Clause-­‐medial par.cles and stance marking in Cantonese
Foong Ha YAP a, Tak-­‐sum WONG a and Winnie Oi-­‐wan CHOR b
The Hong Kong Polytechnic University a and The Open University of Hong Kong b
IntroducAon •  Speakers have various linguisAc resources at their disposal to express their stance, e.g. their feelings, viewpoints, evaluaAons and aJtude. •  It is well known that Cantonese has a rich inventory of sentence final parAcles to express speaker stance •  (see Cheung 1972; Kwok 1984; Law 1990; MaYhews & Yip 1994; Fung 2000; Leung 2005; Sybesma & Boya 2007; inter alia). •  More recently, aYenAon has also been directed to clause-­‐medial parAcles in Cantonese. 2
Clause-­‐medial stance parAcles (< direcAonal parAcles) •  Post-­‐predicate direc.onal par.cles –  Chor (2010; 2013) recently examined the development of direcAonal parAcles into markers of the speaker’s aJtude. •  faan > posiAve –  Cung1 faan1 go3 loeng4 sin1! (‘posiAve evaluaAon related to nostalgic experiences’) •  maai > negaAve aJtude marker –  Gong2 maai4 saai3 di1 gam2 ge3 je5!!! (‘inclusion of the outsider’) 3
Clause-­‐medial stance parAcles (< complement-­‐taking predicates) •  Reanalysis of complement-­‐taking verbs as stance adverbials –  mzi ‘don’t know’ > marker of the speaker’s negaAve aJtude (Wong, Yang & Yap 2011) •  Verb > LP > ParentheAcal aJtudinal marker (preverbal posiAon) •  Verb > LP > RP aJtudinal uYerance tag (infrequent use) – 
– 
– 
– 
Ngo5 m4zi1 nei5 gam3 do1 aa3!!! M4zi1 (keoi5) zou6 gan5 mat1 gwai2 je5! Keoi5 dou1 m4zi1 zou6 gan5 mat1 gwai2 je5! Keoi5 heoi5 zo2 bin1, dou1 m4zi1! 4
Clause-­‐medial stance parAcles (< complement-­‐taking predicates) •  Reanalysis of complement-­‐taking verbs as stance adverbials –  paa ‘(I) fear/worry’ > epistemic marker > pragmaAc hedger (Yap & Chor 2014; see also Endo 2005 on Mandarin pa ‘fear’) •  Verb > LP epistemic adverbial > ParentheAcal epistemic marker (preverbal posiAon) •  Verb > LP epistemic adverbial > RP uYerance tag (very infrequent use; no token in our database) –  Ngo5 paa3 keoi5 dou1 hai1 m4lai4 ga3laa3! –  Keoi5 paa3 dou1 hai1 m4lai4 ga3laa3! –  Keoi5 dou1 hai1 m4lai4 ga3laa3, paa3! –  EvidenAal + epistemic + aJtudinal taipaa •  SubjecAve use (more frequent among the younger generaAon) •  LP stance adverbial > ParentheAcal stance marker (preverbal posiAon) •  LP stance adverbial > RP uYerance tag –  Ngo5 tai2paa3 keoi5 dou1 hai1 m4lai4 ga3laa3! –  Keoi5 tai2paa3 dou1 hai1 m4lai4 ga3laa3! –  Keoi5 dou1 hai1 m4lai4 ga3laa3, tai2paa3!
5
ObjecAves of this talk •  Previous studies have focused on tracing the diachronic development of these clause-­‐medial stance parAcles. •  In this presentaAon, we will focus on the subjecAve and intersubjecAve pragmaAc funcAons of these markers in relaAon to their syntacAc posiAons. •  LP •  ParentheAcal (clause-­‐medial) •  RP 6
Previous studies: complement-­‐taking verb construcAon > stance adverbial
•  Matrix clauses containing complement-­‐taking verbs such as ‘I think/feel’ or ‘(I) fear’ ojen develop into subjecAve and intersubjecAve stance markers with epistemic readings such as ‘probably, perhaps’. –  English I think –  Mandarin wo juede –  Mandarin kongpa (Thompson & Mulac 1991; Kärkkäinen 2003) (Endo 2011; Lim 2011) (Yap, Chor & Wang 2012) 7
Previous studies: from LP > RP and from LP > parentheAcal
•  As adjuncts at the lej periphery (LP), these epistemic stance markers ojen subsequently extend into other syntacAc posiAons—someAmes as epistemic tags at the right periphery (RP), and someAmes as epistemic parentheAcals in clause-­‐medial posiAon (CMP). • 
• 
• 
• 
I think she’s leaving Paris for good. She, I think, is leaving Paris for good. She’s leaving Paris, I think for good. She’s leaving Paris for good, I think.
