Cookies on our website

We use cookies on this website, mainly to provide a secure browsing experience but also to collect statistics on how the website is used. You can find out more about the cookies we set, the information we store and how we use it on the cookies page.




Menu Search
Dokumentation // Documentation ○. at 6.3.2019 2. English 3. Access to wordlist, slips, articles, indexes 3. Orthography

3. Orthography

This is not currently part of the peer-reviewed material of the project. Do not cite as a research publication.

ONP : orthographic norm

ONP's orthographic norm is a reconstruction of the state of Old Norse (medieval Icelandic/Norwegian) ca. 1200-1250 (closer to 1200 than 1250).

Where a choice between an Icelandic and a Norwegian standard is necessary, the more conservative of the two, most often the Icelandic, is chosen. But where the Norwegian form is the more conservative (e.g. in the case of non-lengthening of vowels before certain consonant groups), the Norwegian is selected.

Non-assimilated foreign words (labelled ‘alien.’) are presented in their native spelling, based as far as possible on the orthography of standard dictionaries of the relevant language. Classical Latin is chosen in preference to medieval.

The more integrated loanwords are adapted to the Old Norse norm and are treated on a par with the indigenous vocabulary.

ONP’s orthographic norm aligns closely with the traditional norm, as it has developed over the years in normalised text editions (e.g. Íslenzk fornrit), in dictionaries (e.g. Heggstad, Hødnebø & Simensen: Norrøn ordbok, 1975), and in other handbooks (reference will be made below to Noreen: Altisländische und altnorwegische Grammatik, 4. ed., 1923, and Larsson: Ordförrådet i de älsta islänska handskrifterna, 1891).

The most noticeable deviation from normal practice in editions and dictionaries is the choice of the graphs ǽ and ǿ for the long æ- and ø- phonemes (in harmony with Noreen’s grammar and standard linguistic usage). The choice of ǽ and ǿ (rather than æ and oe) was made partly with a view to the spelling of the manuscripts but mainly on the basis of practical and pedagogical considerations: as a result there is an acute accent over all long vowels in normalised text and unfortunate confusion between æ and oe (esp. in italics) can be avoided.

The orthographic variation which the sources themselves demonstrate arises partly because the texts cover a span of approx. 400 years during which major linguistic changes emerge, partly because of the regional variation, and partly because of the non-standardised spelling of the individual manuscripts.

The orthographic norm which is used in headwords and other normalised text in ONP is therefore an abstraction that is not found in its full extent in any of the single manuscripts which consti tute ONP’s corpus. Especially normalisation of headwords in their grammatically cardinal forms sometimes entails the construction of forms the existence of which is not recorded in the available data.

Normalisation of Old Norse texts removes them from their chronological and regional context and blurs their individuality, but standardisation can bring the literature closer to its readers. When, for example, a medieval text is presented in a Modern Icelandic guise, a bridge is formed to the Icelandic reader. Similarly, standard Old Norse normalisation gives students from all over the world a means by which to approach the texts with the help of dictionaries, grammars and other reference tools. In order to provide this shared and open access this dictionary normalises its headwords and Old Norse phraseology. The citations themselves, however, are retained as far as possible in their unnormalised form; in the edited articles they are presented in a manner that is as close to that of the surviving text as the manuscripts and the editions allow.

(ONP’s normalisation forms the backbone of the standard recommended by MENOTA = Medieval Nordic Text Archive // Arkiv for nordiske middelaldertekster

(The description is not final and is subject to revision at any time. Comments and suggestions for improvements may be sent to

The Old Norse alphabet

(incl. letters in foreign words)

a/á, b, c, d/ð, e/é, f, g, h, i/í, j, k, l, m, n, o/ó, p, q, r, s, t, u/ú, v, x, y/ý, z, þ, ǽ, ø/ǿ, ǫ

Long and short vowels are not alphabetically separated, but in minimal pairs the short vowel is listed before the long; thus hlið comes before hlíð.

Normalisation rules

(In the illustrative examples a headword is rendered in bold type and declined forms are underlined).



All long vowels, incl. ǽ and ǿ, are represented with an acute accent

The lengthening of the vowels a, o, u, ǫ before the consonant combinations lf, lg, lk, lm, ln, lp, ls, which develops in Icelandic and West Norwegian in the course of the period covered by ONP, is not shown, thus:

halfrǫlftalftiralfta-, folginnmjolkholmrǫlnalniralna-, hjalpa,halsbolstr

Lengthening is not shown in arni (dat. of arinn)

Vowels of the preterite of the third class of reduplicative verbs are normalised as short:

blett (blanda), fekk (fá), fell (falla), gekk (ganga), helt (halda), hekk (hanga)

On etymological grounds, length is shown in:



Earlier  is represented as á:


Where nasality has resulted in representations with both á and ó, the ó form has the status of an alternative form

náttnóttánón, ...; ambáttambótt, ǫmbótt

The negative prefix is ó- (not ú-) (cf. Larsson; Noreen § 112, 1)

