geert-jan van bussel the theoretical framework for the 'archive-as-is'
an organization oriented view on archives - part ii
assumption that when a retention period has expired, records have lost their
organizational, legal, and historical relevance and should be irreparably destroyed
(Van Bussel, 2012a). For organizations of local, regional and national governments
the subsequent selection and disposal of records are most often mandatory.
Although not mandatory for non-governmental organizations, disposing of
irrelevant records saves (potentially high) costs for retention and accessibility.
Besides that, irrelevant records make organizations vulnerable to legal proceedings,
for instance in the context of privacy law, fraud or corruption (Van Bussel and
Henseler, 2013). The much disputed 'right to be forgotten' is an essential part of the
discussion on the relevance of records (Mayer-Schönberger, 2009; Stupariu, 2015).
The fourth dimension of information concerns the Survival (4) of records over time.
It pertains to the security and durability challenges, which have to be overcome to
realize access, retrieval, and preservation of records in spacetime (Bearman, 2006).
It stresses the importance of a reliable and durable ICT infrastructure to enable the
continuous and secure storage of records. The features of this infrastructure are
fragile and continuously influenced by the restructuring of organizations (Boudrez
et al, 2005). The challenge of preservation is almost overwhelming. First, hard- and
software configurations are always needed for accessing, retrieving and viewing
information, which means that a solution for technological obsolescence should be
available. Secondly, the large influx of information requires automated archiving
and retrieval functionalities. The ICT infrastructure needs to adapt, transform,
renew and grow, but this enhances the risks for obsolescence. Thirdly, records are
of a diverse nature. There is a diversity of object types, operating systems and
applications. The handling of this diversity is not self-evident, while it is, at the same
time, not impossible to change the content of records, which endangers the trust in
their reliability. Fourthly, records can only be reliably used, when they can be
interpreted by users in their original situational context. A case-based review of this
dimension has been offered by, among others, Hockx-Ju (2006).
4.3.2. The two archival principles (B)
I recognize two fundamental archival principles, an 'old' and a 'new' one, the
principle of Provenance (5) and the principle of (Environmental) Context (6)
respectively. Both principles are closely interrelated. It may even be difficult to
differentiate between them as a result of the intermingling of both principles within
archival scholarly literature. The principles are about the archive as a whole and,
indirectly, about the records within it.
The 'old' archival principle of Provenance (5) is seen as the 'foundation of archival
theory and practice' (Horsman, 1994, p. 51). This 'ambiguous concept' (Sweeney,
2008) has been a topic for scientific discourse since its introduction in the
eighteenth and nineteenth centuries. It still is. According to Shelley Sweeney (2008,
p. 194) 'over the years the principle has been introduced, reintroduced, applied in
part, applied in full, studied, and debated without end'. Giovanni Michetti (2016)
defines provenance (based on ICA definitions) as the relationship between archives
and the organizations or persons 'that created, accumulated and/or maintained and
used [them] in the conduct of personal or corporate activity'. It is also the
relationship between them and the functions that generated their need. The word
'provenance' refers, hence, to 'the origins of an information-bearing entity or
artifact' (Sweeney, 2008, p. 193). That is important, because archives 'should be
arranged according to their provenance in order to preserve [its] context, hence,
[its] meaning' (Michetti, 2016, p. 59). From its early history, the principle of
provenance was meant, first, not to intermingle archives from different origins
('respect des fonds') and, second, to maintain the internal structure of an archive in
its 'original order' ('archival bond') because it is a reflection of the functions of an
organization or an individual. Both are needed for an archive to have evidential and
informational value (Schellenberg, 2003; Posner, 1967; Horsman et al, 1998; Reilly,
2005).
Provenance has become a research object in other disciplines to see how it can be
used and represented in different contexts. In computer science, the interpretation
of provenance is that of data lineage, a description in the ownership history of how a
data object was derived (Buneman et al, 2001). Records can become an aggregate of
several information objects, may be stored in several locations, may be (part of)
databases, documents, spreadsheets, or emails, may cross organizational borders,
and may become part of one or more archives. Along the way, their origin and its
logistic history may become obscure, may contain gaps, or may be lost (Puri et al,
2012). Systems are developed that trace and analyse provenance across distributed,
networked environments, like Chimera in physics and astronomy, myGrid in
biology and CMCS in chemical sciences (Simmhan et al, 2005). In visual analytics,
it is recognized that the need to trace provenance extends beyond computing and
into the realm of human analysis (Lemieux, 2016). In computer science, the focus is
on individual items, while in archival science it usually applies to an archive or an
aggregation of records. Tom Nesmith (1999) associates provenance with the social
and technical processes of inscription, transmission, contextualization, and
interpretation of archives, which account for their existence, characteristics, and
continuing history. It broadens 'the idea of provenance to include its societal
dimensions' (Nesmith, 2015, p. 286). It is a postmodernist interpretation that
unmistakable intermingles provenance and context. Using the principle of
provenance proves to be complex when there is a 'parallel provenance, 'two or more
entities residing in a different context as establishing the provenance of [archives],
even when they are involved in different kinds of action, for example creation and
control' (Ketelaar, 2007, p. 186-187, based on Hurley (2005)).
The object of the principle of provenance is the (business process) archive of an
organization or an organizational chain as a whole and the structure of relationships
within that archive. It is not meant to contextualize archives. It only wants to
ascertain that: [1] archives (or aggregations of records) can be traced back to their
creator(s) and their creation, and [2] the 'archival bond' in which their records are
embedded can be reconstructed (Duranti, 1997b). For EIM the principle means that
metadata about the creation and logistic history of organizational archives are to be
preserved and that their internal structure(s) must always be reconstructable.
Nevertheless, tracing the history of individual records to safeguard the four
dimensions of information seems to be necessary in digital environments (Cui and
Widom, 2003). In reconsidering the archival principle of provenance, this is an
important reason to add data lineage to the implementation of the principle.
archives in liquid times
52
53