Information Structure Types

CHAPTER 2 LITERATURE REVIEW

2.3 Information Structure Types

Lambrecht (2001: 474-476) proposes three kinds of presupposition in discourse:

“pragmatic presupposition” (also called knowledge presupposition) in which “the speaker assumes the hearer already knows or believes or is ready to take for granted at the time the sentence is uttered (the old information),” “consciousness presupposition,” in which “an entity or proposition is consciousness presupposed (C-presupposed) if the speaker assumes that its mental representation has been activated in the interlocutors’ short-term memory at the time of the utterance,” and

“topicality presupposition” in which “the speaker assumes that the hearer considers it a center of current interest and hence a potential locus of predication”. It is claimed by Lambrecht that the notion of “K-presupposition” is similar to the notion “hearer-old”

of Prince (1992) and to the notion “referential” or “uniquely identifiable” in the Givenness Hierarchy of Gundel et al (1993). On the other hand, the distinction between “activated” and “accessible” in the “C-presupposition” is correspondent with the notion between “discourse-old” and “inferable” of Prince (1992) and with the notion between “in focus” and “activated” of Gundel et al. (1993). However, the

It -cleft Author

Constituent clause Prince Decklerck Gundel

1. SF [Type 1] (ex: 10-13) New Given ˇ ˇ ˇ

2. IP [Type 2] (ex: 14-18) Given New ˇ ˇ ˇ

3.Discontinuous cleft [Type 3] (ex: 20-21)

New New ˇ

4. Emphatic cleft [Type 4]

(The independence of new/given) (ex: 28-29)

Given (but relationally new)

Given ˇ

notion of “T-presupposition” does not correspond to any system of Prince (1992) and Gundel et al. (1993).

Based on the above three presuppositions, Lambrecht (1994, 2001) distinguishes three kinds of focus-presupposition or focus categories, which depends on which part of a given proposition is to be in focus in a given sentence: the predicate-focus (PF), the argument-focus (AF), and the sentence-focus (SF) sentences.

The “predicate-focus” sentence relates to “topic-comment” or “categorical” function with a given argument and a new predicate. The “argument focus” sentence relates to

“focus-presupposition,” “argumental,” or “contrastive” function with a given predicate and a new argument. The sentence focus structure relates to “all-new,” or

“presentational” function with both a new argument and a new predicate.

Lambrecht (2001) examines how the three focus categories are expressed by four languages, English, German, Italy, and French. He finds that the four languages would choose different grammatical devices, such as prosodic or syntactic change, and cleft constructions, to express their own meanings. The example of the predicate-focus (PF) sentence, also called topic-comment type, is illustrated in (31) (S stands for “subject,” V for “verb or verb phrase,” italics for prosodic prominence).

Lambrecht points out that in (31) the four languages show similar syntactic and prosodic structure of PF sentences. The topic “foot” precedes the focal predicate

“HURTS.” Thus the PF category is regarded as the UNMARKED focus category both in distribution and semantics.

(31) The predicate-focus (PF) sentence

Context: Have you recovered from your accident? How’s your foot and your knee?

a. English: SV

My foot still HURTS (but the rest if okay).

b. German: SV

Mein Fuss tut immer noch WEH (aber son st geht es).

c. Italy : SV

Il piede mi fa ancora MALE (ma il rest ova bene).

d. French : SV

Mon pied (il) me fait encore MAL (mais le reste ca ca).

According to Lambrecht, the argument-focus (AF) sentence, also called argumental or contrastive sentence, illustrates the most common type of information structure, which is parallel to the stress-focused cleft. The cleft element represents new information whereas the cleft clause conveys old information. Unlike the PF in (31), for him, AF does not exhaust the possibilities in different languages, since alternative structures could be used to present the same meaning, as in (31a) and (32c).

It-cleft is optional in English and Italian, but obligatory in French.

(32) The argument-focus (AF) sentence Context: Is your knee hurting?

a. English: SV/It-cleft

No, my FOOT hurts. / No, it’s my FOOT that hurts.

b. German: SV

Nein, mein FUSS tut weh.

c. Italy : VS / It-cleft

No, mi fa male il PIEDE. / No, e ill PIEDE che mi fa male.

d. French: It-cleft

Non, c’est mon PIED qui me fait mal.

