Bryn Mawr Classical Review 1999.02.02
Jens-Uwe Krause, Gefängnisse im Römischen Reich. Heidelberger althistorische Beiträge und epigraphische Studien, 23. Stuttgart: Franz Steiner Verlag, 1996. Pp. 365. ISBN 3-515-06976-3. DM 96.
Reviewed by Ian G. Tompkins, University of Wales, Aberystwyth (firstname.lastname@example.org)
Word count: 4178 words
K. has produced a very substantial work that considers many aspects of the function and character of prisons in the Roman empire.1 His discussion includes an examination of the organisation of prisons, their living conditions, and the social origin of prisoners. The main focus of the study is on the prisons of the later Roman empire, and to this end extensive use is made of the law codes of late antiquity and a wide range of sources from this period, including Libanius, John Chrysostom, Ammianus Marcellinus, and John Moschus. The central concern of the study is with the changes that came about in the nature of prison life and attitudes to prisons in late antiquity. K. maps these changes onto wider developments in the late ancient world, particularly the rise of Christianity and changes in the structure of the government of the empire, including the development of much smaller provinces, which, it is argued, enabled governors to carry out their tasks more effectively, including their duty to track down, try and punish offenders. K.'s study is thus at heart an argument for how society operated in late antiquity. Changes in the use and character of prisons reflect more fundamental changes in the attitude of late ancient society to the role of government and to religious belief. A number of basic problems arise from K.'s study, which should be addressed ahead of more detailed consideration of his arguments.
K. utilises a wide range of primary sources, but fails to mention much modern scholarship, particularly in relation to late antiquity: for instance, Peter Brown and Alan Cameron are cited once each,2 while no works by Averil Cameron, John Matthews, or Roger Bagnall are cited. K. refers frequently to the Riot of the Statues in Antioch in 387 (90, 177, 221, 275, 294), but does not mention the basic study of Robert Browning, 'The riot of A.D. 387 in Antioch: the role of the theatrical claques in the later Roman empire', JRS 42 (1952), 13-20. Brian Rapske, The Book of Acts and Paul in Roman Custody (Grand Rapids and Carlisle, 1994), which covers much of the same ground as K., presumably appeared too late to be considered.
Secondly, there is some inconsistency in K.'s treatment of the evidence of legal texts. K. does not develop a consistent methodology for assessing the social impact of individual laws. While individual laws may be cited as evidence for particular social patterns, on other occasions the effectiveness of such texts is questioned. Thirdly, K. focuses on a limited number of developments in late antiquity (the implications of the creation of much smaller provinces, primarily in relation to the enhanced role that this development conferred upon the provincial governors [cf. 29, 44, 201, 246-7, 264-5], and on the changes which followed from the rise of Christianity). K. does not pay sufficient attention to other developments, in particular the demise of the Councils in some provinces, changes in the role of the military, and changes in the structure of land-holding, all of which had significant implications for how the provinces were governed and for how justice was delivered. For instance, K. mentions the demise of municipal court prisons (43-4, 248, 267-8), but does not consider the widely attested, although probably not universal, phenomenon of the decline of curial institutions and membership.3 Fourthly, particularly in Chapters 10-13, the model underpinning K.'s discussion of tensions and conflicts between social classes is insufficiently consistent: discussion of official wilfulness and class justice is not correlated with the consideration of the importance of patronage. Finally, K. refers to a range of sources from the end of the Roman West, and from the immediate post-Roman period, but does not set this material within its historical context, for instance the almost unbroken social unrest within Gaul in the time of Sidonius.
In Chapter 2, K. considers some evidence for the use of imprisonment on remand in the Republic. His central concern is to argue that magistrates in the late Republic retained the right to imprison suspects ahead of the trial and that they did in fact exercise this right.
