A metadata model for authenticity in digital archival descriptions

The advent of the information paradigm has shaken many of the principles of archival theory and practice. One key issue is knowing to which extent can digital information be trusted. Digital resources are represented by metadata, and trust consists in demonstrating their authenticity. Since the traditional elements used to verify the authenticity of analog records are not suitable in the digital world, the field faces a major challenge. The use of abundant, pertinent and constantly captured metadata seems to one of the most relevant solutions. This article aims to contribute to tackle this issue by setting the goal of proposing a model that attempts to include the most relevant metadata elements to capture the information that contributes for ascertaining the authenticity of digital archival descriptions. To that end, mixed methods methodology are employed. A qualitative documentary research is used to collect, analyze and interpret a corpus of scientific literature. As a complement, the quantitative technique requirements engineering is used to extract from international description standards the metadata requirements that can assist in the presumption of authenticity. Both approaches are then combined through a critical lens into a single unifying model for authenticity that is deemed as complete as necessary but as simple as possible. The model can be used by organizations or as a contribution to the discussion of authenticity and trustworthiness in digital archival descriptions.


Introduction
As technologies evolve and transform, so do the communication processes enabled by them.A document is information in a medium, so a change in the medium changes what and how information is captured and transmitted.Ultimately, this changes the message.Technology is not neutral since "we shape our tools and thereafter they shape us" (Culkin 1967, p. 70).
A change in information processes -from creation to transmission -requires an equal change in archival theory and practice, at least by constantly questioning its principles and methods in order to assess if they still remain relevant in the new contexts.The transition from an analog to a digital paradigm, i.e., from the management of records as physical entities to digital information management, has scrambled several core notions of the field, such as the principle of provenance and respect des fonds.Another aspect in which cracks have appeared is the presumption of authenticity, which traditionally relied on demonstrating a record's identity and physical integrity.In the digital environment, integrity is not a stable concept but a property of records that is verified by the preservation of a change log over time, since it is inevitable that a digital document is constantly changed and reinterpreted.
As a result, digital information management requires a constant stream of metadata to provide contextual information.Context, as expressed by metadata, can become more important than content, since it allows for content to be reconstructed or, from the perspective of a user query, it allows users to refine and filter search parameters to obtain a relevant corpus, even before viewing a single resource, known as positive search abandonment (Stamou & Efthimiadis 2010, p. 1).Since in the archival domain metadata emerges as a way of documenting context, it thus assists in preserving the identity and integrity of records, demonstrating their authenticity.This was one of the findings of the Inter PARES 1 project (Duranti & Blanchette 2004) that implies that archival description should not be seen as static, but rather as a "fluid, evolving and socially constructed practice" (Yakel 2003, p. 2).
The fluidity and social shaping of archival descriptions becomes particularly relevant the further we transition deeper into an interconnected web of resources, services and agents.The current vision of the Semantic Web aims to transform the web in a global database that enables links to be automatically created between resources as per the request of users (Machado et al. 2019).This intention relies on the resharing of metadata, which in turn requires interoperability between the various representation systems (or archival descriptions), understood as the seamless sharing of information across different systems.However, it is usually noted in the literature that interoperability in the web is hard, and as a result archival descriptions remain isolated (Rolan 2015).Additionally, since representation is subjective, how each community chooses to represent resources varies, further enhancing the problem of interoperability.
In order to facilitate the presumption of authenticity of resources in a shared web environment, and thus to ascertain their trustworthiness, an effort is 1 3 Archival Science (2023) 23:629-673 necessary to bridge these concepts across different communities.This work attempts to contribute to this gap in three different steps: 1) by performing a systematic review of the archival literature pertaining to archival description and metadata over a 10 year period; 2) to conduct a metadata mapping and comparison exercise drawing on ten relevant data structure standards; and 3) to merge both meta-analysis into a consolidated metadata model that can serve to represent the best practices in providing information about authenticity.As a result, it is a pioneer work that integrates the literature review as a critical theory for the analysis of the standards, that contextualizes the metadata elements featured in the standards in light of the previous literature review, documenting how that theory has been applied in practice, and consolidates those two analyses in a proposal of a metadata model that can be treated as a starting point for a reflection of the concepts of authenticity and trustworthiness, while at the same time serving as a representation of the most common metadata elements used or argued for in the field.By drawing input from a broad range of authors and publication venues over a large period of time, as well as from metadata standards from various practice communities and regions, the authors hope that it can be used both in practical application and/or as a starting point for theoretical thinking in how to increase the trustworthiness of descriptions and enhancing their interpretative context.
In this context, it is pertinent to revisit the principles on which the presumption of authenticity traditionally rested, to discuss if they remain relevant in the digital environment and, if necessary, to suggest an alternative.A mixed methods methodology is used pairing documentary research and requirements engineering complementarily in an attempt to rely on the scientific (articles in peer reviewed journals) and technical (metadata standards) literatures as input for the elaboration of the model.The first section of the article discusses the role of metadata in archives.The second section debates the constraints of representations and how authenticity can be conveyed.The third section presents the results of the analysis of the scientific and technical literatures, extracting as requirements the main ideas observed for the presumption of authenticity.The fourth and final section analyses and systematizes the requirements extracted previously, and consolidates them critically under a single model.

Definition and interpretation
Metadata are a form of archival description, which in turn is a form of information representation.Before delving into that discussion, it is important to reach an understanding of what metadata are, since it is a concept used across multiple fields with disparate interpretations.For example, Furner (2020) identified 96 separate ISO standards that provided 46 different definitions.The term 'metadata' was first used in 1968 by Philip Bagley, a computer scientist, who argued that one of the essential components of a programming language is "the ability to associate explicitly with a data element a second data element which represents data 1 3 Archival Science (2023) 23:629-673 However, the interpretation of resources changes because context is not fixed.Metadata are created by certain people at a certain point in time, therefore with some kind of bias.Representations are subjective and metadata, as a form of information representation, do not escape this fate.As Gartner (2016, p. 4) puts it, "there is nothing objective about metadata: it always makes a statement about the world, and this statement is subjective in what it includes, what it omits, where it draws its boundaries and in the terms it uses to describe it".The notion of metadata as statements is not only common in the archival literature, but also necessary to understand that metadata are not objective because they are a human construct and everything that is human is not neutral.Yakel states that "through the process of selection of information for inclusion and choice of access points, archivists reveal and conceal, making finding aids political statements" (Yakel 2003, p. 19), imparting the subjectivity on the finding aids themselves, i.e., on the archival descriptions.
In summary, "metadata is best understood as 'any statement about an information resource', regardless of what it is being used for, which metadata vocabulary is being used, and how the metadata is represented" (Garshol 2004, p. 379).The word 'statement' accounts for the subjectivity of the representation process.The 'information resource' can both refer to records, in the traditional notion of "recorded information produced or received in the initiation, conduct or completion of an institutional or individual activity and that comprises content, context and structure sufficient to provide evidence of the activity" (International Council on Archives 1997, p. 21), as well as any other object that can provide information, such as any entity that establishes a relationship with a record and is captured in the representation system, whether it is a person, an occupation, or a property.

