Non-Binary and Intersex Visibility and Erasure in Roman Archaeology

An issue being tackled in modern, particularly western, archaeology is the assumption that something intrinsic to a scholar’s own culture is an etic in space and time. While many insist our views of sex and gender in the modern world are changing, there has always been variation from the sex-based binary. This article explores intersex visibility in archaeology, and the potential for contemporary non-binary genders. This is achieved through analysis of the issues that prevent non-binary identities and intersex people from being identified in archaeology. This paper argues that gender cannot be explored accurately out of context, as it is strongly linked to the culture they live in and their life experiences. Additionally, the people who buried them might have had a different view of their gender and its presentation. Thus, this paper focuses on the issues of studying gender in Roman Britain, and in the ancient world.


Introduction Background
An increasing problem with western archaeology is the assumption that what we may consider intrinsic to our culture, are an etic throughout time and space (Joy 2011). While some may say our views of sex and gender in the modern world are changing, it can be argued that there has always been variation from the binaries of male/female, woman/man, which are assumed to be traditional values. Research from Redfern, Marshall, Eaton and Poinar (2017), along with Montserrat (2000) have shown the potential for intersex people or non-binary identities in the Roman world. This paper expands from a poster presented at the Theoretical Roman Archaeological Conference (TRAC) 2019 at the University of Kent. The poster reappropriated the works of Redfern (et al. 2017), as well as Cotton (2008) about the Harper Road burial to critique the way archaeologists think about gender in the Roman world, and the wider discipline. Decolonising and activism in archaeology were key themes of the 2019 conference, with the keynote given by a woman of colour for the first time. Dr Zena Kamash's keynote (forthcoming) focused on colonialism, and thus the lack of diversity in the field of Roman archaeology, a topic discussed in the poster and explored further in this paper, even though from a different perspective.

Biases
This paper has been written from the perspective of a White, genderqueer person, from a privileged workingclass background. As a White person, I have a distinct privilege over people of colour, especially transgender and queer people of colour who more often struggle getting into academia and getting their voices heard due to systematic racism (Atkinson et al. 2018;Akiwowo 2019;Ahluwalia 2019;MacDonald 2019;Kamash forthcoming). This privilege contributed to facilitate my access to university education, get funding for my studies along with the opportunity to publish this paper. As such, I have tried to cite a combination of traditionally academic and non-academic sources, and where possible, to cite sources written by people belonging to the group that is written about (e.g. women, Indigenous, trans, non-binary). The significance of such choice is further discussed in section the below, New Methods and Approaches.

Gender in Archaeology
The concepts presented in Figure 1 are often neglected in archaeological discussions of gender. There is a lack of recognition of individuals' self-identified genders being potentially separate from the contemporary public perception of their gender. This contemporary public perception of an individual's gender is much more visible to us today, and as such is considered more important. In his study of gender in imperial Rome, Montserrat (2000) theorises that some people may not have been considered 'male' or 'female'. Among the examples he gives are eunuchs and the Vestal Virgins, thus suggesting that procreation -or lack of it -was a key indicator of gender. Consequently, those who did not, or could not reproduce, were not considered a man or a woman (Figure 2) (Montserrat 2000). Montserrat's theory is congruent with Roman literature where women who wield power were called 'manly'. This is demonstrated in Seneca's Consolatio ad Marciam (Seneca, 2.6) and Consolatio ad Helviam (Seneca, 2.11), in which he discusses gender and ethical virtue (Wilcox 2006). Seneca identifies this as a masculine trait, even when women display it. However, this is not always the case everywhere in Roman literature, as will be discussed further below.
While unheard of in ancient literature, we should not assume transgender, genderqueer, genderbending, and non-binary were are modern concepts; ' a rose by any other name would smell as sweet' (Shakespeare,. However, different cultures may have had very divergent views of the same gender, including those that were not 'boy/man' or 'girl/woman', so applying modern or western terms to them is not always helpful (binaohan 2014: 109-21; World Health Organisation 2017; Monahan 2019).

Sex
'Sex' is the biological categorisation of living things based on certain physical characteristics. For humans, sex is generally determined by five main factors, usually observed in infancy or childhood: (1) the number and type of sex chromosomes, (2) the type of gonads (i.e. ovaries or testicles), (3) the levels of sex hormones such as testosterone, estrogen, and progesterone, (4) the internal reproductive anatomy (e.g. uterus), (5) and the external genitalia (Karkazis 2008;Gendered Innovations 2011).