In Chapter 3, K. discusses the role of the police in detaining offenders. A brief summary is provided of the various police structures in Rome, Italy and the provinces. With regard to Rome, K. notes that several duties, including the maintenance of law and order, fell within the remit of the praefectus vigilum, but that the praefectus urbi, who controlled the cohortes urbanae, more properly considered serious matters (27). The duties of provincial governors normally included the search for and punishment of criminals, for which such military units as they possessed were at their disposal. It is suggested that in late antiquity, the creation of smaller provinces and the transfer of troops from the frontiers enhanced the ability of the governors to police their provinces effectively (28-32). K. emphasises the limited resources and authority outside the immediate confines of their communities which were available to the several forms of police which are attested at an urban and village level, particularly in the Egyptian source material (32-8). The work of policing Rome is also considered by O. F. Robinson, Ancient Rome: City Planning and Administration (London, 1992), 173-95. The wider problems of policing in Cilicia (33) is considered by K. Hopwood, 'Bandits, elites and social order', in A. Wallace-Hadrill (ed.), Patronage in Ancient Society (London, 1989), 171-87. Egyptian police organs are considered by R. S. Bagnall, Egypt in Late Antiquity (Princeton, 1993), 164-9, and by R. Alston, Soldier and Society in Roman Egypt. A Social History (London, 1995), 81-96. Bagnall highlights the role of the nyktostrategoi and riparii, and suggests that the role of the military may actually have declined in late antiquity: beneficiarii are no longer attested after the early fourth century, and complaints were now normally delivered to the civilian praepositi pagi.
In the final section of Chapter 3 and in Chapter 4, K. considers the evidence for the private dispensation of justice. A number of factors are identified which contributed to the bypassing of the state's system of justice: difficulty of access to the governors' courts for many people (43-5); the prevalence of extra-judicial settlements, the scope for which in some cases was eventually legally circumscribed (45-6); the continued existence of house courts under the authority of the pater familias, although the importance of these significantly declined in the course of the empire (46-8); the development of ecclesiastical courts and prisons (48-59); and the use of private prisons, which became a particular problem in late antiquity (59-63). Note also the inscription from Gerasa referring to a prison constructed by the city's bishop, Paul, in 539/40.4
In Chapter 5, K. discusses the three periods in the criminal justice process in which imprisonment was used: the period of remand (64-79); the period between sentencing and the delivery of the sentence (80-3); and the use of penal detention (83-91).5 He highlights the disagreements in the legal sources over whether a loss of social standing was entailed by imprisonment on remand. The main factors in deciding whether a suspect should be detained were the availability of a guarantor, the nature of the offence and the social status of the suspect. Slaves and clients could look to their owners and patrons to stand surety for them, and the evidence of Egyptian papyri shows that the appointment of guarantors was normal practice even among the under-classes. K. concludes from this that distinctions of social class were of minimal importance in deciding whether remand should be ordered. Poverty was not an absolute bar on escaping remand, while wealth was not an absolute guarantee of freedom. K. concludes this section with a consideration of late imperial practice: it is suggested, inter alia, that imprisonment on remand was more common in the later empire.
In relation to evidence for the use of penal detention, K. cites the cases of the detention of members of the Bacchanalian conspiracy and Caesar's proposed lifelong imprisonment for the Catilinarians (83-4).6 Ulpian's emphasis on the role of prison solely as a place of temporary detention and Caracalla's prohibition on the use of lifelong imprisonment may suggest that the practice of penal detention in the empire was not uncommon (cf. Garnsey, 149). The availability of exile and forced labour may have provided suitable alternatives which initially precluded the effective development of the use of penal detention. K. identifies a number of factors which led to the use of penal imprisonment, including the measure of discretion which belonged to governors, and the increasing difficulty which Christian governors found with carrying out executions (contrast Garnsey, 150, for a different emphasis).
