Provisional Semantics: Addressing the Challenges of Representing Multiple Perspectives Within Public Collections

Abstract

This paper presents a reflective account of the AHRC-funded Provisional Semantics project, which focused on how museums and heritage organizations might produce search terms, catalogue entries and interpretation using ethical and equitable practices, particularly in relation to the artworks and histories of racialized and minoritized people, and in the context of a digitized national collection. As the project developed, the emphasis shifted toward a more fundamental enquiry into whether these objectives were achievable, given the structural racism and colonial logics inherent in the museum project and its conventions, histories, and infrastructure. Through a critical discussion of the project’s three case studies, the paper considers the following questions: what happens when we change words and clean data as a putative solution to problematic cataloguing? Can museums genuinely, equitably, and ethically represent multivocality? Why is embedded change in attitudes and behaviors hard to achieve and slow to happen? And is sector guidance for research and documentation fit for purpose in terms of genuine knowledge co-production and engagement?

Introduction

This paper is a reflective account of Provisional Semantics: Addressing the challenges of representing multiple perspectives within an evolving digitised national collection – a two-year project, which ran between 2020 and 2022, and was led by Tate in partnership with Imperial War Museums (IWM), the National Trust and University of the Arts London. The project focused on how museums and heritage organizations might ethically co-produce catalogue entries and interpretation texts, particularly in relation to the histories and artworks of racialized and minoritized people. The project sat within a five-year, multi-project initiative devised by the UK Arts and Humanities Research Council (AHRC), titled Towards a National Collection—Opening UK Heritage to the World (TaNC), and still in process at the time of writing. ¹

With a budget of £18.9 million, TaNC represents a significant investment in UK GLAM (Galleries, Libraries, Archives, and Museums) collections research, although its focus is more on digital infrastructure than on collections content. ² The stated aspiration of the initiative is to explore the challenges of creating “a unified virtual national collection” and to support “research that breaks down the barriers that exist between the UK’s outstanding cultural heritage collections, with the aim of opening them up to new research.” ³ This ambition focuses on the technical barriers between collections databases rather than the ideological barriers that are inherent in them, and depends on an underlying data mass of labels, catalogues, and collections documentation that is at once highly varied, incomplete, frequently impenetrable, and often imperialist in structure, tone, and attitude. ⁴ However, it was precisely these overlooked aspects of the initiative that we focused on within the Provisional Semantics project.

Initially, we sought to explore how museums and heritage organizations might develop ethical, equitable, transparent, and multi-perspectival interpretations of collections objects in order to support a more diverse public to engage with a digitized national collection. But as the project progressed, ours became a more fundamental enquiry into whether these objectives (both those of the Provisional Semantics project and TaNC itself) were achievable at all, given the structural racism and colonial logics that are inherent in museum conventions, histories, and infrastructures. ⁵ This avenue of enquiry developed through three collections case studies, each based at one of our respective partner institutions. Although varied in content and context, all three collections were specifically pertinent to the histories and experiences of people from Asia, Africa, and the Caribbean and their respective global diasporas, and each uniquely exemplified the issues we sought to address.

The project was inevitably affected by its wider socio-political context, namely, the heightened public awareness, and media and institutional engagement with the idea of structural racism that occurred in 2020. Despite decades of anti-racist and anti-discriminatory practice in the heritage sector, it shamefully took the demonstrations and protests of the reignited Black Lives Matter (BLM) movement to bring the explicit connection between the bastions of cultural heritage and systemic, structural racism into a more public and arguably mainstream discourse. ⁶ Subsequent related critiques and challenges to institutional GLAM behaviors and practices amplified the subconscious but systemic racism that was subtly at work within our partner institutions, underscoring the importance and urgency of the kind of reparative work we were attempting to engender through our research. ⁷

We were also concerned and skeptical about using the terms “decolonization” and “decolonial” in relation to our research. The term decolonization was becoming increasingly popular within the UK GLAM sector at the time the project began, but its use diverged from the land-based challenges to settler narratives and behaviors of the decolonization that Tuck and Yang advocate for in their much-cited 2012 paper, and was closer to the metaphorical decolonization that they critique and reject. ⁸ Debate over the use and overuse of the words subsequently surfaced repeatedly in relation to the academy and the GLAM sector during the course of the project, particularly in terms of its conflation with anti-racism, restitution, and repatriation, and also with equality, diversity, and inclusion work. ⁹ In this context, as well as that of BLM described above, we attempted to foreground anti-racism and understood decolonization, within the scope of our project, as an attempt to undo “the legacies and ongoing relations and patterns of power established by external and internal colonialism.” ¹⁰

