IP-final Lengthening and Musical Accent

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

89

Candidates:

a.

| q q q q |

7

q q q q |

8

qU q U qU q |

9

(

(beng-beng)FT (tiao-tiao)FT (wo shi) (xiao tu - zi )FT)FT

)

IP

b.

| q q q q |

7

j e q q |

8

q q q q |

9

(

(beng-beng)FT (tiao-tiao)FT (wo shi) (xiao tu-

zi )

FT)FT

)

IP

LO-to-MI: ALIGN-L(MN, IP),ALIGN-R(SONG,IP) >> ALIGN-R(M2N, IP)

In (154b), the final measure of the song is unaligned so it is eliminated by ALIGN-R(SONG, IP).

The intonational phrasing in (150), (153), and (154) also indicates another phenomenon, namely, an IP-final syllable is musically longer than its preceding syllable.

The following section will discuss musical lengthening at the right edge of an IP

6.3.2 IP-final Lengthening

In Mandarin Children’s songs, an IP-final syllable is usually associated with a longer musical beat, i.e., its musical duration is longer than the preceding syllable. In the database, there are 258 IPs, and 218 of their final syllables are aligned with longer musical beats, found in 84.5% of the data. The example in (155) shows the final lengthening.

(155)

' e e e e e e q '

(

(yu-dao)FT

(lu-deng)

FT

((wang qian)

FT

kai)

FT

)

IP

遇 到 綠 燈 往 前 開

meet green light toward front drive

‘Drive ahead when the traffic light turns green.’

ALIGN-L(MN, IP) ALIGN-R(SONG,IP) ALIGN-R(M2N, IP)

 a. *

b. *!

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

90

The time signature of (155) is 4/4, which means that there are four beats in each measure and a quarter note (

q

) is equal to one beat. The IP final syllable, kai, has one musical beat, which is longer than the preceding syllable, qian, which is set to a demibeat.

(156)

' e U e e e U e e ' e U e e e U e U e '

(

(lan tian)FT

(lan ya)

FT

((bai yun)

FT

bai )

FT

)

IP

藍 天 藍 呀 白 雲 白 blue sky blue FP white cloud white

‘The sky is blue and the cloud is white.’

(157)

' e U e e e U e e ' e e

e e U e U e '

(

(zheng -tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

整 天 沒 事 愛 翻 跟 頭 whole day nothing-to-do love turn somersault

‘(He) loves to somersault when (he) has nothing to do all day long.’

The time signature of (156) and (157) are both 6/8, which means that there are six beats in each measure and an eighth-note (

e

) is equal to one beat. Musical lengthening occurs here. In (156) and (157), the IP-final syllables, bai and tou are associated with three musical beats respectively, while their preceding syllables, yun and gen, are associated with one musical beat each. A constraint of sequential markedness is needed to govern the IP-final musical lengthening, as posited in (158).

(158) *n-1  n]IP

Assign one violation mark for every IP-final syllable whose musical duration is equal to or shorter than its preceding syllable.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

91

This constraint serves to rule out the illegal candidates in (159).

(159)

LO:

(

(yu-dao)FT

(lu-deng)

FT

((wang qian)

FT

kai)

FT

)

IP

遇 到 綠 燈 往 前 開。

meet green light toward front drive ‘Drive ahead when the traffic light turns green.’

Music time signature: 4/4 (There are four beats in each measure.)

| q q q q |

N q q q q |2N

Candidates:

In (159b), the duration of the IP-final syllable, kai, is equal to the preceding syllable, qian, and in (159c), kai is shorter than qian. Both violate *n-1  n]IP fatally. This constraint may interact with the constraints of musical beat assignment, as shown in (160).

a.

'

q q q

q '

(

(yu-dao)FT

(lu-deng)

FT

((wang qian)

FT

kai)

FT

)

IP

b.

'

q q q

q '

(

(yu-dao)FT

(lu-deng)

FT

((wang qian)

FT

kai)

FT

)

IP

c.

'

q q q

q U ' q

q

(

(yu-dao)FT

(lu-deng)

FT

((wang qian )

FT

kai)

FT

)

IP

*n-1  n]IP

 a.

b. *!

c. *!

