On transcribing sound track in Cantonese
By Sam Wong
0. Format of transcription
A sample transcription is shown as follows:-
|
聽講 |
上頭 |
而家 |
派 |
咗 |
|
teng1gong2 |
soeng6tau4 |
ji4gaa1 |
paai3 |
zo2 |
|
hear.say |
superior |
now |
dispatch |
pfv |
|
嗰個 |
姓 |
陳 |
嘅 |
特派貟。 |
|
|
go2go3 |
sing3 |
can4 |
ge3 |
dak6paai3jyun4 |
|
|
that.one |
surname |
person.name |
attr |
special.investigator |
|
|
‘It is said
that the superior has currently dispatched the special investigator, Chan.’ |
|||||
There are four lines:-
Line 1: Chinese character and punctuation
Line 2: romanization
Line 3: gloss
Line 4: translation
To efficiently maintain
a tidy text alignment, a table without gridlines is recommended to store the
contents of the first three lines.
Tabs and spaces, however, also serve the purpose but work with less
efficiency. It should be noted that
all words are left aligned within the same column. If there is insufficient space to put a
long sentence on a single line, the split should be made between phrases that
are more loosely related. For
example, in the above example, a split made between zo2 and go2go3
is better than one made between paai3 and zo2 since the relation
between the latter pair is more closely related than the former pair.
In addition to the four
lines of transcription, background information of the sound track like the
date, venue, speaker(s) and short description of the event should be clearly
stated at the beginning of the file.
The purpose of each line is described in detail in
the subsequent sections:
1. Chinese character and punctuation
Line one is a rendering
of the utterances in the sound track in Chinese characters and
punctuations. In this assignment,
fine transcription is required, so please jot down exactly what was uttered by
the speaker, including the sentence final particles, like gaa3 㗎, laa3 嘑, and bo3 噃, as well as the exclamatives, like waa3
嘩, and ai1jaa3
哎吔, in addition
to the content words.
1.1.Words and phrases
In common practice, the text
in Chinese characters is always written without natural delimiters. On the other hand, it should be noted
that in the course of transcription, in general, word, instead of character, is treated as a unit while a space is
used between words as a delimiter.
In linguistics, word is usually defined as the biggest
element that may be uttered in isolation with semantic or pragmatic
content. For instance, for the word bo1lei1 玻璃 ‘glass’, the component characters, bo1 玻 and lei1_璃, are
meaningless by itself in modern Cantonese, bo1lei1
is thus treated as a word and is written without space in between.
It should be noted that
sometimes there is ambiguity in that a linguistic expression can be interpreted
as either a phrase or a word. For
example, the term sik6 faan6 食飯 can be interpreted as both ‘to eat
rice’ or ‘to have meal’. In the
former case, the meaning of the expression is closely related to its
components, the verb sik6 ‘to eat’ and the object faan6
‘rice’. In this case, sik6 faan6
is considered as a verb phrase and a space is used between the two components. In the latter case, when sik6faan6
is used with the meaning ‘to have a meal,’ in the sense that we not only ‘eat rice’, but also ‘drink soup and eat noodle, dessert, et cetera,’ sik6faan6
is considered as a word and no delimiter is required.
It
should also be noted that this kind of ambiguity is always found in natural
language. There is no rigid rules
on how to resolve such ambiguities so one will need to decide on whether a
linguistic expression is a phrase or a word using his/her linguistic sense based
on his/her understanding of the context as a native speaker. Most of the rules in the following
guideline, which is designed for Mandarin Chinese, also apply to Cantonese and
can serve as a general reference:
http://www.pinyin.info/rules/pinyinrules.html
For details, one can also refer to Yin and Felly (1990). Like any rules, the above rules,
however, should not be followed blindly. The reader can adjust according to
his/her linguistic sense as a native speaker.
1.2.Written form
Mandarin cognates exist
for most of the Cantonese morphemes so in most cases, it is not hard to locate
suitable characters for transcription purpose. Sometimes the etymology of a Cantonese
morpheme is not clear and we are thus not aware of its Mandarin cognate. In the event of this, one can follow the
common practice in daily life experience like those found in newspapers,
magazines, blogs, and other internet resources. For instance, bin1dou6 ‘where’ is
usually written by using the homophones 邊度.
