英语语音学-课件

Sep.
10, 2012
Introduction to Phonetics I
Fall 2012
Instructor: Karen Steffen Chung 史嘉琳
Associate Professor, Department of Foreign Languages and Literatures
National Taiwan University
Unless noted, the course materials are licensed under Creative Commons
Attribution-NonCommercial-ShareAlike 3.0 Taiwan (CC BY-NC-SA 3.0)
1. Class Overview
I. Outline for today:
1. Enrollment and introduction of TAs and to OCW (Open Course Ware)
2. Overview of the course
3. Introduction to some of the areas of linguistics and phonetics
4. Homework
II. Notes
1. Enrollment and introduction of TAs and to OCW
a. Who this class is for:
- Primarily: students enrolled in the education program and translation program students;
- Secondarily: interested DFLL (foreign languages and literatures);
- Finally: other students who have taken or are currently taking Introduction to Linguistics.
b. Introduction of TAs
- Sophie Ding-yi Weng 翁鼎禕 b98102104@ntu.edu.tw
- Mendy Yun-chen Chang 張勻甄 b96102067@ntu.edu.tw
- Ai-ling Chung 鍾艾玲 r99129001@ntu.edu.tw
c. Introduction to OCW
- Students should be prepared to be filmed in class (and therefore shouldn’t be late)
2. Overview of the course

a. Second semester will include acoustics and is recommended for those who are interested in
phonetics. (Former students have formed a phonetics family who speak the same “secret”
language!)
b. General American (GA) or Standard American English (SAE) will be our standard, but frequent
reference will be made to Standard Southern British (SSB) English, or "RP" ('received
pronunciation'; also called 'BBC English') and other dialects of English.
1
Sep. 10, 2012
* In which part of the United States do we find the greatest dialectal variation?
- New England and the East Coast in general, e.g. New Jersey, Boston, etc.
- The South, starting from Maryland in the East, to Florida in the South, to Texas in the West.
* What is “Standard American English”?

- People with a different accent in American English will often “code-switch” when talking to
people from other parts of the country or from other countries for more efficient communication.
The speech style they use will often be some variation of “General American”.
* Estuary English:
- Popular teenage English spoken in the southeast part of the United Kingdom. More and more
adults over 30 or even 40 years olds are using this dialect of British English.
E.g. three: Standard Southern British English interdental [θ]: [θri] vs. Estuary labiodental [f]: [fri];
SSB building ['bɪldɪŋ]; Estuary: ['biudɪŋ]
 For more details, please visit the course website

http://homepage.ntu.edu.tw/~karchung/phon1intro1F12.htm
c. Grading:
- Pronunciation journals: take notes on all pronunciation corrections, class notes, and occasional
notes on your observations of language in use
E.g. How people talk to younger children: 小車車 is used to refer to toy cars; no tone change
when reduplicated;
But with other words, especially familial relationships, a third plus a second tone are often used
in “baby talk” or to show intimacy in Taiwan: 小鞋鞋、哥哥、妹妹、爸爸.
- Participation includes: asking questions, answering questions, bringing up discussion topics, etc.
- Dictations count very little toward your final grade. Mistakes are what help you learn.
- Chapter tests will be given after each chapter is finished, and will count fairly heavily towards
your final grade.
- “Attitude is everything.”
d. Requirements
- Join the class Facebook group at https://www.facebook.com/groups/310197789078925/
(mandatory)
- Class-related correspondence should be conducted via your NTU email account
- Attendance:
If you miss more than three classes or are late to more than five classes without a valid excuse,
you may fail this class.
Students are to be in the classroom by 8:10.
2
Sep. 10, 2012
The classroom doors will be closed and locked at 8:20.
- Cell phones must be turned off during class.
- What you will need:
i. Textbook: Available at Crane's 文鶴 (金山南路二段 200 號 8 樓).
Ladefoged, Peter & Keith Johnson. 2011. A Course in Phonetics. 6th ed. International Student
Edition. Boston: Wadsworth Cengage Learning. 322pp. Paper; includes CD-ROM.
Make sure you have a textbook by Wednesday, September 12.
ii. Notebook:
Class notes will be handed in every Monday in single sheets. You are expected to take notes not
only on the contents of the textbook and supplements, but also on the pronunciation corrections
you and other students receive when reading aloud in class. Tracking your mistakes will help you
identify which areas you need to work on!
iii. Other things:

- A plastic pocket or a pocket folder to save everything
- A plan from YOU to improve your pronunciation
- A computer with a headset with a microphone (for recording assignments) and Internet access
 For more information about tools for taking notes please see the course webpage
http://homepage.ntu.edu.tw/~karchung/phon1intro1F12.htm
* Why is correct pronunciation important?

- Native speakers will feel tired if they have to constantly listen to wrong pronunciations
(especially incorrectly stressed words) or a strong accent.
 The next time they meet you, they may not be as anxious to speak with you.
 Loss of patience from the listeners  Loss of power to you
- Correct pronunciation is:
 Listener-friendly (though tolerating different accents is still necessary!)
E.g. A Taiwanese student’s pronunciation of “TV games” was unintelligible to the instructor
in class, and on a later occasion to all of the native speakers of English attending an international
conference.
3. Introduction to the areas of linguistics and phonetics

a. Visit the webpage of UCL (University College London) to learn about major branches of
linguistics and to know what phonetics is.
http://www.speechandhearing.net/entrance/introduction.php#Linguistics
3
Sep. 10, 2012
*Terms in Chinese:
i. Major branches of linguistics
(1) syntax 句法學 (指語言學中之一支)；語法 (中國慣用)；文法 (台灣慣用)
(2) semantics 語意學
(3) pragmatics 語用學
E.g. Saying “Thank you” is not as established a custom in the PRC as it is in Taiwan and in
English-speaking countries..
E.g. People from Taiwan use “不好意思” a lot.
(4) morphology 構詞學；型態學: How words are put together.
(5) phonology 聲韻學；音韻學 Study of the systems, structure, patterns and rules of language
sounds.
(6) phonetics 語音學
E.g. put [pʊt] vs. [pʰʊt]—is the /p/ strongly aspirated or not?
(7) psycholinguistics 心理語言學
E.g. A translation exam included this sentence: Americans are all fat and no meat. They think
they are the beacon of the world.
 Many people understood beacon as bacon, probably because they were influenced by
context, i.e. the mentions of “fat” and “meat”, (this is called “priming”), so they chose a
food-related meaning of a subsequent word they read. (Another reason may have been because
they were not familiar with the word “beacon”, and simply took it to be a more familiar word,
bacon”.)
(8) sociolinguistics 社會語言學
(9) computational linguistics 計算語言學 (relies much on data collection, or “corpora” 語料庫)
ii. Subcategories of phonetics

(1) articulatory phonetics 發音語音學
(2) auditory phonetics 聽學（or 聽覺、聽辨）語音學
(3) acoustic phonetics 聲學語音學
(4) experimental phonetics 實驗語音學
b. Visit Elizabeth Pyatt’s webpage about linguistics for comparison (optional)

http://www.personal.psu.edu/staff/e/j/ejp10/lingland/faqling.html
4. Homework
a. Take a “before” snapshot of yourself speaking a number of languages; due September 26, 2012:
http://homepage.ntu.edu.tw/~karchung/intro%20page3e.htm
When writing a page of comments on the recording:
i. Talk about how you feel about your pronunciation at this point
ii. Put your comments into an essay (either in double space or single space)
4
Sep. 10, 2012
iii. Do not save your comments as a “.docx” file. Save it as a “.doc” or “.pdf” file.
Mail both the MP3 file and your essay to Ms. Chung at feathermountain@gmail.com
b. Have a look at course webpage 5. Vocal tract and points of articulation

http://homepage.ntu.edu.tw/~karchung/intro%20page%205.htm
Use the image at the second link for learning the names of the articulatory organs
(You will need to be able to produce line drawings of the vocal tract. Practice!)
Pronunciation corrections: framed syllables are stressed; * = tonic stress
Ladefoged [ˈle-də-foʊ -gəd]  [ˈlæ-də-foʊ-gəd] / [ˈlæ-di-foʊ-gəd] (Danish:

phonetics [foˈnɛtɪks]  'læðəfoγə)
[foˈnætɪks]  [fəˈnɛtɪks] (reduced vowel, [ɛ] instead of [æ])
phonetics resources phonetics re*sources  pho*netics resources (compound noun)
speech sounds  speech *sounds (noun + verb)
generated  generated (stress)
vocal tract*  *vocal tract (used as a compound)
…listener, … no rise  rise to signal continuation (continuation rise)
recognise [ˈrɛkənaɪz]  [ˈrɛkəgnaɪz] (the [g] was missing)
them [ðɛn]  [ðɛm] (closure of lips when pronouncing “m”)
called [kəʊld]  [kɔld] (vowel)
articulatory phonetics [ɑrtikju'lætori]  [ɑr'tikjulətori] (vowel)
articulatory pho*netics  ar*ticulatory phonetics (stress: old
information doesn’t have stress)
auditory pho*netics  *auditory phonetics (old information)
speech pro*duction  *speech production (compound noun)
experimental pho*netics  experi*mental phonetics (old information)
*Compound noun:
Combination of two (or more) nouns in which the former modifies the latter; the noun being modified is
NOT stressed.
Read this article (on pp. 12-14) on compound noun stress, entitled 英語教學死角：複合名詞重音,
for more details:
http://homepage.ntu.edu.tw/~karchung/pubs/73_hello_et.pdf
Other articles in the same series are available here. ALL of these contain material that we will cover and
come back to many times in class, so you are strongly advised to read them all now. They are short, in
Chinese, and easy to read:
http://homepage.ntu.edu.tw/~karchung/Karen/Karen_Chung_publications.htm#CET
5
Sep. 10, 2012
Copyright Declaration
Page Work Licensing Author / Source

5 http://homepa Caves Educational Training Co., Ltd / Hello! E.T. Issue No. 70
http://www.cet-taiwan.com/drcet/type.asp?title_type=Hello%
ge.ntu.edu.tw/
20E.T.
~karchung/pu This work is licensed by Caves Educational Training Co., Ltd
bs/73_hello_et for the use of OCW ONLY. The copyright belongs to the above
.pdf mentioned creator(s).
6
Sep. 12, 2012
Fall 2012
Instructor: Karen Steffen Chung 史嘉琳老師
Unless otherwise noted, the course materials are licensed under Creative Commons
2. First Peek into the World of Phonetics

1. Dictation
2. A Course in Phonetics: Chapter 1: Articulation and Acoustics
3. Homework
II. Notes
1. Dictation
a. Test instructions
- Use a sheet of loose leaf paper and write the following information in the upper right hand
corner: your name in English and Chinese, your student number, “Phonetics”, and the date.
- Use brackets [ ] for IPA transcriptions.
- Watch out for vowels and listen to final stops.
- Watch the instructor ’s mouth.
- Scoring: 5 points for spelling, 5 points for IPA, extra 10 points for bonus sentence.
b. Answers and corrections

What was said What was heard and why
(1) bat [bæ t] (1) bag [bæg]: the final stops were unreleased in this dictation,
(2) bid [bɪd] making it hard to identify the place of articulation of the final
(3) but [bʌt] stop
(4) beat [bit] (2) bid [bɪd] heard as: bing [bɪŋ]: the voicing of [d] makes it
(5) bet [bɛt] sound like a nasal
(6) bead [bid] (6) bee [bi]: stop unreleased  hard to hear the final stop
(7) bed [bɛd] (7) bad [bæd]: [ɛ] vs. [æ]- typical Taiwanese vowel error
(8) bud [bʌd] (8) bug [bʌg]: the same as for (1)
(9) bit [bɪt]
(10) bad [bæ d]
Bonus sentence:
The mother raced her children. raised: listener’s perception affected by the context
[ɹeɪsd] [ɹeɪːzd]
1
Sep. 12, 2012
*unreleased stop
- 不除阻塞音: The compressed air behind the obstruction is not released, but the articulators
maintain the gesture for the sound, e.g. for /-t/; the tongue tip remains on the alveolar
ridge, with air pressure built up behind the tongue tip.
- Issues in handwriting:
i. i & j: a dot, not a line (a tone mark)
ii. [ŋ]: not [ŋ] but [ŋ]
Do NOT use Times New Roman for IPA symbols.
2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 2-4)

a. The word articulation: (1) jointed segments; (2) the way we make sounds.
b. Fields that different phoneticians are interested in:

- Learning about all the possible sounds of human speech
e.g. ejectives in Georgian (Republic of Georgia); clicks in some of the languages of southern
Africa
- Pathological speech (correcting speech defects)
e.g. people with a cleft palate (唇顎裂) or hearing disability
- Dialect studies (for performance; to fit into a group; coaching, to avoid being made fun of)
*Why do some people want to learn a particular form of a language so as to avoid being made
fun of?
- People spontaneously have an emotional reaction to strange or different accents, so some
people turn to a dialect coach to minimize such reactions.
- Making computers talk more intelligibly or getting computers to recognize speech
e.g. Siri of iPhone (quite good, but still unnatural due to imperfect intonation and timing)
c. Speech production:
- Most speech sounds are the result of movements of the tongue and the lips
- We humans are able to impart information efficiently:
i. The tongue and lips can make about 25 different gestures to form the sounds of English
- The tongue is especially active – it’s the most flexible, precise, and nimble muscle of the
body
- Look at Figure 1.1 (frames from an X-ray movie) on p. 3
- Watch an X-ray movie of a short English sentence:
http://www.phon.ox.ac.uk/~jcoleman/try_not.mov
Watch for:
(1) when and where the tongue tip touches the palate
(2) the forward and backward movement of the tongue
(3) the raising and lowering of the velum, and the lowering and raising of the jaw, or
mandible
2
Sep. 12, 2012
 To know more about the vocal tract and the point of articulation, visit the course webpage
*About the X-ray movies

- Most of these were filmed in 1950s, when people did not know about the dangers of X-ray
filming
 Many of the people who were filmed to make these movies died of cancer!
 For more information about X-ray cineradiography, please visit the course webpage
ii. The source of energy for nearly all speech sounds = air pushed out of our lungs (talking
while breathing in is much harder than while breathing out)
3. Homework
a. The Audacity assignment (the “before” snapshot homework, due September 26):
Do NOT have sweets or dairy products, but you can drink water with lemon (with no sugar)
before recording.
(Details about the essay can be found in the handout for September 10)
b. Read these two articles (one article in two parts) on the Echo Method and 10 Minutes a Day:
c. Notes for week 1 due September 17. Please write or type on A4 paper.
- Style:
Any style of note taking is fine, but the notes should be comprehensible to someone else
reading them. This is an opportunity for you to organize and consolidate what you have
learned.
- Content:
i. Pay special attention to things mentioned in class that are not in the textbook; also note any
questions you have and things you're not sure of.
ii. The pronunciation corrections are also especially important - these should be organized,
so if the same type of error is corrected more than once, they should be grouped together.
e.g. stressing the wrong syllable, or using a full vowel instead of a schwa, or using the wrong
vowel.
d. Read the Webpages up to 7b
3
Sep. 12, 2012
articulation [ɑɹtɪkəˈleɪʃən]  [ɑɹtɪkjuˈleɪʃən] (vowel)

describing [dəsˈkɹaɪbɪn]  [dəsˈkɹaɪbɪŋ] (nasal)
[bɪŋ] same as Mandarin “冰”
wanting ['wɑntɪn]  ['wɑntɪ̞ŋ] (nasal)
phoneticians [foʊnəˈtɪʃənz]  [fɑnəˈtɪʃənz] (vowel)
ɪn*terested  *ɪnterested (stress) ['ɪntɹɪstəd] or ['ɪntɹəstəd]
looking ['lʊkɪn]  ['lʊkɪŋ] (nasal)
recognize [ˈɹɛkənaɪz]  [ˈɹɛkəgnaɪz] (the [g] was missing)
pur*poses  *purposes
they are talking and… no pause before and  pause before a conjunction
airstream mechanisms airstream me*chanisms  *airstream mechanisms (compound noun)
mechanism me*chanism  *mechanism
speech production speech pro*duction  *speech production (compound noun)
Similar: con*venience store
Doris ['do-rɪs]  ['dɔɹɹɪs] (the [ɹ] starts from the middle of [ɔ])
tongue [toʊŋ], [taʊŋ]  [tʌŋ] (vowel)
particu*lar  par*ticular
of [of]  [əv] (vowel and consonant)
“[of] is wrong” – wrong relative to SAE
e.g. “says” (SAE and SSB – Standard Southern
British English) [sɛz]; (BE dialect) [seɪz]
…information, … no rise  rise to signal continuation (continuation rise)
any *sound *any sound (any is often stressed for
emphasis)
res*piratory  *respiratory
breathing in instead of out  breathing in instead of out (stress the
contrasting elements)
*How unstressed syllables are realized:
- Pronounced with a low even pitch (i.e. flat tone).
*Some words that Taiwanese students often confuse:

- ton [tʌn] tongue [tʌŋ] tone [toʊn]; sore [sɔɹ] sour [saʊɹ]
- bottom ['bɑɾəm] button ['bʌtʔən] or ['bʌʔʔən]
*How to do a continuation rise:

- the last stressed syllable of an utterance receives a high pitch, the voice then falls to a low pitch, and
then rises slightly, to show that the speaker isn’t finished and is about to continue speaking
4
Sep. 12, 2012

1 National Taiwan University/Karen Steffen Chung 史嘉琳
This work is licensed by Creative Commons
Attribution-NonCommercial-ShareAlike 3.0 Taiwan
2 http://www.ph John Coleman, Ladan Baghai-Ravary, John Pybus, and Sergio
Grau (2012) Audio BNC: the audio edition of the Spoken British
on.ox.ac.uk/~j
National Corpus. Phonetics Laboratory, University of Oxford.
coleman/try_n http://www.phon.ox.ac.uk/AudioBNC
ot.mov The supporting annotation and transcription files are Copyright
© 2011 The University of Oxford, and are made publicly
available under a Creative Commons Attribution 3.0 Unported
License.
3 http://homepa National Taiwan University/Karen Steffen Chung 史嘉琳
ge.ntu.edu.tw/ This work is licensed by Creative Commons
~karchung/int
ro%20page%2
05.htm
~karchung/int
ro%20page%2
06.htm
http://www.cet-taiwan.com/drcet/type.asp?title_type=Hello%20
ge.ntu.edu.tw/
E.T.
ge.ntu.edu.tw/
E.T.
5
Sep. 17, 2012
Fall 2012
3. The Vocal Tract and Places of Articulation

1. US accents
3. Homework
II. Notes
1. US Accents
a. Which parts of the country have different accents? (Continued from Unit 1 handout)
- The East Coast: New England
- The South: starting from Maryland down to Florida and west to Texas
* Florida: has many immigrant residents from the North, especially older retired people, who
do not have a Southern accent
* Black English: similar to Southern US English
- Mid-Western: more homogeneous; basis of the “General American” accent favored in
broadcasting
b. How did these variations develop? How does a language change?

- Immigrants with different backgrounds
- Time
e.g. Mandarin in Taiwan and in China: People from all over China moved to Taiwan around
1949
 Mandarin experienced a relatively independent development on the two sides of the
Taiwan Straits from then on
*What is the time differential between the settling of the East Coast, the South and the West?
- In the 17th century, the first groups of immigrants from England and elsewhere in Europe
settled in the East and then in the South  More time for the language to change in these
parts than in the rest of the US, i.e. the Midwestern and Western parts of the country
c. Related studies:
- US Dialect survey
http://www4.uwm.edu/FLL/linguistics/dialect/maps.html
1
Sep. 17, 2012
e.g. We can draw isoglosses (imaginary lines dividing the country according to variations in
speech) showing, for example, what 汽水 (pop/soda/coke/fizzy drink) is called in which
regions
2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 4)

a. The airstream powering speech: Lungs  trachea  larynx  vocal folds
 apart: voiceless sounds
 stretched together tightly  vibration driven by airstream  voiced sounds
*Terms in Chinese:
- trachea 氣管
- larynx 喉
- vocal folds/vocal cords 聲帶
- voiced 帶音、濁音、有聲
- voiceless 不帶音、清音、無聲
 Go over course webpage 5. Vocal tract and places of articulation

b. Places of articulations
- Look at the figure showing the names of the articulatory organs
http://www.music.mcgill.ca/~gary/307/week9/node8.html
(To remember the terms: Treat it like a game of “Concentration”!)
*Terms in Chinese:
- alveolar ridge 齒齦 (the full name is 齒齦隆骨)
- hard palate 硬顎
- soft palate 軟顎
- nasal cavity 鼻腔
- uvula 小舌
- tip of the tongue 舌尖
- blade (of the tongue) / the tongue blade 舌葉
- front (of the tongue) 舌前 (sometimes called 舌尖 in Chinese)
- center (of the tongue) 舌中
- back (of the tongue) 舌後 (sometimes called 舌根 in Chinese)
- root (of the tongue) 舌根
- esophagus [iˈsɑfəgəs] 食道 (the figure uses the British spelling “œsophagus”)
- epiglottis 會厭 (closes over the top of the trachea when eating, drinking, and swallowing)
- pharynx 咽 – pharyngeal wall 咽壁 or 咽喉壁
2
Sep. 17, 2012
*Why does our voice have a nasal sound when we have a cold and our nose is stuffed up?
- The nasal cavity is filled with spongy tissue which can hold liquid; and when our nose is
stuffed, the resonance pitches of the nasal cavity are changed. We can hear this in our speech
even though our nose is stuffed up and we can’t pronounce nasal sounds or breathe
normally.
c. Vocal folds
- Where your vocal folds are: the part on the front of your neck (your “throat”) where you can
feel the strongest buzzing when making voiced sounds
The frequency of vocal fold vibration in ordinary speech is around 200 cps (cycles per second)
or 200 Hz (Hertz); the averages are about 120 Hz for men, 235 Hz for women, and 265 Hz for
children
- Vocal fold movements occur without conscious control
- Watch a video of vocal folds vibrating at
http://youtu.be/9Tlpkdq8a8c
- Both whispering and shouting harm your vocal folds
- Coffee and tea are diuretic (they make you urinate a lot); drinking a lot of coffee or tea will
dehydrate you (cause you to have less water in your body), and this is not good for your voice
- People without vocal folds can use a device which generates vibrations that can be used to
produce speech, but there is no intonational variation in such speech.
 People without vocal folds also have the option of learning esophageal speech 食道語, in
which the patient swallows and then manipulates gulps of air (“burps”) to power their speech
 Go over course webpages 7a. and 7b.

http://homepage.ntu.edu.tw/~karchung/intro_page_7b.htm
3. Homework
a. Read two articles (it is actually one article in two parts) on the Echo Method and 10 Minutes a
Day
b. Make 12 drawings of the head using the model on p.27 of your textbook; due September 19:
Use an A4 paper with six drawings on each side. Three of the 12 drawings are to be labeled.
c. Finish links on course webpage 7a

3
Sep. 17, 2012

windpipe [ˈwindpaɪp]  [ˈwɪndpaɪp]
trachea tra*chea  *trachea (tra*chea out of date)
larynx [ˈlæɹɪŋks]  [ˈlɛɹɪŋks] ( in Midwestern US English – learn one
consistent variety of English  this trains your ability
to imitate any kind of speech)
two small muscular no pause  pause after each word (content words  pause for the
folds brain to process what’s been said)
folds [fɔlz]  [foldz] (vowel; final affricate)
opposed [ə'pɔzd]  [ə'poʊzd] (vowel)
yours [jɔɹs]  [jɔɹz] (word-final s after voiced consonant  [z])
in [iŋ]  [ɪn] (vowel and nasal)
free passage no pause  pause between stressed, monosyllabic content words
vibrate vi*brate  *vibrate (word stress) cf. vi*bration (n.)
alveolar ridge [ælˈvilɚ ˈridʒ]  [ælˈvilɚ ˈɹɪdʒ] (vowel)
(BE) [ælvi'eʊlə]; in AE sometimes [ælˈviəlɚ]
al*veolar ridge  alveolar *ridge (adjective-noun phrase)
hard *palate  *hard palate (when contrasting with “soft palate”)
soft *palate  *soft palate (contrasted with “hard palate”)
pharynx [ˈfæɹɪŋks]  [ˈfɛɹɪŋks]
pha*ryngeal wall  pharyngeal *wall (adjective-noun phrase)
voiced [vɔːɪsd]  [vɔɪst] (length of first vowel in the diphthong)
each [ɪtʃ] [itʃ] (vowel)
them [ðɛn]  [ðɛm] (closure of the lips when pronouncing “m”)
both [bɔθ]  [boʊθ] (vowel)
these sounds_are no pause  pause after the subject
name [nɛm]  [neɪm] (vowel)
*vocal fold vibrations  *vocal fold vibrations (compound noun)
also [ˈoʊsoʊ]  [ˈɔlsoʊ] (vowel)
each of these*words  each of these *words (phrase stress)
[ðɪz]  [ðiːz] (vowel; and do not stress these here)
suggested [səˈdʒɛstəd]  [səgˈdʒɛstəd] (the [g] was missing); (BE) [səˈdʒɛstɪd]
cf. figure (AE) [ˈfɪgjɚ] vs. (BE) [ˈfɪgə]
4
Sep. 17, 2012
* Tonic stress:
The final stress of the utterance (or thought group) is very high and signals the end of the utterance.
e.g. I’m going to the *store.
*Continuation rise (also mentioned on September 12):

The last stressed syllable before a pause receives a high falling pitch followed by a gentle rise, to signal that
the sentence is not finished. It usually occurs before conjunctions, punctuations, and prepositions.
e.g. … the trachea, to use the more technical term
* Stress timing:
English is often described as having a stress-timed rhythm, i.e. there is a relatively fixed interval of time
from one stressed syllable to the next stressed syllable; unstressed syllables are relatively shorter, and can be
made even shorter when more of them are crowded between two stressed syllables, in order to maintain a
regular beat. For example, the time required to say “A BIG BLACK BEAR” is about the same as that
required to say “A BEAUtiful exPENsive Overcoat.” On the contrary, French is an example of a language
that is said to be syllable-timed, i.e. there is less variation in the length of each syllable, and the rhythm is
less governed by the intervals between stressed syllables.
5
Sep. 17, 2012
1 http://www4.u University of Wisconsin-Milwaukee/Bert Vaux
wm.edu/FLL/l This work is used subject to the fair use doctrine of Article 46,
inguistics/dial 52, 65 Taiwan Copyright Act.
ect/maps.html
~karchung/int Attribution-NonCommercial-ShareAlike 3.0 Taiwan
ro%20page%2
05.htm
2 http://www.m McGill University/MUMT 307: Music & Audio Computing II
usic.mcgill.ca/ http://www.music.mcgill.ca/~gary/307/index.html#Outline
~gary/307/we This work is licensed by McGill University for the use of OCW
ek9/node8.ht ONLY. The copyright belongs to the above mentioned
ml creator(s).
3 http://youtu.be Youtube/James Thomas
This work is licensed by Creative Commons Attribution 3.0
/9Tlpkdq8a8c
Unported License.
~karchung/int
ro%20page%2
07.htm
~karchung/int
ro_page_7b.ht
m
ge.ntu.edu.tw/
E.T.
ge.ntu.edu.tw/
E.T.
6
Sep. 19, 2012
Fall 2012
4. Speech Production and Sound Waves

1. Dictation
3. Homework
4. Book Sharing
II. Notes
1. Dictation
a. Test instructions
- Use a normal-sized clean sheet of paper; do not use scrap paper.
- Write the following information in the upper right hand corner: your name in English and
Chinese, your student number, Phonetics, and the date.
b. Answers and corrections

(1) pick [pɪk] (1) pick [pɪk]: contrasted with 9. pig; final stops were released this time,
and the words were thus easier to get right
(2) pan [pæn] (2) pan [pæn]
(3) peek/peak [pik] (3) peek [pik]
(4) pin [pɪn] (4) pin [pɪn]
(5) pain [peɪn] (5) pain [peɪn]
(6) ping [pɪŋ] (6) ping [pɪŋ]
(7) peg [pɛg] (7) peg [pɛg]
(8) pen [pɛn] (8) pen [pɛn]
(9) pig [pɪg] (9) pig [pɪg]
(10) pang [pæŋ] (10) pane [pæn]: final nasals  many Taiwanese students have trouble
distinguishing between alveolar and velar nasal finals.
- IPA symbol [g] not “g”
1
Sep. 19, 2012
* How to distinguish between the nasals [n] and [ŋ] ?