As noted by Lambrecht (2002:507), sentence focus (SF), also called all-new, or presentational sentence, is used to “introduce either new entities or new situations into a discourse,” in which the focus is placed on both subject and the predicate. Like the AF sentence in (32), the SF sentence does not exhaust the possibilities in the four languages, as in (33b) and (33c). Have cleft sentence is optional in Italian, but obligatory in French to convey the same meaning.

(33) The sentence-focus (SF) sentence Context: Why are you walking so slowly?

a. English: SV

MY FOOT hurts.

b. German: SV / OVS

Mein FUSS tut weh. / Mir tut ein FUSS weh.

c. Italy : VS / HAVE cleft

Mi fa male un PIEDE. / Ho un PIEDE che mi fa MALE.

d. French: HAVE cleft

J’ai mon PIED qui me ait MAL.

According to the above discussion, we know that “grammars have special devices to mark the focus articulation of sentences whose information structure deviates from the unmarked predicate-focus type, that is sentences with either argument-focus or sentence-focus articulation”(Lambrecht 2001:487). In addition, it is also observed that Lambrecht’s (2002) three focus types can be compatible with the four function types of clefts: stress-focused clefts, informative presupposition cleft clefts, discontinuous cleft clefts, and emphatic clefts. This issue will be discussed in section 3.2.3.

2.4 Properties and Forms of Chinese Cleft

The above sections deal with the information structure of English cleft sentences. The following sections will turn to the investigation of the syntactic structure of Chinese cleft sentences.

In Chinese, the definitions of the two words shi and de in the cleft sentence vary greatly from different perspectives. The Chinese word shi in a cleft sentence has been treated as a focus marker (Teng 1979, Chan 1990, Zhu 1997, Lee 2005, Shyu 2008), an adverb, (Huang 1982), a verb of assertion (Tang 1983), or a copula (Li and Thompson 1981, Ross 1983, Li Chiu-Ming and Li 1994).

A number of linguists regard shi as a copula verb which is similar to BE in English clefts (Chiu-Ming and Li 1994; Li and Thompson 1981; Teng 1979). They assert that if shi in cleft sentences is indeed treated as a copula verb, Chinese and English cleft sentences might have the same structures, as shown below.

(34) a. It + be + Focus + that/who-Cl

b. shi + Focus (subject here) + zero-Rel.M + Rel.Cl.

(Li Chiu-Ming and Li 1994:55) Li Chiu-Ming and Li (1994) suggest that the structure in (1b) is equivalent to the structure of English cleft sentence, as illustrated in (1a). They assert that the focused constituent, the subject, is introduced immediately by the copular verb, followed by a relative clause. The difference between Chinese and English cleft sentences is that Chinese cleft sentences do not need a dummy subject ‘it’ and an overt relative pronoun in front of the relative clause while English do.

However, the copular analysis of shi is questioned by Teng (1979), Zhu (1997), and Choi (2006) for two reasons. First, Zhu (1997) claims that unlike English cleft sentences, in which the focused element always move to the fixed position following BE, as in (35a), shi in Chinese cleft sentence is put immediately before the focused elements, relatively freely, as in (35).

(35) a. Shi wo zai gongyuan-li zao-dao nide gou de.

SHI I at park-in find-ASP your dog DE

‘It was me who found your dog in the park.’

b. Wo shi zai gongyuan-li zao-dao nide gou de.

I SHI at park-in find-ASP your dog DE

‘It was in the park that I found your dog.’

c. Wo zai gongyuan-li shi zao-dao nide gou de.

I at park-in SHI find-ASP your dog DE

‘It was finding your dog that I did in the park.’

(Teng 1979:102-103)

Second, Teng (1979) and Zhu (1997) also point out that if shi in cleft sentences is a copula, the sentence can be reversible without changing the meaning, as in (36).

However, they claim that shi in Chinese cleft sentences is not reversible, as in (37).

When reversing the pre-and post copula elements in (37a), the sentence of (37b) would become Chinese pseudo-cleft with different meanings.

(36) a. Nei-ge nuren shi Zhangsan de mama.

that-CL woman BE Zhangsan GEN mother

‘That woman is Zhangsan’s mother.’

b. Zhangsan de mama shi nei-ge nuren.

Zhangsan GEN mother BE that-CL woman

‘Zhangsan’s mother is that woman.’

(37) a. Wo shi zai gongyuan-li zao-dao nide gou de.

