of South Australia Email: mervyn.islip@unisa.edu.au Library Phone: +61 8 8302 6722 Holbrooks Road Fax: +61 8 8302 6756 Underdale, S. Aust. 5032 Australia ========================================================================= Date: Tue, 19 Aug 1997 21:13:21 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Abolishing main entry Comments: To: Lbymgi@LBY.PCMAIL.LEVELS.UNISA.EDU.AU In-Reply-To: <1997Aug20.102400.1069.746264@pcmail.levels.unisa.edu.au> Mervyn Islip wrote: >What I am suggesting is that the headings for the author(s) associated with >a work be listed together in the record, rather than separated into main and >added entries. ... >One way around this would be to provide a name-uniform title added entry >where required ... Because of the mixed nature of the responsibility, I can appreciate the arguments for title main entry for performed works, with name-title added entries for the works (music, choreography, play) serving as the basis of performance. But I can see no reason to abandon main entry under the author of a text, composer of a score, painter of a picture, etc. Such a move would create more problems than it would solve, and create redundant information in added entries if the added entry suggestion above were adopted (the same title appearing in 245 or 130 - I assume 240s would vanish - and 700$t). I already object to the $a redundancy of 800$a$t. This $t redundancy would be even worse. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Wed, 20 Aug 1997 10:18:03 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: subsorting (composer main entry) MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT David Miller asks: > > I'm still working on Neil Hughes' posting from this morning, but at the > moment I'll throw out this: what technical reasons necessarily prevent > adequate subsorting via added entries? We could talk about present > limitations of MARC and specific automated systems -- but looking > further ahead, must subsorting be confined to the main entry? > Different queries would be coupled with different expectations regarding subsorting. And then, which of potentially many added entries would be selected for subsorting? With USMARC data as they are now, it is an intractable problem. Consider that the added entries for conductors and performers have no relation (none that any softare could detect!) to the added entries for the works (in the 700 $a$t fields) they conduct or perform. Forget the contents note, its hopeless for this purpose. What we need first before we can talk about subsorting is full analytical records for all works represented in one container, as opposed to what we have now: only one record for the container with a long contents note and a long list of unrelated added entry fields. But even then, analytical entries can also have several added entries, so which one do you want for subsorting? At least some other (new) kind of indicator would be required here, and where is that to come from? B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Wed, 20 Aug 1997 07:50:49 -0500 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Robert Kusmer Subject: main entry vis-a-vis shelf order Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Having just begun reading these postings, I may well have missed this point having been made already. My apologies if that is the case. Has the impact of the elimination of main entry on its function as a shelf order device been considered? If all records were to have title main entry, what would be the device for creating correct alphabetic shelf order? Robert L. Kusmer Cataloger, German/Humanities G253 Theodore M. Hesburgh Library University of Notre Dame Notre Dame, Indiana 46556-5629 Phone: (219) 631-8649 Fax: (219) 631-6772 E-mail: kusmer.1@nd.edu ========================================================================= Date: Wed, 20 Aug 1997 09:06:33 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Re: main entry vis-a-vis shelf order I don't see a great impact, at least with regard to what I've been advocating. I'm not advocating the elimination of main entry, and many works already do receive title main entry. So even if another class of works (recorded performances) was given title main entry broadly, principles for shelflisting should remain as they are. David Miller Levin Library, Curry College dmiller@curry.edu ========================================================================= Date: Wed, 20 Aug 1997 17:08:13 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Re: Abolishing main entry -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of J. McRee Elrod Sent: Tuesday, August 19, 1997 9:13 PM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Abolishing main entry I already object to the $a redundancy of 800$a$t. This $t redundancy would be even worse. [] Oh, I do so enjoy this topic, as it helps to focus one's mind on the skeletal structure that is the basis of AACR. Currently, the 100 field may look like the king on the throne, presiding over the record, with all its attached access points. But the 100 field is understood to be attached to one work and to one work only. This work is either represented by the whole bibliographic item, or the first, main part of the item, with other work headings created with 700 analytical added entries. The 100 author is already attached to one and only one work heading. Never has been attached to more. Never should be. Never will be. The last thing we should be doing is overburdening the already confusing and complex interconnections between the fields in a cataloging record. If we do not treat series as a separate work, just like a separate analytical entry (even if it is the same as the 100 author), precisely how are we making things simpler and easier for cataloguers and system designers? I would think we would want to eliminate multiple, context-sensitive meanings for fields. I certainly have no difficulty tracing a lot of the problems with the AACR translation to MARC and to automated systems back to this one point. In other words, this is not a side issue -- this topic is at the crux of the whole spectrum of confusion and difficulty in updating the code. Recently I catalogued a two-author fiction series by Lucas and Claremont. Based on predominance, Lucas gets responsibility for the entire series, but Claremont has chief responsibility for book 2. I am dealing with two different works, with two tracks of responsibility. In the catalog, I can find the series under both authors, or under a title reference, or from a Related Works screen from each part, so all access points are covered. The role of the name-title main entry heading assists me in collocation, as I can file all books under Lucas (or each book under each main entry author, if I so choose). I will never file this series under series title, as this is not the best way of ensuring high circulation. Therefore, the meaning of main entry has to sway towards predominant access point, which is author access. And this applies equally and separately to series as it does to the individual parts. Most automated systems that I have worked with do not have the sophistication to conjoin disparate data elements to create authority forms, such as series headings. Therefore, eliminating redundancy in name headings will be accompanied by a decrease in functionality and clarity. If anything, the highest degree of accuracy and sophistication on this point can be found in the traditional card catalog design following the rules of main entry for works in Part 2 of AACR. A separate decision is made in creating the main entry for each work. This separation must be maintained, (if not enhanced) in the code to assist system designers. Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Wed, 20 Aug 1997 09:29:10 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: main entry vis-a-vis shelf order Comments: To: dmiller@CURRY.EDU In-Reply-To: <199708201306.JAA06883@hermes.curry.edu> "David P. Miller" wrote: >I don't see a great impact, at least with regard to what I've been >advocating. I'm not advocating the elimination of main entry, and many >works already do receive title main entry. So even if another class >of works (recorded performances) was given title main entry broadly, >principles for shelflisting should remain as they are. I see no problem with performances arranged within class number by main entry Cutters. It would even be fine with me to move to title main entry and title Cutters for works with two or more authors (as opposed to four or more as at present). For works by *one* person, however, I very strongly support main entry under that author, composer, or artist; Cuttering by that main entry for shelf arrangement; and the use of that main entry with title for added entries used with related works. My customers want the works of one author on one subject together, the works of one literary author together; and added entries for works being supplemented, criticized or performed under the author-title of the work being supplemented, criticized or performed. Abandoning main entry for single authors would make all of the above much more difficult to do. David is correct is saying that he did not advocate title main entry for all texts, only performances. But there has been support on this list for abandoning main entry under persons. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Wed, 20 Aug 1997 15:50:22 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: Re: composer main entry Tom Brenndorfer stated: "My take on this issue is that main entry headings must always be constructed because we must always assume a single entry limitation at some point or another. Book catalogs, cutters for shelf arrangement, brief summary displays, citation forms -- all come to mind. In other words, main entry to me means that if there was one, and absolutely only one way to create a heading for a work, fulfilling the needs of unique identification and efficient collocation, what would that form be?" This is well put, and on these terms it seems as though main entry of virtually anything under title (except anonymous classics?) doesn't come off very well. But I do wonder about this premise as it relates to the issues that are being raised in the conference papers, which in turn have been the subjects of discussion (never mind complaint) for a number of years. I accept that systems will, at some point, display a single entry limitation (a unique title requires only one cutter number, a single-line display tautologically consists of a single line). But a "lowest-common denominator catalog" is, probably, inherently unsuited to explicate multiple bibliographic relationships, for example. I'm hardly speaking about this from great institutional wealth! We got our first automated system only two years, and while we're pleased with it, it's pretty straightforward. We're not a beta test site for anything, and you won't see us headlining any LITA preconferences :-). But I think it's important to discuss these issues in terms of what will address the needs. It may be difficult for simpler systems to handle the proposed solutions. Thanks for joining in! David Miller Levin Library, Curry College Milton, mA (MA) dmiller@curry.edu ========================================================================= Date: Thu, 21 Aug 1997 00:35:19 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Re: composer main entry -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of David P. Miller Sent: Wednesday, August 20, 1997 3:50 PM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Re: composer main entry I accept that systems will, at some point, display a single entry limitation (a unique title requires only one cutter number, a single-line display tautologically consists of a single line). But a "lowest-common denominator catalog" is, probably, inherently unsuited to explicate multiple bibliographic relationships, for example. [] Ah, yes, but how would "explication" occur if not through some sort of links between chief identifiers for each work? One alternative would be leading the user from entire descriptive record to entire descriptive record. Or perhaps from cryptic code to cryptic code. As much as the descriptive record forms a nice, neat unit, I don't think this is the unit that most people have in mind when having bibliographic relationships explicated. If we do not promote the idea of main entry on our terms (as a uniform work heading), some system developer, with much less refinement, will undoubtably have to invent the concept (as, in fact, they have -- the "work" identifier in the summary display in our system consists of 100 + 245$a$b$h + 260$c). Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Thu, 21 Aug 1997 09:34:07 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Abolishing main entry MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Thomas Brenndorfer, in a longer message, made this remark: > > ... The role of the > name-title main entry heading assists me in collocation, as I can file all > books under Lucas (or each book under each main entry author, if I so choose). > I will never file this series under series title, as this is not the best way > of ensuring high circulation. Decisions for main entry points should ideally not be influenced by considerations outside the scope of cataloging. Where you FILE a book on your shelves should ideally be a totally different matter from what you assign as main entry. In practice, as we see and as you will all know, these things are intertwined, partly because of automatic cuttering (spine labeling) and whether or not you want to file parts of a series together or separately - for reasons that have nothing to do with cataloging. How many duplicate records in the utility databases result from this? > Therefore, the meaning of main entry has to sway > towards predominant access point, which is author access. No swaying is necessary, rule 21.1 says it all. B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Thu, 21 Aug 1997 13:56:24 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Linking. Part 2. MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT To link or not to link, and how ------------------------------- Part 2: Links between bibliographic records, including work records Like in Part 1, many statements here will be nothing new for readers with a background in library database programming. They can skip to the section on "Links between bibliographic records". It was found necessary to split Part 2 into 2 parts. The first (this one) is now called Part 2, the second will be Part 3 and will deal almost exclusively with the Part->Whole relationship. In Part 1, subfield $3 was used in examples to represent linkage IDNrs. That's a UNIMARC convention, whereas in USMARC, $3 means "materials specified" wherever it occurs. In 76X and 78X, USMARC uses $w for link IDs. Therefore, that is what we use in examples below. "Link" will henceforth always mean "data link" in the sense of Part 1, i.e. controlled link, not "informal link" like in notes. The purpose of bibliographic links must be to enable users to find work records starting from title records, and then to get collocated views of title records linked to work records or to each other in various ways. In what ways? That will be the task of the rules to define, see below. But haven't we had links all the time? And don't the new WebPACs come with all sorts of hyperlinking? -------------------------------------------------------------- Access points in 7XX and 6XX nowadays find themselves elevated to the status of links, or even "hypertext links" in Web OPACs. What these usually do is to make the contents of those fields "clickable" and thus easy to follow for the end-user. More often than not, however, the other end of the link is missing. The 700 and 600 fields were made to refer back (as "added entries") to the record they are contained in, and to collocate records under established standard headings, they were not made to refer to other entries or works. They are not links in the true sense. (The "established