A can be reconstructed rather more closely than we have done so far. We have seen that its backbone is composed of triads, and recovered two such triads from Gildas: the three peoples who invaded Britain and were never ejected, and the three Pictish invasions. Both these triads are set after the end of Roman power.

There is however a third triad which brings together Gildas’ legend of the end of Roman power with his legend of its beginning. He has not only three Pictish invasions and three invaders of Britain, but also three Roman missions to the island: the great invasion that followed the massacre of the rectores by the dolosa leaena, and the two missions against the Picts. The first invasion of this triad is at the beginning rather than at the end of Gildas’ Roman legends; we must therefore look at the possibility that A was the source of all Gildas' Roman legends, forming a unified whole - presumably structured in triads.

Nennius has two separate and incompatible accounts of the end of Roman Britain. The main body of "Roman" lore in his work is a long account of what I call “the legend of the Seven Emperors”, stretching over nine chapters (19-21, 23-28) which form one confident and continuous narration^[1]. Evidently it was the standard account in Nennius' time, confirmed by popular traditions (its "Emperor Constantine" was apparently identified with a local hero of Caer Seint called "Minmanton") and paralleled by some of Geoffrey's Roman legends. It starts from Caesar's supposed conquest of Britain - in which a number of basically historical data drawn from Bede keep uneasy company with a great deal of Welsh legendary material - and ends with the mad adventure of Maximianus (i.e. Maxen), after which the British rejected Roman rule and "killed the leaders of the Romans". The most remarkable aspect of this legend is that these supposed emperors, though imposed by the Romans, seem to have lived and reigned in Britain alone; that is, they were British emperors of Roman race^[2]. The effect of the final British rebellion is that "they no longer accepted the kings of those people [=the Romans] to rule over them" (ch.28); and we are reminded of Elen’s demand that Maxen Wledig, legitimate Emperor of Rome, should come to live in Britain.

Ch.28 concludes the story, drawing a line under the Roman episode. In keeping with Nennius' interest in chronology, it gives a reckoning for the length of Roman rule in Britain - 409 (!) years. There follow two chapters with supplementary material. The first is a collection of annalistic data about the historical Magnus Clemens Maximus; the second - Ch.30 - a most extraordinary jumble of data that claims that the British killed Roman leaders (duces) not once but three times. Clearly, chs. 29 and 30 were put in as alternative information, in the manner of footnotes.

This is typical of Nennius’ methods. Earlier on, he inserts a genealogy of the Western nations (ch.17), certainly non-Roman in origin, as an alternative to the full ethnology of peoples settled in Britain which he has developed in the previous chapters 10-16. He does not much believe in it, but he has found it in ancient written text and feels he must insert it; in effect, this is a footnote. Ch.17 and ch.30 have in common that they cannot be reconciled even with the most basic Roman history, and it becomes clear that this was a Nennian habit. Nennius reconciled everything he could (as for instance in his composite version of Caesar's “conquest” of Britain, slotting together Welsh legend and Bede's historical account), but if he found himself with two alternative and irreconcilable accounts, he placed first, and gave more space to, the one he judged the likeliest (that is, the one which seemed more closely connected with a Roman idea of history), then added the second as if an afterthought, and - to judge from the size of chapters 17 and 30 - gave it as little space as reasonably possible.

This acceptance of Roman historical traditions had its limits. Chapter 29 brings together impeccably Roman and largely historical data; indeed, it seems to single them out and separate them from the body of the narrative. The reason is certainly that it interfered with a genealogical fiction embodied in Nennius’ legend of the Seven Emperors: the division of the features of Magnus Maximus, both historical and legendary, between Maximus the sixth emperor and Maximianus the seventh and last.

In spite of being Gildas’ bad guy, Maximus/Maxen had become the legendary ancestor of some Welsh royal houses. This is reflected in a cleaned-up legend - a form of which we have in The dream of Maxen Gwledig - with features borrowed from the stories of Constantine and St.Helena. To explain the Gildasian tradition’s view, a wicked and nearly homonymous character, Maximianus, had to be invented. "Good" Maximus, separated from "bad" Maximianus, was attributed the white-hat aspects, such as the marriage with Elen of the Hosts. In particular, the historical Magnus Clemens Maximus had a famous encounter with Saint Martin of Tours, who put him in his place quite firmly, and corresponded with the equally illustrious Saint Ambrose of Milan. This was part of a general policy to get the Church on his side, which included the first recorded State persecution of Christian heretics, the Priscillianists (this did not impress Martin^[3]).

Now my point is that although Nennius specifically says that ch.29 is about Maximiano tyranno, the rest of the chapter calls the usurper Maximus; and while ch.29 gathers every scrap of annalistic information he had been able to find about Maximus and reattributes it to Maximianus, it carefully avoids his famous encounter with St.Martin, or his dealings with St.Ambrose, who is mentioned but not connected to him. On the other hand, the only thing that Nennius has to say about Maximus, the “sixth emperor”, is that he spoke with St.Martin! In other words, Nennius understood perfectly well the artificial nature of the division between Maximus and Maximianus and consciously manipulated his material to suit it.

