CHAPTER 2 LITERATURE REVIEW
2.1 Information Properties of English Clefts
Following Hedberg (2000), I will use the term “clefted constituent” to refer to the matrix phrase immediately adjacent to the copula, and the term “cleft clause” to refer to the relative clause such as subordinate wh- or that-clause, as shown in (1). In an it-cleft like It was the vase that John borke, we call the vase the clefted constituent, and that John broke cleft clause.
(1) It was the vase that John broke.
Clefted constituent Cleft clause
The same is true of Wh-clefts, suggested by Geluykens (1988), and Hedberg (2000), since wh-cleft and it-cleft have a similar semantic structure made up of a single proposition (clause) and a semantic gap (constituent), as shown in (2) and (3).
In (2), pseudo clefts can occur with cleft clause and cleft constituent in reverse order, resulting in inverted pseudo-clefts, as in (3).
(2) What John broke was the vase.
Cleft clause Cleft constituent (3) That was what John broke.
Cleft constituent Cleft clause
It is generally accepted that these terms are represented as focus and presupposition on the basis of information structure (Prince 1978, Geluykens 1988, Lambrecht 2001). This type can be comparable with Prince’s (1978) stress-focused cleft which is the most typical type of cleft. The nuclear accent is on the clefted constituent and a weakly-stressed accent on the cleft clause, as in (4). She asserts that the clefted constituent shows new, often contrastive information, and the cleft clause
shows known or old information that is already familiar to the intended hearer, as follows:
(4) So I learned to sew books. They’re really good books. It’s just THE COVERS that ARE ROTTEN. (Prince 1978: 896, ex. 38a)
In addition to the widely known type, there are other types that are paid less attention to. They are “informative-presupposition” cleft, “discontinuous” cleft, and
“emphatic” type. In the “informative-presupposition” cleft, the order of given and new is reversed. According to Prince (1978), the stress is normally on the clefted clause, which bears new information to the hearer, whereas the clefted constituent is frequently old information or short. (5) is cited from Prince (1978 898, ex.41b)
(5) The leaders of the militant homophile movement in America generally have been young people. It was they who fought back during a violent police...
The third type is Declerck’s (1984) “discontinuous” cleft, which has new information in both clefted constituent and cleft clause, as shown in (6). Declerck asserts that, in (6), the focus NP is strongly stressed and long, and it is not related to the previous context in any way.
(6) It is through the writings of Basil Bernstein that many social scientists have become aware of the scientific potential of sociolinguistics.
(Declerck: 1984:263, ex. 28b)
Following Gundel’s independent notion of given and new, I label the forth type as the “emphatic” type in which the clefted constituent could be relationally given, but relationally new (Gundel 2002: 116). For example, in (7B), the pork is relationally given, since it was mentioned in the previous question, but it is relationally new to the topic, what B ordered.
(7) A. Did you order the chicken or the pork?
B. It was the PORK that I ordered. (Gundel, 1985)
As mentioned above, it is clearly that there are four types of information property of clefts. The first and second types are Prince’s “stress-focused” cleft and
“informative presupposition” cleft. The third type is Declerck’s (1984)
“discontinuous” cleft, and the forth type is the “emphatic” cleft, as claimed by Gundel (2002). The following section will deal with the four types of information properties in more detail.
2.2 Studies of Information Structure in English Clefts
The analyses of clefts has been the theme of a large number of studies, for example, Halliday (1976), Chafe (1976), Clark and Haviland (1977), Quirk et al.
(1985), and Rochemont (1986), and it has been observed that the clefted element typically represents new information, whereas the cleft clause represents given information, though the given vs. new definitions and terminology has been addressed from a wide range of theoretical perspectives. In terms of recoverability, Halliday (1967: 204) presents information focus as “one kind of emphasis, where a speaker marks a part (which may be the whole) of a message block as that which he wishes to be interpreted as informative.” He defines “new” information as information which
“the speaker presents it as not being recoverable from the preceding discourse,” given information being defined as recoverable information. Chafe (1967) proposes the definition of the given information from a cognitive perspective. He considers that given information as ‘that knowledge which the speaker assumes to be in the hearer’s consciousness at the time of utterance (Chafe 1967: 30). On the basis of Chafe’s notion of givenness, the only change Prince (1978) makes is from ‘assumes’ to ‘may cooperatively assume,’ since it is unnecessary for the speaker to think that information is in the hearer’s mind. Others define givenness from the long-term ‘shared
knowledge’ between speakers and hearers (Clark and Haviland 1977, 1974).
Sentences (8Ba-c) are typical examples of such clefts:
(8) A. Who broke the vase?
