Inferentials in spoken English

Pragmatics 21:3.307-340 (2011) International Pragmatics Association INFERENTIALS IN SPOKEN ENGLISH* Andreea S. Calude and Gerald P. Delahunty Abstract Although there is a growing body of research on inferential sentences (Declerck 1992, Delahunty 1990, 1995, 2001, Koops 2007, Pusch 2006), most of this research has been on their forms and functions in written discourse. This has left a gap with regards to their range of structural properties and allowed disagreement over their analysis to linger without a conclusive resolution. Most accounts regard the inferential as a type of it-cleft (Declerck 1992, Delahunty 2001, Huddleston and Pullum 2002, Lambrecht 2001), while a few view it as an instance of extraposition (Collins 1991, Schmid 2009). More recently, Pusch's work in Romance languages proposes the inferential is used as a discourse marker (2006, forthcoming). Based on a corpus study of examples from spoken New Zealand English, the current paper provides a detailed analysis of the formal and discoursal properties of several sub-types of inferentials (positive, negative, as if and like inferentials). We show that despite their apparent formal differences from the prototypical cleft, inferentials are nevertheless best analysed as a type of cleft, though this requires a minor reinterpretation of “cleft construction.” We show how similar the contextualized interpretations of clefts and inferentials are and how these are a function of their lexis and syntax. Keywords: Inferential; It-cleft; Spoken language; New Zealand English; US English; Discourse; Just; (not) as if; Like; Discourse markers; Wellington Corpus of Spoken New Zealand English. 1. Why study inferentials in conversation? Ever since Austin (1962), linguists have been musing over how speakers might “mean” more than they “say” and how it is possible for expressions to go beyond their literal meanings into the realm of inference and implicature. Second language teachers and learners are well aware that knowing a language involves more than knowing its words and rules/patterns. One of the linguistic aspects which can account for these facts has to do with the layering of information. Languages present communicators with various means, some syntactic, some lexical, for signalling not only informational content, but also how this content is organised, how it is to be related to co-text and context, and what the speaker's stance towards it is. These devices can help readers and listeners to work out what is new, what is salient or unexpected in a message, and how it is to be related to its discourse context. * AC is indebted to Jim Miller, Peter Collins, Chris Venditti and Christian Koops for fruitful discussion and thanks the New Zealand Foundation for Research Science & Technology. The authors are grateful to the anonymous referees and the journal editor for invaluable comments and suggestions. 308 Andreea S. Calude and Gerald P. Delahunty Clefts constitute a syntactic means for signalling what is new or salient. English is particularly rich in cleft types; some of the main ones are exemplified in (1); see Lambrecht (2001) for a more comprehensive survey. (1) it-cleft: wh-cleft: reversed wh-cleft: demonstrative cleft: all-cleft: presentational cleft: It was a miracle that Obama hoped for. What Obama hoped for was a miracle. A miracle is what Obama hoped for. That’s what he hoped for. All that Obama hoped for was a miracle. There’s a miracle Obama hoped for. Clefts point out larger discourse relations (Doherty 2001: 459-461) and thus tell us something about how text types (or discourse portions) are put together. It can be said that they lie at the boundary between the clause and the larger discourse, connecting the two levels. This, together with the fact that they involve subordination, makes clefts a rather complex construction, a point that becomes relevant in justifying its analysis in conversation. One construction - a term we use in its traditional, broad sense of grammatical type, with no implication that it is paired with conventionalized semantic and/or pragmatic content, (see Goldberg 2006 and Sag 2010) - which has been included in the cleft category is the inferential (Declerck 1992; Delahunty 1995, 2001; Koops 2007; Huddleston and Pullum 2002: 1418). Examples include: (2) (3) (4) (5) It is that the recession is tough on young families. It's that it's tough to get by even. It is just that the Swine Flu outbreak is so unpredictable. It's not that Obama is brighter, he isn't really. It's just he's more charismatic. It's not that I want more money; it's just that I don't want to be forced to look at every penny. Testimony to the need for further investigation of this construction comes from the existence of several competing analyses of the inferential besides the cleft one, namely, an extraposition analysis (Collins 1991; Schmid 2009), a discourse marker analysis (Fraser 1999; Pusch 2006, forthcoming), and a non-cleft copular analysis (Heggie 1998). We know surprisingly little about inferentials despite their cross-linguistic pervasiveness: Bearth (1999) and Delahunty (2001) provide examples from French, German, Irish, Italian, Japanese, Mandarin, Norwegian, Spanish, Swahili, and Akan; Delahunty and Gatzkiewicz (2000) discuss Spanish inferentials. While this dearth of knowledge has been attributed to the construction's comparatively low textual frequency (Bearth 1997: 15, 16), its cognitive entrenchment, i.e., that it is stored as a cognitive whole with very limited possibility of variation (Schmid 2009), the specialized role it plays in discourse, and the disagreements about its analysis strongly suggest a need to revisit this controversial construction. In addition to it-clefts, the following constructions seem, at least at first glance, somewhat related to inferentials: (6) (7) (a) Not that it follows. (Delahunty 2006: 216, ex. 7) (b) Not that I fancied him before but... (Schmid 2009) (a) Well that’s love for you. (García 2007: 66, ex. 1) Inferentials in spoken English 309 (b) Es lo que tiene. ‘That’s the thing for you.’ (García 2007: 72, ex. 13d) The examples in (6) are analysed as reduced versions of inferentials by Horn (1989). However, Delahunty (2006: 221-225) argues against their treatment as (negative) inferentials on grounds that: (1) examples such as the ones above do not have a positive counterpart, (2) they are not synonymous with equivalent negative inferentials (in this case, It is not that it follows, and It is not that I fancied him before but..., respectively), (3) their matrixes do not allow auxiliaries, modals, or adverbial modifiers, and (4) they do not appear in “tandem pairs” - a negative instance followed by a positive one, as is typical of inferentials.1 Furthermore, not that S clauses suggest that the speaker is not being completely sincere or believable, which is not the case with inferentials (Delahunty 2006: 213).2 Therefore, the discussion here does not concern these examples further. García ties her examples in (7) to inferentials on the basis of their structural properties. She claims that both constructions involve a copular matrix which always occurs in third person singular and involves a null subject (2007: 68-69). However, there are also differences between them, as García's examples do not allow the copula to occur in the negative form or be accompanied by modal verbs suggesting possibility or doubt: (8) (9) * Well that’s not love for you. * Well that may be/could be love for you. For these reasons we will assume that these constructions also are not relevant to our analysis of inferentials. Generally, the inferentials previously analysed were collected from written texts (for example, Declerck 1992; Delahunty 1990, 1995, 2001), with the exception of Koops (2007), who addresses inferentials in spoken American English. However, recent work on spontaneous spoken language has shown that many constructions found in this mode are very different from their counterparts in written language. Owing to space and time constraints, we refrain from reviewing the entire literature here, but include several examples below. A classic illustration comes from the work of Miller and Weinert (1998/2009). They show that certain apparently subordinate clauses in spoken English function as discourse-subordinate, even though they are structurally independent of the main clause they are assumed to be subordinate to. (10) (Radio Discussion Programme, from Miller and Weinert 1998/2009: 131, ex. 113a) everyone knows Helen Liddell how hard she works 1 However, it is interesting to note that the New Zealand English inferentials also do not tend to occur in such tandem pairs either (only 2 of the 55 constructions identified take part in such pairs). 2 Schmid (2009) analyses not that sentences (NTS) like those in (6) as instances of extraposition and ellipsis from the form that … is not the case/true for focusing purposes. Delahunty (2006: 218-9) argues that NTSs do not “deny the truth of the propositions their S’s represent,” just their local relevance, thereby rebutting any analysis in which the NTS clause is embedded as an argument of a truth predicate. 310 Andreea S. Calude and Gerald P. Delahunty Because of its position, the clause how hard she works appears to function as a modifier of the noun phrase Helen Liddell. Miller and Weinert claim that this is a case of discourse rather than syntactic subordination. This is because from a syntactic point of view, the how-clause is dislocated from the main clause (compare it with the integrated wh-complement version: Everyone knows how hard Helen Liddell works). The phenomenon of loosely integrated or altogether un-integrated clauses in spoken language was also noted of relative clauses in Russian (Zemskaja 1973: 228ff. cited in Miller and Weinert 1998/2009: 130-131), Finnish että-clauses (Laury 2006), and conditional clauses in Italian, German, Finnish, Japanese, Swedish (Vallaurí 2004), and Australian English (Stirling 1999). An investigation of “object complements” in conversation by Thompson (2002) led her to wonder whether there is any justifiable basis for the term “object complement.” She argues that object complements such as the ones in (11) do not function as subordinate clauses, and are formulaic, encompassing recurring patterns of recycled portions of discourse. (11) (from Thompson 2002: 139, ex. 22-24) I think it'll be interesting. I don't think my brother's so active. I remember talking to him. Calude (2009a, 2009b) examines a construction more closely related to the one investigated in this paper. She argues that demonstratives clefts, such as That's what I am talking about, are formulaic in conversation. She shows that the formula [That's/that was + what/where/how + cleft clause] accounts for an overwhelming majority (80%) of the constructions found. In a similar vein, a recent paper by Hopper and Thompson (2008) argues that (assumed) biclausal structures, including English wh-clefts and extraposed clauses, and German wenn-clauses are restricted in the components they allow, and behave like monoclausal units. They suggest these should be analysed instead as “single, partly formulaic clauses deployed by speakers in managing interactional discourse” (2008: 99). Calude and Delahunty (2010) studied inferentials found in the Wellington Corpus of Spoken New Zealand English, and argue that these are partially formulaic: they are lexically limited, situationally bound, relatively frequent (compared to other lexical bundles) and serve a specific discourse function. However, they are not (perhaps, yet) fully established “fixed formulae” since they are semantically transparent, compositional, and non-fluent. The body of work mentioned above indicates that in spontaneous spoken language, many constructions are simplified and formulaic (Pawley and Syder 1983; Wray 2002, 2008 and many others; see Edmonds 2010 for a review), consisting of a set of predictable patterns associated with specific discourse-related properties and interactional characteristics, e.g., prospection as discussed in Hunston (2006: 60-1) and projectability as discussed in Hopper and Thompson (2008). The full variety of the component forms and structural integratedness found in written versions of these constructions is simply not borne out when the construction occurs in the spoken medium. The most affected expressions are complex constructions, particularly those involving subordination. These are “reduced” to only a few possible patterns, in part Inferentials in spoken English 311 due to the decreased cognitive loading required for their encoding and parsing, and in part, to their acquiring specialised interactional functions. One typical consequence of this becomes the replacement of syntactic subordination in favour of hypotaxis (Halliday 1987) or of beads-on-a-string organisation. With