This is Nennius’ treatment of the then-current legend of the Seven Emperors, and of respectable Roman annalistic material; the duplication of a “good” and a “bad” character out of an originally villainous one is to be remembered, since we will meet it again. However: when we get to Chapter 30, suddenly we are in another world. The chapter poses so many questions - and offers, on the other hand, so many opportunities for fruitful analysis - that I can do no better than reproduce and translate it word for word.

Tribus vicis occisi sunt duces Romanorum a Brittannis. Brittones autem, dum anxiebantur a barbarorum gentibus, id est Scottorum et Pictorum, flagitabant auxilium Romanorum, et, dum legati mittebantur, cum magno luctu et cum sablonibus super capita sua intrabant, et portabant magna munera secum consulibus Romanorum pro admisso scelere; et suscipiebant consules grata dona ab illis; et promittebant cum iuramento accipere iugum Romanici iuris, licet durum fuisset. Et Romani venerunt cum maximo exercitu ad auxilium eorum, et posuerunt imperatores in Britannia, et composito imperatore cum ducibus revertebantur exercitus ad Romam usque, et sic alternatim per CCCXLVIII annos faciebant. Brittones autem propter gravitatem imperii occidebant duces Romanorum et postea petebant; Romani autem ad imperium auxiliumque et ad vindicandum veniebant, et, spoliata Britannia auro argentoque cum aere et omni pretiosa veste et melle, cum magno triumpho revertebantur.

"Three times the leaders of the Romans were slain a Brittannis. For [autem] the Britons [Brittones], when they were made fearful by the tribes of the Barbarians, that is, of Irish/Scots and Picts, would petition for the help of the Romans, and, when they sent ambassadors, would enter in dressed in great morning and with ashes/sands [sablonibus] on their heads, and would bear great compensations for an admitted crime; and the consuls would receive welcome gifts from them; and they would promise with a sacred oath to receive the yoke of Roman jurisdiction, however hard it might have been. And the Romans came with the mightiest army, and placed emperors in Britain, and having made an agreement with [? or "placed together"?] emperor and leaders [?or "having got the emperor to agree with the leaders"?], they took the army back to Rome, and so alternately for 348 years [sic, indeed!]. And so the Britons [Brittones], on account of the heaviness of empire [or: command], would slay the leaders of the Romans, and then would beg; the Romans then would came for empire, for help and for revenge, and, having stripped Britain of gold and silver with bronze [or: money] and of all precious raiment and honey, went back with great triumph."

Now what on God's green earth is going on here? It is all but impossible to get a coherent story from this. The chronology is thoroughly unclear, smothered by an avalanche of imperfects; most verbs are in the imperfect, a tense with no English equivalent for something that "used to happen" over a rather long stretch of time, and that cannot therefore be used to establish a time-sequence. Nor is it possible to understand the relationship of "Emperor" in Rome, "Emperor" in Britain, "leaders" (duces) in Britain, "consuls" in Rome - of which there seem to be any amount, and who may correspond to the Senate - and I will not even try. It is also badly out of keeping with the legend of the Seven Emperors; they are connected by the idea of a massacre of Roman leaders, but otherwise contradict each other in any number of ways. In ch.28 a single Massacre discourages the Romans from ever coming back to Britain; in ch.30 the last (or so it seems) of three Massacres unleashes the mother of all Roman punishment raids. In ch.28, Roman rule is said to have lasted 409^[4] years; in ch.30, 348 - which shows that the two accounts came from different manuscript sources with different chronological computations. And - which ought to make us pay attention - the Picts are hardly mentioned in the legend of the Seven Emperors, but are a major presence in ch.30, where fear of them causes the British to send to Roman help, even after one or more of the massacres have taken place. Ch.30 closes incongruously with a great Roman army coming to Britain for vengeance, and returning to Rome in great triumph, having despoiled the island of gold, silver, bronze, precious raiment and honey. Clearly, chapter 30 needs plenty of disentangling.

However, it is immediately clear that it has a good deal in common with Gildas' pseudo-history. Nothing in the legend of the Seven Emperors has much to do with Gildas' Roman legends, but several points in ch.30 are directly comparable:

1) The beginning of the narrative structure must be in peaceful British acceptance of Roman rule. This is the best reading of the first two sentences: "The Roman leaders were slain three times a Britannis (introduces the subject); for (autem), when/because (dum) the Brittones were dreading the tribes of the Barbarians... they sought the aid of the Romans..." In other words, Roman power in Britain begins with a peaceful British surrender to Rome: in Gildas, as a result of Roman messengers and threats; in Nennius, as a result of the fear - only the fear, mind you - of barbarian tribes. This, especially the Nennian version, conflicts with the Seven Emperors (Nennius ch.20) story of Caesar and Beli son of Minogan, in which Caesar asserts Roman sovereignty simply as a matter of course - the initiative is wholly on the Roman side - but then has to fight for it^[5].

