Part I

The Republic


Intelligence ancient and modern

The Romans built an empire that lasted a thousand years; this book is an attempt to trace the diffuse intelligence-collecting methods used by the Romans to make that empire possible. The Romans offer us a particularly interesting case study, because although they did not establish an institutionalized, unitary, or centralized intelligence organization, it is perfectly clear that they could not have functioned without intelligence, and perforce, an apparatus to collect it. No nation could have assembled the most formidable military machine in antiquity, expanded its dominion over vast tracts of foreign territory, and built one of the largest empires ever known to mankind without possessing the means to obtain and disseminate intelligence.1

The modern concept of intelligence and its appreciation would be as much an anachronism in the ancient world as the modern technology we use to collect it. Intelligence problems in all premodern societies, indeed right up until the nineteenth century, were much simpler than their twentieth-century counterparts. By modern standards, all ancient intelligence organizations seem primitive. Speed of transmission, massive bureaucracies, and technological means of collection were unknown. But despite these limitations the ancient world provides much evidence for intelligence activity. We must also not assume that intelligence activities in all ancient societies were merely replications of one another. There were cultural differences among them that made their approaches to intelligence very dissimilar. Monarchies do not operate like republics, nor do autocracies act like democracies. Also, societies change their intelligence practices as their needs change, from times of war to times of peace. Yet, intelligence activities never entirely disappear. They are as much a part of statecraft as diplomacy, secrecy, and propaganda.

Near Eastern monarchies set up intelligence services early, and each built upon its predecessors. The Babylonians and Assyrians already had military roads, a postal system, a secret service, and elaborate fire signaling operations by the second millennium bc.2 The Persians further developed and refined this system, which Herodotus and Xenophon described with unconcealed admiration.3 But those regimes were autocratic, centralized, and could set up systems of foreign intelligence gathering and internal security without giving a second thought to the rights or sensibilities of their subjects. Greek and Roman governments, on the other hand, had to function within their particular political environments of democracy and republic, which inevitably made their task a more difficult one. Here we must not assume that counterintelligence, in the service of internal state security, is the hallmark of dictatorial regimes alone, while democracies concentrate on foreign intelligence. Democratic governments in reality can, and do, engage in as much truly vital counterintelligence efforts as any autocratic regime, but they stop well short of abusing the domestic security apparatus for narrow party-political purposes. Because of their unique constitutional development, the Romans provide examples of intelligence gathering in both a republican and an imperial setting. In Rome, the twin needs of all states were continually present: how to obtain information about the enemy abroad; and how to keep the government safe at home. In other words, they all needed to collect external intelligence and provide internal security.

Compared to other ancient societies, the Romans did not place a very heavy emphasis on intelligence, because their strategy did not rely on surprise and speed, but rather on strength and consistency. As the strongest military power in the Mediterranean, they were able to rule their empire first by virtue of their armies of occupation, and second by psychological means of control, not the least of which was imperial propaganda in all its manifestations, skillfully designed to keep subjugated peoples in line.4 Many times their conquered enemies were the ones who initiated intelligence operations, because guerrilla warfare was the only practical means of rebelling against Rome’s all-pervasive power.

We should not be surprised that the Roman Republic did not develop an intelligence service. The Roman magistrates who overthrew the Etruscan monarchy in 509 bc were loath to introduce any institution that smacked of monarchical practice. And it is characteristic that, in such a political climate, they would play down the importance of intelligence operations. Once the Empire was established, however, and the Roman government became more autocratic, their preoccupation with internal security increased. They adapted their institutions to the requirements of their new status as world power. Only their deep-seated conservatism and desire to pay lip service to republican tradition kept them from quickly adopting institutions that were either novel or centralized. Aristocratic privilege and lack of bureaucratic specialization precluded a more rapid and deliberate development of an intelligence apparatus in Rome. Generally speaking, the Romans stuck with systems they knew worked, and for the most part they were successful.

The study of military history has witnessed a major revival in recent times,5 and Roman intelligence history touches upon several broader topics of current interest to military historians and to students of international affairs. Strategy, deception,6 surprise operations,7 covert action,8 subversion and resistance, insurgency warfare, and terrorism have all been widely discussed in recent literature. These techniques can all be seen at work in antiquity. For classical historians, intelligence affects frequently debated issues, such as Roman decision-making,9 imperialism,10 the provincial resistance to Romanization,11 Roman grand strategy,12 and the nature of Roman frontiers.13 The study of intelligence is the study not just of espionage, but equally of the entire intricate network of communications, military operations, and frontier policy. For this reason, the larger questions raised in this book have an immediate relevance. Although intelligence methods have changed drastically with the advent of technology, the principles remain surprisingly similar. Political questions, such as “what is the place of intelligence services in a democracy?”, have their roots in the Greco-Roman world.14 There are lessons to be learned about the high costs of intelligence failures, even when they are two thousand years old.

