Академический Документы
Профессиональный Документы
Культура Документы
2
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
3
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
self-pace, regular, and fast, will be controlled during this study. The demographic information of
our participants will be collected to double check the tendency of different speakers.
Mandarin Tones
Chinese tones are divided into two descriptive terms (Chen, 2000, p53): pitch height /
register (high/ low), and pitch movement/ pitch contour (rising, falling, dipping etc.), and thus
render the four Mandarin tones: 55/T1 (level), 35/T2 (rising), 214/T3 (dipping), and 51/T4
(falling); plus toneless suffix (0) as a clause marker. Tones are not just personal emphasis choice
to reflect emotions and feelings. The confusions of tones can actually cause the confusions of
meanings.
1) ai
55
35
214
51
short
love
uh-hun
cancer
To note, among these tone, T3 (214) is long and deep and need more energy and time to
process. Thus, when a T3 is followed by another T3, the first T3 will change into T2 for easier
sound production.
214 35/ __ 214
Here are some word-level tone sandhi--the combination of more than two morphemes
2) Disyllabic tone sandhi
a. mi -mi mi-mi
214 214
35 214
b. gn-jn gn-jn
214 214
Mei-Mei
immediately
35 214
When there are more than two syllables within words, there will be more than one possibilities of
tone sandhi. We could either apply left-to-right, or right to left rule depends on the morphological
structures of the phrases. (Chen, 2000, p100)
Left to right: ABCXBC XZC
Right to left: ABCAZCYZC
4
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
a. li wn -m li -wn m
214 214 214
35
Liu, Wan-Mi
35 214
35 35 214
gall
(brave)
214 35 214
tiger
Right to left:
(coward)
214.214.214214.35.214
d. dng - xio - z
214
214
214
35
Party small
unit
214
35
214
35 214 0
Liu grandma
Here is an example of a polysyllabic word.
f. su -m -l -y su m -l -y
214 214 214 214
Somalia
35 214 35 214
It seems the tone sandhi in Mandarin is restricted (p102). The mechanism the tone sandhi applies
is the morphosyntactic constituents as well as the phonetic effects.
5
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
For the phrasal tone sandhi, short syntactic phrases is considered as a single lexical unit, and thus
follows tone sandhi applications in lexical units. (p490)
4) Short syntactic phrase
jing
zo
zu
214
214
214
Jiang
early leave
jing
zo
zu
35
35
214
jing
zo
zu
214
35
214
(base tone)
For long phrases, morphosyntactic junctures play a key role on tone sandhi.
5) mi -mi
gn -jn
zho-li
zh-zi
214 214
214 214
214 35
35 0
Mei-Mei
immediately
fetch
mi-mi
gn-jn
zho-li
35 214
/ 35 35
214 35
mi-mi
/ gn- jn
/ zho-li
35 214
35 214
214 35
base tone
bamboo stick
zh-zi (Sandhi tone 1)
35 0
zh-zi (Sandhi tone 2)
35 0
It seems there is the minimal rhythmic unit (MRU) mentioned in Chens (2000, p505), as a
tone sandhi unit. It can be sublexical fragment, a whole word, or a whole clause and affects
the applications of tone sandhi and thus cause its variation.
However when we are talking, does the tone sandhi and its variation follows what is listed in the
books and depends exclusively on the syntactic junctures and prosodic boundary strength?
Rationale & Research Questions
Besides the syntactic junctures, we still see the variations of pronunciations, such as
allormorphy, in across-word boundary phonological process (Wagner, 2012, p15) among
English speakers due to locality on production planning, acoustic prominence (Arnold et al,
2013), and alternative stress (Wagner & Klasen, 2015). This proposal intends to explore locality
on production planning on Mandarin tone sandhi and try to locate the variables that contribute
the variation in tone sandhi processes.
6
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
i.
ii.
mi- mi
u- r
mng
214 214
214 214
Mei-Mei
sometimes
mi- mi
u- r
mng
zho
35 35
214 35
35
214
mi- mi
u- r
mng
zho
35 214
/ 35 214
35
214
214
zho
214
214 214
Wan-Min
wn- mn (Sandhi tone 1)
/
35 214
wn- mn (Sandhi tone 2)
/
35 214
xio,
lo
214 214
214
you young, I
n
xio, w
35
214
35
214
old
lo
214
8) a. w
214
(base tone)
(sandhi tone)
(base tone)
yo
214
mi-mi
214 214
bite
Mei-Mei
yo
35
214
9) a. mn
214
35 214
zho
zh (base tone)
214
214
7
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
Min
iii.
mn
zho
zh (sandhi tone 1)
35
35
214
mn
zho
zh (sandhi tone 2)
214
35
gn-jn
yo le
214
214 214
214 0
214 214
immediately
bit
10)
Mei-Mei
gn-jn
yo le
mi-mi
35 35
214 0
35 214
214
/
214
iv.