B. (a) It was Mary who did it.
(b) The one who did it was Mary.
(c) Mary was the one who did it.
In (6), it makes sense that the information in the cleft clause is given, either because it is recoverable from the immediately preceding utterance, or because it is assumed in the hearer’s consciousness in terms of the preceding question, or because the knowledge that is shared by the speakers and hearers. So far, many scholars have pointed out the correlations between the referential expression and the cognitive status is assumed in the hearer’s mind. In what follows, I will discuss the four types of information property proposed by different scholars.
The first type and the second types of information property are what Prince (1978) terms the “stress-focused” cleft” and the “informative presupposition” cleft.
Contrary to the traditional view of the cleft clause as conveying given information, Prince (1978) observes that the cleft clause may contain new information which may or may not be “known” to the intended hearer. The information in the it-cleft clause is defined as in (9).
(9) Known information:
Information that the speaker represents as being factual and as already known to certain persons (often not including the hearer) (Prince 1978: 903) In the first type, as suggested by Prince, the “stress-focused” cleft represents as new and contrastive information and the cleft clause conveys known or old information assumed to be already known to the intended hearer, as exemplified in (10) where (4) is repeated here. In (10), according to prince, the cleft clause represents
old information, since it is known to the speaker that something is rotten, and the clefted constituent is new in relation to the old information.
(10) So I learned to sew books. They’re really good books. It’s just THE COVERS that ARE ROTTEN. (Prince 1978: 896, ex: 38a)
Declerck (1984) refers to this type as “Contrastive” cleft in which the clefted constituent is strongly stressed, thus the focus NP should be strongly contrastive. In other words, the focus is new and the wh/that-clause is old information, as shown in (11). In (11), it is old information that someone DID IT and the new information is that this is A TRAMP.
(11) No body knows who killed the old man. The police seem to believe the it was A TRAMP who DID IT.
(Declerck 1984: 264, ex:29b)
A similar view is held by Gundel (2002) who regards this type of information property as “canonical” clefts where focal stress usually falls on the clefted constituent. According to her, in the canonical cleft, the cleft clause is relationally given and the clefted constituent is relationally new, as exemplified in (12).
(12) A. Did Johansen win?
B. No. It was NIELSEN (who won).
(Gundel 2002:118, ex.12)
From the perspective of “recoverability,” Geluykens (1988) refers to this type as “filler-focus” sentence on the basis of a corpus of spoken English data. In Geluykens’s terminology, the focus of a cleft or pseudo-cleft is called the filler as opposed to the wh/that clause. In such construction the filler is focal, usually new or highly irrecoverable or contrastive, whereas the clause represents background information. Consider,
(13) A: the expression Caucasian meaning white-skinned [12 intervening tone units]
B: so why on earth it was the AMERICANS who started that craze I think.
(Geluykens 1988:828, ex.13)
So far, as we can see in the first type of information property of it-clefts, it is generally suggested that “stress-focused” clefts, (also called “contrastive” cleft,
“canonical” clefts, and “filler-focus” sentence), always have strongly accent on the clefted constituent to present new or contrastive information, and a weakly-stressed accent on the cleft clause that is already known the intended hearer.
Conversely, in the second type, namely, Prince’s “informative presupposition”
cleft, the information in the cleft clause is not presupposed but entirely new because it is not presupposed in the reader or hearer’s consciousness, as claimed by Prince.
Hence, Prince asserts that the cleft clause is used to inform the hearer of the information. For example, (14) would be awkward in canonical order, because it would seem as though the newspaper had just discovered the fact. The it-cleft “serves to mark it as a known fact, unknown only to the readership” (Prince 1978:898).
(14) ‘IT WAS JUST ABOUT 50 YEARS AGO THAT HENRY FORD GAVE US THE WEEKEND. On September 25, 1926, in a somewhat shocking move for that time, he decided to establish a 40-hour work week, giving his employees two days off instead of one.’
(Prince 1978:898, ex.41a)
Declerck (1984) refers to this type as “Unstressed-anaphric-focus” cleft in which the clefted constituent is not strongly stressed, but the cleft clause is normally stressed, and thus the focus NP does not have contrastive interpretation. In other words, the focus is new and the wh/that-clause is old information, as shown in (15). In (15), for Declerck, the cleft clause is entirely new, since it provides the answer to the
question and can not be known to the hearer at the time of hearing the utterance.
(15) A. But why is everybody so interested in uranium?