Metadata interoperability
There are several forms in which metadata can be expressed.The most commonly used data models in archives are the relational model, which includes both tabular (e.g., an Excel sheet) and relational data (e.g., a relational database), the hierarchical model (e.g., XML) and the graph model (e.g., Resource Description Framework).These all have advantages and disadvantages, depending on the goals and constraints of the data curators.Despite emerging across the last decades, each still has a place in archival description, although serving different purposes.Data models can even be mixed.For example, data can be structured in XML but their relationships enhanced by RDF graph representations in the form of triples.
Regardless of which data model is used, it is key that data remains interoperable across different representation systems.Interoperability is one of the five key principles for the construction of metadata suggested by Marcia Zeng and Jian Qin (2016, p. 28), the others being modularity, extensibility, refinement and multilingualism.Interoperability can be understood as "the ability of multiple systems with different hardware and software platforms, data structures, and interfaces to exchange data with minimal loss of content and functionality" (NISO 2004, p. 2).Communicating across different systems so that data is minimally lost is not only a recurring problem for data curators, but it is also one of the greatest hurdles in metadata management.The proliferation of options and standards for describing resources in recent decades exacerbates this problem by multiplying the number of possibilities available.Since these standards are often developed in isolation, seamless compatibility is not automatic.It requires effort for information to be shared efficiently and automatically.In an ideal situation, users "should be able to discover through one search what digital objects are freely available from a variety of collections, rather than having to search each collection individually" (Tennant 2001, p. 118).Databases should allow for cross-queries in order to obtain as many pertinent results as possible from different contexts, thus adding value and meaning to the resources retrieved.In order to achieve this, metadata are "governed by community-developed and communityfostered standards and best practices in order to ensure quality, consistency and interoperability" (Gilliland 2016, p. 2).However, in order for an efficient sharing of metadata, semantic interoperability is required in addition to syntactic interoperability.Not only it is necessary to communicate, it is also crucial to know how to interpret the communicated data.
As a result, there are at least two levels of interoperability.Syntactic interoperability refers to the standardization of communications between both systems (Schaeffer et al. 2012;Sudmanns et al. 2018) and is made possible by a "strict adherence to standard metadata formats and protocols while creating metadata structures and possible only through metadata deployment" (Ramesh et al. 2015, p. 197).Without syntactic interoperability, "data and information cannot be handled properly with regard to formats, encodings, properties, values, and data types; and, therefore, they can neither be merged nor exchanged" (Zeng and Qin 2016, p. 348).In turn, semantic interoperability can be defined as the "capacity for different agents, services, and applications to communicate data, information, and knowledge while ensuring accuracy and preserving the meaning of that data, information, and knowledge" (Zeng and Chan 2010, p. 4646).Without semantic interoperability, "the meaning of the used language, terminology and metadata values cannot be negotiated or correctly understood" (Koch 2006, Interoperability).
Furthermore, interoperability can be observed at various levels, namely at a schema, record or repository level.Since this paper aims to suggest a metadata element set that can be used to represent the authenticity of resources, it focuses on semantic interoperability at the schema level, which comprises the scope of metadata standards.Several types of standards exist that accomplish different goals.Data structure standards govern the elements used in the metadata record.Data content and data value standards clarify semantics, whereas data exchange standards refer to syntax.Amongst these types, this work focuses on structure standards that elucidate about the metadata elements to be used, regardless of how those metadata are expressed by syntax or semantics.Thus, data structure standards (sometimes also called metadata element sets, element sets, schemas or metadata vocabularies) can be seen as "'categories' or 'containers' of data that make up a record or other information object" (Gilliland 2016, p. 3).Those categories are the metadata elements, also known as data elements or data fields.They are the units of information that a standard dictates that should be captured, in the hierarchical order defined by the data structure (Elings & Waibel 2007).

3
Archival Science (2023) 23:629-673 One of the goals of this paper is to suggest a set of elements that can be used to reflect the authenticity of resources across multiple digital contexts, as informed by the meta-analysis of the scientific and technical literatures.Therefore, in the context of this work, the resulting metadata element set is posited as a model for authenticity, in the sense that a model is an abstract representation of an object or a state for analytic ends.Being "an abstractive representation of some object or state of affairs" (Wartofsky 1979, p. 4), the model individualizes a property of an object.It does not require a complete and reliable representation of the object, but merely a similarity to some of its properties.In this case, the object is archival description.The particular property to be studied is authenticity, not all the other characteristics of records.Thus, one of the aims is to create a model for authenticity in digital archival descriptions, manifested in the proposal of a metadata element set.It does not aim to be exhaustive, but rather to be as complete as necessary and as simple as possible.However, before attempting to do so, it is important to understand what authenticity encompasses in order to be able to establish criteria for populating the content of the model.

