Syntactic complexity, discourse status and animacy as determinants of grammatical variation in Modern English

Abstract

The aim of this article is to provide an overview of the syntactic, pragmatic and semantic determinants of word-order variation in Modern English, exemplified by the specific case of the use of long passives as order-rearranging devices. Word order in English and in most other SVO languages is affected by a number of factors such as animacy, semantic role, discourse status and syntactic complexity (Sornicola 2006). In this article, which analyses the influence of such factors in the use of long passives, I will try to show that their effects are construction-specific; in particular, that factors which are crucial in determining word order in some constructions – factors such as the animacy of the constituents involved – are entirely overruled by others in the case of Modern English long passives. Corpus data presented here will also serve to address issues pertaining to the nature of the determinants of grammatical variation, such as their independent versus epiphenomenal character, their interactions, and the locus of their effects on word order.

On the conventionalisation and loss of pragmatic function of the passive in Late Modern English scientific discourse

“…Another important function of long passives as order-rearranging devices is to arrange information in accordance with the principle of end-weight, first formulated in Quirk et al (1972), which predicts that long, heavy constituents tend to occupy the final position of the clause (cf. Biber et al 1999, 940-941;Seoane 2009). Given the functional differences between long and short passives, they were analysed separately.…”

“…If agents are considerably heavier than patients, they will certainly trigger the use of the passive (cf. Biber et al 1999, 940-941;Seoane 2009). There is no consensus regarding what makes a constituent heavy, whether it is length (number of words or syllables, cf.…”

“…Lambrecht 1994). Given-before-new is a simplified representation of a more general category, normally referred to as discourse status (Arnold et al 2000) or pragmatic information status (Hawkins 2004), and concerns the accessibility of information to discourse participants (Prince 1981(Prince , 1992Gundel, Hedberg and Zacharski 1993;Birner and Ward 1998;Biber et al 1999, 940-941;Arnold et al 2000;Huddleston and Pullum 2002, 1444-1445Seoane 2006aSeoane , 2009Sornicola 2006, 470). As far as passives are concerned, long passives tend to be used in order to place in initial position a nonagentive referent which conveys information which is more given, or older, than that of the agent, as in example (15) The subject of the passive in example (15), it, refers back to some writings mentioned in the previous context, while the entity referred to in the by-phrase, some people of judgment, who…, is mentioned for the first time.…”

“…"New" classifies a referent that is mentioned for the first time, absent from the linguistic and extralinguistic context, and not inferable from knowledge of the world (cf. Seoane 2009).…”

“…Animacy has been seen to determine word order preferences, for example, in the realm of genitive variation (Rosenbach 2002(Rosenbach , 2007 and in dative alternation (Bresnan et al 2007). In the case of passives, however, animacy does not have a bearing on their selection over actives, since it has been shown that promoting human noun phrases to initial position at the expense of non-human, inanimate noun phrases is not a conditioning factor in the use of long passives (Seoane 2009;cf. also Ferreira 1994).…”