Microsoft Word - 2012 10 19B submitted to MARS.doc 1 Collaborative Initiatives in Error Handling and Bibliographic Maintenance: Use of Electronic Distribution Lists and Related Resources IAN FAIRCLOUGH Fenwick Library, George Mason University, Fairfax, Virginia, USA This is an electronic version of an article published in Cataloging & Classification Quarterly v. 51 (2013) no. 1-3. Cataloging & Classification Quarterly is available online at http://www.tandfonline.com/doi/abs/10.1080/01639374.2012.719074. ABSTRACT Over the past decade, people working collaboratively have created several electronic distribution lists, each dedicated to notification about a specific issue in error handling for bibliographic and authority records, and other aspects of catalog maintenance. Librarians and others concerned for the accuracy of classification numbers, established headings, and series data can communicate among each other via these lists and related projects. This article documents their history and role in cataloging operations. Subscription information and frequently-used abbreviations are provided in Appendixes. This study was supported by an award of five days of research leave from George Mason University Libraries. The author gratefully acknowledges the kindness of John Zenelis, University Librarian, and the Libraries’ Professional Development Committee (Laura Jenemann, Chair) in granting this award. 2 INTRODUCTION A keyword search in an online catalog fails because of a typographical error in a bibliographic record. A library patron browsing the collection misses a book on the topic of interest because it is misshelved, not by an inattentive page, but because of a mistranscribed digit in the call number. A student who wants a list of all books by a particular author gets titles not only by that person but also others by a namesake with whom the sought person has become confused. A researcher attempting to identify titles in a scholarly series fails to find all but a few, on account of inconsistent application of headings among the various agencies from which the records in a local catalog were imported without examination. These situations have in common the theme that an error has happened in the cataloging and classification process. Often, typographical errors are considered "minor", and perhaps this claim is valid when evaluating the skills of someone being trained in applying cataloging rules to library materials. But in the context of an online catalog, what one person might consider to be a "minor" typo can result in a user's failed search: hardly a minor matter. Contemporary tools such as Web search engines, when faced with nonstandard data, display a "Did you mean?" message with one or more plausible alternatives. Automated spellchecking routines can "correct" data with or without human supervision of each instance. But this practice sometimes results in inappropriate changes, in cases where the supposedly incorrect data is in fact what was intended. Tools such as these cannot perform the task of repairing the damage done by inadequate representation of resources. Human intervention is required. This article describes some of the tools available to librarians for error handling so that they can attend to situations in their local databases and concurrently assist others to do likewise. PREVIOUS STUDIES 3 Literature related to this article falls into two main categories: Work on typographical errors in bibliographic records; and the use of electronic distribution lists as a means of communication among librarians. Terry Ballard documents the origins of numerous interrelated collaborative projects in Jeffrey Beall's "Dirty Database" test1. A seemingly trivial exercise, Beall's test was the spark that set in motion these projects, which are still in full operation two decades later. Published in American Libraries in 19912, the test invites people to examine their local database for accuracy. In private correspondence, Beall has stressed that this "test" is not scientific in nature, a statement that is in no way characteristic of his scholarly output. Ballard performed Beall's test on the bibliographic database at Adelphi University. Although Adelphi performed well, after further investigations Ballard concluded that the situation with respect to library catalogs universally was in need of clean-up. He used the newly- available keyword searching capabilities of Adelphi’s online catalog to compile a list of all misspelled words found therein, a project that gave rise to the database Typographical Errors in Library Databases. 3 This database, which began with a few hundred entries, has now grown to several thousand, and is colloquially known as the Ballard List, a phrase which belies the essentially collaborative nature of the project, with numerous people contributing errors they have found, and assisting with maintenance of the database. In a statistically-based test, Beall and co-author Karen Kafadar examined the extent to which libraries deriving their cataloging from master records in OCLC's WorldCat database had corrected any errors in those records when adding them to their local database.4 Beall and Kafadar's research drew its examples from the Ballard List in a randomized process, thereby reciprocating Ballard's initial use of the "Dirty Database" test. 4 The relevance of Beall's article lies in his findings: Thirty to forty percent of typographical errors in records used for copy cataloging go uncorrected. Beall urges OCLC and other agencies to "redouble their commitment to eliminating typographical errors and develop more sophisticated algorithms to detect and eliminate the errors"5. Whether or not Beall's exhortation was heeded, much work is still needed to reduce, if it is not feasible to entirely eliminate, typos and other errors in bibliographic records. Asking suppliers of records to perform this task, rather than distributing records containing errors that have to be cleaned up by their customers, reduces the workload of all recipients, thereby saving them valuable time that can be spent on other activity. But if those agencies do not do so, the burden falls on individual librarians working with their local collections. Such librarians often work collaboratively, as described below, but not all database recipients may receive corrections. The articles by Ballard, and  by  Beall  and  Kafadar  recount the situation for which electronic mail became the platform for several projects dedicated to the elimination of errors in bibliographic records. A related concern also covered in these projects was improving the quality of bibliographic data that was not technically erroneous at the time of cataloging, but that requires local attention on the part of librarians because of obsolescence or another situation. Bernie Sloan comprehensively discussed the use of e-mail lists in librarianship.6 One concern is terminology: in common parlance such lists are often called "listservs." Sloan cautions: "listserv" should not be used generically to describe electronic discussion lists, as the term is a registered trademark licensed to L-Soft International, Inc."7. He recommends simply using "list". Two further observations to complement Sloan's remarks: 5 (1) Using the term "listserv" to refer to all lists regardless can lead to the mistaken impression that all such lists operate similarly, whereas rival software does not necessarily have the same features as Listserv®. (2) Distribution rather than discussion is a more comprehensive term to interpose, thus: "electronic distribution list". This comment is borne out as Sloan identifies two types of lists: An announcement list serves a function much like that of a newsletter. The communication is one-way. Subscribers cannot post to the list. The list owner or moderator generally is the only person authorized to post items to the list. Subscribers play a passive role, simply receiving information. A discussion list is an interactive forum for communication. Subscribers may post an e- mail message to the list, with all subscribers receiving a copy of the message. Other subscribers may choose to respond to the initial message, with their e-mail reply also being distributed to all list subscribers. To these types, I have added a third: a notification list. This type resembles an announcement list, but with some differences. Rather than just one authorized person, other list subscribers may post. Discussion is not encouraged, but where appropriate it is permitted: for example, when a contributor requests assistance with resolving a problem. Often an initial question results in off-list correspondence, with the resolution of the matter in a subsequent post to the list. This releases other list readers from having to follow intervening stages of the discussion, allowing them to focus instead on the resolution and its application to their own work locally. With discussion lists, on the other hand, it is preferred that all correspondence be made on the list for all to see. 6 These