This paper aims to explore the potentials and limitations of gender and age assignment for archaeological material remains produced by excavations conducted in the period predating modern excavation techniques. The large numbers of artefacts (and their well-documented provenance) available from such ‘older’ excavations are a vast potential resource, but must be interpreted carefully and in line with modern archaeological principles, while conclusions based on dated theories or ill-founded assumptions must be discarded. Osteologically analysed remains from modern excavations, along with the associated grave goods, may also provide valuable data to which the older material could be compared in order to test the interpretative criteria and possibly expand the scope of what can be gleaned from older data.

The cemeteries used as a case study in this article belong to Colonia Iulia Emona, a Roman colony located in the easternmost part of Italy’s tenth Augustan region (Regio X), which lies in modern-day Slovenia (Šašel Kos 2002, 2003). It was built ex novo as an urbs quadrata and discussions about the date of its foundation remain, with opinions coalescing around either 30 BC or approximately AD 14 (Gaspari 2010; Šašel Kos 2012; Gaspari 2014). Tombstones indicate that at least part of the population came from Italy, as several notable Italian families were recorded, especially originating from Aquileia (Šašel 1968). The town flourished in the second half of the first century and in the second century AD, and continued to be in use until the middle of the fifth century AD (Gaspari 2014: 246).

Emona has three major burial areas, all of which are located near main roads, following Roman tradition (Toynbee 1996). As all the cemeteries are located under the modern town of Ljubljana, the capital of Slovenia, structures that once stood above ground are largely absent. The northern cemetery is the largest, while the eastern and western cemeteries combined are substantially smaller, by a factor of more than 20 in terms of burial count. No cemeteries are believed to have existed to the south since the area is too swampy. Parts of the northern cemetery were excavated on different occasions and include significant sites such as Titova cesta, Graisarjev travnik, and Lenarčičev travnik (Petru 1972), the eponymous Northern Cemetery (Plesničar Gec 1972),1 and Emonika-Potniški center (Mulh 2008). A list of all known excavations up to the year 2017 was compiled by the author of this article in 2018, relying inter alia on the summary of excavations up to 1960 provided by Petru (1972).

Of the over 3,000 known burials of Emona, approximately 80% are cremations and 20% are inhumations. The total number of burials is growing with ongoing excavations in 2018; however, these excavations have mostly unearthed later burials consisting predominantly of inhumations (see Stemberger 2018 for a full list).

State of Research and Proposed Approach

As in many other European towns, a large part of the archaeological work on Colonia Iulia Emona predated modern excavation techniques. The excavations from 1683 onwards (Petru 1972: 9) were documented with various levels of precision, with each subsequent excavation report generally more detailed as the discipline developed. The available excavation reports can be roughly separated into two groups: fully published excavations, comprising mostly those until the mid-1980s, and the more recent excavations available as preliminary excavation reports, but lacking details and illustrations. Consequently, osteological analyses of Emona’s burials are only available in a relatively small number of cases. The first and foremost source in this regard is the Potniški center sample of over 400 individuals, with approximately 100 osteologically analysed and published in the form of a preliminary report (Mulh 2008).2 A smaller sample was analysed from the site Štefanova ulica 4 with 20 individuals (Tomažinčič 2011) while a single male burial comes from the Komenskega ulica site (Slabe 1968: 420–421). Statistical data was published for the Kozolec II site, but without related grave goods (Tomazo-Ravnik 2011). Between these reports, most of the osteological material was derived from inhumations, which only constitute a minority of all burials. Furthermore, the descriptions of artefacts are rudimentary, meaning the material could not be used for establishing chronologies.

Both concepts, sex and gender, are in the Slovenian language typically expressed with the same word (spol). Accordingly, the majority of the available documentation rarely makes an explicit distinction. In the two foremost publications on Emona’s cemeteries, Petru (1972) and Plesničar Gec (1972), this ‘sex-gender’ was determined in the following manner: the skeletal remains were in some, but not all, cases classified according to size (for children) and gracility (for women). For the other age and sex-gender groups, comprising the majority of inhumations and all cremations, ‘sex-gender’ was established on the basis of grave goods. Specific artefacts were used to determine the sex-gender of the deceased. Generally, three categories were singled out: male, female, and child. While the first two were covered systematically, the latter only received sporadic mentions in the catalogues.

In the introduction to her catalogue of graves excavated between 1635 and 1960 Petru (1972: 16 and Table 2) briefly establishes the criteria for her determination of gender, which is based on ‘the works of Joachim Garbsch’ (Garbsch 1965). As such, she uses ‘pins, pendants, bracelets, and similar items’3 as indicators of female graves, while ‘one brooch, knife, spear head, axe, or other typically male items’ mark graves as male (Petru 1972: 16). On the same page, however, she notes that at least two graves contain both a mirror and a knife, of which the former she considers a female-related item and the latter a male-related item, and that nine graves with other ‘typically female artefacts’ contain knives as well. I find most problematic her interpretation that one grave belonged to a boy because it contained a golden loop earring, based on the claim that such earrings were worn by firstborn sons ‘to this day’ (Petru 1972: 16), for which she provides no corroboration.

