Missing Queens: Gender, Dynasty and Power in Vandal Africa

This paper reconsiders a curious aspect of the Vandal kingdom of North Africa (439–533 ce): the total absence of women called Vandals in extant sources. It argues that these missing Vandal women are the women of the Hasding royal dynasty. The non‐application of the ethnic terminology to the consorts, sisters and daughters of kings and princes corresponds to a broader dearth of contemporary discussions of their political agency. This reduced visibility reflects particular structural features of Vandal kingship – and, in particular, its succession arrangements – which reduced the significance of queens and princesses in both the representation and practice of governance at the royal court in Carthage.

proximate causes was the denial of an attempt by some East Roman soldiers to claim land by marrying female captives. These prisoners were the wives and daughters of men who had served in the Vandal armies. Procopius implausibly presents this claim as the result of these women persuading their new husbands to seek the land. 7 The women thus (implicitly) have legitimate claims on the 'lands of the Vandals' (τὰ Βανδίλων χωρία). Yet the historian does not unambiguously call them Vandal women, but 'the daughters and wives [of the Vandals]' (Βανδίλοι … τὰς αὐτῶν παῖδάς τε καὶ γυναῖκας). In a later passage, he discusses the shipping to the East of 'the remaining Vandals … and all their women [or perhaps "wives" once more]' (καὶ Βανδίλων τοὺς ἀπολελειμμένους καὶ οὐχ ἥκιστά γε αὐτῶν γυναῖκας ἁπάσας). 8 In both cases, these women are characterised not as Vandals in their own right but as the dependents of adult male Vandals. 9 Our other main historiographical account of Vandal rule, Victor of Vita's History of the Persecution of the African Province, similarly spurns multiple opportunities to present elite women in clear ethnic terms. Victor's History is a polemical account of the reigns of Geiseric and Huneric (r. 477-84), designed to present the kings as tyrants and the Vandals as heretical savages. It both documents and responds to the recurring efforts of both kings -and particularly the latter in the final year of his reign -to enforce their favoured Homoian ('Arian') form of Christianity as orthodoxy within the kingdom and establish a Homoian church, which resulted in the categorisation of Nicene ('Catholic') bishops and their churches as heretical. 10 The History opens with the Vandals crossing the Straits of Gibraltar. 11 At that time, Geiseric ordered his followers to be counted: to pad the numbers and reach a figure of 80,000 Vandals, Geiseric and his agents included 'old men, young men, children, slaves and masters'. 12 The missing category of persons is especially notable since Victor had just writtenwith the rhetorical exaggeration for which he is renowned -that Geiseric had ordered that 'the whole multitude be counted which the womb had brought forth into the light, right up to that very day'. 13 Clearly Victor envisaged women travelling among Geiseric's warband. Yet in this defining moment of Vandal group formation, women are not thought to count as Vandals: whether by our author, by Geiseric and his agents, or both. Like Procopius, Victor went out of his way to avoid labelling these women as such. 14 The obvious explanation for the lack of Vandal women is that only men were considered to be Vandals. As Andy Merrills and Richard Miles have recently put it, in perhaps the most sophisticated current account of Vandal ethnic identity, For the writers of our textual sources, the most interesting features -one might say the definitive features -of Vandal identity were overwhelmingly masculine. 'Vandals' were primarily soldiers, administrators or landlords who held their land by right of (male) inheritance, who fought and governed on behalf of their Hasding kings and who assumed the engendered trappings of the late Roman military aristocracy. 15 The Vandals were not a 'people' in a straightforward sense but a warband which conceived of its cohesion in ethnic terms: one moulded by shared experience of military activity on Roman soil. 16 They exemplify the scholarly consensus that the collective identities of 'barbarian' groups were repeatedly refashioned across the decades they spent on Roman soil, and the shared ethnic affiliations understood to hold them together were -as with all collective identities -contingent, subjective and situational. Vandal identity appears more obviously as a construct than most: the 'Vandals' who gratefully received land in Africa Proconsularis likely included Hasding and Siling Vandals, Alans, Sueves, Goths and Hispano-Romans. 17 These men formed a new elite in North African society whose status was underpinned by service to the king and by the sortes Vandalorum. This is not to say that the civilian aristocracy was supplanted: many members of the traditional Romano-African elite prospered in post-Roman Africa, partly by serving the Hasding dynasty. 18 Contemporary classicising poetry suggests Vandals and Romano-Africans alike participated in classic aspects of Mediterranean elite lifestyles like the construction of villas, bathing and hunting. 19 But what is notable, as Merrills and Miles argue, is a newfound stress on the military prowess of elite men.
