Negotiating hybridity, inequality, and hyper-visibility: museums and galleries’ social media response to the COVID-19 pandemic

ABSTRACT This article examines the impacts of COVID-19 on the digital work of museums and galleries in the UK, 2020–2021. Focusing on social media activity, we explore two questions: (1) How did approaches to, and institutional perceptions of, social media shift during the pandemic? and (2) Looking to the future, what practical and theoretical challenges do social media present for museums and galleries, and what are the related policy implications? The discussion draws on a mixed-methods study including an analysis of 9000 tweets, and reflective semi-structured interviews with 19 digital workers. Our findings can help shape global digital heritage practices as we emerge from the pandemic, enabling more dynamic and meaningful forms of cultural participation, and underpinning more confident and ethical social media trajectories.


Introduction
In this article, we investigate impacts of the COVID-19 pandemic on the digital work of museums and galleries, principally in the United Kingdom (UK).We take social media activity as an entry point, examining a substantive dataset of Twitter posts by museums and galleries (March-May 2020), as well as reflections from Digital Managers obtained via a series of interviews one year later. 1 Working across these datasets we explore two key questions: (1) How did approaches to, and institutional perceptions of, social media shift during the pandemic?and (2) Looking to the future, what practical and theoretical challenges do social media present for museums and galleries, and what are the related policy implications?This second future-oriented exploration is all the more pressing given how visibly debates about inequalities, decolonisation, participation and representation play out for museums and galleries within social networks.
In what follows, we demonstrate the richness of social media activity during this period, but also its volatility.We present an overview of approaches to social media during lockdown, and the institutional considerations that shaped that content.Looking forward, our findings highlight the importance of supporting cultural professionals to: (1) evaluate social media activity with care and reflexivity, (2) appraise the use of internet intermediaries and approaches to datafication, (3) think sensitively about digital inequalities and how they function beyond access, (4) unpack what hybridity means in practice, and what it makes possible, (5) nuance the notion of online community, (6) reinforce museums' activism in the online environment, and (7) advocate for digital practitioners whose hyper-visibility in social networks can mean they feel exposed.These issues are of critical importance to the sector as it emerges from the pandemic, and for researchers in related fields including cultural policy studies, museum studies, and studies of digital culture more broadly.Our insights can help shape global digital heritage practices as we emerge from the pandemic, enabling more dynamic and meaningful forms of cultural participation, and underpinning a more confident and ethical trajectory for institutions operating in the "searchable talk" (Zappavigna, 2011) of social networks.
In the next section, we introduce research into museums' digital work during this period, and take a broader look at digital heritage scholarship, including on social media.In section 3, we introduce our approach, detailing methods used and the operational challenges they presented.In section 4, we discuss findings from the study, before offering some concluding remarks.

Research context
It is common to talk of a "pivot" when describing cultural institutions' digital response to the pandemic (Kidd et al., 2021; Walmsley et al., 2022).Much has been written about the shift to digital practices and the expansion of related capabilities that was required during that period as virtual tours, online workshops and events, podcasts and social media became core activities. 2Ryder et al. (2021) propose that the pandemic "transformed digital content into [institutions'] central message and social media into their primary communication medium" (2021, p. 1), arguing that social media use helpfully shifted institutions' tone, demonstrated transparency, paved the way for dialogue with audiences, and increased visibility ("followers") during that period (2021, p. 7).Kist goes further, maintaining that social media output helped cater for the public's "emotional needs" by providing "positive distractions" (2020, p. 345), and Burke et al. conclude that social media "buoyed up visitor communities" (2020, p. 117).
This begins to suggest that institutions became more audience-centred in their digital outlook during the pandemic, and that there may have been a shift in how senior management teams value and make sense of social media activity in their digital mix, or as part of their "digital estate" (Price & James, 2019).Typically, the success of social media has been understood by institutions through metrics such as likes, shares or number of followers, but Ryder et al. propose that this shifted during the pandemic, as "building connections and communities" became the primary goal of digital content (2021, p. 12), regardless of how extended or enduring those connections and communities proved to be.In this article, we explore the parameters of this emergent audience-centricity, attending to the perspectives of cultural workers, and the particularities of content shared, rather than focusing solely on the metrics.