Sex characteristics can sometimes be observed in other areas of the body, but these are the most prominent and consistent (ibid.). Sex is usually defined as male or female. However, 'intersex' people are born with sex characteristics not typically associated entirely with either female or male (interACT 2020). The Intersex Society of North America, on the topic of defining the intersex category, has highlighted the subjective nature of assigning sex by writing: 'Nature doesn't decide where the category of 'male' ends and the category of 'intersex' begins, or where the category of 'intersex' ends and the category of 'female' begins. Humans decide.' - (Intersex Society of North America 2008b) Among the potential developmental sex differences that intersex people can have is the Klinefelter's syndrome, where an individual has XXY chromosomes, and hypospadias, where the urethral opening is not on the tip of the penis (Intersex Society of North America 2008a; interACT 2020). These biological variations are currently estimated to occur in one of every 100 births (Intersex Society of North America 2008a;interACT 2020). People who alter their sex characteristics in adulthood are not considered intersex (Karkazis 2008;Intersex Society of North America 2008a;interACT 2020). However, this is a purely scientific definition of intersexuality. Influencers from the intersex online community have called for people not to only rely on this definition, but to also consider its cultural, legal, and political components across different cultures, communities, and political stand-points (Figure 3) (Hart 2016;Anick Intersex 2020). Discrimination against intersex people spans throughout history, including bodily mutilation without their consent, but nowadays a global activist movement is calling to end non-essential surgery on intersex children and ending stigma against intersex conditions (Intersex Initiative Portland 2003;Karkazis 2008). In many cultures, intersex people have been stigmatised, being perceived as cursed or unhealthy. Such views can not only lead to mutilation but also infanticide and abandonment (Karkazis 2008;Warne and Raza 2008;Lau 2015;Rohoda 2016). These persecutions have prompted many global human rights organisations to speak on the topic, including the United Nation Organisation which has ruled that medical intervention for intersex conditions should not be allowed until the individual has the maturity and mental capacity to fully consent (European Union Agency for Fundamental Rights 2015).

Sex in Archaeology
Similarly to non-binary and transgender identities, the possibility of individuals being intersex is rarely considered in archaeology. This can be explained by the absence of methods investigating intersex individuals through the examination of skeletal remains. Most commonly, the categories used in osteological sex determination are male, female, unknown or intermediate, and juvenile (Mays and Cox 2000;Bruzek and Murail 2006). While these categories are partially derived from the limited ability of osteological sex determination in fragmentary skeletal remains without aDNA analysis (Skoglund et al. 2013), they also erase the possibil- ity of identifying intersex sexes as there is no category for intersex people of any kind. More recently, some comparisons with skeletal determinations and aDNA results have led to identifications of possibly intersex people (Geller 2017;Redfern et al. 2017). However, in these cases the individuals are merely suggested to be intersex because their aDNA conflicts with the sex originally determined from their bone structure (Redfern et al. 2017). This idea will be further discussed below in section three. There are currently no methods of identifying skeletal remains as intersex. In recent years, there has been much discussion in forensics research of intersex people. Researchers, such as von Wurmb-Schwark, Bosinski and Ritz-Timme (2017), along with Dunkelmann (et al. 2019), have shown that intersex people are incorrectly sexed by amelogenin which identifies subjects' chromosomes but cannot detect polymorphisms of those chromosomes (von Wurmb-Schwark et al. 2007: 28-29;Stewart et al. 2017;Dunkelmann et al. 2019). While they do present trends in the way modern intersex people are treated medically and how this affects the results of amelogenin tests, this does not apply to ancient people who did not have the same medical treatments that are available today.
It is clear then that neither gender, nor sex, is universally binary, and it almost certainly never has been. As such, it is vital to look for these characteristics in order to break the White-western expectation of gender and sex in the archaeological record. The common practice of assigning one of two genders (or sexes) to anyone throughout time and space, especially for the study of an empire as vast as the Roman Empire, makes little sense. The visibility of intersex and non-binary people needs to be considered by archaeologists. As with any marginalised group, representation and normalisation in popular culture hold an important role in ending stigma and discrimination (Elbaba 2019; GLAAD Media Institute 2019). By including discussions about intersex and non-binary individuals in archaeology, their presence in museums, historical documentaries and dramas, and in our lecture theatres will be normalized. Presenting the differences in gender between cultures and across time in heritage media may change how many people view gender, especially for those who believe the world follows one White western idea of gender.