2) This was not just a request for an alliance: it was the surrender of sovereignty. The British promised to accept Roman rule, and while it is Gildas who insists on the shameful and cowardly behaviour of the British, it is Nennius who shows them offering the Romans, unbidden, rule over Britain: promittebant cum iuramento accipere iugum Romani iuris, they promised with a sacred oath to accept the yoke of Roman jurisdiction - underlining the commitment with a pun between iugum, yoke, and ius, jurisdiction, law^[6]. From now on, therefore, the Romans were the legitimate masters of Britain. And it is worth making the point that, even though Roman might rather than the fear of barbarian tribes was the cause in Gildas, the motive power of the surrender of Britain to Roman jurisdiction was in both cases the British inability to defend themselves. (In the legend of the Seven Emperors, the British, under Beli or Casswallawn, stubbornly resist Caesar.)

3) The Romans then placed a number of sovereign representatives in Britain. (In the legend of the Seven Emperors, it is the Seven Emperors themselves who rule Britain, leaving various tokens of their presence from the Wall to "gold seed" in Caer Seint.)

4) However, after accepting the yoke of Roman law, the British found it too hard and massacred the Roman representatives. (In the legend of the Seven Emperors, this did not happen until after the fall of "Maximianus", that is Gildas' Magnus Maximus.)

There is a verbal parallel that proves that Nennius and Gildas were working from a common, Latin, written source. When the Britons, begging for a Roman rescue, make a bitterly self-abasing plea, the authors use singularly similar expressions. Nennius: cum magno luctu et cum sablonibus super capita sua, in great morning and with ashes on their heads; Gildas: legati scissis, ut dicitur, uestibus, opertisque sablones capitibus: "the ambassadors - with their clothes, as they say, torn, and dust poured over [their] heads..." I am especially intrigued to see both authors use the rare word sablones.

As there is no question of Nennius quoting Gildas' scene - the context, so far as there is one, is quite different - we must admit that they are both quoting from a piece of writing known to both. Where they do not agree, it is because Nennius is using a more down-to-earth-expression - cum magno luctu, in great mourning - where Gildas borrows a favourite image from his beloved Bible - scissis uestibus, with their clothes torn. To tear one’s clothes as a sign of mourning was traditional among Jews, but not, we must assume, among Britons; Gildas underlines the unusual nature of the expression with an ut dicitur, "as they say". The two images of mourning are similarly constructed, with a description of the ambassadors' general apparel - deep mourning, "torn clothes" - followed by the detail of a head covered with some sort of detritus, sablonibus, sand or ashes; and the deep mourning and self-abasement are practised by similar personages, ambassadors, seeking the same kind of help from the same power. There can be no chance of error: Gildas and Nennius were quoting.

Now, in Gildas, this embassy is decidedly a part of A. It is the direct result of the Pictish invasion (one of a triad), and leads to the final Roman expedition (the last of a triad); and yet it is, in Nennius, associated with events that, as we have just seen, are directly comparable to the Gildasian account of the beginning of Roman rule, the massacre of the rectores and the avenging expedition. While the structure of this is not clear yet, it shows more reason to think that the account of A covered both the beginning and the end of Roman rule in Britain.

We may have more of A's very words, and significant ones too. The pun between iugum, yoke, and ius, justice, law, jurisdiction, is not at all typical of Nennius, whose Latin is too basic to make a point with such a sophisticated device. What is more, it is not fundamental to his main account, the one to which he is really committed, that is that of the Seven Emperors, which has nothing to say about the heaviness of empire, and where the rebellion of the British has nothing to do with the iugum of Roman law. It is, however, central to Gildas' account, where the British rebel because they cannot bear to have to obey Roman law; and it makes the subsequent massacre of the leaena dolosa a matter of "denial after recognition", thus motivating the Romans' savage punishment; which in turn must have been part of the point of ch.30's original, built as it must have been around three successive massacres of Roman "leaders" (in the legend of the Seven Emperors, revolt scarcely features except after the closing reign of Maximianus, and "denial after recognition" not at all).

By the same token, it would be typical of Gildas' brand of Latin, just as it is not of Nennius, to use a pun, or even a paretymology - the author may have actually thought ius and iugum etymologically related - to make an important point, as ius/iugum/iuramentum does here. The pun shows that the voluntary acceptance of Roman rule is the central point of the story - the British promise with an oath, iuramentum, to bear the yoke, iugum, of Roman law, ius; but it certainly is not in that of the Seven Emperors, that begins with the conquest of Britain by Caesar and in which there never is a question of the Britons voluntarily accepting Roman rule or swearing any oaths.

In other words, the agreement with Gildas in theme, views ("denial after recognition" is as disastrous here as there), style and basic points, strongly suggests that the author came from a similar cultural milieu. But the author is not Gildas, since the narration deviates from Gildas' in important ways - three massacres! - and Gildas himself shows signs of having partly rewritten the description of the ambassadors: that ut dicitur, I think, signals a personal flourish.

I hardly need stress that Gildas and Nennius must have treated their common source in very different ways. Discrepancies leap to the eye. For instance, in Gildas, the reason why the British envoys wear mourning is, that they are in mourning for the ruin of Britain under Pictish blows; in Nennius, that they are atoning for their own people's murder of Roman duces - their fear, anxiebantur, of the barbarous Picts and Irish, seems to be a secondary consideration. And our first remark must be that, granting a common original, it seems probable that Nennius was closer. The abject self-abasement of the legati does seem to suggest something more than a mere plea for help, however tragic the circumstances; and three successive massacres of duces must involve successive attempts to placate the injured Power. We admit that Gildas' Romans had also the stain of a crime against Rome, to wit the rebellion of the "British tyrant" Maximus; the more remarkable, then, that Gildas did not mention it in this scene.

