Rollo

Only in a charter dated March 14, 918, does the name Rollo appear for the first time in a contemporary source. King Charles the Simple confirms therein land “that we have given to the Normans on the Seine, namely Rollo and his companions, for the defense of the realm.”

Before this date, the Frankish sources are entirely silent about him. Of a man who is regarded as the founder of Normandy, only three brief mentions exist from his own lifetime. The rest of the story told about him was not written down until generations later.

Three mentions, and none says where he came from

The contemporary body of evidence can be listed in full.

First, the charter of Charles the Simple dated March 14, 918. It refers to an earlier grant of land and names Rollo by name as recipient, together with his companions.

Second, Flodoard of Reims for the year 925, in connection with the fighting against King Raoul over the fortress of Eu.

Third, Flodoard for the year 928, in the dispute between Charles the Simple and Heribert of Vermandois, where Rollo releases the hostage Odo only after Heribert pledges himself to Charles again.

After that, the sources fall silent. His death can only be narrowed down by the fact that Flodoard names his son William as leader of the Normans in 933. Rollo therefore died between 928 and 933. His burial place is the Cathedral of Rouen.

And now the point on which everything hinges: none of these three passages says a single word about his origin. Flodoard is interested in only one thing – that Rollo is a pirate. Whether Dane, Norwegian, or something else is not worth a line to the only contemporary who writes about him.

A title he never held

Here a distinction is worth making that popular accounts regularly overlook.

Rollo demonstrably held no title. The charter of 918 gives him none. In no document from his lifetime does he appear with one. It is only his son William and his grandson Richard who use comes, consul, and princeps in charters.

Flodoard does, however, describe him as princeps Nortmannorum. This is not the holding of a title but the word choice of a chronicler – and the proof lies with Flodoard himself: he uses the same word for Ragenold, the leader of an entirely different Viking group on the Loire. For him, princeps simply means “leader.”

The later designation “Duke of Normandy” is therefore doubly untenable. The title dux does not appear until the 11th century, and Rollo held no title at all.

The name behind the Latinization

“Rollo” is a Latinized form. Heimskringla reports that Rolf the Walker went to Normandy and ruled there – which is why Rollo is generally regarded as a Latinization of the Old Norse name Hrólfr.

Linguistic support for this comes from Saxo Grammaticus, who in the Gesta Danorum renders the name Hrólfr as “Roluo.” There, however, a different man is meant – the legendary Danish king Hrólfr kraki, not the Norman. Saxo thus demonstrates the sound development, not the person. This is precisely why the argument holds: it does not depend on an identification.

A minority of scholars instead propose a Latinization of Hrollaugr – phonetically this name is in fact closer to “Rollo.” Hrólfr itself is a short form of Hróðulfr, composed of hróðr (“fame, praise”) and ulfr (“wolf”). The German name Rudolf is related.

His byname Gǫngu-Hrólfr, “Hrólfr the Walker,” is explained by Snorri Sturluson in the Haralds saga ins hárfagra, chapter 24, as owing to the man being so large that no horse could carry him – he walked everywhere on foot. This is an origin anecdote, not an established fact, and it comes solely from the saga tradition.

A baptismal name is recorded only by Dudo of Saint-Quentin: Rollo is said to have taken the name Robert at baptism. This information, too, rests on no second source.

Dudo: a house history, written a century later

Richard I of Normandy commissioned Dudo of Saint-Quentin to record “the customs and deeds of the land of the Normans” and the law of his grandfather Rollo. The resulting work, De moribus et actis primorum Normanniae ducum, was completed between 1015 and 1026 – roughly a century after Rollo’s death.

Everything that comes from Dudo about Rollo’s origin and youth is therefore late house tradition, not contemporary testimony.

In Book II, chapters 2 to 4, Dudo makes Rollo the son of a man with possessions in “Dacia.” After the father’s death, the king has Rollo’s brother Gurim killed and drives out Rollo himself.

Dacia is a play on words

Dacia designated, in ancient geography, the region north of the lower Danube, not Denmark. Dudo links the two by way of a learned false etymology: Daci like Dani.

The Danish origin claim – the main pillar of the entire Danish trail – therefore rests in part on an antiquizing play on words.

This fits the rest of the work. Dudo saturates the text with allusions to Virgil’s Aeneid and compares the Norman settlement to the Trojan founding of Rome. He employs prophetic dreams that foretell Rollo’s conversion to Christianity and his legitimacy as a ruler.

This is not sober chronicle. It is literary program, and it can be recognized as such already at the level of the source itself.

Snorri and the Norwegian counter-tradition

In Heimskringla, King Harald Fairhair banishes Hrólfr from Norway – the passage stands in the Haralds saga hárfagra, chapter 24. Heimskringla and the Orkneyinga saga from the 13th century remember him as Hrólf the Walker and make him the son of the Norwegian jarl Rögnvaldr Eysteinsson of Møre.

Snorri has Hrólfr conquer Normandy only after raiding expeditions across the Hebrides, the Irish Sea region, and Francia, and settles many Norwegians there in his narrative.

This version stands in open contradiction to Dudo’s Danish origin story. Both versions come from late sources, and neither can be confirmed contemporaneously.

The origin dispute: Denmark or Norway?

The Danish trail. Dudo explicitly claims Danish origin. Benoît de Sainte-Maure, elsewhere in his work, likewise has Rollo appear as a Dane, although he has him born in a place called “Fasge” – interpreted either as Faxe on Zealand or as Fauske in Norway – and has him flee to “Scanza,” that is, Scania.

The Norwegian trail. The first explicitly Norwegian statement comes from Goffredo Malaterra in the 11th century. William of Malmesbury expresses a similar view in the 12th century and calls Rollo “of noble descent among the Norwegians.” The Historia Norwegie has a “Rodulfus” from Rögnvald’s retinue take Rouen and explains the byname “Gongurolfr” from his body size.

A third thread. The Welsh “Life of Gruffudd ap Cynan” from the 12th century makes Rollo the younger of two brothers of the first king of Dublin – a third version, scarcely reconcilable with the other accounts.

Even the mother is disputed. Snorri names Hild as the mother of “Rolf and Thorer,” while the Orkneyinga saga instead names Ragnhild, daughter of Hrolf Nose.

The scholarly camps. The Danish historian Johannes Steenstrup opened the modern debate by pointing to Dudo’s explicit statement of a Danish origin. British, French, and Norwegian scholarship has since mostly followed the Norwegian-Icelandic tradition, while Danish, Swedish, and German scholarship tends to follow Dudo. David C. Douglas in 1942 combined statements from the Landnámabók with the Planctus on William Longsword. Robert Helmerichs worked out the early title designations in 1997 and, in 2002, considered that the Icelandic genealogy might be constructed, since it gives Hrólfr a brother named Hrollaugr – a name phonetically closer to the form “Rollo” than Hrólfr itself.

The third party: the place names

So far, two late chronicler traditions have stood opposed to each other. There is, however, a category of source that depends on neither Dudo nor Snorri: the Scandinavian place names of Normandy.

Endings such as -tot, -bec, -fleur, -dalle, or -hogue derive from Norse words, and Norse personal names appear as components of place names as well. These forms can be examined linguistically for their origin and dialect features – and they have stood on the map since the 10th century, independent of what chroniclers later wrote.

They have been analyzed since Jean Adigard des Gautries and Lucien Musset, most recently by Elisabeth Ridel. The findings do not settle the question of Danish or Norwegian. They point to a mixed picture – and to the possibility that some of the settlers came by way of England and Ireland, not directly from Scandinavia.

This shifts the question itself. “Danish or Norwegian” is possibly the wrong alternative for a group that assembled itself along the way. The question of origin remains undecidable to this day.

A treaty without a contemporary trace

The Treaty of Saint-Clair-sur-Epte is regarded as the founding act of Normandy. It has not survived in the original – and that is the least troubling piece of information about it.

No contemporary source mentions it at all. Not the ceremony, not the location, not the year. The charter of 918 refers exclusively to an earlier grant of land, without treaty, without location, without date.

Even the year 911, which appears in every schoolbook, comes solely from Dudo. So does the boundary drawn from the Epte to the sea, the baptism by Archbishop Franco of Rouen, the oath on relics, and the marriage to Gisela.

And so, too, does the most famous scene of all. In Dudo, Rollo refuses to kiss the king’s foot as a sign of submission and has one of his men do it instead. The man lifts the foot so high to his mouth that the king falls over backward. The Normans laugh.

It is the best-known Rollo anecdote of all – and it stands in the very same programmatic work that also contains the Trojan parallel and the prophetic dreams.

Marriages, baptism, and a near-contemporary testimony

A near-contemporary testimony is provided by the Planctus de obitu Willelmi, a lament composed shortly after 942 for Rollo’s son William Longsword. It calls William in orbe transmarino natus – born beyond the sea – son of a father who persisted in pagan error and a Christian mother.

This is the most reliable information we have about Rollo’s circumstances, and it says only two things: the father was still a pagan at the time, and the son was not born in Francia. Ordericus Vitalis dates the baptism to 912, performed by Archbishop Franco of Rouen.

William of Jumièges reports that Rollo took Poppa, captured at Bayeux, as wife more danico – according to Danish custom, that is, without a church wedding – then repudiated her for a marriage with Gisela more christiano, and remarried after Gisela’s death.

Whether Gisela even existed is disputed. Dudo calls Poppa the daughter of Count Berengar, while the Annales Rotomagenses instead call her the daughter of an otherwise unknown Guy of Senlis. Pierre Bouet considers the Gisela marriage less fanciful than often assumed, since a daughter of this name who died young may indeed have existed.

The sources also contradict one another regarding Rollo’s religion at the end of his life. Adémar de Chabannes claims that shortly before his death Rollo performed human sacrifices to appease the pagan gods while at the same time making gifts to the churches of Normandy. The Fécamp chronicle, by contrast, insists that he remained faithful to his last breath – a counterstrike by Norman house propaganda against the more inconvenient version.

The bones of Fécamp

On February 29, 2016, researchers opened the lead ossuaries of the graves of Richard I and Richard II in the abbey of Fécamp to obtain teeth for examination. The aim was finally to settle the question of Norwegian or Danish origin scientifically – via the DNA of the descendants.

The result was sobering, and doubly so.

Radiocarbon dating showed that the skeletons cannot have anything to do with Rollo’s descendants: one dates to 250 to 300 BC, the other to around 700 AD. Whoever was buried there, the Richards were not.

And because of the age and manner of storage, no usable DNA could be obtained at all. The question of origin remains open, and the only scientific attempt to answer it did not even find the right bones.

Reception: Vikings and the Gisela legend

The television series “Vikings” makes Rollo the brother of Ragnar Lothbrok. Contrary to this popular image, he was not his brother but lived several generations later.

Something else is noteworthy. The series chooses the very same figure of legitimation as the Norman house history itself – marriage to a king’s daughter, called Gisla in the series, Gisela in Dudo.

It is exactly the motif that stands weakest under source criticism, because it comes almost entirely from Dudo’s programmatic work. Modern fiction and medieval house propaganda reach for the same scene, and for the same reason: it turns a pirate into a son-in-law.

Sources

Contemporary. Charter of Charles the Simple dated March 14, 918, edited in: Recueil des actes de Charles III le Simple, ed. Ferdinand Lot and Philippe Lauer, Paris 1949, no. 92. Flodoard of Reims, Annals, for 925 and 928 (MGH SS 3); English excerpts in Elisabeth van Houts, The Normans in Europe, Manchester 2000, pp. 42–51. There is nothing more.

Near-contemporary. The Planctus de obitu Willelmi, shortly after 942, attests to the origin and baptism of Rollo’s son William Longsword.

Norman house history. Dudo of Saint-Quentin, De moribus et actis primorum Normanniae ducum, Book II, chapters 2 to 4, completed between 1015 and 1026 – source for origin, expulsion, treaty, baptism, foot-kiss, and Gisela. William of Jumièges on the marriages. Ordericus Vitalis on the baptism in 912. Adémar de Chabannes and the Fécamp chronicle contradict each other regarding religion at the end of his life.

Icelandic-Norwegian tradition. Heimskringla, Haralds saga hárfagra, chapter 24, and the Orkneyinga saga, both 13th century.

Further medieval chronicles. Goffredo Malaterra (11th c.) and William of Malmesbury (12th c.) for the Norwegian origin. Historia Norwegie, Benoît de Sainte-Maure, and the Welsh “Life of Gruffudd ap Cynan” (12th c.), each with a diverging version. Annales Rotomagenses on Poppa’s paternity.

Scholarship. Johannes Steenstrup opened the origin debate. David C. Douglas in 1942 combined the Landnámabók and the Planctus. Robert Helmerichs, “Princeps, Comes, Dux Normannorum: Early Rollonid Designators and their Significance,” Haskins Society Journal 9 (1997), pp. 57–77, on the titles; in 2002 on the Icelandic genealogy. Pierre Bouet reassessed the Gisela marriage.

Place names. Jean Adigard des Gautries, Les noms de personnes scandinaves en Normandie de 911 à 1066; Lucien Musset; Elisabeth Ridel on Scandinavian traces in the toponymy and vocabulary of Normandy.

Natural science. Examination of the lead ossuaries in the abbey of Fécamp on February 29, 2016: radiocarbon dating showed that the skeletons cannot be Richard I and Richard II; no usable DNA could be obtained.

Note on research. The German-language Wikipedia states that Flodoard does not name Rollo “by name.” This is not correct – Flodoard names him by name for the year 925.