Breaking through the language barrier – bringing ‘dead’ languages to life through sensory and narrative engagement

ABSTRACT Ancient inscriptions can be difficult to understand and off-putting to museum audiences, but they are packed with personal stories and vivid information about the people who made them. This article argues that overcoming the language barrier presented by these objects can offer a deep sense of engagement with the ancient world and explores possible ways of achieving this. It looks at examples of effective approaches from a range of European museums with a particular emphasis on bringing out the sensory, social, and narrative dimensions of these objects. It argues that inscriptions can change the way that museum visitors view the ancient world and empower them to interpret the past for themselves in new and creative ways.


Introduction 1
This article is about exploring museum objects with writing on them. Inscriptions are a specialist topic and under-recognised in museums, but they carry lessons for other kinds of engagement with more complex objects. When we embarked on the Ashmolean Latin Inscriptions project, on which this article is based, the previous major research engagement with the Ashmolean's collection of Latin inscriptions had been a catalogue written in 1763 (Chandler) and there had been little attempt to make them accessible or appealing to museum visitors. This reflects a wider pattern in which inscriptions are seen as inherently difficult and mostly of interest to experts. Our work challenges this belief. This article begins by exploring some theoretical background on the potential benefits of using objects with writing on them (and particularly writing in 'dead languages') and looking at why such objects prove challenging in a museum space. It then collects best practice from across a range of European museums to show practical approaches to the problems of inscriptions and reflects on why they can be effective in empowering museum visitors to make sense of the Roman past. This article will be relevant for archaeological and historical museums that struggle to connect their audiences with the complex information conveyed by the writing on their objects. It also serves as a broader meditation on the challenges and benefits of deep engagement with museum objects. Nearly all museums house some objects that seem difficult or alienating at first glance, but which can be used to foster deep connection with their subject matter and create a sense of achievement for visitors. This article uses Latin inscriptions as a case study of objects that are information-dense, and difficult for many people to understand without effective interpretation. In doing so, it makes the case for deep engagement with objects that are not instantly appealing, as a way to provide more rewarding experiences for visitors, and suggests ways in which such objects can be used to stimulate creativity and imagination. Its case studies will be relevant to any museum interested in helping a wider public to make deeper connections with difficult objects.

Inscriptions: difficult, but vivid
The word inscription can refer either to the text written onto an object or, by extension, to the whole object including the writing. This can include a wide variety of types of museum object, ranging from massive public documents carved in stone to private notes and names scratched onto personal possessions. It is the writing that turns these objects into inscriptions and this article focuses on the particular challenges and opportunities that this writing presents for museums. However, writing and object are inseparable, and we argue for approaches that recognise both aspects, since our understanding of one can deepen the other. There is a longstanding dichotomy of objects and words in our culture, which can be traced back to the formation of disciplines according to the materials studied in the nineteenth century. This approach tends to treat objects as secondary to texts for understanding history, in the (inaccurate) belief that objects are best at explaining the material conditions of life, whereas complex ideas reside in texts (Moreland 2001, 1-32). Museums challenge this hierarchy by foregrounding objects and objectbased modes of understanding the past. Inscriptions are a clear demonstration that both aspects of the historical record are produced by the same people and processes, often to the same ends. However, inscriptions can present a challenge for institutions that focus on material culture.
As objects that combine two kinds of evidence, inscriptions offer an opportunity to uncover a rich version of the past, including words as well as objects, and evoking the people who wrote and used them. We use the term 'vivid' to describe this version of the past because of its 'lifelike' connotations. No object in a museum speaks for itself (Crew and Sims 1991) and this article will show that there are features of inscriptions that make this especially true for them. However, inscriptions are a communicative medium: they give us (in a mediated form) the real words of ancient people. Because these are words written by ancient people, they give the impression of direct contact with the past and feature details (such as names and explicit accounts of events) that can be preserved only through text. Inscriptions often tell us about named individuals from the past who are not famous historical figures and they provide personal details about those people's lives in ways that emphasise the individuality of people who lived in the past. Inscriptions are objects that lend themselves particularly well to narrative approaches to the past (something we shall explore in detail in the final section of this article), since many were intentionally created to share information about people's achievements, beliefs, and relationships. In the museum context, they can give the effect of direct communication by individuals in the past. While this is only an effect (hence we describe it as 'vivid' not 'alive') it is a powerful one.
