Taiwanese Sī “Be” as a Common Ground Marker
*
Lau, Seng-hian
Ph.D., Institute of Linguistics, National Tsing Hua University voyu.lau@gmail.com
Abstract
This paper intends to investigate a usage of sī “be” in Taiwanese, which occurs higher than speaker-oriented adverb(ial)s and need not adjoin to vP in a wh-question, contrary to its non-copular homonyms (Lee, 2005). I suggest that it can be analyzed based on the denotation of FOR-SURE really, which conveys the meaning: “it is for sure that we should add to Common Ground that p,” proposed in Romero and Han (2004). The analysis not only explains all the empirical disparity observed between this sī and other non-copular usages of sī but also reveals the existence of a pragmatic marker in Sinitic languages, which indicates a result of high degree of grammaticalization of sī, an item whose cognates have also evolved into a wide spectrum of functions / distribution in today’s Sinitic languages.
Keywords: pragmatic marker, Taiwanese, Southern Min, copula verb, focus
◎ 收稿日期︰2016 年 11 月 7 日；審查通過日期︰2017 年 3 月 14 日。
∗ I owe much to questions and comments by Ching-yu Helen Yang, Chung-yu Barry Yang, and three anonymous
reviewers, as well as the audience at IACL-21 and the 18th Workshop of the International Research Project on Comparative Syntax and Language Acquisition (2013 Siena-Tsinghua-Nanzan Joint Workshop). I would like to offer special thanks to Wei-tien Dylan Tsai for his kind support and useful suggestions. Additionally, this research would be impossible without the kind helps from my informants Su-iong Pang, A-bun Pang Ko, Chen Hao, Chaoyi Chen, Man Luo and Hsiu-ling Luo. The research leading to this article is funded by the Ministry of Science and Technology, Taiwan (Formerly National Science Council, Taiwan) (NSC 101-2410-H-007-055-MY2; MOST 103-2410-H-007-026-MY3). All remaining errors are mine.
1. Introduction
In this paper, I argue for a high conversational marker sī “be” in Taiwanese, and expound its differences from other non-copular usages of sī (and its cognate shì in Mandarin Chinese - henceforth MC), especially the so-called predicate-focus marker (Lee, 2005). I will show that the high occurrence of sī can be analyzed à la Romero and Han (2004) and this sī is a realization of a specific usage of “really,” a sometimes covert operator, in English that raises conversational FOR-SURE implicature.
The discussion is arranged as follows. Pertinent data are presented in section 2, followed by a brief review of previous studies in section 3. Section 4 summarizes the empirical observations. In section 5, I try to sort out the properties and the function of the element under investigation. Based on the observations and its properties, I propose to analyze this element and its predicate-adjacent homonyms à la Romero and Han (2004) in section 6. This paper is then concluded in section 7.
2. Data
In this section, the data will be shown in two subgroups. The first one is about sī / shì in declaratives. And in the other I will demonstrate some contrast between Taiwanese and MC with respect to sī / shì in a wh-question.1
2.1. Sī Occurring High in a Declarative Sentence
In contrast to its cognate in MC, which cannot occur higher than epistemics (see (17) in section 4; cf. Lee, 2005: 186-187), it is not problematic for the Taiwanese sī to precede an epistemic that denotes the speaker’s assessment of probability and predictability (Halliday, 1970: 349). (Note that the usage which concerns us here is context sensitive and that this kind
1 In my field work, some dialectal variances regarding the grammatical judgments of the MC sentences were found.
Just as pointed out by a reviewer, the MC sentences marked ungrammatical in this paper are not ruled out in some contexts according to her / his judgment. In my survey, it is true that many MC speakers from Taiwan did not rule out these sentences outright. The judgments vary from being marginal to ungrammatical. Whereas, all my consultants who are MC speakers from Northern China rejected the possibility of putting shì before any epistemics and evaluatives without hesitation. The dialectal variances may be due to language contact and indicate a new usage of shì in MC under development, probably a borrowed conversational marker from Taiwanese, if the analysis provided herein is on the right track.
of sentences does not come out of blue. Please refer to section 5.) 2.1.1 Sī > epistemic
(1) a. A-bîng sī huān-sè bat khì hit-ê sóo-tsāi kuè. (Taiwanese) Abing BE maybe ASP go that-CL place ASP
阿明 是 凡勢 捌 去 彼个 所在 過。 “(As we know,) perhaps Abing has been there before.”
b.* Zhāngsān shì huòxǔ qù guò nà-ge dìfāng. (MC) Zhangsan BE maybe go ASP that-CL place
*張三 是 或許 去 過 那個 地方。 (Intended) “(As we know,) perhaps Zhangsan has been there before.”
To go further, sī can even be followed by an evidential, a grammatical marker which the speaker uses to specify an information source (Aikhenvald, 2004), contrary to shì in MC.
2.1.2. Sī > evidential
(2) a. Tsúi-sūn sī bîng-bîng ū kóng beh tńg--lâi, sī-án-tsuánn Tsuisun BE evidently have say will return why
水順 是 明明 有 講 欲 轉來， 是按怎
lóng bô khuàinn lâng? (Taiwanese) all not see person
攏 無 看見 人？
“(As we know,) evidently, Tsuisun said that he will be back. Why hasn’t he come back yet?”
b.* Zhāngsān shì míngmíng shuō hùi húilái, wèishénme Zhangsan BE evidently say will return why *張三 是 明明 說 會 回來 為什麼
háiméi khànjiàn rén? (MC)
still.not see person 還沒 看見 人？
(Intended) “(As we know,) evidently, Zhangsan said that he will be back. Why hasn’t he come back yet?”
More than this, we can have sī occur before an evaluative which represents a speaker’s evaluation of the fact represented by the sentence proposition content (Ernst, 2009), again, in contrast to shì in MC.
2.1.3. Sī > evaluative
(3) a. Hit kang sī hó-ka-tsài guá kā tàu-sann-kāng, bô, that day BE fortunately I LV help otherwise 彼 工 是 好佳哉 我 共 鬥相共， 無，
i tō tshám --ah. (Taiwanese)
he then miserable PRT
伊 就 慘 矣。
“(As we know,) he was fortunate that I was there to give him a hand on that day. Otherwise, he would be in a big trouble.”
b.* Nà tiān shì xìnghǎo wǒ bāng tā máng, bùrán, tā *那 天 是 幸好 我 幫 他 忙， 不然， 他 that day be fortunately I help him busy otherwise he
ké cǎn le. (MC)
can miserable PRT
可 慘 了。
(Intended) “(As we know,) he was fortunate that I was there to give him a hand on that day. Otherwise, he would be in a big trouble.”
Taken to the extreme, we can also find sī preceding a speech-act adverb which signals the speaker’s communicative intention or illocutionary force (Searle, 1969, 1983).
2.1.4. Sī > speech-act
(4) a. Guá sī láu-st-kóng thiànn kah beh sí, tsing-tsha
我 是 老實講 疼 甲 欲 死，精差
I BE frankly.speaking hurt to.the.extent will die differ
bô háu --tshuat-lâi niā-niā. (Taiwanese) not cry out only
無 吼 出來 爾爾。
b.* Wǒ shì lǎo-shí-shuō tòng de yào mìng, zhǐ chā I BE frankly.speaking hurt obtain want life only differ *我 是 老實說 痛 得 要 命， 只 差
méi kū chū- lái éryǐ. (MC)
not cry out only 沒 哭 出來 而已。
(Intended) “(You see: It’s painful.) Frankly speaking the pain almost killed me. I was close to tears.”
The contrast illustrated above, to my knowledge, is not depicted in the literature and needs explanation.
2.2 Intervention Effect in a Wh-question
It has long been observed that the presence of shì in MC will cause the intervention effect in a question with a wh-adverbial (Cheng & Rooryck, 2002; Soh, 2005; Tsai, 2008; Yang, 2008; see (5a)). Moreover, Yang (2008: 9-10) shows that wh-nominals are not totally immune from the intervention effect, illustrated in (5b). (Examples below are from Yang, 2008: 9 (17a), (16a).)2
(5) a.* Shì Zhāngsān wèishénme / zěnme cízhí? (MC)
BE Zhangsan why / how resign *是 張三 為什麼 / 怎麼 辭職？
(Intended) “Why / how is it such that it was Zhangsan who resigned?”
b.* Shì Zhāngsān chī-le shénme? (MC)
BE Zhangsan eat-ASP what
*是 張三 吃了 什麼？
(Intended) “What was x such that it was Zhangsan who ate x?”
