Clefting and extraposition in English 1

Clefts and extraposition have structural and functional similarities, yet they have been mostly treated separately. An investigation of the Wellington Corpus of Spoken New Zealand English reveals cases where the two constructions appear confounded and difficult to disentangle. The present paper argues that they can indeed be differentiated, and provides a test which can be used to this end. This work offers contributions of a practical nature, in that it provides researchers with an objective criterion for distinguishing between clefts and extraposition, and also in a more theoretical sense, in its focus on two important strategies used in information packaging and discourse organization. 1 The problem of disambiguating between IT-clefts and extraposed clauses The work detailed in this paper arises from a practical problem encountered when attempting to analyze the Wellington Corpus of Spoken New Zealand English2 (henceforth WSC) for cleft constructions, in particular for IT-clefts. The occurrence of IT-clefts was investigated in a subpart of the corpus, namely in 193,000 words of spontaneous, face-to-face conversation data, which was found to contain 150 IT-clefts (this equates to roughly just over 75 constructions for every 100,000 words). Consider the two examples3 given in (A) and (B): (A) DS well margaret started wanting to wear makeup and so on when she was onl MK yeah totally different type of child DS yeah MK or woman now of course DS yeah young women now for sure MK ICAME Journal No. 32 8 DS oh it’s quite good to to hear that faith’s actually experimenting with a few different things MK yes well raelene was she giggled DS yes i’m sure she did (WSC#DPC002) (B) CH well they never got close to me but they were they’re real cowboys and they’re idiots you know and they’re going on about this and that like and they said they wanted an under the table job and i said okay so that’s w we agreed on the price BT should’ve DN mm AL you haven’t paid them anything CH i haven’t paid them a cent and they’re meant to come back well they came back when i was on holiday i left this really clear message saying do not come back till i get back from holiday it’s about six weeks ago they did the job DN mm AL it still fucking leaks (WSC#DPC066) Structurally, the constructions given in boldface are similar to both IT-clefts (exemplified in the following section), and at the same time, (BE-)extraposition.5 In both cases, the clauses have dummy/expletive it in subject position, followed by the copula be, some constituent (the ‘clefted constituent’ in a cleft analysis or the ‘remainder predicate’ of the main clause, under an extraposed analysis), and finally, a subordinate clause (the ‘cleft clause’ in a cleft or the demoted subject/ object clause in extraposition). The problem lies in deciding which analysis to adopt for the two constructions. Should they be analysed as clefts or as extraposition? What is more, this question raises several related issues. At a more practical level, it is worth considering which cases have the potential to cause these disambiguation difficulties; while on a more theoretical plane, it is not clear why these constructions should present such similar structures in the first place, making it difficult for us to distinguish between them. Is it just coincidence or is there more to it? More importantly, is it always possible to distinguish between clefts and extraposition? If yes, what criteria can be used to this end? If not, is one construction a special instantiation of the other or is the boundary between them fuzzy rather than discrete? (See Hopper and Thompson’s seminal 1980 paper and a recent collection of works edited by Aarts et al. (2004) on fuzzy grammar for more details.) If on the other hand, there are cases in which clefts and extraposition Clefting and extraposition in English 9 cannot be disentangled, what do these cases look like and how do they differ from prototypical clefts and prototypical extraposition? Being able to distinguish between IT-clefts and extraposition is beneficial because the two constructions play different roles in the discourse; that is, they are used for different purposes. IT-clefts are a focusing device, highlighting or contrasting particular bits of information. Extraposition, on the other hand, is connected with the avoidance of having complex subjects at the beginning of the sentence and “serve[s] the two principles of end-focus and end-weight” (Quirk et al. 1985: 863). This type of construction is particularly relevant to spoken language, where subjects are almost never complex (not even complex nominal phrases, let alone clausal units). We will see more about the various discourse functions of the two constructions in the following section. However, their discourse functions apart, clefts and extraposition differ in their syntax. Unlike clefts, extraposition may involve predicates other than the copula be, i.e., It helps to think that my mother will be there waiting for us and It surprises me how seriously she takes this stuff. Furthermore, extraposition allows a greater variety of extraposed elements (NPs, PPs, and so on) whereas clefts only allow a cleft clause in the final position (see Table 1 and example 23). The questions formulated above form the basis for the work presented here and will be discussed with reference to examples from spontaneous, unplanned face-to-face conversations from the WSC. In spite of the use of