characterizations are broad generalizations, and do not apply in every instance. Whatever the type, it is important that contributions be made to a list on a fairly regular basis, lest it should become inactive. Another distinction Sloan makes is between "public" and "private" lists, public lists being open to all, while private lists have restrictions on membership. To which I add: with Listserv®, "private" has a specific technical meaning for list operation. "Private" means that only subscribers can send a message to the list, but it does not necessarily restrict who may subscribe. Lists are further distinguished as "moderated" (someone serves as "editor"--again, a Listserv® technical term--to review all messages prior to distribution) and "unmoderated": messages are forwarded without review. To complement Sloan's comments on list moderation: a Listserv® list can be set so that all persons writing to the list must themselves confirm that they intend to do so, a technique to cut down on unwanted messages ("spam"). This initial confirmation can be followed by moderator approval, for double confirmation that the message is pertinent to the list's purpose. Archives are often a feature of list structure. Sloan says, "A good archive will allow you to search for past messages on a given topic, written by a specific person, etc."8 and that unarchived lists are rare. Listserv® as well as rival products mostly allow all messages posted to a list to be stored for future reference. Further comments on Sloan's remarks: archiving does not guarantee a literal rendition of all that was said on a list, for under certain circumstances the archives can be altered. It is also technically possible for a message to be present in the archives without actually having been distributed to subscribers (and with Listserv®, some lists are specifically set up for this purpose). Another means of accessing Listserv® archives is via third- party software such as gmane.com. 7 Sloan's article continues with observations about list etiquette; passive members ("lurkers") versus the active ones who make the list work and provide its personality; a statistical study of list membership and participation (one half of one percent of the number of posters accounted for 22 percent of the total messages sent); the fact that some participants were professionally active before lists existed while others have never known life without them; and the time when commercial activity on the Internet was proscribed (hardly imaginable today, but with a lasting negativity toward vendor participation on some lists). In concluding remarks: Sloan says, "Posting to a list may bring you fame, or it may bring you notoriety. Active participation in a library list allows you to make an impression like nothing else can. Make sure it’s a good impression." Sloan's article is recommended reading for all wanting a general background in list usage and participation. The lists described in this article all have the characteristic of being intended for notification9: an aspect rarely found among other lists, and which distinguishes them as a group. All of them are collaborative endeavors. From the beginning, they were set up as cooperative projects, with several people participating. In their ongoing status, users have both active and passive roles with respect to list participation, as Sloan describes, but only with respect to whether or not they actually contribute to the list. Participation in the projects that the lists support happens as a person reads a list message and investigates the local situation, taking remedial action if necessary. Thus project participation extends beyond actively contributing to the list itself, as "passive" list subscribers follow up on posts. Brief articles announcing the start of these projects were issued in TechKNOW, a publication of the Ohio Library Council's Technical Services Division. 8 INITIATIVES IN ERROR HANDLING The first paragraph of this article gave a description of various types of situations that can result in the failure of an online catalog to assist the user as required. Not all of the situations are errors in quite the same sense, but they must be addressed if the catalog is to function properly. Perhaps some readers of this article believe that the solution to error handling is 100% accuracy during the cataloging process. If feasible, this truly would avoid the problem in most, but not all, of the situations addressed in this article. Nevertheless the fact remains, as Ballard has pointed out: the presence of typographical errors in library databases is a universal problem. Complete accuracy is an unrealistic expectation, and is not feasible. Instead, this article describes initiatives that have been taken to remedy erroneous situations rather than to prevent them. Participation in the projects can result in wholesale improvement of the quality of a local bibliographic database. But improvement will not happen automatically. Human intervention is required. The initiatives and the situations they address can be grouped into three types: (1) Typographical errors, addressed via the Typo of the Day for Librarians Blog; (2) Classification errors, notified via the electronic lists DEWEYERROR and LCCERROR; (3) Access issues, the subject of the electronic lists PERSNAME-L, SERIES-L and SACOLIST. INITIATIVES ADDRESSING TYPOGRAPHICAL ERRORS In his article, Ballard kindly credits this author with asking the initial question that gave rise to the Typo of the Day for Librarians Blog. The blog's introductory statement reads, "We are a group of librarians from all over the world with a common interest - keeping our online catalogs free of errors." This current collaborative project is built on the foundation of 9 Typographical Errors in Library Databases, to which numerous people, notably Tina Gunther and Phalbe Henriksen, have been contributing for many years. The intention of this project is that, by correcting instances of that particular error on that day, databases worldwide can be rid of it: an ambitious goal, undoubtedly, but one which is achieved in part in each location where action is taken. One person from each institution can monitor and act upon blog posts. The blog has blossomed in a way that has sparked the imagination of numerous participants. The daily posts are replicated via an automatic retransmission on AUTOCAT, a general-purpose discussion list for cataloging. The blog is accompanied by a Wiki, used for development of the program of blog posts and as an index to record what typos have already been announced. The LIBTYPOS discussion list (not hosted by Listserv®, but a Google group) allows communication among participants and others interested. Several contributors, notably Carol Reid (New York State Library) have kept the project going. Anyone wishing to participate in writing the daily blog posts is invited to join in the activity.10 The work of correcting typos cannot be automated without danger of introducing an error where one did not previously exist. If a system has a global change function, it is best ignored for this purpose. The word Grammer, for example, featured in Beall's Dirty Database Test, is also a correctly spelled personal name. As such, it is also within scope for the PERSNAME-L list, discussed below. INITIATIVES ADDRESSING CLASSIFICATION ERRORS As with typographical errors, mistakes in call numbers sometimes happen, either from data being mistranscribed, or from the misapplication of classification schemes. When a call 10 number has an error, library users browsing the shelves can miss the materials for which they are looking. Collection development librarians reviewing the materials on the shelf will perhaps wonder why a title that is seemingly out of place is present. Even in a closed-stack environment, a call number error is a problem for a reference librarian attempting to assist with locating materials on behalf of a scholar who has sought assistance.11 Errors in Dewey Decimal Classification (DDC) numbers came to my attention in such frequency that I took action to assist all concerned. The following account describes the circumstances surrounding the establishment of the DEWEYERROR list, and its earlier manifestation, the Dewey Error Notification List. Before starting work at Marion Public Library (MPL) in Marion, Ohio on May 30, 2001 I wondered why a public library serving a community of some 64,000 residents had a need for a second M.L.S.