The Romans handled intelligence broadly in accordance with the four basic steps of the modern “intelligence cycle” as we define it today: direction, collection, processing/analysis, and dissemination. First, they needed to go out and collect the appropriate types of information. Not all information produces intelligence. Intelligence may only be derived from crucial information about the enemy – its strength, location, intentions, and capabilities. Also present is an important time factor: intelligence must be collected, processed, and delivered while it is still valid and useful. As most scholars and analysts know, having too much information can sometimes be almost as bad as having none at all. An information glut makes it difficult to sift through all the available material and pick out that which is most important. Then the information has to be analyzed. The intelligence analyst, unlike a scholar, must produce a conclusion within a certain deadline. When a war breaks out, the commander does not want to hear “We’re still sifting through the data.” The final step in this process is dissemination. Even if intelligence is collected and analyzed correctly it would, in many instances, be of little or no value if the product is not conveyed to the consumer who needs it, and with sufficient time for him to act upon it. As an example, a list of conspirators was thrust into Caesar’s hand shortly before his assassination. His intelligence network had done its job. Had he read the message and acted on it appropriately, he might have survived. Taking advantage of the intelligence product – the decision to act – is not a function of the intelligence apparatus; the two processes are mutually exclusive. Ultimately, it would be up to the commander or statesman to take proper action. If he has all the information yet makes a bad decision, it clearly is not an intelligence failure, but incompetence on the part of the consumer.15 We shall see this particular scenario played out in Chapter 5, with Crassus in Parthia.

Since the Romans did not draw distinctions between political and military intelligence, we must examine all aspects of Roman governmental action and, for our own convenience, sometimes resort to artificial distinctions. Political intelligence, useful to making decisions about foreign policy in which war was often an instrument, became military intelligence on the battlefield once the war had begun. The mechanisms and techniques by which the Roman government accomplished its intelligence tasks will seem not totally unfamiliar to the modern reader, even though, in many cases, there are not analogous modern institutions. Expecting the Romans to organize a centralized intelligence service may be asking too much, but to deny that they engaged in intelligence activities would be naive. Governments and armies do not operate in an intelligence vacuum. They cannot afford to ignore the importance of an intelligence apparatus in providing for national security, nor can they ignore the dangers inherent in its abuse. If Le Carré’s dictum, that each nation’s intelligence service is somehow a mirror of its national soul, is not entirely correct, we can at least say that the service is a true reflection of the regime controlling the government. The factors that affect the form of an intelligence service are more complex. They include the nature of the governmental system, the circumstances in which that government finds itself, the views of factions within the government, and whatever relationships it has with foreign governments. All these interactive variables were as much at play in Rome as anywhere else.

The chapter topics in this book highlight progressive stages in the evolution of Roman intelligence history, from the early Republic through the third century to the reign of Diocletian (ad 284–305). The first two chapters discuss the early evidence of intelligence activities in the Republic (509–264 bc). Part of the problem with telling the story of the early Republic lies in distinguishing between attitude and action: what Romans said and what they did were two different matters. There is evidence that the Romans were mounting intelligence operations quite early, notwithstanding their characterization by Livy as people who would do nothing that was not “above board.” Livy’s idealized picture of the early Romans is more than a little suspect. If his stories were accurate accounts of the way the Romans gathered intelligence to keep themselves safe, then Rome never would have advanced beyond the seven hills. Luckily for the future of Rome, the simple Latin peasants, whom Livy portrays as having utter disdain for anything disingenuous, soon learned that the information they required to survive in a hostile international environment would not be dropped in their laps; they would have to take the initiative and aggressively collect it. The evidence suggests they did master the art of finding out what they needed to know, even if that entailed what Livy might brand “un-Roman” behavior.

