英语主、宾语控制结构的用例辨析及其格位赋值的最简机制研究*
——基于探针-目标关系理论的句法一致性分析
2015-03-20马志刚
马志刚
广东外语外贸大学
英语主、宾语控制结构的用例辨析及其格位赋值的最简机制研究*
——基于探针-目标关系理论的句法一致性分析
马志刚
广东外语外贸大学
英语控制结构能否通过移位的方式生成,其中所假定的空主语PRO能否被拷贝所替代是近期相关研究争论的焦点问题。Landau(2013)基于众多实证语料支持PRO的存在,而最简句法理论不仅为PRO提供了专属的格位,而且将其格位赋值机制纳入到探针-目标一致关系的理论框架中,从而使得主格、宾格和空格的指派机制达到了句法理论的一致化要求。文章通过对英语控制结构以及相关结构的用例分析指出,最简句法的特征传递机制有助于建立功能语类之间的句法依存关系,而语段中心语的(不)可解读特征是未来研究区分限定性和非限定性结构时应该优先考虑的重要因素之一。
最简方案,控制结构,空主语PRO,格位赋值,一致关系
1.引言
英语中作补语的从句可区分为限定句和非限定句。限定从句的典型代表如(1)a中的划线部分所示,其中的主语具有主格;而非限定从句包括由标句词for引导的不定式结构、控制结构、ECM结构和提升结构等,如(1)b-(1)e所示。其中,(1)b和(1)d中的非限定从句的主语均为宾格,而(1)c和(1)e中的非限定从句并不具有显性主语。
(1) a.We think that she likes Jack.
b.He planned for me to stay.
c.Lisa managed to leave immediately.
d.Mary suspected him to be innocent.
e.He appeared to be happy.
通常认为,限定从句和非限定从句之间最为显著的形式区别在于是否允准主格主语(Culicover 1997)。从管约论的视角看,限定从句的中心范畴INFL具有时态特征和一致性特征,因此可以给标示语成分指派主格,而非限定从句的INFL(即不定式标记to)不具有时态特征和一致性特征,因而无力指派格位(Haegeman 1994)。文献中有关非限定从句的争论主要集中在如下几个方面(Davies & Dubinsky 2004):(1)b-e所示的四类非限定从句中是否全都具有主语?(如果有)其主语获得何种格位以及该格位是如何指派的?非限定从句能否和限定从句共享同一套生成机制?二者主语的格位指派是否是基于相同的结构关系得以实现?
对上述问题的回答关涉到生成语法的一致性原则以及功能语类的核心句法地位,而本文正是在探针-目标一致关系的理论框架下(Chomsky 2007,2013,2014),试图提出一种统一的分析方案来探讨上述(1)a类限定句、(1)b类for-to不定式结构和(1)c类控制结构中从句的句法特点以及其中主语的格位指派问题。同时,本文将基于语段理论(Chomsky 2008,2014)对相关的句法现象提出原则性解释。
2. 相关结构的句法表现
2.1 主语控制和for-to不定式
英语控制结构的主要句法特点就是其不定式补语句并无显性主语,依据其中被控制成分的语义指向,可以将其划分为主语控制结构和宾语控制结构。例(2)是典型的主语控制结构(宾语控制的讨论详见下文):
(2) a.John tried / managed to please us.
b.John decided / neglected to please us.
c.John tried / decided [CPPRO to please us]
d.*John tried / decidedhim/heto please us. (2)a和(2)b都属于主语控制结构,其中不定式句的动词please应该具备的外论元在语义所指方面与主句主语一致,因此称为主语控制结构。如(2)c所示,主语控制结构中的空代词PRO属于作主语的隐含性论元。而(2)d显示,该主语如果显性呈现就会导致控制结构形成不合法的句子。(3)中的arrange和plan进一步说明了主语控制结构的这个句法特点:
(3) a. John arranged to visit the dentist.
b.*John arranged them to visit the dentist.
c. John planned to visit the dentist.
d.*John planned them to visit the dentist.