- [n]: alveolar sound (tip of the tongue). Chinese 恩典的恩ㄣ [ən]
- [ŋ]: velar sound (back of the tongue). Chinese 骯髒的骯ㄤ [ɑŋ]
* [n] and [ŋ] in Taiwan:

- ㄣ [ən] and ㄥ [əŋ] are often not distinguished by Taiwan Mandarin speakers
- The vowels in ㄢ [an] and ㄤ [ɑŋ] do not sound alike, but they belong to the same phoneme;
the alveolar (front) nasal results in a more front vowel [a]; the velar nasal [ŋ] results in the back
vowel [ɑ].
- One of Ms. Chung’s former students said that ㄢ and ㄤ are sometimes confused in Kinmen
Mandarin  could not be verified by to two Kinmen natives in the class, a TA and a student.
* Vowels and nasals:

- Vowels are influenced by neighboring sounds
e.g. bans vs. bangs (different nasals  change the quality of the vowels)
- [ɪn] and [ɪŋ] are not interchangeable in US English; however, a final [ɪn] is often used in
informal speech in place of [ɪŋ], usually only in -ing verb forms, e.g. goin’, doin’, for going,
doing, but also for something: something [ɪŋ]  somethin’ ['sʌmθən] or ['sʌmpʔm]
 Please see CET articles numbers 9 and 10 (May/June and July/August 2013 – forthcoming)
for more on English nasals.
2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 4- 8)

a. Speech production:
- Articulators: parts of the vocal tract that can be used to form sounds
- Four components of the speech production mechanism:
i. the airstream process: how air is moved through the vocal tract (air is the energy source that
powers speech)
ii. the phonation process: actions of the vocal folds (voicing)
- vibrating  voiced sounds
- apart  voiceless sounds
iii. the oro-nasal process: controls whether the airstream exits through the mouth or the nose:
- velum lowered  nasal sounds e.g. [m] [n] [ŋ]
- velum raised  oral sounds
iv. the articulatory process: movements of the tongue and the lips interacting with the roof of the
mouth (i.e. alveolar ridge, hard palate, soft palate) and the pharynx
* Clarification of the terms:

- The word process in this context does not mean 過程 but the articulators.
- The compound noun speech event means 講話的事件
2
Sep. 19, 2012
b. Sound waves
- Speech sounds can be the same or different in three ways:
i. pitch 音高: depends on the frequency of the vibration of the vocal folds
ii. loudness 音量、強度: depends on the energy of the sound, how hard the air is pushed out of
the lungs
iii. quality 音質（音的性質、內涵、倍音結構等）: determined by the gestures of the tongue and
other articulatory organs
- The acoustics of speech:
i. Actions of the vocal organs superimposed on the outgoing flow of lung air
 small variations in air pressure occurring rapidly one after another
 these variations, in the form of sound waves, move through the air
 the sound waves reach the ear of a listener and cause the eardrum to vibrate
ii. A graph of a sound wave is similar to a graph of the movements of the eardrum
- p. 8 Figure 1.4
The ordinate 直軸 shows loudness (amplitude 振福)
The abscissa 橫軸 shows the time and the pitch (how many times the vocal folds vibrate per
second = cps or Hz)
Figure A: a sample wave form: “on” [ɑn]

ordinate ↕: the amplitude (loudness) of the sound
abscissa: time → the duration and pitch of the sound
* Definition of pitch:
- how many times the vocal folds vibrate per second = cps or Hz
- cps = cycles per second; Hz = Hertz, named after the German physicist Heinrich Hertz, to
replace the abbreviation “cps”, in 1930
3. Homework
a. The Audacity assignment (the “before” snapshot homework; due September 26):
3
Sep. 19, 2012
Combine the recordings in different languages into one MP3 file!
b. Weekly notes due September 24, Monday
c. Read the article on [i] and [ɪ] at

4. Book sharing:
Stanley Dubinsky and Chris Holcomb, Understanding Language through Humor. 2011.

air passages air *passages  *air passages (compound noun)
Important [ɪmˈpɔɹɾənt]  [ɪmˈpɔɹtʔnt] or [ɪmˈpɔɹʔʔnt] (NO TAP here; should
be [t] + glottal stop or 2 glottal stops)
the oral tract, … rushed through, flat intonation  pause and rise before punctuation (rhythm,
continuation rise)
to *form sounds, to *form sounds,  to form *sounds, (place the tonic stress on the
last content word in a thought group; high
falling pitch with gentle rise if more is coming)
articulators [ɑɹˈtɪkəleɪɾɚz]  [ɑɹˈtɪkjuleɪɾɚz] (palatalization: [ju])
mechanism [məˈkænɪzəm]  [ˈmɛkənɪzəm] (vowel, stress)
speech production speech production me*chanism  *speech production mechanism (compound
mechanism noun)
whole [hɔl]  [hol] (vowel; the same as “hole”)
air stream process air stream *process  *airstream process (compound noun)
articulatory ar*ticulatory process  ar*ticulatory process (compound noun)
process
respiratory system *respiratory system  *respiratory system (as a compound)
vocal folds *vocal folds  *vocal folds (treat as compound)
oro-nasal process oro-nasal process  oro-*nasal process (treat as compound)
sound waves sound waves  *sound waves (compound noun)
so far so far  so far (the final stress is the highest)

speech sounds speech *sounds  *speech sounds (compound noun)
how they are flat intonation  how they are *made, … what we can *hear
made, … what we (stress the contrasting words; tonic stress)
can hear
also [ˈoʊsoʊ]  [ˈɔlsoʊ] (vowel)
4
Sep. 19, 2012
terms [tɜ˞nz]  [tɜ˞mz] (closure of lips when pronouncing “m”)

speech speech pa*thologists  *speech pathologists (compound noun)
pathologists
some sounds some sounds  some sounds (stress some only when it means
‘certain’; here it means ’a few’)
structures/rather [ˈstɹʌktʃəz] / [ˈɹæ ðə]  [ˈstɹʌktʃɚz] / [ˈɹæ ðɚ] (the [ɚ] was missing)
knowledge [ˈnɔlɪdʒ]  [ˈnɑlɪdʒ] (BE [ɒ]; AE [ɑ])
synthesize [ˈsɪnsəsaɪz]  [ˈsɪnθəsaɪz] (stick out the tongue tip for [θ])
recognition [ɹɛkə'nɪʃən]  [ɹɛkəg'nɪʃən] (the [g] was missing)
works [wɜks]  [wɜ˞ks] (no “r”)
often ['oʊfən]  ['ɔfən] (vowel)
x-rays [æks]  [ɛks] (vowel)
same [sɛm]  [seɪm] (vowel)
pitch [pitʃ]  [pɪtʃ] (vowel)
said [seɪd]  [sɛd] (vowel)
scale [skɛl]  [skeɪl] (vowel)
vowel *quality  *vowel quality (compound noun)
a higher pitch or no pause  pause before conjunctions
vibrating  vibrating (word stress) cf. vibration (n.)
variation [væɹi'eɪʃən]  [vɛɹi'eɪʃən] (in Midwestern US English)
When they reach…  do not stress pronouns
father [fɑðə]  [fɑðɚ] (the [ɚ] was missing) cf. (BE) [fɑðə]
0.6 “point six”  “zero point six” (number)
* Pronunciation of [ɔ]:
- It sounds like 芋仔 (taro) in Southern Min pronounced quickly, as a single syllable
- About half of the people in the United States don’t distinguish between [ɔ] and [ɑ]
* How do we define a vowel whether it is a long or short vowel?

- The notions of “long” and “short” vowels proceeded from an analysis of British English, and in British
English, the names are apt: timewise, short vowels have a shorter duration, long vowels have a longer duration.
In American English, however, this correspondence simply does not apply consistently.
We define “short” and “long” vowels phonologically, according to their distribution.
- Definition: A vowel that can only occur in closed syllables (閉音節) is a short vowel.
 Short vowels cannot occur in an open syllable.
open syllable 開音節 closed syllable 閉音節
be [bi] beat [bit]
X [*bɪ] (not a bit [bɪt]
possible word)
e.g. [ɔ]: saw [sɔ] law [lɔ]  [ɔ] is not a short vowel
5
Sep. 19, 2012
- Since by definition, [ɔ] is a long vowel, why do most Taiwanese students think it is a short vowel?
Because our textbooks and teachers say so.
 Don’t take anything for granted. Always question everything!
* Why must we be picky about the seemingly trivial details in pronunciation?

- In some cases, a slight pronunciation error, especially if it makes a word sound like a vulgar or taboo word,
can set off an emotional reaction in the listener and thus distract them.
e.g. If the speaker means to say beach [bitʃ] but says [bɪtʃ] instead, the listener will probably
do a double take and may miss what the speaker says next.
6
Sep. 19, 2012
Page Work Licensing Author/Source

Attribution-NonCommercial-ShareAlike 3.0 Taiwan.


~karchung/int
ro%20page3e.
htm
ge.ntu.edu.tw/
E.T.
4-5 National Taiwan University/Karen Steffen Chung 史嘉琳

7
Sep. 24, 2012
Fall 2012
Unless otherwise noted, the course materials are licensed under Creative
Commons Attribution-NonCommercial-ShareAlike 3.0 Taiwan (CC BY-NC-SA
3.0)
5. Places of Articulation I
1. Phonetic Symbols: KK vs. IPA
3. Homework
II. Notes
1. Phonetic Symbols: KK vs. IPA
a. The Kenyon and Knott (KK) phonetic symbols KK 音標
- Most Americans have never heard of the “KK phonetic symbols”
- KK is a subset of the larger set of IPA symbols, with some specially adapted features for
American English
- Whether learning KK is an advantage is still under debate; as a phonetics teacher, I think it is
useful and important; sometimes the teachers who teach it do not teach the correct
pronunciations, but knowing the symbols is a huge advantage in any case
- KK is no longer taught in many Taiwan classrooms
b. The Daniel Jones (DJ) phonetic symbols

- The phonetic symbol system that was taught in Taiwan before the mid 1960s, when Standard
Southern British (SSB – a variety of English spoken in Southern England; though it now has
fewer and fewer native speakers) was the standard for English teaching in Taiwan
- Differences between KK and DJ:
DJ posits pairings of “long” and “short” vowels, and uses the length mark , a diacritic, or
diacritical mark 區別發音符號, to mark the “long” vowels, for example:
KK DJ
foot [ʊ] [u]
food [u] [u:]
The two dots are actually triangles, with the top one inverted. A regular colon : is often used in
handwriting and also on computers for typographical convenience.
1
Sep. 24, 2012
Here is a chart with the DJ pronunciation symbols:

http://www.inf.fu-berlin.de/lehre/WS05/19617-K/PhonSymb.jpg
c. The International Phonetic Alphabet (IPA) 國際音標 (also called 萬國音標; this Chinese term
has mainly been used to refer to the DJ pronunciation symbol system)
- IPA can also refer to the International Phonetic Association
- The IPA has been revised a number of times over history, most recently in May 2005 with the
addition of a letter for a labiodental flap. See:
http://en.wikipedia.org/wiki/International_Phonetic_Alphabet#History
- There is also an “Americanist” system still in use, especially in works on phonology and in
TESOL texts. It uses, for example, the háček symbol: /š/for /ʃ/ and/č/ for /tʃ/; also /ay/
for /a/.See this link for details:
http://en.wikipedia.org/wiki/Americanist_phonetic_notation
- Key differences between general IPA and KK:
KK IPA
no /o/ [oʊ]
day /e/ [e]
rain /r/ []
water /t/ [ɾ]
Kenyon and Knott considered the sounds [oʊ] and [e] less diphthongal than the other
diphthongs, so they used only single vowel symbols
* These choices may account for some of the pronunciation mistakes in Taiwan:
- IPA [e] KK [e] the vowel in name is often mispronounced as []or [æ]
- IPA [oʊ] KK [o] the vowel in no is often mispronounced as [].
The last symbol is the “tap” symbol, used for a /t/ between two vowels.
* Diphthong 雙母音:
- Definition: two vowels pronounced quickly, one after the other, within a single syllable
- Very often the component vowels used in diphthongs do not exist alone in a given language
as monophthongs (a single vowel sound)
e.g. the [a]in [a] and [aʊ]does not exist as a monophthong (單母音) in English; and [ɑ] is
always a monophthong
* “r” sound:
- The most common “r” sound in the world’s languages is either a tap [ɾ] (e.g. [AmE] water;
Spanish pero ‘but’) or a trill [r] (Spanish perro ‘dog’)
2
Sep. 24, 2012
- The English “r”, written [] in IPA, is a relatively rare sound in the world’s languages;
however, it is found in Beijing Mandarin, and in the dialect of Dutch spoken in Leiden
- KK, and most other pronunciation symbol systems, use a regular [r] for the English “r” for
typographical convenience

a. Reviewing the waveform on p. 7
- A waveform can tell us: 1. whether a sound is voiced or voiceless; 2. its loudness; 3. its
manner of articulation. What it can’t tell us: place of articulation
- The sound [f]: low amplitude not loud; [θ] is even softer
- If the amplitude of the two vowels differs it means one is louder than the other, and this is
often due to stress vs. lack of stress
- The vertical lines of the two vowels are evenly spaced because all vowels are voiced
- The vertical lines of [f] are irregular because [f] is voiceless noise
- Pulse: One vibration = one pulse; the result of the air under pressure pushing the vocal cords
apart and together one time
* Definition of noise:
- Noise: a series of disorganized, random frequencies occurring at the same time; there is no
one single identifiable frequency
Cf. Voicing an identifiable single, regular frequency over a period of time
b. Places of articulatory gestures (starting from p. 8)

- Articulators in the oral tract can be divided into two types:
i. Passive articulators: the articulators that form the upper part of the oral tract; they hardly
move in the articulation process
ii. Active articulators, e.g. the tongue and the jaw, which belong to the lower part
of the oral tract; they can produce different gestures and move toward the upper, passive
articulators when speaking
- Breaking down the word capital ['kæpəɾəl]as an example
[k] – passive articulator: velum; active articulator: the back of the tongue
[æ] – nothing is touching anything
[p] – the lower (active) lip moves up to the upper (passive) lip
[ə] – nothing is touching anything
[ɾ] – passive articulator: the alveolar ridge; active articulator: the tip of the tongue
[ə] – nothing is touching anything
[l] – (for most people) passive articulator: alveolar ridge; active articulator: the tip of the
tongue
3
Sep. 24, 2012
* Two kinds of /l/:

- There are two kinds of /l/ sound in English, the clear /l/ [l]
, which occurs before vowels,
and the dark /l/ [#], which occurs after a vowel (i.e. it’s “postvocalic”).
- Clear and dark /l/ are not as clearly distinguished in American English as in British
English
- Ways to describe language sounds :

i. Articulatory description: this is the method most often used for consonants; you describe
which organs touch which, and how;
ii. Acoustic description: this is more often used for vowels and /r/; you describe what you
hear, since it is difficult to describe the actual places of articulation
- To review the terms for places of articulation, go back to
i. course webpage p. 5 at
ii. Also refer to the class handout for September 17, 2012
* Terms:
- frontal incisors: 門牙 – related to scissors; used to cut food
- protuberance [proʊ'tubɚrəns]: 突出物
– refers to the alveolar ridge in this context
- velic closure: 軟顎封閉 – describes the state of the velum being raised, thus closing off the
nasal tract and allowing the air to go out only through the mouth to produce oral sounds
- The most mobile articulators belong to the lower surface of the vocal tract, the tip and blade
of the tongue in particular
- Breaking down the word peculiar [p'kjuliɚ]to practice describing consonant articulations:
p –two lips come together; then the back and center of the tongue are raised
k –(for most people) the back of the tongue touches somewhere between the hard palate
and the velum co-articulation due to the following [j]
l –(for most people) the tip of the tongue touches the alveolar ridge
- [t] in true and tea: the tongue moves further to the front of the mouth for tea than it does for
true coarticulation due to the following [i]
- [s]vs. [ʃ]:major differences
[s] [ʃ]
acoustic quality higher in pitch louder
articulation a groove (hollow, channel) down the V-shaped
center of the tongue tongue
4
Sep. 24, 2012
- To produce consonants, the airstream must be obstructed in some way as it passes through the
vocal tract
Consonants can be classified according to the place and manner of this obstruction
place of articulation 發音點 manner of articulation 發音方法
- Less specific categories of places of articulation, used more often in phonology than in
phonetics:
i. labial 唇音
ii. coronal 舌頂音 or 舌冠音 or 舌前音
iii. dorsal 舌背音
…vibration (which) rushed through, flat pause and rise before punctuation (rhythm and
intonation continuation rise)
smaller ['smolɚ] ['smlɚ] (vowel; very little rounding before /l/)
variation [væi'eʃən] [vi'eʃən] (in Midwestern US English)
air *pressure *air pressure (compound noun)
called [koʊld] [kld] (vowel) (Many Taiwanese have a strange
pronunciation of [], probably British-influenced
will find out will found out will find outwill
( can’t be followed by a past
participle; this error may be a kind of
hypercorrection 矯枉過正)
names [nms] [nemz] (vowel; voicing of /s/ after voiced
consonants)
principal parts *principal parts principal *parts (phrasal stress)
figure ['fgɚ] ['fgjɚ] (the [j] was missing)
the [də] [ðə] (stick out the tongue)
five [faf] [fav] (voicing; lengthen the vowel)
Cf. fife Ff]
[f 短笛
further ['f67dɚ] ['f67ðɚ] (stick out the tongue)
velum ['vləm] ['viləm] (vowel)
pharynx ['fæiŋks] ['fiŋks] (in Mid-Western AE)
uvula ['juvulə] ['juvjulə] (the second [j] was missing)
cf. uvular ['juvjulɚ] (adj.)
front [frɑnt] [frFnt] (vowel)
center ['snt9ɚ] ['snɾɚ] (tap, or at least no aspiration)
partly ['pɑt9li] ['pɑʔli] (glottal stop)
5
Sep. 24, 2012
position [poʊ'siʃən] [pə'ziʃən] (schwa; voiced [z], not [s]

on [ɑŋ] [ɑn] (alveolar nasal, not velar)
tip [tip] [tp] (vowel)
ridge [idJ] [dJ] (vowel)
other consonant flat rise to signal continuation (continuation rise)
sounds, …
important [m'pɾənt] [m'ptʔnt] or [m'pʔʔnt] (glottal stop)
deep [dp] [dip] (vowel)
cannot represent no stop glottal stop
that shows no stop glottal stop
midline ['mdlãL] [mdlan] (the final “n” was missing)
Bu t we no stop glottal stop
in [in] [n] (vowel)
in order no linking linking (the word-final consonant becomes the
if it
word-initial consonant of the following word
begins with a vowel; also, alveolar, not velar)
conceptualization [kFnsptʃuəla'zeʃən] [kFnsptʃuələ'zeʃən] (pronunciation variety:
-ization in AE usually [ə]; in BE it’s usually [a])
* “Five” issues
- Are the vowel qualities different in five and fife?
Five and fife share the same phoneme /a/, but the [a] in five is longer and the tongue and jaw are
lower.
- Three things about five to pay attention to:
i. voicing of [v]
ii. vowels before a voiced sound should be longer.
iii. [a] + voiced sound tongue and jaw lower.
+ voiceless sound tongue and jaw higher; [F] Canadian raising, also found in many
varieties of US and Scottish English
* About Canadian raising:

- Another diphthong [aʊ] [Fʊ]
e.g. about the house [ə'bFʊt ðə hFʊs] (ICRT broadcaster Terry Engel and DFLL instructor Ted
Partington both have Canadian raising)
- Not all Canadians have Canadian raising; it is found mainly in the eastern part of
Canada.
6
Sep. 24, 2012
See course Webpage (Phonetics II) 2. The American Tap & Canadian Raising at:
http://homepage.ntu.edu.tw/~karchung/Phonetics%20II%20page%20two.htm
* Glottal stop 喉塞音:

- In American English, the “t” has many variations, one of which is the glottal stop [ʔ]. One of the
environments for it to occur is when the word-final “t” is followed by a word beginning with a
consonant.
- To produce a glottal stop, imagine a basketball hitting your tummy hard. What sound do you make?
It should be a glottal stop or two, probably with a vowel between: [ʔəʔ]! Your tummy will be tensed up
when doing this. Your tongue tip doesn’t make the gesture for pronouncing [t], i.e. your tongue tip
does not touch the alveolar ridge, or anywhere else, when producing a glottal stop.
E.g. hit me [hʔ mi]
- There’s also a glottal stop in Southern Min,:
E.g. S Min 藥 [jʔ] ‘medicine’; this originally had a final /-k/which later became a glottal stop
S Min 呷 [tɕjaʔ] (‘eat’)
S Min 鴨 [ɑʔ];this originally had a final /-p/
But these days, whether the final glottal stop is present in these words or not does not seem to affect
Taiwanese Southern Min speakers’ perception of the meaning.
Note: The glottal stop final in SM is dropped before a diminutive 仔 or utterance-final 啊 [ɑ] , e.g.:
i. When saying 藥仔 [j ɑ](’medicine’ + diminutive suffix), the glottal stop disappears.
ii. Some people say “我無愛呷啊啊”; ‘I don’t want to eat it’ – the [a] sound in ‘eat’ is lengthened, with
no glottal stop: [guɑ bo ai tɕia],in order to sound less direct.
Start paying attention to the final stops and nasal finals in Southern Min. Do you pronounce a final
[-m] or final [-n] in the family name 林 Lin? Taiwan Southern Min is changing!
* About mistakes we make in English pronunciation:

- Most of the errors we make are systematic. To learn English or any language well, the first thing we
need is awareness of what we are doing, and then an understanding of what the correct targets are.
7
Sep. 24, 2012
2 National Taiwan University / Karen Steffen Chung 史嘉琳
4 National Taiwan University / Karen Steffen Chung 史嘉琳
5-6 National Taiwan University / Karen Steffen Chung 史嘉琳
8
Sep. 26, 2012
Fall 2012
6. Consonants: Place of Articulation (II)

1. Dictation and syllable counting quiz
2. Issues in students’ notes
4. Q & A
II. Notes
1. Dictation and syllable counting quiz
a. Dictation: answers and corrections
(1) lag [læ g] flag [flæ g]: more familiar word
(2) lack [læk]
(3) pest [pɛst]
(4) pack [pæk]
(5) low [loʊ]
(6) wreck/rec [ɹɛk] *reck (not a word): not familiar with the words wreck/rec
(7) past [pæst]
(8) lug [lʌg]
(9) paste [peɪst]
(10) late [leɪt]
b. Syllable counting quiz:

- Count the number of syllables in each word and put a number after the word to indicate how
many syllables it has.
- Ex. played: 1
1
Sep. 26, 2012
Words Answers and IPA transcriptions

(1) employed 3 2 [ɛm'plɔɪd]
(2) pitted 2 ['pɪtəd]
(3) pleased 1 [plizd]
(4) patted 2 ['pæ təd]
(5) roared 1 [ɹɔɹd]
(6) packed 1 [pæ kt]
(7) hissed 1 [hɪst]
(8) lagged 1 [læ gd]
(9) nodded 2 ['nɑdəd]
(10) dabbed 1 ['dæ bd]
*How to count the syllables in a word:

