ARCH IVAL INFORMATICSNEWSLETTER

Part 1 of Archival Informatics

ISSN 0892-2179 SPRING, 1988 Volume 2, ~ 1

No matter how efficient information scientistsmake their full-text searches. methods to reducethe number of documents thot must be seorched,and the parts of those documents that require"reading", will play on increosingly importantrole in full-text retrieval systems. Archivistshave long employed inferential logic based onthe1r unoorstanding of the Wf!tl in which recordsare created and the structural means by whichdocuments carry their messages, to find recordswith contents relevant to a user query. Nowinformation retrieval specialists are looking forjust the methods long employed by archivists.

Archives and museums have direct contri­butions to moke to the design of im~ and full­text systems if they can formal1ze theirknowledge of image and dooument access methods.This is aworthy challenge, I think.

EXCITING HERESIES

The RIAO-88 meeting provided an occasion toobserve thot archivists and curators hoveknowledges and skills desparately needed bydesigners of image and full-text retrievalsystems. Two particularly exciting heresiesoccured to me. which I had an opportunity toexplore with information retrieval experts atthat meeting (reported further on p.6).

As image retrieval systems become morewidely available, they are being examined byusers who are no longer just "wowed" by theappearance of an image on ascreen. These usersare asking how images can be indexed. usinglangu~ and symbols, so that the desired imagecan be obtai ned. And they are demandi ng systemcapabilities to manipulate and analyse the imagesonce they are found.

Curators have a knowledge, recently formalizedin such products as ICONCLASS and the Art &Architecture Thesaurus, of the differencesbetween SUbject and object indexing and theyhave experience I based on years of work with arthistorians, archeologists, natural historians andengineers, of the tools required to~ an image.

Equally exciting is the realization by full-textoocument retrIeval systems desIgners that theyneed to limit searches for texts toport10ns of thefull archive of office documentation, and that themethods for limiting searches involve exploitingwhat they only vaguely understand about thesources of documents, the systems out of whichthey are generated, and the genre of the record.These concepts of provenance, ser1es and form­of-materIal or oocument type are the stacie intrade of archivists.

TABLE OF CONTENTS

Articles:Advanced Revelation: AReview

J. Penny SmallMachine Reed6b1e Views

Thomas E. Brown

Regular Features:Conferences

ARLIS/NARIAO-88calendar. 2nd & 3rd Quarter

In-BoxLetters to the EditorNewsProjects & ProposalsSoftwareStandardsTechnical Report Summary

2

5

781011151819212224

ADVANCED REVELATIONAREVIEW

by Jocelyn Penny Small, Director. U.S.center.lexicon Icone.Jraphicum Mythole.Jiae CI6SSicae,Rutgers University, College Avenue campus. NewBrunswick NJ 08903 (201) 932-7404

" It was the best of pre.Jrams, it was the worst ofpre.Jrams, it was the age of transparent winOOws.it was the age of opaque documentation... ·inshort ....some of [theJ noisiest authoritiesinsisted on it bei ng received for~ or for evi1.in the super lative degree of comparison only."Advanced Revelation (AREV) provokes extremereactions, and, frequently, simultaneously. The"power" of the pre.Jram is extraordinary;figuring out how to tap that power is alsoextraordinary.

The US liMe Project & RevelationI shall illustrate its features by describing

their implementation in the Computer-Index ofClassical Iconography at the U.S. Center of theLexicon lconographicum Mythologiae Classicae(US L1MC).1 The US L1MC ,as part of aninternational project to publish a multi-volumepictorial dictionary of classical mythology, isresponsible for classical objects (ca.800 B.C­A.DAOO). with mythole.Jical representations. inAmerican collections. Unlike most computerizedprojects, its cataloguing goes below the level ofthe title for each scene to record the individualfigures and elements (animals. plants,archftecture. etc.). their types. and what thefigures are wearing or holding (attributes).

From the outset it was obvious that a relationaldatabase program with fully searchable, variablelength I and repeating fields was necessary.Objects currently range from single figures

1The US LIMC is part of the library at RutgersUniver5ity. ond is very pleosed to ocknowledgenot only the support of the University, but alsothat of the National EnOOwment for theHumanities. Research Tools Div. and the Davidand lucille Packard Foundation. The US L1MCwelcomes inquiries.

(statues and gems) to Roman sarcophagi withnearly forty figures. For microcomputers thesecapabilities appeared only in Revelation (Rev)by Cosmos. now known as RevelationTechnolOJies Inc.. Rev also came with a fullpre.Jramming language.

While Iat times keenly felt the absence of thebookcase of manuals-to-the-manual availablefor dBase and RBase. I was very satisfied withthe results. So why have I switched to AREV?Why have I tortured myself with near paralytictrances over the keyboard? (They changed Allthe keyboard commands in AREV , and prOVided notable of equivalents to Rev.) Three thingsspurred me on: windows. being able to use lowercase at the system prompt (TCl for TerminalControl leve]). and. of course, the compulsiveneed to upgrade. The second release shouldappear this April.

WindowsWindows! Wh~t ~ wondorous tool they are. And

how marvelously they are implemented in AREV.First, some background on my use of them. TheUS liMe has two core files. Objects which havescenes, and nearly thirty satellite files2 • which:( 1) control the words used. and (2) check theirspelling. (3) classify the words (Carnelian is akind of Chalcedony which is aQuart used forGemstones, which are obviously Stone; Myron isGreek and a ScuIptor), and(4) index data from the core files for fastretrieval and relating files.

Whether or not a particular satelltte We doesmore than one task depends on its data.Furthermore. certain statellite files are linkedto other satellite files. Thus Biblie.Jraphy checksall bibliographic references no matter wherethey appear, while Cultures vets the entries forCulture in both the Objects and the Artists file.

The verification of fields in both Rev and AREV1s very simple: put the name of the ver1f1cat10nfile in the input pattern for the prompt for afield. and the prOJram automatically checks the

2 Satellite describes the relative position of thefile to the Core files. "Classification", however.is preferred to "authority" to describe thesefiles, since they are more complex in what theyrecord and what they do.

2 Archival Informatics Newsletter vo1.2.-1

entry against the record identifier of theverification file. In Rev, by indexing(inverting) the data from fields in Objects andScenes into Sl:ltellite files, I could retrieve 011objects made of stone or all scenses from theTrojan cycle, but the process could becumbersome. Now the program "knows" andkeeps track of the linkages.

AS88rchFor a real example, anarcheologist (from

Brown University I naturally) I wanted all vasesin American collections decorated by theProvidence Painter. We started in the Artists'file by first accessing the cross-referencedindex to the individuol word3 in the recordidentifiers (Artist Name) to find all painterswith Providence in their name. From the 11st offour painters we chose the desired ProvidencePainter. Merely by pushing Alt-F5 we switchedto the Object me with all of the objects by theProvidence Painter saved in a list, available forbrowsing backwards and forwards. When aparticular object was of interest, Alt-F6brought up the Scenes on the object, with Alt-F8taking us to further information about the Titlein the Titles file and another Alt-F6 for a "secondcopy" of Scenes with only that Title and then Alt­F5 for a second copy of Objects for more detailedinformation about a particular object. Hypertextis here.

This design speeds corrections for fields wherefull verification files do not yet exist. I invertthe data 1n Objects and scenes into theirrespective classification files, select for allrecords with no record date (all records enteredby human hands are automatically date-stamped)to check the data, call that selected list fromwithin the verification me winOOw so that Icanswitch to 011 the records in Objects and Sceneswith that particular entry with a mere touch ofthe function keys. If you want to see the selectedgroup of records in columnar form, withControl-F5 you ~ to Table Mode, although toomany records with too many prompts w1ll blowthis function.

Two other aspects of w1noows are important.The winoows can be larger than the CRT screen,up to 32K or 180 columns by 180 rows or anycombination thereof. No longer are five

separate, but linked, screens necessary for thetwo core files. One window, or rather template1n AREV terms, suffices for each file, All,nonetheless, is not sweetness and light. Theimplementation of "paging" is jerky, annoy'ing,and absurd in this set up,

Pop-ups & optionsPop-ups are winOOws that appear generally

when you press F2 at pre-defined fields. Whilethey can perform a number of tasks,l use themto prov1de llsts of record IDs from smallclassification files. Techniques and Materials,with a little over a hundred entries, are feasible,but Artists, with nearly two thousand, is easierto consult vi'J the reloted templates. Thus F2 inthe Techniques field brings up the list of allpossible techniques, and all the data enterer hasto 00 is press the enter key at the appropriatevalue. Since this field is repeating (multi­valued), more than one technique can be chosenat one time with F9 saving the selections andautomatically entering them into the field.

Pop-ups are also used extensively by AREV;HELP lists of your last ona hundred commands atthe TCl prompt, an ASCII chart, etc.. The systemalso allows you at any poinno ~to TCl, andthence to DOS. It always remembers where youare, and moves recursively back to the startingpoint. I should a<t1 that l1ke Rev. the program hasnot lost any records in my five months of use; 11has, however, sometimes "hidden" them from,me.

Wh1Je the types of indeXes tn AREV (cross­reference. relational, and b-tree) are morevaried than in Rev, allow for upper and lowercase distinctions, as well as stop- lists system­wide or 1ndividually tailored to particular fields,Ist111 use the Rev program, inyert.all, because Ican ed1t the f1elds contain1ng the inverts. AREVincessantly updates all indexes (in background)by keeping a record of the changes in an indexingtransaction file. Changes made at any level, be ittemplate, an R/Basic (as they call their flavor ofBasic) program, or separately invoked indeXing,will update the indexes.

Here, as throughout the program, the choicesfor w~ to accomplish a particular task aresometimes staggering. Seemingly simpledecisions can have large effects. For example, at

Spring. 1988 Copyright by Archives &; Museum Informatics 3

one point with Rev, Iwas reluctant to reduce thenumber of fields in Objects by increasing the useof the classification files. Because Rev did notautomatically keep track of the cob-web likeconnections between my thirty or so files, it wasdifficult to remember how to traverse them atthe tlme of searching. Now, with the smoothmovement between fi les (templates) I I have nosuch hesitatlons since actual use is much easier.I should odd thot the mointemmce of theconnections and the addition of new ones hasbecome a minor nightmare. Aplain, run-of­the-mill pop-up requires the classification file,its template, the pop-up description stored inthe pop-up file, and the call from the prompt forthe field in the template.