these observations in mind, we set out to investigate the forms and functions of inferentials in spontaneous conversation in order to test whether inferentials display formulaic patterns analogous to those displayed by the complex structures mentioned above. Because we have chosen to use conversational data from New Zealand English, we are in a position to compare the inferential forms and their frequencies in that dialect with the forms and frequencies of their use in US English, as presented in Koops (2007). However, because we identified forms rather than functions, we cannot compare the dialects on the various alternative ways in which their speakers might express the functions inferentials express. The aims of the current work can be summarised as follows: a) to provide a descriptive account of several types of inferentials that occur in New Zealand English conversation, including their frequency of occurrence and structural properties, with the goal of determining whether inferentials conform to the formulaic patterns observed for so many other complex constructions in spontaneous spoken conversation, b) to show that inferentials are best analysed as clefts, c) to describe and account for the discourse functions of the various inferential types, we examine and show how these derive from the interaction of their forms with more general pragmatic principles and the context of their use, with no need for construction-specific stipulations. The paper is organised as follows. Section 2 details the data investigated and the methodology used for identifying inferentials. Section 3 describes the structural characteristics of the constructions found and the speakers who use them (essentially aim a)), and section 4 addresses aims b) and c). Finally, the conclusion summarises the main points made. 2. Data This investigation of inferentials was conducted by exploring excerpts of spontaneous spoken (New Zealand) English found in the Wellington Corpus of Spoken New Zealand English (WSC henceforth). The conversations were conducted in participants’ own homes, without a specified topic or an interviewer present, thereby facilitating as natural a setting and interaction as possible (see Holmes et al. 1998 for a guide). The excerpts were chosen randomly and comprise approximately 250,000 words from 116 conversations, yielding 55 inferentials (see Appendix B for a complete list). This is not a big sample by today’s standards, but the collection was obtained with the aim of comparing it to the study of inferentials in spoken American English conducted by Koops (2007). A typical example is given below. 312 Andreea S. Calude and Gerald P. Delahunty (12) WSC DPC1203 MQ: oh careers yeah <,> reminds me i've got to get fit → FN: <laughs > i want to do it eh it's just i've got to get myself motivated <latch> yeah well how long have you got till MQ: oh i can any time after this this month i can reply The inferentials were found by manual inspection. This was because their elements occur in a wide range of constructions making it impossible to specify a search string that would uniquely identify them. This is also, in part, responsible for the size of the dataset obtained. Under closer scrutiny, it quickly became clear that the element that is not always present since other subordinators may be used, or none at all (e.g., It was not like he even cared about her, It is just I hadn't seen her for ages), and the copula may be accompanied by modifiers (adverbs, modal verbs, or negation). Moreover, expressions which appeared to be suitable candidates as inferentials (containing it, a copula, and that) turned out to be false starts or reformulations, as shown in (13), or simply ambiguous, as in (14), which could be either an inferential or a different type of it-cleft. Such factors make it impossible to automate the search for inferentials. (13) WSC DPC169 SH: not quite a stomp < laughs> <pause > she knows jolly well AD: that plane looks as though only one of its engines is going → SH: <laughs> well shall we start cooking some tea <,> it's just about that rice will just about be cooked i would think the rice is cooked is it okay AD: the rice IS cooked (14) WSC DPC115 AT: anyway we'll look at these some other time BR: <unclear word> yes AT: frequently and yes they have the criteria <reads> rarely sometimes USUALLY frequently and consistently </reads> and the frequently round the other way because to me frequently means often and usually means well that's my HABIT but occasionally i break it if i say i FREQUENTLY have breakfast in the mornings to me that might mean four or five times a week but if usually → have breakfast it means that i DO have breakfast it's just once in a blue moon i forget or or don't moon i forget or or don't have time so to me they're the other way round but you could argue it until the cows came home and i bet there'd be people who would differ on that so the very <laughs> criteria that you set down are so important and and so PERSONAL BR: yes yes 3. Findings from the WSC corpus This section summarises the findings uncovered from the WSC data. We first discuss the structure of the inferentials, followed by remarks regarding the speakers who use them and their distribution in the conversations investigated. One of the most obvious observations to be made is that none of the inferential constructions found in the New Zealand English corpus occurs with a cleft clause, even 3 The WSC examples are annotated for various discourse features, see Appendix A for a list and the explanations of the conventions used. Inferentials in spoken English 313 though examples of this type have been cited in Delahunty (2001) and Huddleston and Pullum (2002) (see the examples in 15 below). The lack of cleft clauses has been interpreted as a potential problem for the cleft analysis, as we will see in section 4. (15) (a) I wonder if it was that they hadn’t room enough for them up in the house that they put them out here in the woods? (Delahunty 2001: 520, ex. 14) (b) It's that he's so self-satisfied that I find offputting. (Huddleston and Pullum 2002: 1418, ex. 14i) Secondly, as noted by Koops (2007), but surprisingly little elsewhere (though see Delahunty 1995: 342ff), in most examples, the copula is modified either by the degree adverb just, or by some negation particle (either not or its contraction). Using the terms and categories included by Koops in his analysis of inferentials in spoken American English, we note the following comparison between the two English varieties in Table 1. Table 1. Comparison between spoken American and New Zealand English Types Example Spoken US English (Koops 2007) Spoken NZ English (WSC corpus) it's not that/as if it's just that what it is that it's precisely that 28 14 4 3 9 39 0 0 it could/may be that 1 1 it is well/you know/ i mean 0 that 2 Modified negation just wh-cleft degree adverbs epistemic modals discourse markers Unmodified plain it is that .... unmodified after negative it's not that... it's that... IC4 TOTALS 0 4 1 0 51 55 As in the American English data, a great majority of the inferentials in New Zealand English conversation are modified (98% in the USE data, 93% in the NZE data). These findings go against the bias of attention directed at unmodified inferentials in articles discussing the construction (cf. Koops 2007: 214 Table 2). Having said this, the NZE data did have four unmodified inferentials, that is [IT IS + (that) S] (e.g., it's that you've got to have them <,> three weeks apart). The most striking difference between the American and New Zealand data is the inverse proportion of negative and just-inferentials. In American English, it is the 4 There is one pair of this kind, but the second inferential is marked by just: It’s not that X, it’s just that Y. 314 Andreea S. Calude and Gerald P. Delahunty negative inferential which is most common (slightly over half of the examples, 55%), whereas in New Zealand English, it is the just-inferential which is favoured (well over half of the examples, 71%). Interestingly, only 16% of the examples found in the New Zealand data were negative. A Chi Square Test reveals that the differences between just-inferentials and negative inferentials in the two varieties of English are significant (2=17.297, p<0.001 for negative inferentials, and ^2=6.087, p<0.014 for justinferentials). We discuss each type in later sections of the paper (see section 5.3 for negative inferentials and section 5.4 for just-inferentials). From these figures we can infer that NZE and USE speakers do indeed differ in their choices from among the inferential types at their disposal. Our sample is random, representative, and unbiased because it consists of well-collected sets of conversations which are balanced for gender, ages and education level of the participants, and people are recorded in their own homes, without an interviewer or a given topic of conversation. The results are therefore generalizable. In fact, with small sample sizes, the Chi Square Test is less likely to detect differences, so because our samples of approximately 50 show significant differences, we should expect that larger samples would show even greater differences.5 However, the test does not tell us why we see these differences and at this point in our investigation we can only speculate: it is to be expected that speech communities that are relatively isolated from each other, as NZE and USE speakers are, would display different frequencies of use of all linguistic elements, but perhaps especially of marked constructions. We should therefore expect that speakers of each of these two dialects would choose alternative ways of expressing what speakers of the other dialect might choose to express using an inferential form. For example, USE speakers, who tend to use just-inferentials less frequently than NZE speakers, might opt for a noninferential construction with the discourse marker just, or a thing-construction; for instance, (16) (b) or (c) rather than (a): (16) (a) It's not that Paris is expensive, it's just that it is full of French. (b) It's not that Paris is expensive, it’s just full of French. (c) It's not that Paris is expensive. The thing is it’s full of French. Complementarily, NZE speakers, who tend to use negative inferentials less frequently than USE speakers, might opt for an unmarked construction with various pragmatic or discourse markers; for instance, (17) (b) instead of (a): (17)(a) WSC DPC007 LL: you have a big role to play I have a role to play it's not as great as yours but i still have a role to play RR: mm LL: right <short pause> <drawls> and um RR: you so you don't want to have to look over your shoulder to make sure you don't → LL: oh it's not that i don't want to have to look over my shoulder NOTHING should HAPPEN should OCCUR during those procedures ANY PART of it 5 The Chi Square Test requires that we make assumptions of expected counts of the various subtypes of inferential across the two dialects. Since there is no a priori expectation that the use of these subtypes will differ in American and New Zealand English, we use the default counts measure. Inferentials in spoken English 315 that's ALL formal the WHOLE lot right from the time the people come onto the marae (Maori: "meeting house") until the time everyone's LEFT the marae all right (b) RR: → LL: you so you don't want to have to look over your shoulder to make sure you don't oh of course i don't want to have to look over my shoulder but NOTHING should HAPPEN should OCCUR during those procedures ANY PART of it that's ALL formal the WHOLE lot right from the time the people come onto the marae (Maori: "meeting house") until the time everyone's LEFT the marae all right Investigating these dialect differences in preferred expression-types would take us far beyond the brief we’ve set ourselves for this paper. Further quantitative investigation of them would require that we begin by identifying specific discourse functions and then identifying the various forms which are selected by speakers of the two dialects for those functions - a difficult but rewarding study, which we leave for another day. The WSC corpus also shows that the so-called that-clause is not always introduced by the subordinator that. Some inferentials contain no subordinator at all (as in 18), and others use the newly-admitted member of the subordinator class like (LópezCouso and Méndez-Naya In press; Romaine and Lange 1991) (as in 19). (18) WSC DPC290 FR: [tells a story] ... like i was very you know when <laughs> when we were kids we were always taught you gotta lock the car before you leave so i locked all the doors and <,,> everything and so the um yeah so i locked his keys in there and so i told him he needed a spare key in the place sec spare set but he didn't you know and <quickly>then the other night</quickly> i know what happened to