COMPARATIVE FRONTIER STUDIES

This paper presents a positive view of comparative frontier studies. It explores the environmental and geographical factors which govern military operations, by comparing Roman practice with lam; better documented, methods of controlling territory. It concludes by examining campaigns conducted in Scotland at widely different periods.

In the interpretation of archaeological remains, the use of comparative evidence from different periods, which have little or nothing to do with the one under consideration, may be deemed a rather useless exercise. Readers of this persuasion will no doubt pass straight on to the next paper, with only a cursory glance at the title of this one. A conviction that comparative evidence has its uses is no real recommendation for its application; it was Goethe who pointed out that talking or writing about anything is a waste of time, since all that usually happens is that opponents of one's views go away confirmed in their opinion, while adherents adhere all the more. Everyone is entitled to an opinion, but this admirably liberal view should be tempered by the fact that opinions can differ so radically even when based on exactly the same set of data. Archaeology is not the most accurate of interpretative tools, and the auxiliary information which comparative evidence can provide ought at least to be taken into account.

ANTHROPOLOGY AS COMPARATIVE EVIDENCE----
The anthropological approach to archaeological interpretation has long been valid for prehistoric archaeologists (or more accurately, archaeologists who study prehistory). This is a form of comparative evidence which accords well with the study of societies whose only legacy is artefactual, with no contemporary documentary sources to enlighten or confuse the researcher. Hence there are books with titles such as A. Mitchell's The Past in the Present, published in 1880, and Ian Hodder's The Present Past published a century later in 1982. The latter must contain something useful, judging by the number of times it has been stolen from our departmental library. In the hundred years which separate these two works, all that has changed is the attitude to ancient and modern autochthonous peoples. Our imperialist ancestors did not hesitate to apply the label 'savages' to such peoples, and the use of such terminology in the following quotation uncomfortably jars the modern mind. These words written in 1880 should be heard by all those who hold an implicit belief in the theory of the ascent of man (Mitchell 1880,214): We can scarcely hesitate to conclude that civilisations are lost as well as gained; that all existing savages possibly are, and some of them certainly are, in a state of civilisation below that which their ancestors occupied; and that there is no intrinsic tendency in human societies separately to pass ever on and ever up to something better, and higher, and nobler.
Comparative evidence from different periods of history has some validity, if only because people remain people; only the outward trappings change. It is these outward trappings which lead successive groups of scholars to reinterpret archaeological evidence in accordance with each individual's background and experience, so that historical and archaeological received opinion follows trends, and acquires a dated look with the passage of time. This is in itself a form of anthropological approach to the study of archaeology. Given a completely unreferenced selection of archaeological writings, would it be possible to date them and guess their country of origin by analysing the ideas expressed in them? Probably. COMPARATIVE MlllTARY STUDIES----In archaeological interpretation there is no single correct solution. Several concurrent hypotheses are all equally valid in most cases, simply because in reconstructing the minutiae of human life in times past, we do not have all the necessary information from which to make any judgement. This is where a theoretical approach can benefit the study of the archaeology of the Roman Empire, provided that the word theoretical is constantly used and understood. Rigidity of interpretation is perhaps much more harmful than the less certain ground of flexibility. Archaeologists should perhaps bear in mind that, on occasion, elements of doubt can be quite healthy, indeed vital, as demonstrated by this en de coeur from someone not perhaps noted for flexibility, Oliver Cromwell: 'I beseech you, in the bowels of Christ, think it possible you may be mistaken.' I have found comparative evidence most useful in military studies. In the social sphere it is perhaps not so easliy applied, since some allowance should be made for customs and social mores which are not necessarily subject to a set of predictable rules. Armies, on the other hand, share the common need to protect themselves, by a limited number of possible means, while on the march or on campaign, or while permanently encamped within their fortifications . They need to supply themselves with food and equipment and to recruit new soldiers when the existing ones are killed or die or simply run away. When they watch over a newly conquered area, their dispositions are surely dictated by geography, and useful comparisons can be made on this basis, in examining how other armies, operating at other periods in similar or the same terrain, set about achieving their goals.