- Count the number of the vowel sounds (not the number of vowel letters); and count the number of
beats you sense in the word
*Pronunciation rules for past tense endings of regular verbs: if the present tense of the verb ends with:
i. /-d / or /-t/ + -ed  (SAE) [-əd]; (BE) [-ɪd] e.g. pat  patted ['pæ təd]
ii. a voiced sound (excluding /d/) + -ed  [-d] e.g. laglagged [læ gd]
iii. a voiceless sound (excluding /t/) + -ed  [-t] e.g. pack  packed [pæ kt]
2. Issues in student’s notes

a. Handwriting issue: [i] written [ī]  force of habit 習慣使然; very typical of Taiwanese writing;
please dot your “I” – don’t write a line above them.
b. What’s the difference between the wedge [ʌ] and the schwa [ə]?
- Phonetically: [ʌ] is an open-mid back unrounded vowel; [ə] is a mid-central vowel (aka the
neutral vowel or reduced vowel). They sound very similar in SAE, but [ʌ] sounds more like [ɑ]
in BE
- Phonologically: [ʌ] can be stressed (e.g. under) or unstressed (e.g. umbrella) while [ə] can
never be stressed (under the particular scheme we’re using for transcribing English)
c. Some terms:
- What is “computational linguistics” in Chinese?
計算語言學 (compute (v.) 計算)
- How do we spell “唇顎裂” in English?
Cleft palate
d. The word of [əv]: as far as Ms. Chung knows, it is the only English word in which “f” is
pronounced as [v]
2
Sep. 26, 2012
e. [ɛ] vs. [æ :]
- Among the most Taiwanese students, this pair is one of the most
difficult.
[ɛ] vs. [æ ] [eɪ]
pet pat pate
tongue higher lower higher than for [ɛ]
jaw higher lower higher than for [ɛ]
duration shorter quite long long (it’s a diphthong)
- Start from the one you are good at: lower your tongue and jaw for [æ ]but do not move the jaw
down for [ɛ]; your mouth should just barely be open.
- If you’re unsure which vowel it is, it’s probably [ɛ], since Taiwanese are usually pretty good at
hearing and producing [æ ,] so you will probably get that one right away without hesitation.
f. How do you write the IPA symbol for the “y” in happy?
- (AE) happy ['hæ pi] not [ɪ]
In AE, every word-final y is pronounced as [i] (as far as we know!)
e.g. happy, city, messy, silly
The same is true of the “-ies” plural ending in GA: it’s pronounced [iz].
- In BE, sometimes word-final y is pronounced as [ɪ], and the plural ending –ies is usually [ɪz]
- The difference between [i] and [ɪ] is not just one of duration but also one of quality.
[i] vs. [ɪ]
tongue higher a bit lower and more back
duration longer shorter
category tense/long vowel lax/short vowel
For more information, please read CET article #3:

3. 大師開講 — 提升聽力祕訣： /i/ 和 /I/ 的辨別 in No. 71, May/June 2012, p. 12-14.

a. Classifying consonants according to their place of articulation
- Two parameters for distinguishing consonants:
(1) Place of articulation 發音位置 (which organs touch other organs)
(2) Manner of articulation 發音方式 (the way in which the articulation is produced and how
close the articulators get to each other)
- Articulators that can obstruct the airstream:
i. lips  labial – [b] [p] [m] [f] [v] ㄅㄆㄇㄈ [ㄪ] (earlier symbol for [v] in 注音符號)
ii. tip/blade of the tongue  coronal – [d] [t] [n] [l] [s] [z] ㄉㄊㄋㄌㄙ (no symbol for [z])
iii. back of the tongue  dorsal – [g] [k] [ŋ] ㄍㄎ [ㄫ] (earlier symbol for [ŋ] in 注音符號)
also includes Mandarin ㄏ [x]
3
Sep. 26, 2012
- The three terms – labial, coronal, and dorsal – are used more often in phonology than in
phonetics to classify sounds. The terms lump too many categories together (e.g. tongue tip,
blade and front are all “coronal”) to be very useful in phonetics.
 We need more specific (less general) terms.
b. Terms for sounds formed with obstructions that involve different parts of the oral tract:
(1) Bilabial 雙唇音: made with the two lips coming together
e.g. buy, pie, my
(2) Labiodental 唇齒音: lower lip and upper front teeth

- As far as researchers know, no language has a labiodental sound articulated with the lower
front teeth and upper lip.
e.g. fie「可惡！」, vie 競爭
(3) Dental 齒音: tip or blade of the tongue and upper front teeth
- For most speakers of Mandarin, ㄉㄊㄋ and ㄌ are dental.
- Most American English speakers protrude their tongues when saying words that have “th”
sounds, voiced or voiceless.
- Some British speakers also protrude their tongues, but some do not. At least one
phonetically-trained Brit questions Ladefoged’s statement that most speakers of British
English put their tongues close behind the upper front teeth when saying words that have
“th” sounds.
- We use interdental 齒間音 to describe sounds in which the tongue tip protrudes from
between the teeth.
e.g. thigh [θaɪ] 大腿, thy [ðaɪ] 【古】你的 – (familiar) ‘you’
(4) Alveolar 齒齦音/舌尖音: tongue tip or blade and the alveolar ridge
e.g. tie, die, nigh【文】接近的, sigh, zeal (n.) 熱心, lie
(5) Retroflex 捲舌音/翹舌音: underside of the tip of the tongue and the back of the alveolar
ridge
Many speakers of English do not use retroflex sounds at all, but some people do, or
sometimes do.
e.g. ire, hour, air
*Postvocalic /r/ in English:

- Postvocalic /r/ is pronounced in Irish English, General American, and some varieties of
British English.
 these are called “rhotic” ['ɹoʊtɪk] dialects. Standard BE is a non-rhotic variety of English
4
Sep. 26, 2012
*Distribution of non-rhotic dialects in the US:

- Non-rhotic accents can be found in many parts of the Eastern US, e.g. New England and
New York City, and in the US South; some are inconsistently non-rhotic.
- Reasons for non-rhoticity:
i. the Eastern US: close relationship with England; perhaps other reasons. But non-rhotic BE
is a variety with high prestige, while non-rhotic varieties of US English tend to have lower
prestige than General American.
ii. the US South: it is said that formerly, many parents sent their children to England; they
subsequently brought their pronunciation habits back with them, and they spread.
(6) Palato-alveolar 顎齦音: blade of the tongue and the back of the alveolar ridge
e.g. shipshape, genre, child, join
cf. alveolo-palatal 齦顎音
e.g. ㄐㄑㄒ in Mandarin and some sounds in Polish.
(7) Palatal 顎音: front of the tongue and hard palate
e.g. you ([j] is the only palatal sound in English)
(8) Velar 軟顎音/舌後音/舌根音: back of the tongue and the soft palate (i.e. velum)
e.g. hack, hag, hang
4. Q & A
a. What is the difference between [ɑ] and [a]?
- The segment [ɑ] exists in English as a monothong (單母音/單元音); the tongue has a low
back position as in the word father.
- The segment [a] exists in English as the first segment in two diphthongs 雙母音/複元音, as
in the word high [aɪ] and the word how [aʊ], but not as an independent monophthong. For [a],
the highest point of the tongue is in the center of the oral cavity, a little bit higher than it is for
[ɑ]. It may be influenced by [ɪ], originally a mid-high front vowel, and [ʊ], a mid-high back
vowel – and [a] is a low front vowel.
*Diphthongs in English
- my [aɪ], may [eɪ], mouse [aʊ], most [oʊ], moist [ɔɪ], muse [ju]
cf. Mandarin ㄞ[aɪ] ㄟ [eɪ] ㄠ[aʊ] ㄡ [oʊ]
b. What is the difference in pronunciation between raised and raced?

- raised [ɹeɪːzd]: the “s” is voiced, so the diphthong [eɪ] is longer, and the final “d” is voiced.
- raced [ɹeɪst]: the “c” is a voiceless [s] sound, so the diphthong [eɪ] is shorter, and the final “d”
is a voiceless [t].
5
Sep. 26, 2012
*Why is the vowel in the two words of different lengths?

- The vowel is affected by the following sound: if it is voiced, the preceding vowel will be
longer; if the consonant is voiceless, the preceding vowel will be shorter. This may be universal
or nearly universal in the world’s languages.
c. Why is the transcription [tōn] for the word tone circled as “wrong” in my class notes even though
that’s the pronunciation given in the dictionary?
- Most US-published dictionaries use ad hoc 將就、隨便湊 pronunciation symbols, and the
symbols adopted generally differ from edition to edition.
e.g. In elementary schools in the US, students learn to distinguish “long vowels” and “short
vowels” using the diacritical marks, e.g. the “short ‘o’” in not is marked with a breve,
meaning ‘short’: nŏt; and the “long ‘o’” in note is marked with a macron, indicating a long
vowel: nŏte.
- Avoid using US-published dictionaries if you want to check the written transcription of the
pronunciation of a word (MW is good for definitions, etymologies and sound files though).
You can use the Cambridge Learner ’s Dictionary Online to check both US and BE
pronunciations of English words in IPA at http://dictionary.cambridge.org/

consonant ['kɒnsənənt](BE)  ['kɑnsənənt] (AE; stick to one variety)
articulators [ɑɹ'tɪkəleɪɾɚ]  [ɑɹ'tɪkjuleɪɾɚ] (vowel: note the [j] glide)
cause [koʊz]  [kɔz] (vowel)
lips [lips]  [lɪps] (vowel)
tongue tip tongue *tip  *tongue tip (compound noun)
blade [blɛd]  [bleɪd] (vowel) cf. bled [blɛd]
labial articulations labial articu*lations  *labial articulations (articulations was repeated)
using (v.) ['jusɪŋ]  ['juzɪŋ] (consonant) vs. use [s] (n.)
similar: house (v.) [z] vs. (n.) [s]
teethe (v.) [ð ] vs. teeth (n.) [θ]
coronal co'ronal  'coronal (word stress)
not need ['nɑ_'nid] (no stop)  ['nɑʔ 'nid] (the word final [t] becomes a glottal
stop [ʔ] when followed by a consonant-initiated
word)
described [dɪs'kɹaɪbd]  [dɪs'kɹaɪːbd] (longer vowel before a voiced
consonant)
feeling ['fɪlɪŋ]  ['filɪŋ] (vowel) cf. filling ['fɪlɪŋ] 補牙的銀粉
these terms *these terms,  these *terms, (tonic stress on the last stressed
syllable of an utterance)
these terms, however, no rise  rise to signal continuation (continuation rise)
6
Sep. 26, 2012
articulatory_gestures extra pause in between  don’t pause; these two words are a unit, and the
final sound of articulatory is a vowel, not a stop)
sufficient detail [sʌ'fɪʃɪən_'diteɪl] (no stop)  [sʌ'fɪʃɪənt 'diteɪl] ([t]; homorganic with /d/, so
/t/ is more likely than a glottal stop here)
arrow ['æɹoʊ]  ['ɛɹou] (in Midwestern US English)
figure ['fɪgɚ]  ['fɪgjɚ] (the [j] glide was missing)
in [in]  [ɪn] (vowel)
Jamie ['dʒɛmi]  ['dʒeɪmi] (vowel)
bilabial [baɪ'læbiəl]  [baɪ'leɪbiəl] (vowel) similar: labial
these; these [ðɪz]; [ðɪs]  [ðiːz] (vowel and consonant)
words [wɜ˞z]  [wɜ˞dz] (consonant; [dz] should sound like 字)
comparable [kəm'pæɹəbəl]  ['kɑmpəɹəbəl] (more common in Midwestern US
English)
American [ə'mɛɹɪkə̃_]  [ə'mɛɹɪkən] (the [n] was missing leaving only a
bare nasalized [ə])
tongue [toʊŋ]  [tʌŋ] (vowel)
called [koʊld]  [kɔld] (vowel)
interdental 'interdental  inter'dental (word stress)
7
Sep. 26, 2012
Pag Work Licensing Author/Source
e

3 http://homepag Caves Educational Training Co., Ltd / Hello! E.T. Issue No. 71,
May/June 2012, p. 12-14.
e.ntu.edu.tw/~
karchung/pubs/ E.T.
71_hello_et.pd This work is licensed by Caves Educational Training Co., Ltd
f for the use of OCW ONLY. The copyright belongs to the above
mentioned creator(s).
8
Oct. 1, 2012
Fall 2012
7. Consonants: Manner of Articulation
2. Homework
II. Notes
a. Quick review: Places of articulation for English (p.11-13)
- bilabial 雙唇; labiodental 唇齒; interdental 齒間 (dental 齒音); alveolar 齒齦/舌尖; retroflex
捲舌(翹舌); palato-alveolar 顎齦; palatal 硬顎; velar 軟顎/舌根/舌後
*Sounds we won’t be discussing:

-marginal sounds in English: this includes extralinguistic sounds that occur in particular
situations and are not part of the phonetic/phonological system of English, for example,
coughs, sneezes, clearing your throat, which are not phonemes in any language.
Some sounds have only special uses in one language but are ordinary phonemes in another,
e.g. in English, a lateral click means “giddy up”, used to tell a horse to get moving; however,
in some languages – all in southern Africa – it is an ordinary phoneme)
b. The oro-nasal process

- Velic closure 軟顎閉塞狀態: The velum (aka the soft palate) is lowered to produce nasals
(in English: rang, ran, ram); the air comes out through the nose while being prevented from
going out through the mouth.
 nasal consonant: when the velum is lowered and an obstruction is formed in the mouth
- Another possibility: incomplete velic closure
e.g. When articulating a nasalized vowel 鼻化母音, the velum is partly lowered, while air is
not at the same blocked from going through the oral tract  air can come out through both
the oral and nasal tracts
*nasalization 鼻化
- in English and Mandarin: not a phonemic feature, i.e. not included in the IPA chart of the
language and does not change the meaning of words, e.g. if you pronounce “say” as [sẽɪ]̃
instead of [seɪ], it’s a bit odd, but it will still be understood as “say”. But:
1
Oct. 1, 2012
- in Southern Min, vowel nasalization is phonemic, i.e., it produces different words with
different meanings,
e.g. 鞋仔 ‘shoes’ [e-a] vs. 嬰仔 [ẽ-a] ‘baby’
- The oro-nasal process: controlled by raising or lowering the velum

 the distinguishing factor between oral and nasal sounds
c. Classifying consonants according to their manner of articulation

- Manner vs. place: manner refers to the way in which a sound is produced, and how close the
articulators are to each other
- The articulators may:
i. close off the oral tract for a very short time or for a relatively long period
ii. narrow the space in the oral tract considerably
iii. modify the shape of the oral tract by approaching each other
- Terms for sounds formed in different manners
(1) Stop 塞音: two different definitions are current in phonetics
i. a complete closure which prevents the air from escaping through the mouth
ii. a complete stoppage of the airflow in both the nose and the mouth
Two types: (from oro-nasal process to articulatory process)
oral stop (aka stop) nasal stop (aka nasal)
velum raised velum lowered
articulators in the oral tract approach articulators in the oral tract approach
and make contact and make contact
no air escapes till release of stop air escapes through the nose
throughout the sound
pressure builds up in the mouth no pressure builds up
when the articulators come apart, the when the articulators come apart, there
compressed air is released in a small is no burst of sound
burst of sound through the mouth
there are changes in sound quality no quality change throughout the
(beginning, middle, end)  stop sound  continuant
e.g. pie, buy, tie, dye, key, guy e.g. my, nigh, sang
(bilabial/alveolar/velar contact) (bilabial/alveolar/velar contact)
*A few things to pay attention to:

- You’ll see the word plosives in IPA tables. Plosives 爆音 are the ordinary kind of stops
we’ve been discussing so far – they have an egressive airstream mechanism, which means
the air leaves the lungs in an outward direction. There are also stops in which the air goes
inward toward the lungs – we’ll talk about these next semester.
2
Oct. 1, 2012
- In English, the velar nasal (i.e. [ŋ] in sang) normally does not appear at the beginning of
a syllable; you can, however, find it between words, e.g. in sing a song [sɪŋ ə sɔŋ] due to
linking, and in the Vietnamese family name Nguyen [ŋʷiə̌ŋ] 阮, though it’s usually
pronounced something [nə'wɪn] in English.
*A comparison of bilabial stops in English and Mandarin

English [b]: voiced (when following a [pʰ]: voiceless, aspirated
voiced sound), unaspirated e.g. pie (unaspirated in spy [p])
e.g. abide (voiceless in bye [b̥])
(diacritic: devoicing)
Mandarin [p]: voiceless, unaspirated (ㄅ) [pʰ]: voiceless, more strongly

e.g. 拜 [paɪ] aspirated than in English
e.g. 派 [pʰaɪ]
*Compressed air in the articulation of stops

- The holding in of compressed air behind an obstruction can be viewed as a key defining
feature of stops; under this condition, nasals do not qualify as “stops”
*Voicing of stops
-In English, voiced stops tend to be devoiced when at the beginning of an utterance or of a
word; they are voiced when they follow another voiced sound.
e.g. *bout (一陣、一回合), guy, day vs. about, a guy, a day
-In Mandarin, all stops (ㄅㄆㄉㄊㄍㄎ) are voiceless. Stops that share the same place and
manner of articulation differ only in the feature of aspiration.
-In Southern Min, there are two voiced stops, /b/ and /g/ (but not /d/), e.g. /b/ as in 醜
[bai], though it is not exactly the same sound as [baɪ] in English
*Minimal pair
-a pair of words that differ only in one sound 最小對比的一對詞; if there are more than
two such words, it is called a minimal set.
e.g. English [paɪ] (voiceless) vs. [baɪ] (voiced)
*Nasal-sounding voice vs. nasal

-Often your voice sounds nasal when you have a cold. Why?
When people have a stuffed nose, the spongy tissues in the nose are filled with mucus
鼻涕 and swell 變得更飽滿, which produces different resonances in the nasal cavity, and
this changes the quality of speech sounds. Some people may say you’re speaking in a
nasal voice, but you’re actually using a kind of anasal voice 非鼻音, if you’re so
congested you can’t breathe normally through your nose.
3
Oct. 1, 2012
(2) Fricative 擦音: close approximation of two articulators  airstream partially obstructed
 audible turbulent airflow
e.g. fie, vie (labiodental), thigh, thy (indertental), sigh, zoo (alveolar), shy (palate-alveolar)
- Sibilants 嘶音: produced by a narrowing between the blade of the tongue and the (back
part of the) alveolar ridge  higher-pitched sounds with a more obvious hiss 嘶聲/咝音
e.g. /s, z, ʃ, ʒ, tʃ, dʒ/. (/tʃ, dʒ/ are classified as affricates; see (5) below)
(3) Approximant 接近音: one articulator is close to another, leaving space for the air to pass
through without causing a turbulent airflow
e.g. yacht (palatal), we (bilabial and velar), raw (alveolar)
(4) Lateral 邊音: obstruction of the airstream by the tongue tip at a point along the center of
the oral tract, with incomplete closure between one or both sides of the tongue and the roof
of the mouth; the air escapes through one or both sides of the tongue
e.g. lie
*Do we always use both sides of the tongue when producing a lateral sound?
-Not necessarily, because the human body is not completely symmetrical.
(5) Additional Consonant Gestures

i. Trill (or “rolled r”) 顫音: the tongue is in free vibration, striking the alveolar ridge at
least three times
e.g. rye, raw (in some forms of Scottish English)
ii. Tap (or “flap”) 閃音: the tongue makes a single tap against the alveolar ridge
e.g. pity, city, butter Although this sound is often called a “flap” in other texts, and
especially in ESL materials, we will consistently call it a “tap”, because we will need the
term “flap” next semester for a quite different kind of sound.
iii. Affricate 塞擦音: combination of a stop immediately followed by a fricative; the
tongue tip or blade and alveolar ridge come together for the stop and separate only
slightly
 a fricative is made at approximately the same place of articulation
e.g. church, judge
*The affricate [tʃ] in phonetics and in phonology

- in phonetics, the affricate [tʃ] (as in church) is analyzed as two separate sounds, i.e. it is
a composite sound 複合音  often excluded in many IPA consonant charts
- in phonology, [tʃ] is structurally one sound because it often represents a single
phoneme, and is often represented with a single symbol with a diacritic: /č/ ( ˇ is a
diacritic called a “háček”; this word and the symbol are from Czech; /č/ is read as
“c-“háček”)
4
Oct. 1, 2012
iv. Glottal stop 喉塞音: a complete stoppage of the airflow in the glottis 聲門 (the
cavity between the two vocal folds); the tip of the tongue does not touch the alveolar
ridge, or anywhere else.
Distribution: at the beginning of words that start with As an allophone of /t/

a vowel
e.g. eek, oak, ark (easy to overlook) button, important
flee east (sometimes obvious) it may, hit me
 See more about the production of glottal stop in handout for unit 5, Sept. 24.
 For more details about the terms phoneme and allophone, please read this page of the
course website: http://homepage.ntu.edu.tw/~karchung/intro%20page%2014.htm
2. Homework:
a. Make and print out waveforms with WASP:
http://homepage.ntu.edu.tw/~karchung/intro%20page%2016.htm ($)
Record: (1) "My two boys know how to fish." and (2) "Tom saw nine wasps."
Use the audio recordings together with the waveforms to check the sentences in the book.
b. Start working on the ch. 1 exercises, p. 25-32.
c. Read webpages 11. Romanization I and 12. Romanization II


Longman ['lɔŋmæn]  ['lɔŋmən] (reduced vowel)
consider [kɑn'sidɚ]  [kən'sɪdɚ] (reduced vowel; vowel)
When you say these When you say these  When you say these (pronouns
function wordsno stress)
sequence ['sik_əns]  ['sikwəns] (the [w] was missing)
velum ['vɪləm]  ['viləm] (vowel) similar: velic [i]
soft [sɒft]  [sɔft] (vowel)
consonant ['kɑŋsənənt]  ['kɑnsənənt] (in Midwestern US English; [kɑn]
sounds like 咖+恩 not like 康 [kɑŋ])
Raising or lowering the no pause  pause after the subject
velum_controls…
oral ['oʊ-ɹəl]  ['ɔɹɹəl] (AE: r-coloring)
articulation [ɑɹtiku'leɪʃən]  [ɑɹtikju'leɪʃən] (the [j] was not clear)
considerably [kɑŋ'sɪdɚɹəbli]  [kən'sɪdɚɹəbli] (reduced vowel; alveolar nasal)
5
Oct. 1, 2012
they [leɪ]  [ðeɪ] (stick out the tongue)

each other each *other  each other (pronouns  no stress)
other ['ɑðɚ]  ['ʌðɚ] (vowel)
modify ['mɒdɪfaɪ]  ['mɑdɪfaɪ] (vowel)
of [of]  [əv] (vowel and consonant)
(of may be the only word in modern English in
which the /f/ is pronounced as a [v])
stop [stɒp] (BE)  [stɑp] (AE; stick to one variety)
articulatory [ɑɹ'tikjulætɔɹi]  [ɑɹ'tikjulətɔɹi] (reduced vowel)
raised [ɹeɪsd]  [ɹeɪːzd] (“s” is voiced, vowel lengthening);
contrast with race [ɹeɪs]
blocked off,… flat  blocked *off,…(tonic stress; continuation rise)
completely [kəm'plɪtli] [kəm'plitli] (vowel)
…will build up_and no pause  pause before conjunctions and punctuation
kind [kãɪ̃d]  [kaɪnd] (tongue tip touches the alveolar ridge)
similar: sound
alveolar [ɛl'vɪlɚ]  [ɛl'vilɚ] (vowel) similar: velar [i]
these [ðɪs]  [ðiz] (vowel; consonant)
sounds [sãʊ̃z]  [saʊndz] (consonant; [dz] sounds like ‘字’)
 similar: words
oral cavity oral *cavity  *oral cavity (adjective ending in ‘-al’ + noun
 often pronounced as a compound noun)
cavity ['keɪvəɾi]  ['kæ vəɾi] (vowel)
the one [ði wʌn]  [ðə wʌn] (one [wʌn] does not begin with a vowel)
close [kloʊz] (v.)  [kloʊs] (the /s/ is voiceless in the adjective form)
illustrate [ɪ'lʌstɹeɪt]  ['ɪləstɹeɪt] (first syllable is stressed)
alveolar ridge al*veolar ridge  alveolar *ridge (phrasal stress)
called [kold]  [kɔld] (vowel)
fricative ['fɹɪkəɾɪv]  ['fɹɪkətɪv] (the /t/ in this word is not a tap)
without the [wɪ'θaʊ_ðə] (no stop)  [wɪ'θaʊʔ ðə] (glottal stop)
the [də]  [ðə] (stick out the tongue)
consonants ['kɑnsənənt_]  ['kɑnsənənts] (the final plural [s] was missing)
region ['ɹiʒən]  ['ɹidʒən] (consonant)
incomplete [inkʌm'plit]  [ɪnkʌm'plit] (vowel)
status [stæɾəs]  both [stæɾəs] and [steɪɾəs] are used
tongue [toʊŋ]  [tʌŋ] (vowel)
that matter [ð æ_'mæɾɚ] (no stop)  [ð æʔ 'mæɾɚ] (glottal stop)
in_English in/English (no linking)  link the final [n] in in to the initial [ɪ] in English
roll and tap no pause  pause before conjunctions
6
Oct. 1, 2012
the alveolar ridge,… flat  the alveolar *ridge,… (tonic stress; continuation
rise)
pity ['piti]  ['pɪɾi] (vowel, and /t/ becomes a tap)
the production of some the production of  the production of *some sounds (when some
sounds some *sounds  means ‘certain’, which implies there is a
contrast, it is stressed)
similar: more than one of the manners
stop closure stop *closure  *stop closure (compound noun)
combination [kɑmbə'nɛʃən]  [kɑmbə'neɪʃən] (vowel)
affricate ['æ fɹɪkeɪt]  ['æ fɹɪkət] (vowel; -ate: nouns and adjectives:
[-ət]; verbs: [-eɪt])
for the stop and then… no pause and flat  pause before conjunctions and after the
transitional words; continuation rise
separate (a.) ['sɛpəɹeɪt]  ['sɛpɹət] (the schwa [ə] in the middle is omitted
when the word is an adjective)
eek [eɪk]  [ik] (vowel)
, etc. (=et cetera) [ɛk 'sɛtʃəɹə]  [ɛt 'sɛɾəɹə] (there should be no [k], should be [t];
tap; even very educated native speakers often
make a similar mistake and say [ɛk 'sɛɾəɹə])
glottal ['glɒtəl]  ['glɑɾəl] (vowel; tap)
beginning [bɪ'gɪnɪn]  [bi'gɪnɪŋ] (vowel; nasal)
*Stress in phrasal verbs think through

-In phrasal verbs, the tonic stress is on the final particle,
e.g. take *off, dream *up, come *on, think *through
-If it is just a verb followed by an ordinary preposition, the preposition is not stressed, e.g.:
e.g. The books were put on the *desks. The bird flew through the *air.
*tonic stress: the final stress of the utterance (or thought group) is very high and signals the end of the
utterance. (See Unit 3 handout)
*Pronunciation of words that end in -ate

Function as a verb noun or adjective
-ait [-eɪt] [-ət]
e.g. separate ['sɛpəɹeɪt] ['sɛpɹət]
affricate -- ['æ fɹɪkət]
graduate ['gɹæ dʒueɪt] ['gɹæ dʒuət]
*r-coloring in American English

-In General American English, vowels before /r/ generally have r-coloring.
e.g. hero [hɪɹ-ɹoʊ], zero [zɪɹ-ɹoʊ], oral [ɔɹ-ɹəl]
7
Oct. 1, 2012
-In some pronunciation styles, not all V + /r/ words have r-coloring.
e.g. ICRT DJ Mike Woodward pronounces hero as [hi-ɹoʊ] and zero as [zi-ɹoʊ].
8
Oct. 1, 2012

~karchung/int Attribution-NonCommercial-ShareAlike 3.0 Taiwan.
ro%20page%2
014.htm
~karchung/int
ro%20page%2
016.htm
~karchung/int
ro%20page%2
011.htm
~karchung/int
ro%20page%2
012.htm
9
Oct. 1, 2012

10
Oct. 3, 2012
Fall 2012
8. Waveforms of Consonants
1. Issues found in students’ notes and dictations
2. Quiz on compound stress
4. Homework
II. Notes
1. Issues found in students’ notes and dictations
a. Some people have trouble distinguishing [e], [eɪ], [ɛ], and [æ ]
(1) bet [bɛt] is often mispronounced as either [bæ t] or [beɪt]
 to correct, raise your jaw and barely open your mouth; put your hand under your jaw and push
up to create a tactile memory to remind yourself of the correct place of articulation for this vowel
(2) bat [bæ t]: Taiwan students usually have no problem with this vowel
 lower your jaw a lot
(3) bait [beɪt] often mispronounced as [bæ t]or [bɛt]
 spread your lips; pronounce [eɪ] just like how you pronounce 注音ㄟㄧ, emphasizing the
second part of ㄟㄧ, the 「ㄧ」
(4) The phoneme [e] does not exist in English as a monophthong! In English, it appears as part of
the diphthong [eɪ]. In a similar way, the monophthong [o] is also a component of the diphthong
[oʊ]. [e] and [o] are used in KK symbols to represent [eɪ] and [oʊ], which can be misleading for
Taiwan students, because these sounds are diphthongs in General American. (The sounds are
however more monophthongal in some dialects of US English.)
b. What fonts should you use for your notes?

