Introduction

Excavation of a single inhumation grave on the isle of Bryher, one of the Isles of Scilly off the coast of Cornwall, at the Hillside Farm site revealed the unsexable and highly fragmented remains of one adult as well as a series of high-quality grave goods including an iron sword with a bronze scabbard and a bronze mirror. Though these items were uncommon grave goods, both were high-quality pieces of metalwork that appeared in mortuary contexts (as well as other contexts) across Britain throughout the Iron Age. These swords and mirrors were considered elite-status markers that indexed binary gender identities, associated with men and women respectively. Thus far, Hillside Farm is the only known burial with both a sword and mirror in Iron Age Britain.

This sword/mirror dichotomy is underwritten by the “two-sex/two-gender” model (Joyce 2008: 18): “the assumption that men and women constitute either/or—and the only normative—social subjects, with status as a man or woman presumed to be a primary and categorically coherent axis of identity” (Ghisleni et al. 2016). These binary expectations tend to mask diversity by reproducing either/or patterns in the data (see also Arnold 2016; Stratton 2016), binding “inquiry toward recognizing only certain kinds of persons, ways of being different, or processes of knowledge production” (Ghisleni et al. 2016). This problem of recognition is compounded in British Iron Age mortuary studies by the assumption that swords and mirrors conveyed equivalent social messages—i.e., elite rank—about the men and women they were buried with. Other potential points of social significance are homogenized or downplayed by fitting all instances of sword and mirror burials within this elite-status framework, whose terms and premises the Hillside Farm site challenges. Binding perceptions of this area of the southwest as a reified territorial unit have also created expectations that local mortuary practices articulated with the regional identity scale, masking potential ambiguity in mortuary practices and the expression of social identities. However, a greater diversity in patterning in the Southwest than has previously been recognized challenges a number of assumptions regarding the coherence of regional identity, as well as the implications of the region as a reified unit for interpretations of local practice and expressions of social identities (Cripps 2007; Fitzpatrick et al. 2007; Johns 2012; Nowakowski 2011).

This article reevaluates how to understand the terms of engagement within mortuary contexts—i.e., the categories and practices that entangle evidence and analytical options at all steps of the interpretive process, often resulting in exclusionary assumptions that bind the questions we ask and the answers we can find. The problematic assumptions that bind analyses of sword and mirror burials center around perceptions of these sites as pan-regional manifestations of a shared social organization structure; the significance of sex and gender in mortuary contexts as well as their relationship to each other; and the unfounded/oversimplified connection between binary gender categories, swords, and mirrors. In critically engaging with the mechanisms that artificially standardized interpretations of sex and gender within Iron Age British sword and mirror burials, the impact of these biases can be better defined. By reframing the terms of engagement, we can begin to reinterpret the Hillside Farm site in ways that do not dismiss variability or ambiguity and that avoid the artificial reification of binary normative traps by considering multiple possible points of significance.

British Iron Age Mortuary Traditions: Social Organization and Regionality Binds

Sex and gender assumptions are often entangled with not only binary binds focused on archaeological approaches to sex and gender but also other biases specific to the culture and archaeological context in which these topics are being examined. These biases can shape the specific manifestations of sex and gender binds, and in better understanding them, we can better plan specific approaches to avoid sex and gender biases in the future. Regarding sword and mirror burials, assumptions concerning the social organization and regionality within Iron Age Britain shaped the manner in which these burials were viewed as a mortuary tradition that crosscut regional cultural divisions on the island.

Until recently, British Iron Age social organization and cultural groups have been discussed primarily in terms of related/homogenous/static hierarchical “Celtic” societies and regional tribal identities and territories, generally thought to have emerged in the Late Iron Age (c. first century B.C. to first century A.D.) and later serving as the basis for the administrative framework of Roman civitates (Roman administrative groups) during the subsequent conquest in c. 43 A.D. (1984: 172; Collis 1981: 67; Cunliffe 2005; Edwards and Pope 2013; Harding 2016; Hill 1989, 2006; Moore 2011: 335–336; Pope and Ralston 2011: 375). Though culturally distinct, the shared Celtic tribal identity was perceived as evidence for homogeneity in the underlying social organization (Moore 2011). The identification of these regionally distinct but socially homogenous Celtic tribes was mainly based on classical historical accounts and later medieval texts discussing local tribal names and regional divisions. These perceptions were then buttressed by nineteenth century stereotypical conceptions of tribes that lingered in Iron Age and Roman archaeology (Edwards and Pope 2013; Hill 1989; Moore 2011). This vague conceptualization of Iron Age and Roman peoples remained mostly unchanged throughout the twentieth century, never clear on whether the term was being used to imply ethnicity, political structure, social organization, or territory (Cunliffe 2005; Fried 1968; Jones 1997; Moore 2011). Consequently, the potential fluidity and multifaceted nature of multiple cultural and personal identities within these groups as well as the potentially complex changes in identities and political structures that likely took place before, during, and after the Roman Conquest were largely ignored in favor of vaguely defined homogenous and temporally and geographically static perceptions of these peoples (Collis 1996; Cunliffe 2005; Fitzpatrick 1996; Jones 1997; Moore 2011: 342, 354).

The perception of tribes as “territorially coherent” cultures (Moore 2011: 342) also resulted in the association of tribal names and locations with distinct material culture distributions (such as ceramic traditions, architectural styles, or burial rites) identified within regions (Childe 1940; Collis 1996; Cunliffe 2005; Diaz-Andreu 1996: 55; Fitzpatrick 1996; Kossinna 1911; Moore 2011: 342). This connection between a tribe, their territory, and material culture evidence also became synonymous with culture, people, race, and/or cultural/group identities like ethnicity during the early twentieth century while simultaneously remaining generalized and only vaguely defined (Harding 2016: 19; Jones 1997; Moore 2011: 342, 354). Iron Age peoples were thus viewed as “as a mosaic of bounded monolithic ethnic or tribal units” (Jones 1997:31).Footnote 1 In this manner, identifiable Iron Age burial practices were equated with ethnic or cultural entities (Harding 2016: 19). This perception of distinct regional cultural divisions based on the same tribal social organization contributed to perceptions that mortuary practices would also demonstrate regional differences (with internal homogeneity) and pan-regional similarities.

During the pre-Roman British Iron Age (c. 800 B.C.–A.D. 43), human remains appear in formal burials—inhumations and cremations often identified as regional traditions based on their geographic distribution—in isolated sites or cemeteries, and in informal burial contexts, such as the placement of (often disarticulated) human skeletal elements in middens or ditches on settlement sites (Cunliffe 2005; Garrow and Gosden 2012: 194; Harding 2016; Wait 1985; Whimster 1981; Wilson 1981). While formal burials appear throughout the British Iron Age, distinct mortuary programs prevalent enough to be described as regional traditions emerge c. fifth century B.C., with some of these traditions lasting through the first century A.D. (Cunliffe 2005: 544–561). The identification of these traditions stems primarily from Whimster’s (1981) comprehensive synthesis on the topic involving the cataloging of every Iron Age burial identified up to that point (Garrow and Gosden 2012: 197).

The regional traditions identified included Durotrigian extended inhumations in southern Dorset; stone-lined cist inhumations in the far southwest (the region associated with Hillside Farm); the La Tène style square ditch and barrow inhumations of east Yorkshire (also called the Arras culture after one of the earliest type sites); the Aylesford Culture late La Tène style cremations in the southeast; and the pit inhumation tradition of central southern EnglandFootnote 2 (Whimster 1981). Whimster also noted that the universal preference for burial in a crouched position with the head directed to the north in the inhumation traditions was indicative of an underlying common tradition (Whimster 1981: 194).

These traditions, however, are not constant contemporaries throughout the centuries, but rather appear more as brief “pulses” (Garrow and Gosden 2012: 194) during the Iron Age within narrow geographic and temporal boundaries (Harding 2016). Whimster also noted that the total number of burials (both formal and informal) in almost all regions on the island, even within limited timeframes, was too small to account for the majority of the population (1981). This was especially obvious in areas with no visible formal burial tradition (such as Wales and Scotland). Consequently, Whimster and subsequent scholars theorized that regional traditions and the informal burials of settlement contexts were likely being utilized by only a subset of the population while the majority of the dead were disposed of using some “invisible rite” archaeologists could not identify (Harding 2016; Wait 1985; Whimster 1981; Wilson 1981). These invisible burials were presumed to involve one or more practices that would decrease the archaeological recovery of human remains such as excarnation, water deposition, or shallow burials difficult to identify on the landscape that had been destroyed in subsequent centuries. As the problem of the invisible rite was found across the island, it was also presumed that, whatever the cause, the underlying function of the unknown burial practice(s) would likely be the same across regions (Garrow and Gosden 2012: 13, 194; Harding 2016; Wait 1985; Whimster 1981).