In one of the longest chapters in the study (Chapter 6), K. considers the various crimes and offences which could lead to imprisonment: murder, robbery, social unrest, attacks, slander, theft, adultery, magic and astrology, and religious offences. K. argues that there were few changes in penal practice regarding murder between the Republic and the empire, but that in late antiquity, detention on suspicion of murder was more readily ordered, and the state was more actively involved in the search for murderers (93-7). Regarding trouble-making (97-103), most of the cases cited by K. come from late antiquity: the burning of the synagogue of Callinicum; the Riot of the Statues; food riots in Rome; and Donatist-Catholic conflicts in north Africa. K. concludes: 'In der Spätantike nahm die Gewaltbereitschaft der städtischen Bevölkerung anscheinend zu' (98). This claim seems justified, but directly contradicts his earlier claim that 'Nun lässt sich eine Zunahme der Kriminalitaet freilich kaum nachweisen' (24).7 K. suggests that imprisonment for injuries and slanders was only imposed in the more serious cases, but that once again the late ancient state was more ready to detain those accused of these offences (103-7). There were significant differences in the state's attitude to theft between the early and the late empires (108-17). In the early empire, theft rarely resulted in detention, since the offence was treated under civil law, and settlements were regularly reached out of court. The role of the authorities was largely restricted to ensuring that a trial should take place, summoning the accused to appear, and imposing detentions when agreements between the parties were not observed. In the course of the empire, theft became a criminal matter, and by late antiquity, the imprisonment of thieves was more common. K. suggests that the prosecution of magicians and astrologers was intensified in late antiquity: particularly intense purges of magicians were carried out under Constantius and Valentinian, but the state's hostility to these groups varied considerably with individual emperors (121-2). K. considers several aspects of the imprisonment of Christians during times of persecution: the requirement on Christians who remained free to visit their imprisoned fellow believers and to provide them with support; the development of cult around Christian prisoners, who were credited with the power to forgive sins; the conditions in which Christians were held; the exemplary rather than comprehensive purpose of the imprisonment of Christians; and finally the imprisonment of Christians by their fellow Christians (122-31). The sources for the martyrdom and persecution of Christians are considerable, but it is of note that no use has been made of Prudentius' Peristephanon, an important text, which has generated considerable modern interest.8
K. examines two aspects of the imprisonment of slaves (Chapter 7): the circumstances under which they were imprisoned, and the responsibility for their punishment. The two central factors in any decision regarding the detention of slaves were the role of the master and the nature of the offence of which the slave was accused. Slave-owners were responsible for the punishment of offences against themselves by their own slaves, and could also settle disputes involving their slaves and a third party, but the state was responsible, both in Rome and the provinces, for the detention and imprisonment of slaves in the case of serious crimes such as public disorder, maiestas and adultery. Slaves were normally arrested, tortured and executed en masse when one had murdered their owner.9 Imprisonment for minor offences would be more likely in those instances where neither the owner nor a third party was present to offer guarantee for the slave. With regard to the punitive stage of the process, the evidence suggests that justice could be dispensed by both the owner and the state as appropriate. Responsibility for punishing lesser crimes generally lay with the owner, whereas the state was responsible for dealing with more serious offences, for instance false claims to have obtained manumission. In some cases, the owner or the injured party could exercise a measure of discretion in deciding whether personally to exact punishment or to entrust the slave to the state. Legal texts also established a different range of punishments for offending slaves sub poena vinculorum (142-3). K. suggests that during the course of the empire there was a shift in emphasis between the role of the state and the private owner: the former became increasingly responsible for the detention and punishment of offending slaves, and hence the numbers of imprisoned slaves will have increased (136-7, 139, 143, 145).
In Chapter 8, on the imprisonment of debtors, K.'s argument is somewhat impeded by a failure to distinguish consistently between private debts between individuals and debts incurred as a result of a failure to pay taxes. In some of the instances cited by K., particularly from the sermons and hagiographical writings of the church fathers, it is unclear whether the references are to private debtors or tax defaulters. K. contends that in the Republic and early empire the role of the state in imprisoning debtors was limited, and that the treatment of debt defaulters was essentially a private affair, although subject to some legal restraints such as the lex Poetelia (326 B.C.). K. suggests that the imprisonment by the state of private debtors became normative during the course of the empire, while private imprisonment attested in the early empire was by the late empire rendered unlawful (154-5). However, the Egyptian evidence implies, as K. notes, that the imprisonment in public prisons of private debtors was common from early in the empire (158-61). K. discusses in some detail the evidence for the imprisonment of debtors, but does not set this discussion against the context of changes in the nature of taxation on the land and land-holding such as the development of the iugatio. A significant body of legislation from the Republic to late antiquity concerns the treatment of imprisoned debtors, but the significance of much of this material is often unclear. Several laws from late antiquity regulate the treatment of debtors, including the grounds for imprisonment and the use of torture, but the repetitiveness of such injunctions is likely to be indicative of their limited effectiveness and not necessarily proof of the effectiveness of the state's 'terroristischen Methoden' (169).
Women were consistently under-represented among the prison population (Chapter 9). There was a widespread reluctance to detain women. Their imprisonment is attested for a number of crimes, in particular poisoning, together with the detention of seemingly innocent women because of the offences of their husbands. The first instance of the detention of significant numbers of women was entailed by the persecution of the Christians. In late antiquity, more women were imprisoned, but the numbers were still small, while the emperors were concerned to ensure that female prisoners were housed apart from men.