In its frequent limitation to adult men, Vandal group identity reflects wider cultural notions regarding ethnicity in the fifth-and sixth-century West. It is not unusual to find late ancient authors predominantly discussing the male members of the 'barbarian' groups who engaged in military activity in the western provinces, settled and established new kingdoms. Julia Smith and Guy Halsall have stressed that membership of these ethnic groups was principally thought the preserve of adult military men. 20 It is rare for contemporary authors to call women Goths, Franks or Burgundians; law codes likewise exclude women and children from membership. 21 As Halsall has put it, 'ethnicity, especially non-Roman, was essentially masculine'. 22 Part of the result of the establishment of the Vandals within the aristocracy of North Africa -as with parallel settlements in other provinces of the fifth-and sixth-century Roman Westwas a militarisation of elite masculinity, of which the reduced visibility of elite women was a by-product.
I do not wish to contest this picture of Vandal identity as principally a form of military masculinity, nor the careful work on gender and ethnicity across the post-Roman West which has produced it. But it represents only a partial answer to the problem of the missing Vandal women. One of the more curious aspects of the textual and material traces of fifth-and sixth-century North Africa is how rarely we come across specific people (as opposed to a generic collective) being called Vandals. There are a few unnamed individuals in Victor of Vita's History of the Persecution: a military commander (millenarius) who tried to force two of his slaves to become Homoian Christians and two Vandals who joined Nicene bishops in exile. 23 Both Victor and Procopius also refer to the Vandals generically as a group, the latter generally in reference to the armies of the last Vandal king, Gelimer (r. 530-34). 24 The History of the Wars correspondingly refers to a handful of military commanders as Vandals. Otherwise, contemporary authors, when seeking to ascribe some form of ethnic difference, tended to use the pejorative lump term 'barbarian' (or, in a couple of cases, 'Goth'). 25 Other texts, seeking to laud specific individuals, do not specify their ethnic affiliation. A Vandal identity is often read into these texts by modern scholars through interpretation of their names as 'Germanic' or because of the qualities valorised (for example, military prowess). 26 The absence of 'Vandal' women is thus a result of the rarity of specific people being called Vandals. In fact, those people called Vandals by extant texts are exclusively members of the Hasding dynasty. 27 The Wandalae we should expect to find are royal women; it is they who are never mentioned as Vandals and, in fact, rarely mentioned at all. 28 These missing Vandal women call attention to the gendering, not just of perceptions of ethnic identity, but of power, dynasty and governance in post-imperial Africa. This paper will seek to explain this dearth of descriptions of Hasding women as political actors. I will argue that this absence is not simply a question (as ever) of partial, fragmentary early medieval texts, but reflects -or, at a minimum, draws attention to -a contemporary lack of visibility. This is not to suggest that Hasding women were never discussed by contemporary authors or used by regimes, nor that they had no influence on the politics of the court or the governance of the kingdom. This paper will set out the various surviving references to the consorts and female children of Hasding kings and princes. Nevertheless, these royal women are remarkably few and flimsily attested. We know of only two queens (of whom one is named) and a slightly larger number of children and princely consorts (of whom two are named). The limited preservation of even basic information about these women reflects, in the first place, their reduced significance in the Hasding dynasty's self-representation as the legitimate rulers of Africa. Beyond that, it suggests the limitations on their opportunities for political agency within the Vandal court. I will argue that this lack of visibility and agency can be understood structurally as a consequence of the idiosyncratic succession arrangements of the dynasty, which made the eldest living adult male descendant of Geiseric the heir apparent. These arrangements made it unlikely that a Hasding prince's first wife would survive to his accession and even less likely that the mother of a king would do so, significantly reducing her political capital as a transmitter of the dynasty. In theory, these arrangements also prevented royal minorities, the situation which, above any other, activated the potential power of imperial and royal women in late antiquity and the early middle ages; in practice, they resulted in a wider distribution of authority through the households of adult Hasding princes, and a series of succession crises which removed influential female members of the dynasty. All in all, the reduced significance of the women of the Hasding dynasty has consequences for how we view government in Vandal Africa.

Vandal queens
There are only two reigning Vandal queens we know about for certain, and only one with a name. We do not even know if there were queens in Carthage for most of the Vandal century in North Africa ( Figure 1).