214
gn- jn
35 214
(sandhi tone1)
yo le
214 0
35 214
11) a. mi - mi
hn-sho
ko do
214 214
214 214
214
Mei-Mei
seldom
beat/ get
me
mi - mi
hn-sho
ko do
w (sandhi tone1)
35 214
mi - mi
35 214
/ 35 35
35
hn-sho
/ 35 214
214
35
214
214
ko do
w (sandhi tone2)
35
214
35
8
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
2. If not, does the locality on production planning, speech rate, syntactic junctures, rhythmic
effects or speakers other L1 background contribute to the variations on tone sandhi
processes among speakers?
Methodology
In order to explore the tone sandhi processes in Mandarin, sentences with following
structures will be created, with three sentences in each structure. Among these three sentences,
one will have short syntactic phrases with one morpheme in each lexical unit, and the other two
will have longer syntactic phrases with two to three morphemes in each lexical unit. In sentence
structure c, d and e, there will be an extra sentence with a u morpheme or phoneme in it.
There will be 18 sentences in total and they will be consisted with most morphemes with T3 in
them.
Sentences structures:
a)
b)
c)
d)
e)
1 short + 2 long
1 short + 2 long
1 short + 2 long + 1 u
1 short + 2 long + 1 u
1 short + 2 long + 1 u
18 phrases in total
Participants:
The Chinese speakers who have been to study in the public school system in either China or
Taiwan will be our interest and 30 of them will be recruited at McGill University. They can be
either undergraduate students or graduate students with any kinds of majors. Their demography
information will be collectedage, sex, origin (which part of China), other first language (L1)
information etc.
Ways of reading:
For testing the locality in production planning in tone sandhi processes, participants will
have to do one self-paced reading, one regular reading and one fast reading. Among the three
ways of reading, regular reading and fast reading can be also used to test the effect of speech
rate.
1. Self-paced reading (one word at a time) all the sentences
2. Regular reading (one sentence at a time)all the sentences will be listed in a
different order than the self-paced reading. (Regular speed)
3. Fast reading after several times of practiceeach sentence will be completed right
after the regular reading after several times of practice. (Fast speed)
Procedure and Data analysis
9
Running head: LOCALITY IN MANDARIN TONE SANDHI PROCESSES
Participants will read the 18 sentences in three different ways and everything will be
recorded. The morphemes between words will be collected with Prat. Each morpheme will be
isolated and played individually for rating. Three other Chinese speakers at McGill with similar
backgrounds with the participants will be recruited as raters. The raters can only hear one
morpheme at a time so that they judge its tone without the effect of other morphemes before or
after it. They will rate the morphemes that they have heard as T2, T3or any other tone.
1. T1 (55) 2.T2 (35) 3.T3 (214) 4.T4(51) 5.T0 6.Not-distinguishable 6 scales
Independent variables will be ways of reading, speech rates, morphosyntactic
constituents, including words with one or two syllables, and speakers other L1 backgrounds.
The dependent variable will be the tone sandhi choices. Most tone sandhi choices will fall on
either T2 or T3. R will be used to display the results of tone sandhi with different independent
variables.
Anticipated Outcomes
The variations of tone sandhi might be affected by speakers L1 backgrounds, speech rates
(regular reading vs. fast reading) and locality on production planning (self-paced reading, regular
reading, and fast reading), and morphosyntatic constituents, including the rhythmic effect of the
words with one or two syllables.
Limitation and Future Studies
Since this study will be complete in a lab and the participants are asked to read the
sentences with different syntactic structures, thus participants natural speech cannot be captured.
We cannot infer our results to participants everyday tone sandhi processes when they are with
friends or at home. We can only draw our conclusion on participants formal speech in formal
settings.
References
Arnold, J. E. and Watson, D. G. (2015). Synthesising meaning and processing approaches to
prosody: performance matters. Language and Cognitive Processes, 30:88-102.
Chen, M.Y. (2000). Tone Sandhi: patterns across Chinese dialects. Cambridge, UK; New York:
Cambridge University Press
Wagner, M. (2012). Locality in phonology and production planning. In Proceedings of
Phonology in the 21 Century: Papers in Honour of Glyne Piggott. McGill Working
Papers.
Wagner, M. & Klassen, J. (2015). Accessibility is no alternative to alternatives. Language,
Cognition, and Neuroscience, 30:212--233