Halliday的子句复合体思想及其理论缺陷

朱   跃

内容摘要:美国语言学家Halliday把子句复合体与传统语法的句子相提并论,且认为子句复合体是由中心子句和其它修饰性子句组成。他从相互依赖性的角度把子句复合体中各子句的关系划分为主从关系和并列关系。他认为在主从关系中子句复合体中的成分子句的次序可以易位,且在逻辑关系上存在不对称性;在并列关系中,各成分子句是自由的,且在逻辑上也具有对称性。Halliday从逻辑-语义关系角度对首要句和次要句进行了破析,指出它们之间存在着扩展和投射系统。本文首先对Halliday的子句复合体思想进行简要概述,然后从中心子句与修饰性子句,主从关系与并列关系以及逻辑-语义扩展关系等方面对他的子句复合体理论进行探讨以揭示其理论上的不足。

关键词:子句复合体  主从关系  并列关系  扩展 

Abstract: This paper aims at the general introduction of Halliday’s clause complex theory and its weakpoints. Firstly, his definition of a sentence as a “ Head + modifier” may produce two clause complexes whose structures are not at the same structural level. Therefore, the use of a comprehensive term “ clause complex” will not effectively reveal the different layers of these two structures. Secondly, in hypotaxis , not all the primary and secondary clauses are reversible in their orders, as Halliday believes they are. In addition, Halliday’s conclusions concerning the differences in the freedom of elements in parataxis and hypotaxis are a result of adopting different evaluative standards. Finally, his clause complex theory , though able to reveal the logico-semantic relations of some of the clause complexes containing conjunctions, is not effective in displaying the relations implied between the constituent clauses in a clause complex.

Keywords:  clause complex,  hypotaxis,     parataxis,    expansion

1.        Halliday “子句复合体”思想的概述

Halliday的功能语法是以子句为基础,并在此基础上揭示英语语言的系统功能。他摈弃了句子作为语言分析的句法层次,以子句复合体代替传统语法中的句子层次。他(1985:193)在功能语法一书中指出:“A sentence will be defined, in fact, as a clause complex. The clause complex will be the only grammatical unit which we shall recognize above the clause.” 他之所以用子句复合体来定义传统语法中的句子是因为他认为只要研究子句复合体就能揭示句子功能的组织结构关系,因此他把子句复合体看成是语法单位的最高形式范畴。为了说明子句复合体之间的关系,Halliday举出了一个他认为很典型的例子: 

(1)   I don’t mind if you leave as soon as you’ve finished as long as you are back when I need you. 

他把这一例句中的各子句关系用下图[i]来表示:

   

            Head                               Modifier

 

          α

 

                 β

 

 Head

Modifier

   

Head

Modifier

 

------------------------------>

   

------------------------>

 

αα

αβ

αγ

   

βα

ββ

 
   

                                                                        ( Figure I )

其中αβγ为希腊字母,表示各子句之间的渐进的修饰关系。这一子句复合体分为中心子句(Head)与修饰性子句(modifier)两个模块,分别用α与β表示,而表中的αβ所揭示的是该子句处在α模块中的第二个子句,且与前面的子句α存在修饰与被修饰的关系,依次类推。具有主从关系的子句复合体中的两个模块可以易位,易位后的两个模块α与β间的修饰关系不变, 如图[ii]所示:

   

            Modifier                              Head

 

          β

 

                 α

 

 Head

Modifier

   

Head

Modifier

 

-------------------->

   

------------------------------->

 

βα

ββ

   

αα

αβ

αγ

 
   

                                                                ( Figure II )

Halliday在研究子句复合体时引进了“相互依赖性”的概念。在相互依赖性的类型中,各成分子句之间存在着主从关系(hypotaxis)或并列关系(parataxis)。他认为语言中的一切逻辑结构关系都可以表现为这两种关系。在对子句间的相互依赖性进行研究时,他又引进了首要子句(primary)和次要子句(secondary)的概念。 在并列关系中, 首要子句为起始句(initiating),次要子句为继续子句(continuing),两个子句间具有逻辑上的对称性;在主从关系中,首要子句为控制句(dominant),次要子句为依赖句(dependent),两个子句间不存在逻辑上的对称性,如下图[iii]所示:      

 

首要子句 (primary)

次要子句 (Secondary)

并列关系 (parataxis)

起始子句 (initiating)

继续子句 (continuing)

主从关系 (hypotaxis)

控制子句 (dominant)

依赖子句 (dependent)

                                             (Figure III )

Halliday又从扩展(expansion)和投射(projection)两个角度对子句复合体中各成分的逻辑语义关系进行了分析。Halliday的扩展概念具有方向性,即,次要子句对首要子句的扩展,它包括阐述(elaboration),延伸 (extension),和发挥(enhancement)。 投射是次要子句通过首要子句表示述说(locution)或思想(idea)。Halliday(1985:196-197)把扩展各成分定义如下: 

Elaboration:   …restating in other words, specifying in greater detail, commenting or exemplifying.