Below, we summarize and critically reflect on each of the three project case studies, and consider whether museums are able to equitably support and represent multiple histories, narratives, and perspectives; what happens in cataloguing practice when changing words and the cleaning up of records is employed as a solution to problematic interpretative texts and labels; and ask why embedded change in attitudes and behaviors is hard to achieve and slow to happen in how we research and write about art and artifacts.

Disrupting the Imperial Gaze in Cataloguing Practices: IWM

The first of the three case studies, undertaken in collaboration with Imperial War Museums, focused on a selection of fifty-three photographs taken in India on behalf of the War Office in 1942 to document and propagandize the recruitment of Indian servicemen during the Second World War. ¹¹ The imperial gaze is inherent in many, if not all the photographs in the series and is further compounded by the original captions, written on the reverse of each image at the time of their production. As Ashcroft et al. (2013) explain, the imperial gaze “corresponds to the ‘gaze of the grande-autre’ within which the identification, objectification and subjection of the subject are simultaneously enacted: the imperial gaze defines the identity of the subject, objectifies it within the identifying system of power relations and confirms its subalterneity and powerlessness.” ¹² We observed that these captions contained several problematic or offensive terms and narrative descriptions that were largely imperialist or racist in tone and attitude, including obviously pejorative terms such as “primitive,” as well as descriptions that seemed innocuous initially, but were revealed as problematic as the research progressed. ¹³ For example, the description “The immemorial ox cart still holds sway in India, just as in Old Testament times” perpetuates the highly questionable notion that India was undergoing a process of modernization at the hands of the British army in service of the civilizing force of Empire. ¹⁴ As such, the photographs were impossible to display or publish without interpretive intervention underpinned by in-depth subject-specialist research, which the IWM had not been able to facilitate between the time the series of images were issued in 1942 and the start of the project in 2020. Our objective, then, was to enable future engagement with these photographs by generating new descriptions and contextual information to inform the catalogue record. It was crucial that these texts critically and sensitively addressed the photographs’ production history and use, the impact and legacy of the war effort in India, as well as the underlying context, purpose, and effect of both individual words and longer descriptions in the original captions. ¹⁵

When the project started, IWM could not offer relevant, in-house, subject-specific expertise and experience to support us in achieving this objective. Further we were keen to move beyond the institution’s usual research and practice methods by collaborating with individuals not already working within the museum in order to foster the re-reading of the images and texts that was required, especially in terms of addressing the inherent racism and imperialism contained within them. We briefly considered community engagement activity, including with Indian veteran groups and their descendants, as a method through which to evaluate the original captions and co-produce new interpretations informed by lived experience, inherited knowledge, or personal response. However, given that we were unable to offer accurate information about the histories and contexts depicted in the photographs, we determined that such an approach would be inappropriate, and that we first needed to establish a critical and nuanced understanding of the photographs through collaboration with subject specialists with expertise in Indian experiences of the Second World War. We therefore invited the historians Aashique Ahmed Iqbal, Diya Gupta, and Ghee Bowman to join us on the project and to undertake this task through collective discussion and individual research. Their brief was to examine and research the photographs and captions, not only as historical documents but also as objects of visual culture, and to produce new texts to describe and contextualize them. These new texts were to be presented alongside a range of additional critical reflections and contextual articles by the artist Annu Palakunnathu Matthew and photojournalist Jess Crombie. The aim was to offer an informed, co-produced, multi-layered, and multi-vocal range of interpretations and resources which would embed criticality in the catalogue records and in the experience of IWM audiences engaging with the photographs.