8
How are clause-­‐medial stance parAcles formed?
•  Strategies for the emergence of clause-­‐medial stance markers (Yap & Chor 2014): –  Derived from postverbal direcAonal parAcles •  Verb serializaAon > suffixaAon –  Derived from complement-­‐taking verb structures •  DisjuncAon > insubordinaAon of complement clause & reanalysis of complement-­‐taking verb structure as stance adverbial at LP •  Right dislocaAon of stance adverbials > stance tags at RP > (SFPs)
•  ParentheAcal inserAon of stance adverbials > clause-­‐medial stance markers 9
Issues to be addressed
•  How do complement-­‐taking verbs grammaAcalize into stance adverbials at the lej periphery and then further develop into parentheAcal stance markers in pre-­‐predicate clause-­‐medial posiAon? •  What types of stance markers tend to occupy clause-­‐medial as opposed to sentence final posiAon? •  How do stance markers in clause-­‐medial posiAon differ in funcAon from those at the lej or right periphery (e.g. discourse parAcles and sentence final parAcles)? •  Are clause-­‐medial stance markers oriented more toward pragmaAc funcAons that are subjecAve as opposed to intersubjecAve in nature?
10
Source of Data
•  Early Cantonese Colloquial Texts: A Database •  Early Cantonese Tagged Database •  A LinguisAc Corpus of Mid-­‐20th Century Hong Kong Cantonese •  Hong Kong University Cantonese Corpus (HKUCC) •  PolyU Corpus of Spoken Chinese
•  Google 11
Distribution on position and function of ngo gokdak Function Position of ngo gokdak Clause-­‐initial Clause-­‐medial Lexical Epistemic 0 0 16 0 Σ 16 1 Google search: 50 tokens of 我覺得, with 16 Cantonese tokens; the
rest appear to be Mandarin or standard written Chinese. Distribution on position and function of ngo gokdak in mid-­‐20th century Cantonese Function Position of ngo gokdak Clause-­‐initial Clause-­‐medial Lexical -­‐ physical -­‐ psychological 3 3 0 Epistemic 14 0 Σ 20 1 A LinguisVc Corpus of Mid-­‐20th Century Hong Kong Cantonese Movies Clause-­‐iniAal Ngo gokdak 我覺得
(1)
14
Clause-­‐iniAal Ngo gokdak 我覺得
15
Clause-­‐iniAal Ngo gokdak 我覺得
16
Clause-­‐iniAal Ngo gokdak 我覺得
17
Ngo gokdak 我覺得
18
Distribution on position and function of taipaa Function Position of taipaa Clause-­‐initial Clause-­‐medial Worry~inferential~probably Worry~probably 4 0 4 0 Worry~inferential Probably~inferential 5 2 8 2 Probably 12 3 Inferential 5 1 Hedger Unknown 2 1 1 0 Σ 41 9 13% Clause-­‐medial parAcles and stance marking in Cantonese
20
94% 21
24% 22
Paa 怕
23
Ngo paa 我怕
24
Ngo tai … paa 我睇…怕
25
DistribuAon on form and funcAon of mzi Position of mzi Form/ function Total no. of tokens Clause-­‐initial (with subject) Clause-­‐initial (without subject) Clause-­‐medial (with 1SG subject) Parenthetical Clause-­‐medial (without 1SG) Parenthetical Lexical verb: -­‐ 1P referent -­‐ 2P referent -­‐ 3P referent 81 59 17 102 3
0
0
262 zi-­‐m-­‐zi -­‐ with 2P subject -­‐ With 3P subject 50? ?
