The Heroic Age, Issue 2, Autumn/Winter 1999

In a previous paper (Malcor 1999),¹ I explored what we know about the life of Lucius Artorius Castus, a Roman officer who lived in the late second century. In this article I intend to show how certain details of Castus's military career in Britain correspond to details in the traditional biography of the legendary King Arthur.² Let me start by taking a look at what we are told about Arthur's military career in Britain.

The Early Arthurian Legends

The primary texts for the legend of Arthur include the Historia Brittonum (ca. 800 of the Common Era [C.E.]; attributed to Nennius), the Annales Cambriae (ca. 960-980 C.E.), William of Malmesbury's Gesta Regum Anglorum (ca. 1125 C.E.), Geoffrey of Monmouth's Historia Regum Britanniae (ca. 1136 C.E.), Wace's Roman de Brut (1155 C.E.), Giraldus Cambrensis's De Principis Instructione (ca. 1195 C.E.), and Layamon's Brut (ca. 1205 C.E.). In addition to these texts, other chronicles, a considerable amount of poetry, massive cycles of romance literature, and some ballads from a variety of cultures give varying accounts of the life of the legendary British king.³

There are few details that all of the Arthurian texts have in common, but several of the manuscripts give biographical details that are generally consistent with each other. The texts tell of a man named Arthur who was a warrior in Britain at some time in the past. The earliest accounts say that he was a soldier, not a king, and that he held the title "dux bellorum" ("military commander during wars").⁴ Often he is associated with Roman legions, either as an ally of the legions or as the enemy of legions fielded by a Roman emperor. The stories often give a sense that this man was the historical leader of a group of armored horsemen who fought using swords, lances and shields. Arthur commanded his troops from a fortress, which is variously identified as Caerleon, Camelot, and other locations. When mentioned, Arthur's standard is usually said to be the "Pendragon." He fought battles in Britain, defending the "civilized" areas of the island against "barbarian" invaders, who are sometimes identified as the Saxons, though most of the texts quickly abandon tales of Arthur's battles against the Saxons in favor of tales of Arthur's battles against Picts and Scots or Irish.⁵ Regardless of whom Arthur fought against, he consolidated Britain, which suffered from infighting among various groups following the death of a previous leader, who is usually said to be Ambrosius Aurelianus or Arthur's father, Uther. Following this consolidation, Arthur made two invasions of the Continent: one to aid rulers in Brittany (e.g., Geoffrey 9.11-12; Thorpe 1966:222-228) and one that resulted from a civil war between Arthur and one of his high-ranking warriors (e.g., Malory 20.10; Cowen 1969, 2:497-504). Several texts also credit Arthur with sending an embassy to the emperor in Rome, who is usually identified as "Lucius."

When studying the primary Arthurian texts, great care must be used. As von Sydow (Dundes 1965:233) pointed out, the earliest known variants of a traditional story are seldom either the most complete or the best. Tales may be far older than the manuscripts in which they appear (von Sydow in Dundes 1965:240). Unlike Märchen,⁶ legends, in the early stages of their transmission, are generally interjected into discussions to prove a point (Ward 1981, 2:373). Legends are told as if they are factual accounts (Ward 1981, 2:374), whether or not the events recounted actually happened. As part of the verisimilitude, legends are attached to places or people familiar to the audience (the Brothers Grimm in Ward 1981, 1:1). The importance of this familiarity is underscored by the fact that, also unlike Märchen, legends do not transmit easily beyond the region of their creation (the Brothers Grimm in Ward 1981, 1:1). Such transmission, however, while uncommon, does happen from time to time (von Sydow in Dundes 1965:222). The Arthurian legends are an example of this type of transmission.

At first glance, the Arthurian legends seem to present an amalgam of various traditions that have had their details updated to match the twelfth- and thirteenth-century medieval world of the Romance authors. Few, if any, of the stories' original details appear to remain. On closer inspection, however, some of those details match what we know about second-century Britain just as well if not better than they fit the Britain of either the fifth or the twelfth century.⁷

The Arthurian legends follow a development pattern that is well known to scholars of British legend.⁸