It is only in recent times that there have been any far reaching changes in transport. Until the nineteenth century, armies moved on foot, or rode on horseback or in wagons pulled by horses. Their problems were possibly not all that different from those faced by the Romans. As far as I am aware, there has been no change in the design of human legs, and as Marshal Saxe succinctly phrased it 'all the mystery of manoeuvres and combats is in the legs, and it is to the legs that we should apply ourselves' (Watson 1988, 55).
In the study of Roman temporary camps, it is useful to have some idea of how far troops can be expected to march in a day, while still having the wherewithal to build fortifications, however slight, at the end of that day. The departure of thousands of men from one location and their arrival at another location some distance away, is not something that can be left to chance, and information on how military authorites organised this procedure can be useful. A study of the armies of more recent historical periods is facilitated by the existence of better documentation than anything which is available for study of the Roman army. Such documentation provides practical information on how many miles of road will be taken up by numbers of marching troops, depending on how many men march abreast, the type of terrain to be crossed, and the distance to be left between the various units on the march. It is possible to estimate the time it would take for a given body of troops to cover a given amount of ground, and to take into account the variable factors which influence these calculations, which many archaeological models ignore.
This practical information can be applied to the study of lines of Roman temporary camps of roughly the same size and shape in north Britain. The usual method of study is to measure the distances between such camps, and on the basis of a theoretical day's march, to predict where the missing camps in the series will be. A brief reading of any of the many surviving diaries left by soldiers serving in armies of different periods will suffice to show that armies cannot operate in such a mathematical way, so that measuring distances can be only a very rough guide at best. Many factors influence the way in which an army marches, such as the constant presence of the enemy, or even more simply, the weather. A day's march can range from over twenty five miles to less than six. Tacitus (Agricola 22) mentions that during one campaigning season, the Roman army was hindered by bad weather. He is not more specific than that, but without his statement there would be not the least hint of any hindrance at all. Archaeology cannot demonstrate such imponderables. Comparative evidence can.
The study of how armies marched could throw some light on troop movements in the Roman Empire, for example from the Rhine to the Danube in times of crisis. It seems that legions were sometimes shunted along from place to place when major wars broke out, for example legionary detachments from Britain seem to have replaced Rhine troops, which in tum marched to the Danube. It is interesting to compare Marlborough's march over the same ground in 1704. From the Low Countries, he progressed steadily but rapidly and arrived at his destination with 40,000 men, with their horses and equipment in good condition, in less than a month (Chandler 1989, 129-30). The march was impressive and exceptional, but it shows what could be done, and demonstrates how rapidly Roman troops may have arrived at threatened points when wars were being waged. Another exceptional march was made by Napoleon in 1805, when he moved troops from Boulogne to Ulm before the enemy had properly mobilised. He used every available means of transport, but he moved nearly five times the number of troops in Marlborough's army, with all the attendant logistical problems of supplying that number of men (Chandler 1966,385).
Comparative study of logistical problems is illuminating. Practical investigations into how much a large army eats, what it eats, how such food is manufactured, where and when it is manufactured, and how it is stored and transported may not always answer similar questions raised by the study of the Roman army, but can help researchers to arrive at a closer understanding of the problem. The restrictions of seasonality on food supply scarcely apply to the inhabitants of the modern western world, but they were of paramount importance to military planning throughout history.

