13385-16099-1-PB

7/30/2019 13385-16099-1-PB

1/23

Finding Balance Between Archial

Principles and Real-Life Practices inan Institutional Repository

ERIN OMEARA and MEG TUOMALA

RSUM Les archiistes daujourdhui ont beaucoup de mal trouer un justequilibre entre la thorie et la pratique dans leurs tches professionnelles, surtoutquand ils doient conceptualiser et mettre en place un dpt darchies institutionnel.Cet article explore les croisements entre la pratique par le biais dune analyse de lathorie archiistique pertinente et des conseils ralistes, puis il examine commentcertaines de ces thories ont seri ou pas au deloppement de la Carolina Digital

Repository (CDR), le dpt darchies institutionnel de la Uniersity of NorthCarolina Chapel Hill (UNC).

ABSTRACT Todays archiists struggle to find a balance between theory and practicein their professional duties, especially when tasked with designing and implementingan institutional repository. This article explores the intersections between theory andreal-life practice through a discussion of releant archial theory and realistic adice,and an examination of how some of these theories were, or were not, applied in thedeelopment of the Carolina Digital Repository (CDR), the institutional repository atthe Uniersity of North Carolina at Chapel Hill (UNC).

Introduction

Is it possible to design and implement a preservation repository founded ontraditional archival theory and principles? This is a question that electronicrecords archiists struggle with as they attempt to find a balance betweenarchial principles, and practices in their professional duties. Although wecan speak only from our own experiences, this article attempts to answerthis question through a discussion of releant archial principles andpractical adice, and an examination of how some of these principles were,or were not, applied in the deelopment of the Carolina Digital Repository(CDR), the institutional repository at the Uniersity of North Carolina at

Chapel Hill (UNC). We close with a discussion of how the institutional

ARCHIvARIA 73 (Spring 2012): 81103Archivaria, The Journal of the Association of Canadian Archiists All rights resered

7/30/2019 13385-16099-1-PB

2/23

82 Archiaria 73

repository1 deelopment process at UNC has shaped plans for the future, ideasand hopes for improements, and how the experience can inform larger issuessurrounding the deelopment of preseration repositories.

Archival Principles and Practices

In the archial profession most practical decisions are based on theory andprinciples. This section explores both traditional and new appraisal and preseration principles and theories that can readily be applied to the preserationof electronic records and digital archial materials within the context of thedeelopment of an institutional repository. It also addresses some of the recentliterature that specifically focuses on issues archiists should consider whendeeloping an institutional repository.

Traditional Approaches

Sir Hilary Jenkinsons A Manual of Archive Administration, originally published in 1922, outlines the history and functions of archies and proidesexplanations for, and best practices on, an array of archial topics such asthe role and duties of archiists, the appropriate selection and acquisition ofrecords, and the idea and importance of custody in recordkeeping. Early inthis seminal work, Jenkinson lays out the fundamental duties of an archiist:firstly to safeguard the records in his or her custody (and their essentialqualities), while secondarily proiding access to these records.2 Jenkinson isexplicit that the order of these roles is of utmost importance and should notbe reersed. This principle is strongly reflected in some archial schools ofthought, and is the approach adopted for the current incarnation of the CDR.

Jenkinson also discusses custody, and its importance within archies andrecordkeeping. He describes cases where there is a need to transfer recordsto another steward, such as when there are records for an organization thatno longer exists. This transfer can be done without the archies losingtheir character ... [as long as] the chain of custody remains unbroken.3 Thisprinciple is especially releant within an electronic records context wherethe custody and state of records can sometimes be in flux. Unlike analoguedocuments, electronic records are not fixed to a medium; digital materials canmoe between computers and storage deices easily, obscuring their historyand releant context. Although he was writing long before digital technolo

1 The authors consider an institutional repository to be not just an open access repository forscholarly output, but also one where digital content managed and/or created by an organization can and should be deposited.

2 Sir Hilary Jenkinson,A Manual of Archive Administration (London, 1922), 15.3 Ibid., 3339.

Archivaria, The Journal of the Association of Canadian Archiists All rights resered

7/30/2019 13385-16099-1-PB

3/23

83Archial Principles and Real-Life Practices

gies existed, Jenkinson describes requirements for ensuring that the chain ofcustody is not broken when transferring records between agencies, includingthe transfer between creator and repository. Requirements include that the newagency follow customary rules for the management of records and ensure thecontinued existence of the archies; and that the new steward take the materials in toto without just selecting what Jenkinson calls the pretty specimens.

As a reaction to Jenkinsons model of appraisal, T.R. Schellenbergs TheAppraisal of Modern Public Records, originally published in 1956, offersmany recommendations and analytical tests that remain useful today forarchiists selecting and working with electronic records in an institutionalrepository. Regardless of todays ery different circumstances and enironments, we can learn from Schellenbergs adice that ery oluminousrecords may need to be reduced before being accessioned into our repositories,and that we must take great care to appraise these records in order to retainthose that hae alue.5 Schellenbergs tests of eidential and informationalalues are still applicable as we aim to presere both digital eidence of anorganization itself and its functions, as well as information on its related people ... bodies, things, [and] problems.6

Schellenbergs test of uniqueness states that the information need not becompletely dissimilar from all other information but that records should contain eidence for, or information about, matters on which other documentary

information does not exist as fully or as coneniently.7 The test of uniquenessis pertinent to the context of an institutional repository, gien the likelihood ofduplication both within its holdings and within and across collections.

Schellenbergs test of form, where both the form of the information heldwithin the record and the form of the record itself is addressed, is also releant to an institutional repository.8 He adises archiists to seek out recordsthat represent concentrations of information, that are in a physical form thatenables ease of use and whose arrangement most facilitates the extracting ofinformation.9 The test of form is releant to an institutional repository where

material can easily be made aailable to researchers in its most useful form.Institutional repositories allow users to iew access copies of files, such asPDFs or JPEGs, instead of extremely large and unwieldy files, such as TIFs oruncompressed audio files stored as preseration master copies.

Ibid., 1.5 T.R. Schellenberg, The Appraisal of Modern Public Records, Bulletins of the National

Archives 8 (October 1956), http://www.archies.go/research/alic/reference/archiesresources/appraisal-intro.html (accessed 28 Noember 2011).6 Ibid., 58.7 Ibid., 63.8 Ibid., 65.9 Ibid., 6566.

http://www.archives.gov/research/alic/reference/archiveshttp://www.archives.gov/research/alic/reference/archives

7/30/2019 13385-16099-1-PB

4/23

8 Archiaria 73

Perhaps most significant is what Schellenberg calls the test of importance, which calls on archiists to appraise records based on their projectedsignificance, utility, and research alue.10 This is the most difficult and imperceptible part of appraisal because it requires the analytical skills of the archiist, and sometimes input from outside agencies and subject specialists.11

Although archiists do aim to presere objectiely and expansiely thehistories of the organizations, constituents, and cultures whose records theypresere, it is prudent to reflect on Shellenbergian appraisal standards whenfaced with the flood of records that todays electronic enironment hasencouraged. Accessions should occur only after achieing a complete assessment and understanding of the records themseles and the repository withinwhich they will be presered. In the case of an institutional repository, thesearchial appraisal guidelines can inform the collecting efforts surroundingdata sets and other non-traditional, scholarly output. They can also help forma sound collection deelopment policy, necessary in light of the large olumeof electronic records archiists are facing.

