 Yn ystod, yma, ymwneud am y panel, ymwneud Elinor Cepaldi ar y byd ymwneud ym Edinbro. Elinor yw'r drwsgwrs hon yn y ddechrau pholedigiau ar y byd ymwneud ymwneud ymwneud ymwneud ymwneud, yn cyrraed o gael o'r Gaelol Llywodraeth. Ymwneud yn cydweithio'r cydweithio ar gyfer yma, oes gyda'r cyflwyllus o'r gynhymhau ac eisiau cynghŷrwch ac yn ystyried i gŵr o'r ffyethau gyda'r ffotogrofi, i gyrwm Llywodraeth i'r cymrydau swyddfa. Felly, o'r gweithio, rydyn ni'n gwybod i Elinor i'r gweithio'r gwaith ar gael cyfryd digital yn 60 sgwm, yn ei gweithio'r narratives ar TikTok. Rwy'n ddweud. Rwy'n ddweud. Rwy'n ddweud, mae'r gweithio'r gweithio. Rwy'n ddweud, Elinor, ac rwy'n ddweud i'r gweithio'r gweithio'r gweithio'r gweithio'r Edinburgh on an HRC-funded collaborative doctoral partnership with National Galleries of Scotland. So some of my areas of interest do include visitor-generated photography and remakes culture. But today I will be talking to you about the concept of audience created digital heritage on the social media platform TikTok by and for LGBTQ plus audiences. So some of the key points that my talk will be revolving around today include the emergence of what I term the creator curator, which is a type of online contributor who not only creates their own content, but curates it in a manner analogous to the traditional act of curating in the museum or gallery. And this is marked the reinterpretation of digitised images of art from a plurality of perspectives, which are then shared with a mixture of audiences, including their peers and more. This sits alongside the repositioning of peripheral voices around art and objects and how they can challenge and even highlight inequities and heritage representation. So what gap is being filled here and why is there a niche dedicated to LGBTQ plus heritage in such a committed way. So I'm trying to understand what difference it makes when LGBTQ plus related art and objects are reinterpreted voluntarily by queer communities online and how this might affect the dynamic between audience and institution if it does at all. So while on the surface it might seem that this is very empowering and freeing for audiences and represents an inroad to rebalancing or even reshaping power structures. I'm thinking that while this could alter current narratives around LGBTQ plus objects, when it comes to the relationships between institution and audience, it could also reinforce them. So I would like to understand what such alterations or divergences can mean when LGBTQ plus communities take the reins of their stories in the spheres of powerful social media platforms where they may have influence, but ultimately don't have control. So in the reuse of digitized images of heritage, the scope for how those images are used and contextualized is extended. There is an inherent openness as these images and the stories they represent are freely circulating online. And as Navar suggests, the value in recycling of art offers a space in which people can reevaluate their histories and contributions to the world. So in this case, audiences may be elaborating upon or from their perspective, reintroducing or elevating histories that they feel have been suppressed or sidelined. In offering their contributions and reevaluating their histories, communities theoretically have the opportunity to democratize heritage and even rebalance the power between institution and audience. However, while these opportunities may be present and even embraced, there is not just uncertainty introduced in terms of how these artworks and objects may be framed, but in whether this wealth of video content has any power beyond the contributor having a platform to curate in and of itself. So the value implications for the contributors and having a voice may present another avenue within the power struggle. Is it enough to speak if you are not heard and heard by whom? So as all these voices online multiply, the questions inevitably multiply around them too. So in March 2020, the month that the UK went into its first lockdown, the platform TikTok, which is a video-based social media app, received the highest number of downloads it received in the UK up until that point. It embodies internet culture, I think, in terms of presenting multiple niche fragments in communities which then comprise a larger entity. And one of these is the home it has provided for independent heritage fans, curators, educators who are directly addressing LGBTQ plus histories. So this, in combination with the rise of the queer TikTok communities in general, led me to investigate a scoping study examining public content uploaded on TikTok to explore the relationship between online audiences and LGBTQ plus art and culture. So I examined videos that used the hashtags LGBT paintings history and queer art history. They were gathered based on the order that the TikTok search results provided them to me. And at the minute it was only 77. And the criteria was basically that they were publicly available and that they included LGBTQ plus art and cultural imagery in some way. So in terms of how the digital objects were used by audiences on the platform and how they contribute to the narratives around those objects, the data site does provide some insight. So of the types of content uploaded audiences use digitized imagery of heritage in two main ways. The first was as supporting imagery. So that would include slideshows or background images which were employed to supplement the story that the presenter is telling a bit like props in their function. So for example, when somebody was describing the events of the Stonewall in riots, there would be archive footage or images behind them. The more dominant represented type was what I've termed subject imagery whereby the images would represent the focus of what the video was centered on and about. So for example, somebody created a queer artist series and the images would be of this kind. So a particular artist, a figure from history, a curious object, particular artwork, etc. So all of the unique accounts in this study, apart from three, were generated by individuals acting in a personal capacity. The contributors are curating and presenting. They're speaking with authority. Previous video introductions on their feed may tell us they are currently students, researchers are working full time in the field of art or archives. But on first glance that you might not know that and it also doesn't apply to everyone. It's not also entirely possible to ascertain people's identities and you definitely can't assume a common agreed signifier or coded language within the community. However, is the inclusion of the rainbow flag in the bio section, which does appear throughout the sample. And if you add that to the drive and the passion with which there is to share these stories, there is a suggestion of authenticity to the content created, which comes from really living in these identities that they are speaking about, although I can't exclusively confirm it throughout the sample. The presenters will follow two patterns of presentation. The first is facing the camera, much like I'm doing just now, like a news reporter or a TV presenter speaking directly into the lens to you, the viewer, with the imagery placed behind them using the green screen feature. And alternatively, the other most common style is for presenters to be heard by voiceover only with the images becoming the sole and only visual focus. So as you'll see from these captions, just a small selection from the sample, there's very much a trend towards revealing that which has not been revealed before. So upon viewing the videos together, the same artists and stories begin to crop up again and overlap. And depending on the knowledge accrued per individual viewer already, this will have varying degrees of surprise or revelation. It appears that the contributors are aiming their videos at queer audiences who are not necessarily in the know already having been deprived by the real world museums and galleries in their view. What the contributors can do here that the majority of institutions unless they specialize in LGBTQ plus history are less likely to do is to offer very frequent regular dedicated content addressing LGBTQ plus histories. So collaboratively the creators when appearing together in the results page of the search function create a channel of sorts of different programming for you to choose from within a specific genre. Alternatively videos of