-degreed professional cataloger. But this need soon became apparent as the extent of catalog and collection maintenance for the existing collection was revealed. I worked mostly with copy cataloging from records found in OCLC's WorldCat database. Additional work was often necessary to correct erroneous data. It seemed counter-productive to make corrections locally without also addressing the problem in the source. Thus it was not long before I began sending reports of errors to OCLC, for records found in WorldCat, and to the Library of Congress (LC), for versions of those records found in LC's own online catalog. This included DDC numbers assigned by LC's Decimal Classification Division (DCD). In giving the MPL nonfiction collection a thorough review, I examined how DDC had been applied, evaluating sections of the collection in tandem with materials selectors, who would advise on priorities and deacquisition items prior to reclassification. Extensive local practices also existed, for which I provided documentation and made additions. Although I insured to the best of my ability that DDC numbers in WorldCat master records were formulated properly, a 11 different number was often assigned locally. Thus investigation of the appropriateness for local use of the DDC number given in the WorldCat master record required a two-stage process: 1. determining that the number given in the master was strictly correct, and 2. entailing steps in local modification. Consequently I evaluated the accuracy as well as appropriateness for local purposes of DDC numbers in records for all incoming nonfiction titles. In the course of this work, records failed the verification process in the first stage sufficiently often to arouse concern about the overall accuracy of DDC numbers in libraries in general. The practice of examining WorldCat member copy records to determine whether a DDC number is correct or appropriate locally is quite common among libraries. But this practice is not the case for bibliographic records originating with LC. It is widely assumed that LC records will be free of error. That is the case for all who routinely assign them to staff charged with accepting the LC record without question. It was the process of first finding out what the DDC number was supposed to be, working strictly as instructed in the schedules and tables, that revealed that errors were indeed to be found in DDC numbers in LC records. Because of the potential benefit to other users, shortly after starting at MPL I began to communicate about suspected errors in LC catalog records. I exchanged e-mail messages with staffers at LC's Decimal Classification Division (DCD). Occasionally a DCD staff person responded with an informative correction to my report, but mostly they confirmed my suspicions and thanked me. I quickly learned that LC staff like to be informed of errors in LC's own catalog12, but are not necessarily concerned with errors in a version of that record found elsewhere, such as in WorldCat. LC staff have no control over other versions of the record once it has been promulgated to other agencies, and often other data (some of it erroneous) is introduced elsewhere. 12 Other librarians came to my attention for posting messages about erroneous records on electronic distribution lists such as AUTOCAT and OCLC-Cat, and gradually the idea formed of separating off these reports into a forum expressly dedicated to communication among all those concerned. A general forum for all kinds of errors seemed too large a project. But one could perhaps address one specific type of error, and DDC numbers seemed to be a logical choice for the project that evolved, first as the Dewey Error Notification List (DENL) and later as DEWEYERROR. DEWEYERROR In spring 2002 I set up the Dewey Error Notification List using Microsoft Outlook. I maintained this list privately on my work computer, adding people to a group of e-mail addresses as they expressed interest, and receiving and forwarding messages from participants. About thirty people participated, many of whom learned about the project from an announcement on AUTOCAT. The number of participants grew to around seventy in the first two years. The list functioned entirely through action on my part in each individual case, though my actions mostly followed those of other contributors. Those who sent reports to LC would include me as a recipient, and I forwarded the reports to the group, in addition to reports of my own. Other people, regardless of whether they actively checked errors or not, received these reports and, to the best of my knowledge, acted upon them. Thus there were two groups for whom DENL was appropriate: 1) Those who routinely accept DDC numbers from LC catalog records without checking them; 2) Those who not only check the DDC numbers, but also advise LC when they suspect a number is in error. 13 This process for forwarding messages was a labor-intensive one. The effort required, plus technical difficulties in maintaining the list using Microsoft Outlook, prompted me to seek assistance. These difficulties entailed handling messages that could not be delivered because of an incorrect address, the non-response of the recipient's host server, and so forth. Furthermore, no facility was available that would create archives of messages contributed. But a Listserv® list can have all its posts archived, and will thus preserve a record of all messages for future use and reference--provided that they are not adjusted (which technically can and does happen). At MPL I did not have access to Listserv® software, but I hoped that someone who did (most probably at a university) might offer to help. My wish was fulfilled when Margaret Maurer (Kent State University) suggested that the Technical Services Division (TSD) of the Ohio Library Council (OLC) could assist with migration of the list to a Listserv® platform. In summer 2004 the council met and readily agreed to assist with this project. The name DEWEYERROR was decided upon as the most suitable name. Since the word Dewey is a registered trademark held by OCLC, an Action Council member (Laura Salmon) sought permission for use of this name, which OCLC kindly granted. Since LC records are in the public domain no authorization is required for their reproduction. But in order to insure that no conflict of interest would arise with LC practices, I contacted DCD to advise them of the new list. Dennis McGovern (then Chief of DCD), spoke kindly of the new venture both at its outset and subsequently, even at one point contacting me for assistance with statistical data on the number of records reported. On October 27, 2004 DEWEYERROR commenced operation, hosted at Kent State University (KSU), with myself, Maurer, and Sevim McCutcheon (also of KSU) serving as co- listowners, plus an initial 77 subscriptions. TSD coordinator Bonnie Doepker (Dayton Metro 14 Library) and I announced the new list in TechKNOW. 13 We also sent announcements to various electronic lists, and the number of subscriptions grew to around 300. Eventually McCutcheon took over as leader, and Tom Adamich became the third listowner when I ceased working with DDC in January 2008. DEWEYERROR was set up under close guidelines for contribution of messages. As with the other lists that are the subject of this article, all messages are reviewed, so no post goes to the readers without first being "approved" by a list owner. The list owner does not, however, check the contents of a message for accuracy. Rather, the message is scrutinized for evidence that LC has already been informed of the suspected error. Most messages contained both LC and DEWEYERROR as recipients, so could easily be identified. But in cases where it was not evident (as when the report to LC was made independently of DEWEYERROR), a listowner would write to the sender to check whether LC had been informed, and if not, to request that the writer tell LC directly. Although LC staff wish to be advised about suspected errors, they have already set up channels of communication by which to receive reports. It was not intended for DEWEYERROR to replace or even complement those channels, although LC staffers are welcome to subscribe and read messages. DEWEYERROR is stricter in adherence to policy than was DENL, because of the special nature of the agreements under which DEWEYERROR was set up. With DENL, I sometimes distributed messages identified as "not exactly wrong" (and therefore not sent to LC), but which reflected a local practice, thinking that some subscribers might wish to use a similar number locally -- particularly if their collection was of similar size and nature to MPL's. But such messages were not in scope for DEWEYERROR. 