One thing we easily recognize is Gildas' invincible army devastating and despoiling Britain in revenge for this massacre. In both accounts, the British slaughter the Roman representatives but are then unable to cope with an overwhelming revenge; the Romans then take over British wealth. In Nennius, Britain is despoiled of auro argentoque cum aere et omni pretiosa veste et melle; in Gildas, the Romans order that quicquid habere potuisset aeris argenti uel auri imagine Caesaris notaretur, "whatever [Britain] could possibly have of coin of silver or gold [or: of bronze, silver and gold] should be stamped with the image of Caesar. Then, in Nennius, cum magno triumpho revertebantur, they would return home with great success; in Gildas, Italiam petunt, suorum quosdam relinquentes praepositos, they go back to Italy, having left some of their own in charge.

The magno triumpho implies that Nennius' army fought a war; on the other hand, he agrees with Gildas that the British had accepted Roman rule peacefully at first; in other words, that the Romans had not fought any war to impose their power in Britain in the beginning. Therefore this great war, along with the massacre that sparked it, must have happened at some point after that; and this agrees with what we have seen, that it was a case of "denial after acceptance" followed by punishment - the Romans were already legal overlords of Britain when the great raid took place.

This is our first step to a reconstruction of A’s original from Nennius’ and Gildas’ heavily rewritten versions. o far, so good; but was it, as Nennius makes it, the last act of the Romans on British soil, or, as Gildas says, the beginning of their direct rule, replacing most of the British nobility with their own rectores, and a whole Host of what one supposes were faithful retainers well able to fight in war?

As Nennius phrases his account, the fact that the Romans as a whole are seen to "go back in great triumph" seems to suggest that this was the final episode in the history of Rome in Britain: the Romans left. But in fact exactly the same thing happens in Gildas: while the rectores and their retainers (which must make up the Host) are left behind to enforce everlasting servitude on the surviving Britons, the bulk of the Roman army Italiam petunt, go back to Italy. And yet this is the beginning, not the end, of Roman direct rule. Arranging a comprehensive form of enslavement for the British, including the "stamping of aes of gold and of silver", the Romans go back to Italy, Gildas says; but Roman power over Britain is established, first by legal submission, and then by the equally legal and legitimate punishment of those who dared rebel against their already recognized overlords.

Now Gildas gives what seems, at first, a reason for the Romans to go home: an ablative absolute - they went away patria uini oleique experte, the fatherland^[7] being deprived of wine and oil. But the ablative absolute can be a very ambiguous form. Commonsense suggests that it means "the Romans returned to Italy^[8] because Britain had no oil or wine", but its position in the sentence suggests a different explanation. Gildas' chapter 7, like the chapter 8 that follows it, is a grand one-sentence passage^[9]which details the measures taken by the victors: establishment of rectores with power of life and death over the natives, renaming of the country - not Britannia but Romania - and stamping of quicquid habere potuisset aeris argenti uel auri. The effect is to suggest that Britain's deprivation of oil and wine is part of the Roman punishment, which is certainly the way I understood it on my very first reading: as Britain is to be placed permanently under the whip of her new Roman lords, so, too, she is to be deprived of wine and oil - without limit of time.

Now Nennius' version of the invincible avenging Roman army also takes wealth over; and we have seen that Nennius has a remarkable list of precious substances. Auro argentoque cum aere et omni pretiosa veste et melle: gold and silver with bronze (or: money) and all precious raiment and honey. The list's wording is notable for a tendency to make it fall apart into groups: gold and silver (where the Latin conjunction -que forces the two words closer together) with bronze/money and all precious raiment and honey. These may be connected with the two groups of substances in Gildas: "whatever it could have of aes of silver and gold" and "deprived of olive oil and wine": the structure of both lists would be the same - five terms, of which he first three are means of exchange, and the last two luxuries.

I can give no certain explanation of the use of aes, bronze or money, which is oddly placed in both lists - separated from silver and gold in Nennius, and, to the contrary, so placed in Gildas that it might mean both "bronze, silver and gold" and "coins of gold and silver". In Gildas' passage, the partitive genitive - "whatever it might have of..." - naturally implies a whole category of objects: "whatever it might have of precious substances". Therefore it is easier to read the word aeris, in aeris argenti uel auri, as an abstract category, aes as money, rather than as the physical substance bronze: whatever Britain might have of money, whether of silver uel (or indeed) of gold (that is to say, the gold is even more subjected to this order than the silver^[10]), was to be stamped with the image of Caesar. It is also possible, though undemonstrable, that the ambiguity of aes may depend in some fashion on the difference between gold and silver and other metals. Gold and silver are precious, means of exchange in themselves; but any other metal only becomes a means of exchange when it is stamped by the State to make coins. We might therefore draw a distinction between aes as such - money in general, and also the non-precious metals most often used to make it, bronze and copper - and gold and silver, which, minted or unminted, retain their value.

