Hakka causative bun1 constructions as a family of constructions
This paper argues that a family of related constructions is needed to account for the syntactic and semantic distributions of causative bun1 constructions in Hakka. A holistic integration of the meaning of all its components lends itself to a better understanding of the causation manifested by the bun1 construction. Five major sub-types of causative
bun1 constructions are presented: coercive causation, purposive
causation, permissive causation, non-preventive causation, and unblocking causation, after cross-investigating such crucial features of event participants as intentionality, animacy, affectedness, and of the verbal features of post-bun1 predicates. It is found that while generalizations exist, idiosyncratic discrepancies are involved in examples of each subclass. Furthermore, corpus data from actual usage demonstrates that the causing event and the resulting event show an asymmetrical ground-figure relation. The pre-bun1 causing event serves to provide background information for the happening of the post-bun1 resulting event.
Keywords : Hakka bun1, causation, intentionality, animacy, affectedness,
figure-ground
1. Introduction
The syntactic and semantic development of verbs-of-giving from a lexical verb to a case marker has been claimed to be a universal cline (Heine and Kuteva 2002; Yap and Iwasaki 2003). Consider the following examples from (1) to (4): hɑÔj in Thai is a benefactive co-verb as in (1) but a purposive marker as in (2); qaoy in Khmer is a causative complementizer as in (3); and dɑin Saramaccan CE is a dative marker as in (4). However, all these case markers have undergone grammaticalization from the meaning ‘to give’.
(1) Thai (Bisang 1998: 771)
Dεεŋ sɔ-ɔn lêeg hɑÔj Sùdaa hɑÔj phŷan
Dang teach arithmetic give Suda give friend ‘Dang taught arithmetic to Suda for his friend.’
(1) Thai (Song 1997: 327)
This paper is partially based on NSC research projects (NSC 97-2410-H-004-110-MY2). An earlier version of this paper was presented at the Sixth International Conference of Construction Grammar (ICCG-6) held by Charles University in Prague, September 3-5, 2010. Special thanks are extended to A. Prezepiorkowski, who brought up the question of whether the phenomena discussed indicated a family or a constructional polysemy, an issue that will be taken up in section 4.
khA-w khiA-n co)tmA-Ay hAÔy khun tOO)p
3SG write letter give 2SG answer ‘He wrote a letter so that you would answer.’ (2) Khmer (Matisoff 1991: 430)
kñom qaoy koət ruət
1SG give 3SG run ‘I had him run (intentionally).’ (3) Saramaccan CE (Veenstra 1996: 101)
mí dɑ́ dí miíi móni
1SG give the child money ‘It is me that gave money to the child.’
In the same manner, Chinese languages such as Mandarin also demonstrate the causative-to-passive development (Norman 1982; Peyraube 1988, 1996, 1999; Xu 1994; Sun 1996; Yap and Iwasaki 2003), which is especially schematized as the cline in (4). For example, yu (與) in the Han Dynasty is a lexical verb denoting ‘give’and then in the Tang Dynasty expresses the dative function, as illustrated in (5) and (6), respectively.1
(4) Lexical ‘give’> permissive causative > reflexive > passive
(Yap and Iwasaki 2003: 420) (5) 即持此寶與諸兄弟《大正藏神經》(Dazhengzang Shengjing)2
ji chi ci bao yu zhu xiongdi
then take this treasure give ZHI-YU brother
‘Then take this treasure to his brothers.’ (Sun 1996: 22) (6) 意欲寄書與人《敦煌變文》(Dunhuang Bianwen)
yi yu ji shu yu ren
intend desire send book to people
1
In Mandarin Chinese, gei (給) is also one of the most frequently discussed morpheme that has the causative-to-passive development. As a diachronic remnant, gei has synchronically different part-of-speech assignments (cf. Her 2006), which conforms to the principle of layering (cf. Hopper 1996).
2 The following abbreviations are used for their corresponding functions: 1PL, first person plural, 1SG, first person singular, 2SG, second person singular, 3SG, third person singular, A, the tentative marker, BUN, Hakka bun1 (分), CL, classifier, COP, copular verb, COMP, complementizer, DUR, durative aspect, EMP, emphatic adverbial, LAU, Hakka lau1 (摎), NEG, negative morpheme, NOM, nominative, POSS, possessor, RESULT, resultative, SF, suffix, ZHI-YU, Mandarin zhi1-yu2 (之於).
‘(One) desires to send a book to the other.’ (Sun 1996: 22) The route of the development proposed in (4) can also be found in Hakka, one of the Chinese languages. Verb of giving in Hakka is illustrated by bun1.3 The basic verbal meaning of bun1 is ‘to give; to separate; to distribute’(Lai 2001; Huang 2005), which typically occurs in a double object construction. In addition, bun1 can occur in various constructions such as dative, permissive, causative and passive constructions. Lai (2001) proposes that bun1 exhibits multiple grammatical functions and has
undergone two routes of grammaticalization. The following examples from (9) to (13) illustrate the various functions exhibited by bun1, and the two clines in (14) and (15) proposed by Lai (2001).
(8) 分佢一領衫
ngai5 BUN gi5 id4 liang1 sam1
1SG BUN 3SG one CL clothing ‘I gave him a piece of clothing.’
(9) 分一領衫佢
ngai5 BUN id4 liang1 sam1 gi5
1SG BUN one CL clothing 3SG ‘Igaveapieceofclothing to him.’
(10) 送一領衫分佢
ngai5 sung3 id4 liang1 sam1 BUN gi5
1SG give one CL clothing BUN 3SG ‘I gave a piece of clothing to him.’
(11) 佢分 去台北
gi5 bun1 ngai5 hi3 toi5-bed4
3SG BUN 1SG go Taipei ‘He let me go to Taipei.’
3 The Hakka data presented in this paper mainly comes from Hakka Corpus (including The NCCU Corpus of Spoken Hakka,http://140.119.172.200/, and written materials). Some examples are constructed to illustrate certain points and some are taken from previous studies with format
modifications for the sake of the unity of this paper. The Chinese characters and Romanization System in this paper are rendered by the Taiwan Hakka Dictionary of Frequent Used Words (臺灣客家話常用 詞辭典 http://hakka.dict.edu.tw/) in 2008 and the Taiwan Sixian Hakka Romanization System proclaimed by National Languages Committee, Ministry of Education (教育部國語推行委員會) in 2009.
(12) 分阿姆當譴
ngai5 bun1 a1-me1 dong1 kien2
1SG BUN mother very angry ‘I made my mother very angry.’
