Peredur: Flower of Warriors, Candle of Knights

Peredur: A Grail Romance? Part V

“Historia Peredur vab Efrawg …. is a microcosm of almost all the problems which can be found in early Welsh prose literature. It is virtually impossible to unravel the manuscript tradition, and because of this it is difficult to settle upon an ‘authoritative’ text; the narrative structure has been analysed in different ways, and widely differing suggestions have been advanced as to the underlying structure which unifies the romance; and the relationship of the story to other versions, notably Chrétien’s Conte du graal and its continuations, appears less than straightforward.”1

The majority of studies of the Middle Welsh tale Peredur, son of Efrog tend to focus on its relationship with Chrétien de Troyes Perceval, or The Story of the Grail. Chrétien’s eloquent writing style and enthralling storytelling has led to a persistence in priority of the French text with the former dismissed as a pale imitation of the latter, Peredur seen as a poorly structured adaption from the French. However, this conclusion fails to give merit to the original tale as can be deciphered from careful analysis of the manuscript tradition instead of rushing into comparisons with the French.

The 'standard' (full or long) version of Peredur son of Efrog, familiar to most of us as found in the collection of tales in The Mabinogion, is found in two manuscripts: the “White Book of Rhydderch” (Aberystwyth, National Library of Wales, Peniarth 4–5, dated c. 1350); and the “Red Book of Hergest” (Oxford, Bodleian Library, Jesus College 111), which is dated slightly later to between 1382 and c.1400. The versions of Peredur found in these two manuscripts are very similar. 

It is certainly true that the popular version of ‘Peredur’ as found in the Mabinogion collection taken from the Red Book of Hergest contains episodes that mirror sections of Chrétien’s tale. However, there are earlier versions of Peredur in which the claimed French influence is noticeably absent from certain sections. 

Valley of the Mills (Alan Lee)

Narrative Structure
In Peredur the hero’s initial journey transforms him from a country bumpkin into a skilful knight; following on from the procession of a bleeding spear and a severed head at his uncle’s castle he embarks on a further series of adventures in which he gains the love of Angharad (Golden-Hand), before winning the admiration of the empress of Constantinople, with whom he rules for fourteen years; finally he embarks on a series of further adventures that leads him to the Fortress of Wonders, where the procession is explained and he gains vengeance on the witches of Caerloyw. 

It was noted in Part III - Peredur: From Caerllion to Constantinople that the so-called standard version of the tale, as found in the Red Book and White Book, contains logical breaks in the story using a standard formula with each section beginning and ending at King Arthur’s court at Caerllion:

1. The first section begins in the Welsh Woods where Peredur leaves his mother to become a knight, goes to Arthur’s court and ends after Peredur befriends Gwalchmai and they return to Caerllion.

2. The next section commences with "The first night Peredur came to Caerllion to Arthur's court" when he met Angharad Law Eurog (Golden Hand). After a series of adventures, he returns to Arthur's court as the ‘Mute Knight’ and wins the love of Angharad.

3. The next section again begins "Arthur was in Caerllion ar Wysg" and ends with Peredur, now known as the ‘Knight of the Mill,’ …. ‘and there did he stayed with the empress fourteen years’. There is a reason why the formula of returning to Arthur’s court is not used here as we will see below.

4. The final section again begins with the same formula "Arthur was in Caerllion ar Wysg" when a black, curly-haired maiden arrives at Arthur’s court on a yellow mule, and after a series of further adventures culminating in the death of the witches of Caerloyw, ends with “And that is what is told of the Fortress of Wonders” breaking the previous links returning to Arthur’s court at Caerllion suggesting closure of the episode of the black, curly-haired maiden, but perhaps leaving the tale open for further addition opening with “Arthur was in Caerllion”.

One can imagine oral delivery of the tale to a court audience with the people knowing a new adventure commenced when the storyteller opened with the words, "Arthur was in Caerllion”.

As we can see the tale has four logical sections divided by events, starting or returning, at Arthur’s court at Caerllion. The exception to this being the very opening episode in the Welsh woods, and the very end of the standard (Mabinogion) version at the Fortress of Wonders and the killing of the witches of Caerloyw.

One exception to the ‘Caerllion rule’ is the third section which ends at Constantinople where Peredur stayed with the Empress for 14 years; the text ends with “And thus ends the Progress of Peredur ab Efrawg”. This ending is very similar to the closure of the Branches of the Mabinogi which use a similar phrase to close the respective branch, clearly implying that this was the end of the original tale of Peredur, apparent by not ending back at Arthur’s court at Caerllion with the final section with the black, curly-haired maiden being a bolt-on episode to align the procession of the bleeding lance and the severed head with the Grail procession in Chrétien de Troyes Perceval as stated in Part IV – Peredur: The Procession. 

However, learned scholars, more qualified than myself, have divided the tale into three parts. For example, Mary Williams2 saw the tale with a tripartite structure:

i) from the beginning to the end of the drops of blood in the snow episode,

ii) the story of Angharad Golden-Hand, Peredur’s adventures in the Round Valley, the Black Opressor, the Sons of the King of Suffering, and the Addanc, the Serpent with the Ring, The Miller and the Empress of Constantinople,

iii) from the arrival of the black, curly-haired maiden to the death of the witches of Caerloyw.  

Williams was surely correct in arguing that the variations in the first part and the last part indicates they were the work of different authors. Williams saw the events described in the last section being added at a later date as they were necessary for the explanation of the first part, and copied directly from the French.

Brynley Roberts also sees three sections to the tale of Peredur which are essentially the same as Williams:

Part A - departure from his mother, to King Arthur's court, the lame nobleman, the procession

Part B - Anghard Golden-Hand, the empress of Constantinople

Part C - opens with the black, curly-haired maiden, ending with death of the witches of Caerloyw.3

The manuscript tradition confirms the separation between these three sections; Roberts claims that in two of the four manuscripts which preserve the tale of Peredur, the section breaks are clearly marked by introducing each with a large capital. These capitals are only used to mark the beginning of these three sections and do not appear anywhere else in the narrative, indicating the importance of these break points.4

We can shed some clarification on the situation when we consider that copies of the tale of Peredur can be found in two further medieval manuscripts; while parts A and B are present in all four manuscripts, Part C is not. 

In addition to the ‘standard’ version found in the White Book of Rhydderch and the Red Book of Hergest there is also two versions that are designated as ‘fragments’ incomplete or deficient of some of the text. One is found in MS Peniarth 14, which breaks off mid-sentence during Peredur's visit to his second uncle, and an earlier version found in MS Peniarth 7, which is deficient of the opening section and ends with Peredur ruling with the empress of Constantinople. 

Notes and References:
1. Ian Lovecy, Historia Peredur Ab Efrawg, in Arthur of the Welsh, Wales University Press, 1991, pp.171.
2. Mary Williams, Essai sur la composition du roman gallois de Peredur, 1909.
3. Brynley F. Roberts, ‘”Peredur Son of Efrawc”: A Text in Translation’, 2000, Arthuriana 10.
4. Ibid.