On 11 September, around AD 100, Claudia Severa invited her friend Sulpicia Lepidina to a birthday gathering in northern Roman Britain. The invitation was written in ink on a thin wooden leaf, then folded for delivery. Beneath the formal message is a greeting in another hand, almost certainly Severa’s own. It was thrown away at Vindolanda, a Roman fort in present-day Northumberland, where damp, oxygen-poor soil preserved her letter for nearly two thousand years.
The Vindolanda tablets record the work and correspondence of a Roman frontier community. They include invitations and officers’ letters, but also food accounts, leave requests, lists of clothing, unpaid bills, and reports recording where soldiers were and whether they were well enough to serve. Most belong to the late first and early second centuries AD, before the construction of Hadrian’s Wall. Their writers dealt with immediate problems: arranging visits, obtaining provisions, paying what they owed and keeping the garrison operating.
These documents come from a military site, so the collection chiefly reflects officers, soldiers, their families and those who traded with them. It cannot represent everyone living in Roman Britain. Its value is the detail it preserves about this particular community, including work and relationships that scarcely appear in histories of emperors or on official monuments.
The Fort Before Hadrian’s Wall
Vindolanda was established around AD 85 beside the Stanegate, a route linking military positions between the Tyne and Solway regions. The first timber forts were dismantled and rebuilt several times as units changed and the army reorganized its presence in the region. Most of the important writing-tablet deposits come from these early occupations, particularly around AD 90 to 120. Construction of Hadrian’s Wall, farther north, began around AD 122. The letters record military life on the frontier before the better-known stone barrier was in place.
The garrison included auxiliaries, soldiers recruited in Rome’s provinces rather than the citizen legionaries of a legion. The First Cohort of Tungrians occupied Vindolanda during an earlier phase, while the Ninth Cohort of Batavians appears around AD 100. Both units were named after peoples from the lower Rhine and Meuse region. Their personnel brought different provincial backgrounds and names to this northern posting. Traders, craftspeople, enslaved workers and members of soldiers’ families also lived and worked around the army.
These letters were never intended for a permanent archive. Excavators first recognized ink writing on very thin wooden pieces in 1973, among waterlogged remains from the fort. Later excavations uncovered more, sometimes in rubbish and rebuilding deposits, including pieces that had been discarded or partly burned. Waterlogged ground with little oxygen slowed the decay of wood and ink. It also protected leather footwear, cloth and other organic materials that would usually have disappeared at a Roman site.
A preserved wooden fragment is not necessarily a readable letter. The Roman Inscriptions of Britain edition covers nearly 780 published texts from the early frontier period, whereas British Museum’s materials research describes more than 1,800 physical wooden tablets. Some survive only as fragments, while others are too damaged to yield a complete text. The totals count different things. Further excavations and editions continue to add material. They also come from different archaeological deposits, not from a single untouched library.
How Romans Wrote, Folded and Read the Tablets
The best-known examples are ink letters written on leaves of wood thin enough to fold. Folding could protect the text, while an address might be written on the outside. Many leaves were roughly postcard-sized, though there was no single standard dimension. Vindolanda has also produced wax tablets, which held writing scratched with a pointed stylus into a recessed wax surface. These are different writing methods, and recovering faded ink requires different techniques from deciphering marks on waxed boards.
Some leaves have notches or holes where a tie could help secure a folded message. The writing itself is Old Roman Cursive, a quickly written script very unlike the capitals cut into public monuments. Being able to read printed Latin does not make this handwriting immediately legible. Palaeographers compare letter shapes, familiar formulas, repeated names and surviving ink strokes, marking where they have restored missing characters. Photography and other imaging techniques can reveal faint traces, though damaged passages may still admit several readings.
The Vindolanda Trust’s description of writing materials sets out how the ink leaves differed from waxed writing boards. British Museum conservation research has also used microscopy and other methods to study how the tablets were made, which woods were chosen and what went into the ink. According to the museum, the ink letters used locally available trees and plant-based carbon inks; some of the stylus boards, however, were made from woods not native to Britain. Not every wooden writing surface had to arrive from the Mediterranean. The garrison used local wood alongside material carried to Britain through longer-distance supply routes.
