Picture the Viking Age and you see ships, swords and rune stones, rarely libraries. There is a reason: the great narrative works about the period, the sagas and the Edda, were written down centuries after the events, on parchment, in the Latin alphabet, in Iceland. What lay in between was a culture of speaking, singing and remembering. This article asks what we can document about that oral world, what is merely plausible, and where scholars still disagree.
Talk of oral tradition tends to carry a romantic picture: a wise elder by the fire passing on, unchanged, what his grandfather said. The sources do not support that picture, at least not in its pure form. "Oral" means only that knowledge was not stored on a written medium but carried in people's heads and renewed in performance. It says nothing about how exact or how free that process was.
The distinction from illiteracy matters too. The Viking-Age North had writing: runes. What it lacked was a book culture in the medieval sense, meaning manuscripts that recorded stories, laws or poems at any length. As far as we can see, runes served other purposes. "The Vikings could not write" is wrong; so is "the Vikings wrote sagas".
A further problem concerns our sources themselves. Almost everything we know about the oral culture of the North comes from written testimony produced by Christians between the 12th and 14th centuries. We observe the oral world through the window of its own writing-down. What changed in the process can rarely be said with confidence, and the following sections return to this again and again.
The skald (Old Norse skáld, "poet") composed skaldic poetry, one of the two main forms of alliterative Old Norse verse. The other is Eddic poetry. Skaldic poems were traditionally composed in praise of rulers; there are also single stanzas called lausavísur ("loose verses"), together with insult, love and occasional verses. The characteristic metre is dróttkvætt, the "court metre", which mostly combines internal rhyme with alliteration.
Skaldic verse is intricately built. It uses kennings, figurative circumlocutions whose solution requires knowledge of mythology, and heiti, poetic substitutes for ordinary words. Modern readers often find the result obscure and ambiguous. The original audience knew the conventions and could solve the riddles. That the poetry aimed at listeners who mastered the rules is a strong indication that it was conceived as performance.
The earliest named skald whose verses survive, Bragi Boddason "the Old", was active in the first half of the 9th century and was Norwegian. His verses are already technically accomplished, which suggests a longer prehistory of the form. What that prehistory looked like cannot be documented. Origins in Norway or around the Baltic are scholarly conjecture, as are possible influences from Irish or Latin poetry.
From the second half of the 10th century on, according to the usual account, all known court skalds came from Iceland or Orkney; the last prominent Norwegian skald was Eyvindr skáldaspillir. Icelanders thus dominated a craft anchored at Norwegian courts, a hint that poetry was also a portable skill that opened doors to rulers. Some skalds, Egill Skallagrímsson among them, are the subject of sagas of their own; see our article on Egill.
The sagas prize unprepared composition. The famous example is Egill's Höfuðlausn ("Head-ransom"), which the saga says he composed in a single night to buy back his life. At the same time, spontaneous composition was probably not the rule in practice. Dróttkvætt, with its demands on rhyme and syllable count, is a poem one thinks about. Whether skalds prepared stanzas in advance, rehearsed or revised them, we do not know. Saga authors had an interest in showing their heroes as masters of the moment; that is literary shaping, not a transcript.
Instruments are not attested for skalds. Some scholars have speculated that verses were accompanied on a harp or lyre, but there is no evidence; on present knowledge we should imagine the performance as spoken or recited text.
How can verses be held in memory across generations? The answer often given in scholarship lies in form. Alliteration, a fixed syllable count and internal rhyme make a line hard to alter: replace a word and the rhyme often fails or the count is off. Form acts as a built-in error check. This is a plausible argument, but a modern one. Nothing in the sources shows that the poets saw it that way, and it does not prove that every surviving stanza comes unchanged from the Viking Age. A metrically correct stanza may equally have been composed later by a poet who mastered the craft.
Kennings and heiti belong to the same toolbox. They form a system that could be learned: anyone who knows how to express "ship", "warrior" or "gold" in the poetic register can understand stanzas and compose new ones. Such vocabularies were later also written down as lists, which leads to the þulur.