(1) Times New Roman 12 pt – for text (for titles, it’s OK to use Arial)
(2) Lucida Sans Unicode 10.5 pt – for IPA symbols (Lucida Sans Unicode is a large font so it
needs to be reduced in size to more closely match the size of New Times Roman 12 pt)
c. Be more complete in writing your notes! Although there is no need to make your notes into an
essay, still please do not be too telegraphic. Your notes should be understandable to another
reader.
1
Oct. 3, 2012
d. How do you say 美國南部 in English?

- the southern part of the U.S. OR: “the U.S. South”: 美國南部 (NOT South America 南美洲)
e. How do you say 沒有重音 in English?

- unstress 不放重音 or destress 拿掉重音. (NOT distress 使沮喪)
f. How do you write the abbreviation of versus?

- vs. – add a period after the s (NOT v.s. or v.s) The period just means “this is an abbreviation”; it
doesn’t mark the exact place where letters have been omitted.
g. Where do you leave spaces when using parentheses?

(1) Space before and after the parentheses, i.e. outside of the parentheses, e.g. Put_(the spaces)_in
these positions.
(2) No space after the right parenthesis just before punctuation e.g. Omit it (before a period).
(3) Leave NO spaces inside the parentheses, i.e. don’t do this: Don’t leave ( spaces ) inside of
parentheses.
h. Easily confused terms:

- epiglottis 會厭 vs. oesophagus (British spelling) or esophagus (American spelling) 食道
2. Quiz on compound stress

a. Instructions:
(1) Circle the stressed syllable(s) of the compound or phrasal noun and indicate the tonic stress of
the item by adding an asterisk * in the upper left-hand corner of the last stressed syllable of the
item.
e.g. con*venience store or con*venience store (in this case, the n is both part of the second
syllable and part of the third syllable, so both are OK; this is called “ambisyllabicity”)
(2) Rule for compound noun stress: When two nouns are put together, the first of which modifies
the second, the modified noun (on the right) receives no stress.
If there are more than two nouns, find the two with the closest relationship to each other, apply
the above rule, then treat the result as a single word. Then apply the rule to that result and the
remaining noun(s).
The basic rules for compound noun stress can be found in CET article #5:
Links to all of the CET articles can be found here:

http://homepage.ntu.edu.tw/~karchung/Karen/Karen_Chung_publications.htm#CET
2
Oct. 3, 2012
b. Answers, corrections, and explanations:
Students’ answers Class Corrections or explanations

1
(1) *urban area compound urban (or urban) *area (or *area)
(2) stringed *instrument phrase2 stringed: past participle 過去分詞  adjective
cf. gerund 動名詞  noun
(3) *paper clip compound *paper (or *paper) clip
(4) *can opener compound
(5) serious *error phrase serious *error (or *error/ *error)
(6) *beach towel compound
(7) *shadow boxing compound or *shadow boxing (more natural)
(8) silly *joke phrase silly *joke (monosyllabic word circle the whole word)
(9) *restaurant manager compound or *restaurant (more natural) / *restaurant manager
(10) sticky *fingers phrase or sticky *fingers (ck  digraph  do not separate)
*Footnotes: compound (compound noun): noun + noun 2phrase: adjective + noun
1
c. More about compound nouns and syllables:

- Sub-rules for compound noun stress: (1) When the modifying noun is the main material or
ingredient of the latter, both nouns are stressed, e.g brick *wall, glass *door.
- When the modifying noun is an (2) a time, (3) a place, or (4) some kind of organization, then very
often both nouns are stressed. Note that you will not find these rules in many textbooks, and
also that there are many exceptions to them. Pay attention when people are speaking to pick up
the correct stress of items like these. Examples: (2) Monday *evening; (3) kitchen *sink; (4)
world *leader (world can also be interpreted as a “place”), family *meeting.
- Ambisyllabicity: the ability of a sound to be part of both the previous and the following syllable
(ambi- means ‘both’), e.g. the “l” in melon

a. Glottal stop [ʔ]: a closing (and often an opening immediately after) of the glottis
(glottis 聲門: a space between the vocal folds)
e.g. hit Robert ['hɪʔ 'ɹɑbɚt] (the vocal folds are held together tightly, temporarily interrupting the
airstream)
b. Address these five parameters when describing a consonant:

(1) State of the vocal folds: whether the vocal folds are vibrating or not, i.e. whether the consonant
is voiced or voiceless 有聲/無聲 or 帶音/不帶音
(2) Place of articulation: bilabial, interdental, alveolar, palatal 舌面音 (e.g. [j]), velar 舌後音, uvular
小舌音, retroflex 捲舌音, palato-alveolar 顎齦音 (cf. alveolo-palatal 齦顎音: Mandarin ㄐㄑ
ㄒ)
(3) Central 中央音 or lateral 邊音 articulation: [l] – the only lateral consonant in English
3
Oct. 3, 2012
(4) Soft palate: raised  velic closure  oral sound; lowered  no velic closure  nasal sound
(5) Manner of articulation: stop 塞音, fricative 擦音, affricate 塞擦音, approximant 接近音
For (3) and (4), central and oral are the default  usually we don’t need to mention them
e.g. [s] is a voiceless alveolar (central) nasal (stop)
*Discussion in Facebook group “NTU Phonetics”

- Please go ahead and post anything interesting
- Rich environment – there are lots of former phonetics students and high profile linguists and
phoneticians in the group
- Topics brought up by students:
i. Video: The woman who says ‘biscuit’ 900 times an hour: Listen to an interview with a Tourette’s
sufferer (submitted by Jamie Yu)
ii. Nasalization in Mandarin: Some Taiwan Mandarin speakers, especially from the South, have a
nasalized vowel when saying 怕 [pʰɑ̃] and 把 [pɑ̃]
 Excellent observation. There’s a reason for everything, and we should try to collect relevant
data and find out more about phonetic variations like these.
 In English, there’s something similar, especially in British English. American English is a
rather nasal dialect of English, and many “extra” nasalized sounds are present (at least true for
many Americans). British English, on the other hand, is generally “drier”, or less nasal, but if
you listen to the BBC, you will notice that the announcers may use nasalization to emphasize
certain points or to sound more authoritative.
 Start listening to BBC and you can collect data like this.
*Clarifying: [ɔ] before –l and -r

- In American English, the [ɔ], although generally described as a monophthong (and it is indeed a
monophthong in standard British English), is actually diphthongal in nature. Its pronunciation is
similar to Southern Min 芋仔 contracted into a single syllable. e.g. law, draw
- Before -l and -r, the second part of the “diphthong” is cut off. e.g. or [ɔɹ], oral ['ɔɹəl]
c. The waveforms of consonants:

- One thing that can’t be seen from a waveform: the place of articulation of a sound
- What is usually visible in waveforms:
(1) the principal manners of articulation – stops, nasals, fricatives, and approximants
(2) whether a sound is voiced or voiceless
4
Oct. 3, 2012
- Waveform for illustration:
My two boys know how to fish.

[maɪ tu bɔɪz noʊ haʊ tə fɪʃ]
amplitude
time
[m] voiced nasal produced behind the closed lips  smaller amplitude (muffled and less loud)
[aɪ] voiced diphthong vowel  larger amplitude (louder); longer duration (longer)
[t] voiceless stop
(1) flat line (a silence or pause before the stop when holding and compressing air in the oral
tract)
(2) a burst (the articulators open and forcefully release the compressed air in the mouth, which
explains why oral stop consonants are called “plosives” in the IPA chart)
(3) friction (air released through the mouth)
[b] voiced stop  small voicing vibrations from the vocal folds; these have a small amplitude;
they are produced in the larynx and with closed lips and are thus less loud
[z] voiced fricative: it has both evenly spaced lines due to voicing, and also irregular lines,
due to the noise made by the turbulent air
[n] voiced nasal produced behind the alveolar ridge  it has evenly spaced lines and a smaller
amplitude than vowels (it is muffled and thus less loud)
[h] short aspiration  with hardly any voiceless interval
to: first element of an infinitive 不定詞的第一個成分  it carries little information value
(1) [t] is pronounced with hardly any closure and has a very short duration
(2) [ə] is a reduced vowel (= a schwa)  only a few vocal fold pulses
[f] voiceless fricative  smaller amplitude
[ʃ] voiceless fricative  smaller amplitude (larger however than [f])
*Vocabulary: to flatline (v.)

- in a medical context, “to flatline” can mean somebody is dead (= no more brain waves display
on an encephalogram)
*Voiced stops
- voiceless when word-initial with no preceding voiced sounds
- voiced when preceded by a voiced sound (the [u] in two before boy, in this case)
*the [h] sound

- has no specified place of articulation; its place of articulation depends on the vowel that
5
Oct. 3, 2012
follows it
- not even included in some consonant charts
- not a fricative; it is simply aspiration or breathing air out
d. The articulation of vowel sounds

- Main difference between vowels and consonants: the articulators do not come as close together
for vowels  there is less obstruction of the airstream in the oral tract
- Description of vowels: vowel sounds can be roughly described in terms of

(1) the position of the highest point of the tongue: it would be difficult to use an articulatory
description for vowel sounds because it is hard to know exactly what our tongue is doing in
what part of our mouth  we can mark the part of the tongue that sticks up the highest;
however, actual measurements are done acoustically and auditorily, not articulatorily!
(2) the position of the lips
- The tongue and the lips are in continuous motion throughout the vowels in saying words like the
following:
(1) heed [i] (2) hid [ɪ] (3) head [ɛ] (4) had [æ ](5) father [ɑ] (6) good [ʊ] (7) food [u].
*Homework: Say the vowels and feel where your tongue is (i.e. what part of the vocal tract it is
close to), and what it does to form each of these vowels. Train yourself in feeling it (this is
called proprioception); it is harder to realize what your tongue is doing when producing vowels
than it is when producing consonants.
- The positions in the figure below are best considered as the targets of the gestures of the vowels
 a “target” indicates where our tongue should theoretically reach, but we don’t necessarily
attain the target every time we say the vowel (vowels are easily affected by neighboring sounds)
In all these gestures:

- the tongue tip is down behind the
lower front teeth;
- the body of the tongue is domed
upward.
6
Oct. 3, 2012
- Front and back English vowels: 前母音與後母音

(1) Front vowels (from high to low): [i, ɪ, ɛ, æ]
(2) Back vowels (from high to low): [u, ʊ, ɑ]
Although both are classified as high vowels, [u] is not as high as [i] because there is less space
vertically at the back of our oral cavity, which is close to the jaw hinge, than in the front.
- The International Phonetic Alphabet vowel chart for English: (circled: sounds discussed here)
4. Homework
a. Practice feeling where your tongue is when saying the vowels in heed [i], hid [ɪ], head [ɛ], had [æ ,]
father [ɑ], good [ʊ], food [u].
b. Finish exercises for chapter one (will be discussed and marked on Monday, Oct. 8).
c. Do the Hanyu Pinyin tutorials.
d. Make and print out WASP waveforms of

- "My two boys know how to fish."
- "Tom saw nine wasps."
e. Make sure you are up to date in your reading.
f. Hand in class notes by Monday, Oct. 8.
7
Oct. 3, 2012

oral ['oʊɹəl]  ['ɔɹəl] (vowel, with “r-coloring”)
articulatory [ɑɹ'tɪkjulətɛɹi] [ɑɹ'tɪkjulətɔɹi] (vowel)
five [faɪf] [faɪːv] (lengthening of vowel before voiced
consonant; voicing of [v])
at_the … rushing through the stop at stops (at [ʔ] the/ not [ʔ] necessary/ not [ʔ]
not_necessary words, failure to stop lateral)
not_lateral at stops
left out, flat left *out (tonic stress and continuation rise
before punctuation)
at_this stage [æ ðɪs] Stop at stops (at [t] this)
not [nʌt] [nɑt] (vowel)
the acoustic [ðə] [ði] (before vowels)
noting ['nɔtɪŋ] ['noʊtɪŋ] (vowel: diphthong)
articulation [ɑɹtɪku'leɪʃən] [ɑɹtɪkju'leɪʃə] (vowel: needs a [j] glide)
manners of articulation… flat *manners of articulation (to show contrast with
places of articulation mentioned before)
apparent [ə'pæɹənt] [ə'pɛɹənt] (system: Midwestern variety)
bottom ['bɒtəm] (BE) ['bɑɾəm] (system: AE) 底部
cf. button ['bʌtʔn̝] 鈕扣
interval [ɪn'tɚvəl] ['ɪntɚvəl] (word stress)
…of the airstream is flat continuation rise (at the end of the long string of
subject)
the x-ray_movie pause no pause between x-ray and movie (a noun
phrase)
8
Oct. 3, 2012
Page Work Licensin Author / Source

g
2 http://homepag Caves Educational Training Co., Ltd / Hello! E.T. Issue No. 73,
e.ntu.edu.tw/~k September/October 2012, p. 12-14
archung/pubs/7 http://www.cet-taiwan.com/drcet/type.asp?title_type=Hello%20
3_hello_et.pdf E.T.
This work is licensed by Caves Educational Training Co., Ltd
for the use of OCW ONLY. The copyright belongs to the above
mentioned creator(s).
2 http://homepag National Taiwan University/Karen Steffen Chung 史嘉琳
e.ntu.edu.tw/~k This work is licensed by Creative Commons
archung/Karen/
Karen_Chung_ Attribution-NonCommercial-ShareAlike 3.0 Taiwan
publications.ht
m#CET
7 WIKIMEDIA COMMONS
(http://commons.wikimedia.org/wiki/File:English_vowel_chart.
png) 2013.1.12 visited.

9
Oct. 8, 2012
Fall 2012
9. The Articulation of Vowel Sounds

1. Discussion of the quiz on compound noun stress, and class notes
2. Hanyu Pinyin
4. Discussion of questions raised during break
5. Homework
II. Notes
1. Discussion of the quiz on compound noun stress, and class notes
a. The quiz on compound noun stress:
- When the item is a monosyllabic (單音節) word and stressed, you have to circle the whole word
e.g. (X) joke (O) joke
- Remember to put an asterisk before the syllable that receives the tonic stress, in the upper left-hand
corner e.g. (X) silly joke* (O) silly *joke
- Make sure that you identify compound noun stress correctly (there is only one stress in compounds!).
b. Class notes handed in last week:

- IPA symbols:
(1) Glottal stop: (X) [ʡ] [ʢ] [ʔ] (O) [ʔ]
- Please do not put a line under the glottal stop, or you may confuse people who are reading it.
- The hook bends to the right.
- [ʡ] (voiced epiglottal fricative); [ʢ] (epiglottal plosive)
(2) [ʔ] vs. [ɾ]: both can be allophones of /t/, but they have different pronunciations, and occur in
different environments
- [ʔ] glottal stop: it is not only an allophone of /t/; it also occurs in other environments in English
i. as the initial of an utterance or a word that starts with a vowel
e.g. _Is he coming? [ʔɪz hi kʌmɪŋ]; _elephant ['ʔɛləfənt] (cf. Mandarin 意義 [ʔiʔi])
ii. between a syllable-final /t/ and a following syllable that begins with a consonant
e.g. hit_me [hɪʔ mi]; partly ['pɑɹʔli]
iii. with nasal plosion (will be taught later in the semester) e.g. important [ɪm'pɔɹʔn̝t]
- [ɾ] tap: between vowels (intervocalic) e.g. water ['wɑɾɚ]; that is [ð æɾɪz]
1
Oct. 8, 2012
(3) Unreleased final stop 不除阻詞尾塞音[ ̚ ] vs. glottal stop 喉塞音 [ʔ]:
If you say hit in isolation, the final /t/ is usually unreleased: [t ̚].
In the phrase hit me, the /t/ usually becomes a glottal stop, because the next word starts with a
consonant: ['hɪʔ mi]. *Oral stops tend to be unreleased when word-final.
- bottom vs. button: often confused by Taiwan students:

 bottom (X) ['bʌɾən] (O) ['bɑɾəm]
 button (X) ['bɑʔn̝] (O) ['bʌʔn] ['bʌʔʔn] ['bʌt̚ʔn]
(1) the vowel: [ɑ] vs. [ʌ]
- [ɑ] is a long vowel, as in father; it is also what children in the US are taught is “the short o” (ex.:
mop [mɑp], hot [hɑt], sock [sɑk]), as opposed to the “long o” (e.g. code [koʊd], boat [boʊt], low
[loʊ]. (note that the KK [o] is written in IPA as [oʊ]!)
- [ʌ] is a short vowel; it is sometimes called “wedge” 楔子 (Taiwan: xièzi, PRC: xiēzi); it is
close to [ə] in US English, and closer to [ɑ] in British English – which seems to account for why
Taiwan students so often pronounce it as [ɑ] instead of [ʌ]. Some 參考書 teach that [ʌ] sounds like
注音符號「ㄚ」!
(2) the consonants:
i. the final consonant:
bottom ['bɑɾəm] (closed mouth for [m], or you may confuse your listener)
button ['bʌʔn̝], with syllabic [n̝]
ii. the allophones of /t/: the /t/ in bottom is a tap (intervocalic /t/), in button it is a glottal stop [ʔ]
(due to nasal plosion – to be taught later)
*About the confusion:

- Although it seems funny that we would mix up two words with such different meanings, this kind of
confusion may also occur in foreign learners’ Mandarin, e.g. they will often get the tones wrong.
- For example, some learners may mix up the two phrases 我要你 (wǒ yào nǐ, meaning ‘I want you’)
and 我咬你 (wǒ yǎo nǐ, meaning ’I bite you’).
 Such confusion is common in language learning – but it needs to be carefully sorted out and
corrected!
Two other pairs to watch out for: sore [sɔɹ] and sour [saʊɹ]; tone [toʊn] and tongue [tʌŋ]!
- Canadian “rising”?
(O) raise (transitive verb 及物動詞): Canadians raise their tongues  (O) Canadian raising
(X) rise (intransitive verb 不及物動詞): the tongue rises
*Vocabulary: raise and rise

- In American English, the noun for 加薪 is raise; in British English, it is rise.
e.g. (AE) I got a raise. (BE) I got a pay rise.
2
Oct. 8, 2012
- American or British English?

Some people seem to be confused when writing about the differences between them. Make sure you
follow the guidelines:
(1) Pay attention to Ms. Chung when she talks about the differences in class.
(2) There may differences of spelling, pronunciation, or collocation 片語. Make sure to clarify
each one precisely in your mind.
(3) If you don’t hear something clearly, raise your hand and ask Ms. Chung to repeat it during class,
or ask the TAs or Ms. Chung during breaks.
- Rules for the –s/-es suffix:

(1) The three situations when we add -s or -es to a word in English:
 nouns with regular plurals, e.g. book  books; bag  bags; bush  bushes
 third person singular present tense of regular verbs (規則動詞的第三人稱單數現在式)
e.g. I look, you look, he looks
 possessive (所有格) e.g. the book’s cover, the bag’s handle, the bushes’ leaves
(2) /s/ is a morpheme 語素 with one of the three above meanings, which has three different
phonologically-conditioned realizations 呈現, called allomorphs 同位語素:
 voiceless [s]: after voiceless sounds e.g. books [bʊks]
 voiced [z]: after voiced sounds e.g. bags [bæ gz],rows [ɹoʊz]
 schwa [ə] + voiced [z]: after sibilants e.g. bushes (AE) [bʊʃəz]/(BE) [ bʊʃɪz]
 the schwa [ə] is added in order to make the final [z] clear and audible – this is a kind of
dissimilation 異化; otherwise plurals for words such as bus “buss” [bʌs-s] would be unclear, because
the final [s] or [z] would be too similar to the preceding sound for it to be heard clearly. [z] is used
rather than [s] because the preceding sound [ə] is a vowel and thus voiced.
(3) We need to learn these rules to speak English correctly, or we may cause confusion or
misunderstandings when talking with others.
*The six sibilants in English:

- sibilants 嘶音 (咝音 in the PRC)
- /s, z, ʃ, ʒ, tʃ, dʒ/
- Palato-alveolar 顎齦音 vs. alveolo-palatal 齦顎音:

 English – palato-alveolar: [ʃ, ʒ, tʃ, dʒ]
 Mandarin – alveolo-palatal:ㄐㄑㄒ [tɕ tɕʰ ɕ]
- IPA symbol [z]: (X) [z] (O) [z]

 Do not put a bar on the symbol! If you do anything to the IPA symbol in your handwriting, it may
change the value. For example, the mark you add to the symbol may be interpreted as a diacritic.
- Spelling: trill 顫音 vs. trial 審判 vs. thrill 戰慄

3
Oct. 8, 2012
*Pronunciation of「顫」音:
-ㄓㄢˋzhàn is the standard pronunciation 正讀; in Taiwan
-ㄔㄢˋchàn is a variant pronunciation 又讀; common in the PRC
2. Hanyu Pinyin 漢語拼音

*Homework: Write out the first sentence of the parrot story “鸚鵡有人性被劫傷離情” in Hanyu Pinyin
on course webpage 3 (see below for the link).
a. Where do we put the tone mark 調符 when there is more than one vowel in a word?
- Diphthongs: always put the tone mark on the main vowel 主要元音 (元音 is a synonym for 母音
‘vowel’; 元音 is used more commonly in linguistic scholarship in the PRC)
(1) For: ㄞ ai/ㄟ ei/ㄠ ao/ㄡ ou  add the tone mark over the first vowel
(2) For all the other diphthongs: e.g. 六 liù (cf. 又 yòu (see b. (2)), 缺 quē  add the tone mark
over the second vowel.
- Triphthongs: the main vowel is the one in the middle (consonants that are not glides)
ㄨㄞ wai/ㄨㄟ wei/ㄧㄠ yao/ㄧㄡ you e.g. ‘飆’ biāo
b. Some things students may need to be careful about:

(1) When consonants are added before the vowel, the Pinyin for triphthongs will look different.
ㄨㄞ wai _uai ㄧㄠ yao _iao
ㄨㄟ wei _ui ㄧㄡ you _iu
w becomes u y becomes i
(2) For some triphthongs, the tone mark will shift to the last vowel because Hanyu Pinyin omits one of
the vowels.
e.g. 為 wèi  鬼 guĭ (“e” is omitted); ‘又’ yòu  ‘六’ liù (“o” is omitted)
(3) Similar changes (as mentioned in (1)) are present in ㄧ yi, ㄨ wu, and ㄩ yu.
e.g. 一 yi  逼 bi
*yi and wu: the asymmetry in Taiwan Mandarin:

