Some time around 1026 the fleet of Cnut the Great met the ships of Olaf of Norway and Anund Jacob of Sweden at a water the sources simply call the "holy river". Little else is firm. An English chronicle has the Swedes holding the field, a court poet hails Cnut as the victor, and Snorri Sturluson, two centuries on, tells of a dam broken at just the right moment. This article checks what each voice actually says about the year, the place, the winner and a famous stratagem of the Viking Age.
The Battle of the Helgeå is not poorly documented. The Swedish scholar Rune Edberg speaks of an unusually rich body of sources, and indeed English annalists, court poets from both sides, a Danish historian and an Icelandic one all have something to say. The trouble lies elsewhere. These voices do not agree, and the fullest of them is the latest.
Anyone trying to understand the battle has to keep them apart. The Anglo-Saxon Chronicle offers a few lines written from an English angle. Skaldic stanzas, the intricate praise poems composed for princes, date from the lifetime of those involved, but they are flattery by trade and survive only in later manuscripts. Saxo Grammaticus wrote in Denmark around 1200. Snorri Sturluson, finally, worked in Iceland in the thirteenth century, and nearly everything now retold as the course of the battle comes from him: the preparations, the flood, the retreat. As far as I could find, the battle has left no archaeological trace. I know of no battlefield, shipwreck or dam remains securely tied to the event.
The background to the conflict and Cnut's rule over several countries stay brief here; they are told in the pieces on Cnut the Great and Olaf the Holy. The subject here is the fight itself.
Under the year 1025 the Chronicle reports that King Cnut went to Denmark with a fleet "to the holm by the holy river". There, in Ingram's translation, Ulf and Eglaf came against him with a very large force by land and sea from Sweden. Many men died on Cnut's side, Danes and English alike, and "the Swedes had possession of the field of battle".
Three things stand out. The note names neither Olaf of Norway nor Anund Jacob, so equating it with the battle of the sagas is an inference, if a very natural one: same river name, same enemy, same years. Second, it speaks from an English standpoint and counts English dead. Third, the report least favourable to Cnut comes from England, the very kingdom he ruled. Whatever else it is, this is not propaganda on his behalf.
Ove Moberg, according to Edberg the most thorough student of the battle, names versions E and F of the Chronicle as carrying this report. He pointed to a misreading with consequences. The editor Thorpe printed "Holme" with a capital, as if it were a place name; Earle and Plummer's edition of E has lower case, and so, Moberg says, does the F manuscript he inspected in the British Library, with a definite article before the word. So the text means "the holm", a small island, and not a place called Holm.
The stanzas stand closest to events. Sigvatr Þórðarson, Olaf's Icelandic court poet, who nonetheless also composed a praise poem for Cnut, mainly supplies the frame. Moberg reads Sigvatr's Knútsdrápa as saying that Olaf attacked Zealand and Anund attacked Scania. Snorri cites Sigvatr several times, for instance on Cnut protecting his land from plunder. Laing's English rendering of these stanzas, which I could only read in summary, seems to shift the roles slightly, and I have not seen the wording of the critical edition. For the fight at the river itself, Snorri turns not to Sigvatr but to two other poets.
One of them is Óttarr svarti, who names the river in his praise poem for Cnut:
"You threw back the Swedes, battle-bold prince, there where it is called the Holy River; the she-wolf got wolf's food in plenty. You held the land, grim staff of terror, against two princes – the raven did not starve there – you are swift in counsel against the race of men."Óttarr svarti, Knútsdrápa, st. 11 ("Svíum hnekðir þú, sóknar …"), my own translation from the Old Norse as printed on Wikipedia (there after the edition Edda Snorra Sturlusonar, vol. 3, Copenhagen 1848–87); word order smoothed, kennings partly interpreted; not checked against the Skaldic Project edition
The stanza says more than its bluster suggests. It names two enemy princes, it makes the Swedes the main opponent, and it says Cnut "held the land". So Óttarr describes a successful defence, not a conquest. It places nothing in Scania or Uppland, giving only the river's name. It also mentions no flood. That alone proves little, because a praise stanza for Cnut had small reason to celebrate losses suffered to an enemy trick. Still, the dam is absent from this voice. That Snorri assigns this Helgeå stanza to Óttarr and a poem on Cnut emerges from Laing's translation.
