Beyond the martial façade: gender, heritage and medieval castles

Gendered interpretations are rare both within castle-studies and heritage discourses on medieval castles. Yet, castles hold potential to inform multi-vocal accounts of the medieval past and to inspire meaningful heritage interpretations to achieve greater societal impact. This article explores the role that gender currently plays in interpretations of medieval castles in Britain, supported by three case-studies written by heri- tage professionals. The enduring narrative of militarism at medieval castles sites is discussed, together with issues of authenticity in relation to the historical record, which is in itself biased and inherently gendered. Outcomes from a collaborative workshop highlight the need to address interpretative issues where gender is considered to equate to ‘ making women visible ’ . Finally, we pose the question: What makes a ‘ good gendered interpretation ’ at a public heritage site?


Introduction
This article explores the context of medieval castle-studies, gender, and related heritage discourse. It arises from a collaborative workshop on medieval castles as part of a larger project which challenges the prevailing notion that the medieval castle was a 'man's world', by telling stories of women's lives through the things (objects) they used and cared about, in the spaces (castles) in which they lived or worked. The project highlights the absence of women's stories in castle-studies but also reflects on the problematic assumption that gendered approaches are synonymous with 'things associated with women' (Smith 2008). Gender is more appropriately understood as the performance of a socially constructed identity that intersects with age, race, class, and which is fluid and changeable over time (Butler 1988;Gilchrist 1999). Museum and Library Studies and other cognate disciplines have explored gender in terms of the absence of representation of women and LGBTQ+ perspectives (Porter 1995;Vanegas 2002;Steorn 2012). However, the critical reflection in these disciplines has not impacted on heritage discourses surrounding medieval castles or castle-studies. A workshop was designed to bring together heritage practitioners and academics from across Britain to jointly discuss approaches to gendered interpretation in confronting the Authorised Heritage Discourse (AHD;Smith 2006;Waterton and Watson 2013) of militarism in the public interpretation of medieval castles. This collaboration helped to identify the overarching aims for the gendered interpretation of castles; participants reflected on what makes a 'good' or 'bad' gendered interpretation in relation to castles; and how and why the AHD of militarism continues to bias public interpretation at castle sites.
Three case-studies presented here draw out key issues in the gendered interpretation of castles that intersects with broader theoretical concerns in gender archaeology and heritage studies. All of the casestudies grapple with issues of authenticity and source criticism, reflecting on how to make medieval women 'visible' in historical sources that are biased towards male agency and masculine definitions of power. They demonstrate a consistent focus on named (elite) women, revealing that public interpretations of castles have not yet fully incorporated wider archaeological sources and approaches to gender. In case-study 1, Stefan Sagrott highlights the relative absence of women's stories from medieval castles in the care of Historic Environment Scotland, emphasising problems around the visibility and representation of women. We extend this theme beyond the case study through an analysis of guidebooks at selected castle sites, examining the changing historiographical context for the omission of women in castle-studies. In case-study 2, Jeremy Ashbee discusses the development of an inclusive approach to interactive interpretation at Goodrich Castle, England, during the occupancy of Joan de Valence (d. 1307 AD). He highlights the tension between authenticity and inclusion in addressing gender at medieval castles, posing the question of what compromises can and should be made in relation to historical and archaeological evidence. The final case study takes an innovative approach to defining heritage value or 'significance' in relation to gender. Samantha Stones examines the role of gender in decision-making in relation to the conservation of building fabric, past and present, at Carisbrooke Castle, Isle of Wight, England, associated with Isabella de Fortibus (d. 1293 AD).

Castle-studies
Castle scholarship is a niche area of medieval studies, the origins of which is enmeshed in military and architectural-history traditions of the late nineteenth and twentieth century. Until the 1990s, scholarship on medieval castles in Britain and much of Europe remained firmly stuck within a descriptive tradition, where architectural details of upstanding masonry structures and related historical dates were prioritised over interpretative narratives that featured the lives of people. Emphasis was placed on investigations of patronage and pinpointing a primary construction phase with a corresponding date. The accepted view in the discipline was that castles were primarily military structures built for, and as a result of, warfare (Hamilton Thompson 1912;Platt 1982;Brown 1985).
During the 1990s and 2000s, medieval archaeologists, historians and architectural historians began to emphasise more social questions and interpretations. The discipline moved away from understanding medieval castles solely in terms of military power (Heslop 1991;Coulson 1993;Gilchrist 1999). Subsequently, it became widely acknowledged that castles functioned in multiple ways, including as residences, centres of administration as well as statements of power, status and wealth, whilst also performing military functions periodically (Speight 2004). Debates shifted from an exclusive focus on the castle's core masonry buildings to include the wider landscape, such as deer parks, routeways and settlements. Buildings were no longer divorced from their surroundings but instead situated within their cultural landscapes (Creighton and Liddiard 2008). New dimensions in scholarship included the spatial arrangement of buildings (Johnson 2002;Marshall 2002;Hansson 2006) as well as related discussions of how spaces can be gendered (Gilchrist 1999;Richardson 2003). Yet, despite these attempts to modernise the discourse (Creighton and Liddiard 2008), the castle-studies communities remain trapped in an ongoing debate of oppositional understanding, where military or functional aspects are pitted against symbolism or social understandings (Creighton 2018). This impasse appears to have prevented the exploration of wider theoretical approaches. Other mitigating factors include the manner in which research agendas were framed, especially rescue excavations at urban centres (Gerrard 2003).
