This thesis is an attempt to address the surprising lack of study into the question of usurpation in late antiquity. During a period defined by a textual corpus (289-389), the thesis looks at how usurpers and usurpation were presented in the panegyrics delivered to emperors and their courts. That usurpation features very heavily in this corpus should tell us something in itself, but it is a feature of these texts which has hardly been observed. The thesis shows how the panegyrics employed aggressive rhetorical tactics that sought not to bury usurpers in silence but rather to glory in their destruction and to create characters for the usurpers and their regimes that were designed to reinforce the legitimacy claims of the victorious emperor. The language of the panegyrics concerning usurpers and usurpation is thus virtually worthless as a tool to reconstruct the historical actualities of the people and times that they discuss. It cannot be used, as some scholars have done, to give insight into the working of particular usurpations. But the study also demonstrates that the panegyrics are far too valuable a body of sources to simply ignore, as many more scholars have tended to do. The panegyrics demonstrate the beginnings of the processes of memory sanction, or damnatio memoriae, that were imposed upon defeated usurpers and, as such, give us a valuable insight into how imperial Romans recorded their history and conceived of the power structures through which they were governed. Panegyrics are vital to our understanding of usurpers and usurpation because they are the first step in the process of understanding why our narrative sources are so unreliable concerning such men.

En continuité avec les ouvrages récents (Veyne 1981, MacMullen 1988 et Kelly 2004) qui tentent de relativiser les effets néfastes de la corruption lors du Bas-Empire, ce travail étudie le suffragium, le processus de nomination des fonctionnaires de bureaux, afin d'évaluer comment les acteurs sociaux du IVe siècle considéraient ce phénomène. Ce système, organisé d'une telle façon que les hauts fonctionnaires devaient fournir des lettres de recommandation aux candidats postulant à des postes au sein de la fonction publique, serait devenu complètement corrompu durant le IVe siècle et les lettres de recommandation auraient commencé à être systématiquement vendues. Pourtant, les lois de Constantin, Constance et Julien ne fournissent aucune preuve tangible que le suffragium était dans tous le cas vénal à cette époque. Bien au contraire, les empereurs ajoutaient la plupart de temps des épithètes au terme suffragium pour spécifier qu'il parle du suffragium vénal. Généralement, les empereurs sont présentés comme farouchement opposés au suffragium et à toutes les tractations qui y sont attachées. Loin d'être aussi hostiles envers les « pratiques corrompues », les empereurs de la dynastie constantinienne firent preuve d'un certain pragmatisme en voyant qu'ils ne pouvaient contrôler toutes les nominations de ceux qui voulaient entrer dans la fonction publique et que ce n'était pas nécessairement à leur avantage de le faire. Les empereurs se concentrèrent plutôt sur les restrictions entourant les promotions afin de faire en sorte que les personnes qui avaient de réels pouvoirs soient celles qui avaient démontré leurs qualités tout au long de leurs années de service. Bien qu'ils n'aient pas concrètement légiféré sur les critères d'embauche des candidats, cela ne veut pas dire que n'importe qui pouvait obtenir un poste. À travers l'étude des lettres de Libanios et de Symmaque, ce travail démontre que les hauts fonctionnaires ne fournissaient pas de lettres à quiconque le demandait, puisque leur réputation pouvait être entachée par le fait d'avoir recommandé un mauvais candidat à un de leurs amis. Les hauts fonctionnaires qui recevaient les recommandations pouvaient également soumettre les candidats à des examens afin d'être certains de la qualité de l'individu. Ce système officieux de contrôle des candidats vint pallier, en partie, les déficits de la législation impériale. Conjointement, la loi et les usages permirent à l'administration de fonctionner en lui fournissant des candidats qui répondaient aux critères de l'époque. / In line with recent literature (Veyne 1981, MacMullen 1988 and Kelly 2004) that attempts to relativize the negative effects of corruption during the later Roman Empire, this paper examines suffragium, the process of appointing officials, in order to demonstrate that this phenomenon is not as bad as the historiography of the twentieth century makes it out to be. This system, which was organized in such a way that the high officials had to provide letters of recommendation to candidates to positions in the public service, became completely corrupted during the fourth century and the letters were eventually systematically sold. Yet the law of Constantine, Constantius and Julian provides no tangible evidence that suffragium was bought in all instances during that period. On the contrary, the emperors often added an adjective after the noun suffragium to specify that it was bought. Typically, emperors are portrayed as fiercely opposed to suffragium and all negotiations pertaining thereto. The emperors of the Constantinian dynasty were not nearly as fierce against those "corrupt practices and they even showed some pragmatism, having understood that they could not control all appointments of those who wanted to enter the public service and that it was not necessarily in their best interest. The emperors focused mainly on legislation surrounding promotions to ensure that people who had real power were those who have shown their qualities throughout all their years of service. Although they did not specifically stated the criteria for hiring candidates, it does not mean that anyone could enter. Through the study of Libanios' and Symmachus' letters, this paper demonstrates that high officials did not provide letters to anyone who asked for them, as their reputation might have been tainted by having recommended a poor candidate to their friends. High officials who received the recommendations could also ascribe certain tests to the candidates to be certain of the individual's quality. This informal system of control of the candidates compensated, in part, for the deficits in the imperial legislation. Together, these two systems enabled the administration to function by providing candidates who met the criteria of the time.

This thesis investigates representations of dynastic legitimacy and imperial power in the later Roman Empire (AD 284-337). It explores the continuity and change in expressions of dynastic legitimacy by, for and about the emperors of this period, which were presented in coinage, panegyrics, and other literary and material evidence. I argue that familial relationships were used throughout this period to make legitimation claims or to counter claims made by rivals, rejecting the notion of clear breaks between the third century, the Tetrarchy and the reign of Constantine. The Tetrarchy’s creation of familial links through adoption and marriage led to a web of inter-familial relationships that they and later emperors used in promoting their own claims to imperial legitimacy. At the same time, the presentation of these imperial colleges as harmonious co-rulership relied heavily on the adaptation of pre-existing strategies, which in turn would be adapted by the emperors of the early fourth century. This thesis proceeds roughly chronologically, focusing on the regimes of individual emperors and their collaborators when possible. Chapter 1 examines the creation of the Tetrarchy as an extended ‘family’ and the adaptation of ideologies of third-century co-rulership. Chapter 2 explores the changes in the Second Tetrarchy, with an especial focus on the ‘Iovian’ family of Galerius and Maximinus Daza. Chapter 3 looks at Maxentius’ claims to both ‘retrospective’ and ‘prospective’ dynastic legitimacy. Chapter 4 examines Licinius’ legitimacy both as a co-ruler and brother-in-law of Constantine, and as the beginning of a new ‘Iovian’ dynasty. Chapter 5 delves deeper into the different claims to dynastic legitimacy made by Constantine over the course of his thirty-year reign. Taken together, these chapters offer a new approach by arguing against the dichotomy between ‘dynasty’ and ‘collegiality’ that tends to dominate scholarship of this period. Instead they focus on the similarities and continuities between the representations of imperial families and imperial colleges in order to understand how perceptions of dynastic legitimacy evolved in the third and fourth centuries.