“了” as an aspect particle and as a modal particle 时态助词“了” 与语气助词“了”的用法 Youxuan Wang 2012-01-23 了 拼音 : liǎo / le 部首 : 乛 笔顺 : 52 This article will first distinguish the two ways in which the word “ 了 ” is used: (a) used as an aspect particle ( 时态助词 shítài zhùcí ), and (b) used as a modal particle ( 语气助词 yǔqì zhùcí ). It will then proceed to explain how they actually used as different types of particle. Non-native speakers find this distinction too subtle to be noticeable, and native speakers who have no training in linguistics find this distinction too intuitive to be worthy of an explanation. However, the student of Chinese as a foreign language is not to be intimidated by the illusory complexity. You only need to know the difference between an aspect particle and a modal particle – The former is a word which is immediately suffixed to a verb to mark the tense of the verb; and the latter is used at the end of a sentence to express an attitude towards what is said in the sentence. By implication, “ 了 ” as an aspect particle and “ 了 ” as a modal particle are actually two different words, and the fact that they look identical and sound identical is merely a coincident. the way each of these two particles is actually used in sentences – You need pay special attention to two issues: (i) word order; and (ii) collocation. In other words, you need take note of where in the sentence each of the two particles is used; and you need pay attention to the question as to whether/how they should be collocated with other words to form a question, an affirmation and/or a negation. This article attempts to explain these two main questions. 1. How to Distinguish Them? 1A. “了” as an aspect particle: • it directly follows the verb; • it expresses a sense of completion , WITHIN the context of the sentence; • it is not used/ required in negative sentences (e.g. negative replies to questions with le ); • it is usually followed by a specific object (eg. something quantified, or not too general), eg.: 我吃 了 三碗饭 wǒ ch ī le sānwǎn fàn 。 I have eaten three bowls of rice. 你到 了 办公室,就给我打电话 nǐ dàole bàngōngshì, jìu géiwǒ dǎdiànhuà 。 Call me as soon as you've arrived in your office. 他洗 了 澡,就去睡觉 tā xǐle zǎo, jìqù shuìjiào 。 He finished taking his bath, and then went to bed. (He went to bed as soon as he finished taking his bath.) 1B. “了” as a modal particle: it comes at the end of the sentence (or sometimes at the end of a clause); it usually expresses a response to a change or unexpected/ undesirable circumstance ; when used with a negative ( 不 or 没有 ) it means, ‘no longer’ . (Note that, by contrast, “了” as an aspect particle, never collocates with “不” or “没有” ). Examples: 我弟弟今年十二岁 了 wǒ dìdi jīnnián shí'èr su ì le 。 My younger brother has celebrated his 12 th birthday this year. (Implication: this is a change in his life.) 对不起,他刚走 了 duìbùqǐ, t ā gāng zǒu le 。 Unfortunately, he has just gone. (Implication: This is something unexpected to the listener.) 我昨天很忙,今天不忙 了 wǒ zuótiān hěn máng, jīntiān bù máng le 。 We was very busy yesterday, but not busy today. (Implication: There is a change in my diary: I’m not as fully booked as yesterday.) 