In addition to discussions regarding regional traditions, informal burials, and the invisible rite, scholars identified what they considered to be another minority rite that crosscut the island, focused on marking high status and gender. The identification of this rite revolved around the presence of inhumation and cremation burials containing iron or bronze swords or mirrors (Collis 1972; Cunliffe 2005; Fox 1958; Stead 1969; Whimster 1981). First identified in Arras Culture sites of East Yorkshire, these burials were considered to be the interments of a particular subset of elite-status individuals, based on the high-quality craftsmanship of these items and the high overall material wealth of these graves (as they often contained other high-quality and/or exotic items such as metal containers, metal or glass jewelry, wagons, or in the case of swords, other martial equipment). Underpinning these assumptions was the fact that where skeletal sexing was possible, the earliest discoveries of Iron Age British graves with swords were found to contain the remains of biological males, whereas those with mirrors contained the remains of females (skeletal sex identifications came primarily from East Yorkshire Arras burials; see “The Sex/Gender Bind” section for more discussion) (Collis 1972; Cunliffe 2005; Fox 1958; Johns 2002–3; Whimster 1981: 129–146).

These perceptions were based on (1) the presumption that Iron Age societies were hierarchical and consequently status differentiation would be evident in mortuary contexts; (2) the notion that grave wealth was evidence of status (Collis 1981: 67; Hill 1989, 2006; Cunliffe 2005; Pope and Ralston 2011: 375); (3) the idea that swords and mirrors were markers of this status, as they were high-quality indigenous (excluding the few Roman examples) pieces of craftsmanship frequently associated with other high-quality and/or exotic items also reflecting wealth (Fox 1958; Garrow and Gosden 2012; Pope and Ralston 2011); (4) contemporary Continental and classical interpretations of the swords and mirrors (see below); and (5) modern sexist biases regarding gendered divisions in martial power (masculine) and the importance of beauty (feminine) (Arnold 2006; Collis 1972; Cunliffe 1996: 116; 2005: 556–557; Johns 2002–3; Giles 2012; Garrow and Gosden 2012; Marshall 2013; Moyer 2012; Pleiner and Scott 1993: 38–59). Consequently, the presence of swords and mirrors in mortuary contexts was also used to underpin the presumed significance of gender and (to some extent) status identities in Iron Age British mortuary contexts. Swords were linked to concepts of elite masculine warrior identities and martial power while mirrors were associated with elite feminine identities, cosmetic functions, and beauty (Collis 1972; Cunliffe 1996: 116; 2005: 556–557; Johns 2002–3; Giles 2012; Garrow and Gosden 2012; Moyer 2012; Pleiner and Scott 1993: 38–59). Consequently, these two items were determined to be equivalent gender and elite-status markers. In other words, this particular manifestation of the “two-sex/two-gender” model (Joyce 2008: 18; Ghisleni et al. 2016) placed swords and mirrors as binary equivalents that should never appear in the same mortuary context because their sole function as elite-status gendered personal items was thought to materially instantiate oppositional sex and gender identities that could never exist simultaneously within a single individual. As swords and mirrors were recovered at mortuary sites across the burial traditions of Iron Age England, it was assumed that these high-status gendered identities crosscut Iron Age regional divisions.

In Iron Age Britain, the practice of placing iron weapons in burials first appears in the second century B.C. and continues into the first century A.D. (Collis 1972; Cunliffe 1996: 116; 2005: 556–557; Johns 2002–3: 64–68). Initial interpretation of these burials identified the sword burials as likely an insular version of the widespread and diverse Hallstatt and La Tène Iron Age Continental European traditions of warrior inhumations, containing the remains of elite-status men and martial equipment (Arnold 2006; Collis 1972; Cunliffe 1996: 116; 2005: 555–559; Johns 2002–3: 64–68; Pleiner and Scott 1993: 38–59; Whimster 1981: 142–143).Footnote 3 On the Continent, swords are not the most common martial item in warrior graves, often appearing in only the materially wealthiest graves with larger war gear assemblages (Pleiner and Scott 1993: 38–59). Consequently, in this British iteration, the indigenous style swords (and sometimes associated martial equipment) were viewed as gender markers of elite-status males (Cunliffe 1996: 116; 2005: 556–557; Johns 2002–3: 65–68; Whimster 1981: 142–143). This argument was strengthened by the similarities between the East Yorkshire barrows and La Tène Continental burial practices in Gaul (Inall 2016; Pope and Ralston 2011) as this tradition contained not only the largest concentration of sword inhumations (predominantly associated with other materially wealthy grave goods) and, more generally, burials with weapons in Britain (Hunter 2005; Inall 2016).

With regards to mirror burials, the five wrought iron mirrors from the East Yorkshire area date to c. 400–150 B.C. and the rest are bronze mirrors from southern England dating c. early second century B.C. through the mid-first century A.D. (Joy 2010, 2011: 470–471; Marshall 2013). They too, particularly in the East Yorkshire sites, appeared predominantly with other high-quality grave good assemblages (Fox 1958; Whimster 1981: 144–145). Whimster is one of the earliest scholars to tentatively describe mirror burials as a corresponding female rite to the sword burials. However, he does acknowledge the more limited sample size of these burials and alludes to the general absence of Continental comparisons outside of the Classical world (Joy 2008, 2010, 2011; Moyer 2012; Whimster 1981: 144).Footnote 4 In this context, their presence in grave good assemblages was long characterized as a female personal item for cosmetic use and a marker of elevated socio-economic status (as they were costly to make in the Classical world as well). This is due partly to Mediterranean iconography and associations between women and mirrors in Greek, Etruscan, and Roman representations (though even here, in these contexts, it is not always clear how these mirrors were being used, and general antiquarian assumptions regarding gender roles that presumed that the connection between women and mirrors was focused on appearance and beauty tended to prevail [Moyer 2012; Pope and Ralston 2011: 404; Spivey 1991]). Mirrors in burials were consequently seen as one of a number of cosmetic grave good items likely to appear in high-status female graves, equated with male warrior burials as defined on the basis of weapons, especially swords (Cunliffe 2005: 557; Johns 2002–3: 68; Whimster 1981: 144). While the Greeks, Etruscans, and Romans in particular produced great quantities of metal mirrors, these seem to have been rarely imported to other regions of Europe for use as grave goods. Instead, it appears that locally made mirrors were preferred, but even these appear far less frequently in mortuary contexts than locally made swords (Moyer 2012: 6, 38–40).

As the number of sites identified grew, it became evident that specific mortuary practices associated with each of the sword or mirror burials were not standardized. While swords appeared exclusively in inhumations, the specifics of sword placement within the grave, associated grave goods, and/or placement of the burial in an isolated location versus within a cemetery varied both within and between regions (including those regions without identifiable formal traditions) to varying degrees. Mirrors also demonstrated much the same patterning in terms of spatial organization and grave goods. And while mirror burials have only been identified in regions with distinct formal traditions, they appear in both inhumations and cremations (Joy 2010; Whimster 1981).

In addition, subsequent synthetic discussions of sword and mirror burials have varied in the total number of sites and burials discussed based on the following: the authors’ perception of the minimum requirements for what constitutes reliable information regarding the presence of a sword or mirror within the same burial context—a particular problem in antiquarian excavations; perceptions regarding whether sword burials should be discussed in context with other weapon burials (see “Mirrors, Swords, and Gender Binds” section); whether the few sites located outside of Britain should be part of the discussionFootnote 5; and the inclusion of non-local style Roman mirrors (Garrow and Gosden 2012; Johns 2002–3; Joy 2008, 2010, 2011; Stead 2006). Consequently, burial totals within Britain range from 37 (Johns 2002–3) to 32 (Inall 2016; Garrow and Gosden 2012:132; Stead 2006) sword inhumations and 33 (Joy 2011: 469) to 19 (Garrow and Gosden 2012: Appendix 4) cremation and inhumation mirror burials.

Despite this diversity regarding the specific contexts surrounding these burials, the appearance of swords and mirrors across regional divisions (including areas without distinct regional traditions) as well as the association with gender and status prompted their discussion as pan-regional burial categories, as well as gender and status manifestations of the same underlying tribal social structure. Consequently, the initial investigations and subsequent interpretations of these materials within one region were often treated as viable explanations for all sword and mirror burials in Britain. In this manner, early interpretations primarily stemming from the East Yorkshire region were extrapolated to the entire island.

