December 17, 1999
(accepted by the Standing Committee on Standards, Jan. 2000)
John Attig; Linda Barnhart(Chair); Sherman Clarke; Stephen Hearn; Rhoda Kesselman; Chris Martyn; Adam Schiff; Manon Theroux
Assuming an environment of computerized bibliographic control, review the provisions for references stated in AACR2r and the associated LCRIs to determine how the current complement of guidelines for references could be adjusted to accommodate efficiency and simplicity in cataloging as well as the change from a manual to an automated environment. Conduct a review of current provisions with a view to:
- identifying a basic set of references appropriate to different types of headings in most library catalogs, and determining when additional references might be appropriate;
- continuing to meet the needs of a broad range of interests of a widely-shared authority system;
- ensuring the integrity of a file that must be subject to a careful set of file management considerations;
- reducing the number of restrictions and exceptions to an absolute minimum;
- determining if any references should be eliminated or added;
- determining the presentation of reference provisions that would best serve catalogers, e.g., associating them with specific rules, associating them with a specific chapter, presenting them combined together in a single chapter, some other presentation, or keeping the current one.
- Recommendations for rule changes to AACR2r for consideration by CC:DA.
- Recommendations for changes or additions to the LCRIs for consideration by LC.
- Recommendations on how best to present the complement for provisions regarding references to assist ease of use by catalogers.
The Task Group would like to thank Brian Schottlaender, Gary Strawn, Kay Guiles, Crystal Graham, and Joan Schuitema for their helpful comments on earlier drafts of this report.
The Task Group recommends five changes to AACR and twelve changes to the
LCRIs. In addition, we have several recommendations for broader discussion
within the profession, and for issues that other groups should pursue.
One major area that cut across several of the recommendations was the need to articulate the differences between shared authority data and data appropriate for local authority files. Another major area that initiated several recommendations was the notion that some references in authority records may serve the purpose of automated heading checking and maintenance. Several changes were proposed out of concern for users in a more global environment, where they might be searching catalogs not in their native language. The Group also made recommendations to include a few small things missing from AACR2, including the addition of a very few clarifying examples to promote global consistency.
Recommendation #1: Add a sentence to AACR2 rule 21.4D1 parallel to the instruction in AACR rule 21.4D2 regarding explanatory references from the personal name to the official name.
Recommendation #2: Adapt LCRI text to add to AACR2 rule 26.2A2 to give more guidance and thus more consistency internationally.
Recommendation #3: Delete AACR2 rule 26.2B (including all subrules: 26.2B1; 26.2B2; 26.2B3; 26.2B4).
Recommendation #4: Add LCRI text to AACR2 rule 26.3A6 regarding written-out abbreviations, including special symbols.
Recommendation #5: Restrict the circumstances in AACR2 rule 26.4B4 for making references from titles proper to collective titles.
Recommendation #6: Form a new group to re-evaluate and re-write the LCRI for Chapter 26, particularly the section on Linking References, to expand the definition of linking references and the circumstances under which linking references can be made.
Recommendation #7: Delete part of LCRI 26.2 in conjunction with moving that text into AACR2 (Recommendation #2).
Recommendation #8: Remove the exception in LCRI 26.2, Referring from Variant Forms, point (4) for German, Dutch, and related languages.
Recommendation #9: Remove the exception in LCRI 26.2, special considerations, point (3) for Hebrew names.
Recommendation #10: Add a new instruction to LCRI 26.2 to prefer forms corresponding to usage over forms corresponding to qualifiers in constructing references.
Recommendation #11: Add a new section to LCRI 26.2 regarding the handling of variant dates.
Recommendation #12: Move LCRI 26.2B2, 26.3B3, and 26.2B4 (pending the recommended deletion of AACR2 rule 26.2B) to LCRI 26.2. Some of the text needs to be revised slightly so that the mention of name-title references is removed.
Recommendation #13: Revise part of LCRI 26.2C to clarify and broaden applicability of see also references from individuals to the group for NACO libraries.
Recommendation #14: Rewrite LCRI 26.3A3, References from Inverted Names of Government Subheadings (including changing this caption), to allow greater cataloger discretion to construct alternate direct order forms and to cover a parallel, currently unaddressed situation.
Recommendation #15: Delete LCRI 26.3A6 (pending the recommended transfer of wording to AACR2 rules 26.3A6).
Recommendation #16: Clarify and expand LCRI 26.3B-C (See also references and Explanatory references) to guide catalogers to make see also references for heads of state (etc.) bi-directional (from official name to personal name, and from personal name to official name).
Recommendation #17: Rescind the practice described in LCRI 26.5 of adding date qualifiers to all references to uniform titles headings of the form "[Name]. Works ", "[Name]. Selections ", and "Bible "
- Explicitly allow the creation of a reference for the "citation form" of a personal name based on cataloger's judgement without ascertaining usage or providing justification from print sources.
- Implement "spell-checker" functionality in library OPACs.
- Form a group to explore further the issues surrounding references from translated titles: the monographs approach and the music approach.
- Monitor further developments to allow the creation of references in non-Roman scripts.
The Task Group conducted its work primarily through email discussion. We began by reviewing the literature, the report of the Joint Task Group on Streamlining Authority Record Creation, AACR2r and the LC Rule Interpretations. We took as our universe all name (personal, corporate, conference), name/title, title, and series headings, omitting only topical subjects from our review.
Our work was guided by the phrase in the charge "to accommodate efficiency and simplicity in cataloging." Not only did we strive to make the creation of authority records easier; we also wanted to make authority records more effective (e.g., users should get better results; records should be more useful). We hoped to provide consistency and simplicity both in cataloging rules and in the resultant syndetic structure, thus providing clearer and more useful references for online catalog users. We also recognized that "see" references are made for disparate purposes, e.g., (1) to redirect user searches, (2) to enable automatic error checking, and (3) to collocate entries for differentiation (e.g., for acronyms, occasionally for series variants). All these purposes seem legitimate to us, and we have tried to value each type of reference correctly, based on its intended use.
- Indexes in online catalogs versus manual card files
An area that requires fuller discussion, and perhaps research, is an exploration of how indexes in online catalogs differ in form and function from the manual card files that were prevalent when AACR was written. This Task Group has tried to work envisioning indexes and displays in various online systems, giving thought to how displays could be consolidated or collapsed to give clearer direction to users.
System designers should be encouraged to think more deeply about how "see" references (and headings generally) should index. Some useful filing conventions long established in card catalogs, e.g., filing parenthetical qualifiers together and ahead of other following data, filing dates such as "b. 1886" chronologically, etc., have often not been carried over into the online indexing of normalized headings and references.