Even a short letter had to be written and carried. Its author or a secretary put the message on wood, folded and addressed it, and passed it to someone who could carry it. The recipient might then keep it, reuse it or throw it away. The route taken by an individual message is usually lost, but the addresses and correspondents show contacts between forts and military households. The tablets reveal something of how communication worked, alongside the information those messages carried.
Claudia Severa’s Birthday Invitation
Tablet 291, the famous birthday invitation, was found in 1985 and belongs archaeologically to approximately AD 97 to 105. Claudia Severa was the wife of Aelius Brocchus, an officer connected with another military post. She asked Sulpicia Lepidina to come for her birthday on 11 September. Lepidina was married to Flavius Cerialis, commander of the Ninth Cohort of Batavians at Vindolanda. Severa also conveyed greetings to Cerialis from herself, Brocchus and their small son. The address on the reverse identifies Lepidina as Cerialis’s wife.
Severa follows the familiar conventions of an invitation: a greeting, the reason for the visit and an expression of hope that Lepidina will attend. Calling her “sister,” Severa signals affection rather than necessarily describing a family relationship. The text supplies 11 September but no year. Editors date it from the archaeological layer in which it was found. The day of the planned celebration is certain, even though its year remains approximate.
The handwriting changes in the final lines. A skilled writer, probably acting for Severa, produced the formal invitation; the closing words were added in another hand. The editors attribute that second hand to Severa, comparing it with her other surviving correspondence, Tablets 292 and 293. It is one of the earliest surviving examples of Latin handwriting that can be linked to a woman by name. The distinction concerns what survives, rather than when women first learned to write Latin.
The exchange places senior officers’ households within the social life of the fort. Severa and Lepidina belonged to households connected with military command, and their relationship involved children, husbands, visits and exchanges of greetings. Their letters moved through the same environment as supply requests and orders. Some women maintained friendships across this frontier community. Yet the position of an officer’s wife was very different from that of an enslaved woman or a farm worker living nearby.
The commander’s residence, or praetorium, combined the functions of a family home and an administrative centre. Letters arrived there, while people in the household arranged provisions, deliveries and contact with other officers’ families. Severa’s invitation illustrates that overlap. Nothing in the surviving text tells us whether Lepidina went to the party, but it establishes that these two households communicated personally at the northern edge of Roman Britain.
Socks, Sandals and the Problem of Keeping People Supplied
Tablet 346 offers a much less complete glimpse of everyday life. Probably written between AD 104 and 120, this damaged letter mentions socks, sandals and underpants, apparently among things sent to someone. It ends with greetings to named people and their messmates. Missing or uncertain letters leave the action and quantities partly unresolved in the editor’s translation. We cannot securely identify the sender or explain exactly why the items were dispatched.
Clothing was a constant practical concern for soldiers on the northern frontier. Equipment meant more than weapons: footwear wore out, garments needed replacing and daily tasks called for suitable clothes. The fragment hints that people obtained things through personal connections as well as official army distribution. Its list of garments and the greetings that follow belong to one message. The parcel carried both material help and the greetings that maintained personal ties.
Excavators have recovered thousands of leather shoes and footwear fragments at Vindolanda, from types worn by men, women and children. The Vindolanda Trust’s collections account explains why this survival of leather is so valuable. Those finds show that the people using the site extended beyond soldiers, though a shoe cannot be matched to the writer of a particular letter. The construction and military use of Roman boots and hobnailed caligae provide further context for the footwear found on the frontier.
A direct request for supplies appears in Tablet 628. Masclus, a decurion or cavalry officer, asked Flavius Cerialis for orders for the following day: were his men to return with their standard as one group or in separate parties? He added that the soldiers had no beer and requested a supply. Masclus was writing within a chain of command and expected instructions from the prefect, with possible help obtaining beer. The tablet does not say whether the prefect sent the beer or issued the requested orders.