While appraisal theory can help institutional repositories decide whatto collect, the core archial concepts of authenticity and reliability can helpdetermine how institutional repositories will presere their content for thelong-term. In her 1995 article, Reliability and Authenticity: The Conceptsand Their Implications, Luciana Duranti defines the core archial scienceconcepts of reliability and authenticity, and their necessity to understanding the true nature of a record.12 According to Duranti, reliability is theauthority and trustworthiness of the records as eidence, the ability to standfor the facts they are about.13 Reliability is achieed through the procedureof creation, which is a way to describe and define the controls in place forcreating and handling records within a recordkeeping system. Naturally, themore rigorous these procedures are and the more routinely they are used inpractice, the more the reliability of the records created in the system isenhanced.1 The authenticity of a record is deried from the guarantee thatthe record is what it purports to be, and has not undergone any alteration orfalsification since its creation.15

Durantis article concludes with a frank assessment of reliability andauthenticity in the modern record-making and recordkeeping enironment,asserting that the easiness of electronic records creation and the leel of

10 Ibid., 66.

11 Ibid., 67.12 Luciana Duranti, Reliabil ity and Authenticity: The Concepts and Their Implications,Archivaria 39 (Spring 1995): 5.

13 Ibid., 6.1 Ibid.15 Ibid., 78.

http:///reader/full/value.10http:///reader/full/value.10http:///reader/full/specialists.11http:///reader/full/specialists.11http:///reader/full/record.12http:///reader/full/record.12http:///reader/full/creation.15http:///reader/full/creation.15http:///reader/full/value.10http:///reader/full/specialists.11http:///reader/full/record.12http:///reader/full/creation.15

7/30/2019 13385-16099-1-PB

5/23


autonomy that it has proided to records creators, coupled with an exhilarating sense of freedom from the chains of bureaucratic structures, procedures,and forms, hae produced the sloppiest records creation eer in the historyof record-making and that electronic records, as presently generated, mightbe authentic, but they are certainly not reliable.16 It follows that in order forrecords to be presered as both authentic and reliable, they must be createdwithin a recordkeeping system that ensures that both of these elements arerecognized and controlled. This leel of control likely reaches far beyond thescope of most preseration-centred institutional repositories that manage content long after its actie life; neertheless the reliable record creator shouldbe apprised of good records management practices and policies.17 To increasereliability, staff can perform pre-custodial interention and proide guidanceto records creators on records management principles applied to the recordsidentified for transfer to the institutional repository.

In their 1996 report, The Protection of the Integrity of Electronic Records:An Oeriew of the UBC-MAS Research Project, Luciana Duranti andHeather MacNeil report on the findings of the Uniersity of British ColumbiasMaster of Archial Studies (UBC-MAS) research project, and proide a morein-depth look at how the concepts of reliability and authenticity can be maintained in the digital recordkeeping and preseration enironment of the day.Additionally, they introduce the concept of integrity for describing both the

reliability and authenticity of a record.The UBC-MAS research project was a deductie research inquiry into

how to identify methods for presering the integrity of records created indigital form.18 Within the context of this project, Duranti and MacNeil proide a ery strict definition of a record. This definition does not exactly alignwith what many archiists recognize to be records being presered as suchin recordkeeping and institutional repository enironments.19 The UBC-MASresearch project suggested a highly controlled recordkeeping enironment

16 Ibid., 9.17 Records management at UNC is focused mainly on the administratie records of the unier

sity, not on faculty papers or records documenting their research related actiities. UniersityArchies and Records Management Serices, UNCs records management unit, also proidesadice and assistance to student groups who wish to presere their organizational records(with the exception of academic output material, e.g., journals and digital scholarship). Someof these actiities are gaining popularity in units such as academic research computing andfaculty assistance centres.

18 Luciana Duranti and Heather MacNeil, The Protect ion of the Integrity of Electronic

Records: An Oeriew of the UBC-MAS Research Project,Archivaria 2 (Fall 1996): 6.19 The definition of a public record in many public records and freedom of information legislations includes much more than the diplomatics-based definition of a record. Examplesinclude some publications, system log files, and other transitory documents. Certain publicrecords that would not normally fit the diplomatics definition are also deemed permanent bythe public and organizational records retention schedules.

http:///reader/full/policies.17http:///reader/full/policies.17http:///reader/full/environments.19http:///reader/full/environments.19http:///reader/full/policies.17http:///reader/full/environments.19

7/30/2019 13385-16099-1-PB

6/23

86 Archiaria 73

that would employ technical actions such as the registration and classification of records.20 These enironments and actions are not always feasible insome enironments such as actie recordkeeping in offices, or een preseration enironments within institutional repositories. According to Durantiand MacNeil, the reason for these controls and the key to preseration isto maintain the reliability of the record as well as the archial bond betweenrecords.21Archival bond is the context and relationships that link all of therecords in a collection together, especially those immediately preceding, andsubsequent to, an indiidual record.22

While Duranti and MacNeil attempt to translate these system requirements and actions into the digital enironment, many institutional repositorieswill not be able to apply all such requirements because of aried institutionalneeds and finite resources. Their definition of a record is too narrow to besuccessfully used by many practicing archiists; howeer, the concept ofarchial bond, and the need to maintain context and relationships of recordsin a collection in order to presere them fully and faithfully, is an attainableand worthy goal for institutional repositories.23 More work needs to be done indemonstrating the contextual linkages found in repository enironments. Thedocumentation from subsequent InterPARES work can help formulate systemrequirements for repositories and proide guidance for records creators.

A New Paradigm

In the mid-1990s a new and sometimes controersial paradigm for understanding and dealing with electronic records emerged. Proponents of the newparadigm argued that the ery nature of electronic records requires archiiststo adopt new ideas that would change or oerturn traditional archial principles such as those presented by T.R. Schellenberg and others.2 In her articleSchellenberg in Cyberspace, Linda J. Henry questions and refutes this newparadigm, adocating for faith in traditional archial theory and principles. In

20 Duranti and MacNeil, 8.21 Ibid., 53.22 Ibid., 9.23 The InterPARES Projects followed the UBC-MAS Project and tried to apply more explicit

requirements to electronic records preseration on an international leel. Luciana Durantiand Randy Preston edited a large electronic publication that represented the actiities andfindings of the InterPARES 2 Project (International Research on Permanent Authentic

Records in Electronic Systems [InterPARES] 2: Experiential, Interactive and Dynamic

Records, [Padoa: CLEUP, 2008]). The InterPARES 2 Project (20022007) was an extension of the first phase of the InterPARES 1 Project (19992001). It expanded the scope ofresearch to include the inestigation of experiential, interactie, and dynamic electronicrecords. The book is extremely detailed and is broken down by the arious teams, domains,focus groups, and task forces that formed the project team.