One such individual, who demonstrates the scope for inscriptions to be vivid evocations of the past is Regina. She is depicted on a slab of buff sandstone (Figure 1), found in South Shields (Roman Arbeia) (https://romaninscriptionsofbritain.org/inscriptions/1065). This stone features a carved relief portrait of a woman (most recently analysed by Carroll 2012) with two inscriptions underneath (Cooley 2012, 304-306;Mullen 2012, 1-5, 32-35). The first inscription, in Latin -'To the departed spirits of Regina, his freedwoman and wife, a Catuvellaunian, aged 30. Barates of Palmyra'tells us that this is the gravestone of a Celtic woman, native to Britain, who had been freed from slavery and married by a migrant from Syria, Barates from Palmyra. The second inscription is a more personalseeming lament in Barates' first language, Palmyrene Aramaic: 'Regina, the freedwoman of Barate, alas'.
This tombstone is one of the most popular objects at Arbeia Museum. It tells a personal story of travel and loss, and represents the diversity of Roman Britain. Inscriptions like this offer tangible evidence of people from all over the Roman Empire who made lives in Britain. Regina is a woman and a former slave, from the margins of the Roman Empireprecisely the sort of person under-represented in the Romans' own histories of themselves as well as in many modern views of this period. As such, she offers scope to challenge preconceptions and to practise the kind of inclusion by representation advocated by Sandell (2002). Archaeologists describe their discipline as being all about people, but individual, personal narratives rarely survive from thousands of years ago (Swain 2007). Inscriptions offer a rare opportunity to connect to ordinary lives. While this object offers an unusually vivid glimpse of the ups and downs of the life of one of the inhabitants of Roman Britain and gives the impression of direct engagement with the words of ancient people, it is not a neutral or unproblematic window on the past. We still depend upon the lens of the husband setting up the memorial for our picture of Regina herself. The words that tell us her name and place in the world cannot be read directly by most people, whilst the Palmyrene Aramaic script, with its more personal message, is even less accessible than Latin. As a result, the vividness of this object depends as much on the interpretation provided by the museum as any intrinsic quality of the writing or image. There is no shortage of creative interpretation to help people understand the story of Regina. Visitors to Arbeia Museum can hear Regina speak at the push of a button and participate in hands-on education activities featuring a replica in the museum's 'Timequest' gallery.' Regina's tombstone has been featured in the BBC series 'A History of the World in 100 Objects' and stars in its own mobile app: The Mystery of Regina's Tombstone, designed and produced by Braunarts for the Roman Society [http://www.braunarts.com/regina/mystery_of_Regina_app.html]. However, for every well-known success story like this, there are many inscriptions still seen as difficult, unappealing, and alienating.
Inscriptions are notorious for being difficult to understand and have been described as an 'instant turn-off' to many museum visitors (Clarke and Hunter 2000). The reasons why inscriptions are off-putting may seem obvious. They are often grey, flat objects with little to catch the eye. We are used to ignoring inscriptions in our own environment as part of the background noise of official architecture. At first glance, it is difficult to tell whether there is anything worth bothering with about the object, and in the atmosphere of most museum displays there are many other objects that will offer more instant recognition or awaken the visitor's curiosity more readily. However, by looking more closely at why these are such difficult objects, we can understand some of the factors that present challenges for museum interpretation and begin to explore possible solutions. At their core, the problems of displaying inscriptions in museums (such as the language barrier, the fact that their appeal is not instantly visible and the need for additional background information to make sense of them) are problems of communication.
Hooper-Greenhill (1994) argued for a model of communication in museums that reflected visitors' role as active participants in making meanings. Much has been done since to demonstrate the truth of this and explore the ways in which visitors make meaning in a museum space. Chronis (2012) has shown how visitors use objects imaginatively to fill in gaps in their understanding of a display and build a sense of personal connection with the museum's subject matter. Objects provide a tangible connection to the past that is particularly fruitful for this model of meaning-making, allowing museum visitors to fill in gaps in their existing narratives of the past, relate it to their own life experiences and issues of current concern, and engage in imaginative reconstruction (Chronis 2012). Schorch (2013) has explored how both the spatial structure of the museum and the content of the displays are brought together by visitors as a process of narrative meaning making. This phenomenon can also be seen in Witcomb's (2003) discussion of how the most effective forms of interactivity in a museum are not technical, but narrative-based, offering to visitors vignettes that they can use to build their own story, rather than a closed, authoritative grand narrative. Museums need to find ways to use inscriptions that expand their narrative potential by giving visitors the tools to understand them in a manner that opens up interpretive possibilities, co-producing meaning with their visitors.