ShortcomingsUnlike Cicero, Icannot pass over the fact that a

number of bUgs are due to too early a release and8 striking lack of communication betweendifferent ~ms of progrmnmers or, worse yet,no attempt at the reconciliation of obviouslychanging methocJs.3 For instance, the routinethat imports Rev screen builds acall to a recordin the HELP f11e that ends with the same fieldname rather than the prompt number for thetemplate, as used for all new templates. Pettycomplaint perhaps, but the absence of all helpand the concurrent creation of four hundred someodd unreachab le entries in the HELP file were nojoy to discover and amend. Some routines, suchas "preserved fields" in the window design, arenot merely half-baked, but oownr1ght raw. Theintricacy of "paint", their program for designingentry and record templates. can be gauged by thefact that each prompt has forty-threeparameters which can (they don't have to be)filled in, and some of these, like patternmatching, allow for multiple entries. Patternmatching can be customized to do just about anypattern you can possibly dream up.

All of which brings usto OOoumentation! Six,count them six, manuals with the poorestindexing seen this side of the mainframe. Notonly does the overall Index leave out maIn

3 A Wise and willy book on software engineeringis: frederick P. Brooks Jr., The Mythical Man-montb, Boston, Addison-Wesley. 1982

references, it still has no entry for "back up".When Iasked why. Technical Support told methat since "back up" was not one of theircommands, there was obviously no need for anindex entry. With that approach, imagine theQuality of the documentation. With thecomplexity of the program added10 the sheerinnocence of what documentation is about,imagine scrambling through the four mainmanuals. With the changes between actualimplementation and the earlier dates for printingthe manuals, think of the results. But do noteven contemplate the fact that since Release E,with which Istarted, every single set ofinstructions for installation has had errors. Norshould you forget that help is not at the cornerbookstore or your local computer center. If theprogram is so goOO, and it is, why the poordocumentation?

But, let us return to what else AREV does do. Itcan be either menu (very easy to set up) orcommand driven or acombination of both. Sinceall fields are variable in length, on ly displaylengths are defined. In templates, the text eithertickertapes for single line displays, or scrolls.You never run out of space. It has 8 mechanismfor doing forms, reports, or- whatever you callthose things that are not columnar lists. Alas, itsprevious forms, whlle not splendiferous, arebetter than the current incarnation, which doesnot adjust dynamically for changing amounts ofdata, such as the 'brief' statue or the "long­winded" sarcophagus. You can, however, printany templote (screen) with 1:1 mere Alt-P. (TheRev program GFORM is on the compatibilitydisk). Anew and improved report program ispromised.

Ull1itiesAREV acknowledges that no one is perfect, and

has prOVided aset of powerful routines for globalupdates. An intermediate transaction file is builtso you can compare the original with the changedversion. One can also do global swapping ofstrings within a field as in word processing; but,of course, it took me an afternoon to figure outhow it worked. Moreover, much, like the pop­ups and the related windows. it is fully automatedso that no programming is needed. The scholar,

Archival Informatics Newsletter vo1.2.·.

if lik.e me, can twiddle to his heart's content untilhe gets it just right or righter.

I have not used their network. version, which issupposed to be excellent. Nor h~e I hll1 anyreason to implement any of their provisions forsecurity. You can define separate users with anykind of ideosyncraUc access you might want. Thesystem automatically resizes f1les for optimumretrieval -- a real plus-- but, at the same time.the data now resides in only two files that cangrow to enormous proportions. My scenes filesare each over 1800K. You must use a DOSprogram that can split fi las between disks. Theprogram now comes on ten diskettes -- some ofwhich you won't need forever (tutorial. Revcompatibility), but do not even kid yourself thatAREV can be used solely on afloppy system.

Whlle Ihave focussed on the new things thatAREV ooes, that Rev does not, it should be notedthat both programs use dictionaries to hold fielddefinitions. "Symbolic fields" are R/Basicprograms up to 32K. I use them for plucking thecity, museum. and inventory number from theObjects file into the scenes f1le. Such "joins" arenot real joins. That is, no transfer of data needoccur (it can, if you want iO, and no third file isbuilt (a great saving of time and space). Thesefields con also be used to do colcuhrtions andspecial formatting, like underlining.

The Bottom lineIf you are already usi ng Rev, shou1dyou

upgrade? Emphatically yes. But emphaticallysteel yourself for an excruciating experience. Ifyou are us'lng something else or not usinganything else, should you buy AREV? Absolutely.Buy AREV at the cheapest price (list $950) you~n find. Then invest your savings in the longdlstance telephone calls to Technical Supportand/or one of the independent consulting firms.(I do both). AREV really will do what you want itto do (all mortals are reasonab1e, you aremortal, therefore... ) Getting there, howeveris half the agony. •

J.c.p

MACHINE READABLE VIEWS

Thomas E. BrownNational Archives &. Records Administration

Washinaton. DC 20408

For the past several months, I have spent a lotof energy as chair of the program committee ofthe forthcoming conference of the InternationalAssociation for Social SCience Informationservice and Technology (I-ASSIST). As anyonewho has ever served on a program committee canattest,it is both time consuming and rewarding.The rewards stem primarily from theopportunity to see the cutting edge of one'sprofession. Because I-ASSIST brings togetherprofessionals interested in the acquisition,administration, use and preservation ofcomputer data files, the last three months haveoffered me insights into the evolving interests ofcolleagues concerned with the archivaladministration of electronic records.

The proposals and final program reflect thegrOWing role of microprocessors or personalcomputers in this, as in other spheres ofcontemporary society. One session is devotedexclusively to the analysis and dissemination ofdata using the microcomputer and severalsessions dealing with archival administration oruse of computerized data include at least onepaper on the applicabllity of microcomputers. Inone series of papers ~ressing the use ofinformation on Individuals in modeling for socialpolicy analysis, two authors propose tutorials onhow to use microcomputers to analyze microdatafor policy analysis.

The I-ASSIST prCYJram also reflects the growthof automated cartographic systems. Whi Ie~~aphic information has long been digttized,10 thlS decade such systems have been riding thetidal wave of the advances in computer graphics.Over the past few years. as developments havetaken place tn the CAD/CAM (Computer AssistedDesign/ Computer Assisted ManUfacturing)arena. the I-ASSIST program has routinelyincluded one or two papers on cartographicsystems. The upcoming conference will devotetwo full sessions to digitized cartographic

Spriol. 1988 Copyright by Archives & Museum Informatics

systemsl The first set of papers will focus onmanaging these systems. One paper will examinethe technical problems associated with theircreation and maintenance and the need forstandards: The second paper w111 discuss onesuch standard emerging within Canada, the Mapand Chart Digital Interchange Format (l'1ACDIF),a telecommunications standard for cart~raphic

information. The last paper outlines acurrentsystem within the U.S. Government, rind relmesthe previous observations to aspecific mappingsystem.

The second session on automated cartographicsystems was designed to be provocative. Eachpaper will discuss whether reference servicesshould provide data to the user in cart~raphic

form rather than (as is now customary) as rawdata or crude tabulations. Michal Peleg of theSocial Science Data Archive of HebrewUniversity in Jerusalem argued in her proposalthat: "clear and meaningful representation of dataIs rI universrll requirement from every scientist,student, and executive. Quantitative data in thesocial sciences is usually presented by tables ofrates, frequencies or percentages, whlle themore advanced data presentations involvestatistical charts. However, geographicallyoriented data needs more sophfsHcatedpresentation techniques. Otherwise the spatialcharacter of the data is largely obscured. Themap then provides the necessary medium forpresenting areal relationships of spatial data,which are so essential in comparative andregional studies. Until the eighties, most of thestatisHcal maps were manually produced by afew geographers and large agencies. Now, due tothe decreased cost of computers and graphicperipherals, as well as the availability ofsophisticated graphics software, this challenge iswithin the reach of public use in a large numberof app lications." As a result, she argues, datarepositories that have data sets with spatialcharacteristics such as census samples and smallarea statististics, household surveys,~ernment macro-data and election resultsprovide their users with informrltion incartographic form.

Peleg's proposal to use cart~raphicpresentation of archival information relates to afar broader question. Today I archives and data

libraries are generally providing their userswith raw data, primarily on magnetic tape, andwith the necessary documentation to interpretthat raw data. Certainly, providing the data onmagnetic tape really does not help the PC user.So what are the alternatives? What should be thestandard for reference service in the future?What level of service should our professionprOVide? These questions remain unanswered.The discussion of CrJrtogrrlphic output as tIreference vehicle is one solution to this broaderand increasingly complex problem which we havenot really addressed or crlequately defined.

The I-ASSIST program reflects a new emphasiswithin the profession on reference service. If wedivide the functions of data archives and datalibraries into three broad categories, acquisition(records scheduling and disposition, appraisal,solicitation policies and accessioning);administration (description, cataloguing,technical documentation standards. andpreservmion); and reference (disseminationtechnology. confidentiality. and end-user needs),conference proposals relating to acquisition werenearly non-existent. Proposals relating toactual administration of computer files within arepository continued to hold their own. Thegrowth area, to the point of eclipsing acquisition,was reference service. And this emphasis onreference seemingly concentrated on theresearch use of the information. What does thissay about our profession?

It may reflect the growing influence within 1­ASSIST of librarirlns who rlre realiZing theimportance of machine-readable records and whohave traditionally had a strong interest inreference service. Originally, I-ASSISTattracted archivists with responsibility forautomated records within a traditional archives,and on ly later, data librarians who usually foundthemselves adm'inistering acollection ofmachine-readable data files outside of traditionallibraries. As a result of the recent influx ofprofessionals working in traditional libraries,management issues may have been overshadowedby reference service. The profession continuesto expand to include new people with newemphases.

6 Archival Informatics Newsletter voI.2.-.

CONFERENCES

Art libraries Society/North America

The (ARLIS/NA) annual meeting in Dallas,February 4-11 , reflected the growing interestamong visual resources managers in automatedsystems, indexing, intellectual accessapproaches, and the MARC-VM format.

