Who holds the key to Holocaust-related sources? Authorship as subjectivity in finding aids

The ﬁ eld of Holocaust studies relies on a wide variety of archives, dispersed all over the world. Identifying the right sources for a speci ﬁ c research question within this ﬁ eld is not easy or straightforward. Yet Holocaust scholars predominately focus on methodologies for source analysis rather than discovery. Archival ﬁ nding aids are among the most important tools to aid primary source discovery, but have hitherto not been considered in methodological discussions on Holocaust research. In this article we will re ﬂ ect on the composition of ﬁ nding aids based on our work for the European Holocaust Research Infrastructure (EHRI). Our premise is that the content of ﬁ nding aids is determined by their authors and the context in which they are creating them. The strongest argument for this subjectivity is that our work – outlined in this article – not only indicates that descriptions of one and the same source differ, but that they can do so quite considerably, and hence can in ﬂ uence research. Our stance is that historians optimize their pro ﬁ t from ﬁ nding aids by becoming more sensitive to the subjectivity and authorship of descriptions. We conclude by showing how an online environment such as the one developed by EHRI can sensitize historians and archivists to the situated and subjective nature of ﬁ nding aids by accommodating a plurality of descriptive voices, and encourage them to share their knowledge and become co-authors of ﬁ nding aids.


Introduction
I think you have to gain the trust of the archivist in some ways andin a lot of archives the cataloguing is very poor so there's not any really decent catalogue so you just have to really trust the archivist to give you the things that you need. You have to go in every day and build a relationship so that they know that you are someone who's serious and then to try and work with what's there. I think I'm using a very difficult place to work. (European Holocaust Research Infrastructure [EHRI], interview with Holocaust researcher) 1 The sentiment expressed abovethe need to establish a good working relationship with archivists in order to get access to relevant primary sourcesis one that will be familiar to most researchers. Indeed, because Holocaust sources are highly dispersed, fragmented, and often insufficiently described, talking to and learning from archivists is often an essential component in a researcher's strategy to find and access relevant archival material. 2 And yet our interviewee implies that the need to rely on the archivist could be lessened if the cataloguing of archival material was of higher quality. It is seemingly merely in the absence of detailed finding aids that a researcher is compelled to talk to an archivist, thereby letting research partially be determined by what is bound to be a subjective reply. If, however, the relevant archival knowledge was neatly and comprehensively encoded in a finding aid, the implication is that the interpersonal and subjective encounter between archivist and researcher could be avoided, and the need to build a relationship of trust lessened.
In this article we will argue that it would be wrong to draw such a conclusion. To create more or, in one way or another, "better" finding aids will not eradicate the problem of archivists exerting a considerable and, notably, subjective influence on Holocaust research. We will demonstrate that finding aids are not neutral guides, but complex documents that cannot but be influenced and shaped by their authors and the institutional, cultural, and political contexts in which they are composed. Or, to put it differently, Holocaust researchers have to deal with the subjective biases of archivists, no matter whether they encounter archivists as interlocutors during reference interviews or as authors of finding aids. The problem to be tackled, therefore, is not how to eradicate subjectivity from finding aids, but how inevitable subjectivity can be conceptualized, acknowledged, and accommodated in the encounter between archivists and researchers.
We will develop our argument by reflecting on our experience of working on the realization of the European Holocaust Research Infrastructure (EHRI). 3 The EHRI project in its first phase (2010-2015) funded by the European Union's Seventh Framework Programme (FP7) and currently under the Horizon 2020 Programme (2015-2019) offers a complex of services for researchers. 4 EHRI's main aims are to integrate information on Holocaustrelated archives and their collections in an online portal, and to investigate how approaches and methods originating in the digital humanities and archival science can be harnessed to innovate Holocaust research. The EHRI portal contains information on more than 1850 archival institutions and on thousands of Holocaust-related collections. As Holocaust-related sources are dispersed across Europe and beyond and held by a wide variety of institutions, the portal also hosts country reports that provide concise information on the Holocaust history and archival situation in 47 countries where Holocaust-relevant materials can be found. Identifying widely dispersed Holocaust-related sources is a major challenge: indeed, large institutions such as the United States Holocaust Memorial Museum (USHMM) and Yad Vashem (YV) have spent several decades surveying archives to discover sources on the Holocaust and are far from concluding their work.
By integrating information about Holocaust-related archives from many different sources, EHRI brings together a large variety of heterogeneous finding aids. 5 Though similarities exist, authors of finding aids employ diverse paradigms and methods in their descriptive work, thus producing finding aids that differ markedly from each other both in terms of structure and content. To overcome these heterogeneities, EHRI defines "archives" and "archival collections" very broadly. Indeed, EHRI has surveyed primary sources from a wide variety of institutions including state and regional archives, memorial sites, tracing services, museums, and private collections. In this article we employ the terms "archival institution," "archival collection," "archivist," etc. in a similarly broad fashion. However, some of the institutions we designate as "archives" and some of the people we label as "archivists" may well not describe themselves in these terms. By the same token, archival institutions use a wide variety of terms such as "archives," "fonds," "record groups," or "collections" to refer to the primary material they hold.
Before reporting on our experience with different finding aids in the context of the EHRI project, a short excursion is needed to concisely frame and define the problem we are seeking to address. Prima facie, we might be tempted to regard the proposition that finding aids are not neutral but reflect the preoccupations of their authors and the wider context in which these operate as simply obvious. After all, finding aids are texts, and the insight that any text can be analyzed as a subjective narrative rather than a simple account of objective fact is hardly a novel one. However, we will show in section two that neither the archival nor the historical communities have traditionally acknowledged this ostensibly simple insight in regard to archival finding aids. The section will locate the reason for this state of affairs in the historical development of the relationship between archivists and historians that has favored an unduly restrictive view of archivists as passive custodians of sources rather than as active participants in the enterprise of interpreting such sources. This section will conclude with a short overview of recent revisionist archival thought that highlights how the descriptive practices of archivists influence historical discourse, and that can help us to contextualize our experiences with finding aids in the EHRI project.