Information Representation
First, it is important to understand that all descriptions in archives are information representation, which in turn are a form of representation.Rosenberg argues that "the essential and characteristic human activity is representation-that is, the production and manipulation of representations (Rosenberg 1981, p. 1).Representations have certain philosophical constraints that emanate throughout the represented objects.One is that a representation is, at its fundamental level, "something that stands, or is believed to stand, for something else" (Yeo 2018, p. 129).The solar system is studied in a classroom using balls that represent planets, with the essential condition that students interpret those balls as an accurate representation of the actual planets.Representations serve a purpose of bringing to attention a distant object, acting as "surrogates for memory" (Jimerson 2015, p. 99).They make present something that is absent.In the case of information representation in archives, archival description enables us to discover and learn about records that are held elsewhere.This aspect is particularly important in the digital realm, in which we interact exclusively with representations in our screens.Resources are searched, aggregated and contextualized using metadata and, if considered pertinent for the objectives of the user, they may consult a digitalized version of the resource (such as a PDF).In archives, records can also be considered a representation of the actions they describe.However, this study focuses merely on archival description as representations of resources, and as metadata as statements about those records.
Additionally, one of the most critical axioms is that a representation is a simplification of an object.Therefore, they are necessarily incomplete.That is a consequence of the reason why we represent -to reduce the complexity of an object so that it can more easily be studied and manipulated.In the classroom example, the balls might have the same shape and relative distance to one another as the actual planets, but do not possess their mass.If some properties of the object are represented but others are not, there is a choice.Such conscious and deliberate choice is made by the creator of the representation.Consequently, no representation is neutral.
The consciousness about the subjectivity of representations has been introduced in archival science by several postmodernist thinkers (Brothman 1993;R. Brown 1991;Harris 1997;Hedstrom 1993;Nesmith 1999;Upward 2000).Modernists, inspired by the French philosopher August Comte (1798-1857), believed that knowledge was limited by what could be experienced, or positively determined.It was believed that science was above any cultural, historical, political or social context, encapsulating the truth within itself, as if it were an objective truth.Post-modernists challenged this notion, arguing that the human perspective conditions the interpretation of phenomena.A consequence is that archivists are no longer seen as "passive guardians of an inherited legacy" but instead as "actively shaping collective (or social) memory" (Cook 2001, p. 4).Archivists shape memory by selecting what to preserve and then by building representations of what was kept.How records are represented by archival description has also been a contentious topic in archival science.The traditional approach has been to the fonds, according to the principles of provenance and original order.The several international standards for archival description published by the International Council on Archives (ICA) throughout the 1990s have prioritized collection-level descriptions of records, according to a hierarchical structure from the general to the particular, from the fonds to the item.The ICA has since then changed its shift and, in its latest standard Records in Contexts, it acknowledges that this focus on the person or group that has accumulated a body of records "often does not reflect the social and material complexity of the origins of records" (International Council on Archives 2016, p. 5).
Bailey also protested against the idea of the fonds as the single method of archival description, since "in a database, objects are related but not ordered.The database logic is nonlinear and there is no original order because order is dependent upon query.… Digital objects will have an identifier, yes, but where they 'rest' in intellectual space is contingent, mutable.The key point is that, even at the level of representation, arrangement is dynamic: access and representation need not depend on the fonds" (Bailey 2013).Similarly, Anne Gilliland reinforces the criticism of using exclusively collection-level, hierarchical metadata arguing that, while they remain valuable for retaining context and original order, it "represents an oversimplified view of the actual complexities of records-creation process and provenance … and necessarily perpetuates a paper-based descriptive paradigm" (Gilliland 2016, pp. 5-6).
As a result, in addition to collection-level metadata, it becomes a requirement for digital archival representation to also include item-level metadata, so that resources can be easily recontextualized in novel aggregations upon query by users.Despite some authors believing that "a relentless focus on the aggregate part is what sets us [archivists] apart from librarians and museum curators " (Greene 2009, p. 24), David Bearman suggests that "item level information is fundamentally more valuable because it can generate more valid collective level data in addition to serving the needs of item documentation" (Bearman 1996, p. 205).Item-level description seems to be a necessary complement to collection-level descriptions.Both serve distinct but crucial roles.While collection-level description communicates the context of a records producers and provenance, item-level metadata facilitates the discovery of resources and their recombination in different aggregates, multiplying their contexts as well as their potential interpretations and meanings.As Terry Cook summarizes: Nothing is neutral.Nothing is impartial.Nothing is objective.Everything is shaped, presented, represented, re-presented, symbolized, signified, signed, constructed by the speaker, photographer, writer, for a set purpose.No text is a mere innocent by-product of action … but rather a consciously constructed product … Texts are all a form of narration more concerned with building consistency and harmony for the author … than they are evidence of acts and facts, or juridical or legal frameworks.And there is not one narrative in a series or collection of records, but many narratives, many stories, serving many purposes for many audiences, across time and space (Cook 2001, p. 7).
It is not the role of archivists and similar data stewards to limit which queries can be made beyond the legal restrictions (such as embargos), but instead to facilitate discovery and reuse of resources through adequate and complete enough archival descriptions that maximize the relationships established by resources, enabling users to construct their queries in a manner that is not pre-determined.This notion reinforces the importance of developing a model for authenticity in archival description that can serve as a guideline for these professionals.

Authenticity, trustworthiness and reliability
The importance of authenticity for the archival profession is attested by its presence as one of the fundamental purposes of an archivist's ethical code: to ensure the authenticity of records under custody (International Council on Archives 1996; Society of American Archivists 2012).Authenticity refers to "trustworthiness of the record as a record, i.e., the quality of a record that is what it purports to be and is free from tampering or corruption" (InterPARES 2 2008, p. 8).It requires establishing the identity and demonstrate the integrity of records over time.Identity is understood as "the whole of the characteristics of a document or a record that uniquely identify it and distinguish it from any other document and record" whereas integrity refers to "quality of being complete and unaltered in all essential aspects" (Inter-PARES 2, 2008, pp. 25-26).To question identity is to ask "was it written by who purports to have written it?",while integrity attempts to answer the question "has it been altered in any way since it was first created and, if so, has such alteration changed its essential character?"(MacNeil 2000, p. 53).
The ICA also defines reliability and usability as two other essential characteristics of records.Reliability means that "a record can be trusted as a full and accurate representation of the transaction(s) to which they attest, and can be depended on in the course of subsequent transaction", whereas usability implies that records can be located, retrieved, preserved and interpreted (International Council on Archives 2008, p. 13).The relevance in establishing that a record is authentic and reliable is related to the ability to demonstrate its trustworthiness as a representation of the actions it manifests.Trustworthiness is "the quality of being dependable and reliable" (Society of American Archivists 2005, p. 388), and is generally defined as hinging on reliability and authenticity.
In the analog paradigm, authenticity was demonstrated by a tight control on recordkeeping processes.In the digital realm, integrity is constantly being challenged as information flows across systems and mediums.Therefore, it has become more consensual that, when managing digital resources, the elements that contribute to the presumption of authenticity "are observable not in the document itself but rather in the procedures" (Eastwood 1994, p. 127) of record creation, maintenance and preservation.Since archival descriptions are basically a metadata element set, they feature as one of the key procedures for capturing the authenticity of records (InterPARES 1 2002), in the sense that they clarify the arrangement of records and the circumstances (the contexts) in which they were created (International Council on Archives 2016;MacNeil 2005).The link between authenticity and archival description is further documented by the Canadian Rules for Archival Description, when stating that one of the goals of archival description is "to establish grounds for presuming the authenticity of archival material by documenting its chain of custody, arrangement, and circumstances of creation and use" (Bureau of Canadian Archivists 2008, p. xxii).Given the evolution in importance of integrated digital descriptions, it becomes increasingly important that metadata is interoperable across information systems and archival collections, so that the chain of custody, arrangement and custody of records are preserved, and their authenticity can be reinforced.
It is also relevant to highlight that authenticity is not a binary question, so that something would be simply authentic or not (Factor et al. 2009).First, authenticity can only be understood in reference to the goals of a user.As Bonnie Mak explains, "the purpose of establishing the authenticity of a record is to position it as trustworthy for a particular purpose, and authoritative within a particular framework" (Mak 2015, p. 122).Furthermore, "the level of confidence one can have in authentication is related to the strength of the evidence, and should not be considered absolute" (Suvak 2015, p. 117).As a consequence, it is observed that there is a gradient in authenticity (Carta 2017, p. 195), a spectrum in which one end is 'more authentic' and the other 'less authentic'.The middle of this spectrum is filled with a grey area of nuance.In practice, this means that archivists cannot claim that a resource or a description is authentic.At most, they can provide the maximum amount of evidence possible so that a user determines if, in their perspective, the object represented is authentic and, therefore, trustworthy.Authenticity is itself a subjective and personal notion, since what might be enough for a user may not be sufficient for other.As a result, the model discussed in this paper cannot have the pretentious goal of ensuring the authenticity of archival descriptions, but merely to provide a guideline to capture certain information about resources so that, when present, they may be complete but simple enough so that users can form their own judgement, and so that such judgement may with greater likelihood conclude that the descriptions are authentic, reliable and trustworthy in as many contexts as possible.

