Digitisation, digital interaction and social media: embedded barriers to democratic heritage

Abstract The democratisation of heritage through digital access is a well-documented aspiration. It has included innovative ways to manage interpretation, express heritage values, and create experiences through the ‘decoding’ of heritage. This decoding of heritage becomes democratised, more polyvocal than didactic exhibitions, and less dependent on experts. However, the decision of what ‘heritage’ is and what is commissioned for digitisation (the encoding) is not necessarily a part of this democratisation. This paper will consider how digitisation reinforces the Authorised Heritage Discourse through the lens of Stephen Lukes’ three (increasingly subtle) dimensions of power: conflict resolution, control of expression and shaping of preferences. All three dimensions have an impact on how public values are represented in heritage contexts, but the introduction of digitisation requires more resources, expertise and training within established professional discourse. Social media may have a positive impact on the first two dimensions, but can reinforce hegemony. Alternatives are subject to epistemic populism. The role of digitisation and social media in the democratisation of heritage needs to be better understood. Questions regarding the nature and process of digital interaction, in terms of whose heritage is accessible, affect the very issues of democratisation digitisation appears to promote.

a lens, this article will consider the disconnection between increased access and increased democracy as we explore the ways public values are shaped according to how heritage is defined, identifying the emphasis on consumption over appropriation.
More than a decade ago, Besser (1997, 118) argued that increased access to digital surrogates in collections would have 'a significant democratising effect' , especially if institutions allow them to be viewed outside their walls. Such positivist views of new media's democratising impact are not limited to Western scholars: Okoth Fred Mudhai, researching digital tools in Africa, extols their potential to act as instigators for social change in a way that 'moves towards democratisation' (as quoted in Wasserman 2011, 3). Specific to museums, Alison Griffiths explains the attraction of digitisation activities and new media in terms of their 'promise to democratise knowledge' (as quoted in Henning 2006, 302). Expectations of the democratising impact of social media platforms on cultural heritage have likewise been discussed previously (Missikoff 2006) in terms of heritage (Waterton 2010) and public archaeology (Richardson 2014). We will further explore these expectations, as well as the digitisation projects built upon them, here.
Although there is good reason to celebrate many aspects of the advent of digitisation and social media in heritage, not least for the opportunities they offer in terms of facilitating greater access to heritage collections and providing spaces for new voices to be heard, these developments themselves are not inherently democratic. In fact, we proceed to argue, certain digitisation activities can subtly reinforce non-democratic structures. While the consumption of heritage is indeed wider and more sophisticated than ever before, as is the ability for people to respond to it with their views and interpretations, the actual decision as to what heritage is, and the implications of some participatory acts on public values, has become increasingly challenging. Indeed, much of the attention given to democracy through digitisation has focussed on the ability to reach larger user numbers, rather than how the discourse itself is created and mediated.

