Theories of Language

There have been different theories of language. Essentially, they are the different perspectives to view the diverse and complex word of verbal communication. Below is a brief chronological description of different language theories.

In the classical Greece and Rome, the focus of language study was the effective use of language, especially on rhetorical structures and stylistic excellence. They had a mystical view of language because they found language had a supernatural power in cursing, blessing, divining and exorcism. They believed language had inherently embedded insights on the nature of reality.

About at the same time there was a different approach to language emerged in India, as exemplified by the Sanskrit grammar of Panini. This grammar, developed sometimes in the period of 352 to 150 B.C., details the features of sound, word formulation and syntax of classical Sanskrit.

During the medieval period language became the tool of philosophy and logic, and Latin grammar became the model for pedagogical grammars of most Western European languages.

At the end of the 18th century the discovery of the Panini’s Sanskrit grandma had greatly changed the scholarly views about language in Western Europe. Scholars realized that Sanskrit was closely related to almost all Western European languages. This encouraged a period of intense and exhaustive study on the history of and differences between various languages. 

Then a different view of languages, a descriptive approach, was established as exemplified in Saussure’s contribution which highlighted the distinction between a historical view of language development (diachronic) and a description of language at a particular point of time (synchronic).

There were important developments in the structural study of language. Weinreinch contributed significantly to the understanding of the role of semantics in language structure. The semantic approach became so popular that it was used to explain almost everything in language.

Then there appeared a great deal of suspicionon the role of semantics. Scholars in Eastern Europe and North American developed different views in the structural study of language. For example, the Prague School focused on phonology and discourse, while Bloomfield emphasized on item and arrangement which was later adapted to be known as Tagmemics – a system focuses on spots, filters and hierarchical structures.

By elaborating networks of relations within and between strata, a stratificational approach to language was prompted by Sydney Lamb and Adam Makkai. Then there was the famous transformational-generative grandma (T-G grammar) raised by Noam Chomsky who argued to look at language from the dynamic perspective – a series of transformations from an abstract underlying base to the surface structure. In this theory, syntax is focal.

The T-G grammar evoked critiques and also inspired developments in language study. A type of generative semantics was developed to start from an underlying semantics level rather than an abstract and underlying syntax. Also, T-G grammar has indirectly prompted the development of sociolinguistics – a theory with emphasis on language as used in a society.

During the 20th century a functional approach was rather popular in anatomizing languages. In the early stage the focus was on how languages work and then it was well developed by Halliday into a theory of systemic grammar which focuses on a dynamic system and can treat numerous phases of language.

Sperber and Wilson developed a new way to explore language which emphasizes the role of relevance in language design and practice. It was rather prevailing at then time to seek single principles and structures to explore language. For example, Chomsky employed “autonomous syntax”, while Halliday and Hasan applied cohesion as a unifying principle.

Semiotics has been the universal approach to languages, with Peirce and Wittenstein as the star contributors. Language has been recognized as a system of signs. Therefore, it is widely believed that semiotics could provide insights to understand how the linguistic code works.

Although no one theory is adequate enough to embody the complexity of verbal communication, all these different theories of language provide important insights into the nature of language.

Functions of language: Sociological

Generally, when people use language to relate to or influence others, they are applying the sociological functions of language. There are basically five types of such functions: interpersonal, informative, imperative, performative, and emotive.

As for the interpersonal function of language, it is about how people use language to negotiate and/or maintain social status. In most languages, there are distinct registers, such as ritual, formal, informal, casual and intimate speech. People make use of language in different registers to help establish themselves in society and maintain relations with one another. For example, people who want to increase their power often try to imitate the speech of those in power. A speaker or writer who wants to develop greater solidarity with the audience usually attempts to use the same register that the target audience uses and appreciates. Besides, it frequently happens that language is used not to say something relevant but just to maintain a relationship, which can be well explained by the cleverly social chattering at cocktail parties.

The use of speech or writing to influence the cognitive content or state of others is the informative function of language. This function is always part of other functions. Besides, to maximize the formative function of language it should be built on the existing cognitive state of the receptor.

The imperative function of language is about the attempts to influence the behavior of receptors. This can be done through commands, exhortations and even more effectively by means of smart jokes, suitable illustrations, searching questions, etc. Its application, with lots of a verbal negotiating, implies a measure of authority or power, although its effectiveness involves the receptor’s self-interest.

