Since brooches in the ancient world were overwhelmingly functional their use is inextricably connected to the clothing they fastened. In many ways they should be viewed as part of the clothing itself, in the same way that buttons are today. The Edict of Justinian shows that cloaks were sometimes sold with brooches (for example, 53-6, 18-9; fibulatorium) and a tomb painting from Silistra of slaves carrying their master’s clothes to be put on after he has bathed shows the brooch left attached to the cloak (Croom 2000, fig. 14.2).

Pictorial evidence shows that brooches could be used to fasten both tunics and cloaks. Men are only ever shown using a cloak brooch, while women, at different times and different locations, use them for undertunics, overtunics and cloaks, and at other times never appear to use them at all. There is, for example, no pictorial evidence that women ever wore brooches with the Gallic coat which was worn throughout Gaul and Britain, which must at least heavily suggest that the majority of brooches found during the period when the coat was in fashion must have been worn by those men who still wore cloaks (in particular, soldiers). Tombstone evidence shows that brooches were used in very obviously different ways by different groups, and so presumably helped to distinguish cultural identities.

Crossbow brooches

Evidence clearly suggests that the late Roman crossbow brooch was worn only by men. The tomb painting mentioned above shows a crossbow brooch, whilst the famous diptych of Stilicho shows the general wearing one fastening his cloak. Of the eight crossbow brooches found in graves at Lankhills, the accompanying grave-goods suggest male burials for at least seven, and the crossbow found at the Eastern Cemetery in London was found associated with a man (Barber and Bowsher 2000, 207, no. 3). It has been argued that they became a symbol of official rank or military status (Philpott 1991, 139-40), and as such would be even more unlikely to be worn by women.

If one style of brooch, in this case the crossbow, was only worn by one sex, can other brooch types also be identified as being either male or female items? Philpott’s study of Romano-British burial practice lists a large number of graves containing brooches and while some of these may be in the grave-fill or be grave-goods that did not necessarily belong to the deceased, their position on the body suggests that some at least were probably being worn when buried (1991, table A30). However, many of these examples come from old reports where details of either brooch type or the sex of the burial is missing, and others from graves where a lack of accompanying grave-goods or the poor condition of the bones meant that the burials could not be sexed. As women are much easier to identify from accompanying grave-goods, more brooch types can be identified for them: penannular (Colchester, p340, no. 647), plate brooches (York, p346, no. cvi), and ‘bow-brooch’ (Normangate Field, p343). Of particular interest is a man who was a possible drowning victim whose corpse was left in the flood-silts. As it can be argued that brooches in burials may have been used as shroud pins, this example is significant since as he was not deliberately buried he must have been still wearing his everyday clothes, in this case including a cloak pinned on his left shoulder by a dolphin brooch (Olivier 1982, fig. 14, no. 4).

Tombstones from Pannonia and Noricum showing tube dresses fastened by brooches.
Figure 1. Tombstones from Pannonia and Noricum showing tube dresses fastened by brooches.

Brooches with attachment loops

The tube dress, a tunic fastened at the shoulders, is known in the ancient world from Denmark to Greece and is a long-lived and wide-spread fashion. Fastening the dress by a pair of brooches is known from Germany, Pannonia and Noricum (Fig. 1) and survives into the pagan Anglo-Saxon and Viking periods. Grave evidence shows that they were usually worn at the shoulders (sometimes as low as the collar bone; Fig. 2 ), as illustrated by a recently published example from a grave in London (Barber and Bowsher 2000, 183, B374). Roman tombstones clearly show that they were used to fasten the overtunic rather than a cloak. The brooches from London were connected by a necklace or chain made up of silver rings, 85 glass beads and possibly also copper fittings or links (ibid.). This is a fashion more commonly known in this country from the pagan Anglo-Saxon period (Owen-Crocker 1986, fig. 30; p55; see Fig. 2) where it is found in many inhumation burials.

Reconstruction of pagan Anglo-Saxon costume using grave evidence.
Figure 2. Reconstruction of pagan Anglo-Saxon costume using grave evidence.

Parallels therefore suggest that pairs of brooches (not always identical) were most commonly worn to fasten a female tunic on the shoulders, and that they could be linked by jewellery such as chains or strings of beads (eg Allason-Jones 1989, fig. 18). A second-century burial at Gulden Morden may well reflect this fashion; an iron brooch and an enamelled bow-and-fantail brooch with attachment loop were found ‘at the throat’, with a glass bead corroded to the iron example (Lethbridge 1935, 117, pl VII). Brooches are also found with a copper -alloy chain of varying lengths (Johns 1996, fig 7.7; Wardle 1998, fig. 19). It has been suggested that the those joined by short chains were worn lower down on the chest in a purely decorative way, such ornamental brooches being shown on some tombstones elsewhere in the Empire. However, where chained brooches are worn on the chest they are shown as being of a different design to the brooches worn at the shoulders. Bow brooches were designed to hold together a quantity of cloth, which is not required when the brooch is simply pinned to the tunic for decorative purposes. Romano-British brooches connected by short chains could have been used by girls, or alternatively the brooches were worn close to the neck where a long chain would not be necessary (Fig. 3.1; compare Fig. 1.2). As this style of wearing the brooches would result in a small neck-hole, a brooch would have to be undone to take the tunic off at night (not necessary when the brooches are worn further out on the shoulders. See Fig 3.2 and compare Fig. 1.1); having it safely linked to the other brooch could then be an advantage.