e. John arranged / planned [CPPRO to visit the dentist]
显然,作为主语控制动词的arrange和plan要求其不定式补语句中的主语只能是(3)e中的空代词PRO,其所指由主句主语John控制(与其同指)。那么能否据此认为,arrange和plan具有完全相同的句法表现? (4)中的语料和分析的确支持这个观察:
(4) a.*They are arranged by John to visit the dentist.
b.*They are planned by John to visit the dentist.
c. John arranged for her to visit the dentist.
d. John planned for her to visit the dentist.
e. John arranged / planned [CP[Cfor][TPher to visit the dentist]]
f. What they arranged / planned was to (for her) visit the dentist.
g.*John tried / managed for Mary to visit his parents.
(4)a和(4)b都不允准被动化,说明arrange和plan都是完整的CP语段,其中心语成分统制域内的任何成分都不能被移动到句首(PIC)(Chomsky 2008)。再者,(4)c 和(4)d表明,arrange和plan都允准for-to不定式句作为其补语。因此,二者补语句的内部构造正如(4)e所示,不定式句主语her的宾格是由及物性标句词for指派的。另外,(4)f说明,arrange和plan的不定式补语句可以是主语控制结构,也可以是for-to不定式结构。这一点,不同于try类控制动词,如(4)g所示。那么其他控制类动词是否可以采用for-to不定式结构作其补语?以下对attempt 和need的用法加以比较。
(5) a. Christofer attemped to reach the summit.
b. Christofer needed to reach the summit.
c. What Christofer attempted is to reach the summit.
d. What Christofer needed is to reach the summit.
e. Christofer attemped / needed [CPPRO to reach the summit].
(5)a和(5)b、(5)c和(5)d说明,attempt和need 都是典型的主语控制动词,因此其补语句的内部结构理应如(5)e所示。但这是否说明二者的句法表现完全一致?如下我们通过为补语句添加显性主语来加以检验:
(6) a.*Jennifer attempted him to please us.
b. Jennifer needed him to examiner her.
c.*Jennifer needed for him to examined her.
d. He was needed by Jennifer to examine her.
e. Jennifer needed there to be an examination.
f.*Jennifer attempted for him to examiner her.
(6)a和(6)b之间的对比显示,need似乎具有ECM动词的用法,(6)c不合法,说明need不能以for-to不定式句为其补语,而(6)d的被动化合法、(6)e的虚指词检验也合法,说明need除了具有主语控制的用法外还具有ECM动词的用法。(6)a和(6)f显示,attempt既没有ECM的用法,也不能以for-to不定式句为其补语,而只有(5)a所示的主语控制的用法。英语中类似于attempt和need的动词还有propose和choose,如(7)所示。
(7) a. Jennifer proposed to cancel the program.
b. Jennifer chose to cancel the program.
c. What Jennifer proposed / chose is to cancel the program.
d. Jennifer proposed / chose [CPPRO to cancel the program].
e.*Jennifer proposed (for) him to cancel the program.
f. Jennifer chose him to be the monitor.
尽管choose也具有和need类似的句法表现,但二者的用法还是有差别的:need具有ECM动词的用法(如上所述),而choose却不具有ECM动词的用法,如(8)所示:
(8) a. The doctor was needed to examine her.
b. Jim was chosen to be the monitor.
c. This spinster badly needed there to be a suitor.
d.*That spinster unwillingly chose there to be a suitor.
尽管choose和need都可以具有被动化形式(8)a和(8)b,但从(8)c和(8)d的对比中可以看出,need可以以带虚指词的不定式句为其补语,而choose则不能。这说明可以被动化的choose不是ECM动词,而是英语中的另外一类动词(宾语控制动词)。在说明英语宾语控制动词的用法之前,有必要对另外一组主语控制动词的用法加以说明。
(9) a. The customer demanded to see the manager.
b. The customer requested to see the manager.
c.*The customer demanded him to see the manager.
d. The customer requested him to see the manager.
(9)a和(9)b显示,demand和request都具有主语控制动词的用法,但(9)c和(9)d的对比说明,request可能有ECM动词的用法,而demand则没有。
(10) a.*He was demanded to see the manager.
b. He was requested to see the manager.
c .*She demanded / requested for him to enter into the contest.
d. She demanded / requested that he should enter into the contest.