In his synthesis, Whimster notes that initially sword burials were only known from this region (1981: 131), and today the East Yorkshire region continues to be the most frequently studied with regard to topics of Iron Age social identities and mortuary practices, including status, sex, and gender (Edwards and Pope 2013; Harding 2016; Garrow and Gosden 2012; Giles 2000, 2012; Pope and Ralston 2011; Stead 1979, 1991). On the whole, this is not surprising, as the local environment is known for its chalk hills, the Yorkshire Wolds, which are excellent for archaeological preservation (Giles 2012). This environment, coupled with the general visibility of sites on the archaeological landscape and prevalence of cemeteries over isolated burials, makes East Yorkshire a well-excavated and well-studied area with strong evidence for a distinct cultural identity (Edwards and Pope 2013; Giles 2012; Pope and Ralston 2011). Consequently, synthetic discussions of the British Iron Age tend to heavily highlight this region in general discussions of mortuary practices and social identities of the deceased (Cunliffe 2005; Edwards and Pope 2013; Harding 2016; Pope and Ralston 2011; Whimster 1981).

The underlying assumption that patterns identified in East Yorkshire can be extrapolated to the entire island (whether intentionally done or not) is problematic for a number of reasons. Regarding the sword and mirror burials, as mentioned earlier, while the East Yorkshire region demonstrates some evidence of intersecting social identities involving sex and gender in the use of these items (see “Mirrors, Swords, and Gender Binds” section), these specific patterns of use within mortuary contexts do not extend outside the region. Furthermore, the emphasis on examples and interpretations stemming from a single region often limits the scope of discussion regarding the identification of general patterning in mortuary practices and social identities across the island, as well as narrowing the focus of preliminary interpretations for less studied regions lacking such a wealth of evidence (see “Re-Assessing Hillside Farm” section) (Brück 2004: 311–313). In addition, more recent scholarship examining the old assumptions regarding regional tribal divisions and pan-regional social organization of Iron Age Britain notes that regional and supra-regional scales cannot be presumed to be the meaningful markers of distinct cultural identities or underlying social similarities. Identities are both nested and scalar, in the sense that one may have local, regional, and supra-regional communal identities, with more than one being referenced in a given context (Giles 2012; Kirkham 2011; Lucy 2005; Moore 2011).

Relatedly, Harding (2016), in his updated synthesis of British Iron Age burial practices, has challenged the idea that there was a single regular and recurrent burial rite for the majority of the Iron Age population (e.g., the invisible rite) with a few minority rites reserved for unusual circumstances (e.g., informal settlement burials) or small subsets of the population (e.g., sparse regional traditions like the southwestern cist tradition). Rather, he argues, “that the British Iron Age was characterized by a diversity of practice in the disposal of the dead, not just between segregated and integrated disposal, but in the variety of practice within each” (2016: 290). In this manner, the identification of a particular burial rite that may predominate within a particular region in a limited time frame has led archaeologists to expect a level of standardization in formal burial practice that in fact does not represent the diversity of practices being used (2016: 267). The seemingly predominant practice of a single rite does not preclude the use of others. The presence of human remains within informal contexts such as settlements and the inclusion of the dead within formal cemeteries or single burials are all potentially well-established practices used regularly for specific individuals or circumstances (2016: 7–8, 267).

These reevaluations of regional cultural groups and mortuary practices make apparent that until we better understand the contextual and scalar parameters that govern Iron Age British mortuary practices, homogeneity in the expression of social identities (like gender) and the meaning of various grave goods (such as swords and mirrors) is an unfounded assumption rather than a stance supported by evidence. Consequently, the early identifications of regional mortuary traditions based on vague similarities may be masking more localized traditions with more varied meanings behind their use of items like swords and mirrors, especially considering the diversity within the mortuary practices these items are found within.

The Sex/Gender Bind

Connected to concerns about the reliability of contextual information surrounding sword and mirror burials were questions regarding the sexing of the skeletal remains. Specifically, these questions were in reference to the frequency (or lack thereof) in which physical anthropological methods were used to determine the sex of the individual and the reliability of the methods themselves. Concerning the reliability of skeletal sexing methods, the antiquity of some of the site excavations and analyses is also associated with the potential use of now outdated physical anthropological methods and/or a lack of standardization in methods used or description of the confidence level that can be ascribed to these determinations (i.e., female vs. probable female). Within modern osteological methods, skeletal sex is assessed via a series of quantifiable dimorphic metric and non-metric traits. Descriptors such as “probable” are explicitly linked to a quantified degree of certainty based on the manifestations of these traits (Buikstra and Ubelaker 1994; Claassen 1992; Geller 2005: 598–599; 2008: 122–125, 2009a; Sofaer 2006; Sofaer and Sørensen 2013; White et al. 2011: 408).

In the case of sword or mirror burials, however, some burial discussions were devoid of a discussion of the methods used to sex skeletal remains (particularly for antiquarian sites), raising the question as to what degree some of these sex assessments were based on reliable skeletal traits as opposed to presumptions regarding gendered use of swords and mirrors (this practice has been a common problem in early sex and gender analyses of mortuary sites generally; see also Stratton 2016). In his synthesis of sword and mirror burials, Johns goes so far as to note that five of the mirror inhumations were sexed as “assumed to be female” in the original site analyses (Johns 2002–3: Table 15), and multiple authors noted the lack of reassessment of skeletal remains from early excavations (Edwards and Pope 2013: 471; Pope and Ralston 2011).

In addition, descriptions of the confidence levels of sex identifications were also varied regarding the standardized methods and terminology used (or lack thereof) to describe the certainty of the sex determination (i.e., female, probable female, probably female, possible female, female?, ambiguous, unknown, ?) (Johns 2002–3: 65–69, Tables 14 and 15; Joy 2011: 474, Table 21.3). Consequently, the sex determinations for these sites cannot be considered a fully comparable dataset that can be used to support the early assertions of widespread correlations between skeletal sex and the use of swords and mirrors as grave goods. Much like the differing totals of sword or mirror burials discussed by scholars, this has led to differing ranges of reliably sexed burials. For mirror burials, totals range from the following: two to five female or probable/possible female cremations (Garrow and Gosden 2012: Appendix 4; Johns 2002–3: Table 15; Joy 2011: 474, Table 21.3); one possible male cremation (Johns 2002–3: Table 15; Joy 2010: 75, Table 10.2; 2011: 474, Table 21.3); and three to seven female or probable/possible female inhumations (Garrow and Gosden 2012: Appendix 4; Johns 2002–3: Table 15; Joy 2010: 75, Table 10.2; 2011: 474, Table 21.3). For sword burials, totals range from 15 to 30 male or probable/possible male sword inhumations (Garrow and Gosden 2012: 203; Appendix 4; Johns 2002–3: Table 14) and 1 possible female sword inhumation (Edwards and Pope 2013: 470; Garrow and Gosden 2012: 203; Appendix 4; Stead 1991: 205). While strong male identifications (all inhumations) have been identified in regions across southern England, the majority also comes from East Yorkshire. In addition, the only three mirror burials with a strong osteological identification of female (all inhumations) are from the East Yorkshire region, again suggesting that, if a correlation is present, it may be regionally specific.

However, within the totals discussed, it is worth noting that the single oppositionally sexed sword and mirror burials, namely, a “male?” cremation mirror burial from southeastern site of King Harry Lane, Grave 13, is from a Romano-British burial containing a Roman style silvered mirror (Johns 2002–3:70; Joy 2011:474; Stead and Rigby 1989: 278; Stirland 1989:243)Footnote 6 and “possible female” sword inhumation (also containing a shield) from the East Yorkshire site of Rudston, Grave 163 (Edwards and Pope 2013: 470; Garrow and Gosden 2012: 207; Stead 1991:205). Acknowledging the potential issues discussed above, the fact that these remains were osteologically assessed after the advent of modern techniques does strengthen the likelihood of their accuracy and suggest that reassessments of remains from early excavations may illustrate more oppositionally sexed burials than were originally identified, potentially strengthening or weakening these correlations (see “Mirrors, Swords, and Gender Binds” section for a discussion of complicating the sex/gender binary). Despite this variety in opinion regarding the number of reliably sexed burials, scholars have generally agreed that comparatively, the correlation between men and swords is a more substantiated assertion than that of women and mirrors (Garrow and Gosden 2012; Johns 2002–3; Pope and Ralston 2011). 