Despite progress towards a modernised, critical discourse of castle-studies, medieval people still remain absent(ed) from discussions (Dempsey 2019). Where people are acknowledged as having been present, it is usually a story of one lord or knight who represents the embodiment of (modern) masculinityimplicitly lauded for his bodily prowess, or appreciated for wielding male power, both economic and political (Johnson 2002;Creighton and Wright 2017;O'Keeffe 2015). Women, when discussed, are categorized as wife-mother-daughter, and often are not named personally even when their names are documented (e.g. Mac Cotter 2016 discussing the fitzGerald family). If women did wield male-power, they are characterized as duplicitous, (both now and in the past): a good example of this is Rhoesia de Verdun (d. 1247), who was rumoured to have killed her stonemason after completion of the building of Castleroache, Co. Louth, Ireland (Barry 2008, 131-2).
Castles-studies has been slow to explore the experience of women beyond their relationship with men, but there is virtually no analysis of the vast range of masculine identities and life experiences in castles, outside the normative of the elite (white heterosexual) male. If a man was not a knight or lord, then there is no space made for his alternative story: these men are as invisible and hidden from history as their female counterparts. The absence of different voices, even within popular narratives of public interpretation, has a profound effect on how we understand the past. For example, within the typical tale of the knight/ warrior there is much discussion of warfare but little exploration of the lived experience of war. The potentially disabling battle injuries that the male body was subject to/a perpetrator of do not feature (Knüsel 2014). This is also seen in the way that the grand narratives of the medieval past focus on seemingly pivotal dates of military action but less so around the daily monotony of life between battles. The experience of soldiers 'passing the time' has tangible expression at some castle sitesgaming boards carved into steps, an abundance of inexpertly carved but functional musical instruments, as well as evidence for gaming pieces and dice, that together indicate a degree of free time (Hall 2018).
Material culture assemblages recovered from castle excavations can be limited, perhaps owing to issues surrounding preservation or the fact that elite buildings were often temporarily occupied, furnished by portable furniture, textiles, utensils, tools and so on (Thorstad 2018). Yet, some sizeable assemblages have been excavated, and it is surprising that discussions of material culture are largely absent from narratives of the knightly and noble classes. Investigations of material culture could raise interesting questions about gender and sexual identities, as well as bringing in potential stories of homosocial behaviour or same-sex relations in castles (see Parkinson 2013). For example, who was the performance of masculinity for, and to what degree did class influence this display? (Nøttveit 2006;Gilchrist 2012). What did courtly masculinity look like in comparison to masculinities of the peasant class? Objects such as spurs and horse harnesses could be used to discuss particular aspects of masculinity, although the latter are less exclusively male. These objects, many of which survive, were often ornate and designed to be visible. There are numerous historical references to spurs as gifts, some given in 'homage' to lords from people lower down the social hierarchy (CDI ii,27,232,403). Other seemingly mundane items such as shears can provide a starting point for medieval lifeways more generally: they can be utilised for craftwork such as spinning or tailoring; they could be used for gardening, whether cutting flowers for display or herbs for medicinal or cooking use; as well as having a role in obstetrics.
'Things' allow for evidence-based, people-centred narratives that account for daily life and lived experience. We must read the archaeological record beyond the 'binary bind' (Ghisleni et al. 2016): objects can, do and did transmit and embody coded messages, including same-sex desire and expressions of gendered identities. Castle-studies has rich potential to create multi-vocal accounts of the medieval past to inform heritage interpretation at castle sites. Such accounts would increase the relevance and accessibility of medieval castles, amplifying their heritage value and contributing to contemporary debate. Creating inclusive, gendered interpretations that account for differences in the past can address historical and systemic inequalities in relation to gender (Kryder-Reid 2018, 691).
The field of castle-studies does not typically acknowledge the vast range of identities within the medieval past: gender, age, social class and ethnicity remain under-explored, as well as the experience of the body and the life course. This is deeply regrettablethese are the powerful factors that shape everyday life: how people make themselves and understand the world around them. There have been important studies on how and why people living in castles organised their space and interacted with each other (Coulson 1976;Fairclough 1992;Gilchrist 1999;Marshall 2002;Mol 2011;Eadie 2015) but little on how and why they used a particular material culture in a distinctive way. In other words, the social and cultural practices of medieval people have not yet been embedded in an interpretative framework for castle-studies. Surveys of the public undertaken by Historic Environment Scotland demonstrate that these questions are precisely the things that people visiting castles want to learn about. This is shown in case-study 1, where Stefan Sagrott reflects on gendered interpretations of medieval castles at Historic Environment Scotland, revealing the public appetite to learn what people's daily life was like.