2 Linda Henry, Schellenberg in Cyberspace,American Archivist61 (Fall 1998): 309.

http:///reader/full/records.20http:///reader/full/records.20http:///reader/full/records.21http:///reader/full/records.21http:///reader/full/record.22http:///reader/full/record.22http:///reader/full/repositories.23http:///reader/full/repositories.23http:///reader/full/records.20http:///reader/full/records.21http:///reader/full/record.22http:///reader/full/repositories.23

7/30/2019 13385-16099-1-PB

7/23

Archial Principles and Real-Life Practices 87

particular, Henry proides specific arguments against three of the key ideasassociated with the new paradigm: the effort to redefine what a record is, theconcept of the records continuum, and the postcustodial model for the preseration of electronic records.

New paradigm supporters require that records, by definition, must proideeidence of business transactions, and thereby exclude personal papers andother documentary materials altogether.25 Henry argues that this new definition of a record is too narrow, and focuses on indiidual transactions thatdo not proide eidence of the big picture, thus ignoring many materials thatmay hae permanent alue such as databases and personal papers.26 She seesthe new definition of a record as an obstacle to archial work and urges,instead of asking whether documentary materials are records, archiistsshould ask if those materials are important.27 This concept is significant indigital acquisitions, as archiists are seeing new types of digital objects (e.g.,large databases, new forms of digital scholarship, and emerging formats) thatchallenge the notion of recordness een further.

A second key concept of the new paradigm is the records continuum,the idea that there should be no distinction between archial and recordsmanagement work. A records continuum replaces the life cycle concept ofrecords that traditionally defined and delineated the responsibilities of recordscreators, records managers, and archiists.28 The records continuum calls for

the archiist to hold responsibility beginning before creation, through maintenance, preseration, and use. Henry interprets this to mean that archiistswould essentially usurp the role of creator making records less genuine,less authentic, and thus sacrifice their highest irtue: neutrality.29

Henry also refutes the idea of postcustodialism, which supports the decentralization of archies, and enisions an enironment where the creators ofrecords take care of their own archial records. She states that an enironmentof non-custody would result in records lost and damaged ... in ast disarrayand archiists left to deal with the aftermath.30 Henry goes on to describe the

potentially deleterious effect of postcustodialism on archies, where historicalrecords in actie systems would be easily destroyed or changed without thecreators knowledge.31 This could easily happen to records stored in enterprisesoftware applications, such as student records at a uniersity. Uniersitieshae allocated significant resources for these systems; this is an area where

25 Ibid., 315.

26 Ibid.27 Ibid., 316.28 Ibid., 318.29 Ibid., 319.30 Ibid., 320.31 Ibid.

http:///reader/full/altogether.25http:///reader/full/altogether.25http:///reader/full/archivists.28http:///reader/full/archivists.28http:///reader/full/knowledge.31http:///reader/full/knowledge.31http:///reader/full/altogether.25http:///reader/full/archivists.28http:///reader/full/knowledge.31

7/30/2019 13385-16099-1-PB

8/23

88 Archiaria 73

archiists can step in and act as stakeholders and consultants, and help definesystem specifications and policies for use in order to ensure reliability.

While pointing out some of the problems with the new paradigm, Henryalso highlights her premise that perhaps it is not impossible to deal with theproblems raised by electronic records with traditional archial theory andprinciples after all. She states that while electronic records present archiistswith new challenges, solutions will come from an examination of what wealready know, not a dismantling of archial theory and practice.32

While most of the writing on electronic records focuses on goernmentaland organizational records, Adrian Cunninghams 1999 article Waiting forthe Ghost Train: Strategies for Managing Electronic Personal Records BeforeIt Is Too Late addresses the issues and challenges electronic records haeintroduced to the field of personal records and manuscripts, and gies somesuggestions for their management. Seeral of these issues and challengesoerlap with those presented with electronic records in all fields, especiallyCunninghams discussion of the tone of the dialogue surrounding electronicrecords, and some of the mentalities that archiists hae adopted toward managing them.

While Cunningham offers some ery useful suggestions on the management of personal records, especially releant to our context is his call for are-examination of the records continuum and the benefits of both continuumthinking and outreach, and some pre-custodial interention.33 Cunninghamasserts that the traditional diision of current records from historicalrecords is an artificial one, and that a record is a record is a record.3 Healso states that archiists cannot afford to be the passie recipients of recordsthat are no longer required by their creators, and that the traditional posthoc approach to record keeping ... is patently inadequate in the electronicenironment.35

Cunningham suggests that while archiists can neer know what willhappen in the future ... there are things about the present that we do knowwill be of enduring interest to society in the future, and that we should notbe derelict in our duty to the future by neglecting those people in the presentwho we know are significant.36 Cunningham also adocates for a proactieagenda when it comes to the design of durable recordkeeping systems as wellas interactions with records creators, stating that we cannot take for granted... that records ... will remain reliable, comprehensible, authentic, accessible,

32 Ibid., 327.

33 Adrian Cunningham. Waiting for the Ghost Train: Strategies for Managing Electronic Personal Records Before It Is Too Late,Archival Issues 2 (1999): 58, 60.

3 Ibid., 58.35 Ibid., 59.36 Ibid., 60.


7/30/2019 13385-16099-1-PB

9/23


and durable without interention.37 While he acknowledges that pre-custodialinterention in all forms can be labour intensie, he does think that it will payin the long run, cutting/eliminating time spent on the arrangement and description of poorly maintained records.38 Cunningham also adocates working withsoftware deelopers and endors to encourage the incorporation of goodrecord-keeping functionality and self-documenting features in their applications and products, a strategy that will not be lost on archiists who haeworked with deelopers and programmers to plan systems.39 The more archiists know about the software deelopment process, the more they can do toadocate for archies-aware software deelopment for personal recordkeeping.

Finally, Cunningham addresses the head-in-the-sand mentality of archiists who ignore the problem or are waiting for someone else to sole it. Heimplies that failing to pursue a more actie agenda will leae us with nothing to satisfy our researchers need for solid, reliable, and authentic eidenceof the past.0

In her article The Long-Term Preseration of the Digital Heritage: TheCase of Uniersities Institutional Repositories, Luciana Duranti proposespractical strategies based on theoretical principles for maintaining authenticity and protecting producer rights within an institutional repository. Mostimportant in establishing authenticity are the integrity of the enironment inwhich a digital entity resides and the processes aimed to maintain them and

to ensure accountability of the person or organization responsible for them.1

This means creating a preseration methodology that allows for mechanismsthat erify source, transmission, and sustainability.2 Duranti also explains thechallenging nature of institutional repositories; their mix of documentationand data, create challenges to continued access and preseration, which isalso the reason why they exist.3

The Need to Incorporate Archival Theory into Technology Development

An understanding of key archial theories and principles set within the electronic records context should inform the design, deelopment, and implementation of an institutional repository. Traditional and new methods for appraisaland selection, determining and presering authenticity and reliability, and sug

37 Ibid., 59.38 Ibid., 60.39 Ibid., 61.

0 Ibid., 63.1 Luciana Duranti, The Long-Term Preseration of the Digital Heritage: The Case ofUniersities Institutional Repositories,Italian Journal of Library and Information Science1 (2010): 158.