When shì occurs in a lower position (adjoined to vP), only the weak intervention effect occurs.
(6) ?Zhāngsān shì chī-le shénme?3 (MC) Zhangsan BE eat-ASP what
?張三 是 吃了 什麼？ “What did Zhangsan eat?”
Tsai (2012) suggests that it is possible to get rid of this weak intervention effect by putting stress on the wh-object in order to emphasize its D-linking effect.
(7) Zhāngsān dàodǐ shì chī-le SHÉNME(, cái huì Zhangsan on-earth BE eat-ASP what only.then would
張三 到底 是 吃了 什麼（， 才 會
dùzi tòng de zhème lìhài)? (MC)
stomach be.painful obtain so serious
肚子 痛 得 這麼 厲害）？
“What on earth did Zhangsan eat? (He has stomachache seriously.)”
What is intriguing here is that no intervention effect is found in a wh-nominal question in Taiwanese.
(8) Tsuí-sūn sī tsih siánn? (Taiwanese) Tsuisun BE eat what
水順 是 食 啥？
“What did Tsuisun eat? (I know that he ate something.)”
The normative understanding of native speakers regarding questions like (8) is that they are employed when an inquirer already knows (or believes) that the event in question did happen and he is curious about the details (Cf. the D-linking effect observed in MC in Tsai, 2012).
The difference between these two languages can be further demonstrated by the (non-)possibility of an intervening adverbial between sī / shì and vP. (9) shows the case where a manner adverb is present; (10) and (11) involve temporal adverbs; in (12) locative adverbial occurs in between; (13) gives an example of causal adverbs.
3 Many MC speakers in Taiwan consider this sentence grammatically correct. This is presumably a dialectal
2.2.1. Sī > X > vP
(9) a. Tsuí-sūn sī hiông-hiông-kông-kông beh khì tó-uī? (Taiwanese) Tsuisun BE hastily will go where
水順 是 雄雄狂狂 欲 去 佗位？
“(We know that Zhangsan hastily went to some place.) Where is Tsuisun hastily going?” b.* Zhāngsān shì huānghuāngzhāngzhāng yào qù nǎlǐ? (MC)4
Zhangsan BE hastily will go where *張三 是 慌慌張張 要 去 哪裡？
(Intended) “(We know that Zhangsan hastily went to some place.) Where is Zhangsan hastily going?”
(10) a. Tsuí-sūn sī tú-tsiah tú-tih siánn-lâng? (Taiwanese) Tsuisun BE a.moment.ago encounter-ASP who
水順 是 拄才 拄著 啥人？
“(We know that Tsuisun just ran into someone.) Who did Tsuisun encounter a moment ago?”
b.* Zhāngsān shì gāngcái yùjiàn shéi le? (MC) Zhangsan BE a.moment.ago encounter who PRT
*張三 是 剛才 遇見 誰 了？
(Intended) “(We know that Tsuisun just ran into someone.) Who did Zhangsanencounter a moment ago?”
4 Ching-yu Yang states that by positioning a pronoun in between Zhāngsān and shì (“Zhāngsān tā shì
huānghuāngzhāngzhāng yào qù nǎlǐ”), the sentence sounds good to her. Aside from the different judgments Mandarin speakers may provide, the cross-linguistic contrast identified in this section still exists. I searched the sequence “shì huānghuāngzhāngzhāng” online, and found the results to be very limited. It is noteworthy that among the cases where huānghuāngzhāngzhāng is used as an adverb most are in a contrastive reading, either with a clausal pair in contrast or a focus marker present (e.g., doū, zhè, gèng). This kind of usage can also be demonstrated by another example from Yang, as shown below.
(i) Tā shì yěxǔ bú huì lái le méi cuò. (MC) he BE perhaps NEG will come ASP NEG wrong
他 是 也許 不 會 來 了 沒 錯 “It’s true that perhaps he will not come.”
It seems to me that the phrase “méi cuò” loads the sentence with an additional verum focus. Nonetheless, the point is that the sī discussed in this paper is not a contrastive one (see section 5).
At any rate, the pre-speaker-oriented-adverbial occurrences indicated previously still needs explanation. Moreover, we should not neglect the dialectal variance aforementioned. This kind of usage is only possible among some MC speakers in Taiwan, according to my survey.
(11) a. Tsuí-sūn sī bîn-á-tsài beh khì bé siánn? (Taiwanese) Tsuisun BE tomorrow will go buy what
水順 是 明仔載 欲 去 買 啥？
“(We know that Tsuisun will buy something tomorrow.) What is Tsuisun going to buy tomorrow?”
b.* Zhāngsān shì míngtiān yào qù mǎi shénme? (MC) Zhangsan BE tomorrow will go buy what
*張三 是 明天 要 去 買 什麼？
(Intended) “(We know that Tsuisun will buy something tomorrow.) What is Zhangsan going to buy tomorrow?”
(12) a. Tsuí-sūn sī tī hia teh kíng beh bé siánn? (Taiwanese) Tsuisun BE in there ASP select will buy what
水順 是 佇 遐 咧 揀 欲 買 啥？ “(We know Tsuisun is there) What is Tsuisun selecting and buying there?”
b. *Zhāngsān shì zài nàlǐ tiāoxuǎn shěme? (MC) Zhangsan BE in there select what
*張三 是 在 那裡 挑選 什麼？
(Intended) “(We know Tsuisun is there) What is Zhangsan selecting there?”
(13) a. Tsuí-sūn sī bô-tāi-bô-tsì tshut-khì tshòng siánn? (Taiwanese) Tsuisun BE without.a.cause out-go do what
水順 是 無代無誌 出去 創 啥？
“(We know that Tsuisun went out.) What is Tsuisun going out to do? I can see no reason.”
b.* Zhāngsān shì wúyuánwúgù zhūqù zuò shénme? (MC) Zhangsan BE without.a.cause out.go do what
*張三 是 無緣無故 出去 做 什麼？
“(We know that Tsuisun went out.) What is Zhangsan going out to do? I can see no reason.”
I summarize the observations in section 4. Before we proceed to that, let’s have a brief review of previous studies in which, to my best knowledge, this usage of sī has not been touched on.
3. Previous
Studies
As mentioned in the beginning of section 2, the data under consideration has not been previously discussed in the literature. Previous studies regarding shì in MC are numerous but none are directly relevant to the phenomena in question. This paper does not intend to review them in detail. The key findings of some of the previous research are discussed briefly in this section.
Shì in MC is described in many different ways, according to its various usages respectively. It has been suggested as a copula (e.g., Wang, 1937; Chao, 1968; Tang, 1979), an identifying verb (Li, 1925; cf. Wang, 1954; Hsu, 1973), a demonstrative (Gao, 1970), a discerning verb denoting affirmation and emphasis (Tang, 1979), a transitive verb (Chao, 1968), or a nominalizing specifier in the “shì...de” construction (Chao, 1968; Li & Thompson, 1981). Some claim that it produces contrastive stress oranassertive reading (Chao, 1968; Lee, 2005), or signals special affirmation (Li & Thompson, 1981). Shì is also entertained to be either transitive or intransitive (Huang, 1988). Based on the different syntactic positions of this element, it is also claimed to be either a Focus head or a vP adjunct (Lee, 2005). A radical proposal is found in Cheng, in which all its usages are argued to involve nothing but a copula (Cheng, 2008).
Stemming from the claim that in MC predicate structure directly determines the topic-comment structure of a clause, von Prince (2012) develops formal definitions of the copula and the so-called comment marker shì. He distinguishes being contrastive from being the comment of an utterance and suggests these two belong to two independent categories and should not be collapsed into the notion of focus. Even though the semantic definition of the copula shì is quite close to the meaning of the comment marker shì, von Prince insists that they are two different lexemes. According to von Prince, the function of the comment marker shì (是) is to interfere with the default predicate structure of a clause and to imply that the comment is contrastive. Syntactically, von Prince suggests that comment marking shì is an adjunct to the constituent which it takes as its first argument.
As noted by von Prince (2012), most previous studies that treat MC shì as a focus marker identify the information-structural particle shì with the copula shì; whereas, no definition which covers both uses has ever been provided in them. Here I would like to point out that an all-copula analysis is not viable. And this can be demonstrated by considering the occurrence
of shì / sī with different kinds of predicates. Compare the relative positions between sī and the adverb ū-iánn “really” in the following.
