The vagaries of subject it: can it serve as a style marker?

The vagaries of subject it: can it serve as a style marker? Helena Fišerová (Prague) ABSTRACT This article analyses the distribution of the different functions of the subject it in two text types: academic prose and fiction. Its objective is to discover to what extent and in which aspects the two examined text types differ. The analysis is based on two hundred examples of it in the subject function obtained from the British National Corpus (BNC), one half of them drawn from academic prose, the other from fiction. The functions of it analysed include anaphoric it, anticipatory it, empty it and it as the subject in a cleft sentence (focusing it). Additionally, the instances of it that did not fit into any of the four main categories are also examined. KEY WORDS anaphoric it, anticipatory it, empty it, cleft sentence, extraposition 1. INTRODUCTION This paper deals with the analysis of the representation of the different functions of the subject it in two text types: academic prose and fiction. It examines four basic functions of subject it — anaphoric it, empty it, and it as the component of extraposition (anticipatory it) and the cleft sentence (focusing it). Focusing not only on the distribution and frequency of the different functions but also on their correlations with other features of the two registers, it aims to discover in which respects and to what extent the two text types differ and whether such a basic, common word as it can play a role as a style marker. 2. THEORETICAL BACKGROUND The pronoun it has the same basic functions as the other third person pronouns, serving deictic and anaphoric uses. Aside from those, it can also have several “special uses” whose exact number varies in different interpretations. The Longman Grammar identifies the following three: — Empty subject/object — Anticipatory subject/object1 — Subject in cleft constructions (Biber et al., 1999: 332). 1 Given the topic of this paper, the discussion of the special uses of it will only include subject uses. 36 LINGUISTICA PRAGENSIA 1/2016 a. Anaphoric it In the anaphoric function, it generally refers to a noun phrase denoting an inanimate entity or an animal, a clause or an entire sentence (Dušková et al., 2006: 392). Most commonly, the meaning of it is identified via an antecedent mentioned previously, but anticipatory anaphora (cataphora) is also possible (Huddleston and Pullum, 2002: 1468). b. Empty it A semantically empty it (also called prop it, dummy it, weather it, ambient it etc.) occurs mainly in expressions of weather conditions, time and distance, where there is no real participant to fill the subject slot required by English sentence structure (Biber et al., 1999: 125): (1) It was not as cold as on the previous night. (2) By the time you get back it’s nine o’clock. (3) It was seven miles to the nearest town and I had to bus or walk everywhere. Traditionally, it in such constructions is considered to be entirely devoid of meaning (Seppänen, 2002: 444). An alternative view is taken by Bolinger, who believes that it is always to some extent referential, and in the above examples it would have some very general referent such as “environment” (1977: 78). In a similar approach, Kaltenböck describes the referential capacity of it as a gradient ranging from “wide reference” of “prop it” to “narrow reference” of “referring it” (1999: 62). Given the types of construction where empty it generally appears, it can be expected that its frequency in academic prose will be significantly lower than in fiction. c. Anticipatory it Anticipatory it functions as the grammatical subject in instances of extraposition, where it anticipates the notional subject, which has the form of a finite or non-finite clause. (4) It was hard to believe that he had become this savage with the bare knife. (5) It really hurts me to be going away. (Biber et al., 1999: 155) There is some disagreement about the level of “referentiality” or “dumminess” of the anticipatory it. Different approaches place it closer to empty it, closer to referential it or establish it as a separate category (Kaltenböck, 1999: 51). An inconsistency in interpretation can be found even within one grammar: Quirk et al. first state that anticipatory it plays “essentially a structural [role]” and “carries virtually no information of its own”, its only function being to fill the obligatory subject slot (1985: 89). They find that it has even less independent meaning than empty it in time or weather expressions, but at the same time, they admit that it cannot be said to be “quite void of meaning, since it arguably has cataphoric reference” to the extraposed clause (ibid.: 349). However, the main difference between anticipatory and truly cataphoric it is that cataphoric it can be turned into anaphoric it, i.e. both the pronoun and its referent have