Extension:    …adding some new element, giving an exception to it, or offering an alternative.

Enhancement:  …qualifying it with some circumstantial feature of time, place, cause, or condition. 

Halliday 把他的扩展理论举例说明如下[iv]: 

   

(i) paratactic

(ii)hypotactic

expansion

(a)

elaboration

John didn’t wait;

he ran away.

John ran away,

which surprised everyone.

(b)

extension

John ran away,

and Fred stayed behind

John ran away,

whereas Fred stayed behind.

( c )

enhancement

John was scared,

so he ran away.

John ran away,

because he was scared.

                                                                 ( Figure IV )

2. 对Halliday子句复合体理论的基本评估: 

2.1 关于中心子句与修饰性子句

Halliday功能语法的特征之一就是使得子句复合体成为子句以上的语法范畴。他认为通过对子句复合体各成分子句关系的研究便可以揭示句子的功能组织。但是Halliday对子句复合体的定性使人难以接受子句复合体可以替代传统语法范畴中句子层次。他(1985:192)把子句复合体定义为: “Head clause together with other clauses that modify it.”,即中心子句和修饰子句的复合体。Halliday所说的中心子句和修饰子句不同于传统语法中的主句与从句的概念,因为在传统语法中,主句是自由形式,在语义结构上可以独立于从句而存在, 而 Halliday的中心子句既可以是自由的,也可以是依附性的,这从Figure I和Figure II中他对子句的分析便可清楚地看出:

(2)           I don’t mind if you leave as soon as you’ ve finished

                              H

                     as long as you back when I need you

                              M

                  I don’t mind  if you leave as soon as you’ve finished

                      H                     M

                  as long as you are back  when I need you.

                          H                M

在他所列的典型的例句中, 子句 “ I don’t mind if you leave as soon as you’ve finished”为中心子句, 子句 “as long as you are back when I need you”为修饰性子句,而在修饰性子句中 “as long as you are back”又是一个中心子句, “when I need you”是修饰性子句。这就产生一个问题。首先,如果我们接受Halliday观点把传统句子理解为“中心子句+修饰性子句”的子句复合体,那么 “as long as you are back when I need you” 就会满足Halliday的定义,然而它的句法功能显然不同于 “ I don’t mind if you leave as soon as you’ve finished,” 因为前者在句法结构上具有依附性,不能独立存在,起着修饰功能,而后者 “I don’t mind” 不具有依附性,且在结构上是自由的。把一个在结构上具有依附性的子句与一个在结构具有自由性的子句同时冠以中心子句,必然会导致语言结构逻辑混乱。

如果说Halliday的中心子句与修饰性子句之间具有层次性,一个是传统语法意义上的主句与从句的关系层次,另一个是Halliday意义上的子句复合体之间的关系层次,且在后一种关系中允许中心子句在层次上低于传统语法中主句意义上的层次,那么就应该通过不同的标识把不同层次的中心子句区别开来。 既然存在不同的结构层次,把它们一味称为 “子句复合体 ”掩盖了子句复合体在结构范畴上的层次性,不利于反映语言层次的真实的面貌,因为Halliday的 “子句复合体”既可以是: 

(3)    I don’t mind if you leave as soon as you’ve finished.

              Head                 Modifier

也可以是:

 ( 4 )     as long as you are back when I need you

                Head            Modifier 

如果说在例(3)意义上的子句复合体可以揭示传统意义上的句子结构功能关系的话,那么例( 4 )意义上的子句复合体显然在层次上高于子句,而又低于传统意义上的句子。这两个子句复合体存在着本质上的区别,因此把子句复合体等同于句子是不合适的。 

2.2 关于主从关系与并列关系 

Halliday把子句复合体中成分子句关系分为主从关系和并列关系。在主从关系中,次要子句可以移至首要子句之前,但是无论置前还是置后,次要子句都在结构功能上依赖首要子句,如:

(5)                 John ran away, whereas Fred stayed behind.    α^ +β

(6)                 Whereas Fred stayed behind, John ran away.    +β^α

Halliday有关主从关系中子句相互易位理论显然以偏概全,因为并非所有主从关系子句都可以相互易位。这从Halliday本人所列举的有关扩展的例句中便可以看出 (见:Figure IV): 

 (7)            John ran away, which surprised everyone.  α^ =β 

这一例句并不符合Halliday的易位理论: 

(8)                                                        *Which surprised everyone, John ran away.   =β^α 

国内一些语言学家把以上例句看成是并列的关系,他们认为首先子句“which surprised everyone”可以独立成句5

(9)                     Every bottle of “Buzz” has been tested in our laboratories. Which is why we are so sure of our guarantee.