In Chapters 10-13, K. develops a theme touched on in the earlier chapter on slaves, namely the position of the upper classes in relation to prisons, including evidence for the detention of members of the upper classes themselves (10), official corruption in decisions on imprisonment (11), the use of imprisonment as a means of controlling the under classes (12), and the means by which prisoners could be freed, including the role of patronage (13). K. reviews the evidence for the limited use of detention of members of the Roman upper classes in the early empire. Imprisonment was largely restricted to the period between sentencing and execution, and detention, if it took place, would normally be in the form of house arrest. K. suggests that the position of decurions significantly declined in late antiquity. The senatorial order still enjoyed a privileged position as attested in C.Th. 9.2.1 (362), which secured senatorial immunity from imprisonment on remand and limited detention to the period following sentencing. The evidence of Libanius shows that imprisonment had become a tool for use against aristocratic rivals, and that governors were now more ready to imprison decurions for relatively ordinary crimes.
In Chapter 11, K. considers the evidence for 'Beamtenwillkuer', including evidence for official corruption and incompetence at the level of both governors and below, resulting in unjust imprisonments.10 A wide range of evidence is cited, from Verres and St. Paul through to 6th century Egypt. K. questions the extent to which the transition from Republic to empire brought an end to abuses in provincial administration (191-2). The prevalence of corruption was a function of the almost unlimited powers of provincial governors. It is suggested that there are more references in late antiquity to 'willkürliche Haftmassnahmen' (196; K. cites a range of sources including Basil, Augustine, John Chrysostom, and Libanius), but the more intense fears of the governors and their officials reflect not a greater measure of corruption but the strengthening of the state apparatus and in particular the creation of smaller provinces. K.'s discussion of late ancient legislation which sought to curb abuses by officials (200-1) overlooks the collection of laws in C.Th. 1.16 (De officio rectoris provinciae), which define governors' duties and seek to curb abuses.11
In Chapter 12, K. considers the use of imprisonment by the powerful in late antiquity to exert pressure on their weaker neighbours. It is suggested that most prisoners in late antiquity came from the under-classes, and that their position had deteriorated in comparison to the early empire (205, 210-1). This reflects the fact that most crimes were committed by the poor, but also that the powerful used prison as a means of exerting pressure. K. concludes that 'In diesem Sinne ist es legitim, die spätantike Justiz als Klassenjustizen charakterisieren' (210). A number of criticisms of this argument may be raised. Firstly, K.'s terminology in this chapter is disappointingly vague: he refers generally to 'die Reichen and Mächtigen' (204), who oppress the 'Angehörige der werktätigen Bevölkerung' (210), but distinctions must be drawn between differing aristocratic groups within the empire just as the ordo senatorius represents a distinct group in the law codes which is treated quite differently than the provincial decurions.12 Without clearer definitions, any conclusion regarding class conflict is insupportable. Secondly, much of the evidence cited by K. actually relates to issues of land-ownership, and any such discussion should be related to wider considerations of changes in the patterns of land-ownership. Thirdly, K.'s emphasis on the construction of a system which gave special privileges to particular social groups is somewhat at odds with his argument in Chapter 5 regarding the class-neutral operation of imprisonment on remand. Fourthly, it is also difficult to reconcile the emphasis on increased class conflict with the claim of the previous chapter that officials were no more corrupt in late antiquity than before. Finally, K. does not map his argument onto the evidence for the use and influence of patronage which forms the heart of the following chapter.
In Chapter 13, K. considers the evidence for the release of prisoners from detention, both those who had been condemned and those who were awaiting their trials. Three aspects of the process are particularly highlighted: the power of governors and other office-holders both to imprison and to release; the influence that a powerful citizen, particularly a patron, could exercise in obtaining the release of clients or friends; and the meaning and significance of imperial amnesties. K. considers the evidence of imperial laws which sought to control the release of prisoners by corrupt officials, but concludes that the laws were often of limited effect and were widely resisted by officials (217-8). K. does not refer to the extensive modern literature on patronage, both of more general theoretical interest and in particular relation to the ancient world.13
K. next examines the evidence for the duration of the periods of detention.14 It is suggested that imprisonment on remand was generally of longer duration in late antiquity (235), but since, as K. notes, average durations are impossible to calculate (224) and references to the length of individual sentences are mostly inaccurate (238), comparisons are difficult. K. identifies several factors leading to extended periods of detention, including the need to extract a confession from prisoners as well as information on their accomplices (239-40), the appeals process (although governors were permitted to refuse appeals in those cases where the accused had confessed and where a rapid execution could be considered to be in the public interest [229-30]), the need in some cases to consult the emperor (230-2), the excessive amounts of work faced by the courts (236-8), requests by participants for delays in the trials (for which limits were set, both in the number of requests, and the maximum period of delay permitted ), and the availability of suitable games in the case of those condemned ad bestias (228). Governors normally left outstanding cases for their successors to complete: when Albinus, procurator of Judaea, insisted on completing all the pending cases, it was considered worthy of note (226). The detention of Christians seems often to have been particularly long: in this case, it may have been so that the prisoners would eventually forswear their faith (232-4). Late imperial legislation sought to restrict the periods of detention, requiring that trials should take place within one month of detention and threatening governors with punishments if they failed to observe this limit, to encourage the rapid performance of sentences and to define the conditions under which appeals could be made (240-6).