A partner of Geiseric is never mentioned, although he had at least three (and possibly four) sons and perhaps a daughter. 29 It is possible that his children were all born before the capture of Carthage in 439. Certainly the eldest, Huneric, was already old enough to be married, in the late 420s or 430s, to an unnamed Visigothic princess. This diplomatic match (and Huneric's seniority) must certainly predate the context in which it is mentioned: Geiseric's violent dissolution of the marriage in 442 so that his heir could be engaged to a member of the imperial family. 30 Huneric was a widower when he succeeded Geiseric in 477. 31 No wife is mentioned for Gunthamund (r. 484-96), Hilderic (r. 523-30) or Gelimer: although Hilderic must have had a consort at some point in his adult life because Procopius reports that the king's unnamed children and grandchildren were rewarded by Justinian. 32 Both the royal consorts whom surviving texts discuss were partners of Thrasamund (r. 496-523). The only named wife of a ruling Vandal king is Amalafrida, the sister of Theoderic, the Ostrogothic king of Italy. Theoderic married Amalafrida to Thrasamund in 500 as part of a series of marriage alliances with other barbarian kings. 33 In discussing the alliance, Procopius reveals that this was Thrasamund's second marriage, that his previous (unnamed) wife had died shortly beforehand, and that they had not had children. 34 The diplomatic calculus is clear from Procopius: Amalafrida brought a chunk of Sicily (contested territory between the two kingdoms) and 5,000 Gothic soldiers. What we know of Amalafrida's time in Carthage comes from two episodes: a diplomatic contretemps and her brutal murder after her husband's death. Some time in the late 500s or early 510s, Thrasamund took in Gesalec, the Visigothic king exiled after his defeat by Theoderic, and furnished him with financial aid to return to Gaul. 35 The incident is described in two letters in the Variae of Cassiodorus, the twelve-book collection of administrative letters which the Italian senator and official wrote on behalf of Ostrogothic kings. Writing in Theoderic's name, Cassiodorus chastised Thrasamund for this decision, playing on the expectation that marriage to Amalafrida and his wife's advice would have led him to act otherwise in relation to his brother-in-law's enemy. Thrasamund backed down and cordial relations were just barely restored. The second breach between Ravenna and Carthage involving her did not receive even this provisional resolution. After Thrasamund's death and the accession of Hilderic, Amalafrida fled Carthage to take refuge with 'barbarian' (presumably, Moorish) allies; she was captured after a military defeat by Vandal forces at Capsa (mod. Gafsa, Tunisia). Hilderic imprisoned Amalafrida and had her Gothic guard killed, claiming she had conspired against him. Konrad Vössing has persuasively argued that this accusation pertained to an attempt to avert Hilderic's elevation in favour of another Hasding, and possibly his eventual usurper, Gelimer. Amalafrida died in prison, possibly as a result of Hilderic ordering her murder. 36 Cassiodorus, acting as the royal mouthpiece once again, wrote a searing letter to Hilderic on behalf of Theoderic's successor Athalaric. 37

Explaining absence: textual representation and dynastic politics
Cassiodorus' letters regarding Amalafrida illustrate the potential aspects of a queen's role -and, more critically, the discourse around it -which are otherwise missing from Vandal Africa. Cassiodorus portrayed Amalafrida as the noble adornment of the Amal royal gens. 38 This dynastic significance was critical to the representation of late ancient empresses and early medieval queens. Images of these women were created to reflect the morality and stability of the regime, centred on the court as a household. Of course, the centrality of images of dynastic women -generally created by and for male rulers and their subordinates -did not necessarily imply actual political agency on their part. As Julia Hillner has recently stressed, it is misleading to suggest a simple relationship between references to imperial or royal women and their actual influence or agency, not least because of the way in which they were used to construct arguments for or against the legitimacy of male rulers. Even in moments where imperial or royal consorts appear to act independently, they were potentially still mere proxies for the will of an emperor or king. 39 Nevertheless, Cassiodorus' expectation that Amalafrida would be a moderating influence and counsellor for Thrasamund finds numerous parallels. 40 The degree to which the position of empress or queen was a 'formal' office, and the extent to which they had access to, or could bypass, more 'public' administrative structures, is a recurring debate across late antiquity and the early middle ages. 41 However their role was theorised within a specific political context, proximity to the ruler lent consorts an obvious significance as an adviser and courtier. This role was often presented in highly gendered terms, as in the image of nurturer and organiser of the royal household traced across early medieval contexts by Janet Nelson. 42 Such activities are by no means the full compass of what empresses and queens can be seen doing. Criticism of specific women for overstepping these marks often reveals these other possibilities, while making plain the practical and ideological restrictions that could limit them. Above all, Cassiodorus' letters show what is fundamentally missing in Vandal Africa: an author interested in the political role and representation of a queen.
This could simply be a function of literary survival. Vandal Africa is particularly poorly served (even by post-Roman standards) by texts which might allow historians to reconstruct the new regime, its self-presentation and its political structures. 43 Our sense of these institutions is cobbled together from heterogeneous (and often polemical) materials. The necessary centrality of Victor of Vita's History of the Persecution is symptomatic. It could be that the invisibility of Vandal queens simply reflects the unsatisfactory evidence for Vandal rule in general. Yet it could nonetheless be expected that work hostile to Vandal rule would highlight the (similarly iniquitous) role played by queens. Late ancient and early medieval authors were not shy of using imperial or royal women as vehicles for criticism of rulers. 44 Such invective was especially prevalent in texts like Victor of Vita's History, designed to excoriate Geiseric and Huneric as persecuting tyrants. Victor's History and the Life of Fulgentius of Ruspe, in particular, recorded the palace machinations which led to the (for them) illegitimate punishment and exile of Nicene bishops. In these texts, rival Homoian bishops tend to be the advisers instrumental in provoking episodes of persecution. 45 By contrast, any number of other late ancient texts highlighted the roles of empresses in ecclesiastical controversies. Gendered invective was directed by the representatives of marginalised church factions against the likes of Justina, Eudoxia, and Pulcheria for their (supposedly) central roles in exiling their champions or favouring their opponents. 46 Contemporary African Nicene authors were aware of the power of such accusations. In texts recounting early fourth-century history, Vigilius of Thapsa and the author of the anonymous Carthaginian Epitome repeated earlier denunciations of Constantia, the sister of Constantine, as the enabler of the rehabilitation of the heresiarch Arius and his followers in the final years of the first Christian emperor. 47 Against that backdrop, the absence of any accusation of persecution against a female member of the Hasding dynasty is striking. It implies that the invisibility of Vandal queens is not simply a matter of fragmentary, hostile evidence.