Methodology
The plurality of perspectives, backgrounds in authors of the field, as well as the different scopes of several metadata standards, calls for a consolidated rethinking of how to represent authenticity in the digital web, a problem acknowledged by the International Council on Archives (2019) when developing Record(s) in Context(s), the latest international archival standard.The main goal of the study is to propose a model that can help bolster the presumption of authenticity in digital archival descriptions.In order to do so, it commits to an extensive review of the scientific and technical literatures as data points for the construction of the model.Documentary research (C.G. da Silva 2021) is employed as a qualitative method, determining the need for assembling a literary corpus and positioning the researchers as the interpreters of the meaning of the texts.The first step in selecting the corpus of scientific works that comprises the scientific literature review was selecting the database Library, Information Science and Technology Abstracts (LISTA) as the source for data collection, due to the relevance of publications it indexes and availability of the texts.Within LISTA, there was an initial query with the terms "Title = 'archiv*', 'arqª' or 'records management'".The goal was to identify the most relevant (high Índice Compuesto de Difusión Secundaria, or ICDS) journals that versed directly about archives or records management, as defined by their presence in the title.
This query yielded on May 2019 a total of 25 results, which were then further filtered by excluding the journals with an ICDS < 6.5, not being peer-reviewed, not having descriptive information, not being active, not being in English, Spanish, French or Portuguese, or being off-topic for the purposes of the research.The result of the application of these filters was the selection of 12 journals deemed appropriate for the research goals: American Archivist, Archival Issues: Journal of the Midwest Archives Conference, Archival Science, Archivaria, Archives [London], Archives [Quebec], Archives & Manuscripts, Archives & Records, Cadernos de Biblioteconomia, Arquivística e Documentação, Journal of Archival Organization, Provenance: the Journal of the Society of Georgia Archivists and Records Management Journal.Once the journals were identified, all articles published between 1 January 2009 and 31 December 2019 were retrieved.Their pertinence and inclusion in the study was analyzed through reading the title and abstract.A total of 213 pertinent articles were extracted from all journals.
As a complement to the scientific literature review, it was considered important to also include in the model the insights from data structure standards for archival description, since these instruments play a key role in influencing archival practice.Since no single standard can adequately capture every context of representation of the several types of collections or resources across different communities (Apostolou 2009;Gilliland 2016), it was decided that the sample should include data structure standards from multiple fields (archives, libraries and museums), geographies (Europe, Americas and Australia) and languages (Portuguese, English or Spanish), and available for free on the Internet.Based on these criteria, the following ten standards were chosen: General Standard for ISAD-(G) and DACS were selected for being international archival standards; AGRkMS for being an archival standard used in Australia; EAD for being an international standard focused on finding aids; e-EMGDE for capturing the Spanishspeaking worldview on public recordkeeping; DCMES for being a multi-purpose international standard with broad application in representing content on the web; VRA Core to include in the analysis an international perspective on description of visual resources and images; MODS to include the bibliographic approach, as it is used internationally; CDWA for its focus in the representation of artworks in museum contexts; and finally, RiC v0.2 was included, albeit being still in draft version, due to its anticipated importance in shaping archival description worldwide, as it is the latest ICA international archival standard, and has been designed specifically with a concern of how to capture the multiple contexts in which resources circulate.The element set should not change significantly in the final version, thus it is expected that the current analysis remains relevant.Thus, the metadata mapping would be incomplete without its mention.Other standards could have been included, but it was deemed that these ten, given their diversity in scope and application contexts, were sufficient for the goals of the study.
These standards were analyzed using the quantitative technique of requirements engineering (Bennaceur et al. 2019).It is a technique borrowed from computer science, in which it is used for software development, since it is useful for defining the requirements that a certain system must have in order to function.Within the scope of this study, the goal is to use this technique to elucidate about the requirements of authenticity, as expressed in the technical standards.Those requirements are first derived from archival theory, according to the scientific literature review, and then critically analyzed in the standards.The expectation is that requirements engineering can assist in identifying the metadata elements employed by the standards most directly correlated with the demonstration of authenticity and, thus, trustworthiness of descriptions.As a result of the combination of qualitative and quantitative techniques, this research follows a mixed methods approach (Creswell & Clark 2013).
The use of mixed methods is particularly useful for this study since quantitative methods allow for a detailed understanding of a problem, although usually lack the ability to extrapolate those results to a broader social context.In contrast, qualitative methods are usually limited in their capacity to analyze large volumes of data, and are more heavily biased by the subjectivities of researchers.Therefore, they are used in a complementary manner, so that the strengths of the qualitative approach compensates for the weaknesses of the quantitative, and vice-versa (Brewer & Hunter 2012;Creswell & Clark 2018;Jick 1979).In the context of this study, it is expected that documentary research illustrates how to represent resources' authenticity through archival descriptions, but by itself would lack a broader context on how authenticity is captured not only by archival standards, but also in other communities of practice.On the other hand, the exclusive use of requirements engineering would provide a detailed, comparative list of each metadata element used, but would be void of the critical thinking to analyze, debate and frame them within the theoretical discourse.Their complementary use as part of a mixed methods approach reinforces the ability to conduct the meta-analysis of the scientific literature, to be able to translate that learning into the specific metadata elements used by the standards and, in turn, to interpret and contextualize those elements in light of the literature review.