An issue related to the accuracy of sex determinations in these burials is the tendency of mortuary archaeology to treat skeletal sex as one of if not the most salient aspect of the deceased for understanding not only the individual but also larger cultural practices of sex and gender differences (Arnold 2016; Stratton 2016). Because the idea that sex is located within specific points on the skeleton is so entrenched (Claassen 1992; Geller 2005, 2008, 2009a, b), one of the primary foci of the analysis of skeletal remains within mortuary contexts is the identification of sex and its implications for gender identity (Geller 2005, 2008, 2009a). Within mortuary and bioarchaeological studies, the two-sex/two-gender model (Joyce 2008: 18) created the presumption that the osteologically defined binary sex categories of the deceased’s remains would correlate with a binary gender categorization, also expected to be present (Ghisleni et al. 2016). The corresponding gender identity of the deceased was assumed to manifest within some aspect(s) of the mortuary program, primarily through the presence of personal items associated with gendered activities, often presumed to align with traditional Western expectations of masculine and feminine gender roles (Doucette 2001; Weglian 2001). Consequently, gender identity was derived from osteological sex designations and mortuary materials consistently associated with these designations. Conversely, the presence of mortuary components identified as markers of gender identity was also sometimes used to extrapolate the sex of the individual when skeletal assessment was not possible (as was discussed earlier in this section; see also Stratton 2016). In this manner, sex and gender were often treated as equivalents within mortuary contexts (Arnold 2006; Geller 2008; Walker and Cook 1998).

The subsequent problematization of this view of sex and gender separated the two into “distinct domains with different ontologies and epistemologies” (Ghisleni et al. 2016). In this sex/gender system, sex was viewed as biological fact and gender as the cultural interpretation or symbolic construct of the body’s physicality (Bolger 2013; Geller 2008, 2009b: 71–72; Ghisleni et al. 2016; Oakley 1972; Walker and Cook 1998). Some scholars argued that this differentiation was better able to recognize the variety of ways in which cultures interpret anatomical differences (Ghisleni et al. 2016; Walker and Cook 1998). Subsequently, scholars have further problematized the “assumption of biological sex as a static and universally salient point of reference around which identity and variation revolved” (Ghisleni et al. 2016). These approaches highlighted the historicity of sex in noting the culturally contingent manners in which the body’s biological manifestations of sexual difference are discussed (Bolger 2013: 6–7; Claassen 1992; Geller 2005, 2008, 2009a; Ghisleni et al. 2016; Laqueur 1990; Sofaer 2006). In this manner, they articulate a concept of the body and osteological remains that are significant beyond fixed sex read from bones.

The points of significance articulated in the skeleton throughout the life course, including the discourses of sex and gender, are not necessarily the salient social aspects marked in the mortuary contexts in which these remains are recovered. The presence of skeletal remains that contain markers of biological sex (that may or may not be discernable to archaeologists) does not necessitate that any gender identity will be marked and/or visible in a mortuary context. This is not to say that the components of personal identities articulated in skeletal remains do not intersect with social aspects marked in mortuary practices, but that these intersections cannot be assumed. Sex and gender manifest in various and contingent ways in the skeleton and material culture throughout the life course and in the context of mortuary ritual; the critical aspects of the body and material culture, how these aspects articulate, and in what register of significance—i.e., sex, age, status, etc.—should be investigated rather than assumed. And yet, analyses of mortuary contexts often begin with the biological sexing of skeletal remains and the subsequent presumption that some form of patterning will emerge along these lines and that this patterning equates to either/or gender identities (Arnold 2016; Geller 2008; Sofaer 2006; Stratton 2016). In this manner, initial discussions of British Iron Age sword and mirror burial contexts also fell into this trap of assuming that the deceased’s gender was a significant component of the mortuary program based on the identification of skeletal sexes that aligned with assumptions regarding the gendered connotations given to swords and mirrors.

Uncritical adoption of assumptions such as these also creates “reductive correlations between objects, the body, and social identity” with the potential to artificially mask diversity in gender configurations (Ghisleni et al. 2016). This artificially binds gender variability to the biological constraints of human osteology, becoming naturalized and transhistorical. Consequently, the binary masculine/feminine options are the only gender configurations considered (Geller 2008, 2009a; Ghisleni et al. 2016). Thus, unexpected or oppositional correlations between gendered grave goods and osteological sex—i.e., a grave good gendered male and a skeleton sexed as female—were discussed as evidence of an exceptional non-normative individual rather than evidence of problematic assumptions regarding gender ideologies (Arnold 1991, 2006, 2012, 2016; Diáz-Andreu and Lucy 2005: 38; Ghisleni et al. 2016; Stratton 2016; see also the “Mirrors, Swords, and Gender Binds” section).

Mirrors, Swords, and Gender Binds

The assumed straightforward correlation between gender, elite status, and swords and mirrors as well as the status of swords and mirrors as equivalent in their social connotations have also been called into question upon closer examination of the mortuary evidence. As briefly discussed earlier (see “British Iron Age Mortuary Traditions: Social Organization and Regionality Binds” section), swords and mirrors appear in mortuary contexts that exhibit diverse patterning, challenging the assumption of a single, island-wide function as equivalent elite-status gender grave goods. While it can be said that both objects only appear in formal burials (as opposed to informal), swords are further limited as to only appear in inhumations, while mirrors appear in both inhumations and cremations. Mirrors also appear predominantly in cemetery contexts while swords most commonly appear in both cemeteries and single isolated burials (Hunter 2005; Joy 2010). In addition, swords have a wider distribution across the island on the whole, as well as specifically within burial contexts. While burials were the most common context for mirrors to be found, this was not the case for swords, which appeared most frequently in ritual depositional contexts (Garrow and Gosden 2012). The most common placement for swords and mirrors within burials (though not standardized and not ubiquitous) is near the remains of the deceased (Garrow and Gosden 2012; Giles 2012; Giles and Joy 2007; Joy 2010, 2011; Marshall 2013), which may support the idea that these items had some relationship with the person, though placement alone is not sufficient to confirm this (see below).

Both swords and mirrors, though commonly appearing with other types of high-quality items, never co-varied with any one particular grave good or set of grave goods. While swords most commonly appear with pieces of martial equipment (e.g., scabbards, spears, shields) in the company of skeletal remains sexed as male, recently, two East Yorkshire burials containing martial equipment (one with a sword and one without) have been identified as containing the remains of women (Edwards and Pope 2013; Pope and Ralston 2011; see below for further discussion). Mirrors rarely appear with other toiletry items (e.g., tweezers, combs) or grave goods more strongly indicative of the feminine gender in Iron Age Britain (e.g., glass beads, rings, bracelets), and as previously mentioned, comparatively few are associated with skeletal remains sexed as female (Johns 2002–3; Joy 2010; see below for further discussion of East Yorkshire).

The temporal, typological, and contextual divisions between the East Yorkshire mirrors and those found in southern England are also worth noting in this regard. As mentioned earlier, the five East Yorkshire mirrors date to the Middle Iron Age (c. 400–150 B.C.), are made of iron and undecorated, and are smaller than their counterparts. These mirrors appear in inhumations, and the skeletal remains associated with three of these mirrors have been identified as female, making East Yorkshire the only region to show any kind of proportionally high correlation between mirrors and women. However, these burials do not include any of the other personal ornaments generally associated with female graves in this region, but three contain the remains of carts, rare items generally indicating a special status for the burial (Edwards and Pope 2013; Harding 2016: 231; Pope and Ralston 2011). The bronze mirrors from across southern England (spanning multiple regional divisions) date to the Late Iron Age (c. early second century B.C. to mid-first century A.D.) and contain curvilinear decorations on one side. These appear in both inhumations and cremations, and whether grouped as a whole or within presumed regional boundaries, their relationship with skeletal remains sexed as female is proportionally much smaller and likely reliant on fewer skeletal markers (as the majority of these remains are cremations and consequently more fragmentary)Footnote 7 (Johns 2002–3: Table 15; Joy 2010). The differences exhibited by these two groups of mirrors suggest potentially different functions within mortuary contexts. The rarity of mirrors within or across regional boundaries also suggests that it is likely that few people ever saw more than one mirror burial in their lifetime. Consequently, the presence of a mirror in a grave does not automatically create any relationship with other burials with mirrors (Marshall 2013: 219). Rather, this association is the result of how archaeologists have approached analysis of these contexts. This rarity and lack of patterning is counter to what would be expected for items believed to mark ubiquitous binary gender and status identities (Joy 2010, 2011; Marshall 2013). This suggests that if a relationship exists between mirrors and women, it is not a straightforward correlation simply marking the gender of the deceased, but one in which other identities of the deceased and/or particulars of the burial are also involved (Harding 2016: 234–236; Joy 2010).

Though it has suffered from some oversimplifications in early scholarship, the association among swords, masculinity, and warfare is stronger than that of mirrors and femininity, but more internally diverse than originally acknowledged. As mentioned earlier (see “British Iron Age Mortuary Traditions: Social Organization and Regionality Binds” section), initial interpretations were based on the identification of warriors in early interpretations of Continental European martial burial traditions in which swords appeared predominantly in materially wealthy graves as part of larger suites of martial equipment, viewed as the burials of elite-status males. Martial burials containing fewer (both in quantity and variety) martial items including an absence of swords were considered to be warriors of lesser status (Hunter 2005: 56; Pleiner and Scott 1993: 38–59). Consequently, while interpretations of British Iron Age sword burials have always been entwined with wider discussions of the placement of martial equipment in burials, their primary function was assumed to be focused on the identification of elite-status warriors (Collis 1972; Hunter 2005; Whimster 1981).