0
0
50 Place-­‐holder 0 0 0 2 2 IntensiZier dou mzi … gei 0 0 0 30 30 (8%) Negative evaluation 8 1 0 25 34 (9%) Σ 378 26 NegaAve evaluaAon reading of mzi construcAons Position of mzi
Clause-­‐initial (with subject) Clause-­‐initial (without subject) Subjective speaker stance Clause-­‐medial (with 1SG subject) Parenthetical Clause-­‐medial (without 1SG subject) Parenthetical NP 都＋唔知＋ ＿唔＿ 5
0
0
0
我唔知＋咁
1
0
0
0
邊個唔知道 1
0
0
0
鬼唔知咩
2
0
0
0
唔知＋-­‐ve polarity＋咩 0
1
0
0
又＋唔知 (+Wh-­‐) 0
1
0
2
都＋唔知＋indeZinite Wh-­‐ 0
3
0
11
都＋唔知＋indeZinite Wh-­‐＋鬼 0
0
0
7
9
5
0
20
Form/function Σ Clause-­‐iniAal Ngo dou mzi 我唔知
28
Clause-­‐iniAal dou mzi 唔知
29
Clause-­‐medial dou mzi 唔知
30
Clause-­‐medial dou mzi V mat gwai 唔知
31
唔知 Neg+Verb: neutral •  Purely lexical meaning, neutral stance
–  我都唔知㗎！
•  Lexical meaning + speaker doubt –  都唔知佢噏乜。
•  ParentheAcal stance-­‐marker •  (+ negaAve aJtude) –  你實在都唔知
Adverb, negative attitudinal 乜鬼！
not.know not.know + doubt (+ a@tude) not.know + doubt + a@tude 32 Hearsay ngo5 teng1gong2 waa6 (‘I heard it said’) (24) 我
聽
講
話
ngo5 teng1 gong2 waa6 1sg hear say say.COMP 佢哋
都
識
唱嘢
㗎！
keoi5dei6 dou1 sik1 coeng3je5 gaa3 3PL also know sing.thing PRT (a) ‘I heard (people) say they are good at singing as well!’ (b) ‘I heard it said (< ‘I heard (people) say) they are good at singing as well!’ Intended meaning: ‘I heard that they are good at singing as well!’ (from the movie Spring Time in the Jade Hall 春色滿璇宮; 1966) 33 Hearsay (ngo5) teng1gong2 waa6 (‘It is said’) (25) 聽講
上頭
而家
派咗
teng1gong2 soeng6tau4 ji4gaa1 paai3zo2 hear.say superior now dispatch-­‐PFV 嗰個
姓
陳
嘅
特派員
go2go3 sing3 can4 ge3 dak6paai3jyun4 DEM.CL surname Chan ATTR special.invesAgator ‘It is said (< ‘I hear’) the superior has currently dispatched the special invesAgator, Chan.’ (from the movie A Deadly Night 死亡角之夜; 1964) 34 Hearsay (ngo5) teng1gin3 waa6 (‘I heard it said’) (*) A:
B:
嗰 個 人
曉得
官話
唔 呢 ？
go2 go3 jan4 hiu2 dak1 gun1waa2 m4 ne1 that CL person know Mandarin NEG Q Lit. ‘Does that person know Mandarin?’ ‘Does he understand Mandarin?’ 我
聽見
話
佢
唔
曉得
ngo5 teng1gin3 waa6 keoi5 m4 hiu2dak1 1SG hear.see say.COMP 3SG NEG know ‘I have heard people say that he does not know.’ (Wade 1867: 15; cited in The Forty Exercises, Exercise VI: Phrase 10, Anonymous, 1877) 35 Hearsay (ngo5) teng1waa6 (‘It is said’) (*) 聽話
好
熱鬧，
teng1waa6 hou2 jit6naau6 hear.say INT exciAng 點解
咁
高興
你 都
唔
去
呀
dim2gaai2 gam3 gou1hing3 nei5 dou1 m4 heoi3 aa3 why such cheerful 2SG FOC NEG go PRT ‘It is said (< ‘(I) hear’) (the event is) very exciAng. It is such a cheerful acAvity. Why don’t you go join?’ (from the drama 血的敎訓 hyut3dik1gaau3fan3 [Lessons from Past Sufferings], mid-­‐20th c., cited in 1999, The Great DicVonary of Chinese Dialects, 2629) 36 Cantonese wo (< wah 話) 齣戲好好睇喎。
Nei ceot hei hou hou tai wo. this CL movie very good see wo ‘(Contrary to your expectaAon), this movie is actually very good.’ 37 Cantonese wah 話 as lexical verb 人地
話
鳥語
可以 忘憂
jɐn.ti
wa
niu.ʮy
hɔ.ji
mɔŋ.jɐu
(reconstructed)
jan4dei6
waa6
niu5jyu5
ho2ji5
mong4jau1
(Contemporary)
others
say
chirps
can
forget.worries
‘It is said that you will forget your worries when listening to chirps.’