Two of the three points, in reverse order, are readily discernible. The third point is shown by the burgeoning of the number and variation of Arthurian texts in the twelfth and thirteenth centuries is beyond dispute, and many of these tales were attracted to the Arthurian cycle after having an independent existence elsewhere.⁹ The development of the core material for the major Arthurian romances can be traced through various sources back to the oldest surviving reference in Gildas's De Excidio et Conquestu Britanniae (ca. 540 C.E.).¹⁰ I have already discussed the development of the Arthurian legends with C. Scott Littleton in From Scythia to Camelot (2000), and I will take another look at those patterns in a subsequent essay. For now, the important point is that what emerged as the twelfth- and thirteenth-century romances was a full-blown tradition, complete with historical elements and folktales incorporated from a variety of sources. This being the case, the sheer number of parallels, often in chronological order, between details in Arthur's fictional biography and details from Castus's reconstructed biography become a compelling argument for an account of Castus's career serving as the seed for Arthur's military exploits. I'll examine some of these parallels in a moment. At this point, however, let us turn to the second stage in the pattern of transmission that I described above, the use of the hero's name.

The Name "Arthur"

While some scholars over the last century have tried to derive the name "Arthur" from Celtic sources, such attempts at etymology have yielded unsatisfactory results.¹¹ Zimmer (1890:785 ff.) was the first scholar to propose that the name "Arthur" actually derived from the Roman gens nomen "Artorius," and many modern scholars have followed his lead.¹² Although most scholars claim that the name "Arthur" is unattested in Britain prior to the late-sixth century, there was one notable exception: Lucius Artorius Castus, who lived and fought in Britain in the late second-century.

Prior to the writing of the Historia Brittonum, the name "Arthur" started cropping up among late-sixth-century and seventh-century Irish immigrants to Wales and Scotland (Green 1999; Ziegler 1999). Padel (1994:24) suggested that the reason the name Arthur did not appear in Britain prior to the use by the Irish was because the name was regarded "with exceptional awe" by the Britons, while the Irish "when they came into contact with the folklore as a result of their settlements in western Britain, need not have felt such reverence or reluctance" and so had no taboo against the use of the name. This hypothesis certainly fits the pattern for legend transmission as we know it in Britain.¹³ That so many people would suddenly start naming sons "Arthur" and making comments such as "although he was no Arthur" (the infamous reference to Arthur in Y Gododdin, ca. 600), indicates that the cycle of legend was no longer in its simple formative stage but rather at a point that the core stories were so well known that Arthur's name had become proverbial in its usage.¹⁴

Yet the name "Arthur" is not the only onomastic parallel between Arthur and Castus. The Historia Brittonum (ca. 800),¹⁵ which was probably compiled by, rather than written by, Nennius, has the dubious honor of being the oldest work to record legends of Arthur.¹⁶ As such, this text is important for establishing the development of a legend at a certain place and at a certain point in time. By the beginning of the ninth century, Arthur was known as both a dux bellorum and a miles ("soldier"), although the legends of Arthur being a king were apparently already in circulation, given the care with which Nennius points out that Arthur was not a king.¹⁷ The two known inscriptions that detail the career of Lucius Artorius Castus place him firmly in Britain as a soldier who at one point in his career held the title "dux" (See Plates 1 and 2):

While Arthur's title dux was eventually supplanted by rex ("king"), dux was clearly his original rank, and the development from dux to rex follows a direct line of transmission.¹⁹

The Battle List

Drawing connections from names alone is a risky prospect at best. But names backed up by story parallels are another matter. When story parallels are present, the similarity of names becomes an important detail, often pointing to the source for the legends. When the parallels match the biography of the figure whose name precedes the gap in usage, that predecessor usually turns out to be the catalyst for the formation of the cycle of legends. Given that the correlation between the names is backed up by over a score of parallels between Castus's actual biography and Arthur's fictional biography, the odds of there being no connection between the historical figure and the legendary hero grow increasingly slim. To see just how extensive the list of parallels is, let's take a closer look at one of the most famous sequences in the biography of Arthur, the battle list from the Historia Brittonum, in juxtaposition to the biography of Lucius Artorius Castus.

Perhaps the most infamous passage in the Historia Brittonum is a list of twelve battles that Nennius says that Arthur fought against the Saxons. This battle list has always posed problems for Arthurian scholars who have attempted to identify the sites that are named in the account of Arthur's military career (Jackson 1945:44). There is absolutely no guarantee that the battles belong together, though independent sources do connect some of these battles with a leader named "Arthur."²⁰ The list (Nennius 56; Morris 1980:55) consists of:

Many of these sites do not match well to locations in southern Britain. Kenneth Jackson's linguistic analysis of the identifications (Jackson 1945:44-57 and 1949-1950: 48-49) produces the following list of likely names:

One thing that can be said with confidence is that the list is certainly not chronological. The end rhymes of the battle names indicate that Nennius took the list from a rhyming-poem (Morris 1980:5; cf. Field 1999). No one selects battle sites so that their names will rhyme when a subsequent poet talks about the great deeds of the commander. Nennius could well have taken names from unrelated sources to fill in a gap in an existing source. Or the poem may have actually said that the commander was named Arthur, and the poem may even have given this Arthur the rank of dux. While such a combination of "leader with battles with title" has not been identified for a fifth- or sixth-century figure, the second-century dux Lucius Artorius Castus most likely fought a series of battles that does happen to fit the battle list.