(13) 佢分壞人 死
gi5 bun1 fai2-ngin5 cii5-xi2
3SG BUN villain kill-die ‘He was killed by a villain.’
(14) Verb > Adposition > Complementizer
(15) Verb-of-giving > Verb-of-causative > Agent marker
Bun1 in (8) and (9), both double object constructions, functions as a lexical verb
denoting ‘to give’, in (10), a dative construction, a goal marker, in (11), a permissive construction, a permissive marker, in (12), a causative construction, a causative marker, and in (13), a passive construction, an agent marker. The abundant functions carried by bun1 have attracted many researchers to investigate the relations between these functions, the semantic and syntactic developments, or cross-linguistic
comparison with Taiwan Mandarin gei and Taiwan Southern Min hōo(予)4 (Lin 1990; Hsiao 1996; Hwang 1997; Lai 2001; Huang 2005; Chiang 2006).
However, since bun1 constructions exhibit high complexities, discrepancies in definitions and classifications arise among previous studies. For example, Chiang’s (2006) classification of bun1 denoting to give or denoting to cause or to permit
indicates overlapping distributions. She claims that bun1 denoting to give can occur in a pivotal construction which usually involves the motion of an entity. However, bun1 denoting to cause or to permit can also occur in a pivotal construction involving a motion event. In addition, according to her definition, VPs in the causative and the passive constructions as in the canonical structure NP1-bun1-NP2-VP both indicate a
4 The Romanization System is rendered by the Taiwan Southern Min Dictionary of Frequent Used Words (臺灣閩南語常用詞辭典 http://twblg.dict.edu.tw/tw/index.htm), proclaimed by National Languages Committee, Ministry of Education (教育部國語推行委員會) in 2008.
result or a state. Without further detailed criteria, to distinguish a causative from a passive becomes difficult. Crucially, the interaction of construction and lexical semantics needs more in-depth investigation to better capture the syntactic and semantic complexities.
If we focus on the bun1-NP-VP construction (i.e. bun1-NP-VP) for the moment, we find out that it has been analyzed by previous studies as a serial verb construction, a pivotal construction, a causative construction, or a passive construction, and its meanings to denote causation, permission, purpose, or beneficiary, as demonstrated in (16), (17), and (18).
(16) 佢會分 去台北
gi5 voi 3 bun1 ngai5 hi3 toi5-bed4
3SG would BUN 1SG go Taipei
‘He would let me go to Taipei.’ (Lai 2001: 139) (17) 佢帶東西分狗仔食
gi5 dai3 dung1-xi1 bun1 gieu2-e2 siid8
3SG bring thing BUN dog eat
‘He brought food for the dog to eat.’ (Lai 2001: 139) (18) 球仔分竻仔刮爛
kiu5-e2 bun1 ned4-e2 guad4-lan3
ball BUN rattan scrape-broken
‘The ball was scraped broken by the rattan.’ (Lin 1990: 68)
Bun1 in (16) is analyzed as a verb denoting to permit or to let in Lin (1990), as a
verb-of-causative in Lai (2001), as a verb to denote ordinary causation or permission in Huang (2005), and as a verb to denote permission in Chiang (2006). Bun1 in (17) is analyzed by Lin (1990) as a preposition denoting a beneficial relation and is followed by a beneficiary NP, and by Lai (2001) as a (purpose) complementizer. The
bun1-NP-VP construction in (17) is claimed to denote ordinary causation or
permission in Chiang (2006). Bun1 in (18) is analyzed by Lin (1990) as a preposition denoting a passive relation and is followed by an instrument, and by Chiang (2006) as expressing unwilling permissive in the non-prototypical passive construction. The
bun1-NP-VP construction in (18) however is claimed to carry adverse passive reading
by Huang (2005). The discrepancies among previous studies indicate the difficulties in depicting the complex conceptual category of causation. This study hence aims to take up such an endeavor in teasing out the causal relation of the bun1-NP-VP constructions. Specifically, a finer-grained examination of its components is carried out so as to depict a clearer classification of the causal relations exhibited by the construction. Although generalizations exist, idiosyncrasies are involved in each of the sub-class. A family of related constructions is argued to better capture their syntactic and semantic distributions. Following the introduction, section 2 presents previous studies on causation. Section 3 provides the classification of the causal relations of causative bun1 constructions. Section 4 argues the classes as a family of constructions, and section 5 concludes the study.
2. Previous studies on causation
Comrie (1989) declares that any causative situation involves two component situations, the cause and its effect (result). There is a three-way typological distinction of formal parameters of causative construction, analytic causatives, lexical causatives, and morphological causatives, as illustrated by the following English examples, respectively:
(19a) The man caused my dog to die. (19b) The man killed my dog.
The causative expression in example (19a) is indicated by two separate verbs ‘caused to die’; in (19b), the sense of causation is included in the basic semantic content of the verb ‘killed’; and in (19c), the causative expression is derived through affixation as in ‘shortened’.
Regarding the semantics of causative constructions, various analyses have been proposed. Talmy (2000: 494), taking a global perspective, considers the basic
causative situation in terms of dynamic oppositions and claims that the resulting event functions as a figure, and the causing event, the ground; the causal relation is “result from,”in which the resulting event takes place during the duration of the causing event. While Talmy’s study focuses on the discussion of English sentences, his insight of the figure-ground relation between the causing event and the resulting event is manifested in Hakka data when larger discourse is examined.
Kroeger (2004: 204ff), teasing out the complex concepts involved in causation, depicts the semantics of causative constructions into several types. The following examples given in his paper can illustrate:
(20a) The captain caused his boat to sink (by drilling holes in the bottom). (20b) The captain caused his boat to sink (by allowing too many passengers to
come aboard).
(21a) John made his daughter watch the rugby match on TV. (21b) John allowed his daughter to watch the rugby match on TV. (21c) John had his daughter watch the rugby match on TV. (22a) John put his (sleeping) daughter into her car seat.
(22b) John made his (*sleeping) daughter get into her car seat.
Example (20a) indicates direct causation whereas that in (20b) indicates indirect (mediated) causation. The distinction lies in the connection of the causer’s action and the resulting event. While in the former, the causer does or says something directly to the causee, usually with the intention of bringing about the resulting event, in the
latter, no such direct action is indicated and the resulting event may be an unintended consequenceofthecauser’sactions. Examples in (21) illustrate the contrast between a coercive causative and a permissive causative, with neutral causation lying
in-between. Coercion as in (21a) and permission as in (21b) differ regarding the degree of initiation and control exercised by the causer, and the degree of control or option retained by the causee. With the neutral causation as in (21c), the initiation may come from the causer, but the causee may have the option to refuse. The last paradigm in (22) contrasts physical manipulation and verbal direction in that the former as in (22a) involves the causer taking direct physical action to bring about the resulting event whereas the latter as in (22b) involves the causer saying something to the causee to achieve the resulting event.