The performative function of language involves a change in the status of the receptor. The speeches themselves are expected to establish a different status of an object, such as the speech used in solemnizing a marriage. Mostly, performative language is highly ritualized and fixed in form which ensures the power and prestige of such speech.

The emotive function of language involves affecting the emotive state of receptors. For this reason, it profoundly exploits the associative or connotative meanings of words. This function may be understood as the most powerful and fearful aspect of language – because the emotions that can be influenced is unlimited. Language can kindle numerous emotions, from religious devotions to hilarious laughters. This function is the object for various groups, from professional linguists to powerful politicians.

In fact, most speeches and writings involve several different functions of language and in different proportions.

Functions of language: psychological

Language has diverse functions which are of two basic types: psychological and sociological. This short essay represents the psychological functions of language.

There are five essential psychological functions of language: naming, stating, modeling of reality, expression and cognition.

Finding names to experience and items is a psychological necessity. However, this need is so obvious that people mostly don’t realize its importance. It is essentially significant to have symbols/names to identify and even control things. In fact, finding the right word to symbolize some object or event seems to give some control over such things.

Stating is to say something about the object and event that are named. This is why there are subjective-predicate statements and topic-comment statements. Besides, since single propositions are too limited to meet certain psychological needs, it is natural to link together strings of sentences.

Modeling of the reality is the more profound function developed by people. In some sense, people instinctively felt words should provide a system for viewing the world – verbal symbols could reflect a reality, although imperfectly. From a different perspective, language is used to model reality in four major semantic classes of lexemes: 1) entities, e.g. daisy, water, tree, 2) activities, e.g. cry, laugh, run, 3) characteristics, primarily qualities and quantities, e.g. good, numerous, quickly, and 4) relations, e.g. in, because, during.

Another psychological function of language is expression – give vent to a person’s own feelings. It could be emotive responses to some event, e.g. ouch, damn it, and oh boy. It can also be a matter of playing with words which small children often love to do. Expressive language might be for the purpose of aesthetics. Words are arranged to form balance, proportion, symmetry and rhythm to formulate a certain psychological atmosphere.

Cognition, using language to think, is possibly the most important psychological function of language. Whatever style the thinking is performed, language is involved. This is a process of manipulating verbal symbols within brain.

Paradoxes of translating

Translating is an intricate and enchanting task.  There are many paradoxes in this particularly complex human practice. Below is a brief note from Language, Culture and Translating, the masterpiece of Eugene Nida, the theoretician of translation, father of dynamic equivalence theory.

It is popularly assumed that close and literal translation equals faithfulness to the source text. But the fact is that literal renderings are often misleading. Because there are many discrepancies between meanings and structures of different languages. Interestingly, while some people, who are concerned with figurative language and complex poetic structures, insist that translating is impossible, more and more translations are done and done well.

There is some contention about the validity of paraphrase (or adaptation if you well) when it comes to translating. Some argue that translating is valid but paraphrase is wrong. In fact, all translating involves differing degrees of paraphrase. It is out of the question that one can successfully translate word for word or structure for structure. The truth is that languages do not differ essentially in what they can say, but in how they say it. Therefore, paraphrase is inevitable. The key point is to ensure the semantic legitimacy of the paraphrase.

Another paradox holds that stylistic editing should be proceeded by a somewhat literal rendering – first produce a literal translating of the source and then improve it stylistically. However, style is the hard core which must be built into the translation from the very beginning. It can be put in this way: A few errors in the correspondences of lexical meanings are more excusable than missing the spirit and aesthetic quality of the source.

Then there is a paradox about translators themselves. It is true that translating is a skill which can be taught and then mastered by considerable practice. However, exceptional translators are born rather than made. They need to have outstanding aptitudefor the creative use of language. Well, this can be categorized into “Nature VS. Nurture”.

As the descending and thriving of computers and Internet, some people find it paradoxical for the existence of human translators. Modern technologies help human a lot, but when it comes to creative contents, such as advertising brochures and lyric poetry, computer printouts are basically useless. Human translators will always be indispensable as long as the text is stylistically appealing and semantically complex – which carry the essential message that is worth communicating in the target language. Human brain is not only digital and analogic but also has an established system of values which gives it a componentially incalculable advantage over machines.