One of the challenges we presented to the subject specialists was to identify the archaic and offensive terms found in the captions, to avoid their repetition when creating the new interpretation texts, and instead, to address the problematic, imperial or racist attitudes and framing from which such terms emerged. Although the specialists were largely successful in this endeavor, we observed that occasionally, certain phrases, such as “the bread basket of India” and “martial races,” were being uncritically incorporated in early versions of the new interpretation, especially in those authored by specialists without lived experience of the long-term impacts of colonial subjugation. ¹⁶ In our view, this was evidence of how imperial concepts and attitudes reside in seemingly innocuous phrasing and descriptions, often appearing as benign to those raised or educated in a traditionally white British setting. ¹⁷ That the veiled racism in object descriptions may sometimes only be obvious to the people being described or the nations being characterized underscored the argument that museums are more likely to avoid the unwitting reproduction of problematic terms and concepts in new interpretation if they collaborate with individuals who can combine rigorous research with their own appropriate lived cultural experience.

Despite the initial difficulty described above, the subject specialists expertly created a range of interpretive texts and explanations that not only addressed the offensive and racist terms found in the original captions, but also elucidated the imperial ideologies from which such terms emerged, the historical circumstances in which the photographs were taken and the ongoing legacies of both in contemporary India and the UK. This vital combination was presented on IWM webpages alongside the original captions, allowing researchers and the wider public to understand and engage with the new texts in the context of, and in contrast with past, problematic collections information. The careful, transparent and detailed specificity of this work seemed to us to operate in stark contrast with some of the more wholesale and hasty efforts we have observed taking place across the UK’s GLAM sector to rename, re-describe and reinterpret cultural objects and artworks. ¹⁸ As we have written elsewhere, we are concerned that problematic terms are being systematically erased from and replaced in some collections catalogues, with little to no record of the changes being made. ¹⁹ In our view, and in the context of the “museums are not neutral” campaign a wholesale and undocumented replacement of offensive, racist and imperial terms with those that are currently deemed decolonial or more politically correct, puts cultural institutions at risk of upholding the false sense of objectivity that they have been critiqued for promoting, with resulting object descriptions that are closer to the authoritative “neutrality” of traditional GLAM cataloguing than might be intended. ²⁰ Although problematic terms undoubtedly need to be identified and addressed, this effort alone does not equate with removing the problematic narratives, histories, and views from which such terms have arisen, and which remain inherent to and evident within the object being described. It is therefore essential that both language and context are addressed, as we did within the IWM case study discussed here. Racist pejoratives should not be the first information a viewer encounters, and preferred terms should be recognized and respected, but legacy titles and captions should be recorded in the interests of transparency and as testament to the attitudes of the time in which they were written.

Building Collections Knowledge to Address Colonial Legacies: The National Trust

The second of the three case studies in our project centered on the so called Clive Museum at Powis Castle, Wales, which is looked after by the National Trust. ²¹ The Trust claims that “The collection of South and East Asian artefacts displayed in the Clive Museum at Powis Castle is the largest private collection of this type in the UK,” with more than 1,000 artifacts including weaponry, textiles and other decorative objects. ²² The collection was accumulated by the Clive family between 1744 and 1839, and its current display has been in place at Powis Castle since 1987 with limited interpretation or information about the origins of the production or provenance of the objects on show. Unchanged since it was first installed in the 1980s, the display of the collection is not informed by current research and promotes a romanticized interpretation of the Clives as collectors and connoisseurs. This is achieved through visual devices including: a grand, heroic portrait of Clive; adoption of the orientalizing, nineteenth-century “indo-gothic” or “Hindoo” style in the dressing of the late-twentieth century display cases; and handwritten labels designed to suggest the curatorial involvement of the Clives. ²³ This is compounded in the accompanying catalogue, Treasures From India: The Clive Collection at Powis Castle. ²⁴ Published in the same year that the display was created, the catalogue presents a largely uncritical perspective on the British Empire and British conduct in India. ²⁵ As such, they do not reflect more recent GLAM and heritage sector attempts to place care and sensitivity at the center of the interpretation of cultural objects acquired through empire. ²⁶

Our objective, then, was to build knowledge around key artifacts in the collection with support from subject specialists, who could shed light on their cultural significance, whilst sensitively addressing the corrupt and violent colonial contexts of their acquisition. In doing this, our aim was to create a knowledge resource from which the Trust could begin to develop new interpretation for objects in the Clive Museum that would center the cultures and histories represented through the objects, and crucially, decenter the biographies and activities of Robert Clive and his descendants. After exploring a range of potential objects to focus on, we selected an eighteenth-century collection of poems, or divan, by the fourteenth-century Persian poet Hafez—a figure who is still of cultural, philosophical, and religious significance across the world, but whose book in the collection languished at the back of a small display case, was presented back to front, and was identified only as a “Persian Manuscript.” ²⁷ To develop knowledge around the book, and to understand its display and interpretation to date, we devised a program of three workshops, involving different stakeholder groups who each held distinct types of knowledge and experience and produced markedly different conversations.

