ARCH IVAL INFORMATICS NEWSLETTER€¦ · ARCH IVAL INFORMATICS NEWSLETTER ... bureau and then...

ARCH IVAL INFORMATICSNEWSLETTER

Part 1 of Arch1val Informat1cs

ISSN 0892 2179 SUMMER 1988 Volume 2, ·2

Remedy for a sick marketThe market for archives and museums

outomation has never seemed so robust, yet I fearit is perilously close to death. On the surface, allthe activity we are witnessing 1S evidence ofhealth, why 00 I worry that it is sick?

During the process of compiling the Directoryof Software for Archives and Museums, I foundmyself wondering why so many of the venOOrs ofG'ollection systems to whom I wrote had theirmail returned unforwarded. After completing thepublication I knew the answer - no one bUySarchives and museum collections systems. Thesales figures returned by venoors of membershipand development packages were relat1velyrespectable - 88 to 500 installed. Generalizedinformation retrieval products like STAR,MARCON and INMAGIC also have sales in thehundreds. But the sales of archives and recordsmanagement systems ranged from none to 45 andof museum packages from none to 15 (excludingthe MDA MODES data entry system and the ISISARKS zoo information system which haverelatively captive markets). The mean numberof sales of 6 purely archiVes and recordsmanagement systems was 17 (even allowing forsome healthy exaggeration on the part ofvenoors) and that of the nine specificallymuseum collections systems was 4.5! No vendorcan survive in such amarket.

The consequences are extremely worrisome.Not only are good ideas and SUbstantialinvestments. being wasted in implementat10nsthat don't nnd asufficient market to supportthem. but vendors starved for their next contractwill be unable to provide the kind of support,Qrowth and enhancement that buyers need,investments made by buyers today ar'e morel1kely tl1an not to tie 1n systems that won't tiemarketed two years from now I

Why do we find ourselves in thIs predicament?Some of it is that systems are easier to OOvelopthan they were in the past, and some is thateveryone wants to recoup their inhouseinvestments by offering asystem for sale, butthe largest part of the problem is that archivesand museums have been teasIng the market forseveral years, but aren't buying. Three yearsago they weren't buying because there were fewif any systems made for them and those that werelacked necessary functlonality. This is no longertrue. several of the new systems on the markett.his year do the job and do It well, and all are .within what should be areasonable price range.Archives and museums that aren't buYing noweither: i) have unrealistic expectations,2) aren't really interested. or 3) are scared.

I believe that fear is now amajor factor. Nowthat archiVes and museums finally have choices.they are paralyzed by having not done adequateanalyses of their own requirements to evaluatethe options available. 0

TABLE Of CONTENTS

Articles:Systems at the MM

by David Bearman 26An Experiment with Research Data &.Standards for MRR Reference service

by Thomas E. Brown 33Copyright &. Optical Disk

by Brian Kahin 35RegUlar Features:Conferences 36In-Box 38Projects &. Proposals 44Software briefs 45Standards 47Techn1cal Report Summary 48

Software vendors at the AAMby David Bearman

The exhibit area at the American Association ofMuseums conference attracted a large numb~r ofvendors of automated systems. Each year thenumber of commercially available museumappl1cations being shown increases, and thefeatures of the systems on the market expand andmature. More than ever. buyers need aclear .sense of thelr own reqUlrements or they will fmdthe array of systems competing for their dollarsoverwhe1ming. .

Quite anumber of the systems being shown atthe MM meeting this year were not displayed inprevious years and have not previously beennoted in this Newsletter. I am discussing themmore extensively than usual because they arejust being introduced. By doing so, I do notintend to sl1ght the established vendors in themarket whose products are being enhanced but donot represent brand new realeases. Systemsdisp laved at the MM meeting, but not reviewedhere, included enhanced products from AccessInternational; Blockbaud Microsystems; CuadroAssociates (STAR); Erros Computing (STIPPLE);Master Software Corp. (MasterPiece andFundMaster); and QL Systems (QL VolunteerManagement). With the exception of STI PPLEand aL, which was discussed in this Newsletter inconunction with the MM meeting last year, theseproducts are retailed in the Directory ofSoftware for Archives and Museums, ArchiyalInformatics Technical ReDQrt vo1.2 # " 1988[available for $45 prepaid, $50 bllled fromArchives 8< Museum Informatics].

The fOllOWing CleSCrlptions are based uponremonstrations I viewed at the MM meeting.Therefore, I can not say anything about how wellthese systems might operate with real databasesand multiple users under real conditions. At thesame time, I am not being non-judgemental. Allthe opinions expressed were formed as I observedthe demonstrations and from vendor responses toQuestions I and others asked while I was present.Potential buyers are strongly advised to checkthese systems out for themselves

Delaware CompuUngS8rv1ces [5700Kirkwood Highway, Suite 205, Wilmington DE19808 302-996-0490] introduced acollection management system, written in COBOLfor DEC Micro-vax and HP3000 systems. Thesystem is offered as customized software. builtaround the model system installed for an"unnamed". nationally known art museum InDelaware, to the specifications of its registrarKarol SChmeigel. Athough it is available as a

. turnkey system or software on ly at pres~nt,Delaware will consider operating aservlCebureau and then mIgrating users to their ownsystems when and if their size justifies it.Delaware Computing is a 1981 spin-off ofHewlett-Packard, based in Belgium with annualsales of about $12M and 180 employees. Most ofits income is derived from custom programming,but it also sells 8 varIety oj packages that 1tmooiffes for buyers in the fashion planned forthe museum application. Two full time and onepart time employee currently support the.museum application, including the converSlOn ofdata from cataloging cards using the PalantirCom pound Document Processor character reader.

Access to the system is security controlled byfunction (menu) and separate permissions can begranted to read, read/write/update, andreoo/write/update/relete. Usual systemsmanagement functions ( including audit trails)and escapes Into locally res1dent utll1tles areprovided from the main menu for systemsadministrators, but the screens are hard-coded.

The main file being shown (subject tocustomization for future clients) consists of 30data elements regarding objects and atextual( notes) f1eId that can be as long as 99 screensand searchable by character string includingwilocards. The data is physical description,source and association, and any field may becontrolled by avalue table, constructed eitheroutside of data entry or through the data entryprocess. Measurement conversIon is provldedbetween English and metric l1near measuresonly. SCroon and field specific help is availableand user mooifiabIe; .

Separate records hold locatlon history (andscheduling forward locations), materialcharacteristics, and ClOnor information. Wh11ethese records are notlinked to auser defined orsystem suppl1ed process, accessloning data entry

26 Archival Informatics Newsletter

can return user deflned prompts and reqUlrespecific fields to be present.

Retrieval is a reporting function, rather than aseparate online catalog interface, but it providesfor some enhanced features l'Ike phoneticsearches for oonor names, afour level thesauruswith scope notes, and the facl1ity to define asetby means of aprior search and execute aschoou1ed locatlon change for such aset, so thateXhibits can be planned or conservation actionsscheduled 'involving anumber of objects at a timeWith the system checking to assure theiraval1abl1it.y and make the necessaryreservations. The search itself displays the dataelements and allows the user to identify up tothree values with EO and NE operators for eachelement. ABoolean OR is imputed betweenelements. Only columnar reports are printed inresponse to Querles, but the elements to beprinted and their table location and length areuser specified. Up t.o five levels of sort arepermitted and the reports can be run inforeground or bookground with the Queryprinting out as a separate page on printedreports. search results may be saved in order tonarrow asearch, but not search logic.

On the whole it is avery usable basic systemwith features l1ke whole/part and object/grouprelations and relations to reproductions,multiple former owners, makers or associatedObjects, but wlthout many features of truecollections management systems that would be ofbenefit beyond the registrar's office.

Milestone 1. Inc. [2625 N. Meridian St.,Suite 43; Indianapolis, IN 46208; 317-9264545] refntroouced the COBOL software forWANG VS systems previously marketed byComsolutions, Inc. The partners, Burland Brownand 11m Prendergast, have enhanced the sixmodules (Fund Accounting,l'1embership &.Development, Collection InformationManagement, SChool &. COmmunlty Tour, Class &.Theatre Registration, and Resource center) andare offering them as software packages, withcustomization of screens but not of theunderlying data def'lnitions and architecture.

Fund Accounting supports 99 funds in its basiolmplementatlon, out has Oeen modifiOO for theBaltimore Aquarium to support up to 9999. Thesystem supports traditional accounts payable,

receipts processmg, monthly f1Oanc1al reportmgand chart of OOGOunts maintenance functionsusing separate vendor files and transaction(order/fulf1llmenO files. Report writing to apsuaoo-spreadsheet view is very flexible but.Quite complex. Budgetting is supported bycopying the year's bUdget and modlfylng lt byapplying increases or decreases in funds lines bypercentages or $ amounts. Some standardoutputs, Hke 1099's, are defined.

Membership &. Development is built around theperson file (member, oonor and the collectionmodule person f11e share this). Person recordsinclude an alternate address, aspousal segment.and the usuallnterest code/membership categorydata. Codes are user defined, outside of the M&.Ddata entry prooess. and exercise validation.Some aspects of the system are more rigid thanwe would expect from applicatlons software. Themooule assumes batch processing of membershiprenewals and the Issuance of cards and notices.Similarly, one can calculate payments due onmultiyear gifts, but afixed number of paymentsmust be set up for all gifts rather than anumberdefined inter~tively for each g1ft.

Unfortunately I didn't have t1me to explore theevents management modules, but I did check outthe so-called Collection Information l'1anagementSystem which consists of 99 fixed length (40character) repeating fields. The informationretrieval technique, because Boolean query is notavallable, 1s to l1s1. one argument, obtain aninterim result and refine the list. Accessionnumbers are 16 characters and no special sortlogic is provided. No collectlons managementfunctions or transaction records are provided. Atbest this is avery primitive facility.

Performing Arts Technology [2115 AFourth St., Berkeley CA 9471 0; 415-5486500] demonstrated t.heir EXPLORER System,consisting of seven modules designed for eventsmanagement, membership and development andaccountlng, that run in avery integrated fasahionon the DEC Vax and Micro-vax. The first fourmodules (Walk-up Ticketting; Advanced sales;Group sales; SChool Group Management) providepowerful tools for active oultural repositories tomanage ~misslons, tlCk.eting for special events,scheduling of groups and educational programsfor groups, and the management. of cash registers

Summer. 1988 • by Archives. Museum Informatics 27

and billing systems assocIated with each oT" thesefunctions.