Heritage discourse on medieval castles
The literature on medieval castles within critical heritage studies is minimal. Where featured, discussions of castles tend to focus on conservation and restoration work (Vodopivec et al. 2014), where to stop and when too much alteration has taken place to fabric (e.g. Château Pierrefronds, Picardy). Less explored are questions of how conservation and restoration work practices are influenced by a desire for a particular authenticity, in other words to enhance the medieval-ness of a building (Coppack 1999;Worsley 2004). The dominant approach is that of a management perspective where heritage is a tool and/or material resource that can be analysed, graded and improved (McGuicken 2006). Visitor experience is discussed in terms of whether tourists prefer guided tours versus self-guided (Edinburgh Castle: Beattie 2018); how easy/hard a site is for visitors to navigate (Leeds Castle : Laws 1998); or the quality of special events such as re-enactments (Caerphilly Castle: Light 1996).
Despite the absence of critical discourse, castles are understood to play an important role in the cultural production of heritage (Chitty 1999;Light and Watson 2016). However, people are absent from the corresponding imagery of public interpretation (Waterton 2009;Waterton and Smith 2009;Högberg 2012;Havery and Waterton 2015). This mirrors the prevailing approach in castle-studies, where people tend to be absented from the narrative and discussions centre on buildings, landscapes or history. Attempts have been made to move away from this narrative at sites such as Helmsley Castle, N Yorkshire (Emerick 2014). However, heritage interpretations or guidebooks still largely comprise unilinear understandings and masculinist perspectives, which often equates with a focus on defence/warfare or power. This then, begs the question, who is the intended audience for heritage at these sites? Are interpretative projects at castles directed at a male gaze, one assumed to be primarily interested in an idealised military past? Are displays intentionally gendered 'male' to invite a particular 'female' gaze? What constitutes appropriate images for these intended gazesis it representations of (passive) womenboth elite and ordinaryas well as knights? Where does this bias originate in heritage practicedo heritage professional bodies have an inherent gendered bias in terms of employment or managerial interest that influences these projects? This needs to be disentangled.

The workshop
The workshop involved heritage professionals from Cadw, English Heritage, Historic Royal Palaces and Historic Environment Scotland, along with a range of academics, professional archaeologists, as well as independent researchers at various career stages. A gender balance was achieved for speakers, but despite sustained efforts, the non-presenting participants were largely female (21:4).
One of the key aims was to share different interpretative approaches to gender and medieval castles, as well as to give as much time as possible to informal discussion. There were six, twenty-minute papers on professional experiences of applying gendered approaches to interpretations of medieval castles. The workshop also included 'breakout' sessions, one of which is discussed here: 'What makes a good/bad gendered interpretation at medieval castles?' In groups, the participants reflected on what makes 'bad' or 'good' interpretations ( Figures 1 and 2). This breakout session was very revealing, with notable disparity in what constituted 'good' and 'bad' gender interpretations as outlined by each group. Suggestions included a call for 'nuanced' accounts, 'everyday life or daily-life experiences' and the presence of 'people' (Figure 2).
'Including women' also featured prominently. Some groups engaged more directly with current theoretical trends of what 'gender' means, noting the importance of the body, life course and cross-societal approaches.
Overall, it seems that imagining what 'good' gendered interpretations looked like, proved a lot easier than discussing the 'bad'. There were some differences of opinion around the fairness of listing gendered interpretations as 'bad' because they were simplistic or did not capture the extents of different gender identities. These were viewed as 'a good effort', reminiscent of second wave feminism and the 'add women and stir' approach (Tringham 1991). There was a consensus across all groups that merely including women in a superficial way was a 'bad' interpretation, despite the fact that most groups had defined a 'good' gendered interpretation as 'including women' (Figures 1 and 2). The results of this breakout  appeared to demonstrate that inserting women into traditional male narratives was still regarded as a valid methodology within professional heritage perspectives. While acknowledging that this approach is problematic, it is important to note that heritage organisations are slow to change but gradual change is happening (Kryder-Reid 2018, 691; see case-study 2).
Case-study 1: rethinking gender at Scottish medieval castles by Stefan Sagrott Historic Environment Scotland cares for and manages 336 guardianship properties. This includes conducting research, undertaking interpretation projects (panels, guidebooks and audioguides) as well as conservation and maintenance work across the sites. Central to all of this is the curation and sharing of knowledge with the general public. The Cultural Resources Team directly delivers or commissions research for interpretation projects on Properties in Care. Within this role, we endeavour to include histories of marginalized communities, such as people of colour, or feature topics such as gender, neither of which have received due attention at our historic sites.