The human bone remains from the older excavations (Petru 1972; Plesničar Gec 1972) also have to be approached with caution since bone gracility was sometimes used as the sole criterion to determine the ‘sex’ of the deceased. Notes on children’s graves are potentially more reliable, as these seem to have been singled out by excavators as ‘unusual’, especially in the reports collected by Petru (1972), in contrast to adults who were not systematically noted as such. In cases where the schematic illustration of the grave’s layout permits it, it can be observed that the skeletons of children were considerably smaller than average skeletons from other graves (e.g. Petru 1972: T.VI).

The results of such older interpretations need to be checked against parallels from more recent material culture studies, paying special attention to the role of individual objects (e.g. Martin-Kilcher 2000; Oliver 2000; Puttock 2002; Migotti 2007; Swift 2009, 2011). Various approaches to establishing gender and age on the basis of artefacts have been proposed and employed by different authors (e.g.; Böhme-Schönberger 1985, 1995; Brown 1993; Martin-Kilcher 1993, 2000, 2003). However, the approach outlined in this article and the one I used to interpret or, in certain cases, re-interpret Emona’s burials is based predominantly on life course studies. In addition to this, I rely on Van Gennep’s (1960) work focusing on rites of passage. Both of these frameworks are little discussed in Slovenian archaeology, and no attempt at interpreting the entirety of Emona’s published burials has hitherto taken them into consideration. Thus, I strived to provide as much external validation as possible for all of the criteria including, but not limited to, studies of Roman cemeteries from across the Empire as well as Classical literature and art. Additionally, I focused on whether finer distinctions can be made within the general gender and age categories. The limited osteological data is generally presented at the end of the respective sections.

The skeletal remains from the older excavations would be of great use if they could be re-examined. Indeed, a special volume focusing on osteological analyses was planned to accompany those by Petru (1972) and Plesničar Gec (1972), as mentioned in Plesničar Gec (1972: 12), but was never published in any form (with the exception of a handful of graves published by Wiercinska in 1978). The remains themselves are unfortunately also unavailable. On the other hand, osteological data from more recent excavations comes in the form of preliminary reports where sex and age were determined based on the measurements (e.g. for Potniški center) of the bones and overall osteological evaluation of the bone remains. While this method is also imperfect, nothing can be done until (or if) DNA sampling becomes the norm for human remains from modern excavations in Slovenia.

Burials of Children and Younger Adults

Child burials can be separated into three categories: 1) the so-called ‘new-borns’ 2) young children 3) children on the verge of adulthood. The distribution of the former two groups is presented in Figure 1. The first category is easily recognisable during excavations due to the size of the skeletons. However, for the excavations from the 1970s and earlier I would not presume that the individuals were necessarily new-borns, but they would likely still fit the modern notion of infants. There are several factors pointing towards this assumption. The first is the fact that the descriptions of the osteological material published in the 1980s (Petru 1972; Plesničar Gec 1972) were done by excavators and not specialists. While the archaeologists were likely knowledgeable on the subject, the excavations were carried out in a hurry (Plesničar Gec 1972) and the determinations of the ages of the deceased are probably better treated as estimates than established scientific facts. Supporting the view that these were preliminary findings is the aforementioned fact that a volume of osteological analyses of these burials was originally planned, but never produced.

Figure 1 

Potential burials of infants (yellow pentagon) and children (red circle) at Emona’s northern cemetery (burials from Potniški center are omitted due to incomplete spatial data).4

The cases of infant burials (Table 1) at Emona are not numerous and are, as a rule, poorly furnished. This seemingly supports the theory that Roman parents were indifferent towards their younger children as expressed also by Roman authors, who were, however, elite and male and are as such considered to have been biased to a degree in terms of expressing feelings since concealing them was considered a virtue (Harlow and Laurence 2001: 118). Despite the lack of grave goods, a certain degree of care is nevertheless expressed in these burials in terms of sheltering the bodies from the soil (Table 1), which contradicts the idea of total indifference and was also observed in other parts of the Empire (Golden 1990: 82; Pearce 2001). The varying treatment of children of different age groups was noted in Roman literary sources, and the variety of items buried with children was explained as related to tooth-growth. In Roman Britain, 18 months was usually the age at which grave goods were deposited in larger quantities (Philpott 1991: 323). The youngest group from Emona likely coincides with this age below 18 months, although it cannot be excluded that other factors, such as the psychological development of the child played important roles. To this group of Emona’s graves we can most likely add at least one of the graves with an example of so-called ‘bird flasks’ that are thought to have played the role of modern feeding bottles, with parallels, for example, from Slovakia (Kraskovská 1981: 16–17). While the glass specimen from Potniški center belongs to an osteologically determined infant, another feeding bottle from the Northern Cemetery is more elongated in shape (Isings 1957: 27, type 11; Biaggio 1991: 8.1.2) and comes from a relatively well-furnished cremation grave. As there were alternative uses for such flasks, for example as containers of liquids or possibly cosmetic powders (Isings 1957: 27; Whitehouse 1997: 122), I would not interpret this Northern Cemetery grave as that of a child.

Table 1

Burials of ‘new-borns’ and infants at Emona.