2 Ibid.3 Ibid., 159.

http:///reader/full/intervention.37http:///reader/full/intervention.37http:///reader/full/records.38http:///reader/full/records.38http:///reader/full/systems.39http:///reader/full/systems.39http:///reader/full/intervention.37http:///reader/full/records.38http:///reader/full/systems.39

7/30/2019 13385-16099-1-PB

10/23

90 Archiaria 73

gestions regarding systems and skills for the modern archies and archiiststo employ, all hae a place within a real-life institutional repository.

Although the initial commitment can seem oerwhelming een unattainable balancing archial theory and practice in an electronic records contextwill improe the chances of acceptance from all stakeholders. Additionally,archiists working within uniersities and larger research institutions oftenfind their approach to managing born-digital content to be entwined with theparent institutions perspectie and strategy on digital infrastructure and institutional repositories. The need to balance theory, principles, institutionalperspecties, and real-life expectations is a challenge for archiists; howeer,it can inspire acceptance and interest from dierse campus, library, and external groups.

Keeping archial theory, institutional perspecties, and achieable goalsin mind, the UNC deeloped a preseration-focused repository that incorporates archial concepts into its architecture. Because of institutional needsand goals, the CDR deelopment process specifically emphasized the need tolook at both the theoretical and the organizational issues surrounding digitalpreseration, a necessary approach gien the dierse collecting streams of theCDR. The repository currently accepts born-digital special collections, digitalscholarly output from the UNC community, and library-generated digitizedcontent.

Convergent Theories in Archival Practice and the CDR

The somewhat disjointed history of the CDR makes it difficult to pinpoint aspecific theoretical framework from which it was deeloped; it is safe to say,howeer, that the repository was based on checklist archial concepts andtheory during its conception and in its deelopment and deployment. Duringthe first couple of years of repository discussions and planning, faculty andgraduate students from the School of Information and Library Science (SILS),and practicing archiists and librarians from the uniersity libraries played alarge role in isualizing the repository that was to be deeloped and setting itsscope.

During this time, SILSs DigCCurr Project was deeloped, and manyDigCCurr graduate fellows were inoled in one aspect or another of CDRplanning and deelopment.5 Concepts being used to deelop the Trustworthy

Richard v. Szary and Erin OMeara, If Not Us, Who? Uniersity Archies and Campus-Based Digital Preseration Repositories, International Council on Archives Section onUniversity and Research Institution Archives Conference, Edmonton, AB, 15 July 2011, p. 8.

5 The DigCCurr Project is an initiatie to deelop a graduate-leel curriculum to preparestudents to become information professionals in the digital era. See http://www.ils.unc.edu/digccurr/ (accessed 15 July 2011). DigCCurr is in its second phase as a project.

http:///reader/full/http://www.ils.unc.eduhttp:///reader/full/http://www.ils.unc.edu

7/30/2019 13385-16099-1-PB

11/23


Repositories Audit and Certification (TRAC)6 also influenced work on theCDR. The TRAC perspectie shifted focus from the authenticity of the recorditself to the larger issues of audit and assessment of the iability and trustworthiness of the repository within its organizational context. Based on theCDR experience, this perspectie is essential; without dedicated resources, thepracticalities of repository deelopment cannot occur.

When assessing the trustworthiness of a repository, TRAC looks beyondthe preseration actiities that a repository can perform; it takes into accountgoernance and financial issues surrounding the repository:

[I]n determining trustworthiness, one must look at the entire system in which thedigital information is managed, including the organization running the repository: its

goernance; organizational structure and staffing; policies and procedures; financialfitness and sustainability...7

The Digital Curation Lifecycle model was also consulted in the conceptualization of the CDR; the model focuses on business planning and sustainability as key factors in an organizations ability to curate digital objects. 8 A lotof time and effort were dedicated to researching the system requirements forthe CDR. Through the work of the Digital Curation-Institutional RepositoryCommittee (DC-IRC), this was accomplished through feedback from dierse

perspecties: TRAC, the Open Archial Information System (OAIS) Referencemodel,9 and the Digital Curation Lifecycle model were all influential whendeciding what technology to choose. The idea of micro-serices had not yetcome about, but the ideas of sustainability, flexibility, and extensibility werediscussed.

During the course of repository planning and deelopment, organizationalshifts within the library changed the reporting roles of the CDRs projectstaff. Between the moe of the project from the Carolina Digital Library andArchies (CDLA)50 unit to the Library Systems department in 2008 and the

6 Trustworthy Repositories Audit and Certification (TRAC): Criteria and Checklist wasdeeloped to assess the trustworthiness of repositories through an external audit process. Seehttp://www.crl.edu/sites/default/files/attachments/pages/trac_0.pdf (accessed 11 July 2011).

7 Ibid., 3.8 DCC Cura tion Lifecycle Model, http://www.dcc.ac.uk/sites/default/files/documents/

publications/DCCLifecycle.pdf (accessed 11 July 2011).9 The Open Archial Information System (OAIS) Reference model is designed as a broad,

conceptual framework for digital preseration enironments. CCSDS, Reference Model

for an Open Archial Information System (Washington, DC, 2002), http://public.ccsds.org/publications/archie/650x0b1.PDF (accessed 22 July 2011).50 The Carolina Digital Library and Archies (CDLA) began enhancing serices to digital

scholarship at UNC in 2007. The CDLA proides digital project management serices andis the main digital production unit in the libraries. See http://cdla.unc.edu (accessed 22 July2011).

http://www.crl.edu/sites/default/files/attachments/pages/trac_0.pdfhttp://www.dcc.ac.uk/sites/default/files/documentshttp:///reader/full/http://public.ccsds.orghttp:///reader/full/http://cdla.unc.eduhttp://www.crl.edu/sites/default/files/attachments/pages/trac_0.pdfhttp://www.dcc.ac.uk/sites/default/files/documentshttp:///reader/full/http://public.ccsds.orghttp:///reader/full/http://cdla.unc.edu

7/30/2019 13385-16099-1-PB

12/23

92 Archiaria 73

dissolution of the DC-IRC, the focus on archial concepts and principles wasde-emphasized; the technical deelopment of a more traditional institutionalrepository became the primary concern. This shift led to decisions beingmade solely on repository functionality, and as a result, preseration actionsthat can be described as good enough and just in time were executed. Itbecame increasingly clear that changes in organizational structure, staffing,and budget allocation could tremendously alter the perspectie and trajectoryof a repository project or program losing your champion during deelopmentcan threaten the iability of the project. It is, therefore, necessary to buildthese relationships in the earliest stages and foster them throughout the deelopment of the project, regardless of any organizational changes that occur.

