[In the following excerpt, Thompson examines ways in which the fragmentary state of the remains of Polybiu's Histories may have affected historians' views of his work.]

There is a romantic allure to collecting and interpreting fragments of lost works. What could be more pleasurable than reconstructing the plot of Prometheus Unbound? And Toni Raubitschek certainly enjoyed himself studying what Theopompus and Theophrastus said about ostracism. But the same interest does not extend to quotations from works which are preserved in their entirety. We know, for instance, who cites Androtion and what sort of information they gain from him, whether it be facts from Athenian history, details of cult, or merely unusual words. But we do not know who quotes Xenophon, nor whether the Hellenica, the Anabasis, or the Memorabilia was the most influential work, nor whether he was cited as the final authority or merely quoted in comparison with Diodorus for fourth century history, Plato for the life of Socrates, and Herodotus for Persian ethnography. Likewise, you can find the fragments of Isaeus without much trouble, but it would be very difficult to find out who saw fit to quote the speeches which have survived and whether they were used as historical sources or literary models. Occasionally it is possible to make some progress along these lines. Hude's editio maior contains a full collection of citations from Thucydides, but very little has been done with it.

In the call to the meeting we are asked "what direction research and explanation should take in the coming decades." It seems to me that one useful project for the future would be to collect the "fragments" of Herodotus and Xenophon—and, for that matter, those of the extant orations—and then to exploit this material and the evidence gathered by Hude for Thucydides.

Even now, without the collection process, it is still possible to learn a great deal about one of the major historians—Polybius—by studying the fragments of those books of his which have come down to us complete. Our text of the remaining, fragmentary books is derived almost entirely from two sources, the so-called Excerpta Antiqua from the first eighteen books, and the collection of Excerpts taken from the whole work, compiled at the command of the Emperor Constantine Porphyrogenitus. The latter is arranged according to subject matter, such as material to illustrate diplomatic embassies, military stratagems, and plots against kings, subjects which no doubt interested the Emperor Constantine greatly. The rationale for the choice of the Excerpta Antiqua has not been adequately determined.

The last half of Polybius' history, as we have it, consists largely of diplomatic history and the internal politics of the Greek cities. One would hardly guess that Polybius was also, perhaps even primarily, a military historian. The reason for this misleading impression is obvious. Most of the text is derived from the Constantinian collection devoted to embassies, and Polybius regularly explains the factional disputes that lead to, or result from, diplomatic missions. Even when there is a war on, our texts of Polybius tell us more about the diplomatic maneuverings than about actual fighting. It's something like preserving Book 5 of Thucydides while throwing away Books 6 and 7. The Third Macedonian War, for example, occupies some fifty-five pages in the Teubner edition. Of these, thirty-three pages are derived from the collection of excerpts De Legationibus.

For the same reason we learn very little about res Asiae from Polybius. Since the Seleucids were further removed geographically and less dependent politically than mainland Greece, there was much less diplomatic intercourse between them and the Romans, and thus much less material to interest the compiler of the De Legationibus. And, of course, he had no concern at all for the fighting which the Seleucids waged. Let us take as our test case this time the Sixth Syrian War, where all but three excerpts are from the De Legationibus. Thus we learn how Antiochus Epiphanes justified his declaration of war (28.20), how the Achaens rejected the call of Lycortas' faction to send troops to the Ptolemies and chose instead to honor the Roman request to mediate a settlement to the conflict (29.23-25), and finally how the Romans ended the war through aggressive diplomacy and their threat to intervene on the side of Egypt (29.27). But we learn nothing at all about how Antiochus came within a whisker of capturing all of Egypt. The closest we come to putting on armor is a fragment preserved by Athenaeus in which Polybius describes Antiochus' victory parade (30.25).

For the purposes of history the selection process of the Constantinian Excerpts is not such a serious matter: we can, after all, obtain a lot of information from Livy and Diodorus about the events which the excerptors omitted. But for the purposes of historiography this method of selection has led to a great distortion, even deformation, of the scope and nature of Polybius' work. It helps to create the impression of a monomaniac, hammering away constantly at the same topic, the Roman manipulation of Greek internal discord, with only an occasional so-called "methodological" digression, in which the author complains mercilessly about the shortcomings of other historians and preaches the need for learning from his own brand of writing history.