Grave Method of sex/gender determination Manner of burial Grave goods

NC 26
(Plesničar Gec 1972: 18)
Excavators’ observations Inhumation Four nails
TC 665
(Petru 1972: 64)
Excavators’ observations Inhumation under a brick Five nails
TC 677
(Petru 1972: 65)
Excavators’ observations Inhumation Glass beaker
LT 56
(Petru 1972: 97)
Excavators’ observations Unknown Pot
LT 64
(Petru 1972: 98)
Excavators’ observations Unknown /
GC 1508
(Petru 1972: 126)
Excavators’ observations Inhumation in a piece of preserved sackcloth, covered with two stone slabs /
PC 169
(Mulh 2008: 121)
Osteological analysis Inhumation in a wooden coffin, surrounded by tegulae Three nails
PC 179
(Mulh 2008: 126)
Osteological analysis Inhumation in grave pit /
PC 331
(Mulh 2008: 187)
Osteological analysis Inhumation in a wooden coffin Seven nails
PC 402
(Mulh 2008: 224)
Osteological analysis Inhumation in grave pit lined with stone slabs /
PC 353
(Mulh 2008: 203)
Osteological analysis Inhumation in grave pit /
PC 143
(Mulh 2008: 109)
Osteological analysis Inhumation in grave pit Bird-shaped flask

The second category of child burials includes burials that were also specifically noted by the excavators, but unlike the first group these cannot be characterised as infant burials. One of the persistent problems with the classification of such children (i.e. ‘non-infants’) is our perception of what it means to be a child. While in modern society, especially in the Western world, a person aged 12 would be considered a child, it was apparently not so in the Roman period, when someone of that age could be legally married (McGinn 2003). Appropriately, the matter is better approached from a sociological and cultural perspective than from a purely osteological one. As demonstrated below, the criteria are related more to social transitions, which are only partially related to physical transitions (e.g. reaching puberty) but are not fixed on reaching a specific age.

These ‘non-infant’ child graves can be further divided into two groups: 1) burials with apotropaic objects 2) burials without apotropaic objects. In general, it was believed in Roman culture that children, prior to entering adulthood, required protection, which was therefore granted to them in the form of magical objects; however, not all grave good arrays demonstrate this. For example, a group of 16 burials from across Emona’s cemeteries classified as belonging to children, in addition to the 15 osteologically determined burials from Potniški center, do not show evidence for any ‘protective’ artefacts. However, grave goods vary among these burials in both nature and number. The Potniški center burials follow the prevalent pattern on the site of nails and oil lamps as grave goods, while the 16 burials from other sites follow no discernible patterns, although many contained dining-related objects. The second group of ‘non-infant’ child burials, however, includes apotropaic objects such as bullae and lunulae, pendants shaped as the sun and moon respectively, and objects that fall under the category of crepundia, a heterogeneous group of objects with supposedly magical properties (discussed in detail below). Traditionally these objects are associated with children before completing the rite of passage into adulthood, which in Roman culture was marriage for girls and receiving the toga for boys (Dolansky 1997). Bullae are related to the latter while lunulae are connected with the former. It must be noted that even though both objects are typically associated with children, especially in Classical literature, they can in some cases be found in the graves of adults as well (Migotti 2007).

Apotropaic objects were not found in large numbers at Emona as only three lunulae5 and one bulla6 are known from graves there. While bullae and lunulae are the most obvious examples of protective jewellery, as other amulets were not found, additional artefacts could be included in the category of magical or protective items. They overlap with the concept of crepundia as discussed by Martin-Kilcher (2000): i.e. curious objects in terms of shape, size, material etc. Amber objects and miniature objects, albeit in Emona not made from lead as was typically the case at other Roman sites (Martin-Kilcher 2000) but rather in glass and ceramics,7 can be included in this group. Amber, which was considered to protect against evil and to possess healing powers, was found in various shapes, but in the majority of cases, it was worked into beads. These were found in graves either individually or constituting a whole necklace. Other outstanding items include amber rings, hairpins, distaffs, and small statuettes. There is a relationship between amber objects and child and, to a lesser extent, female graves, as demonstrated by a study of amber beads from graves across the Roman Empire (Swift 2003: 49–56; see also Allason-Jones 1996: 8–9, 15). The pattern of distribution of amber objects in Emona’s graves coincides with the findings of the abovementioned studies, as only one male grave (PC 27) contained a significant piece in addition to some small shards of un-worked amber (Mulh 2008). A similar pattern of distribution is presented by blue beads and necklaces made of blue glass, which are often found in women’s and children’s graves. In Emona, as elsewhere across the Empire, they occur in children’s graves much more often than in adults’ (Allason-Jones 1995). Although more analyses of wear patterns are needed to provide an answer, for now it seems that amber in the area of Emona follows the usage pattern of jet in Britain in terms of age, sex and gender association, social status, and magical properties (see Allason-Jones 1995: 29, 1996).

Other less secure interpretations of apotropaic objects include seashells and bells. In a few graves seashells and snail shells were deposited around the deceased’s body or buried in the ashes. Of particular interest is the use of shells as jewellery (remade into earrings),8 of which there is only one case at Emona. Literary sources attest that Romans believed that sounds had the power to chase bad spirits away (Martin-Kilcher 2000: 66; Luck 2006: 27). This is one possible way to interpret the bells from five of Emona’s graves9 as protective in nature, as proposed by Dasen (2009: 208; see also Eckardt and Williams 2018).