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

92

(160)

LO:

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou)

FT

)

IP

整 天 沒 事 愛 翻 跟 頭 whole day nothing-to-do love turn somersault

‘(He) loves to somersault when (he) has nothing to do all day long.’

Music time signature: 6/8 (There are six beats in each measure.)

| e e e e e e

_|N

e e e e e e

_|2N

Candidates:

a.

' e U e e e U e e '1 e e

e e U e U e '2

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

b.

' e U e e e U e e '1 e e

e e U e e

'2

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

c.

' e U e e e U e e '1 e e

e U e eU e

'2

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

d.

' e U e e e e e '1

e e U e U e U e U e '2

(

(zheng -tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

LO-to-MI: NOSTRAY, *n-1  n]IP,NOSHARE-(B5) >>NOSPLIT-IC(B)

NOSTRAY *n-1  n]IP NOSHARE-(B5) NOSPLIT-IC(B)

 a. ***

b. *! ***

c. *! ***

d. *! ***

‧

The time signature here is 6/8, where an eighth-note (

e

) is equal to one beat. There is a floating beat in (160b), which is then ruled out by NoStray. In (160c), the duration of the IP-final syllable, tou, is equal to its preceding syllable gen, and thus is ruled out by *n-1 

n]IP. In (160d), the final syllable is linked to five musical beats, incurring a fatal violation of NOSHARE-(B5). Consequently, (160a) is selected as the optimal output.

The constraint ranking in (160) also selects (161a) as the optimal output.

(161)

LO:

(

(beng-beng)FT (tiao-tiao)FT (wo shi)FT (xiao (tu-zi)FT)FT

)

IP

蹦 蹦 跳 跳 我 是 小 兔 子 leap leap jump jump 1SG be little bunny ‘Caper. I am a bunny.’

Music time signature: 4/4 (There are four beats in each measure.)

| q q q q |

N q q q q |2N

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

94

The time signature here is 4/4, where a quarter-note (q) is equal to one beat. NoStray rules out (161b,) which has a floating beat. In (161c), the duration of the IP-final syllable, zi, is equal to its preceding syllable, tu, and hence is ruled out by *n-1  n]IP. As a result, (161a) emerges. The lyric line of (161) is identical to that of (154), but their mappings to music are different. However, their mappings to music are both governed by the proposed constraint rankings.

6.4 M

I

-to-M

O

Mapping: Prosody Removal and Association Faithfulness

In the mapping from the musical input to the musical output, again, the prosodic structure is removed, while syllable-beat association is faithful to the musical input.

(162) MI:

' e U e e e U e e '1 e e

e e U e U e '2

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

整 天 沒 事 愛 翻 跟 頭 whole day nothing-to-do love turn somersault

‘(He) loves to somersault when (he) has nothing to do all day long.’

Music time signature: 6/8 (There are six beats in each measure.)

| e e e e e e

_|N

e e e e e e

_|2N

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

95

Candidates:

a.

' e U e e e U e e '1 e e

e e U e U e '2

zheng - tian mei - shi ai fan gen - tou

b.

' e U e e e U e e '1 e e

e e U e U e '2

(

(zheng - tian)FT

(mei - shi)

FT

(ai fan)

FT

(gen - tou )

FT

)

IP

c.

' e U e e e U e e '1

e e U e e U e U e '2

zheng - tian mei - shi ai fan gen - tou

MI-to-MO: ID-ASSOC,*PROSST >>MAX-PROSST

The musical input is selected from tableau (160). The prosodic structure in (162b) is retained in the musical output, and thus is ruled out by *PROSST.The associated lines in (162c) are different from those in the musical input, incurring a fatal violation of ID-ASSOC. Eventually, (162a) emerges as the optimal output.

6.5 Summary

The intonational phrases in the lyric output are indicated by the punctuation marks. I have posited the constraint Align-R(PM, IP) to govern this pattern. The mapping of the intonational phrases to the musical input is keyed to the measure-to-IP alignment and the IP-final musical lengthening. I have posited two sets of constraints, as shown by the Hasse diagrams in (163-164), to govern the LO-to-MI perception grammar.