If museum communication is a dialogue, Latin inscriptions can give the unfortunate impression of having a lot to say, but not wanting to talk. A Latin inscription is clearly rich with meaning, but that meaning is obscure to most people and cannot inspire the sort of direct, imaginative responses that build engagement and provide satisfying visitor experiences (Chronis 2012). There are still conclusions that visitors can draw directly from an uninterpreted inscription (e.g., that the Romans were good at stone carving) but much of what gives these objects their narrative potential is inaccessible without further interpretation. In this article we describe inscriptions as 'difficult' objects, to explore the extra work that is needed on the part of both museum and visitor to get the most out of them.
The word 'difficult' tends to be used in museum terminology to describe objects and histories that are violent or otherwise evoke painful emotions (e.g., Witcomb 2013). Bonnell and Simon (2007) describe how such difficulty contains both affective and cognitive elements, challenging visitors' ability to understand the material and their emotional capacity to process it. Our definition of difficulty owes a lot to the cognitive half of this distinction, because it focuses on the amount of mental work required for engagement, regardless of whether the histories involved are emotionally demanding. But it is important to note that cognition and emotion cannot really be separated. Smith and Campbell (2015) have demonstrated that cognitive and emotional dimensions are entangled and shape the ways in which museum visitors imagine or remember the past, and argue that museums need to take emotion seriously as something that is socially constituted and has political effects. While inscriptions can evoke a full range of emotions (examples later in this article will include humour and empathy), the particular emotional impact of the difficulty of inscriptions is the feeling of being excluded from a museum as a result of not understanding the objects. Bourdieu and Darbel highlighted the alienation of not having the cultural and educational background to understand museum displays in their influential study The Love of Art (1969; 1991 translation consulted for this article) and these ideas continue to be used to understand the ways in which visitors experience museums (Schorch 2013) or why visitors might feel unwilling to return to them (Falk and Dierking 2012, 208). Inclusion is a major aim for the contemporary museum; displaying objects that might prove alienating carries serious risks for visitor participation (Sandell 2002).
In the case of Latin inscriptions, this problem is compounded by the role which the Latin language has played in the past as a deliberately exclusive feature of elite education (Stray 1998). Many classicists today push back against such exclusive models of engagement, working to acknowledge the role of these languages in delineating class boundaries but taking practical steps to open up access to ancient texts (e.g., the Classics and Class research project: http://www.classicsandclass.info). Because of the potential for Latin inscriptions to alienate visitors, the stakes of displaying them are particularly high. Finding ways to tackle the cognitive challenge of such difficult objects in ways that harness the vibrancy of ancient cultures is vital to ensure they do not do more harm than good with museum users.
At the heart of the difficulty of Latin inscriptions is a language barrier that makes them inaccessible to most people, but they are also difficult to read even for people who do have the language skills, since they are often full of abbreviations and specialist language, and have damaged or missing sections. The first section of this article explores why simply providing a translation is not enough to overcome the language barrier of Latin inscriptions and considers some case studies that show how museums can promote more active models of engagement with inscriptions in spite of the fact that translation (often a necessary starting point for working with these objects) traditionally seems to depend on a transmission model of communication in which an expert gives information to an audience.
While the difficulties of reading inscriptions are the most obvious challenge they pose for museums, their textual nature also has implications for the kinds of meanings that visitors can derive from them and the approaches that museums should take in interpreting them. The final section looks at how the particular kinds of narratives that inscriptions contain can best be used by museums.

Overcoming unreadability
One of the simplest and most common approaches to overcoming the unreadability of inscriptions is to provide a full translation on a nearby label. But this redirects attention from the object to supporting text in ways that undermine the direct, artefactual knowledge that visitors value in museums (Chronis 2015). This section looks at some of the challenges of interpretive text as it relates to museum displays of inscriptions and considers how museums can develop interpretation that facilitates understanding of these objects without undermining the visitor's ability to engage directly and actively with the object.
The public spaces of museums have been shown to be challenging places to read in, despite the important role that interpretive text plays in them. Ekarv (1994) described how museum texts can struggle to compete for visitor attention in a space that presents physical challenges (such as visitor fatigue or poor lighting), and devised compensatory guidelines for making text easier to read. Text on objects is different from the interpretive text added by the museum because it is largely outside the control of the museum, but it still places cognitive demands on its audience. Usually these demands are so great that the museum must add further explanatory text, which (if not carefully structured and composed) can compound the problem.