Overt, covert and latent dimensions of power
We do not claim that organisations, institutions or individuals necessarily aim to marginalise forms of heritage or undermine the democratisation of heritage through digitisation activities and projects. However, the insidious, albeit unintended impact heritage projects, particularly museum programmes and exhibitions, can have in terms of enforcing gross power imbalances between people has been well documented by museologists and heritage practitioners (MacDonald 2006, 3), as well as postcolonial theorists (Mudimbe 1994, 60-69). Furthermore, as Turner (2015) argues, historical decisions made around documenting and cataloguing heritage collections continue privileging certain power-knowledge imbalances and affect what cultural heritage knowledge we now access. In this sense, digitisation of and digital interaction with heritage today cannot be seen as a 'neutral' activity divorced from the entangled power relations of our past (Pickover 2014); indeed we argue that it strengthens many of these elements.
One helpful way of viewing the reinforcement of such power relations between groups is through Lukes' (2005) three dimensions of power. He defines power in three (increasingly subtle) layers: (1) Conflict resolution (A dominates B). Observable (overt) conflict revealed by policy preference and conflict resolution, often where one party dominates or oppresses another (through coercion or force if necessary).
(2) Control of expression (B complies to A's authority). Covert conflict over issues and potential issues, where one party may advance over another through influence, authority or manipulation etc.
(3) Shaping of preferences (A has underlying influences over shared agenda). Decision-making and control over the political agenda dealing with latent conflict of subjective and real interests, through inducement and encouragement.
Conceived in 1974, Lukes examined the concept of power beyond conflict resolution and direct domination to consider more subtle approaches. This built on work such as Bachrach and Baratz (1962) who examined agenda-setting as a means of political control. Part of Lukes' approach to the third dimension was considering why people willingly act in ways that might not be in their interests. This indicated that the most effective form of power is one that avoids conflict, but takes precedence regarding objectively contested interests, such as the expression of cultural identity.
This has clear parallels with Smith's (2006) Authorised Heritage Discourse (AHD) which identifies a dominant narrative connected to national and elite values that works to exclude non-expert views about the nature and meaning of heritage through emphasis of aesthetics, monumentality and 'grand' narrative. Like Smith, Lukes identified the use of authority to exercise influence. However, in this context, there are relevant distinctions. In particular, Lukes notes that a conflict of interest is not necessary for authority to be reinforced through political agenda-setting: the subtle elements of power can shape preferences before they are even expressed in an accessible and seemingly democratic forum. For Lukes, the consequences of power are not necessarily intentional, nor are they consistently identified by conflict or deliberately elevating one party's set of interests over another. As he suggests, 'power is a capacity not the exercise of that capacity (it may never be, and never need be, exercised); and you can be powerful by satisfying and advancing others' interests' (Lukes 2005, 12).
The first edition (Lukes 1974) was not without criticism, demonstrating the difficulty of revealing non-observable phenomena. His second edition, however, provides a nuanced depiction of different forms of power that makes many heritage issues more legible. Part of this is the acknowledgement that everyone who gains or loses from an outcome is not necessarily dominant or dominated. In line with this, he recognises that agents have autonomy, instead of implying that there is no freedom of thought or expression. As such, it provides a useful framework to consider the extent to which digitisation processes can be considered democratic and empowering.
All three dimensions have an impact on how public values are represented in heritage contexts. Heritage Studies has presented many examples of decision-making on issues where there is an observable conflict of subjective interests, for example between local people and the state in Uganda where the former disputes the latter's decision that building mass grave memorials will have a positive impact on national healing (Giblin 2014, 512); such overt disputes have rightly been explored (Fan 2013;Aykan 2015). Contemporary international disputes, like discussion of conflict zone protection within UNESCO, highlight the ways in which more covert decision-making in various contexts can control or prevent on potential discussion of issues (Meskell 2015). However, there is also prior influence and latent control over selecting the decision-making agenda (not necessarily through the decisions themselves). It is this latent conflict of subjective interests that is the most subtle and is explored through Lukes' third dimension. While digital interaction can have a positive impact on Lukes' first two dimensions at a more grassroots level in terms of increasing accessibility to and intensifying exposure of 'bottom-up' heritage activities to, possibly, influence the benefit of digitising heritage is less clear at the third dimension.