The third group of child burials from Emona — children on the verge of adulthood — stands out for the splendour of grave goods and, in certain cases, apotropaic objects (Table 2). While it is a small group of exclusively female burials, the best interpretation so far falls along the lines of the life course approach and argues that these were women who died around the time of marriage (Martin-Kilcher 2000; Oliver 2001). The exact definition of what ‘around the time of marriage’ means has been discussed in detail elsewhere (Stemberger 2014), but the burials are briefly summarised here and certain aspects are discussed for clarity. Typically, these graves are well furnished in terms of personal adornment and contain golden threads either as hairnets or presumably woven into the hair of the deceased. In several cases golden dust was found in the area of the head (in inhumations). Other furnishings, when present, do not form any observable pattern, but all of these burials must have been associated with considerable expenses considering the quality and quantity of grave goods, as well as the burial manner. They are not concentrated in any particular area, but are all located in the relative vicinity of ancient roads (Figure 2). Notably, however, not one of the burials from this group was near the early Christian church unearthed during the 2017–2018 excavations of a presumably late Antique cemetery also located along an ancient road. Furthermore, it seems that such burials, which likely have a pre-Roman origin as suggested by the so-called burial of the Iapydic bride (Teržan 2009), were found across the Empire, albeit in small numbers overall (Martin-Kilcher 2000) and also in smaller concentrations than at Emona. It is therefore possible that we are looking at a cosmopolitan group, as the burials were relatively uniform and recognisable across the Empire. Martin-Kilcher (2000) uses dolls alongside crepundia as her criteria for such graves. At Emona, dolls were clearly present in only one grave with two ivory specimens (Kuret 1984), although in another grave a crude ceramic head was unearthed which could also have belonged to a doll figurine (NC 659; Plesničar Gec 1972: T.CLII).

Table 2

List of graves with prominent jewellery (Stemberger 2018: 191).

Grave Manner of burial Base for sex/gender and age determination Associated artefacts Date

PC 31
(Mulh 2008: 46)
Stone chest, cremation Grave goods Bone hairpins covered with gold foil, an amber spindle whorl, a gold ring Unknown
PC 412
(Mulh 2008: 229)
Simple grave pit, inhumation Osteological analysis Three copper-alloy and one glass bracelet, five gold beads, two gold earrings, and one each copper-alloy, silver and ivory hairpin Unknown
TC 894
(Petru 1972: 83)
Inhumation Grave goods Silver bead, silver lunula, two copper-alloy earrings with seashells, silver bracelet, 2 golden earrings, necklace made of glass beads, earth mixed with golden dust Second century AD
TC 6
(Petru 1972: 120)
Stone slabs, inhumation Grave goods Silver cist, a gold necklace, four gold-plated hairpins, a gold ring, golden hairnets First century AD
LT 75
(Petru 1972: 99–100)
Internal construction, cremation Grave goods Comb and mirror, a golden thread, two silver hairpins, gold flakes, and small gold scales were discovered in the grave, which contained approximately 30 different objects Transition from first to second century AD
TuS 1489
(Petru 1972: 129)
Lead-lined coffin Grave goods, excavators’ observations A golden hairnet which covered four hairpins, 3,000 rectangular golden leaves, golden necklace with beads, two gold rings (one with an emerald), a seashell, a panther figurine and an amber fitting, a copper-alloy patera Transition from third to fourth century AD
(Petru 1972: 122–123)
Stone coffin, inhumation Grave goods, excavators’ observations Two dolls, an amber bracelet and a gold chain consisting of thin links Transition from third to fourth century AD
(Petru 1972: 122–123)
Stone coffin, inhumation Grave goods, excavators observations A necklace made of glass medallions, gold coins of Gallienus, necklace made of amber beads, gold, earrings Transition from third to fourth century AD
SVJ 1315
(Petru 1972: 119–120)
Unknown Grave goods An amber head, a golden ring, two golden earrings, and an oil lamp Unknown
TR 2
(Petru 1972: 118)
Stone chest Grave goods Two golden earrings, silver mirror, silver dress ornaments, glass urn Flavian period
SVZ 1341
(Petru 1972: 120)
Walled grave Grave goods Unguentarium, silver-alloy cist, two hairnets, golden necklace, golden ring, four hairpins Unknown
Figure 2 

Exceptionally rich female burials (red star) across Emona’s cemeteries.

Among examples of potential crepundia from burials that cannot be classified with any certainty, three miniature glass vessels were found in Graves TC 490, TC 789, and TC 791 (Petru 1972: 51, 71). In the latter grave (TC 791), a miniature jug made from blue glass was the only grave good. The only known parallel for this object comes from Poetovio, but in that case it was interpreted as a necklace pendant (Bertoncelj-Kučar 1979: 260). It is curious that these miniature objects were not found in Emona’s exceptionally rich graves, but appear individually in common graves, whereas in Italy (Martin-Kilcher 2000: 66) and the provinces (e.g. for Britain: Philpott 1991: 145–157, 233), such objects were usually found in lavish graves. A further three potential examples of crepundia come from unknown graves. These are all silver-alloy vessels, of which especially the bucket-shaped miniature resembles Eckardt’s (2014: 39) type B bucket pendants and Hansen’s specimen (1995: pl. 30.20).