(14) a. Guá ū-iánn sī hk-sing. (Taiwanese) I really BE student
我 有影 是 學生 “I am really a student.”
b. Guá sī ū-iánn sī hk-sing --ah, (-koh i bô-ài sìn.) I BE really BE student. PRT but he not-want believe 我 是 有影 是 學生 啊， 毋過 伊 無愛 信。
“It is true that I am a student(, but he doesn’t believe it.)”
(15) a. Hong-thai sī ū-iánn lâi --ah. (Taiwanese) typhoon BE really come ASP
風颱 是 有影 來 矣。 “It is true that the typhoon has arrived.”
b. Hong-thai ū-iánn sī lâi --ah. typhoon really BE come ASP
風颱 有影 是 來 矣。 “The typhoon has really arrived.”
c.* Hong-thai sī ū-iánn sī lâi --ah. typhoon BE really BE come ASP
*風颱 是 有影 是 來 矣。 (Intended) “It is true that the typhoon has indeed arrived.”
As shown by the contrast between (14b) and (15c), repetition of sī in a clause is more restricted when the predicate is not nominal. This is not conceivable if we acknowledge that all shì-s / sī-s share the same syntactic status and function.
The fact that there are different kinds of shì-s / sī-s can also be illustrated in another way. Consider the following sentences in which, again, we have shì / sī iterated.
(16) a. Tsuí-sūn sī ū-iánn sī Gîn-khuân ê hk-sing (bô--tih). Tsuisun BE really BE Gin-khuan LK student (not-wrong) 水順 是 有影 是 銀環 的 學生 無毋著。
b.* Sī Tsuí-sūn sī tsa-hng khì Tâi-pak.
BE Tsuisun BE yesterday go Taipei *是 水順 是 昨昏 去 台北。
(Intended) “It is Tsuisun who went to Taipei yesterday and it is yesterday but not any day else.”
Compare (16a) with (16b), it is obvious that double occurrences of sī are conditioned by its positions (and the corresponding functions, probably). If all sī-s are copulas that are identical and distributed in whatever slot in a sentence, the contrast between the examples above would be mysterious.
4. Summary
In a nutshell, shì / sī is a multi-functional element in MC and Taiwanese. Its usages include equatives (identifying reading; equational; specificational), property denoting (attributive reading; classificatory; predicational), existential meaning, subject-focus, adjunct-focus, and predicate-focus (see Lee, 2005). As noted previously, Lee (2005) claims that in MC epistemic and deontic modals can only be dominated by shì in predicate-focus constructions, but not in in subject-focus and adjunct-focus structures (2005: 186-187). In fact, Lee’s examples in regard of epistemics involve only kěnéng “probably,” which has strong verbal / nominal properties in contrast to other MC epistemics.5 Empirically, shì never precedes a typical epistemic no matter under which usages of focus it is. This is evidently demonstrated in the following examples.
(17) a.* Zhāngsān shì yěxǔ qù le Táiběi. (MC) Zhangsan BE perhaps go ASP Taipei
*張三 是 也許 去 了 台北。 (Intended) “Perhaps Zhangsan went to Taipei.”
b.* Zhāngsān shì dàgài qù le Táiběi. Zhangsan BE probably go ASP Taipei *張三 是 大概 去 了 台北。 (Intended) “Zhangsan probably went to Taipei.”
5 Chris I-da Hsieh (p.c.) suggests that the sentences in which shì precedes kěnéng provided in Lee (2005) may
involve an unpronounced yoǔ “have” before kěnéng. This surmise is in agreement with the common nominal usage of kěnéng, which is not available to other typical epistemics in Mandarin Chinese.
As shown in (17), predicate-focus shì (and other focus markers) in MC cannot precede yěxǔ “perhaps” and dàgài “probably,” which are typical epistemics in MC.
The available position and possible co-occurrence of shī / sī and adverb(ial)s are summarized in (18).
(18) a. Available positions in non-subject / -adjunct focus sentences
Taiwanese sī sī sī sī MC speech-act adverbs evaluative adverbs evidential adverbs epistemic adverbs shì deontic adverbs shì manner adverbs shì vP
b. Available positions in wh-questions
Taiwanese sī sī sī MC deontic adverbs manner adverbs shì vP
As attentive readers may have noted, no test involved epistemics and adverbs higher than epistemics is carried out in subsection 2.2. This is due to the conflict of speaker-oriented adverbs and the interrogative mood. Epistemics, especially strong positive polarity items (PPIs), can only occur in a negative rhetorical question (refer to Ernst, 2009, which accounts for it by Giannakidou’s (1999) (non)veridicality theory; also refer to Jackendoff, 1972). The same goes for other speaker-oriented adverb(ial)s.
Taking (18) as a point of departure, we have the following questions to answer: What is this sī under investigation? Why does it behave differently from other non-copular sī / shì with respect to their available positions, the incurring of intervention effect in a question, and (non-)co-occurrence with other non-copular sī / shì?
I will try to answer these questions in the sections to follow.
5. On the Function of the High
Sī
To specify the function of this element in question, firstly we have to identify its characteristics and properties. Foremost, it is noted that this sī cannot be used in a null context.
The context in (19) is designated in order to make the utterance come out of blue.
(19) CONTEXT: All the members of a family gather in the living room. The father is watching TV news while others are chatting. The TV news anchor mentions that a cold front will approach this region tomorrow. No one pays attention to the weather forecast except for the father. And he cuts in on the conversation.
a. Bîn-á-tsài huān-sè beh piàn-thinn --ah. (Taiwanese) tomorrow perhaps will change.sky PRT
明仔載 凡勢 欲 變天 矣。
“Perhaps the weather will turn bad tomorrow.”
b.# Bîn-á-tsài sī huān-sè beh piàn-thinn --ah. tomorrow BE perhaps will change.sky PRT
#明仔戴 是 凡勢 欲 變天 矣。 “(As we know,) perhaps the weather will turn bad tomorrow.”
Compare (19a) and (19b) in this context, (19b), in which the high sī occurs, is infelicitous.
Now with (19a) in mind (added in the discourse), the context is not null any more. (20) provides two replies to (19a). Compare these replies.
(20) CONTEXT: Following up (19). Now imagine that the mother replies the father with something.
a.# Bîn-á-tsài huān-sè beh piàn-thinn --ah, -koh mā sī tomorrow perhaps will change.sky PRT however still BE
#明仔戴 凡勢 欲 變天 矣， 毋過 嘛 是
tih-ài tshut-mg khì siōng-pan, bô thang hioh-khùn. (Taiwanese) have.to out.door go work not can rest
著愛 出門 去 上班， 無 通 歇睏。
“Perhaps the weather will turn bad tomorrow. However, we still have to go to work and cannot take a leave.”
b. Bîn-á-tsài sī huān-sè beh piàn-thinn --ah, -koh mā sī tomorrow BE perhaps will change.sky PRT however still BE
tih-ài tshut-mg khì siōng-pan, bô thang hioh-khùn. have.to out.door go work not can rest
著愛 出門 去 上班， 無 通 歇睏。
“(As we know,) perhaps the weather will turn bad tomorrow. However, we still have to go to work and cannot take a leave.”
With a context in place, the sentence with the high sī becomes the appropriate one. (19) and (20) illustrate that the sī in question is infelicitous in an out-of-blue context.
Before we go on, a reasonable question to ask is whether there is really discrepancy between this high sī and the so-called predicate-focus sī (a vP adjunct; Lee, 2005) for, without the occurrence of adverbs, sentences with either one of them just look the same on the surface. Examples in (21) and (22) are to compare the high sī and the so-called predicate-focus sī in an identical context.
(21) a. Tsuí-sūn án-tsuánn siūnn guá mā bô-khak-tīng, i huān-sè sī Tsuisun how think I also not-be.sure he perhaps BE
水順 按怎 想 我 嘛 無確定， 伊 凡勢 是 siūnn-beh thk tāi-hk. (Taiwanese; predicate-focus)
want read university
想欲 讀 大學。
“I am not sure what Tsuisun thinks. Perhaps he does want to go to university.” b. #Tsuí-sūn án-tsuánn siūnn guá mā bô-khak-tīng, i sī
Tsuisun how think I also not-be.sure he BE
#水順 按怎 想 我 嘛 無確定， 伊 是
huān-sè siūnn-beh thk. tāi-hk. (the high sī) perhaps want read university
凡勢 想欲 讀 大學。
(Intended) “#I am not sure what Tsuisun thinks. (As we know,) perhaps he does want to go to university.”