Key principles: internet neutrality and democracy
Before discussing more specifically the relationship between digitisation and democratisation of heritage within Lukes' framework, we briefly outline key concepts of Internet Neutrality and Democracy, since these related notions are essential to our argument that an apparent increased accessibility to heritage is not synonymous with democratisation of heritage.
Theoretically, the Internet is designed so that all 'packets of information are treated equally' since it is computers that access, retransmit and process information in the network, rather than this being determined by infrastructure at the heart of the network (Croeser 2015, 103). Yet, recent judgements, such as the 2014 ruling that the US Federal Communications Commission can no longer insist that all Internet Service Providers (ISPs) treat traffic equally, threaten to undermine even this notion of networked Net neutrality (ibid.). Our interest for the purposes of this paper is, however, more in how embedded sociocultural and socioeconomic inequalities in the Internet directly affect heritage issues. Commonly debated within the framework of a 'digital divide' , these have been discussed elsewhere, but are worth highlighting to contextualise our discussion (Castells 2001;Van Dijk 2005;Ohemeng and Ofosu-Adarkwa 2014). Richardson (2014, 106) outlines a number of issues that upset the notion of the Internet's neutrality, noting that it can 'perpetuate social divides and generate and actively promote hierarchy and inequalities' , not least in social media. These include: the ability to write code and own server space, allowing only some to create, amend and adjust content at will; the potential for richer individuals and groups to acquire the greatest benefits and growth from stable and faster connections and hardware, augmented by their intellectual access and knowledge capital, such as training to master software and troubleshoot problems, fluency in navigation, unwritten codes of behaviour and safety, as well as an ability to communicate in written English, still the majority language of the Internet. The advent of new technology has altered traditional balances of power, but cannot be isolated from the cultural question of who is best equipped to act upon such changes. As Worcman (2002) points out, even if people are given tools so that they can access the Internet, this means nothing if they are not trained to use them or cannot see how they benefit themselves or their community. Furthermore, capitalising on the Internet's advantages often relies upon an infrastructure that can supply goods and services in 'real life' that are unequally distributed. Srinivasan (2013) draws particular attention to this global inequality in the physical make-up of the Internet, as well as cultural bias in the way data and documentation are organised. Additionally, there are technological factors at play, particularly in terms of how data are organised through search engine algorithms, that fundamentally affect how much and which data are available through the Internet. Those with greater digital literacy are, of course, in a better position to exploit these possibilities. Our argument is that digital literacy tends to be socially and culturally determined, meaning that the Internet cannot be an inherently neutral and democratic space for sharing knowledge and accessing heritage on equal terms.
Arguing this, however, requires us to simultaneously consider our notion of democracy, a complicated topic that eludes simple definition and manifests in different ways. Moreover, 'direct democracy' , incorporating majority votes and populism, may not be realistic, nor even desirable in many cases. As Rodéhn (2015, 95) demonstrates in her recent work, democracy deserves 'reexamination in a global context' since, particularly in the heritage field, it is too often explored from a one-sided perspective that reinforces existing Western power structures. While discussions pertaining to democratising heritage came to the fore in the 1950s and 1960s as a way of representing heritage more fairly and equally to wider society, the specific issue of broadening 'access' as way of achieving this only really emerged in this context during the 1990s; in later years, focus shifted towards the museum's social role, participation and social inclusion (Rodéhn 2015, 98). Notions of open access for all do not always sit easily with these other democratising themes that stress reconciling conflicting groups and engaging with marginalised communities. In fact, in the cases of many Indigenous communities, making all aspects of their heritage accessible online to everyone, regardless of age, gender, initiation or other status, is an anathema to their values (Hogsden and Poulter 2012;Ngata, Ngata-Gibson and Salmond 2012;Boast and Enote 2013). Certain projects, such as the Mukurtu Wumpurrarni-kari Archive (Mukurtu n.d.) and Plateau Peoples' Web Portal (WSU n.d.), do employ Content Management Systems (CMS) that restrict access to culturally sensitive digital information according to Indigenous community protocols. Subsequently, some information on the Mukurtu Wumpurrarni-kari Archive database is only available to women users which seemingly respects Warumungu community protocols around accessing, sharing and creating cultural material and knowledge. Yet, Brown and Nicholas (2012, 311) suggest that it is 'only a matter of time' until display restrictions placed on culturally-sensitive collections by source communities are criticised for being likewise 'undemocratic' .
This issue of open versus restricted access cannot be resolved through technological solutions alone. While nuanced CMS do make restricted access possible, there is not yet general consensus that this is always the most democratic approach to facilitating Indigenous community goals around recovering cultural heritage. Indeed, Morphy's (2015) work with the Yolngu community in Australia's Northern Territory leads him to conclude that the easiest way to facilitate the digital return of visual heritage to this Indigenous community is precisely through open access to digital collections. Moreover, he argues that too often those assumptions on which restricted CMS are organised, such as the perceived significance secrecy plays in creating hierarchies of knowledge or that knowledge is separated by gender, undermine the diversity within Indigenous societies and, worryingly, create an illusion of equality between Indigenous and non-Indigenous communities that does not exist.
Taking a step back to consider the points and scale at which democratic principles are applied in heritage, particularly in terms of public values, is certainly a worthwhile exercise. In terms of understanding the inherent challenges, we suggest exploring democracy in practice more widely so that we can better see the implications that politics have on this matter. As Hague and Harrop's (2013, 42) uncontroversial overview acknowledges, there has been a rise in democratic regimes worldwide but 'despite the spread of democracy, democratic regimes are becoming thinner with the rise of expert and regulatory authority … It is in professional regulation, rather than in authoritarian ideology, that we find the contemporary challenge to democracy' . In these terms, is it not just the process of creating 'heritage' that can seem undemocratic, but the very processes of increasing accessibility that comes with a range of regulations, skills and resources that can implicitly undermine the notion of democratised heritage.