Beyond this job of recovery, there are other possibilities for improvements in the utility of references in OPAC displays. For example, if online systems indexed only subfield a for the typical 400 name data, a more useful collocation of headings to choose from could be achieved, especially if 670 data were also made accessible to make the choosing easier. At the same time, the 400 data could be maintained in its current form on the authority record to catch erroneously formulated headings. Consider which is better for the user looking up "brown thomas":
Brown, Thomas Title Title Brown, Thomas, 1735-1779 SEARCH UNDER Brown, Thomas C. (Thomas Combs), 1735-1779 Brown, Thomas, 1825-1855 SEARCH UNDER Brown, Tom, 1825-1855 Brown, Thomas, 1837-1900 Title Title Title Brown, Thomas, 1899-1960 SEARCH UNDER Brown, Thomas M, 1899-1960 Brown, Thomas, 1933- SEARCH UNDER Brown, Thomas B., 1933- Brown, Thomas, 1940- Title Title Brown, Thomas, 1955- SEARCH UNDER Brown, Thomas W., 1955- Brown, Thomas A., 1837-1900 SEARCH UNDER Brown, Thomas, 1837-1900 Brown, Thomas A. (Thomas Alfred), 1960- SEARCH UNDER Brown, Thomas Alfred, 1960- Brown, Thomas Alfred, 1960- Title or: Brown, Thomas SEARCH ALSO UNDER Brown, Thomas B., 1933- Brown, Thomas C. (Thomas Combs), 1735-1779 Brown, Thomas M., 1899-1960 Brown, Thomas W., 1955- Brown, Tom, 1825-1855 Title Title Brown, Thomas, 1837-1900 Title Title Title Brown, Thomas, 1940- Title Title Brown, Thomas A. SEARCH UNDER Brown, Thomas, 1837-1900 Brown, Thomas Alfred, 1960- Brown, Thomas Alfred, 1960- TitleThe second index presents as much information as the first, with references just as precise, but in significantly less space and with a lot less visual clutter. In short, one could make a case for excluding qualifying subfields from references, and letting the referred-to headings disambiguate references as a general rule.
- Keyword indexes and phrase indexes
An area potentially ripe for substantive change and an area that the Task Group discussed in depth is the notion that keyword indexing obviates the need for certain kinds of cross references. Several members of the Group argued that some references particularly rotated corporate and conference headings were no longer necessary given the current environment in online catalog technology. Because all terms are indexed in a keyword index, such references could be seen as unnecessary, even obsolete.
More members of the Group agreed with the Joint Task Group that preceded us that there is probably not yet enough commonality in our current systems to do away with such references. While keyword title and subject indexes are fairly standard, many systems do not yet provide an "author keyword" index, nor do some libraries see an advantage in providing such an index. In addition, few online public access catalogs include 4xx or 5xx fields from authority records in their keyword indexes; their structure and information does not "fit" with the keyword concept. The syndetic structure that does not function at all in a keyword index is indispensable in an A-to-Z phrase index to fulfill the "collocating" function of the catalog. Therefore, the cross reference structure cannot be simplified because of keywording, and must remain intact to serve the phrase index. This Task Group has not recommended the simplification of any rules or rule interpretations solely because of the keyword function in online catalogs.
- Problems that improved cross references will not fix:
As described by Nirmala Bangalore, there are some problems that even improved cross reference structures will not fix:
- Searching personal names in direct order (i.e., forename followed by surname) in systems that do not have rotated heading indexes
- Typos or obvious mistakes
- Searching the wrong index
- Use of Boolean operators within an index not following the search syntax, or in a system where they're not allowed
Several themes emerged as the Task Group discussed various issues and concerns regarding references in the current online environment.
One major area that cut across several of the recommendations was the need to articulate the differences between shared authority data and data appropriate for local authority files. Members of the Group wondered when it would be appropriate to introduce the concept and wording of "shared authorities" into AACR2. We stopped short of recommending that be undertaken now, but perhaps an opportunity will arise when and if AACR2 is reorganized. There may be more flexibility (and indeed, it may be more appropriate) to gradually add this concept and language to the LCRIs.
Another major area that initiated several recommendations was the notion that some references in authority records may serve the purpose of automated heading checking and maintenance. The need for these references seems to be increasing as online library systems become larger and more sophisticated, and as we develop ever larger files of legacy data. The Task Group sees this need increasing in the future, as demands intensify for cross database links and displays.
Several changes were proposed out of concern for users in a more global environment, where they might be searching catalogs not in their native language. The Task Group thought that users unfamiliar with a particular language might search upon terms that presently are considered too "generic" to make as cross references. Although adding such references will result in longer records, we believe the result will be better user service and fewer exceptions for catalogers to have to remember.
The Group also made recommendations to include a few small things missing from AACR2, including the addition of a very few clarifying examples to promote global consistency.
Recommendation #1: Add a sentence to AACR2 rule 21.4D1 parallel to the instruction in AACR rule 21.4D2 regarding explanatory references from the personal name to the official name. The recommended additional text is shown below in double underscore.
21.4D1. Official communications Enter a work that falls into one of the following categories under the corporate heading for the official (see 24.20 and 24.27B): a) an official communication from a head of state, head of government, or head of an international body (e.g., a message to a legislature, a proclamation, an executive order other than one covered by 21.31) b) an official communication from a pope, patriarch, bishop, etc.(e.g., an order, decree, pastoral letter, bull, encyclical, constitution, or an official message to a council, synod, etc.). Make an added entry under the personal heading for the person. Make an explanatory reference from the personal heading to the corporate heading (see 26.3C1).Rationale: The Task Group noted that AACR2 rule 21.4D1 instructs catalogers to make added entries for personal names in various situations. The Group thought that it was important conceptually to communicate to users through syndetic structure the relationship between these two headings from both the personal name and from the official name, in addition to the actual usage of the headings on bibliographic records. It is perhaps even more important to provide the user with the official name as a reference from the personal name, because of the complex structure of that official form of name. The Group would like to see rule 21.4D1 made parallel to 21.4D2 so the "bi-directionality" of the explanatory references is made clearer. To provide appropriate information, consistency and simplicity, these references need to go both from official name to personal name, and from personal name to official name.
The Group also noted that AACR2 mandates an explanatory reference in 21.4D2, 24.20B2 and 26.3C1 from the corporate name to the personal name. In practice, and as interpreted by LCRI, catalogers construct simple see also references. The Task Group is comfortable leaving AACR2 as written to enable libraries which can create explanatory references to do so. A related proposal to change the corresponding LCRI is made in Recommendation #16.