standard headings" do come very close to a function of work records though.) To clarify matters further: In HTML documents, everything that appears in blue and underlined is called a "link". Originally these always contained the URL of another document. In WebPACs, that is no longer so: what they often do is to start a CGI program which in turn generates a new (dynamic) HTML document, just for this single occasion. The outward appearance of HTML links is always the same in that you see something (not always blue, granted) that invites you to click on it, you don't see what exactly will happen when you do. We are not talking about blue links here but only true links, deliberately implanted into records for the purpose of linking, and linking between bib records only, not linking to primary documents via URL or such. What WebPACs do with these links and what other kinds and types of links they provide is, at present, outside the scope of AACR anyway. (2) As an off-topic remark: To turn added entry fields into blue hyperlinks can prove a mixed blessing. Users may find the catalog to promise too much. That is always the danger with providing nice and easy avenues of access to catalog data: users tend to believe without a moment's reflection that the computer will now find "everything" for them. After the inevitable disappointments, they might conclude there to be one letter too many in the word "hyperlink". The old principles of "garbage in - garbage out" and "something cannot come from nothing" are as valid as ever. Hypertechnology can certainly elevate the appearance, but not the substance of a catalog. Before we can go on to the real issues of bibliographic record linking, one more question needs to be aired: Do links have to be bidirectional? ---------------------------------- It sounds just sensible, doesn't it? But what does it really mean? "Reciprocal"? Or something like "traversable both ways"? It may come as a surprise, but in every properly indexed database, ANY link can be traversed both ways, even if implemented only at one of two ends. So there is no need to have a link reciprocally anchored in fields or subfields at both ends. As an example, lets look at (Deviating from true MARC, we have depicted an "identifier link" ($wa98765) in this example, but a "textual link" using "$aTchaikovsky..." could serve quite as well - provided there is appropriate indexing.) Bib record Authority record 001 b01234 |--------------> 001 a98765 100 10 $wa98765 ---| 100 10 $aTchaikovsky, Peter Ilich... 245 00 $aPiano concerto no. 1 400 10 $aCajkovskij, Petr Ilic ... ... and suppose we have indexed both the 100 $w of the bib record and the 001 of the authority record. Thus, the index would contain entries like a98765 [generated from and pointing to the authority record] a98765 -> b01234 [generated from and referring to the bib record] Using this index, software can traverse the link both ways and in this sense it is bidirectional. It is not reciprocal because the bib record's IDnr "b01234" is not contained in the authority record. To do the latter would be highly impractical: you end up with hundreds of bib record IDnrs in the authority record and have to keep them up-to-date - nobody does that with authority records, but it is just as unnecessary for bibliographical links. IOW, software can easily make every link bidirectional. Presently, not all systems may be prepared for this, however, which is why some people favor or advocate reciprocal links, or mean "reciprocal" when they say "bidirectional". IMHO, software that does not support this kind of linking functionality must be called unfit for the future. Luckily, it is not a complicated or costly feature. Well then, at WHICH end DO we anchor a link if not at both? That's easy. Practically speaking, the situation is always this one: You create a new bibliographic record A and realize you have to link it to another record B (of whatever type). Then the link belongs in A. (When cataloging, you simply don't start out and create a new authority record, or other related record first, but you start with the item in hand. And only THEN discover that another record has to be found or created to link it to! But the one in hand, the one you start with, is the one in which to anchor the link.) Theoretically speaking, the link must always point "upward". That is, from the subordinate to the higher level, from the specific to the general, from part to whole, from small to big, from later to earlier, from manifestation to work, from concrete to abstract. This actually implies the practical rule given above: what you have in hand in the typical cataloging situation is invariably the concrete, not the abstract, it is a part, not the whole, and so on. Links between bibliographic records ----------------------------------- My intention here is not to pile more theoretical chaff on the impressive mound of theory we already have, for it contains as much grain as we will need, if not more. Rather, I attempt to present a minimal implementation of record linking into USMARC without so much as a single new field or subfield, yet capable of handling a wide range of logical links. Hopefully, this will make the suggestions of theorists clearer, and hopefully too, it will show that those suggestions are not extremely difficult to realize. No need, in fact, to wait for a "totally revamped" MARC. This implementation follows S.L. Vellucci's reasoning in her "Bibliographic relationships" paper (p.28/29) and M. Yee's outline suggestions in her paper "What is a work" (p. 25/26). A "conceptual schema" for a full-scale model can be found in a paper by Gregory Leazer (1). To make things really simple, let us use just one tag to accomodate all links, namely the 787 Nonspecific relationship entry (Repeatable) and two subfields : $w Record control number (target to link current record to) $g Relationship information (textual; optional) All other subfields, strictly speaking, are redundant because they all contain fields from the target record pointed to by the number in $w. The other subfields may be used, on the other hand, in situations where the software cannot handle these links or the target record is not present in the database. They can be inserted into exchange records, of course, which would then look as they do now, plus the $w. This way, local software would not be affected if it is not prepared for linking. The subfields are defined repeatable in USMARC, but that makes matters unnecessarily complicated. It is better to have another 787 for every link to be established. Why 787? 787 may presently be used chiefly for serials, but it is the closest thing that can be found to suit this purpose. Instead of using 787, one might think of extending the 700, which currently has no $w. The $6 for "linkage" has nothing to do with $w, for it is meant for intra-record linking to an 880 field in an alternate script! It may be better, however, to restrict the 700 to additional personal name access points and take the name/title references out of it and make true 787 work links instead, or separate analytic records altogether in the case of contained works - but that's for Part 3. On the other hand, the 787 could be made to look almost like a 700, except the $w, so that local systems incapable of handling it could turn it into a 700. But to provide a 700 in addition to the 787 would be a waste. To distinguish between the various relationships, the simple model proposes the use of indicator 2 in 787, hitherto undefined. This indicator might take on the following values (and a full-scale model would not differ in this aspect): 0 Equivalence (facsimile or reproduction) 1 Simultaneous edition 2 Successive derivation, edition, version 3 Amplification (incl. commentaries, illustrations, criticism etc.) 4 Extraction (abridgements, condensations, excerpts) 5 Recordings of performances 6 Adaptation, modification (change of genre or medium, arrangement) 9 Translations a Accompanying relationship (supplements of any kind) p Part -> whole relationship r Review or other descriptive relationship s Sequential relationship (like successive title of a serial) u Unspecific relationship, based on shared characteristics of This list is based very closely on B. Tillett's taxonomy of relationships and Smiraglia's extensions (these being the numbers 1 through 9 above; they are subcategories of Tillett's "Derivative" category which Smiraglia, by empirical evidence, found to be too broad). (1) Differing from Tillett and Smiraglia, the list is ordered, more or less, from very close (identity) to rather distant (unspecific) relationships. One might (but should one?) call this "relevance ranking". (Relevance is subjective, and for the end-user to judge, not for the database producer, let alone the computer.) What if more than one indicator applies? Like, say, for a sound recording of a part (one movement, an aria etc.)? Then use the first in the list that applies. Thus, "5" takes precedence over "p". This appears sensible for a minimal model. A full-size model would have additional work records for parts of a work, or it would define a repeatable subfield instead of the (not repeatable) indicator. Another possibility: use repeated 787s to indicate several different relationships to the same work. Yet another suggestion: add subfields $p for Part and $v for Volume to the 787 definition and use these subfields to relate to parts. But make $v sortable to allow for automatic arrangement of parts in correct sequences! One might think of adding $l to 787, but the language is redundant anyway because of 008 and/or 041. The physical medium or format adds another dimension which we need not discuss here. Of course, to have a work record act as a device to collocate several editions in different media would be a welcome effect. The OPAC or retrieval software can certainly be made to give the user a selection option for media type, in addition to everything else. Work authority records? How can this scheme be applied in cataloging, what will the OPAC user see? This at once raises the question "Don't we need work authority records first, to serve as targets for the links in the 787?" Not necessarily. If we have a record for the original edition of a work, then this may do double duty as a work record. If not, then a uniform title authority record with a $f and $t can be used. For the minimal model, this is sufficient. It means we have no need to define anything new beyond what USMARC already has. The $t subfield is a bit ugly in that the initial article cannot be marked. To simply omit it may have been found tolerable by many, but for work authority records it is not good enough, not for every language at least. But that's another subject. Instead of a name/title authority record one might think of a skeletal bibliographic record with a 100 , 240$a and 260$c to serve as work record. Suppose, by way of example, we have three items to catalog: [1] A translated edition of Shakespeare's "Macbeth" [2] An English and Italian vocal score of Verdi's "Macbeth" opera [3] A sound recording of a performance of the latter. For the two works involved, we use these authority records as work records: 001 a888 100 1 $aShakespeare, William$d1564-1616$tMacbeth$f1605 and 001 a999 100 1 $aVerdi, Giuseppe$d1813-1901$tMacbeth$f1847 787 16$w888 The second has a 787 link to the first, saying it is an adaptation of it. Our bib records will then have these core elements: [1] 100 1 $aShakespeare, William$d1564-1616 240 10$aMacbeth.$lItalian [redundant because of 787?] 245 10$aMacbeth 787 12$wa888 [it is an edition of a888] 787 19$wa888 [it is also a translation] [2] 100 1 $aVerdi, Giuseppe$d1813-1901 240 10$tMacbeth.$sVocal score.$lEnglish & Italian 245 10$aMacbeth :$bopera in four acts 787 16$wa999 [an arrangement of a999] [3] 100 1 $aVerdi, Giuseppe$d1813-1901 245 10$tMacbeth$hsound recording 787 15$wa888 [it is a performance recording] The links in 787 can enable software to find and display the work records upon demand, and to collocate and display records related to a work record in all the various ways. IOW, the 787 as defined here is all that's needed, the rest is "only" software. (At Braunschweig Univ. Lib., we did an implementation into the USMARC database we keep for compatibility studies, and it was only an hour or so of work.) Just a matter of software too is the support of inputting the 787 into new and existing records. One easily envisages a "point and click" method of making a link, and the system would have to prompt for the type of link and the (optional) $g input (plus, probably, $v and/or $p). Establishing links can be made a quick and easy process with this model, and it wouldn't otherwise be feasible. The task, however, to introduce any linking model into a large database remains a daunting one. Maybe one can write software to turn 100/240 combinations and part of the 700 $a$t fields into work authority records and new type links, but these would have to be checked and the indicators supplied. And what will the end-user see? That depends, as always, on the software that presents the OPAC interface. The emerging WebPAC technology can make use of these new elements and index entries to let the user traverse the links. The cautionary remarks about hypertechniques made above remain valid. It will be no difficulty to produce displays like this for a work record: Verdi, Giuseppe (1813-1901) Macbeth. 1847. 1. [Adaptation of:] Shakespeare, William: Macbeth. 2. Adaptations 3. Recordings of performances Here, 1. to 3. may be "blue" links in a WebPAC or just numbers to enter in conventional OPACs. Following link number 1, one might get (depending what types of relationships actually exist in the catalog) Shakespeare, William (1564-1616) Macbeth. 1605. 1. Original edition 2. Later editions 3. Amplifications (incl. commentaries, illustrations, criticism etc.) 4. Extracts (abridgements, condensations, excerpts) 5. Recordings of performances 6. Adaptation, modification (change of genre or medium, arrangement) 7. Translations 8. Other related works Following link number 2 will bring up records like Verdi, Giuseppe: [Macbeth. Vocal score (English & Italian)] Macbeth : opera in four acts ... 1. [Adaptation of:] Verdi, Giuseppe: [Macbeth] (1847) Link 1 here, of course, leads back to the Verdi/Macbeth work record. Final remark. This part talked a lot more about USMARC than about AACR. It is very clear, however, that the whole matter of linking bibliographic records is one that relies on implementation into the format. If the code wants to stay format-independent, it can probably still do so. Work records can be defined in cataloging terms alone and printed on cards as well. Links can be defined as new kinds of "added entries", potentially replacing current added entries (cf. what was said about 700 vs. 787 above). For brevity, these added entries could even be called "links". How much sense all that makes for conventional catalogs, at least in terms of feasibility, is another question. Part 3: The Part->Whole relationship. to follow. References (1) Leazer, Gregory: A conceptual schema for the control of bibliographic works. - In: Navigating the networks. Proceedings of the ASIS mid- year meeting, Portland, Oregon, May 21-25, 1994. (p. 115-135) Medford, N.J.: Learned Information Inc., 1994. ISBN 0-938734-85-7 (2) Amanda Xu of MIT Libraries (zxu@mit.edu) provided me with a very good paper (as yet unpublished?) on the whole subject of hypertext links: Xu, Amanda: Hyperlinks in the WebPAC Systems : A study of hypertext links to a bibliographic record in the Web based OPACs. 