Another Masclus letter came to light in the excavations of 2017. Tablet 892 dates to an earlier period, around AD 85 to 92. Writing to the prefect Julius Verecundus, Masclus asks for leave for five soldiers, requests a cutting knife back and refers to plant cuttings he has sent. He originally requested leave for three men, then crossed out names and revised the total to five. The crossed-out names let us follow an officer revising an actual request, down to the men involved, the tool he wanted back and the permissions he sought.
The Strength Report of the First Cohort of Tungrians
Military administration appears particularly clearly in Tablet 154, a strength report for the First Cohort of Tungrians. The archaeological dating puts it around AD 92 to 97, although the heading gives only 18 May, without the year. It names the prefect Julius Verecundus and was written on an unusually large oak leaf. The figures distinguish the cohort’s total complement from the men detached elsewhere, those at base and those temporarily unable to serve.
Of the cohort’s 752 men, 456 were away and 296 were present. Among those at base, 15 were sick, six had wounds and ten suffered from eye inflammation. Subtracting these 31 from the 296 present leaves 265 fit for active service. That number applies to the men at the fort. It tells us nothing about the health of the 456 serving elsewhere, so it cannot be taken as the number of fit soldiers in the entire cohort.
Coria, generally identified with Corbridge, accounted for 337 of the absent men. Another 46 served as guards to the provincial governor, while smaller detachments were posted in other places. The report even seems to place a centurion in London. Parts of the leaf are worn, and the editors warn that some final digits cannot be read confidently. The overall picture remains clear: large portions of a unit could be working well beyond the fort where it was based.
The strength report helps explain why a fort’s size cannot be equated with its usual garrison. Neither the dimensions of a building nor a unit’s nominal complement tell us how many soldiers slept there on any particular night. Some men guarded other locations, accompanied officials, moved supplies or carried out specialist assignments. By separating those present from those absent, sick or injured, the army kept track of how many people were available for duty. Illness and injury reduced the number available for duty even among men stationed at the fort.
The scholarly edition’s commentary discusses what sort of military return this was. Its layout does not fit neatly into the categories of a routine annual record or a daily morning report, and the date and arrangement point instead toward an interim personnel account. We should therefore be careful about using this one exceptional survival to reconstruct a uniform paperwork system for every Roman garrison.
Flavius Cerialis’s letters belong to a later occupation of Vindolanda than the Tungrian report. They concern the Ninth Cohort of Batavians rather than the unit commanded by Verecundus. Combining the letters into one imagined moment would obscure changes in personnel and organization between occupations. Commanders in each phase relied on written communication. The wider subject of ancient military intelligence and communication makes clear why commanders needed reliable information about troops scattered across the frontier.
Octavius, Candidus and the Cost of Frontier Trade
A frontier garrison needed grain, livestock, leather and many other goods. Officials organized some of this, while traders and private dealings supplied other needs. In Tablet 343, written around AD 104 to 120, Octavius complains to Candidus about the strain these transactions placed on him. He needs cash, is anxious about a deposit and has goods waiting to be moved or collected. His account shows how easily a trader’s money could be tied up in goods and unpaid debts.
Octavius reports buying around 5,000 modii of ears of grain. Each modius was a Roman dry measure holding roughly 8.7 litres, making this a large consignment. He needs at least 500 denarii and is worried about losing the approximately 300 denarii already paid as a deposit. Octavius asks Candidus for money at once: without payment, he could lose his investment and damage his standing with the people involved.
Grain was only one concern. Octavius also writes about animal sinews, hides and a wagon. He has left hides at Cataractonium, usually identified as Catterick, rather than bring them over bad roads and risk injuring the draught animals. He disputes an uncredited payment of eight and a half denarii. The state of the roads affected both transport and payment. A sale could stall when roads became difficult, animals were at risk or someone failed to settle an account.