Helene Roberts, Curator of Visual COllections atHarvard, sounded what was to be a theme of themeeting in the first session when she criticisedexisting reference works that provided sUbjectaccess to visual collect1ons for not beingsufficiently complex, consistent or convenient.Her use and explication of ICONCLASS suggestedthat it is asystem that is sufficiently complex torepresent works of art, but she presented noassessment of its consistency nor any mechanismby which it might prove more convenient. JackRobertson also addressed the convanience of artreference data in his well thought out guicE1inesfor reviewers of reference books, but apresentation by Marshalllapidus on the mergerof RILA and the RAA which followed revealed thenature of these problems in all their concreteugliness. The attempt by the Getty Trust andCNRS to merge the two principal bibliographicresources in the Held is more than three years inthe making. and will represent at best acompromise.

Compromise was the subject of the next sessionIattenlEd on indexing and search strategy usingthe Art and Architecture Thesaurus (MT). AmyLucker presented the use of the MT in theAVIADOR project at Columbia University inwhich only single terms were assigned.simplifying the indexing but 1eoolng to falsedrops and sacrif1cing the hierarchical structureof the vocabulary. Bethany Mendenall examinedthe implementation of expressions. or sets of MTterms in specific systems, and MurrayWaddington presented the way in which the MThas agreed to be represented in MARC usingindexing strings which can be sets of expressions(see also Standards. p23). Inaspirited

discussion led off by commentaries from PatMo1holt and Jim Anderson, the audience debatedthe desirability of complex representation

, versus simpler, more easily implementedapproaches and the barriers to nationalimplementation of hierarchical facetted indexingvocabularies.

The meeting provided an opportunity for ARLISmembers to experience some systems used ordeveloped by their colleagues. Jeanne KeefeWatkinson presented the RPI online public accesscatalog to 65,000 slides. Joy Alexander andJoSchaeffer discussed the'lr experiences automatingslide collections with standard commercialpackages (DBase and SMART). The organizersprovided an afternoon to see other ARLlS/NAmember developed systems, inclUding a DBaseauction cata1a,) file by Beth Dinaff (DallasMuseum of Art), an index to the illustrations inarchitectural books by Henry Pisciotta (CMU),and a PC/Mac Videodisc driver from theUniversity of Iowa. The mooesty of theseimplementations was contrasted with extravagantclaims made by computer systems vendors atInfomart ("the world's largest computershowroom") where Ivisited CPT's full textretrieval, IBM's Infowindows ,and ImlJJB Setsshowing off their use of Pictureware on an ATTsystem, and was im pressed only by the hype.

One late night session was designed to captureh1gh-technologyimag1nat10ns. Karl M111er(U.T.Austin) led off with an enlightening historyof the CRT and the potential of electronic imaging.Madeline Nilsen conducted atour of projects thatcombine image and text ranging from Athena, theN1J Khan, and Boston Architecture Projects atMIT through museum videodiscs to suchcommercial efforts as the Chadwycl<-Healeypublications on American architecture. MertlySnow presented the U.C. Museum (Berkeley)prototype digitizing system described by HowardBesser in a recent issue of Museum StudiesJournal. For me, the most exciting event of theconference was the final fifteen minute talk byMarc Rorvig of Project ICON, at the Universityof Texas. Austin. Dr, Rorvig introduced theconcept of 6 visual thesaurus - Images Hnked inhierarchies with defined relatlons- as a means

Spring. 1988 Copyright by Archives It Museum Informatics 7

of conducting a "wordless search" and ofachieving greater consistency between indexers.He and his colleagues are constructing such avisual thesaurus for NASA. Rorvig suooested thatit be exam1ned in conj unction with the M T,bringing the meeting around in adelicious fullcircle. Whereas we began by struggling withhow to use language to index images and makethem available to new users from diversedisciplines. we ended by using images to indexconcepts in disciplines whose language was aliento us! Although what he clescribed was far fromthe image recognition capab111ty some in theaudience mistook it for. it was elegant for thesimple use it made of human intelligence totransport the user from one image, through aseries of related images. to the one needed.

RIAO 88: Conference on ·User-Oriented.Content-Based, Text and Image Handling-

Over 700 people from 21 countries gathered atM.I.T. March 21-24. for RIAO-88, aconferenceorganized by the centre des Hautes EtudesInternationales d'informatlque Dooumentaire.(Late In the week. Idiscovered that RIAO stoodfor Recherche d'information Assistee pari'Ordinateur ).

Participants were, like myself, largely firstt1me attendees drawn to the conference oy itSemphasis on hypermedia and full-text. Nor werewe disappointed, although the organization ofsessions left much to be desired. The 75 twentyminute papers were delivered much as they readin the three volume conference proceedings. withno opportunity for real discussion. and thedemonstrations of systems (almost allprototypes) were often hard to looate.Nevertheless. the meeting displayed the range ofactivity in these areas by researchers from theArtificial Intelligence. Information Retrieval,and Image Access commun1t1es. The emphas1s ondemonstrabIe systems made the reports broaderthan those ~t slmllar ACM and IEEE sesslons, andtogether wlth a ten page limit, this made thepapers seem somewhat superficial. I left excitedby many of the ideas being explored but

discouraged by the evidence that in hypermediaas in full text. our prototypes are far fromimplementation and scaling up to productionexposes unresolved problems.

Karen Spark-Jones. in her keynote address.characterized the problems as inherent inheterogeneity:

- We have objects of different kinds. Can theybe characterized in common w'(fo{s?

- We have data with different levels ofgranularity. Can they be connected?

- We have different functional uses. Can theybe supported by the same representation?

- We have different relevance criteria. Arethe apposite?Insum I she asked, can we go beyond agJregat1vesystems to integrative ones?

Aweek of looking at large screens, manywindows, bright color im~ and fUll-texts,large optical stores and attractive userinterfaces, could not obscure some very seriousproblems. We still can't capture very largeamounts of text or images from paper formatsvery efficiently. Automatic analysis indexing andanalysis are still in their infancy andcontent/structure relationships are not verywell understood. Freeing users to wander getsthem lost quickly. and even when theyfindthings. we have too few useful analytical toolsattached to retrieval systems to meet users'needs. Iadmit acertain production oriented biasin stating that Ifound the crude tricks moreexciting than the subtle methods.

Among the practical, implemented. conceptswas Lee Hollaar's Utah Retrieval System whichkeeps an index of words in full-text documentsonly. It ooesn't identify relative position orspecific 100ation, but string searches are fastenough. once the document is located, to executesuch sophisticated search, and the overhead forsimple indexes is only 15~. Yaacov Choekareported on acrude method of identifyingphrases, and names, for indexing. which simplylocates all occurences of 2,3,4 and 5 wordphroses in millions of words of text. andprovides an i~dexer with all phrases occuringmore than xtlmes. The resulting lists identified

8 Archival Informatics Newsletter yoJ.2,*1

potential indexing phrases with a precisiongreater (often much greater) that 50%. SusanneHumphrey reported on the National Library ofMedicine's automation-assisted indexing,emphasizing the useful assistance over anynotion of automatic indexing. NLM and OGLC .disp layed "orienting" interfaces that locate "hits"on tables of contents of large texts so that theuser can judge their likely interest based onlocation and clustering of occurences.

This is not to suggest that the prototypes lackeduseful ideas. Howard Besser showed off somevery nice and functional image analysis routinesdeveloped at the University of california atBerkeley and applied to art, architecture andgeography.. Andreas Dangel displayed anautomatic oocument structure analyser thatexploited simple rules about basic structuralcomponents of documents to locate areas ofdocuments with specific information contents(cf., addressee of letter). And Y. Tanaka ofHokkaido University discussed acreativeapproach to cooing characters that can't be OCR'dwith asimple 3 bit per character scheme thatproduced surprisingly high retrieval precisionfor words of more than five letters. Anumber ofclever AI techniques were presented ascomponents of other systems.

An exciting aspect of the conference was thenumber of venoors, especially those new to theUS market, whose products are being turned tothe retrieval of image and full-text. Almosteveryone showed a prototype developed with amuseum in order to show off videodisc or opticaldigital disk stored images in conjunction withtheir information systems. (We joked at onepoint that soon every museum in the world wouldhave asystem, each would be unique, and none oftheir venoors would still be in business.Sophlatec, an offshoot of the University of Nice,displayed English and French versions ofSophjnclocrM (C- DOS for the PC- based) and itspilot project with the Louvre (optical disk) andthe City of Nice (videodisc). SophiOOoorM boastsen integral th998urus generator, variable lengthl'le](1S 1n a re1at10na1 mlXle1, aM tt'le lmage 11nl<.SIGMINITM, a product of the Ecole NationaleSuperiore des Mines showed off aooclaratlve data

structure in which a record consists of numerousparenthetical declarations of the formelement=value, that enabled them to describe alland numerous parts of complex mosaics,including non-hierarchiCal relations (all tied toa videOOisc of mosaies).Honeywell-Bulldisplayed MISTRAln1J the software at the heart ofIMAGO 21M, the system used to manage theFrench national radio and television archives.GECI International exhibited Hyperdocn" itsmult1media OOcumentation package in Cfor PC'swith CD-ROM and the centre de Recherche enInformatique de Nancy displayed avisualthesaurus of mycol~ knowledge calledMycomoUen, to be marketted in July byFUTUR*VISION.

Anumber of US academic off shoot productswere shown as well. Robert Kraft showed off the$25 CD-ROM with full-text of the Bible inGreek, Latin, Hebrew, and English as well asmuch other material for septuagint Studies beingdistributed by the University of Pennsylvania.C<xJnetics Corporation demonstratedByperUesm version 2.3 ($249), developedby Ben Schneiderman at the University ofMaryland for "hypermanual" editing. BrownUniversity demonstrated Intermedia, itshypermedia courseware enVironment, ~ut

similar projects seen at EDUCOM meetmgs, fromM.I.T. itself (Athena), CMU (Andrew), andnumerous other U.S. campuses, were absent.

Now that digital image storage systems arehere, the problems of digital data capture and theabsence of digital analysis capabllities (orneeds?) invites us to re-examine someassumptions. For the first time 1n a few years,it was respectable again to question the need fordigitia1 storage of images. Quite a few of those towhom Ispoke admitted to reconsidering v1deodiscand even microform storage of images withdlg1tlzation on demand, either at the localworkstation or from afile server. Whilearch1vlsts and museum staffs need to rema1nalert to changes in image storage andtransm1ss10n technolog1es, thOSe who are notacquiring systems immediately can use this timeto analyze the uses their clients make of images.

