‘ARMED WOMEN’ IN LBA AEGEAN ART

The criteria used in this article for including an image as an example of the ‘armed woman’ is of a feminine-coded human figure (based on physical characteristics, dress, or artistic conventions) depicted in close association with one or more items of military equipment (swords, spears, archery equipment, shields, helmets, body armour). This methodology yields the following list of examples, ordered chronologically:

1. 1. A sealstone of red sardonyx from Knossos, depicting a figure with a long robe and clearly defined breast. In the figure's right hand is held a sword, while in the left is an object which has been interpreted as a scythe or a plumed scabbard. Date: Late Minoan (LM) IA. CMS II.3, no. 16. Fig. 2a.

2. 2. A sealstone of unknown provenance held in the Staatliche Museen zu Berlin, Antikenabteilung, said to be from Crete. It depicts a figure with a long robe, flat hat, and clearly depicted breasts aiming a bow to the right. A long object at belt level on the figure may represent a sword or dagger. Date: LM I. CMS XI, no. 26. Fig. 2b.

3. 3. A gold ring of unknown provenance held in the Staatliche Museen zu Berlin, Antikenabteilung. In the centre of the field, a figure wearing a long flounced skirt aims a bow at a figure wearing a codpiece or loincloth, who faces left. The pair is flanked to the left by a further figure in a flounced skirt, with the lower body facing left and the upper body facing right. To the right of the central pair a figure in a skirt kneels before an object identified as a rock or sack, leaning on it with their left arm while holding their right arm out behind them. At the far right of the field is a tree. Above the central figures are one unidentified object and others identified as a double axe and two butterfly pupae. Date: LM I? CMS XI, no. 29. Fig. 2c.

4. 4. A gold ring from the ‘acropolis treasure’ at Mycenae. Two figures with breasts, wearing long skirts, approach from the left a third figure, with similar attributes, sitting beneath a tree. Between the two figures and the seated figure is a large double axe. The seated figure is flanked by two companions represented at a much smaller scale, one of whom offers the seated figure flowers, the other of whom stretches their arms out towards the tree. To the far left of the field is a motif of six lion heads. In the upper register are a sun and a crescent moon, separated from the rest of the scene by a wavy double line which may represent clouds. Below and to the left of this set of wavy lines is a figure represented at a small scale, carrying a figure-of-eight shield and a spear or staff – the figure that is the basis for including this item in the list. Date: Late Helladic (LH) II. CMS I, no. 17. Fig. 2d.

5. 5. A white amethyst lentoid seal of unknown provenance held in the British Museum. A central figure-of-eight shield is surmounted by a crested helmet. To each side of the shield are protrusions which may be intended to represent arms, the one on the left holding a pommelled sword. This may be mirrored in the damaged section of the seal to the right. Below the protrusions are objects which have been identified as representations of lion heads. The scene has been described as a panoply, or as a deity in military array (Càssola Guida Reference Càssola Guida1973, 30). Date: LH II. CMS VII, no. 158. Fig. 2e.

6. 6. A painted clay larnax from the North Cemetery at Knossos, showing a figure with an elaborate hairstyle or headdress and a long robe, armed with a spear and a round or figure-of-eight shield. Date: LM IIIA1. Danielidou Reference Danielidou1998, 181–2. Fig. 3a.

7. 7. A painted stucco plaque found in the ‘cult centre’ at Mycenae. Two figures with white skin, dressed in long robes, face a figure-of-eight shield placed between them. Elements of a third white-skinned figure, including parts of a head, left and right arms, and a foot, are associated with the shield, suggesting that it is being worn. A tapering yellow object apparently held in the right arm of this third figure has been interpreted as a sword or a spear. Immediately to the right of the shield, a small altar is depicted. Date: LH IIIA–B. Rehak Reference Rehak1984, 536–8; Immerwahr Reference Immerwahr1990, 121. Fig. 3b.

8. 8. A fresco fragment from the ‘cult centre’ at Mycenae, depicting the head and shoulder of a white-skinned figure facing right. The figure wears a boar's tusk helmet and holds a small griffin. Date: LH IIIB. Immerwahr Reference Immerwahr1990, 121. Fig. 3c.

9. 9. A fragmentary fresco from Room 31 of the ‘cult centre’ at Mycenae. The fresco is divided into two registers. The lower of the two depicts the upper part of a white-skinned figure holding sheaves of wheat (Taylour Reference Taylour1983, 56) or Pinna nobilis shells (Burke Reference Burke, Nosch and Laffineur2012, 175–6) in their raised arms. Part of a pillar and elements of a possible lion or griffin are visible next to the figure. The upper register, composed at a larger scale, depicts two white-skinned figures wearing long robes and facing each other. The upper part of the figure on the left is missing, but it is clear that they hold a large sword with the point resting on the floor. The figure on the right holds a long staff or sceptre. Between these two figures, two smaller figures are visible, rendered in black and placed some distance above the ground level of the composition, giving the impression of ‘floating’. Date: LH IIIB. Rehak Reference Rehak1984, 539–41; Immerwahr Reference Immerwahr1990, 120–1; Rehak Reference Rehak, Laffineur and Crowley1992, 47–53. Fig. 4a.

10. 10. A fresco fragment from Thebes depicting the left-facing head, in profile, of a white-skinned figure wearing a boar's tusk helmet. Elements surrounding the head may indicate that it is looking through a window or battlement, although it is not clear that these elements are intended to depict architecture. Date: LH IIIB. Immerwahr Reference Immerwahr1990, 128. Fig. 4b.