me → MQ: mm yeah yeah <laughs> he does it's just he doesn't trust you that's all cos he knows that you'll get a hold of it and <latch> FR: <laughs> oh yeah MQ: take it for a burn FR: you reckon (19) WSC DPC326 JI: she's very bright yeah she's er she's done law AL: <drawls> mm <latch> JI: and she was the legal advisor for er ronnie burch AL: right <latch> JI: you know <drawls> when when he was yeah race relations officer yeah <with creaky voice> mm </with creaky voice> but um AL: yeah race relations oh <drawls> good so she didn't have a problem getting a job i suppose when her when er <unclear word> → JI: no no but sh she found that particular job very stressful it's like she doesn't think she'd like to go back into it you know cos she was always dealing with problems AL: <drawls> mm right people and problems JI: yeah never really feels happy eh <latch> 316 Andreea S. Calude and Gerald P. Delahunty According to López-Couso and Méndez-Naya (In press), the subordinators as if, as though , and like “can occasionally be found in complementation structures introducing finite declarative clauses,” as like does in “‘cause some people seem like they are older . . .” (Their example 7, p. 4.) López-Couso and Méndez-Naya give five main types of evidence for this claim (pp.5-9): (1) The clauses introduced by these expressions function as arguments of an “appropriate” main verb and so cannot be omitted with out causing ungrammaticality; e.g., *It seemed. (Their 5a.) (2) These clauses can be replaced with that-complement clauses (i.e., by unambiguously declarative clause complements); e.g., and as time passed it seemed that/as if/like the strange little man had never been there. (Their 5b adapted.) (3) These clauses can coordinate with clauses that are unambiguously declarative clause complements (e.g., with that-complements); e.g., Do you feel like your neighbours care about you or that you have any sense of community that way? (Their 11.) (4) These clauses lack the mobility that is typically associated with adverbial complements (i.e., they cannot move around within their clause complex/sentence); e.g., *That/as if/like the strange little man had never been there seemed. (Their 5c adapted.) (5) These clauses and those introduced by that can all be replaced by the pro-form so (McCawley 1998); e.g., It seemed so. (Their 5e.) Some of these tests can be applied to inferentials to demonstrate that as if and like function as complementizers in them too. The clauses cannot be omitted without causing ungrammaticality or at least a change of meaning. For example, omitting the like clause in (19) creates the ungrammatical *It’s, and omitting the as if clause in (22) creates It’s not, which has a different meaning. Additionally, like can be replaced by that without appreciable meaning change in It's not like/that they NEED someone. And like and that can co-occur in coordinate structures: It's not like they NEED someone it's just [that] they're just doing it as a favour because susannah's a mate. (See 45 below.) Additionally, López-Couso and Méndez-Naya cite multiple other works which discuss the uses of like, as if, and as though as complementizers, including Bender and Flickinger (1999), Rooryck (2000), Quirk et al. (1985, esp. 1175 note a), McCawley (1988), Dirven (1989), Huddleston and Pullum (2002: 608 note, 962, 1151-1152, 1158), Dancyngier and Sweetser (2005), Taylor and Pang (2008). It is difficult to ascertain whether the appearance of like in inferentials is specific to New Zealand English since no one has previously mentioned examples of this kind, though this seems unlikely, as speakers of US English readily accept them. This use of like can be problematic since it is not always straightforward to distinguish subordinator like (used in inferentials) from comparative like (as in 20) or quotative like (as in 21) (which are not inferentials). (20) WSC DPC123 TM: accountancy too and i said he said oh what year are you in and i said <laughs> THIRD and he goes oh no peter's first year Inferentials in spoken English 317 → BE: → CH: KM: CH: <laughs> i know yeah cos it's like those um christians that asked me to do that survey for them she goes what year study are you and i said third and she goes oh oh are you <laughs> and i thought that's JUST what i yeah <latch> it's like having a accent <latch> mm (21) WSC DPC262 AN: and i dated the cheque <,> sorry early in the seventh month of this this year and they still haven't cashed it and they said that it's supposed to be coming through on the fourth of the ninth and she said well you've got to comeup and cancel those two cheques and pay for them in <drawls> ca cash <,> send l v → martin and son ch cheques and cash and it's like how am i gonna fix THAT up they are SO stupid BL: oh yeah <quietly>o no don't know because if you cancel it then they could be <,,> still Finally,6 the corpus also exhibits two as if-inferentials, as in (22). They are parallel to like-inferentials, since as if can be replaced by that or like (it's not that/like she has to stay in Kyle's room), though that indicates greater certainty than like or as if, and as if indicates a higher stylistic level than like. Although the corpus does not provide examples of a positive version of the construction, this is possible, e.g., It's as if she has to stay in Kyle's room. (22) WSC DPC032 KT: jan is quite happy to go into mainstream that's what she said she goes yep i'll come to your school you tell rhonda that <quickly> i don't want to go overseas and <laughs> really all this blah blah blah and <quickly> i'm quite happy to do mainstream she says i'll just come along and just slip on in to mainstream and you know blah blah blah RW: <drawls> yeah <laughs> yeah <latch> good on her <latch> KT: and i was sort of thinking <quietly> oh wow cos you see RW: see the thing is too if she does get in there i mean even if she gets in into the school with a permanent position <latch> KT: mm <latch> → RW: it's not as if she has to stay on and STAY in kyle's room <latch> KT: no well that's right and that's what i was thinking IF she won the position It must be noted that although (not) that-inferentials (e.g., it's not that I don't want to have to look over my shoulder), like-inferentials, and as if-inferentials seem to allow the use of modifiers suggesting possibility or doubt (e.g., it could (not) be that/like/as if she has to stay in Kyle's room), examples of these types do not actually occur in our data.7 We discuss these constructions in further detail in section 5.4. 6 A further candidate for the inferential category might be examples such as It’s not to say that X; e,g., it not to say that we don’t want to do the same thing to her wiki (WSC DPC007), but on closer consideration these constructions differ from inferentials in two respects: (1) the to say predicate adds a further level of embeddedness than that found in inferentials, and (2) it often alternates with that (which is referential, not expletive). For these reasons we do not include them in our analysis. 7 A natural follow-up question to ask might be whether, in addition to a wider range of subordinators (such as like, and as if), inferentials might also include a wider range of verbs in the copula 318 Andreea S. Calude and Gerald P. Delahunty Because we believe that inferentials are related to constituent-focus it-clefts, we will briefly address the information-structuring effects of it-clefts and inferentials. Though other information-structuring configurations occur, constituent-focus it-clefts typically have focal stress on (a part of) the focused constituent, so that (a part of) the focus constituent is marked as new or contrastive and the modifying clause typically provides presupposed or backgrounded information. However, because inferentials typically do not cleave a basic clause in two, they clearly cannot structure information as it-clefts do. The data show that while some inferentials bear stress on elements inside the subordinate clause (as in 22 above), in most cases, there is no focal stress, or if there is, it is placed on the modifier of the copula, typically just or sometimes not. In (23), there is no special focal stress assigned, and the element which appears to receive most prominence is the modifier just (in both inferential constructions). (23) WSC DPC037 KT: shoes do you wear you yeah did you do you wear your <,> inner sole BD: yeah yeah i wear them all the time but just she said it's okay if i wear them once every second day like and wear sports shoes <,> one day <unclear word> KT: what is it gonna make it's not you're never gonna be better of it though are you <,> no → BD: better of the way i walk no it's just that you know KT: yeah <unclear word> right cure or whatever → BD: it's just you know how i had those shin splints it's just that i don't get those <,> KT: oh right because it doesn't make it worse doesn't aggravate it BD: because the reason why i got shin splints is because the way i walk so that if when i do sport i have those in and walk or <unclear word> With only one exception, the subordinate clauses (that-clauses, like-clauses and so on) identified in the corpus data contain GIVEN information (Gundel et al. 1993), by and large pronouns and proper names, and in some cases definite noun phrases. However, what is NEW in positive inferentials is the whole proposition expressed by the subordinate clause, as in (24). There, the participant identified as “jim” is mentioned position. In other words, would inferentials extend to include those in examples (2)-(4) below, or only be restricted to those in (1a,b): (1a) It is that he avoided this whole matter by resigning. (1b) It may be that he avoided this whole matter by resigning. (2) It seems like he avoided this whole matter by resigning. (3) It looks like he avoided this whole matter by resigning. (4) It appears that he avoided this whole matter by resigning. We feel that sentences in (2)-(4) involve a richer semantic content than those under (1). The boundary between inferentials and other complex clauses with impersonal/non-referential subjects is, like many other linguistic phenomena, fuzzy and unclear. In order to draw a line between inferentials and noninferentials, most analyses (of clefts in general, not just inferentials) appear to hinge on formal properties of the constructions, that is, the presence of the non-referential it, a copula, and some kind of subordinate clause. However, these semantically bleached elements may be replaced by semantically richer items (with more referential content) or may be modified. When this happens, it becomes unclear what kind of construction we are dealing with. For the purposes of the current work, we exclude examples such as those in (2)-(4) from the analysis, though we do call attention to the fact that establishing whether or not they should be included as inferentials is not a trivial matter. Inferentials in spoken English 319 already in the conversation (and he is known to both participants), but what is newsworthy is AA's assessment that the interpersonal difficulties Jim poses are unintentional. (24) WSC DPC138 BC: that'd be interesting actually if he and mike a pair of them <laughs> AA: oh that's all he said cos it at first i was just saying about how i jim's about the <,,> only one i've had problems with BC: <drawls> oh right yeah AA: cos like jim's the only one <short pause> that won't always <latch> BC: yeah i know he's just harder to get on with → AA: yeah oh it's just that i don't think that jim does it on purpose i think it it's him <latch> BC: it's just him oh yeah AA: yeah it's his personality <latch> BC: personality <softly> yes However, in negative inferentials, the subordinate clause does contain GIVEN or at least information, but the newsworthy material comes from the negation of this information, as in (17a), repeated here as (25). INFERABLE (25) WSC DPC007 LL: you have a big role to play I have a role to play it's not as great as yours but i still have a role to play RR: mm LL: right <short pause> <drawls> and um RR: you so you don't want to have to look over your shoulder to make sure you don't → LL: oh it's not that i don't want to have to look over my shoulder NOTHING should HAPPEN should OCCUR during those procedures ANY PART of it that's ALL formal the WHOLE lot right from the time the people come onto the marae (Maori: "meeting house") until the time everyone's LEFT the marae all right Given the discussion in section 1 of the paper, it might be expected that the inferentials identified in the spoken (New Zealand) English conversations would exhibit formulaic tendencies. While it is true that the constructions involve GIVEN information, and that they do not have overt or otherwise ellipted cleft clauses, it cannot be said that inferentials are formulaic in speech for reasons outlined below. Because we lack clear and