- When we say phrases like 一個人 (yígèrén), we have a glottal stop before the yi [ʔi], and it will
sound weird if we have a glide /j/ before it [ji] (although a few people will have it in their
pronunciation).
- However, when we say 烏來 (wūlái), such distinction is not the case for wu – both [ʔu] and [wu]
are acceptable, which suggests that there is an asymmetry in the phonology of Taiwan Mandarin and
that the sound wu is unstable in Taiwan Mandarin. (Discussion will be continued below.)
- Think about it, and start paying attention to how people in Taiwan say these two sounds.
(4) Do we use an umlaut ̈ in Hanyu Pinyin?
2
- In Wade-Giles Romanization 威妥瑪氏羅馬拼音, the umlaut is used for ㄩ. e.g. ‘餘’yü
- In Hanyu Pinyin, we only use an umlaut to make distinctions between words such as 努 nŭ and 女
nǚ, and 爐 lú and 驢 lǘ. It is only after /l-/ and /n-/ that both /i/ and /y/ can occur.
4
Oct. 8, 2012
- We don’t use an umlaut with other initials: We don’t need an umlaut in 需 xū, because ㄒㄨ is
impossible in Mandarin. ㄐㄑㄒ can be combined with only two vowels: ㄧ/i/ and ㄩ/y/
because these three consonants are (alveolo-)palatal sounds.
cf. ㄓㄔㄕ and ㄗㄘㄙ cannot occur before ㄧ/i/ and ㄩ/y/.
*Homework: Read “Romanization III” on course webpage 13 (see below for the link).
3. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 20)

(Continuing “The articulation of vowel sounds”)
a. Front vowels: the vowels in heed [i], hid [ɪ], head [ɛ], had [æ ](there are other front vowels which are
diphthongs, but we are not discussing these here)
- The tongue is domed and the highest point is in the front of the mouth; therefore they are are
classified as front vowels.
- For [i], the tongue is very close to the roof of the mouth, for [ɪ] a bit lower, even lower for [ɛ], and the
lowest for [æ .]
- We can try to use the method that we learned before for consonants to know where the tongue is when
making these vowels: sucking in air while holding the target position for the vowels. It works the best
for [i] because the articulators are closest to each other here. It doesn’t work that well for the vowel in
hid and most other vowels, since the articulators are too far apart.
- Take out a mirror to see how your mouth becomes progressively more open when saying these words
(from heed to had = from the highest to the lowest).
 [i] a high front vowel / [ɪ] a mid-high front vowel/ [ɛ] a mid-low front vowel / [æ ]a low front
vowel. (Just remember the progression of your mouth when saying these words and you will
remember the terms!) (See Figure 1.13 on page 21 of the textbook)
b. Back vowels: the vowels in father [ɑ], good [ʊ], food [u] (there are also other back vowels not
discussed here)
- The tongue is domed and the highest point is in the back of the mouth; therefore they are classified as
back vowels.
- Among the three vowels, the tongue is the lowest for [ɑ], higher for [ʊ], and the highest for [u].
 [ɑ] a low back vowel/[ʊ] a mid-high back vowel/[u] a high back vowel.
- Just like what we did for front vowels, we can also try to know where the tongue is when making
these vowels by sucking in air while holding the target position for the vowels. We will be able to
feel the cool airflow when producing the [u] which is best for this test, since the path for the air is
narrow enough.
*Long and short vowels in American English: (also see p.5 of Unit 4 handout)
- Long vowels by definition can occur in both open and closed syllables; they include [i] [æ ] [u] ɔ[ ] [ɑ]
- Short vowels can occur only in closed syllables; they include [ɪ] [ɛ] [ʌ] [ʊ]
(we are only talking about the monophthongs here)
5
Oct. 8, 2012
c. Lip gestures: While your tongue is making different gestures for different vowels, your lips are also
doing something. Pronounce the vowels mentioned above and watch your the movements of your lips
in a mirror.
- Start from the front vowels. Going from [i] to [æ ]you will see your lips go from very spread to more
relaxed.
- Next try the back vowels. For [ɑ], your lips are still wide open, though so less than for [æ ;] they are
more relaxed. For [u] and [ʊ], you will notice your lips are rounded.  In general, lip-rounding
圓唇 is typical for back vowels.
 Another way to classify vowels: rounded or unrounded.
*Rounded back vowels:

- It is very normal that many languages have rounded back vowels. For example, in Mandarin, we
also have rounded back vowels: [u] in ‘儲’ [tʂʰu], [ʊ] in ‘純’[tʂʰu̯n], [oʊ] in ‘歐’ [oʊ], and [ɔ] in ‘紅’
[hɔŋ]. (This will be discussed next semester)
- Unrounded back vowels are more marked (有標記、突出的). e.g. [ɯ] in some varieties of Southern
Min (豬, 箸[tɯ]) and [ɤ] Mandarin ‘餓’ (Beijing dialect)
- Rounded front vowels are also marked. e.g. [y] in French, German, Swedish, Mandarin, Cantonese,
Tibetan, and many other languages ([y] is much more common than [ɤ])
d. Summary of this section: three parameters when describing vowel gestures:

(1) Height of the body of the tongue  high/low vowels (first formant/F1)
(2) Front-back position of the tongue  front/back vowels (second formant/F2)
(3) Degree of lip-rounding  rounded/unrounded vowels (third formant/F3)
4. Discussion of questions raised during break:

a. Distinction between [ʒ] and [dʒ]:
- The manners of articulation are different: [ʒ] – fricative (擦音) vs. [tʒ] – affricate (塞擦音)
- [ʒ] is a continuant (連續音, in which the sound quality doesn’t change in the process of articulation)
while [tʒ] is not because it has the stop component [t], which features three stages in the articulatory
process – approach, hold, and release. We can use a waveform to see the difference more clearly.
- To pronounce [dʒ], we can simply compare the sound with the consonant in the word ‘這’, which is
also an affricate [tʂ]. It is sounds more similar to [dʒ] especially in Taiwan Mandarin because we use
the retroflex (捲舌音) less than people who speak the Beijing dialect. e.g. judge [dʒʌdʒ]
- To pronounce [ʒ], just make the [ʃ] sound (as in she) first and then add voicing.
e.g. usually ['juʒjuəli]; genre ['ʒɑnɹe]; beige [beɪʒ] (米色的)
*Continuants in English:
- include vowels, nasals, and fricatives
*Homework: Read the CET article No. 7 “Stop at stops” (see below for link).
6
Oct. 8, 2012
b. Continuing the discussion on ㄧ yi and ㄨ wu:

- yi: (A)/∅i/, (B) [ʔi], or (C) *[ji]?  (A) and (B) sound acceptable to Taiwan Mandarin native
speakers, but the third sounds weird and is marked among Taiwanese.
*Footnote on the marks:

- ∅: zero-initial 零聲母, meaning there is nothing there. It is NOT an IPA symbol, but it is used in
phonological studies on Chinese.
-*: meaning the item is not used or common in the system being discussed
- wu: (A) /∅u/, (B) [ʔu], or (C) [wu]?  They all sound ok for Taiwanese although Ms. Chung is more
used to using (C), perhaps under the influence of English. Children who live in the same
neighborhood with Ms. Chung (in 烏來) seem to say (B) more often.
(The (A) and (C) for wu sound similar, but there is more lip-rounding when we say [wu].)
- Try to say the six sounds on your own and see which one you like the best.
 For yi, most people in the class liked (A); for wu, there was no certain agreement.
- What about the wu in the word for service in Mandarin 服務?
 Ms. Chung’s daughter, who was born and grew up in Taiwan, uses mainly (B); a lot of students
she has met also tend to use (B) more.
- Linguists haven’t figured it out yet, and there is still a lot of discussion about this. Someone should
collect data and do a rigorous study on this topic!
*Homework: Observe which sound people around you and you yourself use.
- You may jump to conclusions based on vague impressions if you don’t pay enough attention to how
people actually say things. Pronunciation varies, especially among young people.
e.g. Some people may say [lɤ] when they mean to say [ʐɤ] for ‘熱’ in the phrase 外面好熱; some
people may say [loʊ] for 肉 in 我要吃肉. (Pronunciation in Taiwan Mandarin is changing, and it is
very possible that the [ʐ] sound will eventually become [l].)
- Sometimes what you think you are doing and what you are really doing are different – if you are not
paying close attention.
e.g. A lot of Americans may claim they never use contractions, saying “I don’t use contractions!”
5. Homework
a. Exercises for Chapter One (will be gone over next class)
b. Write out the first sentence of the parrot story “鸚鵡有人性被劫傷離情” in Hanyu Pinyin on course
webpage 3 http://homepage.ntu.edu.tw/~karchung/intro%20page3e.htm
c. Read “Romanization III” on course webpage 13

7
Oct. 8, 2012
d. Read the CET article No. 7 “Stop at stops”

e. Read the rest of the CET articles that have been assigned, and read ahead in the textbook.
f. Start observing how people pronounce ㄧ yi and ㄨ wu in Taiwan.

bottom ['bɑɾən]  ['bɑɾəm] (close your mouth for [m])
button ['bɑtən]  ['bʌʔn̝] (the vowel; the /t/ allophone)
letter G [dʒy] (like ‘居’)  [dʒi] (there is labialization 唇化, but do not round your
lips too much!)
tongue [tʌn]  [tʌŋ] (consonant)
fairly ['fɛəli]  ['fɛɹli] (the [ɹ] was missing)
the [lə]  [ðə] (stick out your tongue tip)
heed [hɪd]  [hid] (vowel)
this is so_by… rush through  pause before prepositions
holding ['hoʊ_dɪŋ]  ['holdɪŋ] (the [l] was missing)
position, position [pə'sɪʃən], [pə'ziʃən]  [pə'zɪʃən] (consonant; vowel)
Start to look out for words in which the non-final s is
pronounced as [z] and commit to memory!
most [məʊst]  [moʊst] (system: (BE) [məʊst] – be consistent!)
this [lɪs]  [ðɪs] (stick out your tongue tip)
the articulator [li]  [ði] (stick out your tongue tip)
articulator [ɑː'tɪkjuleɪɾɚ]  [ɑɹ'tɪkjuleɪɾɚ] (the [ɹ] was missing)
are [ɑː]  [ɑɹ] (the [ɹ] was missing)
head [hæd]  [hɛd] (the vowel was a little bit lower than [ɛ]; don’t
lower your jaw too much)
these four words, flat  continuation rise before comma
becomes [bɪ'kʌ̃z]  [bɪ'kʌmz] (close your mouth for [m])
remains [ɹi'mɛnz]  [ɹi'meɪnz] (vowel)
as [əz]  [æz] (vowel); “as” is sometimes but not always reduced
to [əz]
called [kold]  [kɔld] (vowel)
mid [mid]  [mɪd] (vowel)
in [in]  [ɪn] (vowel)
mid-high vowel  mid-high vowel (mid and high are both important in
the adjective; vowel is repeated)
vowels ['vaʊʷəls]  ['vaʊʷəlz] (consonant; the plural s is after a voiced [l])
8
Oct. 8, 2012
figure ['figjɚ]  ['fɪgjɚ] (vowel)

for [fɔː]  [fɔɹ] (the [ɹ] was missing)
In all three, In *all three,  In all *three, (tonic stress and continuation rise)
close (n.) [kloʊːs]  [kloʊs] (the vowel before the voiceless sound shouldn’t
be too long) cf. close (v. ) [kloʊːz]
surface ['sɜfəs]  ['sɜ˞fəs] (r-coloring in AE) cf. (BE) ['sɜːfəs]
back vowel back *vowel  *back vowel (back contrasts with front; vowel is
repeated)
food [fʊd]  [fud] (vowel)
vowel ['vaʊʷə_]  ['vaʊʷəl] (the [l] was missing)
high front_vowel no pause  glottal stop between final t and words that start with a
consonant: stop at stops
of [of]  [əv] (vowel; consonant)
the rush of air when… flat  continuation rise + pause before relative pronouns
breathe (v.) [bɹɛθ]  [bɹið] (vowel; consonant) vs. breath (n.) [bɹɛθ]
e.g. Breathe easy! (放鬆) vs. a breath of fresh air
together [tu'gɛd̪ɚ]  [tə'gɛðɚ] (the unstressed o is a schwa here; stick out
your tongue tip for th) e.g. the, this, these
in_a, in_addition no linking  linking (when a final consonant is followed by a vowel)
Try to say “[ɪ] 呢”, and you may get it right faster.
mirror [mɪɹɔɹ]  [mɪɹɚ] (vowel)
that_occurs [ðæʔ ʌ'kɜ˞z]  [ð æɾʌ'kɜ˞z] (/t/ between two vowels  tap); a glottal
stop is now becoming very common in this position in
UK English – this used to be stigmatized
lip [lip]  [lɪp] (vowel) cf. leap [lip] (跳躍)
who’d [hʊd]  [hud] (vowel) cf. hood [hʊd]
*Choosing the variety of English you wish to speak and cultivate:

- In RP, post-vocalic “r” is dropped (unless before vowels), and the [oʊ] is pronounced as [əʊ].
- Choose the variety of English you wish to learn and try be consistent.
*Force of habit:
- Some of our pronunciation habits have become so deeply ingrained that it is hard to change them in a short
period of time. It is easier to pronounce words correctly when they are read in isolation, but in connected
speech, we tend to make the same mistakes over and over.
- We have to stop and think about the pronunciation of the words we’re about to read before we read them.
After doing it for a while, the new pronunciation will be automatic and you don’t have to take extra time to
consciously think about it.
- Some words are so common and short (such as the, for, in), that we may think it’s not that necessary to
pronounce them carefully, but it is important! Regardless of how short and common a word is, it should be
9
Oct. 8, 2012
pronounced correctly. Think of how odd it sounds when foreigners get the tones wrong in Chinese. They are
probably also thinking – we don’t use tones in English, so they can’t be that important. It shouldn’t matter if I
get them wrong sometimes. But you will notice every time, and sometimes it will confuse you!
National Taiwan University/Karen Steffen Chung 史嘉琳
4 This work is licensed by Creative Commons
http://homepag
e.ntu.edu.tw/~k National Taiwan University/Karen Steffen Chung 史嘉琳
archung/intro% Attribution-NonCommercial-ShareAlike 3.0 Taiwan
20page3e.htm
http://homepag
e.ntu.edu.tw/~k National Taiwan University/Karen Steffen Chung 史嘉琳
7 archung/intro% This work is licensed by Creative Commons
20page%2013.h Attribution-NonCommercial-ShareAlike 3.0 Taiwan
tm
Caves Educational Training Co., Ltd / Hello! E.T. Issue No.
http://homepag 75, January/February 2013, p. 12-14
e.ntu.edu.tw/~k http://www.cet-taiwan.com/drcet/type.asp?title_type=Hello%20E.T.
8
archung/pubs/7 This work is licensed by Caves Educational Training Co., Ltd for the
5_hello_et.pdf use of OCW ONLY. The copyright belongs to the above mentioned
creator(s).

8-9 This work is licensed by Creative Commons
10
Oct. 15, 2012
Fall 2012
10. The Sounds of Vowels; Suprasegmentals
2. Homework
II. Notes
a. Summary of “The Articulation of Vowel Sounds”:
- The following three basic parameters are needed to describe vowel gestures (excluding diphthongs):
(1) How high the body of the tongue is:
e.g. [i]  [æ ](front vowels): from high to low
 a resonance is created in the back of the oral cavity, the space just above the vocal folds
 relates to F1 – the first formant (共振峰)
(2) How front or how back the highest point of the tongue is:
e.g. [i]  [u] (high vowels): from front to back
 a resonance is created in the front part of the oral cavity, just behind the lips
 relates to F2 – the second formant
(3) How rounded the lips are:
e.g. [u]  [ʊ]  [ɔ]  [ɑ] (back vowels): from rounded to unrounded
 the lowest back vowel (AE and BE) is unrounded: [ɑ]; there is an additional rounded BE vowel: [ɒ]
 [u] and [ɔ] are more rounded in RP than they are in American English
 English /r/ has lip rounding as a secondary articulation
 Lip rounding and r-coloring are reflected in F3 – the third formant
*About the section summaries

- Usually a succinct (精煉) review of the main points of the section
- If you don’t understand something in the summary, go back to the main body of the chapter and
review it carefully.
- Try the following methods in order to better know the differences between vowels:
(1) To distinguish between the high and low vowels, observe the position of your jaw when saying the
front vowels [i] , [ɪ] , [ɛ] and [æ ]as in the words heed, hid, head, and had. As you go from the highest
to the lowest vowel, you will notice your jaw descending and your mouth opening wider.
1
Oct. 15, 2012
(2) To distinguish between the front and back vowels, contrast the high vowels [i] and [u] as in the
words he and who. You will also notice a change in your lip gestures as you go back and forth
between the two sounds.
- We are not really describing the vowels articulatorily here, and the classification we are talking about is
not very accurate or sufficient. There are four reasons for this:
(1) The high vowels are not equally high, e.g. [u] is not as high as [i]
(2) The back vowels are not equally back, e.g. [u] is not as back as [ʊ]
(3) Many different shapes of the tongue are used in forming the front vowels and back vowels, but we
are not considering these.
(4) There are also variations in the width of the pharynx when we produce different vowels, but we are
not considering these.
 We will use acoustic and auditory phonetics to describe vowels in later chapters.
b. The sounds of vowels:

- Compared to the sound produced by a tuning fork (音叉, which has only one single main pitch), the
structures of vowels are very complex. In fact, each vowel contains several pitches, and that is why
some people seem to have a very rich voice when they sing (this will be discussed further in the future).
- Fundamental frequency (基本頻率, aka 基頻): The pitch at which the vowel is actually heard; the rate
of the vibration of the vocal folds; designated as F0.
- Overtone pitches (倍音): Other frequencies of the voice which are multiples of the fundamental
frequency and have smaller amplitudes; these are not heard as individual pitches in the way that the
fundamental frequency is, e.g. If someone speaks at 100Hz, they will at the same time actually be
producing many frequencies that are multiples of 100: 200Hz, 300Hz, 400Hz, and so on.
- Production of vowel sounds: (disregarding for the time being the different airstream mechanisms)
Phonation The airstream passes through the larynx, causing the  fundamental frequency
process tensed vocal folds to vibrate (with overtones)
Oro-nasal The velum rises, i.e. closes off the nasal tract,  oral vs. nasal sounds
process causing the airstream to pass through the oral tract
and produce oral sounds; a lowered velum produces
nasal sounds
Articulatory The shape of the mouth (formed by tongue gestures,  groups of overtones
process lip gestures, etc) creates particular resonance which correspond to the
chambers in the oral cavity which cause certain natural resonance
groups of overtones to resonate and become more frequencies of spaces in the
prominent oral cavity resonate and
become more prominent
 this causes a vowel to sound different from other vowels (i.e., they have
different vowel qualities)
If you don’t understand the parts about resonance and formants above, don’t worry; these will be taught in
greater detail second semester.
2
Oct. 15, 2012
- Formants: Try using this set of words for the following experiments: heed, hid, head, had, hod, hawed,
hood, who’d.
(1) Say the words in sequence, using a flat intonation, and lengthen the vowels.  They all have the
same pitch – the same fundamental frequency.
(2) Whisper the words (no vocal fold vibration).
 You are mainly hearing the pitches of only one resonant cavity – the one in the front of our oral
cavity/mouth. Whispering enables you to more clearly hear the different F2 pitches of these vowels.
The pitches seem to be going down as you go through the set.
i. front vowels (from high to low vowels): pitches go from high to low – heed > hid > head > had.
ii. back vowels (from low to high vowels): pitches go from high to low – hod > hawed > hood> who’d.
 You will hear a descending pitch pattern when the vowels are read in this order:
[i]  [ɪ]  [ɛ]  [æ ] [ɑ]  [ɔ]  [ʊ]  [u]
∵ The shape of the tongue is making the front cavity of our mouth bigger and bigger and thus we
get lower and lower frequencies. (F2, the second formant, is easier to hear than F1)
 The lowest pitch is in one of the three words hawed, hood, and who’d; which one is the lowest
depends on the speaker ’s accent – regional accents may vary in the pronunciation of these three
vowels.
(3) Whistle from the highest note you can make to the lowest note.
 When you whistle the highest note you can, the part of your tongue sticking up the highest should
be in about the same position that you use to say [i]; similarly, when you whistle the lower notes,
the body of your tongue should be in the positions you use to say [ɔ], [ʊ], and [u]  F2.
 In Chinese phonology, there are two categories of vowels: [i]-like sounds are classified as 細, and
the [ɔ] and [ɑ]-like sounds are classified as 宏. This is a kind of Sound symbolism (e.g. [i] for
tiny things (the opening of your mouth is small;  “Say ‘cheese’!”; and [ɔ] and [ɑ] for large
things (your mouth is open wider); and a large man tends to have a lower voice
 To make a high back vowel, try whistling the lowest note you can. Your tongue and lip gestures
should be in the position and shape used to pronounce [u].
(4) For F1: Use a low, creaky voice.
 Start from the easy ones: say the [ɑ] (as in the word hod), and then the [æ ](as in the word had).
 We are using the lower part of the oral tract, and a resonance is created in the space in the back of
the tongue and above the vocal folds (the pharyngeal area).
 Creaky voice can help us to hear the pitches of F1, the first formant.
 Our vocal folds are vibrating so slowly that you can count how many individual vibrations there
are per second.
 As you go from (A) [i] to [ɪ] to [ɛ] to [æ ,] the pitch you hear will get higher and higher.
∵ The space in the oral cavity between the vocal folds and the tongue is getting smaller and
smaller.
 When you go from (B) [ɑ] to [ɔ] to [ʊ] to [u] the pitch will be lower and lower.
∵ The space in the oral cavity between the vocal folds and the tongue will get larger and larger.
3
Oct. 15, 2012
- Summary:
(1) There is more than one pitch in vowel sounds – there is a fundamental frequency plus groups of
overtone pitches made more prominent through resonances in the oral tract. These are called
“formants”.
(2) The first two formants – F1 and F2 – are major contributors to vowel quality.
(There are other formants in vowels, but the two are the most important ones are F1 and F2; F3,
which is associated with lip rounding, is third in importance; formants beyond F3 are not usually
of concern to us in articulatory phonetics)
 F1 (resonance space between vocal folds up to the highest point of the tongue) can be “heard” by
speaking in a creaky voice; F1 can be used to roughly classify vowels in the horizontal
dimension;
High vowels have a lower F1, low vowels have a higher F1.
 F2 – (resonance space behind lips up to the highest part of the tongue; can be “heard” by
speaking in a whisper or whistling; F2 can be used to roughly classify vowel sounds in the
vertical dimension;
Front vowels – higher F2; back vowels – lower F2.
- Isaac Newton was the first, in 1665 at the tender age of 12, to discover the association between the
second formant and vowels. If we slowly fill a deep narrow glass with water, we can hear the F2 of
the air gradually go up in pitch; you may also be able to hear the falling F1 of the water at the same
time.
c. Suprasegmentals (超音段):
- Definition: features that are superimposed on the syllables
- vs. Segments (音段): collective term for “vowels and consonants”
- Include: variations in stress, pitch, and length
(1) Stress variations:
 Can sometimes distinguish between nouns and verbs  grammatical function
e.g. insult (v.) vs. insult (n.); pervert (v.) vs. pervert (a.); overflow (v.) vs. overflow (n.)
stress the second syllable when it’s a verb and the first when it’s a noun or an adjective
cf. occasional use of tone in Mandarin for a similar purpose: 釘釘子 – dìng (v.) vs. dīng (n.);
電鑽鑽洞 – zuàn (v.) vs. zuān (n.); 處 chù (n.) vs. 處 chǔ (v.).
 Can show contrastive emphasis  semantic function (the third rule of English intonation)
e.g. I want this one, not that one.
*3 basic rules of sentence intonation:

1. Stress content words; do not stress function words.
2. New information is stressed; old information is destressed (the stress is removed).
3. Contrastive elements are stressed.
4
Oct. 15, 2012
 English stress is used to create relative contrasts through the use of greater lung energy or a
change in laryngeal activity.
 The four elements of English stress:
1. Pitch (higher or lower)
2. Vowel length
3. Vowel clarity, i.e. use of a full vowel rather than a schwa
4. Greater loudness
 Try to tap with your pen to find where the stress occurs on the word abominable
There are five syllables in this word: a-----bo----mi----na----ble
(A) (B) (C) (D) (E)
i. Say the word and tap without thinking: On which syllable did you tap?
 (B) – where the stress is
∵You use muscular tension to produce the stress, and the rest of your body will tense at the same
time.
ii. Now say the word again but try to tap on (A), (C), (D), and (E): Is it easy to do?
 It will be especially difficult for a native speaker of English to tap on these syllables because
these syllables are not stressed. Many people cannot tap on these syllables without special
training.
(2) Pitch variations:
 Reflect changes in laryngeal activity, i.e. the fundamental frequency (V0), the rate at which the
vocal folds vibrate per second
 “Pitch of the voice” – what that you alter when singing different notes
 Using a waveform to estimate pitch:
i. Each opening and closing of the vocal folds causes an air pressure peak in a sound wave
ii. We can count how many peaks there are per second in the sound wave to estimate the
pitch  frequency (頻率): Frequency is defined as the “number of complete repetitions of a
pattern/cycle of air pressure variation occurring in a second” (acoustic property of sound).
iii. In the case of the vocal folds, this simply means how many times they open and close per
second. The faster they open and close, the higher the frequency, and the higher the pitch
perceived by our ears. Slower opening and closing means the frequency is lower, and it
produces a lower pitch. (p.24)
 Frequency is measured in Hertz (Hz). 1 Hz = One closing and opening in one second.
 For speech events, we usually measure time in milliseconds (毫秒, 1/1000 second).
 Pitch 音高 is the subjective perception by our ears of how high or low a sound is; it is
subjective because the construction of our ears introduces various distortions in the sounds we
hear; it coincides fairly closely but not exactly with frequency, which is the measurement of an
objective physical event.
 Intonation: the pitch pattern of a sentence
i. Typical intonations: try to say the two sentences: (A) This is my father. (B) Is this your father?
Which syllable has the highest pitch? Your intonation should be like the following:
5
Oct. 15, 2012
(A) This is my father. (B) Is this your father?