B. Because it is URANIUM that YOU NEED TO PRODUCE ATOMIC POWER. (Declerck 1984:263, ex. 26) Gundel (2002) also claims that in the informative presupposition clefts, it is not necessary that the cleft clause should be both referentially and relationally given, it can also be relationally new. Gundel (2002:118) indicates that “the cleft clause is uniquely identifiable, but it is part of the comment, the new information predicated (about the topic) in the sentence. Focal accent in these constructions typically falls within the cleft clause.” Consider the following examples, taken from Gundel 2002:118, ex. (16) and (17).
(16) Bush was elected, but it was Gore who won the POPULAR vote.
(17) Bush was elected, but it was Gore who received a telegram from Queen ELIZABETH.
Likewise, the second type of information property also resembles Geluykens’
(1988) “clause-focus” sentence in which the that/wh-clause is highlighted and irrecoverable, whereas the filler carries background information and is usually short.
In (18B), the filler is recoverable, since the pronominalization he refers to the person
Fuller in (18A), bur the cleft clause is irrecoverable to present a new proposition.
(18) A: did you met Fuller
B: yes it was he who INVITED me.
(Geluykens 1988:828, ex.14)
So far, we have discussed that in the second type of information property of it-cleft, the “informative presupposition” cleft, (also called “unstressed anaphoric”
clefts or “clause-focus” it-clefts) involves old or anaphoric information in the clefted constituent, while bears new information in the cleft clause.
In opposition to Prince’s (1978: 883) claim that it-clefts and wh-clefts are not interchangeable but rather “do different kinds of work and mean different things, at least some of the time,” Declerck (1984) argues that the cleft clause in wh-clefts should be compatible with the idea of an informative presupposition it-clefts, and distinguishes the third type of information property of clefts, “discontinuous” clefts in which both the clefted constituent and cleft clause are new information. Hence he claims that all types of clefts should be regarded as either stressed-focus clefts or informative presupposition clefts, and contends that “it-clefts and WH-clefts have the same meaning and functions” (Declerck 1984: 251). To object to Prince’s claims that only it-cleft, not wh-cleft, has informative presupposition, Declerck offers examples to prove that Prince’s examples of informative presupposition can easily be replaced by wh-clefts, as in (19). For Declerck, the it-cleft in Prince’s example in (11a) can be replaced by a reversed wh-cleft in (19b). A non-inverted wh-cleft can also have the meaning of informative cleft clause, as in (20). In (20), the information in the wh-clause is new, since it provides new information to the hearer.
(19)a. However, it turns out that there is rather interesting independent evidence for this rule and it is to that evidence that we must now turn. (Prince 1978:902)
b. However, it turns out that there is rather interesting independent evidence for this rule and that evidence is what we must now turn to.
(20) A: Those apples are good, aren’t they?
B: So they are! What keeps me from eating all of them is that mother would be furious if I left none for the others. (Declerck 1984: 259) Therefore, both it-cleft and wh-cleft have the property of informative presupposition. In addition to the traditional cleft types, namely stressed-focus clefts and informative presupposition clefts, Declerck distinguishes another type of clefts:
“discontinuous cleft” in which the focus is strongly stressed and the clause is represented as informative presupposition, as in (20) and (21).
(21) It is through the writings of Basil Bernstein that many social scientists have become aware of the scientific potential of sociolinguistics. (Declerck:
1984:263, ex. 28b)
Differ from Prince’s (1978:899) informative presupposition that the focus NP is weakly stressed and “short and anaphoric, as in (19), Declerck asserts that, in (20) and (21), the focus NP is strongly stressed and long, and it is not mentioned in the previous sentence in any way. Consequently, rather than distinguishing two types of clefts, Declerck distinguishes three. So far, the information property of clefts can be divided into three types, as suggested by Declerck (1994:194):
Type 1: clefts with a focus (highlighted element) that is ‘new’ and a wh-clause that is ‘given’
Type 2: clefts with a ‘given’ focus and a ‘new’ wh-clause
Type 3: clefts in which both constituents represent ‘new’ information
Declerck labels the three types as “contrastive clefts,” as in 8, [=Type 1],
“unstressed-anaphoric-focus” clefts [=Type 2], as in (11)-(12), and ‘discontinuous’
clefts [= Type 3], as in (13), respectively. So far, the third type of information property of cleft is what Declerck terms “discontinuous” cleft in which the cleft constituent and cleft clause are both new information. Conversely, the information could be either new or given in the cleft constituent or cleft clause in the traditional clefts, namely, stress-focused clefts and informative presupposition clefts.
Lastly, the fourth type of information property is proposed by Gundel (2002:
116) who claims that “something can be referentially given, but relationally new.” For example,
(22) A. Who called?
