Phonetic Transcription of Modern Swedish in Runes

By and Š 1999 Ingeborg S. Nordén, Edited by Úlfgrim Vílmeiđson


The original inspiration for this project came from a fellow runester and linguist, Douglas W. Smith (aka "Deep Stream"). He has convinced me that the way most Heathens render modern languages in runes is wrong—that people should transcribe a word as it actually sounds, instead of copying the conventional (Roman alphabet) spelling letter for letter. Mr. Smith’s suggestions for transcribing most of the modern Germanic languages are acceptable to me, even though I know little German and no Dutch at all.

Many of the transcriptions that Mr. Smith proposed are acceptable in Swedish as well (especially the consonants, which show fewer differences than the vowels). He did a reasonably good job for someone who does not speak the language, and who has only a general idea of its history. However, every language—even within a tight group—develops some unique sounds and sound-changes: a phonetic spelling system that works well for English, German or Dutch may not necessarily work well for Swedish.

With that in mind, here are my own recommendations for rendering Modern Swedish in the Elder Futhark—one rune at a time, similar to Mr. Smith’s discussion section in his "FUTHARC" article.

Section One: The Vowels


Uruz is a fairly obvious, straightforward equivalent of both u sounds in Swedish. The long u of hus ‘house’ and the short u of hund ‘dog’ should be written with this stave: hus and hund respectively.

The rune had three other sound values during the Viking Age, but I would not suggest these in an Elder transcription:

  1. The consonant v. Rendering Swedish v with Uruz is like rendering Swedish j with Isa—historically accurate in Younger runes, but the Elder runes clearly distinguish consonant and vowel sounds. (I described one worst-case scenario in a note to Mr. Smith: one of the Swedish words for ‘owl’ is uv. Transcribing this word as two Uruz-runes would be very awkward; transcribing it as a single Uruz-rune would be downright misleading, because someone might interpret the stave as a shorthand for its name and read ‘aurochs’.)
  2. The vowel y. Not all y sounds in Swedish derive from an older u; sometimes the original sound was an i, which became rounded because of a u or w sound in the NEXT syllable. (One of the other rune-names is a perfect example of this: the name *Ingwaz in Primitive Germanic ultimately became Yngve in the modern Scandinavian languages.) Furthermore, even some variants of the Younger Runes used a totally different stave for this vowel (see my notes on Eihwaz); distinguishing u from y is just as accurate historically and makes better sense phonetically. (The name Yngve, if I were to write it out in full, would be transcribed INwe…no Uruz in sight.)
  3. The vowel ö. Although the SHORT sound of Swedish u resembles ö closely enough to bother language teachers, the LONG sound is radically different. (Brud ‘bride’ and bröd ‘bread’ are not easily confused; but a Swedish man carving a talisman to attract a fiancée ought to distinguish them in writing. Otherwise, he might end up opening a bakery instead—rune spells can sometimes go wrong because of miscarved staves, as the sagas have shown!) Furthermore, the ö sound almost never derived from a simple u in any of the older Germanic languages; the Elder Futhark allows more accuracy than the Younger, so that using Uruz here is misleading and unnecessary.


The uses of this vowel, too, look simple at first glance; most obviously, it represents both the long and the short a sounds in stavar ‘staves’: stabaz. However, Ansuz is also one possible spelling for the vowel å –and probably the best alternative from a historical point of view, since å did develop from an Old Swedish a: the word gård ’farm, enclosed yard’ can be transcribed gard .


In Swedish, the long i sound resembles the vowel in English "machine"—and writing the word is ‘ice’ without its namesake rune would make no sense. This stave works equally well, though, for a short i (which sounds much like the English or German one): the sequence fina could be read as either fina ‘fine [adj., pl.]’ or finna ‘find’.

Despite Mr. Smith’s recommendation—which has some historical basis—I would not use Isa to represent a schwa in Modern Swedish. That choice does not match actual pronunciation well; even in an unstressed position, the vowel Swedes use is closer to the one in English "ten" than the one in "tin" or "teen".


This rune— not Uruz!— is the one I recommend for transcribing Swedish y. (The name Yggdrasil would be spelled Igdrasil, for instance.) Why do I prefer this solution to the y problem? There are two good reasons, based on the history of both Old Norse and the runes themselves:

  1. Although its phonetic value in Primitive Germanic is unknown, Eihwaz probably stood for a vowel of some kind. (Most runes represent the sound at the beginning of their word-names, if that is normally possible.)
  2. The Ũr-rune in the Younger Futhark (which had a Norse form of the same name) was used for y in some late Viking Age inscriptions.


This vowel-rune has more uses than the other five, which should not be a surprise considering how common e-like sounds are in Swedish.

  1. A stressed e, either long as in hel ‘whole’ or short as in hem ‘home’. (Those two words would be transcribed heland hem respectively.)
  2. A simple alternative for ä, which sounds closer to e than to any Swedish variant of a: häst hest ’horse’; träd tred ‘tree’.
  3. A schwa—if a vowel-rune is needed there at all. I usually indicate unstressed e in a word IF leaving it out would confuse or mislead a reader. There is a huge difference between and and ‘duck’ and ande ande ‘spirit’, for example; the second syllable should definitely be shown in writing!

Sometimes, however, it is safe to omit unstressed e in a runic transcription. If a word ends in –el, -en or –er, it should be obvious enough where the schwa goes. In Old Norse, words like that tended to have one syllable; the vowel was inserted at a later stage, when Swedish had begun to split off—

ON sigl ‘sail’ ā ModSw segel segl

ON vatn ‘water’ ā ModSw vatten watn

ON heiđr ‘honor’ ā ModSw heder hedz

On the other hand, -en sometimes serves a grammatical function in Swedish: the definite article is not usually a separate word, but a suffix added to nouns. Not including a vowel-rune for the e may confuse some readers, especially if the noun ends in n itself. There should be a visible difference between sägen segn ‘legend’ WITHOUT an article, and sägnen segnen ‘the legend’ WITH one. (Double letters are traditionally disallowed in runic writing, so recording the e between those two n-sounds seems like the best option.)


Considering the way the o-sound has evolved in Swedish, using this stave is not as simple as it looks. Three possible sound-values for Othala exist (ranging from most to least appropriate):

  1. The most typical Swedish o, which has no exact equivalent in a non-Scandinavian language; English speakers might think of it as a cross between the vowels in "boat" and "boot". (Swedish is one of the few Germanic languages that has kept the word used to name this rune: the modern form is odal, which contains this changed o sound. In any case, the word certainly OUGHT to be spelled odal if a runester wants to write it out!)
  2. A sound closer to the o (long or short) in most other European languages: sol ‘sun’ can be transcribed as sol, and dolk ‘dagger’ as dolk . Phonetically speaking, Othala could also be used to transliterate å (the most usual Swedish spelling for that sound): my earlier example of gård ’farm, enclosed yard’ could have been written as gord . However, I dislike this usage because it obscures the original connection between a and å. (See the end of this section for a third alternative.)
  3. The vowel ö sometimes (but not always) derives from an o in the older Germanic languages. Although I am enough of a purist to avoid extending the futhark, I am also reluctant to use a simple Othala for this particular sound; it has become different enough to justify another spelling. Short of avoiding words with ö altogether—which is harder than people think, as common as för ‘for’ is!—the best solution seems to be either modifying the Othala rune, or finding a combination of symbols to spell the sound.


In my correspondence with Mr. Smith, he has argued in favor of a minimalist approach to runic transcription: letting a single stave to represent many related sounds, and leaving the reader to puzzle out the context. Purely magical inscriptions may need to be this cryptic, so he does have a valid point in that case. However, some runic inscriptions (ancient and modern) are longer texts that contain mostly ordinary words. This type of inscription ought to be easier for an average Heathen to decipher: that does not have to mean following a Roman-to-runic transliteration slavishly, but it might mean using different spellings for similar-sounding words that might be misunderstood.

What does all this have to do with transcribing Swedish in particular? The conventional Swedish alphabet includes three vowels (å, ä, ö) not found in English. Although å/o and ä/e sound very similar at times, the spelling does distinguish meaning: sed ‘tradition’ and säd ‘seed’ are different words, but they could be confused if a runester spelled both as sed in the same inscription.

Is it possible to distinguish these sounds when the need arises, without inventing totally new symbols? In my opinion, the answer is YES. Even in Viking times, people extended the use of existing runes in two main ways.

  1. A dot could be added to a rune whose sound value was close to the one intended. (For example, dotted Kaun g represents a g-sound instead of the usual k; dotted Iss e represents e instead of i.) The same principle can be used to transliterate the three "special" vowels in Swedish: I often add a dot to Ansuz, Ehwaz and Othala for å, ä, and ö respectively.
  2. A conventional spelling, which often consisted of more than one rune, developed for a few specific vowel sounds. (The combination au was sometimes used for o or ö in Viking Age inscriptions, for example.) Developing multiple-rune spellings for the "special" Swedish letters may be a better solution for people who dislike diacritical marks, or who are unable to show modified letters in text typed on a screen. I have had to do this myself at times—using double Ansuz, Ansuz+Ehwaz, and Othala+Ehwaz for å, ä, and ö respectively. (The sequences aa, ae, and oe seldom occur in ordinary spelling or pronunciation; assigning a sound to the runic equivalents is unlikely to confuse anybody.)


Spelling and pronunciation coincide more often in Swedish than they do in English, so most of the consonant-runes are straightforward even in a phonetically based transcription. Only a few potential misuses arise, such as including silent letters or ignoring "hard" vs. "soft" consonants—but those will be discussed under the individual letters.


This stave has two possible uses in Swedish:

  1. An f-sound (resembling the English one), as in fara fara ‘go’ or frid frid ‘peace’.
  2. A v-sound as in av af ‘off’ or häva hefa ‘lift, raise’. This usage actually reflects older "conventional" spelling in Swedish and Old Norse: f was pronounced as v everywhere except at the beginning of a word. However, I would NOT use Fehu for a v which came at the beginning of a word, or was part of a cluster like sv- or kv-. (See Wunjo and Berkano for other ways of transcribing the sound.)


The sound represented by this rune—th as in English "thing"-- no longer exists in Swedish. The only "mundane" use for Thurisaz would be transcribing Old Norse proper names. Two examples come to mind:

  1. Thor’s name ought to keep the same initial rune it had in Old Norse, even though Swedes now pronounce it with a t-sound: Tor. (Note to Mr. Smith: That final –r is not a grammatical ending, but part of the stem; in Old Norse, the nominative form was spelled with TWO r’s, and the second represents the ending.)
  2. Anyone who works closely with Sweden’s language and culture is going to refer to Sweden fairly often in her rune work…and in a magical/religious context, using the Old Norse name of the country makes more sense. That name, to use a slightly modernized spelling, is Svithjod; because its original pronunciation included a th-sound, the best transcription is swiTjod.


Using this rune correctly is trickier than it seems, especially in a Scandinavian language like Swedish. Old Norse originally had TWO distinct r-sounds: One of them (represented by Raidho) had been r since Primitive Germanic times, and the other (represented by Elhaz) developed from an earlier z-sound. This z-turned-r appeared only in certain grammatical endings, however; it was never part of a word stem. (I will list these endings under the Elhaz rune itself.)

Unless you are writing a word with one of those endings, Raidho is always the correct rune for the Swedish r-sound: rida rida ‘ride’; bara bara ‘only’; finger fiNer’finger’.


Again, Swedish pronunciation makes using this rune harder than it seems. The language has both a "hard" k-sound (the one normally associated with k in English) and a "soft" one used before certain vowels (it resembles the "soft" ch in German ich). Only the "hard" sound, in my opinion, should be transcribed with a Kenaz-rune standing alone. Even then, the rune can correspond to several Roman-alphabet spellings—

  1. The letter k (before most consonants or stressed a, o, u, å): kraft kraft ‘strength’, kung kuN ‘king’. A k that appears somewhere other than the beginning of a word is usually hard as well: rike rike ‘kingdom’; ek ek ‘oak’.
  2. The combination ck: because double letters are traditionally written single in runes, the sequence tak could theoretically be read as either tak ‘roof’ or tack ‘thanks’. (Someone who insisted on distinguishing those words could include a second Kenaz-rune at the end, but this is not historically accurate.)
  3. The combination ch: Other than proper names, Swedish has only one word in which ch represents a hard k; but that one word is the most common in the language. Och ‘and’ should probably be transcribed ok …although some Swedes might argue that a simple Othala-rune best represents the way they say och in casual speech. (That spelling, however, would be taking phonetic minimalism a bit TOO far in my opinion!)
  4. The letter x: Swedes always pronounce x as in English "six" or "tax". A straightforward sequence of Kenaz+Sowilo can be used to transliterate it: yxa Iksa ‘axe’; lax laks ‘salmon’.

As for transcribing the Swedish soft k, I prefer using the sequence Kenaz+Jera--not Hagalaz, as Mr. Smith suggested for the German "ich-sound". (Swedish does not have the corresponding "hard", guttural ch as in German ach; besides, the kj spelling reflects the way this sound developed historically.) Whether or not the present Roman-alphabet spelling has kj- in a word, I use the runic equivalent consistently: kjol kjol ‘skirt’ looks as a Swede would expect, but kärna kjerna ‘core’ gets one extra letter in runes.

One last note to English-speaking people: Swedes DO pronounce k in the combination kn-, so a Kenaz-rune should be retained when transcribing it. The words knekt knekt ‘soldier’ and knut knut ‘knot’ sound as they are conventionally spelled.


This rune, too, is likely to be misused because of conventional Roman-alphabet spelling. In Swedish, the letter g has five possible sound values—only ONE of which should be spelled as Gebo.

  1. A hard g like the one in English "gate" or "gun". This sound is used before most consonants, and before stressed a, o, u, å. It is also the usual g-sound within or at the end of a word: gren gren ’branch’; gammal gamal ‘old’; trogen trogen ‘faithful’; dag dag ‘day’.
  2. A soft g which resembles the typical Swedish j (or the y in English "yes"). See my notes under Jera for examples of transcribing this sound
  3. Before n, the letter g sometimes sounds like ng in English "sing". In this combination the correct rune would be Ingwaz, not Gebo: regn reNn ‘rain’; lugn luNn ‘calm’. Not all combinations of gn sound that way though; at the beginning of a word like gnista gnista ‘spark’, the g is hard and should be transliterated as Gebo. (Yes, Swedes do pronounce that g even though English-speaking people would not!)
  4. In some words borrowed from French, g sounds similar to English sh in "sheep". This sound should be transcribed as Sowilo+Jera, the same way it would be otherwise. The first g in garage (as a Swede pronounces it!) would be written as Gebo, but the second would not: garasj.
  5. Finally, the letter g is totally silent before j; see my notes under Jera for more details.


In Old Norse, two things happened to the w-sound which this rune originally represented. Before some vowels, the w disappeared altogether: this explains the difference in English/Swedish pairs such as "word" and ord, "wonder" and under. Otherwise the w became a v in all the Scandinavian languages (see the examples in the next paragraph).

Because of this sound change, it is possible to use Wunjo as a distinct stave for Swedish v: vilja wilja ‘will[power]’; vatten watn ‘water’; svensk swensk ‘Swede (n.); Swedish (adj.)’. However, if the v is NOT at the beginning of a word (or part of a cluster like sv-), chances are good that the word had either a b (Berkano) or an f (Fehu) in Primitive Germanic. The letter that spelled the original sound might be more appropriate then; but if I had to choose a single Elder Futhark rune to represent v in a Scandinavian language, Wunjo would be that rune.


Finally, a simple rune compared to all the others I have discussed so far! Its only sound value is h, which Swedes pronounce the same way Americans usually do: hem hem ‘home’; hund hund ‘dog’. (For the combination hj-, see my notes under Jera.)


The same rules for using this rune correctly apply in both English and Swedish: it stands ONLY for the n sound of "night" or "thin". (The Swedish equivalents of those words are natt nat and tunn tun respectively.)

For the combination –ng- (as in "sing"), the only correct Elder Futhark equivalent is Ingwaz. The –nk- combination should also use Ingwaz for n, because that IS the way people pronounce it: blank ‘shining’ should be spelled blaNk in runes.


Someone who knows no Swedish might think that using this rune is pretty straightforward. It is the obvious equivalent of j in a word like jaga jaga ‘hunt’; but other ways of spelling that sound exist:

  1. Some consonants are silent before j, and should be skipped in runic transcriptions. (I will deal with each one under the corresponding rune that might be misused.) In theory, the sequence jord could represent THREE unrelated Swedish words: jord ‘earth’; gjord ‘done, made’; or hjord ‘herd’.
  2. A Swedish "soft" g should also be written as Jera, to reflect the pronunciation. The soft sound usually occurs—
    1. before stressed e, i, y, ä, and ö: genom jenom ‘through’; gille jile ‘guild, kindred [in the Asatru sense]’.
    2. in the combinations –lg and -rg: älg elj ‘moose’, berg berj ‘hill’. (This is not always true, however: the word helga helga ‘consecrate’ has a hard g and should be spelled with Gebo.)
    3. in the word Sverige swerje ‘Sweden’. (That transcription shows the way most people would say the name in ordinary speech; to show a very formal, careful pronunciation I might spell it swerije instead.)


As straightforward as its Roman-alphabet equivalent, for once: this rune stands for the p sound in pappa papa (no translation needed here!). Swedish does not have any letter combinations like ph- which might lead to misusing this rune.


As Mr. Smith mentioned in his original paper, this rune occurs ONLY in grammatical endings which had a z in Primitive Germanic. Where have those endings survived in Modern Swedish? All over the language--

  1. The present tense of most verbs: talar talaz ‘speak(s)’; kommer komz ‘come(s)’; är ez’is, are’. (The past tense var waz ‘was, were’ should also be written with Elhaz.)
  2. The plural form of many nouns: dagar dagaz ‘days’; vänner wenz‘friends’, runor runoz ‘runes’. (Plurals that don’t have a vowel-plus-r ending didn’t have a z originally, and should NOT be written with Elhaz!)
  3. The possessive form of most nouns (and a few pronouns as well): hans bok hanz1bok ‘his book’, gudens namn gudenz1namn ‘the god’s name’.
  4. The comparative form of most adjectives: längre leNze ‘longer’, gladare gladaze ‘happier’, mer mez’more’.

Note: The vowel-rune can be omitted in an –er ending (as shown in the examples above). However, I would NOT omit it in any other vowel-plus-r grammatical ending: the unstressed a and o still have distinct sounds in Swedish. I would also retain the vowel in comparative –(a)re, which should be distinguished from –er in some words. (Högre ‘higher’ and höger ‘right [as opposed to left]’ come to mind…)

If a plural –r and a possessive –s occur in the same word, the situation becomes awkward. Using two Elhaz-runes in a row looks wrong, and using a single one to cover both suffixes is misleading. Then, and ONLY then, would I allow a runester to spell one of the endings "incorrectly" (his choice of which, though I tend to pick the –s possessive myself): alla gudars namn ala1gudazs1namn ‘the names of all gods’.

Also, keep in mind that not every grammatical ending with an r in it had a z in Primitive Germanic. The –are suffix used to indicate the performer of an action was originally taken from Latin. It had an r in that position to begin with, so nouns with that suffix should use Raidho (not Elhaz) there: krigare krigare ‘warrior’, älskare elskare ‘lover’.


This is the correct rune for s in most situations (the possessive –s has already been covered above, however). Swedish lacks the voiced s/z sound that exists in West Germanic languages, so transcribing that one really isn’t an issue.

On the other hand, one particular Swedish sound (which foreigners usually have trouble pronouncing!) can be spelled in a number of ways, most of which involve s plus some other consonant: sj, sk, skj, or stj. In a runic inscription based on sound rather than Roman-letter spelling, I would consistently write this sound as Sowilo plus Jera: själ sjel ‘soul’; skilja sjilja ‘separate’; skjul sjul ‘shed’; stjärna sjerna ‘star’.


Using this stave correctly is simpler in Swedish than in English: it represents the sound at the beginning of words like tala tala ‘speak’ and träd tred ‘tree’. Modern Swedish lacks the th-sounds in English "that thing", so transcribing them should not normally be an issue. (Purists might still want to use Thurisaz for spelling some Norse names in an inscription, as I have noted under that rune.)

The combination tj- was originally pronounced as written. However, it now sounds like the "soft" ch in German ich (or the kj combination/"soft k" used in other Swedish words). Theoretically, it COULD be written Kenaz+Jera in runes; but some dialects do have a noticeable t-sound in words that begin with tj-. Not everyone pronounces tjära tjera ‘tar’ and kära kjera ‘dear [pl.]’ exactly alike…


In Modern Swedish inscriptions, Berkano has two uses:

  1. A b-sound like the one used in English: bara bara ‘only’; bok ‘book’.
  2. A v-sound, if it follows a vowel: leva leba ‘live [exist]’; stavar stabaz ‘staves’. The v in such words usually derives from a b in Primitive Germanic, so a runic spelling with Berkano is justified in that case. (As I stated before under Wunjo, not every Swedish v has the same origins: sometimes one rune feels more right historically, sometimes the other does.)


The sound value of this rune is the same in English and Swedish inscriptions: morgon morgon ‘morning’; komma koma ‘come’. English-speaking people should note, however, that Swedes DO pronounce m in the combination –mn (and the runic spelling should reflect that): namn namn ‘name’.


This rune represents the l sound in land land (no translation needed); mellan melan ‘between’; and hel hel ‘whole’. The letter l is silent in a few conventional spellings, however, so NO stave should be used there—

  1. In the combination lj- at the beginning of a word: ljus jus ‘light’; ljud jud ‘sound’.
  2. In the words karl kar ‘man’ and värld werd ‘world’. Note that "Karl" as a proper name, though, keeps its final l-sound (and the Laguz-rune to match): karl..


While discussing MISuses of some other runes, I have already given some situations where Ingwaz would be called for in transcribing Swedish. I will repeat them here for the sake of completeness.

  1. The combination ng usually corresponds to that rune, of course: ring riN (no translation needed); sjunga sjuNa ‘sing’. (English-speaking people take note: Swedes NEVER insert a hard g after the ng sound. Although transcribing "finger" as fiNger matches English pronunciation, the Swedish one would be written fiNer with no Gebo-rune.)
  2. In the combination –nk, the letter n actually sounds like ng and should be transcribed accordingly: blank blaNk ‘shining’; tänka teNka ‘think’.
  3. The g in the combination gn may also correspond to Ingwaz (not Gebo) phonetically. This never happens at the beginning of a word, but is common in other positions: vagn waNn ‘wagon’; hägna heNna ‘enclose, protect’.


In a Swedish inscription, Dagaz represents ONLY the hard d sound: its namesake-word, dag ‘day’, would be written in full as dag. The only potential misuse I can see is in the combination dj- at the beginning of a word. Because the d is silent then, NO stave should be used to represent it: djup jup ‘deep’; djärv jerf ‘bold’.

knotwork01.gif (2510 bytes)

Return to the language page