Translations present challenges for museum label strategies. It can be difficult to make such explanatory text comply with other guidelines imposed by museums' communication strategies, such as maximum word count or reading age (for examples of such guidelines, see Ekarv 1994;Gascoigne 2007). However, a bigger concern for museums that present translations of inscriptions is that they entail transmission of information from an 'expert' to the publicthus undermining the active models of engagement considered most effective in a museum context. The act of reading a translation need not be a passive one. Barthes (1971;1986 translation consulted for this article) compares the act of reading to playing music, and distinguishes between active co-production (using a musical instrument to play a score and adding one's own interpretation to the work of the composer) and more passive consumption (hearing someone else play or listening to a record). He notes that the latter mode of consumption (both for reading and for music) engenders boredom. This section of the article asks what museums can do to promote more active reading when dealing with textual objects.
A useful example of how to integrate translation into a broader communication strategy that does not interrupt visitor engagement with objects or shut down active meaningmaking can be found at the Epigraphic Museum in Rome, located within the ruins of Diocletian's Baths. In the recent refurbishment of its galleries, the whole character of the collection has been redesigned to promote user engagement with the 'Written communication of the Romans'. It focuses on the key questions of 'Who writes? What is written? How is it written?' While the museum uses a variety of communication techniques, labels and other interpretive text form the core of its communication strategy. These texts avoid taking visitor interest for granted and work to promote active engagement with the objects on display. Labels are designed to capture attention and reduce the cognitive effort involved in reading. Figure 2 shows how even simple topic statements and explanations (in this case 'A man in love: dedication on a perfumed ointment vase') can relate an unfamiliar object to an easily recognisable story (personalising a gift to a loved one). This phrase (presented in both English and Italian) does a lot of work: explaining the object, evoking the story behind its inscription (since it is presented alongside varied examples of inscriptions) standing for one possible motive for writing on an object. By not over-explaining the object, visitors are free to relate it back to their own experiences (for example, of writing messages to go with gifts) and integrate it in a larger sense of the ancient world. One simple way to promote active reading is to reduce the amount of effort spent in engaging with the practical aspects of reading, freeing up visitors' attention for more active meaning making. Relating text directly to objects reduces the work of identifying which is the relevant information, such as using numbers to locate labels elsewhere in the case. A direct relationship between label and object is also particularly valuable for inscriptions, since it can help visitors find and read the writing on the objects themselves. With this in mind, several displays at the Epigraphic Museum incorporate words into object mounts (e.g., Figure 3). This highlights the special relationship between these words and the materiality of the objects in question, while still giving visitors the visual assistance needed to process them. Interpretive text echoes the text on the objects in ways that can highlight the beauty of lettering, intuitively fill in gaps, or clarify letters that are difficult to read. Relating objects and explanatory text more closely reduces the friction between object and visitor, even if the difficulty of Latin inscriptions means that this encounter still needs to be mediated.
Accessible interpretation for Latin inscriptions will sometimes mean being satisfied not to explain and translate the whole inscription, something that can be difficult for subject specialists. Museums can only pick one or two facts to convey about a single object and any decision about where and how to display an object opens up some possible readings of it while it closes down others (Clarke and Hunter 2000). While interpretation that does not give access to the whole inscription might be seen as less true to the ancient object, paradoxically it can offer a form of authenticity by reproducing what must have been a common experience in the ancient worldthe inability to read a text. Even in their original contexts, the texts of Latin inscriptions would not have been accessible to all, given limited levels of literacy at the time (Harris 1991). Instead, inscriptions originally derived meaning from their physical materiality as much as from their textual content (Corbier 2006) and there is much to be said for museums acknowledging the different levels of literacy in the ancient world and exploring how inscriptions might make an impression through factors like scale, colour, and imagery.
More comprehensive information can always be provided for visitors who want it: in the Ashmolean, we have found that many visitors want to be able to access full translations and transcriptions, but we have chosen to put these online, along with more detailed commentary (http://latininscriptions.ashmus.ox.ac.uk). While the primary audience for such detailed epigraphic publications will always be a self-selected group, making the edition freely available online and paying careful attention to audience needs can vastly expand their user base (Löser 2014). Technology can also support deeper exploration of inscriptions in the museum space. At Rome's Epigraphic Museum, visitors can choose to explore longer and more complicated (but therefore also more interesting) inscriptions using the interactive computer screens. Different levels of interpretive text that give visitors the basic information to navigate the concepts on display and lets them decide whether to go deeper is a standard feature of museum interpretation (Ravelli 2006, 33-6) but can be particularly valuable for navigating the dense nature of Latin inscriptions and the fact that some of the most involved visitors want not only translation and explanation, but also transliteration to help them make sense of the writing on the object for themselves.