Marinus, Tertius, Fatalis and Frontius also appear in the correspondence, though we cannot always tell exactly how they were connected. Octavius’s use of “brother” for Candidus does not prove they were related by birth; the expression could be social. The text also leaves open whether Octavius traded independently as a civilian or worked within a more involved military supply arrangement. Whatever his formal status, he depended on credit, trusted contacts and correspondence in a sizeable frontier trading network.
Other transactions survive only in fragments. Tablet 182 lists people and prices alongside bacon, pork fat and other goods, and names a brewer, Atrectus. That entry is distinct from Masclus’s request for beer, but both show drink moving through the garrison’s economy. Food purchases and deliveries were recorded alongside names and prices. Many of the people involved are known only because someone discarded a working document. Latin served these ordinary transactions as readily as it served military orders.
The letters reveal several routes by which goods reached Vindolanda: official provisioning, sales by private traders, purchases for households and help obtained through personal contacts. A commander’s household needed supplies, soldiers needed food and clothing, merchants needed payment, and transport required animals and carts. A grain purchase might have been intended for soldiers or for private use; sometimes the text cannot decide the question for us. There was no single, simple supply chain that the documents allow us to reconstruct in full.
Whose Voices Survive in the Letters?
The people named in these letters represent more provincial backgrounds than the label “Roman” alone might suggest. The Tungrians and Batavians were associated with northwestern regions, but Vindolanda’s personnel and detachments were more mixed than those unit names suggest. In Tablet 892, Masclus requests leave for soldiers called Raetians and Vocontians. The editors cannot establish their precise military affiliations. Commanders on the British frontier could be responsible for men whose backgrounds and service connected several provincial communities.
The tablets are written in Latin, though their spelling and grammar often differ from what appears in formal literary texts. These variations give historians evidence for Latin as it was written in everyday transactions. Scholarship on non-literary Latin texts from Britain compares evidence from Vindolanda, London and other collections. These letters cannot tell us how many soldiers were literate, whether Latin was everyone’s first language or how widely the same written habits extended into communities beyond the military settlement.
The local population speaks far less often in this collection. Tablet 164 describes British fighting habits and uses the contemptuous expression Brittunculi, commonly translated “little Britons.” Because the surviving text is damaged and its character is debated, even the immediate context needs care. It gives us a Roman military judgment of the people around them, rather than an impartial account of those communities in their own words.
Women and enslaved people are similarly unevenly represented. Severa’s closing lines show that a named officer’s wife could write at least part of a Latin letter. Leather shoes and other archaeological finds demonstrate that women and children lived at the fort, although few of their own messages survive. Enslaved workers are present in household and commercial life, but we do not hear them on the same terms as officers. This is detailed evidence drawn from a community whose surviving voices were never evenly distributed.
Why the Vindolanda Tablets Survived and What They Cannot Tell Us
The circumstances of burial determine which Vindolanda documents historians can read. A letter thrown into a rebuilding deposit might end up sealed in wet ground, while another written on the same day would rot away. Many recovered tablets are incomplete, with damaged wording and missing letters. The published inscriptions distinguish surviving letters from editorial restorations, allowing readers to see where a reading remains uncertain.
The letters themselves rarely give a complete modern date. Severa named 11 September, but the archaeological setting supplies the range of possible years. The Tungrian strength report similarly records 18 May, not a precise year. Matching names across tablets is useful only when the occupation phases also permit those documents to belong to the same episode. Otherwise two letters could be separated by a change of garrison or several years.
Excavations continue to produce new evidence. Alan Bowman, David Thomas and Roger Tomlin published four letters concerning Julius Verecundus, recovered in 2017, in Britannia in 2019. These letters extended the surviving correspondence of the earlier commander, before Cerialis’s period at Vindolanda. Even a short newly recovered fragment can change the interpretation of a familiar name or a military posting.
The tablets do not give us a continuous history of the frontier or a representative record of its entire population. What they preserve is the daily conduct of identifiable people. Severa arranged a visit, Masclus sought orders and beer, the Tungrian staff counted men away or unable to serve, and Octavius faced losing money on grain while poor roads delayed his goods. Together these letters show how people kept a distant garrison supplied, staffed and connected to other households and military posts.