this type may appear in a persons for you page, which is based on previously liked or viewed videos and is mostly comprised of accounts you haven't followed before. The content is generated as far as can be found by members of the communities in their own voices. So they have selected the artist objects and histories that appeal to them and that they think deserve to be heard. For the majority of these videos the core biographical information pertaining to artists like Frida Kahlo or figures such as Sappho is in the public domain you might only need to look up Wikipedia to find it rather than a specialist source. But taking ownership of this information and the act of saying these words for themselves does have value, even if the institutions are maybe not hearing it. The contributors here might be engaging in an even more radical act of circumventing the institutions altogether and just speaking directly to their peers and communities instead. So while institutions have many forms of undeniable capital and power, if they're not reflecting these audiences as well or as much as they could these audiences can instead turn to their peers, speaking with knowledge about images and objects that these organizations care for but to their own community. So there are two ideas that emerge from the insights of the scoping study which invite more exploration. First, the question of how much audiences reinterpreting heritage challenges the established relationship between heritage organizations and audiences. So, while kid discusses how an organization's digitization and social media efforts intersect with audiences and questions the true capacity to enable more and or better democratic exchanges. As well as alongside the limitations of more people equaling better insight. Based on these audience generated videos there is a steady stream of cultural production by those invested in the stories and what the art and objects represent. So when one creator asks everyone else as the big tagline to their video, which gear you based on Renaissance queers, they are speaking from within a community to their community, something that institutions as outside this periphery are less in a position to do. And in this sense, the mainstream is distanced from the minoritized. So the role of the creator curators and their emergence from within these communities being depicted in relation to the institutions from where their images and objects are housed can be explored further. And secondly, while creator curators are retelling these stories for themselves or telling ones to their perceived audiences and circumventing these institutions, we can see that there are both divergences from and reinforcement of the narratives around these objects. The biographical information is largely sought from shared sources. The figures featured have perhaps been covered by the institutions before. But the images are also used in reminiscent and new ways compared to this mainstream content. So there's a few layered paradoxes at play. But through the examples of how TikTok users employ digitized imagery to support or focus their topic, we can see that there may be a referencing to the forms and styles of mainstream cultural output, but it has undeniably specific and individual elements added throughout. And these are the kinds of discussions that link with wider issues at the moment regarding authenticity, inclusion and representation. So, thank you very much for listening to my presentation. That's where you can find me on Twitter or if you want to drop me an email and very briefly those are some of my references. So I will stop screen sharing now. Well, thank you, Eleanor, for your presentation there. And if you do have questions for Eleanor, do post them in the Q&A and we will come to them at the end of the presentations. I'd now like to hand over to our second speaker of the panel today. So our second speaker is Laura Marsh and she's from the University of Carolina at Chapel Hill. So Laura is a newly minted PhD candidate in information and library science at the University of North Carolina, Chapel Hill. Her work focuses on the intersections of education, creativity and technology. So I'd now like to hand over to Laura for her presentation on dank memes and exclusion themes, cultural institutions, social media and perpetuating cultural homogene online. Hello and thank you for your interest in dank memes with exclusionary themes, cultural institutions, social media and perpetuating cultural hegemony online. My name is Laura Marsh and I'm a PhD student at the University of North Carolina, Chapel Hill School of Information and Library Science. You can tweet me at theartofmarch or email me at lmarch.unc.edu. My research focuses on social media as a space for informal learning opportunities as well as other intersections of education, creativity and technology. I've also created a digital toolkit on this topic for easy reference. Please head to lauramarch.com slash social-toolkit where I have links to many related resources that may be helpful for you or your organization as well as all of my references in this presentation. You're here because you're interested in talking about how power and privilege can manifest digitally. In this presentation I'm going to share how Antonio Gramsci's theory of cultural hegemony can be illustrated through examples of social media content shared by libraries, museums and archives. I'll also describe how to use an adaptation of Gillian Rose's visual discourse analysis to learn who and what appears in content, how they're represented and what's missing. Let's begin by defining hegemony. First off, the word hegemon is ancient Greek for leader and the term hegemony came to be known as a type of authority that still allowed its subjects some autonomy. For example, Athens had a hegemony over other Greek city-states who at that time followed the military, political and cultural leadership of Athens, but remained somewhat self-ruling. Hegemony is not a direct threat of overt coercion, but a more surreptitious organization of consent. And this concept was further developed by Antonio Gramsci many centuries later, after World War I, to better understand and describe why he and other Italian communists failed while fascists succeeded. Gramsci expanded the idea of hegemony from just politics into a broader understanding of how people understand their own existence. In a world where a person views their own subjugation as a natural state, rather than specifically created for the benefit of another, they're unlikely to work towards dismantling the oppressive system they exist within. In other words, Gramsci believed a social group was able to control others, not simply through external influences like rewards or punishments, but also internal influences by molding personal beliefs, convictions, and concepts of reality to replicate prevailing norms. This internal control is cultural hegemony. Libraries have been described as institutions of cultural hegemony because they reflect and reinforce the dominant worldview of the culture they're created within. We can see an illustration of this here where an American library posted a Today in History tweet wishing a Confederate general who fought for the preservation of slavery, hashtag HBD or happy birthday. And if celebrating this man's birthday wasn't enough, the post was made on Martin Luther King Junior Day. In the response below, we see the institution tweet that they regretted publishing the post on MLK Day, but this apology didn't address why wishing a Confederate general a happy birthday would be appropriate for any other day. This could be considered a racial microaggression or a small encounter with racism that usually goes unnoticed by members of the majority race. And this post also shows that social media work isn't just about how to mediate between people or between people and content. These social media communicators are also mediating between the technology that preprogrammed this tweet, along with the people, their institutions, and larger cultural conversations. LaTanya Autry, co-producer of the Museums Are Not Neutral project, similarly described museums as products and projects of colonialism. And, since they stem from and perpetuate conquest, they are by nature not neutral. While many library, archive, and museum staff may perceive themselves and their institutions as neutral mediators, these practices often maintain and reproduce existing power