Spring. 1988 Copyright by Archives. Museum Informatics 9

,,II

[

t,I

li

I~,.'I~ 1

CONFERENCE CALENDAR2nd & 3rd Quarters 1988

April 13-17 Society of ArchitecturalHistorians, Annual Meeting, Chicago [SAH, 1232Pine St.. Philadelphia, PA, 19107]

Apr1l 29-30 Library Descriptive Standards:An Introduction for Archivists, aworkshopsponsored by the Society of American Archivistsin Chicago [600 S. Federal St., Suite 504,Chicago, tL 60605; 312-922-0140] Also to beoffered in Atlanta, prior to the SAA annual mtg.

May 15-18 Artificial Intelligence, the ASISmid-year conference, Ann Arbor, MI [AmericanSociety for Information Science, 1424 16thSt. ,NW, Washington, DC 20036]

May 22-26 Basic Videodisc Design &Production Workshop sponsored by the NebraskaVideodisc Group in Lincoln,NE. [P.O.Box 83111 ,Uncoln, NE 68501; 402-472-3611]. Alsooffered July 17-21.

May 23-24 "The Coming Age of ElectronicText"; Study Group on the Structure of ElectronicText (SGSET) , Conference, Carnegie MellonUniversity, Pittsburgh. PA [Peter Capell. 412­268-8599]

May 26-29, IASSIST Annual Conference­"Public Data: Use it or Lose it", Washington DC.[Pat Doyle, Mathematica Policy Research Inc.,600 Maryland Ave. SW, Suite 550, Washington,DC 20024]

June 3-7 American Association of Museums,Annual Meeting, Pittsburgh, PA [AAM, 1225 EyeSt., NW, Washington, DC 20005]

June 6-10 Association of canadian Archivists,Annual Meeting, Windsor Ont.[secretariat, Learned Societies Conf. 1988,Room 2129 W'indsor Hall North, Windsor, N9B­3P4,CANADA;519-253-4232]

July 8 Management Strategies for DisasterPreparedness, aworkshop sponsored by ALA

Resources and Technical Services Division, NewOrleans, [RTSD, 50 East Huron St., Chicago,lL60611;312~944-6780]

July 19-23 Microcomputer Applications inVisual Resource Collections [F ine Arts Contin.Education, FlneArts Building 2.4, Univ. ofTexas, Austin, Texas 78712; 512-471-8862]

July 20-23 National Association ofGovernment Archives and RecordsAdministrators (NAeARA) Annual meeting,Annapolis, MD [Stephen Cooper, Maryland StateArchiVes, 350 Rowe Blvd., Annapolis, MD21401;301-974-3914]

August 21-28 Preservation of Black & WhitePhotogn'lphs, two workshops [Rochester Instituteof Technology, College of Graphics Arts &Photography, Tech. Ed. center, One LombMemorial Drive. Rochester, NY 14623; 716­475-2757). "Identifying image formingprocesses, hand11ng & stor1ng" August 21-25;"Copying and dupl1cating", August 26-28.

September 11-17 American Association forState and Local History, Annual Meeting,Rochester NY [AASLH, 172 second Ave. , North,Suite 100, Nashville. TN, 37201]

september 21- 24 Second Annual MuseumDocumentation Association Conference,"Terminology for Museum Documentation",cambridge, UK [MDA. 347 Cherry Hinton Rd.,Cambridge, CB 1-4DH, England]

September 29-OCtober 2 Society ofAmerican Archivists, Annual Meeting, AtlantaeA[SAA, 600 S. F~ral St.,Suite 504, Chi~, IL60605]

CALL FOR PAPERSProposals for papers, accompanied by

abstrocts, will be occepled until Apri130, forthe Museum Computer Network AnnualConference, to be held in Los Angeles, OCtober26-28. Contact David Bearman, ProgramCoordinator, 5600 Northumberland S1.,Pittsburgh. PA 15217 (412-421-4638)

i

I••

10 Archival Informatics Newsletter

IN-BOX

REPORTS

Bilfinger, Monica; Buyssens, Danielle; Jost,Kar1j Meles, Brigitte &. Zurcher, Ronald; Versune oangue de donnees culturelles et artlstlguessuisse: Conception de l'informatisation et del'echange de donnees dans le domaine des beaux­arts etdes arts appliQUes, Berne, Academiesuisse des sciences humaines (ASSH), 1987.117pp. also publ1shed in German

This Swiss working group report on thepossibility and value of information exchange inthe arts (commissioned in February 1986 andcom p1etOO in June 1987) sounds many themesthat are famillar from simllar "towards nationalsystems" studles elsewhere. Whlle 11sconclusions are. appropriately. tied to Swisscircumstances, the systematic manner in whichit addresses options, and its application of rigidcriteria of technical and politicalimplementability, should be emulated. As isoften the result ofthe best of such efforts, therecommendations are modest and focus oncommon standards for description and theexp10ration of specific cooperative ventures,ranging from information sharing consortia toshared videodiscs. They emphasize the need forlmprovlng locallnventory control, and creating aunion list over the allure of an online database,and the desirability of shared vocabularies overinteractive authority control. [Available fromMrs. Anne Christine Voge1-CloHu, SwissAcademy of Humanities, Hirscheograben 11.P.O.Box 2523, CH-300 1Bern, Switzerland]

Horvath, DaVid; The Acetate Negative Survey:Finol Report, Louisville, KY, University ofLouisv'l11e, 1987, 91pp. $10.00 from thePhotographic Archives, U.Louisv1l1e, 40292

The final report of this National Museum Actfunded project will be of great value to those whohave responsibility for historical collections ofcellulose ocetate film (many of whompartlcipated in the survey that informed thestudy). Follow'lng adetanoo history of the

problems associated with degradation of "safety"negatiVes dating from 1925-55, and an ~unt

. of the various manufacturing processes Involved,the report summarises the technical literatureon cellulose acetate film stability and factors in .its degradation. Survey findings are summarisedand some conclusions are drawn leading to ,tentati\'e recommendations. Excellentbibliographies and notch reference identificationinformation are appended.

Our Memory at Risk: Preserving New York'sUnigue Research Resources. Areport andRecommendations to the Citizens of New York bythe New York Document Conservation AdvisoryCouncil. [Albany] ,1988. 56pp.

Building on the foundation ora 1984 report"Towards a Usable Past", which stated that"preservation may beregarded as the mostimportant historical records Issue in New Yorktoday" J this report advances a practical programof action at all levels and by an astonishing rangeof tactics. Handsomely and clearly presented,without Jargon or hyperbole, its firstrecommendat1on is to "complete theinitfa1Historical Documents Inventory project,maintain an automated statewide database forhistorical records collections and repositoryacquisition pol1cies, and provide for updating thesystem'"

Vislon 2000: ACooperatlye 10ng-ranlE plan forthe Moine State library, Mojne HistoricPreservation Commission. Maine ArtsCommission and Malne State Museum, [AugustaME, 1987, 80pp).

Aneeds assessment and cooperative planningoocument for cultural agencies In Maine, similarin intent to the 1984 plan developed in New YorkState, but without the synthesis. Here eachagency has daveloped plans to extend its currentfunctions, and juxtaposed them, but there is nostrateqy. Iwas struck by the absence of anyattention to documentation issues, and by theclaim that "no cost" is assoclated wlth theobjective of the museum to "pursue an aggressiveprooram of collections management" or theobjective of the library to study "informationacquisition and data handllng within stategovernment".

Spring. 1988 Copyright by Archives It Museum Informatics 11

-I

SPECIAL JOURNAL ISSUES

AlCARC: Bulletin of the Archives andDocumentation Centers for Modern andContemporary Art, (#25 & 26, 1986/2­1987/1); ; Special issue on "Computers and theFuture of Art Research: Visions, Problems,Projects", 61 pp.

This issue of AICARC brings together the viewsof leading figures in art documentation fromthroughout ~he world, in aseries of tantalizing,short but stlmulating essays on where we are andare go1ng, extencl1ng and 1n some cases rev1st1ngAICARC issues 21/22 "Automation takesCommand" published following the 2ndInternational conference on Automatic DataProcessing of Art History held in Pisa in the fallof 1984. Hans-Jorg Heusser, the issue editor,remarks that this issue is amore sober, if notsomber, view as aconsequence of the hard workof the past several years. While JacquesThuillier, Oskar Batchmann, Vi1 Mirimanov andDimitri Pertsev still herald anticipatedrevolutions in historiography and art criticalmetnoos, most autnors are caut10us, choos1nglike salome SChmid-Isler to assess experts 'systems or like Helene Roberts to explore therequirements of research databases, withoutpredicting that they w"ill radically alter theory.several contributions focus on the critical needfor more data in automated form, expressed byKarl Jost .in his article "Quantity is Quality". andreflected 10 projects reported by Laura Corti,Margrethe F1oryan-Pedersen, and others. Auseful contribution by Brigitte Me1es describesthe "Databases available to Art HistoriansToday", and suooests how much useful researchwe could be dOing with automated assistancea1rea<ty. Afinal set of articles addresses howscho.1ars use, and could use, automated systems,and lOc1udes an exciting testimonial by l'1ary1inAronberg Lavin of intellectual breakthroughs shemade by using such adatabase as well as anannouncement by Marilyn SChmidt of theavailability of the Getty Trust/Brown Universityresearch on the practices of art historianspublished under the title Image, Object and'Inquirv (avai labIe from the Getty Art HistoryInformation Program).

Hum~nistisk.e D&a, 3-87 is aspecial issuedevoted to optical media projects. Englishlanguage articles include areport on "The BBCAdvanced Interactive Video and the DomesdayDiscs" by Phyllis Gave, "The North WestEducational Computing Project" by Ian Robertsonand Mike Picciotto, "Recording the ItalianCultural Heritage" by Ernesto Barto1ozzi, and "ACD-ROM based Geographic Information System"by Erling Maartmann-Moe. Bartolozzi presentsthe plans for implementing avariety of opticalmedia systems through the Italian NationalL1brary Holding System and NationalPhotographic Archives and discusses systemsarchitecture. Maartmann-Moe defines thefunctional requirements of GeographicalInformation Systems in place at the NorwegianComputing Center and explores the implicationsof having such systems on CD-ROM.