11. 11. A fresco fragment from a hunting scene at Tiryns depicting a hand with white skin holding a spear-shaft. Based on the apparent presence of spectators riding in chariots and a figure possibly represented at a larger scale wearing a flounced skirt, Anderson (Reference Anderson1983) has suggested that the hunt has a religious character. Another fragment from this composition depicts what has been identified by Verlinden (Reference Verlinden, Darcque and Poursat1985, 142) as a quiver. Date: LH IIIB. Rodenwaldt Reference Rodenwaldt1912, 121. Fig. 4c.

12. 12. A fresco fragment from Room 27 of the palace at Pylos, showing the hand and wrist of an archer with white skin, set against a blue background. Date: LH IIIB. Brecoulaki et al. Reference Brecoulaki, Zaitoun, Stocker, Davis, Karydas, Colombini and Bartolucci2008. Fig. 4d.

13. 13. A painted larnax from Tholos 2 at Milatos, Crete. A human figure with no obvious garb or sexual characteristics is depicted with raised arms, the left of which holds, or rests against, a figure-of-eight shield. A pair of wavy lines are visible at each side of the figure's head, possibly intended to represent long hair. A fish is depicted below the figure's feet. Date: LM III. Danielidou Reference Danielidou1998, 182. Fig. 4e.

14. 14. An engraved bronze double axe from Voros, Crete, plausibly depicting human figures. Side A of the axe shows a figure-of-eight shield flanked by objects which have been variously identified as ‘sacral knots’ (Platon Reference Platon1959, 387; Càssola Guida Reference Càssola Guida1973, 163–4, no. 54), cult or other garments (Buchholz Reference Buchholz1962b, 170; Verlinden Reference Verlinden, Darcque and Poursat1985, 135–40), or rudimentary human figures (Small Reference Small1966, 104). Side B depicts a further example of this object flanked by objects which have been variously identified as a ritual object like Evans’ ‘snake frames’ (Evans Reference Evans1935, 168–76; Buchholz Reference Buchholz1962b, 170), architectural elements (Small Reference Small1966, 106), ritual tankards (Onassoglou Reference Onassoglou1985, 23–8), or quivers (Verlinden Reference Verlinden, Darcque and Poursat1985, 141–5). Date: LM III. Fig. 5.

Fig. 2. (a) Red sardonyx sealstone from Knossos (CMS II.3, no. 16). (b) Sealstone of unknown provenance, said to be from Crete (CMS XI, no. 26). (c) Gold ring of unknown provenance (CMS XI, no. 29). (d) Gold ring from the ‘Acropolis Treasure’ at Mycenae (CMS I, no. 17). (e) White lentoid sealstone of unknown provenance (CMS VII, no. 158).

Fig. 3. (a) End of a painted larnax from Tomb 107.214 in the North Cemetery at Knossos (after Coldstream and Catling Reference Coldstream and Catling1996, vol. 3, fig. 115). (b) Painted plaque from the ‘Cult Centre’ at Mycenae (Evans Reference Evans1930, fig. 88). (c) Fresco fragment from the ‘Cult Centre’ at Mycenae (photo: David Smith).

Fig. 4. (a) Reconstruction of a fresco from Room 31 of the ‘Cult Centre’ at Mycenae (Barringer Reference Barringer2014, fig. 1:45a). (b) Fresco fragment from Thebes (Morgan Reference Morgan1988, fig. 156). (c) Fresco fragment from Tiryns (Rodenwaldt Reference Rodenwaldt1912, pl. XIV:1). (d) Fresco fragment from Pylos (Brecoulaki et al. Reference Brecoulaki, Zaitoun, Stocker, Davis, Karydas, Colombini and Bartolucci2008, fig. 1). (e) End of a painted larnax from Milatos (Evans Reference Evans1906, fig. 107).

Fig. 5. Bronze double axe from Voros, Side A (top) and Side B (bottom) (Buchholz Reference Buchholz1962b, fig. 1a–b).

Also relevant here due to the possibility that they represent human figures with weapons are a number of painted depictions of figure-of-eight shields surmounted by a small arc, including those on two LM IA three-handled jugs from Xeste 3 (Room 11) at Akrotiri on Thera (Danielidou Reference Danielidou1998, 165–6), an LM IA–IB cup-rhyton from House A at Agia Irini on Kea (Danielidou Reference Danielidou1998, 166), an LM IB cup-rhyton and amphoriskos-rhyton from the Stratigraphical Museum site at Knossos on Crete (Warren Reference Warren1984, 20–1; Danielidou Reference Danielidou1998, 167–8; Warren Reference Warren2000, 462–3), and an LM I three-handled jug from Poros on Crete (Warren Reference Warren2000, 463–4). These arcs have been interpreted as loops for hanging the shields from walls (Verlinden Reference Verlinden, Darcque and Poursat1985, 146), though Peter Warren (Reference Warren1984, 23; Reference Warren2000, 463–4) has argued that they may be intended to represent the head of a ‘Shield Goddess’, as suggested by Small (Reference Small1966, 105) on the basis of parallels with the white amethyst lentoid seal (5) and the painted plaque from the ‘cult centre’ at Mycenae (7). Frescoes of LM / LH IIIA–B date depicting rows of figure-of-eight shields surmounted by a rosette are known from Knossos, Tiryns and Mycenae (Immerwahr Reference Immerwahr1990, 138–40), with the rosette, again, interpreted as the head of a ‘Shield Goddess’ (Mylonas Reference Mylonas1977, 116; Immerwahr Reference Immerwahr1990, 140). However, M. Shaw (Reference Shaw, Nosch and Laffineur2012, 735) notes that the rosettes on the figure-of-eight shield fresco from Staircase 54 at Pylos are positioned between the shields, and suggests that they represent a hanging device rather than a head. Against this interpretation may be set an ivory figure-of-eight shield from the area of the prehistoric cemetery at Mycenae, which is shown either surmounting a boulder or crowned with a disc (Papazoglou-Manioudaki Reference Papazoglou-Manioudaki, Nosch and Laffineur2012, 448–52), suggesting that circular devices atop shields are not necessarily hanging devices. The rosette has also been claimed as a solar symbol with connotations of kingship, on the model of Egyptian religion (N. Marinatos Reference Marinatos, Bietak, Marinatos and Palyvou2007b; Reference Marinatos2009, 27). However, even if all of the examples of solar imagery given are correctly identified, giving it a divine meaning reaches beyond what can clearly be evidenced, as does constructing Minoan religion on Egyptian parallels. It is not possible to determine whether any or all of these depictions are in fact a stylised depiction of a human with a figure-of-eight shield, so they have not been included in the list of armed women but they do represent potential further examples.Footnote ²

