Swearing an Oath in Old Germanic Society

Swearing an Oath in Old Germanic society[1]

The reconstructed lexicon of the prehistoric language called Proto-Indo-European provides the linguist with a limited window on Indo-European concepts of law. From the Proto-Indo-European judicial lexicon Protogermanic preserved some interesting words. The one which concerns us in this essay is the word *aiþaz, the ancestor of modern English oath and the verb which is connected to this noun, namely PGmc. *swarijan-. The Protogermanic form *aiþaz has cognates in Old Irish oeth, Greek οἶτοϛ and Tocharian B aittaṅka but it only has the lexical specialisation “oath” in the western languages, e.g. Germanic and Celtic. In the early twentieth century scholars assumed that the Germanic word was loaned from Celtic, because they thought that the Celts had a higher level of civilization than the Germanic peoples in the early northern European iron age. Nowadays it is acknowledged that both the Celtic and the Germanic word can go back to Proto-Indo-European and there is no need to assume borrowing from one language into the other. The mentioned cognates ultimately go back to Proto-Indo-European *h1óitos which is derived from the root “to go” *h1ei (cfr. Latin ire and Greek εἶμι and Gothic iddja), which points to a meaning “a ritual walking”. Cultural attestations of Indo-European oath-taking by walking between slaughtered animals perhaps colour the etymology somewhat more and is reasonably plausible because of the Old Swedish attestation ed-gång meaning “oath-walking” . The earliest Germanic attestation is Gothic aiþs (glossing Greek ὅρκος) from Wulfila’s fourth century bible translation:

Lukas I, 73
jah gamunan (…)aiþis þanei swor
wiþra Abraham attan unsarana
(and to remember (…) the oath
that he swore to Abraham, our father)

Matteüs XXVI 72
jah aftra afaiaik (laugnida) miþ aiþa
swarands þatei ni kann þana mannan
(and again he denied with an oath
that he didn’t know the man)

Matteüs V 33
aftra haisdeduþ þatei qiþan ist þaim airizam:
ni ufarswarais, iþ usgibais fraujin aiþans þeinans
(again ye heard it said from the elders: don’t
break your oath, but give your oaths to the lord)

Marcus VI 26
jah gaurs warþans sa þiudans in þize aiþe jah
in þize miþanakumbjandane ni wilda izai ufbrikan
(and becoming sorrowfull the king didn’t want to
reject her considering the oaths and the sake of
those who were attending)

In Gothic the word aiþs is in all the four attestations accompanied by the verb swaran[2] in its proximity (cfr. ModE to swear; cognate to Sanskrit svárati, Old Church Slavonic svarŭ, going back to Proto-Indo-European *sṷer meaning “to say, to speak) and is also found with the modal prefixes bi- and ufar- with biswaran meaning “to implore” and ufarswaran meaning “to swear falsely”. The fact that bishop Wulfila used the term in rendering the greek ὅρκος suggest that “swearing an oath” was not purely a pagan religious phenomenon amongst the Goths, otherwise Wulfila would surely have invented a calque. The information we can gleam from the philological context of the Gothic bible translation learns us that in the Germanic languages “swearing an oath” was something that one did against someone (considering Luke I 73 wiþra Abraham). An possible cognate to Gothic aiþs may be Gothic aiþei “mother” (glossing Greek μήτηρ) which is also found in Old High German fōtar-eidi “nurse”, Old Icelandic eiða “morther” and Middle High German eide “mother”. This would mean that this word for mother literally meant “the one with the oath” which probably distinguished the lawful wife from the concubines. The phenomenon that oathtaking across kinshipgroups connected by marriage was common in Germanic society, we will also encounter in an Old English compound furter on.

In the Beowulf epic the combination “to swear an oath” is also used, suggesting that the word was part of the poetic register of the Anglosaxons.

Beowulf (470-473)

Siððán þa fæhðe fēo þingode
Sende ic Wulfingum ofer wæteres hrycg
Ealde mādmas; hē mē āþas swōr
afterwards the feud I settled with wealth
I send to the Wulfings over the back of the water
ancient treasures: he swore oaths to me


Ic on earde bād
Mælgesceafta   heold mīn tela,
Ne sōhte searonīðas ne mē swōr fela
āða on unriht
I awaited on earth
destined events I held my property well
nor did I seek battle-hostilities nor did I swear
many oaths unjustly


Fin Hengeste
Elne unflitme āðum benemde
Finn to Hengest
with undisputed valour proclaimed with oaths

In the third example from the Beowulf an other verb is used, namely benemnan which means to proclaim. The verb is used one other time in the Beowulf and than in reference to the proclaiming of justice by princes whilst facing the last judgment.


Swá hit oð dómes dæg      díope benemdon
þéodnas maére
thus this till doomsday     they proclaimed deeply
the famous kings.

In the Beowulf also the nouns āðsweord  “oath-swearing” and āðumsweoras “father-in-law and son-in-law” are attested. The first is also found in OHG eidswart and biswart seems to continue *swardiz. The second compound, like Gothic aiþei, also refers to oath-taking that accompanied the marriage-bonds between kinshipgroups. Apparantly the bond and the obligations to abandon feuding that a marriage brought along for two kinshipgroups had to be confirmed by oath-taking.  In Old High German another term is found, eidum meaning “son in law”

In the Old Saxon bible epic Heliand, a vernacular retelling of the life of Christ, the word ēð is also used, as is the verb swerian. As we will see this is also the verb used in West Frankish, one of the vernaculars of the Carolingian Franks. The start of fitte 18 in the Heliand begins with a whole sermon against the breaking of oaths and also names the term for a perjurous oath, namely mēnēð < *main-aiþaz, preserved in modern Dutch as meineed. Because it is contemporanous with Carolingian society the passage is worth quoting in full. A connection with the Carolingian programme aimed at reducing the oath-taking in society to swearing oaths of allegiance to the king may be present alongside with the objective of getting rid of the possible pagan religious context that the taking of oaths may still have had in the newly conquered Saxon areas.

Old Saxon Heliand

Ōc is an them ēo gescriban
Uuārun uuordum so gi uuitun alle,
that mīðe mēnēðos mancunnies gihuuilic
ni forsuuerie ina selbon huuand that is sundie te mikil
farlēdid liudi an lēðan uueg
than uuiliu ic iu eft secgan that sān ne suuerie neoman
ēnigan ēdstaf eldibarno
ne bi himile themu hōhon huuand that is thes hērron stōl
ne bi erðu thar undar huuand that is thes alouualdon,
fagar fōtscamel nec ēnig firiho barno
ni suuerea bi is selbes hōbde huuand he ni mag thar ne suuart ne huuīt
ēnig hār giwerkean būtan sō it the hēlago god
germarcode mahtig; bethiu sculun mīðan filu
ērlos ēðworðo. Sō huue sō it ofto dōt
sō wirðid is simbla wirsa huuand he imu giuuardon ni mag
Bithiu scal ic iu nu te uuârun uuordun gibeodan
that gi neo ne suerien suuîðoron êðos,                              

Also is in those books always written
with true words so ye all knoweth
that may avoid perjuries each of mankind
may he not forswear himself for that is a sin too great
it tempts the people to the hateful road
then I want to tell you again that no one may swear
any oathmark of mankind[4]
nor by the high heaven for that is the seat of the lord
nor by the earth beneath it for that is the allmighty’s
beautiful footstool and may any of mankind
swear on his own head for he can not make
any hair black or white except for when the holy god
mightily makes it so: therefore many should avoid
honourless oath-words. Whoever does it thus often
so it will worsen him always because he cannot hold it
Therefore I shall truly order with words
that ye shalt not swear stronger oaths

In connection to Carolingian decrees on oath-taking it may also be interesting to look at the Old High German Mainzer Beicht, a confession list preserved in a tenth century manuscript (dated around 950) but which may go back to a ninth century example.

Mainzer Beicht

Ih gihun gode almahdigen unde allen godes
engilon unde allen godes heilegon unde dir
godes boden allero minero sundino (…) meinero

I confess to god allmighty and to all the angels of
god and to all god’s saints and to you, god’s messenger
all my sins (…) (and) my oaths

Another famous example is the Old High German Priestereid, containing an oath in a religious context.

Priest oath

De sacramentis episcopis qui ordinandi sunt ab eis

Daz ih dir hold pin .N. demo piscophe.
So mino chrephti enti mino chunsti sint.
Si minan uuillun. Fruma frūmenti enti scadun
uuententi kahorich enti kahengig enti statig
in sinemo piscophtuome. So ih mit rehto aphter

That I am loyal to you, the bishop; according to my
strenghts and my wisdom, may I be with my will,
whilst promoting benefit and avoiding damage,
obedient and consistent and steady. Thus I will do
rightly according to the canon
canone scal.

In western Francia it is harder to turn towards the vernacular in search for Germanic terms connected to the swearing of oaths. The Germanic dialects that may have survived into the eighth century in northern Gaul beyond the current language border between French and Dutch were under constant pressure from Galloromance, the vernacular language of the Romance speaking Franks. However, when the Franks settled in northern Gaul in the fifth and sixth century the prestige of their language was still well recognized. A recent lexical study of the Frankish loanwords in Old French makes it plausible that a lot of the idiosyncracies of Old and Modern French are caused by (a considerable amount of) Romance speakers learning Frankish, not the other way around. The Franks ofcourse also brought their legal terminology along. The first and most important testimony to the language and laws of the Franks has been preserved in the Legis Pactus Salicae, an early sixth century law code drawn up by the Salian Franks. In the lawcode vernacular glosses are preserved and because they are introduced by Mallobergo (which you can translate as “on the mountain of the law court”) they are called the Malbergische Glossen. One of these glosses is uuedredo, which doesn’t look Old Frankish but may be a Carolingian (West Frankish?) interpolation of a vernacular law term. The gloss  uuedredo doesnt’t have a satisfactory explanation but one explanation might be that it represents a Frankish nominative plural of the compound *wiþra-aiþa, meaning “counter-oath”, which doesn’t look that weird if one recalls the Gothic expression swaran aiþans wiþra “swearing oaths against one”. In the Langobardic Laws we do have the word aido preserved, connected with a lawarticle on compurgation. In Frankish legal sources we also find the term leudosamium being connected to the taking of oaths. However, considering the meaing of the word *leudi-samjō being something like “people’s peace”, the word clearly had more to do with the desired effect of the oath than the actual oath itself and must have acquired a degree of metanymia in elitist discourse.

The Gallo-Romance language itself preserved the latin word jurare and passed it on to Old French as jurer which doesn’t make clear whether the word endured in the lower strata of the Romance speach community during the entire Proto-French period or that the word was inserted into Gallo-Romance from judicial speech. However, in the Strassburger Eide the verb jurare is used, which does make a continued existence of the verb in the early medieval Romance plausible.

Strassbourg Oaths

Si Lodhuuigs sagrament quę
son fradre Karlo iurat, conservat
et Karlus meos sendra de suo part
non los tanit si io returnar non
l’int pois…

If Ludwigs oath that
his brother karl swore, holds
and karl, my lord, on his part
does not hold it, when I may not
dissuade him from it.

Oba Karl then eid, then er
sīnemo bruodher Ludhuuīge
gisuor, gileistit indi Ludhuuig
mīn herro then er imo gisuor
forbrichit, ob ih inan es
iruuenden ne mag…

If Karl holds the oaths which he
swore to his brother Louis and
Louis, my lord, who swore it to
him, breaks it, when I may not
dissuade him from it…

The vernacular Romance equivalent of Frankish *aiþ is Latin sacramentum which yielded Old French sairement leading to Modern French serment. The verb which accompanies the noun sacramentum  in Old French, however, was loaned from Germanic; Old French escharir goes back to Old Frankish *swarjan via a romanized form *scuarire (the soundchange /sw-/ > /skw-/ is to be expected), possibly with influence from Old Frankish *skarjan. That the noun *aiþ “oath” was in some stage also loaned into Romance is clear from a fourteenth century attestation from Liège; the noun afforat must go back to Old Frankisch for-aiþ, cognate with East Franconian fraaidhi and Middle High German freidi . The fact that for these terms concerning oath-taking an Old Frankish word became current in the whole Proto-French speaking area (Old Provence escharida)  , in my opinion, points to the non-Roman characteristics of oath-taking.


Bourciez, Edouard, précis historique de phonetique française nouvelle collection a l’usage des classes III (Lille 1921).

Braune, Wilhelm, Althochdeutsches Lesebuch (17th edition 1994 Tübbingen; 1875).

Bruckner, Wilhelm, Die Sprache der Langobarden Quellen und Forschungen zur Sprach- und Literaturgeschichte der germanischen Völker 75 (Strassbourg 1895).

Eckhardt, Karl August ed., Pactus Legis Salicae (MGH LNG IV I) (Hannover 1962).

Eckhardt, Karl August ed., Lex Salica  (MGH LNG IV II) (Hannover 1967).

Gamillscheg, Ernst, Romania germanica. (Berlin:1970).

Holthausen, F., Altenglisches Etymologisches Wörterbuch (Heidelberg 1934).

Julia H. Smith, Europe after Rome; a new cultural history 500-1000 (Oxford 2005).

Lehmann, Winfred P., A Gothic etymological dictionary (Leiden 1986).

Mitchell, Bruce and Fred C. Robinson eds., Beowulf; an edition (1998).

Sievers, Eduard, Heliand Germanistische Handbibliothek IV (Halle 1878).

Schützeichel, R., Althochdeutsches Wörterbuch (6th edition; Tübbingen 2006).

Streitberg, Wilhelm ed., Die Gotische bibel (Heidelberg 1950).

Uhlenbeck, C.C., kurzgefasstes etymologisches Wörterbuch der gotischen Sprache (Amsterdam 1900).

Voretsch, Karl, Altfranzöschises Lesebuch zur Erläuterung der altfranzösischen Literaturgeschichte (1921).

[1] All the translations from the Old Germanic languages to modern English are my own, mostly because of a lack of faithfull translations. I prefer to keep my translation of poetry (just as I did with Otfrid’s praise of Louis) as close to the original language as possible.

[2] Gothic swaran is probably secondary to earlier *swarjan < *swarijan- as suggested by OS and OE swerian, OHG swerren, OIc. sverja. This PGmc. *swarijan- is a secondary causative/iterative formation in *-e̯ie- to the root *su̯or- that was adopted into the VI class of strong verbs, which is unusual because class VI mainly has primary *-̯ie-/*-e̯ie-formations.

[3] Litterally: “the business of the assembly” but apparantly it could also be used in a metaphorical sense

[4] the genitival construction depends on neoman

[5] note the stabreim, the original Germanic alliteration model, which strongly suggest that the oath leans on older vernacular formulas.

4 responses to “Swearing an Oath in Old Germanic Society

  1. Dear Peter Alexander,
    i very much enjoyed reading your blog. i was wondering if you could tell me whether you think swaran or svárati may in any way be related to Sanskrit svar, Avestan hvar, Iranian kwar? thank you in advance for imparting your opinion,

  2. Dear Peter Alexander,

    Concerning the Hittite connection, I will take your word for it. My knowledge of Comparative Indo-European Linguistics is very limited, but I understand enough of it to see your point. Thank you.

    At any rate, I was already quite content with my own speculation. 😉

    Interesting blog you have by the way. I’ll be sure to read your future posts.

  3. P.A. Kerkhof

    Dear Olivier,

    Thank you for your comment. You’re right that the etymology may seem what murky, but, as you suggest, the link to greek οἴτος gives us a better semantic transition. That we might want to see a closer original meaning to the Greek sounds very plausible to me. However, if we want to posit a link with Hittite, to my mind, problems pop up. First there is the attestation. Puhvel quotes ḫai- as the form, according to Orel (I dont have Puhvel’s dictionary, so I have to take a leap of faith :))), however, the interpretation of the root as ḫai- is problematic. Most of the attestations point to ḫa-, the one attestation which shows a -i- we dont expect is probably secondary (Kloekhorst 2008). This leaves us with ḫa- itself. This must go back to either *h2VH or *h3VH, since *h1 would not show up in Hittite, so our connection with Greek οἴτος and *h1ei is gone. Secondly, the only IE cognate to the Anatolian root is Latin ōmen, which must go back to *h3eHmen or *h2oHmen, securing the Hittite root as either *H3eH- or *H2eh3-. This root is not compatible to the Germanic and Celtic attestation of the word for “oath”. So it seems, the current etymology is the best we have…

  4. Interesting thoughts…

    You point out that *aiþaz is usually combined with the verb *swaranan. If I may freely speculate from the sideline, perhaps originally the verb bore most of the meaning, in the sense that “to swear an oath” originally meant “to proclaim a fate” and thus, since one’s word was one’s honour, “to guarantee a fate” (in relation to someone else, i.e. wiþra someone else). This would bridge the semantic gulf beween Proto-Germanic *aiþaz and Old Irish óeth on the one hand and Greek oĩtos ‘fate’ on the other, if they really are cognates, and make the derivation of *aiþaz/óeth from the PIE root *h1ei- ‘to go, to walk’ slightly less murky.

    (The Old English pairing of áð with benemnan ‘proclaim’ neither confirms nor refutes this speculative etymology.)

    On the other hand, is the derivation of *aiþaz (and cognates) from *h1ei- really that certain? Orel, following Puhvel, also sees relation to Hittite ḫai- ‘to believe, to trust’ as plausible. “To swear an oath” might then mean something like “to proclaim/guarantee (a) loyalty”.

Leave a Reply

Your email address will not be published. Required fields are marked *