ID-ASSOC *PROSST MAX-PROSST

 a. *

b. *!

c. *! *

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

96

(163) LO-to-MI perception grammar: measure-to-IP alignment ALIGN-R(SONG,IP) ALIGN-L(MN, IP)

ALIGN-R(M2N, IP)

ALIGN-L(IP, MN) ALIGN-R(IP,M2N)

(164) LO-to-MI perception grammar: musical beat assignment NOSTRAY *n-1  n]IP NOSHARE-(B5) NOSPLIT-IC(B)

NOSPLIT-IC(B)

In (163), the top-ranking of ALIGN-R(SONG,IP) and ALIGN-L(MN, IP) ensures that the odd-number measure is left-aligned with an IP, and the final measure of a song is right-aligned with an IP. The medial ranking of ALIGN-R(M2N, IP) requires the even-number measure to be right-aligned with an IP, unless it is followed by the final measure of a song. On the other hand, in (164), the top-ranking of NOSTRAY, *n-1  n]IP, NOSHARE-(B5) and NOSPLIT-IC(B) predicts four patterns. First, no stray syllable or musical beat is allowed. Second, the IP-final syllable must be musically longer than its preceding syllable. Third, a musical beat cannot be linked to five or more musical beats.

Finally, a pair of ICs cannot be linked to different musical beats.

As discussed in Chapter 3-5, the mapping between the musical input and the musical output is governed by three constraints, ID-ASSOC,*PROSST and MAX-PROSST. The Hasse diagram is reproduced in (165).

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

97

(165) MI-to-MO production grammar

ID-ASSOC *PROSST

MAX-PROSST

The ranking of *PROSST above MAX-PROSST avoids the presence of foot structure and IP structure in the musical output. The top-ranking of ID-ASSOC requires the preservation of the musical beat assignment.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

98

Chapter 7

Music-to-Language Mapping: Tone and Term Association

7.1 Introduction

This chapter examines children’s perception of the musical melody in the singing of Mandarin songs. I observe how the children map the musical pitches to linguistic tones, and how they associate the tones to the words or phrases in their mental lexicon.

The music-to-language mapping is generalized by the schema in (166).

(166) Music-to-language mapping: tone

Singing words or phrases are mapped to the linguistic input forms with similar tonal values through the perception grammar. The children associate what they have heard to their mental lexicon through the production grammar. In particular, I examine how disyllabic musical pitches (rising, falling, and level) are transformed into linguistic tones, and how the toned disyllabic strings are interpreted and produced in the output.

7.2 Some Basics and the Proposed Corresponding Principles

In Chao’s (1930) five-scale notation system, Mandarin has four lexical tones, as Musical output

Perception grammar Linguistic input

Production grammar

Linguistic output

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

99

listed in (167), including 55, which represents high level, 35, which represents rising, 21, which represents low, and 53, which represents falling.

(167) Tone notations: numeric values T1 T2 T3 T4

55 35 21 53

Musical pitches show the highness or lowness of a sound. The database provides a variety of musical pitch combinations. For the convenience of discussion, the pitch names are converted into numeric musical notation. The pitch name of a G major scale is given in (168). For instance, the numeric musical notation for G is 1, and C refers to the middle C on the piano keyboard.

(168) G major scale and numeric musical notation conversion

In this study, I propose a corresponding scale between musical pitch and language tone, as in (169).

(169) Musical pitch to language tone corresponding scale

Pitch name G A B C D E F# G

Number

1 2 3 4 5 6 7



Music

Pitch

, 1,

^#

1 2,

^#

2, 3 4,

^#

4, 5

^#

5, 6,

^#

6 7, 1,

^#



Register [-upper] Unspecified [+upper]

Language

Tone 1 2 3 4 5

Register [-upper] Unspecified [+upper]

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

100

There are totally fifteen musical pitches in the database. The distance of each pitch is a half step (半音), which is the distance from one key to the next adjacent key on the keyboard. For example, the musical pitches,



,

1

^{, and #}

1

, correspond to 1 in linguistic tone, the musical pitches,

2

^{, #}

2

^{, and}

3

, correspond to 2 in linguistic tone, and so forth.