structures and relationships. We can easily see some of the byproducts of colonialism in many national museums feeds, like this post here, and what I find particularly interesting about this are the responses acknowledging this history and calling it out. But that's not to say that all conversations happening on an institution's feed are necessarily productive or should even remain on a page without an organizational response. Check out this example from an archive's Instagram, showcasing a publicity photograph taken during World War II. You can see some of the responses to the image on the right. The first says, not snowflakes, and the second says, now those look like real heroes to be looked up to and admired. With hindsight being 2020, we can see how this image of five white men became a dog whistle. The puzzle now, however, is to figure out a method for reviewing or assessing content before it's posted to see how imagery can be related to larger dialogues and the surrounding context. Which brings me to Gillian Rose's Fucodian visual discourse analysis. In her words, discourse of visual culture provides ways of seeing brought to particular images by specific audiences or to the social institutions and practices through which images are made, circulated, and displayed. Rose offers six strategies for interpreting images summarized here. One, use fresh eyes, meaning try to approach your materials without any definitive answers or preconceptions. Two, review sources by allowing yourself the time and the headspace to immerse yourself in all the elements that make up an image. Three, identify recurring references and connections made between those sources and elements which may lead to developing bigger themes. Four, examine the effects of truth, meaning consider how this image works to persuade people to believe a certain thing. Five, note complexity and contradictions. Just as images can highlight the process of persuasion, there can be moments where dissent is acknowledged or not. Six, consider the invisible. Absences can have powerful effects too, which we saw in that post of the World War II pilots. There were plenty of pilots and support staff in the Battle of Britain who weren't white men. Let's return to this image from the beginning of this presentation. Using those visual discourse analysis techniques, we can look at this post with fresh eyes and review its sources as a starting point. This post does a great job of illustrating how private and personal aspects of our lives are often related to the operation of power. It also makes use of Gramsci's description of common sense as a means of perpetuating cultural hegemony. In this case, common sense refers to politically powerful ideas about society and one's place in it, and are the result of socialization and exposure to cultural narratives that justify the values of the ruling class. Pulling oneself up by their bootstraps is a typical example of a common sense notion today, which vindicates capitalism as just and valid, as it fosters a belief in social mobility and economic success as being possible through individual efforts to bring class, racial, and gender inequities. Implying this hairstyle constitutes a bad hair day, maybe common sense, to the person who created the snap, but the United States, where this museum is located, has a long history of discriminating against people of color wearing traditional or protective hairstyles. There was a Louisiana law requiring Creole women to keep their hair covered in public during the 18th and 19th centuries. Indigenous Americans and Asian Americans had their hair forcibly shaved by law enforcement, and even within the past few years black kids have been expelled from school for wearing their hair loose. Honestly, if calling natural hair texture bad isn't an example of perpetuating white hegemony, I just don't know what is. Now, before I move on to any additional examples, I think it's important to recognize that we're all fallible. I'd like to share one of my own tweets. This post shows how excited I was over a manufacturer offering paid vacation to their employees through a donation from a holiday gift I received. Using the effects of truth strategy, I can see how I've been conditioned through cultural hegemony and common sense to believe paid vacation is special and deserving of celebration in America, not something deserved or an inaliable right. It's distressing to realize I was so happy that my gift was used by a very successful company to pay their workers something they ought to have had anyway and paid for themselves. Returning to another example from a cultural institution. Here's a museum advertising their space as a wedding venue. I think the visual discourse analysis strategy of complexity and contradiction is most fruitful here. We see a set table showcased in a space where food and drink are usually forbidden, and this can make us view weddings as so important that they're capable of breaking one of the most sacred rules of museum spaces, no food allowed. If we dig a bit deeper into this, we can see how cultural hegemony is used to make weddings and marriage into such a big deal. Some research suggests that as people grow older, women do better than men when living alone, but men do better when they live with a wife. Moreover, the global wedding industry has exploded into a $30 billion per year business. The cynic in me wonders if the wedding and cultural industries conspire together to ensure women are socialized into aspiring for something that only truly benefits wedding businesses and heterosexual old men. I think it's also important to note that two weeks prior to the day that this message was posted, a small wedding reception and a rural area caused a huge COVID outbreak that led to 177 cases, including seven hospitalizations and seven deaths. We've been so indoctrinated by cultural hegemony to believe a perfect wedding is so important that it's worth dying and killing for. There's one more area I'd like to cover today, the idea that inaccessibility is another means of perpetuating cultural hegemony. The post on the left uses colors that do not provide enough contrast for readability by people who are colorblind. Nor does it use audio or alternative text for the 253 million people who are moderately to severely visually impaired. The post on the right is a screenshot from a what's new this week video post from a library's account in which movie and book covers slide across the screen in between flashing yellow lights designed to look like an animated marquee sign. I'm not playing that video here for you because the yellow flashes on a red background do not pass accessibility guidelines which were made to limit photosensitivity induced seizures. Moreover, this library chose to highlight media overwhelmingly created by white authors. Of the 20 featured, there was only one black author and one Latina author. Both authors books were in the middle of this video and pretty easy to miss. And this is particularly troubling because it was posted only a few days after the library published a stand against racism posts in which they stated. 21st century libraries like ours strive to be and must be pillars of equity, inclusion, diversity and democracy for the communities we serve. We intentionally curate a collection that contains both mirrors and windows for everyone should be able to see themselves reflected in a book and everyone should have the opportunity to view the world from a perspective different than their own. So, if there's any interest and extra time during our Q&A portion, I'm happy to bring up more examples of posts for us to analyze together as a group. But before I end, I'd like to make sure that the takeaway from this presentation isn't that social media is bad or that communicators are racist, sexist, classist, whatever. Far from it, market research shows social media is key to diversifying and increasing visitors to cultural institutions. And people trust these institutions. They're viewed as more credible sources of information than governmental agencies, healthcare workers, newspapers and even personal accounts from family members. And people also expect them to recommend actions related to their missions. So, instead of using social media solely as a one-way marketing channel or passing it off to the least experienced staff members, cultural institutions can use social media to build collective knowledge, uncover the past, and imagine new futures. And with