Another growing area of interest during the pandemic related to the premise and promise of "hybrid" approaches for the cultural sector (Noehrer et al., 2021; Walmsley  et al., 2022).Hybridity 4 is a particularly seductive notion when considering the blurring of digital and physical experiences in our everyday lives, and given the in-betweenness that characterised many cultural encounters during lockdowns.There was increased experimentation with blended approaches to events, games and tours for example, including "the production of digitally-mediated material encounters" which implicated or involved online audiences (Galani & Kidd, 2020, p. 300).As a result, there is energised consideration of "the diverse material implications of digital engagement" (ibid.), which suggests new professional attitudes and approaches, as well as avenues for research enquiry.The notion of hybridity, and how it can help us understand activity in social networks, is something we were also able to investigate further in this study.
The exploration of museums' digital activities is of course not new, 5 and in the sections that follow we draw on scholarship documenting a range of practices, including social media. 6We also connect with research on digital ethics, digital inequalities, and digital culture where helpful.Social networks raise questions about data, voice, ownership and power that cultural institutions cannot ignore, and have often struggled to meaningfully respond to (Kidd, 2020).Risk-aversion within these contexts has been noted, and scholars have been critical of institutions' (lack of) attention to their social media publics, highlighting their propensity for broadcast messaging, and a lack of dialogue in too many cases. 7This can be understood as consequential because it undermines or actively sidesteps the potential of those environments for richer collaboration and exchange.As McGrath notes, at their best, social media interactions might provide "a welcome reprieve from content that aspires to the more didactic tone of traditional museum labels, or the hyperbole of the traditional press release" (2020, p. 168), although these outcomes are never inevitable, and rarely are they predictable.
In what follows, we also connect with literature highlighting cultural institutions' potentialand responsibilityto engage with social justice concerns and to examine their civic purpose. 8This has been a growing focus of academic research and debates within the profession, which crystallised in heated discussions about ICOM's proposed new definition of a museum in 2019.Before the pandemic, campaigns such as #MuseumsAreNotNeutral 9 and #MuseumWorkersSpeak had begun to highlight the potential role of social media for activism and change-making in the sector, and during the period under scrutiny in this article museums responded to the murder of George Floyd and the Black Lives Matter protests.These were also matters that featured in our research then, not least in interviews.
That museums and galleries have become bolder or more confident in navigating these concerns and contexts because of the pandemic is an enticing proposition, but it is one that has not yet been interrogated, or its complexities unpacked.This article takes a longitudinal approach to exploring these developments, as is outlined in the following section.

Methodology
This study was carried out in two stages.Firstly, an analysis of Twitter content across two hashtags March-May 2020, and secondly, a series of 19 reflective interviews May-June 2021.

Analysis of tweets
Data were collected from Twitter on the hashtags #CultureInQuarantine and #Museu-mAtHome during the first six weeks of the UK lockdown. 10These hashtags were utilised CULTURAL TRENDS by museums and galleries during this period as key connective devices.The resulting dataset (n.9000 once duplicates were removed) was quantitatively analysed in Excel using the available metadata from Twitter, before a random five per cent sample was more extensively and qualitatively analysed using NVivo (n.450). 11This more nuanced analysis meant we were able to remain attentive to each tweet's context and the juxtaposition of elements it contained.So, reply tweets could be analysed in relation to original tweets, and those that contained combinations of text, images, links, animated gifs or videos were considered in their entirety. 12This smaller sample was a "thickening strategy" for the analysis in recognition that "data abundance" is not in itself an indicator of research quality or insightfulness (Latzko-Toh et al., 2016).During the qualitative coding we made note of (1) the tone of each tweet (the type of expression or emotional register), ( 2) the theme or themes, (3) the types of content and how they interacted, (4) additional hashtags being utilised, (5) the presence of hybrid approaches, and ( 6) the values being recognised or debated.Using the NVivo sample we have been able to explore these attributes across multiple variableswho is posting and engagement metrics for examplewhich provides incredibly rich insight. 13ith a view to informing future studies we acknowledge the limitations of any analysis of data of this type, and in this way.Firstly, it is evident that our choice of hashtags has shaped the dataset in profound ways.There may have been plentiful and interesting data that fell outside the orbit of these hashtags.Hashtags are useful in practical terms, however, systematically pulling together a sample across varied constituencies and priorities.Secondly, our study is limited to a six-week period at the start of lockdown; a unique context in terms of public mood, levels of anxiety about COVID-19, restrictions and closures, and the types of Government and other support that were in place.Thirdly, institutions were not connecting with a representative cross-section of the UK public through Twitter (Sloan, 2017), 14 so we have analysed interactions with that limitation in mind.As noted previously, cultural institutions typically understand engagement on social media in limited waysnumbers of likes, comments, or shares for example.In this study, we were able to get beyond these "vanity metrics" (Rogers, 2018, p. 454) by using thematic analysis, which, coupled with accounts from the reflective interviews, offers a more substantive insight into social media activities, and their associated challenges.