'Non-Binary' in the 'Roman World' 'Non-Binary' in Classical Sources
Rarely in academia have queer discussions of gender and sex in the Roman Empire happened without mention of Ovid's Metamorphoses. It is important to note here that Ovid's work is not the be-all and end-all of Roman interpretations of gender and sex. First and foremost, his works reflect a specific moment of Roman history, the early Imperial age. He also does not speak for all of the Empire's inhabitants, and certainly not for those who populated it throughout centuries. However, this article is as much an examination of our interpretation of the Roman world as it is of the past, and when attempting to find non-binary and intersex people in the archaeological record we must first consider where our research biases lie. Given the literary renown of the Metamorphoses, it would be a misrepresentation not to mention it here. 1 I will not use Ovid's works to challenge Roman archaeological data, partly because his representation of gender is somewhat incomprehensible by modern people. However, despite its limitations, Ovid's work demonstrates that gender in the Roman world went beyond the assumed binary. The Metamorphoses is a collection of stories about transformations, usually physical, based mostly on already-known myths (Feeney 2004: xiii). Ovid was contemporarily known as a ' deviant' due to the themes covered in his works. He resisted Augustus' 18 BC morality laws, in particular the lex Iulia de maritandis ordinibus and lex Iulia de adulteriis coercendis. These laws restricted people's sexual activity, for example prohibiting relationships between individuals of different social classes such as senatorial men and prostitutes (des Bouvrie 1984: 93). Ovid resisted these laws through writing his Ars Amatoria ('Art of Love'), which could be reduced to a guide to seduction (des Bouvrie 1984: 93). 2 His most controversial section suggested that Augustus' theatres were opportune places to meet rich, high-status women (Ovid, Ars Amatoria 1.89-1.135).
As the title suggests, the Metamorphoses contains several stories about gender switching and mixing in relation to physical sex characteristics, often in retellings of well-known myths. Perhaps the most famous story is that of Hermaphroditus who refused the naiad Salmacis' sexual advances (Ovid,). This ends with him losing his masculinity and becoming a semivir ('half-man'), being forcibly merged with Salmacis by the gods (Ovid,Met. 4.386). He gives up his masculine sexual control, and so loses it forever (Holmes 2012: 78-79). This representation of gender aligns with Montserrat's theorisation of Roman gender where Hermaphroditus, by refusing to use his power, is feminised ( Figure 2). However, in different versions of the myth, Hermaphroditus remains visibly a man, sexually male, and is always referred to with the masculine Hermaphroditus; he merely loses his 'masculinity' (von Stackelberg 2014: 395). On the other hand, Salmacis chooses to give up her femininity but has not become a powerful man. Instead, she becomes part of this semivir, who is not really a man/masculine, or a woman/feminine. This newly formed person is still referred to as 'Hermaphroditus' by Ovid, calling him a nati biformi, a 'son formed of two' (Ovid,Met. 4.387). It is hard to tell whether this is an improvement for her, or a loss, but Ovid makes it clear that this is a loss for Hermaphroditus as he sees himself mollita ('soften'), and compels his parents to curse the waters on which the transformation happened to him so that no one else will share the same fate (Ovid,. What is clear is that it is not clear: gender and sex are linked but not. Masculinity is a force that can be taken away if you do not use it. Femininity is the lack of force and power, so in becoming assertive femininity is reduced or lost but cannot be replaced with manhood. This idea is further exemplified by Ovid's telling of the story of Caeneus (Ovid,Met. 8.305;). Caeneus -originally Caenis -is a female character who is raped by the god Poseidon and then granted one wish as compensation. Caeneus' wish is to be a man and so he becomes. Caeneus is made impenetrable and thus unrapable and perhaps indestructible. Certainly, his skin cannot be pierced as is shown throughout the different stories (Ovid,). Again, masculinity is presented here as a strong force. Of course, Caeneus becoming impenetrable is a metaphor for no longer being able to receive penetrative sex, but there is no explicit reference to a change of sexual organs, as many male people were also sexually penetrated at the time. Many Roman queer sex theorists describe Roman sex as not so much between man and woman but between a top/dominant and a bottom/submissive (Walters 1997;Davidson 2001: 5;Williams 2010: 258). This dynamic has also been linked to the concept of gender. Alike Montserrat linking types of fertility to gender, this theory acknowledges that a male person in a sexual interaction can still be feminine if penetrated. However, it can be argued such a view has a modern homophobic bias, whether a conscious one or not. The idea of the strong penetrator and the passive penetrated implies that gay males who do not overcompensate with hypermasculinity are effeminate or even not allowed to be men, a common homophobic stereotype in the modern world (Hunt et al. 2016;Al-Kadhi 2018). The story of Caeneus presents a far clearer idea of gender that is quite linear and perhaps even binary. After living a woman's life, he became the ultimate man who enjoyed a better existence.