that, I'd like to thank you for your time and attention. My name is Laura March and you can always get in touch with me at the Art of March or through lmarch at UNC.edu. Thanks again. Thank you Laura for your presentation there. And I will remind people that they can post their questions now for both Laura and Eleanor in the chat. We'll come to those at the end of the presentations. But I'd now like to welcome our third and final speaker Caroline Bolton from the University of Leeds. Having qualified as an archivist in Archives and Records Management from the University of Liverpool in 2003, Caroline worked in Data and Records Management before becoming an archivist that the University of Leeds in 2016. She is one of the recent cohorts of TNA and RLUK professional fellows, completing her research on archival catalogs as data in January this year. So I'd like to welcome Caroline now for her presentation on catalogs as data, challenging the narrative. Thank you. Hello, my name is Caroline Bolton and I've been an archivist with special collections at the University of Leeds since 2016. During that time, I've worked on a number of projects where I've had to deal with historically sensitive language and representations in catalog descriptions. In 2019, I began researching archival catalogs as data, reimagining archival practice as part of the National Archives and RLUK Professional Digital Fellowship. So today I want to share with you some of this experience and learning and how its informative approaches that we're starting to take to challenge a bias that can sometimes be found in the narratives of legacy collection descriptions, especially where individuals and communities have been misrepresented, marginalised or ignored, undermining inclusivity and presenting a barrier to accessing these collections today. Also, have transitioning to a more data focused approach to our cataloging practices might help to create more inclusive catalogs. Particularly, I'm going to be looking at methods for identifying and treating offensive terminology. Enriching descriptions with structured and linked identifiers to make the stories of those previously hidden, discoverable and engage in remote volunteers and researchers in these processes. Finally, I'm going to summarise some of the challenges and opportunities that I think these approaches have and might present. For just a bit of background to the fellowship, the aim was to explore the practicalities, benefits and challenges of exploiting the data within archival catalogs for both archivists and researchers with a view to broadening access routes into collections, improving discoverability and facilitating new insights that could support both digital scholarship and our own collections development and management. Really, my application was prompted by the experience of providing information for a funding bid when I realised that the collections data that I was able to access via our backend collection management system, which is key emu, could be manipulated to provide all kinds of insight that weren't available to a researcher on the front end navigating our online catalogue. Just looking at the metadata about locations, subject, date and creators gave me a completely different view and understanding of the collection. So using a couple of collections as case studies, I explored the possibilities for collection metadata to be reused, analysed, mined, enhanced and visualised. This really started with a look at the catalogs and particularly to identify areas within the catalogs which were already quite structured and which weren't and what we might be able to do with those. To start with, it was just a case of extracting the catalogue data and simply playing with it in Excel and pivot tables to see what types of information could be gleaned, where data could be cleaned up or enhanced through standardisation of terminology or references and particularly where there were any gaps. I looked at what we could do to integrate new access points such as people, organisations, place and subject as a way of providing new access routes and making connections across and beyond our own collections. I also looked at using more globally recognised persistent identifiers for people, places or things to start to explore the possibilities for linked data but also as a way to overcome the limitations of simple text formats which can be prone to inconsistency, ambiguity and which then often demands further contextual qualification which is not only time consuming but invites narrative. For example, if we see a reference to Stephen McQueen, is that the American actor or the British film director? Finally, what kinds of tools we could look at to do some of this at scale? Experimenting with spreadsheet add-ons to map geo coordinates for place, we experiment with open refine for data cleansing but also for semi-automating matching some of these persistent identifiers to the terms we had in the catalogue using a feature called reconciliation services. I then looked at where the catalogs contain unstructured data and a couple of areas of the catalogue were especially noticeable in that they were often quite lengthy in narrative and these were scope and content and the administrative or biographical history. Both fields set down by ISADG and common to archival catalogs across the globe. Although as a professional archivist strive for objectivity in their descriptions, I think it's fair to say that where we invite narrative we can't escape that these provide quite fertile grounds for bias that can result for a number of reasons. Whether that's a focus on certain narratives of some or the others, quite often influenced by the interests of researchers, catalogers, funders or benefactors, it might be a lack of lived experience or understanding of the people or events that we're describing. I found this certainly to be the case when I was working on our Gypsy Traveler and Roma collections which I'll mention later. It might also be that the terminology is reflective of the period in which they're written and likely reflect the dominant social attitudes at the time. So we needed a way to analyse these narrative descriptions and we were interested in exploring digital tools to do this. So it's part of an ongoing AHRC project called Legacies of Catalog Descriptions. I was able to take part in some training on a corpus linguistic tool called Anconc which was used to help us mine the data. So you might wonder how this can be relevant to tackling historical bias or problematic language. Well that's because it's given us some new tools and a different way of trying to tackle this at scale. As we learn to adapt to the significant changes and challenges that we've seen, especially over the last year or so. Back in 2016 I was lucky enough to work on a project to enhance the catalogs of the university's Gypsy Traveler and Roma collections. These are a series of mainly 20th century collections that notably were created or brought together by various collectors and activists but who were not Gypsy Travelers themselves. We were keenly aware that these featured representations that were often romanticised of Gypsy Traveler life and culture rather than being bi or about those represented. So we worked with Leedsgate which is Gypsy and Traveller Exchange to engage with local Gypsy and Traveller communities in using the collections themselves to explore and challenge these historical perspectives. It was this collaboration and dialogue throughout the project that really informed our cataloging decisions with the effect that we agreed on Gypsy, Traveller and Roma as a more accurate and respectful way when referring to communities represented in the catalogue. Capitalisation was seen as really important in acknowledging cultural and ethnic identity and we supplemented the catalogue statement to explain this. This also resulted in the renaming of the collections for accuracy as they were previously known as the Romani collections. We also took the decision to retain catalogue's historic descriptions but provided general sensitivity statements to the catalogue that both explained and alerted use of historical language that would now be considered offensive. We also discovered that where our academic audiences might be interested in things like subject, place and family names were really important access points for many of the Gypsy and Traveller communities because of their interesting family history and