During the nineteenth and twentieth century, studies and interpretations of castles were dominated by topics of defence and warfare, which usually exclude all but the most exceptional women. However, our knowledge and understanding of the roles of the women who used, visited, worked at and inhabited castles remains inadequate. The concentration on military aspects is curious as we know that the castle was more than just a defensive structure. They were family homes, and for many noble-women they were a long-stay residence which also brought with it administrative responsibilities. However, is this breadth clear in our interpretation of these sites? Do our guide books and information boards put across a balanced view?
An examination of how gender was portrayed was undertaken across the 73 castles in the care of Historic Environment Scotland. Discussions between staff among a range of teams within the organisation led to an initial analysis of named historical figures in guidebooks for a selection of our medieval castles (Table 1). The results of the analysis are telling: it demonstrates the large bias in how historical events and seemingly key figures are recorded and told, with men dominating the record. It illustrated too, the biases that are present within our workin how we are telling the story to the public today. It is particularly disappointing that the recent Threave Guidebook, which was written in 2017, contains a large gender gap. This is surprising, as we had consciously aimed to include women in the guide, but it seems we had not realised how many named men we had used in the narrative.
Yet, at the same time, Threave is also a success story. Past interpretative narratives had traditionally focused on the castle's patrons and its defensive role during sieges. Two key women in the politics of early fifteenth-century Scotland, Princess Margaret Stewart (d. unknown) and her granddaughter, Margaret, the 'fair maid' of Galloway (d. 1688), had occupied Threave but were largely absent from the story of the castle. Our research confirmed that they were central to the story of Threave and we were cognisant of the previous male-dominated narratives. Simple measures in the reconstruction drawings, such as seating Lady Margaret at the head of the dining table, were means by which we were able to reinforce their importance ( Figure 3). However, it should also be acknowledged that it can be challenging to highlight the role of women in the medieval period, due to the nature of the sources and texts available to us. Perhaps if we placed as much emphasis on the material record as on the historical, we could tell more balanced stories? As heritage professionals we have to continually challenge and question the interpretations that are presented at the sites in our care. The Bechdel-Wallace Test, introduced by Alison Bechdel in her comic strip, Dykes To Watch Out For, is pertinent here (Bechdel 1986). The test contains three criteria that a film needs to meet in order for it to be considered to actively represent women: (1) The movie has to have at least two women in it; (2) who talk to each other; (3) about something besides a man.
The Jaffer-Humphreys Test, created by Laura Humphreys and Aaron Jaffer, asks similar questions of museum/heritage displays: (1) Are there at least two women 'on display' in this gallery?
(2) Are they presented in terms of their relationship to a man?
For a museum display to pass this threshold test, you need at least two women, presented in their own right. If a similar test was applied to our current guidebooks and interpretation panels, they may not be cast in a favourable light.
However, we are aware of this issue and are endeavouring to be more inclusive and representative. Historic Environment Scotland recently commissioned visitor research to better understand what our visitors want and expect from our site interpretation. Of a sample size of 2968 visitors, 49% stated that they wanted to learn more about everyday life at castles and how people lived, with specific topics identified: • What people wore, what they ate, how they found and cooked food, how and where they slept, the work they did, their education, medicine, how they died; • How people living or working at the site managed everyday chores; • How medieval people dealt with/managed issues such as crime, travelling long distances and the harshness of the land and weather. While we cannot immediately address every topic of interest, this survey provides us with a focus when we undertake research and when we produce interpretation panels, to ensure that these important topics can be understood by and shared with our visitors. Perhaps we can begin to disentangle the uniform 'they' into all peoplenot just men, but women, children and those with different gendered identities and roles at various stages of the life course.
Case-study 2: new interpretation at Goodrich castle, Herefordshire, by Jeremy Ashbee The creation in summer 2018 of 'As I was saying', a new piece of public interpretation at Goodrich Castle, Herefordshire, raised a number of interesting challenges for the project team, especially concerning an accurate representation of the gender roles in a medieval castle. Goodrich, one of around 70 medieval castles managed by English Heritage, has been ruined since the Civil Wars of the mid-seventeenth century, but is nevertheless among the best preserved castles in the organisation's portfolio (Figure 4). The greater part of its buildings date to a comprehensive rebuilding in the late thirteenth-century, when the castle belonged to William de Valence, Earl of Pembroke (d. 1296) and his wife, Countess Joan (d. 1307). Subsequent owners retained the groundplan of the thirteenth-century castle with only minimal changes (Shoesmith 2014).