I would instead suggest that the absence of Vandal queens is a structural issue rooted in the idiosyncratic inheritance arrangements of the Hasding dynasty. The starting point for each succession was an order established by Geiseric some time before his death and likely early in his reign. This 'testament' directed that Geiseric's eldest living male relative should inherit the throne. 48 One consequence of the order and Geiseric's own extraordinary longevity was a greater likelihood of middle-aged or elderly monarchs. Huneric in 477 and Hilderic in 523 were most probably in their 60s; the age (and military non-activity) of the latter was heavily remarked upon in the context of Gelimer's usurpation. 49 It was therefore much likelier that the consort of whichever Hasding inherited would have died before their accession, as was the case with Huneric's wife Eudocia, who had both left Carthage for Jerusalem and passed away by 477. 50 These aging Vandal kings may simply not have had living wives.
Even that simple explanation would be of major significance for our understanding of Vandal government. The court in Carthage would not have had a central female figure; a series of practical and ideological implications would have resulted. Yet barring a succession of remarkable examples of royal continence, we would nevertheless expect at least some of these Vandal kings to have remarried -or, at the minimum, to have had female sexual partners. I do not wish to impute too great a formality to marriage in this context, outside of formally recognised diplomatic alliances. The fluidity of political marriage in the post-Roman West is most obvious in the case of Merovingian Gaul, where the status of royal consorts varied greatly. Gregory of Tours describes some Merovingian consorts as concubines without this less formal partnership detracting from their political resources. 51 It is equally likely (given the lack of evidence for royal wives who must have existed) that Vandal kings did have partners, but they were insufficiently significant in the self-representation of the regime or the constellations of the court to be mentioned in extant texts.
The best explanation for this reduced significance is that Vandal succession arrangements placed extraordinary limits on their potential influence and symbolic utility. Path-breaking studies in various contexts have shown the importance of the succession to both the visibility and the power of early medieval queens. 52 In practical terms, the death of the monarch was a key moment around which the power of an empress or queen could coalesce, with the accession of her legitimate son. It was a crucial (though not the sole) motivating factor for the depiction of a royal consort in various media. 53 In a more traditional succession, there was the possibility of a minority, in which the new ruler's mother would act as a key powerbroker. 54 Even during their husbands' reigns, this potential future lent empresses or queens significance within the court.
Given these succession arrangements, and their stage in the life cycle, Vandal kings were (in theory) less likely to need a formally recognised marital partner to legitimate potential offspring. Instead, in surviving texts, it is the dynasty's founder Geiseric who plays this role; descent from him through the male line is key for legitimate succession. In practical terms, the current queen was unlikely to be the mother of the next ruler. There were always adult male relatives of Geiseric whose power was built up using satellite courts in and around Carthage. 55 The fall of Amalafrida provides a neat illustration of this problem. Hilderic's accession produced a course change in Hasding diplomacy, with a distinct turn away from Ravenna and his predecessor's alliance with Amalafrida's brother Theoderic, and towards the East Roman court in Constantinople. If the allegations of conspiracy made against her are accurate, the former queen had to push for an alternative candidate for the throne to maintain her position of influence at court, and her gambit failed. The lack of a direct biological and ideological involvement in the royal succession significantly reduced the opportunities for royal consorts to build up networks of power in Carthage. More narrowly, it reduced the utility for regimes of placing them on display.
