In 1762, a peasant working outside the walls of the Roman town of Bavai (France, Nord) encountered a late Roman inhumation grave lined with stone or tile and associated with the tombstone of a military accountant (scrinarius) named Lucinus. With a consular date of a.d. 404, the inscription has particular relevance to the problem of the relationship between the later funerary inscriptions of Roman Britain and the insular series of fifth-century and later memorial stones from Wales, western Britain and southern Scotland. It has been debated whether the latter derive from late Romano-British epigraphic traditions or represent a new fifth-century introduction from Gaul or from the Western Empire generally. Linked to this is the question of whether it is possible to identify a specifically Christian element in late Roman tomb inscriptions in Britain.

As was often the case in the eighteenth century, accounts of the discovery attracted various unlikely embellishments and it may have been these that led a local historian, E. Desjardins, to claim that the inscription was a forgery. Presumably the original stone had disappeared by this time. In this, Desjardins was followed by Edmond Le Blant, who pointed out that the stone named only a single consul. Otto Hirschfeld accordingly relegated it to the category of ‘Inscriptiones Falsae vel Alienae’ in the appropriate volume of the Corpus Inscriptionum Latinarum. Footnote ¹ Recently, however, the inscription has been rehabilitated by Roland Delmaire, who has re-examined the accounts of its discovery and pointed out that the presence of a single consul is confirmation of the stone’s authenticity, since the eastern consul for a.d. 404, Arisaenetus, was not recognised in the West.Footnote ² In addition, it is highly improbable that an eighteenth-century forger would have had the knowledge to concoct an inscription involving an obscure Roman army rank and place it in the appropriate context of a (then unexcavated) fort and store depot.

In the fourth century, the forum complex of Bagacum Nerviorum, the former tribal capital of the Nervii, was converted into a fortified supply-depot on the Amiens–Cologne highway.Footnote ³ Bavai stood at the junction of two military highways, from Amiens to Cologne and from Reims to Boulogne and the Channel crossing. It thus occupied a key position on the strategic routes between Britain and the Rhine frontier and it is hardly surprising, therefore, that a tombstone bearing the title of scrinarius, denoting an official military role for Lucinus, came to light there. Dated to the sixth consulship of Honorius, the stone is one of a series from Gaul using variants on the memorial formula Depositus. These are generally taken to be Christian, on the basis of some of the phrases used. In the case of the Lucinus stone its Christian nature is made explicit by the presence of a chi-rho monogram beneath the text. Examples with consular dates show that the type was current in the second half of the fourth century and the opening decade of the fifth, in the years immediately before the ‘great invasions’. The Bavai text reads:

The closest parallel comes from the cemetery of the church of St-Ausone in Angoulême (Charente), on a square tile placed in the grave behind the head of the corpse. It dates from a.d. 405, the year after the Bavai text:

As Mark Handley has shown, late Roman consular-dated memorial stones are extremely rare north of the Loire. There are none in the three provinces of Gallia Lugdunensis or, other than the Bavai stone, in Belgica Secunda.Footnote ⁵ The Lucinus stone is thus important for the chronology of Christian memorial formulae and for late Roman Christianity in northern Gaul and Britain. It can also help in assessing those late Roman inscriptions in Britain sometimes seen as Christian, or as having ‘Christian symptoms’, and their relationship to the origins of the insular memorial stones of western Britain.

By the time of Constantine, a form of Christian epitaph had emerged, giving, in varying order, the name of the deceased, a verb such as decessit or depositus, the date of death, sometimes the consuls or consul of the year, and often the phrase In Pace. The earliest example from Gaul is a lost stone of a.d. 334 from Lyon.Footnote ⁶ There are others, of a.d. 347 (Valcabrère, Haute-Garonne), 378 (Autun), and 405 (Angoulême), and undated examples from Lyon, Vienne and Valcabrère.Footnote ⁷ Stones with In Pace or Depositus in Pace, as at Bavai, with the dead seen as resting in peace, awaiting the Resurrection, are specifically Christian, as is the inclusion of the day of death, necessary for annual commemoration (for Christians the ‘heavenly birthday’ — dies natalis), though festivals such as the Parentalia or Lemuria were not of course confined to Christians. In the case of the Lucinus stone, one of the latest examples of this group, the Christianity is explicit. A parallel series from Aquitania uses the specifically Christian Vivas In Deo, more commonly found on metalwork and the like.Footnote ⁸