spoken data, the results obtained are assumed to apply in equal manner to written language, unless otherwise indicated. The paper has three major aims: • first, to support the view that IT-clefts and extraposition6 are distinct and can be distinguished from one another, which is assumed by some (but only stated explicitly by Pérez-Guerra 1998), • secondly, to draw attention to difficulties in distinguishing the constructions in spoken English, and • thirdly, to provide an objective test for distinguishing between them, a test which appears to be informally used by some but not explicitly stated in the literature. The paper is organized as follows. First, extended definitions and examples of IT-clefts and extraposition are given. In the following section, further problematic examples are examined, where it is difficult to distinguish between the two constructions. These are used to show that the only recent criteria (proposed by Pérez-Guerra 1998) for differentiating them are not adequate to resolve ambiguities. Pérez-Guerra’s paper gives a diachronic account of the increase in use of ICAME Journal No. 32 10 clefted and extraposed constructions throughout the history of the English language as grammaticalized versions of right-dislocation. As part of this account, he proposes a set of criteria for distinguishing clefting and extraposition, which will be outlined in section 3.1. However, it will be argued that these criteria are not sufficient for eliminating ambiguities between the two constructions. A ‘transformational-based test’ (word order re-arrangement test) is proposed as a test for distinguishing between IT-clefts and extraposition, which will be applied to further examples from the WSC, including the earlier constructions given in (A) and (B). It will be shown that the test can be used to successfully distinguish between clefting and extraposition, though there are some cases where stylistic difficulties arise. (The test is, however, based on grammatical acceptability, not stylistic judgments.) Such stylistic problems prove relevant only to cases when both the pre-copula material and the post-copular constituents are short/light. This reinstates earlier questions regarding the discourse function of extraposition. Furthermore, I will also discuss problems related to language medium, in particular difficulties in analysing the structure of speech which arise from the lack of syntactic integratedness of spoken data. Finally, the paper concludes with a summary. One final remark concerning the theoretical framework used is in order before proceeding with the analysis. The test proposed here has its basis in the (revised) standard models of transformational grammar, but is not conceived of in this work as anything more than a useful tool for the analyst. The current paper is quite neutral with respect to the debate about the merits or demerits of any particular models of generative grammar, transformational or non-transformational. 2 Background As far as the literature is concerned, cleft clauses and extraposed constructions have been treated separately for the most part, with the exception of a paper by Pérez-Guerra (1998) discussed later in the paper. Furthermore, while there is a wealth of recent research discussing various cleft types (Collins 1987; Hedberg 1988; Delin 1989; Collins 1991; Delin and Oberlander 1995; Oberlander and Delin 1996; Weinert and Miller 1996; Hedberg 2000; Johansson 2001; Lambrecht 2001; Herriman 2003; Collins 2004; Herriman 2004; Delin and Oberlander forthcoming; and others), there is much less current work on extraposition (some of the most recent papers include Pérez-Guerra 1998; Seppänen 1999; and Herriman 2000). Each construction type is considered in turn, starting with clefts. Clefting and extraposition in English 11 2.1 IT-clefts, structure and discourse function IT-clefts are focusing constructions, in which typically a simple sentence (though complex sentences can also be involved) is ‘cleaved’ such that the pronominal it appears in initial/subject position, followed by the copula be, the clefted constituent which expresses the highlighted or focused element, and finally, the cleft clause, modifying the clefted constituent (see descriptions by Hedberg 1990; Weinert and Miller 1996; and Huddleston and Pullum 2002). Example (1a) gives such a simple sentence, and (1b) illustrates the cleft which can be constructed for the purpose of focusing or highlighting the subject noun phrase a bright yellow jumper: (1a) Henry bought a bright yellow jumper yesterday. (1b) It was a bright yellow jumper that Henry bought yesterday. The status of the pronoun it in cleft constructions has caused some debate in the literature. While some argue that it has referential status (Bolinger 1970; Gundel 1977; Declerck 1988; Hedberg 2000), others contend that it is simply an empty place-holder, devoid of any referring role (Postal and Pullum 1988; Haegeman 1991). As far as the highlighted element is concerned, there is still disagreement about what types of constituents are allowed to appear in this position. According to Biber et al. (1999: 959), the slot can be filled by noun phrases, prepositional phrases, adverbial phrases or adverbial clauses. However, alongside these, Huddleston and Pullum (2002: 1418–1419) add the following possibilities: finite and/or nonfinite clauses (It’s that he’s so self-satisfied that I find off-putting or It’s certainly not to make life easier for us that they are changing the