Burials of Women

Emona’s female burials, in this study, were mostly determined by the presence of jewellery, hairpins, and mirrors. As soon as Petru’s criteria (1972: 16) are relaxed, e.g. knives are not considered as necessarily marking male burials, many more graves can be recognised as female graves than in the original publications. However, a certain amount of caution was applied when using typically female objects for the interpretation. I initially followed Martin-Kilcher’s (1998) criteria for the cemetery of Lago Maggiore. It can be expected that the stereotyped depictions on Roman funerary epitaphs that portray women as wool working mothers and dutiful wives (e.g. Hope 1994, 2001; Larsson Lovén 1998, 2003, 2007) should be reflected to a certain extent in grave goods as well. Therefore, I counted textile-working equipment in graves as female-related artefacts. Excavations elsewhere (e.g. the Lankhills cemetery) confirmed that spindle whorls can be safely interpreted as a female-related item (Booth et al. 2010: 245–274). Distaffs seem to have also played an important role in asserting status, and can potentially be interpreted as marks of motherhood (Pásztókai-Szeőke 2011), although in the case of Emona and other Slovenian sites bone distaffs were initially interpreted as large pins by Anja Dular (1979). Another artefact related to the female sphere, especially through its presence and role in various rituals, including marriage and burial, are mirrors. Similarly to distaffs, their appearance in graves is a new phenomenon during the Roman period in Slovenia (Stemberger 2018). Elsewhere in the Empire they were relatively safely interpreted as female items (Wyke 1994: 138; Shumka 2008; Berg 2010: 290–291; Allison 2015: 119–220), similarly to amber objects, although the gender association may be less clear-cut for mirrors since two cases at Emona are ambiguous.

Roman culture placed strong emphasis on the virtues of spinning and weaving as symbols of both motherhood (Pásztókai-Szeőke 2011) and status (Larson Lovén 2003: 65–66). There is a tradition of emphasising wool working in the domestic setting, even recorded for prominent members of the Imperial family such as Livia (Barrett 2004: 115), and a number of epigraphic, literary, and iconographic sources depict women working with wool (Larson Lovén 1998: 85). On the other hand, there is a distinct lack of comprehensive sources that would link women with wool working in terms of profession (Larsson Lovén 2013: 109). Textile working as an occupation was connected with slaves and freedwomen in the context of epitaphs (Larson Lovén 2007), i.e. with the lower classes. This is in sharp contrast with tombstones across the Empire that show women with distaffs alongside cups, mirrors, and other items which seem to indicate a comfortable life (Larson Lovén 2003: 66). Distaffs in graves as well as on tombstones could suggest that the woman was a wool working matrona attending domestic duties who would have had the luxury of not having to provide for the family’s livelihood according to the traditional Roman division of labour between the genders (Larson Lovén 2003: 65–66).

Jewellery and so-called hairpins are a mixed lot. The former I counted as female only in the cases of complex jewellery such as elaborate forms of earrings (e.g. a la baretta and a spicchio di sfera types),10 hairnets, and multiple bracelets. On the other hand, simple bracelets and rings were not considered to mark gender. Similarly, I considered only elaborately shaped ‘hairpins’ as female-related items, as simple pins could well have been used for fastening clothes. Despite Allison’s (2013) premise that certain types of brooches can be linked with male or female attire, I was reluctant to connect brooches with either gender, as the location of the brooches in relation to the body is unclear in most of Emona’s cases, and experts such as Hoss (2016) disagree on the subject in the case of e.g. flat brooches while Ivleva (2016: 69) states that brooches were worn by both sexes and by all classes. The only type of belt that can be relatively safely associated with women are so-called Norico-Pannonian belts, which are extremely rare in Emona with only three cases, but are much more common at another Slovenian site, Poetovio (modern Ptuj; Istenič 1999, 2000). So-called toiletry-related items were excluded from gender determination, as a close inspection of the sets demonstrated the dual role of e.g. spatulae and grinding stones. Toiletry items appeared in combinations closely resembling medicinal sets, in one case accompanied by a scalpel handle (Stemberger 2018: Section 7.3.4).11

Among the total of less than 200 graves determined as female on the basis of artefacts, there is a peak in the Flavian period that coincides with the flourishing of the town and the most expressive fashion trends (Stemberger 2014). These were less lavishly furnished in terms of jewellery, and often contained wool working items and mirrors. This contrasts with the cosmopolitan group from the previous section, which contains more jewellery and almost no items that would express any occupational affiliation. A further 19 osteologically determined female inhumations contain similar objects as the female burials identified on the basis of artefacts; amber beads, earrings, elements of necklaces (golden and glass beads), snails, and seashells. From dining equipment and other items, only nails, coins, and oil lamps, occur in significant numbers in these 19 graves.