Of course, as Merrills has pointed out, Geiseric's 'testament' was challenged on a number of occasions. 56 These succession crises highlight the capacity of a wider contingent of Hasding women (discussed in a later section) to position themselves as influential political actors. Contemporary discussions of these women -and their often grisly fates -imply that their agency was seen as crucial to the assemblage of aristocratic factions to support these (apparent) conspiracies. I have already mentioned Amalafrida's putative attempt to divert the succession and her marshalling of the support of Moorish allies. As with Geiseric's extraordinary longevity, the impact of his succession arrangements on Vandal queenship was contingent on the unpredictable results of dynastic plots, coups and ensuing purges. In other circumstances, these influential princely consorts might have assumed the queenship while bringing with them the patronage relationships developed in the process of their husbands' efforts to short-circuit the succession. As it was, they were killed or exiled in the dynastic purges which resulted. These queens-in-waiting were unable to translate their influence to the royal court. The somewhat unsettled (because never entirely normalised) system of succession acted as an obstacle to female political agency by resulting in the murder, exile or repudiation of powerful (and thus threatening) royal women. 57

Petitions and persecution: ecclesiastical politics without queens
Whether we see it as a result of the terms of Geiseric's testament, or the violent reprisals against royal women which resulted from its contingent application in moments of dynastic stress, Vandal queens do not appear as influential political actors in contemporary texts. The polemical literature produced by Nicene clerics shows a number of occasions when representatives of that church sought to petition their rulers for specific rulings or changes to royal policy. In these accounts, we never see a queen used as a route to access power. Instead, where intermediaries are noted, the recourse of subjects to the king is directed through his male officials and, in particular, the individual in control of his court: the praepositus regni or maior domus. 58 Victor of Vita's portrayal of negotiations between Huneric and the Nicene bishops shows petitions directed to the king, and responses and clarifications transmitted back from him, through the praepositus or a designated notary. 59 The anonymous author of the Life of Fulgentius likewise depicted a single channel of communication between Thrasamund and Fulgentius of Ruspe through the king's specified agents when the former invited the latter to Carthage for doctrinal debates. 60 Victor and the anonymous hagiographer undoubtedly undersold these bishops' options at a court which included Nicene Christians among its staff. 61 This pattern may nonetheless capture something of the shape of the Hasding court. The absence of an influential royal consort took away one of the critical 'alternative routes to power' for petitioners, through whom they might seek changes of policy or clemency in the practical application of royal judgements. It likewise shut down the options of regimes in policymaking, including a manner of receiving appeals favourably without committing too visibly to a change of course. 62 The absence or absent influence of Vandal queens could thus limit what was politically possible in post-Roman Carthage.
The subject of those petitions suggests another potential consequence of the reduced power of royal women. Vandal regimes differed from their Homoian counterparts in the other post-imperial kingdoms in their recurrent promulgation and enforcement of anti-heretical laws against the Nicene Church. This led to the portrayal of Vandal kings by Nicene authors as tyrannical persecutors, in stark contrast to the generally favourable views of Burgundian and Ostrogothic kings articulated by Nicene authors in Southern Gaul and Italy, founded in a more conciliatory relationship between these regimes and the Nicene ecclesiastical institutions within their orbit. 63 The promulgation and enforcement of anti-Nicene policies by Vandal kings was neither inevitable nor consistent. Their application was particularly strict and wide-ranging under Huneric in 484, in the aftermath of which Victor wrote his History. But the regimes of Geiseric, Huneric, Gunthamund, Thrasamund and Hilderic also sought, on occasion, to placate Nicene churchmen; those clerics were likewise, on occasion, able to conceive of their rulers in more conciliatory terms. 64 Nevertheless, a significant contrast remains between Vandal Africa and the rest of the post-imperial West in the nature of royal policies, the relationship between Homoian regimes and Nicene churchmen, and the depictions of rulers that resulted. The causes of this contrast have been the subject of considerable debate. 65 I would suggest that part of the reason for these differences in royal policy and reputation is the absence of a visible and visibly influential queen.
Fifth-and sixth-century regimes across the Mediterranean world used the ruler's consort or female relatives to patronise (and thus contain) opposing factions in ecclesiastical controversy. In the Burgundian kingdom of Southern Gaul, there were both Nicenes and Homoians in the Gibichung royal family. In Ostrogothic Italy, the Homoian Theoderic's mother, Erelieva/Eusebia, was a patron of Nicene churchmen. Perhaps most famously, in the Eastern Roman Empire, Justinian and his consort Theodora worked as a double act to conciliate groups who supported and opposed the formula of the Council of Chalcedon (451). In the latter case, it is clear that these parallel initiatives helped to prolong ecclesiastical dialogue and maintain the political loyalty of the anti-Chalcedonians, even as they began to establish separate church institutions. 66 Imperial or royal women could be usefully deployed to moderate anti-heretical policies, 'capture' ecclesiastical dissidents, and reduce the charge of doctrinal or ecclesiastical differences in relationships between regimes and churchmen.
Without an influential royal consort as a counterweight at court, Vandal regimes may have found it more difficult to balance rival Christian petitioners when they sought to offer patronage to those they otherwise presented as disloyal heretics. The political absence of queens reduced these regimes' ability to encourage usefully openended interpretations of those policies on the part of ecclesiastical dissidents. The History of the Persecution and the Life of Fulgentius recall the conciliatory moves of Huneric and Thrasamund as the result of 'barbarian subtlety', deliberate deception or underlying persecutory agendas. 67 Such disillusionment obviously resulted, in the first place, from the recourse both kings subsequently made to outright coercion. Yet the failure of these initiatives and their later rationalisation also plausibly stem from the nature of the communication between these kings and their Nicene bishops. Coming from a single official channel presenting the king's own personal views, offers to reopen doctrinal questions or make concessions were time limited, and liable to appear to Nicene observers as (at best) disingenuous and (at worst) deviously malicious. Coming from a royal consort -a semi-autonomous figure prominent within the regime, but outside the formal institutional hierarchy -as one of multiple simultaneous initiatives, such encouragement might have more successfully contained Nicene opposition. The political invisibility of queens may thus help to explain the particular ecclesiastical policies enacted by Vandal regimes and the malign reputation of Vandal rule that resulted in the fifth-and sixth-century West. Whatever the value of this (necessarily speculative) hypothesis, that absence -and its impact on the representation and practice of governance -remains clear.