I SHI at park-in find-ASP your dog DE

‘It was in the park that I found your dog.’

b. Zai gongyuan-li zao-dao nide gou de shi wo.

at park-in find-ASP your dog NOM BE I

‘The one who found your dog in the park was me.’

(Teng 1979:102)

Based on the above two reasons, Choi (2006) claims that Chinese and English cleft constructions are structurally different, since shi in the cleft construction is not a copula verb, and Chinese cleft sentence is a simple sentence with only one verb, instead of a complex sentence with an embedded clause in English. In consequence, rather than treating shi as a copula in cleft sentences, Chan (1990), Zhu (1997), Yariv-Laor (1999), and Choi (2006) regard shi as a focus marker to emphasize a portion of the sentence. Hengeveld (1990) also suggests that shi is a morpheme marking focused elements and is inserted immediately before the focused element.

2.5 De’s Function in Chinese Cleft

The Chinese word de in cleft structure has a great number of definitions from different linguists. It could be treated as a past tense marker (Teng 1979, Simpson &

Wu 1999), a nominalization particle (Paris 1978), a sentence-final marker for certainty or speaker’s attitude of assertion (Tang 1983, Lee 2005), or an evidentiality marker (Shyu 2008).

It is noted by Paris (1978) that ‘shi...de’ construction should be treated as a matrix verb shi and a nominalization particle de. Paris distinguishes two types of

shi...de construction: NP (shi) Adj (de), as in (38), and NP shi...de in which shi...de

are compulsory when the NP is the generic or non-generic reading, as in (39a) and (39b), respectively.

(38) 那個人很聰明。

‘That person is intelligent.’ (Paris 1978:54, ex 87) (39) a. The subject NP is not generic

那頂帽子是手做的。

‘That hat is hand-made.’ (Paris 1978:64, ex 123) b. The subject NP is generic

飛機是飛的。(Definition)

‘Planes fly/planes are for flying.’ (Paris 1978:65, ex 135)

The obligatory occurrence of de can also be seen in Lee’s (2005) shi...*(de) structure with non-focus meaning, as in (40). Lee (2005) suggests that the Chinese word shi in cleft sentences has two meanings: copula shi and emphatic marker shi.

The sentence final de is either regarded as a modifier marker when it appears with copula shi in the shi...*(de) structure, or as a final particle when it appears with emphatic marker shi in the shi...(de) structure. She asserts that the possibility of de omission distinguishes focus structures from non-focus structures. In non-focus

structures, de occurs obligatorily in shi...*(de) structure, and should be associated with headless relative clauses, rather than cleft sentences, as shown in (40).

(40) ta shi chi su *(de) he SHI eat vegetable DE

‘He is a vegetarian.’ (Lee 2005:132)

On the contrary, for her, the occurrence of de is optional in shi...(de) focusstructure. Lee also asserts that the possibility of de omission could be categorized into two subgroups: subject/adjunct-focus and predicate-focus structures.

The postverbal de can appear in the subject/adjunct-focus structure, but not in the predicate-focus structure, as illustrated in (41), (42), and (43), respectively.

(41) Subject-focus ‘shi..de’

shi Zhangsan zuotian qu taibei (de) SHI Zhangsan yesterday go Taipei DE

‘It was Zhangsan that went to Taipei yesterday.’

(42) Adjunct-focus ‘shi...(de)’

Zhangsan shi zuotian qu taibei (de) Zhangsan SHI yesterday go Taipei DE ‘It was yesterday that I went to Taipei.’

(43) Predicate-focus ‘shi...(de)’

* Zhangsan zuotian shi qu taibei de Zhangsan yesterday SHI go Taipei DE

‘*It was going to Taipei that Zhangsan did yesterday.’ (Lee 2005:135)

In addition, it is claimed by Lee that predicate-focus structure can be divided into two subtypes according to the acceptance of occurrence of de: stative-predicate and dynamic-predicate focus structures. De occurs obligatorily in a stative-predicate focus structure but it cannot occur in a dynamic-predicate-focus sentence, shown as (44) and (45). Therefore, Lee (2005) asserts that de plays a dominant role in the

shi..de structure, and it cannot be treated as merely an optional element.

(44) Stative-predicate

Zhangsan shi xihuan Lisi de Zhangsan SHI like Lisi DE

‘It is the case that Zhangsan likes Lisi.’

