When you think about code-breaking, the chances are your thoughts will automatically shift to Enigma, Bletchley Park, the Imitation Game and Alan Turing as perhaps the most prominent example. However, encoding messages was by no means an invention of the modern world.
Some form of cryptography itself seems to date back to at least early second millennium BC Egypt, with hieroglyphics were used to decorate tombs, while other examples appear in Mesopotamia and Greece; however, there are doubts over whether some of these encryptions were really used to prevent others from reading the text or as simple literary puzzles for amusement or even as a way to avoid bad omens (Cohen (1995); Kelly (1998); Lateiner (2010)).
The Polybius Square
More clear cut examples of cryptology had appeared by the second century BC. Polybius wrote on a system to be used in fire-signalling “devised by Cleoxenus and Democleitus and perfected by myself (Polybius)” (Polybius X.45.6), now referred to as the Polybius Square.’ Polybius X.45.7-47.11 describes the square in action. Julius Caesar is recorded using a cipher – now known as the Caesar Cipher – which shifted each letter two or three places further through the alphabet (Suetonius, Divi Julius 56.6). Augustus is also recorded using a similar cipher (Suetonius, Aug. 88), while it has been suggested that Caesar may have used an even more complicated system – “there is even a rather ingeniously written treatise by the grammarian Probus concerning the secret meaning of letters in the composition of Caesar’s epistles” (Aulus Gellius, Attic Nights 17.9.1-5).
However, it was not just such cryptology that the Romans resorted to prevention sensitive or important information falling into the wrong hands. We are fortunate to have recorded one such instance of the lengths a Roman diplomat might go to get a message out to the Roman authorities. The basis of this good fortune is the military career of the historian Ammianus Marcellinus. In serving as a protector of the general Ursicinus during the 350s, Ammianus was in a position to see one such message be found, decoded and then deciphered.
The author of this coded message was a certain Procopius. While less well-known that his sixth century namesake, the historian Procopius of Caesarea, this Procopius is a character well worth reading up about. A native of Cilicia, he attempted to usurp the imperial throne from the eastern emperor Valens in 365/366.
Basing his claim on being a maternal cousin of Julian and the idea that Julian named him his successor, complete with an imperial purple robe, at the outset of the disastrous Persian expedition, Procopius’ usurpation was peculiar mix of farce, organisational skill, drama, loyalty, bribery, betrayal and snatching defeat from the jaws of victory (or vice versa depending on the point of view).
Julian and Valens
His improvised coronation with makeshift imperial regalia of what seems like slippers and a napkin, in the dead of night in Constantinople in the hands of Ammianus reads more like the comedy of Aristophanes or the satire of Petronius than a serious assumption of imperial power (Ammianus XVI.6.16-19).
Constantius II and Shapur II
But before he was usurping the imperial throne, Procopius served under the emperor Constantius II as a tribunus et notarius, a position which in itself demonstrates that Procopius was no dummy. He was also felt worthy of being entrusted with one of the more important duties of the 350s: along with the comes, Lucillianus, Procopius was dispatched to the court of the Sassanid Persian King, Shapur II, to negotiate a peace between the two empires (Ammianus XVII.14.3).
In other times, this would not have been as difficult as it sounds as there were prolonged periods where both Rome and Persia were happy to see their shared border remain quiet. Unfortunately for Procopius, this was not one of those times. Shapur II was on the warpath, determined to see large parts of Armenia and Mesopotamia ceded to him either at the negotiating table or through force.
Procopius and Lucillianus had their work cut out for them. Indeed, it could be argued that Constantius II did not intend for them to succeed in obtaining any sort of treaty, merely using the pretext of the embassy to slow and/or discover the Persian king’s plan for the coming campaign season. Ammianus goes as far as to say that Shapur, “armed with the help of the savage tribes which he had subdued, and burning with superhuman desire of extending his domain, was preparing arms, forces, and supplies, embroiling his plans with infernal powers and consulting all superstitions about the future.” (Ammianus XVIII.4.1)
While the exact nature or access of the embassy to the Persian court is not recorded, Procopius and Lucillianus were able to hear or see enough to recognise that even with these ‘negotiations’ still on-going, the Persian army was on the move. And the comes and notarius needed to get the word back to Roman authorities. They achieved this by getting a concealed note to a group of Roman scouts in a scabbard, who then succeeded in delivering this communiqué to Ursicinus at Amida (Frontinus, Strat. III.13.5 advised similar use of a scabbard to conceal secret messages – “some have written on the linings of scabbards”).
The message was not only hidden in a secret place, once it was removed from the scabbard, it was found to be written in code, and even when the cipher was applied, the decoded message appeared nonsensical (Ammianus XVIII.6.17; Blockley (1986) on decoding the letter).
“Now that the envoys of the Greeks have been sent far away and perhaps are to be killed, that aged king, not content with Hellespontus, will bridge the Granicus and the Rhyndacus and come to invade Asia with many nations. He is naturally passionate and very cruel, and he has as an instigator and abetter the successor of the former Roman emperor Hadrian; unless Greece takes heed, it is all over with her and her dirge chanted” (Ammianus XVIII.6.18).