On the basis of consular dates, it would seem that these were replaced in Gaul, after the ‘great invasions’ of a.d. 406, by inscriptions using the formulae Hic Iacet (often, particularly in insular contexts, in the late Roman vulgar Latin form Hic Iacit) or Hic Requiescit. These were the origin of the post-Roman insular memorial stones similarly using Hic Iacit. However, Handley has challenged the generally accepted view that the insular series derives from fifth-century Gaul rather than from fourth-century Britain and argued that Britain should be seen as ‘an integral part of the Late Antique world’, rather than having specific links with Gaul in this period. This might seem a matter of semantics, but Handley’s conclusion that the Gaulish evidence for the early fifth-century introduction of the Hic Iacet formula to Britain is unsound and that specific elements of Christian commemoration can be found on fourth-century Romano-British tombstones has considerable implications, as he points out, for late Roman Britain, for early medieval Western Europe, and in such related fields as palaeography and phonetics.Footnote ⁹

The tombstone of Lucinus, precisely dated and geographically far closer to lowland Britain than other Christian tomb inscriptions from Gaul, can help to provide a new approach to this problem. Our view of fifth-century Britain will be much influenced by whether we visualise a Christian Church in late Roman lowland Britain with a distinctive epigraphic tradition, capable of giving rise to the post-Roman insular memorial stones, or whether the latter derive from fifth-century continental contacts, either by way of the western seaways or via lowland Britain.

Victor Nash Williams thought that the insular series showed direct links between Lyon and Vienne, where dated Hic Iacet inscriptions are known, and the Church of western Britain. However, there are now enough Hic Iacet inscriptions from the geographically closer area of western Gaul to make this hypothesis unnecessary.Footnote ¹⁰ Victorian church historians were attracted by the idea that if the insular Church was descended from that of the Lyon martyrs of a.d. 177, many of whom were from Asia Minor, it did not have to originate from Rome. Nash Williams, a devout Anglican, would no doubt have found such an argument absurd, but it was to be found in a number of influential works on church history and he was probably unaware of the thinking that lay behind this commonly received opinion.

Hic Iacet inscriptions are also common in the Rhineland and north Italy (as Nash Williams pointed out), but the Bavai inscription narrows the gap between the fourth-century series from mainland Europe using Depositus and Hic Iacet and the fifth-century and later insular Hic Iacit stones. The latter are part of a late antique Christian tradition found through much of the Western Empire. The means of transmission was perhaps the historically well attested, but archaeologically elusive, fifth-century Church of southern Britain, which had links with various regions of Gaul.Footnote ¹¹ A parallel might be found with the imports of Mediterranean and Gallic amphorae and fine wares in the same period. Though these do not begin in western Britain until c. a.d. 470, an earlier phase of imports of Palestinian and North African amphorae (Late Roman 4 and Insular B V – ‘Africana Grande’) was reaching urban centres like London, York and Wroxeter earlier, and there is no need to think that the trade was wholly discontinuous.Footnote ¹² An ‘epigraphic habit’ of Christian commemoration could have reached the same cities by the same routes. Bearing in mind the trading links between Britain and the Rhineland from pre-Roman times onwards, there is no real need to deny the Rhineland a share in both introductions. For example, Germanus of Auxerre’s companion to Britain in a.d. 440/5 is said to have been Bishop Severus of Trier, later remembered as the evangelist of Germania Prima.Footnote ¹³

