Polybius has received very great praise for his insistence upon accuracy. Professor Shotwell[7] ends an enthusiastic chapter with the sentence: “But as long as history endures the ideals of Polybius will be an inspiration and guide.” The praise is deserved, especially when we remember that Polybius had behind him in Greece nearly two centuries of extravagant rhetorical history. But when we ask how it happened that he turned his back upon all that tradition, no explanations are offered. It is not an adequate interpretation to say that by living in banishment he was removed from the temptations of historians writing the story of their own people, for he usually succeeds in being quite objective even when he writes of the Achaean League. Is it not likely that his contact with matter-of-fact and legal-minded Roman senators induced him to adopt some of their manners and methods? His respect for the integrity, sanity, and uprightness of Roman senators of the Scipionic period he voices repeatedly[8] in contrasting their qualities with the unreliability, astuteness, and fickleness of his countrymen. It is also to be remembered that the first part of his history is based upon Fabius, who therefore was his first preceptor in historical writing. It would seem at least worth considering whether Polybius did not owe some of his qualities as an historian to the fact that he served his apprenticeship in history among the early Roman annalists and that he adapted his work to the public which had been brought up on those matter-of-fact books. At any rate he is well-nigh unique among the Greeks who wrote history after the classical period.
There is of course nothing to indicate that Fabius and his immediate followers were in any sense great historians. Without any literary background, with only such practice in writing as would come from composing state documents, occupied every day with the concerns of a rapidly expanding state, they recorded only public acts and public discussions. What men did and strove for, outside of the voting, legislating, and fighting groups, was not recorded. Not even within their chosen field does there appear a penetrative analysis of senatorial policy. Fabius, to be sure, enumerated the immediate causes of both of the Punic wars but only with a jurist’s interest in deciding at what point the enemy had committed the breach for which he deserved punishment. As historians these men had the limitations of their qualities and of their occupations. But on the other hand there is no evidence that they knowingly garbled facts.
One may, then, be permitted to object to a common error of judgment regarding the nature of what is called the “scientific method” in ancient history. Students who have to deal with the gullible medieval chronicles seem to assume that historical criticism has but recently succeeded in creating a respect for objectivity and honesty in history, as though the logical processes of the mind were not fully developed in the human race twenty thousand years before the invention of the historical seminar. The incubus of religious authority dominant for centuries in the Middle Ages was a passing phase, as was the overweening respect for dramatic values in the Hellenistic historians and the eagerness to glorify families and the state in the Sullan romancers. But just as Polybius, when transplanted into a soberer atmosphere of action, rid himself with ease of the Hellenistic methods; as Julius Caesar, when occupied with absorbing actualities, could free himself from the habits of his day so far as to record the very crimes for which he was being assailed by Cato in the senate; as Ari Frodi in Iceland escaped churchly influence sufficiently to write the history of his island with the same respect for truth that he used when judging a case at the “thing,” so the early statesmen-annalists of Rome, when recording what was available for the historical period of the Republic, employed documents and personal observations with the same meticulous care that they used when presiding as praetors in the courts or when as senators arguing cases of international relations. Their brief historical notes are largely preserved for us in Polybius, in Cicero’s De Republica, in Diodorus, and in the central skeleton structure of Livy, and the continuous existence of these notes in Roman times kept the legends from ever straying wholly beyond the reach of actuality. This also explains why it is that archaeological knowledge now coming to hand is so frequently found to fit in with what we have been wont to call “tradition.”
The various currents of Roman historiography united in the vast work of Livy, so that, Augustan though he is, he may be taken as a typical product of the several Republican schools. There is no one formula by which the historian may employ Livy without constant caution. Parts contain unadulterated legend, parts that seem at first glance to be sound record are based upon treacherous sources, much is first-rate history; but who has the magic flail that will shell off the husks? There is no more insistent problem in Roman history than the correct use of Livy, for he is, over large areas, our only source, and over periods where he parallels Appian and Cassius Dio he is generally so much sounder than they that he must be threshed through.