AcombInation aT" ticket printing, Visitorattendance/event preference reporting, and cashregister management, makes the walk-upticketing module especially attractive. Specialdiscounts are calculated basec1 on user definedtypes of vlsltors (seniors, cl111dren,membership cat~ries) as well as on eventcombinations (multiple shows on the same day).Venues are booked as far in OOvance as OOsiredand interactive ticketing assures that spaceaval1able ( Including by seating section Ifdesired) is always known to all cashiers. Eachcash register is monitored separately and cashremovals during the day by managers aresecurity controlled. Users control refundspolicies and the daily "set-up" of all activities(which r.an be r.opled from atemplate andmodified occording to the date). Well conceivedvariables in the da1ly set-up procedure allowmanagers to compare prior experience for anydate ( Iike previous year), the day of week, theweather, and by other variables that might affectattendance. AvIsItor survey data gatheringwlnlilw permits the cashier to administer avisitor questionnaire to tracl< Zip cooes, how thevisitor learned of the program, and any otherinformation OOsired by management or themarketing department. The interface tomembershIp enhances the value of reports onwhen each member visited and what specialevents they attended. In oo:lHion to grossattendance and income figures for the day,standard reports detail all catf9)ries of visitorsby all events, distinguishing each show and eachtheatre or exhibit hall.

The OOvanced sales module adds name andoo:lress recording (for b11llng and ma111ng oftickets) to walk-up ticketing and can imposehandling fees based on per person or pertransaction costs. Group sales streamlineticketing by recordIng characteristics of thegroup (contacts, special needs, lunch plans etc.)and allowing for issuance of either separatetickets or agroup voucher. It also permits therepository to store information about thecomposition of the group (minorities, handicapsetc.) to qual1fy for special funds. The SChoolreservation system permits the institution toschedule groups, In advance, for educational

programs that occur only when so scheduled, andto locate such programs as are scheduled for onegroup, and reserve space for other groups to J01 nthem at the same time. It tracks programs thatr.an be offered, interest levels, facilities Inwhich the programs can take place, andInstructors, and reports on each oj these as wellas issuing the necessary tickets or vouchers. Themuseum or archive can define avariety ofeducational offerings, identify instructors, andoofine appropriate spaces for conducting these, Areservation clerk can then schedule schoolgroups totak:e advantage of these Offerings. Theeducation office can control its schedule byblocking out certain programs, Instructors, orfacilities for given dates, and monitor the actualschedule in OOvance with structured data abouteach group (grade, experience etc.) avai lable tothe Instructor. The specIal requirements of aprogram (such as room set up requirements) canbe taken 'Into consideration in scheduling based onpreviously defined limitations, including suchinteraction effects as timing between programsrequired for set up.

The entire events management function, fromwalk-up ticketing through group reservations,is superbly crafted. One is struck by its elegantsimplicity and power, and by the flexibility itprovloos for users to define their own programswithout sacrificing functionality to supportalmost every conceivable reporting need or thedemands of on-site ticketing.

I did not have time during the AAM meeting toexamine Explorer's Membership &Developmentor Accounting modules except to see suchinterfaces to the ticketing function as on-site,walk-Up membership cards, productIOn offinancial reports on attendance related income,and member Interests tracking made possible byrecording membership number when tickets areISSUed.

Questor Systems Inc. [1005 East COloradoBlvd., PesaOOna, CA 91106; 818-356-0808],was showing two additions to Its Argus system: aSlide Libraries module that makes reports,labels and backing cards for shoos and providesfor check-in/check-out, using bar-codes ifdes1red, and aSite Flies module that prOVides forrecording of authority data about architecturaland archeological sites with appropriate links to

28 Archival Informatics Newsletter vo1.2.#2

the Argus lexicon controlled vocabulary. Questoralso announced the opening of an Eastern regionaloffice at P.O.Box 860, Upper Montclair, NJ07043 (201)-746-7333.

Stelwagon Management Systems Inc. [114Forest Avenue, Narbeth, PA19072; 215-6676776] debuted a museum system consisting offour modules: Account1ng, Development, on-l1neReservations, and Curatorial Management,running on an IBM System 36in COBOL and RP6.St8lwagon w111 sell a tur"nkey system or softwareonly, and expects to custom1ze the product for~h cHent. Company President, Jack Stelw~n,

emphasizes integration and shared systems as these111ng point and promises to bid aggressivelywhile declining to suooest any specific prices. Ihoo an opportunity to review the Developmentand Curator1al man8l'J3ment modules only, sothere was little opportunity to verify theintegration of the parts around ashared database.

Development security controlls exercised at aFile, Menu and Field level, but from a userprofile that must be set for each individual sofield level security overheadS may proveprohibitive. The donor record consists of 640bytes in flXed length fields, stored as fixed lengthdata, but aview of the donor record alsoincorporates all transactions associated witheach donor from separate, short. transactionrecords. Codes may be recorded for ma111ngs andcampaigns in the transactions records, allowingreports on oonat1ons associated with speciflcdrives, but there ls no facility for setting up acampaign and managing it as aseparate function.61ft receipts and oonor history reports are theworkhorses of the Development system. otherreports can be written using Executrieve, arelatively straightforward Query language, andStelwagon has made an easy query screen withlimited templates for one client.

Curatorial management provides an objectrecord 11nked to an artist (crootor) authorityrecord and transaction records for exhibition andpreservation historles. No part/whole relationsare reflected in the object record. The catalogueentry field is tied only to atext editor and thereis no query yet available from within thecurator1al management module, althoughExecutrieve can be invoked from outsioo. N:;cesspoints (subjects) are 24 characters in length

and strIng searches are planned. Exh1b1tl0n,preservation and bibliographic references areall text fields without associated functionalityThe system clearly has a long way to go before itsDevelopment and Curatoriall'1anagement (readcatalog) functions begin to be competitive withothers In the market, butthe firm was launchedwith accounting and reservations as itsexpert1se, so I may have missed its strong points.

The WHl1amson Group [ 129 Mount AuburnSt., CambrickJe, MA 02138; 800-992-6848)demonstrated the latest release of ARTIS Collections Management which permits users tooofine acustom data entry screen consisting offields they neeed in the order they need themrather than requiring them to use the ninescreens and see a11145 data elements as in thepast. Th1s is asUbstant1al1mprovement,however, these screensare Hmited to 15 lineseach and must therefore be joined to build a fulldata entry screen. Also, each screens is namedby asingle character, severely l'Imiting thenumber of processes that can be customized andthe ease of use. other enhancements include theab111ty to browse up a 11st w1thout !J)lng out ofthe function, user defined one l1ne displays forhits, user defined 1oan/exhibitioh data records,and true transaction based loon/exhibitionhistory. Now users can set up the loans for aneXhib1t., copy the record and change all relevantdata once when the exhibit travels to 8 new site,rather than having to update each loaned itemwith a new venue.

ARTIS still records all textual data in wordprocessing documents pointed to by name fromthe field to wh1ch they apply. Not only can'tusers invoke the word processing documentdirectly, but calling the document in WP puts theuser outside the ARTIS system, from which hemust log back on and into the record that wasbeing updated. This limitation is so severe thatthe improvements demonstrated at thiS meeting,although substantial, seem insufficient.

Vernon Systems ltd., [P.O. 6909, Auckland,New Zealand; tel, 649-395-230] attended itsfirst AAM meeting and gave me afull daydemonstration of COLLECTION. Recognizing that ademonstration is always less exacting for asystem than 8 hands-on review and that it

Summer. 1988 • by Archives It Museum Informatics 29

emplOYed astngle workstation on a small testdata set provided by Vernon (and cannottherefore tell us about performance), Iwasimmensely impressed. COLLECTION is, verysimply I an elegantly designed system and its"procedural control" features provide it with themost powerful collections managementcapabilities of any system on the market. Myonly serious criticism of COLLECTION is that ittries to do too mUCh, but except for the demandthat places on sophisticated users, this criticismshould be read as praise, since COLLECTION verylargely succee<Js In every case.

After a security controlled (encrypted) log-on,the maIn menu can Identify up to 10 functIons.beginning with Entry/Acquisition, andproc.eeding through cataloging to LoanProcessing. COnservation. Exhlbfts, MovementCOntrol, Generallnqutry and Reports, and finallythe Systems Management and Adm inistrativefunctions. Anicely conceived prompt identit"1edthe next level of the menu at the cursor position.Help is avaj)able both for eoch menu and everyfield, and may be fully tailored by the user. ESCconsIstently enos any functIon and returns theuser to the main menu. Any process can besuspended at any t1me to PERFORM anotheractivity and will be recommenced where it wasleft off after the ootour Is completed.

ACQUISITION: The Entry/Acquisition processdisplays the full power of the system. Any ffeldmay be set up to have a value table; the table mayrestrict entries to terms on the list, or it may be~ to by entries in the Held, and then mOOs tocontrol the new values after it is completed.Value tables may be sImple term I1sts or morecomplex authority files with full name, displayname, notes, broader/narrower relatIonships,use for and use instead indicators, related terms.and status of term. The hierarchical relationsare shown when the data Is entered. so thatdistinctions between ParIS, Virginia, UnttedStates/North America and Paris/France/Europeare Immediately evident. However, since the datais stored 'in only one place. no aciled space isoccupied by the additIOnal information, no matterhow many levels of hierarchy (as in biologicaltaxonomy when thIS can be many) are employed.Some of the implemented authority files arelncredlbly powerful; the person f1le

demonstrates complete deSCendant/ancestorlinkages that display a full family tree.automat1cally making reciprocal updates (chlldrecords generated when mother 1s defined, forInstance). Inaddition to such inherited dataaugmentation, data may be carried forward fromprIor entrIes, or defaulted to a template.

CATALOGING: The cataloging function from thedemonstration menu provides for cataloging ofobjects from 15 disciplines (archeology,costume, decorative art, ethnography, etc.) and11 areas of Interest (1dent1flcation deta1Js,proouction details. field collection ootal1s.association, provenance etc.). These screens aregeneric templates intended to be customized forbuyers of the system according to the nature oftheir collections. It is unlikely anyone wouldhave all 26 SUb-menus.catal~ing ltself can support very complex

descriptions. In addition to the eleven detailed"areas of Interest", a hierarchical structure canbe used to oofine groups of objects, withcomponents of each object and fragments ofobjects as well as pIeces of components. Themultiple independent parenthood of differentparts allows properties of the Whole group, theobject. or acomponent to be inherlted by allpieces. In the oomonstration, Iestabllshed akeyholder with keys as an object, each key andthe I<eyholder as separate components, andassigned the object to the groups of my collectionand of household objects, the keys to agroup ofmetalwork and the holder to agroup ofleatherwork. The characteristics Idescribed foreach group (provenance for my collection,materIal for the metal and leather assIgnments)carried appropriately to the object. its parts andtheir fragments or pleces.