1997. B.E. 19970819 Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Thu, 21 Aug 1997 13:32:00 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "Disipio Mary F." Subject: AACR, Second Edition; Their History and Principles Content-transfer-encoding: 7BIT Please note that the following conference paper is now available on the JSC Web site: Anglo-American Cataloguing Rules, Second Edition: Their History and Principles by Michael Gorman and Pat Oddy. ========================================================================= Date: Thu, 21 Aug 1997 16:20:52 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Mary Grenci Subject: Isn't AACR already format-based? Comments: cc: Lonni Sexton , Christy Carmichael MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII I've been following the discussions on this list with great interest, and have finally decided it's time to jump in. Quite a large number of messages have dealt with library systems and MARC issues, and a number of others have been sent either in support of or arguing against their relevance to AACR. It has been said that AACR is a cataloging code only, that it's not based on the format of the catalog and that revisions/changes shouldn't take into account the format of the catalog. But AACR is most certainly format-based. It is based on the typed 3 x 5 inch catalog card which is then filed (under rules not covered in AACR) in the physical card catalog. Many of the rules were formulated solely because of the limitations of both having to manually type an entire card set, and the size of the cards themselves. If the Library of Congress had not decided to distribute their card sets on 3 x 5 cards, but had used 4 x 6 cards instead, more information could have been accomodated on each card and records could have been longer (except, again, for the effort necessary to type out this added information) There have been catalogs in different formats before the advent of the online catalog, for instance book catalogs and microfilm catalogs. When you look at one of these catalogs it is obvious that what they really contain is a reproduction of the printed catalog card. Why? Because this printed card is the basis for AACR as it now stands. The real question is not 'Should AACR become format-based?' It is 'Should we change the format on which AACR is based?' Or, possibly, 'Should AACR be changed so as NOT to be format-based?' (Which, frankly, I'm not certain can be accomplished since we must always have some basis for decision-making) Mary ______________________________ Mary Grenci Serials Catalog Librarian Knight Library 1299 University of Oregon Eugene, OR 97403-1299 mgrenci@darkwing.uoregon.edu ========================================================================= Date: Fri, 22 Aug 1997 09:55:50 +1100 Reply-To: Hal.Cain@ormond.unimelb.edu.au Sender: "International Conference on the Principles and Future Dev. of AACR" From: Hal Cain Organization: Joint Theological Library, Parkville, Victoria, Australia. Subject: What are we trying to do? Comments: To: AUTOCAT@LISTSERV.ACSU.BUFFALO.EDU Comments: cc: mac@slc.bc.ca MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Transfer-Encoding: 7bit I am posting this message on both AACRCONF and AUTOCAT (since I started the theme there). Apologies to those who get a double helping. On Wednesday (Eastern Australian time) I posted a message on Autocat querying the direction of much of the discussion on AARCONF, hoping for responses from working cataloguers who may not see, or respond, on AACRCONF. I said, in part: > We (I mean cataloguers as a distinct group within the profession) seem > to be heading in two conflicting directions. > > First, the simplify-the-workflow, trim-down-to-core-record-standard > direction ... > > Second, the multi-level, work/manifestation/item pattern which threatens > to impose additional levels of complexity on bibliographic records (in > creation, storage, processing, searching, and interpretation). > > We are already familiar with contemporary cost-cutting administrations > who clip our wings (or amputate them) and with library systems that > don't accomodate all the capabilities built into existing records (even > distort what is there). Few of us now do *all* the authority work that > could be done in our local systems, and quite a lot of us do almost > none. > > How on earth are we to implement a new code, which will incorporate > concepts unfamiliar to many of us cataloguers (let alone to library > administrators and their parent institutions) and entail new > capabilities in USMARC, new designs of library computer systems, new > search strategies, and new display formats? > > And how on earth will we reconcile the millions of existing records with > the new formats? (IMO attempts at automated processing are pretty well > bound to produce results worse than useless because the recorded data is > insufficient for that purpose.) > My thanks to the respondents so far, who ranged from vaguely comforting to vehement agreement. Let me make it clear that I am very much in favour of revision of AACR2. I have no difficulty with the topics of the papers, and the discussion is fascinating. But I think we are getting carried away. The substance of my qualms is expressed above. I have been reflecting about what it is that lies behind them, and have come to the conclusion that the first problem is in the *purposes of the catalogue*. Along with this is a doubt whether the creation of bibliographic records is solely to serve these purposes -- perhaps there are other purposes too? Cutter and others have been quoted, but I think it is time to attempt a comprehensive statement. If nobody else sets about it, I'll try (but it will have to be next week -- other areas of my life are very demanding just now, I can't keep up with this list!). Along with the (real) excitement we should take a good hard look at what we do, starting with why. Hal Cain, Joint Theological Library, Parkville, Victoria, Australia ========================================================================= Date: Fri, 22 Aug 1997 07:27:37 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Antony Robert David Franks Subject: Purpose of cataloging? -Reply Comments: To: Hal.Cain@ormond.unimelb.edu.au, hecain@ormond.unimelb.edu.au I think that Hal Cain is quite right in bringing up the topic of just what, anymore, a catalog is for. Unfortunately, I'm not sure if a "nuts-and-bolts" conference on AACR is the place to do it, since any cataloging code proceeds on the assumption that the catalog it produces is a "good thing". The catalog one wants ought to be the result of the code one uses, and so, we're getting into a vicious (if fascinating) intellectual circle. We seem to be going through a collective period of anxiety over whether or not our activities as catalogers are of any use to anyone. This is reinforced by the current management craze for various buzz words that all translate into "cost cutting." I think it bad for the profession that a period of re-evaluation like this is being forced by a need to do things more cheaply, but that just seems to be the fact of the matter. Perhaps some investigation of what people want from our individual records, and from the collection of records (a catalog), and a clear statement of the results, will give us an idea of what we ought to be providing in those records as directed by the cataloging code. For example, I think that the CORE record continues the past desire for a "generic" record useable by anyone. On the other hand, it offers a nearly infinite malleability in the hands of a local cataloger for reshaping to meet local needs in a local setting. This is something that Hal Cain points out in his original posting when he mentions the proliferation of records for the same item--what is "wrong" with a gerneric record, that it must be reworked so many times to meet so many different needs? Is it the record, or is it the perceived needs for the record on the local level? With costs a rising concern, this discussion runs smack into managements' desires to adopt (copy) such a record with no or little adaptation on the part of the local cataloger. As I said above, these are just some thoughts I have been mulling over in my head the last few weeks as I face some of these issues at my own work. ********************************************************* * Anthony R.D. Franks * * Library of Congress * * afra@loc.gov * * * * Not the official position of the Library of Congress* *********************************************************## ========================================================================= Date: Fri, 22 Aug 1997 11:04:20 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Jennifer Kolmes Subject: Re: composer main entry In-Reply-To: <199708201950.PAA19415@hermes.curry.edu> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii" In response to David Miller's question about why all the resistence to main entry under title for music compositions-- I thought I'd sent a response to the list a couple of days ago, but looking for evidence today, it seems to have vanished off the face of the earth. I'm amazed that no one has pointed out (or am I missing that, too?) that in many, perhaps most, cases, music titles by themselves don't make a whole lot of sense. That is, "Symphony no. 1 in G major" doesn't mean diddly-squat until a composer's name is attached to it. Doesn't matter if it is a score or a "performance," i.e., recording. To me, it just doesn't make sense to have MAIN entry under something that is essentially incomplete. I think this is the reason for most of the resistence you're going to encounter from the music community when trying to move towards title main entry across the board. Jennifer Kolmes Head of Cataloging University of Arkansas Fayetteville, AR 72701 ========================================================================= Date: Fri, 22 Aug 1997 14:21:30 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "James E. Agenbroad" Subject: Why Catalog? & a Change to AACR MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII Friday, August 21, 1997 The first part of this message will be self-evident to most of those on this list but it may help put our activities in perspective. The second part raises an issue where change in AACR is, I believe highly desirable. People organize themselves into groups and subgrousp for for various purposes--municipalities and public libraries, universities and their libraries, churches and their choir libraries, etc. In pursuit of their current and future purposes some groups acquire documents through purchase, gifts, etc. As these documents accumlate they become a valuable asset--a collection--and it becomes harder to find individual items so they must be organized (e.g., cataloged and classified) for control of the asset and for access by group members and possibly others outside the group. Groups want their documents organized efficiently (easy to find) and economically (most groups have limited funds, taxes, tuition, etc., and many valid ways to spend them including buying yet more documents). Since many people use several groups of documents it will often be easier for them if the same organizing principles are used by different groups. Since copies of the same document frequently exist in different collections there are economic advantages if different groups can share the records that organize documents. To improve both the commonality of these organizing procedures and the sharing of the organizing records, representatives of organizers in "Anglo-America" have codified their procedures in certain areas--document description and some of the access points ("headings") used on these descriptions. The following are some possible reasons to change this codification: 1. Expansion of areas of agreement, e.g., topical subject headings, classification, sorting; 2. Expansion of the kinds of documents to be described and organized; 3. Widening of the area represented beyond "Anglo-America"; and 4. Improved means of organization for control of and access to collections such as computers and OPAC software. Thus we have this list and in October the International Conference on the Principles and Future Development of AACR. Nonroman name headings are an example of an improved means of access. Though limiting headings to the roman alphabet is not among the principles mentioned in the paper by Gorman and Oddy, the use of transliteration has been Anglo-American practice at least since Cutter. For most of this period I believe the main reasons have been the limited availability of printing resources to create cards with nonroman text and even more importantly, the absence of staff who could file cards with such headings quickly, accurately and economically. The existence of transliteration tables demonstrates that Anglo-American libraries (both small public and large research ones) acquire documents written in other scripts. If there are those who doubt that transliterated headings deter access to and use of such documents they should either read the articles by S. Spalding (Library resources and technical services, 22:1 (1977) 3-12 and H.Wellisch (LRTS, 22:2 (1978) 179-90, or imagine trying to guess how their name would appear in an unfamiliar script such as Armenian to approximate the frustration of library users wanting books in other scripts. At present AACR proscribes headings in other scripts for all access points except titles proper and geographic names. For several years those using RLIN and OCLC to catalog Chinese, Japanese or Korean items have been able to create and search headings in these scripts. More recently RLG has added the same capability for Arabic, Hebrew and Cyrillic scripts. These extra headings are assigned because they increase the use of these materials and thus the return on the investment in acquiring and housing them. The "Beyond MARC" paper discusses Unicode(tm) as the next step in automated bibliographic control of documents in nonroman scripts. Thus the reasons given above for transliterated headings are no longer valid. Unless xenographphobia, the prejudice against foreign writing systems, is an acceptable excuse, AACR should at a minimum permit headings in other scripts and preferably also provide guidance to promote consistency in their construction. Regards, Jim Agenbroad ( jage@LOC.gov ) The above are purely personal opinions, not necessarily the official views of any government or any agency of any. ========================================================================= Date: Fri, 22 Aug 1997 17:33:48 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Composer main entry Comments: To: jkolmes@COMP.UARK.EDU In-Reply-To: <3.0.1.16.19970822100018.3f9fcc12@comp.uark.edu> Jennifer said: >many, perhaps most, cases, music titles by themselves don't make a whole >lot of sense. That is, "Symphony no. 1 in G major" doesn't mean >diddly-squat until a composer's name is attached to it. Doesn't matter if >it is a score or a "performance," i.e., recording. In terms of scores I agree with you completely. But in terms of performances, most have titles such as "Canadian Brass : Bach : art of the fugue", "A treasury of Gregorian chants", "French ballet music", "Toyohiko Satoh : Chaconne : Johann Sebastian Bach", "Julian Bream plays Grenando & Albeniz", and so forth. I see nothing wrong with entry under the title as on the album, lots of 246s, and *added* entry under 700$aComposer$tComposition title, and 700$aPerformer. I can see David's point. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 22 Aug 1997 22:04:41 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Gorman & Oddy's paper Comments: To: autocat@ubvm.cc.buffalo.edu Comments: cc: VLAFORUM@UVVM.UVIC.CA, BCCATS-L@MALA.BC.CA Michael Gorman and Pat Oddy in their "The Anglo-American Cataloguing Rules Second Edition : Their History and Principles" have written a short (14 p.) but content filled paper which delivers several well aimed wallops. The underlying principles of AACR2 are clearly and succinctly stated. "Unilateral change within AACR2 itself [departing from ISBD] would be a major retrogressive step." "There is need for some change but that change should be gradual, evolutionary, and within the structures and principles of AACR2." "There has been some extraordinary misguided and misinformed discussion on the need to create 'master records' for works that are manifested in many different physical forms ... ; the idea of a 'master record' for several manifestations of the same work is a cataloguing nonsense." (They lay it on the line don't they?) Gorman and Oddy made ten recommendations, with none of which I would take exception. I would only take exception to the position taken in support of one of them. 1) We should get rid of the "special" rules which were imported for political reasons after Lubetzky resigned, e.g., music and laws. (I hope those awful constructed main entries for treaties will be among the first to go.) 2) We should prune descriptive rules which deal with particular cases, and leave these to manuals produced by specialists in the various fields. 