There were practical reasons for our choice of social network that are worth noting.Chief amongst those was the fact that most museums and galleries in the UK have a Twitter presence, even if their frequency of posting might differ considerably.Acquiring a dataset was also more straightforward for Twitter than it would have been for other sites such as Instagram or Facebook.We acknowledge that this too has shaped our data, and that studying interactions on other social networks might have revealed a rather different picture.

Interviews
Another limitation of the Twitter analysis was the fact that cultural institutions' motivations and ambitions in this changed context could only be inferred from short posts.Moreover, there was no way for us to know how they were assessing the success or otherwise of these endeavours.To explore these perspectives further, a series of reflective interviews with Digital and Social Media Managers was carried out approximately 12 months later.Given restrictions in place during the pandemic we used the video conferencing platform Zoom to interview 19 people across the UK, nine working for national institutions, and 10 for independent or city museums. 15These discussions explored priorities for digital work during the pandemic as well as overarching reflections on pivotal moments, successes, challenges, strategy, audience engagement and analytics.

Findings
This section details findings from the study in two parts, working across both datasets.First, we offer an overview of approaches to social media, exploring institutional considerations that shaped content during lockdown, and how perceptions about the value of social media activity changed during that period.In the second section, we highlight practical and theoretical challenges that are likely to intensify as we emerge from the pandemic, and that demand further consideration and support.

How did approaches to, and institutional perceptions of, social media shift during the pandemic?
There was a timeline for social media activity during the first UK lockdown that is visible in our Twitter sample and emerged in discussions with our interviewees.In the first weeks, there was a "reactive" 16 period of intense communications activity to inform audiences about closures and provide entertainment or distraction, whilst museum and gallery staff also transferred to working from home.This period put significant pressures on digital capacity, not least because everybody suddenly wanted a "piece of digital".This was followed by a second phase which saw increased experimentation and the use of more "tactical" approaches, underpinned by shifts in internal processes to support that work, in many cases because colleagues from other departments were by then on furlough.Later, a third phase identified in the interviews saw practitioners thinking about re-opening, with much of their digital work re-oriented toward supporting that ambition.
One of our main findings across both datasets relates to an interplay identified between two modes of engagement.One emphasised the promotional function of social media to provide information and ensure the visibility of institutions at a time when doors were closing indefinitely.The other was concerned with two-way interaction or dialogue between institutions and their publics, as elusively postulated in notions of the "connected museum", and social media in particular (Drotner & Schrøder, 2013).We found instances of a change in tone, reflecting the fact that digital practitioners were able to experiment with more interactive and playful content, as well as a new voice, to speak to the mood of the lockdown.These two modes of engagement sometimes worked in consonance, but at other times were in tension.Interviewees noted, for example, that they were keen to remain "on-brand" and consistent with institutional ambitions, yet they also saw social networks as a good context to pilot new initiatives.Even where these turned out to be unsuccessful, the sense was that the opportunity costs were lower than if they were trialled elsewhere in the digital estate.