It should be noted that we develop our general framework of the relationship between historians and archivists and the impact of this relationship on the composition and understanding of finding aids with reference to historical and archival thought in general. It would of course be interesting to study the changing dynamics between these two actors in the narrower context of the emergence of Holocaust and/or Jewish history and archiving. However, such a detailed investigation would extend beyond the scope of this article. While a detailed reading of the emergence of particular Jewish traditions of historiography and archiving would undoubtedly further nuance our story, we hope that our general account provides a sufficiently adequate framework enabling us to conceptualize and analyze our experience of dealing with (subjective) finding aids in the context of the EHRI project. Sections three and four then provide two detailed case studies that demonstrate the extent to which Holocaust-related finding aids are subjective, and locate the roots for such subjectivity in the wider institutional, political, and cultural contexts in which describing archivists operate. Section three provides an indepth analysis and comparison of seven finding aids of strongly related Holocaust archives that have been composed by four EHRI partner institutions, while section four outlines our efforts to identify and investigate Holocaust-related archives in Ukraine. This section especially focuses on the many different initiatives that have authored and collected information on Ukrainian Holocaust collections prior to, and alongside, EHRI. It analyzes the diverging approaches to archival description such initiatives have employed, and the heterogeneous finding aids they have thereby produced. We will conclude by outlining how we intend to accommodate diverse and subjective descriptions in the context of the EHRI portal. Here we will highlight how an online environment such as the one EHRI endeavors to create can render the subjective character of finding aids explicit by presenting multiple descriptions authored from different perspectives alongside each other, and by enhancing them with in-depth contextual information. We will further present our plans to get Holocaust researchers involved in the process of producing archival descriptions, thereby strengthening collaboration between archival and research communities, and facilitating a multiplication of perspectives from which archival sources on the Holocaust can be approached.
The subject matter of this article is interdisciplinary, bringing together perspectives stemming from Holocaust studies, archival theory, and the digital humanities. It has been jointly authored by an intellectual historian (Speck), a Holocaust historian (Vanden Daelen), and an archivist (Links) who share an interest in the digital transformation of historical and archival practices and theory. While the problem of subjective finding aids will no doubt be of interest to archivists and historians in general, we deem it especially important to raise Holocaust researchers' awareness about its existence and implications. The problem of subjective archival descriptions is particularly pronounced in regard to Holocaust-related sources because of their complexity. In the aftermath of the Second World War many archival sources documenting the Holocaust have undergone complex processes of dispersion and fragmentation, and have ended up in a wide variety of archives that operate in different institutional, cultural, and political contexts. As we outline in detail in section three, such archives consequently approach and describe similar sources from diverse perspectives. Moreover, even though some archival theorists have recently begun to address the complex question of how archival practices may shape historical research, their insights have so far not reached the Holocaust research communities. The online portal EHRI seeks to build, finally, provides us with an opportunity to practically address some of the problems raised by subjective finding aids, thereby allowing Holocaust researchers to approach, and contribute to, finding aids in new ways.

2.
Archivists and historical research: from the neutral, objective, and impartial guardians of sources to active co-creators?
Historians typically do not know much about the work archivists are doing. To be sure, there is much overlap between the activities of the two, and historians often consult the expertise of archivists in order to learn about, locate, and access primary sources. Historians are also avid readers of all kinds of archival finding aids, which are indispensable tools in their quest to unearth new evidence about the past. And yet we can safely assume that only a minority of historians frequently stop in their tracks when perusing a finding aid to ask themselves questions such as: "who has written this text?", "when was it written and what audience had the author in mind?", "what were the archivist's intentions and agendas when writing the description?", "did the archivist employ a thesaurus when describing the sources, and did this thesaurus contain the keywords I am using for my search?", and so on. More often than not, historians treat finding aids as neutral, unproblematic access tools to archival holdings. Of course a given finding aid may serve this function more or less successfully, but it is very rare indeed for a historian to regard a finding aid as an interpretative problem in itself. 6 Such a lack of curiosity and questioning is surprising, not least because the ability to critically analyze texts to uncover their meanings, intentions, biases, contexts of composition, and relationships to wider socio-economic, political, and cultural currents is of course a fundamental aspect of the historian's craft. There is something deeply puzzling about the fact that historians do not seem to apply their usual analytical techniques when confronted with finding aidstexts that play an important role in shaping and furthering their research interests. Indeed, finding aids could be regarded as black boxes that have a determining yet mostly hidden influence on the selection of sources that historians consider in the course of their research.
An uncritical approach to archival finding aids on the part of historians is but one manifestation of a wider trend: a gradually widening gulf between the activities and underlying assumptions of archivists and historians. Over the last two decades a series of studies have appeared that analyze why and how such a distancing between the two professions emerged, and how it impacts upon both archival and historiographical practices. 7 A short outline of some of the key insights emerging from such studies will provide us with a more precise idea of why authorship and subjectivity in regard to finding aids should be regarded as a problem, as well as with initial clues on what can be done to address it.
Recent investigations into the changing relationship between archivists and historians typically proceed from a common point of (modern) origin for both professions. Francis X. Blouin and William G. Rosenberg as well as Terry Cook have traced the roots of both modern historiography and modern archiving back to the nineteenth century. 8 Crucially, at this point of modern formation and professionalization, the two disciplines shared assumptions, goals, and principles. While methodological innovation in areas such as philology contributed to render historical research more "scientific" with the aim to create objective accounts of the past based on original primary sources, innovators in the world of archives sought to assemble authoritative sets of documents that could serve as evidence on historians' mills. Implementation of Leopold von Ranke's famous dictum that historians must strive to recover the past "wie es eigentlich gewesen" [as it actually happened] by examining exhaustively all available primary sources was critically dependent on archives providing access to such sources. Given the positivist assumption, common among nineteenth-century "scientific" historians, that historical documents whose authenticity and authoritative status had been duly asserted could speak for themselves, that is to say provide immediate access to the past by revealing objective truth, it was of course of utmost importance that archivists were not seen to meddle in any way with the pristine evidence under their guardianship. 9 In this positivist environment arose the idea that archivists must renounce all active agency in their role as custodians of the evidence of the past; or as Cook puts it, the ideal archivist was seen as a "neutral, objective, impartial, an honest broker between creator [of documents] and researcher." 10 Archivists became invisible handmaidens of historians; the role they inevitably played in appraising, processing, describing, and classifying documents remained hidden behind a veil of apparent neutrality and impartiality.