Burials of Men

Men are generally much less recognisable in Emona’s graves than women. This is probably due to two factors. First, a relatively large proportion of grave goods linked with men, such as weapons, military belt parts, or military brooches, was dissociated from their original graves. A major contributing factor to this dissociation was the large typological exhibition that took place in the National Museum of Slovenia in the late 19th or early 20th century when the documentation for all exhibited objects was lost (Ložar 1941: 21–22; Gabrovec 1971: 40; Perko 2012: 28); however, many records appear to have been lost even before that. The second reason might lie in the way men were commemorated in the afterlife and the way in which the memory of the deceased was asserted. Since no particularly coherent group emerges from Emona’s burials of men, the discussion in this section can only be of a more general character.

Most obviously male items are weapons. They are rare as individual items, and even rarer in combinations. The most prominent array was published separately by Gaspari et al. (2015) and features an umbo, a gladius, a curved knife, and two spear-heads. Five graves contained equestrian equipment, and three hipposandals were found in unknown graves. The practice of adding weapons into graves ceased by the second century AD (Gaspari 2008; Gaspari et al. 2015). It is important to note that the graves with military equipment are of an early date, and according to Gaspari et al. (2015) related to auxiliary units.

Particular kinds of personal adornment are another potential marker of male graves. One of the most reliable indicators are military belts, confined predominantly to the fourth century AD. Military belts are generally a rare find in Emona, with only five certain cases — the most prominent being the utere felix-type from Grave PC 86 (Mulh 2012) and the belts from ST 18 (Tomažinčič 2011, 2018) and TC 130 (Mulh 2008: 103). Apart from belt strap terminals, belt buckles, and a few belt plates, other belt parts are less well preserved. In particular, propeller plates are the single most common military belt part at Emona’s cemeteries. The majority of (presumably) military belts from Emona are relatively simple and do not exactly correspond to any of the typical military types, making their typology uncertain. Of the more numerous non-military belt parts, only belt buckles are known from Emona (Petru 1972, T. XCIII 17–19, 22–29). Sometimes belt parts are accompanied by other objects like knives and brooches, as in Grave TC 320. Even though belt parts are commonly interpreted as signalling gender and age, this is not always the case as shown in the burial ST 18 (Tomažinčič 2011, 2018). This grave was initially believed to belong to a female, but later DNA analyses revealed the deceased was actually a 15-year-old male (Tomažinčič 2018). The belt from this grave was not worn, but was placed in the grave next to the feet, similar to Grave NC 299, whose sex is undetermined (Plesničar Gec 1972: 59). Specific types of brooches can indicate the sex of the deceased, especially crossbow brooches, which were associated with the military and with social status. Unfortunately, only a small number of them have been found at Emona, and of those at least 20 were dissociated from their original grave units.

Even among the 23 osteologically determined adult males from various modern excavations (Mulh 2008; Hofman 2009; Tomažinčič 2011), no patterns emerge. In fact, the grave goods associated with osteologically determined individuals show rather similar patterns between males and females, with artefacts mainly comprising oil lamps, coins, and nails (Mulh 2008; Hofman 2009; Tomažinčič 2011).

Burials of Mature and Elderly People

Even though Emona’s ager was surely home to many older people, even centenarians according to the epigraphic material (Šašel Kos 2006), these individuals are not traceable on the basis of artefacts alone. The only sure cases of mature burials are found in the preliminary reports for the Potniški center site (Mulh 2008), but even here there is a distinctive lack of the population classified as senilis. The data from Potniški center, which consists of 20 individual burials of mature women and men, provides no specific patterns, except for the lack of excessive grave goods, that could form a basis for comparisons with older data.

The seven mature female graves from Potniški center contained various dining objects (e.g. jug, bowl, plate), of which none appeared in more than one grave. Only a coin and an oil lamp were present in two graves. What these graves do show, however, is a complete lack of wool-working equipment. Jewellery is also scarce, represented by only one ring, two brooches, and one belt part, all from different graves (Stemberger 2018: 202; based on Mulh 2008). Of the 13 mature men predominantly from Potniški center all but one did not show any distinct pattern of grave goods. While the majority had nails in the graves, presumably the remains of the coffins they were buried in, only in two cases was an additional oil lamp found. Other items such as a ceramic tray, a copper-alloy button, and unidentified iron objects were the only items of their type in their respective graves. Several graves contained shards of pottery and glass, but it is unclear at this time if they belonged to backfill debris or if they were deposited in the grave intentionally (Stemberger 2018: 203).

This lack of grave goods can be ascribed to two factors. The first is that all the known graves from the group come from inhumation burials, which traditionally occur in large numbers in Emona after the end of the second century AD and are characterised by scarce grave goods in general. This poses a problem just as in the graves of adult men and women since in the absence of proper osteological data, such graves cannot be aged or sexed. The second issue is that at such an elderly age the person would have been past their peak in life and, in accordance with the theory of rites of passage (see Van Gennep 1960), approaching the ‘other world’, which can be considered similar to how children were not yet completely immersed in this world if they died young and were thus not in need of a standard burial (see Harlow and Laurence 2001: 39).