These extensive buildings are complemented by three remarkable documents, dating to 1295-97: accounts listing the daily household expenditure of Countess Joan de Valence, widowed in May 1296 and now the head of the household (TNA E101/505/25,/26 and/27). The accounts are well-known to social historians and have been mined for information by scholars including Mertes (1988), Woolgar (1999) and most recently Mitchell (2016), author of the only dedicated biography of Countess Joan. The accounts followed a standard formula, listing the outgoings of each household department on every day, and indicating the mistress's location, with the identities of close relations and guests staying with her. The location clauses show that during the period covered by the accounts, Joan stayed at Goodrich Castle for six months between November 1296 and May 1297, and again in September 1297, when the account ends. Most of these daily entries were concerned with the consumption of foodstuffs, but they contained details of occasional purchases, such as a lantern for the kitchen window, and a canopy for the mistress's bed. They also provide information about the activities of household members -Philip, a clerk, travelled to Chepstow to buy fish, Thomas the chaplain bought wax at Monmouth. Finally, and most usefully, the accounts also listed at intervals members of an inner core of servants and officials who travelled with the Countess: the chapel clerk, Humfrey of the mistress's chamber, John of the mistress's wardrobe, the mistress's laundress, Jacquetta of Beatrice's chamber, and so on. Goodrich therefore offers a rare combination of buildings surviving in a 'legible' condition and detailed documentary information about some of the people who lived in them when they were newly built. This information had already been used to good effect in the current guidebook to the castle and in specialist tours given to groups of visitors, who had received it enthusiastically.
In 2017, English Heritage began to work with Matheson Marcault, interpretation designers, on an additional piece of interpretation for Goodrich Castle, targeted at family groups, recognised as a segment needing dedicated interpretation: from the outset, it was specified that this should be focused on the household of Countess Joan. Matheson Marcault devised the interpretation as a card game, to be played by groups of people at specified locations around the castle ( Figure 5). One player would read out a brief biography of a historical character, such as the Countess Joan or one of her household, and would then read out the beginning of a sentence specific to the character. The other players would compete to finish the sentence using words from their cards: some comically inappropriate, some plausible, and a small number correct.
The household described in the documents was certainly not exclusively male, and the team wished to reflect this in the choice of nine historical characters for the game. The unchallenged head was the Countess herself, usually accompanied by her daughter-in-law, Beatrice de Nesles. Female servants were also mentioned, such as 'Jacquetta of Beatrice's chamber', presumably a maid or lady-in-waiting. But named or unnamed female figures were in short supply in the accounts, considerably outnumbered by men. Even within the innermost chambers of the Countess, some of the principal attendants were men, such as 'Humfrey of the Mistress's chamber' as well as the necessarily male preserves of the chapel. The team considered featuring aristocratic female friends of the Countess, such as the Lady of Raglan, who visited Goodrich in the spring of 1297, though she was felt to be too similar to the Countess. The nobly-born Prioress of Aconbury, another friend and occasional visitor, might have held greater interest. For a female servant, it was decided to invent a character, Agnes, assisting John the baker. It must be admitted that no evidence exists for this character, and this is specified in the text on the card. Each historical personage needed to be depicted on their card, and this raised some unexpected further issues of gender representation. The approach we had initially intended to useto base the depictions on references from medieval manuscripts such as the Queen Mary Psalter and the Luttrell Psaltercertainly gave a plausible start. But the process involved the team debating a number of difficult points of detail. For example, the knight Sir Roger de Inkpenne, a versatile and busy member of the Countess's affinity, had to be recognisable as a man of status, but he also represented the social class who might lead soldiers in the castle's defence. Portraying him in chain mail with sword, spear and banner, certainly fulfilled these objectives, but raised eyebrows, both as a cliché, and undoubtedly as an image far removed from his everyday reality. Beatrice was a character only occasionally mentioned in the accounts and other sources, and one of the more endearing details highlighted by Linda Mitchell was her letter to her absent husband, Aymer de Valence (TNA SC1/48/183; see Mitchell 2016, 98, n, 147). The illustrator initially showed Beatrice with the letter in one hand and a quill pen in the other. The assumption that Beatrice had written the letter herself was most unlikelythe manuscript is in the tiny and neat hand of a professional clerk, and such research as was available suggested that the mechanical skill of writing was not widely taught to aristocratic women in the late thirteenth century. To show Beatrice holding only the letter would indubitably have been the best solution: in the event, through a loss of nerve from the writer, the illustrator was instructed to show her with an aquamanile (a hand-washing vessel in the form of an animal or human).
No formal evaluation of the game has yet been carried out, but the castle's staff report anecdotally that it has been well received. The team hopes that the game has succeeded in communicating 'softly' some of the nuances of the varied gender roles in the castle, as well as entertaining visitors, especially young visitors. The team concluded that 'As I was saying' has made a useful start in the re-framing of the debates around this medieval castle, but predictably, much remains to do.