A playful approach to exploring the difficulties of reading Latin inscriptions that acknowledged the difficulties they present for visitors today was 'Antique.trad, lecture de stèles en latin', an event at the Musée Départemental Arles Antique for la Nuit des Musées 2017. This encouraged visitors to record themselves reading Latin inscriptions aloud, and then to compare their reading with other people'sraising questions over what the language really sounded like. Other recordings included Latin teachers, who were able to give a more authoritative reading, but the inclusion of recordings by several professional academics from France, Italy, Germany and the UK demonstrated the variety of possible 'correct' readings. The fact that visitors who do not speak Latin could read the inscriptions demonstrated the familiar elements of these inscriptions (not least the Latin alphabet still used by many languages today). Meanwhile, the variety of individual pronunciations offered an opportunity to reflect upon the flexibility and diversity of language, as well as hinting at the assumptions and guesswork involved in studying the past. Comparing with fellow visitors turned reading (something our culture tends to conceptualise as a private individual activity) into a social experience. While 'Antique.trad' focused on engaging people with the sound of the language, it is also possible, under the right conditions, for museum visitors to engage directly with the meaning of the language on objects.
Inscriptions are designed to be read. In this sense, they are one of the few museum objects that the visitor can still use for their intended purpose (by contrast, they are not usually allowed to light an ancient oil lamp or drink from an ancient cup). As a result, approaches to inscriptions that let visitors read directly from the object offer an unusual form of deep engagement. Many Latin inscriptions have highly formulaic language, so this doesn't necessarily mean years of commitment to learning a new language in order to read and understand them. The Ashmolean's 'Cracking Codes' display in the Reading and Writing Gallery (Figure 4) gives a quick guide to some of the abbreviations and formulas next to simple tombstones that can easily be read using only these formulas. This offers an opportunity for self-directed learning with the objects themselves. This is an example of how reading in museum galleries need not entail a passive, didactic mode of communication, since visitors must play an active role in spotting patterns and bringing them together to create meaning. This serves as a very literal model for the kind of reading which engaged visitors undertake in a museum and introduces the broader themes of the gallery (i.e., language and literacy across cultures). The most direct approach to tackling the unreadability of objects is creating targeted education programmes to help visitors to read them. Programmes targeted at schools can build the skills and confidence that children need to approach these objects, whether or not they have the opportunity to learn ancient languages. At the Ashmolean, we have had particular success with school projects that build on pre-existing curriculum coverage of ancient Rome (such as the requirement to teach 'the Roman Empire and its impact on Britain' at Key Stage 2 of the National Curriculum) and introduce elements of language. The Latin Adventure session for Primary Schools is designed to complement teaching in schools by exploring unusual dimensions of familiar Roman culture topics. For example, many schools study the Roman army as a topic, so the session includes learning about Roman soldiers who fought with slings and their tradition of writing taunts to their enemies on the bullets. Children learn simple formulae to make their own taunts using words and images, and design polystyrene sling bullets with their personalised taunts. The session uses elements of rôle play, investigation and craft to build on the strengths of the museum space. Language learning is central, but its importance is less in the acquisition of vocabulary and grammar than as an intrinsic part of ancient lifeone that we can still participate in and use to learn about the past.
Adult learning classes that give visitors the language and object interpretation skills to make sense of Latin inscriptions for themselves are a popular part of the public programmes of the British Museum and Shrewsbury Museum and Art Gallery. In both cases, small groups are given what is essentially a specialist tour. As with a traditional museum tour, visitors can learn more about the collections from an expert, but with the added opportunity to acquire an unusual skill. The British Museum has turned its tours into a book (Booms 2016), which allows independent learners to have a similar experience (and may prove helpful to museums looking to offer a similar approach). Teaching languages in the museum in this way depends on staff or volunteers with specialist skills. However, there is scope for museums to build collaboration to supplement their skills, whether through contacts in local universities or through the projects dedicated to increasing access to Latin and ancient Greek, including the Iris project (http:// irisproject.org.uk/) and Classics for All (https://classicsforall.org.uk).
Participants in such schemes (while highly engaged and enthusiastic) are only ever going to offer themselves in small numbers. Museums need approaches that also work for visitors who do not have time to learn a new language or do not think of themselves as the sort of people who are interested in Latin. There are still ways to engage such visitors directly with the language of these objects without a full understanding of the words. For another event at the Musée Départemental Arles Antique (for la Nuit des Musées 2016) the artist Dominik Barbier was commissioned to create a projected audio-visual artwork called Dis Manibus. This artwork was projected onto the wall where the Roman funerary inscriptions are displayed, using the interaction between stone and light to play with and rearrange the letters of the inscriptions, converting them into translations and further transforming them into the fire and smoke of a funerary pyre (excerpts can be watched on the museum's Vimeo account: https://vimeo.com/172170457). This demonstrated both the literal meaning of the inscription, and the social context and beliefs behind the simple invocation of 'Dis Manibus' ('to the spirits of the dead').