If the archival and historiographical mainstream thus shared a common mission in the nineteenth century, the close alignment between the two was torn asunder in the course of the twentieth century. The project of scientific history ran into serious trouble in the early twentieth century, and its underlying assumptions such as authoritative history and objective truth became untenable with the epistemological challenges brought about by poststructuralist and post-modernist philosophies. With historians increasingly turning their attention to questions about the intentionality of authors and texts, the situatedness and multiplicity of all historical experience and knowledge, the power of (meta)-narratives, the textual construction of reality, and approaching new fields of enquiry (gender histories, subaltern histories, micro histories, etc.), their concerns and assumptions progressively drifted away from the ones that structured traditional archiving. 11 Archives remained for a long time relatively insulated from these historiographical developments. This is of course not to say that archival thought remained stuck in the nineteenth century, but that archivists have been exposed to different headwinds than the ones experienced by most historians. Most importantly, a combination of rapid expansion of government activities and bureaucracies gathering pace since the aftermath of the First World War and new, especially digital, technologies have necessitated archivists to learn how to operate in a context defined by overabundance rather than scarcity of information. 12 The response to this challenge has, until very recently, been largely a technical and managerial one: while historians have turned their attention to fundamental epistemological issues that have redefined both the ends and means of historical research, archivists have become increasingly preoccupied with questions of standards, system design, efficient archive management, and so on. 13 One consequence of archivists devoting their energies to redefining the means rather than ends of their profession has been that the traditional view of the archivist as the impartial, neutral, and invisible custodian of documents has been able to survive relatively unscathed well into the twentieth century and beyond. 14 According to Cook, for instance, many archivists are still influenced by "a kind of unquestioned professional ethos" with distinctly nineteenth-century roots that posits that "archivists do not interpret, or mediate, or construct social memory." 15 The old veil hiding archivists and their role in shaping the historical record, in other words, has largely remained in place.
Surprisingly, historians have been slow to lift this veil. Whereas a plethora of studies have examined how different institutions and actors influence historical knowledge and social memories, the considerable control archivists exert over the remnants of the past has hitherto been under-explored in historical scholarship. 16 As a consequence, the positivist "unquestioned professional ethos" of archivists has only been questioned belatedly and from the inside by a small but growing number of archival theorists who have endeavored to revise traditional archival thought by employing insights stemming from postmodernist philosophies broadly conceived. 17 Initially, the focus of archival revisionism was chiefly on appraisal and selection. By reexamining these processes, theorists have firmly asserted the powerful control archivists exert over what is knowable about the past, and have made a convincing case that archivists must be understood and acknowledged as important agents within the historical enterprise. Nesmith, for instance, proposes that archivists should be regarded as active "co-creators" and "co-authors" of the sources under their custodianship rather than just their passive guardians. 18 More recently, this active "co-creating" role that archivists perform in shaping the residues of the past has been explored across all archival activities, including arrangement and description. 19 Traditionally, arrangement and description have occupied a central role in archives' quest to preserve authentic sets of documents. Description provides the repository with physical and intellectual control over their holdings, and assists users to gain access to material that may be of interest to them. Guided by the core archival principle of respect des fonds which demands that "the records of a person, family, or corporate body must be kept together in their original order … and not be mixed or combined with the records of another individual or corporate body," 20 description has typically been seen as a complex but ultimately neutral and undisputed process. 21 Indeed, one of the key functions of the principle of respect des fonds is, of course, to diminish archivists' potential to meddle with the materials that are to be arranged and described by attempting to preserve their physical and intellectual integrity, and consequently their authenticity. By assuming that there exists an original order established by the record creators themselves, the archivist's task is reduced to replicate this order through description and arrangement, and relate it to the functions and activities of the creating agency. Description, according to traditional archival theory, is a mechanical act that seeks to make transparent the meaning and significance of archival holdings by revealing characteristics that can be objectively derived from the material itself. 22 This traditional view of arrangement and description has recently become subject to sustained criticism. Most importantly, several studies have shown that regarding archival records through the prism of the principle of respect des fonds is only one possible way of how such records may be arranged, described, and understood. Far from being an objective or natural principle, it is one that privileges a certain view of records that can mask as much as it can reveal about them. For instance, the idea that there exists a neat one-to-one relationship between record creator and records fails to acknowledge the complex provenances and custodial histories most records actually possess. In the same vein, the notion of an original order that can be readily recreated by the archivist has been criticized for masking the habitual disorderliness of records in reality. 23 According to Cook, traditional archival description presents the researcher with a "well-organized, rationalized, monolithic view of a record collection that may never have existed that way in operational reality." 24 Unfortunately, researchers have all too often regarded such monolithic archival views as natural and absolute rather than what they are: authored representations of materials that are subjectively constructed by their author, the describing archivist.
We are now in a position to see why subjectivity in finding aids is not a problem as such, as this is unavoidable. Rather, the fact that inevitable subjectivity has so far largely been disguised for most historians by a nineteenth-century veil of assumed objectivity is problematic. As seen, (parts of) the archival community has recently challenged the presumption of archival objectivity, but their insights have so far not reached the majority of historians. And yet to air the views of revisionist archival theorists as widely as possible is important, because once the veil of positivist archival objectivity is removed, possible theoretical and practical solutions to the problem of subjective finding aids can also start to emerge.