However, in a third 'genderbending' story I will return to these ideas of confusion once again. Iphis is an Ovidian female character (Met. 9.666-796) who is brought up as a boy because their father threatens to kill any girl born of him. Iphis grows and becomes of marrying age, before being paired up with a girl (Ovid,Met. 12.714). The decision sent Iphis into a panic that is relatable for many queer people today, where they lament nec equas amor urit equarum ('no mare loves mare') (Ovid,Met. 9.731), emphasise how a female cannot have sex with another female as no penetration can happen. Many interpret this as the lamentations of a young gay woman who feels too strange in a heteronormative world (Pintabone 2002;Kamen 2012); others regard Iphis as the Roman equivalent of a transgender man, or at least a female masculine person who wishes he could consummate his marriage and live as a full vir (Barish 2018;Judge 2019;Maisel 2019).
This conflict in interpretation is important to the overriding theme of this paper: we do not know what Iphis and his lamentations meant to contemporary people due to the -quite literal -death of the author. Gender is not consistent over space and time, and it is logical that a culture so far in the past would interpret gender in a way we may never be able to fully understand because we cannot witness it. Iphis may just be a woman who loved a woman, who also lived most comfortably as a man; a man who associated with lesbianism because of his physicality; or even just a metaphor for boys being semivir and becoming real men through character progression. This theorising discredits it as a representation of gender that is clearly valid within their own culture.

'Non-Binary' in Archaeology
The case study that brought the issues of this paper to the forefront was the Harper Road burial, originally assumed to be a woman until aDNA analysis showed their sex chromosomes as male (XY) (Redfern et al. 2017: 257-261). The individual's aDNA was examined as part of the 'Written in Bone' project, which led to an exhibition at The Museum of London (Redfern et al. 2017: 253-254). The potential for this person to have been intersex or transgender was discussed on social media and in several major newspapers (Geller 2016;Redfern et al. 2017: 257). Redfern (et al. 2017: 257) has highlighted The Daily Mail's article covering the Harper Road aDNA analysis. The Mail's headline, 'The Roman woman with MALE DNA', illustrates the confusing language used throughout the article (Griffiths 2015). For example, the journalist refers to the individual as a woman throughout, but also states that 'gender is determined by chromosomes', which directly contradicts their description of the individual as a woman (Griffiths 2015). While this kind of reporting of archaeological research is not unusual, it does highlight again the importance of archaeological discoveries on public opinion about gender and sex, as well as other politicised topics. Before Redfern (et al. 2017), the main publication discussing the burial was Cotton (2008) who concluded that the individual was a woman due to their bone structure and grave goods.
The excavation was undertaken in 1977 in Dickens Square, Merrick Square, Falmouth Road, and Harper Road in Southwark, London (Cotton 2008: 151). The excavation was intended to provide topographic information and training for student archaeologists. The Harper Road burial was one of two human inhumations found during the excavation. The subject was buried on their back with their arms by their side (Figure 4). The presence of iron nails suggests that they were buried in a wooden coffin (Cotton 2008: 152). The grave goods included a ceramic flagon at the head, a decorative neck ring, and a bronze mirror at the feet (Dean and Hammerson 1980: 20;Cotton 2008). There were also Samian ware sherds and pig bone but those were displaced during a 19th-century excavation and were not marked on the original excavation plan (Cotton 2008: 152). The subject's left foot was missing again due to this previous excavation (ibid.). Redfern (et al. 2017: 257) have concluded that the skeleton was likely that of an intersex woman, or alternatively a transgender woman. Either way, like Cotton (2008), the work carried by Redfern's team considers the individual to have been a woman, or at least someone who was considered to be a woman by those who buried them, and thus they were presented as such in the museum exhibition (Redfern et al. 2017).