On one hand, posts by museums and galleries in our more focused analysis (see Figure 1 17 ) demonstrated a slant toward a promotional tone in 73 per cent of cases, coupled with cross-promotions (i.e.promoting other venues or activities) in 8 per cent, and providing information in a further 39 per cent. 18This broadly tallies with previous research demonstrating that cultural institutions tend to use social media for marketing and promotional purposes, relying heavily on broadcast messages rather than conversation (Baker, 2017; Kidd, 2014).This shouldn't be surprising given that, in our interviews, respondents almost universally noted that social media was the responsibility of those with oversight for marketing.
On the other hand, 48 per cent of posts by museums and galleries in our sample asked for engagement, 19 and 23 per cent issued a call to action with a response requested. 20his shows there were attempts by museums and galleries to engage audiences through questioning, crowdsourcing initiatives, interactive content and quizzes (for example) and to inspire them to get creative at home. 21Tweets posted by museums and galleries were significantly more likely than others in our sample to ask for engagement in this way. 22One interviewee reflected on this approach: "That was the sort of content that performed incredibly well because people could make and then share that content online … it really tapped into the community interests that were happening online at that particular time".
Our interviewees felt that social networks present possibilities for dialogue with audiences, but they understood all too well that such engagement tends to be unpredictable.Regardless, the change in tone during this period was overwhelmingly described by our interviewees as positive, and even "liberating".Practitioners understood the benefits of using more dynamic or flexible modes of address on social media, including those that were playful and less deferential: There's been a lot more willingness to just let go a little bit and just have a bit of fun with it … before the pandemic it was very corporate and very standardized and everything had to be checked and double checked and triple checked … I think people have really appreciated the tonal shift with museums.These approaches work well given the (digital) cultural logic which underpins, and evolves within, these environments (Hjorth & Hinton, 2019; Leaver et al., 2019).Interviewees reflected, however, that this logic was not universally recognised within their institutions, and that writing for social media channels was a misunderstood and undervalued skill.
Playfulness was a value identified in 26 per cent of tweets by museums and galleries (compared to 12 per cent across the whole sample), indicated through expressed appreciation for light-hearted content, interaction across series, or the use of games.We also coded for expression of emotions in 7 per cent, 23 and for humour, irony and sarcasm in 13 per cent of tweets. 24Across the whole sample (n.9000), tweets with video performed extraordinarily well on attention metrics such as likes, retweets and quote tweets, offering behind the scenes tours for example, often in spontaneous or amusing ways.This is in keeping with Najda-Janoszka and Sawczuk (2021) who report on museums' successes using video content to spark interaction.Tweets featuring animated gifs also did well by these metrics, which is perhaps unsurprising given how prevalent they have become within our everyday digital interactions (Miltner & Highfield, 2017).
Related to these changes, we found that the tone of content was broadly of a benevolent nature.This was a deliberate strategy identified by our interviewees: "our priority really was just to provide some lightheartedness in quite a bleak landscape".Another of our respondents commented similarly, noting that this conscious performance of positivity doubled as a way of touching base with those in similar roles within other institutions, at a point when they were all "sort of treading water".Differing modes of engagement mapped well onto the most frequent themes identified in our analysis (full overview in Figure 2 25 ).
Perhaps unsurprisingly given the broad educational remit of many museums and galleries, "Education" was the most frequent theme, present in 59 per cent of tweets. 26This included a wide range of tweets, from those promoting educational programmes, to those which nodded to lessons we can learn from the past in relation to the specific circumstances of lockdown.These tweets tended to utilise the more promotional or informational tone identified previously.
Tweets coded under the "Education" theme often garnered particularly high levels of interest and traction.Relatedly, tweets aimed at children were also fairly visible in the sample, which is unsurprising given the emphasis during the early weeks of lockdown on providing activities to pass the time or for homeschooling.Interviewees noted that this approach was at times misplaced for several reasons, including families not being part of their core audiences, the timeliness of material produced, or competition from other and bigger players. 27he second most recurring theme for museums and galleries was "Arts as a way of coping" (47 per cent of tweets from museums and galleries, and 59 per cent of all tweets).This included tweets related specifically to COVID-19 or that were expressly inspired by its impactsthe restrictions, loneliness, fear or boredomand that referenced the role of the arts in enhancing wellbeing and making life more tolerable in lockdown.These tweets were more likely to feature the dialogic and more playful tone identified above.