to be present regardless of their sequence: HELENA FIŠEROVÁ 37 (6) When you least expect it, a mistake invariably occurs. (7) A mistake invariably occurs when you least expect it. (Smolka, 2007: 56) Whereas when the subject clause is moved from the extraposed position to the initial one, it becomes superfluous: (8) It is clear that something must be done immediately. (9) *That something must be done immediately is clear it. (ibid.: 57). Based on the findings of Biber et al., it can be expected that extraposition will be more frequent in academic prose than in fiction (1999: 674). d. Focusing it It can function as the subject in a cleft sentence, which serves as a means of focusing on a particular element. In a cleft sentence, one propositional content is divided into two clauses: (10a) I bought a red wool sweater. (10b) It was a red wool sweater that I bought. (Huddleston and Pullum, 2002: 1415) The cleft sentence thus differs from a simpler, non-cleft sentence not by the content but by the FSP structure — it highlights the rheme of the utterance (Dušková et al., 2006: 537). Structurally, the it-cleft (or cleft sentence proper) consists of the pronoun it, a form of the verb be which can be negated or modified by a focusing adverb such as just or only, the focused element,2 and a dependent clause similar to a relative clause, introduced by that, who, which or zero (Biber et al., 1999: 959). The forms whom and which are infrequent, and if they appear after a preposition, only a non-cleft interpretation of the sentence is possible (Dušková et al., 2006: 625). The status of the pronoun it in a cleft sentence is a matter of some debate (Smolka, 2007: 31; Dušková, 1993: 74, note 8). In structurally similar non-cleft sentences where the subordinate clause is a relative clause, it is anaphoric and refers to an antecedent mentioned previously. In that case, it can be replaced by this, that, and the noun phrase it refers to, but this is not possible in the case of an it-cleft3 (Dušková, 1993: 74). In this respect the it in a cleft sentence is more similar to anticipatory it in sentences with an extraposed subject (Quirk et al., for example, group them together (1985: 349)), as it has no semantic content of its own and seems to point forward to the subordinate clause. While the cleft sentence bears some resemblance to extraposition, it here 2 According to Collins, the focused element is only an optional component. He considers examples such as It’s not that Mervyn’s totally unreliable as clefts, although he admits that convincing evidence for this interpretation is lacking (1991: 34–35). 3 There is, however, a minor subtype of the cleft sentence that uses demonstrative pronouns instead of it, e.g. Those are my biscuits you’re eating. The demonstrative is used deictically in this case (Huddleston and Pullum, 2002: 1420). 38 LINGUISTICA PRAGENSIA 1/2016 cannot be considered truly anticipatory because it cannot be replaced by the thatclause, which is possible for anticipatory it (Smolka, 2007: 31). Based on the findings of Biber et al., it can be expected that focusing it will be more frequent in academic prose (1999: 962). 3. MATERIAL AND METHOD The analysis presented in this article is based on 200 excerpts obtained from the British National Corpus (BNC), a corpus of 100 million words of British English from the latter part of the 20th century, accessed using the web-based interface BNCweb (CQP-Edition) available on-line from http://bncweb.lancs.ac.uk/. Two subcorpora of similar sizes were created from this corpus, one of academic prose and one of fiction. The same search for it in the subject position was then carried out in both subcorpora, using the query “it (_{ADV})? _V*[D,Z]”. The result of this query should be it followed by an optional adverb and any verb in either past tense form or third person singular present tense form. This query returned 79,597 hits in the Academic prose corpus and 111,307 hits in the Fiction corpus. The results were then thinned to 120 random hits from each corpus — given that not every it followed by a verb must necessarily be a subject, the sample size was a little higher than the target number. The results were sorted manually and all hits that did not meet the criteria were discarded. (The discarded hits were those in which it was not the subject or those where it occurred in a quotation from a different text type.) The first 100 valid hits from each subcorpus constitute the research sample. The obtained excerpts were then divided into categories and further analysed. In the analysis of anaphoric it, attention was paid to the realisation form and semantic role of the referent, as well as the location of the referent in relation to the pronoun. The analysis of anticipatory it focused on the type of subordinate clause realising the extraposed subject and the realisation form and semantics of the predicate in the superordinate clause. For focusing it, the analysis dealt mainly with the realisation form and underlying syntactic function of the element highlighted by the cleft sentence. Empty it, as well as uses of it that do not fit neatly into any of the four categories, were also examined. The results of the analysis are briefly presented below. 