(22) a. Tsìn-tsîng guá lóng phiàn --lín, tann guá -kánn --ah, Tsuí-sūn before I all cheat you.PL now I not-dare PRT Tsuisun 進前 我 攏 騙 恁， 今 我 毋敢 矣， 水順
láu-st-kóng sī bô kah-ì Gîn-khuân --ê --lah. frankly BE NEG like Ginkhuan PRT PRT
老實講 是 無 佮意 銀環 的 啦。
“I was lying to you. Now I dare not lie anymore. Frankly, Tsuisun in fact does not like Ginkhuan.” (Taiwanese; predicate-focus) b.# Tsìn-tsîng guá lóng phiàn --lín, tann guá -kánn --ah, Tsuí-sūn before I all cheat you.PL now I not-dare PRT Tsuisun #進前 我 攏 騙 恁， 今 我 毋敢 矣， 水順 sī láu-st-kóng bô kah-ì Gîn-khuân ê lah.
be frankly not like Ginkhuan PRT PRT
是 老實講 無 佮意 銀環 的 啦。
(Intended) “#I was lying to you. Now I dare not lie anymore. (As we know,) frankly, Tsuisun in fact does not like Ginkhuan.” (the high sī)
The first halves of these sentences are intended to provide contrastive information for the adverbs in the second halves. As illustrated, only predicate-focus sī but not the high sī is felicitous with this kind of information in place. Therefore, these two sī should be distinguished even though they may look like appearing in the same position in the word string of a simple sentence.
With its non-occurrence in an out-of-blue context in mind, we might wonder if this sī is a presupposition marker. Below several conventional tests are performed to examine this possibility.
The wait a minute test (von Fintel, 2004) is executed from (23) to (26) with different adverb(ial)s respectively.6
(23) a. Tsuí-sūn sī huān-sè bîn-á-tsài beh khì Bí-kok. (Taiwanese) Tsuisun BE perhaps tomorrow will go U.S.A.
水順 是 凡勢 明仔戴 欲 去 美國。 “(We know that) it might be that Tsuisun will go to U.S.A. tomorrow.”
6 In order to exclude the influence of the root phenomenon (main clause phenomenon), this test is carried out
b.? Siánn! Lí kóng siánn! Tsuí-sūn huān-sè bîn-á-tsài beh khì Bí-kok! what you say what Tsuisun perhaps tomorrow will go U.S.A. 啥！ 你 講 啥！ 水順 凡勢 明仔載 欲 去 美國！ (Intended) “What!? What did you say!? Tsuisun might go to U.S.A. tomorrow!?” c. Siánn! Lí kóng siánn! Tsuí-sūn ē siūnn-beh khì Bí-kok!
what you say what Tsuisun will want go U.S.A. 啥！ 你 講 啥！ 水順 會 想欲 去 美國！ “What!? What did you say!? Tsuisun has the intention to go to U.S.A.!” (24) a. (Tsa-hng ê khó-tshì,) Tsuí-sūn sī hó-ka-tsài ū tāi-sing
yesterday LK exam Tsuisun BE fortunately have in.advance (昨昏 的 考試，) 水順 是 好佳哉 有 代先
tsún-pī, (nā bô, it-tīng khó bē kuè.) (Taiwanese) prepare if not definitely take.an.exam not.can pass
準備， (若 無， 一定 考 袂 過。)
“(As for the exam yesterday, we know:) fortunately Tsuisun prepared in advance.” (Otherwise, he would not pass it.)
b.# Siánn! Lí kóng siánn! Tsuí-sūn hó-ka-tsà ū tāi-sing tsún-pī! what you say what Tsuisun fortunately have in.advance prepare #啥！ 你 講 啥！ 水順 好佳哉 有 代先 準備！ “What!? What did you say!? Tsuisun fortunately prepared for the exam in advance!” c. Siánn! Lí kóng siánn! Tsuí-sūn ū tāi-sing tsún-pī!
what you say what Tsuisun have in.advance prepare 啥！ 你 講 啥！ 水順 有 代先 準備。 “What!? What did you say!? Tsuisun prepared for the exam in advance.”
(25) a. Tsuí-sūn sī bîng-bîng ū kóng i beh lâi(, lán koh tán --i Tsuisun BE evidently have say he will come we.INC ADD wait he 水順 是 明明 有 講 伊 欲 來（，咱 閣 等 伊 --tst-ē.) (Taiwanese) one-VERBAL.CL
一下。）
“(We know that) evidently, Tsuisun said he will come. Let’s wait for a while.” b.# Siánn! Lí kóng siánn! Tsuí-sūn bîng-bîng kóng i ū beh lâi!
what you say what Tsuisun evidently say he have will come #啥！ 你 講 啥！ 水順 明明 講 伊 有 欲 來！ “What!? What did you say!? It’s evidently that Tsuisun said he will come!”
c. Siánn! Lí kóng siánn! Tsuí-sūn ū kóng i beh lâi! what you say what Tsuisun have say he will come 啥！ 你 講 啥！ 水順 有 講 伊 欲 來！ “What!? What did you say!? Tsuisun said he will come!”
(26) a. Tsuí-sūn sī láu-st-kóng tsiok kah-ì Gîn-khuân, bô kóng Tsuisun be frankly very like Ginkhuan not say 水順 是 老實講 足 佮意 銀環， 無 講
--tshut-lâi niā-niā. (Taiwanese) out-come only
出來 爾爾。
“(We know that) frankly Tsuisun likes Ginkhuan very much. He just never says it.” b.# Siánn! Lí kóng siánn! Tsuí-sūn láu-st-kóng tsiok kah-ì Gîn-khuân!
what you say what Tsuisun frankly very like Ginkhuan #啥！ 你 講 啥！ 水順 老實講 足 佮意 銀環！ “What!? What did you say!? Tsuisun frankly likes Ginkhuan very much!” c. Siánn! Lí kóng siánn! Tsuí-sūn kah-ì Gîn-khuân!
what you say what Tsuisun like Ginkhuan 啥！ 你 講 啥！ 水順 佮意 銀環！
“What!? What did you say!? Tsuisun likes Ginkhuan!”
The results above are not surprising since speaker-oriented adverbs are well known to be out of presupposition; they also do not contribute to the assertion content. Anyway, it is suggested that the sentence with the high sī in it is not a presupposition itself, in contrast to the presupposed parts.
The second test, which is also common in studying presupposition, is negation test. However, due to the fact that speaker-oriented adverbs cannot be put under negation, it is difficult to come up with sentences with high sī negated. As shown in (27) and (28), these sentences are not grammatical.7 As a result, negation test is not applicable.
7 Sentences in (28) can only be interpreted as rhetorical questions. The rhetorical question reading with high sī is
(27) a.* Bô hit-hō tāi-tsì, bô-iánn Tsuí-sūn sī huān-sè bē lâi --ah. (Taiwanese) no that-CL thing untrue Tsuisun BE perhaps not come PRT
*無彼號 代誌， 無影 水順 是 凡勢 袂 來 矣。 (Intended) “That’s not true. It’s untrue to say: Tsuisun perhaps will not come.” b.* Bô hit-hō tāi-tsì, bô-iánn Tsuí-sún sī bîng-bîng beh khì Tâi-pak.
no that-CL thing untrue Tsuisun BE evidently will go Taipei *無彼號 代誌， 無影 水順 是 明明 欲 去 台北。 (Intended) “That’s not true. It’s untrue to say: Tsuisun evidently will go to Taipei.” c.* Bô hit-hō tāi-tsì, bô-iánn Tsuí-sún sī hó-ka-tsài khó ū kuè.
no that-CL thing untrue Tsuisun BE fortunately take.an.exam have pass *無彼號 代誌， 無影 水順 是 好佳哉 考 有 過。 (Intended) “That’s not true. It’s untrue to say: Tsuisun fortunately passed the exam.” d.* Bô hit-hō tāi-tsì, bô-iánn Tsuí-sún sī láu-st-kóng tsiok kah-ì Gîn-khuân.
no that-CL thing untrue Tsuisun be frankly very like Ginkhuann *無彼號 代誌， 無影 水順 是 老實講 足 佮意 銀環。 (Intended) “That’s not true. It’s untrue to say: Tsuisun frankly like Ginkhuan very much.”
