1
Then Song of Roland is reputedly one of, if not the actual, oldest of the medieval French chansons de geste, or songs of deeds.
The Penguin Classics version I am using is a translation by Dorothy L Sayers. I wonder how much she relished the gory details: the brains dribbling from Roland’s ears, the bowels gushing out of Archbishop Turpin’s side, the scalp cloven to the brains etc etc.
The Song first made its appearance in this form in the thirteenth century, shortly after the First Crusade. This is important because, although it is based on an actual incident in 778, the twist in the chanson de geste version is very important.
The actual incident concerned King Charlemagne, and his being approached by Saracen rulers in Spain for help in dealing with a mutual Saracen enemy. He agreed and entered Spain with them, conquered two major cities, and was besieging Saragossa, when he was called away. He left Spain via the Pyrenees pass of Roncevaux. Here his rearguard was attacked by Basques, who slaughtered them to a man, and left with their goods. It was Basque territory.
The version in the geste has the Saracens the enemy throughout, and the attack on the rearguard an agreement between a renegade Frenchman and the Saracens. The composer of the piece, like his audience, knew next to nothing of Islam, and so we come across some absurdities, some crazy assumptions.
Dorothy L Sayers was well aware of this; it was, as she says, an ignorance shared by Islam of the Christian faith, that the Thousand and One Nights shows similar crazy assumptions about Christian worship.
It is very important to remember that Count Roland of Brittany was the nephew of Charlemagne, and that it was rivalry with his step-father – as in the old folk tales – Ganelon, that caused his death.
The Song of Roland consists of 291 ‘laisses’, that is, stanzas, of varying length. They all follow the same strict metrical pattern, however: this is syllabic verse, and each line is strictly ten syllables in length. The final syllables in every separate stanza are linked through assonance; each stanza has its separate rhyme scheme. All the lines are strongly end-stopped – there are no enjambments. The main stresses of each line fall on the fourth and tenth syllables. As you can see, this was a very tight construction – and to keep it up for 4000 lines was quite a feat.
The first appearance of the chanson was as one of many legends and tales that circulated on pilgrim trails, and in local courts and gatherings. At just about 4000 lines it required quite a feat of memory. And so the tale is structured in such a way, with parallels, repetitions of motifs, events etc, to enable the reciter to be able to tackle the rather detailed subject matter fully, and with skill.
Whether it is a ring-structure or not, I cannot as yet tell. What do you think? -
There are a great many paralleled motifs etc throughout. They are not, however, evenly divided between each half. There are indeed two distinctive halves:
- the first gives us the build up of the Saracen plot,
- and the second, after Roland’s death, the resolution, through the intervention of Charlemagne.
This intervention of Charlemagne is paralleled with the arrival of the Saracen Emir. The first battle consists of a run of descriptions of hand-to-hand combat where Saracens die; this, after the intervention of King Marsilion, is paralleled with a similar run but where the French soldiers die.
When Ganelon receives the commission from Charlemagne to act as envoy to the Saracens, he is offered the King’s glove, but it is dropped. When Roland receives commission to act as rear guard for the troops returning to France through the pass, he is given Charlemagne’s bow, and there is no incident. There are many such patternings in the piece.
The storyline is roughly as follows:
Saracen King Marsilion in Saragossa wanted Charlemagne and his troops out of Spain. He asked for a plan.
The plan was to promise that all the Saracens would convert to the Christian faith because they could not keep the up fight with Charlemagne any longer. They would send sureties, and agree to meet with Charlemagne in his court at Aix, in one year’s time. We find there was a reason for this delay, but later in the tale.
An envoy was sent out to Charlemagne. Roland said No, to the deal, but his step-father Ganelon said Yes. His reasons were more convincing: and Roland had a reputation as a hot-head.
Ganelon was sent out to test the Saracens. He met with Blancandrin, then King Marsilion, and in a long discussion came to an agreement that he would say they are trustworthy, if they attacked and killed Roland for him.
Back with Charlemagne Ganelon wangled Roland into being the rearguard of their exit from Spain.
They exited through the pass; at the last minute Roland and his rearguard were set upon by a huge Saracen army. Roland had a horn to blow in case of need, but refused repeatedly to use it.
12 Saracens boasted they would take Roland singlehandedly. In the fight all but 2 were killed, in single combat. Roland had 12 Peers of the French nobility as backup.
When the battle was joined fully the French won initially; when King Marsilion joined the fray it all turned against the French.
At last Roland blew the horn. Doing so exhausted him, weakened him. Charlemagne heard the distant sound, and Ganelon tried his best to dissuade him from turning back; by this time the battle was going very badly for the French.
Roland blew it a second time, after his second-in-command Oliver, was killed. The third time he blew it, the Archbishop Turpin was killed.
Although the Saracens were routed by then, Roland had received his fatal blow, and died. He broke the horn killing the man who attacked him as he lay dying.
Shortly afterwards Charlemagne arrived on the scene. He set troops to guard the bodies, then set off after the Saracens, and slew most.
King Marsilion was badly wounded: he had lost a hand, and died slowly. Not before reinforcements, sent for sometime previously to this, arrived from the East.
Charlemagne returned to the pass, and funeral and burial rites were recited for the slain. As he was about to depart the renewed Saracen host came upon them. The preparations for the battle were the same as the previous battle.
The fight was long and very fierce; it was decided in a hand-to-hand fight between Charlemagne and the Saracen Emir. Charlemagne was wounded, but slew his adversary. At which the Saracen host fled. They were chased, and slain: ‘No prisoners!’ Those who escaped went to Saragossa. Charlemagne besieged the city, and took it.
He eventually returned to Aix. Ganelon was tried, his case decided by hand-to-hand combat, and he lost. He was committed to torture, and death. Charlemagne’s troops returned to their homes.
St Gabriel appeared once more to Charlemagne, asked him to rouse his troops, there was another fight to be fought, at which:
“God’, said the King, ”how weary is my life!”
He weeps, he plucks his flowing beard and white.
This is the end. You do feel for him at this point!
And so –
You could say: the Song starts with Ganelon’s betrayal, and ends in his death.
You could also say that Ganelon’s betrayal played a major part in the ‘turn’ of the Song – I would place the ‘turn’ on the blowing of the horn: it is from here that Roland is weakened as much by his efforts in blowing as the battle itself.
And Ganelon makes efforts to dissuade Charlemagne from responding to the horn. He was immediately arrested, and his trial and death set from there.
If this is so, then that is a major classic ring-structure turn.
But what about the opening scenes where King Marsilion and Blancandrin cook up their tactic?
If that is the opening of the ring then you could say that delayed promise is the major element:
- the delay of 1 year before joining Charlemagne at Aix – and to allow the armies from the East to arrive,
- the delay of Roland in blowing his horn: Oliver repeatedly asked him before the battle began (all those deaths could have been avoided if help had arrived in time ie when Oliver asked him – but then, Roland was known as a hot-head), and
- Charlemagne’s delay in executing the next battle that St Gabriel demands of him.
This one appeals most because the audience’s empathy is called upon to a greater extent, in order to recognise, and to emphasise the structure.
2
One part that does especially fascinate me about these structures, are the degrees of perception, of listening awareness, demanded of the audience. One can imagine listeners competing to be the best and quickest to spot these structural elements; and to engage in and enjoy the discussions and counter discussions, around them. These would suggest an early, and maybe kinder, form of critique of the pieces.
Once the opening is set then the reciter can lift a strand from it to be the main element of the ‘turn’; and similarly at the ending. The piece can be ‘spun’ for each performance – although keeping all the elements, the re-emphasising of minor themes over the usual major ones could be quite possible.
In this respect, remember in Beowulf, the second ‘funeral’ section is a restructuring of a known tale that we also have, The Fight at Finnsburgh. In Beowulf we see the episode from underneath, as it were: not the glory of the victors, but consequences for the losers.
This, of course, is nothing new: Euripides was doing it in 415 BCE, with his Trojan Women, a drama about the consequences of the loss of the Trojan war on the women of the Trojans. The warriors, of course were out of it, dead, but the women…? Queen Hecuba a slave; to a king maybe, but still a slave. She would maybe have been one of the luckier ones: young women of the enemy: I would not expect much quality of life in the aftermath (to put it mildly).
Of course, there is another possibility: that by the time, 1200 or so, the Song was composed, the structuring by ring had been forgotten, lost or superceded, and what the composer of the piece was left with were half remembered constructive features.
It is certain the Song is constructed in two halves; it is certain there is a ‘turn’ – what is not certain, is that it is constructed as a chiasmus. There does not seem to be a build up and retreat using the same devices, motifs etc.
All the parallelisms and repeated patternings seem to be on a linear basis, rather than circular.
Alternatively the dynamic of the piece could be so well incorporated into the text, and its many details of characters and events, that it is not readily accessible to the reader. As I said above it could well be solely the province of the reciter to bring out structures, whether linear or circular.
This is appealing, because it hands back the work to the people – this is the province of the professional storytellers. And it is ground only recently being re-explored by Jack Zipes, and workers with modern storytelling.