我没有钱 了 wǒ méiyǒu qián le 。 I have run out of money. (Implication: I wasn’t short of cash previously.) (Note that, here, in the sentence “ 我没有钱 了 ” , “ 没有 ” is not an adverb negating the completion of a past action. It is a verb in its own right, meaning “ lack ” or “ having no possession of ” ; and as such, it does not warrant the omission of the particle “了”.) In short, when used as a modal particle, “了” does not mark the tense of a verb, but expresses an attitude towards some change or something unexpected. 2 . How to use them? 2A . As aspect particle (a) When the predicate consists in a verb-object phrase (VOP), “了” is inserted between the verb and object , e.g.: 写字 xiězì 写了字 xiěle zì 上课 shāngkè 上了课 shāngle kè 唱歌 chànggē 唱了歌 chàngle gē (b) As is often the case, if the noun in the VOP is countable, it is to be quantified by means of a measure word: 我 写 了 一百个 字 wǒ xiěle yībǎigè zì 。 I have writ ten down a hundred word s . 我 上 了 一堂 课 wǒ shàngle yītáng kè 。 I have attend ed one session of class. 我 唱 了 一首 歌 wǒ chàngle yīshǒu gē 。 I have s u ng one song. (c) If the countable noun is not quantified, then, (c.i) 了 must be used twice: once between the verb and the object, and once again at the end of the entire VOP (or at the end of the entire sentence, if the VO-phrase is at the end of the sentence). 我 写 了 字 了 wǒ xiěle zì le 。 I have written (down some words). 我 上 了 课 了 wǒ shàngle kè le 。 I have attended class. 我 唱 了 歌 了 wǒ chàngle gē le 。 I have sung. unless (c.ii) the VOP is followed by another statement: 我 写 了 字 就 看 书 wǒ xiěle zì jìu kànshū 。 (I finished writing and then started reading.) I started reading once I had completed my writing. 我 上 了 课 就 打 球 wǒ shàngle kè jìu dǎqíu 。 (I finished my class and then started playing a ball game.) I started playing (basket-)ball once I had finished class. 我 唱 了 歌 就 跳 舞 wǒ chàngle gē jìu tiàowǔ 。 (I finished singing, and then started dancing.) I started dancing once I had finished singing. (d) A case in which the aspect particle “ 了 ” does not appear right after a verb: when the verb is immediately followed by another verb to form a predicate complex with two verbs in series ( 连谓 ). In this case, the actions indicated by two verbs in a series are both completed, and “ 了 ” is used summarily after the second verb. It is not to be used after the first verb; nor is it used twice. 他 去 上 班 了 tā qù shàngbān le 。 He has gone to work. 他们 开始 上 课了 tāmén kāishǐ shàngkè le 。 They have started (to have) their class session. 她 去 同学家 看 朋友 去了 tā qù tóngxuéjiā kàn péngyǒu qùle 。 She went to her (former) classmate's home to meet friends. If you haven't learnt about the construction of predicates in series ( 连谓结构 ) , the easiest way of understanding it at this stage is to think of this English sentence: I came to see you. 连谓结构 is used to write Chinese sentences which translates such English sentences. (e) In two obvious cases, the use of “ 了 ” is not necessary. This is because the notion that the action has been completed or event has taken place has been made explicit by means of a complement of a result indicating the completion, or by means of an adverbial of time spelling out the point of time for the particular action/event in question. For instance, when the notion of completion has been made adequately by the word 完 wán (used as a complement of result immediately following the relevant verb), the “了” can be optionally omitted for sake of stylistic brevity, e.g.: 他吃完晚饭就去看电影。 