rules), and adjectival phrases (It’s not lonely he made me feel – it’s angry and It wasn’t green I told you to paint it). The exact status of the cleft clause has similarly provoked debate, with opinions ranging from those arguing strongly for its analysis as a relative clause (Hedberg 1990; Huddleston and Pullum 2002) to those still holding notable differences between relative clauses and the nature of cleft clauses7 (Quirk and Greenbaum 1985; Miller 1996; Miller and Weinert 1996; Miller 1999; Biber et al. 1999). However, it suffices to say that most studies converge on the idea that cleft clauses are at least reminiscent of, even if not identical with, relative clauses. The discourse function of clefts is to focus new or contrastive information expressed by the clefted constituent (cf. Biber et al. 1999 and Lambrecht 2001 ICAME Journal No. 32 12 among others) and in some cases they can also have a ‘remind me’ role (see Weinert and Miller 1996). 2.2 Extraposed clauses, structure and discourse function We now turn our attention to extraposition. Extraposition is used to ‘lighten’ the load of a sentential subject (and less frequently, an object) by demoting a subordinate clause from subject (or object) position to the end of a sentence. For instance, the subject clause in (2a), That he left in such a hurry, is extraposed to the end of the sentence as shown in (2b), with the help of the inserted pronoun it. Further examples of extraposition are given in (3)–(5): (2a) That he left in such a hurry is no surprise. (2b) It is no surprise that he left in such a hurry. (3) It never ceases to surprise and horrify me how these criminals get away. (4) Well, I believe it to be a crime to let any murderer walk free like that. (5) The professor found it incredible that any student could write such an essay on their own. There is consensus in the literature regarding the type of clauses which can be extraposed: they can be finite (the most frequent type), or non-finite (with gerunds, as in example (6), much less frequent than infinitives, see (7)), and they can be introduced by that, a wh-word or nonfinite to (Collins 1994; Biber et al. 1999; Huddleston and Pullum 2002). However, one aspect which still causes debate concerns the status of it. As with IT-clefts, it has been analyzed in different ways. Some argue that extraposed constructions have two subjects, it being the ‘formal’ subject and the extraposed clause being the ‘notional’, ‘real’, or ‘postponed’ one (Jespersen 1972 and Quirk et al. 1985); others maintain that it is the only subject, the extraposed clause being stripped of its subject-like properties once it is demoted to sentence-final position (Huddleston 1984; Seppänen, Engström and Seppänen 1990; and Seppänen 1999). This issue is beyond the scope of the present paper; what is noteworthy is the similarity between clefting and extraposition regarding the ‘slippery’ nature of it in both constructions. With regard to the discourse function of extraposition, a recent study by Rowley-Jolivet and Carter-Thomas (2005) found two main and partially related uses of extraposition. First, it helps “increase dynamism” by placing new inforClefting and extraposition in English 13 mation sentence-finally. Secondly, it is used to express the speaker’s/writer’s evaluative opinion in a “rhetorically effective” way (Rowley-Jolivet and CarterThomas 2005: 51; also cf. Hoey 2000; and Hewings and Hewings 2002) by introducing the evaluative comments sentence-initially: It is obvious that [...] or It is unusual that [...] so that they are less conducive to being challenged. These findings are in agreement with earlier work by Collins (1994) and Herriman (2000). Despite the fact that extraposition has received most attention in past work, it could be argued that the opposite phenomenon (i.e., non-extraposition, as given in example (2a)) merits equal, if not more, attention. There are two, related reasons for this. First, as shown by Biber et al. for that-clauses (1999: 676), Collins (1994: 14), Herriman (2000: 584), Mair for to-infinitives (1990: 30), and Quirk et al. (1985: 1392), from a statistical standpoint, extraposition is more frequent than non-extraposition. Secondly, extraposition is functionally motivated, whether one adopts a psychologically oriented explanation whereby sending heavy constituents to the end of the sentence allows for easier processing (see Langacker 1974: 653 and Huddleston 1984: 354 for this view), or whether it is simply assumed that end-focus and end-weight principles are at work here (see Collins 1994: 15–16 for a more in depth discussion). In other words, both functionally, as well as statistically, extraposition appears to be the norm rather than the exception. 3 The problem of distinguishing between IT-clefts and extraposition The previous section introduced the two constructions of IT-clefts and extraposition. As mentioned earlier, the present paper argues that the two constructions can be reliably distinguished from each other. It turns out that only one specific type of extraposed clause creates problems in this regard. The term extraposition will be used for the remainder of the paper to refer to this special construction. The label ‘extraposition’ denotes, in this case, sentences whose extraposed subject clauses involve the lexical verb be as the main verb of the superordinate clause, as exemplified in (6) and (7): (6) It is pointless complaining to the head manager. (7) It was very unusual to see someone so intelligent wasting their time in this way. Although extraposition is often contrasted with right dislocation (Huddleston 1984 and McCawley 1988 in Collins 1994: 12–13), with varying degrees of sucICAME Journal No. 32 14 cess (see Collins 1994: 12–13 for a discussion of why the two constructions cannot always be reliably distinguished from each other), the possible overlap between extraposition and clefting is yet to be addressed, with one exception discussed in what follows. 