Violence, conspiracy and satellite networks: locating Hasding women
Hasding succession practices limited both the visibility and the influence of Vandal queens. Paradoxically, the turmoil these practices intermittently provoked is in several cases the sole reason we can glimpse a wider contingent of royal women in extant texts. These women appear, either when the stability of the regime or the succession seemed at stake, or in the context of diplomatic alliances. In the recounting of these episodes, royal women are, once again, more often acted upon than agents. Nevertheless, the moments of political crisis in which they appear do imply significant influence within the elite networks of the kingdom. They also suggest that such influence was not tied to the authority of the king or located at his court, but rather situated in the separate households of Vandal princes. In that sense, once again, the detachment of the women of the Hasding dynasty from the representation and exercise of Vandal kingship is obvious.
Political murders after supposed conspiracies are the inciting incidents for most references to Hasding women, reflecting the more commonplace political violence of the post-Roman West. 68 To locate that agency which was available to them requiresagain, paradoxically -the narration of a series of incidents of violence against them, leading to their murder, mutilation or sexual exploitation. Early in Geiseric's reign, the king had the unnamed widow of his brother, Gunderic, drowned in the river Amsaga near Cirta in Numidia (mod. Constantine, Algeria). He went on to murder her unnamed children, who were, of course, potential rivals to the throne. Victor of Vita does not date the events; they have been placed either soon after Gunderic's own death (so presumably in the early 430s) or in association with the revolt which Geiseric suppressed in 442. 69 Another act of political violence might also connect with that revolt. The sixth-century Constantinopolitan historian Jordanes reports a conspiracy as the pretext for the mutilation and repudiation of Huneric's first wife, an unnamed Visigothic princess, suspected of an attempt to poison Geiseric. The king ordered her nose removed, her ears mutilated and that she be returned to her father in Gaul. 70 The likelihood is that this accusation was concocted -or seized on -to provide king and heir grounds for unilateral divorce, so that a match could be arranged between Huneric and the five-year-old daughter of Valentinian III, Eudocia. Jonathan Conant has drawn out the significance of this betrothal within Geiseric's wider diplomatic manoeuvres in the Mediterranean, seeking recognition from Ravenna of his position within imperial power politics. 71 These machinations would lead, after Valentinian's murder, to the sack of Rome in 455 and the forcible abduction of Eudocia with her mother Eudoxia and her sister Placidia. Some time after she was brought to Africa, Eudocia was married to Huneric: a marriage given imperial blessing after a deal with the Eastern Emperor Leo in 461/462 which saw Eudoxia and Placidia sent to Constantinople. 72 Eudocia was kept in Carthage for perhaps sixteen years, providing Huneric with a son, Hilderic, and other unnamed children, before following the path of previous fifthcentury imperial women and going into retirement in Jerusalem, where she died in the early 470s. 73 Eudocia's enforced sojourn in Carthage is a black box. It exemplifies the analytical and ethical difficulties of reconstructing the experiences of late ancient dynastic women. The imperial princess's value for the Hasding dynasty is obvious. Geiseric and Huneric repeatedly used this family connection to press their case at Constantinople; the Eastern imperial court responded in kind, later using one of Eudocia's family as an envoy. 74 Her symbolic importance is patent, and can be seen in Huneric's (failed) attempts to make his son by her his heir. Once Hilderic finally became king, his regime touted his imperial heritage as a sign of his legitimacy. A short panegyric by the Carthaginian poet Luxorius, preserved in the Latin Anthology, describes images in the king's residence at Anclae celebrating his dual lineage and his descent from the emperors Theodosius, Honorius and Valentinian III. 75 Yet where this leaves Eudocia as a potential political actor -as opposed to a passive receptacle for Hasding imperial grandstanding and diplomatic claims, and an incubator for Hasding offspring -is much more difficult to ascertain. 76 Her marriage to Huneric may have been long awaited (a match made when she was four or five years old), but the context of its forced enactment in captivity robbed her of the political resources and networks which might normally support the (still fragile) position of women used for diplomatic marriages. 77 The imperial treaty which more firmly established her position will have at the same time removed her local support network. The ongoing turmoil at the western imperial court, when combined with Geiseric's continued pursuit of her inheritance as a condition in treaty negotiations, will have isolated her from wider political and diplomatic aid. 78 Eudocia's first recorded active move in Carthage was her last: her supposed flight to Jerusalem. Even then, it is difficult to see how she could have left without royal permission in the manner the early ninth-century historian Theophanes suggests. 79 It may be that Eudocia's position is best understood -as with the parallel case of Galla Placidia recently revisited by Victoria Leonard -as that of 'a war captive and a "spear-won bride"'. 80 Certainly, her lack of consent to this marriage and the sexual intercourse which followed should not be glossed over: in fact, unlike for Placidia and Athaulf, late ancient and early Byzantine authors did not explain away the coercion involved. 81 At the same time, it is important to distinguish her lack of sexual agency from her potential political role. 82 A decade and a half in Carthage, including a decade of imperially acknowledged marriage which led to the production of legitimate heirs, seems a long time to remain entirely politically isolated. A significant parallel is provided by the capacity of both war captives and enslaved women to establish themselves as political players in Merovingian Gaul once their royal possessors had made them into wives or concubines. 83 Another imperial figure who had sought refuge -albeit actively -in Carthage after exile from both the imperial and Visigothic courts, the count Sebastian, was able to become a (threateningly) significant player in Carthage in the 440s. 84 It is possible that, as the consort of the heir apparent, Eudocia was able to attain a position of greater political influence, but this remains unknowable: all we are told of what Eudocia did in Carthage was to give birth and leave.