Most importantly, if we accept that describing archivists present us with one possible representation of an archive rather than objectively reporting an unproblematic objective reality, it follows that archival descriptions should be regarded and analyzed as narrativesstories about recordsthat are authored from a particular perspective. Indeed, an archive can theoretically give rise to any number of stories. It is the describing archivists that shape one particular story, by choosing words and tropes, by furnishing the story with a certain tone, by highlighting certain parts of an archival collection while ignoring others, and so on. Or as Wendy M. Duff and Verne Harris have put it, "Descriptions inevitably privilege some views and diminish others. When archivists describe records, they can only represent a slice, or a slice of a slice, or a slice of a slice of a slice, of a record's reality." 25 Apart from the describing archivist's individual perspective, a finding aid is naturally also shaped by the wider context in which its author operates. Professional guidelines and methods, the institutional requirements of archives, and wider political, cultural, and intellectual currents and pressures provide settings in which finding aids are composed, and must be duly considered and analyzed if we are to understand finding aids as situated narratives rather than natural or absolute descriptions of reality. 26 Crucially, the postulate that finding aids are situated narratives authored from particular perspectives also entails that archivists should effectively be regarded as co-creators of archives. Indeed, finding aids provide essential contextual information about archival holdings that makes it possible for a researcher to locate them, and ascertain their relevance for a particular research question. By significantly shaping this context, archivists have a determining influence on the interpretive possibilities that the materials under their custodianship possess. By highlighting certain aspects in a finding aid while ignoring others, or by assigning one keyword rather than another, archivists gently guide historical interpretations into certain directions. The subsequent stories historians can tell about archival materials are, in other words, not only influenced by the creator's initial inscription, but also by the archivist's interpretation of this inscription that is rendered explicit in finding aids. 27 Recent archival thought provides a theoretical framework that allows us to analyze finding aids as subjectively authored narratives, and furnishes us with clues on how the subjectivity problem can be mitigated in practice. Before demonstrating how we implement such solutions in the context of the EHRI portal, however, we need to measure the extent to which Holocaust-related archival descriptions are subjective in practice. We will do this through two case studies that are related to EHRI's efforts to identify and investigate Holocaust-related archival collections and their descriptions across Europe and beyond. These case studies reveal a wide variety of heterogeneous finding aids, and show that such heterogeneity can to a significant degree be accounted for by subjective choices made by the describing institutions and, indeed, individuals. They thereby demonstrate that the subjective influences describing archivists exert on the traceability of archives are particularly pronounced in the field of Holocaust documentation and thus require our attention.

Finding aids in comparative perspective
In the course of our identification work of Holocaust-related archives for the EHRI portal, we became increasingly aware that different institutions employ a wide variety of paradigms and methods when describing their holdings. To arrive at a better understanding of such differences, we undertook a detailed comparative analysis of selected finding aids produced by three EHRI consortium institutions: the International Tracing Service (ITS, Bad Arolsen, Germany), the NIOD Institute for War, Holocaust and Genocide Studies (NIOD, Amsterdam, the Netherlands), and Yad Vashem (YV, Jerusalem, Israel). 28 For our analysis, we selected seven finding aids authored by these three institutions that describe the same or strongly related archival material. 29 For instance, the archive of the Stichting Sieraden Comité (a Dutch post-war foundation dealing with restitution claims of Jews concerning jewelry) is kept and described at NIOD while YV holds and describes a copy of this collection. 30 By the same token, the YV finding aid of the archive Joodsche Raad Friesland (Jewish Council in Friesland) is very closely related to the NIOD finding aid of the archive of the Joodsche Raad Amsterdam (Jewish Council in Amsterdam). 31 Finally, originals and copies of the archival documents on the transit camp Westerbork 32 are kept by ITS, NIOD, and YV alike, and all three institutions have individually described their Westerbork holdings.
By comparing these seven finding aids, we wanted to determine the extent to which they differ despite the fact that they describe the same, or at least very similar, archival material. Furthermore, we sought to establish whether the respective institutions' missions, cultures, and histories could help us to explain such differences. Overall, we found that the three institutions produce very distinct finding aids whose content and structure are heavily influenced by the institutional context in which they are produced, and, at times, by the preoccupations of individual authors. We thereby found strong evidence about the subjective nature of Holocaust-relevant finding aids, and gained important insights into the factors that determine the structure and content of such finding aids especially in relation to the missions and cultural background of these institutions.
Interestingly, finding aids produced by the same institution already reveal remarkable differences. The three NIOD finding aids we studied provide an excellent example of this, as their internal structures differ from each other. In two of the three casesthe Joodsche Raad voor Amsterdam and the Stichting Sieraden Comitéthe documents are structured according to their provenance and follow the rules of respect des fonds, that is to say they are kept in the order in which they were originally created. By contrast, the Judendurchgangslager aggregates documents from various creators whose content all relate to the Judendurchganglager Westerbork. This aggregation was undertaken by a special department at NIOD in the early years of its existence to support research on prisons and camps.
Further indications that NIOD employs heterogeneous rules are also evident when we consider the levels of detail that are provided in the three finding aids. For instance, the finding aid Judendurchgangslager Westerbork provides significantly more structure and detail in the information provided on the collection level than the finding aid Joodsche Raad voor Amsterdam which compresses all context information into one paragraph. But at the same time there are also commonalities. For instance, in all three descriptions the authors elaborate on the context of the creation of the materials in the introduction to the finding aid and particularly focus on the administrative and archival histories of the collections.
Clearly, individual NIOD describers have considerable freedom when composing finding aids. Such descriptive freedom is apparent in the finding aid Judendurchgangslager Westerbork, where the author frequently highlights parts of the scope and content information that he finds particularly interesting. Likewise, the introduction of the Joodsche Raad voor Amsterdam finding aid focuses on the establishment of the organization and in particular on the activities of one central person, Prof. Dr. David Cohen, the president of the Joodsche Raad voor Amsterdam.