COMPARATIVE FRONTIER STUDIES ----
In the USA, comparative frontier studies have been in vogue for some time, but this approach has only recently been adopted in Roman frontier studies. There is much valuable food for thought in such an approach. Frontiers can be delineated in many forms, ranging from an open stretch of territory patrolled occasionally by a few soldiers, to a continuous running barrier such as Hadrian's Wall. If comparative studies have no other purpose, then they do at least highlight the need for a precise definition of terms.
Not only is there better documentation on the structure of more recent frontiers, and more reliable evidence for their appearance, but there is also more evidence of the purpose behind them and the way in which they worked. Much ink has been spilled on the functions of Roman frontiers, but archaeology alone cannot elucidate purely abstract concepts such as purpose and method of operation. The Russian cherta lines pictured in Christopher Duffy's Siege Warfare (1979, look remarkably similar to the currently accepted reconstructions of the German limes (a convenient word which we are not supposed to use any more). A ditch accompanied by a timber palisade and a string of watchtowers is a very basic form of frontier defence, so it is perhaps not surprising that seventeenth century Russia and second century Rome should employ similar methods to protect themselves from tribesmen living beyond their own lands. The Russian lines are not known in their entirety, but much more is known about them than about the Roman system. For instance, accurate numbers of the troops involved are recorded, and it is known that defence of particular sectors of the Russian frontier was under a single united command, and that there were periods when the lines were temporarily abandoned because troops were called away to deal with crises elsewhere.

So-called Severan Camps
Edward I's 1303 campaign Students of the Antonine Wall please note. To turn to the Roman period in Scotland, it would seem that geographical considerations affect the way in which invasions of this country were planned and implemented. William I for example did not even try to subdue it totally. He reached the Tay, and after Malcolm Canmore submitted to him, or became his man, in feudal terms, William marched back south. The turning point is interesting. Agricola reached the Tay in his third campaigning season, and for a year he advanced no further, either because he was ordered not to do so by the Emperor or because he thought he needed to consolidate the territory behind him. The latter consideration is worthy of note. Agricola had a larger army than William, but it is significant that both military leaders recognised the point beyond which it was not safe to advance unless the lands behind were secure.
The Roman Emperor Severus, on the other hand, probably did advance almost immediately past the Tay into northern Scotland. It is by no means certain that the camps which are labelled Severan do in fact belong to his campaigns, but Dio says that the Emperor penetrated northern Scotland, and he presumably built camps on the way there and back. He did not occupy the territory he overran, even though he campaigned for three seasons before his death and the abandonment of Scotland by his sons. His campaigns may be constructively compa.red with those of Edward I. The source material for Edward's itineraries derives from contemporary records, brought together and published by Gough (1900), with the result that we are much better informed about where Edward went than we are about Severus's movements, but nonetheless it seems that there are similarities in their choice of routes (Figures 23-24). Both monarchs skirted the edge of the Highlands without actually penetrating them, using the more hospitable eastern lands in which to move their armies, probably provisioned by sea. I should add that Edward is not new: Gordon Maxwell invented Edward in a most useful paper delivered in 1983 at Aalen (1986).
There are those who deny that the Romans ever penetrated the Highlands of Scotland, using the analogy of Severns' campaigns and those of Edward I to prove that armies found it far easier to stick to the lowland routes, and that the mountains looked forbidding, and would hinder movement. Given that the Carpathians and the Atlas range did not deter Roman troops, what claim can the Highlands put forward for such superiority? Figures 25 and 26 show Monck's Highland campaign in 1654, and Cromwell's garrisons in the Highlands, revealing that it is possible to maintain troops in the mountains, and to emerge successfully after a series of rapid marches. I believe that the first century Roman forts at the mouths of the Highland glens were not simply guard posts to block exits from the valleys, nor were they part of a linear frontier, but their presence in such locations indicates that troops did go into the mountains. The essential difference between the campaigns of Monck and possibly Agricola on the one hand, and those of Severns and Edward I on the other, is that the last two never succeeded in subduing Scotland and never held it down, despite the fact that they fought battles and won. The campaigns of Monck, and of Agricola, wherever he fought, were successful and longer lasting than anything achieved by a march round the edges of the mountains.
If what I have said seems tenuous in the extreme, because there is no direct archaeological evidence with which to back up the theory, then at least it should have demonstrated that answers to archaeological problems are not single and immutable, and in many cases, what is visible on the ground is not, by itself, enough to construct any hypothesis.