(28) a.* Tsuí-sūn -sī huān-sè bē lâi --ah. (Taiwanese) Tsuisun not-BE perhaps not come PRT
*水順 毋是 凡勢 袂 來 矣。 (Intended) “It’s wrong to say: perhaps Tsuisun will not come.” b.* Tsuí-sún -sī bîng-bîng beh khì Tâi-pak.
Tsuisun not-BE evidently will go Taipei *水順 毋是 明明 欲 去 台北。
(Intended) “It’s wrong to say: evidently Tsuisun will go to Taipei.” c.* Tsuí-sún -sī hó-ka-tsài khó ū kuè.
Tsuisun not-BE fortunately take.an.exam have pass
*水順 毋是 好佳哉 考 有 過。
(Intended) “It’s wrong to say: fortunately, Tsuisun passed the exam.” d.* Tsuí-sún -sī láu-st-kóng tsiok kah-ì Gîn-khuân.
Tsuisun not-BE frankly very like Ginkhuan *水順 毋是 老實講 足 佮意 銀環。
In addition to the two tests, projection in an antecedent of a conditional is observed in (29) and (30). As we can see, there is no presupposition of (29b) and (30b) projected in (29a) and (30a). The result also indicates that sentences with high sī are not presuppositions themselves.8
(29) a. Siat-sú Tsuí-sūn sī huān-sè bîn-á-tsài beh khì Bí-kok, lán if Tsuisun BE perhaps tomorrow will go U.S.A. we.INC
設使 水順 是 凡勢 明仔載 欲 去 美國， 咱
ing-kai tsá tō thiann-kìnn i kóng --ah. (Taiwanese) should early PRT hear he say PRT
應該 早 就 聽見 伊 講 矣。
“Assume that we know it is the case that Tsuisun might go to U.S.A. tomorrow, we should have been told by him.”
b. Tsuí-sūn huān-sè bîn-á-tsài beh khì Bí-kok. Tsuisun perhaps tomorrow will go U.S.A. 水順 凡勢 明仔載 欲 去 美國。 “Perhaps Tsuisun will go to U.S.A. tomorrow.”
(30) a. Ká-sú Tsuí-sūn sī bîng-bîng / hó-ka-tsài / láu-st-kóng ū tāi-sing if Tsuisun BE evidently / fortunately / frankly have in.advance 假使 水順 是 明明 / 好佳哉 / 老實講 有 代先 tsún-pī, án-ne Gîn-khuân bô tsún-pī suah khó
prepare then Ginkhuan no prepare unexpectedly take.an.exam
準備， 按呢 銀環 無 準備 煞 考
ū kuè tō tsin kuài-kî --ah. (Taiwanese) have pass EMPH true strange PRT
有 過 就 真 怪奇 矣。
“If it is evident / fortunate / obvious that Tsuisun prepared for the exam in advance, then it’s strange to find that Ginkhuan passed the exam without preparation.”
b. Tsuí-sūn bîng-bîng / hó-ka-tsài / láu-st-kóng ū tāi-sing tsún-pī. Tsuisun evidently / fortunately / frankly have in.advance prepare 水順 明明 / 好佳哉 / 老實講 有 代先 準備。
8 Thanks to Ching-yu Yang for pointing out that conditionals composed by nā-sī “if”, contrary to siat-sú and ká-sú
employed in the test, cannot accommodate a speaker-oriented adverb(ial) in its antecedent. It is likely due to the different antecedent syntactic structures of these conditionals (refer to the distinction between event-conditionals and premise-conditionals in Haegeman (2003). Note that the projection of presupposition is not effected by the syntax issue mentioned here for the antecedents of the conditionals employed in this test have no problem to contain a speaker-oriented adverb(ial).
“Evidently / fortunately / frankly Tsuisun prepared for it in advance.”
Now let’s turn to presupposition projection in a question. Recall the intuition from my informants about this sī in a question: It is employed when an inquirer has already known (or believed) that the event in question did happen and he is curious about the details. This intuition is confirmed by the test devised in (31).
(31) CONTEXT: In a village, a woman meets her neighbor Madam A-ho, who is obviously going to the evening market. And they have a short conversation. (a) contains the sentence used by the woman to greet Madam A-ho. This greeting sentence can be followed on by an additional question. The point is about the felicity of the following-on questions in (b-1) and (b-2).
a. A-hó-tsím-à, lí khì hông-hun-tshī-á bé tsiah tsē tshài
Aho.madam you go evening.market buy this many cooking.material 阿好嬸仔， 你 去 黃昏市仔 買 遮 濟 菜
--ooh! Ū hî, ū ìng-tshài, ū kuàn-á-tsî... (Taiwanese)
PRT have fish have water.spinach have Chinese.yam... 喔！ 有 魚， 有 蕹菜 有 罐仔薯……
“Madam Aho, it seems you bought so many things in the evening market! Let me see. You’ve got fish, water spinach, and Chinese yams...”
b-1. Lín Tsuí-sūn sī àm-tǹg siūnn-beh tsih siánn? your Tsuisun BE dinner want eat what
恁 水順 是 暗頓 想欲 食 啥？
“(Based on what I see, I speculate Tsuisun wants to eat something.) What does your husband Tsuisun want to eat for dinner?”
b-2.# Lín Tsuí-sūn àm-tǹg siūnn-beh tsih siánn? your Tsuisun dinner want eat what # 恁 水順 暗頓 想欲 食 啥？ “What does your husband Tsuisun want to eat for dinner?”
Compared with (31b-2), which has a disconnected sense in the conversation, (31b-1) is a more felicitous question to follow up the utterance in (31a). Aside from the presupposition in a wh-question, which is commonly assumed in the literature, the presence of the high sī gives rise to an implicature that the inquirer strongly assumes that both parties in the conversation know that Tsuisun does want to eat some specific thing this evening (refer to Romero & Han,
2004).
Lastly, in the vein of Lee (2005), we should not preclude the possibility that this non-copular usage of sī is some kind of focus marker. Nonetheless, unlike the focus markers, including sī / shì used for subject-focus, adjunct-focus, or predicate-focus, which cannot show up simultaneously in a sentence (see (32); as pointed out in Lee, 2005: 99), the high sī has no problem to co-occur with a focus marker, except for a subject-focus marker (see (33)).
(32) a.* Sī A-bîng sī bîn-á-tsài beh khì Tâi-pak. (Taiwanese)
BE Abing BE tomorrow will go Taipei
*是 阿明 是 明仔載 欲 去 台北。 (*subject-FOC > adjunct-FOC) (Intended) “It is Abing who will go to Taipei and it’s tomorrow.”
b.* Sī A-bîng bîn-á-tsài sī beh khì Tâi-pak.
BE Abing tomorrow BE will go Taipei
*是 阿明 明仔載 是 欲 去 台北。 (*subject-FOC > predicate-FOC) (Intended) “It is Abing who will go to Taipei tomorrow and it’s true.”
c.* A-bîng sī bîn-á-tsài sī beh khì Tâi-pak. Abing BE tomorrow BE will go Taipei
*阿明 是 明仔載 是 欲 去 台北。 (*adjunct-FOC > predicate-FOC) (Intended) “It is tomorrow that Abing will go to Taipei and it’s true.”
(33) a.* Sī A-bîng sī huān-sè beh khì Tâi-pak. (*subject-FOC > CO-sī9)
BE Abing BE perhaps will go Taipei
*是 阿明 是 凡勢 欲 去 台北。 (Taiwanese)
(Intended) “It is Abing who will go to Taipei and (we know) it might be the case.” b. A-bîng sī huān-sè sī bîn-á-tsài beh khì Tâi-pak.
A-bing BE perhaps BE tomorrow will go Taipei
阿明 是 凡勢 是 明仔載 欲 去 台北。(CO-sī > adjunct-FOC) “(We know that) it might be the case that it’s tomorrow that Abing will go toTaipei.” c. A-bîng sī huān-sè bîn-á-tsài sī beh khì Tâi-pak.
A-bing BE perhaps tomorrow BE will go Taipei
阿明 是 凡勢 明仔載 是 欲 去 台北。(CO-sī > predicate-FOC) “(We know that) it might be the case that Abing in fact will go to Taipei tomorrow.”
The contrast between (32) and (33) suggests that the high sī differs from its homonymous counterparts as focus markers. As for the ungrammatical (33a), it can be accounted for by saying that the high sī is syntactically higher than FocusP (which accommodates subject-focus and adjunct-focus in Lee (2005); also refer to É Kiss (1998, 1999)). It is common to observe that the wrong hierarchical relation causes ungrammaticality, for example, no volitional or deontic modal can syntactically dominate an epistemic.