The images listed above range in date from LM IA to LH IIIB, a period of around 500 years (c. 1700–1200 BC) which encompasses the Neopalatial and Final Palatial periods on Crete and the early and palatial Mycenaean periods on the Greek mainland. In terms of spatial distribution, eight of the 14 examples (57.14%) are from the Greek mainland and the remaining 6 are from Crete (42.86%). The mainland examples are almost all associated with major palatial centres of southern and central Greece during the LH IIIB period, and Mycenae is a particularly strong focus for finds. The Cretan material is less associated with palatial centres, although Knossos was the find site for 2 of the 6 examples (33.33%), and is limited to the central and eastern parts of the island. As a phenomenon, then, the ‘armed woman’ iconography is very much a product of the latest palatial phases of the Aegean Bronze Age, with Mycenae and Knossos representing the sites most associated with it.

The first and, perhaps, most fundamental question to be addressed is whether, in fact, the figures depicted in these images can be confidently identified as women or female. In contemporary western terms ‘woman’ would be a term for gender, a cultural identity, and ‘female’ would be a term for sex, a biological identification. Such a distinction is comparatively recent (e.g. Stoller Reference Stoller1968; Rubin Reference Rubin and Reiter1975). However, it is understood that even the biological understanding of sex shows it to be a spectrum with clusters around particular nodes, rather than a simple binary (e.g. Nordbladh and Yates Reference Nordbladh, Yates, Bapty and Yates1990, 224; Sofaer Reference Sofaer2006, 92). It is also the case that not all human societies of the past or present thought of sex as binary (e.g. Nordbladh and Yates Reference Nordbladh, Yates, Bapty and Yates1990; Gilchrist Reference Gilchrist1999, 58–64; Geller Reference Geller2005; Voss Reference Voss, Hall and Silliman2006, 107–8; Joyce Reference Joyce2008, 18–19, 43–5). Human definitions and understandings of gender are perhaps even more varied (e.g. Herdt Reference Herdt1994; Gilchrist Reference Gilchrist1999, 54–78; Meskell Reference Meskell1999, 78–82). Given all of this, it might legitimately be asked why we would try and identify the sex and/or gender of figures depicted in ancient art. My preferred answer is that of Skogstrand (Reference Skogstrand2016, 11): because the physical differences that constitute our understanding of ‘male’ and ‘female’ are visible across time, they may be used as an analytic category, a point of departure for research which can examine the diversity mentioned above through seeing whether physical differences were relevant or not, and how differences and similarities were expressed.

In terms of Aegean prehistoric art, previous generations of scholars have often adhered to Arthur Evans’ (Reference Evans1900–1, 94; Reference Evans1930, 208) scheme for frescoes of red-skinned figures representing males and white-skinned figures representing females, but more recent research has extensively critiqued this approach (Damiani Indelicato Reference Damiani Indelicato1988; N. Marinatos Reference Marinatos1989; B. Alberti Reference Alberti and Hamilakis2002; Reference Alberti, Funari, Zarankin and Stovel2005; N. Marinatos Reference Marinatos, Bietak, Marinatos and Palyvou2007a). Figures with black and yellow skins are also present in Aegean fresco painting (Evans Reference Evans1928, 755–7; Immerwahr Reference Immerwahr1990, 96; Morgan Reference Morgan and Sherratt2000, 939), and examples exist of scenes in which figures appear to defy the convention: the white skinned bull-leapers at Knossos and Tiryns who wear clothing otherwise seen on red-skinned bull-leapers (Evans Reference Evans1930, 209–18; Immerwahr Reference Immerwahr1990, 90–2, 110), or the red-skinned figures of the Knossos ‘Campstool’ fresco who wear apparently feminine clothing (Evans Reference Evans1935, 381–95; Immerwahr Reference Immerwahr1990, 95). While skin colour may have relevance to the depiction of gender in Late Bronze Age Aegean art, it is not a straightforward indicator.

While it may be that the people of the prehistoric Aegean had artistic conventions which they used to differentiate male and female, man and woman, it is clear that modern researchers face considerable difficulty in reconstructing what these conventions were. Archaeologists cannot simply make the assumption that human societies have universally had a binary model of sex and gender (Weismantel Reference Weismantel, Stryker and Aizura2013, 322; various contributions to Gaydarska et al. Reference Gaydarska, Rebay-Salisbury, Ramírez Valiente, Fries, Hofmann, Augereau, Chapman, Mina, Pape, Ialongo, Nordholz, Bickle, Haughton, Robb and Harris2023), and being aware of the possibility that we are not dealing with such a society is important when interpreting art. Even if Aegean societies of this period did have strongly binary conceptions of sex and gender, gender systems are not totalising, and societies with strongly prescribed gendered behaviours may in fact generate more alternative gender identities than societies with more fluid concepts of gender (Robb and Harris Reference Robb and Harris2018, 132–3). This must be borne in mind as we proceed to analyse images of ‘armed women’.