 statement  yes-no question
ii. Change of meaning: we can change the meaning of a sentence from a statement to a question
just by altering the intonation. e.g. That’s a cat.
That’s a cat.  statement That’s a cat?  question (astonishment)

cf. French e.g. C’est un livre
C’est un livre. ‘This is a book.’ C’est un livre. ‘Is this a book?’
- The relative nature of suprasegmentals:

 Suprasegmentals are relative! We describe them in comparison with other items that occur in the
same utterance.
e.g. i. Say abominable (A) by shouting; then (B) softly  you stress the second syllable in both
cases, but the volume of the two utterances is different, which makes it difficult to compare them
in terms of absolute degrees of stress.
ii. The average frequency of a child’s voice is about 265 Hz, and children can also use the
same intonation patterns as adults do for the same sentence.
 The absolute values are never linguistically important, but suprasegmentals can show
i. how old the speaker is,
ii. whether the speaker is a female or a male,
iii. what his or her emotional state is right now, and
iv. what kind of attitude he or she holds towards the topic under discussion.
2. Homework:
a. Finish the exercises for Chapter One.
b. Do the Hanyu Pinyin exercises.
body ['bʌdi]  ['bɑdi] cf. buddy ['bʌdi]
[ɑ] is the vowel with the greatest inherent length in
English
front-back position no pause  stop at stops – [p, t, k, b, d, g] (see below for extended
discussion) Make sure you do the text mark-up!
position [pə'siʃən]  [pə'zɪʃən] (consonant; vowel)
points [pɔːɪnts]  [pɔɪnts] (do not lengthen the vowel)
figure caption and… flat  continuation rise before conjunctions and
in vowels, ... punctuation
figure caption figure caption  *figure caption (compound noun stress)
jaw [dʒo]  [dʒɔ] (vowel)
6
Oct. 15, 2012
also ['ɔ_soʊ]  ['ɔlsoʊ] (the [l] was missing)

feel [fɪɫ]  [fiɫ] (vowel)
involved [ɪn'vɔlft]  [ɪn'vɔːlvd] ([l] is voiced  preceding vowel should be
longer; [v] is voiced  the –ed should be pronounced
as a voiced [d])
figure ['figjɚ]  ['fɪgjɚ]
chapter ['tʃ̹æ ptɚ]  ['tʃæ ptɚ] (do not round your lips too much for [tʃ])
cf. children ['tʃɪldɹən]
(more rounded )
#
acoustics [ə'kustɪk_]  [ə'kustɪks] (the final -s was missing)
all [ol]  [ɔl] (vowel)
tuning ['tjunɪŋ]  ['tunɪŋ] (vowel) cf. (BE) ['tʃjunɪŋ] (affrication 塞擦音化)
cf. due (AE) [du] vs. (BE) [dʒju]
containing [kɑn'tɛnɪŋ]  [kən'teɪnɪŋ] (vowel)
Take away the [n] then lengthen vowel to get it right.
of [əf]  [əv] (consonant)
different ['difəɹənt]  ['dɪfɹənt] (vowel; schwa elision  only two syllables)
overtones ['ɔvɚtoʊnz]  ['oʊvɚtoʊnz] (vowel)
from another_by rushed through  pause before prepositions
from [fɹə̃_]  [fɹəm] (close your mouth for [m])
separate (a.) ['sɛpəɹət]  ['sɛpɹət] (schwa elision) cf. separate (v.) ['sɛpəɹeɪt]
the vowel is said, flat  continuation rise before punctuations
When you whisper… When *you whisper…  When you *whisper…(stress content words, don’t
stress pronouns)
a series of sounds_on rushed through  pause before prepositions
regional ['ɹɪdʒənəl]  ['ɹidʒənəl] (vowel)
either ['idɚ]  ['iðɚ] (stick out the tongue for [ð])
hod [hoʊd] [hɑd] (vowel)
which [wɛtʃ]  [wɪtʃ] (vowel)
Accents of English rushed through  pause after the subject is finished
_differ slightly…
These vowels these vowels  these vowels (do not stress function words unless you
are making a contrast)
change [tʃɛndʒ]  [tʃeɪndʒ] (vowel) Try:ㄔㄟㄧ + ㄣ + ㄓ (Forget the n
first or you will produce a strange vowel due to force
of habit!)
order ['ɔ_dɚ]  ['ɔɹdɚ] (the [ɹ] was missing; don’t confuse w/”odor”!)
on [an]  [ɑn] (vowel)
but_they no pause  stop at stops
creaky ['kɹɪki]  ['kɹiki] (vowel)
7
Oct. 15, 2012
known [nɔn]  [noʊn] (vowel)

a_long_time rushed through  pause between stressed monosyllabic words
observed [əb'sɜ˞vd]  [əb'zɜ˞vd] (consonant)
filling ['filɪŋ]  ['fɪlɪŋ] (vowel)
symbols ['simbəlz]  ['sɪmbəlz] (vowel)
best [bæst]  [bɛst] (vowel)
letter ['læɾɚ]  ['lɛɾɚ] (vowel)
handwriting ['hæ n_ɹaɪɾɪŋ]  ['hæ ndɹaɪɾɪŋ] (the [d] was missing)
length [lɛnθ]  [lɛŋkθ] (velar [ŋ] + epenthetic [k] (n.: epenthesis)
(However, some native speakers do say [lɛnθ])
descriptions [dəs'kɹɪpʃən_]  [dəs'kɹɪpʃənz] (the final –s - pronounced as [z]
because the [n] is voiced – was missing)
articulation [ɑɹtɪkə'leɪʃən]  [ɑɹtɪkju'leɪʃən ](vowel)
*About “stopping at stops”:

- Not “stopping at stops” is a typical error of Taiwanese students
- Compare the assimilation of final stops in Southern Min;
it is UNLIKE English, and it is one possible source of failure to stop at stops in Taiwan English!
e.g. ‘一粒’ [tɕi ljap̚] ‘一粒蛋’ [tɕi ljam nŋ̩] (stop  a nasal – continuant – assimilation of manner)
‘茉莉花’ [bak̚ li] [baŋ li] (syllabic )
- Not stopping at stops in English: found in singing by some pop singers (especially Motown singers)
e.g. “I put_my [pʊmːaɪ] hand in my father ’s glove” (Tori Amos: “Winter”)  no stop
Why? Student guesses:
i. because they are singing fast so they don’t stop at stops – but normally native speakers don’t do that
ii. to maintain voicing. We can’t lengthen obstruents consonants when singing, and try to avoid lengthening
consonants in general, including sonorants like [n] and [m]; we usually lengthen the vowels (although final
nasals are often lengthened in Southern Min songs)
Ms. Chung’s theory: iii. They do it for stylistic reasons, to sound “cool”, reflecting a kind of personal power
- Always remember to stop at stops. Also, in the phrase “the front-back position”, the four words are all
monosyllables, which means there will be longer pauses in between. This will be talked about in detail in a future
class.
8
Oct. 15, 2012

6-8 This work is licensed by Creative Commons

9
Oct. 17, 2012
Fall 2012
11. Wrapping up Chapter One

1. Going over the exercises for Chapter One (footnotes)
2. Going over the Hanyu Pinyin assignment (footnotes & Q & A)
3. A Course in Phonetics: Chapter 2: Phonology and Phonetic Transcription
4. Homework
II. Notes
1. Going over exercises for Chapter One (footnotes)
Answers for the chapter exercises are not given in the handout; we expect future students to come up with
the answers on their own. However, you can read things Ms. Chung mentioned during the discussion and
corrections here.
a. Is technical terms a compound noun? If not, why don’t we stress both words like we do for other noun
phrases?
- A lot of noun phrases with an adjective that ends with -al or -ic have compound noun stress
à (X) technical *terms (O) *technical terms
- The reasons may be
i. the modified noun is relatively “empty” (it has a relatively generic meaning and is not semantically
strong)
ii. the adjective implies a contrast with other adjectives e.g. political vs. economic
b. We are easily influenced by writing systems; native speakers can be strongly affected by spellings while
we can be affected by Chinese characters and the Mandarin Phonetic Symbols (MPS 注音符號).
- e.g. ㄢ vs. ㄑㄧㄢ
Why do we write ㄢ for ㄑㄧㄢ although the latter sounds so different from the former? Why don’t
we write ㄑㄧㄝㄣ?
à Place of articulation: Your tongue is a bit higher for ㄧㄢ than it is for ㄢ, because ㄧ is a high
vowel (ㄧ changes the vowel quality of ㄢ).
cf. (English) tea vs. true: your tongue will be more front for tea ([i] high front vowel) and more back for
true ([ɹ] – post alveolar and [u] – high back vowel)
- Other similar questions worth thinking about:

1
Oct. 17, 2012
e.g. ㄒㄩˊvs. ㄒㄩㄣˊvs. ㄒㄩㄥˊ Why do we write ㄒㄩㄥˊwhen it actually sounds more like
ㄒㄧㄛㄥˊ [ɕioŋ] (i.e. instead of a single rounded front vowel [y] (a monophthong), we hear it
separated into the diphthong [io], rather than [ɕyŋ]? cf. (Pinyin) xú 徐 vs. xún 巡 vs. xióng 雄 à the
Pinyin suggests a totally different phonetic structure for xióng 雄 as opposed to xún 巡. In fact, xún
巡 is also somewhat diphthongal – it sounds like [ɕyən].
e.g. 博ㄅㄛˊ Why don’t we write ㄅㄨㄛˊ?
The answer: ㄅㄆㄇ and ㄈ are all labial sounds, and were considered by the designers of the system to
have inherent lip rounding, so they decided not to mark rounding with ㄨ, as is done with: ㄊㄨㄛ,
ㄉㄨㄛ, ㄋㄨㄛ, ㄌㄨㄛ,ㄍㄨㄛ, ㄎㄨㄛ, ㄏㄨㄛ, ㄓㄨㄛ, ㄔㄨㄛ, ㄕㄨㄛ, and ㄖㄨㄛ.
However, although ㄅㄆㄇ and ㄈ all involve the lips in their articulations, they do not actually have
inherent rounding; rounding depends completely on the following vowel, e.g. ㄛ. And ㄅㄛ, ㄆㄛ, ㄇ
ㄛ, ㄈㄛ actually do have an ㄨ glide before the ㄛ! So your grade school intuition was correct!
In addition, you will notice that ㄅㄆㄇ and ㄈ cannot be combined with ㄜ, like most of the other
symbols can – except in the artificially constructed symbol names ㄅㄆㄇ and ㄈ! (They were originally
called ㄅㄛ, ㄆㄛ, ㄇㄛ, ㄈㄛ, but were later called ㄅㄜ, ㄆㄜ, ㄇㄜ, ㄈㄜ so they’d be more
like ㄊㄜ, ㄉㄜ, ㄋㄜ, ㄌㄜ and the rest, even though no characters are pronounced in this way.)
Also: One student felt strongly that ㄢ must be a “diphthong” 雙母音, because it is written with a
single symbol, so in his mind it “had” to represent a “single” sound (but ㄢ [an] contains a nasal, and
diphthongs are two vowels put together into a single syllable; they can’t include a consonant)
à Once we get used to a writing system (or anything else), especially if we learned it in early childhood,
it strongly affects how we think and see the world, and it is very difficult to imagine that any other
viewpoint is possible; it is even harder to change how we think, even in the face of strong evidence.
- Taiwanese learners are less likely to be as influenced by English spellings as English native speakers are,
since most (so far) know KK – though this is now changing.
c. Dental vs. interdental:

- Dental consonants: in Mandarin ㄉㄊㄋㄌ
- Interdental consonants: in English th [θ] and [ð] àAlways say “interdental” for English, because we
need to reserve “dental” for a different articulatory position in Mandarin.
d. Vocabulary (meanings, pronunciation, etymology, etc.)

- Do not take things for granted. Make sure you check the dictionary whenever you encounter
words that you don’t know. Don’t just guess. You can use Merriam-Webster Online to check
standard American pronunciation, and the Oxford Advanced Learner’s Dictionary and Macmillan
2
Oct. 17, 2012
Dictionary for the standard British pronunciation.

- vat (n.): 一缸
- rave (v.): to get very excited over something
- rough [ɹʌf]: the –gh was originally a Germanic sound [x] (voiceless velar fricative) but has become an [f]
or a silent sound in English e.g. tough [f], eight (silent)
- dumb [dʌm]: the final b is silent!
- stomach ['stʌmək]: remember the /t/ is not aspirated!
- wad [wɑd] (n.): 一捲, 一坨, 一絡 (鈔票) a wad of bills
- gate [geɪt]: the vowel [eɪ] is actually a front diphthong
- ether ['iθɚ] (n.): 釔鈦 cf. either ['iðɚ]
- lodger ['lɑdʒɚ] (n.): 寄宿人 cf. larger ['lɑɹdʒɚ] 更大 (These two can be easily confused. For larger,
the [ɑ] is slightly more rounded, and don’t forget to pronounce the [ɹ]! In British English, it is ['lɑːdʒə],
but we should try to be consistent and stick to one variety at one time.)
- mast (n.): 桅杆
e. lull [lʌɫ] – two realizations of /l/

- before a vowelà alveolar “l” = clear “l” [l] e.g. light, love, luscious (美味多汁的)
- after a vowel à velarized (舌根化/舌後化) “l” = dark “l” [ɫ] (the tongue doesn’t necessarily make
contact with the alveolar ridge) e.g. Bill, ball, doll
- recording: the intonation for the original word goes from high to low, but if we play it backward, the
intonation will go up, which makes it sound a little bit funny (students will be able to reverse a
recording and even learn to talk backward after learning to use the software Praat – may be an
assignment); but aside from the intonation, you can clearly hear how different the two “l”s sound when
their positions are reversed.
- Professor Chao Yuen Ren 趙元任 could talk backwards, fluently! (If you don’t know who 趙元任 was,
look him up online – he is certainly one of the greatest Chinese linguists who ever lived!)
f. If you listen to Ms. Chung’s pronunciation of good, does the vowel sound like a pure monophthong to
you?
- In American English, all of the short vowels are actually diphthongs. They are all followed by a short
schwa sound [ə]. e.g. fit, bet (though it may not be as obvious with the very short vowels /ɪ, ɛ, ʌ/)
cf. The vowel [ɔ] is also followed by a little schwa or even [ɑ]; it sounds like 芋仔 in Southern Min
compressed into a single syllable (but it’s a long vowel!).
àan off-glide: the vowel that is less prominent and is at the end of a falling diphthong e.g. [aɪ eɪ aʊ
oʊ ɔɪ]
vs. an on-glide : the vowel that is less prominent and is at the beginning of a rising diphthong e.g.
[ju]
*About the phonetics that we are learning now:

3
Oct. 17, 2012
- Keep in mind that it was largely developed in British English-speaking England.

- Many things we are learning about apply well to British English, but many changes in American
English are not taken into account.
g. In English, the palato-alveolar consonant [ʒ] (as in genre) is seldom at the beginning of a word. except in
loanwords from French or occasionally other languages like Hungarian, e.g. genre. Is has long been in
English word-internally, e.g. pleasure, vision, or as a final, e.g. garage, massage, rouge.
h. In English, there is only one palatal consonant, the approximant 接近音 [j]. It is voiced, so there is no
turbulent flow of air when producing it. In words such as pure, however, [j] is affected by the preceding
voiceless aspirated [p] and it also becomes voiceless. This applies to other approximants as well.
#
Items fricatives approximant (voiced) approximant (voiceless)
Sound friction (audible turbulence) no audible turbulence friction (audible
quality turbulence)
Reasons The articulators are very close The articulators are not that Approximants by
to each other. close to each other. definition have turbulence
à The passage for the airflow à The passage for the when there are voiceless.
is very narrow. airflow is not narrow enough. In the following cases, /j/
becomes voiceless due to
assimilation 同化 by the
preceding voiceless stop
e.g. your [juɹ], lay [leɪ], pure [pʰj̥uɹ], play [pʰl̥eɪ],
ray [ɹeɪ], win [wɪn] pray [pʰɹ̥eɪ], twin [tʰw̥ɪn]
( # voiceless)
i. About the mid-sagittal head diagrams:

- When drawing mid-sagittal head diagrams for non-nasal sounds, make sure to draw the velum raised to
show that air is blocked (so the velum should touch the pharyngeal wall) from flowing into the nasal
cavity.
- When drawing diagrams for interdental sounds, do not make the tongue stick out too much.
- Keep practicing drawing the figures. Use scratch paper – or an iPad! – whatever is handy.
j. About alveolo-palatal and palato-alveolar:

- Keep in mind that they are different!
- Alveolo-palatal: in Mandarin ㄐㄑㄒ and in Polish
- Palato-alveolar: in English [ʃ] [ʒ] [tʃ] [dʒ]
k. About diphthongs:
- The more prominent vowel in a diphthong will not appear in the form of monophthong, and vice versa.
4
Oct. 17, 2012
e.g. bite (X) [bɑɪt] (O) [baɪt]
l. Waveforms: Tom saw nine wasps. (See Figure 1.17, p.32 in the textbook.)
- Pause: flat line (= silence)
- Voiceless stop: flat line (= closure = silence) + burst
- Voiceless fricative: irregular lines (= friction = noise)
- Vowel: regular vertical lines (= voiced), larger amplitude
- Nasal: regular vertical lines (= voiced), smaller amplitude (muffled)
- Voiced approximant: regular vertical lines (= voiced), even smaller amplitude
*Why is there a pause?
- Tom saw nine wasps. (monosyllable content words à all stressed)
- In English, it is not acceptable to have two stresses too close to each other.
ð We need pauses in between (English is stress-timed).
2. Going over Hanyu Pinyin assignment (footnotes & Q & A)

a. 他們 (ㄇㄣ˙) = tāmen (although in Taiwan Mandarin we often say ㄊㄚㄇㄣˊ)
b. The tone mark in ㄞㄟㄠ and ㄡ is always over the left-hand letter. e.g. āi ēi āu ōu
c. The tone mark in other diphthongs is always over the right-hand letter. e.g. ‘會’ huì
d. Rules for separating characters into words can be found on the Internet, but we will not spend time on
these or require you to learn the rules for this class.
e. For 台北, should we write Táipĕi or Táibĕi?

à Táibĕi: Hanyu Pinyin (the way we pronounce it) vs. T’ai2pei3: Wade-Giles Romanization
f. Do we have to capitalize letters for Hanyu Pinyin?

à Yes, we capitalize letters that are at the beginning of a sentence and that are at the beginning of a
proper noun.
3. A Course in Phonetics: Chapter 2: Phonology and Phonetic Transcription (p. 33)

a. There is a lot more to phonetics than simply knowing how to do phonetic transcription. To be a
phonetician, we also need to know (1) how to describe speech, (2) how speech production and speech
perception work, and (3) how languages use these mechanisms. The knowledge of phonetic transcription,
nevertheless, plays a very important and useful role in the description of speech.
4. Homework
a. Do the first of the three tutorials on voicing: links on webpage 7.
http://homepage.ntu.edu.tw/~karchung/intro%20page%2017.htm #
5
Oct. 17, 2012
b. Webpages 14 and 15 “Phonemes and Allophones”


c. Class notes due Monday, Oct. 22.
Pronunciation corrections: framed syllables are stressed; * = tonic stress #

zip [zip] à [zɪp] (vowel)
lull [lʊl] à [lʌɫ] (vowel; consonant)
mother ['mɑðɚ] à ['mʌðɚ] (vowel) cf. brother, another
robber ['ɹʌbɚ] à ['ɹɑːbɚ] (vowel: make it long enough)
leisure ['lɛʒɚ] à ['liʒɚ] (system: most Americans pronounce it this
way) cf. (BE) ['lɛʒə]
wad [wæd] à [wɑd] (vowel)
cop [kʌp] à [kɑːp] (vowel: make it long enough)
good [gʊd] à [gʊəd] (vowel: In AE, short vowels followed by a short
schwa) (see 1. f)
oral ['oʊɹəl] à ['ɔɹəl] (vowel, with r-coloring)
lateral [læt_ɹəl] à [læɾəɹəl] (the [ə] after the /t/ is not omitted; the /t/ is
between vowels ∴ it becomes a tap)
voiceless ['vɔɪslɛs] à ['vɔɪsləs] (vowel: [ɛ] is reduced in the suffix -less)
palate ['pɑlət] à ['pælət] (vowel)
trick [tɹik] à [tɹɪk] (vowel) cf. Treak (a name) [tɹik]
e.g. “Trick [ɪ] or treat [i]!”
fricative ['fɹɪgəɾɪv] à ['fɹɪkətʰɪv] (the /k/ is voiceless,/t/ is aspirated)
five [faɪ_] à [faɪv] (the [v] was missing) cf. Fie! (a Shakespearian
word)
stop [stɒp] (BE) à [stɑp] (system: please stick to General American for
this class)
alveolar [ɑl'vɪlɚ] à [æl'vilɚ] (vowels)
voiced/voiceless [vɔːɪst]/[vɔːɪsləs] à [vɔɪst]/ [vɔɪsləs] (vowel: do not make it too long)
steak [stɛk] à [steɪk] (vowel)
ton [toʊn] à [tʌn] (vowel)
both [bɔθ] à [boʊθ] (vowel)
toes [toʊs] à [toʊz] (consonant; plural –s: after voiced
soundsàvoiced [z])
fished [fɪʃd] à [fɪʃt] (consonant; regular past tense –ed: after
voiceless sounds except /t/ à voiceless [t])
6
Oct. 17, 2012
five [fʌɪf] à [faɪːv] (vowel: longer when before voiced sounds;

consonant) This one comes up very often! Please fix
it!
phonetician [fɑnə'tiʃən] à [fɑnə'tɪʃən] (vowel)
that phoneticians use no pause à pause before prepositions
in…

http://homepage.
ntu.edu.tw/~karc National Taiwan University/Karen Steffen Chung 史嘉琳
hung/intro%20pa Attribution-NonCommercial-ShareAlike 3.0 Taiwan
ge%2017.htm
http://homepage.
ntu.edu.tw/~karc
hung/intro%20pa
ge%2014.htm Attribution-NonCommercial-ShareAlike 3.0 Taiwan
http://homepage.
ge%2015.htm
5-6


7
Oct. 22, 2012
Fall 2012
12. Chapter Two: Phonology and Phonetic Transcription (I)

1. Test on Chapter One (footnotes)
2. Discussion of class notes
4. Homework
II. Notes
1. Test on Chapter One (footnotes)
a. Two ways to produce /r/ in English:
- Post-alveolar /r/ is more front
l The front of the tongue gets close to the alveolar ridge (how many English native speakers normally do
it); the body of the tongue is bunched
l e.g. car [kɑɹ]
- Retroflex: more back
l The underside of the tongue tip turns back in and gets close to the alveolar ridge or post-alveolar area or
even further back.
l e.g. 兒子 (Taiwan Mandarin < Beijing Mandarin) However, the /r/ in both of the dialects is closer to
the post-alveolar /r/ than to the retroflex. The /r/ in the Dravidian languages of Southern India is
much more retroflex.
- /r/ behaves like a lot like a vowel. à It is best described auditorily or acoustically, not articulatorily.
- The IPA symbol for the English /r/ is the upside-down “r” [ɹ].
2. Discussion of class notes

a. IPA symbols: glottal stop, tap, and trill
- [ʔ]: glottal stop (the sound you’d make if a basketball suddenly hit your tummy)
l Regardless of whether you write it by hand or type it, there should be no bar!
- [ɾ]: tap – this is not a backward version of the glottal stop symbol!
l if typed (in Sans font) à no bar; if handwritten à add a baseline bar at the bottom e.g. Spanish pero
['pɛɾo] ‘but’
- [r]: trill (tongue tip makes contact with the alveolar ridge at least 3 times) e.g. Spanish perro ['pɛro] ‘dog’
1
Oct. 22, 2012
b. The empty set (空集合) symbol Ø:

- Mentioned during the discussion on glides in Mandarin (see Unit 9)
- Actually a mathematical symbol – not a legitimate IPA symbol! It is used only in Chinese linguistics (as far
as Ms. Chung knows).
- In phonetics, it’s called a zero initial (零聲母), meaning a bare vowel that is not preceded by either a
glottal stop or glide
e.g. ‘烏來’ /Øu/ vs. [wu] vs. [ʔu] (contrastive function)
- In IPA, there is a vowel that looks similar; it is a smaller circle with a line, and it stands for a mid-high
front rounded vowel. [ø] (search for an IPA vowel chart on the Internet or look it up on the inside back
cover of the textbook) e.g. ö [ø] in Norwegian
- In IPA, there is a consonant which also looks similar and may be confused with the vowel. It is an even
smaller circle with a straight vertical line, and it stands for a voiceless bilabial fricative. To produce it,
spread your lips (like making a smile), make your lips approach to each other (without contact), and blow
air out. (we will learn about it second semester) e.g. ふ [ɸɯ] in Japanese
- If you have questions, make sure that you find an answer to them! Also, please make sure you check NTU
Phonetics – our Facebook group – often.
c. もじばけ (文字化け) ([mo̞dʑibake̞] 亂碼; もじ = 文字, ばけ = 亂七八糟的, or ‘monster’!)