Library Hi Tech #20 (Winter 1987) wasdevoted tO$Pace planning for culturalrepositories and the implications of informationtechnologies. Its ahot topic; Museum Newspromises to focus on architecture in general inits May/June issue. The Society of AmericanArchivists is seeking an author for aSpecHilpubliootion on the topic. And I recently receivedareview copy of Richard W. Boss' InformationIechno1ooies and SDace Planning for Librariesand Information Centers (Boston, G.K.Hall,1987. 116p. + index).

Boss begins by discussing all the relevanthardware and communication systems and theirspace, power and communications requirements.Chapters are cl3voted to automated library(cataloging and circulation) systems.microform, optical media, telefacsimile andcompact storage. Avariety of formula's used bydifferent institut10ns for calculating spacerequirements are presented along with numeroususeful details throughout the text but theconclUding chapter, which recommends that al'ibrary should retain qualified architects to helpit plan, include space for computers andterm'lnals, use standard ratios to plan to storingprinted materia1sand for floor loads and carrelspace and allow 175 ft per staff member forwork space, suggests that information

I 12 Archival Informatics Newsletter vo1.2.·1

,technologies really don't alter space planning atall. The only "tec~lnology forecast" basedrecommendations are to allow for limited growthof microform areas. use no specially constructeddesks, try to minimize glare and use dry pipesprinkling systems to minimize water damage(not even halonl). Did we need another book totell us this?

Fortunately, authors in Library HiTech aremore critical. John Kountz presents the costcomparisions of traditional shelving andindustrial shelving using automated storage andretrieval techniques and Michael Gormanassesses the pros and cons of movable compactshelving, David Michaels presents acase fordiscarding those very tradit10nal planningformula's that Boss recommends. Two superb"roundtables" of librarians and designers addresshow to bul1d the "forgiving" building, one thatperm1ts altered 1nternal space ut1l1zation andaccommodates changing systems reqUirements.case studies of Infomart and other intelligentbuildings, the new OCLC headquarters and theremodel Iing of the california State Library roundout an exceptionally useful issue.

Not surprisingly all the authors agree that amulti-expert team is required to design well,what they don't propose explicitly is how thatteam should record its working assumptions andthe design conclusions that follow in order toSUbject them to scrutiny and critique.Impl1cltly, any argument from functIon to formsuggests that each design element shouldwithstand such scrutiny, but which assumptionsabout potential technological developments shouldbe factored? After considering how to cableevery possible location and all exhibit floorspace, should we ask ourselves what wirelesscommunications within abUllding would mean?After planning for stacks that require power tomove and communications outlets for handheldterminals to transmit inventory informationback to acentral system, should we considerwhat impl1cations communIcating shelves andcontainers, or robotic retrieval would have?Designing bulldings for longterm futures is inmost respects astandard risk managementactiVity - it is odd, therefore, to see no rigorousrisk management methodologies in any of thediscussion cited. Builder beware.

. Museum Studies Journal, vo1.3 # 1, Fall-Winter1987; Special section: "Information Technology& Museums", pp.41- 110

Michael Templeton's introduction to thisspecial feature section of the Museum StudiesJournal raises questions about institutionalbehavior 1n response to technologIcal change thatare addressed in only one paper in acollectionthat seems to have been random ly associated.James Beninger and Georgia Freedman-Harveytake up Michael Templeton's Questions in whatinitially promises to be astudy of the impact ofinteractive technologies on exhibit success andvisitor unoorstanding but degenerates into adiatribe against museums the authors call"chlldren's zoo's", that use techniques intended toinvolve visitors and fal1s to report any researchresults.

Two articles In thIs Issue are Importantreading. Lenore sarasan's sensible "What to Lookfor in an Automated COllections ManagementSystem" can be read with profit by novice andexpert alike. Howard Besser's discussion of"Digital Images for Museums" is an insightfulintroduction to the potential of this technology.Associate Editor Barbara Thorn pson has compiledauseful, if slightly ideosyncratic andincomplete, "resources" directory.

On the other hand, the editors made somedubious decisions. Hubert and Stuart Dreyfushave collaborated with Renee Dreyfus 1n aphilosophical discussion of artificial intell igencethat bowdlerizes their own book, Mind overMachines and doesn't try very hard to beapplicable to museums. The editors have chosento reprint David Williams' relativelyunsatisfactory history of museum automationfrom his more unsatisfactory Guide to MuseumCOmputing.

ARTICLES & BOOKS

Bender, Avi; "Optical disc technology for recordsmanagement: Auser's perspective" , TheElectronic LIbrary, 1987, vo1.5 #5, p276-81

Reports on the partially implemented. or pilot,Nuclear RegUlatory COmmission system forstoring text and image of its l1censing records onWORM.

Spring. 1988 Copyright by Archives It Museum Informatics 13

Genoways, Hugh H.; Jones, Clyde & R033011mo,Olga L. ,ads.; Mammal Collection Management.Lubbock,TX, Texas Tech University Press, 1987219pp.

These papers, given at aworkshop held inAugust 1985, are arranged in three sections.The first three papers define the Importance ofmammal collections, the need for theirmanagement, and the history of mammalspreservation. Seven papers in the second sectionare devoted to computerization issues. The finalsection reports on mammal collections inAustralla, Hungary, Spaln, Lattn Amerlca anOIndia. Terry Yates introduction to the value ofcollections as evidence is auseful contribution tothe literature of museoloqy and as valid for otherdisciplines as for mammal collectors. Apaper byDaniel F. Williams on computer hardwareselection is surprisingly fresh (given its age),but won't remain valid much longer. Afewuseful insights about aspects of mammalcollections and their automation can be salVagedfrom the reports on systems Implemented at theCarnegie Museum, the National l'1useum ofNatural History, the Royal Ontario Museum, theField Museum and Texas Tech, but the guIfbetween 1983/84 and the present is onlymagnified as we read of batch query systems,punched card entry and mini-computersoperati~g with 64K.

Johnson, Susan; "The Birth of an InformationSystem", Information Retrieval & libraryAutomation, Oct 1987.

The LSS Is characteristic of the large,integroted, multi-agency, on-line informotionsystems that are proving so problematic toarchivists; the history of its resign andimplementation is, therefore, of special interest.

Moffett, Jonathan, "Computing in the Departmentof Antiquities of the Ashmolean Museum",Archeolooical Compyting Newsletter, # 13,Dec.1987, p.l 5-20

This account by the Chairman of the Museums'Computing Group of pr~ress in his owninstitution details what an alert professional can00 with few resources.

Veccia, Susan; "Full-Text dilemmas for~rchers and 3Ystems: The Washington PostOnline", Database, April 1988, vol.ll #2,p.13-32

Ninel different versions of the Washington Postare available electronically in as many differentsearching systems. Understanding what works,and doesn't and why is critical to anyonedesigning full-text systems; incidentally thearticle may help librarians and archivists makedecisions about what formats to keep newspapersIn.

NEWSLETTERS

Quite a number of the "Newsletters" I receiveare serially issued promotional literature,available free from commercial firms. Some ofthese contain quite valuable information on aregUlar basis. Among those received recentlyare:FAX Facts [Paper Manufacturers Company•OfficeProducts DIvIsIon, 24 TrIangle Park DrIve,Cincinnati OH 45246]Government PUbl1cations News [BernanAssociates, 4611-F Assembly Dr., Lanham, MD20706-4391]. New In 1988; the first threeissues include quite useful general information.as well as order forms.MAPS Newsletter [Mid-Atlantic Preservationservice, Lehigh University Mountaintop Campus,Bethlehem, PA 18015] MAPS is amicrofilmingservice organization, and its newsletter includesa "Technicallssues" section of general interest topreservation microfi 1miog officers.The Scanner: ABarcode Newsletter [WKM/NIDI,88 Westpark Rd. , Dayton, OH 45459]

Other "Newsletters" issued by commercialfirms that often include information of generalinterest are user group publications. Amongthose recently received are:Up & Running [Questor Systems, 1005 E.Colorado Blvd., Pasadena, CA 911 06]; MUSENews [Julian Humphreys,Corson Hall, CornellUniversity. Ithaca, NY 14853]; Socjety ofAmerican Archivists NOT IS Users GroupNewslettedPatty Cloud, NorthwesternUnIversIty Archives, UnIversIty L1brary 1935

14 Archival Informatics Newsletter vo1.2.·1

Sheridan Rd., EV8nston,IL. 60208]; Network~ (Conservation Information Network, 4503Glencoe Ave., Marina del Rey, CA 90292] ,recently announced avai labi1ity of it up-load anddown-load local systems software and thecompilation of an [-mail directory forconservation professionals.

NOTE:~ suspended pub1ication at the end of1987 (vol.5). According to editor Joseph Raben,it will resume in ~mother form soon.

EPHEMERA:

Directory of Federal Historical Programs andActjvities, Washington DC, Society for History inthe Federal Government, American HistoricalAssociation & National Coordinating Committeefor the Promotion of History. 1987. 84pp.Alist of phone numbers and addresses for an

astonishingly diverse set of people and projects.

1988 Com Duter salary Survey and careerPlanning Guire, 14pp. (free from Source EDPoffices throughout the U.S, and in Ontario] is auseful source of job descriptions and salary datafor employers.

Archival Informatics Newsletter is aQuarterlypublication of Archives & Museum Informatics,5600 Northumberland St., Pittsburgh, PA15217; (412-421-4638). It is edited byDavid Bearman, whose authorship can bepresumed for all items otherwise not attributed.Subscription to the Archival InformaticsNewsletter (ISSN 0892-2179) is available for$24.00 per year pre-paid, US acxJressed,$30.00 pre-paid foreign addresses, and $40.00p.a. billed. Acombined subscription to both theNewsletter and acompanion QuarterlyDub1ication. the Archival Informatics TechnicalReports (155N 0694-0266) is cvailllble for$160.00 p.a. in the US; $180 abroad; and willbe b1lJed at no extra Charge. Ind1vldual technicalreports are available at $W45, each preoaid.