In terms of sex, a set of criteria have been developed by Talalay (Reference Talalay, Donald and Hurcombe2000, 5 and n. 9) to examine Neolithic human figurines. While there are clearly differences in medium, chronology, and contexts of use between these figurines and the LBA material discussed above, the criteria nonetheless provide a useful method which may be applied here:

• • Definite females are identified by the presence of female genitalia and/or breasts which do not appear to represent male nipples.

• • Probable females are identified by the presence of breasts and secondary traits which occur repeatedly on definite females (e.g. design patterns painted onto the surface [in the case of Neolithic figurines] or particular physical attributes).

• • Definite males are identified by male genitalia.

• • Possible males are identified by secondary traits which also occur repeatedly on definite males (e.g. posture).

• • Dual-sexed images are identified by the appearance of male genitalia and breasts which do not appear to represent male nipples.

• • Sexless or sexually ambiguous pieces are identified by the absence of any recognisable sexual attributes or secondary traits that also occur repeatedly on sexable images.

Applying these criteria to the above corpus of images yields 1 and 2 as definite females with weapons or military equipment, 4, 5, 7, 8, 10, 11, 12, 13 and 14 as sexless or sexually ambiguous figures, and 3, 6 and 9 as probable females if dress conventions are correctly understood. At this point we have already begun to discuss gender rather than sex. By its very nature, the process of decoding gender is a far more difficult one than sexing images. A more nuanced gendering of figures which pays close attention to features such as the body, hairstyle, costume, jewellery, colouring, pose, gesture, movement, activities and themes (Rehak Reference Rehak, Casey, Donlan, Hope and Wellfare1998; Alexandri Reference Alexandri and Kopaka2009, 21; Rehak and Younger Reference Rehak, Younger and Kopaka2009, 15; Newman Reference Newman2017; Franković Reference Franković and Matić2022) is required to achieve any meaningful understanding of gender construction in the Aegean Bronze Age. However, there is currently little consensus arising from such studies.Footnote ³ In particular, the possibility that LBA Aegean societies did not perceive gender as a binary system of men/women presents a challenge to modern interpretation, as do those images which may have been composed without any intention of coding for gender, or in the case of figurines perhaps as a genderless ‘core’ which would then have a gender applied to it through the application of perishable clothing. Even if a binary concept of sex and gender did exist in this period, it is entirely possible that ambiguity was deliberately invoked for religious or ceremonial purposes.Footnote ⁴

The position taken in this paper is that, while we may not be able to provide a definitive sex or gender identity for a given image, it is possible to detect some patterning (Chapin Reference Chapin, Nosch and Laffineur2012), with figures coded for gender by varying from male and female prototypes (Alexandri Reference Alexandri and Kopaka2009, 21). Therefore, while we may not able to state with certainty that a weapon is held by a male or a female, or even by a man or a woman, we can say that a figure is coded to invoke masculine or feminine characteristics. While, as has been established above, the general pattern of white being used for feminine figures in fresco painting cannot be used to definitively categorise a figure as a woman, it would add images 7, 8, 10, 11 and 12 to a category of scene in which weapons or military equipment are associated with figures that have been coded for a significant degree of femininity.

As will become clear, what is most striking about the corpus of images which appear to show or allude to armed female figures is the consistent manner in which their meaning has been interpreted by us.

INTERPRETATIONS OF THE ‘ARMED WOMEN’

Images of armed women in the corpus of LBA Aegean art have attracted comment and interpretation from a number of researchers, perhaps because they run counter to both modern conceptions of gender roles and ideas regarding gender roles in the Aegean Bronze Age.

Early interpretations (Gardner Reference Gardner1892–3; Evans Reference Evans1930, 314) identified the figure-of-eight shield as the embodiment of a deity often identified as female on the basis of parallels to the Pallas Athena of later Greek religion. These religious interpretations continued into later periods, with Persson (Reference Persson1942, 92) seeing the shield as a symbol of divine protection and Small (Reference Small1966) suggesting that it was the symbol of a protective goddess in Crete, who was then developed into a ‘warrior goddess’ by the Mycenaeans. Both authors again make a comparison to the later Pallas Athena as part of their interpretation. Vermeule (Reference Vermeule1974, 51–2) believed that the shield was either an attribute of a goddess or her personification. Borchhardt (Reference Borchhardt, Buchholz and Wiesner1977, 12–15) argued that the shield was an attribute of a war or hunting deity on the basis of a ritual interpretation of scenes in which it appears, without specifying a sex or gender for the deity concerned.

What is most striking about these interpretations is the desire to find a Bronze Age version of Pallas Athena, which requires reading back Classical Greek religion some 700–1000 years. The only current evidence for Athena in the Bronze Age is the presence of the term a-ta-na-po-ti-ni-ja (‘Mistress Athana’) on Linear B tablet KN V 52 + 52 bis + 8285 from Knossos, although Athana is plausibly the name of a settlement (Duhoux Reference Duhoux, Duhoux and Morpurgo Davies2008, 276–8; Hiller Reference Hiller, Duhoux and Morpurgo Davies2011, 183). Even if a deity called Athena existed in this period, there is no evidence of her having the attributes associated with her by the Classical Greeks.