3) We should resolve the issue of "unpublished" items. (If this means arriving at sensible information for those ISBD areas now left empty, wonderful.) 4) Include new media, especially electronic, including those accessible remotely. 5) Access issues for new media should be studied without creating "case law" for each. 6) They recommend the commissioning of a consultant to review the whole main entry issue. Earlier they advocated that all access points be equal, with none designated as main. This is my one disagreement with this *very* level headed paper. True, the main entry is outside the ISBD, as are other access points including subject headings. But author/composer/artist as most important (main) access point for their works is in my view a valid concept. For card catalogues, printed book catalogues, single entry lists, bibliographies, and as added entries for related works, a citation composed of single author/composer/artist plus title is by far the simpler method, as well as being a time honoured, solution. Once there are two or more authors (as opposed to four or more as at present) I've no objection to considering the title to be main entry, and all other access points equal. 7) We should resolve the microform issue. (LC is still using AACR1, and most U.S. libraries have followed their practice.) No suggestions are made for doing so. I would favour a compound imprint following AACR2 1.4G4, with the print publisher followed by the microform manufacturer, and a compound collation, with the original pagination in curves following the smd. 8) We should review the examples. 9) We should consolidate a unified MARC and AACR2. (With renumbering of 5XX one hopes!) 10) We should ask LC to review and curtail the LCRI program. Why doesn't JSC simply dominate Gorman to be the Lubetzky of AACR3, and defend him from those who would complicate his clear vision? But make him keep main entry of course :-{)}. __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 25 Aug 1997 11:21:27 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Paper by Gorman & Oddy MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT A few remarks on the recent conference paper, "The Anglo-American Cataloging Rules, Second Edition : Their History and Principles" by Michael Gorman and Pat Oddy Most of my remarks are rather questions, and questions of an outsider. An outsider, however, who is trying to do what little he can to bridge the gulf between incompatible codes and concepts. Yet my perception may be incorrect or incomplete. "Principles" One is wondering why there is no mentioning of objectives before listing the principles of the code. But then the code itself contains no reference to these objectives. Discussions in this list have indicated that there may be a need to talk about objectives, esp. when it comes to collocation. Vellucci's paper (and the work of Tillett and Smiraglia) have made it clear that there are important bibliographic relationships currently not traced in catalogs. Will classical objectives have to be extended, re-evaluated, reformulated for the online age? The main tasks of the code, ever since Cutter, are to locate (item centered) and to collocate (work centered). Access points have to serve both ends, and this is a source of conflict which the principles try to minimize. "The importance of AACR2" Just for historical correctness (re: paragraph 3.) The German Code RAK (Regeln fuer die Alphabetische Katalogisierung) was developed at the same time and gradual introduction in Germany began around 1970. The finalized version was published in 1977. RAK have the same clear distinction as AACR2 between description and access. (Both claim to be based on the Paris Principles of 1961 anyway, and that's where this comes from. ISBD was used in RAK from the start, too.) AACR2 may be called the "first global cataloging code", but the German- speaking countries are uncharted territory still. In the REUSE talks (sponsored by OCLC) we have concluded that a globalized authority file for names would be a big step forward. The name authorities contain numerous German and other non-English name forms already, but these variant forms carry no identification of language, nationality or cataloging code. Is this just a format question? There are incongruences between German and US authority data, esp. for corporate names. These we will try to straighten out, though certain problems will remain. But a consolidated authority file would still have to reflect which of the various name forms are "our" main and reference forms. The other major area of incompatibility is with series and multivolume items, and I'm trying to put something together about this and post it later this week. A third factor are major differences in transliteration of Cyrillic and other scripts. UNICODE is a promise on the horizon, but what are we going to do about legacy data then? Nobody knows it. It is quite surprising to watch (under 6.) Gorman and Oddy foreseeing the "ultimate elimination" of the main entry as an "unnecessary complication of little relevance to the computerized catalogues of today". Did they not follow discussions in this list or in AUTOCAT? I thought we had gotten beyond that, but I may be wrong. "Description/ISBD" The ISBD is praised as the "most successful international endeavor in biblio- graphic standardization". Can it be taken for granted that the ISBD is to be the display standard for online catalogs as much as it was and is the standard for card and list formatting? Should the code not make a clear statement here? Either this, or there would have to be, for example, a list of standardized display labels to be used in OPACs. As of now, displays are less standardized than ever before. One may argue that this is not a fault of the code but of librarians letting things happen. In the age of cards, who would have tolerated non-standard card layouts? Online catalogs have more elements than access points (index entries) and record displays. Of particular importance is the short (one or two line) display for result sets. No two systems seem to be alike in this aspect. Since these short displays have important functions with regard to the objectives of the catalog (identification and collocation), should the rules not at least set a minimum standard or make recommendations as to what data elements should be included and in what order? This list need not be reminded that this is one area where the "main entry" retains its importance. It is now urgent to find a new name for it! IMHO, these are not matters to be shrugged off lightly. The result is that users cannot anymore know a catalog when they see one. One is tempted to say libraries are losing part of their identity in the public's mental image. Before long, a catalog will be, for many users, "just another search engine or something like that, but somewhat fussy, and none of the documents are online". "Where are we now?" Gorman and Oddy vehemently oppose calls for "simplification" where these amount to an abandonment of authority control. Now this is one of the really difficult concepts to grasp for non-catalogers. There was the suggestion in this list, recently, that it would be good to have a "Layperson's Introduction to AACR". It may be too much to hope that administrators could be turned around with this, but on the other hand, you simply cannot present a copy of AACR to an administrator or a controller of the public purse in the hope that they read and understand before they make decisions. Don't have a cataloger write this introduction, but find a science writer (or, why not, someone like Nicholson Baker). They write about everything from quarks to black holes in everyday language, and where is cataloging more difficult than subjects like those? I have experienced a need for this introduction in Web circles too: people recently got fascinated about the "Dublin Core" and completely misunderstood it as being a new and wonderfully simple format-and-code- in-one, and it was well nigh impossible to make it clear that there is and remains the need for heavyweight codes like AACR and XXL-size formats like MARC. These standards are coming under scrutiny by self-appointed metadata experts with no library background. Some of those people are not without influence. How are we going to defend our standards against new competition? "An agenda for managed change" Why are filing rules missing on this list? It bears repeating: computers don't file, they sort. Character by character, in ASCII sequence - that's all they know. Software can improve a few things, admitted, but not enough. Card filers could identify initial articles, filers could correctly arrange numbers, filers consciously made subarrangements where appropriate, they did part of the work necessary for collocation. Filers got dismissed, computers took their place, and we know the results. In the rules for access points, there has to be more regard for filing or you'll never get rid of nonsense arrangements, like "10" between "1" and "2". This is not merely a programming problem. The result of too little regard for filing is inconsistency. People bend the rules until they get what looks right - in their system! Most infamous example: the omitted initial article in the $t. This is now even LC practice without being covered by the rules, and I dare doubt Gorman and Oddy would consent to a rule change that would openly sanction this practice. Filing rules are not a terribly difficult matter IF you get the structure of access points right. If you have, for example, elements like Arabic or Roman numerals as part of certain access points, implicitly making assumptions about the knowledge of (now extinct) filers, you invite rule bending and thus inconsistency in databases. A remark for the historically minded: the North-American preface to the 1967 edition of AACR says, on the matter of filing: "There seems to be no serious bar, however, to purely formal modifications in headings for the purpose of achieving by mechanized means the same order contemplated by the rules." This statement is not repeated in AACRII. But have those "formal modifications" been made? It boils down to this: the rules are giving no guidance for filing (e.g., what elements in a name entry to disregard, and how to arrange the various parts of name and name/title entries), the format is not giving any guidance either (as to which subfields to file on or not, and in what order). What is an implementor to do if librarians want string indexed name and name/title sequences, or sorted arrangements of result sets? Librarians have to think up their own solutions locally, or leave it all to the vendors. As long as a system has basically name and keyword access, there's not much trouble. But as systems leave this simple stage behind and introduce string indexing, sorted result sets, and collocation options, trouble begins. Thank you for your patience. Bernhard Eversberg. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Mon, 25 Aug 1997 09:23:44 -0700 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Dan Kniesner Subject: Paper by Gorman & Oddy -Reply Comments: To: EV@BUCH.BIBLIO.ETC.TU-BS.DE Mime-Version: 1.0 Content-Type: text/plain I would like to pick up on Bernhard Eversberg's good comments on the Gorman & Oddy paper, and expand on his section on filing in online catalogs. I'll also touch on the main entry issue. >"An agenda for managed change" >Why are filing rules missing on this list? It bears repeating: >computers don't file, they sort. Character by character, in >ASCII sequence - that's all they know. Software can improve >a few things, admitted, but not enough. I much prefer online to card catalogs and would never go back to cards, but I want to point out one deficiency of online catalogs. It has to do with filing. Suppose you as a cataloger or catalog user would like to find the journal "Scientific American" in a catalog. The larger the catalog the harder it will be to find it. Why? Because "Scientific American" as title main entry is sorted alphabetically with "Scientific American" added entries. (There are a lot of monographs published by Scientific American.) In the card era, filers put the title main entry first. The title added entries followed. Not so in our online era. Online catalogs make it more difficult in this instance to satisfy Cutter's first objective of locating. We have to do something about this. I don't know how to solve the problem. I find it fascinating, though, that main entry is important to everyone even in a strictly title context, let alone author context. Mixing title main entry with added entries is worse in library consortia union catalogs such as Orbis in the Pacific Northwest (U.S.) or Ohiolink in Ohio (U.S.). To see what I mean, look at: http://orbis.uoregon.edu and search for "Scientific American". Then transport yourself into the mind of a catalog user who has little or no background in catalog structure or fundamentals. (Not to pick on Orbis, the same will be true with almost any online catalog.) If filing in online catalogs is not appropriate to mention in the revision of AACR2, where *is* it appropriate? Can ALA filing rules separately have the same force on catalog designers and vendors as AACR2 does? The question is similar to the labeled displays question. Can ISBD separately have the same force on catalog designers as AACR2? I think at least an authoritative mention by AACR2 of filing and displays would be a good thing. The Toronto conference gives us the opportunity to make our catalogs easier to use, read, understand -- for ourselves as catalogers as well as other catalog users. August 25, 1997 Dan L. Kniesner Oregon Health Sciences University Library Portland, Oregon Internet: kniesner@ohsu.edu ========================================================================= Date: Mon, 25 Aug 1997 08:13:22 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Re: Paper by Gorman & Oddy In-Reply-To: <85AC7732D9C@buch.biblio.etc.tu-bs.de> Normally I dislike posts which quote extensively from other posts, but Bernhard's comments on Gorman and Oddy contain several paragraphs which should be read repeatedly. >It is quite surprising to watch (under 6.) Gorman and Oddy foreseeing the >"ultimate elimination" of the main entry as an "unnecessary complication >of little relevance to the computerized catalogues of today". Did they >not follow discussions in this list or in AUTOCAT? I thought we had >gotten beyond that, but I may be wrong. Let's hope you are not wrong. The suggestion of eliminating the main entry assumes the universality of the automated catalogue. In the small special libraries of North America and Europe for which we catalogue, this is far from the case. Outside the world of the library catalogue, one only has to open to the bibliography in any scholarly book to see a single entry list of works. Hal has already mentioned the danger of having cataloguing practice diverge too far from that of standard scholarly (and I would add legal) bibliographic citation. >The ISBD is praised as the "most successful international endeavor in biblio- >graphic standardization". Can it be taken for granted that the ISBD is >to be the display standard for online catalogs as much as it was and is the >standard for card and list formatting? Should the code not make a clear >statement here? Either this, or there would have to be, for example, a list >of standardized display labels to be used in OPACs. As of now, displays >are less standardized than ever before. One may argue that this is not a >fault of the code but of librarians letting things happen. In the age of >cards, who would have tolerated non-standard card layouts? This is a wonderfully succinct statement of the anarchy of online catalogue display. The unlabeled ISBD should be explicitly stated as the preferred standard for full bibliographic OPAC display, with labels specified as an option (and in time one would hope ignored like the gmd $h[text]). Standards are also needed (as proposed by another paper) for at least the one line display and an intermediate display. How about a committee of three: Michael Gorman, Bernhard Eversberg, and Judith Hopkins to prepare the text of AACR3? Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Mon, 25 Aug 1997 16:55:56 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Judith Hopkins Subject: Re: Paper by Gorman & Oddy Comments: To: "J. McRee Elrod" In-Reply-To: MIME-Version: 1.0 Content-Type: TEXT/PLAIN; charset=US-ASCII On Mon, 25 Aug 1997, J. McRee Elrod wrote: > > How about a committee of three: Michael Gorman, Bernhard Eversberg, and > Judith Hopkins to prepare the text of AACR3? > > Mac > > __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) > {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ > ___} |__ \__________________________________________________________ > Only if it contains a provision for Main entry! Judith ========================================================================= Date: Tue, 26 Aug 1997 16:40:01 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Paper by Gorman & Oddy MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT and there was also a short reply by Judith Hopkins, following Mac: > Date sent: Mon, 25 Aug 1997 16:55:56 -0400 > Send reply to: "International Conference on the Principles and Future Dev. of > AACR" > From: Judith Hopkins > Subject: Re: Paper by Gorman & Oddy > To: AACRCONF@INFOSERV.NLC-BNC.CA > On Mon, 25 Aug 1997, J. McRee Elrod wrote: > > > > How about a committee of three: Michael Gorman, Bernhard Eversberg, and > > Judith Hopkins to prepare the text of AACR3? > > > > Mac > > > > __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) > > {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ > > ___} |__ \__________________________________________________________ > > > > Only if it contains a provision for Main entry! > > Judith > Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Tue, 26 Aug 1997 21:28:01 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Paper by Gorman & Oddy -Reply -----Original Message----- From: International Conference on the Principles and Future Dev. of AACR On Behalf Of Dan Kniesner Sent: Monday, August 25, 1997 12:24 PM To: AACRCONF@INFOSERV.NLC-BNC.CA Subject: Paper by Gorman & Oddy -Reply Suppose you as a cataloger or catalog user would like to find the journal "Scientific American" in a catalog. The larger the catalog the harder it will be to find it. Why? Because "Scientific American" as title main entry is sorted alphabetically with "Scientific American" added entries. (There are a lot of monographs published by Scientific American.) In the card era, filers put the title main entry first. The title added entries followed. Not so in our online era. Online catalogs make it more difficult in this instance to satisfy Cutter's first objective of locating. We have to do something about this. August 25, 1997 Dan L. Kniesner Oregon Health Sciences University Library Portland, Oregon Internet: kniesner@ohsu.edu [] I think this point speaks to a larger issue that has been of some concern to me for a while. For something as basic as title access points, it has always surprised me as to how difficult these have been to me and to other catalogers to get a handle on. We have two tracks for dealing with title access points. We have the "found" titles that are specific to a bibliographic item. These are coded in 246 and 740, and would also include the title proper, 245$a. We also have "controlled" titles, which include 7XX forms, series, subject titles, and main entry uniform titles. For all title access points (except as subjects), we must provide a corresponding descriptive justification. However, what happens when we construct a single index of all titles? Far from behind an outstanding feature, I find this an unwieldy, convulated way of creating a catalog. Qualified uniform titles are mixed with the almost identical titles proper, with the almost identical variant titles. Or maybe they're not mixed at all, such as name-title forms ending up in the author index. This is a breakdown of how title access is treated: Main entry uniform title: -130, 100+240; 100+245 can also be considered in this context Variants to main entry uniform titles -246; 245 if 130 or 240 used Alternate place for variants: -on authority records if used/created Descriptive justification: -245; 500 notes for variant titles Analytical and related uniform titles: -7XX Variants to analytical and related uniform titles: -740 (as found on the bibliographic item) -other variants could be cross-references on an authority reocrd Descriptive justification: -various fields including 245, 500, and 505 Series: -440, 8XX Variants to series: -4XX cross-references on series authority Descriptive justification: -440 or 490 1_ Subjects: -6XX Variants to titles treated as subjects: -4XX cross-references on subject authority No descriptive justification I also noted that the LC Rule Interpretations recently dropped their point on 26.4B that deals with variant title access, such that the variant title would be found under author. Instead of: 100 $aSherwood, Rankin.$tPony express to railways 400 $aSherwood, Rankin.$tBefore the West was won all variant title access would be done through the bibliographic record: 246$aBefore the West was won (the 246 joins the 245 as being title access points to the titles as found on the item) The former method demonstrated how alternate forms of "controlled" names of works could be constructed. The latter deals with titles as found on the item, and is specific to the bibliographic record, not to an authority record. Also, I think there is some confusion on the nature of records for works. Authority records for titles are, in a rudimentary fashion, work records, not just administrative control records to ensure consistency in the spelling of access points. I don't think they are intended to be a front-end substitute for descriptive records of unique bibliographic items--they are "post-factum" records, springing from the need for authority control. Moreover, I think much of the discussion of authority records becoming work records have been on the need to tabulate relationships between works and to outline bibliographic history, not to create a parallel descriptive record encompassing all physical manifestations (which would indeed be impossible to create). But this is the issue, surely, that is being discussed from numerous angles? We have a parallel construction already, with variant titles as found on the item being direct access points to the item, and variant titles as alternate "names of works" and found on authority records, providing a collocating basis for access (with series authority records and its SEE references being the clearest example of this form). On a side note, a series authority record, as a record of a "work," certainly does seem to have a lot of descriptive elements that cover a range of bibliographic items, although this may be an effect due to series' potential alternate treatment as serials (monographic series) or monographs (multipart items). But the issue of title acess comes to a head here, where we have a clear separation between DESCRIPTION and ACCESS, where all issues of access are consolidated under one control mechanism -- the series authority record. All issues of description are handled through the appropriate note fields for series. To extend this mechanism to all titles would be to create a rich, cross-referenced index, where all variant titles would in fact be references, not "added entries" with all their privileges and baggage and potential for confusion for people browsing a title index. With this design we would lose direct access to a specific bibliographic item, i.e. each 245, 246, or 740 is specific to an item, not to a work. Indirect access through a title (i.e., work) authority record would be the same as what we do now for author and subject authorities. As a final point, I wonder if some of the problems with works are not in fact a result of the constraints of the typical cataloging procedure, from Description (Part 1 AACR) to Access (Part 2 AACR). In a card catalog, the flexibility gradually diminishes with each successive choice we make, once we are committed to a particular format (i.e. cards), with logistical limits as to the number of cards that can ideally be used, and in what manner they can be organized. An extended authority record in this environment would face enormous technical hurdles, but it does not seem impossible to create in an automated environment, where "post-factum" reorganization can be done with ease and little more than a flick of a switch. This is a fundamental and inherent quality of an automated environment -- a key distinguishing point between OPACs and card catalogs. We see the flexibility (used more for ill) in the variety of forms in descriptive records in OPACs. This is an attribute, not a flaw, and it need only be directed towards constructive ends!! Any construction can be described, even if the construction is only an abstract idea. The excellent contribution of Gorman and Oddy is flawed by this one logical oversight. Any work can be described by a record -- an annotated bibliography can function as a record for all manifestations of a work. A "cataloging" record for a work is not so much a logical impossibility but a technical nightmare. However, within the confines of authority control records, a great deal of information can be included. The elements of such a work record would derive from the bibliographic records as happens now, but along with spelling and variant access issues, one could certainly add information about relationships and bibliographic history. Difficult for a card catalog? Sure, but not necessarily so with an automated environment, where data can be interfiled and mixed on the fly, and be dynamically updated. Perhaps we are too fixated with the logistical constraints of card catalogs. "MAchine-Readable Cataloging" is a term that highlights only a few narrow aspects of computer technology. The inherent capabilities of computers to reorganize and collect information on a massive scale, provided initial data is coded correctly, inevitably lead one to a concept of "MAchine-Generated Cataloging." A heady concept? Not really, as we have seen numerous complaints about machine-generated descriptive displays for OPACs. Recently, I was in the position of fine-tuning default display options. A flick of switch -- that was all that was required to change the displays of a quarter of a million records. Standards are a good thing for descriptive records, but cataloging standards that take into the account of a computer's ability to create all kinds of records, even functional work records that meet specific needs such as tabulating relationships, would be a truly great thing. Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Tue, 26 Aug 1997 23:55:16 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Author added entry for serials If the proposed expansion of the definition of serials is extended to include the successive editions of major legal works, the absence of author added entry in serial records will be a major problem. (This has been mentioned by Aaron Kupferman in another venue.) Not only are there the original authors' whose surnames are often at head of title, but there are the redactors of the successive editions transcribed in 250$b, which are too often used as main entry in monograph records, in blatant contradiction of legal citation. The most frequent change we make in serial records found on RLIN for annual legal symposia held by The Practicing Law Institute, is to add notes and added entries for the chairpersons. (The same chairs often serve for several years, so this does not require a change to the record each year.) There is other information routinely part of monograph records for symposia not in the serial record, such as the number within the series, the date and place of the symposia, and of course the CONSER omission of 260$c. If the definition of serial is to be expanded to include material now catalogued as monographs, serial cataloguing, which has been wandering off on its own, is going to have to be more closely aligned with monograph cataloguing. Otherwise special libraries in general, and legal collections in particular, will not be well served. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Wed, 27 Aug 1997 09:29:54 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Re: Tom Brenndorfer on Gorman and Oddy MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Writing about title access problems, using "Scientific American" as his example, Tom Brenndorfer said: > I think this point speaks to a larger issue that has been of some concern to > me for a while. For something as basic as title access points, it has always > surprised me as to how difficult these have been to me and to other catalogers > to get a handle on. What Tom Brenndorfer writes about title access doesn't make it clear enough what kind of index he has in mind - it sounds like he is talking about a keyword index all the time. This is not the only, albeit the most frequently found approach. Even worse than his example, "Scientific American", are the two most frequently cited research journals, "Science" and "Nature". As alternative examples, I'm throwing in here several screenshots from our OPAC (containing some 450.000 title records). First: the combined keyword/subject heading index, at "science" 1 sciecles 932==>science 23 science * 1 science -- abstracts -- periodicals -- bibliography * 1 science -- awards -- history * 1 science -- bibliography * 1 science -- bibliography -- catalogs * 1 science -- bio-bibliography * 1 science -- computer-assisted -- congresses * 1 science -- congresses * 1 science -- czechoslovakia -- history -- 19th century * 1 science -- data processing * 1 science -- developing countries -- history -- encyclopedias * 2 science -- dictionaries * 1 science -- directories * 2 science -- directories -- bibliography * 1 science -- europe -- history * 1 science -- europe -- history -- 16th century * 1 science -- europe -- history -- 17th century * 1 science -- europe -- history -- 18th century * 1 science -- experiments -- history * 1 science -- germany -- history * Here, you are not going to find the "Science" journal without some persistence. There is, however, the separate periodical title index, this one being a phrase or string index: (a flick of a button switches over to this) 1 sci rsc medicinal chemistry symposium ; 1983 1==>science 1 science abstracts / series a, physics abstracts 1 science abstracts / series b electrical and electronics abstracts 1 science abstracts / series c, computer and control abstracts 1 science abstracts / a 1 science abstracts physics and electrical engineering 1 science abstracts section a physics 1 science abstracts section b electrical engineering 1 science abstracts series a 1 science abstracts series b electrical and electronics abstracts 1 science agronomique 1 science and agriculture 1 science and culture series ; 5 1 science and its conceptual foundations ; 1988 1 science and its conceptual foundations ; 1993 1 science and its conceptual foundations ; 1994 1 science and its conceptual foundations ; 1996 1 science and medicine 1993 1 science and technology 1 science and technology for the technical man in management 1 science at work ; 1989 here, "science" is only one entry, and the one you are looking for. At the "scientific american" position, we see this: 1 scientia sinica series b chemical biological agricultural medical and ea 1==>scientific american 1 scientific american library 1 