stopping places and travelling routes. So we looked to include these where possible. Finally we ran a workshop which was based on the Collections Trust Revis in Collections Toolkit where we invited members of Leeds Gate to view a series of photographs, paintings and sketches from the collections to create new descriptions that provided such rich insight and context that otherwise would have been totally missed by a catalogue that didn't have that life experience or cultural knowledge. These were then added to the catalogue as an additional description with the contributors acknowledged by their family names at their request. It was a fabulous project to be part of but it was externally funded and the reality is that we couldn't resolve this type of approach for all collections. For a start we might not have the networks or the partnerships with the community to do this so it would likely have very specific applications. Jumping forward to 2020, I began working with the Leeds Archive of Venacular Culture as part of a lottery funded dialect and heritage project. It's a wonderfully rich archive of audio recordings, photographs, surveys and research focused on dialect and fort life featuring the renowned survey of English dialects. It's been extensively described in catalogue during the early 2000s by a research team but for various reasons it still had a quite limited accessibility so the aim was to enhance the catalogue to reconnect it to the communities from which it came. In reviewing the catalogue and also the collections themselves we came across problematic content that included historically sensitive language, opinions and accounts found in the accounts of some of the informants, the lyrics of folk songs, drama and notes of the field workers and researchers. So against the backdrop of 2020 we had the Black Lives Matter movement, the ongoing debates around racism and decolonisation, particularly across the GLAM and HE sectors as well as the COVID pandemic and the switch to remote working and digital delivery. So we had to look at how we could address these issues in a different way. We started to think about access points. Our catalogue was quite heavily skewed towards the creators, that's the stories of the university. So we listed the informants of the survey and the recordings and were able to start challenging the long-held perception that it was men who were surveyed because as the catalogue is shown now we have got the existence of female informants. So it's a small example of how catalogue decisions can uncover those hidden stories. We added places of geolocation which meant that we could plot the collections on a map and immediately it became very obvious and significant when we did this that we had content from Scandinavia and Nigeria which was previously again hidden. We also used antconc, the COPS linguistic tool that I mentioned, to mine the whole catalogue of unstructured and structured data to create simple word lists and counts to help identify any kind of problematic language. One of the hardest things we found was to identify the terms themselves and what we were looking for within the context of that collection. So we only used it in a very basic way but we were very excited by the possibilities and have since secured national archives testbed funding to recruit some linguistic researchers to work with us to look at how we could do more. And I would say that most of that problematic language actually turned up in those unstructured areas that I mentioned before of scope and content and the admin and biographical histories that we previously mentioned but also in the titles. What this means is that it's given us the ability to refine our approach to conducting these types of sensitivity reviews. We've now got a workflow that signposts possible tools and research angles that we or a researcher can undertake to identify problematic language in our descriptions and we're certainly hoping that the results of the testbed will further evolve this. We have a decision tree that makes now a distinction between the voice of a creator which we are more likely to keep and the voice of the curator or the catalogue which we will more likely amend if it's quite excessive but we still retain as part of that audit trail. And then to supplement that we've created additional and revised sensitivity statements that explain this and then we've obviously kept the original descriptions again for future research and aiding the transparency in that. So just to summarise how we can challenge and reduce narrative and the potential for bias by adopting a more data centric approach to cataloging. It's useful to think about structured access points and the opportunities that this provides for discovering the hidden. People's sense of connection to place means that places can be a really powerful route for discovering and engaging new audiences. We also need to look for the gaps in our data and in our collection descriptions perhaps minimising that narratives which are not only prone to bias but lengthy text isn't particularly engaging for digital audiences. We need to maybe think about linking out instead of describing again in a lengthy text looking at classifying and where possible using those globally recognised identifiers because it's much more efficient and reusable. That said it's worth noting that there is even potential for bias in some of these classification schemes. Authorities such as Library of Congress, Getty and Viaff are really useful and widely adapted but they can be domain specific which means they may be limited in scope. And many have their origins in Western cultural heritage and historical significance that's been attributed to aspects or individuals as well as the restrictions on who can contribute new terms to these. WikiData may offer some element of crowdsourcing and democratising these identifiers with the potential to link to online narratives in things like Wikipedia articles. So that's something that we're starting to particularly look at for less well-known parties in our collections. And finally just to say that names of place can be quite contentious whether that's current political situations, historical legacies particularly for ex colonies and colonial names or cultural regions. I went to round off with highlighting some of the opportunities that exist for evolving our approach even further. There's an increasing number of accessible digital tools available out there to explore, not necessarily all that require coding expertise. It's not always possible or appropriate for us to do this alone and this type of catalogue data analysis does lend well to remote research and volunteer and crowdsourcing opportunities. Particularly to engage diverse audiences in co-created enhanced and inclusive catalogues. And as a means of adapting to the reality of the post COVID world where physical access to collections may be limited. I think there is also importance in making collection available of data particularly so that they can be aggregated to enable the kind of analysis and scrutiny that is needed to understand who is representing in our collections and where are the gaps, the inequalities and how that might shape new research and data informed decisions about collections development. And lastly there's lots of research going on in this area to learn from. There's the Towards National Collection, Provisional Semantics and OCLCs, Reimagining Descriptive Workflows, which are both initiatives that we're keeping an eye on. So thank you for listening and thanks to all those who have supported this work and research, especially my colleagues and fellowship mentor. Thank you. Thank you ever so much Caroline for your presentation. That now draws us to a close of the presentations part of this panel. So I would like to invite all the panelists and speakers to turn on their cameras now so that we can move to the second part of the panel where we can take some of the questions that have come through. So please do continue to post questions in the Q&A just at the bottom of the screen and I will keep a look out for them and new questions to come up through. But just as a starting point if I may, I've just got a few questions that I'd hopefully like to kick off with for our speakers today. And my first one is for Eleanor and the question that I have for you is what do you think of some of the barriers and limits are for the queer communities reinterpreting heritage online? There's