The concept of a ‘warrior deity’ has also proved influential in interpreting this iconography. Martin Nilsson (Reference Nilsson1950, 406–12) is unusual in generally viewing the figure-of-eight shield as not a cult object on the basis that he does not see any clear cult-scenes featuring the shield. Instead he sees the shield as part of the equipment of a Mycenaean ‘war goddess’, and armed figures in Minoan art as representing a male hunting god, due to his perception of Minoan art being less violent in nature than Mycenaean art. Mylonas (Reference Mylonas1972, 27–9; Reference Mylonas1977, 115–16) saw the shield as the symbol of a goddess of war who was worshipped in the cult centre at Mycenae, with their function being one related to rites involving animal sacrifices to a divinity. Càssola Guida (Reference Càssola Guida1973, 30) interprets the figure-of-eight shield as a cult object based upon the idea that it was no longer used as a weapon after LH I, and through the hypothesis of a Mycenaean ‘warrior god’. In her later work (Càssola Guida Reference Càssola Guida1975), male and female war deities are identified, characterised by the attributes of the figure-of-eight shield and the boar's tusk helmet respectively. This interpretation is based upon the Aegean colour convention for depicting male or female figures (see above), Linear B texts, and a comparison once again to the later Pallas Athena. Perhaps most influentially, Rehak (Reference Rehak1984; Reference Rehak and Laffineur1999) identified the armed figures as a ‘warrior goddess’ originating in Crete and later being adopted on the Greek mainland on the basis of the generally religious nature and/or context of the depictions, and on the Minoanising elements found in the Mycenaean depictions. In a later work (Rehak and Younger Reference Rehak, Younger and Kopaka2009) it is suggested that the attribute of the figure-of-eight shield is symbolically removed from connotations of hunting or militarism and is, instead, associated with females, the natural world, the practice of bull sacrifice, and sanctuaries. This is on the basis that the shield appears in scenes of sanctuaries with tree-pulling and Omphalos-embracing, and with women in the lactans or breast-holding pose. While Hooker (Reference Hooker, Amory, Considine and Hooker1996) accepts the likelihood of a ‘war goddess’ in Mycenaean Greece, he disputes whether any extant iconography actually depicts such a figure. He suggests, instead, that both the ‘acropolis treasure’ ring from Athens and the ‘cult centre’ plaque from Mycenae depict a hybrid of separate Minoan motifs of a descending divinity and a warrior covered with a figure-of-eight shield by Mycenaean artists unaware of their original significance. The iconography of the figure-of-eight shield without any human elements he sees as a purely decorative motif with no cult significance. Overall, this iconography is said to have developed through Mycenaean misunderstandings of the actual meanings of various Minoan motifs.

These interpretations of the ‘armed woman’ iconography are again striking for their reliance on the later figure of Pallas Athena. What is also apparent is that even the ‘war goddess’ is not always connected to violence; a confused position was perhaps epitomised by Mylonas (Reference Mylonas1977, 115–16), who identifies the place of worship of the ‘Goddess of war’ partly by the presence of frescoes which ‘exhibit subjects of a peaceful and even religious nature’, while even the frieze of figure-of-eight shields ‘was not symbolic of … military prowess and achievement but possessed another special meaning’. In the case of Rehak's later work (Rehak and Younger Reference Rehak, Younger and Kopaka2009), it is particularly strange that he argues for the removal of the shield from contexts of hunting or militarism while simultaneously connecting it to bull-sacrifice and noting that in the LM IB ‘cult basement’ at Knossos a cup-rhyton decorated with figure-of-eight shields was associated with the bones of children bearing cut marks consistent with the removal of flesh (Rehak and Younger Reference Rehak, Younger and Kopaka2009, 12; see also Warren Reference Warren1980–1, 79–92). Once again, violent connotations and contexts for this imagery become muted in our interpretations.

Perhaps representing an opposite pole to interpretations favouring a warrior deity, there is also a body of explanations favouring a connection to fertility. Warren (Reference Warren1984; Reference Warren2000) sees the shield as embodying a nature goddess with connections to fertility, through an identification of stylised sea squills associated with some depictions of the figure-of-eight shield. Later beliefs in the regenerative powers of the plant are used to suggest the possibility of similar associations in the LBA. N. Marinatos (Reference Marinatos1986, 52–8) challenged the association of the figure-of-eight shield with a war goddess, arguing instead that on Crete the shield was associated with males and was a cult implement used in the worship of a hunting god, or related to a ‘vegetation festival’, an interpretation based upon Cretan depictions of the shield being carried by apparently masculine figures, and the hunting or sacrifice required to manufacture a shield.

As with the interpretations of the ‘armed woman’ iconography as a warrior goddess with a curiously pacific nature, the interpretation of the imagery as being connected with fertility once again seems to distance women from concerns of violence. Even if the identification of sea squills is correct, their significance remains uncertain, and it is not clear that they should characterise the whole significance of the figure-of-eight shield. Claiming an association between men and the shield does not help us interpret the shield when it is depicted with women, nor is it apparent what the connection between a shield and a ‘vegetation festival’ would be.

Some interpretations have placed more emphasis on the possibility of connections between the ‘armed woman’ iconography and violence. Kokkinidou and Nikolaidou (Reference Kokkinidou and Nikolaidou1993; Nikolaidou and Kokkinidou Reference Nikolaidou, Kokkinidou and Carman1997) argue that the presence of armed male and female figures in LBA Aegean art represents the arrival of violent trends to Aegean religion, introduced by a greater participation of males in cult activities, in contrast to earlier periods in which such activities would have been more matrifocal. This reading posits that the ‘armed women’ have a more warlike character but retain some part of their long-standing association with fertility (Nikolaidou and Kokkinidou Reference Nikolaidou, Kokkinidou and Carman1997, 184–5). Overall, Nikolaidou and Kokkinidou see the use of symbolically violent imagery in the LBA Aegean as simultaneously supporting and subverting the dominant order. This dialectic is suggested, furthermore, to be an explicitly gendered one, with violent androcentric values being subverted by the strong female element of the LBA Aegean religious domain. If the author correctly understands the discussion of post-Freudian psychology which opens Nikolaidou and Kokkinidou's Reference Nikolaidou, Kokkinidou and Carman1997 paper, these violent men and peaceful women are to be understood in timeless, essentialised terms.