(10)a和(10)b之间的对比进一步说明, 不能执行被动化的demand具有主语控制动词的用法,而request可以被动化,因此隶属ECM类动词。(10)c和(10)d说明,二者的补语均不可以标句词for引导,但可以用that引导。以下讨论英语中具有宾语控制用法的动词及其与ECM动词的关联性,以及二者被动化的用法。
2.2 宾语控制、ECM及其被动化
上文中的用法对比说明,可以以不定式结构为补语的英语动词中包括ECM动词和宾语控制动词,而这两种动词都允准补语句的主语被动化为主句的主语。(11)中的例句是对这两类动词的用法对比:
(11) a. We believe him to be innocent.
b. We told him to leave the campus.
c. He is believed to be innocent.
d. He was told to leave the campus.
e. We believe there to be a party.
f.* He told there to be a party.
英语中典型的ECM动词believe和典型的宾语控制动词tell之间的用法异同呈现在(11)中,特别是(11)f说明tell不是ECM动词。不过有些通常被视为宾语控制类的动词似乎也可以有ECM动词的用法,如(12)显示,order、require、urge就是如此:
(12) a.The general ordered the soldiers to launch an attack.
b.The professor required the students to show more devotion.
c.The Marshall ordered there to be a surprising attack.
d.The general required there to be sufficient ammunition.
e.The president urged the secretary to get ready for the press conference.
f.The president urged there to be a timely press conference.
需要注意的是,英语中的一些宾语控制动词似乎可以表达主语控制句的语义,因而具有歧义性解读,典型代表如(13)所示的ask和beg:
(13) a.I asked the guard to smoke a cigar.
b.I begged the guard to smoke a cigar.
c.I asked / begged the guard [CP[CΦ][TPPRO [Tto] [VPsmoke a cigar]]]
d.John asked/begged the guard to be allowed to smoke a cigar.
通常认为,(13)a和(13)b中主句动词的宾语the guard控制着(13)c中的空代词PRO的语义所指,但二者均可以理解为(13)d中的语义,而(13)d的内部构造如(13)e所示,其中空代词PRO的语义控制者是距离其最近的拷贝成分John。另外,英语中的permit和invite也都具有宾语控制动词的句法表现。
(14) a. The president permitted us to have a luncheon.
b. The president invited us to have a luncheon.
c. The president permitted / invited us [CP[CΦ][TPPRO [Tto] have a luncheon]].
d. We are permitted / invited by the president to have a luncheon.
e.*The president permitted / invited there to be a luncheon.
从(14)a的语义解读看,其中补语句中的空代词PRO(即(14)c)通常的解读均为permit的宾语us;而对于invite而言, (14)c中PRO的解读则应该具有联合控制(joint control)的可能性,即是invited的主语和宾语一起作为PRO的先行语。而(14)e说明,这两个动词都不具有ECM动词的用法,尽管二者都允准(14)d中的被动化。
就英语中的宾语控制动词而言,具有三元结构而且其不定式句的主语必定为主句动词的题元成分(通常为有心智的生命体)是其句法语义方面的定义性特征。因此,在句法检测中,不定式句的主语位置上既不能有虚指词there / it,也不能使用表天气的it。以下对意义较为接近的另一组宾语控制动词convince和persuade的用法加以比较:
(15) a. Tom convinced Jack to give a lecture on Nano-Syntax.
b. Tom persuaded Jack to give a lecture on Nano-Syntax.
c.*John convinced that Bill should give a talk on Mini-Semantics.
d.*John persuaded that Bill should give a talk on Mini-Semantics.
e. John convinced Bill that he(Bill) should study Max-Pragmatics..
f. John persuaded Bill that he(Bill) should study Max-Pragmatics
g. Bill was convinced / persuaded by Bill that he(Bill) should study General-Linguistics.