Multimedia presentation is a common solution to the problems of textual overload that can result from communicating written objects through written labels. It can be used to convey a text in a more accessible way, particularly in busy temporary exhibitions where more people can listen to a reading than can stand around a single label. A good example of this is the way in which the British Museum often uses recordings of its inscriptions to make them more accessible in its temporary exhibitions (e.g., in the Hadrian Empire and Conflict exhibition). As well as improving accessibility, placing vocal performance alongside an inscription can transform the nature of the encounter with it. We have already mentioned how Regina speaks at Arbeia. This is a rich way of conveying information, both through the words spoken and through elements of performance, such as Regina's regional accent. Most such approaches are pre-recorded with little space for the visitor to respond or answer back. As with written approaches, this need not mean that visitors cannot make their own meanings, but this process is not a feature of the medium of the presentation but of the content and the extent to which it leaves space for imaginative engagement (Witcomb 2003). Bubaris (2014) describes the use of sound in museums that is separate from the temporal experience of the visitor as 'non-diegetic' and describes such experiences as an artistic choice with its own rhetorical effects, often to do with authority and directing attention.
However, readings of a text can also have emotional impact by allowing us to imagine the voices of the dead. Recordings in Diocletian's Baths use other relevant texts to add context to inscriptions. For example, Horace's hymn for the Centennial Games, composed to be sung twice during the religious rites by choirs of children, is mentioned in one of the inscriptions on display. A performance of this hymn, recorded by a choir of children, is played alongside the inscription to reconstruct what this part of the festival might have been like. Attention to the artistic quality of the performance gives the space atmosphere in the present, but also evokes the inscription's relevance to Roman lived experience. Reading language aloud communicates implicitly that the words on the stone were a living spoken language. The rhythm and texture of that language were as much a part of a text as its literal meaning and much easier to access when the words are spoken by someone who can pronounce them with confidence. Bubaris (2014) stresses the ability of sound to fill a space and touch the listener in a way other media/sensory expressions cannot. In this case, it fills out the portico of the cloister where the inscription is on display and combats the flatness of the inscription. In this example, the temporal qualities of sound are also important. Listening to a 'dead' language is a powerful and unusual way to experience the past in the present. The fact that this is in the form of a non-diegetic performance has its own rhetorical impact. While it makes the encounter with the past seem less direct than with a costumed interpreter or interactive exhibit, the sound is immersive, but still remote. It models a relationship with the past that stresses its distance from us even as it lets us feel it all around us. The result is more like a haunting than like time travel.
There is no one-size-fits-all approach to making the language of Latin inscriptions accessible in a museum, but there are plenty of possible avenues to explore. This section has looked at how traditional labels can (with careful attention to structure and content) provide vivid impressions of these difficult objects and direct attention back to the objects themselves. Inscriptions themselves are designed to be readable on a number of different levels, and museums can acknowledge the problems of reading them and empower visitors to engage with the language in ways that are meaningful to them. This section has begun to show the meaning-making potential of evoking the past through inscriptions, and how this goes far beyond the language itself. The next section looks at their particular strength for imagining the past through narrative engagement.

Narratives from inscriptions
While the language of inscriptions is the feature that makes them unique, its strength is to convey information. This section explores how best to use that information in a museum context to create rich visitor experiences. It cannot be an exhaustive exploration of the topic (since each inscription has its own unique content with its own range of interpretive possibilities), but it highlights the idea of inscriptions as vivid objects to explore how they can inspire and feature in imaginative, immersive reconstructions of the ancient world. At the heart of all these approaches is the unique narrative potential of inscriptions.
Narrative is a major interpretive tool for museums (Bedford 2001;Glover Frykman 2009;Nielsen 2017) and also an interpretive strategy used independently by their visitors (Everett and Barrett 2009;Chronis 2012). Museums use storytelling techniques to move away from the idea of museums as preserving artefacts 'in aspic', towards encouraging visitors to re-examine and re-contextualise what they see (Black 2012, 9). Stories about the past can be creatively invented by non-specialists as much as painstakingly pieced together by archaeologists or historians. This approach has the advantage of empowering visitors to create their own stories about the objects: no longer is there a single correct version of events that we need to reconstruct and convey (O'Neill 2006), but the sense of a shared learning journey where museum users can be active collaborators in interpreting the objects on display (Black 2012, 10).