Another paradox of translating is that there is never a completely perfect translation. Both the language and culture are in the process of change. Furthermore, language is an open system with overlapping meanings and fuzzy boundaries. Mostly, the biggest problem in translating is not to find the right equivalent in the target language but to thoroughly understand the designative and associative meaning of the source text.

A further paradox of translating is the general assumption that a bilingual person equals a translator or interpreter. The truth is that knowing two languages is far from enough. It is essential to be familiar with the respective cultures. Besides, the capacity to write well is another important quality to become a good translator. As for becoming a competent interpreter, it is imperative to have a quick mind to organize andformulate a response.

There is also a paradox of the language itself. Language not only represents reality but also distorts it. Sunsetand sunrise are the perfect examples of the parallax between language and reality – we all know the sun doesn’t set or rise. Unfortunately, people often don’t recognize such parallax in language and even are accustomed to accept verbal formulation as being the absolute truth. 

In general, the paradoxes of translating are basically the paradoxes of language and of culture.

初试微软机器翻译:惊艳与局限

这几年机器翻译比较热,尤其是神经机器翻译的问世,甚至又出现了“机器替代人”的标题党。由于我从事的翻译领域主要是创意性文案,很少涉及机器翻译,加上多少有些不以为然,毕竟人类的语言太复杂了,因此没有怎么实际体验。前一段时间作为“娱乐”领略了一下微软机器翻译,有被惊艳到,尽管也看到了它的局限性。

我体验的是Translator App上的语音翻译功能以及网页版Bing Microsoft Translator上的文本翻译功能。

对于一般文本,比如下面截图所示,翻译结果还算不错,而且在下图的例子中,它很惊艳地把第二句话隐藏的主语正确地补上了。

对于一般的语句,App的语音翻译也不错,能比较正确地识别语音,断句也还可以。

 

但它的局限性也非常多。简单地说,在语音翻译中,如果连续讲话,语音识别的断句会出现问题,一些词语会识别失误,这直接影响了翻译质量。在文本翻译中,有几类文本是它处理不好的:文化类、语境类、专业文案。

文化类是指包含鲜明地方文化特色或文化寓意的语句,比如:

语境类是指一些词语的意思要根据语境来选择,而机翻选择了错误的意思,或者说选了最字面的意思,但并不符合当时的语境。就比如下面的:

除了选错多义词的语境含义,还有可能出现对整段话的理解太过字面,没有把真正的意思表达出来。

专业文案是指比如法律合同,它会出现语意理解上的偏差甚至做出意思截然相反的翻译。

但有趣的是,正如我们在翻译的时候会经常面对不完美原文文本,机翻面对的也往往是不完美文本。比如在语音翻译方面,很大的概率是人讲出的话本身就是逻辑不够,有重复,不连贯,在不该停顿的地方停顿。由于所录入的文本质量不高,机器无法识别真正的信息,自然输出的译文就一塌糊涂。当然,这并不是像有些观点所说的,要倒逼人讲便于机器识别和翻译的语句;相反,人在讲话时出现的诸多“不完美”正是需要人工译员来处理,甚至是一种机器很难企及“智能”。在文本方面,一些机器很难处理的文化创意文案,实际上对人工翻译也是非常大的挑战;有的即使是“完美原文”,比如一些法律合同中的权责归属关系,但文本含义太绕了,机器很难处理妥当。其实所有这些所谓的局限性既是人工翻译的价值和机遇所在,也是我们对机器翻译的奇幻寄托。

不管怎样,现在的机翻水平是前些年所无法想象的,这是非常激动人心的进步,对翻译从业者来说是福音而绝非威胁。这些日新月异的科技进步,把我们从重复性作业中很好地解放了出来,让我们反思作为人的真正价值所在,也督促我们去实现更加原创的价值。在一次次迭代中,突破局限,领略惊艳。

筷子文化:杜嘉班纳“起筷吃饭”的水土不服

这些天意大利奢侈时装品牌因为“辱华”上了热搜,甚至遭到抵制。其实,导火索“起筷吃饭”系列短片本身只是一个市场营销失误,没有做好文化管理。如果没有杜嘉班纳创始人后续的激烈言论,或者说其社交媒体账号没有被黑,也不会演变地成现在这么严重,当然这是他话,这里主要说说筷子文化这个导火索。