decisive criteria for determining in specific cases whether an expression - or, more accurately, a use of an expression - is formulaic or not (though see Wray 2002, 2008), we cannot say with certainty whether a specific inferential is or is not formulaic. Of course, formulaicity is not a binary feature: As with other grammatical notions, it represents a cline from a more “novel” end to a more “fixed” or “formulaic” end. We locate the inferentials away from the formulaic end of such a continuum, for the following reasons (see Calude and Delahunty 2010). First, the inferential matrix may include modals, adverbs such as just, only, simply, actually (see #15 in Appendix B), and not, which may or may not be contracted, as well as its copula, which is most frequently in the present tense, but may also occur in the past. Second, the matrix may be interrupted by discourse particles such as you 320 Andreea S. Calude and Gerald P. Delahunty know, as in #s16 and 19 in Appendix B, and pause fillers such as um, as in #42 in Appendix B, and Oh, as in # 51in Appendix B, and so on, suggesting that this is not produced as a fixed unit. Third, the inferential clause may be introduced by a variety of subordinators: that, as if, like, or zero (in fact, even more linkers than previously noted in the literature). Finally, though less significantly, the subordinate clause can itself be very complex, consisting of further embedded clauses, as in (24) and (26). The inferential in (26), contains three levels of embeddedness, where the matrix copula contains the subordinate clause she needs to be on a job like that where she doesn't have to do anything, which involves the relative clause where she doesn't have to do anything. All this is hardly formulaic behaviour.8 (26) WSC DPC059 AC: everyone thinks she's an awesome organiser and that's why she got the job on exec but she does she CAN'T organise ANYTHING at all i mean she's a real mess you know she's REALLY NICE and she's really lovely but she's a → TOTAL MESS totally <,,> disorganised and messy and and so i mean it's just she needs <laughs> to be on a job like that where she doesn't have to do anything BS: mm AC: but she can help out the other people and then they In sum, while there may be some legitimate reasons for questioning the biclausal nature of certain constructions in spoken language, as Hopper and Thompson (2008) have claimed for wh-clefts and extraposition, there is no need to conclude that all such constructions are monoclausal. The inferential is one construction that does appear to manifest biclausal status, even in spontaneous conversation. As inferential matrixes seem to allow the range of elements that we would expect in any clause, they should therefore be regarded as generated by the general rules that generate basic clause structures.9 8 In fact, the inferential can be even more complex than shown here, allowing a wider variety of tenses and allowing modals in the matrix, e.g., It may have been that the data collection points were illchosen, or allowing modals in both the matrix and the subordinate clause, e.g., It couldn't be that he would go there again so soon and so on. We refrain from including such examples in our discussion since they were not actually found in the corpus data analysed here, but our feeling is that they are certainly possible though very infrequent in spoken English. 9 A reviewer for Pragmatics suggests that the “onset” of the inferential might be formulaic. The reviewer did not specify what s/he meant by “onset,” but we assume that it denotes the non-referential matrix subject it and the copula. We argued in our main text that the whole inferential matrix clause, including the subordinator, is not formulaic, and we do not believe that the “onset” is either, at least not at this point in the construction’s history. It and the copula are ideal elements for constructions such as meteorological expressions and extrapositives that require a semantically null matrix, so they are not unique to the inferential and so not diagnostic of it. And because they occur in a number of constructions, they are likely to be generated by the grammar and lexis of English, and thus are not fixed. However, while three of the inferential “onsets” in our database are realized as it was, two as it wasn’t, and one as it may be, 89% (49/55) of the “onsets” are realized as the contraction it’s. Though this contraction is readily accounted for as an index of the oral mode, its frequency suggests that we should hesitate before declaring the inferential onset to be always compositional. Moreover, because these onsets are semantically null, they fit Wray’s (2002) “non-salience” criterion, and are therefore preferred candidates for fixation. We believe that the inferential matrix will eventually become fixed and reduced as Bearth (1999) has argued for the Swahili focusing particle ndiyo and the Akan particle nà, both of which derive from Inferentials in spoken English 321 Having discussed structural properties of inferentials in spoken New Zealand English, we now consider two final observations regarding their use. It has been noted by Schmid (2009) that Not that Ss (e.g., not that it matters) were used significantly more frequently by women than by men. This prompted us to investigate the participants uttering inferentials in the New Zealand English data. At first glance, it appeared that women did indeed produce twice as many inferentials as men (45 to 19, respectively); however, on closer inspection, there were also roughly twice as many women speakers recorded as men. So, as far as the New Zealand English data is concerned, inferentials were uttered in approximately equal proportions by men and women. The final observation concerns the distribution of the inferentials in the conversations. All 55 inferentials we identified occurred in only 40 (35%) of the 116 interactions investigated. Several hypotheses emerge: (a) the inferential is a tool which is part of the repertoire of some speakers and not others; (b) the inferentials are appropriate in such a narrow range of contexts that they rarely occur; or (c) the use of the inferential is contagious such that if some speakers in a conversation use it, this prompts other participants in the interaction to use it also (see Calude and Miller 2009 for evidence of contagious behaviour - essentially priming - in spoken grammar involving clefts). The inferentials found in the New Zealand data tend to be uttered by the same speakers within each of the conversations, for example, conversation DPC059 contains four inferentials, all uttered by the same speaker. Overall, only five out of the forty conversations contained inferentials uttered by different speakers (and many of these had at least one or two participants who did not produce any inferentials). This leads to the suggestion that either the use of the inferential ultimately depends on the rhetorical repertoire of the individual speakers involved (their idiolect, see Barlow Forthcoming), or it is functionally motivated to appear in particular discourse contexts whose comparative rarity accounts for the comparative rarity of the construction, which may have to compete in those contexts with other expression types that perform similar discourse functions. However, this issue is beyond the scope of the present work. 4. Inferentials subtypes: Analyses and discourse functions Section 3 detailed the structural properties of the inferentials found in the WSC data, along with information about the participants who use them and the conversations analysed. We now turn to the analysis of the construction with the aim of using our expletive matrix constructions. While the fixation of English inferential “onsets” is most likely to begin in the oral mode with the contraction it’s, it is nonetheless still systematically possible for speakers to produce uncontracted forms, especially when the tense is changed (e.g., it was), or a modal intervenes between subject and copula (e.g., it may be), as well as forms with alternative contractions (e.g., it wasn’t), indicating that the lexico-grammar is still operative, or at least available. However, it is worth bearing in mind that our data contains no instance of uncontracted it is, an absence which might reasonably be taken to support our speculation that it’s, if not merely an artifact of our relatively small database, might very well be the initial locus of fixation. 322 Andreea S. Calude and Gerald P. Delahunty findings to resolve some of the questions raised in the literature. We divide our discussion into structural analyses and pragmatic analyses. 4.1. Structural analyses 4.1.1. The extraposition analysis It has been suggested that inferentials may constitute a case of extraposition (cf. Collins 1991 and Schmid 2009); that is, in its underlying form the inferential clause occupies the matrix subject position but in the course of its derivation it is moved to its surface position and the matrix subject position is filled with expletive it. For the reasons that follow, we believe that this analysis is untenable. Extraposition is understood to be functionally motivated: a psychologically oriented view argues that sending heavy constituents to the end of their constructions allows for easier processing (see Huddleston 1984: 354 and Langacker 1974: 653 for this position, but also Quirk et al’s. 1985 explanation in terms of end-focus and endweight principles). So in spoken language, extraposed that-clauses are the norm, rather than non-extraposed ones (Huddleston and Pullum 2002: 960), particularly in a genre such as conversation, which likes to avoid heavy subjects (Miller and Weinert 1998/2009). Therefore, it is not unthinkable that inferentials could indeed constitute a case of extraposition. However, even though extraposition would be expected in inferentials, we might also reasonably expect that their unextraposed counterparts would occasionally occur just as non-extraposed subjects of other constructions do, but so far we have been unable to find any. Additionally, as Collins himself admits (1991: 35), the fact that the unextraposed inferential is utterly ungrammatical poses a problem for the extraposition analysis: (27) (28) (a) It's just I haven't put on any weight. (WSC DPC064) (b) *I haven't put on any weight is just. (a) It's that you have to have them three weeks within three weeks of each another. (WSC DPC008) (b) *That you have them three weeks within three weeks of each other is. More tellingly, as if, as though, and like can occur as complementisers to inferential clauses, but as Huddleston and Pullum (2002: 962) point out, clauses introduced by these complementisers cannot occur in subject position: (29) That/*as if/*as though/*like oil and water don’t mix is well known. We conclude therefore that the matrix copula of inferentials is like impersonal verbs such as seem and appear which require a complement clause but not a semantic subject and that the rightward placement of the complement clause is not due to extraposition. Inferentials in spoken English 323 4.1.2 The cleft analysis Because prior research on inferentials has disagreed on whether or not inferentials are a subtype of it-cleft (Collins 1991; Declerck 1992; Delahunty 2001; Huddleston and Pullum 2002: 1418; Lambrecht 2001), we address the issue here using as our starting point Lambrecht’s definition of cleft constructions: A CLEFT CONSTRUCTION (CC) is a complex sentence structure consisting of a matrix clause headed by a copula and a relative or relative-like clause whose relativized argument is coindexed with the predicative argument of the copula. Taken together, the matrix and the relative express a logically simple proposition, which can also be expressed in the form of a single clause without a change in truth conditions. (Lambrecht 2001: 467) Lambrecht's definition, like most approaches to clefting, privileges the possibility of being able to re-arrange the focus (predicative argument) and relative-like clause of an it-cleft as a unified sentence. Horn (1989) characterizes inferentials as “sentential-focus clefts,” thus including S amongst the range of phrases that may appear as the focus of an it-cleft, for example: (30) It’s not that the ''intelligent design'' nonsense keeps morphing that’s so frustrating; it’s that so many people believe its misinterpretations and misrepresentations. Nonetheless, because inferentials with modifying clauses are extremely rare in actual discourse, as shown by the data from the New Zealand English corpus, as well as by that provided by Koops (2007) from American English, any analysis of inferentials should attempt to determine whether they are indeed a type of it-cleft. Our position is that they are. This issue is dealt with in some detail in Delahunty (2001: 519-20) and we merely synopsize those arguments here. Delahunty