No CrossRef data available.
Published online by Cambridge University Press: 22 April 2026
Leveling of the radical vowel in the plural paradigm of the consonantal r-stems denoting family relations. In the plural paradigm of the substantive consonantal declension, i-umlaut of the radical vowel originally occurred only in the nom.-acc. form, since only here did the end syllable contain the vowel i. This vowel status is regularly preserved in the monosyllabic consonantal stems; cf. fϕtr (*fōt-iR) fóta fótum fϕtr (*fōt-iR), from fótr (masc.) ‘foot’; merkr (*mark-iR) marka mǫrkum merkr (*mark-iR), from mǫrk (fem.) ‘mark’ (a weight or money value). The consonantal r-stems denoting family relations belonged to the dissyllabic pattern in the singular (faÐer ‘father’, bróÐer ‘brother’, móÐer ‘mother’, dótter ‘daughter’, syster ‘sister‘), but became monosyllabic in the plural (feÐr, brϕÐr, etc.). Nevertheless, contrary to the pattern of the monosyllabic type (fótr, mǫrk), the original radical vowels a and ó in the genitive and dative forms were leveled out in favor of the umlauted vowels e and ϕ of the nom.-acc. forms; cf.
, etc. So far as I know, no explanation has ever been given for this discrepancy between the two types of consonantal stems.
1 In the later period, secondary leveling of the radical vowel occurs in the genitive and dative plural of the monosyllabic consonantal stem dyr ‘door’, resulting in doublet forms for these two cases: dyr(r) dura (dyra) durum (dyrum) dyr(r), with shift of gender from feminine to neuter. The secondary forms dyra dyrum most probably followed the pattern of the neuter ja-stems, where the radical vowel y was phonetically correct throughout the plural paradigm (cf. kyn kyn-ja kyn-jum kyn, from kyn ‘race, family‘).
Conversely, there was secondary leveling of the umlauted vowel of the nom.-acc. form in the plural paradigm of the feminine monosyllabic stem ær ‘ewe’ in the compound slag-ár ‘ewes for slaughter’. Here the reverse vowel leveling was obviously due to the pattern of the contracted monosyllabic ō-stems; cf. the plural paradigms of ó ‘river’ and
, hence
.
In both these examples of secondary vowel leveling, the primary vowel status of the plural paradigm is still preserved alongside the secondary. There are no examples of monosyllabic consonantal stems in which the secondary status has completely displaced the primary in the plural paradigm.
2 The original vowel status survived in Old Norwegian and Old Swedish in such forms as gen. plur. bróÐra (ONorw.): bróþra-nna (OSwed.), dat. plur. bróÐrum (ONorw.): fadhrum (OSwed.). Cf. Noreen, Aisl. Gram.4 §420 Anm. 3; Noreen, Altschwed. Gram. §438.6.
3 Regarding these suffixes see Elof Hellquist, Arkiv för nordisk filologi 7.5–6 (1891).
4 It is quite possible that in the consonantal declension of frændr, the plural form of frændi ‘kinsman’, the vowel leveling (frænd-r -a -um -r) was due to the pattern of the r-stems because frændi (like faÐir, móÐir, etc.) denoted a family relation.
5 The v in fro(u)va ‘Frau, lady’ does not represent a v-suffix but the w of the stem of the foreign word MLG vrouwe, which was introduced into ON; cf. *frawjō > OHG frouwa, ON Freyja.
6 Karl Müllenhoff, DA 5.42 (1908), translates the word vǫlva by ‘stabträgerin’.
7 Cf. Alexander Jóhannesson, Die Suffixe im Isländischen §49.2.
8 For a discussion of the substantive helgr as derived from the adjective heilagr see Bengt Hesselman, Västnordiska studier 2.17 n. 3. Hesselman rightly assumes that helgr does not represent a PGmc. abstract formation (cf. helge < *hailagīn ‘holiness‘) but a specifically ON iō-formation with concrete meaning. He does not, however, explain the irregular loss of the j-suffix.
9 Cf. the displacement of the genitive plural form *læta (from læte, neuter ja-stem) ‘conduct, appearance’ by the form láta (without umlaut) from lát (neuter a-stem), which is synonymous with læte.