The analysis of Kopaka (Reference Kopaka and Laffineur1999), while seeking to place women into the social phenomenon of warfare, relates ‘armed woman’ iconography to the initiation rites of adolescent women in some Aegean Bronze Age societies. In this reading, the transition to adulthood was marked by a ritual performance of war for both males and females, after which they would move into separate adult worlds revolving around warfare (for males) and the home (for females). The role of the home in the female initiation ritual is signified by the presence of architectural elements in the figural scenes featuring ‘armed women’. This interpretation is based upon the idea that maternal care would make women responsible for transmitting cultural ideas about society to children, and on the presence of non-combatant women in militaristic scenes such as the frescoes from Akrotiri on Thera or the ‘siege rhyton’ from Mycenae, in addition to the iconography discussed in this article.

While Kokkinidou and Nikolaidou's analysis is perhaps the most developed and nuanced, it remains based around an essentialised view of masculine and feminine relationships to violence, so the ‘armed woman’ must represent women subverting masculine violence, or ritualised violence must actually be about the control and elimination of violence (Nikolaidou and Kokkinidou Reference Nikolaidou, Kokkinidou and Carman1997, 190). Kopaka's interpretation encounters the problem of why children would receive the same initiation rite connected to warfare when they would subsequently move into separate and highly gendered spheres.

Overall, previous researchers have tended to distance the ‘armed woman’ iconography from violence. While we should be aware that any given object can be, and often is, encoded with a variety of contextual meanings, there appears to be an underlying belief within modern academic discourse that women cannot be associated with violence. Certainly, the variety of symbolic meanings proposed for weaponry and military equipment in Aegean art is much reduced when they are associated with apparently male figures. Is the main determinant for the non-martial reading of a sword, a shield or a boar's tusk helmet simply the observation that it is held by an apparently female figure?

WOMEN, POWER AND THE STRUCTURE OF AEGEAN SOCIETIES

In order to produce new interpretations of ‘armed woman’ iconography which move beyond the consciously, or unconsciously, gendered interpretations of the recent past, there are a number of possibilities which must be considered. Many of the interpretations discussed above revolve around divine explanations, in the form of a war and/or hunting deity. While the context of much of the material does suggest a cult or ceremonial connotation, it has proven difficult to distinguish deities in LBA Aegean art. It is also the case that, as Rehak (Reference Rehak, Casey, Donlan, Hope and Wellfare1998, 192) has noted, by seeing every feminine depiction in Aegean art as a deity, we render the landscape curiously devoid of powerful mortal women.

Should this iconography be taken as a sign that LBA Aegean women and/or people who would in modern western terms be considered of non-binary gender (Herdt Reference Herdt1994) participated in violent acts such as hunting and warfare? The evidence of various historical and ethnographic cultures around the globe (Estioko-Griffin and Griffin Reference Estioko-Griffin, Griffin and Dahlberg1981; Burbank Reference Burbank1994; Koehler Reference Koehler, Claassen and Joyce1997; Alpern Reference Alpern1998; Goldstein Reference Goldstein2001, 59–127; Hollimon Reference Hollimon, Arnold and Wicker2001; Loman Reference Loman2004, 44–53; Pennington Reference Pennington2010; Fabre-Serris and Keith Reference Fabre-Serris and Keith2015; Carney Reference Carney, Heckel, Naiden, Garvin and Vanderspoel2021) makes it clear that we cannot reject the idea out of hand. It might be argued that none of these representations actually depict feminine-coded figures engaging in violent acts, in a manner similar to Nikolaidou and Kokkinidou's (Reference Nikolaidou, Kokkinidou and Carman1997, 182) proposal that in Aegean art the act of animal sacrifice is always depicted being carried out by masculine-coded figures, although feminine-coded figures are seen carrying dead animals and bronze double-axes. This suggests that direct associations between violent acts and femininity were ideologically unacceptable, whether or not they occurred in reality, although Molloy's (Reference Molloy2012, 131) identification of shared elements between scenes of competition, hunting, violence and sacrifice as part of a symbolic grammar of dominance and power means that we should be cautious about making unequivocal judgements. The interment of weapons and military equipment in graves with female-sexedFootnote ⁵ human skeletons is a recognised global phenomenonFootnote ⁶ and one also evident in the LBA Aegean. Examples of weaponry and artistic depictions of martial themes interred with female-sexed skeletons range from Grave Circles A and B at Mycenae to tholos tombs at Archanes on Crete (see Appendix A).Footnote ⁷ Such items should not be regarded biographically (Härke Reference Härke1990; Whitley Reference Whitley2002). But if they are regarded as being linked to performances and displays connected to rank and status when found with male-sexed skeletons, why not also when found with female-sexed skeletons?

Crete provides examples of female-sexed skeletal remains bearing signs of osteological trauma consistent with inter-human violence. The cave of Agios Haralambos was used in Middle Minoan IIB for the secondary burial of remains that had earlier received primary burial at a different location. Osteological analysis of these remains yielded evidence of trauma to the skull in three females, 11 males, one child, and one individual who could not be sexed (Chlouveraki et al. Reference Chlouveraki, Betancourt, Davaras, Dierckx, Ferrence, Hickman, Karkanas, McGeorge, Muhly, Reese, Stravopodi and Langford-Verstegen2008, 581–94; McGeorge Reference McGeorge2011). At Mochlos a basement room in the west wing of Building B2, which was destroyed in LM IB, contained a female-sexed skull with a hole in the cranium (Soles Reference Soles and Laffineur1999, 57–8). By its very nature, osteological trauma may be more likely to identify victims, rather than the perpetrators, of violence, although an individual can be simultaneously both perpetrator and victim (Ferguson Reference Ferguson2021, 118).