Barring Damira, the daughter of the Vandal prince Oageis -recipient of an epitaph from Luxorius on her death aged just three -the remaining known Vandal royal women appear in the context of Huneric's purge of rival family members and their supporters in the early 480s. 85 These included the execution of the unnamed wife of his younger brother (and presumptive heir) Theoderic on charges of conspiracy; after Theoderic's death in exile, the ignominious exile of his unnamed daughters; and the exile of his nephew Godagis with his unnamed wife. 86 As with Vandal queens, so with this wider category of royal women: we simply do not see them in the everyday business of government. The extreme situations in which they figure nevertheless highlight the influence these women were imagined to have in the networks of the court. Of the various accounts of brutal maltreatment after accusations of conspiracy, Victor of Vita's discussion of Huneric's purge is most explicit. Theoderic's unnamed wife was 'cunning' (astuta) and thought able to 'arm her husband and their elder son … with more pointed counsels against the tyrant' (that is, Huneric). 87 These supposed conspiracies highlight the continued resources which these royal women were thought to have, especially in the context of the separate households granted to Vandal princes. These households acted as a locus for the loyalties of wider sets of officials and aristocrats. 88 One of the (multiple) stages of Huneric's purge, as described by Victor, was the execution of a larger group of counts and nobles seen as Theoderic's supporters. As previously noted, the implementation of Geiseric's testament was not inevitable: it could be contested on each occasion a transition of power had to take place. 89 The overthrow of the current king was also always a possibility. This jockeying for position provided opportunities for royal women to act as powerbrokers. But these opportunities (and the accusations they fuelled) were lent a particular shape by the wider distribution of Hasding power: both metaphorically through the royal family and spatially with the establishment of parallel households as alternative powerbases. Vandal royal women are accused of plotting from the outside to overthrow the king, rather than manoeuvring inside his household (whether at his palace within the city of Carthage, most likely the old proconsular seat on the Byrsa Hill, or one of his own suburban residences) to exercise influence against other royal favourites. 90 These reports on supposed conspiracies are the closest surviving texts come to presenting female political agency at the Vandal court. Yet again, they serve to demonstrate the distance of these women from the centre of power.

Conclusion
As part of his account of the aftermath of the reconquest, Procopius of Caesarea describes the triumph which greeted the general Belisarius on his return to Constantinople in 534. The historian notes that, among the spoils of the Vandal war paraded were 'carriages in which it was customary for the king's wife to be carried' (ὀχήματα οἷς δὴ τὴν βασιλέως γυναῖκα ὀχεῖσθαι νόμος). 91 These carriages suggest the participation of Vandal queens in the prestigious display of the court, in the last decade of Vandal rule at the very least. It could be envisaged that, as Amalafrida or the consorts of Hilderic or Gelimer made their way through Carthage, or progressed in state to suburban residences, they acted as a visible reminder of the unique power of the Vandal regime and the Hasding dynasty.
Belisarius' captured carriages suggest that we should not underestimate the investment of Vandal regimes in queens (and royal women more broadly) as manifestations of their power. In discussing the limited surviving evidence for Hasding women, this paper has already drawn attention to a number of facets of their significance both (passively) as symbols and (actively) as agents: the centrality of the Visigothic princess, of Eudocia, and of Amalafrida to the dynasty's forays into Mediterranean power politics, and the resources and networks implied by accusations of conspiracy made against that princess, against Amalafrida, and against the unnamed wives of Gunderic and Theoderic. Moments like Procopius' mention of the queenly carriage or Victor's description of Huneric's purge strongly suggest that royal women could be much more involved in the self-presentation of Vandal regimes and the politics of the court than our texts permit us to see. Yet at the same time, this very absence of evidence points to fundamental limits on female political visibility and agency in Vandal Carthage.