Other routes anchored knowledge as well. Genealogies are everywhere in saga literature; they place people in a web of memory. Place names and landmarks tie stories to the landscape: passing a certain hill, one was reminded of the story set there. Both are visible in the sagas, but again we see them in the written version, and whether things were the same in the 10th century remains open.
The Old Norse word þulr (plural þulir) denotes someone who delivers knowledge; the related þulur are rhymed lists collecting poetic vocabulary, synonym series for terms such as sword, ship or man. Such lists survive in medieval manuscripts, among others in the orbit of Snorri's Edda. I must add a limit here: I read no primary text on the þulur in full for this article. All that is secure is the general statement that this list form exists. Whether the Viking Age had a distinct class of "reciters" with a fixed office cannot be clearly documented. The word occurs in the sources, but a word alone does not make an institution.
What is certain is that we know the Edda, the collection of mythological and heroic poems, only from manuscripts of the 13th century and later. How old the individual poems are is debated for each one; our piece on the Edda goes into this. For present purposes it is enough to say that the poems were probably in oral circulation for some time before being written down, but how long and in what shape differs from poem to poem.
Oral culture is easiest to grasp in law. In Iceland the Althing was founded in 930, by the traditional date, and with it came the office of the lǫgsǫgumaðr, the lawspeaker. Tradition names Úlfljótr as the founder of the laws; according to Ari's Íslendingabók (c. 1130) he brought laws from Norway, most of them modelled on the Gulathing law of the time. Ari gives no exact year; “the 920s” is an approximation in later accounts. That is tradition, not a contemporary record.
The lawspeaker was elected for three years. His task was to recite about a third of the law at the Althing each summer, so that the whole law was heard once during his term. Beyond that the office was mainly advisory: he presided, gave information and sometimes arbitrated in disputes. By the scholarly account it was the only government office of the medieval Icelandic Commonwealth, a polity without king or executive.
One detail shows how physical the task was: the term of Grímr Svertingsson (1002 to 1003) is said to have ended early because his voice was too weak. The law had to be heard, and a quiet voice was a practical problem. The list of lawspeakers is considered probably reliable from Þorkell máni (970 to 984) onward. For the first two successors of Úlfljótr, to whom 20 summers each are ascribed, scholars doubt the numbers. That too is source criticism: even the office lists are partly secure, partly schematic.
What the three-year rotation meant is a matter of interpretation. Part of the law was handed down without writing and publicly renewed, which kept it in the minds of those assembled and let them notice deviations. That the recitation was word-perfect is not stated, and we cannot test it. For how the Althing worked as a place, see our article on the Thing and the page on thing sites.
In the winter of 1117/18 a step was taken that changed the relation of orality and writing in Icelandic law. According to the accounts I read, the Althing decided in 1117 that the laws should be written down. The work was done at the farm of Hafliði Másson over the winter, and the result was presented the following year. The years 1117/18 come from a summary. Ari's Íslendingabók (ch. 10) confirms the core: in the winter after the first summer in which the lawspeaker Bergþórr recited the law, the laws were to be written down “at Hafliði Másson's”. Ari gives no years there.
The surviving lawbook is called Grágás ("Grey Goose Laws"). The name probably does not date from the time of the laws: the account says it appears in a medieval source for Norwegian law and was probably applied by mistake to the Icelandic collection in the 16th century. In the Middle Ages the Grágás never existed as a single complete volume. It is reconstructed from two incomplete manuscripts: the Konungsbók (Copenhagen, apparently written around 1260) and the Staðarhólsbók (Reykjavík, apparently 1280). They partly agree, partly complement and partly contradict one another.
That is instructive. For the law written down in 1117/18 we have no manuscript from the time; the surviving texts are about a century and a half later and differ from each other. Even in law, the best institutionalised part of the oral world, we thus see change and variation. After Iceland came under the Norwegian crown the Grágás stayed in use for a while, until a Norwegian-based lawbook ("Ironside") replaced it in 1271 to 1273; dates vary between accounts.