The music register, as proposed in (169), may affect the tonal perception. Yip (1989, 2007) proposes two binary features, [-upper] and [+upper], to characterize the tonal registers; the tone ranges from 1 to 2 is of the [-upper] register, that from 4 to 5 is of the [+upper] register, whereas 3 is unspecified. The term “register” in fact comes from music. In terms of the pitch-tone correspondence, I propose here that the musical pitch ranges from



to ^#

1

pertains to the [-upper] register, that from

4

^{to #}

i

pertains to the [+upper] register, but that from

2

^to

3

is unspecified. The [-upper] and [+upper]

registers in music correspond to the level tones. The unspecified pitch range can be mapped to either the [+upper] register or the [-upper] register.

The distance between two musical pitches is also a factor that decides the perceived tones. The larger distance between the musical pitches, the larger distance between the perceived tones. The scale in (170) shows the pitch distance that is calculated by the number of half step, namely, 0.5. Two half steps add up to one whole step, which is counted as 1 in distance. For example, the distance from pitch 1 to

5

is 3.5. The distance from a pitch to itself is 0.

(170) Distance of musical pitches

Pitch

 1

^#

1 2

^#

2 3 4

^#

4 5

^#

5 6

^#

6 7 1

^#



Distance 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

101

The distance between two language tones is shown in (171). For example, the tonal distance from 1 to 5 is 4. The distance from any tone to itself is 0.

(171) Distance of tones

The correspondence between the musical pitch distance and the tone distance is proposed in (172).

(172) Pitch distance mapping principle

Musical dist. = 0  Tonal dist. = 0 0.5  Musical dist. ≤ 3  Tonal dist. = 2 Musical dist.  3.5  Tonal dist. = 4

The 0 music distance corresponds to 0 tone distance. When the musical pitch distance is equal to or more than 0.5 but less than or equal to 3, the corresponding tone distance is 2. In Mandarin, 11 and 22, whose distance is 1, are usually perceived as low both. However, 11 and 33, whose distance is 2, are not regarded the same. When the musical pitch distance is equal to or more than 3.5, the corresponding tone distance is 4. This is because there is no limit for musical pitch distance, but the tone distance is at most 4 in Mandarin (and other Chinese dialects).

7.3 Data Design

The data are designed to examine two questions of the music-to-language mapping. First, how is musical pitch interpreted as linguistic tone by children? What

Lg tone 1 2 3 4 5

Distance 1 1 1 1

‧

is the relationship between pitch-tone faithfulness and term association?

The data are collected from normal-hearing preschool children, aged from four to six. The children live in Taipei and their parents are all native Mandarin speakers.

There are totally six children participating in this project. Three of them are female and three are male.⁵ All of the children can precisely repeat Mandarin tones and musical melodies, which indicates that they can correctly perceive Mandarin tones and musical melodies.

The children are asked to listen to the singing of disyllabic terms. The words that children are familiar to are animals, body parts, food and drink, natural vocabularies, common verbs, etc., mainly taken from Liu & Chen’s (2015) study that evaluates children from sixteen months to thirty-six months. The evaluation inventory of Liu &

Chen (2015) is constructed on the basis of MCDI (Mandarin-Chinese Communicative Development Inventories, Taiwan) proposed by Liu & Tsao (2010). Some examples of the disyllabic terms in the present study are given in in (173).

(173) Examples of the familiar disyllabic words/phrases Familiar words/phrases familiar or unfamiliar with. The words or phrases are set to different combination of musical pitches from (173). Each syllable is set to one musical pitch, whose duration is one beat. The syllables are sung by a female native Mandarin speaker with an

5 In spite of the fact that Lin (1968) finds that there are no significant sex differences in language development of preschool children, both sexes of children are included in this study.

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

103

electric keyboard. The singer is well-trained in music and singing so that the musical pitches are produced accurately. The children are asked to tell the possible lyrics that are sung by the singer. The following are some examples.

(174) Examples of the singing words/phrases

The singing output in (174a) display a falling pitch, whereas that in (174b) shows rising pitch. The pitch contour in (174c) is level.

7.4 Musical Pitch to Linguistic Tone Mapping

This section discusses the mapping from the musical pitch to the linguistic tone, and children’s association of the words or phrases to their mental lexicon. I propose a mapping model in (175).