The first workshop involved specialists in Persian and South Asian literature and art history, who shared factual and historical information about the object, as well as how to understand its history and value to multiple cultures over different time periods. Their insights exposed both our own and the Trust’s ignorance of the object and its origins, as well as highlighting the Trust’s problematic interpretations of the book, to date. The second workshop involved National Trust curators who critically reflected on the display and interpretation of the book in relation to practices and policies both within the Trust and across the wider sector. The curatorial staff provided close analyses of how the language in documentation and interpretation had affected public engagement and understanding of the book and the collection and were able to reflect on practical changes for updating curatorial and cataloguing approaches within the Trust. And finally, the third workshop involved Powis Castle staff and volunteers who helped us to look beyond the interpretive texts and revealed how knowledge and oft repeated family anecdotes were uncritically presented to visitors.

Although some amongst the third group were concerned about the veracity of stories being told to visitors and expressed a desire to see new, unexpurgated interpretations of the collection within the museum, others were highly averse to decolonial approaches to history and were defensive of long-standing narratives that are increasingly being challenged and problematized within the GLAM sector. This third workshop also revealed to us how some front-line visitor service staff and volunteers at the castle invoked a narrow definition of looting when communicating with visitors, which served to obscure the contexts and histories of colonial power and violence through which objects in the Clive Museum collection were acquired. ²⁸ Standard definitions of plundering or looting emphasize theft or the use of physical force in settings of war or civil disorder. This does not adequately account for colonial transactions that did not directly involve theft or force, but nonetheless arose in the shadow of colonial violence, mired in the sometimes subtle, but deeply problematic colonial power dynamics of the British Empire. ²⁹ However, that this was denied by site staff and volunteers at Powis Castle, allowed us a clear understanding of how language can be employed by museums and heritage organizations, consciously or otherwise, to deny the violence and traumas of empire, the insidious relationships of power and influence between the colonizer and colonized, and the ongoing legacies of empire for contemporary audiences. We concluded that addressing gaps in knowledge about object provenance and significance through close collaboration with subject specialists with appropriate lived experience could help to reduce the perpetuation of problematic, offensive, and colonial perspectives and interpretations.

Co-Production and Institutional Project Working: The Panchayat Collection at Tate

The third and final case study within the project focused on the Panchayat Collection held by the Tate Library. The Panchayat Collection was founded in London in 1988 by Shaheen Merali, Al’An de Souza, Bhajan Hunjan, Symrath Patti, and Shanti Thomas to respond to the underrepresentation of the art and exhibition histories of artists with African, Caribbean, and South Asian heritage within the British art sector and in wider histories of art in the twentieth century. ³⁰ Its initial impetus expanded to accommodate a widened transnational perspective, but the collection remains characteristic of its time, representing contemporary artists who produced issue-based work, with a particular focus on cultural identity. Although its structure is broadly in concert with traditional archives, some of its most important content includes flyers, pamphlets, and slides, which due to their ephemeral nature are not easily accounted for within the established cataloguing practices of a library and would perhaps be more appropriately preserved within more inclusive and flexible models of community archives. There had been some engagement with the material on its arrival at Tate in 2015, but cataloguing was incomplete at the start of the project. ³¹ Because of this the collection was inadequately served by the library systems, which lacked the information and search terms that would enhance retrievability, and therefore rendered the items within it inaccessible and invisible.

