Dalecarlian runes

The runes that outlived the runes

Christianity and the machinery of the state brought the Latin alphabet to Scandinavia, and by the end of the Middle Ages runic writing had been squeezed out of every part of public life. Except in one place.

In Dalarna — specifically the Ovansiljan district, the parishes of Älvdalen, Mora, Våmhus, Orsa, Sollerön and Venjan — the runes simply carried on. They kept going through the Reformation, through the Thirty Years' War, through the Enlightenment, through the industrial revolution, and stopped only at the turn of the twentieth century. The last person who had learned them natively died in 1980.

They are called Dalecarlian runes, or dalrunor, or Elfdalian runes. Dalarna's reputation as the last stronghold of the Germanic script rests on them, and it is deserved. But the four centuries were not a preservation. The script changed continuously, and what it changed into is the interesting part.

A conservative script for a conservative language

The runes survived here because a great deal else survived here. Ovansiljan was agrarian, largely self-sufficient, geographically awkward to reach, and consciously attached to its own customs — the same solidarity that kept the folk costume kept the alphabet.

It also kept a language. Elfdalian (övdalsk), spoken today by perhaps three to five thousand people around Älvdalen, is the most conservative North Germanic variety there is. It holds onto the old syllable structure, an elaborate inflectional system, the voiced th of English father, and the w sound that standard Swedish lost. Most strikingly it keeps nasal vowels, attested in Old Icelandic in the thirteenth century and gone everywhere else in Scandinavia; modern Elfdalian orthography marks them with an ogonek — ą, ę, į, ų.

Alongside that conservatism runs real innovation. The long vowels broke into new diphthongs, so where Swedish has is and ute, Elfdalian has ais and aute, and where Swedish says solen for the sun, Elfdalian says sola.

The runes tracked this. A carver writing kart for the past participle was recording what people around him actually said — gart, giärt — rather than what the printed Bible said they should. That is the advantage of a script nobody official is supervising.

The wall that showed it had stopped being a futhark

In 2014, inside a preserved wooden house in Älvdalen, researchers found an alphabet cut into the timber of the wall. The building had been dated by its tree rings to 1285; the carving, on stylistic grounds, is from the end of the sixteenth century — the house is old, the inscription is not.

What mattered was the order. Runic alphabets run f, u, þ, a, r, k — that sequence is what the word futhark is. This one ran a, b, c, d.

Someone had rearranged the runes into the order of the Latin alphabet, which is only worth doing if you have stopped thinking of them as a writing system in their own right and started thinking of them as a substitute set of letters for the alphabet you already know. The Dalecarlian runes had become a character-for-character mapping of Latin.

The proof of that is what they did with the gaps. Latin has C, Q and X; the runes had never needed them, because no runic tradition ever had those letters. So the carvers made them up — new runes invented for Latin letters that had no runic ancestor at all. From that point on the script does not answer to Old Norse phonology. It answers to the alphabet.

Bureus writes it down, 1599

The first person to record the script formally was Johannes Bureus (1568–1652), Sweden's first riksantikvarie, royal librarian, and one of the strangest figures in the history of runology — a serious field scholar of runestones who was simultaneously a Rosicrucian mystic convinced that Scandinavia held the remains of a divine original language, and who built his own esoteric fifteen-character system, the Adalrunor.

In 1599, after travelling the country recording inscriptions, he published an engraving titled Runakänslanäs läräspan. Under the heading Nostri secli Dalecarl(is) — the Dalecarlians of our own century — he set out twenty-four runes with their names, in alphabetical order.

The names show what had happened to the script. Some are inherited: is, ar, ur, sol, tir, lagh, birkä, madhär, hagal. Others plainly are not. Qua and äcs exist for no reason except that Latin has Q and X. Knäsol, 'bent sun', was coined for C, echoing the Icelandic knésól used for Z.

Bureus is indispensable and not entirely trustworthy. Some of what he lists was his own tidying — characters borrowed from calendar staffs, such as the bälgturs, pressed into service for Latin letters that the farmers of Dalarna were not in fact writing that way. In 1611 he went further and published Svenska ABC boken medh runor, the first Swedish alphabet book ever printed, hoping to teach the nation to write in runes instead of Latin. That failed. It did leave a legacy of a sort: during the Thirty Years' War, Swedish officers used his runes to encrypt dispatches — the first of two times this script was turned into a cipher.

Wood, because paper cost money

There are no Dalecarlian runestones. Paper and parchment were expensive and the writing was not official, so people wrote on what the forest gave them. Over three hundred and fifty inscriptions survive, and more than nine in ten come from Älvdalen parish alone.

They are on bowls, bridal boxes, chests, kitchen blocks, measuring sticks, furniture and the log walls of buildings. They record ownership, names, dates, and short messages. The Latin alphabet in this society belonged to the clergy, the bailiff and the crown; the runes belonged to the allmoge, the ordinary people, and were used for ordinary purposes.

The oldest securely dated example is a wooden bowl from the village of Åsen recording that Anders made it in 1596. There is a loft at Orsbleck from 1635 naming the two men who built it, another Åsen bowl from 1692 marking a year of famine, and a chest from Kittan in 1720.

Calendar staffs and the village summons

The most sophisticated objects are the rimstavar, the runic calendar staffs. These are sticks carved with the dominical letters and golden numbers needed to work out the church year and the movable feasts — an almanac you could hold, in a society where almanacs were not sold.

About twenty of them carry something else as well: village business. One from Åsen in the eighteenth century puts it bluntly — plikt fölieir den som int går, a fine follows him who does not attend. Another, from Älvdalen in 1818, announces that there is a public meeting tonight at the usual place.

This is a notice board. Whoever cut it expected the village to read runes on sight, in 1818.

Gätrunor: the herders' runes

Dalarna practised fäbod farming — transhumance, moving the cattle up to remote forest and mountain pasture for the summer and living up there with them. That work was largely done by women, alone or in small groups, at a considerable distance from the village for months at a time.

They wrote to each other. On pine trunks, on boulders, on the walls of the huts and shelters, they cut short notices known as gätrunor, herding runes: initials, a year, and the word gät, tended. Who had been at this pasture, and when.

The last genuine Dalecarlian inscription in the unbroken tradition is one of these. It was carved in a pasture hut near Gryvelån in Älvdalen, and by that stage the script had almost dissolved — only a few true runes remain among Latin letters. It reads AAD gät 1900: Anna Andersdotter herded here in 1900.

Two thousand years of Germanic writing end with a woman noting that she looked after the cows.

How the script dissolved

The four centuries divide fairly cleanly.

Before about 1580 the runes still behave like runes. Roughly thirty inscriptions survive from this stage, the shapes close to ordinary medieval forms, spelling phonemic rather than alphabetic.

About 1580 to 1650 is the shift the Älvdalen wall records: the mapping onto Latin, the invented C, Q and X, and the first sign of the alphabet reaching back into the shapes themselves — the g rune is redrawn from three strokes into two, in imitation of a cursive Latin g.

About 1650 to 1750 the Latin hand bleeds through everywhere. The staves of the a and ä runes tilt over to lean like a written a. The runic i, a plain vertical, is displaced by a lower-case Latin i with its dot. Capital Y, L, M and Ö start appearing in runic text with no disguise at all. Meanwhile the language moved underneath the script: Älvdalen lost the /h/ sound by the 1650s, which left the hagal rune unemployed, and it was repurposed to write å — one sound vanishing and shunting a whole letter sideways.

After about 1750 almost nothing is left. Full names give way to initials. Runes survive as isolated habits inside high-frequency dialect words — den, gät — the way an old spelling survives in a signature.

One detail runs through all of it. Medieval Scandinavian writing distinguished extra sounds by dotting a rune: a dotted i gave e, a dotted k gave g, a dotted t gave d. Elsewhere in Sweden that dot eventually stretched into a bar across the stave; in West Norse it curled into a small circle. In Dalarna it grew into a side-stroke off one side — and Dalecarlian is the only script that uses a dotted short-twig a for ä. As that dot lengthened, the letter drifted towards the shape of å, and carvers spent the next two centuries quietly recalibrating which mark meant what.

This is the answer to the Kensington runestone

In 1898 a Swedish immigrant farmer named Olof Öhman turned up a slab of greywacke on his land near Kensington, Minnesota. It carried a long runic inscription: eight Götalanders and twenty-two Norwegians on a journey of exploration, ten men found dead, a prayer — AVM, Ave Virgo Maria, save us from evil — and the year 1362.

Runologists rejected it almost at once, and for a century could say precisely what was wrong without being able to say what it was. The language was too modern. And the letters were impossible: a crossed l, a dotted v, an o with a line through it, a p without dots in its loops. No fourteenth-century rune row contains them. No known rune row contained them at all.

In 2003 the linguist Tryggve Sköld, working through the dialect archive at Umeå, found them. The papers of Edward Larsson (1867–1950), an apprentice tailor raised near Borlänge in Dalarna, contain runic alphabets written out in 1883 and 1885. Larsson's list is a twenty-seven character alphabet for writing modern Swedish, and it matches the Kensington letters — the crossed L, the distinctive Ö and Ü, the odd G. It also contains pentadic numerals, the tally figures from Scandinavian calendar staffs, in the same forms used to write 1362 on the stone.

And Larsson's papers hold a third alphabet, which he describes as the oldest secret writing in the world, the script of the tablets of the Law. It is the Masonic pigpen cipher.

That is the context: late Dalecarlian runes, pentadic numbers and a lodge cipher circulating together among itinerant nineteenth-century craftsmen — tailors, builders, travelling teachers — through Dalarna, Medelpad and Hälsingland. Corroboration followed. A wooden yoke from Månsta in Älvdalen dated 1907 uses the same characters. A memorial plaque from 1868 by Eric Ström, an itinerant writing-master from Medelpad, uses them too, and his lithographed copybook of 1876 prints the whole set — runes, pentadic numerals and Masonic ciphers together. An oak measuring stick from Haverö carries all three, plus twenty-one golden numbers; it is an aln, a unit abolished in Sweden in 1889, which puts a ceiling on its date.

The Kensington stone is not a medieval artefact and not quite a hoax in the ordinary sense either. It is a nineteenth-century Scandinavian-American object written in a real, living, documented script that its finders' generation had actually been taught — carried to Minnesota in the same trunks as everything else.

Why this page has no letter pages

Two reasons, and the second is the stronger one.

The first is technical. Dalecarlian runes have no Unicode codepoints. The runic block encodes the Elder Futhark, the Futhorc and the standard Scandinavian medieval runes; the Dalecarlian characters are not there. Every published table of them, including the ones this page draws on, is showing approximations borrowed from other rune rows or drawn with a custom font. The letter pages elsewhere in this section pair a glyph with the codepoint it actually is. Here there would be nothing honest to put in that field.

The second is that there is no single alphabet to page. The inventory changed shape for four hundred years and varied between carvers and villages; the a of 1600 is not the a of 1750, the g was redrawn to look like Latin, the hagal changed jobs when a sound disappeared, and by the end most letters had been replaced outright. A page per letter would have to pick one moment out of that drift and present it as the Dalecarlian alphabet, which is exactly the false precision this section exists to avoid.

The drift is the subject. It is better told once, in order, than twenty-four times in fragments.