(1828 Jyut6au1 粵謳, by Ziu1 Zi2jung4 招子庸)
38 Cantonese wo as evidenAal marker 唔曾
佢
話
挨
至
和
m.ts‘ang
k‘ü
wá
ái.mán
Chí
not.yet
3SG say
evening
not.until bring
ning
lai
wó (=wo3)
come
EVID
‘No, not yet, but he promised to bring it this evening.’
(Bridgement 1845:158)
39 Cantonese wo as counterexpectaAon marker 你 睇 下 隔籬
嗰
幾
粒 鑽石 ， 正嘢
喎！
nei5
tai2
haa5
gaak3lei4
toi4
go2
gei2
nap1
zyun3sek6
zeng3je5
lai4
wo3 !
2SG
see
ASP
neighbour
table
DEM
several
CL
diamonds
precious
EMPH
CE
‘You see the diamonds in the next table? Precious stuff!’
(Leung 2006: 69, from the movie New White Golden Dragon 新白金龍: 1947)
40 UYerance-­‐final waa-­‐type and wo-­‐type evidenAals Stage I: UEerance-­‐ini.al eviden.al jan4 waa6 ‘people say’ jan4 waa6 ‘people say’ + Complement Clause > ‘hearsay’ evidenAal jan4 waa6 + ‘Insubordinate’ Main Clause Stage II: UEerance-­‐final eviden.al wo3 via right-­‐disloca.on – 
[Complement Clause] [waa6 ‘(people) say’ + o3 emphaAc parAcle] > [‘Insubordinate’ Main Clause] [‘hearsay’ evidenAal wo3] Stage IIIa: UYerance-­‐final waa-­‐type evidenAals + prosodic cues [Complement Clause] [waa6 ‘(people) say’ + sentence final prosody] > [‘Insubordinate’ Main Clause] [‘hearsay’ evidenAal waa2] or [‘Insubordinate’ Main Clause] [‘aJtudinal’ evidenAal waa3] Stage IIIb: UEerance-­‐final wo-­‐type eviden.als + prosodic cues
Stage IV: UEerance-­‐final wo3/wo4 counterexpecta.on marker [‘Insubordinate’ Main Clause] [wo3 + sentence final prosody] > [‘Insubordinate’ Main Clause] [‘hearsay’ evidenAal wo4/wo5/wo6] [‘Insubordinate’ Main Clause] [‘hearsay’ evidenAal wo3/wo4] > [‘Insubordinate’ Main Clause] [counterexpectaAon wo3/wo4] 41 Figure 1. The gramma.caliza.on of Cantonese waa6 from the 18th to 20th century ____________________________________________________________________________ 17th century 19th century 20th century Early Mid Late Early Mid Late ____________________________________________________________________________ UTTERANCE-­‐INITIAL: Development of ‘hearsay’ waa-­‐type evidenAals EvidenAal jan4 waa6 ‘people say’ (early 17th c. opera lyrics) EvidenAal jan4dei6 waa6 ‘people say’ (1828) EvidenAal ngo5 teng1gin3 jan4 waa6 ‘I heard people say’ (1867) EvidenAal ngo5 teng1gin3 waa6 ‘I heard (people) say’ (1867) EvidenAal teng1waa6 (undated)† EvidenAal ngo5 teng1gong2 waa6 (1960s movies) EvidenAal teng1gong2 (1960s movies) -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ -­‐ UTTERANCE-­‐FINAL: Development of waa-­‐type and wo-­‐type evidenAals EvidenAal wo3 (1828) (< waa6 +o3) Hearsay evidenAal waa2 (1856) ReportaAve waa1 (1888) EvidenAal wo4 (1888) EvidenAal wo6 (1888) AJtudinal evidenAal waa3 (1908) AJtudinal evidenAal wo3 (1908) EvidenAal wo5 (1912) CounterexpectaAon wo3 (1947) CounterexpectaAon wo4 (1963) ____________________________________________________________ • † Note: On the basis of textual evidence, the text was probably wriYen ajer 1907. 42 Subtle differences in evidenAal meaning across different dialects (聽 講) 渠 唔
喎
(teng1 kong2) keoi5 m4 lai4 wo5
hear say 3SG NEG come EVID
‘(S)he is not coming, I hear [ < ‘(they) say’].’
