威廉·拉波夫（著） 石锋（译）| 把语言学建立在坚实的基础上——《语言变化原理》中文版序言

欢迎关注我们，一站式分享海量语言学术资源

敬请星标应用语言学研习，喜欢请点赞，真爱请分享⭐

我很高兴石锋教授主持把《语言变化原理》译成中文出版。我希望这将有益于中国的语言演变的研究，使富有成果的汉藏系语言研究的悠久传统得到增强。

语言演变的普遍理论是以在纸上、羊皮纸或竹片上存留下来的文字材料为依据建立起来的。历史语言学家研究因偶发的历史事件而留存的文献残卷，已经在历史语音的构拟中取得了令人惊奇的成功。我怀着敬佩之意把他们的研究称为：对于不完善的资料进行最充分的利用。这套书则是希望说明，他们的结论可以用进行中的音变证据来加以丰富并再现活力，这是基于记录日常生活中即兴语言的研究中得到的。

我第一次到中国是在1973年参加美国语言学会访华团。这次交流之后出版了《中华人民共和国的语言和语言学》（W. Lehmann主编，德克萨斯大学出版社）。我那个团组中的学者们对于中国和中国语言学的了解都比我强得多。我们访问了很多城市，了解到当时为保护少数民族语言和推广普通话，制定出各种结合实际考虑周密的语言政策。我还从他们正式讲座的提要中学了一些汉语，并被夸奖学到“两条腿走路”的意思，我自己工作中也有相似的情况，这可以作为西方语言学中的一个实例。

当时，极为重视语言研究的社会意义。我记得曾听说有一个研究课题是要学生们采访上海港的工人，以便证明革命京剧《海港》中的语言的正确性。然而，由于当时条件限制，并没有像今天这样的方式进行的系统的语言学研究。

在本书中展示的研究工作有两个方面。一方面是显示了日常生活中语言变异的研究如何证实我们对语言总体是一个系统的理解。汉藏系语言的发展在很多方面反映出跟本书各卷中涉及的英语、西班牙语、以其他印欧语言研究得到的驱动力量是相互对应的。例如，省力原则使单词尾部信息量逐渐减少的效果是相同的。汉语的历史上有着大量跟这一过程相对应的重新调整现象。

另一方面，正在进行中的音变的社会分布让我们对于我们所生活的社会有更多的了解。用于在美国划分正在进行中音变的社会经济层级，当然不同于在中国的城市中随机样本的结构。但是我们仍然期望《语言变化原理》书中展现的年龄、性别、以及城市/农村的分析维度，会在汉语的社会语言学研究中有着同样的重要性。

在任何这种社会语言学的中心地带，我们都希望能够发现支配着本书各卷中所报告的研究工作的两个基本原理。

1、方言口语的中心性。一种语言的历史就是在家庭和亲友间使用的方言口语的历史。它从父母传递给儿童，并随着儿童离开父母的影响而发生变化。

2、观察者的悖论。为捕捉到一种方言口语的准确的记录，我们必须去观察人们在没有被观察情况下是怎样讲话的。

这第二个原则突显出形式语言学与语言变异研究之间的区别，前者的语料局限于母语说话人的直觉，而后者则是要找出语言直觉跟日常生活实际说话的不匹配现象。当然，在大多数情况下，我们的直觉跟我们实际讲话是一致的，然而当二者不一致时，重要的问题就出现了。形式语言学抛弃人们在直觉下实际所说的话语，而社会语言学家却走上相反的道路。当研究中包含了系统性的语音变化时，这个问题的性质就变得更为清晰。在音变之初，说话人对于新的形式并没有察觉。随着变化的进行，他可能会有所察觉，但是会归入脱离实际的旧规范。于是，如果直接提问，得到的回答，只是显示出人们对于自己所讲的话所知甚少。

我很抱歉这套书如此之长：内容之多以致我自己都不能够完全记住。然而我却希望本书各卷显示出对日常生活中使用的语言进行客观研究将得到多么大的收获。

本书第一卷把这种研究方法用于探寻语言变化的内部制约因素。书中最后一节集中论述规则性的问题，以及词汇扩散理论的新见解，这是来源于王士元教授跟他的同事们开创性的研究工作。大量的讨论是分析汉语方言中逐字逐词的音变证据，例如，潮州话的声调变化，或是上海话的元音合流。主要的努力是怎样将这些强有力的汉语方言的重要资料去跟新语法学派的规则性音变的观点相协调。

第二卷着眼于决定语音变化的社会因素，大多是报告集中在美国费城一个城市所做的全面研究，但同时也在探索性别和社会阶层制约影响语言变化的普遍原理。在最后一节提出了一种代际变化增量的抽象模式，描述音变怎样在父母传递给儿童的过程中发生。

第三卷先是论述认知因素与语言变化在跨方言的理解中的效应。主要的发现就是特定的音变确实受到理解方面的干预，甚至这就是音变背后的推动力，从而成为更为紧迫的需要研究的问题。随后的一系列章节展示一种语言演变的全部历史：从它的起始到最后的终结。进而提出一种方言分化的普遍模型。最后一节旨在解决另一个历史语言学中长期存在的谱系树模型与波浪式发展的模型之间的对立。书中讲到，谱系树模型是父母与儿童之间一系列不间断传递的结果，是儿童语言学习能力的结果，从而保持了系统的完整性。另一方面，波浪式模型被视为成人之间传播的结果，反映出老年人有限的语言学习能力。

因此，我在本书各卷中试图解决历史语言学领域中两个长期悬而未决的争议。我的基本立场是确认语言学作为语言科学的心智价值，并且认为如果一种观点被几代语言学家所长期坚持，那么它必须建立在一种坚实的基础上。它会跟对立的意见保持联系，并在适宜的范围里保留自己的见解。

威廉·拉波夫

2013年1月4日

PREFACE TO THE CHINESE EDITION

I am very pleased that Principles of Linguistic Change is now available in the translation into Chinese by Professor Shi Feng.  I hope that it will be useful in the study of linguistic change in China, reinforcing the longstanding tradition of linguistic science in which Sino-Tibetan studies have contributed so much.