points out that constituent focus and sentential focus clefts share lexico-grammatical properties, viz., the copular matrix clause with expletive or null subject and a focal constituent; that a clause may modify the focus in both; that in those languages that use focus particles, e.g., Akan, the same particle is used for both constituent and sentential focusing; that they have significant discourse parallels (both occur in “tandem” constructions - a negative followed by a positive one); that both may evoke alternatives to the focus; that both allow the same range of modifiers of the focus, most notably just and only; that both matrixes may be negated; that it-clefts and inferentials may be treated as parallel structures in texts; that both may be used for discourse repair; and that both may be understood in the same way: positive versions assert the greater contextual relevance of the denotation of the focus in relation to other pragmatically determined alternatives of the same type; negative versions deny the contextual relevance of the denotation of the focus in relation to pragmatically determined alternatives of the same type. The context in which the focus is (not) relevant may be left implicit or it may be made explicit by the addition of a relative-like expression whose relativized element is of the same syntactic type as that of the focus phrase, as in complete, canonical it-clefts as defined by Lambrecht: that is, both CCs and inferentials allow the omission of the modifying clause, a subtype of 324 Andreea S. Calude and Gerald P. Delahunty constituent-focus cleft referred to as a truncated cleft (Declerck 1988; Hedberg 2000; Huddleston and Pullum 2002), for example: (31) Who ate the last cookie? It wasn't me [that ate the last cookie]. While it is generally the case that the modifying clause of a truncated cleft can (typically) be recovered from co/context, for most inferentials this is not the case, cf. below: (32) WSC DPC 120 MQ: oh careers yeah <,> reminds me i've got to get fit → FN: <laughs > i want to do it eh it's just i've got to get myself motivated ??that is happening/the case <latch> yeah well how long have you got till MQ: oh i can any time after this this month i can reply In other words, omitted modifying clauses may not hinder a cleft analysis of inferentials, but the impossibility of reconstructing one may. We believe that the following account can resolve the matter. We assume that a relevant contribution to a discourse must be a proposition, so for either a CC or inferential to be relevant, it must represent a proposition. CCs may only occur if an open proposition of the appropriate type (i.e., one whose “gap” matches the type of the focal phrase) is readily accessible from context, a condition that is obviously met by a full CC. A truncated CC may only be deployed when the open proposition is readily derivable from co-text or situational context. The entire proposition represented by a CC must, of course, be optimally relevant in its context, which must be compatible with the information structure imposed by the CC. Because inferentials represent propositions, they do not require a search for an open proposition to complete them. But they do, of course, have to be optimally relevant in their contexts in ways that are compatible with their pragmatic properties. We conclude therefore that inferentials are indeed a lexico-syntactic sub-type of “cleft” construction and that a pragmatic analysis of one should, ceteris paribus, be extendable to the other, a position we have just sketched and which we develop below. 4.2. Pragmatic analyses 4.2.1. Are inferentials discourse markers (DMs)? Pusch (2006, Forthcoming) claims that inferentials are discourse markers: that is, that they should be classed among expressions that modulate in various ways the proposition(s) communicated by the utterance they are associated with. On its face, this is a plausible analysis given that an inferential indicates a marked relationship between its clause and its local context. There has been much research on DMs, but because we wish to answer the specific question of whether inferentials may be classed as DMs, we have opted to base our answer on the framework presented in Fraser (2005), but developed over many publications (see Fraser 1990, 1996, 1999). Fraser (2005: 1) characterizes DMs as a subtype of pragmatic marker (PM). These are “free morphemes, discourse-segment initial, [that] signal a specific message, and are classified not syntactically but in terms of their semantic/pragmatic functions,” Inferentials in spoken English 325 and which “do not contribute to the meaning of the proposition per se.” The “specific message” signalled by a DM may be one of Temporality, Elaboration, Contrast, or Inference. While Fraser’s requirement that a DM be a “free morpheme” excludes inferentials, which must be viewed as syntactic structures, though with specific lexical requirements, we can also reject the analysis of inferentials as DMs on the grounds that they do not signal the messages Fraser claims for DMs. We can immediately reject the possibility that inferentials belong to the Temporality category: None of our examples indicate a temporal sequence between a prior discourse segment and the inferential clause. And we can reject the possibility that inferentials belong to the class of Elaborative DMs (EDM), which - Fraser (1999: 947) describes as indicating “a relationship in which the message of S2 parallels and possibly augments or refines the message of S1.” None of our examples fits this rubric and none of Fraser’s examples of Elaboration can be replaced by inferentials, e.g., The picnic is ruined. . . . The beer is warm. Furthermore/??It’s that/as if/like it’s raining (Fraser’s 31a.) Nor can inferentials belong to the class of Contrastive DMs (CDM): none of our inferential examples allow the replacement of the matrix with the “primary” CDM but without change of meaning. Nor can we substitute an inferential matrix for a CDM, e.g., I like you but/??it’s that I can’t go out with you, (Fraser's 30a) and A: Fred is a real gentlemen. B: On the contrary/???It’s that he’s a boor (Fraser’s 12c). Nor can inferentials be analysed as Inferential DMs (IDMs), which signal conclusions. None of Fraser’s examples with the canonical IDM so can be rephrased as inferentials; e.g., John was tired. So/???It’s that he left early (Fraser’s 29b). More tellingly, the DMs that Fraser (2005) lists semantically specify the relation between the clause marked by the DM and its prior context, whereas inferentials leave that to be worked out in context and may leave the specific relationship quite indeterminate, as one would expect from a device that is so semantically vague. (See Delahunty 1995: 354-5, 2001: 535-537 for details of this analysis.) We conclude, therefore, that inferentials cannot be regarded as Discourse Markers, not only because Fraser excludes constructions from those categories, but because their interpretations do not coincide with any of his categories and because the interpretations of inferentials and PMs and DMs are arrived at very differently. (For a critical overview of discourse markers see Blakemore 2002, 2004.) 4.2.2. A pragmatic analysis of inferentials and clefts Lambrecht (2001: 504) says that if we include inferentials in the set of cleft constructions, then “we have to acknowledge the existence of two semantically and pragmatically unrelated categories bearing the same name.” We believe that the syntactic and lexical characteristics shared by inferentials and CCs, as narrowly construed, are sufficient to demonstrate that languages may adopt the same strategies to mark focus, whether that focus is a clause or a sub-clausal constituent. We do not, however, accept Lambrecht’s conclusion that CCs and inferentials require separate pragmatic analyses; rather, we believe that canonical CCs and inferentials may be subsumed under the same general interpretive schema, which we sketched above and which we now develop somewhat more fully. 326 Andreea S. Calude and Gerald P. Delahunty As Delahunty (1995, 2001) demonstrates, the cleft/inferential matrix is expletive, and so adds nothing to the truth conditions of the overall construction, a position consistent with Lambrecht’s. Because the matrix is expletive it imposes a greater processing burden on the interpreter than its unmarked congener. According to Relevance Theory (and various theories of marking, e.g., Levinson 2001), such extra processing indicates that “that there is something unusual, unexpected or particularly significant about” the marked information, making it of special relevance in its context (Blass 1990: 245). In the case of the inferential, because the complement is a propositional level entity, its special relevance is interpreted as (part of) the context in which some target expression is to be interpreted (Delahunty 2001: 534-535). In the case of complete constituent focus it-clefts (i.e., those with both focus constituent and relative-like clause), the focus is interpreted as special relative to other contextually possible fillers for the missing element of the modifying clause; in the case of truncated constituent focus it-clefts, because the complement is not a propositional level entity, the interpreter must create a propositional level interpretation by identifying an open proposition of the right type from the context (i.e., its missing element must match the type of the focus expression), and the focus is interpreted as special relative to other contextually possible fillers for the missing element of the implicit open proposition. Thus, if in our definition of clefts/inferentials, we background the clefting characteristic of complete canonical it-clefts and foreground what clefts and inferentials have in common, we can provide a unified analysis that explains the formal and discoursal similarities and differences between the subtypes. In particular, this analysis renders moot any concerns about differences in foreground/background or focus/presupposition structure between the cleft and the inferential. Specifically, while complete canonical it-clefts distinguish the typical discourse functions of the focus phrase and the modifying clause, with the former foregrounding information and the latter backgrounding it, inferentials, not being divided in this way, simply foreground the information in the clause relative to some contextually determined background, more or less as truncated it-clefts do. 5. Discourse functions of inferential construction types in spoken New Zealand English As mentioned earlier, our corpus data produced various recurring patterns of inferential constructions. In what follows, we discuss these in turn, with the aim of explaining their roles in discourse. 5.1. That-inferentials We assume that inferentials, positive or negative, with that as introducer of the focal clause are the basic type of inferential and that inferentials without a clause introducer are variants of this type. As an illustration, consider LL’s negative that-inferential in (17), partially repeated here as (33): Inferentials in spoken English 327 (33) WSC DPC007 RR: → LL: you so you don't want to have to look over your shoulder to make sure you don't oh it's not that i don't want to have to look over my shoulder NOTHING should HAPPEN should OCCUR during those procedures ANY PART of it that's ALL formal the WHOLE lot right from the time the people come onto the marae (Maori: "meeting house") until the time everyone's LEFT the marae all right RR’s comment that LL would not want to look over his/her shoulder is prologue to what appears to be a result clause, to make sure . . ., and so functions as a premise. LL recasts this proposition as a negative inferential which denies its relevance, this time apparently as a conclusion, which should follow from the remainder of his/her utterance: s/he won’t want to look over her shoulder because “NOTHING should HAPPEN,” etc. Because the basic inferential can be interpreted without resorting to any construction-specific semantic or pragmatic stipulations, we assume that the other inferential types may also be so interpreted, and that their interpretational differences from the basic type are due entirely to their lexico-grammatical differences from the basic type. In the following sections, we demonstrate that this assumption is wellfounded. 