The 3% diff pronunciations in Chinese Mandarin between Mainland and Taiwan

Given 100 questions, a Pinyin student answered all correctly, but a Zhuyin teacher might only give him a score 97% . Why ?

Mainland Common Tongue (Pǔtōnghuà 普通话) and Taiwan National Language (Guóyǔ 国语)

By “character” : 13~38% Diff

There were many essays related to the topic “Pronunciation differences in Chinese Mandarin between Mainland and Taiwan”. e.g. :

“Comparison of the Chinese pronunciations between Mainland China and Taiwan” (海峡两岸字音比较) by Qingmei Li (李青梅), 1992

 “A Study of the Distinction of Pronunciation Standards between Taiwan and Mainland China” (两岸语音规范标准之差异探悉) by Jihong Nan (南基弘), 2008

Most of them are based on “Diff by character” : The 3500 high frequency words specified in “List of Commonly Used Characters in Modern Chinese” (现代汉语通用字表, 1988).

Experts check the differences character by character. While a mismatched pronunciation was found, essays usually provide several example “words / vocabularies” with that character.

E.g.: 绩/绩 (means “achievement”) is pronounced differently. In Mainland it is tone 4  (jì ㄐㄧ\), while in Taiwan it is tone 1 (jī ㄐㄧ). As an example, the vocabulary “成绩/成绩” :

In this way, a final diff percentage was calculated, based on characters:

“13%~38%” is a large difference. It makes people feel like a student could only get a reduced score of 62%, with his 100% corret study from the other side? A little bit more than realistic.

By “vocabularies” : 3% Diff

In stead of only compare the 3500 characters, ToneOZ tried a new way : To compare all 170,000+ words / vocabularies in the dictionary. The Chinese characters are pronounced differently after they are combined into words / vocabularies, this should give us some data from a different point of view.

By “vocabularies”, there are 513 characters from around 6K words/vocabularies have been identified with different pronunciation between the Common Tongue and the National Language. It is based on a programmatically comparison and then a manual review for around 170K words/vocabularies from several word banks :

CEDICT (the largest open source Mandarin-English Dictionary)

National Language Mandarin official dictionaries “Applied Modern version Jian-Bian-Ben” (国语辞典简编本) and “Literature version Zhong-Bian-Ben” (国语辞典重编本) by the Taiwan Ministry of Education)

“Jeiba” Chinese text segmentation database.

The diff rate is 3.88% : In all 170,580 words, there are 6,623 words are pronounced differently.

We also use word banks to filter percentages from “high frequency words” only. In Australia, there are 2 common Chinese certifications:

Chinese Proficiency Test, Hanyu Shuiping Kaoshi
(汉语水平考试 , HSK)

Test of Chinese as a Foreign Language
(华语文能力测验, TOCFL)

They are both designed for non native Chinese speakers to learn general conversations. As results:

The results filtered by HSK and TOCFL are all close to 3%, which gives us more confidence in this number. And if we filter from basic levels only, the percentage becomes around 1%(For basic conversation ability, HSK level 3 or TOCFL level 4 is required).

In conclusion : There are pronunciation differences in Chinese Mandarin between Mainland and Taiwan, but not too much, only about 3% words are pronounced differently. In general conversation the difference rate could become 1% only. You don’t have to worry too much about these differences if you are learning Chinese.

Toneoz.com/data provides a list for all the Chinese words with pronunciation differences, and it can also check pronunciation differences from a whole Chinese sentence.

Read More

  • 教育注音字典「晓声通」- 为大屏、平板教学设计的「广度拓展」国语辞典

    教育注音字典「晓声通」- 为大屏、平板教学设计的「广度拓展」国语辞典

    晓声通字典源自教育部《一字多音审定表》标准辞典,使用中小学考试标准字音及字词解释。网址:https://toneoz.com/dmap 介绍一些容易上手又好玩的国语字典功能,如「扫描成语造词」「同部首字探索」「课本生词连动」…透过特别设计的使用介面,让老师们在大屏跟平板教学时有效果又有效率。 Web4.0智慧物联网时代的字典 Web 4.0智慧物联网时代,如果只是为了查一个字或查一个词的解释,大部分人会找搜寻引擎或直接问手机的语音助理。字不会写可以直接语音输入,字不会唸可以手写输入,万一不会唸也不会写,还可以直接拍照影像搜寻。 然而一本好的字典,在教育场合有个灵魂功能:「拓展词汇」。 词汇的「广度拓展」:无法取代的查字典乐趣 举例来说: 语言素养的形成,除了「直向」查询字义的深度探索,也需要字词间「横向」连结的广度拓展。透过查字典的过程来发现更多字词左右邻居之间的关联,再连结到生活中的经验,这样的乐趣能提供动机,激发主动学习的兴趣。传统纸本字典在翻找的时候,读者自然而然会接触到许多前一页后一页上额外的字汇,这些都是查询生字时的无形收获。 晓声通字典:教育部标准资料、瞬间呈现横向关联、支援多音注音字型 …
  • 国语课本生词111学年度