In his recent essay, "On Historical Fragments and Epitomes" [in CQ 74 (1980)], Brunt says that "even excerpts [as opposed to mere fragments], unless they are very numerous, substantial, and representative, cannot reveal the scope of an author's work. For instance the Constantinian excerpts from Arrian's Anabasis come chiefly from the volumes on Virtue and Vice and on Gnomai and overweight its moralizing and sententious elements." As I have indicated, the main contributor to our version of the second half of Polybius is the De Legationibus. The next two major sources are these same collections on Virtue and Vice, and on Gnomai. The effect of their contribution to the text of Polybius is the same as Brunt describes in the case of Arrian. The difference is that nobody has to depend on these compilations of extracts for his text of Arrian: he can read it straight. But we can have only the kind of Polybius that the Byzantine anthologists chose to preserve.

Brunt adds, "Epitomators in general seem to have aimed not at producing faithful resumes but at recording, sometimes at length, what they thought of most interest, and their principles of selection are at times impenetrable. They do not necessarily offer a faithful miniature of the original as a whole." He cites the summary of Arrain's Anabasis made by Bishop Photius. "It occupies," he says, "about a hundred lines of the Teubner text, of which nineteen suffice for the first three books; there is not even a mention of the siege of Tyre, By contrast Photius can list almost all the brides and bridegrooms at the Susa weddings. He distorts the whole economy of the work." Brunt concludes that "F̀ragments' and even epitomes reflect the interests of the authors who cite or summarize lost works as much as or more than the characteristics of the works concerned."

Once this is said, it all seems obvious enough, and surely Polybius scholars must be aware of the problem and react accordingly. But I would agree with Brunt that "scholars have often been too precipitate in characterizing and evaluating lost histories on the basis of evidence that is irremediably insufficient, and that in particular too little account is commonly taken of the relevant characteristics of the authors who preserve the 'reliquiae', their reliability in quoting or summarizing, and their own interests and purposes." Even our greatest authority on Polybius, a man who is not only learned but also very perceptive, has not been able to avoid this mistake. At a conference like this one Professor Walbank was asked "whether we have a clear notion of the criteria according to which the excerpts were made: we cannot assume a priori that the tenth century excerptors were interested in the problem of Roman imperialism." Walbank simply replied, "These are important points and the answers are not easy. The interests of the excerptors can, I suppose, be deduced from the actual content of the surviving excerpts." In response, Momigliano suggested a much better way to proceed: to make "a comparison between the excerpts from books I-V and the full text. This, as far as I know, has never been properly done."

This is an excellent idea, to which I return shortly, but first I want to argue that Momigliano has fallen into the same trap which he himself pointed out. In a piece written for The New York Review of Books [reprinted in his Essays in Ancient and Modern Historiography, 1977] he says that "there are at least two basic facts which [Polybius] underrates. One is the Roman conquest of Spain and the other is the Roman organization of Italy. In either case we may suspect that he was misled by his Greek preoccupations and prejudices." In discussing his first contention he observes, "The decision to remain in Spain and to control it was a compound of economic considerations (mines to explore and lands to colonize) and of instinctive pleasure in power … But if there was inducement to plunder and massacre at pleasure, there was also a danger of demoralization of which the Romans themselves soon became aware. It was to fight corruption in dealing with the Spanish provinces that they first instituted special tribunals for malversation in 149 B.C., and there are other signs that they became uneasy about the behaviour of their generals in the Peninsula. Unless we are misled by the lacunas in Polybius' text, he appears to be insensitive to the problems presented to Rome by its conquest of Spain. There is no sign that he realized that the destruction of Numantia by his friend Scipio raised the same moral problems as the destruction of Corinth and Carthage."

Even if the lay reader actually notices the proviso, "Unless we are misled by the lacunas in Polybius' text … ," he will surely ignore it. Clearly Momigliano himself does not believe that in the missing portions of his history Polybius actually did comment on the importance of Spain. He ascribes "Greek preoccupations and prejudices" to Polybius without allowing for the possibility that these are the faults of the epitomators. The military historian M.J.V. Bell [in CQ 66 (1972)] also assures the reader that Polybius "was a Greek, not particularly interested in Spain as the balance of his work shows." But is this really the case? In Book 3 Polybius comments on the significance of Spain to the Carthaginians (10.5-6; cf. 35.5-6) and describes some of Hannibal's fighting there (13.5-14.10 and 17). And in his second preface he announces that, following his exposition of the Roman constitution, he will show how it made possible the reconquest of Italy and Sicily and the subjugation of the Spaniards and Celts (3.2.6). He also promises "an account of the subsequent policy of the conquerors and their method of universal rule, as well as of the various opinions and appreciations of their rulers entertained by the subjects, and finally I must describe what were the prevailing and dominant tendencies and ambitions of the various peoples in their private and public life." Surely this proves that Polybius had an interest in Spain which the excerptors did not share. He goes on to say (3.5.1) that he will describe the Roman war against the Celtiberians and Vaccaei, but once again the compiler of the De Legationibus has omitted this conflict since his task was to record negotiating, not fighting.

To summarize, then, when Oswyn Murray says, [in CQ 66 (1972)], "Certainly Polybius is the greatest surviving political historian of the period, and our accounts [of Hellenistic historiography] will always to some extent be biased towards his preoccupations," I would quarrel with the word "his." Certainly you cannot excerpt what the author himself did not include, and surely Greece was more important than Spain to Polybius, but the almost total purgation of Spain from the text of Polybius shows that our accounts of Hellenistic history writing are biased toward the prejudices and preoccupations of the men who chopped and whittled Polybius down to manageable size.

When we ask "what direction research and explanation should take in the coming decades," I would answer that we should put first things first. Before anyone undertakes an analysis of Polybius similar to the studies of Herodotus and Thucydides which have appeared lately, we need to determine the nature of the bias that underlies our text of Polybius, along the lines suggested by Momigliano. It is necessary to study what the compiler of the Excerpta Antiqua, for instance, chose from the first five books and more importantly—what he omitted from his collection. This should provide a clear indication of what he has omitted from Books 6-18. One must then compare what he did retain from those books to see whether his standards change in the course of his work. One would do the same for the various volumes of the Constantinian Excerpts to discover on what basis items were chosen for the De Legationibus, On Virtue and Vice, Gnomai, and the rest.

This would be a vast undertaking, for the Constantinian Excerpts contain huge chunks of other historians besides Polybius, running to over 1500 pages. Ideally, to understand the mentality of the excerptors one should know his Herodotus so well that in reading through the Constantinian Excerpts of that author he will know what is missing and comprehend the full extent of the loss. And he will know his Thucydides too, and also his Xenophon. The task requires a tremendous power of imagination, for it is one thing to analyze the texts we have before us, but something altogether different to appreciate the significance of what has been lost.

As an example of the sort of research I have in mind, the remainder of my paper will be given over to an analysis of the way the Excerpta Antiqua were chosen. Since we are concerned today with the direction of future research, it seems more appropriate to present an ongoing project rather than a completed one. Even better, this is not a topic where the answer was known in advance and merely needed to be documented: at the beginning of this line of inquiry I had not the slightest idea of what I would find.

First it would be standard procedure to outline current ideas on the subject, but in this case I have only found one scholar who even addresses the question of how the Excerpta Antiqua were compiled. John Moore, who has studied the manuscript tradition of Polybius, asked himself [in Greek Roman and Byzantine Studies 12 (1971)] why the Excerpta Antiqua are drawn only from the first eighteen books. "No satisfactory hypothesis has so far been advanced as to why they should have ended at this point, but consideration of content suggests the following as a possibility. By the end of XVIII the excerpts had not merely covered the history to the end of the Second Punic War, but had also dealt with the conflict with Philip V which was an almost inevitable result of relations between Rome and Philip during the Second Punic War, and had brought the narrative down to a suitable climax with the battle of Cynoscephalae and the Isthmus declaration of 196 B.C. The presence of some material dealing with other areas would be natural, granted Polybius' method of writing history." That is, he assumes that the Excerpta Antiqua constitute, in Brunt's words, "a faithful miniature" of the first half of the Histories. The inadequacy of such an explanation will be apparent as soon as one realizes that it is the De Legationibus, not the Excerpta Antiqua, which preserves the senatus consultum on the freedom of the Greeks and the joyous celebration of it at the Isthmian Games.

To give some idea of the type of material which the selector of the Excerpts omits I have surveyed Book 3, where the events are well known. He does not include Polybius' introduction (1), which is pretty much a repetition of what he has already copied from the beginning of Book 1 (1.1-3.5), nor does he transcribe the list of wars and other topics which Polybius proposes to treat (2-5). He ignores the theoretical discussion of what constitutes the causes of wars as distinct from the pretexts (6-7), and Polybius' specific application of his doctrine to the Second Punic War (8-15). Thus—a point of some significance for Momigliano's case—the excerptor does not notice the growth of Carthaginian power in Spain (13-15.1), which Polybius regards as one of the causes of the war (10.6). The compiler then skips over the dikaiologia, in which Polybius presents the arguments about where guilt for the war lies (20-21) and offers his own interpretation of the three treaties between Rome and Carthage (22-30). The selector also omits Polybius' argument for the value of studying cause and effect (31) and his apologia for the enormous length of his work (32).

When the historian finally reaches the military portion of Book 3, the compiler ignores Hannibal's preparations, including campaigns to secure Carthage's hold on Spain (33.5-35). Nor does he copy Polybius' account of Hannibal's troop strength (35.7-8; 56.4), which—the historian proudly informs the reader (33.17-18)—is based on an inscription set up by Hannibal himself. Likewise he ignores the distances of Hannibal's route, much of which Polybius painstakingly calculated and retraced in person (39; 48.12). He skips over most of the crossing of the Rhone (42-46) and the Alps (50-56).

Once Hannibal reaches Italy, the compiler ignores the basic strategy of the war, including Hannibal's decision to move away from friendly territory and fight the war on hostile ground (78.5), his instigation of revolt throughout Italy, and—in particular—the motive and significance of his attack on the plain of Capua (90.10-12). The excerptor also overlooks the counterstrategy devised by Fabius Maximus (89.2-90.5) and Roman attempts to take the offensive outside Italy (76; 96-99). We search in vain for Polybius' judgment that the cause of Rome's ultimate victory was its politeia (118.7-9).

We also miss some of the particulars of the action, including the first battle in Italy (65), the Roman retreat to the Trebia (66-68), and Hannibal's various marches through the peninsula, especially the grueling three and one-half day passage of the swamp lying before Trasimene (79).

As you can see, the excerptor has badly defaced Polybius' narrative, and we can expect that the result will be the same in those books where the Excerpta Antiqua constitute the main source for our texts.

What, then, has he included? Only matters military and geographical. He transcribes Polybius' discussion of the three continents and their boundaries (36.6-38.5), as well as his geography of the Rhone and the Alps (47.2-4). His military excerpts include the great battles of the Trebia (70.1-75.4), Trasimene (80.1-85.4), and Cannae (108.2-117.6), which might indicate that his work was directed to the general reader who wanted to know the most significant events of the war, without the burden of superfluous detail and the vast quantity of turgid prose Polybius serves up. But in reading the excerpts from Book 3, I developed the suspicion that our man was really more interested in military stratagems than famous battles, and that suspicion was confirmed in studying the other books.

In the very first excerpt from Book 3, Polybius describes the Roman attack on the island of Pharos, in which they first put men ashore at an uninhabited spot and then sail brazenly into the main harbor. When the defenders sally out to meet the ships, the commandoes ambush them. Or take the crossings of the Rhone and the Alps. The excerptor omits all the fighting at the Rhone (42-43) and simply copies out Hannibal's device for transporting the elephants by tying rafts together (45.6-46.7). He does not even include the actual crossing of the panic stricken animals (46.8-12); he is solely interested in the mechane. It took Hannibal fifteen days to cross the Alps, but the compiler has preserved the activity of a single day (50.1-51.11). Hannibal sends troops ahead during the night to seize the peaks above a certain pass. This first discourages the natives from attacking, but when they see the baggage train in difficulty, they make their assault. Hannibal is then able to attack them in turn from the high ground. Other obvious stratagems by the Carthaginian are his staging of a gladiatorial contest between two prisoners of war to demonstrate to his own men that it is better to die fighting than to submit to slavery, and his trick of attaching torches to the horns of cattle so that Fabius Maximus follows the animals while Hannibal's troops escape in the opposite direction.

Hannibal wins many of his victories by ambush, and the excerptor includes two such stratagems based on the principle of knowing the personality of the opposing general. Chapter 81, included in the Excerpta, contains Polybius' reflections on the subject in general: "For there is no denying that he who thinks that there is anything more essential to a general than the knowledge of his opponent's principles and character, is both ignorant and foolish." One must see "what are the weak spots that can be discovered in [the opponent's] mind," and exploit such faults as sloth, drinking, lust, cowardice, and stupidity. "Rashness … and undue boldness and blind anger, as well as vaingloriousness and conceit, are easy to be taken advantage of by his enemy and are most dangerous to his friends; for such a general is the easy victim of all manner of plots, ambushes, and cheatery. Therefore the leader who will soonest gain a decisive victory, is he who is able to perceive the faults of others and to choose that manner and means of attacking the enemy which will take full advantage of the weaknesses of their commander."

Hannibal puts these precepts into practice by learning the character of Flaminius and Marcus Minucius. In a passage from the Excerpta which calls to mind Thucydides' description of Cleon at Amphipolis, Polybius says that Hannibal found out that "Flaminius was a thorough mob-courtier and demagogue, with no talent for the practical conduct of war and exceedingly self-confident withal. [Thus] he calculated that if he passed by the Roman army and advanced into the country in his front, the Consul would on the one hand never look on while he laid it waste for fear of being jeered at by his soldiery; and on the other hand he would be so grieved that he would be ready to follow anywhere, in his anxiety to gain the coming victory himself without waiting for the arrival of his colleague." Flaminius then follows Hannibal into the trap at Trasimene, a passage which the selector transcribed in full. He also chose to include a relatively minor affair in which "Hannibal … was aware of the rivalry [between the two dictators Fabius and Minucius] and of Marcus' impulsiveness and ambition." Now Minucius was also puffed up by a minor victory, so Hannibal devised a plan to take advantage of his over-confidence. He placed men in ambush during the night and the next day began to fortify a hill that lay between his position and the Roman camp, "well knowing that owing to his previous achievement Minucius would instantly advance to frustrate this project." When the Romans came out to dispute possession of the hill, Hannibal unleashed his men from the hiding-place.

Even though some of the excerpts from Book 3 are quite colorful and might interest the general reader, the evidence is strong that the excerptor was preparing a book with a fairly narrow focus for those concerned with military tactics rather than simply a condensed version of an overly long historical classic. In university terms, he was compiling a text for a course in military science, not ancient history.

In the other books where our text is complete, the epitomator is more catholic in his taste. Once again we find the geographical passages: descriptions of Italy (2.14.4-17.8), Byzantium and the Black Sea (4.38.1-45.8), and Media (5.44.3-11). In addition, we find selections that seem to correspond in type to the material which appears in the Constantinian Excerpts. We have character sketches of Aratus (4.8.1-12) and Philip V (4.24.4-7; 5.9.1-12.8), the sort of thing that appears in the collection On Virtue and Vice, and about a dozen general observations that could be classified as gnomai, such as, "So great is the difference both to individuals and to states between carefulness and wisdom on the one hand, and folly with negligence on the other, that in the latter case good fortune actually inflicts damage, while in the former disaster is the cause of profit."

We have the military excerpts, of course, including two set pieces, Sellasia (2.65.6-69.11) and Raphia, and a trio of lesser encounters, which—at first sight—seem to illustrate different types of fighting that a general must master. One is the battle of Caphyae (4.11.1-13.2), where Polybius stresses the importance of terrain, criticizing Aratus for his failure to use his heavy infantry against the Aetolian light infantry in the plain, where he would have had the advantage. In another excerpt (2.25.1-31.7) the Gauls are caught between two Roman armies and accordingly line up with one half of their army facing one way, and the other half facing the opposite direction. Finally, a battle between Philip V and the Spartans (5.22.1-23.6) seems to be about breaking through a strong defensive position.

The selector has also chosen two stratagems devised by Hamilcar Barca (1.75.5-76.9). Noticing that the Bagradas River, which blocked his advance, became very shallow at its mouth when the wind blew strongly upon it at certain times, Hamilcar was able to lead his army across at just the right spot and just the right moment. Thereupon he defeated his opponents, who attacked from two sides, by feigning retreat and—once the enemy broke ranks to pursue—by wheeling around into battle formation.

There are three passages of general historical interest: the preface to Book 1 (1.1-3.5), Polybius' summary of the growth of Roman power up to the First Punic War (1.6.1-7.11), and a gnome emphasizing the importance and difficulty of writing a universal history (5.32.1-33.8). Finally, we have a seemingly inexplicable choice about the arrival of a Numidian defector to the camp of Hamilcar (1.77.6-78.15).

Despite exceptions, it is clear that we have a military, not a political, compendium. The compiler has no interest in the history of the Achaean League, the tyrants of Sparta, or the social question in Greece. Even the military selections are not representative, as they favor tactics over strategy. The selector entirely omits the First Punic War in favor of the attack on Pharos, includes nothing to explain the Mercenary War of which the battle of the Bagradas is part, and has no interest in long marches, such as Hannibal's great treks or the lightning campaigns of Philip V through the Peloponnese which drew the admiration of Polybius. His prime concern is what Polybius calls "plots, ambushes, and cheatery," and this interest grows more pronounced in the remaining books.

Instead of going through the whole of the Excerpta Antiqua, I have chosen to compare the selections from Books 7-11 with those from the first pentad. We find once again extracts about geography and observations on the character of such men as Aratus (8.12), Antiochus the Great (11.34), and Hasdrubal (11.1.2-3.6). The excerptor is especially interested in the character of Scipio Africanus, how he relies on planning and does not trust to luck (10.2-17), and how he wins the support of his troops by his own moderation in the treatment of women prisoners (10.18-19) and his refusal of kingship (10.40). There is one major battle, Philopoemen's victory at Mantinea (11.11-18), one document, Hannibal's treaty with Philip (7.9), and at least one embassy speech (11.4-6). We also find a number of gnomai, including two historiographical passages. In one Polybius again emphasizes the need for universal history (8.1-2); in the other (9.1-2) he justifies his omission of the entertaining in favor of the useful, which is, of course, the excerptor's own creed.

By far the most prominent feature of the second pentad is the large number of stratagems. Their importance to the excerptor is best demonstrated by his inclusion of stratagems which are almost identical to those which he copied from Books 1-5. A campaign of Antiochus corresponds to that portion of Hannibal's crossing of the Alps which the selector included in his book. Hannibal learned that the natives guarded the passes by day and returned home at night, and thus was able to seize the peaks in their absence. Antiochus twice found barbarians guarding passes and by occupying the peaks was able to drive them out with a shower of missiles (10.29.3-31.3). Later in his journey, hearing that some other natives guarded a river crossing during the day but went home at night, he rode ahead and seized the spot when they had retired for the evening. Then he bravely fought off their counter-attacks (10.49). Antiochus' Anabasis lasted some six years, but apart from the capture of a fortified town (10.31.6-13) these exploits of the king are the only actions that caught the selector's eye.

Again in the first pentad the Romans at Pharos, and Hannibal in his duel with Marcus Minucius, first set an ambush and then lured the enemy out of a protected position by provocatively dispatching troops in plain sight. So the Excerpta Antiqua from Book 8 contain a story of how Philip V caused his own men to suffer a defeat in the open field so that the garrison left the Acrolissus in quest of booty, only to fall into his ambush. And the excerptor chose to preserve a battle in which Scipio lured the Spaniards into a meadow by placing cattle there as bait. Just as Hamilcar noticed that the winds made the Bagradas shallow at a certain spot, so Scipio observed that low tide exposed a land bridge to a place from which the citadel of New Carthage could be climbed. He told his men that Neptune promised him in a dream to help the Romans, and the dream seemed to come true when (just on schedule) the waters receded, opening the fortress to attack (10.8.6-7; 11.7, 14).

After taking the trouble to record Hannibal's device of joining rafts together to ferry elephants, the excerptor now copies Polybius' account of how the Romans tied ships together at the siege of Syracuse in order to support giant war machines. He also preserved Polybius' detailed description of these implements as well as the ingenious devices which Archimedes prepared for the defense. He also included the description of Philip's war machine at Echinus, a kind of moving scaffolding with troops stationed on three floors (9.41).

And there are new stratagems. Polybius himself uses the word (11.22.1) in connection with Scipio's two tricks that won the battle of Ilipa (11.20-24): first he brought his army out to fight before the Carthaginians had a chance to eat and then stationed his best men on the wings with the unreliable Spaniards in the middle. The excerptor also transcribed the battle of Baecula, in which Scipio again attacked on the wings while holding his center in reserve (10.33).

In another selection Polybius recounts Hannibal's march on Rome aimed at drawing off the Roman army besieging Capua, and compares it with Epaminondas' march on Sparta, which forced the defenders of Mantinea to leave it unguarded (9.3-9).

One of the collections of Constantinian. Excerpta consisted of stratagems. Although it has not survived, we do know that one of the items was a passage in which Polybius describes how Philopoemen trained his men in all the various cavalry maneuvers. This selection also appears in the Excerpta Antiqua (10.23), as does a passage in which Scipio trains his forces at New Carthage (10.20).

Hannibal's assault on Tarentum (8.24.4-34.13) involves a series of stratagems: the way the traitors arrange to come and go at will by pretending to hunt or pillage (24.9-13; 25.4-11); Hannibal's secret march on the city from a distance of three days, which he capped off by arresting everyone outside the town to prevent detection (26); and his device of hauling ships overland from the harbor to the outer sea in order to blockade the Roman garrison (34).

Finally, the capture of Sardes, based—like many of these stratagems—on the principle of hitting the enemy when and where he least expects it. Buried in Polybius' account of the battle of the Trebia is an observation (3.71.2-3) that teaches this lesson: "the Romans, while very suspicious of thickly-wooded ground, which the Celts usually chose for their ambuscades, were not at all afraid of flat and treeless places, not being aware that they are better adapted than woods for the concealment and security of an ambush." Three times, then, in the second pentad the excerptor found material to illustrate that a citadel can be taken at its strongest point, where no one would expect an attack. Philip took the Acrolissus because, relying on its natural strength, the enemy set only a few men to guard it (8.13.9) and the king was able to fool them. At New Carthage Scipio attacked at a spot which was usually under water and so left unguarded (10.14.13-14). And Sardes was taken when one of Antiochus' men saw that vultures and other birds alighted on a portion of the wall of the citadel. He inferred that the defenders were not standing guard there, relying instead on the sheerness of the rock wall. That, of course, is where the attackers scaled the heights and entered the city (7.15-18).

The key to the selection process of the Excerpta Antiqua comes, I think, in a passage from Book 9: "The accidents attendant on military projects require much circumspection, but success is in every case possible if the steps we take to carry out our plans are soundly reasoned out. That in military operations what is achieved openly and by force is much less than what is done by stratagem and the use of opportunity, can easily be learnt from the history of former wars" (12.1-2). So victory in war is won by the general who—like Scipio—relies on planning and does not trust to luck and knows how to employ stratagems. But the successful leader must master other elements of war, for the quotation continues (12.3-5), "And it is no less easy to be convinced by facts that in those actions depending on the choice of opportunity failure is far more frequent than success. Nor can anyone doubt that most of the failures are due either to error or to negligence on the part of the commander. We must therefore inquire in what such faults consist."

He then proceeds to show the practical importance of astronomy and geometry to a commander, noting, for instance, that Philip V and Cleomenes missed a rendezvous with men who planned to betray their own cities, because each king neglected to figure the time of nightfall according to the season (9.18). Throughout his collection from 7-11 the excerptor has made sure to include other admonitions to would-be leaders. One of the gnomai which he copied from Book 1 says, in effect, that being a soldier is one thing, but being a general is another. Thus he also records Polybius' advice not to become involved in minor skirmishes, such as led to the death of the consul Marcellus (10.32-33.7), but to manage the broader concerns of a campaign, as exemplified in the excerpt on how Philopoemen rode sometimes bravely at the head of the cavalry, but then wisely dropped to the rear or inspected the center of the corps in order to gain a view of the whole situation (10.24). When Scipio fights at New Carthage he has three shieldbearers to ward off missiles (10.13.1-5). The Excerpts also include Polybius' teaching that a general should not rashly entrust himself to his enemies, illustrated by the fates of Ti. Sempronius Gracchus and Achaeus. And while some might prefer to view the episode as an indicator of Scipio's character, I would cite as an example of how to avoid danger the excerpt (11.25-30) in which he puts down a mutiny among his troops through a combination of rhetoric and ruthlessness.

The excerptor includes a long discussion of the Roman method of dividing booty, which Polybius recommends as a safety measure (10.16-17.5). Since under the Greek method, "most of the men start pillaging, commanders cannot maintain any control and run the risk of disaster, and indeed many who have been successful in their object have, after capturing the enemy's camp or a town, not only been driven out but have met with complete disaster simply for the above reason." The selector clearly shares the historian's didactic purposes, for the passage ends, "Commanders should therefore excercise the utmost care and foresight about this matter, so that as far as is possible the hope of equal participation in the booty … may be common to all."

One final selection shows clearly that we are dealing with a military handbook and not a faithful miniature, viz., Polybius' comparison of his own system of fire signals with the one recommended by Aeneas Tacticus (10.43-47). The excerptor allotted almost as much space to this item as to the famous description of the Roman constitution in Book 6 (11.11-18.8).

Is it credible, then, that during the tenth century someone would compile a series of excerpts from Polybius to serve as a textbook of tactics? The answer is clearly, yes, for the great textual critic A. Dain has observed that the common Byzantine practice of reworking classical literature (retractatio) was a response "to the need to adapt an ancient work to new conditions." The Emperor Constantine himself writes that he regularly took certain books with him on campaign: books on generalship, weaponry, and history, especially Polyaenus and Syrianus. Five abridgments of Polyaenus were available during this period, and his stratagems were rearranged according to type to make them more useful. The Excerpta Antiqua are the result of reworking Polybius in the same way.

The topics chosen from Polybius are not merely useful, they are the main concerns of Byzantine commanders. According to Toynbee's excellent analysis of military handbooks of the tenth century [Constantine Porphyrogenitus and His World, 1973], the primary Byzantine tactic was to follow marauding invaders and cut them off at the pass, seizing the heights with light-armed troops who would rain missiles (especially arrows) down on the enemy. This is why the only excerpt from the crossing of the Alps is Hannibal's occupation of the heights, and why the selector recorded Antiochus' seizure of mountain peaks and a river crossing. When the Byzantines raided Muslim lands, says Toynbee, "The most difficult and dangerous of all contingencies for an East Roman Army is to find that a kleisoúra on the Army's line of retreat from enemy territory has been occupied by the enemy. Keep away from any occupied kleisoúra. It is the terrain, not the enemy himself, that creates the danger. Terrain can enable a weaker force to defeat a stronger. If you can, compel the enemy to evacuate the kleisoúra by sending infantry through other kleisoúrai, that he has not occupied, to take him in the rear. If the enemy cannot be forced or tempted to evacuate, and if his position is impregnable, take one of the sideroads." This would explain the excerptor's interest in Hannibal's escape from the plain of Capua with his trick of attaching torches to the horns of cattle. It would also explain why he included a description of Philip V's victory over the Spartans (5.22.1-23.6). The King was returning north after ravaging Laconia (5.19) and found his way blocked since Sparta lay on his left and the Eurotas and some hills on his right. The forces within the city could attack his flank as he marched past, and Lycurgus had occupied the hills. Since there were no side roads, Philip had to force the position, driving his opponents from the heights.

I originally thought that the selector included the battle in which the Romans attack the Gauls from the front and the rear (2.25.1-31.7) as an example of how to manage fighting on two fronts, but now we can see the real reason for this excerpt: the Gauls were heading home with an enormous haul of booty when a second Roman army unexpectedly arrived by sea at Pisa and then seized the pass through which the Gauls had to proceed.

Sir Charles Oman says [in The History of the Art of War in the Hiddle Ages, Vol 1, 1924], "Of the spirit of chivalry there was not a spark in the Byzantine … [He] was equally remote from the haughty contempt for sleights and tricks which had inspired the ancient Romans, and from the chivalrous ideals which grew to be at once the strength and the weakness of the Teutonic West … They considered it absurd to expend stores, money, and the valuable lives of veteran soldiers in achieving by force an end that could equally well be obtained by skill." It is in this tradition that our excerptor copied down Polybius' observation (9.12.2) that "in military operations what is achieved openly and by force is much less than what is done by stratagem and the use of opportunity." The Byzantine, according to Sir Charles, "had a strong predilection for stratagems, ambushes, and simulated retreats." We have already seen the selector's great interest in ambushes and in Hamilcar's feigned retreat at the Bagradas. A favorite device of the Byzantine handbooks, and one actually used against the pretender John Bryennius, is to send out a small force to attack the enemy or to ravage his territory. When they encounter resistance, they flee pell-mell, and lead their pursuers into an ambush. The Excerpta Antiqua include two battles (3.104-105; 8.13-14) in which commanders use their men as bait.

One of the main themes of the handbooks is the need to exercise troops in military maneuvers and discipline well before heading off to meet the foe, and Leo the Deacon on at least half a dozen occasions mentions that a general prepared his forces this way. This interest accounts for the preservation of 10.23, where Philopoemen exercises his men. The Byzantine use of fire signals explains the selection of Polybius' discussion of this device. Another passage which only begins to reveal its meaning in the light of Byzantine tactica is Polybius' account of Hamilcar's conduct in the Mercenary War (1.77.6-78.15), beginning with the arrival at his camp of an African noble seeking to defect. As Toynbee puts it, "The interrogation of prisoners and deserters is also considered to be indispensable," and Polybius shows Hamilcar examining the young man who wants to become his helper, as well as Antiochus scrutinizing the Cretan who offers to betray Achaeus (8.17.2). Finally, we notice the similarity between Hannibal's march on Rome, intended to relieve Capua, and "a brilliantly successful diversionary raid into Cilicia that was made by Nikêphóros Phokás the elder" which forced "the Cilician Moslems to raise their siege of the East Roman fortress Místheia."

It is easy to believe that the Excerpta Antiqua were chosen for their similarity to Byzantine military practices. That in itself does not mean that a Byzantine commander was consciously following Polybius whenever he outfoxed the enemy, but it does mean that the study of Polybius, Polyaenus, and handbooks which go back to the Roman Empire helped shape the overall outlook of a tenth century general.

There are a few topics remaining to complete this research project. We need to examine the character sketches in the Excerpta Antiqua more closely to see whether they correspond to those portions of Byzantine manuals which discuss the qualities of a good general. Professor Pearson, I am sure, will want to know what attracted the compiler to preserve part of the historiographical Book 12. Perhaps the largest problem of all is Book 6. I always used to wonder why anyone would be interested in that tedious description of the Roman camp. Now I know: Byzantine handbooks are filled with similar descriptions which go back—by the process of retractatio—to Polybius. Now, paradoxically, I wonder why the man who showed so little interest in politics chose to preserve Polybius' discussion of constitutional cycles and his description of the Roman politeia.