Nash Williams’ statements on the source of the insular series, the source of much later debate, were reasoned and cautious.Footnote ¹⁴ He says nothing of the means of transmission. There is no reference to the ‘Gaulish missionaries’, introduced by subsequent writers, only of ‘close and continuous contact’ between the British and Gaulish Churches and thus with the wider world of Western Christendom. For the earlier part of the century at least, this is a conclusion with which few would quarrel. From the consular dated epitaphs of the Rhône Valley, it is possible to suggest a chronological sequence from the second quarter of the fifth century to the early seventh.Footnote ¹⁵ However, as Handley has pointed out, this cannot be shown to be valid throughout Gaul owing to the rarity of dated epitaphs elsewhere. Stones with Hic Iacet do not carry a dating clause away from Lyon, and even there dated examples are few.Footnote ¹⁶ The question of whether they represent the beginnings of British Christian commemoration also remains fluid. It is in any case unsafe to use such formulae in isolation. Hic Iacet was common throughout medieval and early modern Western Europe. Similarly, though Depositio is characteristic of fourth-century Christian Gallic inscriptions, a stone from Saintes beginning Depositio Bonae Memoriae has an incomplete regnal dating clause of a.d. 525–610.Footnote ¹⁷

FOURTH-CENTURY BRITISH CHRISTIAN TOMBSTONES?

The Bavai memorial stone of Lucinus thus brings us closer than before to a possible early fifth-century phase of Christian commemoration in lowland Britain. There have been several attempts to identify a Christian element among late Roman tomb inscriptions in Britain, making a post-Roman introduction unnecessary. There is no reason why the considerable Christian communities of fourth-century Britain should not have commemorated their dead with inscribed tombstones. What is problematic is whether we can recognise Christian individuals epigraphically in late Roman Britain. Discussion has usually centred on some three or four inscriptions,Footnote ¹⁸ but Mark Handley has suggested that of some 500 Romano-British tombstones of all periods, excluding those of ‘foreigners’ from other provinces of the Empire, eight of eighteen recognisably late Roman examples using memorial formulae such as Dis Manibus et Memoriae, Titulum Posuit or Vixit Annus Plus Minus show ‘Christian symptoms’. These include a number of recently discovered inscriptions.Footnote ¹⁹

The three or four most frequently discussed stones are all from the northern frontier, though this reflects only the stronger ‘epigraphic habit’ in the military zone. In two cases, the ‘Christian symptom’ consists of a single phrase in an otherwise normal ‘pagan’ epitaph (the expression ‘pagan’ is of course a Christian coinage and in this context could be misleading). A woman from Salona buried at Carvoran is described as ‘sine ulla macula’. As R.P. Wright pointed out, the phrase is frequently, though not exclusively, Christian,Footnote ²⁰ but it need mean no more than that she was a virtuous wife. Similarly, the baby daughter of a centurion buried in York was ‘a most innocent soul’ (anim(a)e innocentissim(a)e). This raises interesting questions as to pagan and Christian views of the souls of infants, but in context is a very natural sentiment, and, as Jocelyn Toynbee pointed out, there are many pagan parallels.Footnote ²¹

The well-known tombstone of the Greek Flavius Antigonas Papias from Carlisle has a number of unusual features and is considered ‘probably Christian’ by the editors of RIB I. It uses the phrase Vixit Plus Minus, common on early medieval Christian epitaphs in the Western Empire and sometimes said to show a Christian disdain for the precise span of life. Often, however, it is used of elderly people, where, in the absence of written records or the testimony of surviving relatives and friends, the precise birth date might not have been available. This might well have been the case with a sixty-year-old (or thereabouts) Greek immigrant buried in Carlisle by Latin-speakers rather than as part of any Greek community. Mark Handley has commented on the ‘very high levels of age ignorance’ in Late Antiquity and suggested that ages on epitaphs were not always a means of recording precise age, but were often rounded off in order to assign the deceased to an age cohort, and thus a role within the community.Footnote ²² In any case, until the formula is found in a post-Roman insular context, it cannot be used as evidence for the origins of such traditions. The other inscription using the Plus Minus formula, from Brougham, was considered as possibly Christian by Toynbee, Watson and Thomas. However, the age of the deceased is not known.Footnote ²³