Heritage production and appropriation
Charting different kinds of digital activity (digitisation, digital interaction) against a simple 'communicative process' model of heritage where meaning is transmitted and received ( Figure 1) allows us to consider latent power (Lukes' third dimension) in terms of heritage. The model is deliberately simple to avoid esoteric and context-specific matters; however, this is instructive in terms of considering how democracy and digitisation are related because it focuses on the point of interaction, rather than the extent of follow-on dialogue. It is also intended to connect with Hall's influential work 'Encoding/ decoding ' (1980) that aligned the study of culture with communication. By considering different kinds of interaction through the point at which users are engaged, one can see how agendas are set.

Stages of interaction, points of contact
Digital interaction has included innovative ways to manage interpretation, express heritage values, and create experiences through the 'decoding' of heritage in museum galleries and beyond. This intensification of heritage communication makes the process of heritage interpretation more democratised, more polyvocal than didactic exhibitions, and therefore less dependent on expertise and institutional structure. However, the decision as to what 'heritage' is, and what is commissioned for digitisation (the encoding), is not necessarily a part of this democratisation. For example, Jones (2014) describes a number of installations that create new forms of engagement, but the engagements required 'a complex of software and algorithmic design, spatial acoustic arrangements and customised hardware engineering ' (2014, 198) that created a novel experience whilst intensifying the strength of the official narrative. Ciolfi (2013) discusses the 'co-creation of heritage' (2013, 72) through digital interaction, which is defined as responding to displays or contributing factual information to the existing narrative.
Substantial interaction can take place at many stages, and dialogue response is often possible. Consultation is frequently laudable and well-intended. However, the issue here is that much of the interaction is purely responsive. These interactions sit at the end of the communication spectrum (Figure 1), but the most profound issues of appropriation and agenda-setting (ideas of what and whose heritage) have already been made. Responding to interpretations and identifying different narratives is valuable, but making this more engaging or elaborate can actually take the public emphasis away from deciding what should be considered worthy of debate or interpretation in the first place. It is not the extent of interaction, but the kind of interaction that is problematic here. At the 'encoding' level, the introduction of digitisation and digital interaction requires more expertise and training within established professional discourse, and more resources, thus displacing values at a more profound and subtle level. Questions regarding the nature and process of digital interaction, in terms of whose heritage is accessible, affect the very issues of democratisation that digitisation appears to promote (without consciously undermining them).
Intensifying the forms and accessibility of engagement supports heritage by designation at the expense of appropriation. The benefit of participation has been questioned, especially when there is a marketing or dissemination-driven result towards which participants are being directed (Irvin and Stansbury 2004). Moreover, until recently cultural heritage digital initiatives have often been overtly driven by technological possibility, rather than societal need. Jones '(2014, 192) investigation into the growing use of augmented reality applications at cultural heritage sites, a situation he argues is driven more by the business sector than consumers, suggests that technological possibility may continue to be the dominant influence in shaping how we engage digitally with heritage in future. Such well-meaning, sophisticated and far-reaching activities can consequently risk exacerbating problems in terms of how heritage can be defined and appropriated by communities.

Reinforcement
The landscape of digitisation and digital interaction shows that the problem of democratisation is embedded in practices at many levels. This underlines the fact that it is not an intended fault, but the emergence of a cluster of factors that influence digital heritage. Connecting to a default position avoids introducing new problems, but reinforces existing ones. It cannot be assumed that this default has always been specifically considered, or even questioned. Indeed, there are several points at which the problem is exacerbated on a broad, implicit level and, transferred to practice, all of these can propagate and intensify the problems discussed. These points will loosely guide some examples of initiatives connected to digitisation and digital interaction. We have deliberately avoided choosing 'bad' examples, as it is not our intention to criticise individual initiatives, but instead to examine a backdrop to the problem at hand. These inter-related factors are: • Research: The direction technology takes has a huge influence on which aspects of digital heritage are researched. The eventual use and implementation of digitised heritage is not always a research priority. Focus on communicating and engaging with existing forms and practices leaves less room for appropriation. • Commissioning: Who funds digitisation and digital interaction activities has a big influence on what is done and whose values are represented. This includes commissioning research as well as specific projects. This frames many of the possibilities for digital interaction, and defines how content flows through any interaction. • Resources: The time and equipment required for many projects requires institutional expertise, teamwork and a solid infrastructure. Many activities require at least one dedicated professional with updated knowledge.
• Training: Expertise required to deliver digital content may vary a great deal, but the advancing of pre-created content out-strips the accessibility to content generation. Despite dynamic platforms for content production, many exciting engagement processes simply facilitate more dynamic consumption. • Choices: The problems above are not insurmountable, and there are efforts to reduce them, but all require some awareness of the political implications. Deferring to the implicit bias of a default situation does not mean that a dominant narrative has been avoided, only that one is not aware of those implications (Lukes 2005). As we have argued above, digitisation mandates to make heritage 'accessible for all' largely means reinforcing existing biases (e.g. an emphasis on material and visual heritage).

