Trust Provenance is a crucial factor of evaluation when assessing the credibility of records on the Internet, therefore it needs to be investigated in order to shed light on the nature and the dynamics of the relationship between records and trust. The latter is a key concept of archival discipline. However, like provenance, it is a multi-faceted and cross-domain concept: trust is about voluntary vulnerability, in that it is based on a voluntary reliance on someone or something that may cause harm; ergo trust is about risk management. In fact, risk can be defined as a deviation - either positive or negative - from the expected (ISO 31000:2009, p. 1). Since trust "falls between hope and certainty," (Dietz, Gillespie, Chao, 2010, p. 13) it requires balancing confidence and control, that is, managing uncertainty, which is the essence of risk management; trust is a process, since the development of trust in systems as well as in people is informed by experience. Trust is built, shaped and assessed by applying known patterns to unknown situations. Therefore, trust changes over time, according to both the ever-changing factors that affect it, and people and systems' reaction to such changes; trust is contextual, because different systems for trust development and assessment are required for different contexts. Tools, agents, procedures, techniques and values vary according to the context; therefore, trust is a cultural thing. The parameters of trust in one cultural context may be very different from those of another context (Ferrin, Gillespie, 2010, p. 42-86). These parameters must be clearly identified and understood if cross- cultural trust - like what is needed on the Internet - is to be achieved; trust is an economic asset. In general, information has become a commodity with economic value. As a matter of fact, when exchanging information we exchange something that we consider valuable. Trust is the framework that allows such value to thrive and be exchanged.3 However, the commodification of data - which includes sale of personal information and other datasets as well as mash-ups of data, which in turn leads to creation of new data and value - is eroding trust and consequently the value of information. This is a crucial issue in the era of open data and big data. Like provenance, trust is a complex concept, this is the reason why it is not simple to deal with it when it comes to records. In fact, records provide evidence of our actions and thoughts, and they allow us to communicate across space and time. Such communication is deeply based on trust, to the point where trust is embedded into records. Records carry tokens of trust: signatures, seals, special signs, the documentary form itself, they all convey trust, not to mention the content, including wording and phrasing. Trust is involved in the transmission process too, since we place a certain level of confidence in the channel, the medium and the transmission service, including any associated agent and technology. (Duranti, 1998; MacNeil, 2000; Yeo, 2013, p. 214-234). The digital environment is no different, rather, it is much more complex. Digital technologies allow us to easily create, use and store documents on the Internet, where they can be de- and re-contextualized with little attention to their authenticity. Users have little control over how and where documents are stored in the Cloud, who has control and jurisdiction, who can access them, or how secure they are. In short, trust is at stake: in the digital realm we can no longer trust documents using the same approaches and tools of the past. Therefore, provenance plays a major role here, since it is one of the crucial factors that support trust. That is why we need methods, tools and metrics - along with a solid theory - to govern provenance and support the evaluation of reliability of digital objects on the basis of information on their provenance. Prior to the digital era, archival materials were trusted because of their characteristics - as we highlighted above - or their placement within a trusted repository, i.e., an archives, with preservation, access and use of documentary objects taking place in an environment or according to processes that were considered trustable. The digital environment has corrupted such belief. The challenge today is to do something similar to what has been done with markup languages: making explicit what is implicit. Archivists and records managers need to retain control of provenance and make it explicit, so that users are aware of the quality of the objects and trust them accordingly. The challenge is to find models, methods and tools to achieve this aim, solid enough to meet scientific criteria, yet easy enough to be managed by users. Preservation Preservation, including digital preservation, is about keeping objects together with the context that provides meaning to them, that is, the complex network of relationships - along with the system of their meanings - in which archival objects have been created, managed and used. Provenance is highly relevant in identifying and determining such context. Consequently, it is key to determining the identity of the objects targeted for preservation, because any definition of provenance, be it narrow or broad, will address at least creation and custodial history (i.e., the chain of agents that held the materials, together with related facts and events). In addition, the provenance of digital objects is itself a digital object that requires preservation. Therefore, provenance, and provenance of provenance are fundamental aspects in any preservation model, theory and practice.4 Access and use Access and preservation are two sides of the same coin. In fact, archival materials are preserved in order to make them available for use. However, "[i]n order to use a record, it must be accessible," (Kozak, 2015, p. 1) which means that policies and procedures should be designed and put in place to serve users' information needs. Provenance plays a role when accessing archival materials, since it is one of the key access points: the names of either the creator or the institution holding the archival materials are among the most common elements used in archival queries. Since provenance is more and more a complex network of relationships - if not a confused archives in liquid times 232 giovanni michetti provenance in the archives: the challenge of the digital 3 As Sissela Bok puts it, "[w]hatever matters to human beings, trust is the atmosphere in which it thrives." Sissela Bok, Lying: Moral Choice in Public and Private Life (New York, NY: Vintage Books, 1999), 31n. 4 Significantly, the OAIS (Open Archival Information System) model - the reference model for preservation adopted worldwide - requires that any object targeted for preservation must be accompanied not only with some Representation Information providing additional, higher-level meaning to the object, but also with some Provenance Information describing the object's history (i.e., origins or source, custodial history, changes, etc.). Provenance Information is in turn a digital object. As such, it must be accompanied with some Representation Information and some Provenance Information that will document the history of the Provenance Information. Such a recursive approach creates a complex network of Information Objects that need to be managed and preserved altogether in order to provide the proper context to the objects targeted for preservation, and to support their preservation over the long term. See ISO 14721:2012 Space Data and Information Transfer Systems: Open Archival Information System (OAIS): Reference Model (Geneva: International Organization for Standardization, 2012). 233

Periodiekviewer Koninklijke Vereniging van Archivarissen

Jaarboeken Stichting Archiefpublicaties | 2017 | | pagina 118