Subject positions and information structure: the effect of frequency on acquisition and change

Abstract

Two subject positions are identified in main and embedded clauses in Norwegian, appearing above or below negation and certain adverbs, and a similar situation is attested in the history of English. This paper investigates the distribution of subject types across several corpora of spoken Norwegian, produced by both adults and children. The frequencies attested are discussed in terms of a cue-based approach to acquisition, according to which children are sensitive to fine distinctions in syntax and information structure (micro-cues). Nevertheless, micro-cues with an extremely low frequency are argued to be vulnerable, and possible language change is discussed with reference to some actual changes which have either taken place in the history of English or which are argued to be in progress in present-day Norwegian. These changes are considered to reflect ''syntactic freezing'' of information structure patterns.

Linguistic variation and micro-cues in first language acquisition

“…In the Tromsø acquisition corpus (Anderssen 2006), the high subject position is attested 81% (1351/1667) in relevant utterances in the adult data. The three children in the corpus are sensitive to this distinction from early on (Westergaard 2008a, Anderssen & Westergaard 2010, Westergaard 2011, typically producing DP subjects in the low position (following negation) and pronominal subjects high (preceding negation), as illustrated in (5) To conclude this section, the data from children's spontaneous production of word order variation, both at the clausal and the phrasal level, show that they produce both options from early on. Furthermore, they generally produce the two word orders in appropriate contexts.…”

The Acquisition of Neg-V and V-Neg Order in Embedded Clauses in Swedish: A Microparametric Approach

“…In line with much previous research, Westergaard also finds that word-order errors are few and that errors are almost always a result of children using less movement than in the adult language. This observation has led many scholars to propose that there are economy principles operating in the process of acquiring a language, principles that make children conservative language learners (Platzack 1996;Zuckerman 2001;Gavarrò 2003;Westergaard 2009aWestergaard , 2009bWestergaard , 2011Waldmann 2012). For example, Westergaard & Bentzen (2007) suggest that syntactic movement does not take place in child language unless it is clearly expressed in the input (see also Westergaard 2008b).…”

“…Further, children are able to make these fine distinctions based on very little input. For example, Westergaard's studies of Norwegian children show that the children distinguish between main and embedded clauses in the acquisition of subject positions and between informationally given and informationally new subjects in the acquisition of subject positions and mixed V2/non-V2 word order in wh-questions (e.g., Westergaard 2003Westergaard , 2008bWestergaard , 2011. In line with much previous research, Westergaard also finds that word-order errors are few and that errors are almost always a result of children using less movement than in the adult language.…”

“…Microparametric approaches to language acquisition (Westergaard 2003(Westergaard , 2008a(Westergaard , 2009a(Westergaard , 2009b(Westergaard , 2011 assert that children are sensitive to very fine distinctions in syntax and information structure in the input and that they are able to produce different word orders that are dependent on minor syntactic distinctions and on distinctions in information structure at an early age. Further, children are able to make these fine distinctions based on very little input.…”

“…Children are endowed with the ability to detect such microparameters in the input. Thus, children are sensitive to very fine distinctions in syntax and information structure in the input, and they are able to produce different word orders that are dependent on minor syntactic distinctions and on distinctions in information structure at an early age (Westergaard 2003(Westergaard , 2008a(Westergaard , 2008b(Westergaard , 2009a(Westergaard , 2009b(Westergaard , 2011Westergaard & Bentzen 2007;Anderssen et al 2010). Within a microparametric approach, it is not surprising that children do not overgeneralize V2.…”