Our objective within this case study was to develop knowledge of the Panchayat Collection, its history, contexts, and contents in close collaboration with its living founders and custodians, and with a selection of the artists whose work and practice are represented in the collection. Our aim in doing this was to make the collection visible, and thereby, to increase awareness and use of it by researchers and the wider public. Within this, we also wanted to surface the ethical challenges of developing knowledge about the collection with its stakeholders. This was particularly critical in the context of the history of the collection and its relationship to Tate. After the acquisition of the Panchayat Collection by Tate in 2015, efforts to catalogue it had been constrained by resources, funding and institutional priorities. Dissemination and access were therefore piecemeal and knowledge of the content of the collection within Tate was also partial and inadequate. Further, Tate had not successfully maintained the involvement of the collection’s keepers, Shaheen Merali and Dr Janice Cheddie, since its acquisition. Working relationships were consequently strained and in need of repair and reparation. Thus, when we asked Merali and Cheddie if they would be willing to produce a variety of materials that would create knowledge about, and insights into the collection, the ethics of co-production within the specific context of museums and galleries as institutions came into sharp focus.

In particular, we observed that the project format did not serve equitable collaboration between the stakeholders and Tate. Attempting to build trust within a time and scope restricted research project was difficult, and the process of co-production of collections information often felt extractive when carried out in an institutional context. For example, recording personal accounts, documenting artistic practices, determining contractual obligations, negotiating intellectual property rights, and agreeing on use, were typically imbalanced in favor of the institution, and not the individuals we were seeking to include through collaboration. These aspects of co-production, as well as the balance between cost and benefit to those positioned outside or adjacent to the institution, as both subject and object in the research process, required significant consideration and navigation, often exacerbating existing tensions rather than ameliorating them.

We found that limited project budgets and timescales, as well as our use of institutional employment contracts, impacted negatively on our attempts to form ethical and equitable partnerships and collaborations within the project. Time and resources to develop networks and relationships of trust with the appropriate prospective researchers and participants from the point of project conception is crucial. Maintenance of such relationships after the project’s end is just as important. That is, for ethics and equitability to be embedded in project design, it is essential to put in place measures to ensure that individuals and relationships continue to be supported by the institution and that research material will continue to be accessible to those involved and implicated, long after projects conclude and after researchers move on to other roles. Our contention is that one of the key ways to do this is to set down co-produced knowledge in the catalogue record. This fundamental work is typically side-lined in favor of digitization that privileges images and limited metadata, but in our view, cataloguing should be regarded as a core output for funded collections-based projects. Without it, the actual and long-term impact of further collections research and the legacy of projects like Provisional Semantics is significantly curtailed. ³²

Discussion and Conclusion

Provisional Semantics set out to explore some of the ethical challenges that may emerge for museums and heritage organizations when attempting to co-produce collections information and interpretation that relates to the cultural and material histories of racialized and minoritized people. This objective was premised on a general understanding that museum histories, conventions, and infrastructures are inflected by imperial logics and structural racism. The project was therefore angled toward a consideration of the current ideological barriers to equitable public access to collections, rather than on the technical barriers that might preclude the development of a future, unified national collection. Our three case studies revealed how language and semantics can serve to deny colonial power dynamics, the violence of the British Empire, and its long-term impact and trauma for contemporary museum audiences. The case studies also demonstrated how problematic terms and concepts require a critical combination of thorough research, expertise and lived experience in order to be identified. Once identified, such terms require documenting and contextualizing to adequately address the colonial and racist circumstances in which many UK collection objects were produced or acquired. And finally, the project case studies highlighted how collaborative co-production of collections knowledge, informed by subject specialism, and lived experience, require time, resources, and alternative commissioning and contractual practices if they are to be carried out ethically and equitably within the space of the museum. The speed and scale of technologically driven approaches are presented as solutions to increasing and improving access to public collections (not least by the research focus and funding priorities of TaNC). However, our research within the Provisional Semantics project has led us to advocate here for a far more people-centered method to disrupt long-standing but problematic historical narratives and racial hierarchies within museums. Such a method requires a significant investment of time to listen, build relationships, and collaborate to produce rich, multi-perspectival object descriptions, and, in our view, a more careful, or slower approach may aid in this effort.