但并非所有的宾语控制动词都具有(15)中的全部用法,比如force就不具有(15)e-f中的用法:*George forced Mary that she should marry nobody.而persuade具有的一种用法为:I persuaded Mary that Bill left the room(Poole 2011:106)。需要特别指出的是,控制结构中(无论是主语控制还是宾语控制),其不定式补语句中的动词通常并不具有时态变化,而表达情态的动词则经常可以出现。(16)b就是典型例证。
(16) a. The general commanded / allow / instruct / encourage the soldiers to charge.
b.*The general commanded / allow / instruct / encourage the soldiers to have withdrawn.
c.*The general commanded / allow / instruct / encourage there to be a cease-fire.
d.*The general commanded / allow / instruct / encourage to forge ahead.
就(16)中的动词而言,(16)b-d显示其并非ECM动词亦非主语控制动词,而是典型宾语控制类动词,如(16)a所示。
最后,英语中还存在既可以有主语控制的用法又可以有提升用法的一类动词。比如,begin和threaten就是典型代表,其主语控制用法呈现在(17)中(其提升用法将另文专题讨论)。
(17) a. The lawyer began to irritate the defendant.
b. The terrorist threatened to drown the informant.
c. They began / threatened [CPPRO to ruin us].
d.*They began / threatened they / them to ruin us.
e. The president started to execute the demonstrators.
如(17)所示,begin和threaten与主语控制类动词try具有基本一致的句法表现,但前者还可以以非生命性的名词作为其主语(提升用法),而这是典型控制动词所不具备的。需要注意的是,在(17)e这种用法,与begin语义基本等同的start也是典型的控制类动词,比如He started to learn Greek。
最后需要指出的是,英语中有些动词的用法并不固定。比如,expect除了ECM、主语控制的用法外,还有宾语控制的用法。比如(18)a是出自英国海军的一句口号,其表达的实质意义是(18)b,也就是说,其所表达的语义更像是宾语控制类动词的语义:
(18) a.England expects everyman to do his duty (Horatio Lord Nelson)
b.England imposes on everyman the obligation of doing his duty.
c.John seems to be a bit under the weather.
d.John seems [PRO to be a bit under the weather].
另外还需要注意的是,通常被视为提升动词的seem在(18)c所示的情况下也可能具有主语控制类的用法,其内部结构的分析如(18)d所示;当然(18)c通常是具有提升用法的,其语义如:John has been reported to be ill。与上述兼类的情形相反,lead在英语中只能作为宾语控制动词来使用,如(19)中的对比所示(Haegeman 1994:281):
(19) a.This analysis led them [PRO to conclude for themselves/*oneself that Poirot was Belgian].
b.*This analysis led [[PRO to conclude for oneself that Poirot was Belgian]].
c.They were led to conclude for themselves /*oneself that Poirot was Belgian.
d.Miss Marple was ordered to go on her own.
e.Miss Marple was instructed to go on her own.
f.Miss Marple was allowed to go on her own.
Bach定律认为(Bach Generalization),宾语控制结构中的宾语内论元是必选论元而且可以被动化(Haegeman 1994)。据此,(19)c-f的合法句和(19)b的不合法都符合Bach定律的预测。最后还需要说明的宾语控制类动词是(20)中所示的ask:
(20) a.We asked them to surrender.
b.We asked that they should surrender.
c.We asked to surrender.
d.We asked for Susan to buy me some potatoes. (Carnie 2011:252)
从(20)中所有的合法句可以看出,使用频率很高的ask不仅具有宾语控制的用法(20)a、虚拟限定句补语用法(20)b和主语控制的用法(20)c,还可以后接for-to不定式句作其补语,如(20)d所示。而更需要注意的是,ask还存在主语控制和任意控制这样两种语义的解读差别,如(20)e中的反身代词必须指称主句主语,而(20)f中的反身代词只能做非强制性的解读。就语义解读而言,英语中的promise也是值得关注的主要动词之一。如下,(21)a和(21)b具有等值的语义,但如果(21)a具有(21)c所示的内部结构,那么PRO因何不以距离更近的Bill为其先行语?
(21) a.John promised Bill to leave.
b.John promised Bill that he (John) would leave.
c.John promised Bill [PRO to leave].
d.John promised Mary to control himself.
e.*John promised Mary to control herself.