Let us not forget that the existence of a list of precious substances is not something I have made up: it comes from the clear correspondence - just about the only clear thing in all the dossier - between Gildas' and Nennius' versions of the conquest of Britain by an overwhelming Roman army, reacting to the massacre of some of its duces or rectores. These two lists represent the final result of the Roman victory: they are what the victorious army takes on Rome’s behalf. As the two lists are at the same place in both accounts, we may a priori expect that they should correspond; and this is the strongest argument to insist that the "olive oil and wine" of which Britain was ambiguously "deprived" when the Romans of Gildas left, were in fact confiscated by them - they were part of the list of precious substances and objects that the conquerors took.

In this light, these are the lists: gold, silver, aes, precious raiment, honey; gold, silver, aes, olive oil, wine. The first thing to leap to the eye is that Gildas' two doubtful terms, olive oil and wine, are not native to Britain, while those of Nennius - precious raiment and honey - are. This suggests that Nennius may have smoothed out an absurdity in his source by replacing the incredible wine and olive with more credible terms. Faced with the same narrative implausibility^[11], Gildas solved the problem differently, hiding the statement away in a deliberately ambiguous thicket of words - an exercise for which Nennius lacked the skill - avoiding a direct statement of the absurd (after all, when a writer of his skill is unclear about anything, it is usually because he wants to be). We know from his polemic against Constantine of Dumnonia that he was well acquainted with classical agricultural vocabulary, surculamen, gleba, imbribus irrigatum; and if he had read classical agricultural manuals such as Columella's, he would know well enough that it took no imperial decree to keep olive trees away from Britain, and that on the other hand grape vines could and did grow here^[12]. Hence the embarrassment with which he buried his source's claim that these plants grew in Britain once, and were taken away, in an ambiguous clause in the middle of an enormous forest-like sentence.

Gildas' list must therefore be closer to the original; and we notice that his substances, unlike those of Nennius, fall naturally into two groups. Gold, silver and money, means of exchange, go together; and so do olive oil and wine, not only as food luxuries, but more importantly, as Christian sacrificial substances. Gildas explicitly mentions the anointing of kings in 21.4, complaining that bad kings "were anointed", ungebantur, without the will of God. The passive seems to me to indicate that he has a clear picture of the king having the ceremony performed upon him, of the oil of consecration being poured over his head. At the time, the list of Catholic sacraments had not yet stabilized into today’s seven. More than one authority regarded the consecration of a king as a sacrament. Oil was probably also used, as now, in such undoubted sacramental acts as Confirmation, Priestly Consecration, Anointing of the Sick. As for wine, it is of course part of the Eucharist. That Rome deprived Britain of both, means that the island was left in a position of radical religious dependency on the Roman Mediterranean lands. (This supports what I have already argued, that the legend of A was formed in an already Christian environment, where the alien juices of the grape and the olive were the most important media of religious ritual. No pagan Celt would have either found any significant and exclusive association between them, or placed them on a level comparable to that of gold, silver and bronze.)

With this goes the Roman claim on gold and silver. Though Gildasian kingship expresses itself in military power such as Maglocunus', gold and silver are what it is made of: "The enormously capacious nets of royal power" that bind Maglocunus and all the "fat bulls of your kind", that is warrior kings, are "gold and silver, and, what is greater than these, self-will"; capacissimis illis quibus praecipitantes inuolui solent pingues tauri moduli tui retibus, omnis regni auri et argenti, et, quod his maius est, propriae uoluntatis. So thunders Gildas; and after all, we have seen that the essence of Celtic kingship is a marriage with the land that places all its wealth in the king's power. And British aes, of silver uel (or even more) of gold, is now stamped with the Emperor's face: that is, the wealth which is the substance of royal power is now radically, intimately claimed by Rome.

Nennius' gold, silver, aes, precious raiment, and honey, have no such clear structure, but they can best be made to make sense as a reworking of Gildas' list. Precious raiment is probably a part of the ceremony of coronation^[13], like coronation oil, and honey is the raw material of mead, an alcoholic drink that can probably be seen as a substitute for wine. If that is the case, it is perhaps worth noting that even the two sacramental substances may be important not for their own selves, but through their relevance to royal power. The Christian Sacraments are accessible to any member of the Church, royal, base or slave; but if Nennius substituted precious raiment for olive oil, the point of the replacement is probably that both are elements in the ceremony of consecration of a king. In other words, the story seems decidedly to centre not on the Church or the faithful, but on the consecration of kings.

It is however difficult to be sure that Nennius understood the meaning of the list. He does seem to have missed the point of gold, silver and aes, speaking not of Rome "stamping" British precious metals, nor depriving Britain of substances that never could have been found there; his Romans simply remove every scrap of British wealth. And this corresponds to another difference: Gildas' Romans establish their rule by the stamping and taking away, but the Romans of Nennius just take everything, and leave.