In estimating the quality of the thirty-five books extant[9]—unfortunately his early work and not the maturest product of his mind—we must distinguish between the results that are due to his own aims and capacities and those that are due to the nature of his varying sources. Everyone now admits with Tacitus[10] that Livy was scrupulously honest, that he was fair, that he did not permit himself to fabricate—as Caelius and Valerius seem to have done—and that he chose good sources when they were available; but a historian needs more than these virtues. What we miss most in this respect is his failure to go insistently to primary sources. To be sure, it was impossible for a man who set out to write a vast popular history—about three times the size of Gibbon’s great life-work—to delve in the archives. Those documents were not then catalogued and classified as they now are. Cato the younger, for instance, when he needed an abstract of the treasury office for a relatively brief period had to pay his assistants some 30,000 denarii to have it made.[11] Ten times the amount would not have sufficed for Livy’s extensive needs. He accordingly made use of what had been published, such things as the Annales Maximi, collected down to the Gracchan period, the magisterial lists as they had been revised by various hands, and collections of laws and senatorial decrees that had been made for the use of lawyers and law-makers. And some of these skeleton bones of history he took from conscientious annalists like Fabius and Piso, who specialized on such matters because they wrote not for the public but for members of the senate and the ruling nobility. Livy’s purpose seems to have been to write a readable and full history of Rome which would displace the unreliable fictionalized history of Valerius Antias by being equally well written but far more reliable. But if he had insisted upon primary sources only he would not have completed one-tenth of his very extensive task. Given his aim and purpose, his duty was to find and exploit the best published documents and histories for each period, and with very few exceptions this is what he did.
It was also his purpose—which a modern historian might well deny himself—to set down the early legends of Rome. Here there were no historical sources, and the question was whether to omit the legends—as Mommsen has done—because they could not be considered worthy of credence, whether to rationalize them and attempt to rescue a kernel of fact as Piso did, and as Pais and Beloch have recently attempted to do, or finally to set them down as found, with a warning that they were legends. Mommsen’s method was facile but we are glad that Livy did not use it. The legends are good literature; they also have a great value in revealing the temper of those who accepted them and passed them to future generations as worth having. Finally they prove upon comparison with archaeological facts to have a sounder basis in fact than Mommsen thought. Even if all their details be legendary, they represent a Rome that could not have been far from the actual state. In fact they prove to be nearer the actuality than the strange and lifeless civilization that Mommsen reconstructed for the early period out of unscientific etymologies and stereotyped conceptions of late legal institutions.
We are also glad that he did not follow Piso’s lead in trying to use them “critically.” Had he done so he would have transmitted them in garbled form and spoiled them, and won nothing in the process. We have learned from recent attempts that this method is a failure. A Charlemagne reconstructed from medieval French epics or a Theodoric shelled out of the Diedrek legend would at best not be accurate history. Thirty years ago our hyper-critical historians tried it, and moved all the early dates of Roman history down a century or two. Archaeology has at least proved this a mistake, and we now are moving the dates back and most of the critics have got into the moving van. After all is said Livy’s method was the soundest. His procedure was the more nearly scientific. It is with exceeding good sense that he says in his preface: “The early stories regarding Rome’s foundation that are handed down to us in poetic romances rather than in sound historical records it is not my intention to support or to refute.” And again in the preface of the sixth book he warns us that very nearly all that he had written in the preceding five books—up to the burning of Rome in 387—rested not on acceptable records but on legend. And even thereafter, throughout his work, whenever for any incident he is limited to the authors who employed legend he is quick to warn the reader of the nature of the source. These passages show that Livy was a sounder critic of Rome’s legends than Polybius was in respect to Homeric stories. Historians who scold Livy for his preservation of legends have not only missed their value but have misunderstood Livy’s cues.