The general theory of language change is well founded on the study of surviving texts, on paper, parchment or bamboo. Historical linguists have had astonishing success in reconstructing the past from the scraps that have survived, by haply historical accident.  I have characterized with admiration their art as“making the best use of bad data”. These volumes hope to show that their conclusions can be enriched and reinvigorated by evidence derived from the study of change in progress, based on recordings of spontaneous speech in everyday life.

My first trip to China was in 1973, with the delegation of the Linguistic Society of America. This interchange led to the publication of Language and Linguistics in the People's Republic of China.  (W. Lehmann, ed.: Austin: U. of Texas Press.1975) I was in the company of scholars who knew much more about China and Chinese linguistics than I did.  We visited many cities and learned a great deal about the pragmatic and well-considered language policies of the time, in regard to the protection of minority languages and the generalization of pu tong hua. I had acquired enough of that language to follow the gist of formal lectures, and was encouraged to learn that the expression “walking on two legs” included some acquaintance with my own work as an example of Western linguistics.

At that time, the social significance of research was strongly brought to the fore. I remember hearing of one research project in which students interviewed workers on the Shanghai docks in order to demonstrate the correctness of the language heard in the revolutionary opera, “On the Docks”.  However, conditions at that time did not favour systematic linguistic research in the way that it can be conducted today.

The work presented in this volume has two faces.  On the one hand, it shows how the study of variation in the language of everyday life can illuminate our general understanding of language as a system. There are many aspects of the development of Sino-Tibetan languages that respond to the same forces that are operating in English, Spanish, and the other Indo-European languages studied in these three volumes. For example, the principle of least effort has similar effects in gradually reducing the amount of information at the ends of words, and the history of Chinese shows massive readjustments in response to this process.

In the second aspect, the social distribution of change in progress tells us much about the society we live in. The socio-economic categories used to trace change in progress in the United States are of course quite different from those that would be used in the construction of a random sample of a Chinese city. But we would expect that age, gender, and the urban/rural dimension would be as important in a Chinese sociolinguistic study as in Principles of Linguistic Change.

At the heart of any such sociolinguistic study we would expect to find the two basic principles that govern the work reported in these volumes.

1. The Centrality of the Vernacular. The history of a language is the history of the vernacular form of the spoken language as used with family and intimate friends, transmitted from parents to children and changed by children as they move beyond the influence of their parents.

2. The Observer’s Paradox. To capture an accurate record of that vernacular we must observe how people speak when they are not being observed.

This second principle underlines the difference between formal linguistics that limits its data to the intuitions of the native speaker, and the study of language variation which is prepared to find a mismatch between those intuitions and what is said in everyday life. Granted that in the great majority of cases, our intuitions correspond to what we say, the crucial questions arise when they do not.  The formal linguist discards what people say in favour of their intuitions, while the sociolinguist takes the opposite road. The issue becomes much clearer when systematic change is involved. At the beginning of the change, speakers have no awareness of the new forms. As change progresses, awareness may arise, but in the form of a stereotype far removed from reality, so that direct questioning only reveals how little people know of what they say.

I am sorry that these books are so long: there is so much in them that I cannot myself remember all of it.  Yet I hope that these volumes will indicate how much can be learned by the objective study of the language used in everyday life. The first volume takes this approach to the search for the internal constraints on linguistic change.  Its final section focuses on the question of regularity, and the new insights on lexical diffusion that stem from the creative work of William S.-Y. Wang and his colleagues. Much of the discussion deals with the evidence of word-by-word change in Chinese dialects, such as the development of tone in Chao-Zhou, or vowel merger in Shanghai. The major effort is to reconcile the Neogrammarian view of regular sound change with this powerful and important data from Chinese dialects.

The second volume looks at the social factors that determine change, largely reporting on the intensive study of a single American city, Philadelphia, but at the same time searching for general principles that govern the effect of gender and social class on language change. Its final section presents an abstract model of the incrementation of change across generations, showing how it is generated by transmission from parents to children.

The third volume turns first to cognitive factors and the effect of language change on comprehension across dialects. The major finding is that sound change in particular does interfere with understanding, and makes even more urgent the problem of the driving force behind these changes. It then presents a series of chapters on the life history of a linguistic change:  from its origins to its final endpoint. A general model for the sources of dialect divergence is proposed. The final section aims to resolve another longstanding opposition in historical linguistics, between the family-tree model and the wave model of development. It is proposed that the family-tree pattern is the result of an unbroken series of transmission from parents to children, preserving the integrity of the system as a consequence of the language-learning abilities of children. The wave model, on the other hand, is seen as the consequence of diffusion among adults, and reflects the limited language learning abilities of older speakers.

Thus in these volumes I have attempted to resolve two longstanding controversies in the field of historical linguistics. My basic position is to recognize the intellectual value of linguistics as the science of language, and argue that if a point of view has been maintained over time by many generations of linguists, it must rest on a solid foundation. It remains to bring opposing views into contact, preserving its insights in its proper domain.