其次,在以上例句中, “which” 可以用 “and this”来替代,如: 

 (10)          John ran away, and this surprised everyone. 

这种观点同样值得商榷。首先所提供的例句只是一种语用意义上的修辞性句子。这种句子是为了强调 “which”从句的结果。如果说 “which”子句是真正意义上的句子,那么以下的例句将是荒唐的,因为它意味着逗号可以连接两个并列句: 

(11)  John ran away, which surprised everyone. 

如果把 “which”等同于 “and this”,这只是一种释义行为。如果说 “which” 真的在意义上等同于 “and this”,这就意味着 

(12)           *And this surprised everyone. 

同样是独立的句子,这显然是不可接受的,更何况释义行为具有一定的开放性。如果用释义来解释句法结构,那么许多主从结构句都可以释义为并列句了,如: 

(13) The afforestation movement rapidly spread, as had been expected.  (主从关系) 

把 “as” 释义为 “and this”后,便成了并列关系: 

(14)       The afforestation movement rapidly spread, and this had been expected.  

同时并列关系句也可以释义为主从关系句,如: 

(15) This is where he spends a great deal of his free time for he has nowhere else to go. 

把 “for”释义为 “because” 后成为主从关系句: 

(16) This is where he spends a great deal of his free time because he has nowhere else to go.  

Halliday(1985:198)在讨论并列关系和主从关系时把子句的自由性作为判别标准。所谓自由性是指子句独立成句的特性。他认为在并列句中,两个并列子句具有相同的地位,因为它们都是自由的,而在主从句中,只有控制子句是自由的,依赖句不能独立成句,因而是不自由的。Halliday的观点显然受传统语法的影响。笔者认为并列句和主从句的不平等性是执行结构分析双重标准所致,请看下列例句:

( 17 )

I bread the poultry

and

my husband looks after the garden.

( 18 )

He was absent

for

he was ill.

( 19 )

He was absent

because

he was ill.

 在以上例句中 “and”, “for” 和 “because”都是连词。在传统语法中, “and” 和“for”引导并列关系, “because”引导主从关系。事实上,如果以连词为中心轴,那么连词左右两个子句均是自由的。如果把连词与第二个子句联系起来看,那么第二个子句都不是自由的:

  (20)             * And my husband looks after the garden.

  (21)             * Because he was ill.

  (22)             * For he was ill.

从这个意义上说,并列关系与主从关系在自由性上并没有本质区别。Halliday认为并列关系中的首要子句和次要子句具有自由性显然把连词排斥在子句复合体的成分以外,而在讨论主从关系句时又把连词包括在子句复合体的成分之内,执行了句法分析上的双重标准。在双重标准的前提下去分析对比并列关系和主从关系的本质其可信性令人怀疑。

笔者认为Halliday在分析主从句和并列句时采用双重标准,并不意味着笔者认为这两种关系在结构分配上没有区别。在并列关系中,第二个子句在结构的分配上自由性较小,如:

(23)                    Because he was ill he was absent.

(24)                  * For he was ill he was absent. 

(25)                    It was because he was ill that he was absent.

(26)                  * It was for he was ill that he was absent. 

(27)                    I left because I was nervous and because I wanted to go.

(28)                  * I left, for I was nervous and for I wanted to go. 

Halliday(1985:198)在阐述子句复合体并列关系时指出: “In principle, the paratactic relation is logically (i) symetrical and (ii) transitive. This can be exemplified with the ‘and’ relation.” ,即并列关系在逻辑上体现了对称性和传递性,例如:

(29)    “salt and pepper”  蕴涵着 “pepper and salt”

(30)    “salt and pepper”, “pepper and mustard”  蕴涵着    “salt and mustard”

同时他指出,这种自然语言的逻辑成分不能概括为非语言性的基本逻辑关系,因此不能把 “bread and butter” 与 “butter and bread”视为同义语。Halliday并没有明确区分自然语言逻辑成分与非语言性基本逻辑关系,但是从他所举的例子中可见,他的所谓自然语言逻辑成分似乎是纯形式上的,而非语言性基本逻辑关系是语义和语用上的逻辑关系。如果说

(31)                     I breathe when I sleep.

 不蕴涵

(32)                     I sleep when I breathe. 