Scientific requirements
In order to facilitate the analysis of the vast scientific literature, it became necessary to group the articles thematically.As a result, the following emergent categories were observed based on data analysis: (1) information representation (IR); (2) archival description (AD); (3) epistemology of information management (IM), sometimes referred to as records management; (4) metadata standards (MDS); (5) epistemology of metadata (MD); (6) authenticity and trustworthiness (AUT); ( 7) and linked data (LD).Table 1 illustrates some of the most relevant works and authors identified in each category.These categories form the thematic pillar of the literature review, in which it is possible to categorize the majority of works identified in the documentary corpus.Articles were aggregated under 'Information Representation' as it was noted that they frequently focused on the information mediation between archive users and how archivists organize collections.The category 'Archival Description', albeit a part of information representation, was chosen to adopt terminological consistency with standards for archival description.These texts tend to question the active and subjective influence of archivists in description of resources and in the recreation of context, revisit the essential functions of archival description or explore the relationship between archival description and digital curation.'Information Management' articles tend to focus specifically on procedures for information management and organization in archives.
The category 'Metadata' was chosen to represent the scientific discussions closest to the concept of metadata, particularly from an archival perspective, although some of the authors included are not archivists, which brings a welcome diversity into the analysis.The category 'Metadata Standards' captures research that focuses on the topics of mapping between metadata standards, difficulties in implementing standards, guidelines for the development of standards and guidelines for adapting metadata standards to the Semantic Web.The category 'Authenticity' captures the works which explicitly aim to debate the authenticity and/or trustworthiness of resources, documents or information.Finally, under 'Linked Data' are included articles that problematize archives from the perspective of concepts such as the Semantic Web and linked data, therefore usually focused on interoperability, as well as in the sharing and reuse of digital archival descriptions.
A detailed list of the bibliographic references of articles that were consulted but do not featured as direct citations is included in Appendix 1.Each single article in Table 1 was read and analyzed under the lens of which criteria it identifies as a precondition for authenticity to be inferred from archival descriptions.A summary of the emerging requirements found in each category is provided in Appendix 2.

Technical requirements
For the ten standards that comprise the technical corpus, previously identified in the methodology, a detailed analysis of their data structure was performed.Each metadata element was considered and, based on the conceptual notions of authenticity as seen in the literature review, a decision was made whether or not to consider them relevant in establishing the authenticity of resources.Since authenticity is comprised of the demonstration of identity and the verification of integrity, the analysis focused on these two parameters.It is noteworthy to observe that authenticity was found in relation not only to resources, but also to the descriptive practice itself.Metadata elements that identify the creator of the archival description, as well as notes or rules followed, were also considered as key for authenticity, 1 3 Archival Science (2023) 23:629-673 which in this case can be seen as meta-metadata, or elements for the control of descriptions.
Thus, a summary of all metadata elements considered relevant for informing about the authenticity of both resources and descriptions themselves is presented in Table 2.The elements are grouped according to their focus: identity, integrity, and description's control, and are separated vertically by a blank space.It should also be noted that, due to time and analytical capacity constraints, not every single metadata element of the standards was considered, but only the first two levels of representation -categories and sub-categories.From the 26 elements of ISAD(G), 16 were identified as potentially relevant to asserting authenticity, among which six were considered mandatory by the standard creators; from AGRkMS's 27 metadata elements and 44 sub-properties, 12 were chosen; regarding EAD, 29 metadata were selected amongst the 165 metadata elements and 85 attributes; in the case of e-EMGDE, amidst its 30 elements, 21 were selected for the mapping; from DACS' 25 elements, 16 were considered related to authenticity; from DCMES' essential set of 15 elements, ten elements and four refinements were selected; for VRA Core 4.0, 17 metadata terms were chosen amongst the 19 categories and their sub-elements; from MODS' 20 elements, 12 metadata elements and four sub-elements were selected; from CDWA's abundant 540 categories and subcategories were chosen 51 elements and sub-elements; and finally, from RiC were extracted 18 entities and attributes based on the drafted 22 entities and 44 attribute.

Analysis and discussion of the scientific requirements
The identification of the scientific requirements in Appendix 2 enabled an overview of the discussions on the concept of authenticity across several perspectives and authors within the field.In order to further facilitate the analysis, the unique requirements were traced and systematized in Appendix 3. A total of 32 requirements were identified.The most common ones are requirements 1, 7, 2, 4, 10 and 17.This seems to indicate that a need for greater transparency and accountability by the creators of the descriptions (requirement 1), as consequence of the growing consciousness of the subjectivity of the representation process (requirement 10) is generally recognized in the archival literature.Therefore, the intellectual control that archivists elaborate can and should be questioned, enabling the possibility that resources belong simultaneously to multiple contexts and collections, depending on the interpretation perspective (requirement 4).Furthermore, it is also generally agreed upon that metadata must be collected across the resources' lifecycle (requirement 7) and must include abundant and detailed contextual information as a way of documenting provenance (requirement 2) and strengthening the verification of their integrity (requirement 17).The fact that these requirements feature in the majority of categories (4 or more) does not mean that less frequent ones are less important.An argument could be made that those are even more relevant, since they are more frequently overlooked.As such, the model considers all requirements equally.
The requirements seem to be structured across four dimensions: (1) theoretical insights on representation; (2) metadata about the resource, (3) about the creator of the resource and (4) about the control of representations.A systematization of this interpretation is summarized in Fig. 1, which serves as the final analysis of the information extracted from the scientific literature.

