We can identify two sorts of secret agents within the Roman world, with some specialist groups alongside them. These are the frumentarii, created by Hadrian, and the agentes in rebus created by Diocletian and particularly active after AD357, when their purpose was modified by Constantius II. Among the agentes were an investigative branch, the curiosi.
Rome had no police force and no prisons for most of the classical period. When L. Sergius Catilina was arrested and charged with treason, in 63BC, he was lodged in a private house of a friend, a sort of bail bond. The lawyers for and against him no doubt used whatever means they could to secure information. We have Cicero’s speeches In Catilinam, but of course not the defence side. The whole process was handled by magistrates. Rome had only two cells, used for lodging accused persons who didn’t have the benefit of rich friends.
Since imprisonment was not used as a punishment, only to hold prisoners before trial, all our information relates to holding, not punishment. For the most part, unless an accused person was cleared, probably due to a bribe, the outcome was a fine for minor offences, execution and exile (social death to anyone who lived in Rome where the right contacts made all the difference). The only quasi-imprisonment available was hard labour in the mines, which was a type of death sentence, since few survived such heavy, dangerous and often poisonous work, and such sentences were reserved for slaves and low-status citizens.
As republican law broke down, anyone could kill almost anyone else with impunity. Julius Caesar had forbidden people in Rome to carry weapons (Lex Julia de vi privata). But he was after all the one who had crossed the Rubicon under arms to conquer Rome in the first place. Under the second triumvirate, anybody could denounce anyone and have them put on the proscription list.
The ‘prison’ at Rome was the Tullianum, originally a spring house, later the site of the medieval Mamertine prison and sometimes referred to as the Carcer (hence ‘incarcerate’), which seems to have included an oubliette from which it was hard to escape and inconvenient captives could be quietly starved to death. There were quarries nearby used to pen prisoners and to one side a long staircase, the Gemonian Stairs (Scalae Gemoniae) down which captives were hurled to their deaths as a form of public execution. It is notable that the Tullianum was across the street from the Senate House and at the top was the Tabularium, where accounts were lodged, suggesting that this was a zone of punishment and death for political and financial crimes.
The Carcer or Tullianum at the foot of the Capitoline Hill
Frumentarii were also used by Balbinus and Pupienus to warn cities across the empire not to support Maximinus Thrax (Hist Aug ‘Balbinus’ 10)
Maximinus Thrax, emperor AD235- 238
and Gallienus to pick up rumours against him (‘Claudius’ (Gothicus) 17). These had originated as grain buyers for the army, hence the name. Everyone needs to eat, so what could be less conspicuous than a grain merchant? Since the poor of Rome lived for centuries on grain handouts, the frumentarii were a good way to pick up news and gossip.
The system of spies was clearly not very successful, since in the third century many emperors faced rebellions and conspiracies, with the average reign of an emperor being 2 years and 6 months. It is reasonable to assume that everyone knew who the legion’s frumentarius was and to feed him false information.
Diocletian of course reformed much else, so the invention of a system of agentes in rebus would be a priority for him to retain his throne unchallenged.
In border regions there seems also to have been a system of double agents; these are known from two sources from Britain. The first is a reference in Vindolanda Tablet 162 to a ‘Miles Arcanus’, a secret soldier. As Vindolanda was abandoned before the creation of Hadrian’s Wall, the system of secret soldiers may have been replaced by the frumentarii, or merely supplemented by it. The idea of a secret soldier is, we assume, to remain secret.
During these outstanding events the areani, who had gradually become corrupt, were removed by him [Count Theodosius, father of the later emperor] from their positions. This was an organization founded in early times, of which I have already said something in the history of Constans [Note: that reference did not survive]. It was clearly proved against them that they had been bribed with quantities of plunder, or promises of it, to reveal to the enemy from time to time what was happening on our side. Their official duty was to range backwards and forwards over long distances with information for our generals about disturbances among neighbouring nations (Ammianus Marcellinus Res Gestae 28.3.8)
Since Latin has no definite article and was not punctuated by commas, it might be that the Latin means ‘areani who had gradually become corrupt’. Areani may be a misreading in the surviving MSS for ‘arcani’; the two seem too alike.
In certain places law and order were maintained by urban cohorts (cohortes urbanae), under the control of the urban prefect, the mayor. Only three places in the western empire seem to have received them, and perhaps they were the ones that needed them: Rome itself, with three cohorts, Carthage and Lyon. Since there were no such bodies in Spain or Britain, it is possible that Carthage (the leading city of Roman North Africa) and Lyon (the capital of Gaul) were headquarters for the Praetorian Prefect and that they relied on the security systems handled by the various frumentarii, later agentes in rebus, the curiosi and arcani, which taken together were probably deemed to be sufficient.
We do not know if the three urban cohorts in Rome were organised by area, by function, or both. There would need to be some idea of ‘the patch’ in which cohort operated, as it would need a building; the urban cohorts were considered to be soldiers and would need to store arms and captives, and possibly recovered loot. The urban prefect would allocate tasks at the highest level and make sure each of the three cohorts knew the others were conducting operations which might cross into the others’ patches.
Operationally, the functions would be catching thieves, stopping disorder and attempted coups, as well as dealing with internal matters such as fraud and unauthorised murder. The one thing they didn’t do was patrol the streets, which was the job of the vigiles, the ‘vigilant ones’, who doubled as nightwatchmen and a fire brigade. Presumably paid fire fighters with a financial incentive to be loyal were preferred to amateurs; Trajan forbade Pliny the Younger from allowing the citizens of Nikopolis (in Bithynia-Pontus) from starting a volunteer firemen’s department on the grounds that such things led to uprisings (Letters of the Younger Pliny, Book 10).
At that time (early second century AD) control over such groups was maintained by the provincial governor, but after Septimius Severus or Caracalla stopped governors having operational control over the army, presumably the same applied to all such groups and the vigiles were controlled by the local army commander.
Local jails did exist; St Paul was arrested and locked up in one on Malta (See Acts 27 and 28; also for his shipwrecks 2 Corinthians 25, which also speaks of him being beaten with iron rods). Acts 28.21 has the Roman authorities say to him ‘We neither received letters out of Judæa concerning thee, neither any of the brethren that came shewed or spake any harm of thee’. This suggests that a system existed for people to condemn criminals discreetly, either by direct testimony or by letter.
This has been a brief exposition of the role of secret agents in the Roman world. I would welcome any comments.