It should of course be noted that similar constraints appear in parallel political contexts across the late and post-Roman world. 92 Recent individual studies have delineated the changing political dynamics for both the representation and agency of empresses and queens in late antiquity. These accounts suggest that various aspects of the wider political landscape of the fifth-century West contributed to an absence of female political actors from contemporary historical accounts and particular constraints on female political agency. In the first place, an important article by Audrey Becker-Piriou has shown the frequency with which imperial and royal women appear simply as 'pawns in the service of diplomacy' traded between regimes in contexts of often unstable power relations. 93 More broadly, the political framework of the earliest successor kingdoms seems closer to a return to a norm of third-and fourth-century military rulership and thus, to a lesser practical and ideological role for dynastic women. As recent work has stressed, the women of the Tetrarchy were rarely exploited for political capital in a homosocial world of (fictive) male familial bonding between coemperors. As the fourth century went on, the women of the Constantinian and Valentinianic dynasties appear similarly intermittently. It is only with the advent of child emperors in the later fourth and early fifth century that imperial women gained defined profiles, since these regimes' protagonists were (inevitably) court bound, and struggled to assert themselves in adulthood. 94 The children of the Theodosian dynasty set a new pattern in Constantinople until Heraclius in the early seventh century: emperors did not campaign in person. 95 The immobility of their western counterparts set the conditions for the rise of powerful generalissimos, a development which fuelled the ambitions of military leaders like Geiseric. 96 This renewed military rulership set in motion conditions less conducive both to the production of literary representations of consorts and to the establishment of their influence through proximity and ceremonial activity. The transition from leader of a warband to ruler of a kingdom, and the peculiar forms of legitimation it engendered, may also help to contextualise the low profile of Vandal women. Particularly relevant here is Smith's recent discussion of how contemporaries constructed the Merovingian Frankish dynasty through the male line from Clovis -an important parallel to the Hasding dynasty's focus on Geiseric, though in a context of far greater queenly influence. 97 It may be that the continued foundation of these kings' power in their armies -even as they formed lasting relationships with provincial Roman aristocracies -led contemporaries to imagine their power as transmitted from men to men. 98 From quite early on, Vandal regimes commemorated their conquest of Carthage in 439 as the origin of their rule. 99 This wider fifth-century Mediterranean context goes some way to contextualising the Vandal case, but it can only take us so far. In the first place, Vandal Africa is where we might expect to see the quickest transition to more traditional (if still fissile) dynastic forms of legitimation. 100 Certainly, the Hasding dynasty were the earliest of the new barbarian leaders to make claims to independent power and autonomous institutions rooted in collaboration with the late Roman provincial elite. 101 Their strategies of legitimation soon included the self-conscious adoption of aspects of Roman political self-presentation. Moreover, the exploits of Vandal rulers in battle do not tend to be the focus of surviving texts. To be blunt, this is partly because kings after Geiseric did not have many successes to trumpet, although military victory certainly figured as part of the staging of Vandal rule. 102 Instead, some of our main accounts are intensely concerned, not with the battlefield, but rather the goings-on of the court in Carthage, because of their malign influence on the Nicene churches of the kingdom. In that sense, it seems much weirder that Vandal government is so consistently presented as the government of adult men through adult men. It looks especially strange if we turn beyond the transitional fifth century to the longer-term perspective of the early medieval West. In fact, the Vandal court looks much more like how we used to view late Roman or early medieval government, according to more old-fashioned models of Staatlichkeit -perhaps part of why this oddity has been so little remarked on. 103 This paper has thus sought to show how the specific structural constraints of the Hasding court and the culture of the dynasty can explain this absence of royal women from contemporary and near-contemporary accounts of Vandal rule. In so doing, it has sought to contribute to wider efforts to reconstruct the lived experiences of imperial and royal women in late antiquity, by producing fine-grained accounts of individual women and political contexts which go beyond the reiteration of their basic marginalisation or exceptional influence. It has sought to delineate the specific constellations of factors which led to the absence of the representation (and, again more cautiously, the agency) of royal women in the specific case of Vandal North Africa. Pursuing these absent Vandal women has highlighted a series of gendered political phenomena that should make us rethink the nature of governance in the kingdom. I have suggested that this absence from surviving texts can be traced, in the first place, to the consequences of Hasding succession arrangements, which depleted many of the usual resources of consorts in late ancient and early medieval palaces -or, more pessimistically, the usual practical and ideological reasons for regimes to use dynastic women. These arrangements also simply made it less likely that the wife of a Vandal heir would live to see her husband take the throne. I have suggested that the absent influence of Vandal queens may have been part of why relations between Vandal kings and Nicene churchmen repeatedly broke down, and Vandal rule became synonymous with persecution. It exacerbated the difficulties Hasding regimes faced in acting plausibly as unifying figures in ecclesiastical politics in the eyes of contemporary Nicene observers on those occasions they offered concessions, while still openly favouring the competing Homoian faction in the kingdom. Where royal women can be seen mediating political activity, it is not at the king's palace in Carthage or one of his other residences, but in princes' suburban households. This wider distribution of Hasding female political agency underlines the importance, from early on in the Vandal century, of suburban royal residences. It may also imply a more diffuse character to court politics, as various members of the royal family established competing (and potential future) centres of power. Above all, attention to the absent Vandal women of post-imperial North Africa reminds us that the blind spots of late ancient authors should not become our own. 104