When one of the authors (OMeara) arried at UNC in the fall of 2009and began work with the CDR project staff, she encouraged a shift backtoward repository deelopment rooted in archial concepts. OMearas perspectie was deeloped while she undertook her Master of Archial Studiesdegree (MAS)51 at the Uniersity of British Columbia. She brought to theCDR concepts that were based in historical recordkeeping, but applied inthe digital enironment. She also has experience with the challenges facingmodern collecting repositories. One of these challenges is that the recordscreation, records keeping, and records preseration actiities all form thechain of preseration needed throughout the life cycle; in practice, howeer,these records enironments are usually ery separate and under ery littlecontrol. Collecting repositories commonly receie records from the creatorat the end of a collections actie life. More often than not, procedural controlwas neer exerted oer these records. In practice, there is a de-emphasis ofthe need for documentary and procedural control that is stressed in theoretical descriptions of both record-making and recordkeeping enironments. Inmodern record-creating enironments,52 registration, classification, and otherprocedural controls oer records are unknown at the creators leel, except inhighly regulated enironments, such as patenting and pharmaceutical drugdeelopment. Explaining the concepts of diplomatics (i.e., persons, form, andactions)53 as they pertain to records does not seem releant to most technolo

51 While OMeara was a student, the MAS program focused on theory surrounding diplomat-ics and research findings from the InterPARES Project. The focus on theory sered as anindoctrination so that in any enironment, one could rely on the core concepts from theMAS program, especially the concepts of authenticity and reliability of records.

52 Examples of creators enironments of objects ingested into the CDR include a research lab

at UNC, materials resulting from student class projects, and special collections that seek outmaterial from personal donors.53 Persons are the entities recognized by the juridical system as capable of haing the poten

tial for acting legally. Key persons in a record are the author, addressee, and writer. Formis the rules of representation that are the eidence of the intent to coney informationand is represented both intellectually and physically within the record. Actions form the


7/30/2019 13385-16099-1-PB

13/23


gists when they are in the midst of actie software deelopment. Because ofthis, OMeara decided to extract the essential meanings of the concepts so thatnon-archiists could relate to them: digital preseration is about maintainingthe context, content, and form of a record. This distillation of preserationconcepts was receied positiely by the project staff. During these discussions,the project team read Peter Hirtles Archial Authenticity in a Digital Age, 5

an article that explains, in clear terms, the concepts of authenticity in adigital enironment using illustratie examples releant to archiists and non-archiists alike. OMearas work to introduce archial concepts to other project staff using clear language and concrete examples enhanced communicationbetween archiists and technologists. In order to foster a shared, constructie,and collaboratie dialogue between archiists and the communities they workwith, it is necessary to communicate using common terms and concepts thateeryone can understand. While the OAIS Reference model states the sharedocabulary between archiists and technologists, the nuances particular toeach profession may need to be further expressed. For example, the reasonsbehind documenting eents post-ingest goes beyond terms such as SubmissionInformation Packages (SIP), Dissemination Information Packages (DIP), andArchial Information Packages (AIP).55

Both the theoretical and practical perspecties of digital preserationaddress the need to inole the records creator or producer during the actie

phase of the life cycle (as it is called in the OAIS model). The methodologyof each perspectie, howeer, is different. One of the core principles of themodel is, keep lifecycle stages simple, and moe complexity into the functions.56 This is somewhat different from the theoretical perspectie of theintricate InterPARES 2 Project Integration Definition for Function Modeling(IDEF)57 models, especially the Chain of Preseration Model, which isuallydemonstrates the complexity that the theoretical perspecties propose for thecreation, management, and preseration of digital objects.58 The records cre

impetus for records creation and come from a will to determine a fact. See Luciana Duranti,Diplomatics: New Uses for an Old Science (Maryland, 1998), 83, 1, 62.

5 Peter Hirt le, Archial Authenticity in a Digital Age, in Authentici ty in a DigitalEnvironment, ed. Council on Library and Information Resources (Washington, DC, 2000),823.

55 These terms refer to packaged objects (Submission, Dissemination, and Archial InformationPackages, respectiely) moing through an archial repository per the CCSDSs Reference

Model for an Open Archival Information System.56 Christopher A. Lee, Helen R. Tibbo, and John C. Schaefer, Defining What Digital Curators

Do and What They Need to Know: The DigCCurr Project, Proceedings of the 2007Conference on Digital Libraries (2007), 950, aailable at http://www.ils.unc.edu/digccurr/jcdl2007_paper.pdf (accessed 3 August 2011).

57 Integration Definition for Function Modeling (IDEF) is a modeling framework used forsystems design.

58 Luciana Duranti and Randy Preston, eds., InterPARES 2 Book, Appendix 1: Chain of

http:///reader/full/objects.58http:///reader/full/objects.58http://www.ils.unc.edu/digccurrhttp:///reader/full/objects.58http://www.ils.unc.edu/digccurr

7/30/2019 13385-16099-1-PB

14/23

9 Archiaria 73

ation and records keeping life cycle stages in the model are ery detailed; thecomplex model proides an excellent paradigm for study, and a springboardfor brainstorming system requirements.

The challenge for the CDR team was to interpret these arious requirements (InterPARES Chain of Preseration Model and TRAC documentation)and deelop a functional system while working with limited resources, understrict timelines, and realizing the need to sere dierse collection materials. The Chain of Preseration Model that came out of the Modeling CrossDomain of the InterPARES 2 Project proides details at all leels of therecords life cycle and helps illustrate the concept of preseration used by theproject team. The model is extremely complicated, but uses the modelingframework to go from the scale of the entire records enironment (creation,maintenance, and preseration) to the packaged and presered electronicrecords.59 Appendices 20 and 21 of the InterPARES 2 Book sere as more-easily interpretable adice for records creators and records preserers.60

Appendix 20 coers concepts that records creators would need to addressfor their enironment. Appendix 21 spells out the preseration requirementsisually represented in the Chain of Preseration Model for archiists. Themodel was used as one conceptual ideal in the deelopment of the CDR.

The main area where theory was directly applied in the CDR reoledaround ensuring that the digital objects would be inextricably linked to thecontext in which they exist within and amongst other records in the collections. We used the object model framework built into Fedora Commonsrepository software to maintain and communicate these relationships. TheCDR has a content model that represents hierarchical relationships betweenobjects (files and associated metadata) that was tested against arious archialand non-archial use cases.

Ensuring the authenticity of a record is one of the fundamental conceptssurrounding diplomatics, the science of the nature and formation ofrecords.61

Archiists can use the methods deised from diplomatics to test the authenticity of a record and identify alterations to it. With existing technology and limited resources, authenticity can become difficult to maintain when proidingan authentic copy of a record for access.

In the future, the CDR could enhance authenticity by employing processes

Preseration Model Diagrams and Definitions, http://interpares.org/ip2/book.cfm(accessed 3 August 2011).

59 The IDEF model is brilliant in its complexity, but would be difficult to use as a requirements document since it inoles rigorous behaiour from records creators, actie recordsstewards, and the archies.

60 Duranti and Preston, InterPARES 2 Book, http://interpares.org/ip2/book.cfm (accessed 23July 2011).

61 Duranti,Diplomatics: New Uses for an Old Science, 2735.

http:///reader/full/records.59http:///reader/full/records.59http:///reader/full/preservers.60http:///reader/full/preservers.60http:///reader/full/records.61http:///reader/full/records.61http://interpares.org/ip2/book.cfmhttp://interpares.org/ip2/book.cfmhttp:///reader/full/records.59http:///reader/full/preservers.60http:///reader/full/records.61http://interpares.org/ip2/book.cfmhttp://interpares.org/ip2/book.cfm

7/30/2019 13385-16099-1-PB

15/23


similar to those in the field of digital forensics. Snapshots of operating systemenironments and explicit descriptions of the state of the object and its enironment can be captured with these tools.62 New methods for incorporatingthis information into the archial description of the records will help to ensurethe authenticity and reliability of records by automated means, all within acontrolled system. The CDR has the ability to do this; howeer, it is a matterof prioritization and resources as to what leel the repository can and will integrate digital forensics tools into the pre-ingest workflow. The recent collaboration between the two fields has brought with it new techniques that mightmake it possible for medium-sized repositories to begin to address issues ofauthenticity and reliability systematically in electronic records preseration.