43 Subtle differences in evidenAal meaning across different dialects 齣戲真好看講。
Chit chhut hi
chin ho
koan kong.
this CL movie very good see kong
‘I’m telling you (from my evaluation) that this movie is
very good.’
44 subtle differences in evidenAal meaning across different dialects 部電影很好看說。
Zhe bu dianying hen hao kan shuo.
this CL movie very good see shuo
‘(I feel that) this movie is very good.’
45 Verbal complementaAon and disjuncAve strategy
46
Conclusions
Clause-­‐medial parAcles and stance marking in Cantonese
47
Future DirecAons
Clause-­‐medial parAcles and stance marking in Cantonese
48
Acknowledgment
•  This study is funded by the Research Grants Council of Hong Kong through a
General Research Grant (GRF PolyU Project No. 551310, 2010-2013) entitled
“Stance Marking Strategies in Asian Languages: Diachronic, Typological and
Discourse Perspectives”.
49
References
Cheung, Hung-­‐nin Samuel. 1972. Cantonese as Spoken in Hong Kong. Hong Kong: Chinese University Press. Chor, Winnie. 2010. Inclusion of the outsider — GrammaAcalizaAon of the verbal parAcle maai in Cantonese. Taiwan Journal of LinguisVcs 8(2): 33-­‐62.
Chor, Winnie. 2013. From ‘direcAon’ to ‘posiAve evaluaAon’: On the grammaAcalizaAon, subjecAficaAon and intersubjecAficaAon of faan1 ‘return’ in Cantonese. Language and LinguisVcs 14(1): 91-­‐134.
Endo, Tomoko. 2011. Epistemic stance in Mandarin conversaAon: The posiAons and funcAons of wo juede ‘I feel/think’. In Chinese Discourse and InteracVon: Theory and PracVce, Yuling Pan & Daniel Kádár (eds). London: Equinox.
Fung, Roxana Suk Yee. 2000. Final ParAcles in Standard Cantonese: SemanAc Extension and PragmaAc Inference. PhD dissertaAon. The Ohio State University. Kärkkäinen, Elise. 2003. Epistemic stance in English ConversaVon: A descripVon of its interacVonal funcVons, with a focus on I think. Amsterdam: John Benjamins.
Kwok, Helen. 1984. Sentence ParAcles in Cantonese. Hong Kong: Centre of Asian Studies, University of Hong Kong. Law, Sam-­‐po. 1990. The Syntax and Phonology of Cantonese Sentence Final ParAcles. PhD dissertaAon, Boston University.
Leung, Chung-­‐sum. 2005 [1991]. A Study of the UYerance ParAcles in Cantonese as Spoken in Hong Kong. Hong Kong: Language InformaAon Sciences Research Centre, City University of Hong Kong. Leung, W-­‐M. 2006. On the synchrony and diachrony of sentence-­‐final parVcles: The case of wo in Cantonese. PhD dissertaAon, University of Hong Kong. Lim, Ni Eng. 2011. From subjecAvity to intersubjecAvity: Epistemic marker wo juede in Chinese. In Current Issues in Chinese LinguisVcs, Yun Xiao, Liang Tao, Hooi Ling Soh (eds), 265-­‐300. Cambridge, UK: Cambridge 50
Scholars Press.
References
MaYhews, Stephen and Virginia Yip. 1994. Cantonese: A Comprehensive Grammar. London: Routledge. Simpson, A. & Wu, X-­‐Z. 2002. IP-­‐raising, tone sandhi and the creaAon of parAcles: Evidence for cyclic spell-­‐out. Journal of East Asian LinguisVcs 11(1): 1-­‐45.
Sybesma, Rint and Boya Li. 2007. The dissecAon and structural mapping of Cantonese sentence final parAcles. Lingua 117: 1739-­‐1783.