This can be seen as symptomatic of a major change between the ages of Gildas (561) and Nennius (834). Archaeology shows that nobody in Britain between Constantine III and the first Christian English kings minted any coins; Gildasian Britain only knew coins as products either of distant Byzantium or of the Roman past, associated, either way, exclusively with the power of Rome. Trade went on and precious metals, whether coined or not, were exchanged, as shown both by massive archaeological record and by the Life of St.John the Almsgiver^[14]; but the mercantile notion of coins as State-guaranteed units of value can hardly have been present when no local sovereign did any coining. Coins were hoarded, faked^[15]and used, but Gildas' contemporaries can have seen no difference, in terms of purchasing use, between them and any other kind of gold and silver. An archaeological find such as the Traprain Law hoard - a mass of silver broken up and clearly intended to be shared out by weight - shows that at least in some areas of Britain weight, and weight alone, measured the metal's value. The only difference, therefore, must have been that coins were stamped with Caesar's image. Gold was simply gold and silver simply silver, of one value whether in bullion, coins or jewelry; but coins with imperial images were not so much precious objects, as tokens of Roman power over wealth.

Gildas' words tell us what ideological construction was put on coins. When Rome took over and, in a mystical sense, married Britain the Bride (inflicting deserved punishment on her unworthy suitors), all Britain's gold and silver (and bronze?) passed into Rome's possession, or rather control. Gold and silver, if we read Gildas correctly, were never taken away from the island; as far as he was concerned, there was if anything far too much of the stuff floating around, corrupting people. Rather, the Romans placed on it their magic sign, to mean that whoever might hold gold or silver in Britain/Romania, they ultimately pertained to Caesar. As the British were, in Gildas' view, the Romans’ slaves, so British wealth was ultimately Rome's, to be disposed of according to Roman law (ius). And yet, just as the "slaves" of the Romans were in fact local lords, of inferior rank but capable of political activity, so too the gold and silver of Britain, while ultimately under Roman power, did in fact circulate among public and private bodies in the island. When the Romans last appear on British soil, they commandeer "public and private" (sumptu publico priuatoque) British finance to build the Wall. They can do what they want with both public and private wealth.

Now, between Gildas and Nennius, coins have returned to Britain. By the ninth century, every English kinglet worth the naming minted silver pennies, and gold coins were not unknown. We know of no Welsh king who minted his own coinage before Hywel Dda, a century after Nennius; but to Nennian Wales, silver and even gold coins can have had neither the exotic aura nor the imperial significance of Gildas' time. And it follows that Nennius would misunderstand what the age of Gildas had to say about minted money. The act of making coins conferred no special right to them; and it was not vested in only one authority, imperial, distant and worldwide, but in every king or kinglet who bothered. The notion of the Romans coming over and stamping precious metals with their image would mean nothing; if Nennius was told that this happened as a result of a great war, to him that would only suggest plunder. He could understand tribute, but connecting it with an ultimate right to all the wealth in the land, confirmed by stamping the imperial image on gold, silver and aes, would be unfamiliar.

When Nennius was faced with this archaic notion, it fell apart in his hands. The Roman claim to the wealth of Britain, though only a part of their royal claim to the island, was fundamental, and, in Gildas, came first - at the beginning of their period of full power. But Nennius just naturally "heard" the idea of the Roman conquest of Britain and of the Roman takeover of British wealth as two completely different events, without even reflecting on them.If the original story said that they Romans "took" all the wealth of Britain, it would not occur to him that they could at the same time claim it all as their own and demand tribute; that British wealth, while remaining in the patria and circulating among its citizens, could still ultimately belong to Rome. In any story he wrote, if everything in a country has been taken, there would be nothing left to fight with or about. It would be the end. What Gildas' version, on the other hand, signifies, is that the Roman claim on British wealth is permanent; therefore the Romans have just asserted, not renounced, their sovereignty. As a necessary precondition of royal power, it belongs at the beginning , not at the end, of their rule, and that is where Gildas places it.

So Nennius made a mess of the story structure. The unity of any story is a function of the cohesion of its basic ideas: if these fall apart, the story falls apart. And if you start from the principle that the prince is only one among many possessors of land and wealth, wealthier, no doubt, than any of his subjects, but not otherwise than any of his subjects, with no overarching claim to land and wealth - then it follows as the night follows the day that if he claims all his subjects' wealth, he cannot at the same time claim tribute from them. In this as in the discovery of real Roman annalistic history, Nennian Wales, in spite of its Celtic and tribal aspects, is several stages closer to Classical, mediaeval and modern Europe than Gildasian Britain with all its unmixed Celticism wrapped in Gildas' splendid Latin.

It is worth pointing out that this has a loud and direct echo in Geoffrey of Monmouth, who inserted in his Roman legend a furious denunciation of Roman greed, wandering all over the world in search of silver and gold (Historia regum Britanniae, 4.2). This is bound to be a misunderstanding; surely the point of the original legend was that the whole world's gold and silver were Roman by right because of Rome's majesty - and in fact, even if Rome claimed them, they would still remain in the hands of their current users, only stamped with Caesar's head! Like Nennius, only more so, Geoffrey had read archaic Welsh accounts, but failed to understand what he read^[16].