Another endeavor in this line of understanding causation is found in Silva
(2007:179ff), in which the concept of letting is in focus. The following examples from the paper illustrate three senses of letting: “not to prevent”, “to let go, to release”and “to allow, to permit”:
(23a) John started fooling around and I let him do it. (23b) John let the bird fly out (by opening the birdcage).
(23c) John asked me if he could go to the cinema, and I let him go.
According to him, the three senses of letting causation categorize three groups of verbs. The first group exemplified by let1as in (23a) and other verbs of
non-preventing expresses a non-interventive or non-preventive causation in which the agent does nothing to stop or prevent an already ongoing event. The second group exemplified by let2as in (23b) and other freeing-exemptive verbs expresses an
unblocking causation in which the agent removes the blockage for the resulting event. The third group exemplified by let3as in (23c) and other permissive verbs expresses a
permissive causation in which the agent makes permissible a future event and has responsibility for its social and moral legitimization.
While previous studies on causation have shed some light on the nuances and the elements involved in such a conceptual category, they also show that no single
situational notion of causation exists as commented by Talmy (2000). Furthermore, their cross-classifications of the complex semantic and syntactic parameters indicate some similarities and yet some discrepancies. In particular, the complexities
encompass the causer’s intention, the causee’s affectedness, the degree of volition of the causer and the causee, and the verbal features of the resulting event. For the present purpose to analyze the causal relations involved in the causative Hakka bun1 constructions, five types will be proposed to classify the data. Specifically, following the stream of thought of constructional approach as in Jackendoff (1997), Goldberg (1995, 2006), and Goldberg and Jackendoff (2004), among others, this study claims that to tease out the intriguing complexities involved in the causal relations, all the syntactic and semantic peculiarities of the components need to be taken into
consideration in that the meaning of the construction comes from the integration of all the components of the construction holistically. The classification of the causal
relations is proposed in section 3.
3. Classifying the causal relations of causative bun1 constructions
Hakka causative bun1 constructions are canonically formalized as
NP1-bun1-NP2-VP, in which NP1 is a causer, and NP2 is a causee. After examining closer corpus data, we classify the causal relations of causative bun1 constructions into five sub-classes: coercive causation, purposive causation, permissive causation, non-preventive causation, and unblocking causation. Whereas the first three types are intentional, non-preventive causation is non-intentional, and unblocking causation can
be either intentional or non-intentional. Each sub-class will be discussed with illustrative examples.
3.1 Intentional causation
Intentional causation5 is in general defined as the occurrence of the first event being brought about intentionally or diligently. In other words, the initiative which brings about the happening of the event totally rests with the causer while the causee has no choice at all. Three sub-types are classified: coercive causation, purpose causation, and permissive causation.
The first type involves coercive causation whereby the causer actively and intentionally works to bring about the resulting event such that the causee is coerced into undergoing the psychological condition expressed by the post-bun1 predicate. Examine the following examples in (24) and (25):
(24) 你(挑試/煞猛)愛分阿姆順心
ngi5 (tiau1-sii3/sad4-mang1) oi3 bun1 a1-me1 sun3-xim1
2SG intentionally/diligently should BUN mother satisfied ‘You should intentionally/diligently please(your)mother.’
(25) 阿文摎阿英都盡增志，無分你失望
a1-vun5 lau1 a1-in1 du3 qin3 zen3-zii3 mo5 bun1 ngi5
A-vun and A-in both very hard-working NEG BUN 2SG
siid4-mong3
disappointed
‘A-vun and A-in are both hard-working so as not to disappointyou.’ In this type, the causer is an animate actor, and the causee is an animate undergoer; the post-bun1 predicate expresses psychological condition that is difficult to control. The causer’s intention can be tested by the co-occurrence of adverbs denoting ‘intentionally’, e.g. tiau1-sii3 (挑試), tiau1-tiau1 (挑挑), gu3-i3 (故意),
5 Theterm “intentionalcausation”isborrowed from Dowty (1972).I
ts definition is similar to Talmy’s (1976) agent causation and Kroeger’s (2004) coercive causation.
ngang3-ngang3 (硬硬), and pien1-pien1 (偏偏), or with adverbs denoting ‘diligently’, e.g. sad4-mang1 (煞猛), fong3-se3 (放勢), da2-biang3 (打拚), nu2-lid8 (努力),
mien1-lid8 (勉力), qin4-lid8 (盡力), and qion5-lid8 (全力), as illustrated in (24). The
causee is an animate undergoer who has undergone the psycho experience expressed by the post-bun1 predicate, which can include unergative psycho verbs such as
sun3-xim1 (順心) ‘to be satisfied’, kien2 (譴) ‘to be angry’, xin3-fug8 (信服) ‘to be convinced’, fon1-hi2 (歡喜) ‘to be happy’, siid4-mong3 (失望) ‘to be disappointed’ and so on. In other words, the intention of the causer brings about the happening of the causee’s psycho experience.
Notice that when more authentic data in a larger context are explored, the causal relation of the causing event and the resulting event demonstrates a figure- ground feature claimed by Talmy (2000). The following examples in (26) and (27) show that the causing event in which the animate causer involves in provides a background so that the animate causee in the resulting event can undergo certain psychological experience. The resulting event led by bun1 represents the figure, the communicative focus in the passage.
(26) 學期一開始 就寫一份教學計劃書，分學生仔親身體驗客家風情
hog8-ki5 id4 koi1-sii2 ngai5 qiu3 xia2 id4 fun3
semester as-soon-as begin 1SG EMP write one CL
gau1-hog8 gie3-vag8-su1 bun1 hog8-sang1-e2 qin1-siin1
teach plan BUN student personally
ti2-ngaim3 hag4-ga1 fung1-qin5
experience Hakka customs and practices
‘As soon as the semester began, I wrote a teaching plan so that students could have hands-on experiences of Hakka customs and practices.’
(27) 前兩日，細舅公个賴仔來 屋下看阿爸，分 試著盡感心
qien5 liong2 ngid4 se3 kiu1-gung1 ge3 lai3-e2
before two CL little grandmother’s brother NOM son
loi5 vug4-ha1 kon3 a1-ba1 bun1 ngai5 cii3-do2 qin3
gam2-xim1
touched
‘The son of my grandmother’s youngest brother’s visit to my father two days ago made me feel so touched.’