LETTERS TO THE ED ITOR

MARCONTed Purr [AIRS Inc., 335 Paint Branch Dr.,College Park, MD 20742] writes:

"The review of MARCON by David Bearman inthe Winter 1987 issue of Archival InformaticsNewsletter reminded ine of the Broadway playFiddler on the Roof. Remember the lines,"On the one hand... , on the other hand... "? Thatwas David's review. There were some lines wewouldli ke to shout at the wor ld; other we wouldjust as soon mumble. Of course MARCON is not aplay; users intend it for serious busi ness.

Overall we feel that David was objective, asusual. We have produced. and are continuallyupgrading. aproduct that, as another reviewer,also from Pittsburoh, said. "ooes everything".From 0 softwore developer's point of view thetruth of that overstatement is not that it ooes itall but that it tries to do so many things thatsometimes, in layperson's terms, it trips.

AI RS spends about half its customer supporttime teaching users the ins and outs of databasemanagement as applied to archives and recordsmanagement. The other half is spent by ourusers teaching us - either about deficiences orabout things they would like, We look on this as apartnership and we thank our many friends.

David's review contained some helpfulsuggestions. His point obout the help screen ondmanual "mess" mne by an external firm has beennoted and new help screens and awhole newmanual are now being shipped. Ahelpful optionsbar will appear in afuture version of MARCON.

David did make one error. He states that in dataentry the screen mas not default to the next .blank template. If you pressShift-Fl0("insteadof f 10), the data is entered and another blanktemplate appears.

The comments about reports were accurate andthis difficult module has been corrected. AIRSsuggests that sometime soon Dcvid might write areview just of report generators, comparingseverlll packl:lges. Our mlljor report adv~nce is togive users two options: ( 1)to print reports asrecord display appears online and (2) to allowusers to "paint" reports with row and column

Spring. 1988 Copyright by Archives It Museum Infor••tics 15

formatting. Sorting is possible for multi-valuedfields, but, ~ noted, intro, not inter collection.

The global edit feature was in the version Davidreviewed but is not in the current version nowbeing shipped., There are some problems of bothdesign and consistency in this feature which weare addressing.

Finally, we expect to have a MARC interface(both for entering and receiving records) in betatest1ng by the summer, We w1llintroduce 1t atthe SAA this fall in Atlanta. We w1ll also offerworkshops in Atlanta prior to the annualmeeting. One w111 be for basic MARCON, and theother for using the U.S. MARC:AMC interfr£e.

MICROFORM YS. OPTICAL DISCC.lee Jones, previously of the Council onLibrary Resources, and now President of theMid-Atlantic Preservation service [ MAPS,Lehigh Univ. Mountaintop campus, 118 ResearchDrive, Bldg.J, Rm.120, Bethlehem, PA 18015],writes:

"At least for bl~k and white and some mono­and duotone print media, the question is notwhether or not to chose optical media for storageond~s, but whether or not it meets the needsof users. There is an option that allows thecapture of print media in more traditlonalmicro-optical (read microfiim) format withconversion to whatever user format may berequired. While the technique is not widelyknown, it is used by some of the l8r~ micro­and disc-publishers.

Aunit known as acomposing/ reduction camerawill take any of a variety of input streams,including 8nal~ or digital or fi 1m, and createoutputs in an equally broad spectrum of formats.The price tog runs to $2 million and the unitsare custom built for the needs of the owner andhis services.

MAPS as you may know was brought into beingwith capital support from the EXXON EducationFoundation for the express purpose of developingand demonstrating preservation content capturestrategies in the not-for-profit environment.The first step was to evaluate options and afterlooking at the various options inclUding opticalmedia, it became clear that the most rationalcurrent appr~h was microfilm, adecision thatcaught many, including EXXON, by surprise. We

continue to keep our ears to the ground and oureyes peeled for any indication that our selectedpreservation format may be replaced. For thefirst few months, we felt certain that we would .end up using some disc technology for capture,preservation and use. However, acombination ofthese strategies now seems the most prudentcourse to follow.

35mm. archival fllm has been demonstrated tobe extremely stab1e when stored under properconditions and systematically monitored; someclaiming as much as 500 years life. While Iwould be happy with 300 or 400 years, thepoint is likely to be moot over those time frames.35mm is also the dominant microformat frompreservation filming, with 16mm adistant andfading second. Microfiche, while lacking thearchival life of roll film ( 100 vs. 300-500years), 1s becoming increasIngly popular withlibrarians wanting to provide the most usefulformats for their readers. Consequently theynow believe that original filming has to be donein a fiche format or 16mm and jacket loaded.These issues forced us to consider what if anyalternatives there might be for prcx1uction offiche from roll film. Cl~rly, there arerelatively inexpensive machines coming on themarket soon that allow us to prcx1uce fiche on1OSmm roll film, but there are then some veryserious material handling problems on thelibrary end of the system.

The composing/reducing camera (CRC), asnoted above, is capable of handling a range of·Inputs and creating outputs in avariety offormats as well. With the professional pressureto continue to explore "electronic/ optical"preservation techniques and the materialhandling problems of 105mm roll fiche, the CRCstarts to become an appealing alternative to thelimits of one content capture format or another.

It is likely that MAPS will put together arequirement statement for such a machine alongthe following 11nes. The primary input stream islikely to be 35mm fllm, probably S8C.ondgeneration in order to preserve the integrity ofthe archive master (the original camera film).However, we are also look1ng at capturing someimages in adigital form prior to laying the imageon film, so we would also like the option of ananalog or digital input stream. This particular

16 Archival Informatics Newsletter vo1.2.·1

feature would be a deduct alternative should itdrive the price higher than resources available,

The output streams required from 35mm rollfilm will include microfiche, 16mm roll film(another ear ly deduct alternative feature), aCD-ROM format (specific technicalrequirements to be determ ined later), andpreservotion photocopy output (yes, even 0

paper option). The engine of this camera is ahigh speed minicomputer which may soon shrinkto asuper micro. Despite that sort of shrinkage,the cost is very high. The first copy will be inthe $2 million range and the second and third inthe $1 mill ion range.

With costs like this, it is not possible for thepresent MAPS to justify such an investment.However, if the Commission on Preservation andAccess is successful in raising $200 million infunds to do the content capture work in libraries,orchives ond museums, then we ore tolldng ot 0much expanded MAPS (with four or five otherregional sites) or several partners interested inthe same business (preferably not-for-profitpartners in order to keep the tax situation assimple as possible) or a combination of thesetwo. The latter is the more likely circumstance.In any case, a CRC will be more re~H1y

amortized if it can be operated two or more shiftsper day, in which case, it will be capabIe ofhandling the output of many 35mm cameras.

All of this just to indicate that there are someoptions thot we expect to come on streom in the·near future that do not force one to decide on alimited archival life in order to provide thepower and flexibility of the user format. It isposs'ible to have a comb'ination of stable film forarchival purposes, even archival photocopy forhands on use, and/or CD-ROM for a "high tech"retrieval and access format. Another virtue ofsuch a combination system is that one does nothave to consider double capture if both archivalfilm and CD-ROM is required for material.

The issue of color will be addressedexperimentolly only after the CRC (ond itsmates) are in operation. There is some light atthe end of the color microform (archival) tunnelwith one group claiming far more stable G1es.Actually, if the dyes were archival for as muchas 100 years, they would probably be deemed anarchival product with aplan to duplicate every

75 years or so. Should that occur I MAPS wouldbe interested in using the CRC in a color mode.So far as I know, there is nothing like that inexistence yet."

SM Automation Program OfficerDonn Neal [Executive Director, Society ofAmeriCl'Jn Archivists, 6005. Federal St., Suite504, Chicago, IL 60605], writes:

"As you may know, Lisa Weber has accepted aposition with NHPRC and will be leaving SAA inMid-May. While we are celebrating with herabout her new opportunity, we are also aware ofthe gap that her departure wi11leave in the SAAoffice: we will be without a Program Officer todirect the remaining months of our NEHAutomation Project, to perform staff duties inthe areas of automation, and to help SAA plan forthe future of this very important initiative.

10m writing to ytou for assistonce inidentifying candidates for this position ...

The NEH grant runs through september of1989. Our first choice, of course, would be tohave someone simply pick up where Lisa isleaving off and see the project through tocompletion while helping us to chart our futurecourse. We realize, though, that few people maybe willing to pull up stakes and move to Chicagofor this purpose, so we are quite open to other,creative solutions. These might include havingsomeone on leave from a regular position, thenegJ1.iotion of ot leost 50:8 releosed time so lhotSM duties might be shared with regular ones, ashort commute, and others ... "

Museum Computer Network DirectorIn asimilar vein, Suzunnah Fabing [Chair.

MCN Executive Director selection Committee,National Gallery of Art, Washington, DC 20565]could have written me that MCN is seeking a fulltime executive director, willing to securefunding for MCN projects, double membershipnext year and expand member services. The MCNoffice will move to the incumbents' locotton. Adedicated individual is sought, especially sincethe salary is only $30K and the future of the jobdepends upon the success of the appointee.Applications accepted to mid-May. Interviews tobe held in Pittsburgh following MM meeting.

Spring. 1988 Copyright by Archives at Museum Informatics 17

NEWSNARA CHALLENGE

Don Wilson was sworn in as archivist of theUnited States on December 4. I look forward toseeing how he will act on the challenge heIdentified in his confirmation hearings onOCtober 20. when he said:

"The ability toadminister public records in thecomputer aoe depends on the ability of theNational Archives to adept to the changes. It mustassume anat10na11eadersh'Ip pos1t10n. Oldarchival theories and practices must bereexam1ned and adapted to present day needs. Thepassive role of archivists waiting for old files ofpaper to be transferred by reluctant agenciesmust be replaced by active participation inrecords management. "

ADVICE FROM NARAAs if in response, NARA has announced the

avallaO'll1ty of two free 1nformat1on ~I<ets:

"Managing Electronic Records" and "RecordsManagement Software Packages". From theRecords Administration Information center(RAIC). Agency services Division, NationalArchives. Washington, DC 20404. Or phone202-724-1471.