Vernon Systems has developed a rich variety ofdating schemes perm1tting objects to be catalogedby a single date or adate range to any level ofprecIsion (year, month, day) as well as uSing thenames of seasons. a number of years/months ordays ago, decades, centuries, circa and suchapproximation qualifiers as pre-; post-; early-;late. All these dating methods are transformedwithin the system to aconsistent internal formatthat can be searched (In general thls 1S veryelegant, but ranges are reflected as their midpoints which 1s not Quite right).


In catalogmg, groups of fields can be enteredwith alternative attributions, allowing thecatalog to reflect different scholar ly opmiOns.Linear and volumetric measures are convertedfrom English to metric and back. Link.s can bemade to other identification numbers, toreproductions (both images and moools) and toassociation histories. And true transactionalrecords are created for all exhibit, loan,conservation, and movement history and can bemade for any other named action.

PROCEDURAL CONTROL: The exceptionalinnovation in this system is called ProceduralControl. Procedural control enables anyrepository to define a process, for instanceacquisition, by specifying its steps and therequirements of each step both in terms of dataand reports. In practice this means that Harepository uses an existing form and procedureit may specify thatcertain data (the form) mustbe completed before it is forwarded to theacqulsition committee for review. and that thecommittee wishes to see the data in acertainformat. This report will then be completed bythe system. awaiting action by Hie committee.When the result of the commlttee deliberations isentered. the next step is defined as notifyingdonors let.t.ers will be written merging datafrom the committee decision and acquisitioninformation. inclUding completed gift agreementforms with donor terms specified. Upon returnof t.he forms. the repository could have furtherprocesses scheduled. Reports of costs inacQuisition would be generated automatically tothe accounting office and of insurance needs to theinsurance off1ce. etc. The entire process iscompletely at the discretion of each repositoryby completlng a Procedure COnt.rol definitionconsisting of: 1) naming the procedure.2) identifying next steps. 3) identifying steps toreverse to. 4) identifying data to be checked attM step and correct1ve actlons if data is missing.and 5) identifying reports to be sent and wr1tingany special calculations to be conducted.

Other aspects of the system also help supportcollections management. locations are attachedto time of day. not just date. so that many movesin the same day wlll record and display 1n order.Lists can be constructed by asearch. and theentire list may then be attached to aprocess. As a

resUlt, Exhi bltlons can be set up. cata10gl ng datafor objects 1n the exhibit imported. labels andcatalogues deflned, packlng mstructlons attached.and the whole can be moved through multiplevenues, shlppers, Insurers. etc. by rekeymgonly the elements of information that change.The entire history is retained. separate steps incomplex procedures. such as conservationactions. may be scheduled In advance to takeadvantage of equipment and people available. andwhen conducted can be recorded en masse.

RETRIEVAL: Finally. but not least, COllectIonsupports full Boolean Queries. building largeexpressions and executing them rapidly. The sets that are retrieved may be saved and researched.but at the moment the expression itself cannot besaved and the iriter1m display of number of hitsper term disappears from the screen too rapIdlyto be used. In image searching. Collection uses aspec1al application of relatlonsh1ps to getassociated biographical information for"portraits of X" while retrieving approprIategeographical data for "landscapes of Y", Thel1nkagesmade 1n this kind of secondary searchroutine are user defined. and thus extensible.

The ent1re system comes complete with thesystems management and adm Inistrative toolsused in its own const.r·uction. giving thesophisticated user a powerful systemsdevelopment environment restricted only 1n itsability to unoo COLLECTION itself. Indeed. the.very power and complexity of COLLECTIONprevents me from recommending it withoutcaveats. Vernon Systems is asmall( 3 person),young (2 yr old) firm in New Zealand (far away)that has yet to make asale. The enthusiasm of thefounders and their evident skills. are not asubstitute for support. whichwill be hard to getunless they open a U.S. office. An IBM PC/386running Revelation under DOS 1s not a largesystem, although it will support multiple users;only benchmarks With a real database will g1veus an idea of how large adatabase it will support.Revelation 1s asuccessful product, but aminoractor in the PC dbms market; its longevlty is anopen Question, Which adds up to COLLECTIONbeing asuperb system for asophisticated userable to assess the risks and provide support.


WI l10ughby AssocIates Inc. [266 limEn St..Winnetka, Il60093; 312-501-4540]previewed their about to be released QUIXIScollections management system. This multi-usersystem, running on an HP3000, is to beWllJoughby's top-of-the-line system. Liketheir PC product MltlSY I QUIXIS is I1censed as asoftware package. without need for customizatfonand with all its capabflities incluOOd for oneprice. Willoughby continues to be willing to bidthe conversion of manual files into its system andencourages new clients to compare the economiesof entering the data themselves wlth those ofcontrocted input.

ARCHITECTURE: QUIXIS is built around anob.lect me that supports object groups and parts.separate records for associations. attributionsand loan/exh1blt transootlons permlt multipleindependenty occurrences of sets of referencedata and object histories. other fields are allvariable in length and occurrence and may beindexed to separate or combined indexes withoutlimit. Rights and reproduction data is drivenfrom aseparate menu to maIntaIn lts securityand an even more limited menu (visiting scholarsystem) is provlOOd for view only access to aportion of the database. Some features of thesystem reflect Willoughby's long experience inmuseums:* the accession number 1s built from four dataelements which can have independent sort rulesso that all the different accessioning systemswithin one institution over many years can beaccommodated and properly sorted;* similar segmentation of locations into sixfields With anote allows the most complexlocations to be acted upon in groups* attr1butlOns records for people andorganizations are full reference databasescomplete with ageneralized foc111ty forrecording events in the life of the indIvidual ororgan1zatlon and relationships or roles.W1110ughby intends to distribute lts system toart museums wah aconsiderable number ofthese records already built. This w1l1 be agreatadvantage to them w'ithout bindlng them lnto anyparticular form of name since the user elects theform to be used and the alternates.

RETRIEVAL: QUIXIS w111improve on MIMSY'sQUE (Query, Update, Entry) screens, using thesame concept to allow for integrated access andentry. but providing both interim hit counts andfull search expression hit counts, Text fields mayb..e searched by character str'ing. With delimitedK8ywords, or for adjacency. with or withoutstoplists. Wildcard searches are fully supportedas is the full range of Boolean expressions. Oneparticularly nice feature is the ability to define adisplay format for records retrieved in aset andto then view adifferent display for the same set.SO that aquery for records having someintellectual charocteristic could display abriefname, title, SUbject, curatorial departmententry. The user ccan then switch to the locationview to see what is on exhibit and where.

COLLECTIONS MANAGEMENT: Although historytracking is supported. no generic actionscapabllity is provided, only loans and exhIbits,Nevertheless. QUIXIS is atrue collectionsmanagement system , not just aspeciallzedinformation retrieval system. Routing for loansand eXh1b1ts. including specIfic flights,Insurances, and packing materials (includingcost data), is supported. Although an on-linetickler, person specific and sensitive to referredoctlons/authorization is stlll lacking at this stageof development. QUIXIS will notify users oftoday's scheduled collections management tasks atlog-on and could notify them at log-off ofincomplete tasks. W1110ughby promises that thesystem w111 come complete with over 200defined reports; those I have reviewed areexceptionally well conceived, reflecting the longexperience of the prInCipals in museums and theuse of their prior systems. MILAM II and MIMSY.QUIXIS which will be avaflable for installation InAugust has been worth walting for.

[)

32 Archival Informatics Newsletter vo1.2. #2

MACHINE-READABLE VIEWS

[The editor wishes to congratulate Tom Brown onhis election as the incoming President of IASSIST.Correspondence regarding this column should beaddressed to hIm at: Administrative Staff, Officeof the National Archives, Washington DC 20408]

An Experiment with R8988rch Data

At the recent conference of the InternationalAssociation for Social SCience Informationservice and Technology (IASSIST), a boldexperiment 1n the preservation of empricaleconomic research data was unveiled. Inempirir.al economics, a researcher acquires 8primary data fi Ie or files from adata repository.Analysis of the information frequently Involvesrecoding values, collapsing variables, inputtinginformation, creating indexes, and alteringvalues for certain variables. This information Iscalled derIvative data because It has been derivedfrom the primary data sources. The researcheranalyzes this derivative data to generate the flnalstatistics that appear in publication. Whileoccess to the derived data is critical to validatingthe methorology, this derived data has not beengenerally availab Ie.

To remroy this, the Economics Program of theNational SCience Foundation has asked allgrantees to indicate the public availability ofderived data in a footnote each time they publ1shempirical research. In addition. journal editorsare being encouraged to adopt editorial policiesrequiring placement of derived data in publicdepositorIes prior to publication. As afinalinducement to researchers, the Inter-UniversityConsortium for Political and Social Research( ICPSR) has agreed to accept deposit ofderivative data. This effort to place derivativedata in an archival repository raises some ..interesting issues. As would be expected. avariety of Questions remain to be answered.

The first broad issue concerns archivalappraisal. The predominant concern in theappraisal of automated records systems hastraditionally been informational value.However, derlvatlve data Is beIng deposited withICPSR in order to permit analysis of themethodology behind a published article. Thus itis being acquired for its evidential value or

evidence of the valtd1tv of the research process.If such acquisitions extends beyond the socialsciences, to include also natural science data, theevidence becomes invaluable as asource for thehistory of science. This argument suggests thevalue of some of the derived data, but not all of it.Yet ICSPR is currently accepting, alleasttemporarily, all derived data sets submitted.Clearly, one complex appraisal Question is howlong to Keep what data. One possible answer is tosave data related to seminal publications.Unfortunately the characterization of aparticular stuay as "seminal" is aJUdgment thatevolves over time. Another possibility is to saveon Iy those derivative data sets wtth highreference demand. Conceivably, a repositorycould accept all derivative data sets; anything notused within acertain number of years would beremoved from the holdings to allow for theacquisition of new data sets. While reference usefrequently reflects the informational value of tMrecords, archivists have traditionally beenreluctant to base appraisal judgements oncurrent reference trends.