By explicating the high sī with a non-focus denotation, we can explicate its being able to co-occur with a focus marker in a sentence, contrary to the non-co-occurrence of two sī / shì both as focus markers. This high sī differs from the subject-focus, adjunct-focus, and predicate-focus sī in that it does not convey contrastive focus reading. Consequently, it is exempt from the non-co-occurrence restriction of non-copular sī and has no problem to co-occur with other non-copular sī. (Except subject-FOC;due to scope / position reason, see above.)10
In addition to the contrastiveness issue, one should keep in mind the following. Compared with shì in (34), which juxtaposes and contrasts two sentences with clausal-initial shì, the high sī under investigation cannot do without a preceding noun. Hence, this high sī cannot be a variant of the clausal contrastive focus marker shì in (34), which takes the whole clause under its scope.
(34) Shì tā lái zhǎo wǒ, bú shì wǒ qù zhǎo tā.
BE he come look.for me not BE I go look.for him 是 他 來 找 我， 不 是 我 去 找 他。 “He came to see me, not I went to see him.” (MC; Cheng, 2008: 256 (46))
To sum up, the high sī in question is not a presupposition marker. Additionally, it is not a focus marker, either. This is not only shown by the fact that MC focus markers shì never precede an epistemic (see section 2), but also illustrated by the contrast between (32) and (33). We also learned in this section that the element we look into herein is higher than the epistemics, the evidentials, the evaluatives, and the speech-act adverbials. Additionally, it cannot occur in a null (out-of-the-blue) context.
10 Multiple focus is possible with some other focus items, like zhǐ (MC; “only”) and in cases of multiple
wh-elements. At this moment, I do not know why sī / shì behaves differently from those focus items. However, it is empirically obvious that focus markers sī / shì never occurs more than once in a sentence.
6. Analysis
Based on what have been shown previously, if the element under investigation has nothing to do with presupposition, and it is not a focus marker, then what is it? To answer the question, I will briefly introduce the adopted methodology in 6.1 before proposing the analysis in 6.2.
6.1. Theoretical Background
The framework adopted in this study is from Heim and Kratzer (1998), which is based on Frege’s (1923) insight of compositionality of language. Frege suggests that semantic composition always consist in the saturation of an unsaturated meaning component. In Frege’s words:
Statements in general, just like equations or inequalities or expressions in Analysis, can be imagined to be split up into two parts; one complete in itself, and the other in need of supplementation, or “unsaturated.” Thus, e.g., we split the sentence “Caesar conquered Gaul” into “Caesar” and “conquered Gaul.” The second part is “unsaturated” - it contains an empty place; only when this place is filled up with a proper name, or with an expression that replaces a proper name, does a complete sense appear. Here too I give the name “function” to what this “unsaturated” part stands for. In this case the argument is Caesar. (Frege, 1891; translated in Geach & Black, 1980: 31)
As the predominant fashion in modern formal semantics, the computation is executed based on semantic types. Basic semantic types include <e>, the type of individuals, <t>, the type of truth-values, and <s>, the type of possible worlds. By combining basic types, we then have some more types. For example: <e,t> type, a type applies to an argument of <e> type. The typical instances of <e,t> type are intransitive verbs, which is saturated by an individual (an <e> type element).
Although some other compositional rules have been proposed under this framework, the aforementioned simple function application that combines each pair of the sister nodes on the syntactic diagram is sufficient for us to come up with an analysis in this study. Whenever the composition is done on a sentence, we would obtain a truth value (<t> type), which is either 1
(true) or 0 (false). And the result of computing a sentence, therefore, would be truth-conditions under which the sentence would be true.
In addition, the proposal presented in the following employs world variables / arguments, which are represented with w. World variables are assumed covert elements. By inserting different world variables, it becomes possible for us to evaluate different parts of a sentence with different possible worlds. For world variables that are unspecified / unbound, we assume them to be world of evaluation.
When computing, we rely on the λ-notation. Here is its general schema (refer to Heim & Kratzer, 1998: 34-35):
(35) [λα<ϕ>. γ]
α is the argument variable, φ the domain condition, and γ the value description.
For example, the formula [λxe. e is a student] denotes a set of individuals and each of
them is a student.
The last mechanism employed in the analysis that follows is lambda abstraction (refer to Heim & Kratzer, 1998: 96):
(36) If α is a branching node whose daughters are βi and γ, where β is a relative pronoun and
i∈ |N, then for any variable assignment function g, ⟦α⟧g = λx∈De.⟦γ⟧g x/i.
In practices, lambda abstraction is not only applied in relative clauses. The analysis provided in 6.2 uses lambda abstraction to bring in an additional <s> argument to dissolve the type-mismatch.
All in all, one of the main issues that the semanticists pursue is the denotations of lexical elements, especially the function words. And this is also the goal of this investigation, in which the function of a special usage of sī is enquired about.
6.2. A Proposal Based on Romero and Han (2004)
observations and proposal are elucidating in explaining the seemingly mysterious element targeted in this study.
According to Romero and Han, English yes / no-questions with preposed negation like “Doesn’t John drink?” necessarily carry the implicature that the speaker thinks John drinks (cf. non-preposed negation yes / no-questions). They argue that it is the presence of an epistemic conversational operator VERUM that derives the existence and content of the implicature.
They further claim that, in English, the implicature / epistemic bias can be spelled out with epistemic adverb really (2004: 624). Romero and Han suggest that inherently focused
REALLY triggers an epistemic bias of the opposite polarity and adds the epistemic (negative) implicature that the speaker believed or expected that the (negative) answer is true. And they designate this operator with the term VERUM.11
It is worth noting that VERUM is not a purely epistemic operator (Romero & Han, 2004: 626). It isn’t used to assert that the speaker is entirely certain about the truth of p, but to assert that the speaker is certain that p should be added to the Common Ground (CG). In other words, the operator is a conversational epistemic operator (Romero & Han, 2004: 627).12
Romero and Han define it in the following.
(37) ⟦VERUMi⟧g x/i = ⟦reallyi⟧ g x/i = λp <s,t>λw.∀ w’ ∈ Epix (w) [∀ w’’ ∈ Convx (w’) [p ∈
CGw’’]] = FOR-SURE-CGx (2004: 627 (43))
Epix (w) is the set of worlds that conform to x’s knowledge in w.
Convx (w’) is the set of worlds where all the conversational goals of x in w’ are fulfilled
and where CGw’’ is the Common Ground or set of propositions that the speakers assume
in w’’ to be true.
As shown in (37), there is no focus involved in it. And, therefore, we do not have to deem this operator a real focus marker and the definition here does not go against our observation that high sī has nothing to do with focus.
11 Romero and Han compare this operator to Hӧhle’s (1992) VERUM. Although Hӧhle claims it’s of focus, the
definition of the operator given by Romero and Han does not involve the notion of focus (see (37)). Recall in section 5 it is suggested that the high sī is not a focus marker.
Following Romero and Han, I propose that the high sī is a realization of the conversational operator in Taiwanese.13 And we can see this by applying (37) to an exemplifying sentence below.
(38) a. A-bîng sī huān-sè bat khì Pîn-tong kuè. (Taiwanese) Abing BE maybe ASP go Pingtung ASP
阿明 是 凡勢 捌 去 屏東 過。 (We know that) “Abing might go to Pingtung before.” b. LF: [CP Opc [IP perhaps Abing has been in Pingtung]]
c.
d. ⟦CP⟧
= λw.∀ w’ ∈ Epix (w) [∀ w’’ ∈ Convx (w’) [λw’’’. perhaps-has-been-in (a, p, w’’’)
∈ CGw’’]]14
=1 iff for all the worlds that conform to the speaker’s knowledge in the world of evaluation, all the worlds in the Common Ground are among the worlds where all the conversational goals of the speaker are fulfilled, and all the worlds where all the conversational goals of the speaker are fulfilled are among the worlds that conform to the speaker’s knowledge in the world of evaluation, the proposition “perhaps Abing has been in Pingtung” is in the worlds among the worlds in the common ground =1 iff it is for sure that we should add to CG that perhaps Abing has been in Pingtung
This proposal can be further evidenced by comparing the high sī with the “really” words in Taiwanese. Unlike English “really,” which can denote FOR-SURE-CGx as defined in (37)
13 Since this sī has no contrastive reading, I do not adopt the term verum, in order to avoid conceptual confusion.
Moreover, in order not to have it be confused with epistemics, I also discard the word epistemic and call it simply conversational operator.