Thompson, Sandra A. & Mulac, Anthony. 1991. A quanAtaAve perspecAve on the grammaAcizaAon of epistemic parentheAcals in English. In Approaches to GrammaVcalizaVon, Volume 2: Focus on Types of GrammaVcal Markers, Elizabeth Closs TraugoY & Bernd Heine (eds), 313-­‐339. Amsterdam: John Benjamins.
Wang, Y-­‐F, Katz, A. & Chen, C-­‐H. 2003. Thinking as saying: shuo (‘say’) in Taiwan Mandarin conversaAon and BBS talk. Language Sciences 25(5): 457-­‐488. Wong, Tak-­‐Sum, Ying Yang and Foong Ha Yap. 2011. A corpus analysis of Cantonese m4zi1 and related negaAve aJtudinal expressions: ImplicaAons for the relaAonship between negaAon, indefiniteness and stance. Paper presented at the 16th InternaAonal Conference on Yue Dialects, Hong Kong Polytechnic University, December 15-­‐16.
Yap, Foong Ha, Winnie Chor and Jiao Wang. 2012. On the development of epistemic ‘fear’ markers: An analysis of Mandarin kongpa and Cantonese taipaa. In Covert Paeerns of Modality, Werner Abraham & Elisabeth Leiss (eds), 312-­‐342. Cambridge, UK: Cambridge Scholars.
Yap, Foong Ha and Winnie Oi-­‐Wan Chor. 2014. Epistemic, evidenAal and aJtudinal markers in clause-­‐medial posiAon in Cantonese. In Werner Abraham & Elisabeth Leiss (eds), Modes of Modality, 219-­‐260. [Studies in Language Companion Series 149]. Amsterdam: John Benjamins. Yap, Foong Ha, Chor, Winnie & Wang, Jiao. 2012. On the development of epistemic ‘fear’ markers: An analysis of Mandarin kongpa and Cantonese taipaa. In Covert Paeerns of Modality, Werner Abraham & Elisabeth Leiss (eds), 312-­‐342. Cambridge, UK: Cambridge Scholars.
51
Clause-­‐medial parAcles and stance marking in Cantonese
52
•  Expanding on previous studies, we will analyze these stance markers by comparing their discourse-­‐pragmaAc funcAons (e.g. aYenAon-­‐geJng, epistemic hedging, solidarity-­‐seeking) at both the lej and right peripheries, and also in clause-­‐medial posiAon. •  Inclusion of the medial posiAon allows us to further compare the development of epistemic-­‐type (‘I think’) complement-­‐taking verbs with those of aJtudinal-­‐type (‘I don’t know’) and evidenAal-­‐type (‘people say’) complement-­‐taking verbs. •  This then allows us to further determine whether different types of complement-­‐taking verbs tend to develop into stance markers that prefer to occupy different syntacAc posiAons and serve different types of subjecAve and intersubjecAve funcAons.
Clause-­‐medial parAcles and stance marking in Cantonese
53
Tengwaa 聽話
Clause-­‐medial parAcles and stance marking in Cantonese
54
Tenggong 聽講
Clause-­‐medial parAcles and stance marking in Cantonese
55
EvoluaAon of Cantonese wo (< wah 話) a.  [CP [IP keoi [VP [V waa6/*wo3 [COMP [TP nei mou lai]]]]]] ‘(S)he said you didn’t come.’ b.  [CP [IP keoi [VP [V tenggong [COMP waa6/*wo3 [TP nei mou lai]]]]]] ‘He heard that you didn’t come.’ c.  [CP [C evidenAal marker tenggong waa6/*wo3 [IP nei mou lai]]] ‘It’s said (< I hear people say) you didn’t come.’ d.  [CP [TP nei mou lai][evidenAal tag construcAon (keoi5/keoi5dei6) (gam2) waa6 o3]] ‘You didn’t come, so (s)he/they say.’ e.  [CP [TP nei mou lai][C evidenAal wo3 (>wo5)]] ‘You didn’t come(,) so they say.’ f.  [CP [TP nei mou lai][C counter-­‐expectaAon wo3]] ‘You didn’t come(,) I’m surprised to hear.’ g.  [CP [TP nei mou lai][C evidenAal/counter-­‐expectaAon wo3][evidenAal tag (ngo) tenggong (waa6)]] ‘You didn’t come(,) so I hear.’ 56