Pitch contours Singing outputs

a. Falling

bing

^#

5 gan

^#

1 ‘cookie’ 餅乾

b. Rising

tai

^#

1 yang

^#

5 ‘sun’ 太陽

c. Level

shi

^#

5 jie

^#

5 ‘world’ 世界

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

104

(175) Musical pitch to linguistic tone mapping: associated terms Musical pitch output

Linguistic tonal input

(lexical) Term assoc. fails Term assoc.

Initial output Final output (faithful) (actual words) Type 2

(postlexical) Term assoc.

Final output (actual words/phrases) Type 1

Two types of operations in term association are observed. Some of the children unexpectedly produce an initial output and then correct that to an actual word, which is referred to the type 1 operation. In the type 1 operation, term association fails at the lexical level, where unknown initial outputs are produced, but succeeds at the postlexical level, where actual words or phrases are produced. In the type 2 operation, term association succeeds at the lexical level, and actual words or phrases are directly produced. The children can mostly produce the actual terms directly, but there are some cases where they produce initial outputs first. The table in (176) presents some statistics.

‧

(176) Statistics of the associated terms

There are 217 tokens of associated terms in the database. Eight (3.7%) of them are produced through certain initial outputs to the final outputs, which pertains to the type 1 operation. 209 (96.3%) of them are directly produced as the final outputs, which pertains to the type 2 operation. In spite of the relative small number of the tokens in the type 1 operation, the fact that the tonal contour of the initial output is faithful to the singing pitch contour makes possible the hypothesis that the initial output is identical to the linguistic input. The existence of the initial output also supports the fact that there is perception in the music-to-language mapping.

However, some of the familiar words are not associated by the children. The unassociated terms will be discussed in Chapter 8.

7.4.1 Type 1 Operation

Based on the singing, some of the children may produce an initial output and then modify it into the final output. The following are some examples

(177) Examples of the type 1 operation Tokens Percentage

Type 1 8 3.7%

Type 2 209 96.3%

Total 217 100%

Musical output Initial tonal output Final tonal output

a. bing^#

5 gan

^#

1

‧

The initial outputs are faithful to the musical output, whereas the final outputs are the words or phrases that the children associate to their mental lexicon. In (177a), the musical output is bing^#

5 gan

^#

1, which is identical to the initial tonal output bing55 gan21. However, the tone of the final output, bing21 gan55 is totally different from bing

^#

5 gan

^#

1. The fact that the initial tonal output is totally faithful to the musical

pitch output supports the argument that the initial tonal output is the same as the perceived linguistic tonal input. Consider the pitch distance in (178).

(178) Musical pitch distance and tonal distance c. shu^#

1 bao

^#

5

Musical output Linguistic tonal input

a. bing^#

5

^gan#

1

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

107

Given the pitch distance mapping principle in (172), the musical pitch distance of bing^#

5 gan

^#

1 in (178a), is 3.5 while the corresponding tonal distance of the

linguistic tonal input, bing55 gan11, is 4, where the musical pitch distance is preserved. The disyllabic tonal contour of the tonal input, bing55 gan11 is falling, which is identical to the pitch contour of the musical output. Since the children fail to associate the linguistic input to their mental lexicon at the lexical level, the linguistic tonal input is produced as the initial output, where 11 is adjusted as 21. The children then successfully associate the term to their mental lexicon at the postlexical level, where they produce the actual NP, bing35 gan55, ‘cookie’. The pitch distance in (178b-c) is preserved in the same way.

In (178d), the musical pitch is level. Mandarin has a high tone, 55, and a low tone, 21, which is like 11. The music register may affect the mapping of level tones.

The music register of shi^#

5 je

^#

5 is [+upper], which corresponds to the [+upper] tone

register of shi55 jei55 according to the corresponding scale in (169). Since the children fail to associate the linguistic input shi55 jei55 to their mental lexicon at the lexical level, shi55 jei55 is produced as the initial output. At the postlexical level, the children successfully associate the initial output with a familiar term, shi53 jei53

‘world’, and the final output is produced.