Another notable feature of the NIOD finding aids is their visual appearance. Both the Judendurchgangslager Westerbork and the Joodsche Raad voor Amsterdam finding aids are richly illustrated with photographs that provide visual context to the archival materials that are being described. Photos embedded in the Judendurchgangslager Westerbork finding aid, for instance, depict important personalities and provide visual clues about the transportations from the camp and the liberation in 1945, thereby focusing readers' attention into certain directions. Moreover, the finding aids also contain various footnotes pointing readers to related archival collections and to further historical literature. Such presentational features serve to give the descriptions something of the appearance of a historical monograph, and readers are clearly invited to read the finding aid cover to cover, like a book.
These findings echo to a considerable extent the history of NIOD and its institutional and cultural self-understanding. NIOD was founded directly after the liberation of the Netherlands on 8 May 1945. As a state institution it had the task to map the history of the Second World War through independent research. Initially, the institution's main task was to collect material about the Netherlands and the Dutch East Indies during the Second World War. 33 In the first years of NIOD's existence these materials were processed in such a way that they could fluently feed into this research and be presented in publications such as Het Koninkrijk der Nederlanden in de Tweede Wereldoorlog (a reference work on the history of the Netherlands in the Second World War) by Loe de Jong. 34 Since the 1990s, NIOD has gradually widened its mission to serve internal and external researchers alike. To this end the existing finding aids were adjusted with increased attention to archival principles. NIOD's double mission to act as a research institute and an archive is reflected in their finding aids where a historical perspective is complemented with an archival one. As a consequence, NIOD finding aids furnish readers with a story that is both about and developed from the described documents.
By contrast, the two ITS finding aids under consideration generally adopt a more unambiguous archival perspective, while still reflecting ITS's historical roots as a tracing service rather than an archive. Like NIOD, the ITS was established in the 1940s. The ITS, called from 1944 until 1948 the "Central Tracing Bureau" (CTB), was formed as an initial effort to trace and register persons who went missing during the war. In the first years of its existence the institution operated respectively under the Supreme Headquarters Allied Expeditionary Forces (SHAEF), the United Nations Relief and Rehabilitation Administration (UNRRA), the International Refugee Organization (IRO), and the High Commission for Occupied Germany (HICOG) until the International Committee of the Red Cross (ICRC) took over the management in 1955. In the same year the International Commission for the International Tracing Service was established as the institution's complementary governing body. 35 To facilitate their tracing work, the ITS collected all sorts of documents that could shed light on the fate of missing persons. Its collecting activities encompassed both original documents and copies of documents from various other organizations. The ITS thereby assembled an immense archival collection consisting of over 30 million individual documents. This collection was, however, closed to external users until 2007. Since then the ITS has put considerable effort into opening up the archive, and has sought to transform itself into an archival institution open for historical research rather than a tracing organization. This transformation is reflected in the partnership of the ITS and the German Federal Archives, and the withdrawal of the ICRC from its management. 36 The finding aid Reception Camp Westerbork is in itself an outcome of the ITS's recent transformation. Authored in 2011-2012, it was created to promote usage of the Westerbork collection by researchers. A striking feature of the finding aid is a detailed archivist's note that gives a thorough account of the methodologies that were followed in its composition. From this note readers learn, for instance, that the finding aid is one out of a series resulting from a joint project with YV and the USHMM that aimed to create basic finding aids for the ITS holdings related to concentration camps. These new finding aids are based in large part on the existing inventories and other tools dating from the 1950s. Consequently, the adopted rules of description are a compromise between applicable usages at the three institutions. It is noteworthy that the extensive archivist's note is displayed very prominently, clearly indicating that it is deemed to be of utmost importance to an understanding of the finding aid as a whole. As such, the note implicitly acknowledges one of our central theses, namely that finding aids are not neutral guides, but complex texts whose contents are strongly determined by their various contexts of composition.
Such a focus on archival processes is also expressed in a long section on the archival history of the collection that provides much detail of what happened to the material since 1945. And yet, in regard to the Reception Camp Westerbork finding aid, an archival perspective on the material co-exists with the historical needs of a tracing service. Indeed, whereas archival concerns predominate in the information provided at collection level, the file-level descriptions hone in on information that is relevant for tracing purposes, in particular lists of names and other personal information. This juxtaposition of perspectives directly reflects ITS's own history: the file-level descriptions were created in the 1950s by people focusing on tracing, whereas the collection level, created in 2011-2012, mirror ITS's new mandate to support research purposes other than tracing by following standard archival practices.
An archival perspective is also reflected in the finding aid Verfolgungsmaßnahmen in den Benelux-Staaten, 1936-1985. It offers extensive information on the custodial history as well as on the acquisition process of the collection. Additionally, much effort has been put into tracking down the origins of the files. The finding aid therefore contains references to archives and institutions that hold the original documents. This provenance-related information complements the way in which the documents were processed in the earlier years of the ITS, when the content rather than the origin of the documents guided arrangement and description practices. Once again, we can clearly detect how institutional settings shape finding aids and pre-determine the ways by which users can locate material within a repository.
The two YV finding aids we studied -The Friesland District Joodse Raad (Jewish Council) Collection, the Netherlands, 1940-1943 and the Stichting Sieraden Comité Collectionshare commonalities with the file-level descriptions produced by the ITS in that they focus on the contextualization of personal information. Both finding aids are meticulously indexed with subject terms and the names of organizations and individuals. The prominence of names of individuals in the finding aids is unsurprising given YV's core mandate "to preserve the memory of Holocaust victims by collecting their names and biographical details, and preserving them in an eternal memorial in the Hall of Names." 37 YV has a long tradition of collecting Holocaust-related archival materials scattered all over the world in order to extract relevant information for its Central Database of Shoah Victims' Names. 38 In this context we can safely assume that one of the key aims of YV's finding aids is to relate information contained in its archives to the database.