Recommendation #2: Adapt LCRI text to add to AACR2 rule 26.2A2 to give more guidance and thus more consistency internationally. The recommended additional text is shown below in double underscore.
26.2A2. Different forms of the name. Refer from a form of name used by a person, or found in reference sources, or resulting from a different romanization of the name, if it differs significantly from the form used in the heading for that person. Always make a reference from each variant that affects the primary elements of the name. For the normal, inverted heading this means variations in all elements to the left of the comma and in the first element to the right of the comma Use judgement in making references from other variants(i.e., those that do not affect the primary entry elements) when it is judged the access to the catalog would be improved, e.g., when the heading is a common-sounding name. Typical instances are: [Additional examples of references for a common name] Williams, Bruce Robbins see Williams, Bruce But NOT: Westgate, Lewis G. (Lewis Gardner) see Westgate, Lewis GardnerRationale: The Group thought that LC's specification of what constitutes a significant variation should be proposed as a change to AACR. Adding such text and examples would lead to a higher level of consistency within the broader AACR constituency. In addition, the Group asserts that by the addition of the examples above, clearer direction is given for decision-making about when references should be made that do not affect the primary entry elements. Recommendation #7 is related to this action; it recommends striking the parallel text from the LCRI.
Recommendation #3: Delete AACR2 rule 26.2B (including all subrules: 26.2B1; 26.2B2; 26.2B3; 26.2B4).
26.2B. Name-title references 26.2B1. If the works of a person are entered under two or more different headings, make a name-title reference when the name appearing in a particular edition of a work is not the name used as the heading for that work. Ashe, Gordon Croaker see Creasey, John (Title page reads: The croaker / John Creasey as Gordon Ashe) Halliday, Michael Edge of terror see York, Jeremy (Title page reads: The edge of terror / by Michael Halliday. A later edition published under the name Jeremy York) 26.2B2. Make a name-title reference from the inverted form of initials entered in direct order for each work entered under those initials. [Examples omitted here to save space] 26.2B3. When two or more persons have used the same pseudonym and one or more is entered under another name, make a name-title reference from the pseudonym for each work of a person that is so entered. [Examples omitted here to save space] 26.2B4. If a pseudonym consists of initials, a sequence of letters, or numerals, make a name-title reference from the real name for each item entered under the pseudonym. [Examples omitted here to save space] In addition, if the initials, etc., stand for a phrase other than a name, make a name-title reference from the phrase in direct order for each item entered under the pseudonym. [Examples omitted here to save space]Rationale: This section on Name-title references was a great mystery to the Task Group members. None of us knew of any instances of these rules being applied in the manner described. We were confused about its existence in the "Personal names" section of this chapter, and why it existed independently of 26.4 (uniform titles). This section consistently has examples showing a name-title reference pointing to a name heading (not a name-title heading). In current LC and NACO practice, name-title references refer to name-title headings; there is parallel construction. The Task Group agrees that this section of AACR is misleading, and recommends that the language be deleted.
26.2B1 is already covered by the first examples in 26.2A1;
26.2B2 is already covered by examples in 26.2A3;
26.2B3 is already covered by the first examples in 26.2A1;
26.2B4 is already covered by the second example in 26.2A2.
We note that there is good information contained in the LCRIs for 26.2B2, 26.2B3, and 26.2B4, but that the interpretation given there is for name headings and references, not name-titles. Is this perhaps what the AACR2 authors intended? See Recommendation #12 for our ideas on moving the RI text should this rule deletion proposal go forward.
Recommendation #4: Add LCRI text to AACR2 rule 26.3A6 regarding written-out abbreviations, including special symbols. Recommended deletions are shown below as strike-through text; additional text is shown in double underscore.
26.3A6. Abbreviations, including symbols If, in the catalogue, abbreviated words are filed differently from words written in full and if the heading begins with an abbreviated word or contains an abbreviated word in such a position that it affects the filing of the heading, refer from the form of the heading with the abbreviated word written in full in the language of the heading. If the heading contains an ampersand or other symbol representing the word "and" (e.g., the plus sign (+)), make a reference from the name using the word "and" or its equivalent in the language of the heading.Rationale: Content from the LCRI regarding the ampersand (or other symbol) logically belongs in the rules. This topic is not addressed in AACR2 at present. Removing the concerns about where in the heading the ampersand/abbreviation falls in the heading simplifies heading construction for catalogers. We doubt there would be many problems created as a result of liberalizing this rule.
Recommendation #5: Restrict the circumstances in AACR2 rule 26.4B4 for making references from titles proper to collective titles. Suggested wording and examples are given below in double underscore.
26.4B4. Collective titles. When a collection of, or a selection from, a person's works is catalogued under a collective uniform title, refer from the name and title taken from the chief source of information or found in a reference source to the name and collective title unless the title taken from the chief source of information or found in a reference source is the same as, or very similar to, the collective title when that name-title might reasonably be sought by the catalogue user. [Additional examples;] Bach, Johann Sebastian Musique de chambre see Bach, Johann Sebastian Instrumental music. Selections BUT NOT: Bach, Johann Sebastian Kremer plays Bach see Bach, Johann Sebastian Instrumental music. SelectionsRationale: The music community has described ongoing difficulties with this rule, and the Task Group thought that more latitude and some additional examples would help resolve problems for that community. We recommend that this particular provision be discussed with the Music Library Association before final action is taken. This problem could also be examined in relation to Recommendation C from the section "Ideas for other groups to pursue:" form a group to explore further the issues surrounding references from translated titles.
References from specific, published titles to a collective uniform title are problematic. The contents of the published entities that fall under collective uniform titles often differ, making them much less interchangeable than editions or versions of a single work. Each catalog will have its own mix of specific titles under a collective uniform title, making it likely that references based on one collection's holdings will be misleading for users of another collection's holdings. Given that specific titles held can be searched separately in each system, it would be better to drop such references altogether from shared authorities, but the Task Group has stopped short of making such a recommendation. At this time, we think that giving catalogers more latitude in determining whether a specific title proper should be made as a shared reference is an appropriate first step. Rather than try to incorporate the concept of "shared authorities" into AACR, we suggest the language above, so that catalogers will feel less compelled than they do presently to add specific titles proper to name authority records.
Another option, not favored within the Task Group, was that additional qualifiers could be added to uniform titles headings under LCRI 25.8-25.11 to break these large authority records into more precise separate records. The point of the concerns is not that uniform titles should be distinguishing more finely between the editions they represent. The utility of these titles lies more in collocating entries under the generalized (collective) heading. The concern of the Task Group is that inappropriate, edition-specific references are automatically integrated into local catalogs where those editions are not held, resulting in unnecessary clutter for users. The root problem here is the need to articulate rules that differentiate appropriate references for shared authority records and appropriate references for local needs.