Alternative, we can say: 他吃了晚饭就去看电影。 Otherwise, stylistically speaking, the word “ 了 ” is used redundantly in the following sentence: 他吃完 了 晚饭就去看电影。 Likewise, if the time of the past action in question has been clearly indicated by an adverbial of time such as “ 昨天下午 ” (or “ 今天下午 ” from the point of view of this evening) , “ 前天 ” , “ 去年 ” , “ 以前 ” etc, “了” would be superfluous. In the Hanyu Jiaocheng , part 1, Book 2 (IIa: p.1) , Zhang Dong says to Tian Fang “ 今天下午 我们 跟政法大学的代表队 比赛 ”, explaining why he failed to call Tian Fang back. Here, the “ 了 ” is, stylistically speaking, quite appropriately omitted. If Zhang Dong were not sensitive to stylistic norms, he could have said something clumsy like the following sentence: 今天下午 我们 跟政法大学的代表队 比 了 赛。 The adverbial of time “ 今天下午 ” warrants the omission of “ 了 ”, if the phrase “ 今天下午 ” in effect refers to a time in the past. (f) If you want turn an affirmative sentence which contains the aspect particle “了” into a negative sentence, you can do so by prefixing the adverb “没有” to the verb whilst dropping the aspect particle, e.g.: 他 吃 了 晚饭 就 去 看 电影。 他 没有 吃 晚饭 就 去 看 电影。 We should never use “不” to negate a verb in the past tense or present perfect tense. For this purpose, we use “没有”; and when using “没有” , we also drop “ 了 ” . To sum up, syntactially speaking, an aspect particle is an integral part of the verb phrase and it serves to indicate the tense of the verb. If the notion of completion is already adequately expressed by the use of the word 完 as a complement of result or by the use of certain adverb of time, “ 了 ” can/should be dropped. 2.B. As modal particle (a) “ 了 ” is usually used at the end of a sentence, e.g: 田芳去哪儿 了 Tián F ā ng qù nǎ'ér le? Where has Tian Fang gone to? (Where is Tian Fang? ) 你下午做什么 了 nǐ xiàwǔ zuò shěnme le? What did you do this afternoon? 我踢足球 了 wǒ tī zúqíu le 。 I went to play football. 你们队又输球 了 吧 nǐmén duì yòu shūqíu le ba ? You lost again, didn’t you? 你学会开车 了 吗 nǐ xuéhuì kāichē le ma ? Have you taken driving lessons and can you drive now? (b) “ 了 ” is not to be used in a negative answer to an affirmative-negative question (…… 了没有? ) , e.g.: 田芳回来了没有 Tián Fāng huíláile méiyǒu ? Has Tianfang come back home? · 回来了。 (affirmative answer.) · 没有回来。 (negative answer.) 张东又给我来电话了没有 Zhāng Dōng yòu géi wǒ láidiànhuàle méiyǒu ? Did Zhangdong ring me again ? · 张东 又来电话了。 (affirmative answer.) · 张东 没有又来电话。 (negative answer.) 你通过考试了没有 tā tōngguòle kǎoshì méiyǒu ? Did you pass your examination? · 我通过考试了。 (affirmative answer.) · 我没有通过考试。 (negative answer.) At any rate, a modal particle is not an integral part of the verb phrase. It does not serve to mark the tense of the verb. Instead, it expresses a response or an attitude towards the proposition expressed by the entire statement. For this reason, it appears at the end of the sentence. 