5.2. Just-inferentials We start by first turning to the dominant (that is, most frequent) type, namely the justinferential, which we take to be an elliptical version of that-inferentials modified by the addition of just. Consider example (34). The speakers are talking about their business, which involves selling cosmetic kits, and are discussing the possibility of increasing their order numbers. They run the business from home, and its administration appears to be done from their bedrooms. (34) WSC DPC293 MK: okay <next utterance directed to person with tape recorder> okay just pause it <,,> can you handle like two kits → FY: oh forgot about that i suppose we could it's just i ca i haven't seen the books so thanks to your sister's fantastic way of cleaning her room MK: because SS: <laughs> MK: because er <latch> FY: suppose we could MK: yeah because i think you know with the er <,,> with <drawls> er kate FY: kate and carmen and see the thing is there's n if carmen and mike can pay for it this week Speaker FY’s response “I suppose we could” (underlined in 34) is a dispreferred response to MK’s question/request “can you handle like two kits.” This becomes clear when FY’s answer is compared with more positive alternatives like “i think we could” or “i know we could” or “we can” or, best of all, “sure.” FY’s actual response does not 328 Andreea S. Calude and Gerald P. Delahunty accede to MK’s request, though it does not deny it either; it is closer to an acceptance of the request than to a rejection of it. It indicates that FY is not sure that they can handle two kits. The inferential presents the proposition represented by “i haven’t seen the books” as a contextual assumption, which in this instance functions as a premise from which it would follow that FY would not know whether they could handle two kits or not. The adverb just may function as a focusing particle meaning merely, exactly, and only. Quirk et al. (1985: 604) claim that just focuses on a particular element in a given clause (in inferentials, the clause following the copula) and “restricts the application of the utterance exclusively (sic) to the part focused.” Aijmer (2002: 158-160) claims that the core function of just as exactly and only is widened to a procedural marker consisting of an “indexical relation to the speaker's attitudes or emotion towards a discourse event” and that just is never semantically neutral, as it always carries evaluative overtones. These overtones are layered through ongoing grammaticalization. Lindemann and Mauranen (2001: 468) characterize just as a “limiting minimizer” which “may be compatible with the connotation that whatever it is limited to is not particularly important.” Consequently, the inferential in (34) indicates that FY’s not having seen the books is the only reason why they might not be able to handle two kits, and it also suggests that this reason is not particularly important, and that MK can reasonably expect a positive response once FY has seen the books. The implication that the expression focused by just is of no great importance underlies the downtoning or softening effect of the inferential in (34) and in other instances, such as (35). (35) WSC DPC240 EL: yeah and the following up is one six one and it ISN'T a reo [Māori word meaning (Māori) language] class <latch> KT: yeah she they don't know how to teach it <,> EL: well i explained to him how we wanted to go to <,,> kuratini in the first year <,> our on our own to do it and he said well can't they see there's the need is there you people wanted you showed them this is what we want and they still haven't KT: they're too lazy and they're too thick to <,> i don't know what it is eh i reckon → it's just that they they feel threatened they're in their nice little cushy position and they don't want it being threatened by anybody else who might be able to do a better job than them cos they know they're not doing as good a job as they should be <,,> they don't know how to teach it <,> EL: well i explained to him how we wanted to go to <,,> kuratini in the first year <,> Speaker KT begins his/her turn by proposing that “they” (referring to the teachers of Māori language classes) are “too lazy and they’re too thick [stupid]” as an explanation for their behaviour described in earlier turns. KT then begins to retreat from this very negative characterization - she pauses, then says “i don’t know what it is,” presumably indicating that she does not know what accounts for their behaviour, then utters “eh,” a commonly used marker of solidarity in New Zealand English, then says “i reckon,” which weakens the force of its complement, which in this case is the just-inferential. The proposition represented by the clause of this inferential functions as a premise from which would follow the behaviour described earlier (particularly in the text underlined Inferentials in spoken English 329 in 35). Here too, just functions as a marker of exclusivity, and is further interpreted as a downtoner. We see the same pair of effects - the clause interpreted as a contextual assumption whose importance is played down by just - in (36). Here, two girls are discussing their student exchange plans. GW has arranged to travel to Russia and Germany, while AN is struggling to decide between going to Japan or to France. GW suggests that AN also attempt to go to both places, but AN’s inferential rejects this suggestion. This inferential also functions as a premise from which it would follow that GW will not spend a year in France and another in Japan. As in the earlier examples, just functions as a downtoner, and so the just-inferential is an indirect and therefore polite rejection of AN’s suggestion. (36) WSC DPC266 GW: and then i think i'd actually be able to <laughs> speak japanese which would be really cool <laughs> AN: i would i would do the japanese one <latch> GW: yeah i know that's what everyone says <laughs> <latch> AN: france would be more fun GW: HEAPS more fun <latch> AN: that would be GREAT GW: yeah <,> but AN: i don't know <,> i mean like japanese is more important for our country GW: yeah <,,> AN: well can't you do one year of each like what i'm doing i mean german's pretty irrelevant really to <laughs> new zealand but i LIKE it so i'm going for a year → GW: oh <laughs> yeah i know yeah yeah i know it's just russia and and germany are a wee bit closer than france and <laughs> japan but do you know the new di the new um director of the alliance is um a japanese <,> speak i mean he speaks fluent <laughs> japanese <latch> AN: really <latch> We note that because the clause of an inferential is interpreted as special in its context, it can be used to counter contextually possible assumptions or interpretations, and that these may arise from the prior discourse created either by the speaker of the inferential, as in (34) (the segment which may be understood as a trigger for the inferential is underlined), or by other participants, as in (35) and (36). 5.3. Negative inferentials Negative inferentials work as one would predict from their linguistic characteristics the inferential form triggers the interpretation that the focused clause is to be interpreted as a contextually special assumption whose relevance is denied by the speaker. For example, in (37), speaker KT describes the Māori classes she chose and the reasoning for her choices. Her first clause complex potentially implicates that she did not remain in the bilingual class beyond her first year, an interpretation consistent with her inferential. This inferential also functions as a premise, which moots and rejects as an explanation for her not continuing in the bilingual class the proposition that she wanted to leave the bilingual class and go to the main stream. This analysis is supported 330 Andreea S. Calude and Gerald P. Delahunty by the clauses following the inferential (underlined), which by the use of “because,” explicitly expresses her actual reason for leaving the bilingual class. (37) WSC DPC240 KT: you know like i was in the bilingual class in my first year you know cos she was she we were just having a chat and um she said have you got a piece of maori in you are you part maori and i said yeah my dad's just under half and um <,> she said oh yeah you know there was a few teachers that were wondering about that some of the parents and stuff and <,> and i said yeah i → was in the bilingual class in my first year it wasn't actually that i wanted to leave the bilingual class to go to main stream in my second year it was because i wanted to take a an advanced maori paper at <,> the varsity because i wanted to you know nurture <drawls> my language Negative inferentials, like just-inferentials, can be prompted by any participant in the exchange, but crucially, unlike the just-inferential, there is no softening or toning down of the force of the rejection. Finally, negative inferentials may be followed by a positive inferential, in a pattern Delahunty (1995, 2001) refers to as a “tandem” inferential. The negative inferential rejects the relevance of the proposition represented by its clause; the positive inferential, in contrast, asserts the relevance of its clause. The inferentials in (41) (below in section 5.4) are an example of this pattern. 5.4. (Not) as if-inferentials and (not) like-inferentials Our default hypothesis is that the discourse effects of inferentials are a function of their lexico-syntactic form and their semantics interacting with general pragmatic principles and local context. That is, unless forced by the data, we avoid stipulating any discourse functions specific to the construction or its variants. It follows that our hypothesis regarding the interpretation of like and not like, as if and not as if inferentials is that their interpretation is the same as that of that inferentials except in so far as (not) like/as if differs from that. We begin with a discussion of (not) as if inferentials and then deal with the (not) like variant. There is relatively little research on either of these variants of the inferential construction, though Huddleston and Pullum (2002) includes some suggestive remarks. According to Huddleston and Pullum (2002: 1146) as denotes comparison, and if “is primarily conditional,” and thus has a “close relation” with though, “which is primarily concessive” (2002: 737). They claim that as if may function as a “single compound preposition,” which, we believe, denotes a sense of hypothetical comparison (2002: 1151). This sense may be quite “attenuated” (2002: 1151) in certain contexts, and consequently, in some instances, as if may be replaced by that or its zero alternant without change of meaning, and so may be interpreted as merely suggesting the truth of a proposition rather than (strongly) asserting it. This is especially the case after appear, feel, seem, sound, and be, which may induce a “medium strength epistemic modality.” Biber et al. (1999: 840-841) claim that with non-finite clauses, as if and as though indicate that the “adverbial clause is showing similarity but is not to be taken factually.” This is consistent with the analysis we developed just above. Inferentials in spoken English 331 Given the very limited number of examples we’ve found, it is impossible to determine whether as if inferentials display the range of interpretations predicted by our analysis. The Wellington Corpus of Spoken New Zealand English data includes only two instances of as if inferential variants, both negative. In order to provide as complete an account as possible we have borrowed a written as if inferential from another corpus of New Zealand English and begin our discussion of (not) as if inferentials with it: (38) WWC SECTION F, F42 186-194 The Sunday News used to be the main proponent of the idea of celebrity in New Zealand. It was in that tabloid beloved of life's losers that we first read about Graeme → Thorne's perm and much other such trivia. It was as if successive editors had a list of so-called personalities from which they never really deviated. It is probably still pasted up in the news-room, slowly yellowing under the harsh fluorescent lights. My guess is that it includes the old names Ray, Bob, Max, Marilyn, the other Ray and Howard. You should know the surnames. They've been around for years. This is a particularly interesting example of an inferential. It can be read in at least two ways. First, one could read it as the author’s positing a hypothetical conclusion to be drawn from the second sentence of the extract, or perhaps from an inference derivable from that sentence. We suggest something along the following lines: celebrity in the Sunday News sense seems to be nothing more than repeated mention of an individual in the tabloid along with such trivia as his/her latest hair-do. One might infer that to create such repeated mentions, the editors acted as if they merely drew names from a list of these “so-called personalities from which they never really deviated.” Alternatively, one could read this inferential as a hypothetical premise from which the celebrity of Graeme et al. would follow, viz., the editors acted as if they drew the names of the celebrities from a list from which they never deviated. If our hypothesis is correct then not as if inferentials should be used to deny the relevance of a comparison with a contextually plausible hypothetical proposition in the interpretation of a target utterance. This is consistent with, though more fully articulated than, Huddleston and Pullum’s remark that “It’s not as if he wasn’t trying . . . is used to deny a proposition that might otherwise have been deduced (perhaps he didn’t perform as well as expected)” (2002: 1152, fn. 36). The inferential in (39) shows that this is indeed the case. (39) WSC DPC032 → AW: well there's only there's only five or six in the race it's not as if they're racing going up three wide ra round fields of eighteen they're only going round fields of six they race sort of there'll be one in the front one <long pause> on the e one on th the trail and one on the outer behind the horse on the on the trail not facing the breeze In the conversation from which (39) is taken, two people are discussing horse racing and specifically the differences between real racing and practicing. AW’s description contrasts “going round fields of six [horses]” with “going round in fields of eighteen [horses],” and rejects the conclusion that in the former the horses are racing. This interpretation is supported in the utterance immediately following the inferential when AW characterizes going round fields of six as only “sort of'” racing. 