Chapter 3 details what happened when Rome, with its primitive intelligence network, found itself at war against a foreign power that was well ahead on the intelligence front. Hannibal’s sophisticated use of intelligence operations proved almost fatal to Rome. The title of Chapter 4, “Diplomat, Trader, Messenger, Spy”, describes the mechanisms by which the Romans collected intelligence in the East during their wars with the Hellenistic kingdoms (200–133 bc). Intelligence disasters in the late Republic are discussed in Chapters 5 and 6. In Crassus’s failure in Parthia (53 bc) and Caesar’s campaigns in Britain (55 and 54 bc), we see how a military commander’s failure to use all his intelligence assets properly can cost the state dearly in money and manpower. Chapter 7 shows how the Roman system had developed by the end of the Republic, and how the crisis which brought the Republic down affected the development of its intelligence apparatus.

With the coming of Augustus also came extensive internal administrative, military, financial, and social reforms in Rome. Its intelligence requirements were now different, and the modus operandi changed accordingly. Chapter 8 discusses this Augustan revolution (27 bc to ad 14), which created the physical network that became the infrastructure of Rome’s secret service and the political climate that nurtured it. Chapter 9 focuses on the institutions that collectively acted as Rome’s military intelligence arm. With all the improvements in intelligence brought by the empire, Rome still suffered from large-scale intelligence failures. Chapter 10 illustrates the most infamous of these failures – the destruction of the Varian legions in the Teutoburg Forest in AD 9. Chapter 12 discusses the frumentarii, a service that has been dubbed “the Roman secret service.” The use of its members for espionage and political assassination made them both loathed and feared by provincials, as well as by members of the court. Chapter 13 discusses the civilian counterpart of that secret service, the agentes in rebus and the notarii. Lastly, Chapter 14 explains the imperial system and how it worked, placing Rome’s intelligence efforts in the context of its frontier policy.

The Romans discovered, as all civilizations do, that the methods a nation must use in order to gain power, and then to maintain it, are not always pleasant ones. And so, although they themselves often denied it, and scholars have sometimes turned a blind eye toward it, the evidence is there that the Romans had no qualms about employing clandestine means to steal a march on their rivals and preserve their own supremacy, first on the Italian peninsula and later in the Mediterranean. The intelligence historian, on the other hand, would conclude that their behavior was normal, and at times even somewhat naive. Later on, however, their expert use of covert action demonstrates a sophisticated understanding of how governments can be subtly manipulated when military force is not an option. This is abundantly clear in the case of Parthia, where even a deliberate show of massive Roman power proved ineffective.

This book will often profile the Romans as oppressive, brutal, and sometimes just plain incompetent. Intelligence history, dealing as it does with military disasters and some of the less palatable aspects of international politics, might appear to be highlighting the negative. Many readers react against this profile and believe it robs the Romans of their due praise. Even those scholars who continue to extol the virtues of Roman imperialism and the invincibility of the Roman legions should consider that the Roman system was not as well adapted to empire-building and conquest as we are often led to believe, nor were their intentions or methods kind. Roman rule could be arbitrary and repressive to its provincials, and internal security measures sometimes excessive. But this was not a natural byproduct of possessing enforcement mechanisms, but rather a consequence of their abuse by corrupt and ambitious officials. Citizens were harassed, intensely scrutinized, and executed in the name of state security; to make matters worse, what constituted state security changed with the mental disposition of each emperor. On the military front, we see how the poor quality of communications and military intelligence in general plagued the Roman legions. As successful as the legions usually were, they also experienced massive intelligence failures, which resulted in needless casualties that numbered in the tens of thousands.

One final caveat: I am not suggesting that intelligence activities were responsible for all of Rome’s successes, nor do they adequately explain all of Rome’s failures. I am focusing on them merely to highlight what has been heretofore an ignored aspect of Roman statecraft. By ancient standards the Romans were neither the best nor the worst at intelligence operations. In the thousand years of their history, we would expect to find some intelligence failures, no matter how efficient their system may have been. Can we cite the longevity of the empire as evidence that Rome’s intelligence services were doing their job? To some extent the answer is yes. Intelligence did play a role in the success and failure of Roman strategy. By highlighting intelligence activities, however, we by no means imply that intelligence is the sole determining factor in military or political dealings. Even timely intelligence carries no guarantees: one may lose even with it, or less likely, one may even win without it.

Admittedly, the study of intelligence does present a special difficulty for historians of all periods. Much of the evidence is unreliable. As intelligence activities are meant to be clandestine, they are not routinely recorded. For this reason, the intelligence component has become, in the words of one writer, “the missing dimension” of most political and diplomatic history.16 In addition, the spies of antiquity, unlike their modern counterparts, did not retire and write memoirs. Classical authors pepper their works with vague phrases such as “he received intelligence that…” or “news arrived of…” but they rarely report who originated or transmitted this information, and how. Indeed, the authors may not have known about sources and methods. The ancient intelligence operative, if unsuccessful in his enterprise, might draw the historian’s attention in a roundabout manner, as a major failure could bring about his execution, or, far worse, a military disaster. If, on the other hand, a man accomplished his assigned intelligence missions successfully, he would remain unheralded and fade into obscurity, unnamed and unrewarded, at least publicly.

There can be no disputing the imperative that all nations need to collect and analyze intelligence in order to make well-informed decisions. Although methods of operation may change from country to country, and from civilization to civilization, the need for the information and the presence of intelligence operations is a constant. The problem is only to find out what each society did and how it was done. Antiquity provides a logical beginning, and not just for antiquarian interest. The question has recently been posed by Neustadt and May: “What could the story of those ancients, armed with spears, propelled by oars, maintained by slaves, deprived of electronics, knowing nothing of air power, convey to men managing a modern war?” The answer is that it teaches us, as Neustadt and May’s title aptly suggests, “to think in time,” namely, to make decisions based on evidence that is both cross-cultural and chronologically diverse.17 Intelligence history of the ancient world shows how, even in more primitive ages, intelligence systems were instrumental in shaping the defense and foreign policy of nations in both war and peacetime. We see how decisions were made, how disasters occurred, and what role covert action played in the foreign policy process. We observe not only a wide range of intelligence activities, but also the weaknesses and difficulties that frequently beset intelligence operations: insufficient resources, errors of judgment, misunderstandings, and inadequate control. As prior to the twentieth-century intelligence was primarily HUMINT, or human-sourced intelligence, and human nature being what it is, certain phenomena associated with the practice of intelligence are ever-present in history. The motives of traitors, for example, have not changed much through the ages.

Historical intelligence provides a meeting ground for academics and intelligence professionals – two groups that have not always been on the best terms since the Second World War. Gone are the days when scholars such as Frank Brown and Bernard Knox had wartime intelligence duties.18 This book is offered to both communities in the hope of overcoming the attitude of scholars who see espionage as an amusing sideline or an “anti-progressive” subject best ignored, or intelligence practitioners who think antiquity is too far removed to matter anymore. In our post-September 11 world, we have witnessed how easily an intelligence failure, such as the one that allowed the World Trade Center bombing to occur, can affect governments and their institutions. The issues of how governments use and control information, and how intelligence services work in a democracy, have never been more timely or more relevant.


1. See the comments of A. Ferrill, Roman Imperial Grand Strategy (Lanham, MD: University Press of America, 1991), pp. 1, 17–18.

2. F. Dvornik, The Origins of Intelligence Services (New Brunswick, NJ: Rutgers University Press, 1974), Ch. 1.

3. Herodotus 5.52; Xenophon, Cyropaedia 8.6.16–18.

4. See K. Hopkins, Conquerors and Slaves (Cambridge: Cambridge University Press, 1978), pp. 197–242, on the imperial cult.

5. See the remarks in Ferrill, Roman Imperial Grand Strategy, p. ix.

6. The literature on deception is immense. To name just a few titles: Abraham Ben-Zvi, “Dynamics of Surprise: The Defender’s Perspective,” Intelligence and National Security 12,4 (1997), pp. 113–44; John Gooch and Amons Perelmutter, Military Deception and Strategic Surprise (London: Frank Cass, 1982); D. Daniel and K. Herbig, Strategic Military Deception (New York: Pergamon Press, 1992); E. Epstein, “Deception and Estimates,” in Intelligence Requirements for the 1980: Analysis and Estimates, ed. R. Godson (New Brunswick, NJ: Transaction Books, 1980); F. Feer, “Coping with Deception,” in Intelligence Requirements for the 1980s: Analysis and Estimates, ed. R. Godson (New Brunswick, NJ: Transaction Books, 1980); M. Handel, “Intelligence and Deception,” Journal of Strategic Studies 5 (1982), pp. 122–54; R. Heuer, “Strategic Deception and Counterdeception: A Cognitive Process Approach,” International Studies Quarterly 25, 2 (June, 1981) pp. 294–327; D. Charters and M. Tugwell (eds), Deception Operations: Studies in East–West Context (London: Brassey’s, 1990); M. Handel, Strategic and Operational Deception in the Second World War (London: Frank Cass, 1987).

7. For a discussion of the literature and definitions of strategic surprise, see: A. Levite, Intelligence and Strategic Surprise (New York: Columbia University Press, 1987); E. Kam, Surprise Attack (Cambridge, MA: Harvard University Press, 1988); R. Betts, Surprise Attack: Lessons for Defense Planning (Washington, DC: Brookings Institution, 1982); M. Handel, “Intelligence and the Problems of Strategic Surprise,” The Journal of Strategic Studies 7 (1988), pp. 229–81; Daniel and Herbig, Strategic Military Deception; A. Ben-Zvi, “Hindsight and Foresight: A Conceptual Framework for the Analysis of Surprise Attacks,” World Politics 28 (1976), pp. 381–95; K. Knorr and P. Morgan (eds), Military Strategic Surprise (New Brunswick, NJ: Transaction Books, 1983).

8. A. Bozeman, “Covert Action and Foreign Policy,” in R. Godson (ed.), Intelligence Requirements for the 1980s: Covert Action (New Brunswick, NJ: Transaction Books, 1979).

9. A.M. Eckstein, Senate and General (Berkeley, CA: University of California Press, 1987); F. Millar, The Emperor in the Roman World (Ithaca, NY: Cornell University Press, 1992).

10. The subject of Roman imperialism is particularly controversial. The field remains divided between those who see the Romans as overt imperialists, as does W.V. Harris, War and Imperialism in Republican Rome (Oxford: Oxford University Press, 1979), and those who believe Rome stumbled into empire, as do E.S. Gruen, The Hellenistic World and the Coming of Rome (Berkeley, CA: University of California Press, 1984) and Eckstein, Senate and General; see also J.A. North, “The Development of Roman Imperialism,” Journal of Roman Studies 71 (1981), pp. 1–9, and A.N. Sherwin-White, “Rome the Aggressor?,” Journal of Roman Studies 70 (1980), pp. 177–81.

11. S. Dyson, The Creation of the Roman Frontier (Princeton, NJ: Princeton University Press, 1985); S. Dyron, “Native Revolt Patterns in the Roman Empire,” ANRW 2,3, pp. 138–75, and “Native Revolt Patterns in the Roman Empire,” Historia (1971), pp. 239–74.

12. On the controversy over whether a Roman grand strategy existed, see Ferrill, Roman Imperial Grand Strategy and E.N. Luttwak, The Grand Strategy of the Roman Empire (Baltimore, MD: Johns Hopkins University Press, 1976); B. Isaac, The Limits of Empire (Oxford: Clarendon Press, 1990); C.R. Whittaker, Frontiers of the Roman Empire: A Social and Economic Study (Baltimore, MD: Johns Hopkins University Press, 1994).

13. Readers are directed to the annual proceedings on the Congress of Roman Frontier Studies. Limes studies have benefited tremendously, for example, from the recent work of David Woolliscroft on signaling. See especially D.J. Woolliscroft, Roman Military Signalling (Stroud: Tempus, 2001), and D.J. Woolliscroft, with contributions by A.T. Croom et al., The Roman Frontier on the Gask Ridge. Perth and Kinross. An Interim Report on the Roman Gask Project, 1995 (Oxford: Archeopress, 2002).

14. See I. Galnoor, Government Secrecy in Democracies (New York: New York University Press, 1977).

15. On the subject of bad decision-making, see B.S. Strauss and J. Ober, The Anatomy of Error: Ancient Military Disasters and their Lessons for Modern Strategists (New York: St. Martin’s Press, 1990). The authors show (pp. 1–15) how competent commanders fail because they choose the wrong strategy: they opt to fight the wrong enemy, at the wrong time, or are unsure how their victory should be defined.

16. See D.O. Dilks, The Missing Dimension (Urbana, IL: University of Illinois Press, 1984), p. 1.

17. R.E. Neustadt and E. May, Thinking in Time (New York: Free Press, 1986), p. 233.

18. For a survey of scholars who served in intelligence in the Second World War, see Robin W. Winks, Cloak and Gown (New York: Morrow, 1987). For the number of scholars offering courses at American universities, see “Syllabi of Intelligence Related Courses,” Symposium on Teaching Intelligence, 1–2 October 1993 (Washington, DC: Center for the Study of Intelligence, 1994).

If you find an error or have any questions, please email us at Thank you!