15 Example of a Typical DEWEYERROR Report Barbara Thiesen (Bethel College, North Newton, Kansas) has provided a model example of a DEWEYERROR post in this recent one14: LCCN: 2010033097 Author: Phillips, Carl Title: Double Shadow Dewey number in record is 813.54. This is a book of poems, so the Dewey number should be 811.54. This message has also been submitted to LC. The message contains a proper identification of the bibliographic record in question, a statement of the problem, and its resolution. The final statement assures the listowner of compliance with the policy of reporting to LC. Since many new subscribers had not seen messages sent via DENL, I began to resend the ones that would qualify under the new DEWEYERROR guidelines, identifying them as reposts with an initial phrase beginning Previously reported. Bryan Baldus (Quality Books, Inc.), who contributes frequently to DEWEYERROR, had maintained a file of all posts to DENL, which came in very useful as we identified which messages to forward. Eventually McCutcheon took over the task of forwarding the messages. A search of the DEWEYERROR archives conducted on November 27, 2011 using the phrase previously reported retrieved numerous such messages, the most recently posted having the date June 8, 2010. It was one that Carl Cording (College of Saint Rose) originally sent on March 22, 2004. One might wonder why a message from six years earlier documenting a DDC error might be deemed of current interest. But for any library, if an item is still misplaced locally after several years because of an erroneous number, the matter can be of concern. By fall 2006 I had become a member of OLC TSD's Action Council, serving (eventually as coordinator) through 2008. We reviewed the status of DEWEYERROR, and decided that 16 further sponsorship was unnecessary. The list was functioning well without any intervention or assistance on TSD's part. Therefore the official connection was dropped, leaving DEWEYERROR in the hands of its listowners for routine operation. Provided that this status continues, and one should bear this provision in mind, DEWEYERROR can continue to operate as under its current arrangements. At present, it has about one post per week: not a heavy flow of traffic, but sufficient to maintain list operation and avoid the danger of becoming inactive. LCCERROR If DDC numbers in LC records have errors, one might think that Library of Congress Classification (LCC) numbers would also have them. It is however more tricky to determine cases of error with LCC than with DDC. And while working with DDC I had no business investigating LCC numbers. It did cross my mind that a companion list to DEWEYERROR, doing for LCC what DEWEYERROR does for DDC, might serve a useful purpose. After moving in 2008 to my current position at George Mason University, an LCC-classed library, and having indeed encountered cases of erroneous LCC numbers, it became practical to consider starting LCCERROR. Roman Panchyshyn (Kent State University) and Sevim McCutcheon kindly agreed to host the list at KSU. Once technical specifications and list policy were established, the list commenced operation on October 31, 2010. It currently has 71 subscribers, far fewer than the other lists. It might seem that the two lists, each addressing one of the major classification schemes in use in libraries worldwide, would have almost identical functions. DEWEYERROR and LCCERROR do share a lot in common in their scope and purpose. But the circumstances of the two classification schemes are different in subtle ways that significantly affect list operation. In 17 part this is from their history, in part from characteristics of the schemes, as well as their ownership and the circumstances in which call numbers are applied to library materials. Practically speaking: the LCC scheme does not apply just to the classification portion of a call number, but also to the shelflisting. In a bibliographic record created by LC, both DDC and LCC numbers can be included, but only the LCC number is carried out to an exact shelf location. Furthermore, libraries other than LC that use LCC create numbers for local use and in the process can integrate them with existing numbers in LC's catalog, in anticipation that other libraries subsequently encountering the number they have created will use that number. Thus the widespread application of LCC within WorldCat falls within the scope of concern of LCCERROR, whereas DEWEYERROR is restricted to numbers found within the LC catalog. The expectation of exactitude applies with LCC numbers regardless of origin; it does not apply with DDC numbers in OCLC member copy. Therefore LCCERROR was set up to include notification of suspected errors in LCC numbers in all sources, not just records originating with LC. No requirement exists that the report must also be submitted to LC. Indeed, in cases of non- LC records, no such report should be sent. As with DEWEYERROR, I was concerned that LCCERROR would cause no misunderstandings or inconvenience with LC. So I wrote to Dr. Barbara Tillett (Chief, LC Policy & Standards Division) to assure her that steps were in place to prevent messages that should properly be sent to LC from being sent instead to LCCERROR. She kindly responded: "Would it be possible for your service to send a notification directly to Mary Kay Pietris [LC Cataloging Policy Specialist] ... as our point person on LCC corrections? We have similar arrangements for personal names and some other alerts, and it really helps us respond more quickly to fix our data. As I am sure you are aware, LCC has evolved greatly over time so what 18 may be considered an 'error' in classification today was not when the book was originally cataloged. The JX schedule, which is now the JZ and KZ schedules is a good example of this. While some libraries might want to reclass their books, I'm sure many others, like LC, don't have the time and manpower. Perhaps on your list you could remind everyone of that and on the LCCERROR homepage. Catalogers should know this about LCC, but sometimes when you're looking at bibliographic records individually you lose sight of the big picture. Thanks for helping us improve the quality of our data."15 I was glad to comply with these requests, and the LCCERROR list thus does have a reader from LC in Ms. Pietris. And in her mention of personal names perhaps Dr. Tillett is referring to PERSNAME-L, which has long had subscribers among the LC staff. In posts to LCCERROR, I differentiate carefully between those commenting on records contributed by LC (which I also communicate directly to LC) and those on records originating elsewhere (and therefore are not reported to LC). Example of a Typical LCCERROR Report Jay Shorten (University of Oklahoma) contributed the following: LCCN 2011006448 The lesson of Carl Schmitt, Expanded ed., is not complete yet, but has an 050 00 of JC263.S34 M44514 2011. This number is not correct; it should be JC263.S34 M44513 2011 to match its previous edition, LCCN 98023580. (LC has changed their call number.)16 This post exemplifies use of LCCERROR to notify about a changed book number. One might question whether such a change is necessary. That question is addressed by the statement that LC changed their number: in other words, it was important enough to LC to make the change. Policy at individual libraries may differ on this matter. 19 If one considers the number of subscribers and messages as a measure of success, LCCERROR has not been as successful as DEWEYERROR. Several months after the start of operations, I asked Bryan Baldus for his ideas about why LCCERROR doesn't get as many posts as DEWEYERROR. Baldus, who contributes to both lists, answered: "Part of the reason could be that many just accept LCCs as they are, without question. Another, that it's more difficult to see the problem with an LCC than in a Dewey, since they are more difficult to quickly parse. After ten plus years working with them, I barely have any memorized, while with Dewey I can usually look at the number and determine what it means; can see standard subdivisions that are or aren't what they should be."17 The mnemonic aspect of DDC is well known, if only to DDC classifiers; whereas LCC lacks such features. The underusage of LCCERROR makes me wonder whether the time and effort spent on maintaining it is worthwhile. As the originator, I have a nurturing attitude toward the list's continuance. Practically, it would be better to close the list down rather than to allow it to become inactive. By alerting the cataloging community to this service, I hope that more people will begin to take advantage of it, make contributions themselves, and ensure that LCCERROR will continue to grow for as long as there is a need for it. INITIATIVES ADDRESSING ACCESS ISSUES Providing appropriate headings in bibliographic records, and assuring that those headings are in the proper form, are tasks with which a cataloger is well familiar. In preparing a bibliographic record for shared usage, catalogers follow rules and hopefully apply them correctly at the time. But changes in practice occur after the fact, as new headings are created, personal names are disambiguated, and series titles are standardized. Sometimes data are updated to 20 comply with revised cataloging rules, an issue about which catalogers are currently wondering as they consider implementation of Resource Description and Access (RDA), a topic that has yet to impact us with full force. Some readers will share memories of the widespread introduction of new headings with the adoption of the Anglo-American Cataloging Rules, 2nd Edition (AACR2) over thirty years ago. It also happens that local practice creeps into shared records. An example of this is encoded in the MARC format. Field 490 (which bears the digit 9, elsewhere in MARC indicative of local practice), recording series data as found in the source, is often coded with a