probably more than one. I think I sort of made reference to the contributors have influence, but maybe not control. So, you know, we talk about algorithms a lot sort of being at the mercy of algorithms, but with Tik Tok, it's a bit more than that, you know, there's teams actively deciding things. So last year, for example, there were conscious efforts to prevent videos that were a hashtag related to LGBTQ plus communities and the disabled communities from being shared once they'd hit a certain number of views. And the explanation was it was to prevent bullying on the platform, because the more popular these videos became the more they attracted trolls, I suppose is what they were getting at and horrible comments. Now you might say well the answer to that would be to look at the accounts that are generating hateful comments. So that was supposedly resolved, but it's things like that that people are up against. There's a lot of talk on Tik Tok about being shadow banned, which is when an accounts views will suddenly decrease and people get the feeling they suspect that the people operating the platform have suppressed their videos, which is probably true. Tik Tok do say that they have local teams that manage the platform in different countries and that they have to abide by the legal requirements in those countries. So if you think you know LGBTQ plus rights are not standardized across the board and they are not as progressive in every country. So, you know, there are certain places where it is actively suppressed because it's a legal requirement, although users will say they find ways around it. So in one of the, so in the slide that I showed with the quotes of captions, someone had purposefully misspelled the word lesbian with an exclamation mark instead of an I so there's a lot of purposeful misspellings that users do to think that they evade the sensors almost. That's that's the feeling that it creates. So there's this really odd kind of paradox of here I am sharing my views. I have a voice. I have thousands of followers or maybe only hundreds but it you know it works for those individuals. But at the same time at any moment Tik Tok could swoop in and suspend your video. And there was one video in the sample where the person had used artwork that portrayed nude female bodies, but by a lesbian artist. So you know challenging the male gaze that kind of thing. And the video was taken down and the only way they could have it up again was to edit out those images. So it, while it can be very empowering in a certain way it's not the yet there's still those hurdles to overcome because ultimately, you know these platforms thrive off your content, but they are in charge, ultimately. And I think from the the barriers maybe for queer contributors from in terms of the galleries and you know I kind of touched upon that relationship between well there this is all happening on social but how does that how does that relate to the galleries in the museums. I think that's a difficult bridge to cross sometimes because I'm sort of torn on the one hand, I sort of want these positions of influence and power that that these institutions are in to to maximize that and some are great at and do a lot of good work. You know in particular it's just been pride month for example so we've been able to see who's been stepping up to the market and who hasn't with their in representative content. So there's this idea of the collaboration and working together and you know it's not the LGBTQ plus audiences are far away from you maybe you're far away from them, and you know bring people in. But at the same time, when I was watching all of this content, I was just so impressed by you know this, you know, really passionate young people who care about their community and care about their history and their heritage. And I sort of think is this maybe a space of resistance, you know, why, why should bigger institutions almost get to capitalize off your hard work. You know, but it's so cyclical, it's so cyclical because the content they're talking about often is held within the collections of the museums and galleries we know. So that's when it comes down to diversifying those catalogs and doing that research so that people are able to take it into their own lives a little bit. So it's definitely a little bit fraught with potholes everywhere because there are barriers and it's an ongoing, I was going to say ongoing conversation but sometimes when I was looking at the sample of videos I thought is this where is the conversation really happening. Are the galleries and museums talking to these audiences? Are they? So that's up for further research and debate, I think. No, thank you for that, Eleanor, and I think it's what was really taken from that is like those potholes that you were saying about these challenges of you getting the content out there, but also having to work with the challenge of the platform, you know, and understand that censorship that you've got to overcome as well, that there are areas in there that the big platforms have maybe put in to protect. But actually work against these communities, so actually having that conversation as you were saying them, or talking to different communities to have that understanding of how we can be supportive and enable a variety of voices to be shared. Nice, thank you for that. I'd like to, if I can do, ask another question then, and this one's directed to Laura, and I think this also touches on the challenges as well that you might say. But are there examples that you came across in your research of the social media content that people are challenging this cultural hegemony? Hygemony, I cannot get that word out today. Fine, the hegemony, it's difficult, and it's just like phenomenology is very, hi everyone, I'm Laura. So there are definitely some great examples of types of content that challenges, the challenges cultural hegemony. And my personal favorite is Brenda Tindall's No Justice, No Peace initiative at the Levine Museum of the New South. And she's social media to solicit stories and exhibition materials for a rapid response program after a police shooting of a black man. I'll drop a link to an article on this right now into the chat. But on a much lighter note, the Uffizi's TikTok is really fun. I'm sure Eleanor has seen some of these as well. I also have a link to a TikTok that has an audio of two guys speaking to each other. And the punchline is that one of them doesn't want anything more in life except a better butt. But that line is given to a portrait of Mary, so a little subversive there. I'll paste my translation in the chat if that's helpful there too. Honestly, medieval TikTok, which talk in general has some great gens really questioning capitalism and patriarchy. It's great stuff. Did you find when you were doing your research as well or that you hit with the censorship issue as well that Eleanor was picking up on or did that not cross into your area of research? I did. And what I find very interesting with talking about producing content is the idea of having outsiders produce social media content and put in unpaid work for that. Featuring content that is free advertising for whatever organization or company. A lot of that happened with Pride in America. I've been seeing a lot of retweets. I think it's one of the models of business that TikTok is using where they're having featured presenters or content creators have sponsored content or ads. And it's difficult to discern what is an ad versus what is actual content that's created without necessarily the desire for capitalizing on that. And it'll be interesting to see probably the UK, the European Union will be the first to discuss harsher guidelines for what is advertised content and what is organic content. And if anybody is interested on the idea of unpaid labor with social media production from audiences, I've got some really great resources on that too. Thanks for that. Now I'd like to ask Caroline a question if that's all right. So my question to you Caroline was what did you see as the main challenges to a data focused approach? I think there's a few different strands. I think one of them is just kind of a cultural shift really towards more co-created catalogs because I think it's kind of challenging the idea of what is an authority of description. I don't know if anybody caught the presentations earlier in the week, Joe Pugh from the National Archives was talking about this idea of led descriptions and