- Rule interpretations
Recommendation #6: Form a new group to re-evaluate and re- write the LCRI for Chapter 26, particularly the section on Linking References, to expand the definition of linking references and the circumstances under which linking references can be made.
26 References Linking References References need not be made from the form used in pre-1981 cataloging to the form used under AACR 2. Such references may be made, however, if judged useful by the cataloger or for specified projects. Although no longer routinely made, linking references are retained in existing authority recordsRationale: While much of the text in this section has historical value and should be preserved (perhaps in a separate document), it needs to be updated to the current cataloging context.
The Task Group thinks that it is time to take a fresh look at linking references, specifically at their utility in automated authority processing.
A re-examination of linking references, both for former headings and for variant headings in large files, raises the issue of the purposes for which references are created. The Task Group recognized several inherent purposes. References provide redirection and information for users. References link variant forms of the name of an individual or a body for identification purposes. References are also made to facilitate automated heading checking and cleanup. It is this last function that the Task Group is focusing on in this recommendation.
Libraries continue to have to deal with batch loaded records, external retrospective conversion services, old cataloging copy, and in the not-too distant future linking to and displaying records from external databases which may use different forms of headings. Increasingly, libraries will also need automated assistance for handling massive changes to legacy data. While the Task Group acknowledges the problems with "loading up" authority records with publicly-suppressed references that contain minor or perhaps even trivial variations, the authority file structure is the only extant mechanism we have to deal with these problems.
The Task Group agrees that adding such references runs counter to calls for simplicity--authority records would become both more complex and more dense. Such records, however, would facilitate more machine manipulation and less human intervention in record correction and maintenance. The Group believes that, on analysis, the benefits outweigh the cost.
A specific technique for adding these references needs to be explored further. Some members of the Task Group thought that with some tweaking the existing conventions of $w/2 might be suitable for these purposes. Other members of the Task Group argued for a new specifically-designed field that could also, like $w/2 of the old linking references, separate the function of file maintenance from that of user access and display. The potential use of the 7xx fields, and their relationship to maintenance needs, also should be examined.
Another valid question that needs to be explored further is how the addition of maintenance-type linking references should be implemented in a shared authorities environment. What references are appropriate for the shared file, and what should be created and maintained locally? How can we be assured that libraries will exercise this discretion in adding references to national-level authority records?
Because of the complexity of these questions and the potential far- reaching impact, the Task Group recommends that another group be charged to examine this issue. We recognize the tension between LC's creation of the Rule Interpretations as a document governing their internal cataloging policy and its use as a standard that is followed by other libraries. We therefore suggest that a joint group with LC CPSO representation as well as from other libraries be formed with the assignment of crafting language to reflect the needs of all stakeholders.
Recommendation #7: Delete part of LCRI 26.2 in conjunction with moving that text into AACR2 Recommendation #2).
26.2. NAMES OF PERSONS Referring from Variant Forms 2) Trace a reference from each variant that affects the primary elements of the name. For the normal, inverted heading this means variations in all elements to the left of the comma and in the first element to the right of the comma. Do not trace a reference that would normalize to the same form as the heading on the same record or to the same form as the heading on another record. 3) Refer from other variants (i.e., those that do not affect the primary entry elements) when it is judged the access to the catalog would be improved, e.g., when the heading is a common-sounding name. 100 Freeman, Robert, 1948- 670 His Hidden treasure, 1980: t.p. (Robert Freeman) 670 Phone call to author, 3/31/82 (Robert Eliot Freeman; b. 6/18/48) (No references) 100 Jenkins, Barbara 670 Jenkins, P. The walk west, 1981: CIP t.p. (Barbara Jenkins) CIP data sheet (Barbara Jo Pennell Jenkins) (No references)Rationale: Recommendation #2 transfers this wording to AACR2. It should be stricken from this LCRI, and the subsequent numbers within the RI should be renumbered. The first two examples in the RI should also be deleted because they contradict the proposed rule (common-sounding names).
Recommendation #8: Remove the exception in LCRI 26.2, Referring from Variant Forms, point (4) for German, Dutch, and related languages. The recommended deletion is shown below as strike-through text.
26.2 Names of Persons Referring from Variant Forms [text omitted] 4) Make one reference from each possible entry element of the name chosen as the heading, including each separate particle or prefix but excluding connectives such as "y" and "und." (In this connection, ignore the "Bure" example on p. 545 of AACR 2; this is based on actual usage, not an arbitrary permutation of the heading.) Exception: In German, Dutch, and related languages, do not refer from the definite article following a prefix.Rationale: This situation is encountered infrequently by generalist catalogers, making the exception difficult to remember. More persuasively, however, the group felt that users do not understand the conventions for what gets indexed for specific languages, and might search under these forms. This is particularly true in the more global environment, where users might be searching catalogs not in their native language.
A question was raised: if"Mac" or "Mc" appeared as a separate word in a heading, would a reference beginning with the following word be created? If the goal indeed is to make authority records work better in a global environment, where users might be searching catalogs not in their native language, then the Task Group thinks that such a reference would in fact be made. This concept may argue for further revision of the above RI (to eliminate the exception for "y" and "und," but the Task Group stopped short of going quite that far.
Recommendation #9: Remove the exception in LCRI 26.2, special considerations, point (3) for Hebrew names. The recommended deletion is shown below as strike-through text.
26.2 Names of Persons. Special Considerations 3) When a Hebrew surname begins with 'Ben-,' do not make a reference from the surname element that follows.Rationale: This situation is encountered infrequently, and as such creates a problem for general catalogers. More persuasively, however, the group felt that users do not understand the conventions for what elements are indexed for specific languages, and might search under these forms. This is particularly true in the more global environment, where users might be searching catalogs not in their native language. One member of the Task Group pointed out that similar linguistic inconsistencies are found elsewhere, and cited the example of the form "Neto" in Portuguese names, which is conceptually akin to referring from "Jr." in an English name. A Judaica cataloger in touch with a member of the group also pointed out that the rules make no parallel exceptions for names containing "Bar" (son of) or "Bat" (daughter of). Removing this restriction would resolve these linguistic inconsistencies. The Task Group recommends further consultation with Judaica catalogers (possibly via the NACO Hebraica funnel), recognizing that their special expertise will provide a very different picture than that of a generalist cataloger.
Recommendation #10: Add a new instruction to LCRI 26.2 to prefer forms corresponding to usage over forms corresponding to qualifiers in constructing references. New language is suggested below in double underscore.