4. RESULTS4 The research clearly shows that there are indeed differences in the distribution of the functions of subject it and their uses in the two text types. Academic prose and fiction differ not only in the frequency of the functions, but also in the ways the functions are used. 4 In the text below, if the referent of anaphoric it is not included in the text of the excerpt, it is provided in square brackets after the pronoun. The context of the excerpt is provided in square brackets if it is needed for the discussion. The letters A or F at the end of the excerpt indicate whether it comes from academic prose (A) or fiction (F). HELENA FIŠEROVÁ 39 First, it is important to mention the most general difference: the total number of instances of subject it in each text type. The query used to look for subject it in the two subcorpora found 79,597 instances in the Academic prose corpus and 111,307 instances in the Fiction corpus. When related to the sizes of the subcorpora, we get the relative frequency of 5044.8 instances per million words in academic prose, and 6942.09 instances per million words in fiction. It in the subject position is, therefore, overall more frequent in fiction than it is in academic prose. Text type Subcorpus size (words) Instances of subject it Instances of subject it per million words Academic prose 15,778,028 79,597 5044.8 Fiction 16,033,634 111,307 6942.1 table 1: Frequency of subject it in academic prose and fiction The numbers presented in Table 1 are of course only approximate, since the query used looked for instances of it followed by an optional adverb, followed by a verb, which does not necessarily have to be a subject, and there can be instances of subject it occurring in different patterns, but this is sufficient for a general idea about the distribution. There are at least two reasons why the subject it is less frequent in academic prose. The first is the lower variety of its functions — as will be seen below, empty it alone makes up almost one quarter of our sample of fiction, but does not appear in academic prose at all. The second reason is that academic prose aims at precision, and it is therefore likely to avoid using a pronoun where its reference might be misinterpreted by the reader, and opts for a nominal referring expression instead. It generally uses personal pronouns more rarely than other registers (Biber et al., 1999: 333). Type of it Academic prose Fiction anaphoric 57 51 empty 0 24 anticipatory 37 11 focusing 6 4 other 0 10 total 100 100 table 2: Functions of it in academic prose and fiction Looking at the frequency of the different types of it in the two registers, summarized in Table 2 and Figure 1 above, we find that anaphoric it is slightly more frequent in academic prose and anticipatory it is significantly so. As expected, focusing it is also more common in academic prose, but since the numbers are very low it is impossible to make any definitive pronouncements about this. Empty it was found only in fiction, 40 LINGUISTICA PRAGENSIA 1/2016 as were the uses labelled “other”, which did not fi t neatly into any of the four main categories and had to be examined separately. a. Anaphoric it Anaphoric it was found 57 times in academic prose and 51 times in fiction, so the diff erence in frequency is not very striking. Th ere are, however, signifi cant diff erences in the use of anaphoric it. Th e most obvious diff erence is the realisation form of the referent. While in academic prose, it is realised by a clause only in two instances (3.5 %), in fi ction, clausal or sentential realisation represents over 27 % (14 instances) (the rest being noun phrases in both text types). Th is could probably be att ributed to the fact that a clausal or sentential referent is somewhat “vaguer” than a nominal one. Especially if the sentence or clause is long, or the referent is realised by more than one sentence or clause, it can be diffi cult to conceptualise it as one specifi c entity. Th is is in a way refl ected in the fact that it is oft en diffi cult to replace it by its clausal or sentential referent without adding to it or rewording it to some extent: (11) [Zoe had expected to be put in charge of the baby.] It was reasonable. (F; BNC: K8R 2152) = For Zoe to be put in charge of the baby was reasonable. An interesting point is that in this sample fi ction uses it to refer only to inanimate entities, while in academic prose, it also refers to animals and people (through collective nouns, such as ex. 12). Th is leads to a wider variety