The coded nature of the letter, in terms of secreting, symbols and allusion, along with the inference of the first line suggest “that the envoys of the Greeks” – Procopius and Lucillianus – had been imprisoned by Shapur or were at least under surveillance.
Mithridates VI of Pontus
The mention of an “aged king” bridging the Granicus and Rhyndacus rivers, both in Asia Minor, was considered an inference to the outbreak of the Third Mithridatic War in 74/73BCE, with the “aged king” being Mithridates VI (Appian, Mithr. 69-71; Matthews (1989), 42-43; Mayor (2009) on Mithridates VI). Respectively, the Granicus and the Rhyndacus are the modern Biga and Mustafakemalpasha rivers in north-western Turkey. The former is most famous as the site of Alexander the Great’s first victory over the Achaemenid Persian Empire in 334BCE, while the latter was the site of two Roman victories over the forces of Mithridates VI, first by Fimbria in 85BCE and then by Lucullus in 73/72BCE (Frontinus, Strat. III.17.5; Plutarch, Lucullus XI.2-3).
The Granicus/Biga and the Rhyndacus/Mustafakemalpasha Rivers
This allusion to Mithridates and his crossing of rivers was considered to be a reference to Shapur’s planned crossing of the Greater Zab and Tigris rivers for an invasion of Roman territory, with the mention of his cruelty and passion likely highlighting that Shapur’s invasion was not some run-of-the-mill raid but a full-scale invasion intent on conquest. A similar sense of the immediacy and size of the threat posed by the Shapur’s latest invasion is also conveyed by the final line of the message.
The Eastern Theatre in the 350/60s
Perhaps the most straightforward piece of information is the allusion to Hadrian’s successor, the emperor Antoninus Pius, as an “instigator and abetter.” This was revealing the presence and identity of a Roman defector, Antoninus, at the Persian court, and his role as an adviser to the Persian king (Ammianus XVIII.5).
Hadrian and Antoninus Pius
It might seem then that that this letter gave Ursicinus valuable insight into the planned movements of Shapur but it in actual fact, the coded message of Procopius is rather short on actionable intelligence or new information. By the time they received Procopius’ message at Amida, Ursicinus and Ammianus had already been confronted with evidence that the Persians were in Roman territory, not only finding a Persian spy at Meiacarire but also being confronted with the initial stages of Nisibis being put under Persian blockade.
While it does hint that Shapur intended to invade en masse, it does not give any real notion of the size of his army or where exactly he intended to cross into Roman territory. Was he going to cross the Greater Zab and the Tigris close together, perhaps at their confluence and then drive at Singara? Or was he going to cross the Zab and then follow the Tigris upriver to the northwest before crossing? And even then where would he cross? Near Nineveh? Bezabde? Amida? Or somewhere in between or beyond?
So Procopius’ message was telling the Roman high command something they already knew – the Persians were invading – but then failed to tell them something of exact strategic usefulness that they did not already know – where was the main Persian army going to cross into Mesopotamia? This cannot be held against Procopius and Lucillianus. The close scrutiny of the Persian court likely limited the intelligence they would get their hands on and the rapidity with which they could get that limited intelligence back to Roman territory.
Desperate to get more firsthand knowledge, Ursicinus sent Ammianus on a mission to contact Jovinianus, the Persian satrap of Corduene, who sent Ammianus on with a guide to a cliff that overlooking the route of march of Shapur’s army. Again harking back to ancient history, Ammianus describes the size of the Persian force with suitably dramatic flair, recalling the great invasion force of the Achaemenid Persian king Xerxes in 480BCE and suggesting that it would take three full days for the entire Persian army to cross the Tigris (Ammianus XVIII.7.1; XIX.6.11 later declared Shapur’s force to be 100,000 strong; Herodotus VII.59-60 on Xerxes’ force).
So the coded message of Procopius might not have wielded any overtly useful intelligence for Ursicinus at the time; indeed, the subsequent Persian campaign saw the epic siege, capture and destruction of Amida. However, this episode does give us some insight into Roman coding techniques of concealment, encoding and allusion in action.
Blockley, R.C. ‘The coded message in Ammianus Marcellinus 18.6.17-19’, Echos du Monde Classique 30 n.s. 5 (1986) 63-65
Cohen, F. ‘A Short History of Cryptography,’ http://all.net/edu/curr/ip/Chap2-1.html (1995)
Crawford, P. Constantius II: Usurpers, Eunuchs and the Antichrist. Barnsley (2016)
Lateiner, D. ‘Signifying Names and Other Ominous Accidental Utterances in Classical Historiography,’ GRBS 45.1 (2010), 35-57
Matthews, J.F. The Roman Empire of Ammianus. London (1989)
Mayor, A. The Poison King: The Life and Legend of Mithridates, Rome’s Deadliest Enemy. Woodstack (2009)
Kelly, T. ‘The Myth of the Skytale,’ Cryptologia 22.3 (1998) 244-260