multiple voices. So I think that's something that we need to kind of bring into the kind of catalogging approach. But I suppose to be able to support that. I think one of the biggest challenges really is the systems that underpin that. So our ability to be able to document that and that's our collection management systems because they are very much geared towards just that single narrative. And we need to be able to accommodate multiple narratives being recorded against these collections and how we also record the provenance of those collections so that when people are engaging with a collection they can see the context of who's written that description to understand if it's any bias. So I think that's going to be a challenge for us in terms of systems development and working with vendors of collection management systems for us to be able to capture that whole audit process of how we do these collections or how we do these catalogue descriptions. And I think really I think the other thing is around scale really. I think there's two issues. I think there's what we want to do with our legacy descriptions that we inherited but also how we want to go ahead going forward. But certainly the legacy descriptions I think sometimes becomes a little bit of a barrier just because there's so much content there for us to kind of reinterpret. And I think that's why we're kind of now starting to look at tools for doing that and different ways of doing that rather than just focusing on a collection by collection basis. What digital tools and who can we engage with with to help us do that in terms of different voices but also different skill sets in terms of data skills digital skills and who can help us to do that. And you mentioned there that it's about the tools. Did you identify any useful tools to be able to assist with improving the data there and that kind of thing as part of your research? I mentioned a couple. I mean in terms of kind of quality of data we worked with things like Open Refine. We've just started to kind of tinker around with that. That was quite useful. Anconc as I mentioned is a corpus linguistic tool and at the moment there are two researchers that have just been appointed that are going to be working with Anconc to do that. And that's really looking at kind of the language and the power language. We were using it very simplistically with just word counts to try and identify some of the terms that might be problematic. But we're obviously interested a little bit more in the relationships of language and where there might be kind of flags to say actually do you know what this isn't appropriate or it's obviously got some bias in there. So those are two particular tools but I think a part of what I was looking at is looking at it from a point of view of not having a lot of personally digital and data literacy. How can we start to build on that and there are lots of you know much more kind of probably advanced open source tools etc. I think it's getting that you know there's so many more digital options but I think we need those partnerships where people have those skills to help us kind of use some of those tools. And thank you for that. So yeah it's interesting that balance of the tools that are open to all and the tools that are open to developers or coders and that kind of thing. Actually on that kind of data mining point and thinking about how you selected your data we did have a question that came in that was to both Eleanor and Laura which was looking at how did you select your sources for your presentation for your research. So was that a series of searching for things across did you create databases, did you use API scrapers and that kind of thing so would you be able to just share a bit about your research process there. I'll go in briefly because Laura gave a great answer so I'll let her recount it. I gave a type a great answer rather. And so I did it the long hand way, I guess I used to hashtag search terms, and I just saw what the what tiktok gave me in the search results. There is a sort of unofficial API, I think for tiktok, but because I was doing a scoping study and because I knew it would be small. I knew that I would be okay to do it long form so I did I have created my own, my own database for my own purposes with all the relevant data you know date account name URL captions that you know I've captured and the relevant information. I haven't downloaded the videos, I have not done that, but I set up an account purely for the purpose of doing this study. So it had no pre liked videos, it was, you know, the algorithm will be starting to learn based on my search terms, but from the point of starting it had nothing to go on really. Unless it was keeping an eye on my other tiktok account that's just mine, I don't know. But yes, it's manageable when it's smaller numbers that way, I think, but if I was going to do it on a bigger scale, or in a different platform, I might look into an API for that. Yes, so I did use APIs to scrape Twitter, because they only allow you to scrape 10 days worth of content unless you have their academic research API, which you have to apply for but it's free. And then I used our you could also use Python to scrape once you have that API password that they give to you after the application. And I was able to capture like over 100,000 at a time tweets in a few seconds through this and it downloaded as an Excel sheet, and then you can use that in whatever data analysis program that you're going to use afterwards for Facebook which owns Instagram and they've also started using having some integration with Reddit as well. CrowdTangle is the application that is most recommended and I have access through my university, but businesses have, I think, even more access if they pay for it. I think they pay through it through the nose for this access. I'll drop in those links of how to get access to both of these programs and let me know if you have any questions about them. It's pretty difficult to use if you're new to programming, but there's plenty of help Google if you need it. Thanks for that. I've got a couple of questions that come in chat, which are for Caroline actually, and one of them is about, you know, did your engagement with the community did that result in the content for the archives or was it really just focusing on the existing catalog items that you had there? No, because at the time they were running a parallel project themselves to catalog their own archives, so that's held within the community. So we were both kind of working on separate cataloging projects so there was never any intention of us taking a collection from the community. We've got another one that's come for you Caroline here, and it's about how did the set of offensive terms come to be and how did the descriptions or content warnings change according to the sources original and collection displays in the work that you were doing? I think when we first started out with the Gypsy Traveler and Broma collections, I don't think our thinking was probably as advanced. I think we were just alerting people to the fact that there was historic descriptions there that may be offensive. On this most recent project, I think we've kind of, it was a colleague of mine that we've had a look a lot more at the different kinds of voices, and I think that's where we started to make that distinction between curatorial voices and creative voices. So that's kind of set us on a separate workflow for doing that. So what that resulted in was that we might make a sort of distinction between those two and do different things with those. And that resulted in slightly different what we call sensitivity statements to alert people to the content. So where we were saying that actually this is offensive language, but we're keeping it because it obviously reflects the attitudes at the time, et cetera, where we'd actually removed it. We would explain that we'd removed it just so that people were aware that there was something else there that existed because we don't want to sanitise it to a point that people can't access those original descriptions. So it was alerting the fact that those original descriptions are still available on request as well. So I think it was just kind of, it's just, they're almost like building blocks. I think the more that we're doing this, we're coming across more variations and we're adding to that. Thank you for that. It's interesting that you want to keep the original record as