Upon necessary, non-Chinese characters can also be used in case an
expression is commonly written in that way, especially the loanwords. Some
examples are shown in Table
1.
|
Romanization |
Common written form |
English equivalent |
|
cok3joeng2 |
chok樣 |
the look when you act cool |
|
ou1kei1 |
OK |
okay |
|
kaa6waai1ji4 |
可愛い |
cute; lovely |
|
kawaii |
Table 1 Cantonese expressions commonly written with non-Chinese characters
For expression with more
than one common written forms, like kaa6waai1ji4 is commonly written as
either ‘可愛い’ or ‘kawaii’, one can select according to his/her own habit but consistency
should be maintained through the whole work of transcription.
Sometimes cognate in
Mandarin does not exist for a Cantonese morpheme but cognate attested in
classical Chinese text does, like 擢樣 for cok3joeng2. In this case, character with
etymological relation with classical Chinese, which is also known as the
‘correct character’, exists although many of them are hard characters. You are encouraged to check these
characters from dictionaries but this should be done only when time is
permitted and in any case should not be the focus of the transcription
work. Some references are listed in
the reference section.
It is always easy to
locate a suitable written form for lexical word but for function word, it is
not the case. The utterance particles
are the hardest among all since one of the greatest contrast between Cantonese
and Mandarin Chinese is the use of this category. There are only 27 utterance particles in
Modern Standard Chinese (Chao 1982: 394−403) but as many as 95 in modern
Cantonese (Leung 2005). For this
reason, Mandarin cognates often not exist for most of these particles. Suitable
characters are thus hard to found to represent these particles. Appendix 1 can be
served as a general reference but the readers are strongly suggested to follow
his/her own habit to avoid inconsistency.
Some operating systems
may lack the necessary font or input method for inputting the special
characters, one can download these tools by following this link:
http://www.ogcio.gov.hk/tc/business/tech_promotion/ccli/download_area/
Last but not the least,
in the case that there is really no suitable written form in your mind at all,
or you even do not understand the meaning of the expression; you can just
directly put the romanization in place of.
2. Romanization
The second line is a
rendering of the utterances in the sound track in romanization. Like what we have mentioned in section 1, word is also treated as a unit when romanizing the
utterances. In other words, space
is not used between syllables within a word. Italic font-face is often used.
2.1.The LSHK system
In
this assignment, the Linguistic Society of Hong Kong Cantonese Romanization
Scheme, a.k.a. the Jyutping 粵拼 system, developed in 1993 is
adopted. To understand this scheme,
the readers are referred to LSHK (2002) and the following web-sites:
The Jyutping Scheme: http://www.lshk.org/node/47
Tutorials on Jyutping:
http://www.cantonese.asia/viewnews-229.html
http://www.iso10646hk.net/jp/learning/index.jsp
http://www.senseasy.net/leeyuiwah/CHS/Jyutping-tutorial.latest.ppt
In addition to Guide
to LSHK Cantonese Romanization of Chinese Characters (LSHK 2002), one can
also check the romanization of a Chinese character via the following databases:
http://humanum.arts.cuhk.edu.hk/Lexis/lexi-can/
(for BIG5 characters only)
http://www.iso10646hk.net/jp/database/index.jsp
For those who are
familiar with other schemes, they can refer to the charts comparing the schemes
in LSHK (2002: 17−20) and the following web-sites:
http://input.foruto.com/ccc/jyt/ap01b.htm
http://en.wikipedia.org/wiki/Hong_Kong_Government_Cantonese_Romanisation
For those who have never
undergone training in phonetics and phonology, nor have not learnt any
romanization scheme in the past, it may be difficult to get familiarized with
the scheme within a short time. The
following web-sites that convert string of characters to romanization are
useful resources:
Chinese Word
Parser: http://www.cantonese.sheik.co.uk/scripts/parse_chinese.php?action=parse
Jyutping Database: http://www.iso10646hk.net/jp/database/index.jsp
JyutPingEasy.Net: http://www.jyutpingeasy.net/scgi-bin/toJyutPing.cgi
HKTV Cantonese to Jyutping: http://hktv.cc/hp/cantonesetojyutping/
Although the above are helpful
resources, one should always bear in mind that since it is common for a Chinese
character to possess more than one pronunciation, the users are strongly
advised to carefully check the computer outputs before utilizing.
2.2.Actual or standard pronunciation?
Living language is always
a dynamic system. There are
sometimes variations in pronunciation from person to person within a community.