传统语法通常把promise视为三元动词,因此(24)a中的Bill属于promise的一个内论元,但是由于John和Bill都可以成分统制PRO,因此二者都有可能成为其先行语,因此只能规定性地认为promise隶属主语控制动词,只能以主语John为其先行语。因此,(24)d和(24)e之间对比性差异在于promise是主语控制动词。事实上,还有一种更有说服力的解释如(22)所示的空介词分析法:
(22) a.John promised (to) Mary to control himself.
b.John promised {to Mary} [PRO to control himself].
c.*John promised {to Mary} [PRO to control herself].
d.*Mary was promised by John to conrol himself.
从(22)所示的用例可以看出,{to Mary}作为附加语嫁接在主句动词promise之上,因而其中的Mary不能成分统制PRO,因此PRO只能以句首主语为其先行语。(22)c不合法是因为herself的语义先行语并不在其约束域内,从而违反了约束A原则。同时,(22)d不合法,是因为promised作为主语控制动词的用法并不允准被动化操作。但promise不允准被动化的观点应该考虑到如下(23)中可以被动化的例句:
(23) a.Jack promised her an I-phone 9.
b.She was promised an I-phone 9.
c.Jack promised her that she would get an I-phone 9.
d.She was promised by him that she would get an I-phone 9.
事实上,promise作为主句动词是否可以执行被动化还受到不定式句结构特点的影响。比如,如下Bresnan(1980:404)的类似例句(24)显示出promise可以被动化的另外一种情况:
(24) a.Joanna was never promised [[PRO to be allowed to smoke]]
b.It was never promised to Joanna[[ to be allowed to smoke]].
c.[PRO to be allowed to smoke]] was never promised to Joanna.
d.Joanna was never promised that she would be allowed to smoke.
e.It was never promised to Joanna that she would be allowed to smoke.
f.That she would be allowed to smoke was never promised to Joanna.
g.There promised to be massive crowds of protestors.
从(24)g可以看出,promise其实还具有提升动词的用法。上述分析显示,promise在有些用例中可被视为是二元动词,而在有些用例中则可被视为三元动词。其实,真正的三元控制类动词基本上都是宾语控制动词,如(25)中的用例说明:
(25) a. Jack convinced Bill to trust himself.
b. Jack convinced Rose to trust herself.
c.*Jack convinced Rose to trust himself.
d. Jack convinced Rose [PRO to trust herself / * himself].
(25)a中的Bill被宾语控制类动词convince选择为内论元,由于在句法结构上,距离PRO最近的先行语为Rose,因此herself只能指称Rose,而John和himself之间的联系被潜在的可能先行语Rose所阻断,因此,(25)c不合法,因为其内部结构如(25)d所示。
综上所述,英语动词want、demand等都是以CP结构为其补语的,而(1)a中以限定句为补语的句法结构也是CP结构:We think [CP [C that] [TP she likes Jack]]。因此值得研究的是,这些补语句是否共享同一套生成机制,其中主语的格位指派是否达到了生成语法所要求的一致性原则? 以下我们从标句词及物性的视角来研究最简句法中的探针-目标一致关系是如何运用在上述结构的生成中的。
3.基于探针-目标一致关系的最简探索
3.1 PRO的分布及其特征组合
原则和参数理论认为(Culicover 1997:85)兼具回指性和代词性的空语类PRO因不受管辖而不能承载格位。据此可以预测的是PRO与显性名词呈现互补性分布,即有格位的位置上不能出现PRO,如(26)a-e所示:
(26) a.The protestors saw *PRO
b.The protestors think that*PRO will win.
c.The protestors believe*PRO to be strange.
d.*PRO is/are singing.
e.*The protestors spoke to PRO.
f.*There occurred three more accidents without PRO being any medical help available on the premises.(Haegeman 1994:279)。
仔细观察,可以发现,(26)f体现出PRO必须以有意愿的论元生命体作为其先行语,而不能以虚指词there作为其先行语。不过,这并不意味着没有格位的位置上就一定可以出现PRO,如下(27)中PRO出现的位置上都不具有格位,但句子依然不合法。
(27) a.* It was captured PRO.
b.*There was captured PRO.
c.*The soldier's destruction PRO was recorded.
d.*The soldiers believe very much PRO to have won.
