Knot runes are an elaborated variety of runes which have fascinated scholars over the centuries. But knot runes have proved to be both elusive and confusing too.

In the year 1600 the founding father of Swedish runology, Johannes Bureus, observed the peculiar knot runes in the Ovansjö inscription, Gs 15, during his visit to the Swedish province Gästrikland in Norrland: “Owansiö […] där äro dubble runor” [Ovansiö has doubled runes]. In 1726 Olof Celsius described Ovansjö as follows: “Dänne är een merkelig Steen, Hwilkens Runor äro bundne med knutor” [This is a remarkable rune-stone with runes tied in knots]. (GsR p. 163).

For many years Ovansjö was the only example of knot runes. In 1901 George Stephens published the so-called Göngu-Hrólf’s horn in Florence (of West-Norse origin) in Old-Northern Runic Monuments, vol. 4. His interpretation “Andrell gart (made) me” was slightly wrong but much better than Halvdan Koht’s “Reinaldr or Reinn made me” from 1928 which later entered Norges innskrifter med de yngre runer (NIYR 5 p. 236 f.). Koht wrote “jeg inrømmer at ornamentikken her er slik at den også bragte mig til å spørre om den ikke hadde til hensikt å fremstille tre runer. Men jeg har ikke kunnet få noen virkelige runer frem” [I admit that the decoration here is such that it also brought me to ask if it could not be meant to represent three runes. But I have not been able to construe and genuine runes.]. Magnus Olsen in NIyR suggests R[agnvald] E[írks]s(on) gerði mik. It was not until 1979 that Aslak Liestøl identified the first runes as knot runes and established the reading “Andres gerði mik”. As Sven B. F. Jansson points out Liestøl seemed to be unaware of the Ovansjö stone (GsR 1981, p. 173). Seemingly both Liestøl and Jansson were unaware of the fact that Erik Brate had given the correct interpretation in his book Svenska runristare in 1925.

In the meantime, an important new-find was unearthed in Greenland. In 1958 C. L. Vebæk published a spoon from Unartoq with the most informative inscription spon carved both in knot runes and ordinary runes. Since then knot runes were reported from Bergen (Liestøl 1979, p. 231) and within the last decades two additional knot rune inscriptions have been found in Greenland, one being a knot runic futhark inscription.

A survey of the, admittedly, limited corpus of knot rune inscriptions reveals that knot runes are in several cases followed by doubled runes which in turn are most commonly used in the Iron Age. Little scholarly attention seems to have been paid to these elaborated runes, although they may in fact have played a significant visual role in conveying the written message. They are not mentioned in Marco Bianchi’s dissertation (Runor som resurs, 2010) or for that sake in the most recent handbook on runes by Michael Barnes (2012).

This raises the question about the origin and meaning of knot runes. Do knot runes signify a specific visual effect or do they have social or stylistic significance? The paper will try to demonstrate that knot runes can both be explained as a coherent runologic feature and at the same time display contemporary mediaeval writing fashion.