7.4.2 Type 2 Operation

In most cases, the children may directly produce the actual words as their final outputs, without producing the initial tonal outputs. Given the fact that the initial tonal output is faithful to the musical pitch output in the type 1 operation, we can posit a linguistic tonal input, which is faithful to the musical pitch output as well. I also posit the faithful linguistic tonal input for the type 2 operation. The following are some

‧

(179) Examples of the type 2 operation

Musical output Linguistic tonal input Final tonal output

a. bing^#

5 gan

^#

1 ‘cookie’ 餅乾

At the lexical level, the linguistic input is successfully associated to an actual word, whose tone is often not identical to the linguistic input. The final outputs are words that the children are familiar with. The associated words are sometimes not the same as the original singing output. For example, the singing output in (179e) is la5 ba5

‘raise’ (拉拔), but the associated word is la21 ba55 ‘trumpet’ (喇叭).

7.5 M

O

-to-L

I

Mapping: Pitch Perception

In both types of operations discussed above, I posit a linguistic input that corresponds to the musical pitch output. This section discusses the perception grammar in the mapping from the musical output (MO) to the linguistic input (LI) through three kinds of disyllabic pitch contours, falling, rising and level.

7.5.1 Falling Contour

The disyllabic tonal contour of the linguistic input is basically faithful to the musical pitch output. For example, the falling musical pitch ^#

5-

^#

1 may be mapped as

the tonal strings 55-11 or 55-33. Four constraints are given in (180-183).

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

109

(180)ID-CTR:

Assign one violation mark for every pitch contour in S2 that is not identical to its correspondent in S1.

(181)MAX-DIST:

Assign one violation mark for every pitch distance in S2 that does not have a correspondent in S1.

(182)No-DIST=1:

Assign one violation mark for every pair of linguistic tones whose distance is equal to 1.

(183)No-DIST=4:

Assign one violation mark for every pair of linguistic tones whose distance is equal to 4.

In (180), S2 can refer to the linguistic input, while S1 to the musical output. The constraint, ID-CTR,in (180) is proposed to ensure the faithful correspondence of the pitch contour. The mapping to the linguistic input is also affected by the pitch distance of the musical output; specifically, the pitch distance in the musical output is preserved in the linguistic input, a fact that can be governed by the constraint MAX-DIST in (181). The constraints No-DIST=1 in (182) and No-DIST=4 in (183) indicate markedness on the tonal distance of the disyllabic output tones. ID-CTR, No-DIST=1,andMAX-DIST must be ranked above No-DIST=4 to preserve the contour shape and the pitch distance but prevent mappings like [^#

5-

^#

1 → 55 44], etc. The

partial constraint ranking is given in (184).

(184) MO-to-LI partial constraint ranking:

ID-CTR,No-DIST=1,MAX-DIST >>No-DIST=4

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

110

Tableau (185) shows the candidate evaluation.

(185) MO-to-LI mapping

Musical output: tian^#

2-jia

^#

1 ‘heaven’ 天家

⁶ Musical pitch distance: 1

Tableau (185) shows that the musical output pitch tian^#

2 jia

^#

1 can be mapped into

two possible linguistic tonal inputs, tian55 jia33 and tian33 jia11. (185c) is ruled out by ID-CTR because its musical pitch contour is rising whereas the linguistic input contour is falling. The tone distance in (185d) is ruled out by No-DIST=1 and MAX-DIST. According to the pitch distance mapping principle in (172), if the pitch distance in music is between 0.5 and 3, its corresponding tonal distance is 2. In this case, the tonal distance in (185e) is not preserved, and thus is ruled out by MAX-DIST.

Tableau (185) shows that the musical output pitch tian^#

2 jia

^#

1 can be mapped into

two possible linguistic tonal inputs, tian55 jia33 and tian33 jia11. (185c) is ruled out by ID-CTR because its musical pitch contour is rising whereas the linguistic input contour is falling. The tone distance in (185d) is ruled out by No-DIST=1 and MAX-DIST. According to the pitch distance mapping principle in (172), if the pitch distance in music is between 0.5 and 3, its corresponding tonal distance is 2. In this case, the tonal distance in (185e) is not preserved, and thus is ruled out by MAX-DIST.

在文檔中 英語與華語歌曲中語言與音樂之關係：音段、節奏與聲調 - 政大學術集成 (頁 97-0)