The increasing scrutiny placed on how and what we catalogue within the context of the “decolonial turn” has coincided with the persistent drive within the GLAM sector to digitize collections images and information at scale, ostensibly to improve and democratize online access through new technologies, on the presumption that more is better. ³³ The TaNC initiative is part of this endeavor that valorizes the use of collections information and the power of digital technologies to solve what are essentially complex socio-political problems. ³⁴ However, the pace, scope, and resourcing of digital dissemination that has proliferated in documentation and collections management work does not, in our view, foster the in-depth, ethical and equitable knowledge production that we argue for in this paper, and may even be fundamentally incompatible with it.

When museums record information about art and artifacts in their collections management databases and catalogues, they organize that information within a standardized format, but often without reference to source, context, history, or authorship. When that information is subsequently published online for widespread dissemination, dislocation from source and context of that information increases. While object records with structured data potentially allow for faster search, more predictable retrieval, replication across multiple platforms, and use in data visualization or machine learning, there remains the now well-established fact that the information and recorded data are not neutral or even accurate, and were created by human beings who are subject to bias, working within systems and institutions that uphold dominant and long-standing social and cultural hierarchies. ³⁵ Yet the promise of the digital for cultural heritage is predicated on the assumption that catalogue records are an accurate, rich and firm foundation. Without significant investment in cataloguing as core museum work that crucially involves a concerted effort to include nuance and debate within the catalogue record, the new technologies for engagement, dissemination, and inclusion that the TaNC initiative seeks to produce will continue to be based on a limited, partial and partisan version of collections information. As the Provisional Semantics research has shown, public collections require cataloguing practices that value depth, focus, context, and ethical engagement over expediency, speed, and scale, and further, that such practices should be embedded as central tasks within organizations as opposed to separate and temporary project-based interventions. This approach is particularly critical in museum collections that contain work by minoritized and racialized makers, where a lack of institutional knowledge, interest and care is often evident within the record, and cements the longstanding neglect and marginalization of both makers and objects by public cultural organizations.

In the comparable field of archival work, but seeking to remedy issues similar to those we identify above, Kimberly Christen and Jane Anderson asked in their 2019 article “Toward a Slow Archive,” “How do we recognize and rebuild archival practices, structures, procedures, and workflows that allow for relational, reciprocal, respectful, and restorative connections to knowledge, kin, and community within their [own] frame?” ³⁶ Their answer is to employ a slow approach, as advocated for more broadly within the “Slow movement.” ³⁷ Slow is perhaps best known in relation to food, and specifically a rejection of fast food and industrial scale food production, as well as taking time to appreciate and understand sources, ethics, and impact. ³⁸ The relation of the slow movement to a people-centered approach to museum cataloguing, and particularly in the context of the Provisional Semantics case studies, is apparent. Ethical and equitable co-production of collections information needs to be conducted at a speed that is appropriate to the specifics of the people, histories, objects, and collections in focus and as Jeremy Hugget suggests “it entails resistance across the range of dominant neoliberal rhythms and values and critiques our practices and ethics.” ³⁹

Ascertaining where collections information and data comes from, who created it and why, and understanding its lifecycle and impact, is just as important in cataloguing as it is in food production. As Christen and Anderson assert, by slowing down, and by placing value on slowing down, a space opens out in which we can consider how knowledge production operates in public collections, and how it is presented, contextualized, and used. Further, they argue that an approach in which listening, decentering colonial attitudes, and acting ethically is emphasized, disrupts the embedded structures of ownership and attribution, and instead allows complex networks of relationships between people and objects to surface. These principles can be applied within a rethought cataloguing practice that is considered, reflective and careful, and it is encouraging that this has begun to be recognized in recent cataloguing guidance produced for libraries, archives, and museums. ⁴⁰

Digital dissemination and technological developments based on GLAM collections information and catalogue records undoubtedly have benefits in terms of widening access to cultural heritage (though perhaps only to those who have access to and an inclination to use digital technologies). However, if we pause and redirect our focus from mass aggregation, artificial intelligence, and short-term efforts to remove problematic terms from collections databases, to carefully examine and slowly catalogue our collections with and in support of the people represented in them, we can begin to achieve the widened engagement that initiatives such as TaNC seek to engender. ⁴¹ The benefit of the Provisional Semantics project was that it simultaneously allowed us the time and space to test out slow approaches that are not always employed in core work, and the importance of embracing the wealth of messy, human-centered knowledge as part of a radical rethinking of collections cataloguing.