14 The details of the epistemic, tense and aspect are ignored.
CPt perhaps-has-been-in sī <<s,t>,s,t> Abing e w’’’1 s Pingtung e <s,<e,<e,t>>> <e,<e,t>> 1 t <s,t> ws <s,t> <e,t>
(refer to Romero & Han, 2004), no “really” word in Taiwanese is parallel to the high sī. See the “really” words including ū-iánn, tsin-tsiànn, tsiânn-st, st-tsāi, khak-st and tik-khak in the following example, none of them can precede the epistemics as the high sī does.
(39) a.* Tsuí-sūn ū-iánn / tsin-tsiànn / tsiânn-st / st-tsāi / khak-st / tik-khak huān-sè
Tsuisun really perhaps
*水順 有影 / 真正 / 誠實 / 實在 / 確實 / 的確 凡勢
beh khì Bí-kok. (Taiwanese) will go U.S.A.
欲 去 美國。
(Intended) “Perhaps Tsuisun indeed will go to U.S.A.”
b. Tsuí-sūn huān-sè ū-iánn / tsin-tsiànn / tsiânn-st / *st-tsāi / *khak-st / Tsuisun perhaps really
水順 凡勢 有影 / 真正 / 誠實 / *實在 / *確實 / *tik-khak beh khì Bí-kok.
will go U.S.A.
*的確 欲 去 美國。
“Perhaps it’s true that Tsuisun will go to U.S.A.”
Moreover, the grammatical contrast in (39b) agrees with the observation in Romero and Han (2004) that there are different kinds of “really” (2004: 624-625, see especially fn.11).
Regarding the term VERUM, note that Lee (2005) suggests the predicate-adjacent shì in MC is a VERUM focus marker (which is dubbed predicate-focus). As I have demonstrated previously, the high sī behaves differently from the predicate-adjacent sī / shì. Following Romero and Han (2004) in distinguishing different kinds of “really” in English, here I further propose that high sī and predicate-focus shì / sī are realizations of different kinds of English “really” in Sinitic languages.15
Apart from other kinds of “really” words, Romero and Han suggest the following test to distinguish be-sure “really” from “really” of FOR-SURE-CG.16 Parallel examples in Taiwanese
15 Regarding the so-called predicate-adjacent sī / shì, we may ask another question: Does it pass the tests with
respect to presupposition? A quick test seems to indicate that it passes wait-a-minute test but fails to project in the embedding tests of negation and the antecedent of a conditional. So far I have no conclusion on this issue.
are given in (41).
(40) a.? I am sure I am tired. (Romero & Han, 2004: 626 (41)) b. I really am tired.
(41) a.? Guá st-tsāi / khak-st thiám --ah. (Taiwanese) I really tired ASP?
我 實在 / 確實 忝 矣。 (Intended) “I am really tired.”
b.* Guá sī thiám --ah. (no stress on sī) I BE tired ASP
*我 是 忝 矣。 (Intended) “I am really tired.”
c. Guá ū-iánn / tsin-tsiànn / tsiânn-st thiám --ah. I really tired ASP
我 有影 / 真正 / 誠實 忝 矣。 “I am really tired.”
Romero and Han point out that the be-sure “really” in (40a) asserts certainty about the speaker’s own inner sensations. And the sentence is a bit odd (as if the speaker could be confused about that). On the other, they suggest that (40b), instead, is perfectly fine, and the presence of “really” simply emphasizes or insists that the addressee should take the proposition as true. In contrast to (41c), st-tsāi / khak-st in (41a) and the unaccented predicate-adjacent sī in (41b) correspond to the “really” of be-sure in (40a), defined in Romero and Han (2004: 626) (42), which is reduplicated below.
(42) ⟦be sure⟧ = ⟦tik-khak⟧ = λp <s,t>λw∀ w’ ∈ Epix (w) [p (w’) =1]
As for ū-iánn, tsin-tsiànn, and accented predicate-adjacent sī, I propose that they are of in-actuality reading (refer to Romero & Han, 2004: 624 fn.11). Against Romero and Han (2004) (they liken (37) to VERUM focus in Hӧhle, 1992; see their section 3.3), I suggest that it
is these in-actuality elements that are of polarity focus / VERUM focus, but not the FOR-SURE
conversational operator. These in-actuality elements are realizations of the operator suggested for declaratives in regard of focal stress in Hӧhle (1992).
Hinterwimmer, 2011 (33)), the in-actuality of polarity focus is defined as follows:17
(43)∀ r[r∈R ∧ r ≠ p → false(r)]
R is the focus semantic value of a sentence S.
The denotation indicates that the propositional operator is applied to the ordinary semantic value p, an operator emphasizing on the truth-value of the proposition.
In this line of reasoning, I suggest that the so-called predicate-focus / VERUM focus in Lee (2005) should be further categorized into be-sure“really” and in-actuality of polarity focus (defined in (42) and (43) respectively). The proposed categorization is summarized as follows.
Item Status the high sī a FOR-SURE “really” conversational operator
unstressed predicate-adjacent sī / shì and tik-khak
a be-sure “really” operator
stressed predicate-adjacent sī / shì and ū-iánn, tsin-tsiànn, tsiânn-st, st-tsāi
an in-actuality of polarity focus marker (i.e. VERUM focus in Hӧhle, 1992)
Now let’s turn to the cases where the conversational operator high sī (henceforth
CO-operator) occurs in a question.18
17 In Taiwanese and MC, only (Taiwanese: kan-tann; tk-tk; MC: zhǐyǒu) can be reiterated in a sentence. However,
in a single sentence, when sī and shì are used as focus markers, iteration is not allowed. The prohibition of multi-foci seems to apply on contrastiveness but not exhaustivity, and focus elements are probably not homogeneous and should not be analyzed in the same way. For comparing zhǐ and shì, also refer to Lee (2005: 89-94); Lee also mentioned that sentences with multiple focuses are not accepted if the focuses are marked by the focus marker shì (2005: 98-99).
18 I will not discuss yes-no questions in this paper. The reason is that those questions with high sī give rise to an
additional rhetorical reading (e.g., (i)), which is quite different from what we observe in wh-questions. The sī involved in these questions probably differs from what we see in this paper. The analysis in Han 2002 is a possible way to approach this sī.
(i) Lí sī bîn-á-tsài ū beh lâi bô? (Taiwanese) you BE tomorrow HAVE will come Q
你 是 明仔載 有 欲 來 無？ “You will come tomorrow, won’t you?”
Compare (i) with the counterpart sentences in MC:
(ii) Nǐ shì míngtiān yào lái ma? (MC; yes-no question) you BE tomorrow will come Q
你 是 明天 要 來 嗎？ “Is it tomorrow that you will come?”
Assume the partitional approach for questions. The semantic computation and partition of an example of questions including a CO-operator are given as follows. The denotation of Opc is
from (37).
(44) a. Tsuí-sūn sī tú-tsiah tú-tih siánn-lâng? (Taiwanese) Tsuisun BE a.moment.ago encounter-ASP who
水順 是 拄才 拄著 啥人？
“We know that Tsuisun ran into someone a moment ago. Who is that guy?” b. LF: λjs.λis. [Opcλxe [personi(x) ∧ encounter-a-moment-agoi(x)(Tsuisun)]
= Opcλxe [personj(x) ∧ encounter-a-moment-agoj(x)(Tsuisun)]]
c. ⟦(44a)⟧ = {“it is for sure that we should add to CG that Tsuisun encountered a a moment ago”, “it is for sure that we should add to CG that Tsuisun encountered b a moment ago”, “it is for sure that we should add to CG that Tsuisun encountered c a moment ago”,…}
This analysis captures the intuition that the question inquirer has already known (or believed) that the event in question did happen (being for sure that it should be added to Common Ground) and he is curious about the details.
7. Concluding
Remarks
Now we are in a position to answer the questions raised in the end of section 4. Regarding the high sī in question, it is a conversational operator which denotes an implicature from the speaker that it is for sure the content of the proposition should be added to Common Ground. This sī is speaker-oriented and conversational; and it has wider scope than other non-copular sī, including the predicate-adjacent sī (either functions as “really” of be-sure or a polarity focus marker on the truth-value). By distinguishing it from the focus markers, we account for the acceptability of its co-occurrence with either adjunct-focus marker or predicate-focus marker, in contrary to the ungrammaticality of co-occurrence of any two sī / shì as focus markers in a single sentence.