As expected, the interpretation of the data from older excavations was limited by several factors. The prevalence of cremations, which hinders the determination of sex and age even in modern excavations, and the lack of anthropological determination provided for the skeletal remains unearthed in the older excavations are among the foremost. Yet despite recent osteological analyses, the data from modern excavations currently provides a less useful base for comparisons than one would hope, primarily because the available preliminary reports provide only a selection of the data and no comprehensive publication of the recently excavated sites exists. It is likely that when thorough descriptions of all graves from these sites become available with the results of osteological analyses, more patterns will emerge. However, it is also important to note that Emona’s inhumations date to the third and fourth centuries AD, thus coinciding with a period characterised by fewer grave goods than graves from the first and second centuries AD, when cremation was more prevalent. Even without taking into consideration the variation of burial customs within such a long time span, there may prove to be little usable overlap between the older and newer datasets. Regardless, various new conclusions can be reached for all discussed groups.

Despite what the data seem to show, it can hardly be concluded that the population of Roman Emona comprised such a small proportion of elderly and infants. Reaching mature and old age must have been less common than in modern times, but the notion that the graves of the mature and elderly are nevertheless underrepresented at Emona is supported, if nothing else, by the epigraphic record (Šašel Kos 2006). For children too, it is highly likely that the identifiable burials are only a fraction of the actual number since it is estimated that 60% of children died before their first year and that approximately 31% of all burials should belong to this age group (Carroll 2012: 44). It is particularly unfortunate then that in both osteologically determined groups (i.e. children and the elderly) grave goods do not form any recognisable pattern that could be applied to graves from older excavations.

The apparent lack of child graves in the funerary record could be related in part to the fragility of their not yet fully formed bones, which are very prone to decay before the age of four and can decompose without a trace (Redfern and Gowland 2012: 114). The second reason may be that the remains of infants and younger children are known to have sometimes been buried intra muros. Although this practice was widespread in the Alpine regions, in Gaul, and in the Eastern Mediterranean (Carroll 2012: 45), to my knowledge only two such graves in Emona were found inside a building, but the building was located extra muros (Badovinac et al. 2011: 22). Additionally, the fact that child burials at Emona contained relatively few grave goods in general hinders not only gendering, but also the dating of these burials. It is known that the rituals concerning child burials were less elaborate than those of adults in the Roman world, and that there were also laws restricting and even prohibiting certain elements of Roman funerary practice being applied to the burial of children. This is confirmed by a parallel study from Britain which shows that children of six months and older had the same artefacts as adults, while younger children did not (Dasen 2009: 209). It is also possible that some of Emona’s burials of children belong to a later period — i.e. the third and fourth centuries AD — when the number of grave goods declines sharply in the majority of burials. This however does not explain why there are so few child burials from the first and second centuries AD, when Emona was flourishing, especially when taking into account Carroll’s (2012: 44) statistics. Plotting identified child graves onto a map showed that there was no special location reserved for these burials, but rather that the graves were scattered amongst the general population.

Child graves with apotropaic items of the type that emerged from the old excavation reports are, as of yet, absent from the more recently excavated material. The infrequent appearance of apotropaic objects could suggest that only a handful of parents would have considered that their children needed symbolic protection. On the other hand, it is also possible that protective items were made of perishable materials and were thus not preserved. Several other objects such as feeding bottles and bells potentially signal the presence of children’s graves. Perhaps more so than with the other groups discussed in this article, interpreting the excavators’ notes plays a crucial role when studying children’s burials at Emona. Especially in the grave descriptions published by Petru (1972), one can even trace individual styles of writing and, with help of illustrations, reconstruct to a degree the ‘internal typologies’ that were used by different excavators.

Graves identifiable as either male or female, osteologically or otherwise, amount to less than 300 graves out of around 3,000 total graves at Emona — i.e. less than 10% (Figure 3). Based only on grave goods, women’s graves are generally more easily identifiable, in line with Larsson Lovén (2003: 66) who found women’s social roles to have been more stereotyped in the Gallo-Roman epigraphic record. Thus, ‘the notion that ancient women left behind little trace in the archaeological record (Milnor 2005: viii) is completely out of tune with the grave assemblages of Emona where female graves are much more easily discernible [on the basis of artefacts] and thus appear in larger numbers than securely male graves’ (Stemberger 2018: 209). Concerning men, greater emphasis was put on professional affiliation at Emona, again echoing Larsson Lovén’s (2003) findings from Gaul. However, establishing any kind of identity in a burial, regardless of gender, relied significantly on the financial means of the people organising the funeral, which is related to social status. Wealth and status must therefore be taken into consideration when interpreting the funerary record.

Figure 3 

Male (green triangle) and female (red diamond) burials at Emona’s northern cemetery (burials from Potniški center are omitted due to incomplete spatial data).

For example, many of Emona’s female graves contain items of personal adornment or wool-working objects, both indicating wealth and the latter also suggesting that these women had the luxury of not having to provide for their livelihood. The analysis of the older material from Emona showed that there are at least two different groups of female graves. The smaller group, which arguably overlap at least in part with child burials, as discussed in Section 3, is characterised by more prominent artefacts of what can be summarised as items of personal adornment, e.g. jewellery and clothes, and a general lack of other grave furnishings. The deceased of this group are not concentrated in one part of the cemetery, but are scattered throughout and were, as a rule, buried in expensive containers, such as lead lined coffins, stone chests, and stone cysts. They seem to fit the profile of cosmopolitan elites, as similar types of burials have been noted in small numbers across the Empire (Martin-Kilcher 2000). Even though Emona has one of the highest concentrations of such graves, they were only recently noted and discussed as a coherent group (Stemberger 2014), although they were certainly previously noted due to their lavishness by the excavators. The second group of female graves is distinguished by objects like wool-working tools, mirrors, and dining related artefacts, whereas jewellery makes far less of an appearance. These burials align with the ‘lanam fecit’ representation found on tombstones (Larsson Lovén 1998); like the ‘cosmopolitan’ group they were only recently addressed as a group (see Stemberger 2014, 2018). The notion that women are ‘weavers of life’ and that distaffs especially must therefore be placed into their graves was an idea introduced into the area following the Roman occupation (Stemberger 2018: 310).