We have perhaps a fairer quarrel with him for following the Greek custom of inserting fictive speeches in the body of his work. To the modern reader many of them are tedious and create a suspicion of being unreliable. It is never quite safe to quote a line from these speeches as indisputable evidence on any event, though most of them contain the gist of an actual speech delivered on the occasion stated. All we can usually be sure of is that they give Livy’s conception of what was likely to have been said by the speaker in the situation. That is often worth having, for Livy usually knew more of the pertinent conditions than we do and he possessed a sympathetic penetration into pristine characters and events that enabled him to make valuable reconstructions. One has only to read the several speeches attributed to Scipio Africanus to see that they make a consistent and vivid portrait. If we have the patience to read these speeches with Livy’s purpose in mind we shall know how to profit by them. The convention was of course understood, and was no more misleading than the equally artificial convention of modern historians who employ a kind of fictional mind-reading, a “stream of consciousness” device, which may be found on almost any page of Mommsen or De Sanctis. Mommsen could hardly have made a silent character like Caesar real without constantly conjecturing as to his intentions and motives, as when he writes: “Evidently here too it was Caesar’s intention,” or again “When Caesar projected the plan for a new code, it is not difficult to divine his intentions” (and he puts down a page of divining), or again, “So far Caesar might say that his object was attained.” These musings of a great historian of our time are cast in a different form from the invented harangues of Livy and the Greeks, but we read them with the same caution, knowing that they are surmises. The historian, who like Livy and Mommsen must deal with tantalizingly fragmentary sources, must have the liberty to bridge the lacuna by some such method. But we will be on our guard when reading such matter.
Thus far I have spoken of Livy’s work as affected by his aims and methods. What is even more important for the reader who uses Livy is to comprehend the varying quality of the available sources. For the long period before 200 B.C. there was of course no writing of history at Rome. Very meager records existed for most of the obscure period, 500-280, and these had been exploited by Fabius, but they made no story that could be told in a consecutive form. Hence their data were welded together with the help of legend during the second century before Christ. Of the story of the Samnite wars the mere skeleton is all we can accept as firm history. And that was as true before Livy wrote as after. Neither he nor anyone else could mend matters. For the Pyrrhic and First Punic War the sources were good but the corresponding part of Livy is lost. For books 21-30 the sources were full. Here two responsible participants, Fabius and Cincius, told the story from the Roman viewpoint, while three companions of Hannibal told the same story as they saw it from the Punic camp.[12] Any tendency to exaggerate on either side could at once be checked from the reports that came from the other, and the excellent Greek historian Polybius came soon after and did a great deal of checking. Here Livy had only to be diligent, fair, and honest to be able to write reliably. The third decade of Livy is accordingly as dependable history as we have of any ancient war. It is only in the brief Spanish portion, for which there was no Punic account, and where Polybius himself had written too enthusiastically of Scipio’s work, that we touch quicksands.
Books 31-45 are not quite so firm. The chief difficulty here is that there was no contemporaneous historian at Rome for this period except Cato, who wrote a very brief account of a part of it. Polybius was the first to compose the whole story, but excellent as he was, he came some years after the events, had observed, so far as he did, only from his home in Greece, depended largely upon biased Rhodian writers, and knew so little about Rome’s activities outside of Greece that he omitted much, in fact all of Rome’s internal and western history. Campaigns in Gaul and Spain, for instance, did not get recorded at all until they were well permeated with legend, and there was no account available from the opponent’s side. Hence it is that here Livy is of necessity exceedingly uneven, treading on a fairly firm corduroy for most of the important events in Greece, but on a marshy ground of semi-legend when he has to deal with western campaigns. Fortunately the somewhat scanty documents of the state archives had been well culled before him by reliable men like Piso, and these usually kept the legends from dangerous extremes. It is a complete misunderstanding of Livy to suppose that he did not know the weaknesses of Valerius Antias when he used him. Livy knew them all along, but in some portions of this period he had no good source, had nothing available but Valerius and his kind, who had set things down as they heard them, fables and all. Livy’s frequent citation of Valerius Antias does not betoken a gullible love of this writer, but is intended as a danger signal. Here there are several boggy spots. But fortunately the period deals largely with eastern affairs and for that portion the sources were fairly good. It is fair to say that Livy did as well as was to be done in his late day with the material and time available, and that nine-tenths of this portion is acceptable history.