scientific american library ; 9 1 scientific american library ; 20 1 scientific american library ; 34 1 scientific american library ; 35 1 scientific american library ; 42 1 scientific american library ; 49 1 scientific american library ; 59 1 scientific and engineering computation ; 1992 1 scientific and technical aerospace reports 2 scientific and technical information publication 1 scientific and technical journal fujitsu 1 scientific and technical memoranda european space agency 2 scientific and technical memorandum 2 scientific and technical report 1 scientific and technical reports // euratom 1 scientific and technical reports european space agency 1 scientific and technical review eldo esro 1 scientific annual report gbf, gesellschaft fuer biotechnologische fo and you get the idea. Given the right software with the appropriate switches to flick, it is all a matter of indexing. Changes are possible anytime, too, just like with displays. (Index reorganisation can take a few hours, though) Tom Brenndorfer then goes on writing about the topic of work records and series authority records, > Also, I think there is some confusion on the nature of records for works. > Authority records for titles are, in a rudimentary fashion, work records, not > just administrative control records to ensure consistency in the spelling of > access points. I intend to get back to this in Part 3 of my "Linking" series, later this week (hopefully). B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Wed, 27 Aug 1997 09:52:45 -0400 Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: "David P. Miller" Subject: super work records Tom Brenndorfer spoke very well to a question that I've been mulling over, regarding the idea of "work records": "Also, I think there is some confusion on the nature of records for works. Authority records for titles are, in a rudimentary fashion, work records, not just administrative control records to ensure consistency in the spelling of access points. I don't think they are intended to be a front-end substitute for descriptive records of unique bibliographic items--they are "post-factum" records, springing from the need for authority control. Moreover, I think much of the discussion of authority records becoming work records have been on the need to tabulate relationships between works and to outline bibliographic history, not to create a parallel descriptive record encompassing all physical manifestations [...] Any work can be described by a record -- an annotated bibliography can function as a record for all manifestations of a work. A "cataloging" record for a work is not so much a logical impossibility but a technical nightmare. However, within the confines of authority control records, a great deal of information can be included. The elements of such a work record would derive from the bibliographic records as happens now, but along with spelling and variant access issues, one could certainly add information about relationships and bibliographic history. Difficult for a card catalog? Sure, but not necessarily so with an automated environment, where data can be interfiled and mixed on the fly, and be dynamically updated." This idea, if I understand him correctly, of creating a master record for different manifestations of a work, based on machine-collated information drawn from bibliographic records, is intriguing. As I read the "super work" record idea, though, it goes beyond this to include works with derivative and shared-characteristic relationships, to use Tillett's categories. These are not generally considered manifestations of the same work (leaving aside, for the moment, the disagreement about performances of music). Fattahi's Prototype Catalogue of Super Work Records includes an entry for the _Arabian Nights_ , an anonymous classic. Here are several works related to the _Arabian Nights_, which might be usefully related in a catalog display: 1) The musical composition _Scheherazade_, both its scores and performances. This piece has a shared-characteristic relationship with the _Arabian Nights_, in that it takes the latter as its subject. (It is not a setting, abridgement, edition, translation, amplification, etc.) 2) The short story "Dunyazadiad," by John Barth. This is a work of metafiction, told from the perspective of Scheherazade's younger sister "Doony". It also has a shared-characteristic relationship, in that it takes the telling of the _Arabian Nights_ as its subject. 3) The unpublished play, "Scheherazade's Sister", an adaptation of the Barth play, performed in the mid-1980s by TheaterWorks, a Boston theater company I was associated with. Not an adaptation of the _Nights_, but does have a shared-characteristic relationship with it. 4) Recorded performances of the playscript. These would not be performances of 2), may or may not be separate works from 3), and have nothing to do with 1). These recordings are three degrees of separation from the _Nights_. The relationships between these works could be made clear to the user through a kind of "super work" record -- one which is neither a descriptive bibliographic record, nor an authority record. This sort of "super work" record would be a kind of switching mechanism, establishing links between members of a bibliographic family. It could indeed express the "abstract work" to an extent that a descriptive record can't -- because, as I stated here before, an "abstract work" cannot be delimited, or described in any detail apart from a manifestation. But the "super work record" as switching mechanism could express something of what a progenitor work -- the _Arabian Nights_ -- has come to represent in world culture by aggregating its manifestations and related works. It could provide, for the end user, the basis for the type-of-work display that Martha Yee posits on p. 26 of her paper. I'm addressing this because I think that the need to aggregate members of a bibliographic family is real, and the activity would serve a useful end -- but the proposal to create descriptive records for "abstract works" doesn't seem to meet that need. David Miller Levin Library, Curry College dmiller@curry.edu ========================================================================= Date: Thu, 28 Aug 1997 11:10:35 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Jurisdiction in imprint The AACR2 provision which allows cataloguer judgement to be used in deciding whether a city needs jurisdiction transcribed or supplied in imprint has lead to great variety in practice. The rule provides the same rationale for transcription as for supplying, so logically the jurisdiction being on the prime source should not affect what is given in the description. In fact, cataloguers almost always transcribe the jurisdiction if present, but supply it only when they feel it is needed. This one often sees "Boston, Mass.", but one usually sees just "Boston" rather than "Boston [Mass.]", while if the rule were being applied as written, they would occur equally. Further variety is caused by the LCRI which instructs that a postal code is to be transcribed rather than supplying the AACR abbreviation for the missing jurisdiction, thus adding "Boston MA" to the mix. For purposes of standardization, easier Boolean searching, and international exchange of records, shouldn't AACR2 1.4C3 be simplified to instruct that jurisdiction be transcribed or supplied using the abbreviations? Postal codes should not be considered jurisdictions. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ ========================================================================= Date: Fri, 29 Aug 1997 14:50:52 METDST Reply-To: B.Eversberg@tu-bs.de Sender: "International Conference on the Principles and Future Dev. of AACR" From: Bernhard Eversberg Organization: UB der TU-Braunschweig Subject: Linking. Part 3 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT To link or not to link, and how ------------------------------- Part 3: The Part->Whole relationship Contents 1. Introduction 2. Remarks on "work" and "item" 3. Situation A : Work manifested in several physical parts 4. Situation B : Several works manifested in one physical volume 5. Synthesis : A and B are the same 6. A note on authority records vs. bibliographic records 7. "Work records"? A new suggestion 8. Relief for Germany 9. Examples 1. Introduction This part has two purposes. First, to explore new ways of dealing with Part->Whole relationships and second, to show how exchange between German and Anglo-American agencies can benefit. In actual fact, the latter was the primary motivation for writing this series of postings. As it turns out now, it is not the dominating aspect. I was tempted to use some German examples, but the problems we have with USMARC data can be made clear enough from within the AACR2 and USMARC framework, pointing out where and why changes and improvements are desirable, esp. when taking the emerging new concepts of relationship linking into account. It appears that these concepts could be implemented with very few and very small measures - which would help remedy our trouble with USMARC as a side effect. The handling of Part->Whole relation- ships is largely a format question. Both rules and format do make provisions for almost everything needed, but as in many cases, there are useful options or alternatives which have never been used. Discussion of "multilevel hierarchies" is not included. This would make up Part 4 - which presently I have no intention of writing. Of course, as soon as you can link something to something bigger, you can extend this to more levels and construct tree-like hierarchies. There may be errors in my interpretation of rules or USMARC data. Feel free to correct me, it is like a foreign language for all of us here. Beginning with: 2. A few remarks on "work" and "item". The time seems to have finally arrived now where the concept of the "work" has become a real focus of attention in the world of cataloging. Whatever the wording of the various suggested definitions, for example by Martha M. Yee (on p. 33/34 of her paper "What is a work"), they all revolve around the idea of the work being a "product of intellectual or artistic activity ... which can stand alone as a publication...". A publication can thus be a physical embodiment or manifestation of a work, but it IS not itself the work. Cataloging has focused on "the piece in hand" for a long time, which is always a "copy of a manifestation of an expression of a work" (IFLA "Functional Requirements for Bibliographic Records (FRBR)" study). The result is that catalog records (USMARC or other) always contain elements from all these levels. Asking "What's the work in this publication" often reveals that either the piece in hand is only a part of what one would call a work, or it may be looked upon both as self-contained AND as part of some bigger intellectual product, or it contains more than one product of intellectual activity, where each one could conceivably stand alone as a publication and sometimes does. Should cataloging become serious about work orientation, we have to conclude that the unwritten principle of "one book - one record" is inadequate. It may remain the most frequent situation, but two other situations need to be distinguished: A) One physical item is a manifestation (or part of a manifestation) of more than one work Typical case: Volume of a series B) One physical item is host to (contains manifestations of) more than one work Typical case: Audio CD containing recordings of several pieces. These situations are being catered for in AACR2 chapter 13 "Analysis". The way these rules are translated into MARC records is still directed by card-oriented, not work-oriented thinking. One item - one main entry card, that's the underlying principle. Additional cards (added entries) are then made as needed, using the appropriate fields for headings. A "multivolume item" is regarded by AACR2 as one item. 3. Situation A : Work manifested in several physical parts ---------------------------------------------------------- First, some terminology : Series and multivolume monographs ----------------------------------------------------------- English and German terminology are awfully incongruent. German catalogers' jargon uses "mehrbaendige Werke" (multivolume works) to comprise both of the above (!). The English "series" is often misunderstood and equated with "serial", because the German word for "serial" is "Serie", and there is simply no exact equivalent for "series". One would have to say "mehrbaendiges Werk mit Stuecktiteln" - multivolume work with distinctive title volumes. But that's not quite the same. Because: Works by personal authors, since their lifetimes are all finite, are always regarded in the sense of "multipart items" in the sense that they will have a finite number of separate parts ("mehrbaendiges begrenztes Werk") whereas AACR2 quite often treats these as series, just because there is no indication as to exactly how many volumes there will be. The traditional term "mehrbaendige Werke" should rather be abolished because a "work" (= "Werk"), in current understanding, cannot consist of volumes. The work is an abstract entity. Only a manifestation of a work is physical and thus can appear in several volumes. The term "mehrteilige Veroeffentlichung" (= multipart publication) has been suggested as a replacement, but is not widely used yet. The term "collection title" for the general title (found on every part) of a series, multivolume monograph or serial translates as "Uebergeordneter Gesamttitel" (superimposed or superordinate collective title). Shortcomings of AACR2 cataloging, as we perceive them in Germany, boil down to the following, and I indicate possible solutions in terms of USMARC instead of describing them epically or using any of the German formats as examples: [1] Volumes with no titles or indistinct titles are listed in a note (MARC 505, based on 13.4). This note, however, can contain other information like tables of contents, and there is not sufficient formatting in the note text, nor other coding in other fields, to make it apparent for software that this record describes a multipart item. German catalogers mostly think the use of the 505 for multiparts is not a good idea at all, and no German systems use this technique. Instead, we have linked subrecords for ALL parts of a multipart, whether they have a distinctive title or not. Why? In shared cataloging, you can attach holdings to the volume records instead of just one location symbol to the title record. This way, the union catalog database can show exactly which volumes are located where, which is supposed to make interlibrary loan more efficient. If the MARC world cannot warm to this idea but wants to retain the 505, what can be done? The easiest solution appears to be the use of indicator 2 (presently undefined and always blank): a value of '2' could mean "multivolume". The formatting (punctuation) within $a might be made unambiguous to enable software to extract volume designation, date, and pagination for every volume listed (provided someone puts them in). AACR2 talk summarily of "contents notes". Maybe the term "volume list" or "list of parts" or something to this effect should be introduced to denote the case of a multipart publication. On the other hand: AACR2 leave it to the "cataloging agency" (13.1A) to decide if and when they want to make contents notes. Specifically, AACR2 do not strictly dictate a distinctive/nondistinctive title differentiation. Thus, nothing in the rules would stop an agency from abolishing the contents note for multiparts and introduce something similar to German practice. In card printouts, a contents note could still be generated by software. No rule change is called for, but changes in MARC encoding are desirable. [2] Volumes with distinctive titles are very often (but in our view should always be) cataloged under these titles (i.e., given their own, separate records) with a 400 or 800 referring to the title of the series. Of course, there is the occasional argument over whether a volume title is distinctive or not. Cases