事实上,PRO定理(PRO必须不受管辖)完全是经由约束原则推演而得出的:在管辖域内,PRO的[+anaphoric]特征决定它必受约束,而其[+pronominal]特征又要求它必不受管辖。因此,解决方案就是PRO不具有管辖域。此外,(28)显示出PRO的任指性语义特点:
(28) a.To err is human,to forgive is divine.
b.[PRO to err] is human,[PRO to forgive] is divine.
c.To praise oneself is always unwise.
d.To praise him is always unwise.
e.[PRO to praise oneself / him] is always unwise.
f.Swimming away upset John.
g.PRO Swimming away upset John.
(28)b说明,(28)a中具有PRO主语,而其语义指称具有任指性(arbitrary)。基于约束原则A和B以及(28)c和(28)d中的语料,可以证明这两个句子中必然具有PRO作主语,否则无法解释oneself在约束域内是如何获得语义指称的,以及him为何必须要与篇章中其他的名词短语同指。可见,(28)e中的分析是符合PRO定理和约束原则A、B的。那么PRO作为名词短语是如何获得格位指派的呢?
3.2 三类名词短语的句法分布共性
Davies & Dubinsky(2004)认为,want的补语句具有for的空形式,而与此相关联的是类似于John wanted her to please us中的her的宾格自然也就是标句词for所指派的,而非由可以跨越语障的want指派的。Rizzi(1997:304)曾指出,PRO作主语时的空格应该是由C来指派的,而限定性标句词理应为主格指派者的观点已经被普遍接受(Chomsky 1995:35)。问题是宾格指派与主格、空格的指派机制是否具有一致性呢?以下(29)是对相关的三种情形的分析:
(29) a.John tried / consented [CP[C] [TPPRO [Tto] please us]
b.John arranged / desired[CP[C][TPPRO [Tto] please us]
c.John preferred / arranged[CP[Cfor][TPher [Tto] please us]]
d.John wanted / preferred[CP[Cfor][TPher [Tto] please us]]
e.John thought / said[CP[Cthat][TPshe [Twould] please us]]
f.John thought / said[CP[Cthat][TPshe [Twould[ please us]]
可以看出,(29)已经呈现出格位指派的一致性模式:及物性不同的三类标句词成分决定着相关名词短语的格位拼读形式:空标句词[C ]具有空一致特征(null agreement features),传递给[T to]后由其执行空格指派;及物性标句词[C for/ for]具有宾格一致特征,传递给[T to 后由其执行宾格指派;而不及物性标句词[C that/ that]具有主格一致特征,传递给[T would]之后,由其指派主格。Collins(2005:104)认为,PRO主语的空格是由C指派的机制理应和主格、宾格的指派机制具有一致性,而(29)所呈现的一致性完全可以简化为(30):
(30) a.[CP[CΦ ] [TPPRO [Tto ] [VP]]] MODALITY
b.[CP[Cfor / for] [TPACC [Tto ] [VP]]] TRANSITIVITY
c.[CP[Cthat / that] [TPSUB [TAf] [VP]]] TENSE
Chomsky(2008)在语段论中进一步提出,任何类型的标句词都属于语段中心语,承载着所有的不可解读特征,但标句词并非是真正执行格位指派的探针;在句法结构的推导机制开始运作之前,标句词C必须把一致性特征和EPP特征移交给紧接其下的功能语类T,并由T来执行主格指派。而上述(30)中的主格、宾格和空格的指派机制符合语段论对特征传递的要求,因而是最简句法中较为优化的格位指派机制。另外,前文对主语控制、宾语控制以及ECM等动词诸多用法的辨析表明,依据其是否具有显性主语、是否显性标句词并不能完全识别某种语言中的限定性结构和非限定性结构,因为和限定句中既可以有实主语也可以有空主语一样,非限定性也同样既可以有实主语,也可以有空主语;如果限定句的标句词可以为实或为虚一样,非限定句的标句词也可以或为实或为虚。至关重要的是,在非限定句的辨识中,时间信息是否呈现并不能作为唯一的标准,因为通常认可的非限定句是允准时态变化的,如(31)所示:
(31) a.We believe him to be lying. (含有现在时态特征的非限定句)
b.We believe him to have left. (含有过去时态的非限定句)
c.We are hoping for him to take her out. (含有将来时态的非限定句)
d.She seems to be leaving.
e.She seemed to have done her bit.
f.She seems to leave for home soon.