    国语课本生词111学年度

    新学期国语生词,「播报」的「播」是一声还是四声? 教育部官网刚刚放上了新学期的国语生词资料。可以看到一个有趣的变动:2021年底的新闻中被名作家批评为「弱智浅语」的翰林二年级11课「门斗大仙」,被换成了新课文「奇怪的门」。 新闻播报的「播」字是新课文生词,提醒关心时事的同学们要注意新闻用语喔! 晓声通是线上中文字典,支援「点任意词快速连查」「教育标准注音字型」。 资料源自教育部「简编本/重编本/成语典」及「国语课本」。标准楷体注音,可免安装线上使用,也可 下载字型安装,注音字可复制贴入Office。 字典网址 https://toneoz.com/imez/?lang=zh-TW
  • 修正缺字型的康熙字集

    修正缺字型的康熙字集

    常用字部首字如:「工、长、门、行…」字型不协调,换任何字型都无效? 别处取得的文字档,贴到 Office 中要换字型,这时发现有一些常用字部首字如:「工、长、门、行…」字型不协调,换任何字型都无效?这种情况,有可能是原始的文字档中使用了特殊的「康熙部首 Kangxi Radicals」字元集,(参见维基百科 https://zh.wikipedia.org/wiki/Unicode部首列表 ),如下图: 为何别人给的文字档会有这种情况?可能是原始作者使用的某些版本Office、PDF转换程式、或者是输入法有字元相关的问题。 以前遇到这种情形,只能认命一个字一个字改。本文章介绍另一个快速解法,自动将所有康熙部首字元转换成常用Unicode区段的字元,换字型再也不缺字失调。 …
  • 找出生词标注音,依照小学国语年级课程。输入任意文章10秒完成!

    找出生词标注音,依照小学国语年级课程。输入任意文章10秒完成!

    课堂上已经教过的字,希望隐藏注音,让学生自己练习唸,只有新的生词才标上注音,提醒学生正确的唸法。随着学习进度不同,哪些才是「新的生词」?同一篇阅读测验在不同年级的教学中,需要标注音的生词也不同。ToneOZ拼音注音编辑器有办法帮助老师,10秒内完成找生词标注音的任务! 输入一篇文章,指定出版商(康轩南一翰林均可)、年级、以及上到哪一课?自动帮你找出还没教过的生词,标上注音,适合需要自己制作教材的老师或家长。 使用方法(注音编辑器网址:ToneOZ.com/imez) 键入或贴上文章的瞬间,ToneOZ注音编辑器会自动标上注音,并做破音校正。 找出生词标上注音:点选上方工具列「课本分级选字」选单,直接勾选出版社、年级、课文,可一次全选整个年级,或者是依照上课进度一课一课勾选。 选好字后,利用上方工具列「注音」调整。例如先选出一二年级的字,再关闭注音,则一二年级教过字的注音都会隐藏。 也可以反过来查询出版商年级课程。点选「寻找」后输入生词,会瞬间列出哪些课文有包含这个词。例如查询「鲸鱼」这个词,可以发现 “翰林” 3年级下学期的【飞行员和小王子】及 “康轩” …
  • 文章的「字」自动分组成「词」表格,贴入Excel, WordWall, 或Quizlet

    文章的「字」自动分组成「词」表格,贴入Excel, WordWall, 或Quizlet

    中文的「词」是由多个「字」组合而成,许多情境都会需要将一整段文章「分词」来制作教材。 使用 拼音注音编辑器 toneoz.com/ime 自动分词 ToneOZ的分词功能是使用内建的 17万词辞典,词库涵盖两岸教学字典、检定考HSK(汉语水平考试)、及TOCFL(华语文能力测验),适合教材制作,全自动处理省时又简单。 以下范例示范如快速将整篇中文文章分词,汇出到Microsfot或Google试算表、Quizlet或WordWall互动教材工具。范例汇出时用到的软体均为举例,实际使用时老师们可自行灵活更换: 整篇文章自动分词,做成中文 、拼音 、注音三栏Excel表格 …

Leave a Reply

Your email address will not be published. Required fields are marked *