AB Mandarin Onomatopoeia— Consonant + /l/

國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

3.1.4.5 Ranking argument and Hasse diagram of AA Mandarin onomatopoeia A

LIGN

(R,P

R

W

D

)-R>>A

LIGN

(R,P

R

W

D

)-L

A

NCHOR

B

R

-L(

SEG

),>>A

LIGN

(R,P

R

W

D

)-L A

NCHOR

B

R

-R(

SEG

) >>A

LIGN

(R,P

R

W

D

)-L A

LIGN

(R,P

R

W

D

)-R>>A

D

-B

Y

-S

EG

3.1.4.6 AB Mandarin Onomatopoeia— Consonant + /l/

In this section, I will take AB Mandarin onomatopoeia with Consonant+/l/ strucrue‘s input as a grammatical monosyllabic onomatopoeia. The second onset /l/ of the syllable is a pre-linked onset. The reason why /l/ is the pre-linked segment is not only due to its unmarkedness, may also because the /l/ pre-linked segment can be traced back to ancient Chinese.

The Emergence of the Unmarked (TETU) (McCarthy and Prince 1994a) is a common condition in the generation of languages in the world. In this analysis fixed

segmentism occurs in the output rather than copying from the base segment to

decrease phonological markedness is fall under the OT rubric of TETU. The constant occurrence of /l/ can be taken as a fixed segmentism in Consonant + /l/ group

derivation since the /l/ is a coronal consonant. That the coronal consonant is a fixed

(23) Hasse diagram of AA Mandarin onomatopoeia

A

LIGN

(R,P

R

W

D

)-R

A

LIGN

(R,P

R

W

D

)-L

A

NCHOR

B

R

-L(

SEG

) A

NCHOR

B

R

-R(

SEG

)

A

D

-B

Y

-S

EG

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

segmentism is the default choice (Paradis and Prunet 1991) according to the Place-markedness hierarchy (Prince and Smolensky 1993, Lombardi 1997). The Place-markedness Hierarchy will also take part into my examination since Clements

& Hume (1995), Alderete et al. (1999) and Yip (2001) stated that there is a tendency when choose fixed segment in reduplication. Prince and Smolensky (1993; see also Smolensky 1993) proposed a universal non-permutable constraints hierarchy when inserting a segment. She extend this hierarchy at the lower end, designating a pharyngeal (which includes laryngeal) place as less marked than a coronal.

(24) Place-markedness Hierarchy (Prince and Smolensky 1993, Lombardi 2002)

*P

L

/L

AB

,*P

L

/D

ORS

>>*P

L

/C

OR

>>*P

L

/P

HAR

This constraint ranking indicates when insert a segment, a less marked segment will be chosen first. It is to make sure that the fixed vowel in the output will be the proper optimal candidate.

In the study of ancient Chinese consonant clusters, Consonant + /l/ are the biggest and most reasonable one (Zhu 1995, Lin 1998, Li2003). Not only because the occurance of Consonant +/l/ is supported by much modern Mandarin evidence, but also because the phonological structure Consonant +/l/ follows the Sonority Sequencing Principle (SSP) and obeys the Universal Grammar. In the category of AB Mandarin

onomatopoeia, Consonant +/l/ is the biggest group. We also found that in this

category the /l/ must occur as the onset of the second syllable. Every onset of the first syllable can be /p/, /t/, /k/, /

j

/, /h/, /

x

/, etc. No matter which consonant is the onset of first syllable, it must be less sonorant than /l/. There is no AB onomatopoeia in which the onset of the first syllable is a glide while the onset of the second one is /l/. For instance, there are AB Mandarin onomatopoeias such as ‗danglang‘ (sound of a metal thing falling), ‗xili‘ (sound of raining), ‗gulu‘ (sound of drinking), ‗xonglong‘ (sound

‧

of thunder). However, there are no AB Mandarin onomatopoeias such as ‗*yili‘ and

‗*wala‘ since I assume that in ancient Chinese the consonant cluster must obey the SSP. In addition, because of the Sonority Dispersion Principle the less sonorant the first consonant is the more sonorant the second consonant is. Mandarin onomatopoeia prefers coronal /l/ over glides such as /j/, /w/ and // since in the intervocalic position a consonant is better than vocoid. Therefore /l/ became the most popular choice of second consonant not only because its higher sonority but also its [+voiced] and [-cont] features (Yip, 2001).

The input in the Consonant+/l/ group is a monosyllabic onomatopoeia. Due to the unmarked /l/ in Modern Mandarin, this monosyllabic input reduplicates its vowel right-ward as a partial reduplication with a fixed /l/ and generates the grammatical Mandarin disyllabic onomatopoeia. Given the presence of the segment-sized reduplicant vowel and the fixed segment /l/, I propose thatA

D

-

BY

-σ and *P

L

/D

ORS

are high-ranking constraints. Thus I expect that A

NCHOR

B

R

-R(

SEG

)will outrank A

NCHOR

B

R

-L(

SEG

) because of the vowel reduplication rather than a consonant one. I will examine ‗pala‘ (sound of something falling) first.

(25) ‘pala’ (Sound of something falling)

‧

candidates such as Candidate b ‗papi‘ and Candidate c ‗paka‘. However, Candidate d

‗papa‘ least violate the high-ranking constraints than candidate a. and it cannot be decided which one is the optimal candidate. In addition, Complex onset such as ‗ppa‘

is not allowed in Modern Mandarin. The reduplicant size is syllable size in Mandarin disyllabic onomatopoeia reduplication so that candidate with complex onset such as

‗ppa‘ will not be chosen as candidate in tableau in this thesis. I would like to give Alignment constraints to analyze the generation of Consonant+/l/ Mandarin onomatopoeia and rule out candidate such as papa.

(26) ‘pala’ (Sound of something falling) (Revised version)

In Tableau (26), there is no correspondent right reduplicant segment and reduplicant is aligned to the right edge of prosodic word so that Candidates b. violate dominant constraints A

NCHOR

B

R

-R(

SEG

)and A

LIGN

(R,P

R

W

D

)-R

T

and be ruled out even when candidate b. is preferred by dominated constraints such as A

NCHOR

B

R

-L(

SEG

), A

D

-

BY

-S

EG

and *P

L

/C

OR

. Candidate c. is not violate the low-ranking constraints

‧ 國

立 政 治 大 學

‧

N a tio na

l C h engchi U ni ve rs it y

A

LIGN

(R,P

R

W

D

)-L

T

and *P

L

/C

OR

. However, candidate c. align its reduplicant to the right edge of the prosodic word so be ruled out by A

LIGN

(R,P

R

W

D

)-R

T

. Candidate d.

is ruled out by constraint *P

L

/D

ORS

because of the dorsal consonant although it is preferred by the low-ranking constraint *P

L

/C

OR

. Candidate e. rule out by constraint A

LIGN

(R,P

R

W

D

)-R

T

since the reduplicant is not aligned to the right edge of the prosodic word though it is preferred by low-ranking constraints A

NCHOR

B

R

-L(

SEG

) andA

D

-

BY

-S

EG

.

The constraint ranking of AB Consonant+/l/ Mandarin onomatopoeia—

*P

L

/D

ORS ,

A

LIGN

(R,P

R

W

D

)-R

T

, A

NCHOR

B

R

-R(

SEG

)>>A

NCHOR

B

R

-L(

SEG

), A

D

-B

Y

-S

EG

,*P

L

/C

OR

3.1.4.7 Ranking argument and Hasse diagram of AB Consonant+/l/ Mandarin

在文檔中 從優選理論分析華語擬聲詞結構 - 政大學術集成 (頁 45-49)