Case-study 3: thoughts on conservation practice at Carisbrooke castle by Samantha Stones
In England, contemporary conservation practice in the historic environment is guided by the use of 'Conservation Principles' (English Heritage 2008). The application of this framework facilitates an assessment of the importance, or 'significance', of an historic asset by weighing identifiable historical, evidential, aesthetic and communal values. Condition surveys and assessments of risk or vulnerability are also critical, but assessments of significance contribute to the decision-making process regarding the allocation of resources. If a structure or building element can be assigned one or a combination of heritage values, it is considered to have significance and is prioritised for conservation. Whilst carrying out minor re-pointing, consolidating stonework or even maintaining gutters may not at first appear to contribute to the narrative that a site presents, these interventions are indicative of an ascribed contemporary value and the prioritisation of a particular narrative (Coppack 1999;Chitty 1999, 94). This process of prioritisation is driven by assessments of 'heritage significance' as judged by remnants of the building or structures, interpretations of documentary evidence and other aspects of material culture (Paris 2000). Therefore, conservation of building fabric of historic assets must be considered as a medium of interpretation as buildings do not 'speak for themselves' (Chitty 1999). It is necessary to ensure that modern conservation practice incorporates scrutiny of the values that underpin and influence assessment of significance, including contemporary issues of gender, to enable critical assessment and perhaps revision of received and embedded narratives at historic sites (Arnold 2003).
Carisbrooke Castle, Isle of Wight (Hampshire) provides an interesting example of a highly significant medieval site that both demonstrates and embodies a multitude of assigned heritage values as well as, unusually, having a female-dominant narrative (Gilchrist 1999). The site has many important international and national historical associations, demonstrates evidence of multiple phases of activity and can be assigned both aesthetic and communal value (Kerr 2000;Gibbs 2004). Amongst these competing values and narratives, the association with Countess Isabella de Fortibus (d. 1293) is foregrounded by English Heritage. The associated guidebook refers to Carisbrooke as 'Countess Isabella's castle' (a rare association of a medieval castle with a woman without qualification) and she is frequently included in digital content (Young 2013). In addition, there is 'Isabella's window'; a visible building element on the visitor route ( Figure 6). The naming of this feature can be traced to at least the late nineteenth-century, its good condition can perhaps be attributed to this early assignment and acknowledgement of historical importance. This window opening is located in the north curtain wall and once formed part of Countess Isabella's great chamber. Of the extant buildings and structures of the castle, it is not the only one that can be associated with this period and there are other buildings that could be considered to be more significant, such as the earliest phase of the gatehouse, the great hall and remnants of a chapel. However, in contemporary assessments of value, the level of significance that the window currently enjoys ensures that it is a priority for conservation. It is always presented to the highest standard: clear of vegetation to ensure legibility and structurally stable to ensure longevity. For modern visitors, the window's name is a signifier of a prioritized story and building element. Ultimately, it conveys a message that this element is noteworthy, as is the associated individual. The window provides the visitor with a tangible connection to Isabella's life where they are encouraged to imagine her presence at that particular location, thereby ensuring that her story remains prominent. This example of conservation and presentation informed by one woman's narrative is perhaps exceptional. It also neatly illustrates how easily narratives can be created by extant structures, but also reinforced by repeated recognition of significance, followed by acts of preservation or conservation. The opposite is true for the vast majority of historical women and marginal groups for whom the association with the built record itself was unrecorded, has been allowed to decay, or has now been lost entirely.
In her critique of architectural history, Dana Arnold (2003) identified the ongoing and subjective nature of the 'dialogue' between the past and the present, through which meanings, associations and narratives are attached to buildings. For medieval castles, these layers are often inherently gendered male. We must consider the patriarchal social and cultural context in which medieval castles were constructed as well as the largely male institutions who created the contemporary documentary archive. Frustratingly, this male-dominated infrastructure continues in the modern world where bodies such as the Ministry of Works, shaped the narratives of many sites in the twentieth century (Emerick 2014). Perhaps future gender-attuned archivists and curators will shape the archive and collections differently (Paris 2000).
The combined result of historic and modern interventions has created the monuments we now conserve as professionals and enjoy as visitors, and it is necessary to acknowledge that these interventions, including those of the twentieth century, were always reflections of contemporary principles and priorities. Historic sites are multi-vocal; amongst other things, they reflect the priorities of the past and the present. Whilst it is one role of the heritage professional to identify, then balance, assigned and competing values, the point remains that significance is influenced by and reflective of contemporary social context (Chitty 1999). Contemporary conservation practice more generally requires an awareness of the diverse ways that historic values and events, including past conservation approaches and interventions, have influenced and continue to shape our current understandings (Paris 2000). This includes the presentation of gender at historic sites or indeed the lack thereof (Grahn 2011;Grahn and Wilson 2018). Without such critical awareness, historic biases and a desire for authenticity can be compounded or amplified, through even minor intervention.

