In Old English, genitive nouns in apposition are both marked for case. King Arthur's men, for example, would be Arthornures cyninges menn, Arthornures menn thornaes cyninges, among a few other possibilities. Non-agreeing genitive appositives such as Arthornur kinges men and Arthornures men (thorne) king began to appear in the Early Middle English period. The analysis in two versions of La?amon's Brut suggests two possible pathways through which they were established. One is the development of the name and the title as a compound-like unit, which marks genitive only at the end of the second element as in Arthornur kinges men, and the other is ambiguous appositive or predicative expressions such as Arthornures men (thorne) king, interpretable as either King Arthur's men' or men of Arthur, (who is) (the) king', which seem to have led post-head constituents of split genitives to appear in the nominative or unmarked case, resulting in genitive marking only on the pre-head components. These new usages are thus originally and essentially independent of the developments of the case-marking system itself.

This article investigates variant forms of the demonstrative 'that' and the adjective in the Caligula Manuscript of La.amon's Brut and concludes that they to a large extent maintain the traditional category distinction between the accusative and the dative by the use of the suffixes -ne and -Vn (where V stands for any vowel) respectively. There are, however, factors that potentially compromise the status of these terminations as case markers: phonetic reduction (which has often been invoked), capricious addition or deletion of final e, occasional doubling or simplification of nasals, and simply unexpected choices of forms in the paradigm. These seem to be consistent with the sort of mistakes that scribes might occasionally make when faced with an original that has a different orthographical and morphological system from their own, and they are not as disruptive of the case-marking system as at first sight they might appear. For one thing, they occur rather rarely and are generally outnumbered by historically expected forms; for another, the resultant unexpected case forms, usually with the stem vowel preferred by the historical forms in the case of the demonstrative, are predominantly accompanied by a historically motivated case form of their head noun. The status of accusative -ne and dative -Vn, in fact, appears to be stable enough to lead to the development of a system or subsystem of indicating case regardless of gender considerations in their respective case contexts. These suffixes can therefore be treated validly as independent case markers, although in concrete cases the possibility always exists that there is an optional final e, an unhistorical double or single n, or an unexpected choice of inflectional forms.

An examination of early examples of the his-genitive in the Otho manuscript of La3amon's Brut suggests its orthographical origin as a variant of the historically expected s-genitive, but at this stage they are already different morphologically, though they remain identical syntactically. Their orthographical and morphological independence as well as their lessened degree of attachment with their head noun appears to have advanced their clitic-like nature, and when the process is complete, the group genitive becomes available. Once the clitic status is attained, his, with its more syntactic freedom, may be said to have become more prone to reanalysis as a pronoun, and the resultant construction is further generalized to the feminine and the plural to develop her- and their-genitives. This reanalysis could have happened even before the cliticization of his, however, given its obvious formal similarity to the pronoun, and, in fact, they appear to have been associated with each other earlier on to lead to the morphological independence of his from the inflection. It may be just that the incorporation of the reanalysis into the system or sub-system of grammar has to wait until the cliticization of the genitive or possessive marker. Thus it appears that the necessary components are all present in Otho to be further exploited for the later development of his- and other related genitive constructions.