The concept of record in interactive, experiential and dynamic environments: The view of InterPARES

Archival Science

Volumn 6, Issue 1, 2006, Pages 13-68

  Duranti, Luciana ^a   Thibodeau, Kenneth ^b  

^a UNIVERSITY OF BRITISH COLUMBIA   (Canada)

^b National Archives and Records Administration   (United States)

Author keywords

Archival theory; Concept of record; Digital records; Diplomatics; Dynamic systems; E government records; Electronic art; Electronic science; Electronic systems; Experiential systems; Interactive systems

Indexed keywords

EID: 33750538869     PISSN: 13890166     EISSN: 15737519     Source Type: Journal    
DOI: 10.1007/s10502-006-9021-7     Document Type: Article

References (77)

1. See the InterPARES website at http://www.interpares.org.
2. There have been numerous projects that have examined the characteristics of digital documents. Most notable is the Open Archival Information System (OAIS) Reference Model, available at http://www.ccsds.org/docu/dscgi/ds.py/Get/ File-143/650x0b1.pdf.
3. The information model articulated in the OAIS standard has been the foundation of analysis of the characteristics of digital documents in several other projects, such as the CEDARS, PREMIS and Persistent Archives projects, respectively accessible at http://www.leeds.ac.uk/cedars/, http://www.oclc.org/ research/projects/pmwg/, and http://www.sdsc.edu/NARA/. However, these projects have developed their characterizations of digital documents with a view towards addressing the practical challenges of preserving them. In contrast, this article seeks to identify the characteristics of digital documents that are records. The goal is to describe these records in themselves. Such description must be independent of and transcend any and all approaches to preserving them.
4. Completeness here is not mentioned as a characteristic of the record, because an incomplete record is still a record, albeit a bad one, but as a characteristic of a fixed form, according to which a form that is fixed is one that does not lose any of its original elements in the process of being stored and retrieved.
5. The stability of the record, as determined by its fixed form and its unchangeable content, is only implied in the part of the archival definition that reads that a record is a document (i.e., rather than just data or information), but it is explicitly stated in the diplomatic definition and concept of record. (See Luciana Duranti, Diplomatics. New Uses for an Old Science (Lanham, Maryland, and London: The Scarecrow Press, Inc., The Society of American Archivists and the Association of Canadian Archivists, 1998), pp. 41-58.
6. This characteristic corresponds to the archival bond, which is implied in the archival definition when records creation is linked to an activity, but it is made explicit by archival theorists of all times and cultures. See Luciana Duranti, "The Archival Bond," Archives and Museum Informatics 11, nos.3-4 (1997): 213-218.
7. While characteristics 4 and 6 can be deduced from the archival definition, characteristic 5 derives from the diplomatic concept of record: it was considered important in order to distinguish records from digital objects resulting from simply querying a database. The author is the person issuing the record, the writer is the person determining the articulation of the discourse in the record, and the addressee is the person for whom the record is intended. As a record must participate in an action and any action must fall on somebody, the addressee is necessary to the existence of the record. See the Appendix 2 of the book The Long-term Preservation of Electronic Records: the InterPARES Project on the InterPARES website at http://www.interpares.org/book/index.cfm.
8. This book is also distributed in print by the Society of American Archivists: Luciana Duranti ed., The Long-term Preservation of Electronic Records: the InterPARES Project (San Miniato, Italy: Archilab, 2005).
9. The reason why the constituent parts of the record ended up in the template that is supposed to represent the ideal form of a record is that all identified constituent parts used to be regarded as necessary extrinsic elements of form by traditional diplomatists. It was important to show their presence, definition and purpose, and the fact that they are now independent of form.
10. In a previous research endeavour commonly known as the UBC-DoD project, the parts constituting the records were identified as: medium, form, action, persons, archival bond, content and context. See Luciana Duranti and Heather MacNeil, "The Preservation of the Integrity of Electronic Records: An Overview of the UBC-MAS Research Project," Archivaria 42 (Spring 1997): 46-67;
11. and Luciana Duranti, Terry Eastwood and Heather MacNeil, Preservation of the Integrity of Electronic Records (Dordrecht: Kluwer Academic Publishers, 2002): Chapter 1. In the context of InterPARES, it was decided that action, persons, archival bond and content, contrary to the other constituent parts, continue to manifest themselves in formal elements and are inextricable from them, so they do not have to be identified separately from the form. As it regards the annotations, which in the UBC-DoD project were included among the elements of form, they were added to the constituent parts because they are often linked to the record rather than embedded in it, and need therefore to be looked at separately from the record form.
12. Defined as "The rules of representation according to which the content of a record, its administrative and documentary context, and its authority are communicated."
13. Defined as "The elements of a record that convey the action in which the record participates and its immediate context."
14. Defined as "The elements of a record that constitute its external appearance."
15. See "Template for Analysis,"Appendix 1 in The Long-term Preservation of Electronic Records: the Inter P ARES Project, cit.