Dimensions of power in practice
In complex situations, there is not always a clear distinction between Lukes' second and third dimensions of power; this difference tends to be more obvious if examining specific events. Both dimensions deal with the notion of 'authority' , distinguishing between observable and latent conflict of interests.
In terms of digitising heritage, numerous apps generated by heritage institutions promote a way of identifying and interacting with heritage, but could not necessarily be considered exclusive or undesirable, as per the second dimension. However, the expense required to produce these apps, the bespoke development required for many platforms, and the consequent expertise they entail (Jeater 2012) all contribute to the hegemony. Furthermore, and pertinent to the third dimension, this can intensify the passive, receptive element of the heritage process for the user.
Other initiatives directly reinforcing the AHD can be seen when digital interaction has been deployed to help form ideas about heritage and nationality. For example, the ICONS of England project, where members of the public can nominate and vote for symbols of England's national heritage, is envisaged as an online collection of the country's most cherished cultural treasures (Mason and Baveystock 2009). Although nominations are possible, the concept of iconic heritage representing national unity was already predetermined. Furthermore, celebrities and institutional partners, such as the National Trust and English Heritage, could comment before a 'top tier' of nominations were decided by an Advisory Board. The 'approved' icons contained information about visits and history (Mason and Baveystock 2009). Despite 8000 suggestions, only 1165 were listed and of these, just 100 were approved, demonstrating the potential diversity but the restrictive discourse in the most profound stages of concept and selection, with a confusing and dissonant middle. The nature of the medium, which couldn't connect the fragmented values expressed during bottom-up phases of the project, was a significant factor affecting this (Mason and Baveystock 2009). The power to determine what is to be decided (or not decided) demonstrates significant power over the outcome, not by outright dominance but through subtle agenda control, which connects with Lukes' second dimension.

Commissioning research
The direction digital heritage research takes is often determined at high strategic levels. For example, the 7th Framework Programme 'Learning and Access to Cultural Resources' states that its 'aim [was] to exploit Europe's vast and exclusive cultural resources and learning traditions as a source of innovation and creativity for businesses, researchers, educational organisations and the general public' (European Commission 2013, 91). High-level investment in development here focussed on top-down approaches for educating a public about existing values and heritage.
The recent Reflective Call more openly acknowledges this facet of digitisation, noting in particular that producing 3D models, perhaps a pinnacle of digital achievement, is time-consuming and expensive 'not least because the modelling is carried out for individual objects rather than for entire collections' (European Commission 2014b, 90). The expertise required to deliver this kind of work, and the need for the profound decision-making of what is to be digitised, further widens the gap between those who have resources and those without. The aim for European digitisation of heritage is to be a representation of heritage 'forever, for everybody, from everywhere' (European Commission 2014a, 40), which ultimately reinforces the power associated with existing collections and narratives.
Top-down approaches become increasingly problematic in cases where digitisation projects are commissioned in the supposed interests of marginalised communities, but are largely funded and directed by the institution or government. Too often, these collaborative digitisation projects between institutions and source communities overlook the ways projects might realise local goals that support learning and engagement with cultural heritage; instead, they simply 'broadcast collections' and the accompanying catalogue information already held by the institution (Srinivasan et al. 2009, 162). As Boast makes clear, even digitisation projects that incorporate these marginalised voices in digital spaces frequently fail to move contributed comments from the online periphery into the central institution catalogues where knowledge is produced and defended (Boast, Bravo, and Srinivasan 2007, 401;Boast and Enote 2013) since this is not pre-defined as an intended research outcome.
Project funding has also gone to crowd-sourcing, which explicitly induces people to engage with heritage. However, this broad and accessible engagement is often intended to add value to future research projects (Blaser 2014), scholarly references (Causer and Terras 2014), and government records (Leon 2014). These projects rely on participation, and provide enjoyable or interesting experiences, but they simultaneously reinforce the latent power of Lukes' third dimension since agendas are set for the benefit of specific groups and preferences shaped before engagement by the crowd begins. Smith (2009) points out that encouraging social inclusion within heritage organisations can engage the public with authorised versions of historical and archaeological narratives that reflect the interests of expert-driven principles. Yet, intensifying this activity with rapidly emerging technology holds the danger of only making this problem more pervasive and more significant. Attempts to encourage a sense of place often ignore existing community relationships and attempts to be inclusive, such as the UK's Big Society, are often top-down in practice (Waterton 2005).