The second broad Issue concerns the extent ofprocessing. ICPSR has developed four classes todescribe the level of processing a primary dataset receives. In aXii tion I acquisitions areannounced in the Quarterly ICPSR Bulletin, andincorporated into the annual Guide to Besourecesand services. Such descriptive efforts prOVidefor intellectual control associated withprocessing an accession. For derivative data,ICPSR has adOOd a fifth class (Class V). Thisclass is for data sets or related documentationthat the repository has decided not to validate.describe or otherwise process. At the presenttime, most of the derivative data is received onmicrocomputer disketttes containing both thedata and the technical documentation. Thedepository is acquiring acollection of diskettesbut has little or no knowledl}l of their technicalspecifications or intellectual content. Given thislack of processing, the derivative data seems tobe at the periphery of ICPSR's archivalcollection.

Finally, avariety of Questions concerning theresearchers and reference service remainunresolved. Because of the lack of descriptivework, the only public indication that ICPSR hasacquired the derivative data is afootnote in the


r:~t,

f.

puOllshed artIcle. ThIs Is clearly In lIne wlththe major objective of the program J namely topermit valIdation of published results, howeverresearchers will have difficulty identifying andlocating derivative data sets generated from thesame primary data for compara!ive stucty.Because of the lack of processIng and valIdation,researchers are also likely to encounterproblems in using the data. Researchers seloomhave the time or resources to organIze the dataand documentation to make it rea11ly usable byothers. Thus users of derivative data can expectdata dIscrepancIes, and Incomplete or evenerroneous oocumentatlon to interfere with theirefforts to verIfy methodologies.

Certainly as the program unfolds, answers willbe found to most of these unresolved issues. Theanswers may result inunforeseen innovations indata archives and data 11brar1es. Derivative datafrom empirical economic research offers a newtype of information: the archival management ofit can easily involve the data repositories withother disciplines in the social and naturalsciences that generate derivative data. The lack ofprocessing might lead to a re-evaluation of theessential elements in acquisition of data files forarchival custocty. Asimilar re-evaluation couldalso focus on reference services. Thus theexperiment may have benefits beyond theimmediate goal of providing access to derivativeeconomic data to validate published researchresults. [J Thomas E. Brown D

Standards for MRR Reference SBrv1ceFor the past two decades, most repositories

administering data files for research use haveprovided their users with technical dooumentatlon and acopy of the data on tape for users tomanipulate and analyze on computers to whichthey have access. This approach to referenceservice arose when the only way to analyze suchdata was by batch processing on mainframes.Technology now offers us other options.

The (pal of reference service for data inelectronic form should be to prOVide no less thanminimally acceptable reference service. In thepartnership between the researcher and thearchivist, 1t is the researcher who initiallydefInes criteria for evaluatIng services. The

criterIa that are typICally cited are thatreference services should provide informationabout or from the records: 1) where It is wanted,2) when it is wanted, and 3) in the format inwhich itls wanted, Reference services could,therefore. be,evaluated according to thesecrIteria on acontinuum from unacceptable tofully acceptable, as represented by the followingmatrix:

WHERE WHEN HOWFully Acceptable

Min. Acceptable

Unacceptab1e

In reference service for machine-readablerecords, each element has a range of options:WHERE: at only one locatlon

at multiple locations within theinstitution

at selected institutionsanywhere the researcher des1res

WHEN: Receivedon ly during ceratain times of daytwenty-four hours 6 day.E1lledwithin seconds of receiving the requestsame day servicewithin ten work ing days

FORMAT exact tape copy of the original recordtape copy to user specificationsnew format as requestedon-line access

To the researcher, if a repository cannotprovide data when, where, and how it is needed(one element is unacceptable), the whoIe isunacceptable. Thus the archives must meet atleast the minimum acceptable standards in allelements if it is to provide minimally acceptablereference service. An archives may determinethat, due to other considerations, it will not meetthe researcher defied minimum for one or moreelements. In this situation, Ibelieve thearchives should not provide reference, even ifthe other elements meet researcher definedminimally acceptable standards. Rather, 1tshould propose an alternative reference service


that could sattsfy all three user definedrequ jrements.

Archivists should be aware that what userswill consider unacceptable, minimallyacceptab Ie or fu lly acceptabIe is affected by theoriginal records and the nature of research use.Thus appraisal and accesslonlng of mochlnereadable records must include establishing areference strategy for the record.

[) Thomas E. Brown (]

Fa1r Use of Electronlc ArchlvesOptical storage technologies offer remarkable

opportunities for providing efficient access totext and im8{13S. Information stored in digitalform, whether primary text or image codes, canbe located by computers with extraordinaryspeed. By rel1evlng scholars from tedlum andrepetition. the computer increases productivityand allows analySts much closer to the natural"speed of thought...

However, the process of recording informationon optical disk necessarily involves copying,typically by scanning (in the case of text) oranalog recording (in the case of im~ forvideo:Hsc). While it is widely assumed thatcopying onto optical media for archival purposesis fair usa under Sect.ion 107 of the CopyrightAct, there is no consensus on how far the fair useprlnciple extends or should extend.

For example, once an archival optical disk iscreated, how can tt be made aval1abIe to thescholar ly community. Can the disk be offered forcirculation by mail? -- or is that aviolation ofthe distribution right? An owned copy cannormally be dlsplayed publicly at the placewhere it is located (Section t09(d», but ooesthat principle apply to acopy created under fairuse? [Consider the same issue in adifferentcontext: Is a time-shifted television recordingpermitted under the Betamax case law an "ownedcopy"? If so, it can be sold under sectlon109(a»).

can the archlVal copy ever by COPled -- 1nwhole or in part? Can ascholar oownloadportions for his or her prlVate use? can the disk.be mount9d on a localar98 network or a bulletinboard for access by more than one user? can ltbe mounted on Bltnet or other wide areanet.works, so it can be made available to

researchers In other locatIOns? What tf theaval1abi1tty for remote use is publ1cized?

In short, the more accessible the archive is,the more likely it is that fair use principlescannot be stretched to cover the situation.Securing permission from copyright owners is aprohibItively burdensome task for archlvalreproduction in the first instance. The morepublication paths are contemplated, and the morevalue to more users, the more impossible itbecomes.

There is virtually no case law on theapplicability of fall' use to archival copylng orpublie display. However, considerable funds canbe expended to develop mochine-readablearchives before it becomes apparent that whatmakes these archives valuable may also renderthem useless. Unpublished material, even if it isof very narrow lnterest, presents spec1aldifficulties because the authors may not wanteven l1mited publication.

Fair use guidelines have been developed inseveral other areas. Guidelines for libraryphotocopying have been incoprporated into theCOpyright Act as sect10n 108. Guidelines forphotooopying and off-air taping for classroomuse have been developed by representatives of theeducation community and copyright oweners. Thelatter gUidelines, although lacking the force oflaw, have been enoorsed by the House Committeethat deals with copyrlght.

Some initial assessment is called for todetermine the scope of the problem. To whatextent is the development of machine readablearchives inhibited by the copyright problem? Towhat extent is the problam baing addressed byexisting archives. How is it being ao:Jressed? Isthere sufficient consensus to do anything aboutthe problem? What are possible solutions? Theauthor invites reaoors to share their reactionsand experiences with h'im.

[] Brian Kahin. J.D. [)[Research Aff'lliate, Research Program onCommunications Policy, M.I.T., 254 ConcordAve, cambridge, MA 02138J

Summer. 1988 • by Archives & Museum Informatics 35

CONFERENCESAMERICAN SOCIETY OF INDEXERS. 20thAnniversary meeting. May 13

The ASI meeting was an intensive I day longsession, held in conjunction with the On-l inemeeting in New York. The opening talk was givenby Hans WeJlisch on the 'Literature of Indexing',followed by Dorothy Thomas on 'Book IndexingPrinciples and Standards', These exposed aweakliterature with inadequate and conflictingstandards; Ms. Thomas argued that 2 of the 9criteria advanced by the Chicago Manual of Style(the limIt of 3 sub-heOOlngs anl1 the deletion of"trivial" items from an index) are plain wrongand noted that the major texts in the field and theNISO standard 2.39.4 all use differentterminolcgy for the basic elements of an index,Needless to sery, she cal led for oction.

L1nda Fetters updated her list of softwarecurrently available on PCS for indexing andsuggested atypolcgy consisting of two maincategories: "Indexer Controlled" and "SoftwareCOntrolled", Indexer controlled systems includestandalone, imbedded and hypertext systems inwt11Ch the lnl1exer selects terms; tt10se In whichthe software selects terms include text retrieval,automatic indexing and AI systems. She suggestedthat these systems can help do the clerical workof indexing, are raising the consciousness ofpublishers to the subtleties of~ indexing, andprOVide arealm for thesaurus development.

The impact of new distribution media, bothonline data bases and CD-ROM was atopicad:lressed by many speakers. Barbara Preschelof the Publfc Affairs Information serVice (PAIS)spoke to it directly, As an axiom, she stated thatIndeXing methods must be dictated by ~he. methOdof retrieval, therefore, the charocterlstJcs ofprint and CD-ROM were contrasted. Printassumesbrowsability; this could be provided inthe CD or online, but rarely is, Print assumesone view; CD or on line may support numerousviews, In print indexes, there 1s af1xed list ofheadings whose depth and specif1clty is limited;in CD the depth and specificity and number ofheadings are open-ended. In print, the index isassumed to be in proxImity to the text; online it

usually isn't but m CD's it can be even closerPrint indexes are always pre-coordinated; inonline and CD envlronments they can be postcoordinated. Print indexes employ naturallanguage from the texts; in CD's, acontrolledvocabulary may be required to provideconSistency given the potential size of thedatabase.

Everett Brenner I apioneer of abstracting andindexing, reviewed the history of the use ofcomputers 'in indexing from the 1950's to thepresent and concluded that controlledvocabularfes and thesauri have fa1Jed as asubstitute for natural language but so has fulltext retrieval and called for research on how tomake retrieval work, rather than making morethesauri. Like Bruce Croft who spoke onautomatic indexing later in the program,Brenner saw promise In the automatic indexmgtechniques he and his colleagues have used at theAPI and proposed similar analysis of the Queriesprovided by researchers.

James Anderson and Tefko saracevic of RutgersUniversity presented research frameworks foranalysis of retrieval in which It 15 evlclent thatthe number of variables are such that we candraw few large conclusions other thanestablishing, as saracevic and his colleagues doin the Journal of ASIS, that it doesn't work well,Anderson did provide some guidelines on fileorganization and display, using library f11ingrules as an example of "pernicious display".saracevic focussed on the results of their studywhich showed that when multiple searchersretrieve the same items they are more likely tobe ju();jed relevant by users, suggesting that ifmoney Is no object I we should hire second, thirdand fourth opinions in our database searches.

