Null Subjects in Early Icelandic

Abstract

This paper investigates the possibility of subject omission in the history of Icelandic, including the syntactic and pragmatic conditions under which it could arise. Based on regression analysis of substantial data drawn from the IcePaHC corpus, we provide robust quantitative support for Hjartardóttir's (1987) claim that null subjects persist until a very late stage in Icelandic. We also argue, contra Sigurðsson (1993), that only one licensing mechanism is needed for null subjects in early Icelandic. Moreover, on the basis of the position of the null subjects and their person features, we argue that the modern stage, where (predominantly 3rd person) pro-drop yields to a system permitting topic drop of all persons, arises in Icelandic in the early twentieth century.*

Null arguments in old Norwegian

“…e.g. Sigurðsson 1993 andKinn et al 2016 English, as well as the comparative studies of Rosenkvist 2009 andWalkden 2014). In this paper I propose a new analysis of null arguments in Old Norwegian, an understudied variety whose null argument properties are not immediately captured by previous accounts.…”

“…Håkansson (2008Håkansson ( , 2013, Walkden (2014) and Sigurðsson's more recent (2011) crosslinguistic study of null arguments, all assume a single (though not the same) way of deriving null subjects. This has advantages in terms of theoretical economy; Kinn et al (2016) argue that the unified approach can be supported also on empirical grounds. One of Sigurðsson's arguments for the distinction between topic-drop and pro-drop is an alleged difference in antecedent relations: Sigurðsson (1993:251-252) proposes that pro always requires an overt DP antecedent in the preceding linguistic context (an NP antecedent in his terminology), whereas dropped topics in verb-initial main clauses may occur without an overt DP antecedent.…”

“…One of Sigurðsson's arguments for the distinction between topic-drop and pro-drop is an alleged difference in antecedent relations: Sigurðsson (1993:251-252) proposes that pro always requires an overt DP antecedent in the preceding linguistic context (an NP antecedent in his terminology), whereas dropped topics in verb-initial main clauses may occur without an overt DP antecedent. 5 The corpus study of Kinn et al (2016) argues that antecedentless null arguments in Old Icelandic may occur in not only verb-initial main clauses, but also in subordinate clauses and non-verbinitial main clauses, i.e. in the contexts where Sigurðsson (1993) 'Then the king was told about the rock on which the holy king Óláfr fell.…”

Topicality in Icelandic: Null Arguments and Narrative Inversion

“…However, for expository ease, I will occasionally refer to Type A and Type B drop as pro drop and topic drop, respectively. 5 Kinn et al (2016) argue that this is an incorrect characterization and that there are some cases of referential pro without a coreferential antecedent. However, the nulls in question are either arbitrary/generic or expletive or found in idiomatic expressions, and nulls of these sorts are also found in Modern Icelandic texts, in contrast to clearly referential nulls like the ones in (5a-d).…”

“…While type (21a) has disappeared, types (21b-c) seem to be getting more frequent in the written language (cf. Kinn et al 2016), presumably as a side effect of much increased use of informal written style. These drop types cannot always be easily distinguished from one another when the null argument is a subject.…”