The Caledonian Invasion

At the time of the 183-185 C.E. Caledonian invasion, Castus was praefect of fort Bremetennacum (Ribchester in Lancashire).²⁶ The Romans had regarrisoned many of the forts between Hadrian's Wall and the Antonine Wall, although they had not regarrisoned the Antonine Wall itself. In 183 the Caledonii breached the Antonine Wall and flooded south, across territory held by the Dumnonii, Selgovae and Votadini. There is evidence at this time period of war-related destruction at both Blatobulgium (Birrens)²⁷ and Burnswark,²⁸ near the western end of Hadrian's Wall (Salway 1981:211 and 1993:154, 156) as well as at other forts both on and just south of Hadrian's Wall. The western end of the Wall, however, held, as shown by an inscription at Luguualium (Carlisle), which tells how a cavalry regiment slaughtered a "band of barbarians" ("manu barbarorum").²⁹

The east was a different matter. In 183 the Roman forts--Trimontium,³⁰ Habitancum,³¹ Cappuck,³² and Bremenium³³--collapsed along the road that became known as Dere Street (Salway 1993:157). Cassius Dio (73.6; Cary 1932:89) said that in 184 C.E. the invaders breached Hadrian's Wall itself, perhaps still following Dere Street.³⁴ They mostly by-passed the territory controlled by the Carvetii and attacked Eboracum (York),³⁵ where they killed a Roman "general."³⁶ Between this incident and the arrival of Ulpius Marcellus as governor of Britain (Salway 1993:157), something happened that turned the tables for the Romans. By the time Marcellus arrived, the fighting was no longer south of Hadrian's Wall but in southern Scotland, north of the Forth-Clyde Isthmus, and he ordered the successful commander to pursue a punitive campaign against the invaders, that is, if possible, to exterminate them all (Salway 1981:211 and 1993:157).³⁷ In 185 Lucius Artorius Castus was promoted to the rank of dux, a position that very few equestrians held prior to the time of Diocletian. These known details plus the pattern of destruction from 183 to 185,³⁸ attested by archaeological finds, suggests that Castus was the victorious commander who reversed the disaster for the Romans, and the troops he used were the Sarmatian cavalry of Bremetennacum (Ribchester).³⁹ Consider the following scenario, which is based on the battle list Nennius ascribed to Arthur:

After sacking York, the Caledonii continued south through the territory controlled by the Brigantes. The invaders used the Roman roads, turning west at the first major fork and crossing the Pennines. The fort Olicana (Ilkey) lay along the most likely road and was "restored" sometime between February or June 197 through May 2, 198, which suggests that the invading Caledonian invasion did indeed follow the route from Eboracum past Olicana.⁴⁰ That road led them straight to the crossroad guarded by Bremetennacum (Ribchester). Nennius says that one of Arthur's battles happened at Mount/Hill/Rock Breguoin, which can be identified as Bremetennacum (Ribchester).⁴¹

Castus and his Sarmatians defeated the Caledonii at Bremetennacum (Ribchester; Nennius's Mount/Hill/Rock Breguoin) and pursued them west along the river Ribble to the tidal estuary (Nennius's "Strands" of the river Tribruit).⁴² At the estuary near where the Douglas and the Dow join the tidal portion of the Ribble, another unit of Sarmatians defeated the Caledonii again.⁴³ The Caledonii then fled south along the river Douglas. The four battles fought on the river Dubglas in Lancashire would have been within the region controlled by Bremetennacum (Ribchester).⁴⁴ So far the Sarmatians were precisely the troops who should have been responding to an invasion in this region. Then something happened that caused them to leave their normal area of operation. Castus chased the Caledonii along the river Douglas, then angled them back across the Pennines toward Eboracum.⁴⁵ Perhaps Castus continued his pursuit beyond his boundaries because of the collapse of the VI Victrix. Perhaps the acting-governor, M. Antius Crescens Calpurnianus, ordered him to continue since he alone seemed to be having success in routing the Caledonii. Whatever the case, continue Castus did.