Even when sword burials were discussed in conjunction with other forms of martial equipment, during these early initial investigations, it was assumed that the presence of martial equipment was sufficient evidence to designate the burial as that of a male who had actively participated in warfare (Collis 1972; Hunter 2005; Whimster 1981). However, much like the assumption that the presence of grave goods designated as masculine could stand for skeletal sex, the assumption that the presence of martial equipment in a burial could be used as sufficient evidence to identify an individual who had actively participated in warfare during their life has been debunked (Arnold 2006; Härke 1990, 1992; Stoodley 1999). This has been most notably demonstrated in archaeological contexts where young children were interred with martial grave goods they would have been unable to hold, let alone use (Härke 1990, 1992; Stoodley 1999). The identification of an individual as a participant in warfare requires more than interment with a weapon. Osteological evidence of martial-related traumas or muscular skeletal and skeletal markers potentially indicative of lived martial activities, such as archery or horseback riding, offers the best line of additional evidence that individuals buried with weapons may have used them in life (Arnold 2006: 145; King 2010). Such analysis should be undertaken to substantiate, rather than simply assume, warrior status. Though direct correlations between individuals interred with weapons and skeletal evidence of violent trauma are uncommon, this form of osteological evidence is visible within the larger skeletal populations in East Yorkshire (Giles 2012), the south central areas of Wessex (characterized by predominantly pit and settlement burials) (King 2010), and Dorset (containing an extended inhumation tradition) (Redfern 2008). In addition, Giles notes that in East Yorkshire, weapons appear more commonly in the graves of young adults, who would have been the primary subset of the population to be engaged in warfare activities, and it is possible they died from violent injuries that were not osteologically visible (2012: 168). While the evidence is suggestive of a need for martial equipment in the lives of British Iron Age people, it is not widespread and continuous, implying geographic and temporal shifts in the frequency of such violence and the need for individuals actively engaged in warfare.

The steadily increasing number of burials containing martial equipment has also revealed growing evidence for diversity in the funerary practices that created these interments. Currently, 80 martial burials can be confidently dated to the British Iron Age, with more than half from East Yorkshire (largely due to the Arras Culture tradition). An additional 22 may also be from this time period (Inall 2016: 44).Footnote 8 The amount and form of diversity suggests a possibly more complex and nuanced use of martial equipment beyond that of masculine martial identities of various social statuses. Martial burials appear in isolated find spots, clustered or isolated in cemeteries, and in some instances as the primary burial of a cemetery (Hunter 2005: 52). The weapons included were defensive (e.g., armor, helmets, shields) and offensive (e.g., sword, dagger, spear, ferrules, sling-stones). In addition, scholars continue to debate the possibility that some of these items may not have been signifying warfare or martial attributes as they were defensive (e.g., shields) or may have been used more commonly in hunting activities rather than warfare (e.g., spears) (Hunter 2005; Inall 2016). While there are no ubiquitous patterns with regards to the association between specific martial items in mortuary contexts, some trends have become visible. The majority of burials contain only one or two types of martial items. Spears, found in 36 burials, are the most common (Inall 2016: 44–49). In addition, at least 15 of the spear burials (possibly as high as 24) involved a “speared corpse” in which “one or more spears were thrust or thrown into the grave, sometimes piercing the corpse” (Inall 2016: 49). All but two of these came from East Yorkshire. Nine burials with similar spatial organizations contained a sword, spear, and shield. They are also located in regions (five from East Yorkshire) with marine or riverine Continental connections (Cunliffe 2005; Inall 2016: 44–46). With regards to the association of martial items and concepts of grave wealth, only the East Yorkshire barrows and the southeast cremations show any patterningFootnote 9 (Pope and Ralston 2011: 375–376).

This diversity of practice in the mortuary deposition of martial equipment without clear geographic, temporal, or contextual parameters suggests we are still missing some of the critical governing principles, some of which may not involve gender or status identities. Consequently, while masculinity, warfare, and weapons have vague, undefined connections in some of the patterning of martial burials, we cannot assume that gender and status are the primary governing principles of these burials or that the meanings of these items are homogeneous. The lack of equivalence between sword and mirror use in mortuary contexts and the lack of clear connections between these items and binary gender configurations speak to the need to address other recent additional considerations in the study of social identities in mortuary contexts. Gender identities do not exist in isolation, but with other social identities of the deceased such as “age, social role, status, and shifting identities linked to the life course of the deceased” (Arnold 2016). Gender configurations should be approached from a holistic perspective that references these other potential intersectional identities that may impact how these identities materially manifest (Arnold 2016; Giles 2012).

The intersectional nature of social identities is further complicated by the nature of the mortuary context itself. The deceased, like any grave goods (s)he is buried with, is an object interred by the living. It is they who create a burial. Consequently, it cannot be assumed that all grave goods were personal objects of the deceased or that their purpose in a mortuary context was directly related to their function among the living, as they could also be markers of community identity or other symbolic components of the funerary ritual (Arnold 2016; Giles 2012: 125; Harding 2016; Parker Pearson 1999: 84; Pope and Ralston 2011: 402). Assumptions that grave goods reflect aspects of the deceased’s social identity are undergirded by concepts of ownership stemming primarily from our own modern cultural context related to the display of one’s wealth, status, and identities through personal possessions (Brück 2004: 309). The role of grave goods within a burial and their relationship with the deceased must therefore be investigated rather than assumed (Brück 2004; Harding 2016: 13; Marshall 2013: 216; Pope and Ralston 2011: 400). The underlying implication of this is that even if sex and gender identities are being marked with swords and mirrors in some capacity, other factors such as non-binary manifestations of gender, other identities of the deceased, and other functions of these objects (see below) can intersect with these identities in manners not yet fully understood. In addition, the mortuary context is not just an arena for the lived or idealized social identities of the deceased and the community to play out. Aspects of mortuary programs can take on independent meanings, and grave goods can have ritual functions and salient aspects independent of the deceased in these contexts.

Recent analyses of gender configurations in East Yorkshire Arras Culture burials seek to address some of the complexities of social identities in mortuary contexts. Despite the use of East Yorkshire’s burials in the creation of early overly simplistic interpretations of sword and mirror burials generally, more recent interpretations have attempted to highlight the intersectional and contingent nature of gender identities (Edwards and Pope 2013; Giles 2000, 2012; Pope and Ralston 2011). For example, it appears that gender in mortuary contexts was not strictly binary and more fluid for high-status women. Men of a higher status could only appear with martial items, while women could appear with martial items, exotic jewelry, and grave goods linked to ritual ideologies (mirrors and spoons)Footnote 10 (Pope and Ralston 2011: 409). It is also worth reiterating that East Yorkshire contains the largest concentration of sword burials, a distinct subset of mirror burials, and the largest concentration of sexed human remains associated with these burials (Edwards and Pope 2013; Inall 2016; Giles 2012; Johns 2002–3; Pope and Ralston 2011). These distinct aspects of Iron Age East Yorkshire support its designation as a regional cultural group with distinct high-status gender configurations. It also highlights this region’s influence over the early discussions of sword and mirror burials more generally, resulting in the artificial focus on their function as high-status gender equivalent personal items, ultimately masking the diversity within these rites. The presumption that certain grave goods had to highlight certain identities of the deceased, and that gender, sex, and status identities were the likeliest to be marked, becomes a self-fulfilling prophecy. Assuming that these items were gendered personal items prevented scholars from exploring other options regarding their potential function within mortuary contexts or considering the possibility that gender and sex identities were marked in non-binary and/or non-equivalent manners or simply not marked at all within these mortuary contexts.

The initial connection made among swords, mirrors, and the deceased has also seemingly barred discussions of other possibilities with regard to their function in mortuary contexts. While the above discussions have investigated these assumptions as well as other possible connections these items may have had to social identities (real or idealized) of the deceased, they do not discuss the possibility that these items may have had qualities of their own being utilized in mortuary ritual. In the last 15 years, the scholarship surrounding complex metalwork items such as swords and mirrors specifically, and burial contexts in the British Iron Age generally, has grown considerably. Rather than presuming to know the function of such items in various contexts, recent approaches have explored various physical attributes of these items in discussions of object biography and materiality to better understand how their construction, appearance, and destruction can illuminate their function within various archaeological contexts (Garrow and Gosden 2012; Giles 2000, 2012; Giles and Joy 2007; Joy 2008, 2009, 2010, 2011; Marshall 2013). In the case of mortuary contexts, scholars have focused their efforts on discussing the new material insights of these items in conjunction with the specific social and material relations of their deposition.