16. Defined as "Additions made to a record after it has been created."
17. Defined as "The framework of action in which the record participates."
18. For details related to annotations and contexts, see the Template for Analysis referenced above.
19. Defined as "The physical carrier of the message."
20. An additional reason for the InterPARES team to dissect the concept of record was to identify what parts or elements contribute to the authenticity of the record and to the ability to verify it.
21. There may be only minor differences between the form in which a record used in interactions between systems, rather than presented to a human, is stored on a digital medium and the form in which it is used in automated transactions; nevertheless, it remains true that the way the digital data which constitute the record are inscribed on a physical medium, any physical medium, and the form in which they are transmitted between systems or the form in which they are stored in a computer's memory during transactions are never identical. In contrast, a traditional, analog record is inscribed on paper and transmitted and read in exactly that form.
22. 'Presentation' in this analysis means the act of materializing the overall and specific presentation features of an electronic record or the result of this action.
23. A defining characteristic, or attribute, of the record element "seal" may be its legend.
24. A record profile is an annotation inextricably linked to the record that includes several fields, which are either automatically or manually filled in with the record's metadata.
25. Traditionally, works of art and recordings of performance art would be regarded as end products of artistic activities, rather than records. However, the traditional definition of 'record' in archival science poses no restriction on the type of information object which may be a record. Provided it satisfies the requirements for records summarized in section 1 above, a work of art or recording of an artistic performance may be a record.
26. IEEE. Standard Glossary of Software Engineering Terminology. In IEEE Software Engineering Standards Collection. IEEE, 1990. Std 610.12-190.
27. See: http://www.interpares.org/display_file.cfm?doc = besser_eci.pdf.
28. Of course, bank officers/employees are involved in producing the record of the transaction in so far as they are responsible for establishing and enforcing the bank's ATM policies and procedures (which would make them the de facto writers of the record), and ensuring that these policies and procedures are translated into a system that in turn outputs appropriate transaction records in response to user requests/actions.
29. Again, there is human involvement via consideration of who the competent writers and authors of the records are, as well as who the addressees are, all of whom are persons, not computers.
30. Clifford Lynch. "Authenticity and Integrity in the Digital Environment: An Exploratory Analysis on the Central Role of Trust." In Authenticity in a Digital Environment (Washington, D.C.: CLIR, 2000), pp. 32-50, available at http://www.clir.org/ pubs/reports/pub92/pub92.pdf.
31. Ramesh Jain. "Experiential Computing," Communications of the ACM 46(7) (2003): 48-54.
32. Seamus Ross. 2000. Changing Trains at Wigan: Digital Preservation and the Future Scholarship. London: NPO Preservation Guidance Occasional Papers.
33. Summit Strategies. The 2005 Summit Seven: Dynamic Computing Gets Down to Business. Market Strategy Report 4EC-07. December 2004, available at http://www.summitstrat.com/store/4ec07detail.
34. A. E. Eiben and J. E. Smith. Introduction to Evolutionary Computing (Berlin: Springer, 2003).
35. Janos Sztipanovits, Gabor Karsai and Ted Bapty. "Self-adaptive software for signal processing: Evolving systems in changing environments without growing pains," Communications of the ACM 41 (5) (1998): 66.
36. Narinder Singh. "Unifying heterogeneous information models: Semantic tags support knowledge webs," Communications of the ACM 41(5) (1998): 37.
37. Jeff Sutherland and Narinder Singh. "Application integration and complex adaptive systems Association for Computing Machinery," Communications of the ACM 45(10) (2002): 59-64.
38. The distinction among these three systems/attributes remains a matter of debate among InterPARES 2 researchers, primarily because they are more arbitrarily and subjectively than logically and objectively defined. As noted earlier, experiential and dynamic systems are types of interactive systems. However, in light of the need to compartmentalize them, another way of distinguishing them is by their relative levels of direct, real-time, human user involvement; with experiential and dynamic systems situated at opposite ends of the spectrum, and interactive system falling somewhere in between. In fact, from the definitions here provided, it seems that the only real difference between the interactive and experiential attributes is that only the latter ipso facto requires direct, real-time, human user involvement. One cannot help wonder whether the nuanced, largely subjective, distinctions among these three types of system attributes are: (a) actual and workable, (b) necessary, or (c) ultimately helpful.
39. As established in the first phase of InterPARES, a computer system may be said to contain an electronic document when it has the capability of reproducing that document. It may, but does not necessarily, do this by storing the document as a single data object.
40. Archival literature frequently describes records as consisting of content, context and structure. The discussion above describes the relationship of the three types of data to content and structure, but not to context. This is because, as described in section 1 'Findings of InterPARES 1' above, the significant context is external to the record. It is constituted by the record's relationship to other records, the administrative environment in which the record is created and maintained, the action in which it is involved, and the persons involved in its creation. Some of the content of a record may indicate or reveal its context; nevertheless, it is made of content data.
