Parameters of variation between verb–subject and subject–verb order in late Middle English

Abstract

This article sets out to clarify the contribution of syntactic properties and subject weight for variation between verb-subject and subject-verb order in a database of fourteenthand fifteenth-century prose. It sets out the syntactic structures which are assumed, and investigates the impact on ordering of a set of factors, using established quantitative methodologies. A series of conclusions includes the continuing distinct status of initial then, the systematic importance of clause-final position, the different impacts of subject length in different contexts, and the presence of a definiteness effect for the late placement of a subject after a nonfinite unaccusative. 1 I am delighted to acknowledge a research readership granted by the British Academy which gave me the time to collect the data investigated here, and the comments of audiences atIn today's English the choice of order in such clause types seems to involve both syntactic and other factors. The reversed order is clearly required with BE when its predicate is fronted as in (1a), and it is promoted by the presence of an unaccusative verb as in (1b-d). It has also been suggested that it involves the structuring of information. Birner (1994: 233) claimed that the ordering of examples like (1) above in a present-day corpus 'depends on the relative discourse-familiarity of the information represented by the preposed and postposed constituents', so that information that is more familiar precedes information that is less familiar, in line with the general view (variously defined, and expressed in different terms by different theorists) that 'given' information precedes 'new' information. But it may not only be the relative information status of constituents that is involved here. Wasow (2002) discusses factors which are relevant to the choice of a particular order of postverbal constituents in several constructions in Present-day English where variation in order is found. He notes that the comparative grammatical weight of a constituent (potentially defined in terms of its length or its grammatical complexity) plays a major role, with a strong tendency for less weighty constituents to precede more weighty constituents, and he goes on to suggest that utterance planning may be a crucial factor underlying this tendency. He also presents evidence that in at least some constructions weight and information structuring are distinct parameters, though in general their contribution is difficult to distinguish (2002: 69-81). Both of course are likely to reflect the planning and production, and the parsing and perception of utterances. Wasow notes more generally that the question of what factors motivate the selection of a particular order is a complex one with 'no simple answer ' (2002: 109). See Culicover & Levine (2001) for some further recent discussion and analysis of inversion in Present-day English, Green (1980) and Stockwell (1984) for earlier discussion.It is clear that information structuring and grammatical weight also had considerable importance in earlier E...

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

“…Cardinaletti (2004) and Westergaard & Vangsnes (2005). This corresponds to the analysis of the two subject positions in OE found in van Kemenade (2000) or Warner (2007), referred to as SpecF(unctional)P and SpecTP in the former, and as SpecAgrSP and SpecTP in the latter. Here I adopt the view that both subject positions are in the IP domain and argue that the high position, which is the Spec of what I call the In(ner)Top(ic)P, is reserved for informationally given subjects, while the low position (SpecTP) is used for informationally new or focused subjects, marked as [+/− foc ] in (12) 3 .…”

The consequences of the loss of verb-second in English: information structure and syntax in interaction

“…More recently, data like (2) and (3) have been interpreted as demonstrating the existence of a separate projection for clitic pronouns (van Kemenade 2000;Haeberli 2002), with distinct structural positions for pronoun subjects and full nominal subjects, and two positions for the fronted finite verb. This means that the finite verbs of (2) and (3) are in different positions after all, as argued by Pintzuk (1991): in C in (2) (see the tree structure in (4)), and in the head of a lower functional projection which Haeberli (2002) and Warner (2007) identify as AgrSP in (3) (see the tree structure in (5)). Note that AgrSP and TP now host information that used to be situated in IP in earlier versions of the theory (following the proposal of the Split IP in Pollock 1989).…”

“…This construction allows the subjects that are informationally new to appear in the end-focus position (presentational focus). Warner (2007) notes that a sentence like (9) is derivationally ambiguous:…”

“…Any 'late subjects' like (11) encountered in lME and eModE, the crucial period for the loss of verb-second, should be kept separate and not counted as a possible instance of verb-second; they do not involve verb-movement. Cases with a single lexical, and unaccusative, verb like (12) are even more difficult to deal with because they are derivationally ambiguous (Warner 2007; see also Haeberli's (2002: 253-5) discussion of what he labels Type I and Type II inversion in Middle English). They might be instances of late subjects, like (11), but they could also be instances of verb-second if they are the output of a grammar which retains verb-second and which does not yet recognize auxiliaries as a separate category.…”

Subject Positions and Information-Structural Diversification in the History of English

“…i.a. van Kemenade 1997, Tanaka 2002, Warner 2007, the rise of Transitive Expletive Construction during the ME period (cf. Jonas 1996, Tanaka 2000, Links 2010 and the phenomenon of Post-finite Misagreement (cf.…”