Unlike the first type, the second type involves purposive causation in that the causer intentionally does something in order to bring about the happening of the resulting event in which the causee performs some action. The causer is an animate actor, the causee is also an animate actor, and the post-bun1 predicate expresses action. Consider the following examples in (28) and (29):
(28) 佢帶東西分狗仔食
gi5 dai3 dung1-xi1 bun1 gieu2-e2 siid8
3SG bring thing BUN dog eat
‘Hebroughtfood forthedog to eat.’
(29) 緊看電視還會緊講客家人个故事分 聽
gin2 kon3 tien3-sii3 han5 voi3 gin2 gong2 hag4-ga1-ngin5 ge3
DUR see TV EMP can DUR speak Hakka people NOM
gu3-sii3 bun1 ngai5 tang1
story BUN 1SG listen
‘He could tell Hakka stories to people while watching TV at the same time.’ In (28), the causer is the subject gi5 (佢) ‘he’, but the causer is implicit in (29). Both can pass the intentional test with the adverbs tiau1-sii3 (挑試) ‘intentionally’or
sad4-mang1 (煞猛) ‘diligently’. In both cases, the post-bun1 predicates are action verbs for the causee-actor to perform. This type illustrates a pivotal construction where the pre- and post-bun1 verbs share the same patient argument. For example, in (28), the purpose of the causer to bring food to the dog, the shared patient argument, is for the dog, the causee, to eat food. Since the causer’s bringing about the first event is for the purpose of the happening of the second event, a deictic verb such as hi3 (去) ‘go’or a relative phrase ge5 mug4-did4 ciu3 he5 (个目的就係) can be inserted before
bun1. Example (30) can illustrate.
gi5 dai3 dung1-xi1 hi3 / ge5 mug4-did4 ciu3 he5
3SG bring thing go / NOM purpose EMP COP
bun1 gieu2-e2 siid8
BUN dog-SF eat
‘The purpose of his bringing food is for the dog to eat.’
In addition, in this type, a nominal clause that specifies an eventive subject can also be found. The following example in (31) illustrates:
(31) 大家共下來學客家話，分客家話流傳千年萬年
tai3-ga1 kiung3-ha3 loi5 hog8 hag4-ga1-fa3 bun1 hag4-ga1-fa3
everyone together come learn Hakka BUN Hakka
liu5-cun3 qien1 ngien5 van3 ngien5
hand down thousand year ten thousand year
‘Let’s learn Hakka together so that Hakka can be handed down generation after generation.’
Notice that two distinctive features can be detected between the first type and the second type. For coercive causation, the causee is an undergoer that undergoes the psychological condition indicated by the post-bun1 psycho predicate. For purposive causation, the causee is an actor that performs the action indicated by the post-bun1 action predicate. A causee has argued to be the least volitional whereas a causer is the most volitional along a hierarchy (cf. Van Valin and LaPolla 1997, Park 1993).6 The two types of Hakka data indicate that causee’s volition can differ in degrees
depending on whether it functions as an actor-caussee or as an undergoer-causee. The distinction of an actor-causee and an undergoer-causee in Hakka can be tested out by means of the (non)co-occurrence of the post-bun1 verbs with the tentative aspect V-V-a2 (V-V-啊) ‘to try on’. Cases of coercive causation whereby the causee is an undergoer cannot co-occur with the tentative aspect, but cases of purposive causation
6
Van Valin and LaPolla (1997) propose the Actor-Undergoer Hierarchy in Role and Reference Grammar. There are only two roles in RRG: actor and undergoer. The assignment of a macro-role, i.e. generalized semantic roles across the thematic relation, to a given argument is subject to the hierarchy. Park (1993) proposes that an agent or an effecter is the unmarked choice for actor, while a theme or a patient is the unmarked choice for undergoer.
whereby the causee is an actor can co-occur with the tentative aspect. Contrast the following two examples:
(32) * 分阿姆當譴譴啊
ngai5 bun1 a1-me1 kien2-kein2-a2
1SG BUN mother angry-angry-A (33) 佢帶東西分狗仔食食啊
gi5 dai3 dung1-xi1 bun1 gieu2-e2 siid8-siid8-a2
3SG bring thing BUN dog-SF eat-eat-A ‘He brought food for the dog to eat some.’
The third type involves permissive causation whereby the causer socially or morally grant permission for the causee to bring about a future event. The causer is an animate actor whose authority makes permissible a future event performed by the causee. The causee is also an animate actor, and the post-bun1 predicate indicates an action for the causee to perform. Examine the following examples in (34) and (35):
(34) 阿叔分 跈去台北
a1-sug4 bun1 ngai5 ten5 hi3 toi5-bed4
uncle BUN 1SG follow go Taipei ‘Uncleallowed me to follow him to Taipei.’
(35) 姐婆愛分阿舅討餔娘了
jia2-po5 oi3 bun1 a1-kiu1 to2 bu1-ngiong5 le2
grandmother will BUN uncle get wife PRT ‘Grandmotheris ready to let Unclegetmarried.’
In both cases, the causer’s authority can be tested by higher predicates such as future-having verbs without dative alternation (cf. Levin 1993: 139), e.g. ti5-ngi3 (提 議) ‘to advance’, tung5-i3 (同意) ‘to grant’, bo2-ziin3 (保證) ‘to guarantee’, dab4-in3 (答應) ‘to promise’, and hen2 (肯) ‘to be willing to, and say verbs (cf. Levin 1993: 209), e.g. xien1-bu3 (宣布) ‘to announce’, zu2-zong1 (主張) ‘to claim’, gien3-ngi3 (建議) ‘to propose; to suggest’, gong5 (講) ‘to say’, and sang1-min5 (聲明) ‘to state’. The post-bun1 verb is an unergative motion verb, which indicates the causee’s
movement, such as ten5 (跈) ‘follow’in (34), or a transitive activity verb, which indicates the causee’s action, such as to2 bu1-ngiong5 ‘get married’in (35). Notice that the occurrence of a higher predicate helps establish the deontic status of
permissibility of an action in the resulting event. This causation however emphasizes the causee’s will to perform the action. What the causer does is nothing more than exerting its authority so as to legitimize the action to be performed by the causee socially or morally. Irrealis modality denoting future can co-occur with the post-bun1 predicate. The following examples illustrate that zo3-ded4 (做得) ‘can’denotes deontic modal in Hakka can be used to indicate the directive modality of permission.7
(36) 佢分 去台北
gi5 bun1 ngai5 hi3 toi5-bed4
3SG BUN 1SG go Taipei ‘He let me go to Taipei.’