The preposition/prefix is órór- (not úr, úr-) (cf. Larsson)

An etymological distinction is maintained between ǽ and ǿ

An etymological distinction is maintained between ø and ǫ


The weakly stressed palatal vowel is i (not e)

The weakly stressed velar vowel is u (not o)


The standard suffix is -endi (not -indi) (alternative forms: -yndi, rarely -andi)

The standard suffix is -ligr, -liga (not -legr, -lega)

The standard suffix is -semi (alternative: -simi)

The definite article is represented by inn (not enn) and hinn (<> dem. pron. hinn)

A choice between word final -un/-an is made on the basis of an examination of all ONP occurrences of a given noun, both simplex and compound


i(í)/j, u(ú)/v

The letters i and j, which in the manuscripts are non-phonemic graphic variants, are distributed as follows:

i/í for the full vowel:

binda, innaní

but j at a syllable margin as the consonant or half-vowel:

jǫrðfjár (fé), hjartaveljanesjum (nes)

There is a similar distribution for u/ú:


as opposed to v:




Representation of a long consonant in the root morpheme is upheld as a double consonant in compounds:


thus in declined forms:

allriallt (all-), falls (fall), felldi (fella), kenndi (kenna), manns (mann-), minnstr (minn-),vinnrvannskvannt (vinna), þynntr (þynna)

and derivatives:

áttján (átta), byggð (byggja), grennd (grannr), minnka (minn-), skemmtaskemmtun(skammr), þurrka (þurr)

In forms like hugði (hyggja), lagði (leggja) the root morpheme is originally short, but there is lengthening (double consonant) in the infinitive.

In the following forms there is a single consonant:

framfrum-, konstrkunst

ð/d/t (principally) in declension endings and derivative suffixes

The following rules apply to the representation of the dental obstruent in declension endings and derivative suffixes:


1. after a vowel, where the dental precedes a vowel or r, or is in word-final position:

kallaðikallaðrkǫlluð (kalla)

2. after rð, where ð in the root assimilates with the suffixed dental to form a single ð:

virði (pret. of virða)

3. after root-final b, f, g, ng, k, m, p, r (regardless of syllable length):

kembði (kemba), helfð (halfr), horfði (horfa), uggði (ugga), lengð (langr), þyngð (þungr), sekð (sakr), spekð (spakr), vakði (vaka), merkði (merkja), kvǽmð (koma), dreymði(dreyma), glapði (glepja), dýpð (djúpr), kippði (kippa), steypði (steypa), stýrði (stýra), gerðr(gera)

exceptions: þurfti, keypti (→ t item 4, below), gift etc. (→ t item 5, below)

4. after a short root syllable ending in l, n:

fjǫlðihulði (hylja), vesǫlð, þolði (þola), unði (una) (cf. Larsson lþ; debated origin)

exceptions: seldi, skyldi etc. (→ d item 4, below)

5. The following word which is not covered by the rules in this section is written -nð:

synð (cf. Larsson nþ)

On aldr, einfaldr, farald, grund, heldr, hold, -kund-, stund, verǫld, ǫld → d item 5, below


1. after root-final d, where the radical d assimilates with the suffixed dental to a single d:

eldi (pret. of elda), sendi (pret. of senda)

2. after root-final ð, where the radical ð assimilates with the suffixed dental to dd:

breidd (breiðr), eyddi (eyða)

exception: after r (cf. ð item 2, above)

3. after a long root syllable ending in l, n:

efldi (efla), felldi (fella), girndisk (girnask), hvíld (hvíla), sigldi (sigla), sǽld (sǽll), brenndi(brenna)

exception: mǽlti etc. (→ t item 3, below)

In cases where l or n are preceded by another consonant, the addition of a suffixed dental can result in a phonetically complex three-consonant group. A group of this type can be simplified in two ways: by loss of a consonant or by metathesis; thus we find forms like girðisk for girndisk (girnask), elfði for efldi (efla) and silgði for sigldi (sigla), i.e. the resolution of the consonant group allows for a pronunciation with ð rather than d. In ONP these forms are nonetheless normalised with d (girndisk, efldi and sigldi), in accordance with the rules above.

4. in the following declined forms after a short root syllable ending in l, n (cf. Noreen § 238, anm. 5):

seldi (selja), skyldi (skulu), vildi (vilja), mundi (munu)

(cf. ð item 4, above).

5. The following words which are not covered by the rules in this section are written with ld, nd (cf. Larsson l(l)d, nd):

aldreinfaldrfaraldgrundheldrhold, -kund-, stundverǫldǫld


1. after root-final t, where the radical t assimilates with the suffixed dental to a single t:

gifti (pret. of gifta), lyfti (pret. of lyfta)

2. after root-final s:

brosti (brosa), kyssti (kyssa)

3. after root-final l, n which were unvoiced in proto-nordic (cf. Noreen § 238, 2, b):

hellti (hella), mǽlti (mǽla), rǽnti (rǽna), vǽpnti (vǽpna)

(cf. d item 3, above)

4. where the declension ending after f, p demonstrated t already in the earliest manuscripts (reflecting Proto-Germanic t, cf. Noreen § 238, anm. 7):

þurfti (þurfa), keypti (kaupa)

and where the derivative suffix reflects an earlier t:

gift (*gefti), greft (*graftu), gröftr, haft (*haftu), -skaftr, -skeftr, -skiftrþrift (*þrifti), þurft (*þurfti)

Dental + s : z

d, ð, t in the root before an s in the declension ending or derivative suffix most oftenresults in an assimilation:

ds, ðs, ts > z

1. the reflexive ending -sk:

kvezkkvazk (kveðask), lízkleizk (lítask), sezk (setjask), komizk (< komit+sk), látizk (< látit+sk)

2. the 2. pers. sg. ending -st:

veizt (vita)

3. superlatives in -str:

helztr (heldr), neztr (neðri), beztr (betri, *bat-), ýztr (ýtri)

4. derivatives:

benzl (band), íslenzkr (Ísland), breizl (breiða), gerzkr (cf. Garðar, Garðaríki), góz (sb. n. ~ góðr adj.; evt. eng. låneord), greizla (greiða), gǿzka (góðr), herzla (herða, harðr), jázi(játa), leizla (leiða), lýzkalýzkaðr (ljóð), vestfirzkr (fjǫrðr), beizl (bíta), veizla (veita),vǽzka (vátr)

dsðsts is written as an exception
e.g. where a genitive -s does not affect the orthography of the root:

lands (land), góðs (góðr), báts (bátr)

z is not written in forms where ts is the result of an assimilation of tns:

vatns-, botns-

In consonant groups llsnns no allowance is made for a parasitic (i.e. non-etymological) stop before s; z is therefore not written:

ellstr (ellri), kallsamannsmennskrminnstr

beisk- rather than beizk- is written, cf. Larsson; basis for this normalisation is a moot point.


1. In syllable-initial position f is written for f:


and v for v:


2. In other positions v is written after an initial consonant:

dveljahverrkvernsvala (cf. item nekkverr)

otherwise f:


3. v in root-final position (of the type which is omitted before certain final vowels) is written v:

hǫggvasjóvar (sjór), ǫrvar (ǫr), tívar (pl.)

4. v in derivative endings is written v:


ft/pt, fs/ps

In the consonant groups ft/pt and fs/ps the norm is f regardless of whether or not a historical p can be traced.

ft and fs are therefore written, for example, in:

aftannaftrgrǫftr (grafa), skifta (*skiptian), skrift (~ ags. scrift, nty. schrift, lat. scriptum), svifta (*swipatjan); refsa (*rafisjan), glefsa (*glapisjan), ufsi ‘storsej’ (*ufisan), ufsi ‘tagskæg’ (*oupselo)

except where a stem remains unchanged before an ending, resulting in pt and ps:

djúpt (djúpr), steypti (steypa), skips- (skip)

just as the original consonant in a number of clearly recognisable cases incl. loanwords is retained:

glópska (glópr), glǿpska (glǿpr), egipzkr (egiptaland)

In ONP 1-2 a more traditional normalisation was applied with a high frequency of
pt (andps), thus for example aptann, aptr, áraskipti (and bops), while in ONP 3 a new rule was applied: the consonant groups were represented in accordance with their etymology where this could be proven; thus the majority spelling wasexcept where a historicalcould be traced. This distinction, however, lacks a clear phonological foundation, and all material on ONP's new Internet portal will show normalisation with ft and fs.

ks, gs : x

x is written for ks < *hs:

lax (*lahsa-), víxla

and also in foreign words and loanwords:

crucifixus, crux, dextilldextilla

The genitive -s, however, does not affect the orthography of the root; thus ksgs:

þaks (þak), loks (adv. = gen. of lok), heilags (heilagr)

and gs is written in derivatives in accordance with the etymology:

fylgsni (~ folginn, pret. part. of fela), vígsla (~ vígja)


The phonetic development nnr > ðr has a variety of manifestations, and ONP’s normalisation reflects this diversity.

In a number of words and wordforms ðr is written throughout:

maðr (mann-), suðr (sunn-), aðrir (annarr)

while in others variation is accepted nnr/ðr:

brunnrbruðr (alternative form), munnrmuðr (alternative form), brennr/breðr(alternative declined form)

yet the following word, which is only attested with -nn-, is allotted a single cardinal form in nnr (not ðr):


Reflexive forms are written with -sk

Capital/lower-case letters

In ONP’s norm all headwords and parts of headwords are presented in lower case:


One or more word(s)

(random examples)

ONP writes as a single compound:


but separates:

upp á, á móti, at fram, í gegn, í hjá, í mót, með fram, út borinn, vel borinn


Log in

This service is only available to members of the relevant projects, and to purchasers of the skaldic volumes published by Brepols.
This service uses cookies. By logging in you agree to the use of cookies on your browser.