We found that institutional size, structure, governance, and digital maturity mattered when it came to pandemic response.We were unable to discern this from the social media analysis itself, but it was very evident in the reflective discussions.All our interviewees worked in institutions with at least a rudimentary digital strategy, and those with an extant infrastructure were able to make the shift online more seamlessly.In terms of that infrastructure, our interviewees noted that pre-existing digital assets, a skills base which had been nurtured over time, and a defined and widely understood trajectory, were far more important than access to technology (devices, storage or connectivity) in and of itself.Digital teams in institutions that were undergoing periods of change in structure or governance felt the impacts of the pandemic greatest.
Naturally, the structure and size of digital teams also impacted what was deemed achievable during this period, and there was often expansion of teams over time through new appointments and internal secondments.Despite an existing skills base being important, respondents were clear that a view of digital competence needed to be taken across institutions, rather than attempts made to upskill everybody.Again, digital capability was understood not as a technology issue, but as a mindset or approach which could be characterised as collaborative, open, participatory, and, above all else, audience-centred.
Interviewees reflected on content strategies during lockdown, remarking that some social media activity (by no means all) had been a notable success story in terms of levels of interest.The importance ascribed to social media by interviewees is difficult to overstate: "Whilst we're closed you know, our social media platforms aren't promoting the museum, they are the museum".Digital practitioners were wary about "spamming" followers, however, and at times felt the need to push back against perceptions from elsewhere in an institution that increased volume of output was in and of itself a good thing during this period.Interviewees understood the need to be considered in their engagement, and to remain attentive to online community norms and logics which differ by platform, 28 appreciating how these might best be navigated or indeed "exploited".We heard strong evidence of a shift, in many cases a profound one, in how institutions' senior managers perceived, and valued, social media activity over the course of the pandemic: We were at a tipping point anyway, but I think [the pandemic has] accelerated it … real positives have come out of our work over the last year that we can take forward and it's not going to be a case of going back to where we were.
As might have been expected, the biggest shifts in perspective were being felt in less digitally mature organisations.Such changes in perceptions are significant because they may lead to a strengthening of capacity for this work in future, increased interest in contributing to social media output within institutions (outside of the initial flurry of activity experienced during the pandemic), and, in time, more scrutiny of that activity too.

What practical and theoretical challenges do social media present going forward?
In this section, we explore a series of practical and theoretical challenges which emerged across our study.Grounding our discussion in evidence from our samples, we highlight the importance of supporting cultural professionals to: (1) evaluate social media activity with care and reflexivity, (2) appraise the use of internet intermediaries and approaches to datafication, ( 3) think sensitively about digital inequalities and how they function beyond access, ( 4) unpack what hybridity means in practice, and what it makes possible, ( 5) nuance the notion of online community, ( 6) reinforce museums' activism in the online environment, and ( 7) advocate for digital practitioners whose hyper-visibility in social networks can mean they feel exposed.
In reflecting on the activities highlighted in section 4.1, interviewees reported that their evaluation of social media activity was often done in rather simplistic ways, understanding success solely in terms of the available "vanity metrics" (e.g. from Google Analytics or Matomo).There are a number of concerns related to the use of social media metrics in this way, including about what limited insights might be derived from the rudimentary demographic data they surface, and digital practitioners were aware of these challenges.Despite reservations, however, practitioners do use these metrics to help justify and explain their work to managers and funders, who, they say, can otherwise find these activities quite difficult to grasp.In this sense, the metrics become a kind of performance within the "success theatre" of social media (Rogers, 2018, p. 454), a "projection" of success or "klout" which serves a purpose, but problematically so.Practitioners were clear that more expanded and qualitative assessments such as that presented in this article would be helpful in interrogating those metrics and articulating the more subtle or expansive ways in which engagement functions, but they do not have the resources to do that work.They are keen to find ways of moving away fromor nuancingdata from third party providers which they maintain also present ethical challenges.Datafication, and attendant questions about power, exploitation, and stewardship of data, are pressing issues for museums as they increasingly rely on "data-intensive practices", although these practices remain somewhat obscured at present, and we would do well to understand them better (Pruulmann-Vengerfeldt, 2022, p. 63).Such exploration should be coupled with further consideration of the role that social networks (the platforms themselves) play as intermediaries; implicated within global flows of information and often disinformation, yet reluctant to accept that responsibility or regulation should follow from that fact (Iosifidis & Nicoli, 2021).