Ben Ami L1petz rounded up the conference Withaprovocative assessment of the usefulness ofindexes. He has concluded, from an analysis ofbook reviews, that indexes don't matter to eventhe most critically Informed readers, and hesuggested that back of the book indexes won'ttherefore matter to pubJishers unti I someonecreates aproduct to cumulate published indexes,which serves the interests of publishers byselling sections of books (on-demand printing)to users who now search only the periodicalliterature. Perhaps the back of the book index


has afuture as a free standing informationproduct

I came 8WtfY impressed bv how little explicit('.onsideration indexers have given toimplementation environments in the traditionalfocus on the terms, the structure of indexes andthe techniques of index construction. BarbaraPreschel, who came closest. ignored the fact thatCD's carry their own retrieval software. JimAnderson noted that agood system makes its rulesapparent to users and that pernicious systemshide them, and Ev Brenner tried to turn attentionto the WeN the mdex works in practice ratherthan the vocabularies, but the meeting as aWholenever ('.ame to grips with it. Probably we assumethat we all know how printed indexes work andforget that retrieval is system dependent 0

AMERICAN SOCIETY FOR INFORMATIONSCIENCE. mld-year meet1ng. May 15-18

The ASIS mid-year conference on ArtificialIntel11gence: Expert Systems and OtherApplications demonstrated how routine AI rl6Sbecome mthe past few veal's. Speaker afterspeaker presented real systems, smartersystems that. used AI concept.s. but, they stressed,these systems were just plain old programming.Each speaker seemed to be engaged tn asystematic demystif1cation of this deliciousmystery, in part to explain why theirapplication was not smarter and in part tomdlcate how AI can be appropriated by anysystem. As Mitchell Waldrop stated in hisplenary addess, the most important influences of,ll.! w11l be hidden in programs that operate asintellloent secretaries. off1ce manaoors, travelagents and Ule like. .

Iwas unable t.o attend the entire meeting, but Ileft with strong impressions of afew "nice littleIdeas", Chuck. Goldst6ms' presentation of usercordial (user specific) interfaces written withthe Borland Turbo Prolog ($160) tool set that'imitated the Grateful Med system developed inexpensive mainframe environments with manyman-years of programming, not only sugestedthat such personalized systems can indeed bemade and wHl inooed be with us soon, butunderscored the democratization of AI. Goldstein

showed that If ahandful of production rules aredefined for each database and aseparate set foreach utility. users can exploit the primitives of ageneralized envIronment to conduct searches onnumerous systems and combine them usefullv,while maintaining their own personal interactionWith their local SYstem.

Everett Brenner described the iterativeimprovement of the Machine Automated IndexingSystem implemented at the American PetroleumInstitute since 1982. Initially implemented toassist indexers by proposing terms for theirconsideration, It has now displaced them. Thesystem began with only 40~ hits and an equalnumber of noise terms but has beenprogressively reflned and now achieves over70% good assignments with little noise, It ISnoteworthy, especially to those who equate termselection with professionalism, the PC AT basedsystem contains painfully few rules.

Martin D1llon reported on work underway atDele to weigh title terms according to theirscatter across LC classifications in order toprovide more strategic retrieval when usingkeywords from titles. The results suggest thatphrase extraction and matching against such astructure can vast.ly improve full text retrieval.

I chaired asession in which Toni Petersenrelat.ed the development. of t.he Art andArchitecture. Thesaurus to the problems ofsemantic analysis and Pat Molholt representedthe state.ments that can be made using thecontrolled MT vocabulary 1n natural languagephrases. The concept that the multi-faceted MTis alanguage and that its helrarchies are tr18grammar of adisciplinary discourse leads toexciting 'Implementation ideas, Pat Molholt, forexample, analyzed the types of relations otherthan "Whole-part" or "is-a" that are representedin natural language statements mapped to the AATand considered the opportunitles we have tooper-ate on these extended relationship links. 0

Readers will note that the editor attends manyconferences. He will be attending all or part ofthe conferences starred 1n the list on the nextpage, however, he welcomes any readers topropose themselves as reporters for any of thesemeetings and/or to suggest other meetings ofint.erest to readershio of this newsletter.


"-I!,i

'I

i1

CONFERENCE SCHEDULES

July 20-23, 1988; Natlonal Assoclatlon ofGovernment Archivists and RecordsAdmimstrators, Annapolis, MD *[Steve Cooper, Maryland State Archives, 350Rowe Blvd. ,Annapol1s MD 21401 ]

sept. 14-17, American Association for State andLocal History I Annual meeting, Rochester NY[AASLH I I 72 Second St. Suite 102, NashVIlle TN37201 J J

september 21-24,1988; InternationalConference on Terminology for Museums *[sponsored.iointly by the Museum DocumentationAssociation and the Getty Art History InformationProgram I cambridge, England; Contoct: MDABuilding 0,347 Cherry Hinton Rd., cambri~CB 14DH ENGLAND]

sept. 29-OCt 2, Society of American Archivists,Annual Meeting, Atlanta *[SM, 600 Federal S1., Suite 504, ChicagJ, 1160605)

october 3-6. 1988, Association of RecordsManagers and Adm inistrators, Baltimore[Cont~t: ARMA, 4200 Somerset. Suite 215,Prairie Village, KS 66208]

OCtober 2-6. 1988; Library and InformationTechnology Assocl(~tlOnJ Boston *[ lITA, 50 E. Huron St., Chicago, IL 60611]

OCtober 4-7, 1988; International Council onArchives Meeting of Experts on DescriptiveStandards. Ottowa. canada *

OCtober 23-27; American Society for Information Science. Annual Conference, Atlanta GA *[ASIS, 1424 16th St., NW JSuite 404,Washington, DC 20036]

OCtober 26-28. Museum Computer NetworkAnnual Meetlng, Santa Monica CA *[MeN Conference, 5600 Northumberland St. ,Pittsburgh, PA 15217]

IN- BOX

REPORTS

DevelopIng Deser1pUve Standards: A Callto Action; Jean Dryden & Kent Haworth for thePlanning Committee on Descriptive Standards,Bureau of canadian Archivists. occasional Paper# 1, Ottowa ,Bureau of Canadian Archivists, 1987

As the title implies, this report is acall toaction, pointing to benefits to be gained fromstandards and sU9'J9Stlng how and why thecanadian professional community should proceed.

Informat1on Architecture ProgressReport; Rae Thompson, Smithsonian InstitutionOffice of Information Resource Management, May'1988 67p, plus lengthy appendexes.

Although this is an interim draft of aworkingl))Cument 1n an on-goIng proJect, th1s progressreport on the Information architecture of acomplex of museums and on the process employedby users and systems analysts at the Smithsonianfor defining an information architecture isexceptionally valuable for anyone involved inmodelling information systems for culturalreposHories. By deftnHion, Informationarchitectures are not directly transferabIe toother institutions, and even the process ofdefining them wi11look different in othersettings, but the progress report has much toteach nevertheless.

Public Record Office Optical DiskProject: Interim [valuation Report,London, Public Record Office ,May 1988 .

In March 1987 the Public Record Office of theU.K. launched ap110t project to explores thefeasibility of storing its data archlVe on optlCaldisk due to the costs of controlled environmentsand recopying of m~netic tapes. The objectiVesof the pilot are to: 1) test the suitab'ility ofoptical disk as an archlval medium; 2) explorethe implication of public access to computerrecords; 3) establish astandard of documentatlOnfor computer records transferred to the PRO; 4)assess the costs of est~blishing and running anoptIcal disk archive, and 5) share their findings.The pro1ect was conflned to traditionally


structured dlgltal data, and dld not lnclude freetext I sound or image.

The interim report for November 1987 toApr i1 1988 represents the ear1iest results of theproject (following training). The PRO isencouraged by costs and ease of use. Still to be~onducted are tests of the medium longevity,mcludmg one using optical disks that have beenstored both vertlcally and horizontally onregular shelves in the paper storage stOOKs sinceDecember 1987. The only serious issuesencountered to date are that the PRO was initiallygetting only 65% of the promlsed IGB storageper disK. Their vendor, Data General, has sincerectified the problems and the PRO is getting95% of the promised utilization. Furtherreports can be expected.

Reachlng Out: A Revlew of Educat10n &Public Programs In New York CityCultural Institutions. Barbara KaNertDeputy Director for Cultural Institutions, 'February 1988, New York, New York City Dept.of Cultural Institutions, 1988, 64pp.

HandSomely illustrated, well presented, usefulrecommendations for coordination andstrengthening of the diverse cultural institutionsin acity. Amodel of how community planningefforts can be presented to win broad support.

Technology & U.S. GovernmentInformat1on Pol1cles: catalysts for NewPartnerships, Association of Researchlibraries, Washington DC,ARL, 1988; $5.00argues that technology has left policy behind andthat this has severe impllcatlons for culturalrepositories that have traditionally providedaccess to government information includinglibraries & archives. '

Videodiscs in Museums: A Project andResource DIrectory, Roberta Binder, FallsChurch VA, Future Systems Inc., January 198825p. plus appendixes 11stlng proJects &resources; $49

T.hls s~iral boun~ report extends the directoryprinted 10 the April 1987 issue of VideodiscMonitor, provlOlng longer entries for eootlproJect reported and ashort introduct10n to thetechnology and its uses.

JOURNAL ART ICLES &BOOKS

Abell-seddon, Brian; Museum Catalogues: AFoundation for Computer ProcessingLondon, Clive Bingley, 1988, 224p. '(ISBN 0-85157-429-7)

Brian Abell-Seddon presents acoherent casefor building museum systems on museum doto inthis important book. The proposed ReferenceFramework for OrganiZing Records in Museums(REFORM) is aprogram for the incrementalimprovement and mlgratlOn of manual cataloguesto automated forms which turns out to be thearchitecture of his commercially avallablemuseum Information system, REFORMATIONbut his analytiC perspective has valldltY bey~ndany specific implementation. The book as awholemakes an eloquent case for sYstem analysis anddata modelling, but without ever employing anyC?mputer jargon. In lts plooe, Abell-seddonl1berally illustrates his points using actual datafrom awide variety of museums, demonstratingconcretely how museums can make theirinformatIOn more consIstent and useful bvooveloplng and applying conventions and'terminology controls. He also clearly describeshow they can exploit the information theyalready have for scholarship, collectionsmanagement. and administration. Museumcurators and archivists wi 11 find his discussionsand illustrations of authority control,vocobulory development and descriptlOnconvention germane, even when the examplesdon't directly apply to theIr situation.

