Saga ábendingarfornafnsins sjá
ÚtdrátturThe History of the Demonstrative Pronoun sjáKeywords: historical morphology, language change, declension, analogy
The Old Icelandic demonstrative pronoun sjá ‘this’ had 3 genders, 2 numbers and 4 cases and there were a few alternative forms. The paradigm was the result of a complicated prehistory where e.g. two types of deictic particles, the strong declension of adjectives and the declension of the demonstrative pronoun sá played a part. The paradigm of the oldest Icelandic manuscripts is presented in table 1. It is, perhaps, not surprising that it has proved unstable. Several earlier observations on the chronological development of sjá are summarized in fig. 1.
A collection of examples from a number of texts is presented here to illustrate the development of the declension of this pronoun, but the comparative rarity of examples presents a considerable difficulty. An attempt is made to provide a more accurate chronology and especially to explain the changes of the paradigm. The chronological development is shown in fig. 2. It is suggested here, on the basis of the available evidence, that the changes of the paradigm took place in several stages. In every case there was a close interplay with various features of the declension of other pronouns and adjectives. The main hypothesis is that every step in the development tended towards simplification. There was a direct connection between some of the stages (see fig. 3), and one can possibly speak of a chain reaction.
The main changes, apart from the disappearance of the archaic alternative forms þvísa and þeima, were the following, according to the present analysis:
1a sjá -> þessi (nom.fem.sg.)
1b sjá -> þessi (nom.masc.sg.) 2a þessi -> þessari (dat.fem.sg.) 2b þessar -> þessarar (gen.fem.sg.), þessa Æ þessara (gen.pl.) 3 þenna -> þennan (acc.masc.sg.) 4 þetta -> þettað (nom.acc.neut.sg.). The modern paradigm is shown in table 10 .
Old Icelandic sjá was an infrequent pronoun compared with sá ‘that’ (approximately 1:6.85), which also has a declension that would seem to be unstable. The difference in frequency may partially explain why the former changed but not the latter.
| ||||