Where interviewees were able to comment on online audience demographics during lockdown, there was a sense that despite a growth in numbers, audience composition hadn't changed in any meaningful way.By and large, they reflected, online audiences tended to mirror in person ones, and lacked diversity: In terms of other audiences, diversifying, I'm afraid no.I think we failed to do that and it's something we really have to work on doing.But the Museum has to work on it as well and it's not just a digital thing, the two things go together.
In interviews, reflections on audience diversification sometimes blurred with discussions about access as is demonstrated here: we were definitely conscious that we had the ability to reach more people than we had been able to, but were also conscious that not everybody has access to the Internet … of families in particular not having, you know, four laptops.
Access itself was an issue sometimes understood only in terms of "haves" and "have nots", to the point where, as one interviewee told us: "Because our remit was digital, the people that weren't accessing digital were kind of someone else's problem".
Several respondents offered a more nuanced account, however, considering how age and socio-economic background for example might impact engagement, and how disparities can be further polarised online.As one interviewee noted, "I think we shouldn't underestimate the scale of the problem right … .This is not an arts sector problem to solve.This is a socio-economic problem, it's not a cultural problem".Such insights indicate efforts by a number of practitioners to better understand how digital inequalities function in practice, and demonstrate an awareness that binary assessments of access/no access underplay the "social, economic, and cultural factors" that shape participation (Wilkin  et al., 2017, p. 333).According to Robinson et al. (2015) a consideration of "first-level" digital disparities (access) needs to be coupled with acknowledgement of "secondlevel" inequalities "such as those related to skills, participation, and efficacy" even for those who we might consider to be "users" of technology.As seductive as it might be to assume that social media have a democratising function, we know that these networks do not constitute a diverse and open conversational sphere that anyone can be a part of.Inequalities are not simply technological or economic phenomena, but ones impacted by our everyday social circumstances and struggles (Helsper, 2021; Ragnedda & Muschert,  2018), giving rise to distinct practices, or what Watkins calls "techno-dispositions"& nbsp; (Watkins, 2018, p. 2), which a few of our respondents wish to further understand and interrogate in relation to their own institutional contexts.These respondents have moved beyond understanding access largely as compliance with laws or best practice, and instead, echoing Kudlick and Luby (2019) in their work on "access as activism", have begun to centre inclusive and universal design and development principles, an outcome which it is hoped will benefit all users in time.
One way in which museums and galleries tried to diversify content and audiences during the pandemic was through the increased use of hybrid or blended approaches.
As noted in section 2, discussions about hybridity gathered steam during the pandemic (Galani & Kidd, 2020; Noehrer et al., 2021) and we saw evidence of such approaches in our content analysis, where complex interplays between materiality, digitisation and remediation were observed in 85 per cent of tweets. 29This included behind the scenes tours, downloadable activities, use of digitised collections, and calls for audiences to get creative at home, like this example from the Ashmolean Museum, re-produced here alongside a sample response tweet (Figure 3).
These attempts sought to connect users not just with ideas, or with an institution, but with the materiality of culture and collections in ways that go beyond representation.These practices extended Miles' notion of "digital staging" (2018, p. 306), demonstrating ways of producing, constructing and co-curating outputs that are given a "generative force" through digitality and (sometimes) liveness, an approach embraced by several of our interviewees, who noted that they were "trying to gradually move towards having a much more hybrid approach" "so the two [online and on site] can support each other".