Bakewell, Elizabeth, William O.Beeman &CorolMcMichael Reese ; Object. Imago. Inquiry:The Art Historian at Work, Report on aCollaborative Stut1y by the Getty Art HistoryInformation Program and the Institute forResearch in Information & SCholarship, BrownUniversity; Marilyn SChmitt, General Editor.santa Monica, Getty Trust,1988(ISBN 0-89236-135-2)

This long awaited report on how art historianswork ,and what n~ they have for images,doesn t answer eIther Question. We learn abitabout how some art hlstonans pursue research J

but without much of aconceptual frameworkwith which to make sense of the data, And we are


left without guidance about how Image retrievaland manipulation systems might be designed tosupport art hIstorical use of images.

Blackaby, James, Patricia Greeno. and theNomenclature Committee; The RevisedNomencl8ture for Museum C8t8Ioging: ARevised 8nd Exp8nded Version of Robert6. Chenh811's System for ClassifyingM8n-M8de Objects, Nashv1l1e, TN, AASLHPress, 1988

The revised Nomenclature consists of 34 pagesof introductory material on what it is, how to useit and the sources for each section, along with a228 page classified list, a242 page alphabeticallist, and anine page bibliography. It is aheftybOOK in part due to use of typewritten lists.

The classified listing consists of ten categJri.eswith up to two levels of inverted terms beneaththem. SCope notes are only provided for the twohighest cat~ries. Terms are "~neric objectterms used for inct3xlng purposes" (p.II-3), notcommon names or manufacturer names forobjects. The authors warn that "users ofNomenclature will encounter some difficulties.Despite c.ontinuing efforts to add to the list, notall objects have object terms. Some objects areplaced ambiguously in the system. Some obiectsin museum collections are parts of things or setsof things. Some things are difficult to classifywithout referring to their varietal names. Somethings have been modified from their originalpurposes".

What they 00 not SfJY is that the classificationcontinues to have greater ldeosyncracy because ithas not been constructed according to strict rulesof thesaurus construction. Object terms areusually binomial. The term Itself is anoun, butit may be modified by a range of terms that obeyno obvious grammatical rules. MACHINEBATCHER/JOGGER; MACHINE, BEATING;I"IACHINE-BRICK; all follow closely in thealphabetical listing! Nor does the same Qualifiernecessarily operate the same Wf/Y: BOX,NEEDLEWORK and CASE, NEEDLEWORK are"Textileworking Tools and Equipment" under(categ:>ry 4), while PICTURE, NEEDLEWORKappears under "Communication Artifacts: Art"(catetJlry 8) and needlework itself, though acommon usage, doesn't even get across reference.BOX, BREAD holds bread and BOX, CIGARETTE

holds Cigarettes but what of BOX, SHADOW (norelation to shadow boxing)? BOX, LUNCH is thecontainer, not abox lunch which isn't here andBOX. WOOD is a "Temperature Control Device"for which YOU should useWOODBIN! How is itthat NET, FLY is an animal husbandry tool whi IeNET, MOSQUITO 1s bedding and general NET,PROTECTIVE is clothing -- headwear?

All of which is just to SfJY that this system canbe used to classify objects, but It is not asaccessible an access language as it should be.Hopefully this w11l be <rl1ressed by: 1) addingvarietal and manufacturers terms as lead-invocabulary, 2) prOViding the file in amachinereadabIe, searchabIe form, and 3) 111ustratingthe terms with scope notes and pictures. Thisshould engage the history museum community forthe next decade or so. Therefore, the AASLH isselling Nomenclature as asubscription serviceincluding quarterly updates and atechnicalreport on converting from "Old Nomenclature" to"The Revised Nomenclature".

Cook. Michael ed., Approaches to ProblemsIn Records M8n8g8ment II: ComputerGener8ted Records, Society of Archivists,Computer AppIlcatIons CoinmIttee/Recordsl'1anagement Group, 1987 60p.

These proceedings of a conference held inLiverpool In september 1986 Include papers onvarious aspects of the identification I appraisal,accesslOnmg. and use of machine readable andcomputer generated records including an initialreport of aceTA experlment with optical dISks,an outline of the PRO optical disk study and someassessment of the legal sltuatlOn of computergenerated evidence in British case law.

Inachallenging mtroductlOn, S.C. Newton, theconference chairman, observes that "perhaps thereal issue is not computer generated records, butthe records which are not generated by thecomputer, I.e. those WhiCh remam forever 10 theheart of the machine." He expands upon thissubsequently, noting that "if someone gIVes youmagnetic data filed without aDBMS or aquerylanguage, it is the equivalent of glVing you aboxof words - how do you know what sentences theyare suppOSed to make?; iMeed, how do yOU knowwhat sentences they Qk1 make when In operationaluse?"

40 Archival Informatics Newsletter voI.2.#2

Dearstyne, Bruce; The Management of localGovernment Records: AGuide for localOff1cials, Nashville TN, AASLH, 1988, 146p.(ISBN 0-910050-91-0)

Buy your local government official acopy ofthis book! Bruce Dearstyne addresses h1m or herpersonally from the flrst page in thisreadablepractical, and pithy gU103. l1ke the rest of the'book, ~he 10% devoted to automated systems formanagmg records and the management ofautomated records can be counted on to distill thebest adv1ce from arange of sources. Aniceoverview indeed.

$aracev1c, Tefko et.al.; "A Stuc1y of InformaHonSeeking Behavior", Journal of the AmericanSociety for Information Science, vol.39, 1988p.161-216

ThlS three part artlcle on an extensive study ofwhat makes for effectlve information retrieval isthe most importont b6Sic reset'lrch to bepublished 'in this critical area for years andshould be required reading for anyone involved inproviding information services.

Stone. Nidia; "Infofind: Apractical tool formanaging information", Information ManagementReview, v,3 #4, 1988, p,39-46

Dlscusses an IRM directory developed 1n newJersey to identlfy records systems in stategovernment. The 41 00 entries includedisposition reqUirements.

Weber, Lisa, "Educating Archivists forAutomation". Library Trends, Winter 1988,p.501-S18

The former automation officer of the SAAprovides acomprehensive overview of what hasbeen going on in archival automation for the pastseveral years and how 11 1S affect1ng archivaleducation and the skills required to be aprofessional archivist.

NEWSLETTERSCAD Information CDA (JSSN 0709-4620)publ1shed oj-annually for the Commlss1on onArchival Development by the National Archivesof canada, Bernard Wei 1brenner editor, 395Wellington St., Ottowa, KIA ON3, canada; free.

Government Publications News (BernanAssociates, 4611 Assembly Drive, Lanham l"'ID20706-4391) report.s that the lona-awaltedOTA study of electronic publishing anddissemination of government documents w1l1 bedistributed through GPO in late summer or earlyfall.

Heritage Education Quarterly, The PreservationL'lbrary & Resource center, 429 South Main S1., 'Madison. GA 30650 $ 12 p.a.

This one year old newsletter on culturaloutreach programming is aboon to museum andarchives educators.

Humanlst.iske Data,vo.l.1 , 1988 pA-47,contains the proceedings of aconference onoptical media including short reports. some inEnglish and (unfortunately for me) some inNorwegian, which provlde atour of numerousexciting projects in the humanities usingvideodisc and optical digital d'isks.

Library Systems Newsletter, vol. 8, #3 &4,contains the Annual Survey of Automated LibrarySystem Vendors, in two parts: turnkey vendorsand software only vendors. Updates developmentsin each system over the past year. .

Museum Association security CommitteeNewsletter, vol.7 # 1, Spring 1988 [available tomembers of the Committee; AAM dues plus $1 0.]reports that the National Fire ProtectlonAssociation's Practices #91 0 "Fire Protectionin Museums and Museum Collections" and #911"Fire Protection in Libraries and Archives", willbe revised next year. Contact Jack Watts, Firesafety Institute in Manchester VT, 802-4622663 if you want to be lnvolved or cancontribute ideas,

Museum Management Institute Newsletter, vol. I#2 (American Federation of Arts, 270 SutterSt., San Francisco, CA 94108) contains ascriptfor some role playing that has great possibilitiesfor developing some better approaches topersonnel management for anv l:X1mln1stratorStrongly recommended - but follow directionsand don't read both scripts!


Optical Memory News, Issue 53, November1987, Dubl1stled the text of the Texas law onadmissability of Optical data in courts of law. Inarecent issue of INFORM (April 1988 p.43-44)the law is reviewed by Douglas Allen, CRM, andfound to be seriously flawed. Others promotingsim ilar legislation should note his critique.

Publishing and New Medi8 Technology Newsletter(ISSN: 0885-6214) bi-monthy from PergamonPress, £ 130 p.8.

This expensive, stapled, 12 page. newsletter of"company news, market trends, statistics,product innovations, technological developmentsand media competition" consists almostexclusively of press releases by "new media"firms. purchase offers and conference schedules.

Registrar. vol.5 #2 (MM Registrar Committee)is devoted to storage environments 8< planningreqUirements. Auseful treatment.

EPHEMERA

The Association of Research Libraries [ 1527New Hampshire Ave. ,NW. Washington Dc20036J will issue the report of its Task Forceon Government Information in Electronic Formatthis summer. Drafts of the report, Technology8< U.S. Government Information Pollcies:catalysts for New Partnership, have beencirculating since OCtober J987 and have nowbeen endorsed by members. The report defines aframework for government policy and privatesector roles that contrasts sharply with thataoopted by the Reagan administration, and itdevelops the implications of these policies withrespect to changing technologies of informationcreation and dissemination in wcrys that areextremely important to archivists andlibrarians, Thereport also presents ataxonomyto categorize the types of governmentalinformation in electronic formats that shouldadvance our discussion of their handling.

Mellon Bank Corporation, Discover TotolResourceS: AGuide for Nonprofits, free fromCommunity Affairs Division, Mellon Bank

Corporatlon, One MelIon Bank center I

Pittsburgh.PA 15258 or call Sylvia Clark,(412) 234-3275.

Aself-study check list to assure that a 000prof1t is doing everything possible to acquire anduse resources. Full of goOO ideas.

University of Toronto Archives, Principles andGujdelines for the Description of Records in theUnjversity of Toronto Archives, January 1988

Kent Haworth, Archivist of the Universlty ofToronto, and a leading member of the Canadianworking groups on descriptive standards, haswritten an ·internal manual for description,based on Steve Henson's manual for cataloging,Archives. Personal Papers 8< Manuscripts ,andthe rules for description of graphics recordsproposed by Elizabeth Betz Parker. Haworth.and Heather MacNe1J I who shares responsibilityfor the U.T. guidelines, have a sophisticatedappreciation of the informative value ofprovenance and the importance of integratingarchival descriptive tools that makes this anexceptionally valuable manual. Hopefully theseviews wlJl inform the emerging Canadian·standards, providing aclear path for theacceptance of mutual standards throughout NorthAmerica.