Romano-British brooches used to fasten a tube dress in a number of ways
Figure 3. Romano-British brooches used to fasten a tube dress in a number of ways. Note (with Fig 2) the variety of ways the deceptively simple tube dress can be worn.

Reconstructions have shown that single strings of beads can be attached to brooches fastening a tunic by either a small loop at both ends to slip onto the pin of the brooch, or larger loops to go over the whole of the brooch (Fig. 2). A number of Romano-British brooches were designed with a very obvious, integral loop (headloops) for the attachment of strings of beads or chains, and this design may suggest that wearing these brooches with jewellery was more common than wearing them without. This use of brooches with jewellery may be significant. Philpott’s study of Roman burial practice has shown that very little jewellery was worn by men, and what there is confined to finger-rings and (occasionally) bracelets (1991, 144). Therefore it could be suggested that brooches designed to be worn with a necklace (ie all those with a loop) should be considered a female style of brooch.

Could a single brooch with a loop still have been worn by a man to fasten his cloak? This seems unlikely if the brooch was designed to both fasten a style of tunic worn only by women, and to incorporate a necklace, also used only by women. In the Mediterranean Roman world men could be ridiculed merely for wearing colours or fine cloths that could be considered ‘feminine’, let alone for wearing female garments. The Emperor Commodus, for example, was condemned for his ‘complete indifference to propriety’ in wearing female clothing in public (SHA XIII.4). Whilst Roman women could wield great political influence and have immense wealth, they were still considered to be weaker creatures who were not the equal of men. To be female was to be inferior both physically and mentally, which made it of some importance for men to avoid appearing feminine. In Celtic or Germanic societies it is possible that women were held in higher esteem and that therefore there may not have been the same fear of appearing feminine, but their societies were as much warrior societies as the Romans, and I would argue that any society that honours physical prowess above all else is unlikely to genuinely consider women as equals. It seems likely that their menfolk were just as unwilling to be associated with female clothing or jewellery as the Mediterranean Romans.

Unisex brooches

It is possible that men and women could wear the same type of brooch for fastening the cloak as there is nothing in the cloak itself to identify it as either a male or female item of clothing. However, women are not often showing wearing the cloak in the same way as men (i.e. covering both shoulders and fastened on the right shoulder) and as well as being worn in a different fashion it may also have been fastened with a different form of brooch. Cloaks worn by men, whether civilian, soldier, Roman or barbarian, are always shown fastened by a single brooch on one shoulder.

The penannular brooch may be one such form of brooch worn by either sex. Grave evidence shows penannular brooches could be worn by women; another example to add to this list may well be the penannular brooches reported on in Lucerna 25 since they were linked by a chain (Hill 2003, 11; although since these are a pair they must have been used to fasten a tube-dress). A penannular has also been found in a grave of a possible male in the Eastern Cemetery of London (Barber and Bowsher 2000, p175, B329.1; table 7). This brooch type might therefore have been acceptable for either sex, but there could well still be minor differences in the design between male and female styles. Even nowadays items used by both men and women such as buttons, belts and watches have differing designs for the two sexes, the differences sometimes being slight but obvious to those who know the conventions (consider also such subtle differences, such as buttoning coats etc on opposite sides for the two sexes).


Brooches must surely be a greater source of information than they are currently. It is likely that many subtle differences between who wore what and how have been lost to us (and unless we find a large number of convenient flood victims we may well never be in a position to untangle all the proprieties of brooch-wearing), but there is still much work to be done to see if at least some brooch types can be divided by tribal fashions, function (cloak/mantle/tunic) and use by the different sexes.

Alex Croom
Arbeia Roman Fort
Baring Street
South Shields


My thanks to Margaret Snape and Wulfingas 450-550 Society for permission to use their photographs as the basis for the sketches of dresses worn with brooches.


Barber, B & Bowsher, D, 2000, The Eastern Cemetery of Roman London. Excavations 1983-1990, MoLAS monograph 4 (London)

Croom, A T, 2000, Roman Clothing and Fashion (Stroud)

Hill, , J D, 2003, ‘A pair of silver penannular brooches from Wheathampstead’, Lucerna 25, 11-2

Olivier, A C H, 1982, ‘The brooches’, in T.W. Potter and C.F. Potter. A Romano-British Village at Grandford, March, Cambridgeshire, British Museum Occasional Paper 35 (London)

Owen-Crocker, G, 1986, Dress in Anglo-Saxon England (Manchester)

Lethbridge, T C, 1935, Further excavations in the early Iron Age and Romano-British cemetery at Guilden Morden, Proceedings of the Cambridgeshire Antiquarian Society 36, 109-20

Philpott, R, 1991, Burial practices in Roman Britain, BAR 219 (Oxford)

Wardle, A, 1998, ‘Roman London: recent finds and research’, in Roman London: recent archaeological work, J Roman Archaeology Supplementary series 24, 83-9