In the past two years British archaeologists have discovered and in some part deciphered more than 240 fragments of 1st century A.D. Roman cursive writing on thin slivers of wood in a far corner of the Imperium where the survival of such material would have been thought most unlikely: in a fort on Hadrian’s Wall.
The material was brought to light in 1973 and 1974 by Robin Birley, Director of Excavations of the Vindolanda Trust, controllers of the fort, a mile or so south of the Wall, and of the civilian settlement (vicus) adjacent to it. The site, known since the 18th century, was first seriously investigated by Eric Birley, father of the present excavator. His son’s work makes Vindolanda one of the most important and exciting digs now under way in Britain.
Of itself, and quite apart from the written fragments, Vindolanda (perhaps meaning “white lawn”) is a remarkable site. More than 18 acres in all, in the Tyne Valley of Northumberland near Hexham, it has revealed no less than five different periods of occupation, from the first epoch of Roman presence in the north of England after the conquest of Claudius, i.e., from A.D. 80 to ca. 125, to their
withdrawal in the 4th or 5th century. Originally a defensive position on the Stanegate, the Roman road crossing Britain at the narrow point running roughly from Newcastle to Carlisle, it predated the great Wall by more than 40 years. It was apparently abandoned during the Hadrianic period, 125 to 160, its garrison probably moving a few miles away to the well-known fort of Housesteads, on the Wall itself. It was reconstituted in the second half of the 2nd century and by the dawn of the 3rd century was rebuilt into a typical cohort garrison, 1,000 or so strong. Abandoned at the end of the 3rd century, it was later replaced by a new fort, A new vicus was superimposed, cruder, irregularly laid out and oblivious of the drainage and street systems of the earlier one.
The excavations of the later, post-Hadrianic, constructions reinforce, with abundant evidence, the falsity of the picture firmly believed until 60 years ago of a lonely, all-male, all-military occupation of the northern frontier. Instead, the soldiers (from all corners of the Empire, but probably mostly Gauls and indigenous recruits) lived with their wives—unofficial but no less conjugal in Hadrian’s time and quite legal thereafter.
The 3rd century vicus at Vindolanda covers some 15 acres, the fort only three. Its civilian population of wives, children, merchants and artisans was perhaps two to three times that of the regiment in residence (probably Cohors III Gallorum throughout). It included a mansio (inn for travelers), married quarters, a bath house and many other buildings and shrines, Within the fort itself, with its typical playing-card shape, were headquarters and office buildings, barracks, stables, storage depots and the house of the commanding officer.
Later-day excavations of the site were only in small scale until 1969 and of substantial extent mostly in the last two seasons. These latter, however, have already produced the finest collection of Roman leather goods in Britain—some very elegant and splendidly preserved women’s slippers from the pre-Hadrianic period, for example—an abundance of cloth in almost perfect condition, and a good many wooden artifacts. The wet soil under a clay packing served, it seems, as a preservative rather than a disintegrator.
The Vindolanda forts and vici are not unique, but nonetheless superb in themselves and valuable extenders of the knowledge of Roman occupation gained from such other forts along the Wall and the Stanegate as Housesteads, Corbridge, Chesters, etc. What is unique is the collection of writing tablets. They are doubly precious in coming from the pre-Hadrianic fort, just outside the later fortifications, under at least two superimposed timber buildings of the later period. They date almost certainly from the earliest period—that between the conquest of northern Britain in the second half of the 1st century and the time when Hadrian’s Wall was begun in 122—the exact time span which remains the least known of all the four centuries of the Roman presence and hence the most needful of clarifying information.
A word first about Roman writing in general: Until the advent of vellum (not common in the Roman world until the 3rd or 4th century) most writing—excluding of course inscriptions on marble and stone—was on papyrus, very little of which could survive except in the dry climate south and east of the Mediterranean. But there was also writing, much less of which has been preserved to us, on wood. Most of this was on tablets, rectangles with a small rim, covered with wax and inscribed with the sharp end of a metal stylus and erased when necessary by the opposite blunt end. There are also occasional examples of writing with pen and ink on the wood itself of an unwaxed tablet. Finally, there was writing with pen and ink on slim wooden sheets, but examples are like papyrus writing, very rare except where the dry climate of Egypt and the Near East preserved them. Several tablets or strips could be bound together in a kind of book by thongs threaded through holes drilled through each piece of wood.
To find a trove of them. even in the fragmentary form that the waterlogged soil of Vindolanda has conserved, is exciting indeed, and the more so since they are almost entirely of the relatively unknown post-Flavian, pre-Hadrianic period in Britain.
The fragments are on sheets from one to two mm. thick, some as small as postage stamps and none, so far, larger than 12 by 4 cm. The wood is extremely fragile and the writing tends to disappear almost at once on exposure. A triumph of the excavators and technicians in several Britial laboratories and universities has been the development of a preservative for the wood and a second triumph, by Miss Alison Rutherford at Newcastle University, has been a technique for bringing out the writing under infrared photography.