杜嘉班纳原计划于本月20号在上海发布其品牌的时装大秀The Great Show,为了预热,发布了“致敬中国文化”的美食系列短片《起筷吃饭》。如果要列举中国和意大利的共同之处,美食无疑是最显著最安全的切入点。

这套美食系列短片每支两三分钟的样子,展示的是用加长版筷子吃夸张版意大利传统食物:披萨饼、意大利面和香炸甜卷。争议点之一是短片设计成模特一开始不会使用筷子吃意式美食,做了很多尝试,比如,吃披萨的时候一手一根筷子切披萨,把筷子竖插到披萨上,用筷子直接夹超大号甜卷。

我们小时候都有被家教不能把筷子竖插到饭里,更不能在饭桌上一手一根筷子拿着。筷子竖插到饭里,基本上等同于给死去的人上香,在饭桌上做这个动作无疑是非常晦气的。要是一手拿着一根筷子张牙舞爪,更是不可接受,一来这本身是危险的动作,容易戳到眼睛之类的,二来这是很显然的不得体的餐桌礼仪。如果食物太大,我们不会直接去夹,我们会用到勺子,或者提前切小,实在大块头的是可以直接用手拿的。

细想,我们对筷子是充满敬意的。我记得小时候每次帮忙摆饭桌,都是要先给爷爷奶奶拿筷子,最后才分给自己,筷子的粗头要朝外方便直接拿起来使用,潜移默化中也被教育在饭桌上不能一边吃饭一边拿着筷子指指点点,不能把筷子含嘴里说话,当然更不能把筷子竖插到任何食物里。相信这些对用筷子吃饭的我们来说,都不陌生。

有观点称杜嘉班纳这支广告是为了迎合千禧一代,他们独立个性,与众不同,打破传统。但我相信,就算是更年轻更独特的零零后,他们在餐桌上也不可以不尊重筷子文化。但是,如果是用筷子吃西餐呢?在餐桌上,筷子是万能的,用筷子的人是充满创意的,几米长的兰州拉面都能用筷子挑起来,还拿不下意大利面?所以基本可以说,这几支起筷吃饭的广告太尬演了,没有哪个用筷子吃饭的人会那样用筷子。

另外从文案本地化的角度来说,也是有不少问题。比如作为一支在中国播放的宣传片,不需要解释“小棍子样的餐具”,因为国人都知道那是什么,直接说“筷子”就好了。当然,也许这根本不是翻译的问题,在这种项目中,翻译往往处于被动的地位,能不能产出漂亮的本地广告文案,关键在于整个项目团队的运作。比如,在进行创意的初期,中国本土团队在文化方面的建议是不可缺少的;在文案本地化过程中,强调创造性翻译是必须的;后期成品的质量控制也不能缺席。

不管怎样,文化与经济的搭档关系,不是那么简单的工程,不小心就会水土不服。做足功课,虚心学习。

笔译也要保持口语水平

长期伏案码字,加上现实中口头交流的场合不多,难免会生疏了外语口语,甚至一开口会感觉怪怪的,腔调、流畅度、语感等等都找不准了。

虽然笔译主要从事笔头工作,但难免会遇到口头交流的场合和需求,如果口语拖后腿儿,会是很遗憾的事情。比如,新项目启动,客户需要开会交代交流一下,或者是接到一个新客户,需要进行面试。虽然这些场景也许不常见,但无法自如地用工作外语交流无疑是潜在的发展障碍。再比如最实际的场景,家里有熊孩子,第二代的外语教育难免忍住不要亲自上阵,至少参与其中。不管怎样,口语不能丢,口语水平要保持。

对于笔译而言,我认为要找回和提升口语水平,首先要明确对口语的本质需求是什么,然后找到自己的问题所在,继而寻找解决方案。

对口语的本质需求
作为笔译,我认为我们对口语的本质需求基本就是两点:表达观点、分享信息。我经常问自己,对于要表述的观点和信息,能用中文高质量地展示出来吗?——论点明确,论据充实,逻辑条理。现实情况是,有时候用中文都说不明白,所以很自然地用英语也表达不利索。换言之,不管是外语还是国语,语言是皮,内容和观点才是骨。先理清自己想说的、想分享的,然后再考虑如何去说。

问题所在
至于问题所在,就是自查口语表达的障碍。比如我个人而言,长时间不说英语,再讲遇到了以下问题:

– 流畅度不够,表现在容易卡壳儿,恩啊filler太多
– 语法容易乱,比如s/he、单复数、时态会混了
– 逻辑性偏差,容易前言不搭后语

解决方案
出现这些问题,当然“有历史的原因”——口语本来就不是强项,但解决方案也不是那么难找。

  1. 刻意训练——雅思口语

IELTS Speaking是我发现的非常棒的自我训练方式。雅思口语以话题为主导,考察日常语言交流能力,强调逻辑性、流畅度、语法、词汇。而且,有充足的训练资源,比如智能对练软件、各种雅思口语资源网站、专业的雅思口语培训书籍。我个人认为,用雅思口语来自我训练,能比较迅速地找回口语表达的感觉。比如以下话题,用一分钟准备,两分钟内讲完。

Describe a place you would like to travel to that is far away from where you live. You should say:
– Where is it
– How you would like to go there
– What you would do there
– And explain why you would like to go there

这样的话题也许有些刻板和“考试主义”,但通过参照考试标准进行自我训练,会短期内高效唤醒语力,比如减少恩啊filler,提高表达条理性,掌握日常实用词汇,增强语法意识,最基本的语言元素都系统地涵盖在内。

  1. 实际应用
    我有两种方式:

跟家里人讲。我家人不懂英语,但我们每天会预留出半小时左右的时间,让我用英语加汉语翻译跟他们交流——先说英语,然后翻译成汉语给他们听。一开始觉得好玩,但很快就发现这是一箭双雕的事情,既是有趣的家庭交流,对我来说又是不错的学习和训练机会。我们聊的都是些天南海北、街长里短的事情,很多时候我会捉襟见肘不知道怎么说,事后查阅这些难点,做成笔记,个人感觉效果非常不错。

复述。这个目前我还没来得及实施,但我的计划是,以话题和问题为导向,用录音的方式做记录,比如,看了一篇文章,用复述的方式条理地讲出其观点内容,然后回听自己的录音,发现问题所在,比如流畅度、逻辑性、语法、发音。

  1. 英语写作
    听说读写是相互促进的关系,尤其是写作和口语。写作注重语法、词汇、连贯性和逻辑性训练,通过写作提升这些语言能力,对口语表达会起到积极的促进作用。

总而言之,作为语言服务者,口头和笔下都要做足功夫。

作为评估者的审稿

这几个星期,连续做了几批Assessor & Test Reviewer的工作,有的是帮助英国翻译协会评估新的认证译员,有的是帮助翻译公司筛选新译者或者帮助赢取试译的新项目。除了一两份非常精彩,一看就是老译骨手笔,其他大部分都是领域新手的样态。

如果只修改还不算太痛苦,更痛苦的是作为Assessor的身份。

为了所谓的客观,要对每一处修改进行错误类型归类,并标记每处问题的严重程度(比如Minor, Major, Critical),然后逐一列出原文、译文、修改、修改原因,最后再就整体质量写一段小结。一开始比较怀念吃掉的语言学和语法课,写评估内容也不太适应。不过好在每种错误类型都有明确定义和规范,温故知新起来也比较有的放矢,另外也可以对号入座。但有意思的是,有时正是因为规范地太明确了,反倒入不了座。比如原译文就是啰嗦,可没有哪个错误类型是啰嗦(排除其他硬伤,比如术语,可以归入Style或者Rewording的范畴)。

另外在帮助赢取试译的新客户项目时,有一点很有意思。这个新项目的几百字试译文本让五个人来做,每个人的措辞造句风格必然有所不同。但除非有原则性硬伤,否则不要修改原作业,这样是为了能呈现多元化文案的提供能力。这其实是非常微妙的要求。毕竟,译稿不是出自名家,本身是不完美的,甚至问题斑斑,那如何界定原译文的措词造句风格是否值得保留呢?