3. Conclusion: The distinction between the use of the word “ 了 ” as aspect particle and “ 了 ” as modal particle: As aspect particle, it indicates that an action (referred to by the verb it immediately follows) is completed. Hence, it is used right after the verb. (It modifies the sense of the verb which it follows.) As modal particle, it expresses the speaker's judgement or opinion about the occurrence of a new situation that the entire statement describes. Hence, it is used at the very end of the sentence to modify the meaning of the entire proposition. (It modifies the meaning of the basic statement it is attached to.) They both share the sense of completeness, though. The aspect particle indicates the completion of an action, whiles the modal particle indicates a change in situation which the entire sentence describes and also expresses a response on the speaker's part. Exercises: Wang, Youxuan. 'Exercises: “了” as an aspect particle and as a modal particle'. Intensive English Blog. 2012-01-25. URL: http://intensive-chinese.blogspot.com/2012/01/exercises-in-using-particle-liao3.html . http://intensive-chinese.blogspot.com/2012/01/used-as-aspect-particle-and-as-modal.html H. Modal particle le 了 used to sum up after a series of actions The modal particle le 了 can be used in the narration of a past situation to mark the end of a series of actions before going on to a new one. In order to maintain the continuity of the narration, it is used at the end of each series of actions, not after each action within the series. The following example shows how the modal particle le 了 marks the end of each series of actions within the bigger picture of what happened after the woman made the phone call. 8.1 Tā dǎwándiànhuà, liúxiàqián, náqǐshūbāo, jiùchūqu le . Zǒudàoxuéxiàodàménkǒur, tāpèngjiànZhāngMěiyīng le .GēnXiǎoZhāngshuō- le yìhuǐrhuà, tājiùqùgōnggòngqìchē-zhàn le . Zǒudàochēzhàn, chēháiméiyǒulái, tājiùdàopángbiānr de xiǎoshūdiànqùmǎibào.Mǎi- le bào, chēyělái le , tājiùshàngchēhuíjiā le . 她打完电话,留下钱,拿起书包,就出去 了 。走到学校大门口儿,她碰见张美英 了 。跟小张说了一会儿话,她就去公共汽车站 了 。走到车站,车还没有来,她就到旁边儿的小书店去买报。买了报,车也来 了 ,她就上车回家 了 。 When she finished the phone call, she left the money, picked up her bag and left. When she got to the school gate, she ran into Meiying Zhang. After taking with her for a little while, she went to the bus stop. When she got to the bus stop, the bus had not arrived, so she went to the little bookstore nearby to buy a newspaper. After she got the paper, the bus came. She got on the bus and went home. Note that the first modal particle le 了 marks the end of a series of actions before the woman went out. This series of actions presents a picture of what she did after making the phone call and before she went out. Le 了 may not be used after each action is completed, or it would break the continuity of the narration. The next series of two phrases presents the picture of her going toward the school gate, where the action was stopped by an accidental meeting with a friend. Therefore, the second modal particle is used there to mark the end of that sequence. The third series of actions presents the picture of what happened after she ran into her