3.1 Differences between IT-clefts and extraposition: Pérez-Guerra (1998) Despite their separate treatment in the literature, IT-clefts and extraposition exhibit a number of affinities. First, as Pérez-Guerra correctly points out (1998: 8–9), they are both thematically marked, in that the theme slot is occupied by the pronoun it. Secondly, the two constructions have similar structural properties: they both have it as their initial element, followed by the copula be, an additional constituent (the highlighted element in clefts, and the remaining part of the predicate of the main clause in extraposition) and a subordinate clause. Compare the following pair of examples taken from the WSC: (8) FE but he does seem to flit around doing little bits and pieces MJ well he was over in england a few years ago and he went on a special eye course there to qualify it was very hard for him to get taken because of the old school tie in england and he being from the colonies but er after the course he got top marks so you know (WSC#DPC002) (9) AW well you’ve been um going out so much on friday nights you wouldn’t have taken much notice of that friday night meeting it’s always auckland anyway auckland always meets on fridays it’s a very only the very odd occasions someone else works you know actually someone else actually um (WSC#DPC032) As discussed previously in relation to examples (A) and (B), it is not straightforward to work out whether the two constructions given in (8) and (9) are IT-clefts or extraposition. In both cases, the clauses involve the pronominal it, the copula verb be and a sentence-final subordinate clause (to get taken and someone else works). A recent paper by Pérez-Guerra (1998: 10–11) claims that the two constructions are distinct. Three criteria are cited as distinguishing between them: Clefting and extraposition in English 15 (1) In IT-clefts, the element which follows the verb and precedes the final constituent in the superordinate clause is compulsory, whereas in extraposition, this element is optional. (2) In the case of that-clauses, the subordinate that-clause is ‘complete’ in extraposition, but ‘gapped’ in IT-clefts. The gapped element in IT-clefts is coreferential with the obligatory element following the verb and preceding the sentencefinal constituent mentioned in (1). Huddleston and Pullum (2002) exemplify this with the cleft construction It’s the president [I’m referring to __ ] (ex. 11i: 1418), which has a gap where the clefted constituent the president would normally occur in the unclefted equivalent I’m referring to the president. Another example from the WSC Corpus is given in (10), where the clefted constituent human eyes functions as an argument of the cleft clause he was working on and thus leaves a gap inside it which would be He was working on human eyes in the unclefted counterpart. Contrast this with example (11), where the extraposed clause was so lucky does not leave such a gap (that is, I got this is ‘complete’ without it): (10) FE i got a postcard from him the other day it looked a BEAUtiful university and um it’s obviously their break over there for about nine weeks and that’s when this course is running and he said there were vets and doctors evidently cos it was human eyes he was working on when he wrote from all over the world there (WSC#DPC022) (11) MD unle unle as long as you’ve booked to at i mean i’ve we haven’t i mean we haven’t it was so lucky i got this cos the the ballot closed three months ago just for September (WSC#DPC023) (3) According to Pérez-Guerra, and as noted previously by Quirk et al. (1985), there are fewer types of clauses which can be clefted than there are clause types which can be extraposed. And indeed, these need not be necessarily clauses, but can also be complex phrases. IT-clefts can have that-clauses, WH-clauses, and very rarely infinitive clauses, whereas extraposed elements can be: that-, whether-, if-, and why-clauses, adverbial clauses, NPs, and PPs. These possibilities are exemplified in Table 1. ICAME Journal No. 32 16 Table 1: Examples of the types of clauses which can be clefted and/or extraposed 3.2 Counterexamples from the WSC to Pérez-Guerra (1998) While Pérez-Guerra (1998) is right in claiming that clefting and extraposition are distinct from each other and that they can be distinguished, his criteria are not particularly successful in doing this, at least as far as some of the data in the WSC are concerned. I consider each of three criteria presented above in turn, and give examples showing their inadequacy in resolving ambiguity between the two constructions. First, while it is true that many IT-clefts exhibit an overt element between the predicate of the superordinate clause and the subordinate clause, these elements need not be present. Some clefts have no such element in that position, as in examples (12) and (13): (12) WE oh it’s so easy i’m sure it was that well i’ll i’ll get my b a in maori tomorrow IB but the thing is it was the teachers that made us speak er er