The most common late Roman memorial formula in much of the late Roman West is Dis Manibus et Memoriae. The phrase Dis Manibus ‘to the spirits of the departed’ was a conventional, religiously neutral phrase, used by pagans and Christians alike. There is no need to regard its later variant as specifically Christian. The five examples from Britain occur both in the South, at London, Colchester and Caerleon — major Roman centres where an above average number of non-local people might be expected — and in the North.Footnote ²⁴ Not all belong to the latest period of Roman Britain. That from Catterick was re-used as a door-sill in an early fourth-century bath-building, and is presumably third-century. That from York was later re-used for a secondary plaster burial.Footnote ²⁵ More relevant for ‘Christian symptoms’ are three epitaphs from London and York using the simple Memoriae, perhaps avoiding Dis Manibus, which, though religiously neutral, might have retained some residual pagan connotation for rigorous Christians. The sarcophagus of Valerius Theodorianus from The Mount at York commemorates an Italian from Nomentum, north of Rome, and the same formula appears on the tombstone of Bassaeus Julius, also from York. The sarcophagus of Valeria Amandinus from London, now in Westminster Abbey, has also sometimes been regarded as Christian.Footnote ²⁶

In contrast, the seven British examples of the formula Titulum Posuit are all from the northern frontier zone, with three from Brougham and others from Hadrian’s Wall, Old Penrith, Templeborough and York.Footnote ²⁷ One from Brougham also uses Vixit Plus Minus. Again, some were re-used — that from Templeborough as a drain-cover beneath the wall of the later fort, and that from Hadrian’s Wall cut down for re-use as a hearthstone. They show an almost exclusive concentration on military sites, and two specifically commemorate soldiers. The formula, with its variant forms, is common on the Rhine and Danube and could have been introduced to the northern frontier zone as a result of late Roman troop movements. Some of the soldiers involved might have been Christian, but it is difficult to see a specifically Christian element in the statement that a particular person had erected the tombstone.

It is thus more realistic to look, as has been done in Gaul, for ‘Christian symptoms’ rather than for ‘Christians’, in what would undoubtedly have been a very fluid situation.Footnote ²⁸ It is a commonplace that funerary commemoration of all kinds represents the concerns of survivors rather than the views of the dead person. Possibly an analogous problem in a later period of radical religious change can help put the problem into focus. The wills of Reformation-period Britain have sometimes been scrutinized, in the same manner as late Roman tombstones, for evidence of religious belief. Wills, unlike tombstones, were produced by their subjects, not by surviving relatives. Whilst some will formulae, like references to ‘the Blessed Virgin and all Saints’ or conversely the Calvinist wish ‘to be counted among the Elect’ seem specific enough, the great majority are less so and the latter merely reflects Elizabethan Calvinist orthodoxy. The formulae used may reflect local usage, prevailing local orthodoxy, or the views of the lawyer or priest to whom wills were normally dictated, rather than the beliefs of the testators.Footnote ²⁹ Similarly, late Roman tombstones may reflect a range of concerns and local fashions rather than the religious views of the deceased. It is more realistic to look for ‘Christian symptoms’, rather than attempting to guess the religious beliefs of the individual commemorated.

In addition to formulae, possible continuity between late Roman and post-Roman memorial stones can be approached by way of the script used. Carlo Tedeschi has argued for epigraphic continuity between late Roman and post-Roman Britain on the basis of specific features of late Roman lettering which recur in the post-Roman insular series. He has identified a system of ligatures and of reversed or inverted letters, separate from the similar ‘vernacular’ features found in other Western provinces. He has also argued that any influences between manuscripts and inscriptions on stone were from the latter to the former, not vice versa.Footnote ³⁰ However, of the seven inscriptions whose script is singled out by Tedeschi as ‘specifically British’, four have Imperial or consular dates in the first half of the third century.Footnote ³¹ This may reflect nothing more than the notorious paucity of dated fourth-century inscriptions in Britain, but apart from two inscribed metal plaques from southern England,Footnote ³² all those specifically cited are from Hadrian’s Wall and its vicinity, save for a milestone from north Wales.Footnote ³³ There is only one tombstone that is without any ‘Christian symptoms’ and none of Tedeschi’s inscriptions figure among those sometimes singled out as potentially Christian on other grounds.