We saw in the introduction how the presence of personal names on many inscriptions allows us to people the Roman world with a much wider variety of individuals than is often imagined, going well beyond the history of 'dead white males'. This can be used as a foundation for living history programmes, such as the costumed interpreters at the Roman Baths, Bath, whose profiles are taken directly from individuals named on inscriptions in their collection. Linking such names with objects can make them much more than evidence for the existence of the named person. For example, a set of panpipes in the Ashmolean from Shakenoak Roman villa in Oxfordshire (AN2005.34: AshLI 423 = http:// latininscriptions.ashmus.ox.ac.uk/xml/AN_2005_34.xml) can still be played, connecting us directly with the sounds made and heard by the two people whose names are inscribed on them. However, even in this unusually rich object there are gaps, and these gaps make excellent storytelling prompts. We ask visitors on tours and education events how they think two people came to own one set of panpipes (a question as puzzling to us as to them) and the resulting suggestions are often insightful and amusing. Chronis (2012;2015) has discussed how gaps in the explanations of museum displays prompt narrative imagining from visitors and this process can be seen to be particularly powerful as visitors explore the gaps between the textual and material dimensions of objects and the lives of the people they name.
The desire to fill the gap between inscription and ancient context can also inspire visual engagementfor example the Voci di pietra exhibition at Catania Civic museum invited school student collaborators to produce colourful images and immersive reconstructions to restore lost context to the inscriptions on display. The students added a silhouette and paintings of a missing statue of Venus to make the function of an inscribed statue base instantly recognisable (https: //www.merton.ox.ac.uk/news/'voci-di-pietra'-dr-jonathanprag-curates-new-exhibition-roman-inscriptions). Such visual reconstructions allow visitors to reflect on inscriptions as objects with a practical use and a rich social context. For example, The Ashmolean Latin Inscriptions Project reconstructed a Roman columbarium (multiple burial chamber) in the Reading and Writing gallery ( Figure 5). This evokes the hand painted decoration of the columbarium, adding colour to the display. Colour is a major feature missing from ancient objects as we see them today (Bradley 2009) and has been used to transform the way that museum visitors see ancient objects through painted replicas (e.g., the Gods in Color travelling exhibition) or light projection (e.g., special events at the Ara Pacis in Rome: http://en.arapacis.it/mostre_ed_eventi/eventi/l_ ara_com_era). Our columbarium was designed to evoke not only the visual experience of this ancient environment, but also its social role in commemoration, with labels foregrounding the people mentioned in inscriptions and activities like making offerings at graves. Displays like this can provide a vivid context for creating immersive stories about the past.
As part of the Ashmolean Museum's Dead Friday event marking Halloween in 2015, our project performed a Roman funeral, complete with a procession of the body on a bier, professional mourners, freed slaves, family, and musicians (http://blogs.ashmolean.org/la tininscriptions/2015/11/23/the-walking-dead-staging-a-roman-funeral-at-the-ashmoleanmuseum/) (Masséglia 2016). The event was started by the Funeral Director, who provided some contextual information about the strange world of Roman funerals and addressed the audience as participants in the procession. The procession then wound its way into an atrium-space, where the eulogy was delivered by the deceased's son, followed by an off-stage cremation and narrative of that spectacle. The procession then ended up at the columbarium display, where the real ash urn of the deceased had now been put on display (the ash urn of Tiberius Claudius Abascantianus: AN Loan 88 = AshLI 169 = http://latininscriptions.ashmus.ox.ac.uk/collections/gallery/detail/?a=AN.Loan.88), and where the procession was dismissed. The project leader gave a brief talk about the ash chest to provide further information about Roman burial customs and explain the ways in which storytelling decisions had been extrapolated from the archaeological evidence to make a coherent Roman funeral. Visitors were invited to commission their own replica tombstones and engage further with staff and volunteers. Over five hundred epitaphs were produced in Latin by our scribes in collaboration with visitors who created their own Roman identities. The story of the funeral encouraged visitor participation, with a range of possible degrees of involvement, from simply absorbing the atmosphere to creating a character and joining in the noisy spectacle.