The aDNA, mitochondrial DNA, isotope, and grave goods analyses brought to light much more context unlike the gender confusion found in the media. This individual lived through the conquest of Britain, the burial dating from AD 50-70 (Redfern et al. 2017: 257). Thus, their community's culture would have been at the intersection of native and continental influences, something that is arguably reflected in their grave goods (Redfern et al. 2017: 257). The grave contents included a decorative neck ring, a mirror, Samian pot fragments, and animal bones including pig bone (Figure 4) (Redfern et al. 2017: 257). Many of these objects were deemed native to Britain, and others identified as 'Roman' by both Cotton (2008: 154-156) and Redfern (et al. 2017: 257). The subject was determined to have been between 21 and 38 years old, and was a White European who likely grew up in Britain (Redfern et al. 2017: 257?;Cotton 2008: 154-56?).
In this section the burial goods from the Harper Road burial, what they can and cannot tell us about the subject's gender, will be discussed. Each item or category of item that was used by Cotton (2008) and Redfern (et al. 2017) to infer the subject's gender and status in their society will be further reviewed.
The bronze rectangular mirror is one of the main points of evidence used by those discussing the gender of the subject. With plain bevelled edges, it was likely encased by a wooden frame. None of the reports of the burial discuss any decoration on the mirror, leaving very little analysis of it to be made in terms of its cultural meaning and significance other than purely from its existence. However, we know this mirror was identified as a very popular form and examples can be found across the empire (Lloyd-Morgan 1981: 3, group a). Dean and Hammerson (1980) have suggested the mirror was produced in Northern Italy, even though the concentration of this mirror type in Nijmegen (Netherlands) could denote the existence of a workshop there as well (Dean and Hammerson 1980: 20;Lloyd-Morgan 1981: 3, group a).
For decades, mirrors in burials have been accepted by the archaeological community as denoting womanhood, though more recently archaeologists have questioned such interpretations (Joy 2011: 468;Jordan 2016: 871). Joy (2011) has concluded that, outside of East Yorkshire, Iron Age mirror burials had no strong association with women. He further suggests this reflects a preconception taken from the classical worldwhere Greek women in art are often depicted with mirrors -and our own modern conceptions of gender (Joy 2011: 275-276). The portrayal of these rare objects as purely a representation of feminine vanity is a representation of the marginalisation of women in Iron Age and Roman research (Giles and Joy 2007). Aside from the bone structure, the presence of a mirror is the main argument for the individual being a woman in both Cotton (2008) and Redfern (et al. 2017) works. The mirror was buried at the subject's feet which Redfern (et al. 2017: 258) has interpreted as a ' deliberate construction of identity through grooming' and a female activity. Redfern has additionally highlighted grooming practices as a way to present cultural affiliation, although she notes that it is unclear whether the mirror was used in an 'indigenous' manner or a 'Roman' one.
However, these generalisations are likely oversimplifying complex social constructs and cultural practices of a variety of Iron Age communities (Giles and Joy 2007). Joy's research is not without its issues, especially when solely looking at the sex of skeletons without considering that it may not 'match' the subjects' gender (Giles and Joy 2007). As is highlighted by this case study, the osteological analysis and aDNA can only provide the probable sex of the individual, and gender does not equal biological sex. Mirror burials are rare, particularly in Britain. As such, burials with a mirror connote high status, or the individual having an unusual role in their community, rather than any specific gender (Joy 2011;Jordan 2016).