first indicator that represents the application of a local situation rather than the decision recorded in a series authority record (SAR). Thus when LC decided in 2006 to no longer provide an added entry for a bibliographic series, but simply to use field 490 to record the series data as found, they were in effect implementing a local practice, "local" in this context referring to LC. And in so doing LC followed a MARC coding practice already observed elsewhere. If the previous sentences were made in a post to a discussion list rather than in a published article, an outburst of responses would likely ensue! But this observation is made here, not to provoke controversy, but in order to document the background against which PERSNAME-L and SERIES-L came into existence as notification lists. These lists can differ from DEWEYERROR and LCCERROR in what is considered to be an error. In some situations no error existed at the time that the original cataloging was done. In many cases, changes in headings that were correct when assigned subsequently took place for legitimate reasons. Some agencies can receive notification of the change via a third party, but for others, particularly those for whom such services are prohibitive, either for reasons of 21 expense or because of the staff time involved in attending to the reports, the PERSNAME-L and SERIES-L lists can come in handy. PERSNAME-L PERSNAME-L originated largely through the efforts of Jay Shorten to communicate information specific to personal name headings via a dedicated list. For some time, Shorten had posted notices to AUTOCAT announcing the creation of a name authority record (NAR), the modification of a name heading, or some other issue concerning a personal name in bibliographic and authority records. My initial role was to ask for his agreement to collaborate on this project. I also provided documentation for policies and procedures and wrote the welcome message. Shorten set up the list, and PERSNAME-L started operations on July 30, 2007. Whether fully detailed or a brief comment, one of the PERSNAME-L listowners approves for distribution all posts that fall within the list's scope of interest. My active role as a listowner ended in January 2008. At that time Roger Miller (Public Library of Cincinnati and Hamilton County) and Wanda Gunther (University of North Carolina) became listowners and began to share with Shorten, on a rotating basis, the responsibility of approving messages.18 Most posts to PERSNAME-L consist of reports of a changed name heading. Some subscribers including myself have on occasion asked if another subscriber has information about a person they are trying to contact for the purpose of establishing an authorized heading, or whether an existing NAR represents the individual whose book is being cataloged. In this respect, the more people participating in PERSNAME-L, the more likely it is that someone will actually know the individual being investigated, or perhaps will work at the same institution and 22 have access that is not available to outsiders. Thus PERSNAME-L exemplifies a form of "crowd-sourcing". A Notification on PERSNAME-L Sometimes a cataloger from an institution that is not a participant of the Name Authorities Cooperative Project (NACO) seeks help from a NACO member in getting a NAR created, so as to document research, provide a reference, or to record other pertinent information for those interested. Such work is illustrated by the following exchange. Rich Aldred (Haverford College) wrote: "My system reported to me that Napoleon, Art (trumpeter, cornetist, leader, and writer) is a 400 to Sudhalter, Richard M. However, there was a movie director, producer, writer Art Napoleon. There is no authority record for him, but, according to International Movie Database (IMDb), his dates are 1920-2003, so I added that to our catalog record, and the few OCLC records for his works."19 To which Deborah Tomaras (New York Public Library) kindly responded: "I've made the authority record for director Art Napoleon (ARN 8934581). I'll control all the headings with his name on them." 20 Note that Tomaras took the additional step of controlling the headings in bibliographic records, a feature of OCLC Connexion functionality. Such work, which goes beyond the scope of PERSNAME-L activity, greatly assists all future users of the records involved. Thus NACO participants can and do engage in the kind of differentiation work that lends itself to reporting about changed headings for individuals. Such changes form a substantial portion of the messages that have been contributed to PERSNAME-L since its inception. PERSNAME-L has no formal connection with the NACO project, although it receives 23 occasional mention on PCCLIST, the Listserv® list associated with (and restricted to participants in) the Program for Cooperative Cataloging (PCC) of which NACO is a component. Not all situations necessarily receive all recommended attention, however. On OCLC- Cat a correspondent inquired about an apparent mismatch in a WorldCat master record between a heading and the NAR to which it was controlled.21 That message prompted me to alert PERSNAME-L readers to a change in death dates for author Thomas G. Frothingham from 1937 to 1945, as documented in the third 670 field of the NAR (n 90606903). Another correspondent wrote pointing out that the first 670 field gave a usage different from that on which the heading had been based (middle initial G. vs. middle name Goddard), and that therefore the heading in the NAR should be revised.22 As of February 20, 2012 no such revision has taken place. Many catalogers will find themselves in the position of knowing that edits to records are appropriate, but will not do so for various reasons.23 I corrected the dates locally, in the interests of our catalog users. But owing to conflicting work responsibilities, I did not go further. The concern over the middle name/initial was in my judgment not of sufficient concern to take upon myself the additional tasks to retrieve our local item so that I could have it in hand while editing the NAR. The data that concern catalog users are not factually incorrect. In considering what action to take, the accuracy of the data plus the needs of the user are the principal criteria. PERSNAME-L has derived or incorporated related projects. An example of this is a set of data files from Gary Strawn (Northwestern University). His message List of Newly- differentiated Headings from LC Names ..."24 drew my attention to this resource, which Strawn created independently of the PERSNAME-L project. It is updated weekly and has proved a valuable source for headings that had recently been created with qualifying information that 24 differentiated one person from another namesake. Strawn says: "The reports of newly- differentiated headings are indeed kept up to date by weekly postings. It may be of interest to note that these are now accompanied by weekly listings of several other kinds of changes: authority records for undifferentiated names whose 'author of' 670s have changed; authority records deleted; authority records whose 1XX field has changed (available both as text and in MARC format)".25 The file of newly-differentiated headings formed the basis of a local project. With assistance from a student employee, I identified headings of local concern and investigated whether an update was required. In searching locally for the undifferentiated heading, sometimes I would find the one referenced in the file. In other instances, I would find yet another namesake who was unqualified locally, but is now represented by a differentiated heading in the WorldCat master record. The procedure for each heading under investigation in the local catalog entailed: 1) Perform a "browse by author" (in some cases, also by subject) search. 2) If an undifferentiated heading is found, display the bibliographic record and search in WorldCat for the current version of the master record. 3) If the heading in the master record has qualifying information, import the NAR to the local database and edit the local heading to match the master record. 4) Send e-mail to PERSNAME-L advising about the changed heading. An important caveat: One should not assume that a heading in a WorldCat master record is necessarily correct. In my posts to PERSNAME-L I avoid where possible statements concerning the accuracy of found data, preferring merely to point out the discrepancy. Instances have 25 occurred where a local record has the correct heading and the master record is in error. Sometimes a local heading contains a qualifier, but the NAR is unqualified. No standardization of message format is required, and contributors are free to write messages in any style they choose. For many messages, I use a template stating the data found locally plus that found in an WorldCat master record, with the associated NAR included in full. The posts that are the most useful for PERSNAME-L are those that can be acted upon. Such posts typically give advice concerning situations that have been resolved already. Among the issues entailed in posting a message to a list such as PERSNAME-L are the following: • The likelihood that the message will interest most subscribers • The amount of detail to be included • The probability that the information presented is accurate. Sometimes a group of PERSNAME-L contributors write several messages in succession with information on various individuals, in a chain reaction resulting in multiple disambiguations. One contributor, Stephen Arnold (University of Oxford) takes an additional step of researching the likelihood that a heading he is documenting will be found in other catalogs, checking it first against a set of other catalogs, and thereby increasing the likelihood that someone looking for that heading locally will find it.26 Thus messages from Arnold are particularly likely to result in updates to local records. On September 7, 2011 I asked PERSNAME-L readers for their opinion about including OCLC holdings in posts, so that readers can see whether their own institution will be one affected by the information in the message. Formerly, displaying holdings for WorldCat records incurred a transaction cost, but since that no longer applies, the principal expense entailed is that of the time taken to do so. For titles with many hundreds of holdings, including them in an e-mail 26 in this way would not necessarily help, but for titles with few holdings, it might cut down on users following up unnecessarily for headings they won't likely have. Opinions varied as to how useful this information will be. It is technically feasible, but not in great demand. Moreover, other bibliographic records can represent other editions of the same work with different holdings, and it proved impractical to include all holdings. One reader commented that it is worthwhile for him to check each heading brought to his attention. The catalog might not have that particular heading, but in many cases that of a namesake is affected. In September 2010 I asked subscribers to PERSNAME-L for feedback in connection with a presentation on related topics given at Library Research Seminar V (LRS-V), held in October 2010 at the University of Maryland University College (College Park). The purpose of this request was to illustrate how PERSNAME-L is used in people's work. Here are some of the testimonials provided at that time: " As I receive messages I … determine what changes need to be made … an invaluable cataloging tool." Elizabeth Heffington (Lipscomb University).27 "It’s far more manageable than the LC Authority records feed, and altogether a very useful item." Teague Allen (California Institute of Technology).28 " When I receive a message … I check our database and correct bibs (merging, change dates, initials, etc) then I check the LC auth list and import the corrected auth record. I have been able to clean up and correct many headings using the PERSNAME-L." Barbara Stampfl, (Polk County Library Cooperative, Florida).29 "I have, on occasion, been able to provide information … that one of my list-reading colleagues can use to create an authority heading that the rest of us can then use." Dennis Reynolds (Madison Public Library, Wisconsin).30 27 " I go through each message and check our database heading by heading. It is time consuming but quite worth while. When it comes to undifferentiated headings and headings used in old records I was able to update those headings in our database and make sure the vendor includes us the corresponding authority records in our next batch… The other great thing … experienced NACO catalogers. I learned a lot by reading how they investigated to break conflicts. …" Tzu-Jing Kao (Multnomah County Library, Oregon)31 " … these posts often led to a discovery of multiple people referenced on a single local name heading. … I hate to do major work on records with only my library being the beneficiary. It can be time consuming to determine whether or not NARs are duplicates or not, and to spend this time is hard to justify if only my library has the information." Jason LeMay, (Gwinnett County Public Library, Georgia).32 " I use it to disseminate some info that we've garnered for our own purposes but which may be useful for others and may not otherwise be readily available to them. … any credit Oxford builds up … is amply repaid. … Roger Miller was invaluably helpful over a tricky series … which I'm sure was in part because Cincinnati has benefitted from Oxford posts." Stephen Arnold (Oxford University).33 I asked a similar question on AUTOCAT and, in addition to comments like those above, also requested and received some responses from individuals telling why they do not participate in PERSNAME-L. Most telling was the following, from a person who requested not to be identified: " I do feel that I'm doing some really good work that would come in handy to others, but, alas, I have been told explicitly that our mission is not to take extra time for the larger cataloging community. I just wish I could spend more time going back and fixing our own messes." 28 The number of positive contributions received outweighed the negatives. Nevertheless, the negative contribution is well grounded. It is not only an economic issue but a philosophical one. That is to say, a cataloging agency might not have the financial resources to be able to participate in PERSNAME-L. But that is a separate issue from whether one should or should not take the extra steps required to participate in the projects described. Such agencies do receive some of the benefits the projects confer, because they will encounter and benefit from records on which other agencies have worked to provide corrections and so forth. The following comment from Anthony Franks (formerly Cooperative Program Section Head at LC) is a resounding endorsement of our collaborative activity: "Do keep up, however, this list strikes me as a far more efficient mechanism than waiting for LC or OCLC update reports."34 As of this writing PERSNAME-L has about 350 subscriptions. Posts to the list occur at a rate of two or three per working day (the archives for October 2011 have 51 entries). It has significantly altered the bibliographic landscape, enabling wide-scale disambiguation both in local catalogs and in national databases. In the event that name headings in bibliographic records are eventually linked, not to authority records (as with OCLC Connexion's "control headings" functionality) but to other resources, PERSNAME-L will have greatly assisted. Envisioned are links from personal name headings in bibliographic records to: biographical and bibliographical tools; dictionary and encyclopedia articles; personal web pages and social media sites. Linking to such resources requires accurate data in bibliographic records. In particular, the currently still sanctioned practice of allowing a NAR to represent first one person, then subsequently (through undifferentiation and redifferentation) for that same NAR to represent a different person, must be deprecated. 29 SERIES-L SERIES-L, dedicated to issues with bibliographic series, started operation in 2009, with Wayne Sanders (University of Missouri-Columbia), Kathleen Schweitzberger (University of Missouri-Kansas City) and myself as co-listowners. The host site is the University of Missouri- Columbia Listserv® installation. SERIES-L was intended primarily as a notification list, but actually the phrase "action list" was used in the initial publicity and list documentation, in part intended as a gentle hint to users that posts upon which a user could not act were to be avoided in favor of those giving directions for immediate database maintenance. But in the event, this phrase might have discouraged people from posting. It was my expectation that SERIES-L would be comparable in performance to PERSNAME-L, with regular posts from a band of dedicated contributors. But in comparison, postings were few and far between. Nevertheless, the list unexpectedly found a special niche. Roger Miller and Bryan Baldus effectively made SERIES-L viable by contributing lists of SARs, primarily for adult fiction and children's literature, and often having a heading entered under a personal name. In fact Miller mentioned that he had been in discussion with Margaret Maurer with a view to setting up a list expressly dedicated to such SARs. The appearance of SERIES-L fulfilled their purpose, and the contributions of Miller and Baldus have been the mainstay of the list's existence. Joan Condell (Dallas Public Library) wrote, "I work at a public library, and series, especially kids/YA series, are very important to the public services librarians. … I found a lot of series that needed controlling."35 It had also been my expectation that academic librarians would take advantage of SERIES-L to collaborate on getting their indexes of series holdings in order. One instance which was awaiting my attention at the time of list set-up was Biblioteca de autores españoles. The 30 concern with this publication was its complicated volume numbering, with several subseries interspersed amidst the general enumeration, and requiring special description in analytic bibliographic records and their associated holdings. The details of this project were more complex than could be accommodated in a message to SERIES-L, so instead I offered to send the documentation (kept locally in a Microsoft Excel file) to anyone wishing it. In the event, one or two people wrote, a disappointing but understandable rate of response. Perhaps other libraries already had these volumes in order and required no further attention. But a more probable expectation is that this series was not a priority for catalog editing and maintenance at the time that the list posting appeared. One consolation is the fact that the post remains in the list archives and can be retrieved, eventually, by anyone knowing that this source is available. One might think that in the wake of LC's decision not to trace series, SERIES-L would serve as a means whereby those librarians who considered tracing series in accordance with existing and new SARs would take up the opportunity to communicate with each other. But despite the outcry of protest against that decision, SERIES-L has not served in this way. Since at the inception of SERIES-L it was made plain that discussion of the LC decision would be inappropriate, perhaps people shied away from the list for that reason: although it was discussion of the decision itself, not dealing with its consequences, that was out of scope. Why has not more interest been shown in SERIES-L, compared with PERSNAME-L? Any number of reasons might account for the low response. In order for an electronic distribution list to continue in service, a regular flow of posts is required, lest people simply forget about its existence. A recent post of mine to OCLC-Cat listed all the lists described in this article: and although announcements have been made for all these lists, plus occasional reminders, one correspondent replied to OCLC-Cat that he was unaware of them. With 31 DEWEYERROR, I used the previously reported messages not only to bring those messages to more people's attention, but also to generate a regular flow of postings in order to keep the list in people's active memory. For this reason, although discussion was not encouraged on SERIES-L, I have learned not to attempt to quash it when it arises. It's better for a list to have a regular flow of posts, some of which might not qualify as "actionable," than for the list to wither through inaction. Thus SERIES-L has somewhat more posts that qualify as discussion than was originally intended. SACOLIST This list is one with which I have had no direct involvement beyond subscribing and reading the messages posted. It is dedicated to issues about LC Subject Headings (LCSH) and shares many characteristics with the lists already described. It differs from the other lists and projects desribed in this article in that LC hosts and administers it directly. SACOLIST has its own web page, which states: "The listserv may also be used as a vehicle to foster discussions on the construction, use, and application of subject headings. Questions posted may be answered by any list member and not necessarily by staff from the Coop Team or CPSO." 36 Applicable here are the comments from Sloan's article, reviewed above, concerning the "personality" of the list. It has functioned mostly for announcements, notably of LC's Subject Editorial Review Meeting, and provides valuable feedback concerning the meeting's rationale in cases where a proposed subject heading was not approved. Recently, list subscribers have responded critically to some of the announced decisions, giving the list more the character of discussion than of notification. The web page designates it as The SACO Listserv, and it is associated primarily with the Subject Authority Cooperative Program. Subscribers might thereby conclude that posts should be related to that program. But the web page clearly indicates 32 that a broader scope is intended. Notification of errors in the manner and practice of the lists discussed above has not occurred--yet. It remains for people to make use of the opportunity afforded. ISSUES AND CONCERNS All of the projects described in this article are collaborative in nature. Yet with the exception of SACOLIST, none of them has required the approval of a professional body. Otherwise, the only involvement of any such body was the sponsorship mentioned in the establishment of DEWEYERROR, which after two years was discontinued, leaving the list to continue to operate without any noticeable change. My role has been that of an initiator, but without the cooperation of all those mentioned--and many others, to whom I apologize for not having named them individually, they are simply too numerous--these projects would never have come to fruition. People continue to post to general discussion lists, such as AUTOCAT and OCLC-Cat, on matters that are within the scope for notification lists. They are entitled to do so, for the messages are also in scope for the general lists, and it is entirely a matter for the individual to decide in which forum to post a notice. Perhaps those who post elsewhere do so because of the greater number of subscriptions that the general lists have (each has several thousand, compared with a few hundred at most for the notification lists). Or perhaps people are simply not mindful of the existence of the pertinent notification list. When feasible, if I see someone posting elsewhere then I write to advise them of the notification list's existence. But this is a time- consuming activity, and does not always get the desired results. Another option is to repost the information. This can be done by requesting the writer of the message for permission to forward it. When doing so, you can also ask the person if they 33 would prefer to send it on themselves. Such messages do not always get responses, however. Reposting messages without permission is to be avoided since it is a generally deprecated practice. Therefore when cooperation is not forthcoming, I write a fresh message to the list on which it belongs, mentioning that it was occasioned by the previous message, and giving the name of its author and list. POSSIBLE FUTURE PROJECTS A vision for a future notification list dedicated to a particular aspect of cooperative quality control arises once people have identified the concern which it will address. Some readers might have guessed that, if a list dedicated to personal names has prospered, a companion one for corporate names can be established. "CORPNAME-L" is indeed a possibility. Here is why such a list does not yet exist. First, no one has yet started one, a simple enough reason. Anyone can set up a list if they so choose, but a host site is a prerequisite, as are co-listowners. Although technically an individual acting alone can get a Listserv® list set up and serve as the sole listowner, doing so is contrary to the collaborative spirit that has been expressed in the existing projects, and runs the risk of a domineering personality unduly affecting the nature of the list, as well as trouble if the person stops performing the tasks necessary to insure its ongoing viability. When setting up lists I have sought to get different host sites with two or three people. Hopefully doing so avoids consequences such as would occur if one Listserv® site hosting all the lists should become dysfunctional, thereby rendering them all inoperable simultaneously. Having multiple listowners also shares the responsibility among several colleagues, and allows for a succession of leadership in administering the list. 34 A more substantial reason why CORPNAME-L does not yet exist is that corporate names present different challenges and opportunities than personal names. Corporate bodies change their names for different reasons and at a different rate, with earlier and later headings to be accounted for in NARs. Some of them have hierarchical structures, which must be accommodated according to the rules for subordinate bodies. Even the type of publication with which they are associated is different: corporate headings are more likely than personal names to be associated with serials, and in many libraries serials cataloging has traditionally been the domain of a specialist. Hopefully, CORPNAME-L will eventually be realized : any reader with a mind to participate in it is invited to take the initiative. It is also hoped that the list would be successful. A listowner sometimes needs to take steps to insure that it does not become inactive. Nothing is worse for a list than to set it up, announce it, have a group of people subscribe, and then for no posts to appear. Effectively, such a list is a "silence" list, not qualifying under any of the categories described above. It is understandable that a once thriving list might become inactive: all it takes is for those making posts to stop doing so, without people stepping in to take their place. But to set up a list that from the start is inactive is to be avoided. Furthermore: the Listserv® manager at a host site can notice that a list is inactive and take steps to discontinue it, along with the archival record. To prevent possible dysfunctionality of this nature, a prospective listowner of a new list is recommended to have a store of situations laid up so as to provide substance for a number of messages to get the list started. Once those messages start to appear, hopefully other readers will begin to make their own posts. Eventually, a "critical mass" of list contributors can emerge, assuring the list's viability. 