Analysis and discussion of the technical requirements
The data extracted from the international standards was analyzed in a similar way to the articles.The metadata elements identified across every standard, as shown in Table 2, were standardized and the unique requirements were identified.Appendix 4 provides a list of the unique requirements, serving also as a crosswalk between the 10 standards that were scrutinized.The distribution of the metadata elements also sheds some light on the differences between the standards.ISAD(G), as an ICA international standard for archives, includes several elements directly related to the archival perspective.It is also a reflection of the time of its creation, in the 1990s, revealing a documentary inclination for the management of records as physical entities and for a multilevel, hierarchical structure.As for AGRkMS, a standard for records management in public administration, it shows some similarity with archival practices, but is more focused on digital information when compared with ISAD(G).It is also notable for the absence of metadata elements for control of the descriptions.
The nature of EAD as a standard for the codification of archival descriptions is clearly made evident by its metadata structure.Some of its elements, like archival description, biographical history, custodial history and appraisal are characteristic of elements used in information management in archives.As such, it is closely related to ISAD(G) and AGRkMS, but distinguishes itself by also focusing on the transparency of the descriptions.The standard e-EMGDE is a product of the juridical context that governs record creation in Spanish public administration.In addition to the traditional elements used in archival description, it distinguishes between public and private documents, and shows some characteristics of a focus on the medium.Here it is also found the absence of metadata for the control of descriptions.
The simplistic approach of DCMES is reflected upon its vaguer designations and omission of some of the most usual elements used in archives, such as administrative history.Instead, it seems to be an essential set of elements necessary for the basic identification of resources.VRA Core is another standard that is not from the archival community, but is instead used for the representations of artwork.Therefore, it includes some elements important for this community, such as cultural context.There is also an almost absolute absence of control metadata.Likewise, MODS is a standard for the description of bibliographic resources commonly used in libraries.It also omits certain elements of the archival domain but, contrarily to VRA Core, emphasizes the control of descriptions by documenting their author, language, sources and footnotes.Another standard used in artwork is CDWA, which focuses on artistic objects such as architecture.It dedicates great detail to the physical characteristics of works, such as techniques and construction methods, but seems to disregard metadata related to the digital management of resources.Lastly, RiC v0.2, although still in consultation draft, as the latest international standard for archival description, seems to consolidate the representation of information communities that are close but sometimes distant, such as archives, libraries and museums.As a result, the metadata elements do not seem to be exclusive to any of these communities, but rather serve as a flexible conceptual model that each information manager may tailor to their needs.However, there is a lack of explicit metadata for the control of descriptions, which could be a useful addition.
Despite the differences between some standards, some requirements stand out as most frequent.Considering the universe of 10 standards, the creation date of resources (T4 in Appendix 4) features in every single one.The resource's title (T2), creator (T3), physical dimensions (T5), administrative history (T16), current location (T19) and related resources (T20) feature in 90% of standards.The use of an identifier for resource description (T1) and the custodial history occur in 80% of standards.With less frequency albeit in the majority of standards is also observed 1 3 Archival Science (2023) 23:629-673 in 60% of the sample the description of the physical characteristics of resources, such as documentary form or type of resource (T11), description level (T13) and the classification system (T14).In 50% of the standards are found the physical medium of the resources (T6), information about appraisal, including destruction (T22), additions to the collection (T23), the author of the descriptions (T27) and their date (T28).
As with the scientific requirements, all requirements were considered equally when constructing the model, even the least frequent ones.The final assessment of the metadata elements to include is presented in Fig. 2, which seems to orbit around four dimensions: metadata about the resource, the aggregate, the creator, and the control of representations.

Proposed model
A comparison between the requirements extracted from each approach reveals some similarities.For example, both perspectives include arguments for the description of resources, creators, or control of the representations.Regarding the creators, both technical and scientific literatures suggest the indication of their name and biographical history.As for meta-metadata, there is also a juxtaposition in recommending the creator of the description, the data, a changelog and creator notes.However, the scientific literature is usually more summary, repeating a smaller number of essential requirements.As a result, as far as control metadata are concerned, the technical literature includes all these and a few more, so is more complete.To counterbalance, the scientific literature details more completely some considerations for the representation process, including theorizing the objectives of description and some implementation guidelines.
Once the contributions of each portion of the corpus have been clearly articulated, a comprehensive analysis that uniformizes them is now possible.It is worth remembering that the ultimate goal of the model is for the resources it describes to be considered trustworthy, which is made possible when authenticity is established.The analysis is driven by an overarching concern to capture not only the identity and integrity of the descriptions of resources, but also the identity and integrity of the descriptions themselves.Based on these premises, the interpretation of the data yielded 42 metadata elements that comprise the model proposed, reproduced in Fig. 3.It is believed that these elements capture the most relevant dimensions of authenticity.It does not simply combine all metadata elements from both data sources, since that would result in a long list that could be hard to implement.Instead, it aims to be a trade-off between simplicity and completeness, in the sense that it intends to be simple enough to be easily implemented, but also complete enough to not overlook any element that might be important for users when appraising authenticity.
As such, concerning the resources, the indication of an identifier, title, dates, creator name and physical characteristics are crucial to manage resources as physical objects.The description level, subject and representation form contribute to describe the structure of the aggregate, the activities that are reflected in the resources, and the nature of the resources in that aggregate.Furthermore, the description of the resources' location and custodial history is essential to inform about the chain of custody over time.The declarations of integrity also assist in clarifying the history of transmission of records, highlighting the physical manipulations that they underwent.A description of the interventions of archivists in the management of resources is enabled by the elements of classification history, appraisal, and physical interventions.Regarding provenance, language can also be an important information.Metadata about the creator contribute to strengthen provenance by indicating an identifier, a name and a biographical history.Still referring to resources, the inclusion of information about related resources is important to establish bonds with resources from other aggregates, contextualizing the interpretation and meaning of resources.
As for the aggregation, a description of new additions may aid in clarifying interventions by the curators.Other information may be provided at the aggregationlevel, such as physical characteristics, description level, subject, location, amongst others, and automatically inherited by lower description levels.However, they feature in the model underneath the resources as a reminder that this information should be available at an item-level in order to facilitate retrieval, enhancing discovery, interpretation and contextualization.Lastly, it is believed that all metadata elements suggested for the control of descriptions -or meta-metadata -can be important to attest the authenticity of representations by elucidating about their provenance, promoting transparency of description procedures and holding the creators of the descriptions accountable.A more detailed list of every element that includes their definition is included in Appendix 5.
Reference must be also made to the elements that were chosen not to be included in the model.Regarding the requirements identified in the scientific literature, the suggestions about the inclusion of use rights for records and the Curriculum Vitae (CV) of curators were excluded. 1As for the technical literature, it was decided not to include the elements referring to the target audience, the documentary form, the orientation/disposition of visualization and the origin of resources.Indicating a targetaudience seems like a subjective and unnecessary prediction that, although not forgetting that many organizations serve a primary community, it is not absolute in the digital realm where resources may be retrieved in unexpected ways.It is also admitted that documentary form could be useful for the description of analog records, but harder to justify in the digital.Information about the orientation of resources seem more appropriate for artworks, therefore questionable for archival description.Finally, the origin of the resource originates in a Spanish standard for recordkeeping in public administration, due to the need to distinguish between private and public records.Although sometimes useful, it was considered that this requirement was not essential for the demonstration of authenticity.