A broader approach is that of the late Gifford Charles-Edwards on the palaeography of early medieval insular inscriptions in relation to that of earlier Roman scripts. She emphasised that whether or not the habit of erecting inscribed tombstones survived the end of Roman occupation, literacy in a variety of other media certainly did and that ‘contact with the wider world might have continued, even though the inscriptions show a decline in high epigraphic standards and a growing use of low-grade letter forms’ and a less formal script, as seen for example in the mid- to late sixth-century Springmount Bog tablets from County Antrim.Footnote ³⁴ This conclusion removes any over-simplistic arguments about whether the script of the post-Roman memorial stones was derived from late Roman tombstones, or was a fresh introduction from fifth-century Gaul and prevents possible confusions between the formulae used on the memorial stones and the script in which they are cut.

THE LAST INSCRIPTIONS OF ROMAN BRITAIN?

An inscription from the church of Llanerfyl in Montgomeryshire, 50 km west of Wroxeter, at a river crossing on a Roman road running north from the fort of Caersws, can be regarded either as one of the last inscriptions of Roman Britain, or as one of the earliest of the post-Roman insular series. One of Charles Thomas’s ‘extended latinate’ series, it is, like others of the series, cut horizontally, in normal Roman style, unlike the vertically-cut texts of most insular memorial stones. It retains a number of features found in the fifth-century epigraphy of Gaul and the wider Western Empire, which are soon dropped in the insular series. These include punctuation, the age of the deceased (a thirteen-year-old girl named Rustica), and a final In Pace. Footnote ³⁵ These features, taken together, are characteristic not of the late Roman tombstones considered above, but of the fifth-century Gallic series.

Though the Bavai tombstone of a.d. 404 retains the earlier usage Depositus, Lanerfyl, like another of the ‘extended latinate’ series, from Hayle in Cornwall, uses a more developed variant of the same formula beginning Hic…

There is no trace of any of the earlier Romano-British formulae sometimes thought to show ‘Christian symptoms’.

The Llanerfyl stone is sometimes dated to the third quarter of the fifth century, on the basis of Gallic parallels. This has implications for the date of the beginning of the insular series of inscriptions and hence for epigraphic and other continuity between late and sub-Roman Britain. However, the memorial of Lucinus from Bavai suggests that the two British inscriptions need not be much later than the early to mid-fifth century and is indeed something of a ‘missing link’ between the fourth-century Gallic series and the fifth-century insular stones.

The evidence of another mid-Wales stone can be added. This, from Llywel in Breconshire, discovered in 1954, is a palimpsest, with a horizontal Latin inscription and a secondary Ogam inscription cut over the preceding text. The primary Latin text is badly damaged and incomplete, but seems to be a personal name such as <Lu>cius <An(n)>ius <T>aur<i>anus or <L>aurianus. Roger Tomlin has shown that the short horizontal lines, purely Latin character of the word-endings, and punctuation (there is a stop after the second line) suggest that the inscription is sub-Roman rather than early medieval in character and Charles Thomas has suggested a date ‘not long after the middle of the fifth century’. However, a date in the first half of the century would be possible.Footnote ³⁷

One surprising feature of these ‘extended latinate’ inscriptions is their geographic location in Cornwall and in remote areas of mid-Wales. Two factors might account for this. The two mid-Wales stones are from the area of the early medieval kingdom of Powys. The name Powys is thought to derive from the Latin Pagenses, ‘dwellers in the countryside’ around some Roman centre, presumably Wroxeter, which has itself produced a fifth-century inscribed memorial stone.Footnote ³⁸ A line of transmission from late Roman Britain is thus possible. The other factor, long ago pointed out by Raleigh Radford, is the similarity of many insular memorial stones to late Roman milestones from the same areas. These would have been one of the few forms of inscription familiar to an untravelled British chieftain, by which he could have seen how a Roman Emperor could display his name and status in stone. Any epigraphic continuity between late and post-Roman Britain may therefore owe as much to milestones as to tombstones.