The discussion of burials with weapons in Aegean prehistory and in archaeology more broadly has in recent decades been concerned with the question of whether such interments can be described as ‘warrior graves’ (Härke Reference Härke1990; Treherne Reference Treherne1995; Whitley Reference Whitley2002; Frieman et al. Reference Frieman, Brück, Rebay-Salisbury, Bergerbrant, Montón Subías, Sofaer, Knüsel, Vandkilde, Giles and Treherne2017; Georganas Reference Georganas, Horn and Kristiansen2018), with a broad consensus arising that these burials should be seen as an expression of a social status. To use Georganas’ (Reference Georganas, Horn and Kristiansen2018, 191) phrasing ‘In the Bronze Age, as in later periods, being a male member of the aristocracy usually meant being a “real man”. A “real man” meant being identified as a warrior.’ This is not incorrect, but we should consider whether being a ‘real man’ – or possessing at least some of its defining characteristics – was limited to males. Molloy (Reference Molloy2012, 89) has proposed that in the Bronze Age Aegean a warrior identity was enacted through competitive displays including bull-leaping, boxing, hunting, combat training, and fighting, while Franković and Matić (Reference Franković and Matić2023, 56) believe that such activities were how elite boys acquired masculinity. We have already seen above that the white-skinned figures present in some depictions of bull-leaping have led to suggestions that women took part in these displays, and that several of the depictions of armed women – particularly those with archery equipment – have been interpreted as women engaged in hunting.Footnote ⁸ For boxing, at least one of the figures engaged in the activity in a fresco from Building Beta at Akrotiri on Thera has been suggested to be a girl on the basis of a slightly lighter shade of red used for the skin and the presence of a gold earring worn by women in other depictions.Footnote ⁹ It is also the case that, as noted by Parr (Reference Parr2014), the comparative lack of osteological studies carried out in Minoan archaeology means that any of the numerous LM II–IIIA weapon-burials near Knossos (L. Alberti Reference Alberti, Cadogan, Hatzaki and Vasilakis2004) – held to be of men largely because they are weapon-burials – could be the burial of a woman. The possibility that at least some of the violent, risky lifeways that were part of elite status for LBA Aegean men were also open to LBA Aegean women must surely have consequences for our understandings of political power in the period.

Any consideration of women and power in Aegean prehistory inevitably exists in the shadow of ‘prehistoric matriarchies’, a position advanced in Aegean archaeology by Evans,Footnote ¹⁰ part of a more general intellectual movement concerning prehistoric matriarchy and the ‘Mother Goddess’ or ‘Great Goddess’ which extended into the middle of the twentieth century (Goodison and Morris Reference Goodison, Morris, Goodison and Morris1998, 7–9). More recently, the matristic ‘Old Europe’ reconstructed by Marija Gimbutas (Reference Gimbutas1974; Reference Gimbutas1989; Reference Gimbutas1991) has proven influential, particularly at a popular level. Such views have now been largely rejected by archaeologists (Nixon Reference Nixon, Krzyszkowska and Nixon1983; Talalay Reference Talalay1994; Conkey and Tringham Reference Conkey, Tringham, Stanton and Stewart1995; Meskell Reference Meskell1995), and to posit the existence of one or more matristic systems in the Bronze Age Aegean based on the prominent artistic depiction of women (Thomas Reference Thomas1973; Rehak Reference Rehak, Casey, Donlan, Hope and Wellfare1998, 197) is methodologically naïve.Footnote ¹¹ Recent work by Driessen (Reference Driessen and Pullen2010; Reference Driessen, Panagiotopoulos and Günkel-Maschek2012) has argued on the basis of architectural studies that Cretan societies of the Neolithic down to Neopalatial were ‘house societies’ on the model of Lévi-Strauss (Reference Lévi-Strauss1982) in which society is composed of corporate groups materially represented by a physical structure and objects associated with it. Many house societies are matrilocal (married couples reside with the wife's family) and matrilineal (kinship is traced through the mother), which often results in a higher social status for women. While a matrilocal and matrilineal society would meet some definitions of matriarchy (Göttner-Abendroth Reference Göttner-Abendroth2007, 4), Driessen explicitly argues against Bronze Age Crete being matriarchal, noting that in most matrilineal societies it is men who hold power (Driessen Reference Driessen, Panagiotopoulos and Günkel-Maschek2012, 141).

While the universality of male domination in human societies has been a matter of debate (e.g. Rosaldo Reference Rosaldo, Rosaldo and Lamphere1974; Schlegel Reference Schlegel1977; Atkinson Reference Atkinson1982; Bell Reference Bell1993), the discussion of women in ‘state’ societies has tended to follow Engels (Reference Engels1884 Reference Engels[1972]) in arguing that such societies feature a universal subordination of women (e.g. Leacock Reference Leacock and Engels1972; Gailey Reference Gailey1987). The most significant deviation from this pattern has been the work of Ortner (Reference Ortner1978; Reference Ortner, Ortner and Whitehead1981), which has put forward the view that the rise of ‘state’ society in fact enhanced the status of women. Certainly it is possible to find examples of ‘state’ societies in which women attain positions of power, wealth and status (Wheatley and See Reference Wheatley and See1978; Rohrlich Reference Rohrlich1980; Bay Reference Bay1998; Nelson Reference Nelson, Hays-Gilpin and Whitley1998). Rather than adhere to any universalised scheme, it may be most productive to focus on the culturally and historically specific developments which occurred in the gender relations of past societies (Rapp Reference Rapp1977; Silverblatt Reference Silverblatt1988; Pyburn Reference Pyburn and Pyburn2004). While most known societies may be patriarchal, there is significant variation in women's power in those societies.