We can find further illustration of such a focus on the names of individual victims through an analysis of the composition of the finding aids studied. Both are multiauthored and contain legacy informationfor the Stichting Sieraden Comité, YV took over the original collection-level description from the NIOD; for The Friesland District Joodse Raad a previous collection-level description from the acquisition periodas well as content produced by YV. Crucially, the content YV has produced is strongly geared toward retrieval of information by personal names. In the case of The Friesland District Joodse Raad, very detailed descriptions on file and, at times, document level have been produced that enable precise searching of the finding aid. It is telling that document-level descriptions have only been created in cases where the underlying documents contain lists of personal names. 39 Similarly, in the case of the Stichting Sieraden Comité finding aid, YV did not follow NIOD in describing all intermediate levels between collection and file level, but focused exclusively on the file level where ample information about individuals can be found.
We may therefore conclude that, unlike the ITS and NIOD that both explicitly describe the archival and, intermittently, historical context of collections, YV contextualizes as far as possible at file level, and focuses on the collections' social context. In cases where information about the archival context was available from third parties, for instance from the institutions that provided YV with copies, this information has been taken over.
The comparative analysis of finding aids composed by ITS, NIOD, and YV demonstrates that commonalities as well as differences are to a large extent explicable by the respective institutions' missions, histories, and self-understandings. Indeed, such an analysis brings to light a variety of perspectives, not only between institutions but also within the same institution. Given that the considered finding aids describe the same or strongly related material, the very existence of such variety clearly indicates the considerable extent to which finding aids are never created within a vacuum; on the contrary, they are authored by individuals operating in complex institutional, cultural, and political contexts that cannot but influence the descriptive act. Moreover, such contextually situated descriptive practices have a potentially significant impact on historical research. Historians employ finding aids as gateways to archival sources, but these gateways are not neutral or unproblematic. On the contrary, reflecting the preoccupations of their authors, finding aids inevitably direct the historian's gaze into certain directions. A given finding aid may highlight one part of a given collection while obscuring another, or it may firmly situate a collection in one particular historical context while keeping silent about alternative settings. Using archival finding aids, in other words, inevitably entangles a historian's research in the complex cultural, political, and institutional milieus in which the gateways were produced.
While variety and heterogeneity are thus characteristic of the finding aids we studied, this does of course not necessarily entail that there exist outright contradictions between the descriptions produced by the three institutions. On the contrary, from a researcher's point of view, the availability of several different finding aids for the same material may well be a distinct advantage. As each finding aid approaches the material under consideration from a different vantage point, each enables different methods of retrieval and preliminary analysis. The next section of this article will further elaborate on the complementary character of parallel finding aids, thereby further strengthening our argument that descriptive diversity and subjectivity should be embraced rather than treated as a problem.

Identifying Holocaust sources in Ukraine
The previous section demonstrated the influence of institutional settings on the structure and content of finding aids through a detailed analysis of seven carefully chosen examples. This section will show that the descriptive diversity caused by such influences is not exceptional but characteristic in regard to Holocaust-related finding aids. We will develop this argument by reporting on EHRI's work to comprehensively identify and investigate Holocaust-related sources in Ukraine.
With approximately 1.6 million Jewish victims (1.5 million of the local Jewish population, plus about 100,000 Jews from other areas murdered on Ukrainian territory), assembling descriptions of archival sources that document the murder of Jews in the Ukraine is of key importance to the EHRI project. 40 An important part of the archival sources that relate to Ukrainian victims are stored in Ukrainian archives. However, the identification of Holocaust-related sources in these archives is not a straightforward undertaking, and multiple sources of information were employed to arrive at a complete overview. The sources used by EHRI encompass information produced by the Ukrainian state archives themselves, outcomes of surveying and copy activities by USHMM and YV, published finding aids composed by Project Judaica, outputs from projects such as the Records of Einsatzstab Reichsleiter Rosenberg (ERR) digitization project in Ukraine, and EHRI's own surveys undertaken by a researcher of the Ukrainian Center for Holocaust Studies (UCHS).
The starting point for our investigation was the Ukrainian archives themselves. The archival system in Ukraine is centralized, headed by the State Archival Service of Ukraine. Two categories of archivescentral and regional (there are no municipal archives)are administered in this structure. In addition, several archives, libraries, and museums operate outside this centralized system. 41 In 2005, the Ukrainian state archives published a guide on its sources on the Second World War in Ukrainian. 42 This guide is supplemented by other state archives finding aids written in Ukrainian or Russian. 43 However, the centralized archival system works with very large collections (or fonds), often comprising tens or hundreds of files, some even exceeding 50,000 files within one collection, which makes the identification of the right materials challenging. Moreover, these guides rarely explicitly mention the Jewish population, the persecution of the Jews, or the Holocaust, which makes the identification of Holocaust-related collections via keyword searches difficult.
The absence of specific identification of Holocaust-relevant material in archival finding aids may at least in part be explicable with reference to Ukraine's post-war history. Indeed, Jews became an "invisible group" in many former Soviet Union countries, especially in the post-Stalin era. 44 All victims of Nazi repression were treated under one denomination, as citizens of the Soviet Union, without making a distinction between nationalities. Under Soviet rule all were Soviet citizens, but with different nationalities, such as Ukrainian, Russian, or Jewish (which was considered a nationality in the Soviet Union), and nationality was not a distinction made in the description of victims of Nazi oppression. Jews were only identifiable in the documents as Jews via their names or in descriptions of special Jewish fonds, such as from the interwar period. This invisibility did not fully disappear with the fall of the Soviet Union. Also, the events of the Second World War are referred to in regionally specific ways. For instance, in the former Soviet Union, the Eastern Front of the Second World War is generally known as the Great Patriotic War. However, this is only a partial explanation as the non-identification of Holocaust-relevant material is a general problem across archives that hold potentially relevant sources but do not possess a specific interest in the event. Indeed, Holocaust-related sources are frequently held by national archives and other repositories that do not employ "the Holocaust" or the "persecution of Jews" as the prism by which they assess and describe their materials.