Impact of increasing accessibility
The British Museum has been very proactive in digitising collections and engaging through digital interaction. One aspect of this is the recent release of files for printing museum objects in 3D (Vincent 2014). The downloadable collection provides worldwide access, which is commendable, but places the user firmly in the position of passive recipient of a complex and resource-heavy (albeit impressive) process. These dynamics already exist in museums but are intensified by the focus on users receiving a small collection already pre-selected by the museum. As well as reinforcing producer-consumer roles, the bias towards visual culture encourages reading heritage as 'a narrative of identity, politics and power' (Watson and Waterton 2010, 84).
In the case of the British Museum, online focus has been community building and direct interaction, with the museum accepting social media as 'a mechanism for broadcasting cultural propaganda, to influence target audiences and create new relationships' (Pett 2012, 83). Consequently, the emphasis is on engagement with content or on community sharing of content to help create social capital (Pett 2012, 83). Pett argues that 'the centrepiece of the BM's website is the Museum's collection, with an emphasis on making it more accessible to the public, documenting its condition and presenting the ongoing research connected with it' (2012,87). While this is not unwelcome, we might question whether this truly fosters democracy. Certainly the benefits of moving away from one-way didacticism towards a culture of dialogue are evident, but the mode itself is profoundly responsive. Although commercially driven, there is no overt intent to 'dominate' and encourage relevance or participation where it didn't exist before. Instead, the museum enjoys the resources to facilitate and encourage institutionally desirable activities that reinforce existing relations, although perhaps at the expense of making heritage that is not online, event driven, or marketable comparably less accessible or palatable. As Lukes states, the domination of defenders of the status quo may be so secure and pervasive that they are unaware of any potential challengers to their position and thus any alternatives to the existing political process whose bias they work to maintain. (2005,25) It is not a matter of one set of values trumping another, but a shaping of preference and appropriation. Additionally, Facebook discussions for the British Museum are regulated for commentary deemed offensive or 'off-topic' , partly due to areas of contention that some objects elicit (Pett 2012). Consequently, the possibility of democratic exchange is further undermined; the matter is not so much about uneven conflict, but that some items do not really appear on the agenda.

Self-authored narratives
There are a number of developments within social media that have provided space for a larger number of voices, appearing to offer 'a more open and democratic attitude to heritage' (Economou 2016, 222). The potential for this is extraordinary in terms of democratising and appropriating heritage. However, the practice does not necessarily translate into deconstructing existing hegemony. Social media may have a positive impact on the first two dimensions of power, or the further end of the spectrum in Figure 1, but subsequently reinforce hegemony in the third dimension, or make engagement with the creation of heritage a more remote possibility.
In high profile projects such as Wiki loves Monuments, for example, the public can upload their own images of heritage sites. The project itself is 'led and driven by a global network of volunteers' and 'open to absolutely everyone, and participation is completely free' (Wikimedia 2014). However, Wikipedia's convention to avoid original research (Wikipedia 2014) meant that the appropriate kinds of 'heritage' are defined by top-down lists (e.g. government designation). Wikimedia (2013) required that images 'contain an identified monument' thus creating a specific, authorised group of sites to which the public could respond. Since everyone is invited to interact with a specific, monumental form of heritage, exciting initiatives for user-generated content can still demonstrate problems through the lens of the third dimension of power. By deferring from this, and creating something open, flat and participatory, the public values are expressed within a pre-defined framework of relevant heritage, thus subtly reinforcing the AHD rather than democratising heritage. Again, there is no coercion or outright directive given by an institutional authority and in many cases participants are actively compliant. It does, however, provide a means of shaping preferences by reinforcing the value-laden status quo in terms of Lukes' third dimension (2005). Even bottom-up counter-narrative initiatives, such as alternative, irreverent guides to museums of art (Thomas 2015), involve a pre-selected form of heritage whose significance is propagated by intensifying its exposure above other forms of heritage. It is a difficult situation where many alternatives are subject to epistemic populism. As such, well-meaning disengagement from the political consequences means such volunteer-led initiatives can reinforce existing problems.