Runes contradict the picture of a writing-free society, yet they do not confirm the opposite either. The Younger Futhark, the rune row used in the Viking Age, developed from the Elder Futhark. By a common account the transition began in the 6th century; in the early 9th century both rows were in use side by side, as the Rök runestone shows, and by the 10th century only the Younger Futhark remained. The number of known Younger Futhark inscriptions is given as "close to 6,000", dated to the 8th to 12th centuries, many on runestones. These figures come from an unreferenced table; treat them as an order of magnitude, not a measurement.
What does this mean for the question of writing? Runes were cut into stone, wood, bone and metal. They are a utility script for short texts: names, ownership marks, memorial formulas, blessings and protective words. Narratives or laws of any length are hardly found in runic writing. That does not exclude longer inscriptions; the Rök stone is considered unusually long, which I give only as a pointer, not as a checked figure. Why books were not written in runes cannot be said with certainty. It is plausible that there was neither material nor demand, but that is a scholarly explanation, not a source statement.
On how widespread rune knowledge was, the overview page I read gave only estimates for the earlier Elder Futhark period (it cites Fischer for several hundred active literate people in total, peaking at 1,600). These figures cannot be transferred to the Viking Age. How many people could read or cut runes then, we do not know. For more on runology and the state of research see the runology hub, Runes and Science and the basic article on runes.
The Latin alphabet came north with Christianity, and with it the parchment manuscript. In Iceland, where Christianisation is traditionally dated to around the year 1000, a vernacular written culture arose that begins in the 12th century. Law, as the Grágás tradition shows, was among the early subjects of writing; the great wave of narrative texts came later. According to saga scholarship most sagas were written between 1190 and 1320, perhaps two thirds of the Icelanders' sagas in the 13th century and the rest in the 14th and 15th; kings' sagas date from the 12th to 14th centuries.
The gap needs spelling out. The events narrated in the Icelanders' sagas lie in the 9th to early 11th centuries. The texts were written in the 13th and 14th. Between event and record there are roughly two to three hundred years, a calculation of my own from these figures, not a number from a source. For the so-called contemporary sagas the gap is shorter: they were mostly written 20 to 70 years after the events, and most survive in the compilation Sturlunga saga, assembled around 1270 to 1280. For reliability these are two different cases.
For an overview of the sagas and how they are classified, see Icelanders' sagas and the list of the ten most important sagas; individual texts are treated in Njáls saga, Grettis saga and Laxdæla saga. Snorri Sturluson belongs here too: his Edda, written around 1220, is a handbook of poetic diction and one of the main stores of skaldic verse.
How much of the sagas is "oral" and how much "book" has occupied scholarship for decades. Two labels sum up the poles: freeprose and bookprose. The freeprose position holds that the stories were told for a long time in fixed or semi-fixed oral form and that scribes essentially recorded them. The bookprose position stresses the creative achievement of the 13th-century authors, who shaped literary works from fragments of tradition, genealogies, skaldic verse, place names and their own invention. Which scholars held which line in detail I did not verify against their works for this article; I give only the terms and the two lines of thought. Anyone who wants to follow the debate closely has to read the specialist literature. The English Wikipedia bibliography at least lists Knut Liestøl's The Origin of the Icelandic Family Sagas (1930) without describing his position.
The present state can be put in one sentence taken from an overview: most scholars now see the sagas as conscious artistic creations based on both oral and written tradition. The simple alternative "either oral or written" is thus outdated. The question has shifted to which elements come from which tradition and how that can be shown. It is hard to show, because the oral precursors are missing.
There is no blanket answer, and anyone who promises one should raise suspicion. Scholarship distinguishes kinds of tradition. Verse, as described, is better protected against change by its form than prose; still, the authenticity of many attributions is uncertain. Of the roughly 5,500 surviving skaldic verses, most are found in sagas and in Snorri's Edda, and many are fragments of longer poems or by unknown authors. The reliability of skaldic accounts as historical sources is also debated. Stanzas that sagas ascribe to women have traditionally often been regarded as inauthentic; few female skalds are known by name.
Prose is more vulnerable. Genealogies can partly be compared with other sources, events hardly at all. That sagas contain details confirmed by archaeology or other texts does not mean every scene happened as told. Scholarship rates the sagas as valuable and unique witnesses to the society, religion and heroic age of medieval Scandinavia; their authors are mostly unknown. Both hold at once: they are irreplaceable and they are late.
A sensible approach grades statements by their evidence. What appears in several independent sources is more likely to hold. What turns up only in a late saga and fits the drama too well deserves caution. And where sagas quote verses, it pays to separate the verse from the narrative frame; the two can be of different age.
One counterpoint to scepticism: people in cultures without books often remember things that matter to them remarkably well. That is a general observation, not evidence for the North. For the Viking Age it shows mainly that feats of memory are not impossible; whether one was actually performed in a given case is decided case by case.
That this world was about more than information is shown by a stanza of the Hávamál, the "Sayings of the High One" in the Edda. It concerns lasting fame, what remains of a person after death.
Deyr fé, deyja frændr, deyr sjalfr it sama; en orðstírr deyr aldregi, hveim er sér góðan getr.
Cattle die, kinsmen die, you yourself will die likewise; but fame never dies for the one who wins a good one.Hávamál, stanza 76 (as numbered by Bellows and in the edition on heimskringla.no; the sister stanza 77 ends with “dómr um dauðan hvern”), Old Norse text; own translation
The verse says something about a culture of remembrance: what has value lives on in the words of others. A society with few books holds to this idea in practical terms too; whoever is not talked about disappears.
Anyone who wants to understand Viking-Age motifs, runes or mythological images should keep the layering of sources in view. A rune on a stone is a witness of its time; a saga about the same region is a 13th-century witness about the past. Both are valuable, they are simply different kinds of evidence. In our workshop in Alsdorf we laser-engrave Nordic motifs, and the question of what is documented and what is later interpretation goes with us. For the wider setting see Typically Viking?; for places of writing and trade, Hedeby and Birka are good next steps.
Were there books in the Viking-Age North? Vernacular manuscript books are known only from the High Middle Ages, after Christianisation. None survive from the Viking Age itself. Rune inscriptions, by contrast, were common.
Who was the lawspeaker? By tradition the office was created in Iceland in 930. The holder was elected for three years, recited a third of the law each year at the Althing, and advised, arbitrated and presided.
When were the sagas written down? Mostly between about 1190 and 1320, probably two thirds of the Icelanders' sagas in the 13th century. The events they tell of often lie centuries earlier.
What do freeprose and bookprose mean? They are two poles of the scholarly debate: the sagas as largely fixed oral stories that were later recorded, or as independent book works of the 13th century. Today they are usually seen as a mixture of oral and written tradition.
Can skaldic verse be trusted? More than prose, because form makes change harder. Authorship, dating and attribution remain uncertain, and many verses survive only as fragments.
The Thing · Thing sites · The Edda · Snorri Sturluson · Icelanders' sagas · The ten most important sagas · Egill Skallagrímsson · Runes · Runes and science · Runology hub
The Thing · The Edda · Snorri Sturluson · The Icelandic Sagas · Runes & their meaning.
All History & Finds articles →
Hold a well-preserved blade from the Migration or Viking age in the right light and you notice something strange: a …
Read more →Of the Naharvali we have a single paragraph: around AD 98 Tacitus describes in the Germania a grove of ancient worship …
Read more →The Namnetes (Latin Namnetes) lived on the lower Loire, and their name survives in Nantes. Of the people themselves we …
Read more →On 18 September 887 a Doge of Venice fell in battle off the Dalmatian coast, fighting men from the Neretva. It is a …
Read more →