of doubt are probably more often decided pro-nondistinctive (it's less work), but the opposite is more reasonable. And there are those cases, where the first volume is non-distinctive, but after a while a distinctive volume 2 comes out... AND there is the whole issue of classification decisions influencing the cataloging treatment decisions (in our view, a most distressing aspect). Well, this jagged borderline could be completely eliminated. New suggestion for USMARC (not for AACR2): The series as such should be considered a work in its own right and cataloged as such (like a serial record). Presently, AACR2 do not prescribe a bib record of its own for the series (13.3). Instead, in USMARC, an authority record is usually made for the series on the basis of 26.5. (More on this, since it is quite fundamental, see under 6.) We find this quite exotic. And annoying, because this way we never get to see the series record. Why not? The LC authority file is not loaded into any German systems for reasons of incompatibility and incongruence, and we would have no use for the vast majority of the records anyhow. That means USMARC bib records, when converted and merged into our systems, are essentially incomplete: we lack all the references contained in the authority records. [3] Subfield $p in 245 is a bad compromise. Software cannot determine if the part title given in $p is a distinctive one. This technique, rarely applied as it is, had better be abandoned. On what rule is it based anyway? Some background: The solution typically implemented in Germany consists of -- One main record for the series or multipart item (This is a bibliographic record, not an authority record!) It contains no links to the subordinate records. No holdings are attached to this main record. It is not a "work record", for every manifestation gets its own record. -- One subordinate record for every part or volume, whether it has a distinctive title or not. These subordinate records are linked (upwards only!) to the main record via its IDNr. In the union catalog databases, holdings are attached to these records so as to accurately reflect which library has which parts of the publication. Volumes with non-distinctive titles just have no title field - as you may well guess. All other fields, like date, pagination, names of persons relating to that volume, whatever, it can all be there. For us, it is never relevant for cataloging whether or not titles of a series are classified together. Shelving decisions must not, or so we think, influence cataloging decisions. (Different libraries can make different shelving decisions, which would compromise the shared database. But who am I telling this?) Cataloging or OPAC software can present the comprehensive work with all its parts (via the index of IDNrs) but also, when a subrecord is hit, the main record can be displayed with just the subrecord in question. And it may come as a surprise: This solution exists in USMARC as well, at least in theory: Field 773 can be used to link a volume record to a collection record: 773 $w IDNr of collection [ $a heading $t collective title ] $g volume information For all intents and purposes apart from this, the volume record is a regular USMARC bib record. A paper authored by Sally McCallum describing this in all detail was made available to me ("Multilevel descriptions in USMARC", 20 Jan 1997). But to the best of my knowledge, nobody is using this technique. The approach of least change for USMARC users would be to go on using 4XX and 8XX for series access points, but then change the series authority records into bibliographic records. This would make the indexing easier too. To establish real links ("data links") there's the possibility of introducing $w into the 4XX and 8XX. Another solution: (see Example 2) If that's found too difficult, one may consider using the 787 as described in Part 2, and use 'p' for 2nd indicator. This way, one would have all relationships between bibliographic records implemented in a uniform way in just one additional field. And all multipart publications could be treated alike in this concept - the cumbersome contents note could be altogether abolished for multiparts. (Software can, of course, using the links, assemble a contents note for card output or display. Software could also produce an added entry out of a 787 just like from a 400 or 800 for the structure is virtually identical for this purpose. This means one could even avoid redundancy.) Navigating the relationships (i.e., to write software for this purpose) would be made easier this way than with any other solution (see Part 2). Every physical part would have its own record in this solution, linked to a common main record. Besides, every physical part can relate to a separate work, and/or represent a part of a larger work, to which the main record would be related. The 787 being repeatable, every physical part can have links to (be a part of) more than one comprehensive work. Circulation (copy) records can be, quite naturally, attached to the subrecord describing the physical part. 4. Situation B : Several works manifested in one physical volume ---------------------------------------------------------------- The most typical examples are in music, but festschriften or conference volumes can be used as examples too. For the latter two, however, hardly any library is doing analytics for all contributions in such volumes. Everybody will know the structure of USMARC music records: (showing only those parts relevant to our discussion) 100 10 Composer of first piece 240 10 Uniform title of first piece 245 10 Title of container 505 Contents: composers and titles (not in authority form!) 511 Performers and conductors, not in authority form 700 11 $aPerformer (R) 700 11 $aConductor (R) 700 12 $aComposer of 2nd piece $tTitle of 2nd piece (R) 700 ... For the 700$a$t fields, indicator 2 is set to 2 (analytic). These 700 fields can serve for analytical added entries (13.2) as well as for "In" analytic entries (13.5). In practice, only the former is done, because the "In" analytics would be somewhat deficient: The pain in the neck is, of course, that no program can determine from this which conductor and performer(s) belong to which of the pieces listed in the 700$a$t entries. Results of boolean searches for performer AND composer may therefore be irrelevant because the performer and composer are in fact unrelated and only happen to be listed on the same CD. Keyword searches for composer and title word are even more disappointing. The simplest solution of this dilemma would be to make one separate record for every piece on the CD, with its own 100, 240, 245, and 700s for the performer(s) and the conductor belonging to this piece. These analytic records would be linked (upward) to the main record for the CD. This main record would not have a contents note but just a 245 and the descriptive data necessary to identify the CD. Unfortunately, the existing, badly convoluted USMARC music records cannot be dissected by software into a main record plus analytic records, for the reason mentioned. It is only possible to produce incomplete analytic records with composer/title in a 100/240, out of the 700$a$t fields, the other 700s and the 511 would have to remain in the main record. We went through this exercise and produced a music database of some 40.000 records arranged in this way. The advantage is that the anyword boolean search for composer AND title word then yields only relevant titles. Otherwise, when keyword indexing original USMARC, you get a hit when searching for "mozart and trio" when in fact a CD contains a Mozart piano sonata and a Beethoven trio. (You can access this database by telnet if you want.) The new, work oriented solution could be almost exactly like the link structure of Situation A: here, however, the main record represents the physical volume, whereas the upward linking subrecords are relating to separate works, representing manifestations that do not stand alone as a publication. Holdings (or copy records in local circulation systems) would be attached to the physical volume, i.e. the main record. No rule change is called for: Chapter 13 on analytics says nothing on how to implement analytic records in any format. 5. Synthesis. In other words: If we turn to a work-oriented approach, we have to focus on logical entities rather than physical items, on identifiable intellectual products rather than "pieces in hand". If we do that, Situations A and B become essentially the same. Two different solutions are no longer needed. As early as 1989 Patrick Wilson stated that ".. the control of items is achieved at the expense of the control of works." ("The Second Objective" in: "The conceptual foundations of descriptive cataloging" / ed. by E. Svenonius. - San Diego: Academic Press, 1989, p. 8) That can be rectified. The cost in terms of rule changes is zero (not rule usage and interpretations!), the cost in format implementation is rather low. The real problems are in the legacy data. No complete conversion is possible, whatever model is chosen. (One sometimes gets the impression we are far beyond the point where ANY change is possible.) 6. A note on authority records vs. bibliographic records (see 3 [2]) ----------------------------------------------------- Authority records owe their existence to the card concept of references, AACR2 chapter 26. Faithful to the letter of rule 26.5, authority records are created for series. The online equivalent of a reference need not look like a card reference, however. It can function in different ways, depending on the software. The same effect can be achieved by having a bibliographic record for the series instead of an authority record. (Other than names and subjects, or "works" for that matter, a series is a bibliographic entity!) This would avoid the awkwardness of using different fields and subfields in authority records and in bib records (100 $t instead of 245, 643 instead of 260, etc.). Which, in the view of a database programmer, is a design blunder in USMARC. Rule 26.5B (references for serials) is analogous to 26.5A for series. Yet, for serials with no distinctive title volumes, no authority record but a bib record is made - for obvious reasons: there would otherwise be no catalog entry for the serial. Or the authority record would more or less duplicate the bib record, only with different tagging (ouch! says the database programmer). And of course, one cannot attach holdings to an authority record. Thus we have two perfectly analogous sections of a rule (26.5A/B), yet their treatment in the format is very different. The only reason will be that this was the easiest way to produce the appropriate card headings for series volumes (100$a$t directly provides the heading). It is not the only possible way: if we had a bib record for the series as such, the heading could still be produced, using the 100/240 or 100/245 of the series record. In terms of linking, the volume record would ideally contain the record ID of the series record only, which is enough to produce the heading when needed. Where this kind of data linking is not possible, the volume record can retain exactly the structure it has now, with the parts of the 400 or 800 composed out of 100 and 240/245 of the series record instead of the 100$a$t of the authority record. It should not be difficult to convert series authority records into series bibliographic records. The difficulty will be (ouch! again) that these records, like most authority records, are doing double duty as subject authority records. To have, conversely, a series title record (as bib record) double up as authority record is questionable and would surely be rejected. To have two records of equal content but different structure to serve these different purposes is not attractive either. That, however, is what we frequently have with uniform title records. Maybe, if it comes to a work-oriented approach, the whole dichotomy of bibliographic vs. authority records should better be re-evaluated. This is now speculative, but authority records could be restructured to look largely like bibliographic records, lacking a 245 and 300 etc. That would eliminate the difference in designator definition between the two formats, always VERY annoying for implementors. Eventually, the authority format could be phased out altogether. All kinds of links are made easier. Speculation!? 7. "Work records"? A new suggestion. A work record cannot be a bibliographic description because a work, by any definition, is not a bibliographic entity. Certainly, what is not envisaged is a combined description of various manifestations in different physical forms - a "cataloguing nonsense" as Gorman and Oddy would have it. It is the elusive quality of the "intellectual product" that needs to be pinpointed, to provide an anchoring point for new types of links, as described in Part 2, that is to group records of manifestations in useful ways. Uniform title or series authority records have aspects of what a work record might be in that they provide standard names for works. A bibliographic record for an original edition can also be regarded as representing a work. This function of "representing a work" does not call for a new type of record then, but it can be made an additional feature of existing records. The simplest implementation would be to define an indicator, and I'm bold enough to suggest position 8 or 9 in the leader (hitherto always blank), to say "this record, among other things, represents a work". This byte would then make the record eligible for all those links, anchored in 787 fields in all kinds of other records, each of which, in turn, could also have this same feature of "representing a work". This solution appears quite alluring because, all of a sudden, we would have work records for most anything we need one for. We just have to flick that indicator switch to make a record an official "work record". And then, we could go ahead and get serious about linking, straight away... Will there have to be a whole new chapter for work records and linking principles in AACR? Nice though it would be to have it, who is going to formulate it, given the present pre-experimental stage of things. And to set a standard too early is sometimes not a good idea. Just what additional elements will be required in a work record, what additional access points will there have to be? The discussion has only just begun. The Toronto conference will certainly be a good opportunity to discuss options and problems, but not to hammer out rules. A definition of "work", maybe. ------------------------------------------------------------------------- 8. Relief for Germany: Series authority records have an 'a' in position 16 of the 008 fixed field and thus are easily selected. Position 13 even indicates whether or not the series is numbered ('a' = numbered, 'b' = unnumbered), and position 12 has an 'a' or 'b' for series / multipart. Therefore, even if nothing else happens, at least we in Germany can go ahead and restructure those records for our databases, i.e. turn them into series main records! Conversely, series main records produced in Germany could be restructured into series authority records for USMARC, BUT (ouch!) we have no indicator saying the series is a multipart item without distinctive titles and would thus have to become a bibliographic record with a 505. In most cases, this fact can be derived from the subrecords (having no equivalent of a 245 then), but this is not quite straightforward to program. 9. Examples ----------- Much of the following is probably utterly unrealistic, but be that as it may. Regard this material as something to chew on, and as illustrations for the above. (All examples are from real life, but shortened to the essential parts.) Ex.1 : for Situation A[1] ------------------------- a) Bib record of a multipart item without distinctive titles 001 67026020 100 10$aKnuth, Donald Ervin,$d1938- 245 14$aThe art of computer programming$c[by] Donald E. Knuth. 260 0 $aReading, Mass.,$bAddison-Wesley Pub. Co.$c1968- 300 $a v.$billus.