University of Liverpool, Archival DescriptIonProject, Archlyal Description: AMAD UserGuide, Test Draft I! 1.25 september 1987

Margaret Proctor and Michael Cook offer usthe first draft of ausers gUide to the Manual forArchival Description. Intended as astandard forthe U.K.• MAD takes "levels" of records as acentral principle. The users manual is,however, more phi losophical than exp licit aboutwhot informotion should be incorporated intodescription of each level. Although examples areprovided, they have few discrete and identifiabledata elements, and consist largely of relativelyloosely structured text. The User Guide ISreminiscent of David Gracy's guidelines forconstruction of registers and InventorIes, exceptfor advancing what seems like asomewhat rigidnumbering scheme for levels. On balance, themanual does not address Questions of source ofinformation, access points and vocabularycontrol that have come to the fore in the NorthAmerican diolog obout description.


(Dis: a publication note

It Is now evident that CD's are coming toarchives, not just as media generated by theirparent institutions, but as reference resourcesfor archIvists themselves.

The Library of Congress CatalogingDlstr10utlon 5ervlce 1S selling L10rary OfCongress Subject Heooings on CD-ROM(complete through March 1988) for $300,including the original, two fully cumulatedupdates and manual. The CD will play on anystandard CD drive with aPC using DOS 3. 1+.Library of Congress Name Authorities wfll followsoon.

Chadwyck Healey Inc. intends to market theindexes to the National Inventory of DocumentarySources on CD. No publlcation data has yet beenset.

AIRS Inc. has announced the publlcatlon of thecomplete NUCMC Index ( 1958-1988) on CDROt''!. To be available in september for $595 theCD is being offered at apre-publication price of$495 bv July 15. Annual updates will be offeredfor $225 to original purchasers. The productw1lJ feature Boolean search1ng, saved Queries,hypertext browsing, and the printing of reportsto disc or prmter.

A large number of archives are associated withlJnJversJties and state libraries that are engagedin major retrospective conversion/CD-ROMpuol1cat10n proJects. Anyone 1n the positionwould be well advised to reoo aconsistentlyuseful series of articles on CD-ROM PublicAccess Catalogs (PACs) that has been appearingin Library Hi Tech. The first article, by LindaHelgerson, was entitled "Acquiring aCD-ROMPubllcAccessCatalogSystem" (vol. 5,#3,p,49- 75). The second article, by Linda Bills andLinda Helgerson, on "Database Creation andMaintenance" examined how the diferent vendorsprepare the databases for CD recording(vo1.6# 1, p.67-S6. The third article, also byBlIIs and Helgerson, Is entItled "User Interf~for CD-RO~1 PACs". It is asuperb critical reviewand compar1son of the commerc1ally avallablesystems, important not only to potential CD-ROMusers, but to anyone involVed in user intertacedesign or in the use of systems (vol.6 #2, p.73115). Real these reviews.

Archival Informatics Newsletter is aquarterly publication of Archives & MuseumInformatics, 5600 Northumberland St.,Pittsburgh, PA 15217; (412)-421-4638. It1s edited bY Dav1d Bearman, whose authOrshipcan be presumed for all items not otherwiseattributed. Subscription to the ArchivalInformatics Newsletter (ISSN 0892-2179)1s avaIlable for $24.00 per year, pre-paid, toU.S. OOOOsses; $30.00 per year, pre-paid toforeign OOdresses; and $40.00 per year billed,worldwide.

Asubscription to both the Newsletter and itscompanion quarterly publication, ArchivalInformat1cs TechnIcal Report ( ISSN 08940266) is aval1able for $160.00 p.a. in the U.S.;$180.00 abroad; and w1ll be billed at no<lIi:Iitional charge.

Individual technical reports are ava11able at$45 each, prepaid; $50 billed. Titles currentlyavailable include:

o Optical Media for Archives 8< Museums

o Collecting Software: ANew Challenge forArchives and Museums

o functional Requirements of CollectionsManagement Systems

o Acquiring & Implementing Automated Systems

o Directory of Software for Archives 8<Museums

Forthcoming issues inclUde:o Smithsonian Seminar on Authority Control in

Archives (August 1988)

o Archival Appraisal of Onl1ne InformationSystems (Nov.SS)

o Functional ReqUirements for ExhibitsManagement (Jan 89)

Summer. 1988 • by Archives" Museum Informatics 43

PROJECTS &PROPOSALS

F1orel1o H. L86uard1a Photo ProjectArecent blitz of pub1icily unleashed by the

Archives of the City University of New Yorl<regarding its computer system for smallarchives or museum photo collections led me toask them for some documentation. Dr. Richard K.Lteoerman, proJect dIrector, provIded me wlth acopy of the final report he submitted to the NEH,documenting the system that was developed tocatalog 2800 photographs of New York Cityduring Mayor LaGuardia's administration. Hispress release states that "in the absence of anaccepted standard for the computer catalcglng ofphoto collectfons under 50,000 items. theArchives has designed its own informationretrieval system. The system allowsresearchers to seek items by site. persons In, orforty-five topics... " .

Ttle system Is an extremely OaslC, sIngle me.catalog. Each record in the system consists ofeleven fields: caption. year, id:lt. subjectcategories (from the 45 terms avaflable),subject headings (from eight broad categories),oonor. site, negatlve f1~, copyright holder,partlclpants and location. Users, we are told,may review the subject categ:>ries and SUbjectheadings, as well as all sites and participantsfrom printed Iists prior to using the system.searches, based on character string occurrenceonly, prodUce either acount or aseria1displayOut cannot provIde Oottl.

Apparently the system lacks value tab Ie orauthority validation, provides for repeatingoccurrences of fields but not repeating fieldgroups, and requires all data to be keyed at theHem level even when the same donor record in amultl-record database could have served as thesource for anumber of Item records. Theexamples of records given in the report Hselfillustrate avariety of undesirable practices suchas citing LaGuardia himself when he appears in aphotograph as FHL and employing an exceptionally ldeosyncratlc lIst of 45 "cat~rles" thatInclude terms such as Agriculture. Candids,Funeral, Leisure, Out-Of-State, and UNRRA![Archil/es, LaGuardia Community College, CUNY,31-10 Thomson Ave., Long Is. City, NY 11101]

Further on Form from North Carol1na

Barbara Cain, C'.omputing Coordinator of theNorth carol ina State Archives, shared with me amemo she drafted to the internal computercommittee at NCSA in which she makes anumberof oOsarvaHons that w11l be useful to otherarchivists. I am gu11ty of extracting a fewpassages out of context from her lengthy report,so any lack of coherence should be laid at mydoor:"Some of our difficulties with form in the pasthave resulted from confusion between form andformat. (This mistake was made in the 1985RLiN list of Form Terms we stUdied, compiled byHickerson and Engst). Amap is a map whether itis an original, amanuscript copy, axerox, aphotostat. aphotographg or on microfilm. Andany one looking for maps will want the search tobe across the board. It Is important, therefore,that MAP be used w1th all maps, In whateverformat the appear. The same would be true ofdiarIes, whether the format be abound volume, alooseleaf notebook, typed transcrlptfons, xerox,or microfilm. The Information on format. then,I}Jes in the DESCRIPTORS field In FAIDS. It 1simportant to remember that the physical formatwe have in hand (microfilm, volume.photographic negative, electrostratic COPy, etc.)wl11 be found in DESCRIPTORS, not FORM... "

Elsewhere In her memo, Barbara states:"Our use or non-use of subdivisions of forms hasbeen very SUbjective so far, depending on theimportance each archivist attaches to the morespecific variety of form. and depending on thecontext in which it is found. We now have inFAIDS both certificates and ElacHon certificates;Resolutions and (in the General Assemblyrecords) Simple Resolutions and JoIntResolutions; Photographs and AerialPhotographs; Deeds and Sheriff's Deeds; Maps andProfile Maps. We have Bonds now and may wishto add Marriage Bonds. otherwise. I would thinkwe would enter Bonds plus Apprentices,OrdinarIes, Officials, etc. as SUbjects. Do we addDeath Certificates. Birth Certificates, andMarriage Certificates as separate forms? Acasecould be made either way.... It is important toremember ttlat the FORM Is not desIgned to defineeverything we have, but to assist with retrieval.For example, it may not be necessar-y, and may

44 Archival Informatics Newsletter vo1.2. 112

even be d1sruptlve, to distinguIsh JoIntresolutions from simple resolutions, since onewould normally be look1ng for aresolutionrelating to acertain person or subject, notspecifically for ajoint resolution... ".

Now that Barbara has launched this discussionInside North CarolIna, I hope that others will .become involved, either directly with her orthrough these pages.

Function & the RUN 7 States Project

The first grant funded project involving sevenstate archives in cooperative exploration of theuse of the Research LIbrarIes InformationNetwork (RLlN) has been completed. Aproposalto expand the community of users to 17 statesand the National Archives and RecordsAdministration is pending before the NHPRC and,if funded, could be underway by December 1.

Aparticularly detailed evaluation of the use offunctional access in the first phase of the projectis being conducted by Kathleen Roe (New YorkState Archives) and Alden Monroe (Alabama Dept.of Archives & History) this summer. Roe andMonroe were awarded NEH/Mellon Fellowships atthe Bentley Historical Library to assess by casestudy, and test searches the utility of thefunctions vocabulary and prepare guidelinesbased on this analysis that will improve itseffectiveness.

Clearinghouse on Art DocumentationProposes to go online on RUN

Pat Barnett, DIrector of the Clearinghouse onArt Documentation and Computerization [ThomasJ. Watson LIbrary, Metropolitan Museum of Art,Fifth Ave. @82nd. St., New York, NY 10028]has submitted aproposal to the Councll onLibrary Resources to enabIe the Clear inghouse,whic~1 has previously been available only by mailand phone, to make its resources known throughthe R~"8arch libraries Information Network.

The Clearinghouse is also Involved in acooperative proJect with the Art & ArchitectureThesaurus to maintain vocabularies relating tocomputeriZed doCumentatton.

SOFTWARE BRIEFS

Crowninshield Softwore Incorporated[98 Crowininshield Rd., Brookline, MA 02146;617-232-1488] is distributing demonstratlOndiskettes of its optical publishing softwaresystem, Media8ose. MediaBase allows users tobuild local multi-media (including motionimages and sound) databases, index them andoutput them in High Sierra format on 9 tracktape for direct transfer to CD-ROM. Thedemonstration diskette doesn't show much I butthe literature looks promising and I will bepursue it.