(37) 做得去台北
ngai5 zo3-ded4 hi3 toi5-bed4
1SG can go Taipei
‘(After getting permission,) I can go to Taipei.’
The first three types all involve intentional causation. Non-intentional causation is found in the next type.
3.2 Non-preventive causation
Non-preventive causation, as defined by Silva (2007:179), expresses that the causer does nothing to stop or prevent the subsequent already ongoing event from going on.8 In other words, the causer, although unintentionally, does not take an
7 The deontic modal in Hakka, zo3-ded4, behaves like ke3-yi3 (可以) in Mandarin. As identified in previous studies (Chao 1968; Liu et al. 1996; Huang 1999; Wu 2009), ke3-yi3 denotes the directive modality of permission.
8 Non-preventive causation is also called non-interventive causation in Silva (2007:179). It is somewhatlikeKroeger’s(2004)neutralcausation as well as Talmy’s(1976) author causation, i.e. unintentional causation, and autonomous causation.
initiate to prevent an ongoing event such that the causee that owns the autonomy undergoes the resultant state. The causer is an animate actor, the causee is an undergoer that can be animate or inanimate, and the post-bun1 predicate is a resultative verb complement. Consider the following examples in (38) and (39):
(38) 伯姆毋記得餵豬仔，續分豬仔枵死
bag4-mei1 m5 gi3-ded4 vi3 zu1-e2 sa3 bun1 zu1-e2 iau1-xi2
aunt NEG remember feed pig-SF hence BUN pig-SF hungry-die ‘Auntdidn’trememberto feed pigs,letting thepigsdieofhunger.’ (39) 醫生(無細義)分病人(自家)死忒
i1-sen1 (mo5 se3-ngi3) bun1 piang3-ngin3 (cii3-ga1)
doctor NEG careful BUN patient by oneself
si2-ted4
die-RESULT
‘Thedoctor(unintentionally)letthepatientdie(by himself/herself).’
Cases in (38) and (39) illustrate non-preventive causation, in which both the causer and the causee are animate, and the post-bun1 verb is a resultative complement that indicates the causee’s resultant state. The non-prevention and the unwittingness of the causer can be tested by mo5-se3-ngi3 (無細義) ‘unintentionally’, and the autonomous naturalness of the causee can be tested by cii3-ga1 (自家) ‘by oneself’.
Notice that similar figure-ground distinction can be detected when authentic data are examined. In example (40) below, the delineation of the doctor’s behavior and attitudes is stated, giving more specific background information regarding the causer’s unintentionality so as to explain the happening of the resulting event.
(40) 醫生無細義个態度摎行為分病人自家死忒
i1-sen1 mo5 se3-ngi3 ge5 tai5-tu5 lau1 hang3-vi3 bun1
doctor NEG careful NOM attitude LAU behavior BUN
piang3-ngin3 cii3-ga1 si2-ted4
patient by oneself die-RESULT
‘The doctor with his unintentional attitudes and behaviors had the patient die by himself/herself.’
In addition, cases as in (41) and (42) in which the causee is inanimate are also found. Even though the causee is not animate, its own autonomy of leading to the resultant state can still be observed. The post-bun1 verb is either a resultative compound or a resultative complement led by the complementizer do3 (到).
(41) 醫生分痣仔 發大/發到當大
i1-sen1 bun1 cii3-e2 bod4-tai3 / bod4 do3 dong1 tai3
doctor BUN pile-SF swell-big swell COMP very big ‘The doctor let the pile become bigger/very big.’
(42) 佢分火摎屋仔燒忒
gi5 bun1 fo2 lau1 vug4-e2 seu1-ted4
2SG BUN fire LAU house burn-RESULT ‘He let the house burned into the ground by fire.’
The non-preventive causation differs from the previous three types in three aspects. The first and most important distinction lies in the causer’s intentionality: the first three types involve intentional causers whereas non-preventive causation
involves unintentional causers. Second, the causee of intentional causation is always animate whereas the causee of non-preventive causation can be either animate or inanimate. Third, the post-bun1 predicate of intentional causation denotes a psychological condition undertaken by the causee or an action performed by the causee whereas that of non-preventive causation indicates the causee’s resultant state.
3.3 Unblocking causation
The last type in the taxonomy is unblocking causation whereby the causer intentionally or unintentionally removes the blockage of the causee for the happening of the resulting event. The causer can be an animate or inanimate actor, the causee can be an animate or inanimate undergoer, and the post-bun1 predicate is a resultative verb complement. Consider the following examples in (43) and (44):
(43) 頭下該陣大雨分河壩水淰出來
moments ago that CL big rain BUN river water
nem1-cud4-loi5
full-outside-come
‘Theriverwateroverflowed because of the heavy rain.’ (44) 巡察分該賊仔走忒了
cun5-ca5 bun1 ced8-e2 zeu2-ted4 le2
cop BUN thief-SF run-RESULT PRT
‘Thecop (un)intentionally let thatthiefrun away (from theprison).’ In example (43), the heavy rain has the primary responsibility for the overflowing of the river, and in example (44), the cop has the primary responsibility for the thief’s running away. In the heavy rain case, it is the circumstance that brings about the resultant state of the undergoer. However, whether an animate causer
intentionally or unintentionally brings about the resultant state of the causee can be up to contextual construal. Contrast the following two examples in (45) and (46). The (un) intentionality of the subject causer can be explicitly identified by the adverb
mo5-se3-ngi3 (無細義) ‘unintentionally’as in (45) or by the adverb tiau1-sii3 (挑試) ‘intentionally’as in (46). The intentionality of the causer in (46) can be further emphasized by the purposive clause followed. The same contrast can also be
illustrated in (47) and (48), in which the cop is specified unintentionally in the former, but intentionally in the latter.
(45) 佢無細意分鴨仔食忒多
gi5 mo5 se3-ngi3 bun1 ab4-e2 siid8 ted4 do1
2SG NEG careful BUN duck-SF eat too much ‘He unintentionally had the ducks eat too much.’
(46) 佢挑試分鴨仔食忒多，就係想愛遽遽摎鴨仔拿去賣錢
gi5 tiau1-sii3 bun1 ab4-e2 siid8 ted4 do1 ciu3 he3
3SG intentionally BUN duck-SF eat too much EMP COP
xiong2 oi3 giag4-giag4 lau1 ab4-e2 na1 hi3 mai3 qien5
think want quickly LAU duck-SF take go sell money ‘He intentionally overfed ducks so that he could quickly sell them for money.’