The second poet is Þórðr Sjáreksson. In Laing's version he praises Olaf for fighting the king of the Jutes with a blue blade, and in his lines too the wolf howls over the dead. Two skalds, one river, two heroes. A poem celebrating a victory proves, first of all, that somebody wanted it celebrated.
Only Snorri gives a connected narrative, in his saga of Olaf the Holy. I follow Samuel Laing's English translation and its chapter numbers here; other editions number differently. Olaf sails to Zealand and plunders, Anund falls upon Scania, and the two meet intending to subdue Denmark. Many people submit out of fear, and the land is laid waste (ch. 155). The skald Sigvatr himself arrives in England from Rouen that summer and only with difficulty obtains Cnut's leave to travel on, because the king has forbidden merchant ships to sail until his fleet is ready (ch. 156).
When news comes that Cnut is approaching from the west in strength, both kings fall back on the Helga. According to Snorri, Cnut's force is half as large again as both opponents together (ch. 159). Further numbers come later, spoken by the kings themselves at an army assembly at Barvik: Anund had 350 ships and now has only about 100, and Olaf has 60 with the best men (ch. 161). They cannot be checked. These are figures from a thirteenth-century author, placed in the mouths of two kings, and they serve his story: the Swede loses patience, the Norwegian keeps his nerve.
The most famous scene is often shortened in retelling, so here it is with care. According to Snorri, Olaf goes upstream to the lake from which the Helga flows. His men build "a dam of timber and turf" at the outflow and dig a deep ditch that leads other waters into the lake, so that it rises. In the riverbed they lay great logs (ch. 159).
In the evening Anund's scouts sight Cnut's fleet. Olaf has the dam broken and moves his ships during the night to join Anund. With little wind, Cnut enters the harbour with only part of his fleet. The next morning the flood arrives. It carries trees that drive in among the ships, drowns many men and scatters the fleet. Cnut's great dragon ship can hardly be turned and comes under attack from the ships of both kings. Then Earl Ulf appears with his vessels, Cnut's fleet gathers, and only now does the real fight begin. Olaf and Anund decide they have won "all the victory that fate permitted them to gain" and row away east along the coast (ch. 160).
It cannot be ruled out entirely. Moberg writes that unusual interventions in the landscape are not rare in Old Icelandic prose, Snorri's saga of Olaf included. That is also the objection, since the motif fits Snorri's picture of a clever king all too well. Moberg's verdict on the episode is that the dammed river harming Cnut's ships seems "föga trovärdig", hardly credible.
The episode can be tested at three points. First, the Chronicle lacks it: it gives Cnut's losses but no cause. Second, the two stanzas on the battle that I read or saw in summary lack it too. Neither absence weighs much alone, since terse annals and praise poems rarely dwell on engineering. Third, there is no archaeological evidence that I know of. That leaves the dam as what it demonstrably is: a story first found in a source some two hundred years later, with no contemporary confirmation. Whether a real memory of flooding or driftwood lies behind it is something that cannot be clearly shown.
How scholars handle the scene is telling. Bo Gräslund uses it as an argument for a different site, while writing himself that this does not prove the flood manoeuvre happened. Moberg objects that Gräslund accepts the Icelandic tradition except for its location in Scania.
The traditional site is the Helge å in the east of Scania, Danish territory at the time, which flows out into Hanö Bay. In favour of this stand Snorri's framing story (a raid across Scania, then a retreat to the river), Sigvatr's report of Anund attacking Scania, and the voyage eastward along the coast afterwards.
What Moberg treats as decisive, however, is a report by Saxo: Cnut had the fallen buried at Åsum, near the river mouth. This, in Moberg's words, "talar avgörande", speaks decisively, for Scania. He suspects Saxo knew the area from travelling with Archbishop Absalon, who built the church at Åsum, and that the detail rests on local tradition. It is a good argument but not proof. Saxo writes nearly two centuries after the event, and a local tradition can attach itself to a suitable river after the fact. I have seen Saxo's wording only through Moberg.
Moberg also has a simple answer for the Chronicle's "holm". A map by Buhrman from 1684 shows a small island in the old course of the Helgeå, which vanished in the 1770s. The explanation is neat, but it assumes that a seventeenth-century island already lay there in the eleventh. The map cannot show that.