B. Pat said SHE called. (Gundel, 1980) (23) A. Did you order the chicken or the pork?
B. It was the PORK that I ordered. (Gundel, 1985)
From different perspectives, Gundel (1988, 1999) treats the traditional notion of given and new information as referential givenness/newness and relational givenness/newness. There are many representative examples of referential givenness/newness concepts, including “activation and identifiably statuses” of Chafe (1994) and Lambrecht (1994), and “hearer-old/new and discourse old/new statuses” of Prince (1992), but these scholars seldom classify the cognitive status of referents. In view of this, Gundel Hedberg and Zacharski (1993) propose the Givenness Hierarchy to clarify the controversial issue of cognitive status of referents, as suggested by Chafe 1976, 1987, Gundel 1978, and 1985 Prince 1981b, and assert the relationships between the referential status and the cognitive status by the givenness Hierarchy, as shown in the Givenness Hierarchy in (24):
(24) The Givenness Hierarchy (Gundel, Hedberg and Zacharski 1993)
in uniquely type focus > activated > familiar > identifiable > referential > identifiable
{it } {that,this,this N } {that N } {the N } { indefinite this N}{ a N } According to Gundel et al. 1993, as in (24), there are six cognitive statuses associated with the determiners and pronouns. They interpret that a speaker can signal her assumed cognitive status to the addressee by using a particular form. Gundel (1988, 1999) regards the relational givenness and newness as the relationship between two complementary parts, topic and comment. Topic is relationally given to comment, while comment is relationally new to topic.
Following Hedberg (2000), Gundel (2002) regards the cleft clause as a part of a definite description, and thus the cleft clause is always referentially given at least uniquely identifiable, and often familiar, activated and in focus. Gundel asserts that in a canonical cleft, or stress-focused cleft, the clefted constituent is typically relationally new, and the content of the cleft clause is relationally and referentially given, as in (25), which is (12) repeated here.
(25) A: Did Johansen win?
B: No. It was NIELSEN (who won). (Gundel, 2002:113, ex 2)
Similarly, the relationship between relational givenness/newness and referential givenness/newness can also be applied to Prince’s (1978) informative-presupposition cleft. Gundel asserts that in such construction the cleft clause is uniquely identifiable but it is part of the comment with the focal accent, and is thus relationally new, as shown in (26) and (27), in which (16) and (17) are repeated here
(26) Bush was elected, but it was Gore who won the POPULAR vote.
(27) Bush was elected, but it was Gore who received a telegram from Queen ELIZABETH.
Apart from the above properties of the two cleft constructions, stress-focused clefts and informative presupposition clefts, Gundel asserts that the relationship between referential givenness/newness and relational givenness/newness are logically independent. Hence, “something can be referentially given, but relationally new.”
(Gundel 2002: 116) For example, (28) A. Who called?
B. Pat said SHE called. (Gundel, 1980) (29) A. Did you order the chicken or the pork?
B. It was the PORK that I ordered. (Gundel, 1985)
In (28), according to Gundel, if SHE refers to Pat, it is referentially given, but it is new in relation to the topic, namely, x called. Similarly, in (29), the PORK is referentially given at least activated or even in focus, since it is raised in the preceding context. But it is relationally new to the topic, namely, what B ordered.
To sum up, according to the above discussion, basically, there are four types of information properties in it-cleft, as in (30) below. Type 1 and Type 2 are what Prince’s (1978) terms ‘stressed-focus’ it-cleft and ‘informative-presupposition’ it-cleft, respectively. According to Prince, the clefted element in Type 1 conveys contrastive information, whereas the cleft clause in Type 2 informs the hearer of the information.
In addition, Declerck (1984) labels Type 3 as ‘discontinuous’ it-cleft,’ in which the clefted element and the clause are both new to convey a piece of new information.
Last, modified by Gundel ‘independent feature of relational and referential given/newness,’ Type 4, I suggest, is considered as an ‘Emphatic cleft’ in which the given information is mentioned again to emphasize what has been chosen in the previous context. Although the cleft clause in the stressed-focus it-cleft [=Type 1] is generally given or presupposed, and the clefted constituent is typically new, the cleft clause and clefted element could also be either new or given. The following table is shown as a summarized checking list of the assertions discussed by the authors above.
(30) Four clefts types of information properties
Type 1: clefts with a focus (highlighted element) that is ‘new’ and a wh-clause that is ‘given’
Type 2: clefts with a ‘given’ focus and a ‘new’ wh-clause
Type 3: clefts in which both constituents represent ‘new’ information Type 4: clefts in which both constituents represent ‘given’ information
Table 1 A comparison of information property with it-cleft
Besides these four main properties of clefts, the study of information structure
Besides these four main properties of clefts, the study of information structure