 那么下面的主从句在逻辑上却是对称的:

(33)                     John ran away while Fred stayed behind.

(34)                     While John ran away, Fred stayed behind. 

(35)                     As he came in, I went out.

(36)                     As I went out, he came in.

由此可见,逻辑上的对称性并不是并列句与主从句的区别性特征。 

关于逻辑语义关系:扩展

Halliday认为首要子句和次要子句之间的关系可以从扩展与投射两个角度加以描述。其中扩展逻辑语义关系又可以进一步分为阐述,延伸和发挥。但是他对阐述,延伸和发挥的判别标准却是模糊的,语言使用的灵活性使得这些判别标准无所适从,如:

(37)                  John ran away, which surprised everyone. 

Halliday认为这个子句复合体属于阐述的范畴,因为它们两个子句之间的关系是“i.e.”关系,即次要子句是对首要子句的评述性言语,但是评述本身也是一种新信息的提供,那么我们如何在阐述和延伸的新信息之间找出它们的区别性特征呢? 

(38)        I said you looked like an egg, sir; and some eggs are very pretty, you know.

                         ( 延伸 )

(39)        They decided to cancel the show, which upset everybody alike .

                         ( 阐述 ) 

Halliday把例(38)看作是延伸是因为他认为例(38)中的两个子句不存在语义关联,而在例(39)中两个子句在语义上是有关联的,这种关联是通过 “which” 把第一个子句中的内容联系在一起。再看以下例句:

(40)              I have some news for you; John has arrived 

例(40)是阐述关系。在例(40)中, “John has arrived”  是对“news”的阐述;但是问题是有些子句语义的关联十分微妙,很难说它们属于何种类型:

(41)              That clock doesn’t go; it is not working.  ( 阐述 )

Halliday把例(41)看作为阐述类型,即它们之间的关系是 “i.e”的关系,事实上这一例句中标点符号 “;”的使用本身又允许这两个子句在语义上存在如下关系:

(42)                That clock doesn’t go, for it isn’t working.

(43)                That clock isn’t working, for it doesn’t go.

换句话说,这两个子句之间的关系可以是发挥的关系。分号这种特殊的引导因果关系的功能随处可见:

(44)                Joan and Mary were obviously weary; I got up to go. 

再看以下两组例句: 

(45)            I bread the poultry and my husband looks after the garden. ( 延伸 )

(46)            I bread the poultry, so my husband looks after the garden. ( 发挥 ) 

Halliday认为例(45)是延伸关系。问题是 “and”作为一个并列连词所具有的语义往往受两个子句的内容的影响,如:

(47)             Jane ! You have perplexed and pained me. 

“perplexed” 和 “pained”之间显然蕴涵因果关系,因为如果把两个动词的位序颠倒过来,语义便不自然:

(48)             * Jane! You have pained and perplexed me. 

所以例(45)的关系完全可以是一种发挥的关系。Halliday(1985:.207)认为例(45)中两个子句之间不存在任何形式上的因果和时间关系,它们之间纯粹是信息的叠加关系。这种观点令人费解。也许Halliday是从纯语言角度来分析两个子句之间的关系,而没有考虑人所处的经验世界或说话的语境,但是正是由于缺乏说话时所处的语境才使得子句间的关系有可能有多种解释。

总之,Halliday的子句复合体思想并不完善,在语言子句复合体结构分析上逻辑混乱,以偏概全;在主从关系和并列关系的研究中执行双重标准,自相矛盾;在逻辑-语义分析中,阐述,延伸和发挥的概念还不具有区别性的特征,使其理论的操作具有一定的局限性。


注释

[i] 见Halliday .1985.  “ An Introduction to Functional grammar.” Edward Arnold. 194页. Figure 7-2

[ii] 见Halliday .1985.  “ An Introduction to Functional grammar.” Edward Arnold. 194页. Figure 7-3

[iii] 见Halliday .1985.  “ An Introduction to Functional grammar.” Edward Arnold. 195页. Table 7(1).本表略

  有改动。

[iv] 见Halliday .1985.  “ An Introduction to Functional grammar.” Edward Arnold. 197页. Table 7(2). 本表略

  有删节。

见《现代外语》1998,第1期,第4页。 

                                               文献目录

Halliday, M.A.K 1985. An Introduction to Functional Grammar. Edward Arnold.

Halliday ,M.A.K.1985. An Introduction to Functional grammar.( 2nd ed. ) Edward Arnold.

Quirk, R.,S. Greenbaum, G. Leech & J. Svartvik. 1985. A Comprehensive Grammar of the English.

      London:Longman.

黄国文 (1998)《英语“Wh-继续分句”的功能分析》《现代外语》. 第2期,2-10页。