The multistep process required to construct these complex items highlights the variety of potentially significant social, material, and conceptual connections among individuals and materials involved in their creation and subsequent use. The composite nature of swords, scabbards, and mirrors also highlights the potential for the replacement of various components. In particular, differing decorative styles and imperfect refits between sets of swords and scabbards paired in mortuary contexts signal objects with separate use-life histories being brought together specifically for deposition. The identification of physical wear, repair, and evidence of intentional breakage or reshaping prior to repair illustrates that many of these items obtained histories of their own prior to their final deposition, some of which appear to have lasted multiple generations. These histories and assemblages of attributes entangled in social relations act as potential points of significance that could be drawn upon in the final deposition of these items (Garrow and Gosden 212; Giles and Joy 2007; Joy 2008, 2009, 2010, 2011; Marshall 2013).

The visual components of swords and mirrors also offer insights into how these materials affected viewers and functioned within specific contexts. For example, the complex interplay of interlaced shapes found in the curvilinear decorative patterns on these items has been discussed in terms of capacity to draw viewers into a visual maze. Furthermore, swords, scabbards, and mirrors are capable of reflecting light, affording the potential to be visually powerful for viewers. They also have the capacity to show viewers their own, albeit likely distorted, reflections, something not readily available to Iron Age peoples except in water surfaces. The similarities between the reflective powers of mirrors and water surfaces, which were construed as liminal supernatural places during this period, have also been noted, highlighting the potential for mirror use in ritual and supernatural endeavors, including acting as a supernatural weapon intended to threaten, awe, and intimidate (Garrow and Gosden 2012; Giles 2012; Joy 2010, 2011; Moyer 2012).

Swords and scabbards, particularly with regard to their decoration, have also been discussed in a similar vein, with scholars noting that the decorative elements of swords, scabbards, and shields may have been at play in spiritual/supernatural warfare, where they may have acted as apotropaics or sources of supernatural power for the user, and in turn as hostile entities for an enemy. The designs of the Grimthorpe and Bugthorpe scabbard chapes have been discussed as stylized fish faces, also possibly suggesting a water connection (Giles 2012: 164–165; Stead 1979:61) These visual aspects can also be concealed in different manners, limiting access to who can view and/or access the visual qualities of the items. Swords can be contained within their scabbards, while mirrors can be wrapped or bagged by fabric, as is evidenced from the remains of textiles found still adhered to these items (Giles 2012; Giles and Joy 2007; Stead 2006).

Approaches such as these reveal the capacity of any element of a mirror, sword/ scabbard, or person to become a point of significance within a context, but no quality is inherently so (Joy 2010, 2011; Marshall 2013). It is the job of the archaeologist to investigate the “terms of engagement”(s) (Marshall 2013: 218) among materials and human persons within a specific context or action, rather than assuming them. To that end, some scholars have begun exploring other potential functions for burials exhibiting a variety of uncommon attributes such as isolation or primacy within a cemetery and the use of a variety of rare grave goods. These discussions have included particular sword or mirror burials. In their microscale approach to creating burial biographies meant to explore the specifics of each burial, Garrow and Gosden (2012) discuss the burial process, object and human histories, object-human connections, and connections beyond the grave. In their discussion of the warrior burial at Mill Hill in Deal, Kent (in the southeast), they note that the burial contains one of the largest collections of metal artifacts in Iron Age Britain (Stead 1995). In addition to an intentionally bent shield, the assemblage included a bronze headband; an iron sword; a wood, leather, and copper scabbard placed face down next to the deceased with stylistic and morphological differences suggesting that the sword and scabbard had not originally been made for each other. Overall, the stylistic elements of the metal pieces in the burial suggested they were from differing spans of time. The burial was also the first placed on the hill during the Iron Age, aligned with a Bronze Age barrow nearby. The site would then continue to function as a cemetery for the next few centuries. The heirloom aspects of the metal items and the inversion and breakage associated with some of them, as well as the primacy of the burial’s placement, indicate a function alternative or addition to simply identifying a warrior, possibly related to using or manipulating the warrior image to serve as a foundation or mythic ancestor for the creation of this cemetery (Garrow and Gosden 2012: 226–241).

Marshall (2013) adapts and builds on Brück’s (2004) approach of discussing burials as biographical objects that are also composed of objects, each with their own biographies as well as Joy’s (2010, 2011) object biographies of Iron Age British mirrors in her exploration of new approaches to mortuary analysis that avoids the reification of binary gender assumptions. Marshall advocates for an approach that moves from identifying the specifics found within individual burial contexts, to subsequently condensing this specificity into a generalized summation of patterns among associated burials, followed by a subsequent revisit to the specifics of each burial with a focus on reinterpreting them with the conclusions reached in identifying normative patterns (2013: 210). Consequently, this approach avoids beginning analysis with predetermined conceptions of if or how social identities will emerge and considers all points of possible significance. The marking of the deceased individual’s personal identities, then, is just one of a variety of options to be explored. In her discussion of the Portesham mirror burial, Marshall notes that the complexity of the mirror both in its attributes and creation point to a rich variety of potential relationships to be marked in the mortuary context. She also notes that Joy’s discussion of the material complexities of mirrors and their meanings in mortuary contexts highlights likely interpretations but never dismisses the possibilities of others. Consequently, “the possibility that the Portesham mirror was interred with an adult female emerges as one of many equally valued possible points of significance. The sex of the skeleton is not primary to analysis and does not precede and shape all other analytical possibilities” (2013: 219).

Moving Forward from the Binary Binds

Among the tangled web of binds discussed above, a number of recurring problems and solutions are visible. In more closely examining these trends, we can better identify the best solutions to avoiding the reification of these unfounded assumptions.

Not Starting with Sex and Gender

In developing new approaches to interrogate questions of sex and gender, we should, in fact, begin by not beginning with a search for sex and gender patterns as evidence of male and female dichotomies (Ghisleni et al. 2016; Marshall 2013; Stratton 2016). In framing our questions and methods around the assumption that these particular identities will be present and visible, we reify binaries and fill them with our own perceptions of what these categories constitute. This then leads us to ignore the intersections of these identities with others and to miss the other potential points of significance within a given context, ultimately excluding the possibility of identifying difference before it can even be contemplated (Marshall 2013).

Beginning with sex and gender can also afford them primacy in mortuary analyses (whether intentionally done or not) that subsequently requires an a priori binary division of burials before additional analyses can be performed. Often, those burials that do not fit into these binary divisions of skeletal sex and/or aspects of the mortuary program deemed to mark gender are left out of subsequent analyses (e.g., age, health, grave goods, spatial organization). This exclusion or artificial division of a potentially large portion of burials from various analyses has the potential to create false patterns based on an artificially limited subset of data (Stratton 2016).

Regarding this discussion, our inability to determine if gender, sex, or any aspect of the personal identity of the deceased is being marked in a particular manner, with particular materials, and/or within a particular context, should be treated as a challenge to engage in different analytical approaches that do not presume these aspects will be present. In this way, we are open to context, complexity, and nuance in our narratives rather than reifying the normative patterns that we think we already know. We must instead unbind our thinking and allow patterns that may or may not primarily be organized by sex and gender identities to reveal themselves in our analyses.

Maintaining All Points of Significance

It is important to consider that not only swords and mirrors but also potentially any material aspect of the burial, including the skeleton, could have been drawn upon as points of significance within the mortuary context, but no object or aspect is automatically predisposed to be more significant or socially salient than others (Marshall 2013; Sofaer 2006). “There is no necessary correlation between the burial of an object with a specific human body and the use, meaning, or personhood of that object or human being during their lifetime” (Marshall 2013: 216). Considering all possible points of significance when analyzing mortuary contexts, it is also worth remembering that any aspect can also have multiple meanings, and the discernment of patterning along a particular line of inquiry does not preclude engagement with other lines (Marshall 2013).

Unlike binary gender and sex paradigms that propagate and are propagated by a host of assumptions down the line and, in the process, preclude other kinds of narratives, the more recent analyses focused on various approaches to burial biography discussed earlier illustrate that the identification of one form of engagement does not prohibit the potential for the presence of others (Garrow and Gosden 2012; Giles and Joy 2007; Harding 2016; Joy 2008, 2009, 2010, 2011; Marshall 2013). Approaches such as these highlight a number of significant considerations applicable to reframing the terms of engagement with sword and mirror burials in ways that do not dismiss variability or ambiguity, avoiding the artificial reification of binary normative traps. In this manner, multiple interpretations, often intersectional, can be drawn out from individual contexts. Archaeological analysis can move between a wider and more general discussion of sword and mirror material aspects and patterning across depositions to the specificity of a particular burial context. Analysis comes full circle, from the consideration of a wider pool of possibilities for the generation of personhood—object or human—to the specific depositional context of a particular object/burial/person (Marshall 2013).