OPTICAL MEDIA IN GOVERNMENTSuch advice is going to need to 8CXjress the

impact of write-once optical digital media ingovernment accord1ng to Karen M.G.Howell."Federal Government Applications of Write-OnceRead-Many(WORM) Optical Disk Systems inLibrary Hi Tech News. January 1988). Inaddition, government archivists need to payattention to CD-ROM publications. E.J.McFau1,in CD Data Report Feb, 1988, pp.32--36. listsU,S. Government applications agency by agency,and there's hardly an agencY untouched. This ontop of Lindo Helgerson's report in January on theUSPS Zip+4 CD, and Robert Williams' report inDecember on the U.S.Navy Paperless Shipprooram should wake Government archivists up!

LIABILITY FOR ELECTRONICMISlNfORMAliON

The New York Times reported on March 6 thatTerry Dean Rogan, arrested five times in Texasand Michigan for crimes he did not commit. wasawarded $55,000 from the City of Los Angeleswhich had failed to remove his name from theirpolice files where it had fed into anational crimenetwork. The network is aCalifornia basedversion of the New York State system that AlanKowlowitz has appraised, and which he haswritten up for the f~1l1966 Archiv~l

Informatics Technical Report. At $55K ashot,we had better devise !J)Od retention and disposalprocedures!

NEWS FROM STATE ARCHIVESThe NAGARA Clearinghouse reports that the

Delaware State Office of Information Systems infunding aMachine-Readable records appraisalplanning project. Hawaii's new archivist JolynG. Tamura is exploring automation to improverecords management. Mississippi has received agrant to automate the trackIng of newspaperconservation data during its statewidenewspaper project. North Carolina is addingfolder level data on over 200,000 land grantrecords to its FAIDS system. The State ofWashington is integrating all its recordsmanagement, archives and microfilmmanagement systems using GenGatTM, softwaredeveloped by Eloquent Systems.

HISTORY OF INFORMATION TECHNOLOGY

Background papers from the Symposium on theHistorical Context of the Evolution of InformationTechnology held at the Notional Museum ofAmerican History sept. 11 J 1987 are availablefrom the Museum upon request. Asummary ofthe meeting was published in "News of theInformation Exhibition" aquarterly newsletterfrom the staff of this large new exhibit plannedto open in 1990. Contoc:t David Allison J Curator,National Museum of American History,Washington, DC 20560 (202-357-2038).

18 Archival Informatics Newsletter

I

PROJECTS ~ PROPOSALS

RlO/Stote Archlves ProjectThe Research Libraries Group project in

which the State Archives of Alabama, california,Minnesota, New York, Pennsylvania, Utah andWisconsin have been involved for the past threeyears has now completed Hs NHPRC and CLRfunded effort to incorporate state and localgovernment records into the RLI Ndatabase,utilize RLI Nfor collections actions management,develop a list of functions terms to describe theactivity of governmental agencies and test RUNas a mechanism for assisting In appraisal ofgovernmental records. RLG will submit a finalreport on the three year project to the NHPRC inJune, along with aproposal for an even moreambitious second phase.

Having demonstrated the potential benefits ofRUN use by State archives, RLG Is expanding theproject to Include the National Archives. andmunicipal archival agencies, and pursumg thedevelopment of form-of-materlal terminologyand authority lists to be used in conjunction witharchival appraisal, description and retrieval.Their proposal is, in effect, to augment thedatabase being built by the initial projectparticipants and explore the full value of anational system for archives. One would expectthe conclusions of this second phase, if funded, todetermine not only the costs and benefits of suchexchanges but also prOVide the standards for databeyond those being developed by Steve Henson forarchival descriptive cataloging. In particular,the project should test the viability of inter­~ernmental records strategies, such as thoseproposed by the States and echoed recently byFrank Evans at the National Archives.

SCholorly Text £ncoolngThe NEH funded conference on the requirements

of scholars for text encoding standards held atVassar College in November (vol, 1, p.62),produced the follOWing framework for thepreparation of text encoding guidelines:

"1, The guidelines are intended to provide astandard format for data interchange inhumanities research.

2. The guidelines are also intended to suggest. principles for the encoding of texts in the same

format.3. The guidelines should

a. define a recommended syntax for theformat,

b. define a metalanguage for the description oftext encoding schemes,

C. describe the new format and representativeeXisting schemes both in that meta-language andin prose.

4. The guidelines should propose sets of codingconventions suited for various applications.

5. The gUidelines should include a minimal setof conventions for encoding new texts in theformat.

6. The guidelines are to be drafted bycommittees on

a. text documentationb. text representationc. text interpretation and analysisd. metalanguage definition and description of

existing and proposed schemes, coordinated by asteering committee of representatiVes of theprincipal sponsoring organizations.

7. Compatibility with existing standards willbe maintained as far as possible.

8, Anumber of large text archives have agreedin principle to support the guidelines in theirfunction as an interchange tool. We encouragefunding agencies to support development of toolsto facilitate this interchange,

9. Conversion of existing machine-readabletexts to the new format involves the translationof their conventions into the systax of the newformat. No reqUirements will be made for theaddition of information not already coded in thetexts."

Based on this declaration from the planningconference, the Association for Computers andthe Humanities, Association for ComputationalLinguisitcs and the Association for Literary andLinguistic Computing, along with a large numberof other scholarly organizations, have proposedto the NEH 8 three year project to draft andapprove standards for: .- documentation of encoded texts- representation of texts at the

typograph1callevel

--_.Spring. 1988 Copyright by Archives It Museum Informatics J9

- representation of scholar ly analysisand interpretation in on encoded text

- formal descriptions of the syntax ofthis and other encoding schemes.

[contact; project director, Dr. C. M. Sperberg­McQueen, Computer Genter (M/C 135),University of Illinois at Chicago, Box 6998,Chicago, 11 60680]

Cornell Un1v. MP Sl1de UbraryFor several years, Cornell University's

Architecture, Art 8< Planning (MP) SlideLibrary has been stu~ing computerization ofaccess to its holdings of approximately 350,000slicBs. It recently issued a report on the nearlycompleted "specifications" phase in which itcXx:uments its decision to use videodisc overoptical digital disc and acommercial full-textretrieval package.[conttX;t: Nancy Humphries, consultant,E-TECH,607-539-6220]

Art and Architecture ThesaurusIn January 1988, the MT distributed its 17th

hierarchy, FUNCTIONS, to its user community,together with revised versions of:Architectural componentsSingle built works and open spacesBunt complexes and areasSettlements, Systems 8< Landscapes

. Document TypesVisual GenreDrawingsStyles and PerioclsMaterialsPeople and OrganizationsDesign Elements and AttributesProcesses and ToolsDisciplinesAssreiated ConceptsEvents

TheMT is now nearly half completed,although it will never be finished in the sense ofbeing frozen. Plans are underway to publish thecompleted hierarchies in a number of formatsand to make them available on-line.

The structure of the FUNCTIONS hierarchy asdrafted is quite simple. seven sUb-cat8lJ)riesare defined beneath which all functions terms arearr~ed. These are:

<analytical functions)<: collections management functions)<economic and financial functions><governmental and legal functions><information handling and

communications functions><organizational functions><travel and recreational functions>

MT Project Director Toni Petersen hasrecently proposed that the MT Functionshierarchy be broadened to include termsrequired by archivists. She has also made theoffer to expand and then maintain the MT'sDocument Types hierarchy (similar to archivalform-of- material), and the Processes andTechniques and Drawings hierarchies need byusers of the MARC:VM format. One possibility isto mount all the MT hierarchies on RL IN

Archival Descriptive StandardsThe Canadian Planning Committee on

Descriptive Standards has announced afree,"occasional" paper, entitled "DevelopingDescriptive Standards: ACall to Action". Anewlyappointed working group on indexing as appliedto archives met in Jlmuary and plans to submit afinal report by the end of March. Other groups tomeet in 1988 will develop descriptive standardsand rules for graphic, sound and moving imagesat the series, file and item level. Now that thearchives format has been integrated into thecanadian MARC Communications Format (newedition available from the National Library ofCanada) and the Canadian Committee on MARC hasaccepted the ACA as a permanent member(represented at present by Hugo Stibbe),descriptive standards activity has an acXIedurgency.

In the U.S., concern expressed in a resolutionpassed by the Description Section at the SAA.annual meeting over the absence of archivaldescriptive standards hasled to the submission ofa proposal to NHPRC for a working group toidentify where and why standards are required,the scope of standards to be developed, criteriaby which to define benefits, and mechanisms formonitoring and changing standards. The proposalasks that the one year project be coordinated byLarry Dowler at Harvard University.

20 Archival Informatics Newsletter vo1.2,·1

SOFTWARE BRIEFSCOLLECTION

MUSEMUSE is described by its creators, Julian M.Humphreys and Barry Chernoff as amicrocomputer collection management databasesystem: It ~as ~veloped for ichthyologicalcoll~tlOns In whIch numerous specimens areacquIred from one locality, taxonomichierarchies are important. and loan managementand label generatlon are a must. It was deslgnedto be used as amultiuser system with largedatabases (up to 300 MB) on PC/AT or 286 classmachines on asoftware foundation OOveloped bySoftcraft Inc. of Austin. TX. MUSE has beenaround for quite awhile as a piece of consultware(essent1ally free but reQu1rlng the serv1ces ofthe creators to make it run). Now it has anewsletter and a users group. Contact theauthors at Cornell University I Ithaca NY 14850.

MARC Catalog Records Data EntryUltracard/Marem and MITINET ImarCr" belong

to an unusual category of software that facilitatesdata entry of MARC records on PC's, but ooesn'tdo much with them (beyond printing cards) oncethey are made. Intended to allow institutions toexport records made off-line from the util1ties,these packages both accommodate all the formats.MITI NET /marCr" has substantial validation and"expert" help built in; Ultracard/marc doesn't.Contoct:Ultracard/Marc: Small Library Computing Inc..619 Mansfield Rd., Willow Grove, PA 19090MITINET /marc: Information Transform Inc.,502 LeonardSt., Madison, WI 53711

Artist AuthorUies wUh Images!

Choowyck-Hooley Inc. has onnouncedpublication of the New York Public LibraryArtists File on microfiche, with a name index tobe published in machine-readable form(presumably on CD-ROM). Details about thedat~ to be cantai ned in the name index are not yetoval lab16: I 5U5pect that Chadwick- Heoley Inc.could be mfluenced by interested potentialbuyers if there was an expressed need for anartists name authority f11e and/or forincorporation of existing artist name files.