$c25 cm. 505 1 $av. 1. Fundamental algorithms.--v. 2. Semi-numerical algorithms.--v. 3. Sorting and searching. b) New 2nd indicator for the 505: (saying "this is a volume list") 505 12$av. 1. Fundamental algorithms.--v. 2. Semi-numerical algorithms.--v. 3. Sorting and searching. c) A better solution: series main record + title record (This is what might be produced out of German records) 001 67026020 100 10$aKnuth, Donald Ervin,$d1938- 245 14$tThe art of computer programming 260 $aReading, Mass.,$bAddison-Wesley 300 1 $av. 1- 001 85028675 //r955 100 10$aKnuth, Donald Ervin,$d1938- 245 00$aFundamental algorithms /$cDonald E. Knuth 260 0 $aReading, Mass. :$bAddison-Wesley,$cc1968. 300 $axxi, 634 p. :$bill. ;$c24 cm. 800 1 $aKnuth, Donald Ervin,$d1938-$tThe art of computer programming ;$vC. or, better (instead of the 800) 787 1p$w(DLC) 67026020$v1 Ex.2 : Situation A[2] : A multipart (or series?) WITH distinctive titles ----------------------------------------------------------------- a) The authority record, as it is now: 001 n 84717754 100 10$aKnuth, Donald Ervin,$d1938-$tComputers & typesetting 400 10$aKnuth, Donald Ervin,$d1938-$tComputers and typesetting 640 1 $aComplete in 5 v.$zCIP t.p. verso of v. A 643 $aReading, Mass.$bAddison-Wesley b) The series main bib record, as it might be: 001 n 84717754 100 10$aKnuth, Donald Ervin,$d1938- 240 10$tComputers & typesetting 245 10$tComputers and typesetting 260 $aReading, Mass.,$bAddison-Wesley 300 1 $aComplete in 5 v.$zCIP t.p. verso of v. A c) One of the 5 (distinctively titled) volumes: 001 85028675 //r955 100 10$aKnuth, Donald Ervin,$d1938- 245 14$aThe METAFONTbook /$cDonald E. Knuth ; illustrations by Duane Bibby. 260 0 $aReading, Mass. :$bAddison-Wesley,$cc1986. 300 $axi, 361 p. :$bill. ;$c24 cm. 490 1 $aComputers & typesetting ;$vC 800 1 $aKnuth, Donald Ervin,$d1938-$tComputers & typesetting ;$vC. The 800 is a textual link to the main rec (its 100 and 245 combined) To introduce a data link, the 800 might be extended like this: 800 1 $aKnuth, Donald Ervin,$d1938-$tComputers & typesetting ;$vC. $w(DLC) 85028675 Or, more radical, a data link of the Part->Whole category: 787 1p$w(DLC) 85028675$vC The 800 would be, in principle, redundant, but might still be supplied, minus the $w, for those systems that don't understand the 787. ............................................................................. Consoling remark (to whom it may concern): The USMARC flaws cannot be blamed on the designers. Their task was to design a format for data EXCHANGE on magnetic tape, not a DATABASE FORMAT. Later on, it was adopted as a database format largely by default because no database standard format existed, and it appeared as an advantage to have the same structure locally and thus less conversion programming effort. And now: have fun cataloging this series! (Or is it a 3-part item?) B.E. Bernhard Eversberg Universitaetsbibliothek, Postf. 3329, D-38023 Braunschweig, Germany Tel. +49 531 391-5026 , -5011 , FAX -5836 e-mail B.Eversberg@tu-bs.de ========================================================================= Date: Sat, 30 Aug 1997 01:31:07 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Extended Markup Language (XML) and Internet metadata The following article appeared in the PCWeek Internet site. Some of the points raised here, and on the followup sites at Microsoft and Netscape extend upon some points in Mark Ridley's contribution "Beyond MARC." Of particular interest is the concluding call for more specific metadata formats, although it is equally intriguing to note the parallels between the development of a cataloging code with that of an electronic metadata format for the unruly Internet. Tom Brenndorfer Guelph Public Library ********************************************************************* August 25, 1997 Intersights Developing a card catalog for the expansive Web By Eamonn Sullivan Making the Web more like a library and less like most bookstores has been a goal of Web researchers for years, but we'll get closer to that goal in the next few months with the development of a tool called Resource Definition Format, or RDF. The difference between a good library and a bookstore is the card catalog. When looking for a book in a bookstore, you have to make do with the usually simple organization imposed upon the information by the owners. Books are organized in broad categories, such as fiction and nonfiction, history and philosophy, or science and science fiction. If you know exactly what you're looking for, you can ask the salesperson. Otherwise, you have to browse, using the order imposed by the seller as a guide. The only improvement on this system offered by the Web is the full-text search, which is of limited value. With most search engines, pages on Barney and the Smithsonian's dinosaur pages have equal weight. In a library, the card catalog (whether electronic or not) gives you a tool for more sophisticated searches, using author, title and subject to narrow down your search. The card catalog is called metadata--information about information. The ability to add metadata to Web content has been available for a long time, using things such as metatags. But the approaches, with the possible exception of the PICS rating system, have been somewhat haphazard. The emergence of XML in a more or less solid form earlier this year has provided a more comprehensive framework for metadata, prompting several organizations to propose solutions based on XML. The main proposals have been XML-Data from Microsoft (which is available at www.microsoft.com/standards/xml/xmldata.htm) and MCF (Meta Content Format) from Netscape (available at www.w3.org/TR/NOTE-MCF-XML/). Both proposals provide for a sophisticated method to describe the structure of information, such as properties about authorship and relationships between objects. This week, a working group under the auspices of the W3C organization will meet in Redmond, Wash., to begin hammering out a specification that will take the best parts of XML-Data, MCF and PICS. The resulting RDF specification, if used widely, will enable more efficient searches and exchanges of information between organizations. For example, detailed descriptions of information structures and content will enable Web browsers to provide more useful site maps for navigating through large sites and allow for more customization of the way information is displayed. The key, of course, is widespread use. RDF will likely be open-ended; each publisher could use RDF to create its own terms to describe information. Until there is something like the Dewey Decimal System for the Web, looking for information across sites will likely remain a somewhat manual process. The bookstore and library metaphor is also true in another way for describing the Web: The content of a bookstore is more fluid than a library, making card catalogs a lot more difficult to maintain. Maintaining metadata on a large, rapidly changing site would be an arduous task. RDF is just a first step, but an important one. After that step, it will become necessary to agree upon more specific metadata formats and terms, such as creating something like Microsoft's Channel Definition Format for metadata. With an agreed-upon metadata format, your Web browser may be able to do more for you automatically, making the Web a much more useful place. Do you use metadata now? Tell me about it at esullivan@zd.com. ========================================================================= Date: Sat, 30 Aug 1997 02:45:36 UT Reply-To: "International Conference on the Principles and Future Dev. of AACR" Sender: "International Conference on the Principles and Future Dev. of AACR" From: Thomas Brenndorfer Subject: Beyond MARC [... et al.] A number of interesting issues arise out of Mark Ridley's contribution "Beyond MARC," along with parallel work being done on creating metadata for the Internet, using SGML or XML, etc. What are the objects? While we separate concepts such as works and bibliographic items and physical copies, I think their treatment as "objects" (as Mark Ridley indicates) is an interesting point worth getting a handle on. Despite the physical connotation, the term "object" covers all those things in cataloging which can be broken down into discrete, defineable units, whether abstract or not. Of particular concern is the way different objects have been tagged in overlapping ways. "Entries" generally meant physical cards entered into the catalog, with one given predominance as the main entry. "Main entry" is also used in different ways to describe the heading for the item or work represented on the card. The "main entry heading" marks the card meant to be the main entry in the catalog. However, while a catalog card can be the object so identified in this way, the general confusion seems to be around the concept of the main entry heading as also identifying the "work" in the bibliographic item. In particular, author, uniform title, and title proper are also seen to identify a discrete, specific unit -- the main work. As such, I find the idea of eliminating the "main entry" to be terribly confusing. We could eliminate the "main entry" for the cataloging card (i.e., no card entry is given predominanc), but we can't eliminate the "main entry" of the work, since that is how the work is identified (every work has a "main entry" in this sense). And then we have the bibliographic item, but it too seems to have something that looks like a main entry!!! Priority is given to one title (the title proper), and we often see this in conjunction with the author as providing a unique identifier for the physical item, almost as an inventory label. What an extraordinary confluence of concepts at the top of a cataloging record!! But once we identify the specific, separately defineable objects involved, what is the "main entry" really? Is the main entry heading not a naming convention? Is it not like a naming convention on the Internet where every site a numeric designation and a corresponding spelled out form that is generally intelligible. Data processing requirements vs. cognitive requirements. Any record can have a control number attached, but is not a name required to make sense of what the record is all about? And what if that name also dovetails with other requirements in a catalog such as filing and sorting and collocating? Is this not the raison-d'etre of a name-title heading? [Of course we could have alternate headings for different purposes, but this would increase unnecessary duplication and redundancy, and lead to an increase in confusion]. From this point of view, one could very well say that the main entry is not lost and forlorn in a computerized world -- the main principle behind a main entry (as a chief identifier written in an intelligible form, whether it be for entry card, work, or item) is similar to the core elements uniting different components in a computerized world, such as the recipient addresses for this e-mail. Some of these concepts work well together with points made about SGML and XML -- the tagged formats for designing metadata on the Internet. One point that struck me as significant is that relationships can only be defined after the content of the objects has been defined. We do that to a large extent in cataloging, moving from a description of an item to an identification of the works involved, but we don't really define or label the linked result of our activities -- a catalog card seems like such a unitary thing, even though there are most certainly layers of "objects" involved. Another interesting point about SGML and XML (and I think Ridley alluded to this) is that of the "tree" structure of this type of tagged code, with its potential for nesting, subsumption, containment, and contextual relationships. Even while a tree structrure does this, it shares with the more flattened MARC format a linear structure that a program can parse incrementally to a defined endpoint. An additional benefit for a tree structure would be incremental displays (programmed or selected by the user). A tree could structure both the elements found in objects, and the relationships between objects. (I suppose a "web" concept would extend this to deal with more complex relationships between objects, hyperlinks joining together farflung catalog records). The filing and collocating function of a card catalog seems to do this to some extent. We can define a block of cards under a common uniform work heading as being a complete, dynamically created (and ephemeral) work record which draws together through specified relationships different manifestations of the work. Extending and adding to this elementary structure does not seem to be a big deal, really. And an automated environment would seem to one that would gladly absorb more complex interrelationships that cannot be easily expressed in a card catalog format -- namely through tree structures or something of that sort. On a final point, another term that I thought worthy of focus is "lexical data type," a term applied to the myriad date and number formats, and the way they are parsed or interpreted by computer programs. I would certainly like to see filing rules, originally defined in an analog environment, made with at the very least a reference to the computerized world, perhaps even with a reference to particular ISO and other standards. The undifferentiated, character by character filing done in most computerized situations seems like an unnecessary complication all because certain "lexical data types" were not translated in any way. Any revision of AACR2, as far as I can tell, would be of limited use if there was no attempt to take into account what can and cannot be done with a computer. Tom Brenndorfer Guelph Public Library thomasb@msn.com ========================================================================= Date: Fri, 29 Aug 1997 23:33:34 -0400 Reply-To: mac@slc.bc.ca Sender: "International Conference on the Principles and Future Dev. of AACR" From: "J. McRee Elrod" Organization: Special Libraries Cataloguing, Inc. Subject: Seriality and Main Entry (fwd) I have Aaron Kupferman's permission to forward this to the aacrconf list, since it relates to a topic which has been discussed here. As most of you know, Aaron catalogues law at LC, and through his participation on autocat, has helped give LC a much more personal face. Mac __ __ J. McRee (Mac) Elrod (mac@slc.bc.ca) {__ | / Special Libraries Cataloguing HTTP://www.slc.bc.ca/ ___} |__ \__________________________________________________________ -------- Forwarded message -------- Date: Fri, 29 Aug 1997 10:36:12 -0400 From: Aaron Kuperman Subject: Seriality and Main Entry One of the adverse effects of increasing the definition of what is a serial would be that serial catalogers traditionally are reluctant to make entries (main or added) for human authors or editors. This is somewhat reasonable since a "serial" may be immortal whereas humans are generally accepted to be mortal. However there is no rule at present that justifies this interpretation, but rather it is an "unwritten" tradition that need changing (especially at LC, where serials are NOT cataloged by catalogers who work primarily with legal materials). To users of legal materials, it is totally irrelevant if the author of "Corbin on Contracts" or "Benjamin on Sales" has been dead for a generation. The book is cited based on the original author, and that fact alone requires a serial catalog to make added entries for human authors. The citation to the original author of a work is common to many countries legal publishing patterns, and has to be taken into account in serial cataloging. What is needed is to educate serial catalogers to the importance of including a heading for the human responsible for the work if it is prominent on the title page, since that is what users cite to for legal materials. This doesn't require a conference in Toronto, or yet another cataloging code. Aaron Wolfe Kuperman (LC, Social Sciences Cataloging Division, Law Team) This is a private note and is not an official communication from the Library of Congress.