Cuodra Associates [11835 W. Olympic Blvd.,Suite 305, Los Angeles, CA 90064; 213-4780066] celebrated its 10th anniversary thisSpring. STAR NEWS, vo],1 # 1made its debutwith an announcement of its STARe/PC productwhich supports six simultaneous users. Two newusers groups are organizing - in Los Angeles andWashington DC.

Electronic Text Corporation [5600 NorthUniversity Avenue, Provo, UT 84604],publishers of CD-ROM versions of Shakespeare.the Bible, Goethe, and such American classics asEmerson, Twain, Melville, Lonoon, and James, ispUbliciZing its offerings in News etc.. Vol.l # 1reports on the WordCruncher software it uses,in conjunction with WordPerfect™.

Moster Software Corporation [8604Allisonv1lle Rd. Suite 309, Indianapol1s, IN46250; 317-842-7020] reports in MasterEi.l.es. (vo1.3# 1) on new features of FundMaster 6.0, due for releose ot the end of thesummer. Included are spousal donation accounts,the maintenance of "letter" histories to warn youbefore you send the same form letter twice to thesame oonor, and automatIc calculatlOn ofmembership r.ategories based on donation levelare among the new twists. The issue alsodescribes MosterPiece, their new collectionsmanagement system, apIan for regional usersgroup meetings, and the recently released FundMaster 5.0 which includes numerous desktoppub11shIng features.

Summer. 1988 • by Archives It Museum. Informatics 45

Michigan Microcomputer catalogingSystem [University of Michigan Software,Intellectual PropertIes 0ffjce. 225 W.Engineering Bldg.• Ann Arbor. MI 48109-1092;313-936-0435] is being adverised as aslidecataloging system for PC/XT /AT systems runningDOS 2.0T using compiled FoxBASE T. For $1000per site thesystem provides 16 fields up to 110characters 10 length and stores variebIe lengthrecords and prints slide labels or shelf cardlists.

Time-Space Systems Inc. is working withthe Metropolitan Toronto Archives and Records •centre on an on-line catalog system with aunique cartographic/chronological interface tobe navigated by mouse. Company PresidentRichard Hill sees the interface as being uniquelyable to provide noive users with immediotefeedback about any aspect of Toronto area historyas wel1as I1nl<ing to machine readable dataconcern Ing the city and Its demography [formore information call 416-699-2003].

Image Online Inc. [P.O. Box 1124, Topanga,CA 90290j 213-455-3883] announces theElectronic Library, amer~r of imagingsystems and library bibliographic systems toprovide libraries of print, audio, and stillimages,on-11ne. Image OnL1ne is anew firmbut its principols hove many yeors of experl~ncein libraries and sol1d experience serving as anintegrator.

8lackbaud MlcroSystems Inc. { 160 EastMain St., Huntington, NY 11743; 516-385-1420] has introduced Planned Glvlng II, anupgraded version of th~1r earlfer planned giVingsoftware pack/l93 th~t features companson , onone screen, of remamder values for up to sixgifts or variations on agift, calculations of cashflows for various types of Unitrusts, data fordetermination of the gift ond estate toximplication of Inter-Vivos Charitable leadTr.usts, and calculations to determine if gifts w111trlgger Alternative Minimum Tax and themaximum gift that will not trigger AMT.Numerous other features make the package a'must-see' for planned giving professionals.

AmembershIp ao:11tlon to the Raiser's EdgeMembership and Development software was alsobeing shown at the MM. It is described In alargepacket of documentation and reports available •from Blackbaud. Unfortunately, I didn't havetime to examine the system personally.

NEW SOFTWARE DIRECTORIES

Ame~ican Society of Indexers; A Guide toIndeXIng Software, 2nd edition, Linda Fetters.compller, lOp. from ASI J $7.50 members$10.00 non-members '

Listing of nine In():lxlng packages with astructured one page description of each. Theeditorial ju~ments expressed in this"compllation" are relatively pronounced and maybe useful gUidance.[Ed. note: Among the IBM PC indexIng packagesD.Qi 11sted in the ASI directory are: IN>SORT[Kensa Software. P.O. Box 49. cathedral Stat Ion I

New York, NY 10025 - $79.95 post paid];wlNDEX [Watch City Software, 24 Harris St.,Waltham, MA 02154 - $109.95]; MACREX[Bayside Indexing serVice, 265 Arlington AveKensington, CA 94707, $325]. J

Archives & Museum Informatics [5600Northumberland St., Pittsburgh, PA 15217J haspublished its Directory of Software forArchiVes &. Museums, Ist edition I 1988. Thedirectory contains two page structured datacomparing 35 packaoes along more than 100dimensions and twenty pages of tables comparingthe systems ~rding to an additional 300characteristics. Indexes by system name.hardware, operating system, vendor, appl1cationand ut1l1t1es are provided for easy reference.The report is available for $45. prepaid, $50billed, from Archives & Museum Informatics.

National Archives & Records Administration I

Records Adm1nlstration Information Center isdistributing Records Management SoftwarePackages. 1987". lOp., free. This latestNARA-RAIC Informatlon Package list 9 softwarepackages for records management along with aone page prose description of each. Agood placeto start for ao:1resses.

46 Archival 1morDlaticsNewsletter

I

1iI

i

l.

STANDARDS

Framework for Internat10nal Exchange ofMuseum Information

At the AAM Meeting, Peter Homu]os, Directorof the Candian Heritage Information Network(CHIN) and Acting President of the MuseumComputer Network (MCN) presented a paper ,onInternational Exchanges of Museum InformatIOnthat has significant implications for museum datastandardization efforts,

He laid out aframework, apD1icabIe to anyInformation exchange, that bulldS fromogreements to standards to networks.

Recapitulating the steps taken by CHI Nover thepast several years. I:lomulos detailed the terms ofan International AQreement on MuseumInformation Excha-nge, presented by the Canadiangovernment as adraft international conventIOn,The Draft convention addresses ethical, legal •.er.onomic, and technical issues involved in suchexchanges, The technical issues were, by far. theeasiest to resolve; they bon oown to astandardformat for information interchange. EthicalIssues such as agreements to reveal the problemsknown to exist in the data. the uses of data bynon-signatories, exchange of research data, andnon-compliance are more complex. Legal issues.such as national regulations, data ownership,duration and llmltations on data use. andcopyright involve national sovereignty and canonly be accommodated within the framework of ageneral agreement when coupled with amechanism for actual data interchange thatpresumes all such interchanges reflect specificbilateral contracts between institutions involved.

Moving to standardS, Homulos noted that theactivity taking place at anational level, such ashas been sponsored by the MDA in England, CHINin Canada, the MCN and AASLH in the UnitedStates, and MARDOC in Holland, is now beingdealt with at the international level by theCommittee on Documentation (CIDOC) of theInternational Council of Museums (ICOM)through its database survey, oocumentationcentre, terminolcqy and terminology controlworking groups.

Finally, Homulos noted, if we have agreementsand standards. we can exchange data, and hepointed to the success of CIN, the ConservationInformation Network. mounted on computersowned by CHIN and accessed by almost 150 "member instltutions from around the wor ld. The(;IN which is less than two years old. was able toutilize CHIN as autility because it built uponexisting standards and existing physicalconnection methodS. Observing the success ofboth the Conservation Information Network andCHIN itself in the years since its establishmentIn 1972 as the National Inventory ProJect,Homulos suggested that further exchanges ofinformation between museums were onlYawaiting agreement from afew potentialparticipants, and they too could take off on thefoundation of an existing network and standards.

Photographic Th8S8urus ProgressDiane Vogt O'Connor. author of the draft

photographic thesaurus circulated last spri~g bythe Smithsonian Archives Photo Survey Project(see vo1. 1# 2), reports that she has joinedforces With RIchard Pearce-Moses. PhotoCurator at the University of Arizona, to extendthe thesaurus and convert it toa strict ANSIthesaurus structure. They will be adding moretradenames, vernacular terms, generic termsand analytic terms for processes and structuringthe data. Users of the draft who have comments,suggestions, or additions are welcome tocontribute them. [Smithsonian Archives, A&I2135, Washington DC 20560]

Art & Architecture Thes8urusThe Art and Architecture Thesaurus has issued

User Update # 14 (April 1988).TheMT is making final arrangements to

distribute its terminology in machine-readableform and to give software vendors an opportunityto incorporate the vocabularies into theirsystems. Project Director Toni Petersen reportsthat this distribution will take place as soon asreasonab Ie terms can be worked out to protectthe integrity of the MT.


SMITHSONIAN SEMINARON AUTHOR ITV CONTROL IN ARCH IVES

Edtted by Avra Miche1son

Archival Informatics Technical Reoortvo1.2, #1, Spring 1988

In October J987, the Smithsonian Institution Blbllographlc Information SystemArchives User Group held a one day semlnar on Authorfty Control to which theyinvited archivists and museum E:urators from throughout the Smithsonian. Invitedspeakers included:Jack ie Dooley, Special Collections Llbrarian, University of Cal ifornia, San DiegoTom Garnett, Systems Administrator Smithsonlan Institution LibrariesMarion Matters, Senior Records Analyst, Minnesota Historlcal SocietyLIsa Weber, Automation Program Officer, Society of American ArchivistsRichard Szary, Systems Administrator, OIRM, Smfthsonian InstitutlonDavid Bearman, Consultant Archives & Museum Informatics

Papers introduced the concept of authority contrOl, discussed its application inl1brary systems, examined lts use in an hIstorical society both in manual andautomated implementations, reviewed the position of the profession vis-a visauthority control, identified the functional requirements of systems of authoritycontroI required by archives and museums, and suggested directions for futureapp lications of authority control. Audience questions and an extended groupdiscussion at the conclusion of the day were transcribed and edited for thisvolume.

The volume is introduced by an original paper, by AvraMlchelson, editor of theProceedings, that presents the state of standards in the archival profession andidentifies the place of terminology control within the framework of archivaldescription standards.

Individual copies of technical reports are available from Archives &Museum Informatics, 5600Northumberland St., Plttsburgh, PA 15217 for US$45. prepaid. Subscriptions to the Technical ReportserIes, lnC1U(llng also four Issues of the ArCt"l1Val InformatIcs Newsletter I are US$160. p.a. for oomestlcad1resses' US$ 180 for foreian addresses

Archival Informatics Newsletter

Date post:	25-Aug-2018
Category:	Documents
Upload:	trinhnhan
View:	216 times
Download:	0 times

ARCH IVAL INFORMATICS NEWSLETTER€¦ · ARCH IVAL INFORMATICS NEWSLETTER ... bureau and then...

Documents