Vernon Systems has annolmced severalenhancements to its COLLECT ION system since theinitial september release. Among these are thatall data entry and query screens now have thel'bility to sel'rch via multiple cross referenceindexes, group as well as individual permissionsets may be defined, users may define their owndata fields, and the extended ASCII chara::ter setis supported. The most extensive enhancementis a parameter driven procedural "harness"which permits a local definition of procedureswithin ageneraliZed framework. Specifically,the procedural "harness" is said to permit aninstitution to:1) name a step (and make user notes)2) identify next step permitted3) identify reversal steps permitted4) check for presence or absence of data in adefined field (approval code or prior action)5) check to assure integrity of transaction6) def~ne forms ge~erated by completion of step7) defme other actlOns required by taking step8) update status and action history

.W~ile I haven't seen.this feature. the concept issl~llar to the generahzed action processor Idefl.ned for CMASS (the Collections ManagementActlOn Support System) at the Smithsonian inlogical models developed in 1983/4 but neverimplemented there. It could potentially resolvethe problems created by local proceduraldifferences within categories of actions such asloans, conservation treatment, or acquisitioncommittee review and will definitely be worthexamining closely.

Willoughby Associatos

Willoughby Associates announced the opening ofa new office, staffed by Jane Sunderland, MarcyReed. l'nd Lynn Remington l't 11 61 9 Ohio Ave. ,Penthouse, Los Angeles, CA 90025 (213-444­8994). The ChiC89J office is still reached at312-866- 7996.

Spring. 1988 Copyright by Archives. Museum Informatics 21

STANDARDS

FACETTED CLASSIFICATION IN MARC

At its meeting in December, MARBI, theadVisory committee for the MARC formats, gave agenerally favorable hearing to one of the most farreaching proposals it has ev'er received, Theproposal, by the Art and Architecture Thesaurus(AAT), and supported by the National Library ofMedicine, permits catalogers to assign post­coordinated 1ndexing terms in place of pre­coordinated Library of Congress SUbject HeadingsIn MARC records,

In this way, a cataloger using the AAT, couldconstruct an "expression" (to use some examplesproposed by Murray Waddington), like'construct1vlst arch1tecture' , or 'GothIcrose windows' , or 'Embossed vellumbookbindings' and modify them by ageographical modifier and achronologicalmodifier in the equivalent of an LCSH headIng,while keeping each term component In itsappropriate facet or h1erarchy. Acompletestring, as these headings are called, might be:Modernist Industrial bUl1dlngs - France- Paris. Adaptive reuse.orGothic Churches - Italy. Tympana.Photographs - 20th century,

The, proposed format of such 8 field (tenatfvelydes1gnated as a65x field in MARC) would be:65X

Indicator 1undefined2 thesaurus 1d,

$a focus term$b other term$c facet/hierarchy designation$y chronological term$2 geographical term$2 thesaurus cooo

Aproposed coding pattern would be:SaScSbScSbScS2

but the "focus" term need not occupy the firstposition, as in the example of "Gothic churches"where churches is the focus. Another plausi bIepattern would draw terminology from avarietyof thesauri, and would be coded:$a$o$2$b$c$2$y$c$2$z$c$2

If this proposal is adopted by MARBI and theLibrary of Congress. its initial implementationwould be as avariant or LCSH and other pre­coordinated subject headings, but itsimplications are far more revolutionary. Such asystem w,ould have the benefit of preservingh1erarchlCal placement of each term, thus inprinciple permitting narrowest possible facetassignment but searches for terms on parentvalues, Thus, if I were interested in churches.and aterm "cathedrals" or "cloisters" was afacet of some assigned indexing terms, aretrieval system could be implemented thatsearched down the hierarchy from churches,lrent1f1ed these narrower terms, and retrievedthem together with my request for churches,without rendundant term assignment on the partof the indexer/cataloger, The potential"explosion" of terms provides many more accesspoints for users and, if implemented for both"preferred" and "non-preferred" uses, obviatesthe need to standardize language in index termassignment or in query,

While this proposal st111 has to be approved,the issues it raises for archives and museumsare sufficiently important to warrantintrooucing the possibility here. Comments andsuggestions can be addressed to Toni Petersen atthe AAT, Phyllis Bruns at the Library ofCongress, or Lisa Weber at SM.

GEOGRAPHICAL AUTHORITIES

The J. Paul Getty Trust Art HistoryI~formation Program Vocabulary Control Group,d1rectoo by Eleanor Fink, is exploring thesources of machine-readable geographicalauthority files in order to build on some existingdatabase the internat10nal, h1stor1cal and mult1­disciplinary vocabulary for access to humanistic~ta, ,A, recently completed internal study' has1OOnt1f1OO some plausible bases for such an effortand negotiations are underway to acquire therights to an appropriate foundation. Because noextant databases have the historical cross­references required, or systematically locatesuch humanities goo-concepts as linguisticregions, ethnic groups, literary/mythologiclocations. bullding names etc" much elaborationof this basic file w'ill be requ'ired for it to serveall the needs or art and cultural research.

22 Arch1val Informat1cs Newsletter vo1.26 1

ARCHIVES .The Society of American Archivists has begun

offering aworkshop on "L ibrary DescriptiveStandards: An Introduction for Archivists". Theworkshop is intended to teach archivists aboutthe Anglo-American Cataloging Rules and Libraryof COngress Subject Headings, issues they areconfronting as they implement the MARC:AMCformat In library utl11tles. The workshop w1ll oeoffered for the first time April 29-30 inChicago, and again in Atlanta in september inconj unction with the SM Annual meeting. SM isstill holding its "Understanding the MARC AMCFormat" workshops, w'lth the next onesscheduled for June 2-3 at Old SturbridgeVillage.MA and June 20-21 in Jackson I

Mississippi.

RARE BOOKS & MANUSCRIPTSAvery useful list of all the rare books and

manuscrIpts standards and how to get them hasbeen published by John B. Thomas III. see his"Standards and Guidelines Prepared by the RareBooks and Manuscripts section of the Associationof COllege and Research Libraries", Rare Booksand Manuscripts Librarianship, vo1.2#2, Fall1987p.109-112.

GOVERNMENT RECORDSNICLOG, the Notionollnformation Center forLocal Government Records, is now available to11­free on 1-800-284-5456 to give tldvice onstandard methods for manaoing historicalrecords. NICLOO will also respond to mail sent toNICLOG, AASLH, 172 second Ave. North, Suite102, Noshville, TN 37201

HISTORY MUSEUMSThe "Common Agenda" project, including its taskforce on documenting collections which isdeveloping approaches to definition of standardsfor historical collections is seeking input fromthe profession. Write of call the projectcoordinator, Mary Alexander, at the NationalMuseum of American History, 1'16B-66,Smithsonian Institution, Washington DC 20560,(202) 357-4573 or contact the documentationtosk. group choi r mon, James Blockaby, curator ofthe Mercer and Fonthill Museums. The projectwill report on its progress at the AAM meeting.

LIBRARIESThe Linked Systems Prolect: ANetworking Tool

for L1brarles, Dubl1n OH, OCLC. 1988 Is acomplete history, technical introduction , andapplications/impl1cations analysis of the LSPprotocols (which support computer to computercommunication of MARC data) by leaders in thefield. It should be considered must reading forlibrary and archives technical servicespersonnel.

PUBLISHING & DOCUMENT MARK-UPAlthough they've been aval1able for some time,

I just received copies of three very useful textson SGML by Joan M. Smith. These studies wereconducted for the British Library and publ1shedas British National Bibliography Reports #22,26,8<27. They hove the titles: The StondardGeneraJ1zed Markup Language and Related Issues(#22); The Standard Generalized MarkupLanguage (SGML): Guidelines for Editors andPublishers (#26); The Standard GeneralizedMarkup Language (SGMU: Guidelines forAuthors (#27)

The Guideline volumes are appropriately nittygritty, with all the necessary detalls and lists.The issues volume covers the gamut from history(with adistinctly British slant), through theparticipants, and on to impl1catlons. All threeare recommended reading.

It is increasingly clear that SGML will becomean Important constItuent of all electronic textsand that as text creators and users, it wHl beuseful to understand them even though the actualembeckiing of codes will. increasingly, be oonewithout direct knowledge of the writer, as It is insuch word-processing packages as SoftQuadPublishing Softwar8rM by SoftQuad Inc. [720SpadinaAve.. Toronto, eanem. M5S-2T9].

Those interested in the latest updates shouldconsider attending MarkUp'88, the sixthInternational Conference and Showcase on SGMLsponsored Oy the Graphics CommunlcaUonsAssociation in Ottowa, May 24-26. COntact: GCA,1730 North Lynn 51. 5ulte 604, Arl1ngton, VA22209-2085 or call (703-841-8160),

Spring, 1988 Copyright by Archives & Museum Informatics 23

DIRECTORY OF SOFTWARE FORARCH IVES &MUSEUMS

Archival Informatics Technical ReoortVo1.2, # I, Spring 1988

The Directory of Software for Archives and Museums is organized in threesections with a short introduction to the role of such a directory and its use.

The first section describes systems, in alphabetical order by system name,according to general characteristics such as hardware and software environment,size limits, standards, integration, support, utilities and types of applicationssupported based on information provided by vendors. Pricing and acquisitionoptions, including availability of consultative support and customization, aredescribed.

Section two compares systems by application and sub-systems or utilities.­Application comparisons tables are presented for: cataloging & description.collections management, conservation management, education & interactivetraining, events management, exhibits management, information retrieval.membership/development & fund raising, publications management. recordsscheduling & disposition, space management. travelling exhibits and volunteermanagement. Utility or sub-systems tables are presented for: authority control,data dictionary, data entry, graphics, help, query & retrieval, report writer, screenwriter, security and text editing/word processing. Each table compares datasupported by the system, its functions and features of the application.

Section three indexes the systems by vendor name, hardware and operatingsystems. A bibliography cites selected published reviews of the products anddiscussions of systems selection processes.

Individual copies of technical reports are available from Archives & Museum Informatics. 5600Northumberland St., Plttsburah PA I,ZI7 for USS-t'.prepaid. SUbscriptions to the TechnicalReport and four issues of the Archival Informatics Newsletter are US$160 p.a.. for addresses inthe U,S, & Canada, USS180 p.a, overseas.

24 Arch1val InformaUcs Newsletter