(45) Dynamic-predicate

Zhangsan shi da-le Lisi *de Zhangsan SHI bet-Asp. Lisi DE

Zhangsan did beat Lisi.’ (Lee 2005:163, ex59a,b)

Likewise, the appearance or disappearance of de can be seen in both cleft and noncleft structures suggested by Paul and Whitman (2001, 2004). As for cleft constructions, in the first pattern, when both shi and de are both present, the focused constituent is the immediate constituent following shi, resulting in the subject and adjunct focus reading, as in (41) and (42) above. In the second pattern, sentence initial bare shi without de results in the entire proposition reading, as in (46) or the subject focus reading with phonological stress, as in (47).

(46) Shi xia yu le, bu pian ni. (Lu et al. 2000:500) SHI fall rain PERF NEG trick 2SG

‘It really is that it’s raining, I kid you not.’

(47) Shi Aku he Xiao-D he – le hongjiu (Tsai Wei-Tian 2004:00-100) SHI Akiu and Xiao-D drink-PERF red wine

‘It’s Akiu and Xiao-D who drank red wine.’

There are two types of noncleft structures suggested by Whitman and Paul.

The first pattern is “propositional assertion” with both shi and de but no focused element after shi, as in (48). The second pattern is “sentence-medial bare shi” without

de resulting in focus association reading, as in (49a, b). Any constituent following shi

can be focused.

(48) Ta shi gen ni kai wanxiao de 3SF SHI with 2SF open joke DE

‘It is the case that) he was joking with you.’

‘He was just joking with you that was what he was doing.’

(Chao’s translation 1968:296) (49) a. Ta shi zai Beijin xue yuyanxue Ø,

3SG SHI at Beijing learn linguistics bu shi zai Shanghai xue (yuyanxue)

NEG SHI at Shanghai learn linguistics

‘He studies linguistics in Beijin, not in Shanghai.’

b. Ta shi zai Beijing xue yuyanxue Ø, 2SG SHI at Beijing learn linguistics

bu shi zai Beijing xue fawen NEG SHI at Beijing teach French

‘He studies linguistics, not French, in Beijing.’ (Whitman and Paul 2007)

The above discussion shows that the Chinese shi...(de) structure can be regarded as either cleft constructions or non-cleft constructions in terms of the possibility of de. However, contrary to the traditional view that the final marker de can be an optional element in shi...(de) structure as claimed by Lee (2005) and Whitman and Paul (2007), Shyu (2008:2) suggests that de should be an obligatory element for “attributing property to and predicating with the verbal nominal and the clause, respectively,” and serve as an “evidentiality marker” connected with the speaker’s asserted utterance, inference, or belief. For Shyu, the concept of evidentiality marker is in accordance with the “situational de” of Chao (1968),

“affirmation mood” of Zheng et al. (1992), and the “propositional assertion” of Paul&

Whitman (2007), as in (50).

(50) Zhangsan shi han ni kai-wan-xiao de Zhangsan SHI with you open-joke DE

‘He was just joking with you that-was-what-he-was-doing.’

(Chao’s translation 1968: 296)

Based on Declerck’s (1984) distinction of specificational reading of it-clefts from the predicational it-clefts, Shyu (2008) considers that Chinese shi...de, in addition to the identification function of focus, strongly prefers predication reading of it-clefts. For her, Chinese shi...de (with de) can express three meanings:

property-denoting sentences, as in (58), predicate focus, as in (59), and subject or adjunct focus, as in (62i)-(62iii). As for Chinese shi....(*de) (without de), similar to Whitman and Paul’s (2001, 2004) focus association type, any constituent, including the cases of subject, adjunct, or verb, following shi can be focused. For this, Shyu (2008) casts doubts on the previous claims that Chinese has it-cleft made up with

shi...(de) structure, and contends that Chinese shi...(de) cleft construction can not be

parallel to English it-cleft, but is regarded as a predicational sentence (Lambrecht 2008). The way to express the specificational cleft is by the means of Chinese pseudo cleft with de shi structure, as in (21b) below. Therefore, she questions the existence of Chinese it-cleft with the following problems, and then resorts to Lambrecht’s (1994, 2001) three focus types: Predicate Focus (PF), Sentence Focus (SF), and Argument Focus (AF) to account for the shi...(de) structure, which should be considered as a predicational sentence with an optional emphatic marker shi.