- When you print out a Word file, sometimes the words/symbols do not look like what you originally
inputted, (for example, you may see little boxes instead of the symbols you inputted), or they do not appear
at all. Those words/symbols are what we call 亂碼 or mojibake.
- Solve the problem by saving your word file as a pdf file and then printing it out – although you may need
to download some special fonts for Japanese characters, etc. for the pdf to display and print correctly.
- Always proofread your notes before you hand them in. If any symbols are incorrect or missing, you can
use correction tape and write them in by hand; if you have more time, reprint the file in pdf format.
*Footnote about the diacritic in the transcription for もちばけ

- # = lower tongue position

d. “Palato-alveolar”/“Alveolo-palatal”
- When you combine these two terms together, either for 顎齦音 or 齦顎音, the former one will become
the modifier of the latter and has to change to an “-o” ending .
- à palatal + alveolar = palato-alveolar [dʒ] [tʃ]; alveolar + palatal = alveolo-palatal (ㄐ[tɕ], ㄑ[tʰɕ], ㄒ[ɕ])
e. Rules for adding an -s to words:

- Some people may need to review the rules.
- The words that -s can be added to include regular verbs ( à third person singular present tense), nouns (à
regular plurals), possessives (’s), and contractions with is, e.g. John’s going.
Please see CET article No. 8: http://homepage.ntu.edu.tw/~karchung/pubs/76_hello_et.pdf
2
Oct. 22, 2012
f. The two IPA symbols for English “a” sounds

- [ɑ] is used for the monophthong. (and mind the spelling of monophthong!)
- [a] is used only for diphthongs in General American English [aɪ] [aʊ]
- Useful tools for typing IPA: i2Speak http://www.i2speak.com/
for typing Pinyin: Pinyin Tone Tool: http://toshuo.com/chinese-tools/pinyin-tone-tool/
g. Translation for “sibilant”:
- 嘶音 – Taiwan
- 咝音 – PRC (this is now used in Taiwan, too)
- There are six sibilants in English – [s], [z], [ʃ], [ʒ], [tʃ], [dʒ]; the last two in the list are affricates.
h. [ʒ] vs. [dʒ]

- [ʒ] fricative ( = [ʃ] + voicing)
- [dʒ] affricate = a very short stop + fricative; similar to the ㄓ [tʂ] consonant in 這, which is also an affricate
i. [ə] [ɚ] [ɜ˞]

- [ə]: The schwa is not really a phoneme but a reduced vowel (the vowel was originally a full vowel). Many
vowels in unstressed syllables are pronounced as a schwa.
- [ɚ]: unstressed rhoticized (兒化) vowel e.g. (AE) teacher ['tiːtʃɚ] cf. (BE) ['tiːtʃə]
- In the KK system and in the conventional IPA representation of English, the schwa cannot be used in
stressed syllables, and therefore we use a different symbol for stressed syllables.
à [ɜ˞]: stressed rhoticized vowel e.g. bird (AE) [bɜ˞d] cf. (BE) [bɜːd]
(Standard British English is a non-rhotic dialect, in which post-vocalic (母音後的) /r/s are not pronounced,
but instead usually become a schwa, or the vowel is lengthened e.g. fire [faɪə], car [kɑː], word [wɜːd])
(You can check our Facebook group for a related discussion.)
*About rhotic (兒化韻)
- # = rhoticized
- Some people say ['ɹoʊɾɪk] while Ms. Chung sometimes says ['ɹɑɾɪk] (but it may sound funny when we say
“a rhotic”). For this class, we will say ['ɹoʊɾɪk].
j. Spellings
- 延續音’ (X) *continuent (X) continent (大洲) (O) continuant
- ‘一段話; 一個聲音’ (X) *oterin (O) utterance (to utter: 說出; 出聲)
- Although many native speakers make this kind of spelling mistakes (e.g. “independant” for independent),
you should still make sure you always use correct spelling. The red underlining in Word can help – don’t
ignore it!
3. If there are things you don’t know or are not sure about, you can (1) look it up, (2) ask Ms. Chung or the TAs,
or (3) post your questions on Facebook.
3
Oct. 22, 2012
3. A Course in Phonetics: Chapter 2: Phonology and Phonetic Transcription (p. 33)

a. Bold: Items in bold are important and often appear in tests.
b. Citation style of speech/citation form: (In Chapter Two, we will focus on the phonetic transcription of this
style of speech)
- the dictionary form; the style (very clear articulation) you use when you read a word in isolation (in careful
speech) à very different from phonetic descriptions of connected speech (see below)
e.g. a [eɪ] book à this sounds odd because function words (the a is an article 冠詞) are usually not
pronounced in their citation form.
e.g. (Mandarin) ‘程’ vs. ‘計程車’ (has been written in a paper by Ms. Chung)
- useful in language documentation 語言紀錄, fieldwork 田野調查, and lexicography 字典編纂.
*Applying what you learn

- Some DFLL students may end up better at English than Chinese in some arease! Although it is a good thing
that our English is good, still we need to maintain our native language. Much of the knowledge we are
learning now can be applied to Chinese, and other languages.
-The concept of “citation form” is a good example. If we want to have a language exchange or teach our
native language to foreigners in the future, we may pronounce words according to the dictionary when we
read them in isolation. When we say the words in a compound, phrase, or long sentence, however, we may
pronounce them in different ways (e.g. Mandarin 計程車; the initial of the second item in a three-syllable
expression is often elided.. For a study on contraction in Taiwan Mandarin, please see:
http://homepage.ntu.edu.tw/~karchung/pubs/Contractions_Chung.pdf)
c. Connected speech: used in normal conversation
4. Homework:
Do the Tutorial on Plosives (Part 1) on webpage 7.
Pronunciation corrections: framed syllables are stressed; * = tonic stress #
voiced [vɔːɪst] à [vɔɪst] (vowel: [ɔ] is very short)
stop [stɒp] (BE) à [stɑp] (system)
come [kʌŋ] à [kʌm] (the [m] was missing à close your mouth for
[m] and don’t substitute [ŋ]!)
In this *chapter, à In *this chapter, (this is stressed; it is being
contrasted with the previous or other chapters)
we will be à we will be (function word à unstressed)
concerned [kɔn'sɜ˞nd] à [kən'sɜ˞nd] (vowel) cf. confirm; see CET 86:
http://homepage.ntu.edu.tw/%7Ekarchung/pubs/C
ET86.pdf
consider [kən'sidɚ] à [kən'sɪdɚ] (vowel)
4
Oct. 22, 2012
five [faɪv] [faɪːv] (vowel: make it longer)

we will we will (function word à unstressed)
Stressed function words:

– Although function words are usually not stressed, we do stress them when they carry contrastive meaning.
e.g. The word some is not stressed when it means a few (一些), but when it is contrasted with other (and in this
case, it means certain 某些), we stress it.


http://homepage.
ge%2017.htm

5
Oct. 24, 2012
Fall 2012
13. Phonology and Phonetic Transcription (II)
1. Pre-announcement of assignments
3. Homework
II. Notes
1. Pre-announcement of assignments (see below for all the links)
a. Go to webpage 17 and do the third tutorial on VOT (voice onset time) and aspiration
b. Go to webpage 10 “Writing Chinese in IPA and the International Phonetic Association”

- Read this page carefully
- There are several schemes for transcribing Mandarin into IPA, but please pay special attention to the one
suggested by Professor Li Wen-Chao (李文肇). We will basically be using his system to transcribe
Mandarin into IPA in the future. Print out p. 3 and p. 4 (with tables of the consonants and vowels of
Mandarin) and look them over carefully. Link to Jimmy Lin’s useful comparison table:
http://homepage.ntu.edu.tw/~karchung/ChineseIPA.pdf
*What are the biggest differences between Romanization and IPA? Focus on the purposes each is typically
used for.
- Both have varying degrees of objectivity and subjectivity.
- One purpose of Romanization is to enable people who don’t read Chinese characters to access personal
and place names and other words in Chinese in a written form that they can recognize. For example, in
Taiwan, place names were formerly written mostly in the Wade-Giles system; then Tongyong was used in
Taiwan in some contexts for a while; but now Pinyin is most commonly used. But there are often many
errors, regardless of the system used, since Taiwanese are mostly unfamiliar with Romanization, and
usually don’t pay much attention to it.
- In earlier times, mostly Wade-Giles was used to spell Chinese loanwords in English; currently Pinyin is
most often used, e.g. kungfu 功夫, t'ai chi ch'uan 太極拳 vs. qi 氣, fengshui 風水.
c. Go to webpage 13 “Romanization III”

- Read the two essays by Professor Li Wen-Chao about the Romanization systems used in Taiwan.
1
Oct. 24, 2012
2. A Course in Phonetics: Chapter 2: Phonology and Phonetic Transcription (p. 33-36)

a. Broad transcription: aka phonemic transcription
- Uses phonemes rather than allophones à enclose in slashes e.g. water /'wɑtɚ/
- Assumes the reader knows the phonological rules and therefore does not provide a lot of phonetic detail
b. Narrow transcription: (applies allophonic rules; uses a phonetic rather than phonemic representation)
- Uses brackets, represents allophones à e.g. water ['wɑɾɚ]
- Will show various levels of detail depending on the purpose of the transcription. Usually we don’t want
the transcription to be too narrow, or we may end up exhausted. Also, some details are not really
necessary.
- For this class, a transcription that indicates the main phonological rules which can help a reader to
pronounce a word correctly is narrow enough.
c. A quick review of citation style of speech and connected speech:

- The citation style of speech is (1) what we find in dictionaries and (2) usually used for words uttered in
isolation.
- Connected speech is speech in context.
d. Phonology: the study of systems and patterns of sounds that occur in a language.
- Phonemes 音位/音素 vs. allophones 同位音/音值:
(1) A phoneme is a category (abstract concept/ideal); under this category there are variations (different
phonetic realizations – what one actually produces), i.e. allophones. e.g. /t/ is a phoneme, and its
allophones include [ɾ] (as in bottom) and [ʔ] (as in button).
(2) If replacing one sound with another would change the meaning of a word entirely, then the two
sounds are different phonemes. e.g. pie /paɪ/ vs. die /daɪ/. If a different word with a different
meaning does not result, then the two sounds are allophones. e.g. bit [bɪt] vs. [bɪtʰ] (unreleased 不除
阻 vs. released 除阻 final voiceless stop)
In some languages, [t] and [tʰ] may be different phonemes.
à Whether two sounds are phonemes or allophones is language-specific.
e.g. pop [pʰɑpʰ] vs. [pʰɑp̚] (aspirated/released vs. unreleased)
(In English, voiceless stops that are in word-initial position are aspirated; they are usually
unreleased when in word-final position.)
*Sound changes: perception and production

- It is very common for younger children whose first language is English to have difficulty making
the [ɹ] sound; some will say [w] instead. After they grow up, at some point they will switch to
saying [ɹ]. e.g. A child used to call Ms. Chung ['kɛwən] instead of “Karen” ['kɛɹən].
- Although these children cannot make the [ɹ] sound, they can often distinguish whether adults (or
older children) pronounce this sound correctly. If we pronounce the /r/ the same way they do, we
may get corrected. e.g. If we say write [waɪt] a letter, such a child might correct us by saying “No, I
2
Oct. 24, 2012
want you to write [waɪt] a letter.”

- For some adults, it is also common that they can hear the difference between [ɑ] and [ɔ] although
they may not distinguish them in their own speech.
*Homework: Read course webpage 15 (see below for the link)
(3) Analogy:
i. A person with different aspects to their personality #
phoneme allophones
you the different ways you talk/behave when you are with different people, e.g. your
teacher, your friends, your little sister/brother, a department store cashier, etc.
/t/ word-initial position – [tʰ]; between vowels - [ɾ]; word-final position – [t̚], etc.
ii. A family with different members: A phoneme is more like a family; it is not a single sound but a
group of sounds. It is like a family name that family members – the allophones – share.
(4) The concept of a “phoneme” is basically a phonological (the structure of sound systems) notion. We
need the concepts of phonology to do phonetics, and we also need phonetics to do phonology, though
they are separate sub-fields of linguistics.
- Allophones:
(1) Spelling is not a reliable clue for us to judge whether two sounds belong to different phonemes or not.
e.g. car vs. key: Though spelled differently, the two initial consonants belong to the same phoneme
/k/. In addition, the /k/ in car is lower and more back because of the following low back vowel [ɑ];
the /k/ in key is higher and more front because of the following high front vowel [i]. You can hear this
difference by whispering just the /k/ in these two words (difference in the second formant).
Please see Unit 10 for the discussion of formants.
(2) Other examples of allophones:
e.g. true vs. tea: the /t/ in tea [tiː] is more front (high front vowel [i])
e.g. ten vs. tenth: the /n/ in ten is alveolar; it is more interdental in tenth [tɛn̪θ] (interdental fricative
[θ] à dentalized /n/ [n̪])
e.g. pit [pɪt̚] (unreleased) vs. pity ['pʰɪɾi] (between vowels à voiced tap 閃音 [ɾ], sounds like a very
short d)
e.g. lay [leɪ] (voiced) vs. play [pʰl̥eɪ] (aspirated stop [pʰ] à voiceless [l̥ ])
e.g. listen ['lɪsən] (clear /l/) vs. pull [pʰʊəɫ] (word-final + after vowels à [ɫ] –velarized /l/; dark /l/)
These differences are due to the effects of coarticulation/assimilation 同化.
*pity: a comparison between Standard American English (AE) and Standard British English (BE)
- In AE, the short vowel [ɪ] is followed by a schwa offglide and is thus lengthened, and the /t/ is a
voiceless tap. Please see Unit 11 for the discussion of diphthongization of short vowels in AE.
(It is not very common for voiceless stops to become voiced, like the tap in English. And many other
dialects of English do not have an allophonic tap.)
3
Oct. 24, 2012
- In Standard British English, there is no schwa after [ɪ] (which makes the vowel shorter than it is in
AE), and there may be some degree of affrication in the stop /t/. à ['pʰɪtʰi]
The tap is, however, becoming more and more common in the speech of young BE speakers.
(3) Allophonic variations and speakers’ awareness

l Native speakers of any language are often not aware of the allophonic variations they have in their
speech; e.g. many of us do not notice the allophonic rules of Mandarin.
e.g. 一萬 vs. 萬一: If said carefully, our tongue will touch the alveolar ridge for the 萬 in the two
phrases, but often in connected speech, we pronounce 萬 with a nasalized vowel in 萬一 (à
[wãji]). Such allophonic variation also applies to other situations where a final n is followed by a
glide, e.g. 一萬五 [iwãwu].
e.g. 計程車 [tɕìʈʂə́ŋʈʂə̄] (citation style) vs. [tɕìə́̃ʈʂə̄] (connected speech): contraction
l The allophonic variations of Southern Min can be more complicated, especially due to tone sandhi
轉調 (the tone of a word changes when followed by different words). e.g. goa 我 in isolation is
pronounced with a high falling (41) tone; in a phrase like 我是, 我 is pronounced with a high level
(44) tone.
Most people do not know about the rules. In most Southern Min classes we have for children now,
these rules may not be taught at all, probably because many of the children already speak Southern
Min and do the tone sandhi very naturally. Mainlanders 外省人 who do not speak Southern Min
can be very confused if they have not learned the rules, and may often use incorrect tone values
when speaking Southern Min.
l It is universally the case that although we are often not aware of the allophonic rules in our native
language, we can apply them very naturally. In addition, we may unconsciously apply these
allophonic variations to other languages that we are learning, resulting in incorrect forms.
e.g. Ms. Chung once applied an allophonic rule of English –unstressed vowels are often reduced –
to Georgian when she wanted to say “arian” (in Roman letters) ['ɑɾiɑn] but said ['ɑɾiən] instead, and
was corrected by her teacher.
e.g. Many Taiwanese students will say [wʌ̃'jiɹ] for the phrase one year [wʌn 'jiɹ], applying the
allophonic rule we have (when we say 萬一) in Mandarin but that is not used in English.
- Phonemes and writing systems of languages:

(1) When designing a writing system for a language, we need to identify all of the phonemes of that
language to make the system well-organized and systematic. The phonemes serve as symbols that can
convey meanings unambiguously.
(2) Other applications:
i. Making up a new language for science fiction:
e.g. Elvish languages in The Lord of the Ring (maybe partly based on Old English)
e.g. the Klingon language in Star Trek (Some people are learning it, and a father even tried to make
his son bilingual in English and Klingon. Find these stories on the Internet!)
4
Oct. 24, 2012
*Interesting facts about bilingualism

- It is not always easy to bring up children to be bilingual. Peer pressure and society as a whole will
influence your children’s choice of the languages they want to use.
- Some children of Taiwan immigrants in the U.S. or Canada can only speak English and Southern
Min. While some children may be cooperative and go to (Mandarin) Chinese school on Saturdays,
some may end up hating Chinese if they are pushed too hard, because most of the people they are
connected with do not speak Mandarin. If they come to Taiwan, maybe to study, they may have no
trouble communicating with local people in southern Taiwan, but they will need to learn Mandarin
fast if they want to live in northern Taiwan.
ii. Regarding choice of a Romanization system, or of an orthography for an unwritten language, or

for a language that has many competing writing systems that haven’t yet stabilized:
e.g. Choosing a Romanization system for Mandarin: Tongyong Pinyin 通用拼音 (Taiwan) vs.
Hanyu Pinyin (China) à Support for Tongyong may have been due to purely political reasons.
e.g. Many of the aboriginal languages of Taiwan (which belong to the Austronesian family of
languages 南島語系) now have writing systems, but these have not yet stabilized completely.
e.g. Some languages in Southwest China, South East Asia, Nepal, and India which belong to the
Tibeto-Burman 藏緬 branch of the Sino-Tibetan family 漢藏語系 have not yet been written down
or studied. These languages need linguists to design good writing systems for them.
*Linguistic field work

- A lot of outstanding linguistic field work in unwritten languages has been done by American or
European missionaries 傳教士. An organization, SIL (Summer Institute of Linguistics), has a
great abundance of material on unusual languages, many of which were collected by missionaries.
SIL’s website: http://www.sil.org/
(3) When we design a written system for a language, we need symbols that record all and only the sounds
that can differentiate meaning – the phonemes. e.g. We do not use a separate tap symbol in English
spelling; it will not affect the meaning of the word water if we say ['wɑtʰɚ] instead of ['wɑɾɚ].
We only want as many symbols as we need to distinguish words that have different meanings, i.e. we
want an economical system to describe the language. A broad/phonemic transcription can fulfill this
goal. e.g. The writing systems of many languages of Africa are very close to a phonemic transcription
because they were created relatively recently, i.e. in the past century or so.
Some languages have undergone spelling reforms and thus have a more phonemic writing system. e.g.
Spanish.
*Spelling of Spanish
- If you know how to speak Spanish, you will probably make very few spelling mistakes since
Spanish orthography is largely phonemic; however, in Mexican Spanish. e.g. double l and y are
homophones, and are often confused.
5
Oct. 24, 2012
English has existed as an independent language since about 450 A.D. At first, English was spelled
pretty much the way it was spoken. However, while English pronunciation changed a lot after ca. 800
or 900 A.D., the writing system did not change as much.
à The spelling system of English is not very phonemic; letters are used that are not pronounced at all,
and some letters have more than one sound.
*Why have efforts to reform English spelling been largely unsuccessful?

- A conservative spelling system connects the users of the language better to their past.
cf. traditional characters of Chinese 繁體字 vs. 簡體字.
- English has many dialects. If we wanted to change the writing system, we would need to choose
one variety as the standard, which is very difficult – no variety can represent the entirety of English
if we want to have a more phonetic writing system.
- We have an emotional attachment to the current system. e.g. fight
cf. Chinese character components 偏旁 for Mandarin speakers e.g. 果 in 棵, 裸, 踝, etc.
à Keep in mind the reasons why things happen, why they change, and why they are kept.
e. Transcription of consonants:
- To find the distinctive consonant sounds of English, we can find words that rhyme 押韻: we keep the
vowel and only change the consonant.
- We can search for words that rhyme with pie and have only a single consonant at the beginning, i.e. we
can’t use spy, try, or spry – they have more than one consonant sound at the beginning! We can use tie,
die, bye, etc.
à Sets/pairs of words that only differ in one sound are called minimal sets/pairs 最小對比組/對. Look
at Table 2.1 on p. 36, and you will see a list of minimal sets for English consonants. When you are listing
words to contrast, make sure that you are not confused by the spelling!
- Some words that rhyme with pie start with two consonant letters but in fact begin with a single consonant
sound. e.g. thigh [θaɪ], thy [ðaɪ], shy [ʃaɪ]
à digraph 二合字母
- We can also use words that rhyme with pea and have only a single consonant at the beginning,
considering that some consonants do not occur in words that rhyme with pie. We can use words such as
tea, key, bee, etc. Examples are illustrated in the third column of Table 2.1.
(The letter z is pronounced differently in British/Canadian/Australian/New Zealand/South African
English: [zɛd]; in American English, it is [zi], and as far as Ms. Chung knows, the U.S. is the only place
where people say it this way.)
- There are still some consonant sounds that we cannot find in the set of words that rhyme with pea, such as
the digraph -ng as in rang [ɹæŋ], which only occurs at the end of words. The consonant sound in the
middle of the word vision ['vɪʒən] usually does not occur in word-initial position – except in French loans
(such as genre ['ʒɑnɹə]). The two words vision and mission ['mɪʃən] can form a minimal pair, among the
very few pairs of words that we distinguish through a contrast between [ʒ] and [ʃ] in English.
6
Oct. 24, 2012
- Sometimes the symbols we use in transcription are not the same as the letters we use in spelling.
e.g. cake [keɪk], city ['sɪɾi], success [sək'sɛs]
([c] is an IPA symbol that stands for a voiceless palatal stop, but we do not use it when transcribing
English sounds!)
e.g. guy [gaɪ] vs. age [eɪdʒ]
à Each IPA symbol represents only one single sound.
*Pronunciation of -cc-and -gg-

- Taiwanese people often pronounce success as [sə'sɛs], dropping the [k] sound.
- e.g. accent ['æksənt], accept [æk'sɛpt], suggest [səg'dʒɛst], etc. (However, in British English, we do say
suggest [sə'dʒɛst]!) Please see CET 81: http://homepage.ntu.edu.tw/%7Ekarchung/pubs/CET81.pdf
- We need to add a few symbols that the English alphabet does not provide us with. We will be using
phonetic symbols from the set established by the IPA, the International Phonetic Association.
*Founding of the IPA

- The International Phonetic Association was founded in 1886.
- The earliest symbols were designed by leading phoneticians from France, Germany, Britain, and
Denmark.
(Do not mix up Germany and German! Germany is the country, and German is the language people in
that country speak! So do not say “I am going to German” or “I am learning Germany”!)
- Additional IPA phonetic symbols:

l [ŋ] – angma, a velar nasal. e.g. rang
ü Writing the symbol: Do not put a bar under the vertical line on the left like [ŋ]!
l [θ] – theta, a voiceless interdental fricative. e.g. thin, three (AE), thigh, thimble, ether, breath, mouth
ü Writing the symbol: Make the symbol long and thin! If it is short, it becomes a mid-high central
vowel [ɵ].
l [ð] – eth, a voiced interdental fricative. e.g. thy, then, them, breathe
ü Writing the symbol: The symbol goes above the line of writing – it is an ascender. (So is [θ]!)
θ ð ŋ # ([ŋ] is a descender)
l [ʃ] – esh or long s, a voiceless post-alveolar (or palato-alveolar) fricative. e.g. shy, sheep, rash
ü Writing the symbol: Make the symbol go both above and below the line of writing – it is a
combination of ascender and descender.
l [ʒ] – yogh or long z, a voiced post-alveolar fricative. e.g. vision, measure, leisure; at the beginning of
foreign loans (especially French loans) as in Jean, gendarme, Zsa Zsa
ü Writing the symbol: Remember that it is a descender!
l [ɹ] – upside down r, a voiced post-alveolar approximant. e.g. rye
ü Writing the symbol: In the textbook, it is written as the regular [r] for convenience and also because
two major dictionaries of American and British English pronunciation use it. We will continue
7
Oct. 24, 2012
using the upside down r for this class as we did when using earlier editions of the textbook.
3. Homework
a. Do the third tutorial – Tutorial on VOT and aspiration on webpage 7
b. Read about writing Chinese in IPA on webpage 10 (Writing Chinese in IPA and the International
Phonetic Association
Carefully read the paper by Professor Li Wen-Chao (李文肇), print out p.3 and p.4 and bring them to
class. You will be asked to transcribe Chinese sounds using IPA.
c. Read the two essays by Professor Li Wen-Chao on webpage 13 (Romanization III)

Find out the differences/ the purposes of using Romanization and using IPA to transcribe Chinese.
d. Read about phonemes and allophones on webpage 15