The Caledonii turned to face the Sarmatians at York (Nennius's "city of the legions"; Field 1999).⁴⁶ Castus won again, and the Caledonii retreated along Dere Street with Castus in pursuit.⁴⁷ At fort Vinovium (in Durham near Binchester), Castus engaged the Caledonii again, giving rise to Nennius's battle at castle Guinnion (e.g., Crawford 1935:287).⁴⁸

The Caledonii retreated beyond Hadrian's Wall this time, but they left Dere Street and headed for a site near Flodden, northwest of Wooler in Northumberland, that is renowned for battles in that region. The site happened to be on the river Glen, which corresponds to Nennius's river Glein.⁴⁹

By now the remnants of the Caledonian forces were in a panic and intent simply on returning to their homeland north of the Antonine Wall. Castus pursued, possibly now under orders from Ulpius Marcellus to exterminate all of the invaders.⁵⁰ The Sarmatians caught the Caledonii in the Cat Coit Celidon, Caledonian forest, and defeated them again. The few Caledonii who survived, fled for home. During this flight they may have turned to fight another battle at a river known as Bassas.⁵¹

In any event, the last battle was at a place that became associated with Mount/Hill/Rock Badon in later traditions, though the original battle possibly took place near Dumbarton Rock in Strathclyde.⁵² Though Dumbarton was a major political center by the time Bede wrote his Historia Ecclesiastica Gentis Anglorum (Snyder 1997), to date no evidence has been found to suggest occupation in Castus's time.⁵³ In Geoffrey of Monmouth's Historia Regum Britanniae (9.6; Thorpe 1966:218-19), Dumbarton Rock appears as the place where Arthur killed the Scots and Picts by the "thousands,"⁵⁴ "treat[ing] them with unparalleled severity, sparing no one who fell into his hands."⁵⁵ This scenario of Arthur's definitive battle against the Picts and Scots fits quite well with what we can surmise happened in the second century.⁵⁶ Sometime early in 185 the punitive campaign against the Caledonii came to an end. Given Marcellus's orders, this could only have happened if all of the invaders had been exterminated or if the remnants of the invading force had been driven north of the Antonine Wall in such a state that they were not worth pursuing.⁵⁷ We do know that Castus returned to York, was appointed dux and found himself leading a force of cavalry units to Armorica by the summer of 185.⁵⁸ If Castus was the Roman commander who pursued the Caledonii into Scotland, he would have fought his final punitive battle against them somewhere in the vicinity of the Antonine early in 185. Coins confirm that a victory of this nature did take place in 185 (Salway 1993:157), and the position of Dumbarton Rock on the border of the territory held by Rome and the territory held by the Caledonii make it as likely a place as any for the battle to have occurred.

In Geoffrey's Historia, he has Arthur fight battles against the invaders at Eboracum (York) twice, on the Duglas (river Douglas), Lincoln in Lindsey, the Caledon Wood, Bath,⁵⁹ and at "Alclud" (Dumbarton Rock; Wright 1984:101-106). This is roughly another description of what happened in the second-century, with the Caledonii killing the governor at York, continuing to raid as far as the river Douglas in Lancashire, then being driven back north of the Antonine.

Castus's campaign, like Arthur's, consolidated Britain following a period of infighting (in Castus's case, this involved the mutinee of the VI Victrix), and Castus's connection with the VI Victrix might be what prevented the legion from being disbanded after attempting to name "Priscus, a lieutenant," as emperor (Cassius Dio 73.9.2a; Cary 1932:89). In any case, the victorious campaign conducted by Castus with the Sarmatians of Bremetennacum (Ribchester) preserved Britain for the Romans and certainly sky-rocketed Castus's career into realms rarely seen by an equestrian of his day.

Additional Parallels

Following Arthur's Scottish campaign, Geoffrey had the king return to York (Thorpe 1966:220), then proceed to campaign abroad, including a protracted campaign against the Romans on the Continent (Thorpe 1966:223 ff.). This pattern precisely describes Castus's movements following the Scottish campaign: He returned to York and then sailed to Armorica with his cavalry to put down a rebellion there that was being led by ex-Roman soldiers (Malcor 1999). I will explore these and other parallels between Arthur and Castus in more depth in another paper. For now, let us return to an examination of Arthur's campaigns in Britain.

