Michael Wu, 20, a student at Peking University, grew up in Shanghai. But when he wants to talk to his cousins in Hainan, he needs to bring his mother along to interpret the conversation.

今年20岁的Michael Wu是北京大学的一名学生，他从小在上海长大。当他想和海南的那些亲戚们交流时，他得拉上自己的妈妈来当翻译去解释谈话内容。

The cousins in Hainan speak two kinds of Hainan dialect. 'I actually cannot understand either of them,' Wu says. 'It's actually not much good for me to [try to] communicate with them.'

海南的亲戚们说着两种不同的当地方言。Michael Wu说：“他们说的那两种话，我真的都听不懂。让我（试着）跟他们交流，我实在是做不了。”

In China, that's a common problem: The differences in dialects are so vast they amount to different languages--possibly more than 3,000 variations, according to some estimates. It's one of the reasons that standard Beijing Mandarin has become the lingua franca of schools, businesses and government in China. But that uniformity comes at a cost: the rapid loss of many of these dialects.

在中国，这是一个普遍的问题：地方方言差别如此之大，它们实际上是完全不同的语言──据统计，全中国可能有超过3,000种的语言变体。这也是标准普通话变成中国的学校、商界和政府通用语言的原因之一。但这种融合统一是有代价的：其中很多方言都快速消失了。

Now two Americans have taken on a daunting task: trying to get an audio record of all of the thousands of China's languages and dialects before they disappear.

而现在，有两个美国人担起了这项艰巨的重任：赶在这数千种中国语言和方言消失之前，试着将它们全都录下来。

Linguists Steve Hansen and Kellen Parker are enlisting volunteers to canvass the country to capture both the languages and the stories of all of China's 2,862 counties and 34 provincial areas. Phonemica, founded last year, now has about 200 Chinese and Chinese-speaking foreign volunteers lined up to record their friends, parents and grandparents, telling a story in fangyan (regional speech).

语言学家史蒂夫•汉森(Steve Hansen)和凯伦•帕克(Kellen Parker)正在招募志愿者，希望他们能深入到中国各地，记录下全国2,862个县、34个省区的语言和故事。去年建成的乡音苑(Phonemica)网站现在拥有约200名中国志愿者和会说中文的外国志愿者，他们都在等着录下自己的朋友、父母还有祖父母用方言讲的故事。

'The idea is that we want to record it all,' says Mr. Hansen. 'And the only way to do this is through a crowd-sourced approach. We're trying to get people involved who will go to their hometowns and record friends and relatives.'

汉森说：“我们的想法是把所有的内容都录下来。能将这一想法付诸实践的唯一办法就是通过众包的方式来完成。我们想方设法让那些会回到故乡记录亲友说话的人都参与进来。”

'It is absolutely unique,' said Victor Mair, a professor of Chinese language and literature at the University of Pennsylvania, by email. 'No one else is attempting to do this for Sinitic' (the languages of China).

宾夕法尼亚大学(University of Pennsylvania)中国语言文学教授维克托•梅尔(Victor Mair)在邮件中说：“这种方式真的相当独特。目前还没有别的什么人为了中国的语言而尝试去做这些。”

Phonemica is nearly out of time. Scholars say that a few generations from now, all of China will speak as a first language standard Mandarin, the Beijing dialect that is taught in schools and used by new migrants to cities as well as businesspeople in every province.

但乡音苑所剩的时间不多了。学者们说，从现在起往后过不了几代人，全中国都将把标准普通话当成第一语言来说。普通话是以北京方言为基础的，在学校里统一教授，也是迁往城市的新移民和各省商务人士的通用语言。

Richard VanNess Simmons, a Rutgers professor of Chinese, says that as China's economy has taken off over the past 20 years, 'Mandarin has become the language that gets you somewhere and the language that parents want their kids to learn.' Even parents who speak regional dialects prefer that their children speak Mandarin at home.

罗格斯大学(Rutgers University)的中文教授理查德•范内斯•西蒙斯(Richard VanNess Simmons)说，过去20年间，中国的经济已开始起飞，“普通话已经成为能让你有所成就的一种语言，也成为父母希望子女去学习掌握的一门语言”。即便是那些说地方方言的父母也更喜欢自己的孩子在家说普通话。

'It's happening so fast it's almost too fast to document,' he says.

他说：“这一切发生得如此之快，快得几乎无法将其记录下来。”

The Chinese government also has taken on the task of recording the country's dialects, but its Chinese Language Resource Audio Database ( ) is still in the 'fieldwork' stage, says Mr. Simmons, and 'no results have yet been published as far as I know.'