of semantic roles of the referents in academic prose, as we also fi nd roles generally att ributed only to animate entities, such as agentive or experiencer. (12) Th is point has been reinforced further by the fi ndings of Goldsmith and Newton (1986) who show that, while the � atcher government has been very directive figure 1: Functions of it in academic prose and fi ction HELENA FIŠEROVÁ 41 towards local authorities on council house sales, by contrast it hardly involves itself in matters of environmental health (a reflection of the much lower priority the Thatcher government gives to this policy area). (A; BNC: G19 1194) For the analysis of semantic roles, I used a classification based on Quirk et al. (1985: 741–7) and Biber et al. (1999: 123–5), distinguishing the subjective roles of agent, external causer, instrument, recipient, experiencer, positioner, affected, identified, characterised, locative, temporal, and eventive. In both text types, the most common semantic role is “characterised”, i.e. the role assigned to the subjects of copular verbs in qualifying and classifying constructions. This is not necessarily surprising, since both registers provide ground for describing and characterising things. It is, however, significantly more common in fiction, where it represents almost 55 % (as opposed to only 28.1 % in academic prose). What might be a contributing factor to this is the fact that while academic prose can only present objective descriptions, fiction also includes subjective views of the characters, thus the amount of things that can be characterised and the ways it can be done is theoretically larger. This is supported by the fact that in fiction we also find perception copulas such as look or sound, which present the content as relativized (Dušková et al., 2006: 206). (13) It was simply a happy coincidence that it [= her honest opinion] sounded like an insult! (F; BNC: H97 290) Another feature of anaphoric it that I examined was the distance of the pronoun from the referent, and in this respect the differences between the two registers are not very striking. The percentage of cases where the referent is found either in the same sentence or in the immediately preceding one is exactly the same in both text types (82.4 %), and academic prose shows only a slightly stronger tendency for the referent and the pronoun to appear in the same sentence (45.6 % vs. 41.2 % in fiction). The only significant difference lies in the “extreme” cases, in terms of how far the referent can be. In academic prose, the highest number of sentences separating the referent and the pronoun was 8, the referent being referred to by other instances of it four times in the intervening sentences. In two examples of fiction, however, the referent was found as far as 16 and even 19 sentences back, without a clear referential chain in-between. In both these cases, both the referent and the pronoun were found in dialogue and separated from each other by both dialogical and narrative passages; however the referent and pronoun are much “closer” to each other if instead of sentences we count only the intervening dialogue turns. This seems to suggest, therefore, that if the anaphora occurs within a dialogue, intervening narrative passages are irrelevant for determining the distance of the pronoun from the referent, in the sense that they do not make the interpretation of the pronoun more difficult — no ambiguity arises whether it could possibly refer to something mentioned in the narrative passages. Cataphoric reference is rare in both registers. It did not occur at all in the sample of academic prose and was found only twice in fiction. Both cases are found in dialogue and refer to clauses, so the reason for using cataphoric rather than anaphoric reference seems to be the ease of expression and the end-focus principle: 42 LINGUISTICA PRAGENSIA 1/2016 (14) I know it sounds silly but I think Miss Hatherby is too interested in Constance. (F; BNC: CEY 854) This phrasing is both neater than the variant with anaphoric reference I know Miss Hatherby being too interested in Constance sounds silly, but I think so, since it avoids a long subject clause, and it places the most informationally prominent part at the end of the sentence. In this respect the use of cataphoric it here is similar to anticipatory it. b. Anticipatory it There is a significant difference in the frequency of anticipatory it in the two registers. In fiction, anticipatory it occurred in only 11 instances, while in academic prose, the number was more than three times higher: 37 instances. The analysis of anticipatory it registered the realisation form of both the extraposed subject clause and the predicate of the superordinate clause, but it has shown that the type of extraposed clause is not particularly interesting in this respect: in both registers the extraposed subject is realised