Conclusion
Metadata is understood as data about data that adds context and structure in human and machine readable form, so that it can be meaningfully interpreted and used by both humans and machines.It can elucidate about the identity and integrity of resources, thus strengthening the presumption of authenticity and consequent trustworthiness of archival descriptions.However, as a form of representation, archival description and, by extension, metadata, are not neutral.Every representation implies a simplification of the represented object, which means a choice of which elements to include or exclude from the representation.Thus, metadata is not objective, but rather a subjective human construct shaped by the context of its creator.Furthermore, authenticity is not an absolute concept, but equally subjective in the sense that it depends on the goals and the reference points of the user.What is authentic for one might not be enough for other, so authenticity should be seen as a nuanced concept that plays out across a grey scale.It is not the role of archivists to ensure authenticity, but merely to provide as much information as necessary so that users can form their judgement.
In a digital context in which information can be retrieved by multiple representation systems, it is key that no metadata is lost in translation, in order to preserve resources' context and meaning, as well as to maintain their trustworthiness.This requires a semantic interoperability of the metadata elements used to describe resources that is facilitated if information management communities articulate together to standardize description procedures.The sheer diversity of approaches, standards, and professional backgrounds in the information management field may create confusion, compromising the ability for researchers and practitioners to dialog, and for information systems to communicate.This article's proposal of a model for authenticity aimed to contribute to filling that gap, not only performing a thorough meta-analysis of the literature of the last decade, but also a mapping of how metadata standards have expressed authenticity, combining both inputs in a suggested a set of metadata elements firmly anchored in relevant literature.The model strived to capture the authenticity not only of resources, but also of the descriptions themselves as a postmodern acknowledgement of the archivist's subjectivity in the representation process.A compromise was made between simplicity and complexity, so that the elements included in the model followed a rationale of being as complete as necessary while as simple as possible.It can be used by data practitioners as a benchmark for application, as well as by archival scholars or standards researchers as a starting point for the discussion of authenticity and trustworthiness in digital descriptions.

Archival description
According to the More Product Less Process methodology, it is suggested that description processes should be simplified with essential metadata in order to lower the description costs of large document volumes (Anchor 2013) Description must be centered on user needs (Anchor 2013) Description should be more honest, acknowledging the transformative processes that influence archive formation over time.Notes and/or citations of the creator of the description should be included, indicating creatorship (Douglas 2016;Meehan 2009) and even their CV (MacNeil 2009) Descriptions must be more transparent, possibly adding footnotes that document archivists' decisions, in addition to explaining description (Douglas 2016;MacNeil 2009;Meehan 2009) A significant focus should be placed on the custodial history of records.Suggestion of an autonomous description field, subdivisible in three elements: 'name(s) of the custodian(s)', 'custodial history', and 'Arrangement history of records' and related finding aids' (Douglas 2016).These should include a detailed log of the records' history until acquisition by the archival institution, including 'conversation history' (MacNeil 2009) The possibility of creating an independent field for creator's history with 'creator's name' and 'administrative history', as well as a field for archival history with the fields 'custodian name', 'custodial history', 'history of the arrangement of records' and 'history of the arrangement of finding aids' (MacNeil 2009) Archival descriptions are subjective constructs created by archivists in a given context (Douglas 2016;MacNeil 2009;2012) The suggestion for users to contribute to archival description through comments, tags, annotations or reviews (Douglas 2016) Archival descriptions should include: File name, location in the storage system and technical hardware and software specifications that would assist in information retrieval (Hedstrom 1993).As for metadata, 'codification format' and 'physical material' should be included (Llanes-Padrón 2019).Physical changes to records should also be documented (MacNeil 2009) Information that documents the relationship between records' context and content (Hedstrom 1993;Meehan 2009) A log about records' destruction.This information may be included in the elements 'administrative history', 'appraisal' and 'planned destruction and eliminations' (Niu 2013) Information about records management throughout their life-cycle, including audit trails (Hedstrom 1993; Niu 2013) The inability of a single hierarchical structure to capture sufficiently the complexity of relationships that records establish during their use, as they may be interpretable across different contexts (Hedstrom 1993; Llanes-Padrón 2019; Meehan 2009; Zhang 2012c) Resources should be described independently, at item-level, in order to enable the elaboration of several aggregates by web users (Moyano Collado 2013; Zhang 2012c) Metadata should be added gradually and progressively throughout the records' life-cycle (Niu 2013) Category Requirements