friend. This series was concluded by her leaving to head for the bus stop. Therefore the third modal particle le 了 is used. The next series of actions presents the picture of what she did while waiting for the bus. This sequence is concluded with the coming of the bus and was therefore marked by the fourth modal particle le 了 at the end of lái 来 . The whole narrative was concluded by her getting on the bus and finally heading home, which was marked by the last modal particle le 了 . 8.2 Tā dǎwándiànhuà le , liúxiàqián le , náqǐshūbāo le , jiùchūqu le . Zǒudàoxuéxiàodàménkǒur le , tāpèngjiànZhāngMěiyīng le .GēnXiǎoZhāngshuō- le yìhuǐrhuà, tājiùqùgōnggòngqìchē-zhàn le . Zǒudàochēzhàn le , chēháiméiyǒulái, tājiùdàopángbiānr de xiǎoshūdiànqùmǎibào le .Mǎi- le bào, chēyělái le , tājiùshàngchēhuíjiā le . 她打完电话 了 ,留下钱 了 ,拿起书包 了 ,就出去 了 。走到学校大门口儿 了 ,她碰见张美英 了 。跟小张说了一会儿话,她就去公共汽车站 了 。走到车站 了 ,车还没有来,她就到旁边儿的小书店去买报 了 。买了报,车也来 了 ,她就上车回家 了 。 The example above shows how the continuity would be broken by using the modal particle le 了 at the end of each action. Consequently, such narration is unacceptable. Different languages have their own ways to deal with this problem. The following example shows one of the ways English maintains the continuity of a narrative. Note that in 8.3 the subject ‘she’ is not repeated as it is in 8.4. It shows how one sentence alone is used to present a series of actions, which maintains the flow of ideas. 8.4, however, repeatedly uses the subject, ‘she’. As a result, 8.4 is broken into many sentences and becomes a poor and unacceptable narrative. 8.3 When she finished the phone call, she left the money, picked up her bag and left. 8.4 She finished the phone call. She left the money. She picked up her bag. She left. “This sense of summing up a situation or bringing a particular topic to a close before going on to a new one by the use of end-of-sentence/clause le 了 may also be found with nominal comments. Compare the following pairs of sentences: 孩子今年五岁。 Háizi jīnniánwǔsuì . 孩子今年五岁了。 Háizi jīnniánwǔsuì le . The child is 5 years old. 今天星期六。 Jīntiān xīngqī-liù . 今天星期六了。 Jīntiān xīngqī-liù le . It’s Saturday today. The first example of each pair only expresses a fact: ‘the child is 5 years old’ or ‘today is Saturday’. The addition of end-of-sentence le 了 conveys the sense of eventually reaching the present situation or position: the child is (now) 5, and today is (finally) Saturday.” (Yip Po-Ching and Don Rimmington: 2004, p. 321) http://comet.cls.yale.edu/mandarin/content/Lele/grammar/lele-3h.htm
证伪法的改造与语言研究 ——西方语言哲学研究之九 钱冠连 摘要: Popper 的证伪论(或证伪法)长于重大的理论创造,能够挽救有意义的科学命题,但有重大缺陷(完全抛弃归纳法等)。考虑到中国语言学界的传统特点与现状,本文提供两个可操作的“证伪最简模型”,模型之一指出:何时启动证伪程序,如何避免提出明显荒唐的猜想。模型之二是对证伪法的改造:它加进了局部性归纳,又保住了证伪法长于重大理论假设或猜想的优势,还便于具体操作。 关键词: 证伪、最简模型、改造 今天提出科学哲学家Popper的证伪论,对其进取精神与重要缺陷进行认真的吸取与改造,对于突破我们自己的语言研究,有着一定的推动作用。 一、证伪论的内容简介 Popper (Popper, K.R.)指出: 作为与假科学相对的科学,其真谛不在于它提出了一些假设等待证据在某种高的程度上被证实,而在于它的假设能够被证据反驳。这就是说,它们真正地要面对这样的可能性:被验证,因不合经验终被否定。如此说来,科学的方法不是从积累的材料中机械地归纳出一般来,而是形成大胆假设,然后(此假设)服从严格的验证,这是一种猜想与被反驳之法。科学家提出一个有趣的猜想随后被推翻并不为错,错的是提出一个猜想而不允许被反驳或者面对反驳的证据而死守那个猜想。 (Popper 1959; 参见Blackburn 1994: 136) Popper 尔(Popper, K.R. 1959)证伪论是“科学哲学的核心概念”。