The notion of hybridity allows us to articulate assemblages produced in the relationality of digital and analogue things, their co-dependency and (often) in-betweenness (Galani & Kidd, 2020).In doing so, it can go some way toward alleviating concerns that institutions and/or some users might have about the seeming intangibility or "ephemerality" of social media interactions, an anxiety evidenced by one interviewee as follows: "everything is vanishing over kind of 48 hours … it feels instinctively odd to me for museums to do that".Such unsettlement, steeped in notions of museum and gallery purpose centring permanence and physicality, may seem anathematic to social media users whose everyday experiences of "the digital" and "the material" are now indistinct (Farman, 2012).Hybrid encounters can embrace that fluidity.Well-conceived, it is possible that they could help mitigate the "digital divide" in some instances (Child et al., 2021) or encourage participation in audiences that might instinctively feel "digital ambivalence" (Macfarland et al., 2020) about interactions not (on the face of it) "in real life".In our interviews, we heard evidence of increased engagement by local audiences, with content appealing "to people that have maybe more interest in their local environment, their local community" because of the pandemic.Significantly, local and hyperlocal communities were mentioned more frequently in discussions than international audiences or "visitors": "We were like, we are a community resource … we are here for you"."Place" as a theme appeared markedly in tweets by museums and galleries (29 per cent of tweets 30 ), often indicated via the use of hashtags which situated tweets within (hyper)local contexts.This was sometimes in tension with content reflecting on the more universal value of arts and culture as a source of solace, and echoes other research demonstrating the importance of localising events even within the context of global cultural programmes and engagement on social media (Ozgul et al., 2022).
These findings resonate with broader developments in the sector 31 which centre museums' social and civic purpose (Chynoweth et al., 2020; Janes & Sandall, 2019), 32 and which were accelerated during the pandemic as institutions re-evaluated their relevance to local communities as part of a "pivot to purpose and people" (Robinson, in  Walmsley et al., 2022, p. 64).Our research shows this happened in relation to their digital work also.Amelia Wong (2015) has called for a more nuanced consideration of what specifically organisations mean when they talk about "community" within the context of social media, and our research suggests this would be a timely discussion.Pre-pandemic, museums and galleries (certainly those in big cities, or with national profiles) may not have thought it helpful to tie their social media ambitions too much to their geospatial realities, yet during lockdown, they appear to have become more mindful of those realities and their possibilities, not least because tourism and travel had become impossible: "It was pretty clear that the recovery program out of lockdown would rely on local audiences".
How related debates about museum activism and social justice translate into the online environment is yet to receive extensive examination, but this issue was raised by all our interviewees, not least in relation to the Black Lives Matter protests in the summer of 2020, and the compulsion museums felt to respond on this and other "contentious" issues.Some felt well supported by management in taking a strong stand: "We could be strong in our [belief] as well … That came right from the top".Whereas others had struggled with the limits of their institution's response.All our interviewees felt strongly that cultural institutions must take an unequivocally anti-racist stance and were keen to see that connect meaningfully with work to decolonise and diversify narratives and collections.Where interviewees felt such connections were not thoughtful or purposeful, professional and personal discomfort followed: "I just got very frustrated and disenchanted with senior management's just kind of spinelessness … And it just sort of made me feel like I was on a different value trajectory to the management".
These developments were a consideration for those working in digital roles especially, as they fielded public responses to institutional statements or actions, often negotiating adverse reactions and trolling.Our data show that practitioners were keenly aware of the ways social media might enable both "solidaristic, mutual, collegial and parasocial relationships" as well as "hateful communication", including "dehumanising, discriminatory, threatening and/or abusive messages" (Banaji & Bhat, 2022, p. 119).Navigating this duality can be exhausting, and interviewees connected that professional reality with broader challenges related to work/life balance, stress and mental health which they felt had been exacerbated during lockdown. 33In our research, these concerns were often linked with a particular apprehension about those working in social media environments being hyper-visible spokespersons for their organisations: "Being a social media person makes you quite exposed and can impact on your mental health … the member of my team who does social media has felt very much in the front line … particularly when it's been quite horrible abuse".That feeling of being "exposed" was no doubt amplified during periods of closure whilst many other museum and gallery activities ceased.Practitioners felt hyper-visible in comparison to their co-workers, and thus under increased scrutiny from colleagues and managers.This scrutiny was deemed more troubling where practitioners felt their work was not well understood, or that they were not trusted: "there was definitely tension of people not quite understanding how social media works … You have to have faith in me doing my job and doing it to the best of my ability".