41. Jean-Francois Blanchette, Francoise Banat-Berger, and Genevieve Shepherd, Computerization of Alsace-Moselle's Land Registry, InterPARES Case Study CS18. 21 September 2004. See also: http://www.interpares.org/display_file.cfm?doc = ip2_alsace_characterization.pdf.
42. The Long-term Preservation of Authentic Electronic Records: Findings of the InterPARES Project, available at http://www.interpares.org/book/ interpares_book_f_part3.pdf, p. 6.
43. The Long-term Preservation of Authentic Electronic Records: Findings of the InterPARES Project Ibid. pp. 7-8. For a fuller discussion
44. see How to Preserve Electronic Records, Appendix 6 in The Long-term Preservation of Authentic Electronic Records: Findings of the InterPARES Project, available at http://www.interpares.org/book/interpares_book_o_app06. pdf.
45. If the system does keep an audit trail, the relevant concern is not necessarily whether the system can reproduce every sequence and consequence of every "input → computation → output" process, but rather whether the system can reproduce only those process sequences and/or consequences that correspond to what the user (or the author/system, depending on the perspective involved) identifies as the documents/ records that correspond to the user's interaction with the system.
46. There is some disagreement among music theorists on whether the score is also a musical work in its own right, but this does not invalidate the the ideas presented here.
47. The digital environment enables an artist to record an artwork in different documentary forms. In addition to the documents enabling performance and the recording of a performance, the InterPARES 2 case study Waking Dream is documented on a website, which is regarded by the principal author as part of the art work, and which includes the 'script' along with information about the performance space and the gadgets used, résumés of the artists, and videos and images from performances. The web page is a document whose digital components include the web page itself and the objects accessible from the page, such as documentation in textual form, audio-visual recordings of actual performances, and sample still pictures of imagery projected during performance. The computer code is neither included nor described on the website. The web page is neither a recording of a performance of Waking Dream, nor a document enabling a performance, such as the script of a play or score for a musical work. Rather, it is an alternate form of presentation of the work. For a complete description of the case study, see: http://www.interpares.org/ display_file.cfm?doc = ip2_waking_dream (complete).pdf. The Waking Dream web page illustrates a third type of experiential document, one that enables a user to experience a work by interacting with the system: the nature of the interaction is in fact more active than passive, or more participatory than observational. This is a more complex form of an interactive system, possibly including heterogeneous types of data and correspondingly more varied possibilities for presenting those data, and providing more flexibility or sensitivity in responding to user input.
48. Söke Dinkla. The History of the Interface in Interactive Art. 1994: http://www.kenfeingold.com/dinkla_history.html.
49. See also: http://a.parsons.edu/~praveen/thesis/html/wk05_1.htm1
50. and http://www.artmuseum.net/w2vr/timeline/Krueger.html.
51. For example, "Obsessed again..." an interactive piece for bassoon and computer by Canadian composer Keith Hamel. See http://www. interpares.org/display_file. cfm?doc = ip2_obsessed_again(complete).pdf.
52. InterPARES 2, Waking Dream Case Study Final Report, cit.
53. David Reim. "Online behavior: A brand builder's best friend," Pharmaceutical Executive 22(4) (2002): 104-108.
54. In this and many similar cases, the most likely and workable solution to the absence of a clearly identifiable entity that can be kept as a document is to establish intermediary 'termination' points (these could be predefined or randomly chosen to facilitate statistical analysis of the results, if desired, or they could be triggered whenever a predetermined set of user interaction criteria are met, etc.) in the ongoing process, when documents are produced that attest to the state of the system at those points.
55. Of course, one should not be induced to believe that the content and form are unique entities created on the fly from an infinite pool of possibilities, as they are simply manifestations of predetermined combinations of content and form that are then selected from a finite pool of such combinations in response to user behavior.
56. To a degree, this is precisely what the history cache in a web browser does, which users are able to reproduce by using the browser's back and forward buttons.
57. Because the documents in which the performance is captured are 'static' and/or are incapable of capturing all aspects of the actual performance, they ipso facto must embody less variability than is permitted by the rules describing the components, context, preconditions, or other requirements for the performance.
58. This is consistent with the definition of 'experiential' by InterPARES researchers, but broader. The project's definition is restricted to 'sensory experience,' but an experience could be intellectual or affective, as well as sensory.
59. Nicola Ferrier, Simon Rowe, and Andres Blake. "Real-Time Traffic Monitoring," In Proceedings of the 2nd IEEE Workshop on Applications of Computer Vision, Sarasota, Florida, 5-7 December 1994, pp. 81-88, available at http://ieeexplore.ieee.org/iel2/998/ 7985/00341292.pdf.
60. Noel Baisa. "Designing Wireless interfaces for patient monitoring equipment," RF Design April 2005: 46-54, available at http://www.rfdesign. com/mag/504rfdf4b.pdf.