他从演绎主义的立场出发,反对逻辑实证主义的归纳主义。他认为科学知识不是个别知识,而是普遍知识,不是单称陈述(如“这只天鹅是白的”),而是全称命题(如“所有的天鹅都是白的”,省略形式“天鹅是白的”。即:“任意的X,假定X有性质P” )。单称陈述仅是个别的经验判断。也就是说,科学知识是普遍有效的全称陈述,不能用归纳法从个别事实或单称陈述中得出。他的一个极具震撼力的观点是: 从有限不能证明无限,从过去不能证明未来( 参见Blackburn,S. 1994: 136 ;涂纪亮 1996:330-331)。 未来与无限充满了随机因素。 当某事物运动的一个变化周期(尽管这个周期非常非常长)尚未完结时,不能给那个事物定性,不能对它作出全称陈述。 能满足“普遍有效性”的经验事实必须攘括从过去到未来的全部的或穷尽性数量的事实。但这样做是不可能的,从这里就注定了证实一个全称陈述是毫无希望的。所以,他断言:“理论要得到经验的证实在逻辑上是不可能的。”(Popper, K.R. 1959,参见涂纪亮 1996:330-331) 爱因斯坦在赞同Popper的信中说:“理论不会从观察的结果中构造出来,它们只能被发明。”(参见赵南元 1994:76) 休谟指出:“单个的观察陈述不管数量多大,它们在逻辑上不可能蕴含无限性的普遍陈述。”(Bryan Magee 1988) —— 不能认为有限的实践(经验)可以代替无限的事实,过去的实践(经验)不能代替未来的事实。归纳法的不可靠之一就在如此。等待整体呈现完成过后再去归纳,归纳就得无限期地被迫搁置起来。因为某些事物的整体呈现的过程也许是无限的。这是归纳的局限性之二。 归纳是对准经验事实的归纳,而经验事实总是随时间或地点的变化而变动不定的。这就是归纳法的局限的根源。 那么,科学发现与科学理论的出路在哪里?Popper认为,科学理论虽然不能被事实所证实,然而可以被经验事实所证伪。证伪论的名称由此而来。这意味着科学规律尽管不能证明,但可以检验:通过反驳它们的一系列尝试,它们得以检验。 这就是说,一个猜想或假设被证实尚无法安身立命(不是科学真谛),最可靠的倒是被反驳。一个假设被反驳掉,反而是科学真谛,这一点使许多科学家无法接受(下面还要提及)。 提出大胆的理论假设(猜想)或科学发现的假设,然后等待证伪,如不能被证明为伪,在证伪之前的一刻,这个假设的理论或科学发现就应该被认为是可行的或为真的(以上简介参见钱冠连 2002)。科学家和语言学家有了这点幸运就足够了。 二、它的进取精神与重大缺陷 Popper 认为,归纳推理从来就不曾理性地被证实过。钠经过随机选样一万次,燃烧时吐出黄色火苗,不能据此推论出,所有的钠燃烧都吐出黄色火苗。……归纳论证简直就没有理性的力量(Inductive arguments simply have no rational force),——归纳论证过程中,前提不能衍推出结论,只是意在支持结论。于是,Popper 就寻求完全依赖演绎论证。一方面,我们从来没有哪怕一点点儿正面的理由认为一个假设为真或者可能为真;我们却只能用演绎论证来表明那个假设为伪。只要我们观察到一只黑天鹅,于是我们就可演绎地推测出“所有的天鹅是白的”为伪(此处引号为本文作者所加)。这便是Popper科学哲学的重点与难点。够得上认可的推理,就是抛弃所相信的东西或抛弃一些假设(It is only the rejection of beliefs or hypotheses that can have the sanction of reason. Popper 1959; Newton-Smith 2001:110;黑体为本文作者所置)。 对不了解Popper科学哲学的证伪论的人来说,或者对于只习惯于归纳法的人来说,上面用黑体标明的一些论述,是不可思议的,甚至是非常荒谬的。下面的论述中,这样的不可思议的论述更有甚之,但却放射出进取精神的光芒。 Popper 认为,我们不能也不应该靠归纳推理推进工作。他宣称,一个好的科学家从来不如此。Popper式的科学家,应有丰富的想像,只须作出一个大胆的猜想,而且,越大胆越好(The Popperian scientist, equipped with a fertile imagination, simply makes a bold conjecture and the bolder the better.) 此后,他就以观察和经验寻求反驳,反驳那个猜想。如能找到一个反例(a contrary instance)那个猜想即被证伪,随即便被抛弃。在此情况下,科学家又开始一个新的猜想。如果一个猜想在实验中未能被证伪,那么它就被认为“证实”了(If a conjecture is not falsified in a test, it has been “corroborated”)。按Popper定义,证实不会提供任何理由以便认为那个假设在将来有任何站得住的可能性。那个假设只是报告,它尚未被证伪。(Popper 1959; Newton-Smith 2001:111) 为什么许多科学家会认为证伪法是不能接受的呢? 理由之一:研究者一辈子活着的希望在哪里?你说科学活动的方法,是提出大胆的理论,越大胆越好,此后,科学家的任务就是寻求反驳(反驳这些理论)。一辈子仅仅是献身于找到理论然后发现此理论为伪,这个科学家不是发疯了吗?这是一个什么样的倒霉的使命呢?哪一个科学家会以这样的任务为他一生的使命呢?如果某一位科学家不为科学的真理而只为某一个自己的理论不被淘汰而活着,那他就不会赞成证伪论。哪一个科学家希望自己最珍爱的理论最终也将被证伪?如果科学家发现的全部东西只是某一理论为伪,那么他殚精竭力的研究还有什么样的希望呢?窃以为,科学家的这些责问是可以理解的,但这些说法并不是真正的科学精神。理由很简单:科学家发现某一个理论为伪,也是功劳。而且,抱着一个珍爱的但错误的理论到底,这绝不是科学家的幸福,而是与科学相悖的愚昧。 理由之二,Popper的证伪论可能导致怀疑主义。 理由之三,许多我们认为是科学的东西,原来并非是可证伪的( much of what we count as science turns out not to be falsifiable , Newton-Smith 2001:111 ) 。 Popper 的证伪法的重大缺陷在于完全拒绝归纳法。这一点是不符合人类的思想的规律的。人类科学史表明,一,归纳是一种可行的推理方法:由一系列具体的事实概括出一般原理,与演绎法相对;二,人们通过归纳法也能得到相对的真理。 窃以为,证伪法的进取精神在于,“如能找到一个反例(a contrary instance)那个猜想即被证伪,随即便被抛弃。在此情况下,科学家又开始一个新的猜想。” 一个真正的科学家会从被证伪的前一个理论中学会更谨慎,更科学的态度,开始另一个新的猜想。其实,我们许多科学家其中包括一些中国的语言学家,有的自觉,有的则在不自觉的状态下,不断地猜想——提出一个命题,不断地以少许反例去对自己的命题进行证伪,不断地抛弃,不断地尝试提出新的命题。抛弃了的命题是大多数,相对而言,尝试成功(如果以发表之后在相当长的历史时期之内还没找到反例为标志的话)的是少数。只是他本人并未觉得这就是证伪法。如此而已!同时,Popper的证伪论还是给了人许多希望。他指出:“如果一个猜想在实验中未能被证伪,那么它就被认为‘证实’了”。这就够幸运的了。这就足够了!我们一生还能怎么样呢?