Maintaining Ambiguity

The acknowledgment of variance and the maintenance of ambiguity in interpretations can help recoup a sense of complexity, contingency, and context. Only through the open recognition and preservation of the ambiguity of data can we identify new areas of investigation and possible points of significance that can lead to more robust understandings of archaeological contexts (Gero 2007), particularly in areas such as the study of sex and gender, which have long been plagued by binary binds. Evidence that will not yield to a normative defined by a priori patterning should be honored, not erased. The firm facts that archaeologists strive for should be attained through appropriate “mechanisms of closure” for the context in which they are working (Gero 2007). “Every phase and feature of archaeological research requires archaeologists to make difficult or even impossible interpretive decisions on the basis of incomplete, unfamiliar, indeterminate or bewildering complex evidence…Most often, the confusion, uncertainty and ambiguity are left out of our conclusions, overlooked, ignored, forgotten or erased” (Gero 2007: 312).

The evolution of interpretations of sword and mirror burials has benefited from the recognition of ambiguity buried within over-simplistic and unfounded assumptions that plagued these sites. Early interpretations suffered predominantly from methods that masked ambiguity in what Gero describes as the cleaning and stretching of data. Cleaning artificially reduces ambiguity by dampening the variance within data “using a set of semantically broad but conceptually limiting categories into which all evidence is accommodated” (2007: 320), obscuring an unspecified amount of variation and creating the illusion that “homogeneous meanings hold across different times and contexts” (2007: 320). Data stretching “means finding the general in the specific, or generalizing to make sweeping claims from specific cases” (2007: 321). In this practice, a “plausible functional or causal relationship in a local context is said to hold widely or to stand for a larger class of interpretations where the same relationship would hold” (2007: 321). Rather than crafting over-generalized statements that allow vague similarities to be construed as robust patterns in support of outdated and unfounded assumptions, the maintenance of ambiguity allows scholars to avoid reifying false patterns and points the way toward new approaches to analyses. Maintaining ambiguity “preserves options against such time when clarity may be better achieved” (2007: 323), acknowledging that a strong interpretation may not currently be possible. With the growth in interpretations that embrace the potential for multiple points of significance and multiple interpretations, this also opens up the possibility of exploring different ways of examining the data to answer a wider variety of questions.

Reassessing Hillside Farm

In acknowledging ambiguity, not beginning assessment of the burial with presumptions about the demarcation of sex and/or gender, and being open to all potential points of significance, a number of salient points regarding the Hillside Farm site and burial practices in the southwest come to light (Fig. 1). The cist inhumation grave at the Hillside Farm site on Bryher, Isles of Scilly, revealed the fragmentary remains of a crouched (flexed) individual lying on their right side with the head pointed to the north. The cist was oval in shape and composed of boulders, consistent with the local style found in the islands (Porthcressa type—see below). While the human remains were too fragmentary to be sexed or provide a viable DNA sample, molar wear patterns indicated that the individual was likely in their mid-20s (Johns 2002–3). While grave goods are not uncommon in Iron Age southwest Britain, the number, variety, and quality of items from this grave make it the materially richest grave of the period in this region (Johns 2002–3: 70–72). These items included an iron sword, bronze scabbard, sword belt ring, shield fittings, brooch, spiral ring, a shattered unidentified tin object, fragments of animal skin and woven textile, and a bronze mirror, all in close proximity to the body (Johns 2002–3: 1). The scabbard and sword belt ring were on the west side, and the scabbard had been placed face down in the burial while the sword hilt was to the north near the belt ring, just south of the mirror. The undecorated reflective mirror surface was turned toward the face of the interred individual while the spiral ring and tin object were placed beneath the mirror, and the brooch was recovered from the area near the feet. The shield fittings were recovered primarily from the center of the burial (Johns 2002–3: 15–16). Thus far, it is the only known burial with both a sword and mirror in Iron Age Britain. In addition, it is the only sword burial and one of only two indigenous-style mirror burials within the presumed Iron Age tribal region associated with this part of the southwest peninsula (see below). Because of the inclusion of these two supposedly oppositionally gendered items, excavators also investigated the site for signs that the grave had been a double burial. The cist was too small to hold two people laid next to each other, and consequently if there had been two individuals they would have had to have been placed vertically. Neither the skeletal evidence (though highly fragmentary) nor grave good assemblages gave any indications that more than one person laid in the cist. Calibrated radiocarbon dating of the remains (200–45 cal B.C.) coupled with the typology of the fibula place the burial closer to the first century BC, near the end of the pre-Roman Iron Age in Britain (Johns 2002–3). Consequently, this makes the Bryher mirror the oldest decorated bronze mirror in Britain. In addition, this mirror contains unique motifs not found elsewhere (Joy 2008: 89). Nearby, a second (unexcavated) cist was discovered, suggesting the presence of a possible cemetery (Johns 2002–3).

Fig. 1
figure 1

Hillside Farm Site. Reprinted with permission from Charles Johns (2002–3: Figure 13): Plan of the cist interior showing the position of the skeletal remains and grave goods. 1 Sword, 2 mirror, 3 shield selected 3a corner knobs, 3b fittings shield binding clips, 3c studs, 4 brooch, 5 sword belt ring, 6 spiral ring, 7 tin object. Illustrations of grave finds not to scale

Full size image

Some scholars have discussed Hillside Farm only in general terms (Cunliffe 2005), cleaning and stretching the data (Gero 2007) in a manner that focuses on the site’s vague similarities to both the presumed equivalent sword and mirror traditions as well as the southwestern cist burial tradition. However, as the ambiguities in the social, geographic, and temporal parameters of the mortuary practices and the cultural divisions of the far southwest have become more visible (see below), scholars have embraced this ambiguity in their discussion of Hillside Farm as well and acknowledged that the development of a robust interpretation for the site remains out of reach while still discussing some of the unique or uncommon aspects of this site and positing multiple points of significance and interpretations (Harding 2016; Johns 2002–3; Joy 2010). Most notably, scholars highlighted various features that support the burial’s connection to ritual practices beyond the funerary deposition. The deliberate breaking of the sword, shield, and tin object at or prior to deposition could mark the ritual killing of these items, allowing them to leave the realm of the living. The placement of the scabbard face down is an inversion of the most common placement of these items elsewhere in Britain, in which the decorative side is facing upwards. Inversions of normative practices in burial contexts are often noted to function as separations between the living and the dead (Johns 2002–3: 71; Joy 2010; Parker Pearson 1999). Coupling these practices with the placement of the mirror’s reflective surface toward the deceased, Joy posits that these actions of ritual killing and inversion may have been taken because the mirror, “with its reflective surface capable of reversing the lived-in world,” may have also been viewed as a bath into a realm not for the living (Joy 2010: 64). Johns (2002–3): 71, also noting the supernatural and ritual functions of mirrors in other contexts, suggests that while the grave goods could be personal items from life, they could also be meant to serve some purpose in death, perhaps as magical/protective items for the deceased to use in the afterlife to protect the community (2002–3: 71). He notes a Roman account that describes the inhabitants of the Isles of Scilly as having knowledge of the future (2002–3: 70)Footnote 11 and points out that the site’s island location may be connected with the choice to place other warrior burials in high points on the landscape with views of the coast or island locations. In this manner, the “mirror could have been perceived as an extra weapon or powerful magic for the warrior in the Otherworld” (Johns 2002–3: 71). Harding (2016), also drawing on the site’s placement on one of the smaller inhabitable islands at the edge of the Atlantic, describes a possible site parallel. On Lambay Island, County Dublin (in the Irish Sea), several crouched inhumations were identified but destroyed before they could be properly excavated. Reports note the presence of a sword and shield covered by an iron plate, which may be the mirror that was also discovered there. Unfortunately additional contextual evidence to confirm the association of these items is lacking (Rynne 1976).