YV archive surveyors provided EHRI with a basis for a first selection as they had indicated all the fonds in the Ukrainian guide that potentially contain Holocaust-related materials. The selected collections were added to the EHRI portal with their basic Ukrainian-language descriptions as provided by the state archives' online finding aids. However, this often resulted in short descriptions with no indication of why the selected materials were Holocaust-relevant. We therefore enriched the basic descriptions with ancillary information from Project Judaica. 45 This survey project aims at providing detailed overviews and descriptions of Jewish historical sources in the former Soviet Union. For Ukraine, multiple guides have been published or are currently under publication, most of them in Russian. EHRI was allowed to add these descriptions to the ones gathered from the Ukrainian state archives, and in certain cases Project Judaica pointed EHRI to so far unselected or unknown collections, especially in Southern Ukraine. 46 The Project Judaica descriptions are typically one book page in length for each relevant fonds. Unsurprisingly, given the project's focus, these descriptions explicitly address the topics Holocaust researchers are looking for.
In addition to descriptions of the original documents in Ukraine, EHRI also collects information via descriptions made by archives that hold copies from Ukrainian archives. Both YV and USHMM have a long tradition of surveying archives in order to copy the Holocaust-relevant materials and make these accessible in their respective reading rooms in Jerusalem and Washington DC. Copied materials from Ukraine are described in the English and/or Hebrew finding aids. Whereas the aim of these finding aids is to make the copied material available for researchers in the two institutions' reading rooms, they often include references to the original documents. Therefore, the descriptions of the copy collections can also be helpful for Holocaust researchers who wish to consult the originals. As we will outline in more detail below, the EHRI portal establishes links between the original Ukrainian collection descriptions and the collection descriptions of the copied materials, thereby supplementing local and general descriptions with English and Hebrew descriptions composed by Holocaust experts. 47 Hence, the EHRI portal hosts descriptions of Ukrainian sources authored by many different institutions and projects. A few concrete examples demonstrate the breadth of the information that will thus become available for researchers. Collection "P-2390 Претура Піщанського району, с. Піщани," for instance, is held by the State Archive of the Odessa Oblast. 48 The Ukrainian description from the state archives describes it as "Рахунки та фактури на отримання товарів та сільськогосподарського знаряддя" [Billing and invoice on receipt of the goods and agricultural implements] and indicates that the language of the material is Romanian and Russian. The English-language Project Judaica description provides an English translation of the title of the collection, namely "District Pretura, Pishchany District, Village of Pishchany," and states: These were district organs of Romanian executive power that functioned from 1941-44 in the occupied territory included in the newly-created Governorate of Transnistria. They were headed by praetors and subordinate to their respective county prefectures; and in turn held jurisdiction over village and city primarias. The preturas' functions included, among other things, the organizing of ghettos. 49 The explicit mentioning of "ghettos" makes a crucial difference to catch the attention of a Holocaust researcher, and clearly marks this description as relevant to the main audience of the EHRI portal.
Important to note, though, is that EHRI does not merge the various descriptions gathered from different institutions and projects into one "master description". Rather, all individual descriptions that relate to the same collection are inter-linked and treated as parallel descriptions. As each description is authored from within a specific context and thus possesses unique characteristics, we consider it important to maintain the integrity of each individual description. By bringing together and inter-linking all available descriptions pertaining to one collection, we offer researchers a variety of perspectives. Thereby we hope to enable access to sources via different pathways and, at the same time, raise researchers' awareness about the subjectivity of the finding aids they use.
To enhance researcher access to Ukrainian sources, having parallel descriptions is important since they reveal diverse information in various languages. For example, for the collection "P-23 Костопільський гебітскомісаріат, м. Костопіль Рівненської області" held at the State Archives of Rivne Oblast, EHRI provides a Ukrainian description from the archives itself, a Russian-language description from Project Judaica, and links to an English-language description on the copies made from this collection, stored and available at USHMM. 50 All descriptions mention the census of the Jewish population in the Bereznovsky district and the selling of Jewish property, albeit in different wording. The Ukrainian state archives' guide on its sources on the Second World War also mentions that "permission [was granted] to sell the former Jewish homes." USHMM gives the same information, but further gives the dates 1941-1943 for "information on confiscated property" and provides the number of pages on each subject (31 and 48, respectively). Project Judaica is the only one that added the date of 1943 to the census information, and includes much more qualitative information on the topic useful for Holocaust researchers. A translation from its Russian description reads: Provided information about the number of Jewish population in districts [rayons] and the materials of the census of the Jewish population in Berezno district [rayon] (1943); orders on resettlement of Rivne Jews to the ghetto and correspondence on this subject between the Rivne mayor and Gebietskommissar (1942); disposition of the Rivne Gebietskommissariat concerning organizing the service by Jews to establish order in Jewish homes, lists of houses where the Jews lived, permits to sell the former Jewish homes; correspondence between the Rivne Judenrat and mayor on the distribution of "Jewish labour force"; order by Kostopil town administration (10 October 1941) on the use of clothes and shoes left by Jews; ordinance by the commandant of the Ukrainian militia in Klevan (17 September 1941) on the use of the Jews at the construction and general works for the improvement of the city and others. 51 For a collection with German, Russian, and Ukrainian as the language of the archival documents, providing this multi-lingual and multi-faceted approach can be very helpful. Moreover, it can even enable Holocaust researchers to learn about freely available digital copies. This is, for instance, exemplified by collection "3206: Рейхскомиссариат Украины, Ровно" (Reichskommissariat Ukraine, Rovno). This collection is part of the ERR project, a digitization initiative funded by the Conference of Jewish Material Claims against Germany (Claims Conference). 52 Via this project, ERR sources stored in Ukrainian archives are digitized, described in Russian and Ukrainian, and freely available online. For our example, fonds 3206, the EHRI portal contains a link to the digital ERR surrogate, as well as the Russian collection description from Project Judaica and an English copy collection description from USHMM. 53 All these pieces of information play their role. As the collection contains German-language documents, the link to the English-language description of the copy material may form the key for a German reader to identify this material (otherwise, without mastering Russian and/or Ukrainian, the ERR scans remain hidden). 54 In a last step of our identification work, Mikhail Tyaglyy, a researcher at the Ukrainian Centre for Holocaust Studies (UCHS), verifies and further completes information integrated into the EHRI portal. He functions as EHRI's general feedback person as his institution is involved in, or is aware of, the above named projects. 55 Thus far our work has resulted in the identification of 47 repositories and about 850 collections across Ukraine. As a minimum these collections have a basic (Ukrainian) description, but often this basic information is supplemented by a Russian or English Project Judaica description and linked to English or Hebrew descriptions of copy collections held at USHMM and YV. On its portal, EHRI shares all this information. In this way, these multi-lingual sources not only become more visible to scholars with different language skills, but are also described from a variety of perspectives and according to diverse descriptive methodologies. It is our premise that each of these descriptions contributes in its own way to enhance our knowledge about the content of the archives, and, therefore, that having access to multiple and contrasting finding aids for the same source is beneficial to archivists and researchers alike.