well as that historic and we're not trying to rewrite history here, but we are trying to be inclusive in our approaches now as we for people finding that content and ensuring that it speaks in the right voice. One of the things that I thought came across in all the presentations was this theme of authenticity and actually enabling a voice from the community. So actually how we diversify in terms of our social media content in terms of our catalogs, how we work with diverse groups to ensure that authenticity. And I wondered if you had tips from your research that you would be able to share with us of how you can engage a diverse audience to contribute to make things more inclusive. I think one of the things I would mention is certainly when I was working with the jiu jitsu traveller Roma communities with the danger of thinking that certain people represent those communities because I think there's a tendency to think of people as kind of homogeneous groups, but they're not their individuals that have different cultures and lived experiences. So I think sometimes it's kind of thinking quite broadly rather than just specific groups. We were always very keen with the jiu jitsu traveller Roma community that we worked with to say that they were just one of many communities that have very different perhaps histories, et cetera, rather than them becoming the voice of the community because they aren't the voice of the community, they can only obviously give them their perspective. Yeah, I would really, I think that's great. Just to build on that, it's like when we talk about LGBTQ plus communities because every single identity within that is so diverse and specific and unique in its own ways. And we are linked by something overall, like there's shared struggle to different and varying degrees and there are different levels of privilege, et cetera, within the group, but yeah, it's recognising the multiplicity. And I think something that would be really beneficial is to lose this idea of more diverse groups being an afterthought. So we've got our default content, which speaks to our core audience, which is, you know, unconsciously or consciously, you know, your white middle class, et cetera. Oh, and yeah, then we've got to remember that we need to do Pride Month and it's Black History Month coming up. Oh my God, have we like booked enough speakers, you know, like stop that. Like this needs to be like integrated. This needs to be in your social media planner and in your strategy and, you know, thinking about it and yeah, this thing between the sort of constant tension between the centre and the periphery. And I feel like that is a very tricky tension to navigate, but I think if you want to make people feel included and represented, don't let them be an afterthought. I agree so that when you're looking to provide services for your community. Instead, maybe we're favourite as what is your community looking for and how can you work with them to provide that. Nina Simon's book The Art of Relevance is free online has a really great chapter on this. So I'd recommend that. There's also some scholarly research saying that as white women, we are not in a position to speak for everybody and we know and glam. That's who we are mostly and so to sit back and listen to other people and hear what they have to say first and what they want from us and provide that without trying to put our values on it as well. Now I think that openness and acknowledging that we are a group that might not be as diverse and how we how we engage with people and I think in picking up on Eleanor is saying as well as the diversity within groups as well, that is a community and hearing multiple voices from multiple places. We just need to show that we are being inclusive in our approach and not saying that just that, you know, not projecting really is the thing isn't it that we want to ensure with this kind of thing. There was another question that's come through to Caroline, which was about your workflow for using open refine. So I wonder whether you will be able to just expand a little bit more on how you worked with how you started using that for your data cleaning. I don't think it's particularly a workflow. I think it depends on the collection that we're working with. We started to experiment with the features of open refine and we were just simply using there's lots of online tutorials with open refine. I think again, going back to the idea about accessibility of these tools and open refine is another good one. And obviously the kind of learning and personally for me the kind of video tutorials are really useful, but we were certainly using it to tidy up a lot of our data. So the kind of standardisation about our data. But I think the biggest benefit that we were coming across was this idea of looking up identifiers, whether that was wiki data identifiers or VF. We were looking at their reconciliation services, which is an add-on service within open refine. And I think that was one of the things that we were kind of most excited to try and progress because I think if we're looking at embedding kind of linked data or the potential for linked data. One of the biggest challenges is how we do that and we started doing that manually and it's just obviously not sustainable right across collections. But open refine was able to do that quite quickly on several hundreds of parties that we came across. So I think it really has just been trial and error with open refine and I don't think we've anyway touched kind of, you know, all the features of it. It's something that's still ongoing. But yeah, I think the reconciliation services was something that we definitely want to kind of further. Thank you. We've got another question that's come in, which is to all speakers, which is looking at what implications do you hope these intersectional approaches, especially within fairly homogenous archive sector, will have not just for the users but also staff in the future. So I find it very interesting that social media managers in particular and that's what my dissertation work focuses on social media managers at cultural institutions are one of the shortest tenured professionals at their place they believe their job faster than any other type of professional and that they are usually paid the lowest and that they are also a voice of their institution. So the executive director who pours over a brochure that is seen by maybe 1000 people isn't necessarily taking the time to check out or create a strategy for social media posts that reaches millions. So to consider where your resources are, and to make sure that the staff who has that responsibility is prepared and is creating a strategy and fulfilling that that works with what you're hoping to do in your organization. I think I thought it isn't somebody's niece or cousin that needed a job that it's an actual position that deserves respect, and that it deserves to be recognized as a channel that's probably the most important channel for diversifying visitors and for connecting with community members. I really hope for staff that for example so whatever sort of minoritized community they may be a part of that they don't have to shoulder the burden of being the one voice to represent that community, because it's really hard to do it on your own, and structural change embedded across an organization and if we're talking about organizations that have long histories that are maybe slow to change, it can feel like you're tripping away at a huge monolith sometimes. I think I would hope that there is more diverse staffing, and that those staff are supported, and that they are heard and listened to and that they have some power to make the decisions, because what I've often found is that you get a lot of people with the right ideas. If they don't have as much experience, they want to do the right thing, or they really are geared up and ready to go, but their power only stretches so far. And ultimately there's a bit of a, you know, eventually they have to sit down with the senior management or what have you and mount a case for why diversity activities should happen. So, yeah, more to have that intersectionality reflected more in staffing choices and in the culture of organizations and for people to be supported and empowered, I think, to be able to deliver the change that they might imagine. And that requires effort, top plan, you know, bottom up and top down, and then you hopefully meet somewhere in the middle.