For instance, in modern Hong Kong
Cantonese, virtually /n/ is missed among the initials in the phonological
system while a number of young speakers have lost the /ng/ and /k/ coda in
their speech. These two
phonological developments are the so-called sloppy speech. Specifically, there also exists free
variation for the same word in identical context. The same person may even choose different
pronunciations in two consecutive utterances, e.g. hung4dau2bing1
~ hung4dau6bing1 ‘shaven ice with red bean.’
For this reason, the
actual rendering of the same word may be different within a passage of
transcription. To provide more
information about the variety of the speaker, actual rendering of the
pronunciation is preferred. However,
transcribing in the standard variety or the transcriber’s own idiolect is also
acceptable.
Most of the syllables in
modern standard Cantonese can be transcribed with the LSHK system. In case you
find some escaped from the net, the IPA system can be used for these syllables.
3. Gloss
To
help non-native speakers understand better the literal meaning and syntactic
properties like word order of the utterance, an English word-by-word rendering
of the utterances in the sound track is provided in line three. For translation with more than one
English words, a period is inserted between the words, like ‘have.meal’ for sik6faan6
and ‘special.investigator’ for dak6paai3jyun4.
For
content words, it is always easy to find an English equivalent of the Cantonese
words. For function words, however,
it is often hard to find equivalent because different syntactic system are
often found in different languages while polysemy is frequently observed for
function words.
For
the former problem, for example, in Cantonese, classifier must be used between
numeral and noun but in English, there is no such category while measure words
are optional. It is often not easy
to find an English equivalent of the classifier while the translation of the
classifier always does not help the reader understand the text better since the
classifier system only reflects how we classify the objects in the world, which
is somewhat similar to gender in the European languages.
For
the latter problem, for instance, the word you in English is both a second
person singular pronoun and a second person plural pronoun. If we put you as the gloss of
both nei5 ‘the second person singular pronoun’ and nei5dei6 ‘the
second person plural pronoun,’ those who have inferior knowledge of Cantonese
will have no way to tell the different meanings between the two.
For
these reasons, for function words in some special categories, special
abbreviations in small capitals are usually adopted in place of the English
equivalent. The option of small capital is available in Format
> Font (Figure
1). Table 2 shows some abbreviations commonly adopted among
linguists.
Figure 1 The option of small capital is available in Format > Font
Table 2 List of common abbreviations
For function
words falling out of the above categories, the English equivalent is used. For
proper nouns, translation is not necessary but italic ‘person.name’ and
‘geographic.name’ are used.
4. Translation
Line
four is a translation of the utterance in plain English. It should be noted that colloquial
English should be used to match the genre of the sound track. Upon necessary, if there is a huge
distance between the literal meaning and the translation; in other words, if
the words used in the translation are very different from those in the
word-by-word gloss, the literal translation should be attached before the
translation. The following shows an
example:
|
嗰 |
個 |
人 |
曉得 |
官話 |
唔 |
呢 |
？ |
|
go2 |
go3 |
jan4 |
hiu2dak1 |
gun1waa2 |
m4 |
ne1 |
|
|
that |
cl |
person |
know |
Mandarin |
neg |
q |
|
|
Lit. ‘Does
that person know Mandarin?’ ‘Does he
understand Mandarin?’ |
|||||||
For more fine details on glossing rules, you are referred to the following resources:
Leipzig Glossing Rules: http://www.eva.mpg.de/lingua/resources/glossing-rules.php
Interlinear morphemic glosses: http://www.ling.hawaii.edu/ldtc/website/syllabus/sp06/LehmannGlossing.pdf
Comments and suggestions are welcome! For other questions concerning
transcription, please direct to Mr Sam Wong Tak-sum at egwts@polyu.edu.hk
.
Reference
Chao, Yuen Ren. 1968. A
Grammar of Spoken Chinese. Berkeley: University of California.
Guan Jiecai. 1990. A Dictionary of Cantonese Colloquialisms in
English.
Hutton, Christopher M. and
Kingsley Bolton. 2005. A Dictionary of
Cantonese Slang: The Language of
Lehmann, Christian. 1983. Directions for interlinear morphemic translations. Folia Linguistica 16: 193−224.
Lo, Tam Fee-yin. 2007. Cantonese Colloquial Expressions. Hong
Kong: The
Matthews, Stephen and
Virginia Yip. 1994. Cantonese: A Comprehensive Grammar.