One of the most distinctive aspects of the Arthurian legends is the appearance of the knights. Arthur's men were armored cavalry who fought with swords, lances and shields. While in Britain, Castus commanded Sarmatian numerii from fort Bremetennacum (Ribchester; Malcor 1999). Numerii were auxiliary troops that were attached to the legions rather than actually part of the legions.⁶⁰ These Sarmatians were horsemen who wore scale mail and who fought using swords, lances and shields (Nickel in Lacy et al. 1996:12-15).⁶¹

Roman military units had their own standards in addition to the standards of their legions, particularly auxiliary units like the Sarmatians.⁶² The standard used by the Sarmatians happened to be a bronze dragonhead with a windsock banner attached to it (Nickel in Lacy et al. 1996:13; Dixon and Southern 1992:60-61, fig. 29, plate 10).⁶³ Such banners would have been used by Castus's Sarmatians when they rode into battle, and victories won under the dragonhead banners could have given rise to the connection between the name "Artorius" and the cognomen "Pendragon" (Littleton and Malcor 2000:101).

In referring to Britain, Gildas uses the word "regio" to describe an area of military command that crosses provinces (Higham 1994:94). In Gildas's experience, Britain was divided into multiple provinces by the Romans, something that happened ca. 295 under Diocletian. Gildas seems to be unaware that Britain was ever a single province (Higham 1994:94). What is interesting about this choice of word is that it matches the very strange title for the command at Bremetennacum (Ribchester): praepositus numeri et regionis ("commander of the troops of foreign origin and of the region"; Collingwood and Wright 1965:194-197, nos. 583, 587). Castus's inscription says that he was a praefectus while in Britain, the title that the commander of the fort at Bremetennacum (Ribchester) would be expected to have. During the invasion, however, Castus's duties expanded considerably as the VI Legio Victrix collapsed and mutinied.⁶⁴ Castus took on the defense of Britain, campaigning well beyond the boundaries of his fort. So regio in the sense that Gildas uses it would be a perfect description of what Castus was defending in the 180s. It is possible that the odd description of the commander's rank at Bremetennacum (Ribchester) developed in response to Castus's protracted campaign and then persisted through administrative inertia.

Gildas's references to several governors ("rectores") of Britain have puzzled many scholars, some of whom have speculated that Gildas was talking about a governor for each of the five provinces in Britain with which he was familiar (Higham 1994:151). It is possible, however, that Gildas was working from a source that mentioned multiple governors being involved in the invasion of Britain by the Picts and Scots. Historically, there were three governors involved in the second-century Caledonian assault on Britain: the one who was initially killed at York, M. Antius Crescens Calpurnianus who served as temporary governor until the permanent replacement could arrive from Rome, and the governor, Ulpius Marcellus, under whose orders Castus carried out the punitive campaign in southern Scotland (Salway 1993:157).

A Brief Note on Camlann

While Camlann, Arthur's last battle, is not part of the battle list, it was a famous battle attributed to Arthur that was fought in Britain. The battle of Camlann is first mentioned in the Annales Cambriae (ca. 960-980; Ashe in Lacy et al. 1996:8-9). The name Camlann most likely derives from that of fort Camboglanna (Castlesteads) on Hadrian's Wall (Jackson 1945:56).⁶⁵ This was one of several Hadrian's Wall forts that saw heavy fighting during the Caledonian invasion.⁶⁶ Perhaps the heavy fighting and defeat at this fort combined with legends of Castus's campaign against the Caledonii. Perhaps a later commander named Arthur did die at Camboglanna in the sixth century. Whatever happened at Camboglanna in the second century, though, it was not Castus's death. Castus went on to have a long and prosperous career after he fought his battles in Britain, a career that continued to involve the Sarmatians of Bremetennacum (Ribchester) and that eventually led him into combat against them as they took opposite sides in a Civil War in a battle that does have strong parallels to Arthur's battle at Camlann (Malcor 1999).

Conclusions

While the biography of Lucius Artorius Castus cannot account for every detail in the Arthurian legends, the parallels between Castus and Arthur are striking not only in their number but also in the variety of levels on which they occur. Castus meets all of the qualifications required for him to be the historical catalyst for the legends and lived at a time that would make the development of the Arthurian legends consistent with what we see in analogous cycles of legends from Britain. That, combined with his connection to known transmitters of folktales that were later incorporated into the legends, makes Castus a powerful candidate for the title of the historical Arthur.