Textual research into LBA Aegean societies has recently demonstrated the existence of elite women with access to some forms of power in the LM / LH IIIA–B periods. At Knossos, while men appear to have controlled more property, personnel and commodities than women, the Linear B texts attest to the presence of non-priestly women who owned land and controlled foodstuffs, slaves, livestock, textiles and luxury goods. These women are recorded in the documents in a manner analogous to that of male landowners (Olsen Reference Olsen2014, 257). At Pylos the situation is different (contra the more utopian account given by Billigmeier and Turner Reference Billigmeier, Turner and Foley1981), with economic status limited to female cult officials who controlled leased (rather than owned) land, and control over foodstuffs, textiles, and commodities including bronze (Olsen Reference Olsen2014, 255). These ka-ra-wi-po-ro (‘key-bearers’) are linked by Younger (Reference Younger, Laffineur and Crowley1992, 272–3, 276) to a number of depictions of feminine-coded figures which he identifies as wearing the administrative technology of lentoid seals, including the figure holding the sheaves of wheat or Pinna nobilis shells in the fresco from Room 31 of the ‘cult centre’ at Mycenae (9). This control of economic resources must have implications for our understanding of power in LBA Aegean societies, at least at the level of political economy (Sanday Reference Sanday, Rosaldo and Lamphere1974; Leacock Reference Leacock, Bridenthal, Koonz and Stuard1987, 33–4) and for particular segments of society.

To return to the ‘armed woman’ iconography, it is clear that we should avoid a totalising explanation given the variation which exists in both the medium of representation and the nature of the scenes depicted. Such is well exemplified by the high number of archery depictions in the corpus, compared to the generally low presence of the bow in LBA Aegean iconography (Snodgrass Reference Snodgrass1967, 17–18). It is, therefore, possible that archery had strongly gendered connotations in the LBA Aegean world, although we should remain aware of some depictions of military archery in the art of the period (e.g. Sakellariou Reference Sakellariou1974; Reference Sakellariou1975), and that archery equipment could form part of weapon-burials.Footnote ¹² The use of boar's tusk helmets, figure-of-eight shields, and swords in palatial iconography suggests that these items form part of a language of power. As such, the concentration of such items iconographically associated with feminine-coded figures in the ‘cult centre’ at Mycenae (7, 8, 9) is of particular interest, given the combination of a comparatively circumscribed social and political role for women and their strong presence in Mycenaean cult activities. The concept of the ‘cult centre’ as an entity separate from the palace has been articulated by a number of scholars,Footnote ¹³ and while westernised concepts of a separate church and state are inappropriate (Nikolaidou Reference Nikolaidou, James and Dillon2012, 42), it may be productive to consider the ‘palace’ not as a monolithic entity but rather as a location in which competing forms of power were articulated. In such a context, depictions of ‘armed women’ may have served to lay claim to particular types of power, and perhaps more explicitly those connected to coercive authority. As Jensen and Matić (Reference Jensen, Matić, Jensen and Matić2017, 14) have recently pointed out, cosmologies and belief systems are not innocent alternatives to violence, as they are deeply entangled with the material realities of power. Nikolaidou (Reference Nikolaidou, Davis and Laffineur2020, 9) has recently spoken of the use of violent, competitive imagery by women in the LBA Aegean as representing their appropriation of the dominant ideology, but it is also possible that women were always involved in the creation and maintenance of that ideology.

The maintenance of such a view would, perhaps, require reassessment of the political structure of Mycenaean palatial society, often still described in terms of a pyramidal vertical control hierarchy (Kilian Reference Kilian1988). The work of Crumley (Reference Crumley1979; Reference Crumley, Patterson and Gailey1987; Reference Crumley, Ehrenreich, Crumley and Levy1995) has introduced concepts of heterarchy into anthropology and archaeology. Heterarchy describes a situation in which elements are either unranked or have the potential to be ranked in a number of different ways (Crumley Reference Crumley, Ehrenreich, Crumley and Levy1995, 3). As such, heterarchy is not a model of how a particular society operates, and it is possible to describe a variety of social systems as heterarchical (Moore and González-Álvarez Reference Moore, González-Álvarez, Thurston and Fernández-Götz2021, 127). Rather than a pyramid, societies might be better represented by a trapezoid (Hill Reference Hill, Moore and Armada2011, 254–5) or as a network of nodes (Mytum Reference Mytum2018). Concepts of heterarchy have been applied to the LBA Aegean, with one recent study using Cretan and mainland societies as poles on a continuum between ‘corporate’ and ‘network’ societies, contrasting ‘corporate’ theocratic Cretan heterarchical societies with ‘network’ mainland hierarchies based on economic control (Parkinson and Galaty Reference Parkinson and Galaty2007). This, however, ignores the potential for more horizontal or heterarchical relationships of power in the Mycenaean world. We have already discussed the possibility of the ‘cult centre’ at Mycenae operating separately from the palace, and should resist the assumption that religious power must be inferior to political power.Footnote ¹⁴ Traditional models of the ‘state’ (e.g. Flannery Reference Flannery1972; Service Reference Service1975; Rousseau Reference Rousseau, Claessen, Van de Velde and Smith1985) have demonstrated little scope for the theorising of female political action, but heterarchical concepts (at least in the broader descriptive sense accepted by Saitta and McGuire [Reference Saitta and McGuire1998]) may prove useful in its future analysis. Heterarchical concepts offer one way of moving beyond strictly binary conceptions of gender, and of exploring the way in which gender may be cross-cut by multiple other aspects of a social order (Levy Reference Levy and Nelson2006; Hutson, Hanks and Pyburn Reference Hutson, Hanks, Pyburn and Bolger2013). As such, heterarchical concepts might offer the most appropriate method for further exploring the position of women with regards to violent power in LBA Aegean societies that has been outlined in this article.