Referring from Variant Forms 6) In constructing references, prefer forms corresponding to usage over forms corresponding to qualifiers. 100 1 Mills, R. A. $q (Rachel A.) 400 1 Mills, Rachel 670 [title]: $b t.p. (R.A. Mills) pref. (Rachel Mills) NOT 400 1 Mills, Rachel A. Note that if another Rachel Mills were already established in the authority file, the form of the cross reference in the above example would then become 400 1 Mills, Rachel $q (Rachel A.)Rationale: It was brought to the group's attention that this policy has not been made clear either at LC or at the British Library, and that these groups have informally agreed upon the above instruction in the course of harmonization talks. The Group suggests that this decision be recorded.
Recommendation #11: Add a new section to LCRI 26.2 regarding the handling of variant dates. New language is suggested below in double underscore.
4) When variant birth or death dates which are in widespread use are evident and have been used in a heading in another large file, construct a linking reference to include the variant date(s), coded not to display to the public. 100 0 John, of Rodington, ca. 1290-1348? 400 0 Joannes Rodingtonus, d. 1348 $w nnoa (NUC pre-56 form)Rationale: This topic was raised on PCC-L in July 1998, and it was evident from the discussion that there are conflicting concerns about handling this information. Catalogers want to have this information included in the record in a form that is useful for automated database maintenance, yet such references would probably be confusing to users in a public display. The current practice of recording such information in a 670 field is not satisfactory for automated maintenance.
The Task Group noted that catalogers should not make references from all the variant dates found in sources; for some historical figures this could easily get out of hand. Therefore we suggest wording for the LCRI ("dates which are in widespread use") to provide boundaries for this practice.
This issue is related to Recommendation #6. If this recommendation is not acceptable as LC practice, additional wording as "LC/NACO practice" should be considered here to allow NACO libraries the latitude to incorporate such references.
Recommendation #12: Move LCRI 26.2B2, 26.3B3, and 26.2B4 (pending the recommended deletion of AACR2 rule 26.2B) to LCRI 26.2. Some of the text needs to be revised slightly so that the mention of name-title references is removed. This recommendation is related to Recommendation #3, which would delete corresponding rule 26.2B.
26.2B2. If a personal name heading consists entirely of initials, make a reference from the inverted form alone(not a name-title reference) beginning with the last initial. If the heading contains the full form of the name for which the initials stand as a parenthetical addition (cf. 22.18A), add the parenthetical addition to the reference as well. (Do not invert the names used in the addition.) 100 X Y Z 400 Z, X Y 100 H. D. (Hilda Doolittle), 1886-1961 400 D., H. (Hilda Doolittle), 1886-1961 26.2B3. When two or more persons have used the same pseudonym and and one or more is entered under another name, make references to the names alone, not name-title references. from the pseudonym to the established names. 100 Enriquez, Colin MetcalfC. M. $q (Colin Metcalf),$d 1884- 400 Theophilus, $d 1884- 100 Dickinson, Jonathan, $d 1688-1747 400 Theophilus, $d 1688-1747 26.2B4. If a heading consists of initials, a sequence of letters, or numerals, make a reference from the person's real name alone, not a name-title reference. in inverted order. 100 H. D. (Hilda Doolittle), 1886-1961 400 Doolittle, Hilda, 1886-1961 If the initials used in the heading stand for a phrase other than a personal name, make a reference from the phrase alone(indirect order), not a name-title reference.Rationale: This is useful information that should be kept in the LCRIs. It could be incorporated to the text and examples in LCRI 26.2, either in the section entitled "Referring from Variant Forms" or "Special Considerations."
Recommendation #13: Revise part of LCRI 26.2C to clarify and broaden applicability of see also references from individuals to the group for NACO libraries.
26.2C. "See also" References See Also References from Individuals to the Group When the name of a group contains the name of one or more of its members, make a see also reference from the heading for each person to the heading for the group (but not from the group to the person). LC/NACO practice: NACO libraries may make such references for any group which contains the name of one or more of its members. The Library of Congress is limiting this practice to collections of special materials (in lieu of making multiple added entries on individual bibliographic records) and to music performing groups, when the information is readily available and cataloging resources permit. Note: Before 1995, references were also made from the group to the person. Do not delete such existing references.Rationale: Several members of the Task Group agreed strongly that this note regarding LC practice must be revised to clarify the difference between NACO practice and LC practice. We have heard a need in some cataloging communities that they would like to apply this RI in other circumstances (e.g., architectural firms) but will not do so now because the statement above, although clearly labeled as "LC practice," is interpreted to mean "LC/NACO practice."
One Task Group member thought that the parentheses in the last sentence of the LCRI above should also be deleted and replaced with the words "and vice versa" so that the see also references could be bi-directional. As another recommendation suggests that bi- directional references be made for heads of state, adopting this suggestion for bi-directional references would promote internal consistency. However, others on the Task Group disagreed, thinking that the workload of adding personal name "see alsos" to every group heading would be prohibitive and not sufficiently beneficial to offset the cost. Concerns about corporate body name changes as they relate to personal names also add complexity in decision-making, and add weight to the argument not to make these references bi- directional.
Recommendation #14: Rewrite LCRI 26.3A3, References from Inverted Names of Government Subheadings (including changing this caption), to allow greater cataloger discretion to construct alternate direct order forms and to cover a parallel, currently unaddressed situation. Suggested changes are shown below in double underscore, following the strike-through text that shows the current version.
26.3A3. Different forms of the name References from Inverted Names of Government Subheadings References from Generic Terms and Keywords in Names of Subordinate Bodies 1) Make an inverted a reference(under jurisdiction) from the generic term in the name that indicates the type of agency (e.g., dept., board, committee) when this term is preceded by a word or words that may not be recognized as part of the name. 110 Michigan. State Dept. of Education 410 Michigan. Dept. of Education, State 410 Michigan. Dept. of Education 2) Make an inverted a reference (under jurisdiction) from the first key word in the name and from the key word of an English form of name of an agency entered in a foreign language(regardless of the grammar of the language involved).followed by an indication of level in the hierarchy. If these two elements can be combined grammatically in the reference, construct a "made up" reference; if not (as may be the case in languages other than English), make the reference a simple inversion. 110 India. Ministry of Health 410 India. Health, Ministry of 410 India. Health Ministry 110 Spain. Subdirección General de Archivos 410 Spain. Archivos, Subdirección General de 110 Rockville (Md.). Dept. of Planning 410 Rockville (Md.). Planning, Dept. of 410 Rockville (Md.). Planning Dept. Make such inverted references only for headings entered directly under the jurisdiction. Do not make the inverted references if the government body is entered under its own name rather than as a subheading of the jurisdiction. 3) When the name of a subordinate body begins with a keyword, make a reference from the name of the subordinate body beginning with the term in the name that indicates the type of agency if such a form is readily apparent. If the form the reference should take is unclear, do not make the reference. 110 Chester County (Pa.). Health Dept. 410 Chester County (Pa.). Dept. of HealthRationale: Subbody references could be made clearer in public displays, and would be less artificial (and confusing) by allowing more cataloger judgement to drop and/or rearrange terms in this circumstance. (The Task Group recognizes that straightforward inversions are sometimes simpler for cataloger to formulate, however, our primary concern here is a sensible display for the user.) One Task Group member commented that Forms like "Dept. of Education" and "Education Dept." were such common grammatical variants in English as to be interchangeable. This applies whether these are government subheadings or departments of universities. If the established form of a heading contains such a phrase, shouldn't we automatically be able to make a reference from the other form without justification? There was also general agreement that government subheadings should not be treated any differently than subheadings for other corporate bodies.