Furthermore, the neck ring uncovered suggests this person held a specific role in their community, as its decoration is quite unusual, with a wreath of feathers or laurel leaves (Figure 5). Cotton (2008: 155) links the feather pattern to peacocks which connoted immortality in Roman religion and were associated with Juno, the goddess of female fertility and marriage. Cotton's assertion is of interest in regard to the nonfemale identification of the sex as a result of the aDNA analysis and therefore, the likely infertility of the subject, even though this interpretation has been hardly accepted. The identified feathers could be instead a victory wreath, mimicking the Roman military armillae found in Britain (Redfern et al. 2017: 260-261), thus supporting the theory that this individual was influential in their community. Redfern (et al. 2017) has agreed with the wreath theory, concluding that this neck ring was either an accidental or intended symbol of subversion of male military symbolism and apparent 'Roman' culture of the time (Redfern et al. 2017). Her research additionally suggests that the wreath testifies of the greater power and influence held by women in Roman Britain, especially compared to their counterparts in Italy during and after the invasion (see also Pope and Ralston 2011). However, in the context of Montserrat's theory on the gender system in Rome, the subject could have been classed as 'powerful' and thus 'masculine' (Montserrat 2000). If the individual considered themself to be Romano-British or similar, then it could be argued that they were considered to be some kind of non-binary gender outside of a man and woman definition. Montserrat (2000) further implies that people who did not procreate would have been identified to a third gender. This suggestion could correspond with our subject, as they were unlikely to be able to reproduce in a way typically associated with  Redfern et al. 2017: 260) women, assuming we discount Cotton's Juno torc theory (Redfern et al. 2017: 257). All of these theories are equally likely without further investigation into the cultural affiliation of the individual. Neither Redfern (et al. 2017) nor Cotton (2008) discuss the ceramics in the burial in much detail. However, the inclusion of Claudio-Neronian ceramics and Samian goods is another indicator of wealth and high status, as they were known to be repaired and restored more often than any other goods that were typically just thrown away when broken (Willis 2011: 171-180). It could be argued that the inclusion of a Mediterranean pottery type suggests the individual felt in some way affiliated with Roman culture. Redfern (et al. 2017: 258) has suggested this could signify an adoption of continental types of feasting. This is supported by the prevalence of Roman style feasting wears associated with client rulers' burials, native rulers recruited by the Roman invaders, in this period (Fitzpatrick 2007). However, we know that Samian and other foreign goods were already imported to Britain before the invasion, so it would not be that unusual for native people not associated with the Roman invaders to have made use of them (Willis 2011: 226-227). Another interpretation could be a continuation of Redfern's (et al. 2017) theory that the neck ring was a subversion of the Roman male ideal, taking a high-status Roman item to add to this symbolism. Nevertheless, as mentioned above, there is evidence that these Roman ceramics were associated with burials of client rulers who were either under the control of or working with the Roman invaders (Fitzpatrick 2007). Cotton (2008: 159-160) and Redfern (et al. 2017: 258) comment on a potter's stamp bared by all the ceramics with the name Vitalis. They suggest this to be a play on words for life and vitality, and thus underline the mourners had some understanding of Latin. However, it could have also been the name or the pseudonym of a potter, and so just a coincidence.
To summarise, the grave goods of this individual have been interpreted in various ways as belonging to a woman, or someone considered to be at least feminine by their community. Nonetheless, there are plenty of equally plausible interpretations that would suggest otherwise, or not suggest any specific gendered narrative. Little is known about the way sex and gender was seen in different communities across Britain, and so we cannot make confident conclusions about the gender of individuals before further research has been done on the subject. Even then, as discussed in the literature review on Ovid's heritage, we shall never understand fully the different cultural meanings of sex and gender of these communities as we can never be part of them. However, case studies such as the Harper Road burial raise important issues about how we should reflect on gender in the past, and what meaning that has for people in the present.

Issues and Problems
Rare are the archaeological analyses of remains and burials that includes the explicit consideration of transgender and intersex people like in Redfern (et al. 2017). This work was sparked by the apparent disparity between the aDNA and skeletal structure of the Harper Road person. Arguably, one of the reasons why the 'Written in Bone' exhibition was so impactful was due to the aDNA analysis being still relatively new and underused in representations of the past. It is therefore understandable that research projects explicitly discussing the possibility of an individual being transgender, non-binary, or intersex remain scarce. This is especially true when considering the lack of public knowledge of these concepts (Jones et al. 2018: 161-162). Kamash (forthcoming), after analysing all papers that have been presented at any TRAC (Theoretical Roman Archaeology Conference), has found that, overall, 47% of papers were about Roman Italy or Britain. This is representative of European Roman archaeology, especially theoretical Roman archaeology, meaning the most easily accessed case studies available focus on those countries and not the wider empire. Of course, a longer research project could investigate the burials of individuals that have not previously been considered to be transgender, intersex, or nonbinary, but doing so was outside the scope of this article. Such a project would involve in-depth analysis of many different burials until some suitable ones could be found, followed then by further research to legitimise the possibility of identifying someone as not having a binary gender or sex. I hope to be able to offer new answers by taking up doctoral research or another long-form project.