Rome's disposal of the precious/sacred liquids goes along with the "stamping" of gold, silver and "whatever the country had of aes". Gold and silver pertain to the king, indeed they are the "nets of royal power" in which fat bulls like Maglocunus are ensnared; oil and wine are holy, they serve the Church's sacraments (oil especially the consecration of kings). This is not a question of wealth alone, but of the functioning of society's two highest functions, the royal and the priestly. They depend on these substances: the substances are are precious not for themselves, but for their part in royal/terrestrial and priestly/otherworldly power. This is what Rome claims. Their confiscation stands for a Roman claim on everything royal and sacred in Britain, and signifies, not the end of Roman power in Britain, but its fullness. It is absurd to imagine that Rome would claim the stuff of royalty, gold and silver, and actually remove from the island the stuff of sacred ceremonies, and then never interfere in the island again. Their taking is an act of majesty, part of the process by which Britannia becomes Romania.

Most important is the fact that, while gold, silver and aes are stamped and then allowed to stay in Britain and circulate, the substances of the Mass, oil and wine, are effectively and for ever taken away from the patria. From henceforth, Britain will be dependent on Rome for sacred activities. From the point of view of the legend's origin, this is certainly a reflex of the experience of having to bring in these precious liquids from abroad in the bottoms of trading ships which, archaeology tells us, travelled as far north as Strathclyde and Ireland. The urgent need of northern Christian communities for regular quantities of wine and oil was surely the motor that moved this trade, although, of course, once it had started, all sorts of other luxuries will have been introduced to northern lords by merchants eager for a return on their lengthy and uncomfortable journeys. But from a religious point of view, I think there is little doubt that it reflects a complete and, what is more, untroubled religious dependence on Rome. Unlike political dependence, there is no indication that it ever ceased, that it was regarded as in any way a iugum, or that any break was ever envisaged; there is no trace whatever of the religious nationalism that several authors have found good or less good reason to suspect in early fifth century Britain.

It is probably to the point that we are told that the final victory over the Picts was won after the British consecrated themselves to God, transferring, as I pointed out, to God the royal functions that the Romans would no longer exercise; and God gave them the victory, providing a more reliable and less heavy iugum than the temporal power of Rome. In the fifth century, many Western Roman writers felt that the religious domination of the Church of Rome was replacing the failed temporal empire, and not only succeeding to it, but enlarging it; such views can be found in St.Augustine of Hippo, Pope Leo the Great, Orosius and Prosper of Aquitaine, for instance. This seems, in a provincial and mythological manner, a reflection of the same idea: that the sacred majesty of Papal Rome not only outlasts but surpasses, in the name of God, that of the empire of old. (And dare anyone say that this has not proved to be the case, then and for fifteen centuries since?) A new British nation has been born, but not a new British religion; that is the same, the religion of the Palestinian Bread and (imported) Wine.

Notes

^[1]Apart from ch.22, which inserts two separate characters from a different line of tradition: the "British" "King Lucius" and "Pope Eucharistus" of Christian Welsh legend, both thoroughly apocryphal, associated with the first coming of Christianity to the island. The reason for the sudden apparition of an autonomous British sovereign during a period in which, according to all other traditions, the Romans ruled (unless of course it had no original chronological dimension at all and was arbitrarily assigned to the second century) is surely to make Christianity the free choice of a supposedly autonomous local figure - that is, the ancestors of the Welsh owed their conversion to the True Faith only to their own free will. It had nothing to do with the legends of Roman power in Britain and may be a good deal later. In the fourth century, St.John Chrysostomos knew of a legend that the Apostles had first taken Christianity to Britain; this was likeliest to have come from Britain itself and suggests that, like contemporary Gallic and Hispanic Romans, the Romano-British Christians wanted to trace their Christianity back to the very first generation. The invention of a British king and an unhistorical Pope suggests a far more particular and nation-centred relationship to the universal Church, dating surely to after the sundering of the British from the notional Roman world and their self-recognition as a separate nation. Bede explained away the unhistorical Eucharistus by identifying him with the historical second-century Pope Eleutherius; the name of "King Lucius" is suspiciously close to that of the great pagan god, Lug or Lleu, who may be shown to have become incarnate, at one point or another, in several heroes (cf. my Gods of the West I: Indiges, Brussels 1999).

^[2]That the Seven Emperors were understood to have lived and died in Britain is shown, not only by the way many of them were associated with local features and legends, but also by a Nennian remark: that "the ancient tradition of our elders" knows of seven emperors, but that the Romans know nine; the extra two, he says, are "another" Severus and one Constantius. Now, the historical Septimius Severus and Constantius Chlorus are the only two Emperors to have died in Britain; this is obviously a malentendu between native savants looking for Roman Emperors resident in Britain, and Continental traditions about Emperors of Rome sojourning and dying there. The Welsh wanted to know about Roman Emperors who stayed in their own country; Roman tradition mentioned only two - and those two unknown to native legends; so Nennius, or whoever first brought the records together, assumed that these two were extra. (Curiously enough, Nennius states that the "second" Severus died in Rome.)

^[3] But it impressed other Church writers, alas. Except for Britain, Magnus Clemens Maximus left behind a rather better impression than most pretenders of the period: Orosius, no friend of usurpers, called him uir... strenuus et probus, et augusto dignus, a man of vigour and honesty, worthy of being an emperor. OROSIUS, Historia adversus paganos, 7.3.4.

^[4]This clearly depends on the Anno Domini date of the Rescript (410AD), but fails to understand it, taking it for the years from the Roman conquest of Britain, presumably by Caesar. That is, at some point in the transmission of the account, standards of Latin, chronology and simple intelligence had seriously slipped - after all, dating from the birth of Our Lord should not exactly be a hard concept to understand! But as the legend of the Seven Emperors is much the later of the two, I cannot connect this to the ignorance of chronological science I diagnosed in Gildas.

^[5]See Appendix I.

^[6]Plus a secondary one with iuramentum, oath, which I do not underline because iuramentum comes from ius and so is hardly a proper pun at all. It is also far less meaningful than that between iugum and ius, the "law" as a "yoke".

^[7]Gildas says the patria, the fatherland, by which he certainly means Britain. It is significant that even at this point of deepest national humiliation, he still naturally thinks of Britain as "the land of our fathers": it is the Romans who "leave" the "fatherland" to go "to Italy".

^[8]As an Italian in Britain, I cannot help responding personally to this passage. It sounds to me like an unexpected sudden touch of realism in this high-flown legend of conquest and terror: Italians going back to their precious wine and oil, convinced, like Italians at all times, that the British can't cook to save their lives and that the best food and the best wine are at home in Italy! Did Gildas ever meet and speak with a homesick Italian? It is a thin, even frivolous consideration; but it does incline me to believe that Gildas wanted us to understand that the Romans went back to Italy because they had found Britain deprived of the precious vegetables, rather than because they had deprived the island of them.

^[9]Since we know that both 7 and 8 were recast from an earlier source, 7 from A, 8 from Rufinus, it seems likely that Gildas liked to recast sources in this grandiose manner, both to underline their basic unity and to show he could do it. On the other hand, he did not do so regularly: his chapter 21, for instance, is certainly derived from a single source, but it is cast in fairly regular prose with no attempt to form a long sentence.

^[10]The emphasis on gold may have to do with the fact that, in Gildas' time, gold coins were regarded as peculiarly an attribute of the Roman (Byzantine) emperor: "the gold solidus continued to be regarded as an imperial prerogative. The presumption of the Frankish king Theudebert I [reigned 534-548] in issuing solidi in his own name was censured, and the experiment was not repeated in Gaul until the concept of Empire had weakened a century later." M.ARCHIBALD, M.BROWN & L.WEBSTER, Heirs of Rome, in L.BROWN & L.ALCOCK (ed.), The transformation of the Roman world, London, British Museum Press, 1997, p.210; cf. also P.GRIERSON & M.BLACKBURN, Medieval European coinage I: the early Middle Ages (500-1000), Cambridge 1986.

^[11]This suggests that Nennius and Gildas must have been in receipt, not only of a common narrative tradition about the Roman conquest, but also of some subsequent criticism. While Gildas had certainly enough Latin reading to know about olive trees and grape vines and where they grow, there is nothing to suggest that Nennius did; the man who treated the Bristol Channel as an alien region full of legends and miracles cannot be expected to be very clear about the climatic and farming differences between his native country and distant Italy. I suggest that he found a criticism of the wine-and-olive passage embedded in whatever account of A reached him, and it did not improve his opinion of a source for which he does not anyway seem to have had great respect.

^[12]Traces of a Roman vineyard have been found in Northamptonshire; Current Archaeology 150 (Nov.1996), 212-215. Then as now, growing decent grapes so far north would be difficult at best. What is more, if A originated in the north of the island, near or beyond the Wall, that decidedly is no place for either grapes or olive trees; even if vineyards did grow in Northamptonshire, they might not know of them in Stirlingshire.

^[13]In Ireland, special dress was a specific feature of the inauguration of kings great and small, and was afterwards granted as a gift either to a leading ecclesiastic, or, more frequently, to the chief poet who had taken part in the ceremony. F.J.BYRNE, Irish Kings and High Kings, Dublin 1973, 20-22.

^[14]A Byzantine life of an Egyptian saint, which includes an account of a trading expedition to Britain. E.DAWES and N.H.BAYNES, Three Byzantine saints, Crestwood, NY 1996, 216-218 (Life of St.John the Almsgiver §10). SNYDER op.cit.310 (note 38 to page 113) gives a brief account of the stages of its writing, including reference to skeptical scholarly views as to the historical worth of the journey to Britain. In my view, the journey to Britain sounds perfectly credible, save for the final miracle - which takes place in a North African port - of the transformation of British tin into silver, which might well depend on a simple misunderstanding of a statement that silver was paid for tin.

^[15]"Individual finds in Britain dating after 411 tend to be copies or counterfeit, not official issue"; SNYDER op.cit. 134 and note 12. That post-Roman and Gildasian Britons apparently engaged in the faking of contemporary coins suggests that they had a sufficiently clear idea of their value.

^[16]This, of course, has a bearing on the uexata quaestio of Geoffrey's sources. I will have something to say on the matter in Books 6 and 7.

History of Britain, 407-597 is copyright © 2002, Fabio P. Barbieri. Used with permission.

Comments to: Fabio P. Barbieri