The third son of Mannus will occupy us even longer than his brothers. Ermino's posterity completes the cycle of the three main races of Germany: Ingaevones, Iscaevones, Herminones. The order in which they stand seems immaterial, in Tacitus it merely follows their geographical position; the initial vowel common to them leads us to suppose an alliterative juxtaposition of the ancestral heroes in German songs. The aspirate given by the Romans to Herminones, as to Hermunduri, is strictly no part of the German word, but is also very commonly retained by Latin writers of the Mid. Ages in proper names compounded with Irmin. In the name of the historical Arminius Tacitus leaves it out.
[Arminius is a latin cognomen and is not related to Irmin or to the Teutonic name Herman. The native name of Arminius is not recorded, but the names of his family are known: Segimerus, his father; Inguiomerus, his uncle; Segestes, his father-in-law; Segimundus, his brother-in-law; and Thusnelda, his wife. Since his father's name is Segimerus, it is safe to say that the native name of Arminius also had the "Sigi" element. This of course, brings to mind, Siegfried. WW ed]
As with Inguio and Iscio, we must assign to the hero's name the otherwise demonstrable weak form Irmino, Ermino, Goth. Aírmana it is supported by the derivative Herminones, and even by the corruptions ‘Hisicion, Armenon, Negno’ in Nennius (see Suppl.). Possibly the strong-formed Irman,
Irmin, Armin may even be a separate root. But what occurs far more frequently than the simple word, is a host of compounds with irman-, irmin-, not only proper names, but other expressions concrete and abstract: Goth.
Ermanaricus (Aírmanareiks), OHG. Irmanrîh, AS. Eormenrîc, ON. Iörmunrekr, where the u agrees with that in the national name Hermundurus; OHG. Irmangart, Irmansuint, &c. Attention is claimed by the names of certain animals and plants: the ON. Iörmungandr is a snake, and Iörmunrekr a bull, the AS. Eormenwyrt and Eormenleáf is said to be a mallow, which I also find written geormenwyrt, geormenleàf. Authorities for irmangot, irmandiot, OS. irminthiod, irminman, irmansûl, &c., have been given above, p. 118. A villa Irmenlô, i.e., a wood (in illa silva scaras sexaginta) is named in a deed of 855, Bondam's charterbook, p. 32. silva Irminlô, Lacombl. 1, 31.
In these compounds, especially those last named, irman seems to have but a general intensifying power, without any distinct reference to a god or hero (conf. Woeste, mittheil. p. 44); it is like some other words, especially got and diot, regin and megin, which we find used in exactly the same way. If it did contain such reference, Eormenleáf would be Eormenes leáf, like Forneotes folme, Wuotanes wec. Irmandeo then is much the same as Gotadeo, Irmanrîh as Diotrîh; and as irmangot means the great god, irmandiot the great people, iörmungrund the great wide earth, so irmansûl cannot mean more than the great pillar, the very sense caught by Rudolf in his translation universalis columna (p. 117).
This is all very true, but there is nothing to prevent Irmino or Irmin having had a personal reference in previous centuries: have we not seen, side by side with Zeus and T_r, the common noun deus and the prefix t_-, (p. 195-6) ? conf. p.339. If Sæteresdæg has got rubbed down to Saturday, Saterdach (p. 125), so may Eritac point to a former Erestac (p. 202), Eormenleáf to Eormenes leáf, Irmansûl to Irmanessûl; we also met with Donnerbühel for Donnersbühel (p. 170), Woenlet for Woenslet, and we say Frankfurt for Frankenfurt [Oxford for Oxenaford, &c.]. The more the sense of the name faded out, the more readily did the genitive form drop away; the OHG. godes hûs is more literal, the goth. guþhûs more abstract, yet both are used, as the OS. regano giscapu and regangiscapu, metodo giscapu and metodgiscapu held their ground simultaneously. As for geormen = eormen, it suggests Germanus (Gramm. 1, 11).
It is true, Tacitus keeps the Hermino that lies latent in his Herminones apart from Arminius with whom the Romans waged war; yet his famous ‘canitur adhuc barbaras apud gentes’ applied to the destroyer of Varus, might easily arise through simply misinterpreting such accounts as reached the Roman ear of German songs about the Mythical hero. Granted that irmansûl expressed word for word no more than ‘huge pillar,’ yet to the people that worship it it must have been a divine image, standing for a particular god. To discover who this was, we can only choose one of two ways: either he was one of the tree great divinities, Wôdan, Thonar, Tiu, or some being distinct from them.
But here we must, above all things, ponder the passage partly quoted on p. 111 from Widukind, himself a Saxon; it says, a heathen god was worshipped, whose name suggested Mars, his pillar-statue Hercules, and the place where he was set up the sun or Apollo. After that, he continues: ‘Ex hoc apparet, aestimationem illorum utcumque probabilem, qui Saxones originem duxisse putant de Graecis, quia Hirmin vel Hermes graece Mars dicitur, quo vocabulo ad laudem vel ad vituperationem usque hodie etiam ignorantes utimur’. From this it follows, that the god to whom the Saxons sacrificed after their victory over the Thuringians was called Hirmin, Irmin, and in the 10th century the name was still affixed in praise or blame to very eminent or very desperate characters. Apollo is brought in by the monk, because the altar was built ad orientalem portam, and Hercules, because his pillar called up that of the native god; no other idol can have been meant, than precisely the irminsûl (pp. 115—118), and the true form of this name must have been Irmines, Irmanes or Hirmines sûl. The Saxones had set up a pillar to their Irmin on the banks of the Unstrut, as they did in their own home.
The way, Hirmin, Hermes and Mars are put together seems a perfect muddle, though Widukind sees in it a confirmation of the story about the Saxons being sprung from Alexander's army (Widuk. 1, 2. Sachsensp. 3, 45). We ought to remember, first, that Wôdan was occasionally translated Mars instead of Mercurius (pp. 121. 133), and had all the appearance of the Roman Mars given him (p. 133); then further, how easily Irmin or Hirmin in this case would lead to Hermes, and Ares to Mars, for the Irminsûl itself is connected with Eres-burg (p. 116). What the Corvei annalist kept distinct (p. 111), the two images of Ares and of Hermes, are confounded by Widukind. But now, which has the better claim to be Irmin, Mars or Mercury? On p. 197 I have pronounced rather in favour of Mars, as Müllenhof too (Haupt 7, 384) identifies Irmin with Ziu; one might even be inclined to see in it the name of the war-god brought out on p. 202, ‘Eru, Heru,’ and to dissect Irman, Erman into Ir-man, Er-man, though, to judge by the forms Irmin, Eormen, Ermun, Iörmun, this is far from probable, the word being derivative indeed, yet simple, not compound; we never find, in place of Ertag, dies Martis, any such form as Ermintac, Irminestac. On behalf of Mercury there would speak the accidental, yet striking similarity of the name Irmansûl or Hirmensûl to E r m _V and _r m a = prop, stske, pole, pillar (p. 118), and that it was precisely Hermes's image or head that used to be set up on such _r m a t a , and further, that the Mid. Ages referred the irmen-pillars to Mercury (p. 116). In Hirmin the Saxons appear to have worshipped a Wôdan imaged as a warrior.
If this view be well grounded, we have Wôdan wedging himself into the ancient line of heroes; but the question is, whether Irmin is not to be regarded as a second birth or son of the god, whether even as an ancestral hero Irmino is not to be distinguished from this god Irmin, as Hermino in Tacitus is from Arminius? So from thiod, regin, were formed the names Thiodo, Regino. It would be harder to show any such relation between Ing and Ingo, Isc and Isco; but I think I can suggest another principle which will decide this point: when races name themselves after a famous ancestor, this may be a deified man, a demigod, but never a purely divine being. There are Ingaevones, Iscaevones, Herminones, Oescingas, Scilfingas, Ynglîngar (for Ingîngar), Völsûngar, Skiöldûngar, Niflûngar, as there were Heracleidae and Pelopidae, but no Wôdeningas or Thunoringas, though a Wôdening and a Kronides. The Anglo-Saxons, with Wôden always appearing at their head, would surely have borne the name of Wôdeningas, had it been customary to take name from the god himself. Nations do descend from the god, but through the medium of a demigod, and after him they name themselves. A national name taken from the highest god would have been impious arrogance, and alien to human feeling.
As Lower Saxony, especially Westphalia, was a chief seat of the Irmin-worship, we may put by the side of Widukind's account of Hirmin a few other traces of his name, which is not even yet entirely extinct in that part of Germany. Strodtmann has noted down the following phrases in Osnabrüvk: ‘he ment, use herre gott heet Herm (he thinks our Lord is called H., i.e. is never angry); use herre gott heet nich Herm, he heet leve herre, un weet wal tóte-gripen (knows how to fall on)’. Here there seems unconcealed a slight longing for the mild rule of the old heathen god, in contrast to the strictly judging and punishing christian God. In Saxon Hesse (on the Diemel), in the districts of Paderborn, Ravensberg and Münster, in the bishopric of Minden and the duchy of Westphalia, the people have kept alive the rhyme:
Hermen, sla dermen
sla pipen, sla trummen,
de Kaiser wil kummen
met hamer un stangen,
wil Hermen uphangen.
Hermen is challenged, as it were, to strike up his war-music, to sound the catgut, pipe and drum; but the foe draws nigh with maces and staves, and will hang up Hermen (see Suppl.). It not impossible that in these rude words, which have travelled down the long tradition of centuries, are preserved the fragments of a lay that was first heard when Charles destroyed the Irmensûl. They cannot so well be interpreted of the elder Arminius and the Romans. The Striking and the staves suggest the ceremony of carrying out the Summer.
"Teutonic Mythology," by Jacob Grimm, trans. by James Steven Stallybrass, vol. 1, page 351-355.