The runic inscription on the lance head of Mos (dated between 210/20‑250/60 [cf. L. Imer, The oldest Runic Monuments in the North, in: NOWELE 62/63 (2011), p. 197]) has long been a puzzle.
It can either be read as sioag or as gaois (Photo).
Other readings - gaŋis < PGmc. *ǥanǥia‑ 'he, who belongs to the armed encounter' (E. Seebold, Die sprachliche Deutung und Einordnung der archaischen Runeninschriften, in: K. Düwel [ed.], Runische Schriftkultur in kontinental-skandinavischer und -angelsächsischer Wechselbeziehung, Berlin/New York 1994, p. 73) and gais < PGmc. *ǥasa‑ 'spear' (M. Giertz, Mos - Sveriges äldsta runinskrift. Förslag till ny tolkning, in: Gotländskt arkiv 63 , p. 105) - can easily be disregarded here.
The reading sioag that is based on the direction of the a-rune (and also of the s-rune, which is not that decisive), is out of two reasons improbable:
1. The sequence sioag cannot be interpreted in any way (cf. W. Krause, Die Sprache der urnordischen Inchriften, Heidelberg 1971, p. 155: "keinen Sinn ergebend"). However, this argument is not that decisive, because there are more inscriptions that show an unclear sequence of runes (cf. eg. the inscription on the chape of Vimose [ca. 210/20-310/20]: ttnþ).
2. More important is the direction of the inscription. The runic inscriptions on lance heads tend to have a reading direction towards the point.
The last point shows that the most probable reading is gaois. In this case the a-rune (as the s-rune) must be seen as a Wenderune.
As a single word this is most likely a word in the nominative singular. The final -s (and not -z) immediately suggests that the language at hand is East Germanic.
Etymologically the word gaois was by W. Krause / H. Jankuhn, Die Runeninschriften im älteren Futhark, Göttingen 1966, vol. I, p. 81) interpreted as: "Bei der Annahme, das keine weiteren Runen ausgefallen sind, könnte man in dem Worte gaois (= gaujis) ein Nomen agentis vom Typus got. ubil-tojis 'Übeltäter' erkennen", which he connected to the verb PGmc. *ǥae/a‑ 'yell'. This verb is continued in OE gēgan, WFris. geije, OIcl. geya (another derivation of this verb in the runic inscriptions is most probably also found on the strike-a-light of Illerup ådal [210/20-250/60]: gauþz 'yeller').
As was noted by R. Nedoma, Schrift und Sprache in den ostgermanischen Runeninschriften, in: NOWELE 58/59 (2010), p. 23 this interpretation suffers from two difficulties:
1. There seems to be a j lacking.
2. One would expect the spelling -au- and not -ao- of the diphthong *a (cf. e.g. lance head of Øvre Stabu [150/60-250/60]: raunijaz < PGmc. *rania- 'tester').
However, both objections are in fact void.
The spelling -ao- instead of -au- has parallels in the SouthGmc. runic material: fibula of Lauchheim I (ca. 551-600): aono < PGmc. *anan-; ivory ring of Pforzen II (ca. 600): aodliþ < PGmc. *ađa-. That there were other spelling possibilities for this diphthong is also shown by the inscription on the belt fitting of Nydam (ca. 210/20-310/20): rawsijo < PGmc. *rasia/ōn‑.
The assumption that one would expect a spelling with -j- is based on the situation in Gothic; here the type PGmc. *χara- 'army' results in nominative singular harjis. However, it is undoubted that in this form the -j- is the result of an analogical process (cf. R. Schuhmann, Zum analogischen Ausgleich bei den got. ja-Stämmen, in: Th. Krisch e.a. [eds.], Indogermanistik und Linguistik im Dialog, Wiesbaden 2011, pp. 508-516). According to the sound laws PGmc. *χaraz would have developed through *χariz > Goth. **haris, which was replaced by harjis. This means that the sequence PGmc. *ǥaaz developed through *ǥaiz to *ǥais, which would have resulted by analogy in Goth. **gaujis.
It is therefore clear that the form gaois represents exactly the stage *ǥais and dates from the time before the analogical instrusion of -j- in the nominative singular.
The inscription on the lance head of Mos gives therefore a small glimpse in the EastGmc. language that predates the transmission of the Gothic Bible.