Chapters 15 and 17 consider the organisation of the prisons, including responsibility for their control in both Rome and the provinces (248- 64), the development of these structures in late antiquity (264-70), the role of prison guards (305-8), and the measures taken to prevent the flight of prisoners (308-15).15 K. assembles a mass of detail beginning with prisons in Rome. He highlights the shift in responsibility between municipal authorities and the governors and their staff for the control of prisons. The early empire attests the existence of municipal prisons alongside those controlled by the governors and their retinues, but later the municipal authorities came to play an increasingly less significant role in detaining prisoners, while the governors' responsibilities grew, particularly as the provinces were increasingly subdivided into smaller units. K.'s discussion of prisons in Rome (248-51) does not go beyond Pliny, and the particular circumstances of Rome are not alluded to in the separate discussion of the late ancient evidence (264-70). As suggested earlier, the discussion of the role of the municipia lacks the wider context of the demise of the Councils. Evidence is cited from Augustine in relation to the Circumcellions and Donatists for the continued vitality of the municipal institutions (268-9), but the north African situation may be considered unusual in that urban institutions may have continued to thrive here into late antiquity.
In Chapter 16, K. considers the evidence for living conditions within prison,16 including the location of prisons, their physical structure and the number of prisoners. K. notes that sources regularly refer to the cramped conditions, the hunger, the chains, and the filth, and that these could be worsened by particular incidents, such as campaigns against religious groups and wide-scale social unrest such as the Riot of the Statues.17 Late imperial legislation sought to ameliorate some of these conditions, partly under the influence of Christianity, but real improvements are hard to discern (282-3, 286, 288, 296). Variables which could intermittently result in improved conditions such as freedom from chains and access to bathing facilities included the possession of high social status, but such considerations were without binding validity as the case of Verres shows. K. briefly considers the evidence for the use of torture (291-4); it is noted that its main purpose was to extract a confession and information on accomplices, and that it was, in the absence of forensic methods, an essential part of the criminal justice process.18
In Chapter 18 K. draws together his diverse observations on prisons in late antiquity. He discusses four main aspects of the role of prisons in this period: the charitable activities of the church (316-23); the rejection of the operation of state justice by the church (323-30); late ancient penal practice (330-40); and late ancient legislation on prisons (340-4). Late imperial legislation on prisons was generally concerned to accelerate trials, to improve the conditions in which prisoners were held, to regulate the conduct of prison guards, and to promote the governors' and bishops' oversight of the prisons, but the extent to which these laws actually affected the realities of prison life is clearly another matter (340-4). The involvement of Caesarius of Arles and Avitus of Vienne in ransoming prisoners also merits consideration.19
1. The author's previous works include Spätantike Patronatsformen im Westen des Römischen Reiches, Vestigia: Beiträge zur alten Geschichte 38, Munich, 1987.
2. 298, n. 32, and 327, n. 64. K. refers to Brown's seminal study in JRS 61 (1971) on the role of holy men, but Brown has considerably developed his views since then: 'The saint as exemplar', Representations 1 (1983): 1-25; 'The saint as exemplar in late antiquity', in J. S. Hawley (ed.), Saints and Virtues (Berkeley, 1987): 3-14; Authority and the Sacred. Aspects of the Christianisation of the Roman World (Cambridge, 1995): 55-78.
3. Brief accounts of the demise of the curiales in J. H. W. G. Liebeschütz, 'The end of the ancient city', in J. Rich (ed.), The City in Late Antiquity (London, 1992), esp. 25-8, and Averil Cameron, The Mediterranean World in Late Antiquity AD 395-600 (London, 1993), 168-9.