@Pronunciation corrections: framed syllables are stressed; * = tonic stress #

voiced [vɔːɪst] à [vɔɪst] (vowel: [ɔ] is very short)
stop [stɒp] (BE) à [stɑp] (system)
come [kʌ̃_] à [kʌm] (the [m] was missing à close your mouth for [m]!)
In this chapter, In this *chapter, à In *this chapter, (this is emphasized as opposed to the
previous chapter and therefore is stressed)
we will be we will be à we will be (function word à unstressed)
concerned [kɔn'sɜ˞nd] à [kən'sɜ˞nd] (vowel) cf. confirm
consider [kən'sidɚ] à [kən'sɪdɚ] (vowel)
five [faɪv] à [faɪːv] (vowel: make it longer)
we will we will à we will (function word à unstressed)
For example, flat à continuation rise before punctuation
the [li] à [ði] (stick out your tongue tip for th)
others ['ɑðɚs] à ['ʌðɚz] (vowel – put your tongue more to the center of
your mouth and make it short; plural s after voiced
sounds à [z])
the_other [ði_'ʌðɚ] à [ði 'ʲʌðɚ] (linking and the transitional glide)
cf. Younger speakers now often say [ðə 'ʔʌðɚ]
usually ['judʒʊəli] à ['juʒʊəli] (consonant)
8
Oct. 24, 2012
what_we/ rushing through the à stop at stops (what [ʔ]) – remember to stop every time
not_realize words, failure to stop you see /p, t, k, b, d, g/ at the end of a word
at stops
what we don’t, flat/ what we don’t, à tonic stress and continuation rise (what we *don’t,)
the systems and rushed through à pause after the subject is finished
patterns of sounds
_that occur…
that_occur no linking à linking (when a final consonant is followed by a vowel)
it_involves
distinctive sounds, flat à continuation rise
is [ɪs] à [ɪz] (consonant)
for example, for example, à tonic stress and continuation rise (for ex*ample, )
and [æ̃] à ways that we can reduce “and”: [æn], [ən], and [n̩]
realize ['ɹɪəlaɪz] à ['ɹiəlaɪz] (vowel)
these [ðɪz] à [ðiz] (vowel)
two distinct sounds two dis*tinct sounds à two distinct *sounds (tonic stress)
belong [bə'lɔŋ] à [bi'lɔŋ], [bɪ'lɔŋ], [bə'lɔŋ] are all used by native speakers.
phonemic [fo'nimɪk] à [fə'nimɪk] (vowel: schwa)
difference ['dɪfəɹəns] à ['dɪfɹəns] (schwa elision)
white/right [waɪːt]/[ɹaɪːt] à [waɪt]/[ɹaɪt] (vowel length: make the vowel shorter
when it is followed by a voiced consonant)
single sound *single sound à single *sound (adj. + n. à stress on the noun)
For the first of these For the first of *these à For the *first of these sounds, (tonic stress on the word
sounds, sounds, with contrastive meaning; no stress on the function
words and repeated information
these [ðɪz] à [ðiz] (vowel)
lips [lips] à [lɪps] (vowel)
consonant ['kɒnsənənt] (BE) à ['kɑnsənənt] (system)
not_necessary ['nɑtʰ 'nɛsəsɛɹi] à ['nɑt̚ʔ 'nɛsəsɛɹi] (We don’t need to release the final /t/,
but we have to stop. Be careful of the timing of pause.)
could [kʊ_] à [kʊd] (the [d] was missing)
the last word you rushed through à pause before conjunctions
said_before…
Both consonants in Both consonants in à Both *consonants in this word/ between them
this word/ this *word/ (intonation; function words and repeated information
between them between them are unstressed)
different/difference ['difɹənt]/[ 'difɹəns] à ['dɪfɹənt]/['dɪfɹəns] (vowel)
change/same [tʃɛndʒ]/ [sɛm] à [tʃeɪndʒ]/[seɪm] (vowel
of [əf] à [əv] (consonant)
begin [bɪ'gin] à [bə'gɪn] (vowel)
9
Oct. 24, 2012
can [kæn] à [kən] (style – in connected speech, we usually reduce

this vowel in can)
feel [fɪɫ] à [fiɫ] (vowel)
in_a/of_a/ [in_ə]/[ əv_ə]/ à [ɪnə]/[əvə]/[ɪn'ɪŋglɪʃ] (vowel; linking)
in_ English [ɪn_'ɪŋglɪʃ]
subtle ['sʌbɾəl] à ['sʌɾəl] (the b is not pronounced in this word)
*The pronunciation of other in Taiwan English:

- The first vowel in the word is wedge (or “upside down v”) [ʌ], which is pronounced more like a low back
vowel [ɑ] in British English; in American English, on the other hand, it is closer to a mid-central vowel.
- The reason why many people in Taiwan pronounce English [ʌ] as [ɑ] (e.g. cup [kʌp]) may be due to the
previous use of the British English-based DJ system in Taiwan English teaching.
- However, the wedge in British English is not as long as the [ɑ], and therefore pronouncing other as ['ɑðɚ] is
still is not correct.
- In addition, some English reference books 參考書 published in Taiwan say that the wedge is close to the
Mandarin sound ㄚ[a], which is not correct; it is closer to ㄜ[ə].
- It takes effort to fix problems like this since some inaccurate information from previous textbooks or teachers
has taken root in our brains. If we want to sound more natural and consistent, we should stop operating on
automatic pilot and start making an effort to be extra careful of our pronunciation.
*Reduction of and
- The conjunction and is a function word and can be reduced. We can say [ænd], [æn], [ən], or [n̩], but not [æ̃].
[æ̃] is not an acceptable reduction of the word for native speakers of standard English.
Practice saying the following phrase: salt and pepper.
- When we start reducing this word, the first thing we can omit is the final /d/. If we say [ənd] – adding the final
/d/ back – it will sound a bit strange. Therefore, it’s better to pick one of these realizations.
- If non-native speakers of Chinese randomly reduce Chinese words, they will also cause a similar reaction in
native speakers of Chinese.
- No matter which language we are learning, we have to be careful in using phonetic reductions of words.
10
Oct. 24, 2012
Page Work Licensing Author/ Source

7 http://homepage. National Taiwan University/Karen Steffen Chung 史嘉琳
ntu.edu.tw/~karc This work is licensed by Creative Commons
hung/intro%20pa
ge%2017.htm
hung/intro%20pa
ge%2010.htm
hung/intro%20pa
ge%2013.htm
hung/intro%20pa
ge%2015.htm
11
Fall 2012
The textbook used for this course is: Peter Ladefoged, Keith Johnson (2011). A Course in Phonetics (6th ed.) The accompanying
handouts include only limited material from the textbook. Viewers are requested to purchase their own copy.
14. Transcribing English Sounds
(10/29/2012 class)
1. Pre-announcement about assignments
3. Homework
II. Notes
1. Pre-announcement about assignment
a. Start working on the (1) written exercises and (2) performance exercises for Chapter 2. If you encounter
questions that you don’t understand, go back to the text to find the answers.
2. A Course in Phonetics: Chapter 2: Phonology and Phonetic Transcription (p. 37-40)

a. Things to consider when transcribing English consonants:
- Transcription systems: We are using the standard IPA system. There are other styles of transcription, each
serving different purposes or being more suitable in particular circumstances, e.g. TESL (Teaching
English as a Second Language) books, phonology, etc. (Some books related to phonetics do not use IPA
at all!)
e.g.
[č] vs. [tʃ] choose, church, touch phonologically one single inseparable phoneme,
[ǰ] vs. [dʒ] judge phonetically two sounds (a stop + a fricative)
[ž] vs. [ʒ] genre
[š] vs. [ʃ] shy
[y] vs. [j] yes, yet, yeast [y] reserved for /ü/ sound in French, German, Mandarin, etc.
For the purpose of this class, it is better to use the phonetic transcription.
- The affricate [tʃ] vs. the consonant cluster [t] + [ʃ]
why choose [waɪː tʃjuz] white shoes [waɪt̚ ʃjus]
fricative onset more abrupt less abrupt
the timing of the stop and the fricative more rigid less rigid
Pause between words shorter longer ∵stop at stops
1
(We can use the ligature symbol [ ͡ ] to indicate that two phonetic sounds form one phoneme.
Ligature: binding/tying two things together.)
*Is a diphthong one sound or two?
- It depends on how we define a diphthong.
(For us, a diphthong is two vowels put together in the same syllable.)
- When we are talking about diphthongs, remember that a diphthong is composed of two elements, and
the two elements form a segment – the diphthong.
- In English spelling we also use two symbols to indicate that a vowel is a diphthong.
e.g. rate [eɪ] (vs. rat [æ]); pail
- Some phonetic systems don’t use two symbols for diphthongs.
e.g. single [e] in K.K. system
àMany people in Taiwan monophthongize 單元音化 diphthongs! e.g. cake *[kæk] or [kɛk]
- Glottal stop: Does it count as a phoneme in English? Or is it just a context-dependent allophone of a full
phoneme?
Our concern is the limited distribution of this sound in English:
(1) In American English:
l word-initial position for words with a vowel initial
e.g. flee east [fliʔist] (although sometimes [flijist] in casual speech)
vs. fleeced [flist] 除去羊毛; 騙走他人錢財
l (in casual speech) word-final position in place of or together with final [t] when followed by a word
that begins with a consonant
e.g. hit me [hɪʔmi], bat [bæʔ] or [bæʔt], cat [kæʔ] or [kæʔt]
(2) In London Cockney: in intervocalic position in words
e.g. butter ['bʌʔə], bottle ['bɒʔəl]; button ['bʌʔn̩]
cf. butter ['bʌɾə], bottle ['bɑɾəl]; button ['bʌʔn̩] (General American English)
à The glottal stop doesn’t change the meaning of words ∴ it is not a phoneme in English.
cf. ㄢ and ㄧㄢ in Mandarin: the two ㄢ s are actually allophones; the reason ㄧㄢ sounds so
different from ㄢ is that this sound has changed its quality under the influence of the ㄧ, which is a
higher sound. (See p. 4 for extended discussion.)
*London Cockney
- It is a variety of working class British English spoken in the East End of London.
- Cockney is becoming an endangered variety of English. It is often used in parodies and jokes, but
there are not that many actual native speakers. Author James Hunter is a genuine native speaker of
London Cockney. Look him up on webpage 26. (See p. 7 for the link)
*The glottal stop: a marked sound in British English in some contexts

- This sound is considered typical of Cockney, a less prestigious variety of British English, so
therefore many Brits are very sensitive to this sound and may look down on speakers who use it a
2
lot because of its association with the working class. However, it’s now becoming increasingly
common in Estuary English.
cf. Americans’ notion of how people say asks: When a person says [æks] instead of [æsks], the
listener will immediately link it to Black English, a less prestigious and sometimes even
stigmatized variety of American English, and be distracted. Social class in Great Britain is as
important as race in the U.S. People can’t help but be very sensitive to these sociolinguistic
phenomena.
*Another example of allophonic variation in English

- Adding /t/ to tense [tɛns] à [tɛnts] and to dance [dæns] à [dænts].
- A related phonetic joke: A man complained to his doctor that he kept dreaming that he was a teepee
and then a wigwam (two kinds of Native American tents). The doctor said, “I know what your
problem is – you are too tense [tɛnts] (two tents)!”
- Epenthesis: Adding an extra sound (adjective: epenthetic)
In the case of tense, a voiceless stop is inserted between a voiceless fricative and a nasal. The place
of articulation of the stop is the same with the nasal (i.e. they are homorganic).
e.g. strength [stɹɛ̝ŋkθ] (some native speakers say [stɹɛnθ], but the spelling suggests that -ng should be
(raised)
a velar nasal.)
- The contrasts between consonants: different speakers have different ideas about whether some consonants
contrast with each other.
e.g. which vs. witch; why vs. wye; whether vs. weather – The consonant sound [w] does not contrast with
voiceless [ʍ] in most forms of American and British English, but for some English speakers they are
contrastive. Actually, the voiceless [ʍ] is gone in RP, i.e. standard British English. If you hear an RP
speaker says it, the speaker may be doing it on purpose. In many regions of the U.S., with some
individual cases, people still have the [ʍ] sound. (A friend of Ms. Chung who is from Iowa has adoptesd
the /w/ vs. /ʍ/ distinction – e.g. she says which [ʍɪtʃ] one – just for fun.)
*Writing the voiceless /w/
- In KK, we write it as [hw], which may be inspired by Old English, in which wh was spelled as hw.
e.g. the first word in Old English epic poem Beowulf is hwæt, meaning “listen!”
- In IPA, we can write it as [ʍ], or [w̥].
- Things Ms. Chung brought up when going through Table 2.1, “Symbols for transcribing English
consonants” (on p.36): (excluding repeated contents)
(1) Voicing and tone: When we pronounce a voiced sound, usually the pitch of our voice will fall.
(2) The IPA symbol for the voiced interdental fricative eth [ð] is also an Old English letter.
(3) The approximant [j] is contrastive with the vowel [i] in English.
à Minimal pairs: yeast [jist] 酵母 vs. east [ist] 東邊; year [jiɹ] 年 vs. ear [iɹ] 耳朵
cf. Mandarin: Saying [iŋ] or [jiŋ] for 英 in 英國 doesn’t affect the meaning.
3
(4) The affricate [tʃ]: e.g. chai 印度茶 (from Mandarin chá 茶)
(5) Example of the sound [z]: mizzen, the third mass of a ship
- The writing system deeply affects how we think about language. Here are some examples:
(1) Average English speakers often believe that ng is two separate sounds. If someone says going
(standard) instead of goin’ (casual), they will think that the person dropped the g when the person
actually replaced the [ŋ] with an [n] .
Originally ng was not a separate phoneme; people used to say singer [ˈsɪŋgɚ] instead of [ˈsɪŋɚ]. This
also explains in part why its distribution is so limited.
(2) Native speakers also tend to think of th [θ] as two sounds. (Both ng and th are digraphs.)
(3) Some Mandarin speakers may think of the following as phonemes instead of two separate sounds:
ㄢ [an], ㄧㄢ [jɛn], ㄤ [ɑŋ], ㄧㄤ [jɑŋ]. We are highly influenced by Mandarin Phonetic
Symbols (注音符號) à make sure you look at the sounds of Mandarin phonetically.
*The allophonic realizations of [ɑ] in Mandarin

- These vowels are all allophones of the same phoneme /ɑ/ㄚ, which is a low back vowel.
(The nasals [n] and [ŋ] are two different phonemes.)
- [ɑ] à [a] / + [-n] ∴ ㄚ+ㄋ= [an] à ㄢ (low back [ɑ] to low central [a])
- [ɑ] à [ɛ] / + [j-] ∴ ㄧ+ㄢ = [ian] à ㄧㄢ [jɛn] (low back to mid-low front)
- [ɑ] à [ɑ] / + [ŋ]# ∴ ㄚ+ㄥ= [an] à ㄤ [ɑŋ]; ㄧㄤ [jɑŋ] (no change in original [ɑ]ㄚ; the
preceding palatal glide [j] doesn’t much affect the vowel here)
à Both the sounds before and after the vowel can change the quality of the vowel.
*Mandarin Phonetic Symbols: alphabet? syllabary?

- Alphabet (one symbol or digraph per phoneme) e.g. English – a, b, c, d, e, ….
- Syllabary ['sɪləbɛɹi] 音節文字 ([C]V-structure; one symbol for one or two phonemes forming a
syllable) e.g. Japanese –かきくけこ…
à Mandarin Phonetic Symbols: a mixture – ㄅㄆㄇㄈ…..ㄢㄣㄤㄥ
alphabet syllabary
(Homework: Look into the Korean writing system and decide which category it belongs to: is it an
alphabet or a syllabary?)
b. Vowel transcription
- English vowels are harder to transcribe than consonants because
(1) varieties of English differ greatly in the vowels (∴ we have to choose a dialect for transcribing);
cf. Spanish –few variations in vowels but some in consonants:
ella (she) letter c and z
Spain [eʎa] [θ]
Latin America [ˈeʝa] [s]
(2) different people have different ideas about how to represent the vowel sounds.
- Table 2.2, “Symbols for transcribing contrasting vowels in English”, on p.39
4
(1) How to read the table:
l Column 1 and 2: Column 1 for General American English (AE) vs. Column 2 for standard British
English (RP, aka BE)
*Current situation of British English
- Now fewer and fewer people speak RP, and many varieties of British English are changing at the
same time.
l Column 3 and 4: closed syllables vs. open syllables à We won’t find the short vowels in Column 4.
l Column 5 and 6: words ending with a voiced stop /d/vs. words ending with a voiceless stop /t/
l Column 7: the voiceless velar stop /k/ as initial and the voiced alveolar stop /d/ as the final sound
(2) Going through the table (Column 1-4):
(AE)
l hayed [heɪd]: how Singaporeans pronounce head (which is [hɛd] in AE and BE)
l hard [hɑɹd] (AE) as the example for /ɑ/:
It is not good to have an r in any examples of vowels, for the [ɹ] often changes the quality of the
vowel in AE, e.g. the r-coloring in the vowel of bird [bɜ˞d].
vs. BE: hard is pronounced as [hɑːd] and so it is fine to have the r. In other cases, the r may be a
schwa and thus creates a separate syllable, e.g. fire [faɪə]
à Use words like spa [spɑ], father ['fɑðɚ], or spot [spɑt] as examples.
l hod [hɑd] (AE) as the second example for [ɑ]:
In AE, vowels spelled with an o (i.e. “the short o”) are often pronounced as [ɑ], e.g. stop. The book is
using this example in contrast to the BE [hɒd], but it is confusing to have two examples for the same
symbol. In fact the vowel categories of /ɑ/, /ɔ/ and /ɒ/ have been in flux for over a century, and
phonological change is still in progress, particularly in US English, so the picture is a muddy one that
has still not, in Prof. Chung’s opinion, been satisfactorily analyzed, classified and clarified.
l [ɜ˞]: two phonemes /ɜɹ/ in AE; one single phoneme /ɜ/ in BE (no post-vocalic /ɹ/)
Also listed here but not phonemes: [ɪɹ, ɛɹ, aɪɹ]/ (BE) [ɪə, ɛə, aɪə]
l [aʊ]: example: how’d (i.e. how did, would, or how had)
l [ɔɪ]: example: void 空的、無人的
l [ju]: the only rising diphthong, with the on-glide [j] (the less prominent component as the initial of
the diphthong) cf. falling diphthongs [aɪ, aʊ, ɔɪ] with the off-glide [ɪ]/[ʊ]
(BE)
l hid [hɪd]: Don’t put a schwa after the vowel vs. AE [hɪəd] (In AE, there’s a schwa after every short
vowel!)
l [æ]: is a bit lower and more back in BE
l [ɒ]: short (not in open syllables!) and rounded vs. AE [ɑ]: long (e.g. pot) and unrounded
When saying it, make sure your mouth is open wide enough, or it may sound like [ɔ].
n [ɔ]: long and without a diphthongal quality vs. a sound like Southern Min 芋仔 elided into a single
syllable (‘taro’)
l [əʊ] vs. AE [oʊ]
l [ʌ]: short; sounds more like [ɑ], which is why many people get a false impression that [ʌ] equals to the
5
Mandarin Phonetic Symbol ㄚ [a], e.g. cup.
l [ɔɪ]: the [ɔ] may be longer than AE
l [ɪə]: only in BE; e.g. here [hɪə] two syllables vs. [hɪɹ] one syllable in AE (An American may mentally
automatically add in “r”s when they are not actually there phonetically.)
l hire: There can be a triphthong 三母音 [haɪə] or a diphthong [haə]
l hue [hju]: pretty much the same as AE, but there may be more friction as in other consonants.
(Consonants in BE tend to have more friction than in AE.)
(3) Other things we need to keep in mind:
l We are only considering the two representatives from the two major categories, i.e. not every speaker
of AE or BE speaks like this.
l When we say British English is a non-rhotic [nɑn'ɹoʊtɪk] dialect, we are referring to RP. Not all
British dialects are non-rhotic; in many parts of England you can still hear a post-vocalic /ɹ/.
l History of the post-vocalic r in BE:
trill /r/ in Old English à /ɹ/ in modern AE à dropped, - Ø, sometimes with vowel lengthening; or
schwa - in RP (which occurred only during the past few centuries)
l Something BE has but AE does not: the three diphthongs [ɪə, ɛə, aɪə]; and the monophthong [ɜ]
l About half of the population in the U.S. has merged the two sounds ɑ and ɔ.
(Homework: Read webpages 25 and 26 on the tut/taught merger. See p.7 for the links.)
l Distinction between the modal auxiliary 情態助動詞 can and the noun can 罐頭 is made in some
parts of the U.S. (mostly in the East Coast). Some people make the noun more like a diphthong
[keən].
l The IPA symbols for vowel sounds we are using mainly reflect Continental European values, i.e. the
sounds they represent in French, Spanish, and Italian;
e.g. If you see the letter “a” in Spanish, it is usually pronounced /ɑ/ (cf. English: letter “a” is often
pronounced/æ/).
Similar phonetic values are often found in other languages written in the Latin alphabet as well (e.g.
Swahili, Turkish, Navajo). à These symbols are useful in creating an orthography for newly
described languages.
* Changes in writing systems

- Turkish used to be written in the Arabic alphabet; it is now written in a Latin alphabet with some
extra symbols. Many countries with a large population of Islamic people have adopted the Arabic
alphabet as their writing system; Turkey adopted the Latin alphabet with additional diacritics in 1929
following the Turkish War of Independence in 1923.
- Uyghur 維吾爾語, the Turkic language spoken in Xinjiang 新疆, originally used an Arabic
alphabet, then in the 20th century switched to a Cyrillic and then a Latin alphabet; and then changed
back to the Arabic alphabet in 1982.
6
3. Homework
a. Work on the written exercises and performance exercises for Chapter 2.
b. Look into the Korean writing system and decide if it’s an alphabet or a syllabary.
c. Read webpages 25 and 26 about the ɑ-ɔ merger:

(See webpage 26 for more information about London Cockney.)

unfortunate [ʌn'fɔɹtʃjuneɪt] à [ʌn'fɔɹtʃjənət] (vowel)
–ate [ət] (n.)/(adj.) vs. [eɪt] (v.)
e.g. graduate ['gɹædʒjuət] (n.) vs. ['gɹædʒjueɪt] (v.)
diphthongs ['difθɔŋz] à ['dɪfθɔŋz] (vowel)
east [ɪst] à [ist] (vowel)
contrasting vowels ( contrasting vowels à contrasting *vowels (stress of adj.+ noun phrase)
flat à continuation rise before punctuation marks
in_English no linking à linking ∵ final consonant + initial vowel
sounds/lists/ [saʊnd_]/[lɪst_]/ à [saʊndz]/[lɪsts]/[ɹi'kɔɹdɪŋz] (plural –s)
recordings [ɹi'kɔɹdɪŋ_] à watch for plural endings
authorities differ_in failure to pause à pause ∵ before a preposition
their views_of
table ['tæbəl] à ['teɪbəl] (vowel)
appropriate (a.) [ə'pɹoʊpɹieɪt] (many à [ə'pɹoʊpɹiət] (vowel) vs. (v.) [ə'pɹoʊpɹieɪt] 占為己有
would say)
same approach same *approach à same *approach (stress of adj.+ noun phrase)
consonants, flat à continuation rise before punctuation marks
contrast [kən'tɹæs_] à [kən'tɹæst] (the final [t] was missing)
hud [hʌ̠d] (retracted) à [hʌd] (spread your lips a bit more)
7
6 http://homepage.ntu.ed National Taiwan University/Karen Steffen Chung 史嘉琳
u.tw/~karchung/intro% This work is licensed by Creative Commons
20page%2025.htm Attribution-NonCommercial-ShareAlike 3.0 Taiwan
6 http://homepage.ntu.ed National Taiwan University/Karen Steffen Chung 史嘉琳
u.tw/~karchung/intro% This work is licensed by Creative Commons
20page%2026.htm Attribution-NonCommercial-ShareAlike 3.0 Taiwan

英语语音学-课件

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

英语语音学-课件

Uploaded by

Copyright:

Available Formats

Sep.

2. Overview of the course

* What is “Standard American English”?

 For more details, please visit the course website

iii. Other things:

* Why is correct pronunciation important?

3. Introduction to the areas of linguistics and phonetics

ii. Subcategories of phonetics

b. Visit Elizabeth Pyatt’s webpage about linguistics for comparison (optional)

b. Have a look at course webpage 5. Vocal tract and points of articulation

Pronunciation corrections: framed syllables are stressed; * = tonic stress

Ladefoged [ˈle-də-foʊ -gəd]  [ˈlæ-də-foʊ-gəd] / [ˈlæ-di-foʊ-gəd] (Danish:

Page Work Licensing Author / Source

2. First Peek into the World of Phonetics

b. Answers and corrections

2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 2-4)

b. Fields that different phoneticians are interested in:

*About the X-ray movies

d. Read the Webpages up to 7b

articulation [ɑɹtɪkəˈleɪʃən]  [ɑɹtɪkjuˈleɪʃən] (vowel)

*Some words that Taiwanese students often confuse:

*How to do a continuation rise:

Page Work Licensing Author / Source

3. The Vocal Tract and Places of Articulation

b. How did these variations develop? How does a language change?

2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 4)

 Go over course webpage 5. Vocal tract and places of articulation

 Go over course webpages 7a. and 7b.

c. Finish links on course webpage 7a

Pronunciation corrections: framed syllables are stressed; * = tonic stress

e.g. I’m going to the *store.

*Continuation rise (also mentioned on September 12):

e.g. … the trachea, to use the more technical term

4. Speech Production and Sound Waves

What was said What was heard and why

- IPA symbol [g] not “g”

* How to distinguish between the nasals [n] and [ŋ] ?

* [n] and [ŋ] in Taiwan:

* Vowels and nasals:

2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 4- 8)

* Clarification of the terms:

Figure A: a sample wave form: “on” [ɑn]

abscissa: time → the duration and pitch of the sound

b. Weekly notes due September 24, Monday

c. Read the article on [i] and [ɪ] at

Pronunciation corrections: framed syllables are stressed; * = tonic stress

so far so far  so far (the final stress is the highest)

terms [tɜ˞nz]  [tɜ˞mz] (closure of lips when pronouncing “m”)

* How do we define a vowel whether it is a long or short vowel?

* Why must we be picky about the seemingly trivial details in pronunciation?

Page Work Licensing Author/Source

3 National Taiwan University/Karen Steffen Chung 史嘉琳

3 http://homepa National Taiwan University/Karen Steffen Chung 史嘉琳

5 National Taiwan University/Karen Steffen Chung 史嘉琳

b. The Daniel Jones (DJ) phonetic symbols

Here is a chart with the DJ pronunciation symbols:

2. A Course in Phonetics: Chapter 1: Articulation and Acoustics (p. 8-10)

b. Places of articulatory gestures (starting from p. 8)

* Two kinds of /l/:

- Ways to describe language sounds :

Pronunciation corrections: framed syllables are stressed; * = tonic stress

position [poʊ'siʃən] [pə'ziʃən] (schwa; voiced [z], not [s]

* About Canadian raising:

* Glottal stop 喉塞音: