ARCHIVAL INFORMATICS NEWSLETTER Part 1 of Archival Informatics Newsletter &. Technical Reports
ISSM 0892-2179 Sua.or.1987 Yoluao 1.· 2
welcome. Firstbecause theyare chall~ THIS ISSUE inthemselves. and are the kinds of opimon Actually. Canada has been justnorth ofus for a pieces which would nothave found their longJime. even though itmay seem from this issue of way to J!rin.t ex~ through a vehicle such as theNewsletterthatwe justdiscovered it! this. an(l second oecausethey demonstrate a As~ throu~outthisissue. archival automa need for the Archival InformaticsNewsletter tionactivityinCanidahasreachedacrescendorecent more than anvt.bin.e I can;:r' ly. Tom Brown's column (p.5-7) reports on the ex The first isSue ofArchL Informatics citing I-ASSIST and ACAmeetings. The article by Technical R~. on Optica1Media in John"McDonald (p.14-15) exploresthe challeJ:Jges archives and museums. was published in being faced bythe new National Archives ofCanada May. Issue 2. on Colle~ Software in and declares the rights of archivists to be enfranchis archives and museumswill e availablein ed inthe requirements definition t>tocessesforinfor August. The fall issue will be devoted to the mation~ms serving all aetiviueswitbintheiror requirementsforcollectionsm~ement. It ganizations. On pages 15-16. I introduce the Plan will discuss not only the information needs ning Committee on DescriptiveStandardsand their ofcollectionsmanagers and define software plans (whatmy Smithsonian colle~es were fond of ~ms to meetthem. butalso examinethe Chiding me byca11iJJg "plan-plans"). relationship between co11ectionsmanage But Canada doesn't have a monopoly- on debates mentand otberinstitutionalmissions. about descriptive standards. On pages 9-13. Usa Weber. the automation officerofthe SM. presents a TABLB OF CONTENTS discussion ofthe confusion and confiietSUlTOunding description of microforms (and. by extension. any Museum Automation at AAM 2 "copies" inanothermedium) within theframework of ARTFL and Textual Archives 4 the MARC Formats for Bibli~hic Descri¢on. If the detai1stend attimes tobehairsplitting. itfefleets Tom Brown on: the ~lem oftaying to define a ~icdescriptive I-ASSIST Conference 5 cataloging rule to meetthe needs Of a V8riety ofuser ACA Conference 7 communities. No single. "logically consistent" view seems to do for all. Letters to the Editor 8 This. of course. isthe challenge faced by museum Standards 8 informationprofessionals whose situationis the sub jectofmyreflections on the AAM me~ which was Lisa Weber on: JUst held inSan Francisco (p.24). Amidst some Describing Microforms: 9 promising developmentsininformationstandards and some tentativeofferings of computer software. the John McDonald on: chaos of Babble. each with his or her own language. Electronic Records &. the new reigned. And yet. there was a sense of promise inthe National Archives of Canada 14 air. of a new beginning I am delighted to be running letters to the editor. Canadian Descriptive Standards 15 The interest. controversy. and information exch~e Conferences 16 potential ofthe first issueis immenselygratifying.-1 In-Box 17 hope thatthe requests for dialog issued by each ofthe Software Briefs 20 aufhors of pieces in this issue Will find an equally responsive audience. Capturing Rich Contexts 22 The contributions I received to this issue are doubly Report on Col1ecting Software 24 Museum Automation at the AAM only important if we want to communicate with by David Bearman anyone, and then reported on how the project she directs. AVIADOR, has make use of both AACR2 I think we will come to see the year that the (library community cataloging standards) and American Association of Museums met in San the AAT. Eleanor Fink, reported in the use of Francisco as a watershed in museum automation. MARC by the Index of American Sculpture and Although there was a certain immaturity expos the development of shared data definitions be ed in the balance of sessions (only one session tween that project, the Index of American Paint devoted specifically to mainline automation con ing and several other art database projects at the cerns out of four overall, with the others dedicat Smithsonian National Museum of American Art. ed to CD. digital color imaging, and artificial The MARC format was serving well as a means for intelligence), automation emerged as a full inter-system data sharing. fledged partner in the professionalization of Lenore Sarasan of Willoughby Associates was registrars and collections managers. The same to present the "devils advocate" position. But as immaturity was evident on the exhibit floor. she argued that standards could be developed by where systems accessing videodisc or digitized users of commercial systems. and presented the images attracted more attention than a five digit work her firm was doing in developing common numeric field in a database deserves. Yet some of data dictionaries and authority files for clients. the systems being displayed were practical and it was evident that this sphere of standardiza could be implemented essentially as is. And al tion did for those users exactly what profession though it was clear that most AAM members are wide standards might do for all museums. Lenore still lost in the present of automation. it was evi revealed that any argument about the need for dent that an adequate number can not only see standards in the museum community is no longer the outlines of a future but are taking actions to about ends, but about tactics. assure that it evolves according to a profession Two promising tactical developments were ally dictated plan. featured in sessions. The first. already briefly mentioned, is the revision of Nomenclature by a Sessions: committee (albeit self-appointed) representing a While many attendees wlliiong remember the range of museums which have used the classifica extraordinarlly crisp color images presented by tion system. More important than the second HowlU'd Besser and the experimental digitizing edition itself (although it will improve on the systems at the University of California. or the de first). is the self-conscious discussion which the lightful tour of artificial intelligence pastures group generated about the purposes of an in Stuart Dreyfus' keynote address. the museum hierarchical classification system and the differ profession will be influenced over the longer run ence between those ends and the aims of an index by the dialog begun in three sessions devoted to ing vocabulary. While the discussion at the ses discussions of evolving information exchange sion itself did not resolve the purpose of the new standards. edition. it paved the way for increasingly sophis The one session explicitly devoted to museum ticated discussions of the aims of different types information standards, chaired by Deirdre Stam. of standards. focused the issues. Toni Petersen (editor of the The third session was a report to the commun Art and Architecture Thesaurus) spoke to the ity by participants in the spring Conference on a very concept of a standard as something support Common Agenda for History Museums. sponsored ed by numerous agencies and not subject to uni by the Smithsonian Institution. The meeting was lateral change. Recognizing that standards are held to identify actions required in four areas: expensive to maintain, she insisted that they are collections, collaboration. interpretation and essential in our complex society. She then pre documentation. The working groups will be con sented the AAT as a potential vocabulary stand tinued under a broader umbrella of the AASLH ard. Jim Blakaby reported on the efforts of a in the next year, but their two day output is it committee involved in revision of Cbenball's self exceptionally promising. especially in the Nomenclature. a topic to which he had previously difficult area of common databases and document addressed himself in a session devoted strictly ation where the group launched a survey of the to the revision process. He presented Chenhall data fields currently used by history museums as a classification system, which could be a as a first step towards normalizing the data in a standard while being open to local addition of data dictionary which could help museums to specific terms at the lower levels. Angela Gira" plan for common documentation. The modesty of of the Avery Art Index. noted that standards are the effort bodes well for its ultimate success. as
2 Archival Informatics Newsletter voL 1, #2 does the openness of the subcommittee to a per collection or a modest history or archeology mus missive data standard. The conference report eum. Only the US Park Service DBase III System, will be published in fuB by the AASLH this fall. which is too limited and unsupported to be taken seriously as an option for others, and the Oracle Systems: based MYMSY (Willoughby) are built on commer Six museum information retrieval/cataloging cial DBMS's. The others have a few startling systems were exhibited at San Francisco along weaknesses as a consequence. with numerous membership/development. finan At present the vendors of museum software cial, ticketing. shipping and other applications. products are not doing a very good job of differen With the exception of a new offering from QL tiation themselves, in part. it appears, because Systems (The Volunteer Management System), I they do not tnow the capabilities of their compe will restrict this report to comparing these six. tition. Each claims to be better able to support a First. because it is too good to pass up. I want museum because of their experience in museums. to share my enthusiasm about the QL product but none can point to more than a handful of in ($399; $299 from AAM to AAM Members). The stallations of their present system. There is a Volunteer Management System is, simply, the definite tendency to mislead naive users about best human resource management package devot the virtues of operating systems (Plct vs. Concur ed to this tedious but critical function I've ever rent DOS, vs. MS-DOS vs. S-38 OS) which are irre seen. It performs in an elegantly simple fashion, levant to evaluation of the applications. Questor retaining information abour your volunteers and made a splash with the videodisc link in ARGUS their skills, the availability and the cumulative (they can have a five chuacter field and a cable experience, reporting with ease on who can fi 11 a to a videodisc!) and Willoughby wowed the crowd need. It even sends year end thank you letters with Its digital Images in the QUiXIS demonstra with total hours neatly merged using a word pro tion file (using Picture Ware for imaging in a cessing function. Most archives and museums I demo which had no other real functionality). Ob know of can use this IBM/PC based product as is, viously other vendors could have done the same. and have needed it for years. I was encouraged by one important shift - while AAM exhibit goers could try out ARGUS (by some vendors demonstrated their systems on Questar), ARTIS (by the Williamson Group). PC's, the commercial applications are desilned MYMSY (by Willoughby Associates), STAR (by to run on mini-computers for multiple users. Cuadra Associates ), STIPPLE (by Erros Comput ARGUS, STAR and STIPPLE showed thesauri and ing) and a U.S. National Park Service D-Base 111 Willoughby promised one in Super Mimsy and system. ISounds a bit like the monkeys got my QUiXIS in the fourth quarter. This is encourag keyboard for a moment there.I In addition. they ing. although must museums are probably not could hear about MILAM and QUEXIS (both by generally ready to use a thesaurus well. Willoughby). but not use the actual products Both Willoughby and Questor emphasized the because QUEXIS is not yet developed and MILAM importance of quick entry for retrospective data was not on-line to the exhibit floor. and their ability to custom make a data entry The most interesting observations about this screen for clients (although neither demonstra crop relates not to what they are, but to what ted any functionality in this respect not shared they all aren't, None of them are essentially col by STAR and STIPPLE). ARTIS, apparently aware lections management systems. They all lack some that its eisht screens could inhibit data entry minimal functionality in this respect (e.g. tick severely. promised to reform. What is more in lers, generalized collections actions statistical teresting to me is the insistence on the part of reporting and life-cycle tracking). Instead they museum oriented vendors and museum staff on are information retrieval, or cataloging, systems. retrospective conversion. rather than on simply They tended to be thin for experienced users starting up the system and using it progressive none of them provide a direct command Interface ly. This reflects the fact that the systems are in addition to menu driven capabilities, none pro not very capable in collections management (and vided for easy record redefinition or systems ad that information retrieval functions are suspect ministration within the confines of the applica with incomplete databases) and the assumption tion itself (features present, for instance, in the made by museum staff that all the records of the public domain ILS or MIN ISIS packages ). Nor are museum will be (should be?) automated, rather any capable of processing or generating MARC than simply "pointed to· by the system. While records to interface with library based systems. both these assumptions seem to me invalid, this None yet has an installation larger than about issue deserves a full treatment. I welcome com 100.000 items. the size of a tiny natural history ments for a future issue. I will address this
Summer, 1987 Copyright by Archives & Museum Informatics 3 further in the Fall Technical Report. Reauire the Universit.y of Chicago and the Centre Nation ments for Collections Management Software. A al de la Recherche Scientifique. The database. comprehensive comparative analysis of museum which now consist.s of over 150 million words, and archives software packages is scheduled for was initiated by the French Government in 1957. Archival Informatics Technical Reports. Spring as as step in the creation of the Tresor de Ja 1988 (v. 2 #1). Langue Francaise. a new dictionary. Si nee J984. ARTFL has been organized as a consortium which VENDORS: can be joined by any degree granting inatH-ution Cuadra Associates. Inc. (STAR) ($750 p.a. for PhD granting and $400 for others); 11835 W. Olympic Blvd .. Suite 305 more than 20 major American Universities are Los Angeles. CA 90064 (213-478-0066) members of the program, contributors of texts, and users of the ARRAS. (ArChive Retrieval and National Park Service Analysis System). ARRAS. a fUll-text system Curatorial Services Division currently being converted from an IBM3081 t.o P. O. Box 37127 UN IX. does not analyze a text in the sense of Washingtion. D.C. 20013 (202-343-8138) interpretation. but it does provide statistics on word occurrences. concordances. occurrences Questor Systems Inc. (ARGUS) within contexts. and indexes. It can produce 844 Colorado Blvd. graphic distributions of results and do proxim- Los Angeles. CA 9004 I (2 13-258-5174) ity searches. Materials in the database range from trouba STIPPLE Database Services dor poetry and lyric poems in the old Provencal Warren Farmhouse language. to a body of texts from the revolution rhame Lane. Culham of 1848 and modern works by Apollinaire. Oxfordshire. UK OXI4-3DT (44-235-24676) Bonnefoy and Meschonnic. It has been used for research into neologisms. the significance of The Williamson Group (ARTIS) certain streets in Parisian novels. t.he origin of 129 Mount Auburn St. concepts such as "opinion publique" and "admin Cambridge. MA 02138 (617-497-6848) istration" in intellectual and political history. and for teaching. Recognizing that the database Willoughby Associates. Ltd. (MIMSY & QUEXIS) could be more useful to scholars. ARTFL is cur 2800 Sheridan Place rently rewriting the search systems to permit its Evanston. IL 6020 I (312-328-3284) use on smaller systems and cooperating with the Textual Information Retrieval and Analysis (TIRA) project at the University of Chicago in a cooperative software development effort. THE ARTFL SCANNER AND The acquisition of the Kurzweit Optical Char TEXTUAL DATA ARCHIVES acter Recognition device. which was paid for by the Packard Foundation. is part of this effort by A recent announcement in the ARTFL Newslet ARTFL to serve its community in new ways. Jt ter that the long established project on American seemed to me that this kind of project has poten and french Research on the Ireasury of the tial for bringing those special items held by ar french bangange had acquired a Kurzweil scan chives. items of historical significance. to the at ner and was offering ASCII encoding of printed tention of researchers who can make use of them or typed texts as a free service to members, sug in new ways. As we begin to consider the role of gested a number of opportunities for fruitful. if primary materials in teaching and scholarship not artful. cooperation between archives and in the "hypermedia" learning environment many scholars interested in text analysis. It also ser major universities are planning for the 1990's, ved to remind me of the risk of overlooking text the role of textual databases (along with the ual data archives. and increasingly graphic and image bases and sound bases of museums). will even performing art data archives, in our focus become more critical. Machine-readable does on the more traditional social science and govern not just mean social science. ment data archives. IARTFL. Department of Romance Languages & ARTFL is a textual database ·of 17th - 20th Literature, Universit.y of Chicago. 1050 East century french language texts in literature. 59th St., Chicago, IL 606371 philosophy the arts and sciences. maintained by David Bear.an
Archival Informatics Newsletter vol.t, #2 NORTH OP THE 49thPARALLEL to have decided the early turf war in favor of By"[honw E. Brovn those arguing that a major purpose of data archi Events during the last few months have rein ves is to facilitate access. In this connection. forced the leadership image of Canadian archiv Nasatir stated that there has been a convergence ists in the management of automated information of data archivists and traditional librarians. He concluded that the influence of the librarians' systems. inviolable commitment to access underpinned the progess which data repositories have made in I-~ promoting access to computerized information. As it does every four years. Canada hosted the Despite the keynoter's acknowledgement that 13th annual I-ASSIST conference at the end of other unnamed information professions had been May. The International Association for Social involved. session chair and Association Presi Science Information Service and Technology. is dent. Judith Rowe of Princeton U., noticed some the professional association of data archivists consternation among the audience and twice ex and data librarians. As one would expect. four plicitly acknowledged those outside of the libra days of workshops and format sessions covered a ry community who have helped data archives ad variety of topics important to archivists in tradi vance. Following Nasatir's presentation, she not tional institutions who administer computerized ed the role of the SAA Task Force on Automated files. Topics included trends in the use of mach Records and Techniques and in her conference ine readable data. data interchange standards. ad summary, Rowe detailed the contf"ibutions tf"adi vances in storage and dissemination technology. tional archives have made to the data repository and tools to train data archivists. effort. She specifically mentioned appraisal cri The keynote speaker. David Nasatir of Califor teria, standards for research use of administra nia State University--Dominguez Hills. outlined tive fHes, and preservation. the history of the social science data archives. One of the conference workshops focused on He contrasted two philosophies which have com CULDAT, the Canadian Union List of Machine peted since the mid 1960's when these archives Readable Data Files. Paula Mitchelt and Edward were beginning to flourish. One school argued Hanis of Tycho Research. Associates descf'ibed that the purpose of the data archives was to facil CULDAT as a nationwide conputerized inventory itate and promote access. This group went so far of machine-readable data in Canada, created as a as to propose that aU data created with govern source for information products and to help re ment grant or contract funds should be available searchers identify and locate machine readable to the research community for the cost of dupli files. CULDAT wlU support an 00-line service cation. The other approach concentrated on try available nationally and a reference periodical. ing to insure the proper use of the data. This CULDAT has standardized its record format meant restricting the information to social scien based on the MARC format for machine-readable tists associated with major research institutions. data files. developed cataloging rules with auth since only they would be in an environment in of"ity lists, and constructed an online cataloging which the data could be properly analyzed. The system. Tycho estimated that up to 10.000 re tension between these approaches led to the de cords will be in the database when completed. mise of the Council of Data Archives. the first or The final of the program session was a 150 ganization concerned with data archives. whose minute marathon of five papers. Organized by members were North American repositories and Sue Gavrel of the National Archives of Canada. it whose purpose was to establish standards for presented a glimpse of the successes and fail archives of computer files. I-ASSIST emerged to ures of traditional archives in dealing with auto replace the Council. as an organization of indivi mated infotmation systems. In the first presenta duals concerned with data archives and their tion a records manager for the government of operations. Nasatir argued that I-ASSIST has British Columbia. Rueben Ware, outlined his been instrumental in the progress which data aggressive program to effect the economical de archives have made in the last two decades. Dur struction of computer records of tempOf"ary value ing this time, he noted: (t) the expansion of the and to identify those records which have archi scope of research use of materials from data ar val potential. Not suprisingly. the Provinceadop chives. (2) increased use of data archives in soc ted the "systems approach" to inventorying and ial science research projects (3) additional ways scheduling which the national repository has of disseminating information. and (4) standards pioneered in Ottawa. Ware discussed some crea for data description and data interchange. All of tive means to get the information sYstem mana these trends have improved access and thus seem gers involved with records management but the
Summer. 1987 Copyright by Archives & Museum Informatics program illustrated a common disjunction be information in computerized form was of great tween ideals and realities. The Provincial Archi interest to social scientists, it was not easily ves may be able to identify records with archival accessible. First. some of the data remained in potential, but British Columbia does not have a the hands of RDA's contractor and was unavail program to accession them. Thus despite the pro able for analysis by the agency. Since the data ductive efforts of the records managers to inven was organized for administrative purposes. it tory and schedule the disposition of the compo needed to be reorganized and adapted to differ nents of the province's automated information ent kinds of software for statistical analysis. systems, conputerized materials with archival Furthermore, it was evident that the information value may still be lost. In her paper, Margaret would be more valuable if it could be linked to Adams took up the archival management of auto Norway's Central Statistics Bureau. Leipart out mated information in the Kentucky project which lined how his data library acquired. reorgan was discussed in my last column. She again call ized, and linked the RDA data into a valuable ed on records managers and archivists to rethink information resource. In the final paper. Mario approaches developed before the advent of elec F. Lopez-Gomez of the National Archives report trostatic copiers and microfilm, let alone com ed on the researchers interested in the computer puters. In the earlier era, records were static ized datafUes accessioned by his organization. and information sharing was not possible. Since Using survey data for the period January 1985 to technology has advanced, she urged the archival June 1986, he reported that most researchen profession to focus on information sharing in the did not write but used the telephone to obtain present rather than reference service in the information about the records in the National future. However, she had to report that no action Archives. The focus of attention was on finan has yet been taken about her recommendations cial, transportation, and military records. Final (or Kentucky to move in those directions. In the ly, he reported that his researchers were equally third presentation, William Deimer outlined his divided among other government agencies, acade efforts to establish a municipal data library for mic institutions, and private firms. While it the City of Los Angeles. Beginning around 1970, would be quite unusual for most archives to have the Community Analysis Bureau within the muni one-third of its clients come from private firms. cipalgovernment acquired a variety of computer it is not for repositories of computerized mater files relating to Los Anseles. Federal Census ma ials. terial was supplemented by a materials from city This session had a variety of themes and sovernment, includins tax information. crime re counter themes. currents and countercurrents. ports, ambulance runs, morbidity, natality. etc. Each paper outlined a different function of the From a reservoir of over 1500 data tapes in the archival administration of computerized informa city'S central computer center. the Bureau provi tion -- inventorying, scheduling, accessioning , ded analysis (or policy and administrative deci processing, preserving, and disseminating sions. But in 1982, the Bureau suffered a severe machine-readable data files. These different budset reduction. As a result. use of the data functions were reported from different penpec declined. Because of the infrequent use, the data tives-- records manager. archivist, policy files were routinely blanked until none of the analyst. and academic social science researcher. city's records remained. When Deimer contacted And the five presentations focused on different the municipal archives (or assistance to prevent levels of government: two from the national the destruction of the data, the response was level, two were from the state or provincial level, sympathetic but ineffective. Undeterred, he is and one from the municipal level. Yet in this now workins to establish a Statistical Bureau diversity, a coherent whole emerged. The first within the city government to acquire and three papers candidly reported that while indivi analyze data related to Los Anseles. duals can create a foundation for the archival After these three disheartenins reports. Jorn administration of automated information sys Leipart of the Norweigan Social Science Data Ser tems. these will come to nought if, as in these vice provided a welcome balance. He reported on cases, the responsible archival repository seem his organization's effort to acquire and dissemin ingly fails to respond. Just when we began to ate data produced by the Norway's Regional De wonder whether to bother. the final two speakers velopment Agency (RDA). This agency provides demonstrated that with proper institutional economic assistance to businesses interested in support, the information had research value and locating in rural areas of the country and. in the could be of interest. process, acquires a variety of information on these businesses and their activities. While the
6 Archival Infocmatics Newsletter vo1.l, #2 ASSOC. Of CANADIAN ARCHIVISTS and corporate perspective. While each of these During tbe last four or five years, countless control systems implemented Gillis' prerequi formal and informal meetings have discussed site that organizations know what information is holding a conference on archives and automation. under their control. not one of the systems had I have participated in no less than six or seven incorporated their respective archival repositor such meetings. Invariably, someone will com ies into the design of the control systems. As ment that the last conference devoted to archives one person observed from the audience, automat and automated records systems was held in Ann ed control of the tife cycle had stopped at the Arbor in 1979. The Association of Canadian archives' (ront doorl Archivists changed that. At its annual meeting Sessions covered a variety of other topics deal during June. every paper of every session con ing with automated records. One session present cerned either the archival administration of auto ed case studies by two archivists involved in the mated records or use of automated techniques for appraisal. acquisition. and preservation of valu the control of archives. We can now say that the able information generated in automated offices. last meeting devoted to archives and automation Their perspectives were dramatically different; was the ACA conference in Hamilton, in 1987. one shaped by a large corporate structure and Under the theme"Archives and the Informa the other by the needs of a small operation. I tion Age", the program explored the implications participated in a session on appraisal in the for archivists, both in theory and in practice, of information age. Five separate sessions dealt new, repidly changing information technologies. witb automated control of materials in various In the opening keynote, Hugh Taylor argued that sizes of institutions. These dealt with both the changing nature of information technology microprocessor and mainframe systems, strate has transformed the archives through its impact gic data design information system planning, and on culture, records. the computer, the research approaches to networking and descriptive stan er. and the archivist. For the next three days. dards. One interesting sessions on archival auto every paper of every plenary and concurrent mation asked bow researcbers view such sys session discussed Taylor's transformations and tems. Do our systems meet the users' needs? Do practical responses to them. the users even care? One theme which ran through several ses The program committee cottected about half sions was information resources management of tbe papers into an informal set of proceedings. (IRM). Presentations examined whether or not For information on haw to acquire a copy, write: archivists have a role in information resources Association of Canadian Archivists, P. O. Box management and if so what should that role be. 2596, Station D, Ottawa, Ontario, KIP 5W6. This discussion exposed the disparity of mean ings of the term, and while no clear definition ACA fOOTNOTE: REPORTS AVAILABLE emerged during the conference it was clear that In his ACA paper, Jay Atherton (Director working definitions of IRM were reshaping pol General, Historical Resources Branch) reported icy. Peter Gillis. of Canada's Treasury Board, tbat the Public Archives of Canada has been outlined how the Canadian government was re experimenting with ways to control the disposi vamping its government-wide information poli tion of records in automated information systems cies after deciding that "the Hrst crucial deci and has tested several pilot projects. The find sion that we had to take was that most of our cur ings indicate a need to take a fresh look at how rent policies in this area are obsolete. This was records are scheduled and suggest a methodology somewhat difficult to admit since the most re based upon a system approacb: analyzing and cent policy, records management, dated only scheduling all data and information used and from 1983." (How many of us with records man produced within a definable administrative pro agement responsibility will conclude that our cess. A report and other information on these current policies are likwise obsolete as a result pilots is available from John McDonald (as of the information age?) A representative from reported by him, pp.14-151. Also available Canada's Department of Energy, Mines and Re tbrough Mr. McDonald is a fascinating study on sources outlined how she had implemented the interchange standards for a variety of different new requirements into a system for control of purposes and for records containing a variety of the information resources within the depart different types of information. It is called "Data ment. Her theme was echoed in another session and Document Intercbange Standards". This entitled"Automating the Life Cycle." In it. the study, produced under contract to the Canadian speakers outlined automated records manage government. paints a clear portrait of the ment control systems -- from federal. municipal, complex family of interchange standards.
Summer. 1987 Copyright by Archives & Museum Informatics 1 LETTERS TO THE EDITOR STANDARDS
Frank Burke (NARA) wrote on April 30, Common Command Lan ~uale: that in spite of my reservations about Louise R. Levey reports, 1n the Bulletin what OPTIRAM claims to be doing (v.1 # 1) of the American Society for Information "The National Archives has been impres Science v.13 #5 (June/July 1987) on the sed with the results of its tests of OPTIRAM status of the proposed standard for a on a limited number of documents that it common online command language, The has subjected to the process. These consist proposed language. which uses the typical ed of holograph letters, 19th century ships verb-object structure of most command passenger lists, printed forms with hand languages (DISPLAY, FIND etc,) contains written insertions, typewritten 3x5 cards 20 primary commands and the syntax in French, and a variety of typed, printed rules for t.hem, The standard wiU be sub and handwritten documents. The tran mit.ted for review and vote shortly, Ifit scription accuracy rate was very high. passes, a future in which a user could The Archives is now trying to find out if reasonably expect to search a variety of anyone else has developed a comparable databases without having to learn each of system for scanning these mixed formats." their command languages is imaginable. [ On May 29, NARA released Solicitation To review the standard, contact Pat Harris, NASP-N2-P-0046 for an "Indefinite quan Execut.ive Director. NISO, National Bureau tity contract for optical character recog of Standards, Administration lOI/Library nition for handwritten materials", with E-106, Gait.hersburg, MD 20899, bids due on 7129/87, so we may find out whether OPTIRAM or anyone else has the MARC FORMAT: technology it advert.ises. ed.] At its June 1987 meeting, MARBI approv ed the SAA proposal for defining a sub Elizabeth Betz Parker (LC) wrote on May 17 field in 851 (locat.ion) and 853 (location that. the "tC Thesaurus for Graphic Mater of Originals/DupHcates) to permit sorting ials: Topical Terms for Subject Access. com repositories by count.ry or state/province piled by Elizabeth Betz Parker, introduc of the U.S .. Canada. U.K. and Soviet Union. tion by Jackie M. Dooley. will be available It accepted a proposal to drop some form of (I hope) in late June. 617p." reproduction and media codes in 008122 "The Prints and Photographs Division and 008123, keeping those which indicated has begun entering records for groups of how the information was to be played back photographs in the MARC format.,. The or read. In a move of importance to ar records will be distributed to subscribers chives and museums, MARBI a.dopted, with of MARC tapes and will therefore be avail modifications, a proposal to use the 583 able in RUN and OClC. (However. none of (Actions) field to record information about our records have been verified yet so as to preservation actions. Now that libraries initiate distribution. Should be pretty have adopted this approach to collections soon)" management. data recording we can hope to see expanded support for it by the Glen McAninch (KY State Archives) wrote bibliographic networks and extensions of on June 16 that he would welcome contri the concept in loca.l software systems. Two butions of references and articles to be days of preliminary discussion by MARBI used in compiling a revised and updated of the complex proposal for "format bibliography on archival automation integration", produced equally pre incorporating some citations (perhaps as liminary consensus around the proposal to many as 300) from Richard Kesner's pre view "seriality" and "archival and vious bibliographies and all post 1983 manuscript control" as ways of looking at materials. Glen can be reached at the cultural materials which are distinct from Kentucky Department of Libraries and bibliographic item description. Further Archives, P.O.Box ~37, 300 Coffee Tree Rd., discussion can be expected in January, but Frankfort, KY 40602 I see little propect for a final agreement before July 1988.
a Archival Informatics Newsletter .. vo1.l, #2 DESCRIBING MICROFORMS AND THE MARC FORMATS A Discussion Paper Usa B. Weber, Society of America Archivists
describing microforms. Only when this I. INTRODUCTION slippery question is answered, can we ask Both archivists and librarians are hav how to use a specific format (AMC, Books, ing difficulty using the MARC family of or Serials) to describe the microform. formats to describe microforms. Some of Although the focus of this paper is on their difficulties reside with the format. microforms, these same issues present But, underlying the technical problems themselves with respect to material copied are deeper confusions and thornier issues by other techniques, including xeroxing. which need to be addressed before we can Archivists need to be able to use the MARC develop satisfactory solutions. formats to describe these kinds of mater Ubraries and archives approach micro ials consistently, and archivists and lib forms differently. Ultimately, the reason rarians must face these same descriptive lies in the fundamental concepts which issues associated with technologies, such distinguish archival from library mater as machine-readable magnetic records ials. For a discussion of microforms, we and the various optical disk formats. The need to make two distinctions; library and National Archives and the Library of Con .archival materials differ both in the inten gress are already experimenting with opti tion associated with their creation and the cal disk technology as means of preserv methods we employ to control them. U ing information. And, an increasingly brary materials result from deliberate, large number of indexes and publications intellectual acts, with a creative purpose. are being distributed on CD-ROM. There Archival materials are the residue of dis fore, any solutions to handling microform parate activities which often span long within the MARC family of formats must periods of time. Ubrary items are usually take these, and future, technologies into in a single format. Archival materials are account. often collectivities comprised of materi8.1s in a variety of formats. This paper out II. MICROFORMS: LIBRARY POINT OF VIEW lines the issues as a first step towards a Not surprisingly, the kinds of micro solution which both the archival and lib forms librarians encounter are primarily rary communities can accept. "micropublications" or microforms that Historically, what distinguished archival are created for wide distribution. Their from library materials, has been the con emphasis is, therefore, on bibliographic cept of publication. Ubrarians collect control. Unfortunately, changes in the items that are published and exist in many cataloging rules have confused the issue identical copies while archivistsand manu of library microform cataloging. script curators collect unique records The Anglo-American Cataloging Rules, created during the course of daily activity. First Edition (AACRl) established the prin These "clear" distinctions are becoming ciple that a microform was to be described increasingly obscured by technological in terms of the original work, so micro changes in the production and distribu form publication details were relegated to tion of information. And, although archi a note. This rule assumed that microforms val materials are usually unique, modern cataloged by libraries were primarily cop collections often contain "near-print" and ies of already existing published entities. other published materials such as books The second edition, AACR2, took a different. written or collected by the creator of the tack. Since a microform requires special collections. equipment for its use, under AACR2 rules Yel, answering the question ofwhether microforms are re8arded as a special type the materials in hand are published or un of library material. AACR2 rules require published ( library or archival) is crucial ~hat the cata~oger de~cribe th~ microform to u' the M Cf . of for a
Summer, 1987 Copyright by Archives & Museum Informatics 9 original item in a note. chivists view microfilming as a process The AACR2 approach to cataloging micro in the life-cycle of information. as such, forms provoked much controversy in the microform is a means to an end (or a tooD, library community. To understand why, it not a new item to catalog. Often archival is importantto know that librarians distin repositories either run their own micro guish between reproduction and original form laboratory or have easy access to microforms. A "reproduction" is a micro one. (Of course, librarians may also pro form which is a copy of a pre-existing bib duce microforms and many reasons cited liographic entity O.e. a book or serial). An here for archival microfilming also apply original microform is more difficult to de to libraries. but I'm making distinctions to fine. Glenn Patton, ofOCLC. explains that contrast the approaches). As such, archiv original microforms include items or ists use microfilming ( and other forms of collections without a previous bibliogra "copying") in a variety ofways that are phic identity. For example. items brought not mutually exclusive. It is not unusual together specifically for the purposes of for archivists to film the same materials producing a microform publication would for several reasons. produce an "original microform." A. Filming for Users The Library of Congress took an official 1. Reference or scholarly copying stance contrary to AACR2. Its rule inter To provide offsite researchers with pretation 11.0A states that for microforms unique (and therefore non-circulating) that are reproductions, LC continues to materials. For a fee, it is not unusual for a follow the AACRI principle of describing repository to microform part of all of the original in the title and statement of collection for a researcher who cannot responsibility, edition. publication. distri travel to the institution. bution and physical description areas. 2. Copying for publishing Information about the microfilm is placed , This is the category that mostdirect in field 533 (reproduction note). For origi ly relates to the library dicussion ofmicro nal microfilm. LC follows AARC2. The moti publications. Commercial publishers are vation behind LC's position stems from the interested in archival materials to make benefits of derivative cataloging and the them available for wider dissemination at economic inefficiency of AACR2 for "re a profit. Sometimes micropublications are production" microforms. Following AACRI partially funded by a granting agency allowed the cataloger to use the orginal (most often the NHPRC) and distributed by cataloging record to derive a new copy, a micropublisher. Commercial microfilm and merely add anote. Following AACR2 publishers film collection editions (a sin would require that the cataloger create an gle coUection from a single repository entirely new record or modify the record such as the Draper Manuscripts from the of the original work extensively. State Historical Society of Wisconsin) or Whether they follow AACRI or AACR2 coHected editions (materials about a topic. rules, library catalogers must record the event, or person that are gathered, select same information in the record. The ed, and filmed from a number of repositor difference in approach will be apparent ies.> only for "reproduction" microforms. 3. Publication of holdings Closely tied to commercial micropub MICROFORMS: THE ARCHIVAL VIEW lications (and often overlapping) is in house (or out-of-house) microfilming to Archival concerns vis-a-vis microforms disseminate a repository's own holdings. are very different. First. since archivists This category is also related to reference administer "unique" materials, derivative or scholarly copying only the purpose is cataloging has never been an issue in the for wider distribution and the products are archival community. Related to this more polished. These "publications" may situation is that fact archivists have not, be no more than duplicates of a master until recently, been concerned about negative retained from a reference re universal cataloging standards. Previous quest. But established programs such as ly, each archival repository"cataloged" the ones at the Library of Congress and its coUections as it saw fit. the National Archives. issue these kinds of But the primary difference is that ar- products are of high quality. This may
to Archival Informatics Newsletter vo!.1, #2 confuse matters further. as some catalog Who made the reproduction ers t.hen treat. LC and NARA as commercial 1. Inhouse publishers. 2.0ut-of-house B. Acquisition copying Availability source To acquire archival materials which 1. Own repository are owned or housed at other repositories, 2. Other institutions, or are in private hands. Ownership of master negative C. Preservation 1. Own repository To preserve the intellectual content. of 2. Other deteriorating original documents and to protect original documents from the wear IV. DESCRIBING MICROFORMS USING MARC and tear of use. THE CURRENT SITUATION D. Bulk Reduction Having discussed the background to the To save space, The bulk of many library approach to cataloging micro modern archival collections is so forms, and the archival use of microforms. tremendous that archivists choose to film let us examine how librarians and archiv the records and dest.roy the originals. ists currently catalog microforms of archi val and ma~uscriptmaterials using the Other archival concerns: MARC formats. Because t.hey see microfilming as a pro A. Librarians: cess suited to any or all of the above ends, 1. Which format to use? archivists often film, or acquire on film, The choice of "which format to use7" only parts of collections. Scrapbooks, clip should be straightforward. The MARC For ping files. case files, and other forms of mats for Bibliographic Description offer materials are routinely filmed as partof clear guidelines in Leader/06 (Type of processing to J?reserve information that Record) which states that: exists on fragIle media. reduce the bulk of "Microforms. whether original or repro the collection, or acquire part of the col ductions, are not identified by a distinctive lect.ion held by anot.her inst.itution. These type-of-record code. i.e. the type of mater microforms are very different indeed ial characterisitics described by the codes than materials produced by micropublish take precedence over the microform char ers and intended for wide distribution. acteristics of the item." Finally, in addition to making microform Therefore. according to MFBD. all micro copies in the course of managing their forms of manuscripts materials should be collections, archivists administer archival described in the AMC format. However. collections in which microforms are the this is not the practice of library catalog "original" documents. Many government ers. Both OCLC and RLG users are catalog agencies and large institutions with ing some microforms of archival and man voluminous documentation responsibili uscripts materials in BOOKS, SERIALS, and ties create COM (computer output micro AMC formats. form) as the original record. (When I a. Library catalolers Use BOOKS first heard of the library distinction be and/or SERIALS format for"commercial tween reproduction and original micro ly" generated microforms (see discussion forms, I immediately thought of COM as concerning the question ofwhat is a "original" microforms). "commercial publisher"). Proponents say Reflecting their orientation, archivists these belong in either BOOKS or SERIALS are concerned the following data elements because they are "published" Summer. 1987 Copyright by Archives & Museum Informatics 11 530-Additional physical forms available is were faced with cataloging archival mater not. present. in BOOKS, Catalogers using ials in accordance with library standards BOOKS or SERIALS either leave this infor to create descriptive records that could be mation out of the record or put it. in t.he integrated into library catalogs. Their 500, general note, field. concerns in cataloging microforms were, b, Library cataloaers who use t.he however, different from those of the lib AMC format. following the MFBD's, encoun rary community. For example, archivists ter problems with this format as well. Be are more often faced with cataloging cause commercially generated microforms microforms created for a variety of rea often exhibit "publishing" kinds of infor sons (see discussion on page 4-5) . Catalog mation, fields not valid in AMC are needed. ing micropublications of archival mater For example, commercial microform pub ials is just one aspect. lishers often produce series of microforms The four archival repositories taking and therefore catalogers need the 4xx and part in the initial RLIN AMC implement 8xx series fields. (Although not valid in ation (Cornell. Hoover. Stanford, and Yale) MFBD, OCLC makes 4xx and 8xx series field developed a seriesof guidelinesfor catalog available in AMC). ing reproductions in AMC. These suggest 2. Reproduction vs. original aspect ed guidelines have become the de facto Regardless ofwhich format the cata standards and are what SAA teaches in the loger chooses, the question ofwhether the MARC AMC format workshops. The RLG microform is a reproduction or an origin implementation group decided to catalog al must be faced. OCLC, based on LC policy, micropublications of archival materials in gives their users guidance to distinguish AMC. However, they were much more between t.he two. RLG has not issued any concerned with cataloging the other kinds guidelines, although their Archives and of microforms and concentrated on Manuscripts Task Force is examining t.he developing guidelines for these cases. issues. RLG has. however. been more 2. Reproduction vs. original aspect involved in helping archivists describe Instead of concentrating on reproduc reproductions in AMC (see below). tion versus original aspects, the RLG "OCLC advises users to catalog micro archivists' major concern was to develop forms of archival collections. or parts of standard ways to use 530/533/535 within archival collections. as reproductions us the library context and at the same time to ing the AMC format when the collections answer archival needs. The kinds of infor exist prior to filming and as originals mation archivists want to include in the when the collections are brought together records answers the question "does my re to generate original editions in micro pository own or have custody over the ori form. LC uses similar guidelines. but tends ginals that were filmed? if not, who does?" to catalog more of these items as originals. 3. Use of 531 note fields For example, in cases where part of a col Following the RLG guideline, archivists lection is filmed, OCLC has advised librar using RLIN AMC catalog microforms in the ies to catalog the item as a reproduction following way: while LC would probably advise libraries 530- Additional physical form available to catalog the item as an original." Ifall or part of the archival mat OCLC tells its users that. within AMC, erial is available in a different physical if it is a reproduction, describe the origin format ( microform. photocopies, publish al in 245-300 with a note in 530, 533, or ed book?) and your repository holds the 535. Ifitis an original. describe the micro originals, describe the originals in 245 form in 245-300 with appropriate notes in 300 and use 530, to note the additional 53x. These guidelines are based on LC physical form. guidelines. ParentheticaHy, LC does not 533- Reproduction note presently use the AMC format and describ Ifyou own only the reproduction es microforms of archival and manuscript and the originals were I) owned by you materials in BOOKS so their rules about use but destroyed or 2) owned by a different of 53x notes differ. entity and either still extant or destroyed, B. Archivists then describe the material in hand in 245 1. Which format to use? 300 and note the reproduction informa- With the advent of AMC, archivists 12 Archival Informatics Newsletter vo!.1. #2 tion in 533. Although dictated by the de 3. his solution violates current MFBD scription of field 533. archivists do not distinctions. always describe the originals in 245-300. 4. Itwould confuse searching the data This field is used in conjunction with 535. bases if researchers/reference librarians 535 - Location of ori~inal/dup1icates search under rather than books Use only in conjunction with 533 when 5. How would this solution effect an describing location of originals (because archivist who has responsibility for cata you use 530 when YOU have the originals). loging this kind of "commercial" material Note the location of the originals or note if but has software that only creates AMC the originals were destroyed in 535. Ifa records (e.g. MicroMARC:amc). repository tracks the location of duplicat B. Put all microforms of archival mater es. note those locations in 535 as well. ials in AMC. C. Issues and Problems Problems: 1. The distinctions between "original" 1. LC would need to validate fields for and "reproduction" is difficult to apply to publication information including the microform materialsofmanuscripts co11ec series fields. tions. 2. This goes against fundamental 2. Confusion stems from trying to use archival principles. fields origina11y developed by library cata Possible solutions to the problems of de logers for other formats (533 and 530) in scribing reproductions in AMC are not as archival ways. For example, archivists dis straightforward. To redefine fields 530 tinguish between the use of 530 (addition and 533 so that they are more 10gical11y al forms) and ~33 (reproductions) both by consistent, is not possible because 530 is who owns (or has custody) of the materials valid in the VISUAL MATERIALS, SERIALS. and whether the materials still exist. The and AMC formats and ~33 is valid in the RLG guidelines are not lo~ica11y consistent BOOKS, VISUAL MATERIALS, MAPS, MUSIC, because library practice Isn't; therefore, SERIALS and AMC formats. they are confusing to apply. An option suggested by Max Evans is to Specific problems occur when you try create one large field (~3~ since it is only to describe the followin g: valid for AMC?) and use it to hold all the - a collection which a repository owns, necessary subfields archivists require. has filmed and has destroyed a part of; - copies of your material, available for VI. CONCLUSION purchase from another institution It is easy to get bogged down in the com - the distinction between preservation plexities of this topic. What is most impor negatives and positive copies. tant to keep in mind, (and what we te11 the 3. There is inconsistency in what SAA MARC-AMC workshop participants) archivists are describing in 300. Some is to create a catalog record that expresses repositories are describing originals and to the user what you want it to. However, microform and some are describing just this is often easier said than done. originals. SAA NEWS V. POSSIBLE SOLUTIONS The NEH awarded the SAA most of the The two principal issues discussed in this funds it requested to continue offering the paper are which format to use to describe MARC AMC workshops, a workshop on de "commercially produced microforms" and scriptive standards for the next two years, how to describe various. kinds ofreproduc- revise Steven Henson's archival catalog tions within the AMC format. Two possible ing manual and publish a book of descrip solutions are discussed below. A. Catalog all "commercia11y produced" tions in AMC format, as well as to support microforms of archival materials in BOOKS the SAA Automation Information Center. or SERIALS By now, most American archival repos Problems: itorieshave (hopefu11y) returnedquestion 1. LC would need to validate some of naires concerning automation inevtments the AMC fields for BOOKS/SERIALS. to Lisa Weber at the SAA office. Data is be 2. How to define"commercially pro ing entered into an SAA database where it will become part of a clearinghouse on duced?" archival automation. Summer. 1987 ., Copyright by Archives & Museum Infonnatics .3 PLANS & PROPOSALS: tion Archival Informatics Newsletter vo1.I, #2 scheduling should be incorporated into is policies, and procedures, that govern the sues that have already been identified management of information in a given within the institution, These can range organization. Suggestions concerning the from the space problems associated with establishment of (a) model(s) of the infor tape storage. to the impact of legal and mation universe from the perspective of other accountability requirements. to the the issues raised in this article would be need by senior management to establish useful. comprehensive views of their institution's Among other activities related to the Na information holdings (possibly through tional Archives' involvementwith federal the linkage of corporate finding aids such government data holdings, steps have as automated records systems and data dic been taken to develop guidance concern tionaries), ing the application of General Records Given that the data holdings of an institu Disposal Schedules to data in automated tion are often scattered across a variety of information systems. Similarly, a project diverse program activities (operational is underway to produce a retention and and administrative), it has been found use disposal authority for so-called transitory ful to build these data scheduling and data records or those inconsequential records conservation functions within those areas that are normally of temporary value and that will offer the greatest return for the are considered to be neither corporate nor energy invested O.e, through the selec part of the official information system of tion of highly valuable corporate data the organization. This latter project has holdings which best reflect issues that can been particularly challenging and any be identified and owned at the senior cor ideas or suggestions regarding the criteria porate level of the institution). Finally, if that could be used to define this body of re data scheduling and data conservation con cords would be welcome. siderations are to figure prominently in Finally, a recently completed study has the design steps leading to the installation presented options for potential National or major modification of systems/surveys, Archives involvement in national and in then archivists and records managers will ternational data and document interchan be required to view themselves as corpor ge standards activities. This study describ ate users - users who have a right to ex esthe objectives, structure, and responsibi press functional requirements that ought lities of national and international stand to be respected by systems designers to the ards organizations and assesses the appli same extent as the requirements expressed cability of certain standards to the con by the primary users ofthe systems or sur cern of the National Archives for the on veys. going care of digitally recorded informa With respect to the nanagement of elec tion, particularly as it is managed in fed tronic documents. a setofdraft (very preli eral government institutions. minary) functional specifications for the Copies of this report, as well as other in management of information in integrated formation associated with the issues raised office support systems has been developed. in this articles, may be obtained from: In contrast to the data scheduling efforts John McDonald, Director, Automated Infor described above. these specifications focus mation Systems Division, Government on the management of documents (un Records Branch, National Archives of structured text created, transmitted, used, Canada, 395 Wellington Street, Ottowa, retained and disposed of in a non-systema Ontario, KIA-ON3 or call (613) 996-0969. tic manner - essentially a 'document man agement' environment supported by the tools and procedures associated with the CANADIAN DESaUP'I1VH STANDARDS field of records management>. It is my privilege as editor to introduce a The rather simplistic division of the in new Canadian colleague -- the Planning formation universe into data management Committee on Descriptive Standards. and document management worlds raises a Following the publication ofToward Des: number of questions concerning the man criptive Standards. in 1986, the Bureau of ner in which archivists should view the Canadian Archivists established a commit information universe. For the order one tee to ensure that the recommendations of gives to this universe, will determine the that report were acted upon. It consists of Summer. 1987 Copyrightby Archives & Museum Informatics IS two representatives each of the ACA and CONFERENCES the Association des archivistes du Quebec plus the Secretary General of the Bureau of Canadian Archivists (jacques GrimaI'd) July 12-14 and a PAC observer. When it met for the International Conference on Databases first time early in 1987, it defined its objec in the Humanities & Social Sciences. AUM. tives as: Montgomery. Alabama Establishing descriptive standards and rules for the intellectual control of archi July 22 - 25 val materials a) at the fonds d'archives National Association of Government level (to accomodate all media). and b) by Archives and Records Administrators. medium at the series. file unit and item Colony Square Hotel, Atlanta. Georgia level, Working with the Canadian Committee August 2-5 on MARC and the National Library of Cana Recognition Technologies Users Associa da to adapt the format. [CanMARC is only tion, Hyatt Regency. San Francisco slightly different from USMARC. but it is politically quite another beast.! September 1 -5 Studying a) existing names authorities Society of American Archivists. Grand and rules and b) the problem of subject Hyatt. New York. SAA. 600 5, Federal St.. indexing of archives and archival finding Chicago. IL 60605 <312-922-0140) aids, and adapting them to archival prin ciples and practices. October 4-8 This breathtaking agenda is supported ASIS 50th Anniversary Conference, by the Canadian Coun cil of Archives Sheraton Boston Hotel, Boston. ASIS, which provided a grant to enable the 1424 16th St., NW. Washington, DC 20036 committee to hire a project officer and to (202-462-1000) hold its meetings. Two short term projects were undertaken. The first is to advise the October 12-13 Canadian Committee on MARC about the Museum Computer Network, Royal view of Canadian archivists towards MARC Sonesta Hotel, Cambridge. Mass. format development. The second is to pro $90. incl, 1 yr. membership. toP.O.Box 111, duce a guide for archivists on construc East Winthrop. ME 04343 tion and use of name authorities (drafts will be ready by the end of summer). October 19-22 The main project for the year has been ARMA Annual Conference, Anaheim, CA assigned to a subcommittee of media spe Int. Rec. Mgmt. Council, 22243 Miston Dr" cialists who will establish the rules for Woodland Hills. CA 91364 control of archives at the fonds d'archives level. They are expected to report by the October 21-23 end of March 1988. The second group has Local Television News Archives Confer also been appointed. and while it will be ence, Madison WI. gin its discussions this year. it cannot pro Sponsored by the National Center for ceed far on rules for description at the Film and Video Presentation, American series, file unit and item level until the Film Institute, with funding from the first group reports. NHPRC. For archives. libraries and The Planning Committee on Descriptive museums that preserve local television Standards will keep interested parties news and public affairs broadcasts, $25, informed of its progress through an registration deadline August 1. AFt 2021 occasional newsletter. For subscriptions, North Western Ave., Los Angeles, CA 90027 write to: (213-856-7637) Diane Beattie. Project Officer, Planning Committee on Descriptive Standards, c/o November 11-13 Public Archives of Canada, 344 Wellington Optical Publishing and Storage. Penta St.. Room 4101, Ottowa. Ontario K1A-ON3, Hotel. New York CANADA or call (613)-99)-2372 Learned Information. 143 Old Marlton Pike, Medford. NJ 08055 16 Archival Informatics Newsletter vol.l, #2 send updates and recommendations. Because the IN-BOX Library will maintain the list, we will aU bene REPORTS: fit; we need {.o put aside other quibbles and adopt the language. We can work to make it work. American Association of Colleagiate Registrars and Admissions Officers (One Dupont Circle NW, Miller, David C.• Soecial Report: Publishers. Washington. DC 20036), Retention of Records: A Libraries & CD-ROM: Imolications of Djgjtal t Guide for Retention and Disposal of Student Re Optical Printing. A report for the Fred Meyer ~ 45pp.. 1987 ($8), is an exceptionally valu Charitable Trust. March 1987, 99pp. I able report of a very productive AACRAO Task Millers' style is very chatty and the structure of Force. Not only does it treat the legal issues and the report is somewhat telegraphic. If you requirements for a records retention program, it already know the basics. Millers opinions are provides excellent advice on micrographics, com intriguing, if debatable. puter readable records and electronic imaging systems. A fascinating appendix reports. state New York State. Governor's Office of Manage by state, on the policies governing disposition of ment and Productivity, State Archives & State academic records of closed schools. Education Department: Computer and AudJo visual Records in State Government: Preliminary Coopers & Lybrand, Information & Image Report of the Special Media Records Project, Management: The Industry &. The Technologies. April 1986, 69pp. Coopers & Lybrand, NY, 1987, 67pp. & glossary. This is a report on a cooperative study of 19 This study. commissioned by the Association NY State Agencies conducted by the authoring for Information & Image Management, is a fore organizations In 1985/6. Their findings. that the cast of the market and trends in this dynamic use of computer and audio-visual information field. Significantly. the study concludes that was increasing dramatically and that the state micrographics will not be made obsolete by other government was ill equipped to handle them are means of dense data storage in the near term. Not not surprising but their examples of the import surprisingly it looks towards greater integration ance of such records (their non-routine quaHty) of imaging with other office capabilities. The re and the range of actions proposed. are fresh and port is loaded with pretty graphs but the sources worth further study beyond the state boundary. of the data are usually AIlM itself, so one is for ced to wonder whether the purpose of the study Vogt. Diane: Smithsonian Archives Photo Survey is to legitimate Information and Image Manage Project: A Draft Photograohic Thesaurus, 3/87, ment, and if so, for whom (given that AIIM wants 117pp. is the bi-product of three years effort in $~'n a copyl) Do you believe that in 1986 the cataloging over 6,400,000 photographs in 1.500 U.S. produced 1.175 Billion pages of paper (or Smithsonian collections for The finders Guide to film equivalents) and that 34% of this was on Photographic Collections at the Smithsonian COM?I or that scanning and OCR costs will fall Institution a planned five volume work to be 20% a year for the next decade?1 published beginning in 1988. It is likely to be more important to photographic archives and re Library of Congress. Prints & Photographs positories with photographic collections. than Division, Descriptive Terms for Graohic Mater even the impressive volumes it was constructed ials: Genre and Physical Characteristics Head to index. The draft, billed as incomplete and pro ings, compiled and edited by Helena Zinkham mised soon in an edited and widely available ver and Elisabeth Betl Parker (available from the sion. is, in its present state, the most complete. Cataloging Distribution Service) Washington DC, authoritative and usefully organized reference 1986. is the long awaited list (fami liarly known work for photographic cataloging which exists. as gmgpc to your subfield 2 of MARC 655 and Its importance lies in its definitions of photo 755 which have been hankering after it). It is graphic proceses, formats and techniques, where everything we waited for and more. Not only the discussions are extended and provide clear does it contain most of the terms we could want, criteria by which to distinguish different types. it is and excellent thesaurus with helpful scope The draft contains a few modest errors and occas notes for the pUblic and catalogers. The intro sionally differs in use of preferred terms for duction contains good clear advice and reason document types from the LC DescriPtive Terms able suggestions for level of specificity and ex for Graphics Materials and the AAT. HopefulJy, haustivity in indexing. Perhaps the most import these differences can be worked out in final ant information in this book is the address to draft making this a truly definitive work. Summer. 1987 Copyright by Archives & Museum Informatics 17 Walch. Victoria Irons. Information Resources for Traveling Exhibition Information Service Archiyists and Records Administrators; A Newsletter. published bi-monthly by The ReDort and Recommendations. Albany. National Humanities Exchange Inc. (P.O.Box 1608. Largo Association of Government Archives and Records FL. 34294) is more a classified listing service Administrators. 1987, 42pp. IAvailable, free, than a journal, but its lists of 50 or more from NAGARA, NY State Archives, Room lOA75, available exhibitions from its members is very Cultural Education Center. Albany, NY 122301 useful. For tbe past two years, NAGARA bas been study ing bow best to meet the information needs of ar ARTICLES It Bool:S chivists and records administrators; this report presents three options ranging from status quo to full-service information center. It advocates Blake, Monica; "Aspects of Electronic Archives", the middle course of establishing a modest clear Electronic publishing Revin" vol.6 '3. 1986. inghouse function within the NARA library. Al p.151-158, is a report of a study by the British though the report is well written and its conclu National Bibliography Research Fund on estab sions well supported. its very conservative re lishing a national archive of electronic publica commendations are somewhat anticlimatic. tions similar to the National Sound Archive and the National Film Archive in concept and form. The paper reports on the extent of electronic NEWSLETTERS: publication in the UK by format (videodisc, CD ROM. on-line databases. videotext) and discusses Access ReDorts: Freedom of Information some archival implications of dynamic change in Newsletter (lSSN 0364-7625) is published bi datalimagebases which are only available elec tronically. The problem of archiving Prestel is weekly by the Washington Monitor Inc., 130 I Pennsylvania Ave. NW, Suite 1000, Washington discussed. The paper does not propose actions. DC 20004 $250 p.yr. has been carrying a series of reports on computerized government databases Cloud. Patricia, "RLIN, AMC and Retrospective and privacy issues and closely following Conversion: A Case Study", Mid-Western legislative hearings on "computer matching". Archivist, v. 11(2) 1986, p.125-134, is the first real discussion of the costs of doing a retrospec Charles Babbage Institute Newsletter (Center for tive conversion in RLiN for MARC AMC. It is ex the History of Computing, 103 Walter Library. ceptionally useful for that reason as a planning 117 Pleasant St. SE, Minneapolis, MN 55455) framework for others. The results, simply, are reports on a wide variety of activities having to that it took Northwestern 2.7 hrs per record with one full time project coordinator, one part time do with the history of information processing, most recently on the formation of a "National archivist and a part time student assistant. Con Archives for the History of Computing" in the siderable time was devoted to authority checks U.K. (by John Pinkerton at Manchester Univer and reviewing the records themselves; in addi sity, Dept. of History). tion it took 41 minutes to code and 22 minutes to enter a record. Library Conservation News (ISSN 0265-04 I Xl is a free quarterly publication of the Preservation Cook, Michael, "An Introduction to Archival Automation: A RAMP Study with Guidelines", Service of tbe British Library, Great Russell St.• London WCIB 3DG. I found October 1986, which UNESCO General Information Program, September 198649pp. reported on the digitization of sound recordings at the British National Sound Archive, partiCUlar As tbe title implies, a primer, complete with a ly intriguing, but it is always full of useful basic bibliography. information and is very international in scope. Johnasson, Stig; "Machine Readable Texts in Library High Tech News (published II times p.a. English Language Research" . Humanistiske Data. by Pierian Press, P.O.Box 1808, Ann Arbor, MI #3-86. p.27-34, reports on the status of the In 48106, $65 p.a.) continues to be the best source ternational Computer Archive of Modern English of basic bibliographic references for archives (lCAME) and the authors' use of the Laocaster and museum automation, even though it includes Oslo/Bergen Corpus which, together with the Brown University Corpus, the London-Lund no archives or museum periodicals lltt.,R. The Corpus and several minor corpa is distributed bibliography is printed in a machine readable o t i . h's e e. by ICAME. 18 Archival Informatics Newsletter voJ.1. #2 Logan, Robert K.; The Alphabet Effect: The Welsh, Peter H. and Steven A LeBlanc. ·Computer Impact Q( the phQnetic Alphabet Qn the Literacy and CQllectiQns Management". Museum DevelQoment o( Western CiVilizatiQn, New York, N.mi, VQl.65 '5. june 1987 p.43-5 L is a solid. William MQrrQw & CQ., 1986. It prQbably takes general introductiQn. tQ the kinds o( questions having cQ-authQred with Marshall McLuhan tQ which should be asked in a computerization write that: "It is the transitiQn from the impres prQject. The authors dQ not rigorQusly address sed tablets with a vocabulary Qf two hundred the requirements Q( a cQllections management token signs tQ the incised tablets with a proli(er system ~ nQr, I'm afraid. do they fUlly de atiQn Qf pictographic signs, created with the use clare their interest in Questor Systems and its of a stylus, that marks the advent Qf writing. man software ARGUS which is their exemplar through kind's greatest breakthrough in data process out. They insist upQn the CQmmon (& dUbiQus) ing." Other observations are equally (ascinating view of museums that it is essential .. to enter 11/1 and equally facile. adding up and infuriating in rt!CortIs cO.tJ1p/ele/y IIOd lIS theystllOtI' (their the manner of Michel Foucault. emphasis) rather than build the collections management database through use. because they Marx,Peter; ·State Public Records: Database equate collections management (unctionality Goldmine Qr Landmine", In(ormation Times. with information retrieval and are reallydiscuss April 1987 p.21 ,28, explores the prospects (or ing cataloging/registration systems rather than commercial distributiQn o( state databases in collectiQns management. They seem overly im cluding legislative and regulatory data, conclud pressed by the impQrtance of operating systems ing that efforts to date have been hampered by for end-users (and with their selection of a PICK position taken by the states with respect to com operating system based application). but their mercial use o( public data. Urges the develop observation on the importance of system tools ment o( "uniform and reasonable" standards (or certainly holds true for developers and Jf multi access to state data as an impetus (or investment user. multi-tasking functionality is not consider by private firms in its dissemination. ed a concern which would be voiced at the appli cation level. their discussion of it is well taken. Nath. sandra; Research Study (or a National Documentation Organization in the United States. May 1987, MS. Thesis. Museum Studies, john F. EPHEMBRA Kennedy University. 172pp. It is a coincidence that the 50th anniversary year o( the American Kerr &. Downs Research; ARMA International Documentation Association (now ASIS) is the Membershio Survey. 1987. 22pp. is based on a tenth anniversary o( the Museum Documentation 41 %response of members polled. ARMA is per Association, the British group which inspired ceived tQ be meeting the needs of members. What this thesis, but the fact (which went unnoticed might interest archivists and museum stalf is by the author) might have contributed to this that well over 80% of ARMA dues are paid entire study which examines the MOA and the Canadian ly by employers with exceptions being almost all Heritage In(ormation Network and asks. in effect, retirees, consultants (who are also their own em how we can make it happen here. The author con ployers) and government workers. When a trade cludes that what is required in the US is an certifies practitioners. and employers hire the organization which maintains a database. data certified. the employers in effect laurantee the standards. consultative services and training continued membership of the association. programs. as CHIN and the MDA do in their respective (but quite different) ways. One might Derwent Guide to Patents , free from: Derwent. also conclude that the (ailure o( that model in Inc. (USA). 6845 Elm St.• Suite 500. McLean. VA the U.S. in the past bodes iJJ for it in the (uture. 2210 I. is a usefuJ Introduction to the patent process and to patent records and their contents. Perry, Meg Woollen, •An Inside look at a LAN Data Archive System"•~ july 1987, p.l 69 0&0, Directors'&. Officers' Liabiltvj A Crisis 176. reports on a home grown LAN archiving (not in the Making, A study of Museum Directors. backup. but true archiving) system in use at Peat Marwick. 1987. discusses the changing re Brecton Dickinson Research Center in Research sponsibilities Qf directors of non-prQfits and re Triangle. NC. Useful (low charts Hlustrate user PQrts on a museum survey which documents the and system decisions in a real time environment extent to which they have not adjusted to the new with an archiving option. A wQrking model o( context o( legal liability. Applicable tQ all kinds how data archiving can work in an OA system. of cultural institutions this study complements Summer. 1987 Copyrightby Archives & Museum Informatics 19 other Peat Marwick studies of the liability of SOFTWARE University Presidents & Boards. Government officials, and others. Available free from Peat Mason Barnett (Assistant University Archivist.• Marwick. try your local office. Duke) reports that after considering MARCON. Micro-MARC AMC. and In-Magic. the Duke Univ We Are Losing Our Past is a popular pamphlet ersity Archives selected Revelations by Cosmos calling attention to the Preservation Needs of and is in the process of writing the routines to State Archives published by the National Assoc create MARC AMC leader/header and record AMC iation of Government Archives & Records Admin descriptive data within that package. istrators and available free from the Council of State Governments. Iron Works Pike. P.O.Box Caesar lacovone (Director. Div. of Archives & 11910, Lexington. KY 40578 Records Management. NJ) reports that t.he NJ micrographics accounting system is propriet As part of an NHPRC grant which enabled the ary. but for the benefit of those considering simi transfer of the archives of the National Federa lar applications. he describes it as consisting of: tion of Abstracting and Information Services. to "System Code Tables - for identifying employ Temple University Archives. NFAIS publlshed!!... ees. labor houf'S. supplies. fixed assets. and hard Model for Donor Organization and Institutional copy/microfiJm location; codes are used to drive Repository Relationships in the Transfer of data entry and reporting on all system modules. Organizational Archives. by Miriam Crawford. Job Costing - individual and aggregate cost NFAIS, Philadelphia. 1987. 2.5pp. INFAIS, I 12 S. totals for seven-phase production process 16th St.. Philadelphia. PA 191021. Job Specification - complete service specifica tions for microfilm jobs UNESCO's (COM Documentation Center (Maison Productivity Tracking - based upon entry of de L'UNESCO. I Rue MioJlis. 7.5732 Paris) will be daily labor hours and supply usage; used for em glad to send, upon request. the fourth edition of ployee productivity evaluation. budgeting. sup its Basic Museum Bibliography (J 986 ). ply inventory depletion. and billing Work Flow Tracking - reporting modules for The NARA Archival Research and Evaluation monitoring work-in progress and turnaround Staff has circulated Prospectus for Access by performance Function and Process for the National Archives Bill Out - itemi2ed billing based upon input of and Records Administration. 23pp plus numer productivity data ous appendexes. within the National Archives. Accounts Receivable - driven by billing; inclu The document discusses the potential use of a des client budget status and open amounts (aged) functions/processes vocabulary to support user Accounts payable - records payments due to access to materials across provenance in line vendors; tracked by vendor. account. and minor with the proposals being discussed in the RLG object of expenditure Seven States Project and often advanced by this Inventory - status tracking of supplies; inclu editor. des reorder report and current inventory level /value report Financial Executives Research Foundation has Fixed assets - records information about equip pubHshed "EDGAR: The SEC's Pilot Program and ment and fixtures; includes depreciation track Its Impact". an introduction to the electronic ing and book value reporting filing project and a discussion of its potential. General ledger - general purpose application ($8. from FERF. 10 Madison Ave., P.O.Box 1938. for recording and reporting on all fiscal & finan Morristown. NJ 07960) cial transactions; includes various financial reports - e.g. balance and income statements" Technical Data Publishing Company (91 North Bertrand Rd .. P.O.Box 4,58. Mt. Arlington NJ 07856) is distributing a free electronic publish Peter Sigmund (Director, Rijks Archiefschool. ing glossary including the terminology used by the Netherlands) sent me a description of MAIS. printers, typesetters. writers. editors. and terms a new Oracle based. PC system developed by the from data communications and computing., Dutch Ministry of Interior. "MAIS stands for 'Micro Archief Inventarisatie Systeem'. MATS is a fully menu-driven program designed to support the description of archives and the production of inventories and guides. 20 Archival Informatics Newsletter vol.t. #2 Functions of the Svstem: (such as carry forward of field values and vali Data Entry: The screen lay-out contains fields dation by MDA vocabulary and syntax) and that cover all relevant elements for arranging prints 3x5 cards. MDA claims that the input and describing archives. such as 'form of mater and output formats can also be tailored. It sells ial'. contents. dates. etc. Value tables speed up the system. without updates or support. for I 15 the input and control the uniformity of the de pounds sterling. One year of software and docu scriptions. In the contents text field. words can mentation updates cost an additional 33 pounds. be marked for indexes of persons or subjects. Ex tra descriptors can be added. There are special Copies of the ADAPSO/EDUCOM brochure functions for linking series descriptions or split "Using Software: A Guide to the Ethical and Legal ting series descriptions into sub-records. The Use of Software for Members of the Academic system keeps track of correct numbering. Community" which includes the text of the Arrangement: Once entered. descriptions may statement on software and inteHectuai rights. be automatically arranged for a guide according are available free. from ADAPSO. 1300 North to classification codes assigned by the archivist. 17th St.• Suite 300. Arlington. VA 22209. The descriptions are sorted by classiffcation. according to criteria chosen by the archiVist. Paradigm Press announces that DiscSCOPE. a The system sorts descriptions in chronological version of SCOPE: Humanities Computing Update. order within the arrangement selected. although which includes selected items from Software, one can choose other fields for a secondary sort Courseware and Software Reviews sections of as well. The system provides several methods recent issues. "has been placed on a floppy disc for assigning classification codes and designing with a sophisticated menu program to provide a classification scheme (such as record groups. access. Tbe disc may be freely copied for circu classes, series. subseries). For instance. it is lation amoung colleagues and uploaded to bulle possible to store classification codes with their tin boards. Updates are scheduled every four definitions in separate tables in advance or dur months. Copies of discSCOPE can be purchased ing the description process. A special function for $5. Organizers of conferences and user ser enables the archivist to view the distribution of vice personnel at computer centers who wish to descriptions by class, or to get a synposis or the distribute copies (with their own labels if the documents that are not yet classified. desire) may request free masters on their letter Reports: MAIS produces six standard reports: heads." Prepaid orders or requests for free mas I. A preliminary inventory (including both tbe ters should be addressed to DiscSCOPE. P.O.Box preliminary and final numbering) 1057. Osprey. FL 33559. 2. A finished inventorty ( in which only the final numbering is printed) 3. Indexes by SUbjects with keywords from the contents field. as well as attached keywords IN CONGRESS 4. Indexes on names (with the same options as 3) Proposals to ban. Digital Audio Tape re 5. Concordances corders without copy protection systems 6. Lists of annotations made up by the archivists are still being pressed by the music indus while arranging and describing can be printed try and optical publishers. Ifthey succeed, with reference to tbe number of the description. digital audio tape (Whats DAT?) will not do Technica~ specifications: to CD's what videotape did to home video MAIS runs on microcomputers with MS DOS discs and what audiotape threatened to do version 2.1 or later on the relational database to phon08raphic records before CD's got management system Oracle which requires at there, and archives & museums will lose a least 5 I 2KB memory and a 10MB hard disk. A potentiaUy valuable technology. ( see facility is provided to edit MAIS reports in Nancy Herther. "Much ado about DAT", Wordstar or Wordperfect. Database v.10 #3, June 1987 p.,116-120) The Museum Documentation Association At this writing, Congress had over (Building O. 347 Cherry Hinton Rd .• Cambridge whelmin8ly passed legislation to remove CBI-4DH. ENGLAND) has released its Museum authority for control of non-classified but Object Data Entry System (MODES) which sup "sensitive" information from the military ports entry of the standard MDA data records on effectively rescinding President Reagan's an IBM PC or a system running CP/M 80 version notorious executive order of November. 3.0. The system contains data entry features 1986, although it remains to be signed. Summer. 1987 Copyrigbt by Archives & Museum Informatics 21 Capturing Rich ConteIt the air. If I want to document what the file actu by David Bearman ally looked like (what data it contained) when I made an important decision based on it, I need not just a snapshot of its contents at some later For the put several months I've been preoccup date. and not even that snapshot plus a complete ied with two problems which arise from the goal audit trail (with which to back in and out all the of attempting to document culture accurately and changes), but also an actual record of what I look richly. The first issue is how to identify and ac ed for and how the software processed it. Pro quire evidence of processes, not just of products. cessing rules are increasingly replacing proce The second is how to represent processes and dure manuals and regulatory interpretations as relationships, not just entities. As usually automated systems take on the tasks of deciding happens when we get preoccupied. I would Iike who is eligible for what treatment. lirve don't you to take these concerns as seriously as I do. document thesoitrvare itself, Jle cannot recon We have long relied on tacit knowledge of how structtheprocess. things are done (by humans, in particular types Perhaps some radical examples can help. To of social situations which we can envisage document electronic music, or computer art, do ourselves in) as a surrogate for documenting how we want the compositions or the programs or things are in fact done. This luxury is receding both? If the bridge collapses, and I don't have in a number of areas as a result of automation. both the stress mode II ing algorithms which oper For example. since the advent of bureaucracies ated in the system in which the bridge was desig and postal systems, archivists have assumed ned and the assumptions which the designers with reason that if a document is in my ri ling actually fed into the system, where will I assign system, I received it shortly after the date it was the blame for the collapse? Would the calcula written and referred to it in connection with the tions (the record) help? This last example is topic under which it is filed. Based on such in typical of the kinds of issues which are arising ferences from the evidentiary remains of the in the evolving case law of software liability; office of a policy maker, we can make reasonable issues which are familiar to archivists concern guesses about how a policy decision was formed. ed with documented accountability. In the past, In fact, the validity of the inference began to an engineer who used a faulty method of deter break down with the widespread use of tbe mining stress would clearJy have been culpable. photocopier, because many records in my files as would a government official who mistook the arrived at times long past the dates on them and intent of a regulation and misapplied it. It is not may never have been referred to by anyone, It clear today, and what is more, we cannot know is clear that in electronic office environments without documenting aspects of process we have our assumptions completely break down, either previously been able to live without. because the document isn't flied with me at all Some of these problems inspired the work I or because my query for information recovers recently completed for the Computer Museum in numerous documents which I subsequently do Boston and the Technical Report on Collecting not examine. The computer attached printer, Software which grew out of that study. especially laser printer. and CRT viewed docu My second set of challenges arises from a need ments. have merely exacerbated the problem by to represent the contents of our cultural collec eliminating any clues we might have had about tions. be they paintings, shards, or documents. what constitutes the "original" (a problem which and is closely associated with the first problem. Lisa Weber addresses at length elsewhere in this sharinR a focus on the wholeness of social pro issue). The result is that if we want to know cesses. The inadequacy of keyword indexing, what went into a decision, we may need to have which leaves undefined the relationship between contemporary. explicit. process oriented doCU keywords, is well known. A variety of methods mentation. This kind of analysis is also the of avoiding false associations have been adopted foundation for the efforts described by John in many systems. inclUding keywords in context McDonald. and keywords Qualified (as they are in most two This problem is more serious with electronic level back of book indexes). The computer has databases. In its last issue, the Archival Infor introduced an easy variant on qualified key matics Newsletter published a technical leaflet words - keywords in rotatable phrases - which on how to transfer records from a DBMS to a soft permits one keyterm to Qualify another and each ware independent format for retention. This sol to appear first in the proper place in a printed ves some physical problems, but it leaves the fun index. Finally. there is increasing pressure to damental question of what we are documenting in adopt natural language searching, with proxim- 22 Archival Informatics Newsletter vo1.t, #2 ity of terms serving in lieu of constructed head ed index. The search issue becomes how to ident ings. ify "frames". or social situations. which are im Unfortunately, none of these methods repre plied by various semantic models and allow for sents relationships between terms very well. searching by frame matching. That is to say. how This problem is one which increasingly attracts to recognize that People (Samuel Adams) with artificial intelligence researchers who can now Social Roles (Citizen) participate in Events (t.he take advantage of the substantial knowledge of American Revolution) Which occured in Time semantic relations built by linguists over the (date ranges) and were perceived by persons past two decades. And it is engaging me, in some (George Ill) with Social Roles (King of England) work Jam currently doing with the National as Concepts (Uprisings) etc. in such a way that if Security Archive (a non-partisan organization I have an item associated with a frame called which acquires declassified documentation of uprisings I will be able to provide values to fill American foreign and security policy for the implied process roles of instigators (Person analysis and dissemination). The challenge is to participating in Event of type uprising). of use insights from semantic analysis in a practic public authorities (Person or Corporate Name al indexing system which must drive printed against which Event of type uprising is direc indexes but which is able to retain as much of ted). of consequences (Event or Concept resulting the context of term use as possible both in print from occurence or anticipation of Event of type products and when used as a part of an on-line uprising). etc. so that if my query invokes the searching system. Indeed. the ultimate aim is to uprising frame. and my document description move beyond the limitations of natural language. invoked the uprising frame. I will retrieve the full-text. based searches by employing context document even if I did not request any of the defined "frames" particular to the foreign facets specifically indexedl This is similar to affairs/security knowledge domain and derived the extension which takes place with thesauri semi-automatically from the indexing system. when I use a different term (not preferred or All this sounds heady. but it is really quite broader) for the same concept. and can retrieve simple. Imagine that we write a sentence (or the correct document. except that instead of multiple sentence abstract) describing an object operating simply on the term level, we can now or document. The sentences are comprised of move to the level or the meaning of the document words which belong to specific potential vocabu description and the query as a whole. lary lists. Such a set of linked vocabularies. for As we struggle with designing document and example. are being constructed by the Art & Ar object surrogates in our information retrieval chitecture Thesaurus which has 27 hierarchies systems. we should consider how to better cap of terms. each controlling a single "facet" of de ture the relationships between descriptive terms scription. The relationship between facets in a in order to support researchers with a variety of description within anyone domain of discourse perspectives and more subtle research problems. is limited. In art criticism. we can not logically (Instigators need to be searched without getting talk about Rococo smelting or about terra cotta victims and objects of revolutionsl) I should note architects. We COUld. logically, talk of materials that there is some impressive applied research and processes. even if feather smelting is not an going on in these areas. For a taste of it. 1recom actual combination. because the definition of mend a May 1987 article by Suzanne Humphrey processes is that they act on materials. What and Nancy Miller of the National Library of this means is that we can construct vocabulary Medicine in the Journal of ASIS, and a July 1987 controlled descriptions with complex relation article by Natasha Vieduts-Stokolov of 810SIS in ships and qualifiers. Can we derive printed the same journal. I am personally indebted to indexes from these and can we search them Pat Molholt whose insights in a recent unpubl ish automatically in such a way as to retrieve on ed paper on the "Development of term relation relationships as well as on entities? ships for the enhancement of semantic networks At the NSA we are launching a project to do and hierarchically structured thesauri" convinc both. as far as possible. with off-the-shelf small ed me that practical application of research system software and document analysts whose findings in this area was well within our grasp training is in the subject area. not in indexing. and that thesauri such as those developed by the The indexing issue becomes how to semi-automa AAT can best be used as vehicles to this richer tically parse a complex sentence whose semantic context and relationship capturing approach to structural variants are known. in order to gener document and object description. Indeed. Pat ate index phrases consisting of 3 or at most four graphically represents the AAT as authority lead terms which can then be rotated in a print- control over each term in a complex description. Summer. 1987 Copyright by Archives & Museum Informatics 23 TECHNICAL REPORT ON COLLECTING SOPTWARE For forty years software has been an important creative product of our soc iety. Its intel1ectual, social, economic and political impact has shaped the con temporary world and lent its name to an age. yet the community of culture preserving institutions has failed to document the evolution of software. Not a single archive or museum devoted to software exists. No substantial collecting of software history has taken place. Yet software is being written every day which defines the way in which we work. Bureaucracies (including govern ments) are entirely dependent upon software to faithfully execute the policies (including laws and regulations) which they have established, yet archives, those guardians of bureaucratic accountability, don't retain software. Popular culture and the arts have both been transformed by software, yet museums have yet to collect it. To archives and museums, software is still alien and insubstantial. This report examines the history of software and its influences on our society and addresses the barriers to collecting software as a cultural record. It identifies essential policy distinctions which administrators will need to consider between software collections and other collections of archives and museums. It examines the ways in which software can best be described, made available to researchers, and exhibited and it proposes a framework for a de scriptive vocabulary, And it identifies the physical requirements and manage ment issues associated with the retention and storage, retrieval and use, of soft ware in cultural repositories. An earlier draft of this report was prepared for the Computer Museum in Boston as the framework for a discussion with staff of the Smithsonian Institution and the Charles Babbage Institute on establishing a national software collecting consortium. It became clear in the course of that project that no single institution, or even group of institutions as prestigious and well situated as the sponsors were, could expect to collect the entire corpus of software related materials, to support research on technical. social, financial and cultural impacts of computer programming. The report therefore consid ers approaches to multi-institutional collecting issues such as collections policy, cooperative acquisition and information sharing. Most importantly, it provides the concrete guidance needed by every cultural repository, for all archives and museums which document any part of modern culture should be considering acquisition of some software as part of their collections, since software is an integral part of that culture. AvailableinAugust. Archives & Museum Informatics. 5600Northumberland St.. Pittsburgh. PA 15217 $45 pre-paid orby subscription to ArchivallnformaticsNewsleuerandTechnicalRepOlt Tosubscribe. send maili..De and billing address with subscriptionrequestto: Archives & Museum Informattcs, 5600 Northumberland St.• PlttsburgJi. PA 15217 ARCRJVALINFORMATICSHEWSI,6rtE1t&Tf.CRNlCALREPOIn'S Part 1 only: ARCHIVAL INFORMATIa NBWsunwfBlt US$24 prepaid, US$30foreign address airmail. US$40 billed .Parts 1 & 2: ARCHIVAL INFORMATIQ; NBWSLHITER It.ARCHIVAL INFORMATICS TECHNICAL RBJORTS US$160 billed or prepaid; US$180 foreign address airmail