那些认为Popper证伪论不过瘾,使人失望的看法,正好表明,他们总是以为自己能发现一个永恒的真理。Popper 向我们推荐一个没有过高奢望的、也是一种科学的诚实态度:一个假设尚未被证伪,就是被证实了,就是一个可信度高的理论。还要怎么样呢?无论是西方的上帝还是中国的观音都不能向任何一个人施舍一个永恒的真理。如果一个科学家(一个语言学家)相信自己能创造一个永恒的真理,那他当初实在是不应该立志当科学家或者语言学家。Popper主张大胆猜测,严格反驳,原因就在于理性可能是错的。所以科学的发展必须经过严格的证伪与反驳。“这样科学知识的增长就是猜想与反驳的过程。这一过程为表示为:问题(P1)—各种尝试性的理论(TT)—消除错误(EE)—新的问题(P2)。即科学家就某一问题(P1)提出各种尝试性的理论(TT),然后通过排除错误而筛选出具有较高逼真度的理论(EE),最后随着科学的发展,这一新的理论被证伪,又出现新问题(P2),从而开始新一轮的循环。科学的发展实质上就是一个不断证伪、不断批判、不断革命的过程,其根本精神就是批判与理性。”(转引自程志民江怡 2003:200)这就是证伪论的进取精神。应该说这是一个以科学进步为第一生命的过程,而不是以自己的声誉为第一生命的过程——以自己的声誉为第一位的人,理论上很可能是抱残守缺者。 本文对Popper的证伪法的改造,就从容忍归纳入手。 三、两个“证伪法最简模型”及其要点 为了观察归纳与演绎的区别,观察归纳与证伪方法(属于演绎法)的区别,我们首先看看归纳的最简要的过程(见图1): 解说:归纳法:从一套经验性事实(以○表示)出发,得到一个新结论( 以◎表示 )。这个新结论可以是经验性的,也可以是理论性的。Popper认为归纳法不可靠因而完全抛弃归纳,是一个失误。归纳法也会得到相对的、局部的认识与真理,因而实证性研究是相当有用的。丢掉它是毫无根据的。后来有人指责“Popper犯了一个逻辑错误,……这使归纳法恢复使用。因此归纳法好像是不可或缺的,当人们把它从门口抛出去,它又从窗户钻进来。”(布鲁诺·雅2000: 176或见Bruno1992) 我认为这个嘲笑是偏激的。本文提出的第二个最简模型中,因为容忍了局部的归纳才能避免提出一个明显荒谬的假设。 两个可操作的“证伪法最简模型”的前提是,都要从一个已知的理论出发(去提出一个新的理论假设或猜想)。这一点很重要。“爱因斯坦相对性的理论不是出于实验,也不源于经验,而是基于一个原理。”(布鲁诺·雅2000: 91或见Bruno 1992) 如果一个研究者不接受从理论到理论这个路子( 爱因斯坦相对性的理论就是这个路子,正是我们中国学者最不敢、最不习惯接受的一条路子。当然,这只是人文科学研究中的 一条路子而已,不是全部),他是绝不敢也绝不会使用演绎与证伪方法的。 为了说明一个已知理论的重要,我们还要指出一个事实:“实验总是以一种思想作为出发点的。伽里略使一个球在斜面上滚动,以发现落体定律,如果没有一种 预想的观念 ,就不可能进行这样一种试验。”(布鲁诺·雅 2000: 182 或见 Bruno1992 ,黑体由本书作者所置)请见证伪法最简模型之一 : 解说:从一个理论前提出发,一步到位地(不必等到搜集经验事实之后)猜测出一个新的理论结论(一个假设或猜想),等待被少许的反例去证伪,即等待被反驳(左右两个问号代表可能出现的少许的反例,随时准备对新的理论结论进行反驳 );在这个新理论结论之下,还可以再演绎出(衍生出)一套新范畴与新概念,它们也等待着被少许的反例去证伪,即等待被反驳。可以看出,可能出现的少许反例站在新假设或者猜测旁边, 此时证伪程序才开始启动。也就是说,证伪程序在一个假设形成之后才启动。 反驳可能使这一新的猜想被证明为伪。一个谬误的全称命题因此而得到纠正。正因为一个假设面临着时时被反例推翻的考验,那些居然没有被推翻的理论猜想才是真正有生命力的理论。而且这种理论往往是重大的理论创造。“令人折服的重大科学理论都要首先经过事实的反驳” (布鲁诺·雅 2000: 176 或见 Bruno1992 ,黑体由本书作者所置) 下面,我们再看证伪法最简模型之二: 解说:(1)证伪法最简模型之二,在起点与终点(形成一个猜想及此后的再衍生)的解释上,在证伪程序何时启动与怎么证伪以及证伪的结果上,与上图(证伪法最简模型之一)完全一样。( 2 )两个模型 不同之处在于,模型之二在一个猜想最终提出之前,增加了少许事实的支持——这里体现了对证伪法的改造,从而可能弥补了也许是完全谬误的一步到位 (因为在着意搜集经验事实之前就一步到位地提出了猜想)。为了保证不出明显为谬误,猜想的提出者在脑子里总会有 早先保存的少许相关事实 (以※号表示)支持这一新理论假设。这里我们发现了, 在演绎中混合着局部归纳,在猜测中结合着局部归纳的过程。 这样提出的假设或者猜想更经得起证伪的考验。 两个证伪法最简模型的 两个要点 :( 1 )两个模型中,证伪程序都是在一个理论假设提出之后才启动。( 2 )为了保证不轻易出现一看就荒唐的猜想,(语言)研究者不应该抛弃归纳法,反而应该运用早先就保存在大脑中的少许的相关事实支持猜想,从而形成了在演绎中混合着局部归纳,在猜测中结合着局部归纳的过程(所以嘲笑“归纳法又从窗户钻进来”是错误的)。 掌握两个要点的好处是:理解如何具体操作证伪法:知道何时启动证伪程序,懂得如何避免提出一看就是荒唐的猜想。 说明最简模型之二的语言理论猜想的一个例子 Chomsky 猜想,人类习得与使用语言的能力的普遍性特征表明,人类语言习得一定有某些遗传成分( some genetic component ),也就是说,婴儿有学习复杂语言的先天倾向( human babies are predisposed towards ),而动物没有(参见Baghramian,M.1999:290)。 他的猜想所根据的事实是极为有限的。他认为,有如下几点可显示人类习得与使用语言的能力的普遍性特征。第一,没有任何其他的生物体使用如人的语言那样的东西。第二,几乎所有的人都是在五岁之前就学会了他们的母语。第三,纵观世界上各种语言,种种语言能力的发展和语言技巧的习得的阶段,看来或多或少都是步调统一的(uniform)。第四,所有语言的说话人看来都是没多大困难地、接触语言使用甚为有限、极少或者几乎没有得到纠正的情况下,习得了复杂的语言技巧的。这样,年岁的幼儿,在极少训练的情况下,一听到复杂的句子就能理解。第五,任何一个语言说话人,从非常年轻的时候起,对于他们以前从未碰到过的句子,就能做出正确的判断。 他做出这个猜想(人类语言习得一定有某些遗传成分)时,做过实验——把一个个的脑袋打开察看过遗传成分并找到过这些遗传成分吗?没有。他不是也绝不可能通过穷尽性的经验事实归纳出一个结论,做出了这一猜想。但是,如果没有在使用演绎之中的局部的归纳即用五个事实支持他的猜想,这种猜想就是一个明显的荒唐。证实的事实太有限,猜想毕竟是猜想。所以剩下的事情就是等待反例去证伪。 还要补充的是: 证实与证伪之间的不对称性 ( 解释见下 ) 有着巨大的优越性,有利于科学研究,当然也有利于语言研究。例如,为了证实“所有天鹅是白的”将世上所有的天鹅都搜集到面前的做法,是一个愚蠢的幻想。怎么办呢?聪明的办法是以几个有限的反例去证伪,例如,一旦发现世上有一只天鹅是黑的,“所有天鹅是白的”结论便推翻了。于是,修改的判断便会是另外的样子:天鹅既有白的,也有黑的。这实质上挽救了一个有意义的命题。这就用证伪论帮助自己做出了一个正确的结论,避免了一个谬误。 