Where runes came from
The runic alphabet was invented once, somewhere in northern Europe, in the first or second century AD. Beyond that, almost every detail is contested.
Runes are not a spontaneous invention. The shapes, the order of the characters and the principle of one-sign-one-sound all point to a Mediterranean alphabet as the model. What nobody has settled is which one.
The three main theories
| Theory | Argument for | Argument against |
|---|---|---|
| Latin origin | Rome was the dominant literate power on the Germanic frontier, and several runes closely resemble Latin capitals: F, R, B, H, I. | Some runes have no plausible Latin source, and the runic sound values do not always follow the Latin ones. |
| North Italic origin | Several alphabets used in the Alpine region before Rome share distinctive shapes with runes, and the geography fits contact with Germanic peoples. | The North Italic scripts were dying out by the time the earliest runes appear, leaving an awkward gap. |
| Greek origin | Goths had contact with the Greek-speaking world, and a few correspondences are suggestive. | The chronology is difficult and the geography is wrong for the earliest finds, which are northern. |
The Negau helmet, and what it is not
One object comes up constantly in this argument and is almost always described wrongly. The second of the Negau helmets, from a hoard found in what is now Slovenia, carries the inscription harikastiteiva…, cut somewhere between roughly 350 and 100 BC, centuries before any rune. It is usually read as a Germanic personal name, Harigast, and Markey takes the following word as a title, giving "Harigast the priest".
The helmet is not written in runes, and it is not an early form of them. The consensus is that its script is North Etruscan proper and predates the runic alphabet outright. What it does show is the one thing the North Italic theory actually needs: a Germanic speaker having his name written, in an Alpine alphabet, well before runic writing existed. That is evidence of contact, not of ancestry, and the difference matters: this object is regularly produced as "the oldest runic inscription", which it is not by any reading.
The origin of the runic script is an open question in the field, in the strict sense: no answer commands agreement. Any source that states flatly that runes derive from a particular alphabet is presenting one position as settled fact. The Negau reading is contested too: that the first element is the Germanic name Harigast is widely accepted, and the rest of the inscription much less so.
The earliest inscriptions
The oldest objects generally accepted as runic date to around the middle of the second century AD. The Vimose comb from Denmark, carrying the single word harja, is usually cited as the earliest, at roughly AD 160. The Meldorf fibula from northern Germany is a century older but its inscription may be Latin rather than runic, and it is argued about accordingly.
What these earliest finds share is a fully formed alphabet. There is no archaeological trail of experimental or half-developed runes. The script appears essentially complete, which is one of the arguments that it was deliberately devised rather than gradually evolved.
The order of the alphabet
The Kylver stone from Gotland carries the oldest complete futhark row we have. Its order is not alphabetical in the Mediterranean sense (f-u-þ-a-r-k is not derived from alpha-beta-gamma), and the reordering is itself unexplained. The Vadstena bracteate, a century later, gives nearly the same sequence, with dots dividing it into three groups of eight.
Nearly, but not quite. The two rows disagree about two pairs, four positions of the twenty-four, leaving twenty in common: Kylver reads ...j p ï z... and ends ...ŋ d o, while the bracteate reads ...j ï p z... and ends ...ŋ o d. Kylver is the outlier both times. The other early rows side against it: the Grumpan bracteate agrees with Vadstena on both pairs, and the Breza column in Bosnia and the Charnay fibula from Burgundy agree on eihwaz-before-perthro before their rows break off. The Mariedamm bracteate agrees too, but it is struck from Vadstena's die and so does not count separately. The order used on this site and in every modern chart splits the difference anyway, taking eihwaz-before-perthro from the majority and dagaz-before-othala from Kylver alone. It is a scholarly convention, not an ancient row anyone actually carved.
How the script changed
| Period | What happened |
|---|---|
| c. AD 150–400 | The Elder Futhark in use across a wide area, from Denmark to the Balkans. Inscriptions are short and rare. |
| c. AD 400–700 | The Migration Period. The greatest geographic spread, and the bracteate inscriptions. The alphabet begins to diverge regionally. |
| c. AD 400–800 | England and Frisia expand the alphabet into the futhorc as Old English develops new sounds. |
| c. AD 700–800 | Scandinavia reduces the alphabet to the 16-rune Younger Futhark. Nobody knows why. The Ribe skull fragment, c. 725, is the clearest object from inside the change: already a younger futhark, but still using two Elder Futhark letter shapes. |
| c. AD 800–1100 | The Viking Age. Thousands of runestones, mostly Swedish, mostly memorials. |
| c. AD 1100–1500 | Dots and additional characters restore the lost distinctions. Medieval runes are used for everyday writing. |
| After 1500 | Runes survive in rural pockets, notably Dalarna in Sweden, into the 20th century. See Dalecarlian runes. |
Why the Younger Futhark shrank
This is the strangest event in runic history. Between roughly 700 and 800, Scandinavian carvers dropped eight of their twenty-four characters, at exactly the point when Old Norse was developing more distinct sounds, not fewer. The result was a writing system markedly worse at recording the language it was used for.
Explanations have included a deliberate reform, pressure from carving practice towards fewer strokes, and a shift in what runic writing was for. None is generally accepted. It is worth sitting with the fact that one of the most consequential changes in the history of the script is simply unexplained.
What we do have is an object from inside it. The Ribe skull fragment, cut around AD 725, is written in a younger futhark but still forms its h and m with the Elder Futhark shapes ᚺ and ᛗ. It does not explain the reduction, but it dates it, and it shows that the change went through in stages rather than at a stroke.
Questions this guide answers
- What is the oldest runic inscription?
- The [Vimose comb](/inscriptions/vimose-comb/) from Denmark, inscribed harja and usually dated to around AD 160, is the oldest generally accepted runic find. The [Meldorf fibula](/inscriptions/meldorf-fibula/) is older but disputed: its inscription may be Latin rather than runic.
- Who invented runes?
- Nobody knows. The script appears fully formed in the second century AD, modelled on a Mediterranean alphabet, but the individual or community responsible left no record.
- Are runes older than the Latin alphabet?
- No. The Latin alphabet is several centuries older, and is one of the candidate models for the runic script.
Sources
Full references on the sources page. Our methodology explains the evidence labels used throughout.