There are concerns about this approach. Foremost among them are the grammatical problems caused for languages other than English. The Group has suggested language above to deal with that situation. Also of concern is that, under this technique, true variants will be obscured. There may also be more conflicts with earlier/later forms of name. The Group recognizes the validity of these concerns.
One Task Group member thought that this entire section of the LCRI should be deleted, and stated that this part of the RI has extrapolated AACR2 and obscured its original meaning, resulting in some odd-looking references.
The Task Group had quite a discussion about inverted references, both for this LCRI and for other parts of the rules (e.g., AACR2 26.3A3). For this RI, we elected to do away with the concept and terminology of "inversion" but in other parts of the rules the direction was less clear. In general, the Task Group could not come to consensus on the way such references should be formulated. There did seem to be general agreement that the goal of these references is variant entry under a different name element. Some thought inversions should be formulated in an alternate form rather than a simple inversion; others thought that the inversion should not be a "made up" form.
The Group also recognized that this RI also contains instructions for the inversion of conference name headings. We felt conferences were different, and that inversion in that circumstance was more "user friendly" and thus acceptable.
Recommendation #15: Delete LCRI 26.3A6 (pending the recommended transfer of wording to AACR2 rules 26.3A6).
26.3A6. Abbreviations. If the name used in the heading contains an ampersand or other symbol representing the word 'and' (e.g., the plus sign (+)), make a reference from the name using the word 'and' or its equivalent in the language of the heading whenever the symbol occurs within the first five words of the heading. Make a reference from the full form of other abbreviations only if the abbreviation 1) occurs within the first five words of the headings, 2) is not listed in Appendix B, and 3) does not represent a proper name.Rationale: This action is proposed in conjunction with Recommendation #4 above. Content from the LCRI regarding the ampersand (or other symbol) logically belongs in the rules. In addition, the Group advocates removing the restrictions upon the location within the heading/reference of the abbreviation/ampersand in the interest of simplicity. We doubt that there will be significant impact or workload in liberalizing this rule.
Recommendation #16: Clarify and expand LCRI 26.3B-C (See also references and Explanatory references) to guide catalogers to make see also references for heads of state (etc.) bi-directional (from official name to personal name, and from personal name to official name). This recommendation is related to Recommendation #1, which adds a relevant sentence to AACR2 rule 21.4D1.
26.3B-C. See also references and Explanatory references Explanatory References 2) Private communications of heads of state, etc. Connect headings for heads of state, ecclesiastical officials etc., to the corresponding official heading with see also references instead of the explanatory reference called for in 21.4D2 and illustrated in 26.3C1. Make see also references both on the personal heading authority record and on the official heading authority record. authority record 100 Mohammed Reza Pahlavi, Shah of Iran, 1919- 510 Iran. Shah (1941-1979 : Mohammed Reza Pahlavi) ($w, position 1 = n) AND authority record 110 Iran. Shah (1941-1979: Mohammed Reza Pahlavi) 500 Mohammed Reza Pahlavi, Shah of Iran, 1919- ($w, position 1 = n) resulting reference Iran. Shah (1941-1979 : Mohammed Reza Pahlavi) search also under Mohammed Reza Pahlavi, Shah of Iran, 1919- Use the following guidelines to determine when to make the references: a) If the official heading is one established without names or dates (24.20C-E), do not trace the reference. b) If the official heading has been established, make the see also reference. c) If the official heading has not been established,establish it and make the see also reference from it to thepersonal heading only if the item being cataloged is not clearly the work of private authorship.Rationale: The Group noted that AACR2 mandates an explanatory reference in 21.4D2, 24.20B2 and 26.3C1 from the personal name to the official name. In practice, and as interpreted by LCRI, catalogers construct simple see also references. The Task Group is comfortable leaving AACR2 as written to enable libraries which can create explanatory references to do so. However, the Group would like to see the LCRI rewritten so the "bi-directionality" of the see also references is made clearer. For the sake of consistency and simplicity, these see also references need to go both from official name to personal name, and from personal name to official name. We understand that at one time, NACO libraries made these references in both directions; before and after this time the reference were made only from the corporate to the personal heading. Records in the NAF today reflect this vacillation, which is confusing both to catalogers and to end users.
The Task Group also noted that AACR2 rule 21.4D1 affects this situation, as it instructs catalogers to make added entries for personal names in various situations. A change to that rule is recommended in Recommendation #1. The Group thought that, regardless of the bibliographic records that utilize either the personal name or the official name headings, it is important conceptually to communicate to users the relationship between these two headings from both the personal name and from the official name. It is perhaps even more important to provide the user with the official name as a see also reference from the personal name, because of the complex structure of that official form of name.
The group also notes that some of the language in the earlier part of this same RI may no longer be necessary, and some streamlining is in order. For example, the "resulting references" examples, which follow tagged examples, are redundant and are probably no longer necessary.