So, Siu-hing Simon. 2002. A Glossary of Common Cantonese Colloquial
Expressions. Hong Kong: The
Swofford, Mark. Basic rules
of Hanyu Pinyin orthography (Summary). Pīnyīn.info: a guide to the writing of
Mandarin Chinese in romanization. Modified 2010. Accessed 2nd
October, 2012. http://www.pinyin.info/readings/zyg/rules.html
Yin,
Binyong, and Mary Felly. 1990. Chinese Romanization: Pronunciation and
Orthography. Peking: Sinolingua.
Bai, Wanru 白宛如 1998：《廣州方言詞典》（《現代漢語方言大詞典．分卷》，李榮 主編），南京：江蘇教育出版社。
Chao, Yuen
Ren 趙元任 著，丁邦新 譯 1982：《中國話的文法》，香港：香港中文大學出版社。
Cheung,
Hung-nin Samuel 張洪年 1972：《香港粵語語法的研究》，香港：香港中文大學出版社。
Kong,
Zhongnan 孔仲南 1933：《廣東俗語攷》，南方扶輪社。1992年經上海文藝出版社重新影印出版。
Leung,
Chung-sum 梁仲森 2005：《當代香港粵語語助詞的研究》，香港：香港城市大學語言資訊科學研究中心。
LSHK 香港語言學學會粵語拼音字表編寫小組 2002：《粵語拼音字表》，第二版，香港：香港語言學學會。
Li, Xinkui 李新魁、黃家教、施其生、麥耘、陳定方 1995：《廣州方言研究》，廣州：廣東人民出版社。
Mai, Yun 麥耘、譚步雲 1997：《實用廣州話分類詞典》，廣州：廣東人民出版社。
Rao, Bingcai
饒秉才、歐陽覺亞、周無忌 1997：《廣州話詞典》，廣州：廣東人民出版社。
Wong, Shek
Ling 黃錫凌 1941：《粵音韻彙》，香港：中華書局。
Yu, Xuepu 虞學圃、溫岐石 1915：《新輯寫信必讀分韻撮要合璧》，原著於1782年成書，近年經香港陳湘記書局重印。
Zheng, Ding’ou 鄭定歐 1997：《香港粵語詞典》，南京：江蘇教育出版社。
Appendix 1:
A list of Chinese characters for transcribing Cantonese expressions
|
Chinese
Character |
Pronunciation |
|
喔 |
[ɔʔ22] |
|
嚡 |
[hai353] |
|
喀 |
[hɐʔ] |
|
嘞 |
[lɐʔ] |
|
吖 |
a1 / aa1 |
|
呀 |
a3 / aa3 |
|
啊 |
aa2 |
|
呀話？ |
aa6waa5 |
|
哎 |
ai1 |
|
呃 |
ak3 / aak3 (utterance
particle) |
|
罷 |
baa2 |
|
嚊 |
be3 |
|
誒 |
e6 |
|
欸 |
ei3 |
|
咖嘛 |
ga1ma3 |
|
㗎 |
ga3 / gaa3 |
|
㗎嘑 |
ga3la3 /
gaa3laa3 |
|
噉 |
gam2 |
|
咁 |
gam3 |
|
嘅 |
ge3 |
|
吓 |
haa2 |
|
下話？ |
haa6waa5 |
|
唏 |
hei1 |
|
愾 |
hei3 |
|
噷 |
hm |
|
嘑 |
la3 / laa3 |
|
喇 |
laa1 / la1 |
|
嗱 |
laa4 |
|
呢 |
le1 (utterance
particle), li1 (pronoun) |
|
唎 |
le3 |
|
咧 |
le4 |
|
褦 |
le5 |
|
囖 |
lo1 |
|
囉喎 |
lo3wo3 |
|
囉 |
lo4 |
|
咪 |
mai6 |
|
緡 |
man1 |
|
咪嘢 |
mi1je5 |
|
哦？ |
o2 |
|
哦 |
o4 |
|
卅 |
saa1aa6 |
|
聽日 |
ting1jat6 |
|
哇 |
waa1 |
|
喎 |
wo3 |
|
啝 |
wo4 |
|
𢰸 |
wo5 |
|
咋嘛 |
za1ma3 |
|
咋 |
zaa3 |
|
咋？ |
zaa4 |
|
唧 |
zek1 |
Last Updated:
10 March 2018
4:40 PM
Copyright © Department of English,