In discussions of gender, scholars noted that if the old presumed gender equivalent status of the sword and mirror items was at all accurate, the presence of the mirror in association with two martial items (sword and shield) could signal an inversion of the normative gender of the deceased or mark gender duality or deviancy (Johns 2002–3: 70). It has also been suggested that the Hillside Farm mirror may illustrate a contextually dependent, polysemic use for Iron Age items in which the appearance of a mirror with a sword may cause both objects to have a different meaning from when these items appear alone, transcending any gender associations (Harding 2016: 234) or any association with the social identities of the deceased altogether. However, there are often underlying uneven gender assumptions within some of these discussions. The Hillside Farm burial has been represented predominantly in the literature as a warrior buried with a sword and a mirror, with the mirror likely serving as a divination tool or supernatural weapon of some sort. Consequently, despite the fact that no mirror and martial item has ever appeared together in Iron Age Britain, the mirror has been viewed as a potential weapon, while the sword (and shield) are not discussed in terms of any non-martial interpretations for their presence. In these scenarios, the mirror is assumed to be polysemic while the sword is not. This is not to say that some aspect of a martial/masculine identity could not be at work, but that at this ambiguous stage in our archaeological understanding, solely discussing this burial as that of a warrior of some sort can potentially serve to simply close off discussion of other aspects of the burial or other potential interpretations. The rarity of both of these items throughout the southwest does not automatically suggest a correlation with sex, gender, or an identity shared by a significant portion of the population, and their placement together is enough to suggest the exploration of other possibilities.

Further discussion of the presumed southwestern cist burial tradition can also aid in contextualizing the Hillside Farm site and exploring other points of significance and ambiguities. This region of the southwestern peninsula generally and the mortuary landscape specifically have been plagued by the social organization and regionality bind that have impacted the interpretation of Iron Age contexts across the island. Classical sources identified the far west of the peninsula to be the territory of the Dumnonii tribe, encompassing modern-day counties of Cornwall (including the Isles of Scilly) and Devon (Cunliffe 2005: 201; Hencken 1932; Holbrook et al. 2007: 158; Rive 1964; Todd 1987). The stone-lined crouched inhumations identified throughout the early twentieth century in the southwest were dubbed the South-Western Cist Burial tradition and dated to roughly the fourth century B.C. to the second century A.D. based predominantly on stylistic assessments of grave goods (mostly bronze or iron fibulae) (Cunliffe 2005: 551–552; Johns 2002–3; Hencken 1932; Whimster 1981: 60–74). Located mainly in Cornwall and the Isles of Scilly with a few sites in Devon, the generally small cemetery sizes, isolated burial finds, and almost complete absence of children were interpreted as evidence that this was a practice for only specific subsets of the population (Todd 1987; Whimster 1981). Elite-status individuals were generally considered some part of this subset, based on general patterning identified elsewhere on the island as well as the limited identification of rarer items within some graves, including three bronze mirrors. Perceptions of this area as a unified tribal region created expectations that local mortuary practices articulated with the regional identity scale, masking potential ambiguity within the burial practices. More recent recognition of greater diversity across multiple lines of archaeological evidence challenges assumptions regarding the coherence of a southwest regional identity as well as the construction of social identities in mortuary contexts (Cripps 2007; Fitzpatrick et al. 2007; Johns et al. 2012; Nowakowski 2011). Most notably, it has been suggested that rather than a regional scale, society in the far southwest was organized at the level of household or community (Cripps 2007: 153). While investigation of this possibility is still ongoing, a potential shift in scalar focus could drastically impact our understanding of mortuary practices, especially since new patterns in burial practice have been emerging as new mortuary sites are excavated.

Currently, there are approximately 140 stone-lined cists from five cemeteries and isolated inhumations from the pre-Roman and Roman Iron Ages known in the presumed territory of the Dumnonii.Footnote 12 Late nineteenth and early twentieth century excavations are the basis for this presumed regional tradition (Cunliffe 2005: 551–552; Hencken 1932; Todd 1987; Whimster 1981: 60–74). Characterized by stone-lined cist burials constructed of local slate in cemeteries close to the coast, these single crouched (flexed) inhumations are generally oriented in a north-south direction and the grave goods used to date them mainly consist of personal ornaments (fibulae, bracelets, or pins of bronze or iron), rare metal items such as mirrors, and local pottery (Cunliffe 2005: 551–552; Quinnell 1986: 118–119; Whimster 1981: 60–74). On the Isles of Scilly, the Stone-lined Cist Tradition is also referred to as the Porthcressa-type Cist Tradition after the type-site of Porthcressa on the island of St. Mary’s. On the islands, cists were constructed of local granite instead of slate, are generally oval or rectangular in shape, and have been positioned in a north/south alignment (Ashbee 1954, 1979; Johns et al. 2012: 100).

A systematic synthetic discussion of all dimensions of the mortuary practices between or within sites has yet to be attempted, and little to no radiocarbon dating has been conducted on the antiquarian sites. In addition, the skeletal preservation is generally poor. Analysis instead focused mainly on the presence or absence of stone cists, a north-south body orientation, and the presence of Late Iron Age grave goods (Quinnell 1986: 118–119; Whimster 1977: 81; Whimster 1981: 60–74). Any briefly noted differences in the mortuary treatment within or between sites were considered simply a variant of the regional tradition. In addition, the primary type-site of this tradition, Harlyn Bay, is an exceptional rather than typical cemetery in this region with regard to its large size and evidence for mortuary practices not found elsewhere in the region (Quinnell 1986; Whimster 1977, 1981).

Overlooked diversity within the stone-lined cist burials as well as more recent mortuary excavations over the last few decades, including the Hillside Farm site, have produced evidence of a more complex picture of this area (Fitzpatrick et al. 2007; Johns 2002–3; Johns et al. 2012; Nowakowski 2011). Trethellan Farm in Cornwall, dated to c. 200 B.C. to A.D. 100, is the first un-lined pit crouched inhumation cemetery to be discovered (Nowakowski 1991). It is also significant because it contained the remains of men, women, and children, a more complete cross section of the population than the more typical subset (Nowakowski 1991). For many years, cists on the Isles of Scilly with highly fragmentary or no human remains were assumed to be those of adults, likely because this was the trend of cists on the mainland and those on the island in which skeletal remains were identifiable. However, the excavation of a child’s cist grave from the isle of St. Martin’s, Isles of Scilly (Churchtown Farm), and the discovery of another small cist with no remains (Lunnon Farm, St. Mary’s) has sparked a reexamination of the dimensions of cist graves across the Isles (particularly Ashbee’s smaller type II graves at the site of Porthcressa), and it has been suggested that upwards of 10 % may contain children (Charles Johns, personal communication). In addition, a recent cist burial from another area of Harlyn Bay, radiocarbon dating indicates, is of Bronze Age rather than Iron Age (Andy M. Jones, personal communication). This calls into question the presumption that all cist burials in this region can be dated to the Iron Age based solely on the presence of a stone-lined grave which is especially problematic for some antiquarian excavations lacking any typological or scientific dates.

Beyond contradicting assumed gendered patterning in mortuary contexts, the diverse set of metalwork recovered at Hillside Farm also contradicts the notion that this region was culturally poor throughout the Iron Age (Johns 2002–3). Coupled with the fact that more recent stylistic analyses of the decorative motifs of the Bryher mirror and the St. Keverne mirror from another stone-lined cist site, Trelan Bahow in Cornwall, have noted that some designs were distinct from bronze mirrors recovered in other regions, this suggests the potential for a local metalworking tradition (Joy 2008: 89; Johns 2002–3).

These recent contributions to mortuary research in this region highlight the need for more in-depth and systematic examinations of previously excavated materials to disentangle the possible significance of this emerging evidence of diversity from the vaguely defined cultural and mortuary parameters grafted onto this region based on normative patterns identified elsewhere in the British Iron Age. The Hillside Farm site is thus not necessarily a disruption of a normative/variant gender axis mandated by regional and cross-regional patterning. The site’s significance and place within the wider mortuary landscape is not yet known. Rather than shoehorning our interpretation of sex, gender, and identity into a standard pattern continuous with the scale of the regional ideal type, analysis should modulate the articulation and discontinuities among multiple possible intersections and loci of significance. In this manner, the ambiguous diversity of burial practices can be assessed to determine the likelihood of regional/localized burial tradition(s).

Final Thoughts

Ultimately, these are preliminary speculations on the Hillside Farm site, but they highlight potential future avenues of research to be explored in addition to any attempts to further investigate sex and gender configurations. Assuming that the patterning of any one burial context will accord with a particular scale of analysis might not only mask local variability, but also obscure the potential points of significance of the mortuary program, where an object’s meaning may have resonated beyond the sphere of sex or gender. Even if items have seemingly strong gendered associations in one context, the gendered component should not be automatically translated as primary, or translated at all, to another region or scale of analysis within a defined region. Beginning by not beginning with men and women (Marshall 2013) may allow us to open up possibilities that maintain simultaneous differences. Unbinding sex and gender from simplistic paradigms entails unbinding how patterns become constituted and how those patterns are related to one another at multiple scales and in multiple archaeological contexts. Hillside Farm as a starting point highlights the need to keep in mind the potential significance of any aspect of the burial, not to deny that men and women may have existed, or that certain artifacts may have conveyed a gendered meaning, but to refrain from excluding other routes to identity, personhood, or significance before analysis begins.