4.结语
英语控制理论研究的焦点都是与PRO有关的一些问题:控制语如何选择、确定和转换;被动化、焦点化能否执行;补语结构中的侧向移位;反身代词的隐性控制;分列式控制结构中语义解读如何实现;部分控制中的成分统制能否实现等(Boeckx et al.2010)。Hornstein(1999)一直主张废弃PRO而采用移位来统一分析提升和控制,但Landau(2013)则基于跨语言的语料证据支持PRO存在的假设。最简句法基于后一种观点认为,限定句是不允准PRO为其主语的,PRO是与非限定性相关的句法主语(Radford 2004),而本文所择选的诸多用例显示,非限定性结构不仅包括各种控制结构,而且还包括其他并非以PRO为主语或者根本就不允准主语的结构。本文仅仅尝试为限定句、for-to不定式以及主语控制结构中的格位赋值机制提供统一分析,而未来的研究更应该考虑如何对其他非限定结构作统一处理。
Boeckx,C,N.Hornstein & J.Nunes.2010.ControlasMovement[M].New York:Cambridge University Press.
Bresman,J.1980.TheoryofComplementizationinEnglish[M].New York:Garlund.
Carnie, A.2011.ModernSyntax[M].London:Cambridge University Press.
Chomsky,N.1995.TheMinimalistProgram[M].Cambridge,Mass:MIT Press.
Chomsky,N.2007.Approaching UG from below[A].S.Uli & H.M.Gartner(eds.).Interfaces+Recursion=Language?[C].New York:Mouton de Gruyter.
Chomsky,N.2008.On phases[A].M.Kenstowicz(ed.).KenHale:ALifeinLanguage[C].Cambridge:MIT Press.
Chomsky,N.2013.The problems of projection[J].Lingua(130):33-49.
Chomsky,N.2014.Minimal recusion:Exploring the Prospects[A].T.Roeper & Speas,M(eds.).Recursion:ComplexityinCognition[C].Switzerland:Springer International Publishing,1-14.
Collins,C.2005.A smuggling approach to the passive in English[J].Syntax(8):81-120.
Culicover,P.1997.PrinciplesandParameters[M].London:Oxford University Press
Culicover,P.2009.NaturalLanguageSyntax[M].London:Oxford University Press.
Davies,W.& Dubinsky.2004.TheGrammarofRaisingandControl:ACourseinSyntacticArgumentation[M].New York:Blackwell. Haegeman,L.1994.IntroductiontoGovernmentandBindingTheory[M].London:Oxford University Press
Horstein,N.1999.Movement and Control[J].LingucsticInquiry(30):69-96.
Landau,I .2013.ControlinGenerativeGrammar:AResearchCompainon[M].New York:Cambridge University Press.
Poole,G.2011.SyntacticTheory[M].Palgrave Macmillan.Rizzi,L.1997.The fine structure of the left penphery[A].L.Haegman(ed.).ElementsofGrammar[C].Boston:MA:kluwer,281-337.
Radford,A.2009.AnalyzingEnglishSentenceStructures[M].Cambridge:Cambridge University Press.
徐烈炯,1984,管辖与约束理论[J],《国外语言学》(2):1-15。
(马志刚:广东外语外贸大学语言学及应用语言学研究中心副研究员)
通讯地址:510420广东外语外贸大学语言学及应用语言学研究中心
*本文系国家社科基金一般项目“汉语特色句式的优化生成及其中介语加工机制研究”的阶段性成果,项目编号15BYY090。
H03
A
2095-9645(2015)03-0028-07
2015-05-03