First, Shyu raises a question regarding Lee’s (2005) observation that the obligatory appearance of de can turn a dynamic predicate into a stative predicate, as in (17). According to Shyu, (52) has the property of cultivating land which is predicated with the subject ta ‘he,’ meaning He is a farmer. But for Shyu, why the active verb 來 ‘lai’ in (51) can occur with an adjunct 從中國 ‘cong Zhonggo’ in Chinese

shi...(de) structure, whereas cannot change into a stative verb and stay alone in

Chinese shi...(de) structure, as in the ungrammatical sentence (53a). In addition, Shyu further points out why the verb phrase喜歡看電影 ‘xihuan kan dianying’ can change into a stative verb with (53b-i) reading in Chinese shi...(de) structure, rather than with

the it-cleft reading in English, as in (53b-ii).

(51) Ta shi cong Zhongguo lai de. (Chao 1968:719; Paris 1979:112) He SHI from China come DE

‘It is from China that he has come.’

(52) Ta shi zhongtian de.

he SHI cultivate-paddy field DE

‘Lit: (What he does) is cultivating land./ He is a farmer.’

(53) a. *Ta shi lai de.

he SHI come DE

‘*It is coming that he did.’

b. Zhangsan shi xihuan kan dianying de.

Zhangsan SHI like see movie DE

(i) ‘It is true that Zhangsan likes to see movies.’ Lee (2005: 203) (ii) ‘*It is loving to see movies that John does.’

Second, for Shyu, the longstanding problem why the object cannot be cleaved in Chinese shi...de pattern, as in the ungrammaticality of (54a), but can only be cleaved in Chinese de shi pattern, as in (54b) still remains unanswered. It is noted by Shyu that the identification cleft is manifested by Chinese pseudo cleft.

(54) a. *Zhangsan he shi putaojiu de.

Zhangsan drink SHI wine DE

‘It was wine that Zhangsan drank.

b. Zhangsan he de shi putaojiu.

Zhangsan drink DE SHI grape-wine

‘What John drank is wine.’ (Shyu 2008)

Shyu further elaborates that English it-clefts like (55) should be viewed as Chinese de shi pattern with VP focus reading, as in (56), instead of shi...de pattern with adjunct focus reading, as in (57i) or proposition assertion reading, as in (57ii).

(55) It was writing a book that John did during the break.

(56) Zhangsan fangjia zuo de shi [xie yiben shu]

Zhangsan have-holiday do DE SHI write one-CL book

‘What Zhangsan did during the break was writing a book.’

(57) Zhangsan shi [[fangjia de shihou] xie le yi-ben shu] de.

Zhangsan SHI have-holiday Gen time write Asp one-CL book DE (i): ‘It is during the break that Zhangsan wrote a book.’

(ii): ‘It is the case that Zhangsan wrote a book during the break.’

(Shyu 2008)

So far, Shyu asserts that de as an evidentiality marker is an obligatory element in the shi...(de) structure, and contends that Chinese shi...(de) structure is not equivalent to English it-cleft, since it is considered as a predicational sentence (Lamrecht 2008) similar to Lambrecht’s three focus types: SF, PF, and AF.

In the light of Lambrecht’s three focus types, it is denoted by Shyu that shi...(de) pattern denotes “speaker’s presupposition relevant to the assertive proposition via the demonstrative/deitic grammatical function and the speech act evaluative/evidential function of de” (Shyu 2008:8). As regards the predicate focus, Shyu distinguishes two types of predicate focus: “property-denoting” predicate focus and “proposition assertion.” In the property-denoting predicate focus, as in (58), Shyu suggests that

Zhong-tian-de should be equated with predicative nominals, with individual-level or

generic reading, attributing the property of cultivating land, contrary to Lee’s (2005) headless relative analysis of predicate meaning He is a farmer.

(58) Zhangsan shi zhong tian de.

Zhangsan SHI cultivate land DE

‘Zhangsan is a famer.’

On the other hand, the proposition assertion has two functions. First, according to Shyu, the emphatic marker shi “affirms and asserts the proposition.” Second, the evidentiality marking de can not only predicates with the pre-shi subject but also

relates to previous inference or event. That is, the speaker’s utterance or belief can be traced back to his personal experience or to a familiar context known to the hearer.

Consider (59) where (50) is repeated here:

(59) Zhangsan shi han ni kai-wan-xiao de Zhangsan SHI with you open-joke DE