Much of the scholarship on gender in the rest of the empire have concentrated on the Roman view of 'Easterness' as effeminate, unlike a 'proper' Roman man (Icks 2017: 66). This idea -generally accepted as traditional -is discussed at length by Icks (2017) who compares literary depictions of 'Eastern' rulers with Roman rulers who have been described to act like women in some way, despite being known to be male (ibid: 67-78). Icks (2017: 77-8) finds a trend in which Eastern rulers are described as suffering from mollitia, a softness of mind and body, while Roman rulers are described as 'theatrical' and ' dramatic'. Icks then concludes that feminine rulers were considered 'un-Roman', aside from Roman rulers presenting as such for theatrical effect. This potentially supports the Montserrat theory of gender (2000) relating to power in Rome but neither Icks nor Montserrat explore this in other parts of the Empire, nor explore ideas of gender for its more ' ordinary' inhabitants who wield less power. This can be primarily explained by the state of the literary evidence which only presents the thoughts and ideas of writers and high-status people they discuss (Hallett 1997: 194;Roller 1997: 543). Hence, discussions of the 'non-elite' are often considered from the perspective of a more privileged person who has received an education (Hallett 1997: 194?;Roller 1997: 543?). Consequently, the euro-centric Roman research discussed above, have affected -and continue to affect -the scholarship on gender in the rest of the empire.
Compiling data from Google Scholar searches, most discussions on the Roman world and 'transgender ', 'transvestite', 'transsexuals', ' cross-dressers' and 'non-binary' or 'third gender' are based on classical literature (Figures 6 and 7; Table 1). Searches were made on the 17th January 2020 and results from the front page  of each search were counted. 3 Classical literature has thus been the natural starting point for archaeologists researching these topics, which unfortunately only adds to the eurocentrism in Roman research. While not all classical writers were based in modern Europe, most were, and almost all were, writing from a place of power or as a coloniser (Hallett 1997: 194;Roller 1997: 543;Beard 2017: 13). Redfern (et al. 2017)'s work on the Harper Road burial presents another issue: it is assumed that all people are cisgender, heterosexual, endosex (the opposite of intersex), heteroromantic, and fit the gender binary until something 'unusual' is found. In the case of the Harper Road burial, no one had considered the buried person to be any gender other than a woman until Redfern's research found out that they had 'XY' chromosomes, which in modern Western culture is taken to connote manhood or boyhood. This trend in archaeology that does not consider ' queerness' until something clearly unusual is identified has been made into a meme, a widely shared and adapted joke on the internet by the LGBTQ+ community (Figure 8).
LGBTQ+ people have used this meme to criticise heteronormativity in archaeological, historical, and classical research, and stress the mistrust and betrayal they feel towards 'historians' (i.e. any person working in heritage) perceived as continually erasing LGBTQ+ people from the past. The meme exemplifies the trend of only including LGBTQ+ identities in interpretations of that past if no other explanation can be found. On one tweet (Figure 8), a transgender person wrote xhe was going to carve 'trans' into xer bones so

New Methods and Approaches
It is clear, then, that a change of approach needs to happen if we want to improve our understanding of gender in the Roman world, as well as the past overall. Accordingly, the final section of this paper will explore methods and approaches that are beginning to be used to address these issues. To anyone reading this paper who is involved in academic decolonisation campaigning and research, it shall appear clearly that these issues are deeply rooted in colonialism. As discussed above in the section Gender, the assumption that cisgender and binary genders are the norm is due to colonialism and the suppression of the ' other' in both colonised and colonising countries (binaohan 2014: 109). Decolonising the curriculum, then, is surely one step towards this issue of erasure and eurocentrism. However, the key aims of the ' decolonise the curriculum' movement are to diversify the academy by changing what schools and universities teach, what is currently on their reading lists, and who is teaching and learning (le Grange 2016; Kamash forthcoming). While this change would be a positive one, it does not address the problem at its core. Heritage research -and the academy in general -is founded upon colonialism and racism, and by extension heteronormativity. Decolonising the curriculum does not decolonise the core of academia, and thus does not make any research into gender more accurate, as it still relies on the idea of 'normal' and ' other'. While we cannot dispose of the colonialist foundations of archaeology and heritage, as researchers we can become more aware of it. We need to not only know about colonialism in the discipline but think about it whenever any research is done. When analysing any excavation records, data, or academic writing, we need to ask where the colonialist foundations are. Normalising this practice of acknowledging colonialist foundations in research is one step towards studying gender more accurately and with less bias.