4. P.-L. Gautier, 'Nouvelles inscriptions de Gerasa', Syria 62 (1982), 297-312, esp. 297-307, with the comments of P. R. L. Brown, Power and Persuasion in Late Antiquity. Towards a Christian Empire (Madison, Wisconsin, 1992), 152.
5. On these issues, see P. Garnsey, Social Status and Legal Privilege in the Roman Empire (Oxford, 1970), 147-52, and Rapske, 10-20, 32-4.
6. Cf. Garnsey, 148-9, and R. A. Baumann, 'The suppression of the Bacchanals: five questions', Historia 39 (1990), 334-48.
7. Cf. T. E. Gregory, Vox Populi. Violence and Popular Involvement in the Religious Controversies of the Fifth Century A.D. (Columbus, 1979), and 'Urban violence in late antiquity', in R. T. Marchese (ed.), Aspects of Graeco-Roman Urbanism. Essays on the classical city (Oxford, 1984), 138-61, cf. also Rapske, 41-2. To the literature on crucifixion in n.30, p.98, should be added M. Hengel, Crucifixion (London, 1977).
8. R. Henke, Studien zum Romanushymnus des Prudentius (Frankfurt am Main, 1983); M. Malamud, A Poetics of Transformation: Prudentius and Classical Mythology (Ithaca, 1989); A.-M. Palmer, Prudentius on the Martyrs, Oxford, 1989; J. F. Petruccione, Prudentius' Use of Martyrological Topoi in Peristephanon (unpublished Ph.D., University of Michigan, 1985); M. J. Roberts, Poetry and the Cult of the Martyrs. The Liber Peristephanon of Prudentius (Ann Arbor, 1993), reviewed in this journal by E. Vance and this reviewer (94.4.17 and 94.7.3).
9. Cf. K. M. Hopkins, Conquerors and Slaves (Cambridge, 1978), 120-1, G. E. M. de Ste Croix, The Class Struggle in the Ancient Greek World (London, 1981), 409.
10. No reference is made to Ramsay MacMullen, Corruption and the Decline of Rome, New Haven and London, 1988.
11. On governors in late antiquity, cf. Brown, Power and Persuasion, 22-34.
12. On the importance of such distinctions, cf. the study of D. Schlinkert, Ordo senatorius und nobilitas. Die Konstitution des Senatsadels in der Spätantike, Stuttgart, 1996, 84-116.
13. General: S. N. Eisenstadt and L. Roniger (eds.), Patrons, Clients and Friends. Interpersonal Relations and the Structure of Trust in Society (Cambridge, 1984), E. Gellner and J. Waterbury (eds.), Patrons and Clients in Mediterranean Societies (London, 1977). Particular: B. K. Gold (ed.), Literary and Artistic Patronage in Ancient Rome (Austin, 1982), R. P. Saller, Personal Patronage under the Early Empire (Cambridge, 1982), Wallace-Hadrill, cited above. The case of the workers on the imperial estate in Agabeykoey in Lydia (214-5, cf. 193) is also considered by S. Mitchell, Anatolia. Land, Men, and Gods in Asia Minor. Vol. 1, The Celts and the Impact of Roman Rule (Oxford, 1993), 229-30. M. considers a much wider range of evidence for the negative impact of the military on the civilian population of Anatolia in the third century.
14. Also considered by Rapske, 316-23.
15. Cf. Rapske, 20-5 (evidence for prisons both in Rome and outside), 28-32 (military custody), and 244-76 (prison personnel).
16. Cf. Rapske, 196-225, who considers the issues of overcrowding, darkness, chains, food, clothing, illness and mortality.
17. Note also Prudentius' vivid depiction of prison conditions in the Peristephanon. The term catenae symbolises the whole of the incarceration and sufferings of the Christians, cf. Roberts, Poetry, 46-7, 59-61 and 79-91. On the use of chains, cf. also Rapske, 25-8.
18. Comparative Greek material is to be found in P. Dubois, Torture and Truth (London, 1991). On images of torture in Prudentius, Peristephanon, cf. Roberts, Poetry, 55-68.
19. Cf. W. Klingshirn, 'Charity and power: Caesarius of Arles and the ransoming of captives in sub-Roman Gaul', JRS 75 (1985), 183- 203, and S. Baumgart, Die Bischofsherrschaft im Gallien des 5. Jahrhunderts. Eine Untersuchung zu den Gründen und Anfängen weltlicher Herrschaft der Kirche (Munich, 1995), 166-8.