61. For more information about the VabMap case study see: http://www.mterpares.org/display_file.cfm?doc = ip2_vanmap_characterization.pdf.
62. Dale Thomas. "Aesthetic selection of morphogenetic art forms," Kybernetes 32(1-2) (2003): 144-155.
63. One might argue that dynamic documents are not necessarily a subset of interactive ones. It is possible to conceive of dynamic documents that do not include any interactive features but generate varying displays from algorithms that alter themselves. However, such cases would be properly characterized as pseudocs or even applications, rather than documents.
64. 'Author' and 'writer' are terms used here as defined by diplomatics. See Footnote 6. When the author is an individual, it usually coincides with the writer. When the author is an organization or a collective or collegial entity, the writer is the person(s) who articulate(s) in writing its will, usually the signatory(ies).
65. Sometimes, rather than on a corner, they would write the data on the back of the medium.
66. Technically, the very first was an original, while the subsequent ones, lacking primitiveness, were copies in the form of original, although they had the force of an original. However, as they were all produced directly from the imbreviatura rather than from one another, and, having a different date of transmission and, possibly, a different writer (a notary who has legitimately succeeded to the original one) and a different addressee (the addressee of the action would remain the same, but the addressee of the record could be a descendant of one of the original parties), were different records, they can be all regarded as originals.
67. There is a subtle, yet important, distinction here that makes the correspondence to the Medieval analogy less than exact. The digital records are not being stored in documentary forms that are different than originally intended, rather, they are being stored in forms that are different than the forms in which they were originally created. The imbreviaturae (notwithstanding that they are complete records unto themselves) are, in one sense, incomplete 'stand-in' records for complete transaction records that, in most cases, were never created. This is not the same process that is occurring with the digital records, because the digital records are, in fact, created as complete records in their final form, and then saved in a form that differs from the form in which they were originally created. Thus, unlike the potential 'records to be' that are associated with the imbreviaturae system, the products of the digital systems discussed here are complete records created in their final form prior to being stored (except in those cases, as noted, where the minimum requirements for an e-record are not met).
68. Future analysis of how the different types of data are mapped into digital components should address how the system recognizes and processes the different data types in different mappings.
69. Such analysis, however, would only be partial. It is also necessary to determine, through additional analysis, whether the requirements related to context, action, persons, archival bond, and intrinsic elements of form are satisfied.
70. Prospective records still retain the basic function of remembrance: they enable subsequent actions and actors to remember what to do and/or how to do it in accordance with prior decisions.
71. Strictly speaking, computer code is not stored in the form in which it controls or shapes processes. It needs to be translated into machine code at the time of execution, but that translation is analogous to the translation of a musical score into signals processed by the human brain during performance of the work.
72. The situation is reversed when such records become inactive. In order for a human to understand what these records did in their original technological, documentary and administrative contexts, it is necessary to convert them from the form in which they were stored and functioned as records to a form that humans can read; for example, instructions must be converted from the binary form in which they were executable to a textual form. In most cases, this conversion will involve translation from machine language to a humanly readable one.
73. The InterPARES 1 report suggested another possibility, that of trading the record characteristics of stability of content and fixity of form (including completeness of content and form with respect to the first and to any subsequent instantiations of the record) with the ability of the system containing it to track and preserve any change to the record. See http://www.interpares.org/book/interpares_book_d_part1.pdf, p.24. In other words, the researchers were inclined to shift the requirements of stability and fixity from the record to the log of the changes to the record once the record was no longer active; in this context, the object identified as the record and to be kept intact would then be the last instantiation of the fluid object, plus the complete log of changes, and the metadata of both. This option is conceptually sound only if the creator uses this set of objects as its record, but this scenario is very unlikely because it would be highly impractical.
74. Tracey P. Lauriault, Peter Pulsifer and D.R. Fraser Taylor. The Cybercartographic Atlas of Antarctica Project. See http://www.carleton.ca/gcrc/ caap.
75. It might be argued that such a system does not necessarily need to maintain records of earlier steps in a process or earlier transactions, but that it would suffice if the system kept the data about those steps or transactions. However, if such a system does not demonstrably satisfy requirements for keeping records, it should not be relied upon to carry out business over time.
76. See http://www.si.umich.edu/CAMILEON/.
77. See: http://www.loc.gov/standards/mets/.