by a dependent declarative content clause in more than a half of the cases, followed in terms of frequency by infinitive clauses and interrogative content clauses. The type of extraposed clause is often dependent on the predicate of the superordinate clause, and that is where the main difference between the two registers lies. In academic prose, the most common type of predicate is adjectival (54.1 %), followed by verbal predicates in the passive voice (29.7 %). Fiction, in contrast, uses mainly verbal predicates in the active voice (54.5 %), followed by adjectival predicates (36.4 %). Aside from the difference in frequency of the types of predicate, the semantics of the predicates also plays an important role, both for the verbal and adjectival predicates. In adjectival predicates, academic prose uses mostly certainty adjectives (e.g. (im)possible, doubtful, apparent…) , in order to specify the certainty level of the facts presented in the subject clause. They can also be used when the author wants to avoid committing him-/herself fully to the truth value of the statement contained in the subordinate clause, i.e. they can function as hedging devices: (15) It is also likely that this view is a departure from earlier law, in which it is not attested. (A; BNC: B2P 1290) In contrast, adjectival predicates in fiction contain in the vast majority of cases evaluative adjectives (e.g. appropriate, wrong, surprising…), expressing a character’s opinion of the content of the subject clause: (16) It was ironic to Victor and the colleagues sitting round the table that morning that Gorbachev had been the KGB choice as leader. (F; BNC: FSF 1744) The fact that the information is presented from a character’s point of view can be made explicit, as in the example above (to Victor), or it is obvious from the fact that the sentence is a part of a character’s inner monologue. HELENA FIŠEROVÁ 43 In general, extraposed subject clauses with adjectival predicates allow academic prose to present a view that is not directly attributable to anyone (Biber et al., 1999: 721), which contributes to their high frequency in this register. A similar tendency can be observed in the semantics of the verbal predicates. In fiction (ex. 18), they include the verbs matter, occur (to someone) and cross (someone’s mind), i.e. verbs expressing mental processes and attitudes. They are in the active voice, while the vast majority of verbal predicates in academic prose (ex. 17) denote opinions, decisions and observations (e.g. assume, decide, show) and are in the passive voice — a major reason why anticipatory it appears so frequently in academic prose thus seems to be its tendency to avoid mentioning the agent. (17) At the outset, it was decided to use direct observation of how clients spent their time as the principal measure of effectiveness. (A; BNC: GWJ 146) (18) For a silly moment it crossed her mind that he must be the Nazi spy. (F; BNC: FPX 2019) The major difference in the use of extraposition in academic prose and fiction is therefore that in academic prose, extraposition is a means of presenting information as objective and/or avoiding attributing it to a particular source, while in fiction, it presents subjective information filtered through the consciousness of a character. c. Focusing it The number of cleft sentences was rather low in both academic prose (6) and fiction (4). The higher frequency of cleft sentences in academic prose is in accordance with the observations made by Biber et al. (1999: 961); we can therefore assume despite the low occurrence in the sample that this result is not an aberration. Additionally, some tendencies can nevertheless be observed in the realisation form and underlying syntactic function of the focused element. In academic prose, the underlying syntactic function of the focused element was an adverbial of place in half of the excerpts and subject in the other half. In fiction, most of the focused elements were subjects, one was a direct object. Additionally, the underlying subjects in academic prose differ from those in fiction — in academic prose, the subject is inanimate, expressed by a noun phrase and has the semantic role of external causer, while in fiction, it is animate with the agentive semantic role and it can also be realised by a personal pronoun or a proper name. Fiction has overall a high number of personal pronouns and proper names, it is thus not surprising that they appear as focused elements more frequently here. (19) It is literariness and not mimesis which interests the Formalists. (A; BNC: H8V 174) (20) It was she who turned away first and when she rested her arm against the wall to support the linen whilst she put the key in the lock, he came forward, saying eagerly, ‘Let me.’ (F; BNC: AT7 1957) Given the very low frequency in both samples, these results can of course be considered only tentative.