Before a repository can proide full preseration and curation serices, itneeds to demonstrate basic preseration behaiour. With the CDR, there wasa decision to focus on basic, bit-leel preseration. Because of this, some curation actiities were not always fully addressed and for a time, this was a contentious topic within the steering committee. If the collecting scope had beennarrowed and more resources had been dedicated to building the technical andingest infrastructure, full curation serices at the onset would hae been morefeasible. Some steering committee members felt that the curation of some ofthe content (e.g., the digitized surrogates of historical photographs), would notneed the full curation serices that a collection of born-digital personal papers

would need.This ariety of content (and the dierse needs associated with it) has been a

common challenge across institutional repository deelopment projects. Whilethe format-drien approach to defining preseration policies can be a solutionfor institutional repositories that aim to collect traditional content (such astextual documents that hae been generated fairly recently), it can be problematic for repositories with a wider collecting focus, such as the CDR. Policiesfor leels of preseration and format support similar to those employed bythe Uniersity of Michigans Deep Blue63 institutional repository is one way

to delineate preseration serices among arious formats and content typesingested. The CDR is looking at the three main collecting streams (borndigital special collections, digital scholarly output from the UNC community,and library-generated digital content) as a way to deelop tailored preseration

62 For more information on the application of digital forensics in an archial setting, seeMatthew G. Kirschenbaum, Richard Oenden, and Gabriela Redwine, with research assis

tance from Rachel Donahue, Digital Forensics and Born-Digital Content in CulturalHeritage Collections (Washington, DC, 2010), http://www.clir.org/pubs/reports/pub19/pub19.pdf (accessed 15 July 2011). The authors proide a thorough oeriew of the state ofthe art.

63 Deep Blue Preseration and Format Support Policy, http://deepblue.lib.umich.edu/about/deepbluepreseration.jsp (accessed 15 July 2011).

http:///reader/full/tools.62http:///reader/full/tools.62http://www.clir.org/pubs/reports/pub1%1A9http://deepblue.lib.umich.edu/abouthttp:///reader/full/tools.62http://www.clir.org/pubs/reports/pub1%1A9http://deepblue.lib.umich.edu/about

7/30/2019 13385-16099-1-PB

16/23

96 Archiaria 73

serices. For example, specific preseration actiities could be performed as astandard eent for one collecting stream but not another.

The Carolina Digital Repository (CDR)

The CDR is an institutional repository for digital format materials producedby members of the UNC community. The main goal of the CDR is to keepUNC digital scholarly output safe and accessible for as long as needed. It alsoseres as a repository of historical materials that broadly support the uniersitys academic mission. More specifically, the CDR aims to acquire UNC digital material, and ensure it is accessible, searchable, and safe from alteration.The CDR is a partnership between the UNCs uniersity libraries, the Officeof the Proost, and the School of Information and Library Science (SILS). TheCDR proides a defined serice to the UNC campus that directly aligns withthe uniersity librarys larger role as a trusted steward of information.

Background and History

Seeral years before the beginning of formal repository work, researchand inestigation began that suggested the need for a repository for facultyresearch at UNC.6 Repository deelopment began formally at UNC in 200with the creation of the Digital Curation/Institutional Repository Committee(DC-IRC). A large stakeholder group with membership spanning the entirecampus, the DC-IRC had an ambitious mandate.65 The group deeloped a proposal that called for the funding of an institutional repository. This proposalwas submitted to the Proost who allocated funding to hire programmers andbegin the technical side of repository deelopment.

In 2007, the uniersity library took on the physical deelopment ofthe repository (now called the CDR), superising all project staff. Shortly

6 The Minds of Carolina project was an initiatie led by Helen Tibbo and Paul Jones,faculty of the School of Information and Library Science at UNC. The project explored howto enable faculty to self-archie or prepare their materials for deposit into a repository. Seehttp://www.ibiblio.org/minds/innoation.html (accessed 15 July 2011).

65 The DC-IRCs mandate was to [d]eelop a feasible plan that will both sere the UNC-Chapel Hills curation needs and will place the Uniersity in the forefront of such effortsin the Triangle, nationally and internationally; design a pilot institutional repository anddigital preseration program in partnership with Information and Technology Serices, theUniersity Library, and the School of Information and Library Science that will support

ongoing research; deelop policies, procedures, and long-term digital preseration strategiesto benefit the entire campus. This will include strategies to educate the campus community. Carolyn Hank,A Progress and Recommendations Report from the Digital Curation/

Inst itut ional Reposi tory Commit tee, 2005 07: Informing a Successful Inst itut ional

Repository Deployment at the University of North Carolina at Chapel Hill (Chapel Hill,NC, 200), 10.

http:///reader/full/mandate.65http:///reader/full/mandate.65http://www.ibiblio.org/minds/innovation.htmlhttp:///reader/full/mandate.65http://www.ibiblio.org/minds/innovation.html

7/30/2019 13385-16099-1-PB

17/23


after, the DC-IRC was dissoled and a smaller group formed the CDRSteering Committee.66 This committee is composed of faculty members fromSILS, representaties from library administration and the Data IntensieCyber Enironments research group (DICE), and seeral library staff whoare directly inoled in repository deelopment. In April 2010, the librarydeployed a beta ersion of the CDR, proiding access to three pilot collectionsthrough a web access portal.67 Building such a robust and ambitious repositorywould certainly benefit from broader partnerships; howeer, they can slow thepace and make project management more complicated.

Repository Architecture and Technical Specifications

The underlying architecture of the CDR uses Fedora Commons repositorysoftware68 connected to an iRODS data grid69 to form a complex preserationenironment. A Solr70-indexed, custom web access portal is used for searchand retrieal.71 The CDR employs custom METS and PREMIS profiles anduses MODS72 as its primary descriptie metadata standard. The library collaborates with the campus Information Technology department for distributedstorage and backup.

66 The CDR Steering Committee seres in an adisory role to the project and helps determinepriorities for both repository and collection deelopment.

67 The pilot began with collections from the Research Labs of Archaeology, the SouthernFolklife Collection, and the African American Performance Art Archie. As of October2011, there were oer 8,000 objects in the repository (see https://cdr.lib.unc.edu/). Many ofthe collections hae access controls that allow for collections that contain sensitie materials,or that hae priacy constraints or copyright restrictions to be securely posted to the repository. The repository uses Shibboleth, the uniersity authentication serice, to grant authorized access to restricted collections housed in the CDR.