Additionally, we must acknowledge that a modern person can never fully understand a past person's understanding of gender or anything else, because a modern person cannot be a part of their culture or community. This can be achieved by avoiding absolute conclusions in research on gender, and collaborating with those in relevant communities outside of academia. Collaboration in a non-hierarchical environment is key, as until recently public archaeology has predominantly involved 'us' telling 'them', which has disallowed any perspective outside of the colonialist academy (Bollwerk et al. 2015: 179-180;Mitchell and Colls 2020: 32). Recently, Bollwerk (et al. 2015) has discussed non-hierarchical community collaboration techniques that come under the ' co-creation' umbrella. 'Co-creation' public archaeology schemes aim to create research, museum exhibits, and other work traditionally dominated by academics, with equal input from 'non-academic' people (Bollwerk et al. 2015: 180-82). Similar community-led projects are becoming the norm in public archaeology. For instance, Mitchell and Colls (2020) note that the Heritage Lottery Fund (HLF) grants are more frequently going to non-academic communities working on heritage projects (Mitchell and Colls 2020: 32). Both of these approaches are positive steps forward.
All of these are important steps that should improve archaeological discussions of gender and create more ethical museum exhibits that are meaningful to people. This could significantly improve the lives of people who are marginalised because of their gender and/or sex, and increase public understanding and acceptance. However, none of these methodological changes will reveal for certain the gender of someone who lived and died hundreds or thousands of years ago, nor what gender their peers thought they identified as, nor what it meant for them. Yet, it does not mean it is pointless to try to answer those questions, but it shows that we have to re-think complexly about gender, and never assume we have a definitive answer. Notes 1 All the stories from the Metamorphoses discussed in here have also been discussed by modern scholars through the study of statues or other media (Ajootian 1995;Graumann 2013;Stackelberg 2014). These are not covered in this article due to lack of space but should be acknowledged. 2 All translations by the author unless otherwise specified. 3 Searches were made on 17th January 2020, and results from the front page of each search were counted. Searches were for '"roman empire" "transgender"', '"roman empire" "transsexual"', '"roman empire" "transvestite"', '"roman empire" "cross-dress"', '"roman empire" "third gender"', and '"roman empire" "non-binary"'. The total results counted were 58, with 27 being counted as Classical or Historical, 6 as Archaeological, and 25 as ' other'. Searches were categorised as 'Classical or Historical' if they focused mainly on art and literature, or if they identified themselves as such in the text; as 'Archaeological' if they focused mostly on sites or artefacts, or identified themselves as such in the text; and as ' other' if they were not about gender in the Roman World and therefore not relevant here.

Acknowledgements
I am so proud of myself and my achievements that have brought me to this point in my career, my first published article. Less than a year ago, I could barely manage to attend a class due to disability and illness, let alone attend conferences and write papers. It is therefore very important to acknowledge those who got me here. Thanks so much to Dr Zena Kamash for guiding me through an academic world that feels like it wasn't built for me, for encouraging and platforming my ideas, and reminding me to rest and take care of myself. Thank you to D. Lewis for supporting me and caring for me so I could continue to do research in spite of my ill health. Thank you to the countless colleagues and friends online that answered my questions and cheered me on. Finally, many thanks to the guest editors, Sophie Chavarria, Karl Goodwin, and Philip Smither for inviting me to submit this paper, with special thanks to Sophie for guiding me through the process and having so much patience with me.