Grassroots projects
There are, however, encouraging examples of digitisation projects that adopt a more nuanced bottom-up approach to digital heritage. The Zuni A:shiwi A:Wan Museum and Heritage Centre (AAMHC) in New Mexico, USA, is one case in point. Originally conceived as a digitisation project that might enrich Cambridge University's Museum of Archaeology and Anthropology (MoAA) database of Kechiba:wa items with contemporary Zuni views, the active involvement of the Zuni community meant the project developed in such a way that the digital database has been constructed according to Zuni principles and is housed within the Zuni pueblo (Isaac 2009;Boast and Enote 2013). The power to share and add knowledge to the database rests with the Zuni community, suggesting grassroots democratisation of heritage is really taking place. However, restrictions placed on access to certain items in the collection does run counter to national guidelines guaranteeing equal access to heritage which again illustrates the tension Brown and Nicholas (2012) flag between access and democracy.
As well as the Zuni AAMHC digitisation project, the Tribal Peace Project, also in the USA (Srinivasan 2007), the Museum of Anthropology -University of British Columbia's Reciprocal Research Network, co-developed with the Musqueam Indian Band, the Stó:lō Nation/Tribal Council, the U'mista Cultural Society in Canada (Rowley 2010;Gibson and Turner 2012), and the Ara Irititja digital database in Australia (Christen 2006;Hughes and Dallwitz 2007) are other projects where experts and institutions facilitate or respond to community heritage. This is achieved by building accessible databases that are constructed around cultural meaning rather than operating systems (Srinivasan et al. 2009;Srinivasan 2013) and facilitating social history groups in constructing their narratives around their local environment (Giannachi et al. 2014). Consequently, the same actors are able to find new ways for values to be expressed and communicated, and for the public to appropriate and communicate their heritage values. While even these projects that focus on community developed content are still subject to criticism, further exploration of the issues and areas with which they engage would bring to light the impact digitisation has on 'democratising' heritage and how heritage might be 'better' represented.

Conclusion
The aim of this article is certainly not to tell people, or heritage institutions, what they should do. Instead we suggest exploring a diversity of approaches that incorporate alternative forms of participation which bring different parties' values into focus. There are already very exciting activities in digital heritage which could scarcely have been imagined in the recent past. These have undoubtedly allowed more people to consume heritage in new ways and on new terms. Mobilising people to engage with heritage is valuable and interesting, but the role of digitisation and social media in the democratisation of heritage needs to be better understood. Understanding the implications, particularly unintended, that come from digitising or digital engagement becomes increasingly important.
Examining this through the lens of power dimensions illustrates that even when some groups' interests are being met through dissemination or access, these welcome actions may implicitly promote or reinforce more fundamental interests of other parties. As Lukes' (2005) framework shows, impressive and satisfying polyvocal experiences can in fact reinforce subtle or latent forms of power. Space is provided for other voices in a seemingly democratic fashion, but this only occurs within a pre-selected frame of reference that consequently promotes a more limited range of interests. Given that rules and regulations are not neutral selectors of outcomes, in many cases all that is required for Lukes' third dimension is the continuation of a pattern of behaviour that brings about results desirable to the powerful or privileged (Dowding 2006). Consumption of and participation in heritage through response do not permit the most fundamental aspects of heritage: creation and appropriation. In this case, what is not discussed can be more important than those decisions and interpretations in which people can participate. As such, simply expanding participation is not enough; we need to think more about the kinds of participation.
This leads us to question whether heritage institutions can move from their position as authoritative expert towards empowering public values at this subtle level. Museums do invite participation but, as discussed above, the collections have often been pre-selected and the objects chosen so that the public responds to these, rather than creating a situation whereby people are defining their heritage according to their own values. However, there are points in the process where participation could increase and arguably there is need to make greater room for this at the beginning of the heritage process. Indeed, Worcman (2002) suggests that the creation of the digital collection itself might not be as important as the group's 'engagement in the process' . Can the public assert or express their identities and values in a way in which museums can respond or engage? Moreover, is democratisation through open access always a definite goal for cultural heritage? As Buchanan points out, we increasingly find that there is no universal model of access, dissemination, control and content construction that can support and 'respect cultural specificity, subjectivity, and values ' (1999, 200).