68 Fedora Commons is an open-source repository tool. The Fedora object model concept helpsintellectually arrange files and maintain relationships between files and metadata. See http://fedora-commons.org/ (accessed 15 July 2011).

69 iRODS is an integrated, rule-oriented data system deeloped by the DICE group. The CDRdeploys iRODS-based rules to automate preseration actiities such as checksum erificationand other file-leel alidation actiities. See https://www.irods.org/index.php/IRODS:Data_Grids,_Digital_Libraries,_Persistent_Archies,_and_Real-time_Data_Systems (accessed 11July 2011).

70 Solr is an open-source, web-indexing platform that allows for faceted search and browsefunctionality.

71 The CDR web access point is focused on download capabilities. Traditional IR functionssuch as impact rankings for articles, social media-based sharing capabilities, or faculty

profiling hae not been implemented.72 METS, PREMIS, and MODS are all established metadata schemas that allow for the standardization and management of the description of content. The CDR uses METS for packaging contents for ingest into the repository. PREMIS is used to describe preseration eentsthat happen to objects oer time. MODS is used to describe digital objects stored in therepository.

http:///reader/full/Committee.66http:///reader/full/Committee.66http:///reader/full/portal.67http:///reader/full/portal.67http:///reader/full/retrieval.71http:///reader/full/retrieval.71http:///reader/full/https://cdr.lib.unc.eduhttp:///reader/full/fedora-commons.orghttps://www.irods.org/index.php/IRODS:Datahttp:///reader/full/Committee.66http:///reader/full/portal.67http:///reader/full/retrieval.71http:///reader/full/https://cdr.lib.unc.eduhttp:///reader/full/fedora-commons.orghttps://www.irods.org/index.php/IRODS:Data

7/30/2019 13385-16099-1-PB

18/23

98 Archiaria 73

CDRFedora

IngestSIPPreservationPolicies

AIPPREMISMODS

iRODS Access

DIP

SearchQueries

Figure 1. Technical Diagram of CDR System Architecture

Curation and Preservation Actions

The transfer of digital objects from their creator to the CDR is mediatedthrough a repository staff member who consults with the potential depositorsbefore accepting any content. At this time, the nature of the digital objects isdiscussed and an ingest surey is conducted to alert repository staff to issuesconcerning metadata, rights, or storage that may arise. Additionally, a depositagreement73 is signed before the transfer of materials occurs, ensuring thatboth repository staff and depositors are made aware of the rights each holdin regards to content deposited into the CDR. This pre-transfer mediationprocess allows repository staff to uncoer foreseeable challenges to proidinglong-term access, and discuss these upfront with depositors in order to deisestrategies for meeting both parties preseration expectations and goals.Repository staff realized that this leel of indiidualized serice could proeto be unsustainable once the number of depositors to the CDR increases. Itwill certainly be necessary to moe to more efficient modes of transfer, suchas self-depositing and drop-boxes. This is not to say that mediation and quality control procedures will be completely abandoned, howeer they could becarried out in a more streamlined fashion.

Once the materials are physically or digitally transferred to the library,ingest preparation begins. First, materials are staged to an iRODS grid wherethey are held before metadata is linked to the objects and the collection is prepared for ingest into the CDR. Pre-ingest actiities include selection at the fileand object leel, arrangement of files and objects, and linking user-supplied orlibrary-generated metadata to the objects. Ingest occurs mainly through a web

73 See Appendix A.


7/30/2019 13385-16099-1-PB

19/23


portal, where a METS manifest is generated through the CDRs ingest tool,Curators Workbench.7 This manifest is uploaded along with a brief, top-leelcollection description. Curators Workbench also generates unique identifiers,checksums, and a METS profile that maps out the structure and arrangementof objects (files and associated metadata) within a collection. Upon ingest, theCDRs custom Fedora ingest serice generates another checksum for each fileand erifies it against the checksum from the pre-ingest METS manifest. TheFedora ingest serice also generates the relationships described in the METSfiles within the pre-defined Fedora object model.

Currently, the CDR proides bit-leel preseration of the digital objectsingested into the repository,75 i.e., the bitstreams of the files are preseredexactly as they are deposited. In the future, the CDR hopes to moe beyondbit-leel preseration and integrate a suite of preseration actiities such asfile normalization, refreshment, and migration serices. These serices will beperformed on select content and their deployment will ultimately depend onaailable financial and personnel resources.

Priorities for preseration actiities within the CDR hae been enisionedin a layered approach. The first layer of preseration actiities, which has beeninstituted, is keeping the actual ingested files at the bit-leel safe from alteration. The second layer of preseration actiities will increase the functionalityof the system so that it can perform better integrity checks upon, and monitor

the oerall health of, the objects held within the system. A third layer of preseration actiities will include better documentation of digital collections andobjects from the time they are acquired by the library to the point of ingestinto the CDR. This will strengthen the chain of custody of the objects and isderied from the Jenkinsonian theoretical perspecties discussed preiously.We hope to implement further preseration actiities for the CDR, specificallyaddressing the issue of object and system integrity described by Duranti andMacNeil.

It is unclear if library staff will eer be successful in increasing the docu

mentary and procedural controls of recordkeeping systems while the recordsare actie and in the creators custody. Without exerting sufficient leels ofpre-custodial control oer the records in these enironments, will we eer hae

7 The Curators Workbench is an open-source, pre-ingest tool deeloped at UNC to support theappraisal and processing of digital materials. The open source code and wiki is aailable onGitHub, https://github.com/UNC-Libraries/Curators-Workbench (accessed 15 October 2011).The tool is designed to improe efficiency when processing large numbers of files with

custom, non-standard metadata. It assists with description, file staging, and the crosswalkingof custom metadata to a standardized format.75 For more on bit-leel preseration as compared to full preseration serices, see Priscilla

Caplan, The Florida Digital Archie and DAITSS: A Working Preseration RepositoryBased on Format Migration, International Journal on Digital Libraries 6, no. (July2007): 307.

https://github.com/UNC-Libraries/Curators-Workbenchhttps://github.com/UNC-Libraries/Curators-Workbench

7/30/2019 13385-16099-1-PB

20/23

100 Archiaria 73truly authentic and reliable records to ingest? Een though it is now more thanfifteen years old, John McDonalds article Managing Records in the ModernOffice: Taming the Wild Frontier rings true:

[I]n many ways the modern office enironment is not unlike the wild frontier of thelast century. Instead of horses and wagons, our organizations hae proided us withcomputers and software, telling us to charge off into the great unexplored plains ofcyberspace where supposedly we can work more effectiely.76

Pre-custodial interention can be used to offset this issue to an extent.Training or working with content creators earlier in the records life cycle willproide more reliable records and better metadata. Pre-custodial interention

does take extra resources from library staff, but with large streams of regulardeposits from specific donors, the benefits may outweigh the costs. While wecannot fully exert documentary and procedural controls before the recordscome to us, we can maintain the authenticity of deposited records that are inboth the library and the CDRs custody. Communicating this fact to the usersof the CDR who may not realize that the content they are accessing was notcreated by the library is a challenge that remains to be addressed.