Recommendation #17: Rescind the practice described in LCRI 26.5 of adding date qualifiers to all references to uniform titles headings of the form "[Name]. Works ", "[Name]. Selections ", and "Bible"
26.5A. Series Types of See References 1) Alternative forms not selected as series heading c) Heading is name/uniform title a. Give a reference from title proper. 100 1# $a James, Henry, $d 1811-1882. $t Selections. $f 1983 430 #0 $a Selected works of Henry James, Sr. $f 1983 100 1# $a Sterne, Laurence, $d 1713-1768. $t Works. $f 1978 430 #0 $a Florida edition of the works of Laurence Sterne. $f 1978 b. Give a reference from name/title proper when the title proper differs from the uniform title except when the difference is the addition of a language name, a date, or a parenthetical qualifier. 100 1# $a James, Henry, $d 1811-1882. $t Selections. $f 1983 430 #0 $a Selected works of Henry James, Sr. $f 1983 400 1# $a James, Henry, $d 1811-1882. $t Selected works of Henry James, Sr. $f 1983 100 1# $a Sterne, Laurence, $d 1713-1768. $t Works. $f 1978 430 #0 $a Florida edition of the works of Laurence Sterne. $f 1978 400 1# $a Sterne, Laurence, $d 1713-1768. $t Florida edition of the works of Laurence Sterne. $f 1978Rationale: Current LC practice when constructing references to the collective uniform titles "Works" and "Selections" and to Bible headings is to include the subfield f date required in the heading in each of the references. The rule apparently being followed is LCRI 26.1, "construct a reference in the same form in which it would be constructed if chosen as the heading." A specific instruction regarding the addition of dates in these references has not been found; however, examples of the practice are included in LCRI 26.5, and in many examples from the LCNAF (e.g., n81135121; n87883625; n95075432; no97045342).
The interpretation of the general rule is open to question. A specific edition title is already different in form from one of the three uniform titles notes above, and if a specific edition title were established, the rules would not require that the subfield f date be added to it. Further, the rules already permit the omission of other kinds of heading qualifiers from references (e.g., the omission of (Firm), and of the conference name qualifiers in the case of rotated conference name references). The presence of the date in the heading is by itself sufficient to differentiate these references in standard index displays if the dates in the references were dropped:
[Name.] Selected poems Search under [Name.] Selections. 1954 [Name.] Selections. 1973
- Presentation to and use by catalogers
Discussions are underway about the potential reorganization of Part I of AACR2 into an ISBD-based structure rather than the current format-based structure. Little has been said about the possibility of reorganizing Part II. The Task Group agreed that, for the moment, we should keep the present structure of rules and rule interpretations and not recommend a major Part II reorganization. It is premature to take any action at this point. We (the PCC) should monitor the developments in this area. Should the opportunity for reorganization arise, another group should be formed to examine this issue more closely. We should be prepared to take advantage of the opportunity large-scale reorganization would present.
The decision to reorganize this part of the rules should not be made lightly. On the one hand, it would be very logical to have all instructions concerning references located in one place, rather than scattered as they are now throughout chapters 22-26. On the other hand, though, as catalogers establish headings they might want to determine potential references at that time; interpolated instructions about references could give them the appropriate information at the appropriate time. Further study to identify the most useful approach needs to be undertaken.
In addition, major reorganization of Part II could offer the opportunity to incorporate concepts and language into the rules about "shared authorities." Perhaps the time has come for this notion, mentioned elsewhere in this report, (as well as shared cataloging) to be formally recognized.
At the present time, the effort required to independently streamline and reorganize the rules and corresponding rule interpretations for access points would not be cost effective. This could be offset somewhat (efficiencies of scale) if there was a larger movement promoting broader reorganization. The Task Group was not troubled about leaving the rules and rule interpretations as they are until a larger initiative gets underway. While the duplication of content and the elements of unpredictability as to whether instructions for references are in chapters 22-25 or chapter 26, are bothersome, the Group does not see this as an urgent problem that needs to be addressed now. The Group also agreed that no new "cataloging aid" (in other words, a tool separate from AACR2 or the LCRIs) should be developed for the presentation of authority-related information.
- Explicitly allow the creation of a reference
for the "citation form" of a personal name based on cataloger's judgement
without ascertaining usage or providing justification from print sources.
The Task Group would be interested in promoting a broader discussion of the idea of optionally creating a reference (without literary warrant or corresponding justification) for the "citation form" of a personal name (e.g., surname, initials). Typically on the verso of a scientific publication one will find the phrase "Cite as"; this almost always gives a citation using only the author's initials. In other disciplines, scholars are cited by initials only in footnotes (L.W. Smith). One could argue that references from such forename initials should be made as a matter of course, without citing usage.
One member of the Task Group thought that these should be optional rather than automatic, and expected that they would be more useful in headings for academic authors than for other personal name headings. Another Task Group member suggested that for "common sounding names" (or for files in which the initial and the name would likely file nearby each other) this technique would be useful (L.W. Smith), but that it does not seem necessary for names like J. Schuitema or R. Penkiunas. One option is to remove the restriction that, in general, forms that are shorter than the chief source form of a personal name (e.g., Larry W. Smith on t.p., vs. Smith, L.W. in the bibliography) are not cited or referred from. Rather than craft specific language at this point, the Group thought broader discussion was warranted to ascertain interest, impetus, and consensus.
- Implement "spell-checker" functionality in
The Task Group felt that the library community needs to mobilize to systematically lobby OPAC vendors for basic spell-checker functionality within name (and perhaps title?) indexes. There are kinds of variants (misspellings, etc.) which many users would benefit from if their failed searches could be matched against a file of such variants. Librarians and system designers should consider ways of standardizing, building, sharing, and maintaining such files, separate from the formal authority record, and incorporating them into machine-mediated "search helper" functions, separate from the formal OPAC index display. Examples might include such entries as:
Smith see also Smyth Alan see also Allen McLean see also MacLean Department see also Dept.This is not data that we think should be coded into individual authority records (unless, of course, usage warrants), but on a more general level could be quite helpful if built into our online systems.
- Form a group to explore further the issues surrounding
references from translated titles: the monographs approach and the
Uniform titles (and name-titles) are complex and deserve a separate task force of their own. The current reference practice is problematic, and the solution is not obvious. Specifically, the Task Group discussed the issue of name/title references for monographs being created on different "parent" records than they are for music materials; each community apparently interprets rule 26.4B1 differently. (There is no LCRI for this rule, but there is a MCD.) This is a fundamental discrepancy in both principle and practice. Examples of the problem may give a clearer description than written text:
The current rules for monographs dictate:
100 10 Dostoyevsky, Fyodor, $d 1821-1881. $t Prestuplenie i nakazanie. $l English 400 10 Dostoyevsky, Fyodor, $d 1821-1881. $t Crime and punishmentThe current rules for music dictate:
100 10 Mozart, Wolfgang Amadeus, $d 1756-1791. $t Nozze de Figaro 400 10 Mozart, Wolfgang Amadeus, $d 1756-1791. $t Noces de Figaro 400 10 Mozart, Wolfgang Amadeus, $d 1756-1791. $t Hochzeit des Figaro 400 10 Mozart, Wolfgang Amadeus, $d 1756-1791. $t Marriage of FigaroThe monographs pattern is particularly problematic in subject indexes, where language-qualified headings are rarely used, and accordingly references to foreign language titles from familiar English titles are disallowed. Also, embedded in this problem are the issues of "work" versus "edition" or "printing," which the Task Group thought were out of its jurisdiction. Nevertheless, the Group did want to draw attention to the problem and to the need for more in-depth analysis and problem solving. It is our hope that another group will be assigned to look more deeply into the issues surrounding uniform title and name-title references, in the name, title, and subject indexes.