证伪论的聪明之处就在于,以极少避开极多。 Popper 认为,从假说与事实相符(即所谓的证实)进而推论假说正确的方法是极有危险性的,这很可能 暗含事先肯定结果 的错误。而从证伪角度,只要找出任何一个事实与假说不相符的情况就可推翻假说。这就是证实与证伪之间的不对称性。——简单地说, “ 证实,所需证据量大(大到无法穷尽)、找证所费时间长(长到无穷);而证伪,所需证据量小(有限,甚至几个反例就够了),找证所费时间短(有限的时间)。”(钱冠连 2002)所以,只要一个反例就能证伪而没有被证伪,那个假设就值得很大程度的信赖了。“以极少避开极多”之法,肯定是聪明之法。 面对一个要否定的全称科学命题,如何挽救它呢? 严格奉行可证实性原则就会把科学理论中的全称命题摈弃于有意义命题之外(因为你得不到穷尽性的事实,就会被迫摈弃这个命题)。Popper为了挽救科学命题,提议将证实原则改为证伪原则:一个命题是有经验意义的,即在原则上可证伪的。上面白天鹅的例子就是一个挽救的例子。这里再举一个小例子:先有全称命题“汉语中所有的句子都包含主题与述谓”,如果将汉语中所有的句子都搜集拢来走证实之路,注定是劳而无功的。用证伪法,找到一个反例,如汉语里的“大姑娘了!”(词组加句调即可成句,“大姑娘”之后没有后续的述谓或评论),就能证明原命题为伪。但原全称命题还是有意义,予以挽救。于是将其修改成:“汉语中,句子既可包含主题加述谓,也可仅持有单一成分。”如果这种挽救熟练地运用于语言研究中,会避免许多谬误,提出可信度较高的命题。事实上许多人成功地使用了证伪法,只是不自觉而已。 Popper 认为,既然任何假设的证实都是不可能的,科学假说永远不能被证实为真,我们能做的 只是增加一个假说的可信度,或者提出更多的假说去与某个假设竞争,淘汰可信度低的假说。 结论: 证伪的功能在于,以极少避开极多(表现了证实与证伪之间的不对称性)的办法去证伪一个全称判断,并将这个被证伪了的全称命题(如果它有意义)修改成一个可信度较高的命题(即挽救了一个有意义的命题)。 考 虑到中国语言学界的传统特点与现状,本文提供两个可操作的“证伪最简模型”。模型之一指出:何时启动证伪程序,如何避免提出明显荒唐的猜想。模型之二是对证伪法的改造:它加进了局部性归纳,又保住了证伪法长于重大理论假设或猜想的优势,还便于具体 操作。 值得重申的是,中国语言学家如果用证伪论去改造自己的语言研究,利用证伪论长于重大的理论创造的特点,在应用归纳的同时也尝试演绎与猜想,对于中国语言学界说不定是一个有意义的推动。 参考文献 Baghramian, M. 1999. Modern Philosophy of Language, Counterpoint, P.O. Box 65793,Washington, D.C. Blackburn,S. 1994. Oxford Dictionary of Philosophy . Oxford University Press. Bryan Magee. 1988. 《开放社会之父》,南砚译,湖南人民出版社。转引自赵南元, 1994: 72. Bruno J. 1992. Invitation a la philosophie des sciences, ( c ) Editions du Seuil. 见(法)布鲁诺·雅:《科学哲学》,张莹 译,北京大学出版社, 2000 年 4 月版。 Newton-Smith W.H. 2001 Karl Popper, in A Companion to Analytic Philosophy by A. P. Martinich and David Sosa, Blackwell Publishers Ltd 108 Cowley Road Oxford OX4 1JF UK. Popper, K. R. 1959. The Logic of Scientific Discovery . Hutchinson, London. 程志民 江怡 2003《当代西方哲学新词典》,吉林人民出版社。 涂纪亮 1996《现代西方语言哲学比较研究》,北京:中国社会科学出版社,。 赵南元 1994《认知科学与广义进化论》,北京:清华大学出版社,。 钱冠连 2002 证伪论与语言研究——西方语言哲学研究之七,现代外语,2002(2) The Reconstruction of Falsification Linguistic Researches in China Q ian Guanlian ( Guangdong University of Foreign Studies, Guangzhou, 510420, China ) Abstract: The falsification method proposed by K. R. Popper is good at grant theoretical creativities, able to rescue some significant propositions of sciences, although fraught with great disfigurements shown mainly in utterly rejecting induction and so on. On the basis of the traditional features and the status quo in linguistic researches at home, this paper provides the readers with the two operating “models for falsification in the simplest way”, making an attempt on reconstructing the falsification. One of the two is to demonstrate when a falsification procedure starts and how an obviously absurd hypothesis can successfully be avoided. The other is the reconstruction of the falsification, which adds local induction to the falsification procedure and keeps on the strong point inherent in the falsification to put theoretically important hypotheses. Key words: falsification, models in the simplest way, reconstruction 证伪法的改造与语言研究 ——西方语言哲学系列研究之九 作者简介:钱冠连(1939—)男, 教育部人文社科重点研究基地(广东外语外贸大学)专职教授, 博士生导师,全国语言文字标准化技术委员会外语分会;中国英汉语比较研究会副会长。研究方向:理论语言学、西方语言哲学、语用学。主要作品: 《美学语言学:语言美和言语美》、《汉语文化语用学》、 《语言全息论》。现正主持教育部 人文社科重点研究基地 2002 年项目“语言:人类最后的家园:人类基本行为的哲学与语用学研究”。