all say English (WSC#DPC004) Clause Type IT-clefts Extraposition That-clauses It was a play that I saw. It was amazing that he did it. WH-clauses It’s their arrogance which they hated most of all. It was amazing how he spoke. Nonfinite clauses It’s still Mark to come. It is impossible to please him. Adverbial clauses n/a It seemed hours before she arrived. (ex.10, 11) Whether-clauses n/a no examples given by Pérez-Guerra If-clauses n/a no examples given by Pérez-Guerra Why-clauses n/a no examples given by Pérez-Guerra NPs n/a It is trewe euery word that is wretyn in Brides boke. (Kempe, The Book of Margery Kempe I, 1438, 47, ex. 14, 118) PPs n/a It is better vppon a scaffolde than vppon the grounde. (Fitzherbertm The book of husbandry, 1534, 38, ex. 15, 11) Clefting and extraposition in English 17 (13) AR well there’s auras people have auras and things like that so that er that didn’t frighten you either that thing BT read it mm yeah NO it was just so strange but i didn’t TELL the kids cos i didn’t want THEM to be frightened and there so it wasn’t like robbie was saying it because he’d heard me talk about it because i was very careful not to let them hear about it and so once he said it i told him what i’d seen too er so i don’t know and i don’t know where the ball of energy comes from or anything about it (WSC#DPC121) The examples above give a particular type of cleft, often termed ‘inferential cleft’ (see Delahunty 1981; Lambrecht 2001); what is highlighted is not the clefted constituent, but rather the cleft clause, namely i’ll get my b a tomorrow [b a = BA, Bachelor of Arts] in (12), and robbie was saying it was because he’d heard me talk about it in (13), respectively. Some (Hedberg 1990; Delahunty and Gatzkiewicz 2000) have analyzed the subordinate clause following the copula as being the clefted constituent itself (rather than the cleft clause). The debate as to whether the subordinate clause should be analyzed as a cleft clause or as a clefted constituent goes beyond the scope of the present work; however, it serves to draw attention to the fact that the analysis of the inferential construction is not a closed case and deserves further investigation.9 Even if the subordinate clause is analyzed as a clefted constituent, the construction still remains problematic for the test proposed, because under this analysis, the final constituent coincides with the obligatory constituent (it is not that the element following the verb and preceding the final constituent is obligatory as the test would predict of clefts, but rather, there is no distinct final constituent). Conversely, in the case of extraposition, the element occurring between the superordinate clause predicate and the subordinate clause is typically not optional, but rather obligatory, as given in (14a) – note that removing it produces ungrammaticality, as indicated in example (14b): (14a) MK i don’t know what it does to the cells of the things it’s good for cooking and eating but it dena CY you’re saying it would be better to put it into that thing and put hot water round the outside and that way it would be sitting in water MK microwaves (WSC#DPC077) (14b) * It would be to put it into that thing and put hot water round the outside. ICAME Journal No. 32 18 As far as the second criterion is concerned, clearly it is not only that-clauses that can potentially cause ambiguity problems; but other clauses (such as non-finite, WH-clauses, etc., see examples (B), (8) and (9)) can also be involved in problematic cases. However, leaving this point aside, it is not always the case that clefted constituents function as arguments of the cleft clause which they relate to. That is, in some cases, the cleft clause is not exactly a relative clause and the clefted constituent is rather an adjunct of the cleft clause. In such cases, the gap test does not hold in the same way as we have seen earlier, since the cleft clause is ‘complete’ without the ‘missing’ clefted constituent. Consider example (15a): (15a) TS and i think i’ll be i’m sure i’ll get maturity onset diabetes KA it’s for the sugar that it has to secrete the insulin TS yeah but i think your pancreas isn’t that what they th that’s one of the theories why they think people get maturity onset diabetes just from years of pancreas getting tired (WSC#DPC024) The clefted constituent for the sugar is an adjunct of the cleft clause it has to secrete the insulin since it is optional and the clause is complete without it (admittedly, the cleft clause does allow the PP to be present, but it does not require it). Hence in examples such as (15a), the gap test may be considered a weaker and perhaps not entirely convincing means for establishing the desired cleft classification. Similar examples can be found of cases where the clefted constituent functions only as an adjunct, and not as an argument in the cleft clause, and thus leaves a ‘weaker’ gap in it. Three additional ones are given in (15b)–(15d) from the WSC: (15b)BT oh oh AL and just at that point you suddenly get the shits you know you think i’m in a car with a mad man BT that like de da de da de da AL that’s right but you s it’s just at that point you realise how out of how little control you’ve got as a passenger in a car and um so then i’m starting to think about what to do next because what’s going to happen is ac is a another curve coming up (WSC#DPC049) Clefting and extraposition in English 19 (15c) LR and do you all have a um a support group time QT my chest yes yep