(47) 摝恁久，巡察還係無細義分該賊仔走忒
lug4 an2 giu2 cun5-ca5 han5 he3 mo5 se3-ngi3 bun1
put-effort so long cop EMP COP NEG careful BUN
ge3 ced8-e2 zeu2-ted4
that thief-SF run-RESULT
‘After putting efforts (into this matter) for such a long time, the cop still unintentionally had that thief run away.’
(48) 看著該賊仔恁衰過，巡察包尾答應分該賊仔走忒
kon3-do2 ge5 ced8-e2 an2 coi5-go3 cun5-ca5 bau1-mi1
see-RESULT that thief-SF so miserable cop finally
dab4-in3 bun1 ge5 ced8-e2 zeu2-ted4
promise BUN that thief-SF run-RESULT
‘Considering that the thief was so miserable, the cop promised to let him run away.’
To sum up, the causative bun1 constructions are classified into five classes: coercive causation, purposive causation, permissive causation, non-preventive causation, and unblocking causation. The key features involved in the classification are intentionality, animacy, and affectedness of the participants as well as the verbal features of post-bun1 predicates. The taxonomy of the five types of causative bun1 constructions can be summarized as follows: When the causer in the causing event intentionally exerts the efforts for the happening of the resulting event, it is intentional causation. When the resulting event indicates the causee’s psychological condition, it is coercive causation; when the resulting event denotes an action performed by the causee, it is purposive causation; and when the resulting event indicates a permissible action performed by the casuee, it is permissive causation. On the other hand, when the causer does not exert any efforts to prevent an ongoing event from going on, it is non-preventive causation. Finally, either an intentional causer that is animate or an unintentional causer that is inanimate can be involved in unblocking causation. In either case, the causer removes the blockage so as for the happening of the resulting
event that indicates the resultant state of the causee.
4. Causative bun1 constructions as a family of constructions
The classification of Hakka causative bun1 constructions indicates that regardless of the generalizations, they do not exhibit homogeneity but rather show a great deal of syntactic and semantic variations. It is hence better to treat Hakka causative bun1 constructions as forming a family of five sub-types of constructions, sharing important syntactic and semantic properties but differing in certain syntactic and semantic specifics.
A perspective that serves better for the account of such intriguing complexities exhibited by Hakka causative bun1 constructions is that of the constructional approach (Kay 1995, Goldberg 1995, 2006, Goldberg and Jackendoff 2004, Croft 2003, Boas 2003, 2008abc, among others). Taking usage-based viewpoints, the constructional approach holds that the meaning of language is derived from the context in which it arises. It recognizes form and meaning as parts of each
grammatical element rather than as separate components of the grammar. All levels of grammatical analysis involve constructions: learned pairings of forms with a semantic or discourse function, including morphemes or words, idioms, partially lexically-filled phrasal patterns and fully-generalized ones, as illustrated in Table 1 (Goldberg 2006: 5).
Table 1 Examples of constructions, varying in size and complexity
Morpheme e.g. pre-, -ing
Word e.g. avocado, anaconda, and
Complex Word e.g. daredevil, shoo-in
Complex Word (partially filled) e.g. [N-s] (for regular plurals)
Idiom (partially filled) e.g. jog <someone’s> memory, send
<someone> to the cleaners
Covariational Conditional The Xer the Yer (e.g. the more you think
about it, the less you understand)
Ditransitive (double object) Subj V Obj1Obj2(e.g. he gave her a fish
taco; he baked her a muffin)
Passive Subj aux VPpp (PPby) (e.g. the armadillo
was hit by a car)
All of these constructions are organized in a particular way in a speaker’s mind. A construction is any formal element that is directly associated with a particular
meaning, pragmatic function, or discourse context. Idioms that are usually deemed as peripheral in the generative model not only exhibit patterns and regularity but also represent a syntax-lexicon continuum (Kay and Fillmore 1999).
While it is convincingly argued that constructions carry meanings of their own, language phenomena manifest that subtleties of verbs still play significant roles when fusing with constructions. The nuances of verbal meanings should be more carefully represented and incorporated into constructional meanings, as maintained by Croft (2003), Iwata (2005a, 2005b), and Boas (2003, 2005, 2008a, 2008b, 2008c). They caution that the nuances of verbal meanings need to be explicitly specified in
constructions so as to avoid overgeneralization. For instance, Croft (2003: 58ff) points out that the different senses of ditransitive constructions as claimed by Goldberg (1995) in fact are associated with different verb classes. He maintains that the
following two sets of examples from Goldberg (1995:130) indicate that constructions are sensitive to particular verb classes:
(50) Sally refused/denied/*prevented/disallowed/*forbade him a kiss. (51) Tessbaked/*broughtBillacake,buthedidn’tgetit.
Croft (2003) holds that modulation of each constructional sense matches a semantic component of the verbal meaning in that construction. The different senses of the ditransitive construction are very closely tied to the verb classes that each sense occurs with (Croft 2003: 56). Examples (49) and (50) indicate that the constructional sense ‘agent enables recipient to receive patient’is tied to the co-occurring verb classes, that is, verbs of permission, e.g. permit and allow, rather than verbs of refusal, e.g. refuse, or verbs of future transfer, e.g. leave. In (51), it is the crash of the sense of
bring and the sense ‘agent intends to cause recipient to receive patient’carried by the construction that blocks the occurrence of bring.
Accordingly, Croft (2003) proposes the concept of verb-class-specific constructions and verb-specific constructions to account for the highly-related correlation between lexical idiosyncrasy and constructional generality. According to Croft, the semantics of the combination of the verb and the ditransitive construction is divided into three components. The first is verbal constant, the core of the meaning that differentiates verbs of the same semantic class. The second is the transfer of possession, the meaning associated with the verbs when they occur in the ditransitive construction. The third is the modulation, the subtleties of the meaning of the transfer of possession, whether it is actual, conditional, or intended.
Croft’s verb-class and verb-specific constructions are in principle comparable to Boas’(2003, 2005, 2008a, 2008b, 2008c) mini-constructions, which are defined as lexicalized representations necessary for predicting the exact distribution of a verb in constructions to encode conventionalized senses of verbs including syntactic,
exhibit a uniform distribution in syntactic behaviors. For example, although talk and
speak are communication verbs, they show discrepancies regarding their syntactic
distributions in declarative sentences as in (52) and (53) and in resultative constructions as in (54) and (55).
(52) Miriam talked (to Joe). (53) Miriam spoke (to Joe).
(54) Miriam talked herself blue in the face. (55) *Miriam spoke herself blue in the face.