35 Finally: it is not known what the "tolerance level" for these projects is. Several people participate in more than one, or indeed all of them, and one does not wish to set up a situation in which someone finds the number of projects to be especially burdensome. For this reason, it is recommended that monitoring each project be a responsibility shared among several co-workers. CONCLUSION The projects discussed in this article constitute a radically different approach to cooperative quality control than has existed prior to the electronic era. Communication via electronic mail allows notification to large groups of people. On some discussion lists for cataloging, notably AUTOCAT and OCLC-Cat, people ask for help in doing their jobs, and receive it: and in the process, most other subscribers read the messages, and the aggregate amount of time expended is considerable. The dedication of PERSNAME-L and other lists to notification about specific situations entails a much closer correspondence between the time spent reading a message and the benefit received. There  are  many  ways  in  which  staff,  technical  services  managers,  administrators,  and  other   organizations  can  work  collaboratively  to  improve  the  records  that  are  contributed  to  shared  databases.   It  would  be  advantageous  to  all  stakeholders  if  employees  are  encouraged  by  managers  and   administrators  to  contribute  to  these  collaborative  efforts  that  have  the  potential  for  far-­‐reaching   positive  effects.  The  role  of  cooperative  organizations  such  as  OCLC,  and  vendors  that  supply   bibliographic  records  to  libraries  should  also  not  be  overlooked.  Together,  all  of  these  stakeholders  can   work  together  to  improve  the  effectiveness  of  bibliographic  control  and  increase  library  efficiencies. APPENDIX A Contact and subscription information for the discussed Listserv® e-mail lists. 36 In some cases, registration with the host Listserv® site might be required. DEWEYERROR: Subscription address: listserv@listserv.kent.edu Address for posting messages: deweyerror@listserv.kent.edu Address for contacting listowners: deweyerror-request@listserv.kent.edu Web page for subscription: https://listserv.kent.edu/cgi-bin/wa.exe?SUBED1=DEWEYERROR Web page for archives search: https://listserv.kent.edu/cgi-bin/wa.exe?A0=DEWEYERROR LCCERROR Subscription address: listserv@listserv.kent.edu Address for posting messages: lccerror@listserv.kent.edu Address for contacting listowners: lccerror-request@listserv.kent.edu Web page for subscription: https://listserv.kent.edu/cgi-bin/wa.exe?SUBED1=LCCERROR Web page for archives search: https://listserv.kent.edu/cgi-bin/wa.exe?A0=LCCERROR PERSNAME-L Subscription address: listserv@lists.ou.edu Address for posting messages: persname-l@lists.ou.edu Address for contacting listowners: persname-l-request@lists.ou.edu Web page for subscription http://lists.ou.edu/cgi-bin/wa?SUBED1=PERSNAME-L Web page for archives search https://lists.ou.edu/cgi-bin/wa?A0=persname-l SERIES-L Subscription address: listserv@po.missouri.edu Address for posting messages: series-l@po.missouri.edu Address for contacting listowners: series-l-request@po.missouri.edu 37 Web page for subscription https://po.missouri.edu/cgi-bin/wa?SUBED1=SERIES-L Web page for archives search https://po.missouri.edu/cgi-bin/wa?S1=SERIES-L The SACO Listserv (sacolist@loc.gov) To subscribe, etc. follow the instructions on the SACO Listserv web page http://www.loc.gov/aba/pcc/saco/sacolist.html. APPENDIX B ABBREVIATIONS AACR2 Anglo-American Cataloging Rules, 2nd edition. ARN Authority Record Number (OCLC) DENL Dewey Error Notification List DDC Dewey Decimal Classification DCD Decimal Classification Division (Library of Congress) KSU Kent State University LC Library of Congress LCC Library of Congress Classification LCSH Library of Congress Subject Headings MARC Machine Readable Cataloging MPL Marion Public Library (Marion, Ohio) NACO Name Authorities Cooperative Project NAR Name Authority Record OCLC Online Computer Library Center OLC Ohio Library Council 38 PCC Program for Cooperative Cataloging RDA Resource Description and Access SACO Subject Authority Cooperative Program SAR Series Authority Record TSD Technical Services Division (Ohio Library Council) NOTES 1 Terry Ballard, "Systematic Identification of Typographical Errors in Library Catalogs", Cataloging & Classification Quarterly 46, no. 1 (2008), 27-33. 2 "AL ASIDE--IDEAS," American Libraries 22, no. 3 (March 1991): 197. 3 "Typographical Errors in Library Databases" is available online at http://librarytypos.blogspot.com/. 4 Jeffrey Beall and Karen Kafadar, "The Effectiveness of Copy Cataloging at Eliminating Typographical Errors in Shared Bibliographic Records," Library Resources & Technical Services 48 no. 2 (2004): 92-101. 5 ibid., 97. 6 Bernie Sloan, “Electronic Discussion Lists,” Journal of Library Administration 44 no. 3-4 (2006): 203-225. 7 ibid., 204.   8 ibid., 210. 9 The remarks in this paragraph, and elsewhere in this article, might not apply exactly to SACOLIST, "The SACO Listserv", discussed below, for which the author had no responsibility, and which merits consideration because it falls within the scope of concern. 10 An index to past postings, with sign-up instructions for contributing, is available at http://libtypos.pbworks.com/w/page/17113321/FrontPage. 39 11 Thomas Mann, Reference Librarian at LC, has described such activity in his essay "What is distinctive about the Library of Congress in both its collections and its means of access to them, and the reasons LC needs to maintain classified shelving of books onsite, and a way to deal effectively with the problem of 'books on the floor'", Paper prepared for AFSCME 2910, The Library of Congress Professional Guild, November 6, 2009. http://www.guild2910.org/Future%20of%20Cataloging/LCdistinctive.pdf . 12 LC's catalog is available online at http://catalog.loc.gov. 13 Bonnie Doepker and Ian Fairclough. "DEWEYERROR Discussion List has a New Home -- with OLC" TechKNOW :11 issue 1 (March 2005):1-2. This announcement and those of other lists described in this article are available at http://www.library.kent.edu/page/11234.   14 Barbara Thiesen, e-mail message to DEWEYERROR, January 6, 2012. https://listserv.kent.edu/cgi-bin/wa.exe?A0=DEWEYERROR 15 Barbara Tillett, e-mail message to author, November 9, 2010. 16 Jay Shorten, e-mail message to LCCERROR, November 1, 2011 https://listserv.kent.edu/cgi-bin/wa.exe?A0=LCCERROR 17 Bryan Baldus, e-mail message to author, April 1, 2011. 18 Miller retired in April 2012 and no longer serves as a listowner. 19 Rich Aldred, e-mail message to PERSNAME-L, August 4, 2011. https://lists.ou.edu/cgi-bin/wa?A0=persname-l 20 Deborah Tomaras, e-mail message to PERSNAME-L, August 4, 2011. Tomaras is now a cataloger at South Portland (Maine) Public Library. 21 Adger Williams, e-mail message to OCLC-Cat, October 5, 2011. http://listserv.oclc.org/archives/oclc-cat.html 22 Jack Hall, e-mail message to PERSNAME-L, October 5, 2011. 40 23 In several workshops and conference presentations with Brenda Block (Manager, OCLC Quality Control Section) I have reviewed numerous reasons why people do not correct errors -- as well as reasons why they should. 24 Gary Strawn, e-mail message to PERSNAME-L, November 6, 2008. 25 Gary Strawn, e-mail message to author, December 18, 2011. This file and others are available online at http://files.library.northwestern.edu/public/AuthLoadReport/Names/ . 26 Stephen Arnold, e-mail message to author, May 9, 2012. 27 Elizabeth Heffington, e-mail message to author, September 28, 2010. 28 Teague Allen, e-mail message to author, September 29, 2010. 29 Barbara Stampfl, e-mail message to author, September 28, 2010. 30 Dennis Reynolds, e-mail message to author, September 28, 2010. 31 Tzu-Jing Kao, e-mail message to author, October 5, 2010. 32 Jason LeMay, e-mail message to author, September 28, 2010. 33 Stephen Arnold, e-mail message to author, September 28, 2010. 34 Anthony Franks, e-mail message to PERSNAME-L, May 20, 2010. 35 Joan P. Condell, e-mail message to author, April 2, 2009. Quoted in Ian Fairclough, "SERIES-L: A New Tool for Cooperative Quality Control," TechKNOW :15 issue 1 (June 2009):11. http://www.library.kent.edu/files/TechKNOW_June_2009.pdf . 36 "About the SACO Listserv (sacolist@loc.gov)" http://www.loc.gov/aba/pcc/saco/sacolist.html