Presentation of gender in past guidebooks of castle sites: Goodrich, Stirling and Carisbrooke
We examined a series of guidebooks to further analyse changing professional practice in the public interpretation of castles in relation to gender. This study focused on guidebooks of three castles that feature within the case-studies presented in this article -Goodrich, Stirling and Carisbrooke. There are inherent biases with this selection for an analysis, as they have historically documented and long-accepted associations with elite women who feature prominently in their narratives -Joan de Valence (d. 1307), Mary Queen of Scots (d.1587), and Isabella de Fortibus (d. 1293), respectively. These documented associations are likely to explain why these castles were chosen as case-studies for a workshop on gender, which, as noted earlier, is often misunderstood as 'things associated with women'. Therefore, any analysis of 'gender' or 'women's stories' from these castles is not representative more widely across medieval castles sites.
This analysis includes all of the official guidebooks for these three sites published by governmental agencies (Table 2). There have been at least three official guidebooks published for each site since the early twentieth century, with a new or updated edition typically produced every two to three decades, as well as numerous reprints per edition (Table 2). At least half of the updated guidebooks were written by a different author. All authors were male, apart from Margaret E. B. Root, assistant inspector of Ancient Monuments for Scotland, who co-authored Stirling Castle Guidebook (1936). Each guidebook follows a relatively similar format in that an historical narrative of the castle's patrons or occupiers is recounted and either pro-or preceded by an architectural description. This is completed as a guided tour approach, or as a list of features, e.g. 'barbican', 'gatehouse', 'hall'.
We anticipated that there might be a growth in explorations of gender over time; however, this is not exactly the case. There are no explorations of gender per se: it is more the case that men dominate the stories of these castles and women were either present or not. A quote from Carisbrooke Guidebook (2003, 23) communicates this: 'the last de Redvers lord of the island was in fact a woman . . . . . . . . . .to achieve this in a masculine and often dangerous world she must have been truly formidable and of very strong character' (Young 2003, 23).
It is unclear if the phrasing is intended to emphasise the unusual situation whereby a 'lord', usually a pronoun used for a male, is a woman, a position often denoted by use of the title 'countess' or 'lady'. Or, is there an implicit statement of surprise at the property being in the hands of a woman? The inference of female fragility in a 'dangerous', male world certainly transmits a particular gendered ideology, but it projects present social norms rather than those of the past.
The Goodrich Castle, Herefordshire (1933) guide by C. A. Ralegh Radford does not dwell on Joan de Valence, and cites the countess only as the conduit by which the castle passed to Aymer de Valence. Optimistically, rather than assuming that this is a deliberate omission, we might suggest that Radford may not have known about Joan's detailed household accounts. However, he had excellent medieval Latin and used medieval sources at other sites such as Glastonbury Abbey (Gilchrist 2013). Derek Renn's guidebook, Goodrich Castle, Herefordshire (1993) features only a summary of the accounts, which were certainly well-known by this time (e.g. Mertes 1988). Therefore, it appears that previous guidebook authors, consciously or unconsciously absented Joan de Valence, despite her known association with the castle and the wealth of historical documentation as well as architectural evidence. While the guidebooks for Goodrich have moved on from dismissing Joan de Valence as a side character, they are still primarily architectural accounts of the castle. Selected historical figures and their household accounts are included to cement the narrative of the building phases. Within the Stirling guides, Mary Queen of Scots is mentioned a great deal, as to be expected of a monarch closely associated with Scottish national identity. Isabella de Fortibus is also foregrounded in the Carisbrooke Castle guides, as well as the later Princess Beatrice (d. 1944) andQueen Victoria (d. 1901), but in comparison to stories of men there remains a significant gap.
There is a continuing assumption in castle guidebooks that we can only talk about the past through historical documentation, as outlined by Jeremy Ashbee in his concerns with an interpretative game. Equally interesting is the adherence to particular historical sources for use in site interpretation; usually they are descriptions of rare military action, favoured over the seemingly mundane but more plentiful household accounts (Mertes 1988), etiquette manuals (Philips 2005) and colourful manuscript imagery (Porter 2000). These restrictions further limit our potential to tell stories of the past that account for many different people who are not often represented in the historical narratives. However, the problem of over-reliance on limited primary sources can be viewed positively, as it means there is much room for future discussions inclusive of gender. One approach would be to explore how and why different groups of people from various places in the  1993 Derek Renn;reprints 1998reprint 2007revised reprint 2009Carisbrooke 1934second edition 1948second edition , reprint 1950second edition 1956second edition 1972revised reprint 20082010revised reprint 2013 social hierarchy used the spaces of the castlelandscape and buildingsin a particular (gendered) way. We should also experiment with multi-vocal biographies of objects in relation to castles, as noted above.

Conclusion
Gender, as is now well-accepted, is the performance of a socially constructed identity (Butler 1988). The case-studies presented above emphasise that within the context of public heritage and interpretation of castles, gender is perceived as the visibility and authentic representation of women, as defined by historical sources. The discipline of gender archaeology challenged the constraints of 'visibility' by developing new conceptual and methodological frameworks that expanded the definition of gender (e.g. Moore and Scott 1997). For heritage studies, a comparable approach might be to redefine heritage value in terms of gender, and to reflect on gender as part of both social and evidential heritage values. Gender could be regarded as a factor in assessing the social importance of a castle and in shaping its management, alongside historical and aesthetic values and the castle's relevance to a national story (Emerick 2014). Gendered interpretations of castles remain focused on the exceptional stories of named, elite women, privileging historical sources and placing maximum value on authenticity as defined by the written record.