A handful of writing tablets and wooden sheets has turned up over the last fifty years, mostly in southern Britain, and it has been possible to read some of them quite successfully. But even with that background, the decipherment of the Vindolanda material has been extraordinarily difficult and the work, like that of the whole excavation itself, is only in its early stages.
The so-called cursive script, presumably the form in common use at the time, ranges at Vindolanda from extremely crude to attempted elegance. Among the problems of reading it: The letters are not nearly so carefully formed as the capitals universally inscribed on stone; sometimes two or more letters are joined in ligatures; several of the letters are virtually indistinguishable, one from another, to modern readers, e.g. an a from an r, a p, e and s from each other; solving one man’s handwriting helps almost not at all in deciphering another’s.
Worst of all, the principal normal aid to comprehension, that is to say the context which gives the reader an idea of what might be expected, is almost totally lacking in such small fragments. Furthermore, there is never any use of capitalization or punctuation. Word division, according to the main code-breakers, A. K. Bowman and J. D. Thomas, “may be bizarre or simply not used at all.” Abbreviations are frequent, often without any indication that they are abbreviations.
A Roman reader, knowing what to expect in a piece of writing, familiar with the standard abbreviations and knowing Latin far better than any man today, would presumably read a sheet with the same facility as a modern quartermaster would read a truncated, jargon-filled memo from a subordinate, the ease depending an the latter’s penmanship or lack of it, Not so the 20th century scholar.
In such circumstances, it is a wonder the decipherers have made any progress at all. But they have found several lists or accountings and some letters that can be made out with considerable certainty. One complete tablet is an accounting of supplies dispensed to the garrison over three consecutive days: barley, garlic, a surprisingly large quantity of wine-73 modii or 1,160 pints—celtic beer, hay, vinegar, brine and pork fat. Some letters of recommendation, common in Roman times, from some person of influence or importance to a garrison commander or officer, have been deciphered, as well as one letter telling of goods the bearer is carrying or perhaps requesting, consisting of shoes, socks and underpants.
Quite as fascinating and still remarkably puzzling is the identification of the room in which the writing tablets were found. The locus is almost certainly a building within the earliest fort, below a three-foot layer of dumped clay on which later constructions were built. Beneath were five feet of occupation material, waterlogged, but sealed by the clay. The combination, curiously, made for preservation of the material.
Sections of walling, of timber and wattle, a large gate lying on its face, heavy oak uprights, and the timber base of what may have been a small catapult or hoist were uncovered in good condition.
The occupants of the building had the unpleasant—and one would have thought un-Roman—habit of putting down a fresh carpeting of bracken, straw and gorse over the dirty floor without removing the rubbish. Layer after layer was put down, one on top of the other. It is now compacted into a mat 18 inches thick which must be cut like peat and hoisted in blocks 8 or 10 feet to the surface and then taken apart, almost strand by strand, to disclose the huge wealth of organic material in it.
There is a scatter of pottery and glass; quantities of oyster shells (expensive luxuries); lost items such as a brooch, two iron keys, ballista bolts, bone gaming counters, a spatula, a bronze lion-headed phalera and iron bolts and rings; a multitude of food bones of pig, sheep and hare and skulls of oxen pierced by round holes suggesting that they were used for target practice; a huge deposit of leather—of more than 2,000 items, including worn-out boots, portions of garments and tents and some leather goods in undamaged condition; tools for working leather and material for treating hides; 60 textile pieces, some in almost loom-fresh condition and two or three still showing a purple stripe which suggests they were from the tunic of the oyster-eating commanding officer.
The straw and bracken floor was suffused with the pupae of stable flies, an estimated one million of them, a finding strongly indicative of urination, in turn often associated with leather tanning. There are also quantities of excrement, some probably human. And in this ugly mixture are the 240 wooden writing tablet fragments, 13 apparently from the wax tablet variety, and 14 stylus pens, only one of which was unserviceable; the rest must have been dropped and lost, not thrown away.
What are we to make of the room in which this motley and noisome assemblage was found? One’s first thought is a garbage heap. Yet it was within walls and the articles such as the undamaged metalware, the usable pens, the good leather and the women’s adornment certainly suggest living quarters. The evidence seems also ample that leather-working and the treatment of hides were carried on. The excreta—and some of the textiles in 4- or 5-inch squares—proclaim a toilet.
The likeliest answer is that it was some sort of workshop-cum-writing room within the commanding officer’s quarters, used by occupants with a very casual attitude about hygiene. The whole find calls into question the notion of the Romans, with their aqueducts scores of miles long to obtain unpolluted water, and their invariable installation of baths, as the original cleanliness-next-to-godliness folk.
Be that as it may, the find of the writing tablets remains all important and suggests even greater discoveries to come. For the soil and water conditions at Vindolanda are not unique along the Wall: they must he duplicated, more or less, at other forts and settlements. Accordingly, future diggers can expect to find other tablets elsewhere equally well preserved, that will throw light on a period and a technique about which all too little is now known.