In the case of males, grave goods provide little basis for determining gender, with the exception of objects associated with the military. Contrary to the predominantly domestic role of women in Roman culture, public roles were strongly associated with men. Rather than through grave goods, it is likely men’s social roles and status were commemorated with monuments, funerary or otherwise, meaning that the approach proposed in this study is less suited to identifying male burials. The epigraphic record on display at the National Museum of Slovenia (Šašel Kos 1997) attests to occupations only in the case of men, whereas grave goods reliably indicating occupation are missing for both women and men. A conspicuous exception are male burials identifiable by military-related objects; they are concentrated in early graves with military equipment and late graves with belts and brooches. In contrast with what can be recognised as female graves, most of which date to the Flavian period (Stemberger 2014), identifiable male graves cluster at the beginning (transition from the first century BC to AD, and first century AD) and end of funerary activities at Emona (fourth century AD).

There are two more groups discernible from Emona’s material that are worth mentioning, although they cannot be aged or gendered. The first group is composed of relatively few graves containing tools of trade that presumably belong to adults, yet in most cases there is little to no material to support such an assumption. A potential exception is a female grave with a necklace and a merchant’s weight. The most common occupation-related objects at Emona are elements of medical sets (see Stemberger 2018 for details), but other tools such as chisels and shears were also found. The second group is characterised by an abundance of dining equipment and a general lack of other types of grave goods. These were previously interpreted as middle class burials (DeMaine et al. 1999), but it seems more likely that the dining equipment signals wealth through the practice of banqueting.

It is important to note that grave goods serve as indicators of gender only in the statistical sense (Pearce 2013), but it is entirely possible for individual graves with supposedly female artefacts to have belonged to a biologically male burial or vice versa; however, in comparison with the British material, deviations are not identifiable on a large scale (e.g. Philpott 1991). In Emona there is one grave that, not only according to Petru’s standards, would have passed as an adult male burial, but in fact belonged to a boy less than 15 years of age: the excavations at Štefanova ulica 4 have unearthed a grave that contained among other grave goods a complete military belt (Tomažinčič 2011, 2018: 438). While the majority of known belts from Emona’s inhumations were worn on the body, there are two cases where the belt was definitely not worn, one of which is this belt from Štefanova ulica 4.12

A military belt appearing in such a grave might indicate the practice of expressing status through adult male relatives connected to the army or civil service. The status of soldiers was argued elsewhere to have been special and different from that of the civil population (Tomažinčič 2018; Stemberger 2018). Being married to or descending from a member of the military presumably improved the social status of an individual. Such cases demonstrate that not only should the combinations of grave goods that are used for determining the gender of the deceased be rethought, but also that the position of the artefacts in the graves should be closely observed. The inherent problem with this method in the case of Emona is that the vast majority of burials from the early period, i.e. from the first and second centuries AD, are cremations, where the position of the grave goods in relation to the body is unrecoverable — although it can sometimes be inferred from whether they were burnt at a pyre or whether their skeletal remains were mixed with the pyre remains.13


Despite the inherent imprecision of establishing the age and gender of the deceased (Pearce 2013), there are a large number of cemeteries, not only in Slovenia, where this approach to data analysis is still the only option. While large collections, such as Emona’s, have plenty of drawbacks, they still offer a good basis and potential for further studies. Osteological data can serve as part of the approach to untying the Gordian knot presented by older documentation; however, it is also true that the ‘old’ excavation data can be used for much more than just establishing typologies and chronologies. It can offer a glimpse into social relations and dynamics as expressed in the funerary record.

The majority of the data upon which the findings presented in this article are based comes from the older excavations without osteologically determined sex and age, but that is not to say that the findings are merely small improvements obtained by the same methods used already by archaeologists in the 1970s and 1980s. The data may be the same, but the approach is significantly updated. It now incorporates many more theoretical frameworks and parallels, and above all has a much broader focus — rather than discussing individual burials in isolation and generalising from ‘typical’ cases, it attempts to address graves as belonging to coherent groups. New burial patterns have emerged, most notably among burials of women. As it stands, the data from older excavations can be used in its own right to deliver new insights even in the absence of more recent material to compare it to.

When trying to understand the past through old excavation reports, we are faced with the more recent past — i.e. the time in which the excavations were carried out and documented. By taking into consideration the worldview of researchers from different periods, and the circumstances in which they produced their work, much can be learned about their perception of the Roman period artefacts, and the practices related to them in what is now already a historical concept. Working through the documentation, or rather, meta-documentation, between the Roman period and the modern reader in turn influences the way we understand the Romans and ourselves.