The evocative power of stories can also be used to encourage visitor participation outside the museum space (especially on social media) through engaging marketing strategies. Lively video trailers recorded for the exhibition Made in Roma at Trajan's Markets in Rome are a good example of the role Latin inscriptions can play in advertising an exhibition (https://youtu.be/UJfOnGXNqAU). The actual objectsstamped inscriptions on pieces of brickseem unlikely stars of the show, but by extracting characters from the individuals named on the bricks and by setting up a fictional conversation between a high-ranking senator and a freedwoman brick-manufacturer, the exhibition became such a success that it was extended for further weeks. One of the keys to this success was the humorous flavour of the trailers: in Roman reality, no freedwoman would ever have had a conversation with a senator on an equal footing. The possibility of an ordinary woman speaking back to a pompous and powerful individual is genuinely funny in this instance because of the importance of character to stories. Humour is a dimension of the past that museums almost never touch upon, and it also makes videos like these much more likely to go viral on social media. The way that sharing is built into social media sites not only increases museums' reach (if they produce effective content) but also gives users a way to respond actively to the content in a public space tied to their own identity, promoting deeper engagement.
Perhaps the deepest form of narrative engagement with inscriptions that museums can provide is the opportunity for visitors to create their own narratives about the past through creative writing programmes. The Museum of My Life programme at the Museum of Antiquities at the University of Newcastle (now incorporated into the Great North Museum) used poetry writing as a way to explore its collections. A group of adult learners worked with the poet Maureen Almond and museum staff to learn about objects from the collections through guided questioning and close sensory engagement. They were then encouraged to write poetry that explored the personal dimensions of these objects (Almond 2013). Inscriptions were an important part of this, offering just enough of the personal stories of ancient Romans, while still leaving space for the writers to imagine the emotional and personal dimensions that might have been lost: By using poetry as a medium, people are encouraged to develop their own thoughts and reflections on objects in museum galleries which results in a more personal relationship with those objects. It's all to do with putting the human back into the objects. If there is an inscription, that's great, but it is a starting point; something to be built on if one is to really grasp that an item belonged to/ was made by/ used by / treasured by a real person. Their hands, their lips have touched that object. They lived and shared the same humanity as we do. -(Maureen Almond, email to author) In this case, the specific role of creative writing (which need not be perfectly accurate to be effective) freed people who did not consider themselves as experts to think deeply about objects and come to their own conclusions about possible meaning and function. Like the more informal gap-filling narrative engagement discussed earlier in this article, this can allow visitors to make complete meaning from a patchy archaeological record, including restoring many of the things that make the past vivid. It creates space for museums to speculate about features of the past that are difficult to prove, without undermining trust in their expertise. It also allows experts to talk freely about the relationship between imaginative speculation and close study of the evidence in generating knowledge about the ancient world. Both archaeological theorists (Mickel 2012) and excavators (Andrews, Barrett, and Lewis 2000) recognise the value of narrative in bridging the gap between the raw evidence of ancient sites and objects and the bigger interpretive conclusions that can be drawn from them. Storytelling in museums opens up the creativity and subjectivity of understanding the past to the public, both by making experts' accounts of the past more compelling and by empowering non-experts to explore the possibilities raised by museum objects for themselves. Because of their evocative combination of text and object, the deep engagement prompted by their 'difficulty' and the gaps still left to be explored, inscriptions are an ideal starting point for such engagement.

Conclusion
This article makes the case for giving inscriptions a more prominent place in museums. These are difficult objects, and their value to visitors is strongly influenced by the quality of the interpretation that museums provide. We have discussed a wide range of possible tactics for overcoming the language barrier, maintaining focus on the objects and getting at the real people and personal stories that make these objects so effective. Personal names that can be imbued with a narrative richness and used to attract visitors create immersive experiences of the past and encourage visitors to take an active role in forming their own sense of history. The personal stories drawn from inscriptions often offer features that visitors can relate to but can also challenge preconceptions. There is much to be said for using stories like that told on Regina's tombstone of the integration of her Palmyrene husband within the society of north-eastern Britain to show some of the complexities of ancient lives and challenge the racist narratives that some visitors will have been exposed to (Zuckerberg 2017).
The unintelligibility of Latin inscriptions to most people might usually act as a turn-off, but it has scope to deepen engagement if handled correctly. Beard (1992) has written on the surprising appeal of the Rosetta stone (an object which would be seen as boring without its pivotal place in the story of modern Egyptology). She sees it as a metonym for the intellectual work of solving the mysteries of the past. By this reasoning, the difficulty of the Rosetta stone is an essential part of its appeal, without which it could not work as a physical embodiment of the challenge and excitement of engaging with the past. The Rosetta stone could be seen as an exception, but we argue that this is not limited to a few famous and influential objects. With the right approach, any inscription can be a personal Rosetta stone for visitors that not only stands for solving the mysteries of the past but lets visitors participate actively in making their own understanding.