Boas holds that the purpose of mini-constructions is to bring the information about the types of event participants that may occur with a specific sense of a verb. The underlying semantics shared by all verbs in the verb frame, and the semantic bundles of lexical characteristics and idiosyncrasies will help determine the degree of verb descriptivity. Thus, the power of constructions is delimited and
ungrammatical sentences will not be over-generalized.
To recapitulate, recognizing the existence of meaningful constructions, Croft (2003), Boas (2003, 2008abc) further argue that lexical nuances and constructional meanings interact and that the syntactic, semantic, and pragmatic information of the co-occurring lexical entries and the construction have to be taken into consideration at the same time to resolve form-meaning discrepancies. The
discussion of causative bun1 constructions in the previous section accords with this stream of thought. Specifically, the constructional approach not only provides a plausible account for the intricate complexities but also better our understanding of the causal relations of causative bun1 constructions, as has been discussed in section 3. If we extract from the examples their particulars, we can get the generalization for Hakka causative bun1 constructions as follows:
(56) The syntax and semantics of Hakka causative bun1 constructions account for the causal relations between two sub-events, the causing event and the resulting event.
To capture the idiosyncratic discrepancies of the data, five major sub-types as given from (57) to (61) below are posited to encompass the intriguing complexities exhibited by the data.
(57) Coercive causation
Syntax: [S/VP1/NP1]causing eventBUN [NP2 VP2]resulting event
Semantics: thebringing aboutofthecausee’spsychologicalcondition in the resulting event is coerced intentionally by the causer in the causing event Causer: animate actor / Causee: animate undergoer
Post-BUN predicate: psychological predicate (58) Purposive causation
Syntax: [S/VP1/NP1]causing eventBUN [NP2 VP2]resulting event
Semantics: the causer intentionally initiates the causing event so as for the causee to perform a certain action in the resulting event
Causer: animate actor / Causee: animate actor Post-BUN predicate: action predicate
(59) Permissive causation
Syntax: [S/VP1/NP1]causing eventBUN [NP2 VP2]resulting event
Semantics: thebringing aboutofthecausee’saction in the resulting event is permitted by the authoritative causer in the causing event
Causer: authoritative animate actor / Causee: animate actor Post-BUN predicate: action predicate
(60) Non-preventive causation
Syntax: [S/VP1/NP1]causing eventBUN [NP2 VP2]resulting event
Semantics: thebringing aboutofthecausee’schange-of-state in the resulting event is due to thecauser’sunintentional non-prevention of an already ongoing event
Causer: animate actor / Causee: animate or inanimate undergoer Post-BUN predicate: resultative verb complement
(61) Unblocking causation
Syntax: [S/VP1/NP1]causing eventBUN [NP2 VP2]resulting event
Semantics: thebringing aboutofthecausee’schange-of-state is due to the causer’sremovaloftheblockage
Causer: animate or inanimate actor / Causee: animate undergoer Post-BUN predicate: resultative verb complement
5. Concluding remarks
The endeavor of the study has attempted to offer a comprehensive descriptively plausible account of Hakka causative bun1 constructions, hence providing a more systematic characterization of the complex conceptual category of causation. In accordance with the constructional perspective, it is claimed that all the components of the construction needs to be integrated holistically to better understand the causal relations associated with bun1 constructions. In particular, features of the event participants such as intentionality, animacy, affectedness, and nuances of post-bun1 predicates are crucial to explicitly tease out the possible relation between the causing event and the resulting event. Five major sub-classes are posited, with cases of each sharing generalizations and differing in specific idiosyncrasies. As a result, the five types are claimed to form a family of related constructions.
This study has also demonstrated that the causative bun1 constructions evidence one of Talmy’s (2000: 494) depiction of a basic causative situation: “[t]he caused event functions as the Figure and the causing event as the Ground of the whole situation…,”in particular when authentic data from a larger discourse are examined. How the causal relations associated with causative bun1 constructions interact with the information structure of the discourse is an issue that is worthy of further research. Furthermore, Talmy (2000) employs the primitive experientially grounded notion of force dynamics to analyze causation. Whether the five types of bun1 constructions exhibit different degrees of force dynamics is an issue worthy of future study.
In addition, more data of bun1 can be observed. For instance, the V-bun1 constructions as in (62) and (63) also involve a causal relation.
(62) 轉去好煮分孫仔食
zon2-hi3 ho2 zu2 bun1 sun1-e2 siid8
go back in order to cook BUN grandson eat ‘Go back to cook forthegrandson to eat.’
(63) 你摎吾爸帶轉來， 就嫁分你
ngi5 lau1 nga1 ba1 dai zon2-loi5 ngai5 ciu3 ga3
1SG LAU 1SG:POSS father bring come back 1SG EMP marry BUN 2SG
bun1 ngi5
‘As soon as you bring my father back, I’ll marry you.’
The two cases are variations of bun1 constructions that manifest the complex category of causation. More data need to be investigated to build up the network of
References
Bisang, Walter. 1998. Adverbiality: The view from the Far East. Adverbial
Constructions in the Languages of Europe, ed. by Johan van der Auwera, 643-812.
Berlin: Mouton de Gruyer.
Boas, Hans C. 2003. Constructional Approach to Resultatives. Stanford: CSLI Publications.
Boas, Hans C. 2005. Determining the productivity of resultatives: A reply to Goldberg and Jackendoff. Language 81.2: 448-464
Boas, Hans C. 2008a. Resolving form-meaning discrepancies in Construction Grammar. Constructional Reorganization, ed. by J. Leino, 11-36.
Amsterdam/Philadelphia: John Benjamins.
Boas, Hans C. 2008b. Determining the structure of lexical entries and grammatical constructions in Construction Grammar. Annual Review of Cognitive Linguistics 6: 113-144.
Boas, Hans C. 2008c. Towards a frame-constructional approach to verb classification.
Grammar, Constructions, and Interfaces (Special Issue of Revista Canaria de
Estudios Ingleses 57), ed. by Eulalia Sosa Acevedo, and Francisco José Cortés Rodríguez. 17-48.
Chao, Yuen-Ren. 1968. A Grammar of Spoken Chinese. Los Angeles: University of California Press.
Chiang, Min-hua. 2006. Grammatical characteristics of tung and bun in Dongshi Hakka and the relatedness of the two markers. Language and Linguistics 7.2: 339-364.
Croft, William. 2003. Lexical rules vs. constructions: A false dichotomy. Motivation
in Language: Studeis in Honour of Günter Radden, ed. by H. Cuyckens, T. Berg,
R. Dirven, and K.-U. Panther, 49-68. Amsterdam/Philadelphia: John Benjamins. Dowty, David. 1972. Studies in the logic of verb aspect and time reference in English.