The Group noted that the JSC has charged LC to write a discussion paper on uniform title added entries for translations for continuing resources, but we see a broader need than addressing only continuing resources. The two groups might want to coordinate their efforts.
- Monitor further developments to allow the creation
of references in non-Roman scripts.
The Task Group is aware of the work of James Agenbroad and the new rules (26.1B2 and 26.1C2) regarding non-Roman scripts which are being proposed to CC:DA. The Group did not have complete consensus, but decided not to formulate a specific recommendation in this report regarding references in non-Roman scripts, since this topic is already being addressed by another group. The prevailing thought of the Task Group at this time is that these principles are already implicit in the rules as they currently exist. Nothing in the rules specifically disallows references in non-Roman scripts, although one member pointed out that AACR2 assumes a "romanized" environment and doesn't specifically mention vernacular forms. We are also aware of the work of the MARBI task force on multilingual characteristics of authority records, and think that their input will help resolve some of the technical difficulties in this area.
- Using the fullest form of name as the heading
One thought that occurred to a member of the Task Group was that changing the rules so that the fullest possible form was established as the form of the heading in an environment of implicit right-hand truncation would in fact obviate the need for many cross references. Because there is no standardization between online systems, the benefits derived from automatic right-hand truncation (which is certainly common but not universal) cannot at this time be translated into simplification of cross reference structures. This very radical idea, which would require rethinking the basic principles of AACR as well as the recoding of massive amounts of legacy data in both bibliographic and authority records, was of course instantly rejected.
- Literary warrant
Arlene Taylor recommends that we rethink our assumption that cross references should only be made from forms of name that have appeared in print. Even though the definition of publishing (with the evolution of the World Wide Web) seems to be broadening, perhaps it is time to make a radical change and open cross references up to cataloger judgement. While we may want to impose some restrictions to avoid keying references for obvious misspellings, guidelines with examples could perhaps be developed to support a more liberal use of cross references. The Task Group was not ready to take on this challenge at this time.
One Task Group member strongly supports the principle of describing the source of the reference in a 670 source citation, even if only to note that it is a form known by the cataloger to be frequently sought by users. References on older authority records, from a time when this principle was not in effect, present problems for catalogers.
- References with "tiny variations"
26.1H. "Do not make a reference if the reference is so similar to a heading (name and/or title) or to another reference as to be unnecessary."The Task Group talked about this rule in the online environment, wondering if it needed to be changed to accommodate the literalness of computers in searching for user terminology. The consensus was that we don't want to make catalogers crazy by forcing them to provide a reference for every wayward search term. We think there is both enough guidance and enough flexibility as written for catalogers to be able to adequately interpret what is appropriate for a given situation. We therefore opted to leave the wording of this rule alone. This rule was also the impetus for our proposal regarding "spell checker" functionality in the search interface of online systems under Section 5 above.
- The "Inheritance" Problem
The Task Group noted another difficult problem, which we described as the "inheritance problem." An example is not being able to retrieve "American Library Association. Council" when searching the form "ALA. Council"; the acronym is not necessarily carried through in the authority record for each subbody. The Group elected not to recommend a change in practice allowing such constructions to be routinely added as cross references in records for subordinate bodies.
- Language-specific circumstances
In general, the Task Group would like to lessen language-specific circumstances in the rules. There are fewer decision points for catalogers and better access for users if references are made from all personal name elements. If one purpose of references is to generalize access vocabulary for searchers with different linguistic backgrounds, then there's no need to specify them for each language. We intended to recommend to simplify the decision points in AACR2 rule 22.5C1 for references for complex personal names in various languages, but in the final analysis decided not to propose the following change in the rules. We thought that, as stated, the rule was clear enough without the additional wording. The recommended additional text is shown below in double underscore.
22.5C1. Preliminary rule The following rules deal with the entry of surnames consisting of two or more proper names (referred to as "compound surnames") and names that may or may not contain compound surnames. Apply the rules in the order given. Refer from elements of compound surnames not chosen as the entry element regardless of language.
- Inverted headings
The Task Group could not come to consensus on the way inverted references should be formulated. Some thought inversions should be formulated in an alternate form rather than a simple inversion; some thought that the inversion should not be a "made up" form. There did seem to be general agreement that the goal of these references is variant entry under a different name element. There was also general agreement that government subheadings should not be treated any differently than subheadings for other corporate bodies. We discussed the possibility of deleting several examples in AACR2 rule 26.3A3 but could not reach consensus in doing so (The Jackson (D.G.) Advertising Service; Jackson Advertising Service; College of William and Mary). The issue for us for government bodies was clearer, and resulted in Recommendation #14. So in a sense this is not a "rejected idea," but rather an idea with which we had a hard time reaching complete agreement.
Bangalore, Nirmala S. "Authority files in online catalogs revisited." Cataloging & Classification Quarterly , vol. 20, no. 3 (1995) pp. 75-94.
Jamieson, Alexis J., Elizabeth Dolan, and Luc Declerck. "Keyword searching vs. authority control in an online catalog." Journal of Academic Librarianship , vol. 12 (November 1986) pp. 277-283.
Library of Congress. NACO Participants' Manual. 2nd ed. Washington, D.C.: Library of Congress, Cataloging Distribution Service, 1996.
Taylor, Arlene G. "Authority files in online catalogs: an investigation of their value." Cataloging & Classification Quarterly, vol. 4, no. 3 (Spring 1984) pp. 1-17.
Thomas, Catherine M. "Authority control in manual versus online catalogs: an examination of 'see' references." Information Technology and Libraries, vol. 3, no. 4 (December 1984) pp. 393-398.
Watson, Mark R. and Arlene G. Taylor. "Implications of current reference structures for authority work in online environments." Information Technology and Libraries, vol. 6, no. 1 (March 1987) pp. 10-19.
Wilson, Mary Dabney. "Key issues and priorities for authority work in the online environment." Convergence: Proceedings of the Second National Conference of the Library and Information Technology Association, October 2-6, 1988, Boston. Chicago: American Library Association, 1990, pp. 35-47.
Yee, Martha M. and Sara Shatford Layne. Improving online public access catalogs. Chicago: American Library Association, 1998.
Younger, Jennifer A. "After Cutter: Authority control in the twenty-first century." Library Resources and Technical Services, vol. 39, no. 2 (April 1995) pp. 133-141.