西蒙斯还称，中国政府也开始担负起记录各地方言的任务，但政府的中国语言资源有声数据库(Chinese Language Resource Audio Database)还处在“调研”阶段。西蒙斯说：“据我所知，迄今为止还没有什么结果公布。”

An earlier effort by the Chinese Academy of Science in the 1990s, he says, produced a small series of about 20 recordings of dialects on audiotape and in book form. The academy then put out a DVD version that required users to install an old browser for Windows, so 'it didn't end up being very useful in digital form.'

西蒙斯表示，此前中国科学院(Chinese Academy of Science)在90年代也做过一些工作，做出了一个包含了约20份方言记录的小系列，这些内容被人用录音带和书本的方式保存了下来。中科院后来又出版了这些资料的DVD版本，但需要用户安装一个老版本的Windows浏览器，所以“这些努力最终并未以特别实用的数字形式呈现出来”。

Virtually unlimited storage in the era of cloud computing means Phonemica can record as many dialects as it wants without worry about technology becoming obsolete. The bigger challenge has been in getting drumming up enough volunteers to make recordings.

在云计算时代，存储几无上限意味着乡音苑可以记录下它想记录的所有方言，而无需担心技术会过时。乡音苑面临的更大挑战是难以招揽到足够多的志愿者来制作方言记录。

The site's original model was the Speech Accent Archive created by George Mason University in Virginia, which records hundreds of English speakers reading the same English-language paragraph with different accents, from Brooklynese to Dutch. But Messrs. Hansen and Parker felt that there was also value in recording family stories, in the style of StoryCorps, an archive of 45,000 stories told by Americans to each other.

这个网站的原型是由弗吉尼亚州乔治梅森大学(George Mason University)创建的语言口音档案(Speech Accent Archive)，该档案记录了几百名说英语的人用不同的口音──从布鲁克林人到德裔宾夕法尼亚州人所说的各式口音──朗读同一段英语文章的情形。但汉森和帕克认为，以“故事团”(StoryCorps)的方式将家族故事记录下来也很有意义。“故事团”是一个档案库，包涵了美国民众口口相传的45,000个故事。

So far, the site has 50 stories on record, ranging from young people talking about their favorite Chinese pop stars to grandparents remembering being sent to work in the countryside during the Cultural Revolution. Mr. Hsieh, a 61-year-old from Pingtung County, Taiwan, tells a story in the Hakka Sixian dialect. There are more than 30 million native speakers of Hakka, Mr. Parker says, but the Sixian dialect group is much smaller, mostly in the northwest part of Taiwan. Song Hongcheng, 50, from Fangxian in Hubei province, tells a story in the Fangzian dialect, a subset of Central Plains Mandarin, which has about 300,000 speakers.

迄今为止，乡音苑已收录了50个故事，话题从年轻人谈论他们最爱的中国明星到祖父辈们回忆文化大革命时期下乡劳作的往昔。来自台湾屏东县61岁的谢先生就用四县客家话讲述了一个故事。帕克说，母语为客家话的人口超过三千万，但说四县方言的人就少得多了，他们大部分都生活在台湾西北部。来自湖北房县50岁的宋宏成用房县方言讲了一个故事，房县话是中原官话的一个分支，目前有30万人在说着这种方言。

So far the project has relied too heavily on young people--students interviewing fellow students--with not enough contribution from older generations, according to Mr. Simmons, who serves as an informal adviser to the project. He advised Phonemica to try to record people over the age of 30, and preferably over 60. Younger people, even if they speak in a dialect, tend to throw in standard Mandarin words more often than their grandparents would.

以非正式顾问的身份服务于乡音苑的西蒙斯说，到目前为止，这个项目还是太过依赖于年轻人──学生采访同学──没有获得年长一辈足够的支持和贡献。西蒙斯建议乡音苑尝试去收录那些年过30的人，最好是年过60的老人所说的话。毕竟，那些年轻些的人，就算他们说的是一门方言，也会时不时掺杂着一些标准普通话的词语，而他们爷爷奶奶讲的方言会更地道。

Leonardo Liu, 40, who has already recorded his parents, his wife and several colleagues, says the project made him wonder what it would have been like to hear the voice of Confucius. 'Maybe 1,000 or 2,000 years later, we'll be the ancient people. If the website can keep our voices and our accent, it will be great.'

现年40岁的Leonardo Liu已录下了自己的父母、妻子和几位同事讲述的故事。他说，这个项目让他不禁想象孔子的声音会是什么样子。“1,000年或2,000年以后，我们就全都变成了古人。如果这个网站能将我们的声音和口音一直保留下去的话，那就太棒了。”