Contributors to the workshop stressed the importance of strengthening interpretative emphasis on material culture and the archaeological record, to develop more balanced accounts of the daily life of all people in the past. There was also a call for more nuanced appreciations of medieval masculinity in castles and understanding that gender is not synonymous with women. There is interpretative scope to move beyond 'women equalling gender', to take account of different gendered identities, from queer histories to homosocial bonds among knights or ladies-in-waiting.
Recently there has been a welcome recognition of diverse stakeholders and their importance to participatory heritage to foster inclusiveness beyond experts (Kiddey 2018;Emerick 2014). However, heritage professionals still play a large part in how knowledge is shared, curated and accessed; it is important for them to be part of this democratising process. At the same time, we recognise that the Authorised Heritage Discourse (AHD) (Smith 2006) is deeply flawed: it is patriarchal, de-peopled and imbued with an unequal value system that prioritises a material heritage at the expense of others (Smith 2006;Högberg 2012). In relation to medieval castles, there is a risk that we may be complicit in (partly) perpetuating the AHD, by encouraging new gendered interpretations within the accepted management structures of heritage. Nevertheless, we believe there is value in trying to work with multiple stakeholders in order to present the past in varied and complex ways that show the 'messiness' of humans, their relations and emotions, as well as the wide spectrum of people at different points of the life course.
While we note that gradual change is happening within professional heritage practice, there remains much scope for including, challenging and changing gendered interpretations at medieval castles. This collaboration has highlighted that more active reflection is required on what we want from gendered interpretations and what assumptions we make in projecting a gendered gaze. We might ask the following questions of ourselves and current and future archaeological/heritage practices at medieval castles: (1) Do we want to examine only the gender identities of medieval people?
(2) Should we aim to demonstrate the spectrum of gendered identities and roles of people in the past? (3) Do we want to reveal the diverse range of gendered identities and roles of the medieval people in the past in order to share knowledge with those in the present?
(4) Or/And should we challenge visitors to consider how the gender identities of medieval people presented at castle sites may reflect or reinforce the gender ideologies present in the Western world today? (5) Is it acceptable in terms of authenticity to create biographical accounts from material culture to portray an historical approximation that does not always exist as a named figure in the historical recordsuch as the laundress or the female scribe?
What is very clear from the workshop and resulting article is that to encourage and achieve change requires collaborative working and sharing ideas, as well as listening to a public audience who express the desire to understand daily life in the castle beyond its martial façade.

Acknowledgments
Thanks to National Heritage Agencies and all participants in the workshop for their support, collaboration and feedback. This collaborative article was written as part of my Marie Skłodowska-Curie Fellowship, entitled, 'HeRSTORY', based at the Dept. of Archaeology, University of Reading, UK.

Disclosure statement
No potential conflict of interest was reported by the authors.

Funding
This project has received funding from the European Union's Horizon 2020 research and innovation programme under grant agreement No: 746406. Support was received from the Open Access fund at the University of Reading.

Notes on contributors
Dr Karen Dempsey is a Marie Skłodowska-Curie Fellow at the Dept of Archaeology, University of Reading. Her work blends the theoretical perspectives of prehistoric, historic and anthropological archaeology as well as material culture and heritage studies. She is interested in how present societies utilise past material culture to negotiate and legitimise their views or ideals, something that is very visible in the world today. One of the core themes of her research is that space and place are intrinsic to how communities make themselves; therefore, she believes that communities' engagement with cultural heritage enables people to make meaningful contributions to their individual and shared identities today. Her current Marie Skłodowska-Curie Fellowship focuses on feminist practice in archaeology as well as creating new gendered understandings of medieval castles with a focus on medieval women. Stefan Sagrott is an archaeologist in the Cultural Resources Team at Historic Environment Scotland. With a background in prehistoric archaeology, he finds himself increasingly caught up in the, sometimes confusing, medieval world (he is still in awe over the number of named people in the historical record). In the last year, Stefan has written the guidebook for Threave Castle, provided expert advice on the content for a range of interpretation panels across the national estate and undertaken a range of research, from geophysical survey and excavation to desk-and-archive based research. Stefan has a keen interest in ensuring that research does not just follow the status-quo and investigates poorly understood and under-represented groups at HES properties.
Samantha Stones is a Properties Curator at English Heritage and advises on the conservation and presentation of over 60 historic properties in the southeast of England. She has nearly 10 years experience at English Heritage and has studied history and conservation of the historic environment at postgraduate level. In 2015, she received the Institute for Historic Building Conservation (IHBC) Gus Astley award for her research.