332 Andreea S. Calude and Gerald P. Delahunty The inferential variant we discuss next is the one in which the clause is introduced by (not) like. Here again our hypothesis is that we can account for their interpretations compositionally and without stipulation. Even though our negative occurrences outnumber our positive ones, we begin here also with the latter as it is more basic than the former: (40) WSC DPC326 JI: and she was the legal advisor for er ronnie burch AL: right <latch> JI: you know <drawls> when when he was yeah race relations officer yeah <with creaky voice> mm but um AL: yeah race relations oh <drawls> good so she didn't have a problem getting a job i suppose when her when er <unclear word> → JI: no no but sh she found that particular job very stressful it's like she doesn't think she'd like to go back into it you know cos she was always dealing with problems AL: <drawls> mm right The basic meaning of like is similarity, though it is being grammaticalized as a marker of reported speech and thought, used most frequently by young people in their casual speech (Romaine and Lange 1991). Our analysis is consistent with this. In (40), the proposition represented by the inferential clause is presented as similar to a proposition which is to function as a conclusion which would follow from the proposition represented by the immediately prior clause, “she found that particular job very stressful.”10 This cause and effect relationship is made explicit by the conjunction “cos” that introduces the clause that follows the inferential, “she was always dealing with problems.” Not like inferentials reject the relevance of the proposition represented by the inferential clause as a more or less faithful interpretation of the proposition entertained by the speaker. Thus the negative inferential in (41) rejects the potential characterization of the situation as “they NEED someone.” Because it is presented in inferential form, this proposition functions as an explanation, that is, as a proposition from which “They wouldn’t want Thomas” would follow. This is consistent with the positive inferential that follows the not like one, “It’s just they’re just doing it as a favor because Susannah’s a mate.”11 We interpret this as a premise from which it would follow that the Wilkins would have AN “come over there every week.” (41) WSC DPC059 AC: well missus wilkins said i could do it over there i mean every week but they they wanted me to do it every week BS: yeah pity in some ways isn't it because it's quite good money AC: mm BS: do you think thomas would do it 10 It is worth considering whether uses of like such as that exemplified in (40) are markers of “interpretive use” (Sperber and Wilson 1986/1995: 224-231). That is, the proposition represented by the clause introduced by like is to be interpreted as, to one degree or another, resembling a proposition from which relevant contextual effects would follow. Crucially, the proposition represented by the inferential clause in (40) is not a proposition assumed by the speaker; it merely resembles some such proposition. 11 That these two occur as a tandem pair supports our claim that the types they represent should be regarded as inferential variants. Inferentials in spoken English 333 → AC: BS: AC: BS: AC: BS: AC: they wouldn't they wouldn't want thomas mm it's not like they NEED someone it's just they're just doing it as a favour because susannah's a mate <drawls> yeah <,,> not like they need anyone i've OCCASIONALLY thought that you could actually do some work for kelvin and sharon but i'm not sure but they i should just do it for them for free It is our intuition that in the examples above, (not) like may replace (not) as if, with only a stylistic shift; we find the (not) as if variants to be slightly more formal than the (not) like variants. For example, (38) is from the written portion of the Wellington Corpus of New Zealand English and so we must assume that the author (and perhaps editors) chose their words and constructions carefully when they chose as if instead of like for their inferential. But given that substituting like for as if seems to have no effect on the interpretation of the text, we might assume that the choice merely reflects different stylistic levels, which we explain below. We note here that not like and not as if inferentials, like that in (42), have a contrastive role but lack the down-toning effects we described for just-inferentials. (42) WSC DPC156 SH: they did try but <,,> they decided but you were engaged and they decided because you'd probably be on the phone to either me or a or matthew <laughs> may as well just come round cos it'd be AGES before you got off the phone <laughs> BE: yeah cos it was really freaky cos i was there all by myself i mean yeah → matthew had just rung up and there was this BASH on the door it wasn't like it was just a knock there was a BASH and i thought <laughs> i don't want that to be for me doesn't sound friendly so i went and opened the door and i thought i should put the chain on <laughs> before i open the door Assuming that these substitutions do not in fact alter the text in any way except perhaps to alter the level of style, we must now ask how it is that (not) as if and (not) like have such similar effects. This is how it works: 1. As can clearly be used to indicate comparison - As big as a barn. 2. So can like - Big, like a barn. 3. If indicates conditionality, either realis - If I finish this by noon, I can take the rest of the day off - or irrealis - If I were only three inches taller, I’d be able to reach the top shelf. In both cases, if assumes that the conditional clause is not true of the discourse world at the relevant time, though this assumption may be cancelled - If I committed the crime, and I neither admit nor deny that I did, I wouldn’t tell anyone about it. So an if clause can be entertained as a hypothetical proposition. 4. Consequently, as if compositionally denotes both similarity and conditionality/ hypotheticality. That is, the proposition represented by an as if clause will be entertained as potentially comparable to a contextually relevant assumption. 334 Andreea S. Calude and Gerald P. Delahunty 5. An as if inferential, like inferentials generally, indicates that the proposition represented by its clause is to be interpreted as relevant as a contextual assumption to be considered in the interpretation of a target utterance, 6. Like, whose basic meaning is similarity, is being grammaticalized as a marker of reported speech and thought (Romaine and Lange 1991). The proposition represented by an inferential clause introduced by like is not a proposition assumed by the speaker, but is merely similar to such a proposition. (Not) like inferentials reject (the relevance of) the proposition represented by the clause as a more or less faithful interpretation of the proposition entertained by the speaker. 7. Because this use of like has its origins in casual conversational use by teenagers, it cues a casual frame of reference. As a result we find this use of like in informal, unplanned discourse and in representations of such discourse, for example, in fictional renditions of speech. As if on the other hand cues a higher stylistic level, and so we find it in more formal contexts, such as academic prose. 6. Conclusion We began our discussion of inferentials with a description of the variant forms of the construction that we discovered in the WSC. While these variants had occasionally been noted before, their interpretations and discourse functions had not been described. We hypothesised that their interpretations and functions could be accounted for simply by virtue of their lexico-grammatical properties, general pragmatic principles, and local context, without the need for any construction-specific stipulations. We provided accounts for that inferentials with modifiers such as just in their matrixes, as well as those whose clauses are introduced by (not) as if and (not) like. In each case we were able to show that the variants functioned as we predicted without resorting to any stipulations. Because the discourse functions of the various types of inferentials follow from the interaction of their forms with general pragmatic principles and local context, there is no need to analyze them in the way that discourse markers are analyzed by Fraser (2005: 5), namely as denoting two place predicates which take as their arguments the expression introduced by the marker and its relevant prior context - predicates that “impose a relationship between some aspect of the discourse segment they are a part of, call it S2, and some aspect of a prior discourse segment, call it S1.” This follows from their linguistic properties and general principles. While we were successful in accounting for the contextualized interpretations of the various inferential types, we were unable to arrive at any robust conclusions that would allow us to account for the statistically significant difference in occurrence of certain inferential types in the New Zealand corpus and in Koops’ corpus of spoken US English. We believe that this topic is worth fuller investigation, though this would require considerably larger corpora, ideally structured so as to allow cross-register comparisons in several dialects of English. Inferentials in spoken English 335 References Aijmer, K. (2002) English discourse particles: Evidence from a corpus. Studies in corpus linguistics 10. Amsterdam: John Benjamins Publishing Company. Austin, J.L. (1962) How to do things with words. Oxford: Clarendon. Barlow, M. (2010) How to distinguish individual speakers: A corpus-based investigation of idiolects. Manuscript. Bearth, T. (1997) Inferential and counter-inferential grammatical markers in Swahili dialogue. In E.M. Beck, T. Geider, W. Graebner, and I. Heine (eds.), Swahili forum. Cologne: Universität zu Köln. Bearth, T. (1999) The inferential gap condition. Pragmatics 9: 249-288. Bender, E., and D. Flickinger (1999) Diachronic evidence for extended argument structure. In G. Bouma, E.W. Hinrichs, G.M. Kruijff, and R. Oehrle (eds.), Constraints and resources in natural language syntax and semantics. G. Stanford, CA: CSLI, pp. 1-19. Biber, D., S. Johansson, G. Leech, S. Conrad, and E. Finegan (1999) Longman grammar of spoken and written English. Essex: Longman. Blakemore, D. (2002) Relevance and linguistic meaning: The semantics and pragmatics of discourse markers. Cambridge, UK: Cambridge University Press. Blakemore, D. (2004) Discourse markers. In L.R. Horn and G. Ward (eds.), The handbook of pragmatics. Malden, MA: Blackwell, pp. 221-240. Blass, R. (1990) Relevance relations in discourse: A study with special reference to Sissala. Cambridge Studies in Linguistics 55, Cambridge, UK: Cambridge University Press. Calude, A. (2009a) Cleft constructions in spoken English. Berlin: VDM Verlag. Calude, A. (2009b) Formulaic tendencies of demonstrative clefts in spoken English. In R. Corrigan, E.A. Moravcsik, H. Quali, and K.M. Wheatley (eds.), Formulaic language: Volume 1. Distribution and historical cxhange. Philadelphia: John Benjamins Publishing Company, pp. 55-76. Calude, A., and G. Delahunty (2010) Inferentials: Fixed or not? Paper presented at The international conference on fixed phrases in English. October 22-24, University of Perpignan, Via Domitia. Calude, A., and S. Miller (2009) Are clefts contagious in conversation? English Language and Linguistics. 