Conclusion: finding aids in the EHRI portal
Our case studies above indicate some of the complexities involved in identifying Holocaust-related sources and highlight how diverse institutional perspectives have resulted in highly heterogeneous finding aids. Such complexity and heterogeneity pose significant challenges in regard to the integration of identified finding aids, and their presentation in a unified interface to researchers. By way of conclusion we will outline how we tackle some of these challenges in the EHRI portal 56the main gateway by which Holocaust researchers can explore and access all the information assembled in the EHRI project.
When designing the portal, we endeavored to leave institutionally diverse descriptions intact. Rather than attempting to homogenize finding aids by means of standardization or bringing all available descriptions together into one "master description," we want to preserve the various institutional perspectives on Holocaust archives. 57 As we have seen above, institutionally idiosyncratic descriptions can, at times, reveal much about the describing institutions and their collections, and as such constitute in themselves a valuable information resource for researchers.
Indeed, far from attempting to hide descriptive differences, the portal is designed to highlight them by rendering explicit connections between related descriptions of archival material held at diverse repositories. The availability of parallel descriptions and links to related items for a given collection is prominently displayed in the portal, enabling users to easily switch between them. Thus, a researcher can conveniently access, assess, and compare all the ancillary information that, for instance, Project Judaica and the Ukrainian state archives have produced about a given Ukrainian collection without losing sight of the fact that different pieces of information were written by different people operating in diverse institutional settings.
Given that we decided to preserve institutionally diverse descriptions in the portal, it is crucial to provide contextual information about the describing institutions. If we accept that finding aids are subjective, and that their character is shaped by the institutional, cultural, and political contexts in which they are composed, access to reliable information about this context is crucial for a successful interpretation of finding aids. 58 In the EHRI portal, descriptions of archival materials are linked to detailed descriptions of the holding institutions that encompass narrative descriptions of their missions, policies, and histories. 59 Moreover, if a description is not authored by the holding institution, information that ascertains its authorship is included in the description itself.
By preserving institutionally diverse finding aids and by providing contextual information about their authors, the EHRI portal alerts users to the fact that any given archival description only offers one particular perspective on the described material. In addition to raising awareness of this fact, we also want to give users the opportunity to participate directly in the authoring of archival descriptions. The EHRI portal facilitates this by allowing registered users to annotate archival finding aids. By authoring annotations, users can add their own knowledge and perspectives on archival materials. In the EHRI portal annotations are by default private to the author, but they can also be made visible to all portal users. Such public annotations hold the promise to enhance existing descriptions in a number of ways: they could, for instance, point to particular items in a collection that are pertinent for a given research theme, correct factual mistakes in the original description, or provide an alternative reading of the significance of a given collection for Holocaust research.
We envisage that Holocaust researchers will be the primary authors of annotations, thereby generating research-driven perspectives on Holocaust-related collections in addition to the ones prepared by holding institutions and archive survey projects. However, archivists are of course also invited to participate in the composition of annotations. Reference archivists continuously make new discoveries about their own collections when answering researchers' queries, and annotations provide them with a convenient means to encode newly accumulated knowledge and make it available to everyone. This way, annotations provide a flexible framework that can accommodate multiple voices and perspectives and progressively express new readings and revisions as our understanding of, and methods to interpret archival materials develop over time. 60 The keys to Holocaust-related sources are clearly held by many and have different shapes and forms according to the specific contexts in which they were forged. The EHRI portal recognizes this fact by presenting finding aids not as static, monolithic, and ostensibly neutral descriptions of archival materials, but as fluid, plural, and situated representations thereof. By integrating the descriptive voices of archivists, researchers, and other relevant projects and initiatives, it also holds the promise of fostering new ways of collaboration between all the stakeholders with an interest in Holocaust documentation. As seen, the concerns of archivists and historians have gradually drifted apart in the course of the last century, and this widening gulf has left both with an impoverished and problematic understanding of finding aids. By enabling archivists and historians to co-create finding aids in a collaborative environment, EHRI can also contribute to a much needed reinvigoration of the ties between the two professions for the benefit of both.

Notes on contributors
Petra Links has a master's degree in Modern History from Utrecht University and a master's degree in Archival Science from the University of Amsterdam. She works as an Archivist at NIOD Institute for War, Holocaust and Genocide studies. Petra has five years of experience working on international, interdisciplinary archival projects, including the European Holocaust Research Infrastructure (EHRI). She also provides consultancy to the Kigali Genocide Memorial on the expansion of the Genocide Archive Rwanda.
Reto Speck is a Research Fellow at King's College London and a Researcher at NIOD Institute for War, Holocaust and Genocide Studies, Amsterdam. He holds a PhD from Queen Mary, University of London, and his research interests encompass the intellectual history of the European Enlightenment, the history of historiography, and the digital transformation of archival research. He currently co-ordinates EHRI's dissemination work and leads the conceptualization and development of the EHRI portal.