依照流程客观评估自然是非常重要的,但作为评估者的审稿所作出的主观判断对项目也非常重要的。比如,原始作业的措辞造句风格到底要不要保留,保留多少,这就不是可以通过量化实现的,而是需要作为评估者的审稿进行主观判断。

另外,评估者不是审判者,没有一丝一毫的居高临下,每个改动都要有成熟的理由去解释和支撑。这也反过来督促我们去琢磨该如何把项目做好,把翻译做好。

最后,就是作业效率,也可以说收费问题。不管是按什么标准来结账,最后都落实在时间上。一开始没有经验,一份预期2.5小时的评估实际用了5小时多。这里面自然有不熟悉流程、首次作业生疏等因素,但后来分析,即使熟悉了流程和操作,2.5小时也很难够用。正所谓,不仅要提高工作效率,更要熟悉自己的工作效率。

翻译不完美原文

这两天在做一个培训类翻译,原文有的地方太不漂亮,这引得我思考该如何看待和处理不完美原文。

不完美之殇

原文只要是由人来书写或讲述/录入,最终呈现的文本总是容易不完美。比如会出现基本的拼写、语法和录入错误,也常会出现冗余的词句表述。人的主观能动性可以创造出名著,但更多的是普罗大众的平庸失误。

作为译者当事人,深深时时地体会到这样的原文事实所带来的干扰、困扰:该怎么翻译很有问题的不完美原文。

这就有质疑了:译者有资格判定原文不完美吗?毕竟,译者本身也存在各种不完美因素,不能用假设的完美译者去定论原文不完美。但如果就是任性地假设译者充分发挥主观性,接近了完美,那么他们有否可以判论原文,并修正所谓的不完美,产出一个接近完美的译文呢?

完美磁场

稍微冷静一下,就意识到这样的任性追问是没有实际意义的。要解决问题,应该首先明确问题是什么。不完美原文的问题无非有二:怎样分析不完美原文,以及怎样处理不完美原文。

原文的不完美。书写、语法和录入错误,是基本的不完美体现。事实信息错误(比如把2017/4/20记成周二),是另一种硬伤的不完美表现。无谓地消耗读者精力的冗余语句,或者采用扭曲的句式结构,也是一种不完美。至于再深入的,比如论点论据论证等方面的问题,虽是不完美,但已不是翻译的课题了。

这个不完美还太抽象单薄,因为还没有考虑作业类型和配套资料等因素。比如一个文本本身没有问题的操作指南,但由于缺乏充分的背景资料支持,导致原文的含义隐晦难以捕捉,很难说这样的项目文本是完美的。所以项目的成熟度也是影响原文文本的重要因素。有趣的是,原文语句冗余、句式结构扭曲这一点,在翻译课题内也很微妙。如果原文繁冗,译文可以简练吗?

我想,在处理这些不完美原文问题的时候,既不能凭空理论,也不能追求一劳永逸的方案,而是要根据项目特点和文本内容做具体问题具体分析。更有效的方式是明确不同题材、不同类型的完美文本应该是什么样子,然后与项目实际原文文本进行对比分析,并列出原文实际存在的问题,有理有据,再就处理方案达成一致并付诸实施。

相对的,成功文本至少要覆盖两个方面。一个是基本的语言信息水平,比如拼写、语法、事实信息,这不言而喻。另一个是项目类型,从另一个角度讲就是文体文风,这有不少文章可做。

在处理不完美文本之前,先要明确文本所属的类型,然后再分析该类型文本的完美标准是什么。比如确定了一个文本属于企业内部培训题材,那就要明确完美的内训文本要采用什么样的文体和文风。除了基本的题材写作通用原则,也要兼顾周边因素,比如文本的投放平台和投放方式等,而如果是第三方合作,最重要的参照标准莫过于客户的既定风格指南。完成这样的前期工作,就来观察原文,如果确定有明显的不完美,那就留出足够的时间和精力进行对比分析。

脱离具体项目来讨论完美是不可行的。要成功应对不完美原文问题,需要的是一个完美磁场,它能分解嫌疑原文,包容完美原文,吸纳全部项目因素,兼顾预期译文质量。

愿这个磁场你我共拥有。

五笔进度

经过这些天的练习,五笔最快码字速度27字/分钟,而且这个速度只是对照语篇进行码字,没有考虑实际作业中可能记不起字形的情况,要投入实际工作还有很大的差距,至少得提高到90字/分钟左右才行。

这两周以来,最集中的训练阶段是周末,工作日平均抽半小时训练。不过前几天”浪费”在理论学习上,强记成字字根、键名字根以及一二三级简码,到了后半段时间才发现这些字根理论/技巧先有个印象就可以了,关键是直接去打字,在实际打字的时候再去领会那些技巧/理论。

未来一星期的目标是篇目训练速度达到50字/分钟。努力实现这个脑袋拍出来的数字 :)