Challenges

Een with stakeholder support and dedicated staffing, there were numerouschallenges that tested the success of the repository and how it would fulfill itsmission. In the fall of 2009, the Uniersity Librarian gae project staff a deadline to hae a working repository with a small set of pilot collections ingestedby April 2010. This tight deadline inhibited the realization of certain preseration actiities within the CDR for the soft launch of the project. Focus wasplaced on how to ensure that core repository functions, such as basic ingest,storage, display, download, replication, and disaster recoery, would be car

ried out. To moe forward with the deelopment and launch of the repositoryand meet this deadline, repository staff members made compromises. Staffdecided to prioritize basic functionality across the repository and build furtherenhanced serices in collecting, preseration, and access oer time.

Traditional archies hae a clear collecting mission, scope, and audience.With institutional repositories, howeer, collecting areas, scope, and stakeholders are sometimes blurred, ill-defined, or not defined at all. There is anidentified need for stewardship of digital scholarship within the UNC community, both within the library and externally from faculty and students, and

the CDR is trying to hone in on its prioritized collecting tracks based on these

76 John McDonald, Managing Records in the Modern Office: Taming the Wild Frontier,Archivaria 39 (Spring 1995): 71.

http:///reader/full/effectively.76http:///reader/full/effectively.76http:///reader/full/effectively.76

7/30/2019 13385-16099-1-PB

21/23


needs.77 As more objects are ingested into the repository, CDR goernance willneed to decide how to narrow the collecting scope. By focusing on selectedcontent areas or formats, the repository can demonstrate alue and releance tousers and other stakeholders. Haing a broad collecting scope means that noteerything that is ingested into the repository fulfills the definition of a record,een when a broader definition is being used. As the repository matures, sowill the collecting areas. These areas will help determine the leels of preseration needed for each category.

Both the theoretical and organizational approaches support the idea thatarchiists should hae a proactie and inoled role in the deelopment ofrepositories, including policy and technology decision-making. Howeer,with only one archiist on the project staff and a few more in the SteeringCommittee, it is highly likely that decisions affecting the preseration framework will be made with little input from archiists. Although the CDR hasbuilt in mechanisms for consultation by archiists before large technicaldecisions are implemented, separate departmental affiliations and physicallocations of archiists and technologists encumbers this process. While largeinstitutions can hae the adantage of haing more resources, archiists cansometimes get lost in the organizational structure where they will not be ableto make an impact. Access to technology and clear communication channelsto technologists is the key to making sure that this does not happen. The CDR

has been a great example of the benefits of keeping organization-wide dialogue between technologists and archiists clear and open.

Conclusion

The CDRs current priorities include enhancing existing preseration actiities and using a measurable standard to assess performance.78 Repository staffmembers continue to build on existing preseration actiities within the repository to demonstrate the integrity of the system itself, and ensure the reliability

and authenticity of the records it preseres.In order to remain true to themseles and the profession while continuing

to moe forward and stay releant, it is necessary for archiists to reflect onarchial theory while remaining open to practical innoations. This is justone of the many challenges archiists face when tasked with designing andimplementing institutional repositories. The case of the CDR reflects on this

77 Collecting tracks currently include faculty and student scholarship such as datasets, digital research materials, and published material that the UNC community deposits into theinstitutional repository; institutional electronic records that Uniersity Archies collects;born-digital special collections that are transferred to Wilson Library; and digitized specialcollections within Wilson Library.

78 For example, using TRAC as a guideline to form an internal framework for assessment.

http:///reader/full/needs.77http:///reader/full/needs.77http:///reader/full/performance.78http:///reader/full/performance.78http:///reader/full/needs.77http:///reader/full/performance.78

7/30/2019 13385-16099-1-PB

22/23

102 Archiaria 73

challenge and others, including finding balance between finite institutionalresources and conceptual ideals presented in archial theory and research.

Returning to our original question, Is it possible to design and implementa preservation repository founded on traditional archival theory and prin-

ciples? Yes, to an extent, and depending on repository aims and institutionalperspecties. It is possible to reach some goals, and it is necessary to compromise on others. The CDR is an example of these goals and compromisesrealized.

Compromise was necessary in three areas. First, with respect to preseration actiities, we found that digitized content does not need the samepreseration enironment as electronic records, and consequently chose touse a layered approach to preseration actiities. Second, as far as descriptiepractices were concerned, we chose not to use Encoded Archial Description(EAD) within the repository, but where appropriate to link EAD finding aidsto born-digital collection objects that are in the repository. Third, we chose afairly large and undefined collecting scope oer a more thorough, formal, collecting policy.

Unexpectedly, we found that this large collecting scope did hae its benefits. For example the CDR was able to obtain more funding opportunities thana smaller repository, such as one for just born-digital objects, would hae beenable to secure. Additionally, because of the dierse nature of the CDRs collections, it receied more recognition within the library and the professionalcommunity.

In conclusion, we offer some thoughts on how to sustain a preserationrepository based on our experience with the CDR. Start, and continue, tobuild the repository in staged layers. If you currently cannot build an actualrepository, begin to think about future requirements, and in the meantimebuild a storage space that incorporates bit-leel alidation and other basicpreseration actiities that can be performed oer a file system. Base yourrepository architecture on theory and best practices, but do not strie for perfection or unattainable goals. Acknowledge, and be able to communicate, thata repository deelopment project requires a serious inestment of resources.Stress to administration and project staff that archiists need to be there whenpolicy and technical decisions are being made. Work to get more library staffengaged with the repository by making it part of their daily workflow. Finally,balancing preseration and user needs with a shrinking budget is challenging,but demonstrating use and alue can help build a case for the ongoing commitment of resources.


7/30/2019 13385-16099-1-PB

23/23


Appendix A: Carolina Digital Repository Non-Exclusive Deposit Agreement

Please take a moment to read through the terms of this agreement. Your

signature is required for the Uniersity Library to reproduce, translate,and make your submission publicly aailable through the Carolina DigitalRepository.

By signing this agreement, you (the author(s) and copyright owner(s)) grantthe Uniersity of North Carolina at Chapel Hill (UNC) the non-exclusie rightto reproduce your submission, translate the submission to any medium orformat for the purpose of preseration and public access, and/or publicly andglobally distribute your submission in electronic format.

You also grant that UNC may keep more than one copy of this submission

for purposes of security, back-up, and preseration.You agree that the submission is your original work, and/or that you hae

the right to grant the rights contained in this agreement. You also agree thatyour submission does not, to the best of your knowledge, infringe upon anyones copyright.

If the submission contains material for which you do not hold copyright,you agree that you hae obtained the unrestricted permission of the copyrightowner to grant UNC the rights required by this agreement, and that such third-party owned material is clearly identified and acknowledged within the text or

content of the submission.If the submission is based upon work sponsored or supported by an agency

or organization other than UNC, you agree that you hae fulfilled any right ofreiew or other obligations required by such contract or agreement.

UNC will not make any alteration, other than as allowed by this agreement,to your submission.

n I ACCEPT the terms of this non-exclusie deposit agreement

n I DO NOT ACCEPT the terms of this non-exclusie deposit agreement

Name (Please print):_______________________________________________

Signature: _______________________________________________________

Date: ___________________________________________________________

Documents

13385-16099-1-PB