(iii) * Nǐ shì míngtiān yào-bú-yào lái? (MC; A-not-A question) you BE tomorrow will-NEG-will come
*你 是 明天 要不要 來？ (Intended) “Will you come tomorrow?”
Additionally, the interpretation of this high sī as a conversational marker also explains why, unlike focus marker sī / shì, no intervention effect is observed in questions containing it. Based on the prevalent assumption that question forming involves focus operation, the intervention effect then is only expected if there’s another focus marker occurs in the interrogative.
Last but not least, we should note that the interpretation and function of either the conversational marker or a focus marker depends on the context heavily. Without a proper context, in a simple sentence, it may become difficult to tell the conversational marker apart from a focus marker. Nevertheless, this should not prevent us from telling them apart.
References
Aikhenvald, A. Y. (2004). Evidentiality. Oxford: Oxford University Press.
Chao, Yuen Ren. (1968). A Grammar of Spoken Chinese. Berkeley: University of California Press.
Cheng, Lisa Lai-shen. (2008). Deconstructing the shì...de construction. The Linguistic Review, 25, 235-266.
Cheng, Lisa Lai-shen & Rooryck, J. (2002). Types of Wh-in-situ. Leiden, Netherlands: Leiden University. Unpublished manuscript.
É. Kiss, Katalin. (1998). Identificational Focus versus Information Focus. Language, 74(2), 245-273.
É. Kiss, Katalin. (1999). The English Cleft Construction as a Focus Phrase. In Lunella Mereu (Ed.), Boundaries of Morphology and Syntax (pp. 217-229). Amsterdam, Netherlands: John Benjamins Publishing Company.
Ernst, Thomas. (2009). Speaker-Oriented Adverbs. Natural Language & Linguistic Theory, 27, 497-544.
Frege, Gottlob. (1891). Funktion und Begriff [Function and Concept]. Lecture presented at the Jenaischen Gesellschaft für Medizin und Naturwissenschaft [Jenaic Society for Medicine and Science]. Jena, Germany. January 9.
Frege, Gottlob. (1923). Logische Untersuchungen, Dritter Teil: Gedankengefüge [Logical Investigations, Third Part: Thought Structure]. Beiträge zur Philosophie des deutschen Idealismus [Contributions to the Philosophy of German Idealism], 3, 36-51.
Gao, Ming-kai (高明凱). (1970). Guóyǔ Yǔ Fǎ [Mandarin Grammar]. Taipei: Letian Publishing (樂天出版社).
Geach, P. & Black, M. (1980). Translations from the Philosophical Writings of Gottlob Frege (3rd ed.). Oxford: Blackwell.
Giannakidou, Anastasia. (1999). Affective Dependencies. Linguistics and Philosophy, 22, 367-421.
Haegeman, Liliane. (2003). Conditional Clauses: External and Internal Syntax. Mind & Language, 18(4), 317-339.
Halliday, M. (1970). Functional Diversity in Language as Seen from a Consideration of Modality and Mood in English. Foundation of Language, 6, 322-361.
Han, Chung-hye. (2002). Interpreting Interrogatives as Rhetorical Questions. Lingua, 112, 201-229.
Heim, Irene & Kratzer, Angelika. (1998). Semantics in Generative Grammar. Hoboken, NJ: Wiley-Blackwell.
Hinterwimmer, Stefan. (2012). Information Structure and Truth Conditional Semantics. In C. Maienborn, K. von Heusinger & P. Portner (Eds.), Semantics: An International Handbook of Natural Language Meaning, Vol. 2 (pp. 1875-1907). Berlin & New York: Mouton de Gruyter.
Hӧhle, Tilman. (1992). Über Verum-Fokus im Deutschen [About Verum-Focus in German]. In J. Jacobs (Ed.), Informationsstruktur und Grammatik [Information Structure and Grammar] (pp. 112-141). Opladen: Westdeutscher Verlag.
Hsu, Shi-ying (許世瑛). (1973). Zhōngguó Wénfǎ Jiǎnghuà [Remarks on Chinese Grammar]. Taipei: Taiwan Kaiming Shudian (台灣開明書店).
Huang, C. -T. James. (1988[1990]). Shuō “Shì” hàn “Yǒu” [Say “Shì” and “Yǒu”]. Bulletin of the Institute of History and Philology Academia Sinica, 59(1), 43-64.
Jackendoff, Ray. (1972). Semantic Interpretation in Generative Grammar. Cambridge, MA: MIT Press.
Lee, Hui-chi. (2005). On Chinese Focus and Cleft Constructions. Doctoral dissertation, National Tsing Hua University, Hsinchu, Taiwan.
Li, Jinxi (黎錦熙). (1925). Xīn Zhù Guóyǔ Wénfǎ [New Mandarin Grammar]. Shanghai: The Commercial Press (商務印書館).
Li, Charles N. & Thompson, Sandra A. (1981). Mandarin Chinese: A Functional Reference Grammar. Berkeley, CA: University of California Press.
Romero, Maribel & Han, Chung-hye. (2004). On Negative Yes / No Questions. Linguistics and Philosophy, 27, 609-658.
Rooth, Mats. (1985). Association with Focus. Doctoral dissertation, University of Massachusetts, Amherst, USA.
Searle, John. (1969). Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press.
Searle, John. (1983). Intentionality. Cambridge: Cambridge University Press.
Soh, Hui-ling. (2005). Wh-in-situ in Mandarin Chinese. Linguistic Inquiry, 36, 143-155. Tang, Ting-chi (湯廷池). (1979). Guóyǔ de “Shì”-Zì Jù [Shì Sentences in Mandarin Chinese].
Guóyǔ Yǔfǎ Yánjiù Lùnjí [On the Study of Mandarin Grammar] (pp. 133-142). Taipei: Taiwan Student Publishing Company (台灣學生書局).
Tsai, W. -T. Dylan. (2008). Left Periphery and How-Why Alternations. Journal of East Asian Linguistics, 17(2), 83-115.
Tsai, W. -T. Dylan (蔡維天). (2012). Lùn Jùdiào Zhòngyīn tuì Yǔfǎ Quánshì Jīzhì de Yǐngxiǎng [On the Influences from Intonation and Stress on the Syntactic Interpretation]. Hsinchu, Taiwan: National Tsing Hua University. Unpublished manuscript.
von Fintel, Kai. (2004). Would You Believe It? The King of France Is Back! Presuppositions and Truth-Value Intuitions. In Marga Reimer & Anne Bezuidenhout (Eds.), Descriptions and Beyond (pp. 315-341). Oxford: Oxford University Press.
von Prince, Kilu. (2012). Predication and Information Structure in Mandarin Chinese. Journal of East Asian Linguistics, 21(4), 329-366.
Wang, Li (王力). (1937). Zhōngguó Wénfǎ Zhōng de Xìcí [The Copula in Chinese Grammar]. Qīnghuá Xuébáo [Tsing Hua Journal of Chinese Studies], 12(1), 1-67.
Wang, Li (王力). (1954). Zhōngguó Yǔfǎ Lǐlùn [The Theories of Chinese Grammar]. Beijing: Chung Hwa Book Company (中華書局).
Yang, Barry. (2008). Intervention Effects and Covert Component of Grammar. Doctoral dissertation, National Tsing Hua University, Hsinchu, Taiwan.
台灣話的「是」做為共知背景標記
劉承賢
國立清華大學語言學研究所博士摘 要
本文以台灣話的「是」為目標，特別是在結構上高於言者中心狀語（speaker-oriented adverb(ial)），且於疑問詞問句當中無需鄰接輕動詞詞組（vP）的用法；基於其位置，此 一用法實有別於其他非繫詞用法（non-copular homonyms）（Lee, 2005）。文中提議可由 「FOR-SURE really」的語意做為基礎來進行分析，根據 Romero 與 Han（2004）的主張，「FOR-SURE really」的語意為「我們應該把 p 命題加入共知背景（Common Ground）裡」。 本文的分析不僅可解釋該用法與其他非繫詞「是」的用法在語言現象上的差異，同時也 明確指出漢語當中語用標記的存在，而這也佐證了「是」的高度語法化，其同源成分在 今日的其他漢語裡頭，也於功能及句法分布上有了相當分歧的演化。
