Appendix III

The Roman Manpower Question

Throughout this work, the issue of the shortage of Roman manpower for military service has been a fundamental one. Yet despite all the evidence, the issue of available manpower and the reasons behind it have been a bone of contention amongst historians throughout the twentieth and twenty-first centuries. As it is such a fundamental issue, it is worth reviewing the principal problems that confront us.

The Ancient Evidence

The widely-held assumption that underlined the ancient accounts of the manpower issue that affected Rome in the second century BC, was one of a recruitment crisis caused by military overstretch. In short, the ever-increasing length and complexity of the wars in the second century BC, especially in Spain, meant that the soldiers were away from their farms for ever-increasing amounts of time, and that this neglect led to them falling into ruin and being bought up by the rich landowners who created great estates run by masses of slaves. This, they believed to lie behind the proposed reforms of Ti. Sempronius Gracchus in 133 BC, and was only solved by Marius’ supposed abandonment of the property qualification, which allegedly cut the ties between the landed farmer and military service.

As we have already seen, not only were Gracchus’ proposals more complicated than that, but Marius’ actions did not mean a permanent abandonment of the property qualification needed to serve in the Roman military. As well as the narrative sources attributing these reasons to Gracchus’ actions and the underlying problems at the time, there are two other contentious areas of ancient evidence: the Roman census figures and whether the level of the property qualification for military service was lowered during the second century.

Roughly every five years, the Roman Senate elected two senior members from their ranks, distinguished ex-consuls, to serve as censors. These two men served for up to an eighteen-month period and their primary duty was to conduct the census of Roman citizens.⁴⁷⁷ For the latter part of the second century BC, the given figures are as follows:⁴⁷⁸

There are immediately two problems with these figures. Firstly, we have no clear idea of the criteria used to register the citizens on this list, whether it was all male citizens or simply those of the first five property classes (those eligible for military service). Nor do we know how accurately the census recorded all the Roman citizens in Italy. The ancient sources record the penalties for failing to make oneself available to register, but it has been questioned how enforceable these were in a region the size of Italy.⁴⁷⁹

The second problem is that we have no direct surviving source for these figures, merely the surviving narrative works, primarily Livy’s histories and its epitomes. This immediately leaves them open to mis-transmission throughout the 1,500 years between when they were written and the early modern period when the surviving manuscripts were analysed and standardized. Thus, despite the fact that the figures we have show a declining citizen population until 131 and then a massive jump, which is most widely ascribed to the Ti. Sempronius Gracchus’ land reforms, there have been enough reasons to argue against them.

The other area of ancient evidence is centred on discussions over what the figure was for the lowest of the five census classes and thus what level of wealth a man needed to be eligible for military service. Here we have three different figures: 11,000, 4,000 and 1,500 asses. The figure of 11,000 asses comes from Livy and dates back to the supposed origins of this system, back in the sixth century BC (thus some 500-plus years before he wrote about it). The figure of 4,000asses comes from Polybius. Several late Republican sources, including Cicero, give us the figure of 1,500 asses, and seem to backdate it to the period under discussion, the second century BC.⁴⁸⁰

The most common arguments presented by modern historians of Rome are that at some point, most likely around the Second Punic War or soon after, the figure needed for military service was dropped from 11,000 to 4,000 and then again to 1,500 after Polybius, but predating Marius. The central problem with these theories is that there is no tangible shred of evidence to support them in the ancient sources. Again, as with Marius’ army reforms, much of the modern argument is built from an assumed event that took place in our sources. However, of the three figures, only Polybius’ can be treated as contemporary, and even that has the problem that the figure was given in drachmas (as he wrote for a Greek audience, in Greek) and has to be calculated back into asses by modern historians of Rome. We do not know how accurate Livy’s figure was for the sixth century, or even if the system dated back to the sixth century. Cicero’s figure may have been accurate for his day but we have no actual evidence that it was in force prior to Marius.