Despite the potential negative effects of this over-exposure, there were also new opportunities for experimenting with different forms of engagement, with audiences and others in the sector.In our dataset, we found that 12 per cent of those tweeting on the hashtags were cultural workers themselves (half employed in institutions, mostly museums).There is perhaps a museum equivalent of the "employee influencer" phenomenon emerging, as we have seen in other sectors (particularly retail).This could present both possibilities and challenges going forward.

Conclusion
In this article, we have explored the richness of social media activity during the pandemic, but also its volatility.Varying levels of digital maturity have informed the practices of museums and galleries within the digital environment, but all have seen the benefits of re-affirming their commitment to social media.We have highlighted the interplay between different tones and modes of engagement that prevail, sometimes in tension, within these contexts, and have found evidence that social media interactions became more dialogic and playful during the pandemic.Museums' social media communications were intriguing.No sooner had institutions used their social media feeds to announce (physical) closures, than they launched into a period of quite frenetic activity on those same platforms: a performance of their openness.Throughout this period they played a part in creating and shaping many people's experiences of an intensely challenging present.
Both institutions' digital maturity and their audiences' digital competencies emerged as important issues in this research.While some cultural institutions had been on clearly defined and widely understood trajectories in their digital work pre-pandemic, others lacked vision, extant infrastructure, and support.Notably, our interviewees were universally in favour of understanding digital capabilities not in relation to technology per se, or institution-wide acquisition of digital proficiencies, but as a mindset or approach.Going forward, they want to prioritise the nurturing of audience-centricity, coherent user experiences and rich storytelling opportunities.
We also identified issues worthy of further research and fuller consideration by policy makers and funders.The first of those relates to the above point about the resourcing of digital activity.Most content discussed by interviewees or featured in the Twitter sample was free to access. 34This was sometimes because output used pre-existing digital assets or was produced at speed, and, in other instances, was due to uncertainty about the appropriateness of monetising content during an unprecedented time of crisis.Our interviewees felt strongly that the time has come for a conversation about whether, when and how museums and galleries might monetise digital content.
Secondly, our research demonstrated efforts by practitioners to use social media to support institutions' civic goals around community, wellbeing, and the environment.There was a benevolence in evidence across our datasets, a desire to be helpful, compassionate, and public-spirited, as digital practitioners were at the forefront of a "pivot to purpose" (Walmsley et al., 2022).This gentle civic-mindedness will no doubt need nurturing as we emerge from the pandemic if it is to translate into a more sustainable and confident humanitarianism, connected with efforts across the sector to champion social justice and reform.What steps do we need to take to ensure the legacies of this "pivot"whatever those might beare well-supported, just, and fair?
A third connected strand in our research documented an expansion of possibilities when hybrid practices were embraced.Such practices are significant as they have the potential to envisage and assemble institutions differently.The notion of hybridity usefully draws attention to in-betweenness and transition, highlighting the complexities of cultural participation during the pandemic.It assists our thinking about how old and new things, or digital and physical things, are blended or enmeshed, but also about how organisations, platforms and publics interact.It can help interrogate and explain the complex interplay between the manyoften competingpurposes and values of museums and galleries.This examination is long overdue, as has been made clear in this article and in the final reflection of one of our interviewees: "What is a museum when it's not something you go and visit?What is it?Is it the people, the expertise, the collection?It's become something different that you're talking about".A more expansive embrace of a notion of hybridity, understood not as an either/or; but rather as a "not only, but also" (Chadwick, 2017, p. 4) may enable more creative and urgent answers to the question of what a museum is in the wake of the pandemic and, moving forward, what role digital practises in arts and culture should play in the post-pandemic recovery.

Figure 1 .
Figure 1.Total tone frequency for museums and galleries.

Figure 2 .
Figure 2. Total theme frequency in museums and galleries.

Figure 3 .
Figure 3. Tweet from Ashmolean museum encouraging users to be creative at home.Sample response tweet from a user.