In 1986 Bo Gräslund proposed another solution in the journal Scandia. He moved the battle to south-eastern Uppland, to the area of the Närtuna waterway, where the watercourse may once have been called Helgå. By his own account that form is not attested; he infers it from names such as Helgö, recorded from the thirteenth century. His other arguments, which I read only in extracts through a retrieval tool: Snorri's description of the terrain fits there much better than in eastern Scania. The damming, which could not work on the lower Scanian Helgeå, would be possible there. The Chronicle's "Holm" is the village of Holm recorded in 1324. And the coins naming Cnut king of the Swedes point to a theatre of war in the Mälaren region.
It has not prevailed. Moberg answered with a debate piece in Scandia in 1987, calling the identification "långsökt", far-fetched. The holm is no place name, because the definite article precedes it. Icelandic ignorance of Scanian geography is no argument for Uppland. And Saxo's Åsum remains unexplained. In 2011 Edberg called the relocation "svår att försvara", hard to defend, because it depended on the Sigtuna coin argument, itself now heavily doubted. Most of the literature I read favours Scania. The site cannot be proven.
Each date has its own origin.
1025 is in the Anglo-Saxon Chronicle. The catch: the same Chronicle puts Cnut's Rome journey in 1031, though it took place in 1027, and according to Moberg most Anglo-Norman chronicles put the battle itself in 1031. Its dates for Cnut's foreign ventures are hardly beyond doubt.
1026 is the year most often given today, in reference works and in this site's neighbouring articles. I found no medieval source that states it outright. It is a scholarly inference: in 1985 Moberg thought summer 1026 likely, because Cnut's letter from Rome presupposes the attack already repelled, and Gräslund adopts the year citing Moberg's work of 1941.
1027 sits in Laing's Snorri translation. There a parenthesis says Sigvatr came "this summer (A.D. 1027)" from the west, meaning the year of the campaign. As far as I know Snorri does not date by the Christian era; whether the figure is Laing's I could not check against the Old Norse, because the retrieval tool could not reach heimskringla.no. The Dictionary of National Biography cites Icelandic annals for 1027; I have not seen them myself. Moberg uses 1027 for something else, a second campaign against the Mälaren region.
The one fixed point is Cnut's letter from Rome. It is undated, but his visit for Emperor Conrad's coronation on 26 March 1027 places it. In it he announces that he will go home by way of Denmark to make peace and a firm treaty with the peoples who had wanted to deprive him of his kingdom and his life. This sounds like the war of the Helgeå, but alone it establishes no sequence: it could look back on the battle or on a conflict still open. The most cautious wording is therefore around 1025 to 1027, usually placed in 1026.
The Chronicle names Cnut's opponents as Ulf and Eglaf. Who they were bears on what actually happened. The common reading takes Ulf as the earl and Cnut's brother-in-law, husband of his sister Estrith, and Eglaf as Ulf's brother Eilaf, which would mean a rebellion within his own circle. According to Wikipedia, M. K. Lawson regards this identification as common but uncertain; I have not read Lawson himself.
Snorri tells the opposite: Ulf rescues Cnut in the battle. Saxo in turn puts Ulf on the enemy side, and Moberg suspects Saxo confused the brother-in-law with a Swedish commander of the same name. In Scandia in 1985 he had set this out, identifying the two as the brothers Ulf and Eilif Ragnvaldsson, sons of a Swedish magnate and thus Anund's fleet commanders. The reasoning: Cnut's brother-in-law was with him at Roskilde straight after the battle, and Eilif later ranked among his closest advisers. Neither fits a turncoat.
The echo survives in Snorri's best-known scene (ch. 163). Cnut and Ulf play chess at Roskilde, the king makes an illegal move and the earl overturns the board. Cnut jeers, "Runnest thou away, Ulf the coward?" Ulf retorts, "Thou wouldst have run farther at Helga river, if thou hadst come to battle there", and adds that the king did not call him a coward when he came to his help "while the Swedes were beating thee like a dog". Soon after, Cnut has him stabbed in the church. It is a literary scene shaped by later knowledge, not a record, but it is revealing: even in Snorri, the Helgeå does not appear as a shining victory for Cnut. Snorri, incidentally, has Cnut ride to Roskilde the day before Michaelmas, while other accounts give Christmas 1026 for the death. This is unresolved.
The answers can be laid side by side. The Chronicle: the Swedes held the field. Óttarr: Cnut threw back the Swedes and held the land. Snorri: a surprise success for the allies, then withdrawal before the intact main force, with no pursuit by Cnut. The Dictionary of National Biography once stated that the Danes were worsted. Moberg decides for Cnut, relying on Óttarr, and Gräslund agrees. Thomas Lindkvist, according to Edberg, also credits Cnut. Edberg and, by his account, Axelson (1955) before him read the sources rather in Anund's favour. Edberg calls the whole affair a possible Pyrrhic victory for Cnut: even had he won militarily, it counted for nothing in the long run.
The voices exclude each other less than it seems, since they measure different things. Whoever keeps the field can claim the win, and that is what the Chronicle reports. Whoever holds the land and watches the invader leave can too, and that is what Óttarr sings. A fleet may suffer heavy losses and still win the campaign. That reading is interpretation, not tradition.
Only the broad outcome is firm. The grab for Denmark failed, Anund went home, and Olaf lost his ally and soon his throne.
In Snorri the alliance comes apart at Barvik and shortly after. Anund's men slip away home, a further council is held in autumn, the Swedes do not want to be caught by frost, Anund returns with his army, and Olaf is left alone (chs. 161, 164). In 1028 Cnut came to Norway and drove Olaf out, in Snorri almost without a blow. The road on to Stiklestad in 1030 has its own article.
Whether Cnut then also ruled in Sweden is the debate's real long-range effect. In his letter of 1027 he calls himself king of all England, Denmark, the Norwegians and part of the Swedes, and there are coins with the legend CNUT REX SW that have been assigned to Sigtuna. In 1910 Lauritz Weibull inferred from this a rule over part of Sweden, but according to Moberg this won little hearing among scholars, partly because the letter survives only in a late Latin translation and scholars such as Lappenberg and Liebermann took the phrase for an addition. Moberg himself reads letter and coins as evidence of a second campaign in 1027, separate from the Helgeå.
Recent numismatics is more cautious. Kenneth Jonsson, according to Edberg, judges the legend almost certainly a meaningless formula. Brita Malmer traces all specimens to a single die from a Danish die chain, probably from Lund, and reads it as propaganda, not as a trace of power in Sigtuna. Edberg accordingly rejects the idea of Danish rule over Sigtuna. For the Helgeå this matters, because the strongest supporting argument of the Uppland theory falls away.
Where was the Helgeå of the battle? It is usually identified with the Helge å in eastern Scania, then Danish. In 1986 Bo Gräslund proposed a river of the same name in Uppland. The theory won little support. No find has refuted it; the objections rest on the texts.
What year was the battle? The Chronicle says 1025, most accounts say 1026, and Laing's translation of Snorri gives 1027 for the same summer. None of these can be clearly established.
Who won? That remains disputed. The Chronicle gives the Swedes the field, Óttarr praises Cnut. In the outcome Cnut kept Denmark, the alliance dissolved, and Olaf lost Norway two years later.
Did Olaf really dam a river? That is first told by Snorri in the thirteenth century. It cannot be ruled out entirely, nor can it be shown, and Moberg finds it hardly credible.
Why is the river called "holy"? The name is already there in the oldest texts: á hin helga in Óttarr, "holy river" in the Chronicle's English rendering. I could not establish where the name comes from.
Cnut the Great · Olaf the Holy · The Battle of Stiklestad · Óttarr svarti · Snorri Sturluson · Saxo Grammaticus · Emma of Normandy · Longships
Cnut the Great · Olaf the Saint · Stiklestad – the fall of Saint Olav · The Battle of Svolder: What Skalds, Adam of Bremen and Snorri Really Know About the End of Olaf Tryggvason.
All History & Finds articles →
Five names stand in a single sentence of the Germania: Harii, Helvecones, Manimi, Helisii, Naharvali. Four of them come …
Read more →Five names stand in a single sentence in Tacitus, and one of them belongs to the Helvecones. After that, the record …
Read more →A whole people burned its villages and set out to find a new homeland, and ran straight into Caesar's campaigns: the …
Read more →The Helvii lived between the Rhone and the Cevennes, in what is now the Ardeche. They were small, unspectacular and …
Read more →