Studies in Linguistics, Department of Linguistics, University of Texas, Austin, Texas.
Goldberg , Adele E. 1995. Constructions: A Construction Grammar Approach to
Argument Structure. Chicago and London: The University of Chicago Press.
Goldberg, Adele E. 2006. Constructions at Work: The Nature of Generalization in
Language. Oxford: Oxford University Press.
Heine, Bernd, and Tania Kuteva. 2002. World Lexicon of Grammaticalization. Cambridge: Cambridge University Press.
Her, One-Soon. 2006. Justifying part-of-speech assignments for Mandarin gei. Lingua 116.8: 1274-1302.
Anthropology 25: 217-236.
Hsiao, Yun-hsiu. 1996. On the Hakka Pun. M.A. Thesis, National Taiwan Normal University, Taiwan.
Huang, Han-chun. 2005. Causation types and the path to passives: A comparative study of Hakka Pun and Mandarin Rang. Paper presented at 2005 National Conference on Linguistics. July 3-4. National Chiao Tung University, Taiwan. Huang, Yu-Chun. 1999. A Semantic Study of Modal Verbs in Chinese. M.A. Thesis,
National Taiwan Normal University, Taiwan.
Hwang, Jya-Lin. 1997. A comparative study of Mandarin Gei ‘give’,TaiwaneseHo
‘give’and HakkaPun ‘give’.Paperpresented atNACCL-9. May 2-4. University
of Victoria, Canada.
Iwasaki, Shoichi. 2002. Japanese. Amsterdam/Philadelphia: John Benjamins. Iwata, Seizi. 2005a. Locative alternation and two levels of verb meaning. Cognitive
Linguistics 16.2: 355-407.
Iwata, Seizi. 2005b. The role of verb meaning in locative alternations. Grammatical Constructions: Back to the Roots, ed. by Mirjam Fried, and Hans C. Boas, 101-118. Amsterdam/Philadelphia: John Benjamins.
Kay, Paul. 1995. Construction Grammar. Handbook of Pragmatics: Manual, ed. by Jef Verschueren, Jan-Ola Östman, and Jan Blommaert, 171-177.
Amsterdam/Philadelphia: John Benjamins.
Kay, Paul, and Charles J. Fillmore. 1999. Grammatical constructions and linguistic generalizations: The What’s X doing Y? construction. Language 75.1: 1-33. Kroeger, Paul. 2004. Analyzing Syntax: A Lexical-functional Approach. Cambridge:
Cambridge University Press.
Lai, Huei-ling. 2001. On Hakka Bun: A case of polygrammaticalization. Language
and Linguistics 2.2: 137-153.
Li, shih-min, and Huei-ling Lai. 2010. Understanding and classifying the causal relations of Hakka causative bun1 constructions. Paper presented at the Sixth International Conference of Construction Grammar, September 3-5, Charles University, Prague.
Levin, Beth. 1993. English Verb Classes and Alternations: A Preliminary
Investigation. Chicago/London: The University of Chicago Press.
Lin, Ying-chin. 1990. Remarques sur “PUN”與 et “LAU”(/”THUNG”同) dans les dialects Hakka. Cahiers de Linguistique Asie Oriental 19.1: 61-89.
Liu, Yue-Hua, Wen-Yu Pan, and Wei Gu. 1996. Shiyong Xiandai Hanyu Yufa (Practical Modern Chinese Grammar). Taipei: Normal University Publishing. Matisoff, James A. 1991. Areal and universal dimensions of grammaticalization in
Bernd Heine, 383-453. Amsterdam/Philadelphia: John Benjamins.
Norman, Jerry. 1982. Four notes on Chinese-Altaic linguistic contacts. Ting Hua
Journal of Chinese Studies 14: 243-247.
Park, Ki-seong. 1993. Korean Causatives in Role and Reference Grammar. M.A. Thesis, The State University of New York, Buffalo.
Peyraube, Alain. 1988. Syntactic change in Chinese: On grammaticalization. Bulletin
of the Institute of History and Philology, Academia Sinica 59.3: 617-652.
Peyraube, Alain. 1996. Recent issues in Chinese historical syntax. New Horizons in
Chinese Linguistics, ed. by C.-T. James Huang, and Y.-H. Audrey Li, 161-213.
Dordrecht/Boston/London: Kluwer.
Peyraube, Alain. 1999. On the modal auxiliaries of possibility in Classical Chinese.
Selected Papers from the Fifth International Conference on Chinese Linguistics,
ed. by Samuel H. Wang, Tsao Feng-fu, and Lien Chin-fa, 27-52. Taipei: Crane Publishing Co.
Silva, Augusto Soares da. 2007. Verbs of letting: Some cognitive and historical aspects. On Interpreting Construction Schemas: From Action and Motion to
Transitivity and Causality, ed. by Nicole Delbecque, and Bert Cornillie, 171-200.
Berlin: Mouton de Gruyter.
Song, Jae Jung. 1997. On the development of MANNER from GIVE. The Linguistics
of Giving, ed. by John Newman, 327-348. Amsterdam/Philadelphia: John
Benjamins.
Sun, Chaofen. 1996. Word Order Change and Grammaticalization in the History of
Chinese. Stanford, CA: Stanford University Press.
Talmy, Leonard. 1976. Semantic causative types. Syntax and Semantics Vol. 6: The
Grammar of Causative Constructions, ed. by Masayoshi Shibatani, 43-116. New
York: Academic Press.
Talmy, Leonard. 2000. Toward a Cognitive Semantics Volume 1: Concept Structuring
Systems. Cambridge: The MIT Press.
Van Valin, R. D. Jr., and LaPolla, R. J. 1997. Syntax, Structure, Meaning and Function. Cambridge: Cambridge University Press.
Veenstra, Tonjes. 1996. Grammaticalized verbs in Saramaccan: The interplay between syntax and semantics. Changing Meanings, Changing Functions: Papers Relating
to Grammaticalization in Contact Languages, ed. by Philip Baker and Anand
Syea, 95-112. London: University of Westminster Press.
Wu, Chun-Hui. 2009. Polysemous Modal Verbs in Mandarin Chinese. M.A. Thesis, National Chengchi University, Taiwan.
Xu, Dan. 1994. The status of marker Gei in Mandarin Chinese. Journal of Chinese
Yap, Roont-Hua, and Shoichi Iwasaki. 2003. From causatives to passives: A passage in some East and Southeast Asian languages. Cognitive Linguistics Research:
Cognitive Linguistics and Non-Indo-European Languages, ed. by Eugene H.