There is one other possible reference to changes in recruitment practices and that comes from a fragment of Cato, which has been interpreted as advocating the recruitment of the proletarii (those who fell below the minimum qualification) into the military, but the source is too fragmentary to allow any clear meaning.⁴⁸¹ Thus many of the arguments over this issue are built on houses of sand, and there is no actual evidence that the Romans did lower the minimum property qualification needed for service in this period.

The Modern Theories

With all these problems with the ancient evidence on this issue, many modern historians of Rome have searched for alternative fields of evidence to validate or invalidate the ancient evidence, notably archaeology and demography. For those who look to archaeology to shed light on this issue, the central argument is whether we can determine evidence of these large-scale slave estates that Gracchus talked of.⁴⁸² In short, the answer appears to be a negative one, but that does not automatically mean that they did not occur, due to the vagaries of finding sites that would both give us the detail and the scale needed for such an exercise.

Recent years have seen a growth in the use of demographic models of population growth and decline to see if they can shed light on this situation.⁴⁸³ Whilst such theoretical constructs do provide us with a new approach to studying the limited evidence we have, we must always be aware of their limitations, namely that they are indeed simply theoretical models, and should exercise extreme caution if they clash with our ancient testimony. Such an example can be found in one of the latest works in this field, by Keaveney, who criticizes an earlier work by Morley:

There have, however, been attempts to discredit Tiberius on the part of scholars who share Rosenstein’s view that there actually was a population increase. Morley (2001) is one of those. Approaching it from the standpoint of the demographer, he comes to the conclusion, based on his own calculations, that Tiberius has got it wrong. I believe we may have a fundamental difference of approach. Morley’s calculations can sometimes seem complex, but this is because essentially they are mathematical exercises and cannot have anything more than a hypothetical value. Recognising this I believe, unlike Morley, that when offered a choice between clear-cut unequivocal ancient source and a fragile modern construct, we must surely choose our sources.⁴⁸⁴

Amongst the many factors that may have been involved in this manpower issue in the period under review, two stand out. Evans pointed out the fundamental difference between the manpower theoretically available to Rome for military service, as seen in the census figures and the numbers who actually made themselves available to serve. In short, he focused on the problem of draft dodging.⁴⁸⁵ Throughout history, it is a given that there will always be a large percentage of men who will avoid compulsory military service, made all the easier by the vastness of a pre-industrial Italy, no matter the harsh penalties for those caught.

Rosenstein in a recent work has also done much to dispel the simplicities of the ancient model of extended military service resulting in the loss of farms and thus eligibility, from at least two areas. Firstly, he argues that extended military service was not unique to the second century BC and had in fact been a long-standing issue as Rome fought more complicated wars in Italy in third century BC. Secondly, he theorizes that Rome’s system of inheritance, whereby a father divides his property up equally between his children, meant that whereas the father might have enough wealth to be eligible for military service, by dividing it, his sons would all be ineligible. This in itself is an excellent theory, though he does then go on to postulate that despite the census figures the Roman citizen population actually rose in the second century BC, rather than falling.⁴⁸⁶

These few examples are merely presented to illustrate what is one of the most complicated and historically fertile issues in Roman history. Two clear conclusions can be drawn. Firstly, it is clear that the picture presented in our few surviving ancient sources can present too much of a simplistic representation of what must have been an extremely complicated socio-economic phenomenon, one that the Roman recognized was having implications for their military system. The second conclusion is that modern commentators on the ancient world must always watch themselves for constructing false arguments where there is no ancient evidence or dismissing the testimony we already have.

Ultimately, we can rely on the basic ancient premise that Rome indeed did suffer a problem in finding enough quality soldiers for her ever-increasing wars. Yet we must be careful not to assume that this was the sole reason for any presumed decline in her military efforts. As the Gallic Wars of the 120s showed, Rome in this period was more than capable of conducting a successful war against a tribal enemy. That they failed against the Cimbri for so long had more to do with more immediate factors, such as Roman commanders and Cimbric numbers rather than any deep-seated flaw in the Roman system. If there was no such flaw then there was no need for such a radical cure, such as the assumed abolition of the property qualification. Rome in this period did face a crisis on a number of issues, but these were temporary ones with temporary solutions; crisis did not mean inevitable collapse.