Information management
Representation should occur at item-level in order to enable users to recontextualize resources during their individualized queries (Bailey 2013) Records can be interpreted in multiple ways since their meaning and context vary over time (Bailey 2013;Cumming 2010;MacNeil 2017) Archival description should be a permanent process throughout records' life-cycle (Cumming 2010) The physical conditions of materials should be documented as a critical part in the perception of their meaning, context and authenticity (Acker 2017) Recordkeeping challenges need to be tailored to organizational needs, and there is no single, universal and magical solution (McLeod 2014) Metadata should reflect, above all, the context of the activities of which records are a subproduct (McLeod 2014; Tough 2016; Wright 2014) The creator of the representation should be identified (Wright 2014) The object of representation in archives should be (Yeo 2011), to allow semantic relationships to be established with resources from other collections (Wright 2014) Archivists are active agents in the construction of the representation and their meanings (Ridener 2009) Metadata The identification of who owns the metadata, as well as a log of changes over time, are part of a basis for trustworthiness (Apostolou 2009) User annotations, comments and evaluations may be included to boost content and context (Apostolou 2009) Acknowledging that each community has different requirements and needs, therefore there is no universal metadata structure, nor there is a common ground on which to define 'essential' metadata (Apostolou 2009;Gladney 2007;Haynes 2018) Metadata creation should be seen as an incremental process with shared responsibilities (Baca 2016) In order to simplify and accelerate metadata management, their creation must be as automatic as possible (Baca 2016) Metadata should be created by organisations so that they are easily sharable and reusable (Baca 2016; Gartner 2016) Faceted metadata can be used to improve access (Gartner 2016) and to fulfill the query goal by simply looking at the results retrieved (Beyene and Godwin 2018) The relationship between provenance and authority of the creator, according to the demonstration of its identity, is highlighted: For the explanation of provenance, the following metadata elements are considered important: 'Title', 'Identifier', 'Format', 'Subject', 'Rights', 'Editor' (Chen et al. 2011), as well as 'Creation date e 'Description' (Haynes 2018) For descriptions to be complete, as many elements as possible should be included, namely 'Creator', 'Contributor', 'Format', 'Editor', 'Source' and 'Coverage' (Chen et al. 2011), in addition to 'responsible entity' and 'update frequency (Zeng and Qin 2016) For the assessment of the resources' integrity, it is suggested to include metadata such as 'source', 'relationship', 'version/edition' and 'digital signature' (Zeng and Qin 2016) Metadata that can serve audit trails should be included to encourage the verification of resources' provenance and integrity, ultimately reinforcing trust in institutions (Gladney 2007; Li and Sugimoto 2017) Provenance may also be strengthened with metadata that answer the following questions: Who created the metadata?When were they created or modified?What are the circumstances that led to their creation?(Haynes 2018).Accruals, elimination and substitution of metadata elements and their definitions should be included in the answers.(Li and Sugimoto 2017) Metadata about retention/elimination policies should be included, such as 'Elimination act', 'Temporal interval of the elimination', 'Expected date for the elimination' and 'Elimination authorized by' (Haynes 2018) Each resource should have one and only description, according to the one-to-one principle (Zeng and Qin 2016) 1 3 Archival Science (2023) 23:629-673

Metadata standards
The mapping of metadata across standards is not linear.There are several inconsistencies and imperfect or non-existent matches (Andrade et al. 2014; Woodley 2016) There is a distinction between archival description standards and information exchange standards.The former attempt to ensure records' authenticity across time, contexts and domains, whereas the latter serve as safety vests that dictate how information is shared (Bunn 2013) The most successful models tends to be free and are written clearly, ensuring that they are easily read and understood (Dryden 2009) The creation context of records should be captured, and descriptions should be shareable and reusable between several communities in the web (Llanes-Padrón and Moro-Cabero 2017; Llanes-Padrón and Pastor-Sánchez 2017) Metadata creation is seen as n iterative process throughout the life-cycle of the records.
Metadata for appraisal, control, preservation, retrieval, access, use and an audit trail should be included as part of intellectual control (Mckemmish et al. 1999) If a standard includes too many metadata elements, it becomes harder to implement, generating costs in time and money (Youn 2015) Authenticity Mechanisms should be included to document the variability of digital objects through provenance metadata, fixating their physical integrity.Some possibilities are checksums or file size verification (Bountouri et al. 2017).An object's integrity may also be ascertained using cryptographic authentication (Gladney 2009; Rogers 2015) Provenance information, which are essential to document a record's life-cycle, should be collected since objects are created (Bountouri et al. 2017) The assertion of authenticity and trust are subjective acts shaped by various interpretations (Duncan 2009;Duranti et al. 2019;Engvall 2019;McLeod and Gormly 2017;Rogers 2015;Yeo 2013).A starting point may be to anchor the reference point of the "other" according to whom the authenticity degree is measured (Duncan 2009) The criteria used for trust in digital records are necessarily distinct from the principles that govern analog record creation.In the digital realm, enhanced transparency and accountability of records' creation and management processes, as well of their metadata, is required (Duranti et al. 2019;Engvall 2019;Rogers 2015; Yeo 2013) The explanation of the origin, provenance, chain of custody and a log of changes and audit are suggested to be indispensable aspects for transparency and to establish the grounds for trust (Duranti et al. 2019;Gladney 2009; Price and Smith 2011) Provenance information should be associated to the resources so that descriptions are selfexplanatory (Gladney 2009).Some of the key metadata elements of an abstract concept of record, according to a Diplomatic perspective, might be: 'action', 'bond', 'addressee', 'writer', 'author', 'adminContext', 'digitalObject', 'ingestDate' (Jansen 2015) Each record, or even each version of each record, should have a unique identifier (Gladney 2009) Collections from various institutions should be aggregated together in order to enrich context and better reflect society (Price and Smith 2011) Metadata creation should be made as automatic as possible in order to save costs and make description viable (Yeo 2013) As for authenticity indicators, description should include information about the software that generated the records, about actions carried out on the records, retention and elimination schedules, audit traits and cryptographic validation techniques (Rogers 2015) Blockchain technology is pertinent to recordkeeping as a safeguard for the integrity and identity of records, since it documents provenance permanently (Bhatia and Wright de Hernandez 2019;Hofman et al. 2019;Lemieux 2016) Category Requirements

Linked data
Interoperability is facilitated by a higher granularity in descriptions (Gartner 2015;Jones 2018;Niu 2016) It is important to use unique identifiers to define entities and their relationships (Gartner 2015) Some metadata elements may be useful for aggregators of multiple repositories, such as those that describe the collection (identifiers, titles), the components ('physical location', 'material description', 'content description'), holding institutions, dates, relationships, rights, use impediments, language and subject (Gartner 2015) Each representation model (tabular, relational, hierarchical and graph) has intrinsic advantages and disadvantages, so that each has its own value in being implemented, according to context (Hooland and Verborgh 2014): The use of RDF as a metadata sharing format has the disadvantage of requiring a high number of triples for description, as well the problem of intellectual control and curation rights (Gartner 2015) Nevertheless, it is still advantageous to make descriptions more flexible by expressing metadata in a machine processable manner, for example through linked data that facilitates the integration of collections (Gartner 2015;Hooland and Verborgh 2014;Jones 2018;Machado et al. 2019; Niu 2016) The focus of description should be in the relationships built within networks, not in rigid hierarchies built upon documents, collections or organizations (Jones 2018) Knowledge should be made explicit in metadata that facilitates users to apprehend the meaning of resources, who should not be required to ask the archival staff for further information (Jones 2018).Users should also be able to build their own narratives and interpretations (Rolan 2015

Fig. 1
Fig. 1 Metadata elements and orientations derived from the scientific literature.The number indicates the requirement fulfilled, according to Appendix 3

Fig. 2
Fig. 2 Metadata elements identified in the technical literature, with the indication of the requirement fulfilled, according to Appendix 4

Table 2
Summary of the metadata elements for authenticity identified in each standard