13: 127-132. Collins, P. (1991) Cleft and pseudo-cleft constructions in English. London: Routledge. Dancygier, B., and E. Sweetser (2005) Mental spaces in grammar: Conditional constructions. Cambridge: Cambridge University Press. Declerck, R. (1988) Studies on copular sentences, clefts and pseudo-clefts. Foris: Leuven. Declerck, R. (1992) The inferential it is that-construction and its congeners. Lingua 87: 203-230. Delahunty, G. (1990) Inferentials: The story of a forgotten evidential. Kansas Working Papers in Linguistics 15: 1-28. 336 Andreea S. Calude and Gerald P. Delahunty Delahunty, G. (1995) The inferential construction. Pragmatics 5: 341-364. Delahunty, G. (2001) Discourse functions of inferential sentences. Linguistics 39: 517-545. Delahunty, G. (2006) A relevance theoretic analysis of not that sentences: “Not that there is anything wrong with that.”Pragmatics 16: 213-245. Delahunty, G., and L. Gatzkiewicz (2000) On the Spanish inferential construction ser que. Pragmatics 10: 301-322. Dirven, R. (1989) A cognitive perspective on complementation. In D. Jaspers, Y. Putseys, W. Klooster and P. Seuren (eds.), Sentential complementation and the lexicon: Studies in honour of Wim de Geest. Dordrecht: Foris, pp. 113-139. Doherty, M. (2001) Discourse functions and language-specific conditions for the use of cleft{-like} sentences: A prelude. Linguistics 39.3: 457-362. Edmonds, A. (2010) On the representation of conventional expressions in L1-English and L2-French. Ph.D. dissertation, Departments of French and Italian and Linguistics, Indiana University. Fraser, B. (1990) An approach to discourse markers. Journal of Pragmatics 14: 383-395. Fraser, B. (1996) Pragmatic markers. http://people.bu.edu/bfraser/ Fraser, B. (1999) What are discourse markers? Journal of Pragmatics 31: 931-952. Fraser, B. (2005) Towards a theory of discourse markers. http://people.bu.edu/bfraser/ García, F.G. (2007) That’s a construction for you/las construccionnes es lo que tiene(n): Grammaticalization via subjectification in attributive clauses in English and Spanish. Journal of English Studies 7: 65-99. Gundel, J.K., N. Hedberg, and R. Zacharski (1993) Cognitive status and the form of referring expressions in discourse. Language 69: 274-307. Goldberg, A. (2006) Constructions at work: The nature of generalization in language. Oxford: Oxford University Press. Halliday, M.A.K. (1987) Spoken and written modes of meaning. In R. Horowitz and S.J. Samuels (eds.), Comprehending oral and written language. San Diego: Academic Press, pp. 55-82. Hedberg, N.A. (2000) The referential status of clefts. Language 76: 891–920. Heggie, L.A. (1998) The syntax of copular structures. Doctoral dissertation, University of Southern California. Holmes, J., B. Vine, and B.G. Johnson (1998) Guide to the Wellington corpus of spoken New Zealand English. Wellington, New Zealand: School of Linguistics and Applied Language Studies: Victoria University of Wellington. Hopper, P., and S.A. Thompson (2008) Projectability and clause combining in interaction. In Laury Ritva (ed.), Crosslinguistic studies of clause combining: The multifunctionality of conjunctions. Amsterdam: John Benjamins Publishing Company, pp. 99-123. Horn, L. (1989) A natural history of negation. Chicago: Chicago University Press. Huddleston, R. (1984) Introduction to the grammar of English. Cambridge, UK: Cambridge University Press. Inferentials in spoken English 337 Huddleston, R., and G. Pullum (2002) The Cambridge grammar of the English language. New York: Cambridge University Press. Hunston, S. (2006) Phraseology and system: A contribution to the debate. In G. Thompson and S. Hunston (eds.), System and corpus. 55-80. London: Equinox. Koops, C. (2007) Constraints on inferential constructions. In G. Radden, K.M. Kopcke, T. Berg, and P. Siemund (eds.), Aspects of meaning construction. Amsterdam: John Benjamins Publishing Company, pp. 207-224. Laury, R. (2006) On subordination, Finnish-style: Questioning the category of finite clausal complements in spoken Finnish. SKY Journal of Linguistics 19: 310–321. Lambrecht, K. (2001) A framework for the analysis of cleft constructions. Linguistics 39: 463–516. Langacker, R. (1974) Movement rules in a functional perspective. Language 50: 630–664. Levinson, S. (2001) Presumptive meanings: The theory of generalized conversational implicature. Cambridge, Massachusetts: MIT Press. Lindemann, S., and A. Mauranen (2001) “It’s just real messy”: The occurrence and function of just in a corpus of academic speech. English for Special Purposes 20: 459-475. López-Couso, M. and B. Méndez-Naya (in press) On the use of as if, as though, and like in present-day English complementation structures. Journal of English Studies. McCawley, J.D. (1988) The syntactic phenomena of English. Chicago and London: The University of Chicago Press. Miller, J., and R. Weinert (1998/2009) Spontaneous spoken language. Oxford: Oxford University Press. Pawley, A., and F.H. Syder (1983) Two puzzles for linguistic theory: Nativelike selection and nativelike fluency. In Jack C. Richards and Richard W. Schmidt (eds.), Language and Communication. London: Longman, pp. 191-225. Pusch, C. (2006) Marqueurs discursifs et subordination syntaxique: La construction inférentielle en français et dans d'autres langues romanes. In M. Drescher and B. Frank-Job (eds.), Les marqueurs discursifs dans les langues Romanes: Approches théoriques et méthodologiques. Frankfurt: Peter Lang, pp. 173-188. Pusch, C. (forthcoming) Pragmatic markers involving subordination in Romance: Do they structure discourse or comment on it? Quirk, R., S. Greenbaum, G. Leech, and J. Svartvik (1985) A comprehensive grammar of the English language. London, New York: Longman. Romaine, S., and D. Lange (1991) The use of like as a marker of reported speech and thought: A case of grammaticalization in progress. American Speech 66: 227-279. Rooryck, J. (2000) Configurations of sentential complementation: Perspectives from Romance languages. London: Routledge. Sag, Ivan A. (2010) Sign-based construction grammar: An informal synopsis. In Hans C. Boas and Ivan A. Sag (eds.), Sign-based construction grammar. Stanford, CA: Center for the Study of Language and Information, pp. 39-160. http://lingo.stanford.edu/sag/papers/theo-syno.pdf 338 Andreea S. Calude and Gerald P. Delahunty Schmid, H-J. (2009) Rare but contextually entrenched: The English not-that construction. Paper presented at the International Conference on the Linguistics of Contemporary English, University of London, UK. Sperber, D., and D. Wilson (1986/1995) Relevance: Communication and cognition. Cornwall: Blackwell Publishing. Stirling, L. (1999) Isolated if-clauses in Australian English. In P. Collins and D. Lee (eds.), The clause in English. Philadelphia: John Benjamins Publishing Company, pp. 273-294. Taylor, J.R., and K.Y. Pang (2008) Seeing as though. English Language and Linguistics 12: 103-139. Thompson, S.A. (2002) “Object complements” and conversation: Towards a realistic account. Studies in Language 26: 125-164. Vallaurí, E.L. (2004) Grammaticalization of syntactic incompleteness: Free conditionals in Italian and other languages. SKY Journal of Linguistics 17: 189–215. Wray, A. (2002) Formulaic language and the lexicon. Cambridge: Cambridge University Press. Wray, A. (2008) Formulaic language: Pushing the boundaries. Oxford: Oxford University Press. Zemskaja, E. (1973) Russkaj razgovornaja reč’. Moscow: Nauka. Appendix A. WSC annotations <drawls> <latch> <laughs> <,,> <,> <reads> ... </reads> <quickly> <quietly> <softly> <unclear word> <with creaky voice> speaker drawls overlapping speech speaker laughs longer than 2 second pause 1 second pause portion given between tags was read by the speaker speech portion is uttered quickly speech portion is uttered in a quiet voice speech portion is uttered softly speech is inaudible or incomprehensible to transcriber speech portion is uttered with a creaky voice Appendix B. Inferentials found in the WSC data, 250,000 words across 116 conversations 1. it's just i've got to get myself motivated <latch> 2. well it's just that <latch> 3. yeah <,> it's just i'm getting all the heavies <,> 4. <,> it's just that they don't know what happened 5. yeah oh it's just that i don't think that jim does it on purpose 6. it wasn't like it was just a knock 7. it's just you don't know what they're talking about 8. it's just that she don't listen eh 9. it's just that if you've been given the money you should be Inferentials in spoken English 339 10. it's just that hawaii's more commercialised not that bali 11. it's just like people go there you know they <clap> on the beach... 12. it's just that that impression has been given politically 13. it's like that's only where food goes you know 14. i reckon it's just that they they feel threatened 15. it wasn't actually that i wanted to leave the bilingual class to go to main stream in my second year. 16. it's just you know <,> it was like <leave me alone>... 17. yeah yeah i know it's just russia and and germany are a wee bit... 18. it's just he doesn't trust you 19. it's just you know you want to find the right person 20. it's just that we know a lot of people with the same 21. it's just that well [incomplete] 22. it's just i ca i haven't seen the books so 23. it's just that <,> they've never stood on the marae atea and done it 24. it's just they don't have to worry about board and those sorts of things 25. mm <,> i'm not doing too bad with my fundraising it's just that it's just that i keep dipping into the my big one 26. it's like well kataraina didn't actually make 27. oh it's not that i don't want to have to look over my shoulder 28. it's that you've got to have them <,> three weeks apart 29. it's not like the really has got much semantic weight to it 30. it's just that it makes the spaghetti bigger 31. it's just that you know [incomplete] 32. it's just that i don't get those <,,> 33. yeah it's just i'm interested in case you get caught in an avalanche 34. it's not like they NEED someone 35. it's just they're just doing it as a favour because susannah's a mate 36. so i mean it's just she needs to be on a job like that <laughs> where she doesn't have to do anything 37. it's not like she doesn't work very i mean she doesn't ... 38. it's just i i haven't put on any weight <latch> 39. it's just <,> i haven't been playing soccer you see 40. it's just understanding the language 41. it's just that she likes that little spot because it's very warm 42. it's um that i mean public transport is always regarded as something which was expected to make a LOSS because it was the responsibility of the local 43. yeah i think it's those families have so much character and so much fun 44. it's just that i can't [incomplete] 45. it's just there's nowhere for anyone to sit 46. so it wasn't like robbie was saying it because he'd heard me talk about it 47. it's just that i yeah you've got your slab poured and you've put some boxing there for your wall 48. it's just that i kind of think <,,> um <latch> 49. it may be that HE needs to say that she can't live at home on her own any more 50. yeah it was just that [incomplete] 51. yeah yeah well to me it was just oh he was just pissing around 340 Andreea S. Calude and Gerald P. Delahunty 52. it's just the weather was just like this 53. it's not as if she has to stay on and STAY in kyle's room 54. it's not as if they're racing going up three wide round fields of eighteen 55. it's like she doesn't think she'd like to go back into it GERALD DELAHUNTY is Assistant Chair and Associate Professor of Linguistics and English in the English Department at Colorado State University, Fort Collins. He has also taught at the University of California, Irvine, and at San Diego State University. He received his Ph.D. in Social Sciences from the University of California, Irvine, and has published on syntactic theory, English syntax, the pragmatics and discourse functions of several non-canonical structures, sociolinguistics, and Irish archaeology and landscapes. With the late James Garvey, he is the author of “Language Grammar and Communication” (McGraw Hill 1994) and “The English Language: From Sound to Sense” (Parlor Press/WAC Clearing House 2009). Address: Department of English, 359 Eddy Hall, Colorado State University, Fort Collins, CO 805231773, United States. E-mail: gerald.delahunty@colostate.edu ANDREEA CALUDE is a Research Fellow at the University of Reading, UK, funded by the New Zealand Foundation for Research Science and Technology. She received her Ph.D. in cognitive linguistics from the University of Auckland, New Zealand, and published her thesis as a book entitled “Cleft Constructions in Spoken Language” (VDM Verlag, 2009). She has published in areas including machine translation, morphology, cognitive grammar, the middle voice, formulaic language and the philosophy of mathematics. Address: School of Biological Sciences, University of Reading, Lyle Building, Level 4, Reading Berkshire, RG6 6AJ, United Kingdom. E-mail: a.s.calude@reading.ac.uk