CHAPTER 3 : Corpus-based Analysis
3.2 Adaptations in Violation of Mandarin Phonotactics …
3.2.2 Adaptations in Conformity to an Emergent Interlanguage Grammar
As in (20), /nʌ/ and /no/ are replaced by [no], following Korean grammar. Other dental sounds, such as /t, th, t’/ and /l/, did not incur any phonotactic violation with the [-high, +back] vowels; while the dental /n/ with [-high, +back] vowels showed 100% (31/31) tokens of phonotactic violation. The overall statistics of the adaptations following Korean grammar are shown in (21) below.
(21) Adaptations following Korean grammar Korean
CV sequence Kong-er result Number Total Percentage
/k, kh, k’, h/ + /i/ Korean grammar 51
51 100%
Mandarin grammar 0 0%
/p, ph, p’/ + /ʌ, o/ Korean grammar 20
51 39.22%
Mandarin grammar 31 60.78%
/n/ + /ʌ, o/ Korean grammar 31
31 100%
Mandarin grammar 0 0%
As shown in (21), the sequences with the velars /k, kh, k’, h/ + vowel /i/ all follow the Korean phonotactics while the sequences with labials /p, ph, p’/ + vowels /ʌ, o/ show Korean and Mandarin phonotactics. Lastly, the dental /n/ + vowels /ʌ, o/ follows Korean phonotactics in all of the tokens (100%).
3.2.2 Adaptations in Conformity to an Emergent Interlanguage Grammar
Among the phonotactic violations found in the KM Kong-er corpus, there are also some sequences that follow neither Mandarin nor Korean grammar. This type of token occurs in 39.29% (66/168) of the tokens, using a Zhuyin/English form. The sequences are
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
grouped into non-coronal consonant and vowel /jʌ/, and the coronal alveolars /s, s’/ and the vowel /e/.
3.2.2.1 Non-coronal Consonants and Vowel /jʌ/
In the previous section, we observed that the combination of the velars /k, kh, k’, h/
with the vowel /i/ follows Korean phonology, mapping to [ki]/[kh]/[xi]. However, when the [-high, +back] vowel /ʌ/ follows the [+high, -back] vowel /i/, the sequence follows neither Korean phonotactics nor Mandarin phonotactics. This condition arises because the vowel sequence [+high, -back][-high, +back] in Mandarin appears only in [jou], which is only allowed to occur after coronal consonants. Thus, the interlanguage phonology based on phonotactic violation is observed when non-coronal consonants, such as the velars /k, kh, k’, h/ and the labials /p, ph, p’, m/, combine with the vowel /jʌ/.
100% (19/19) of the syllables with the velars /k, kh, k’, h/ and the vowel /jʌ/ show outputs following interlanguage grammar. For example, ‘kjʌ’ in /kjʌ.wul/ ‘winter’ is replaced by [kjou] in the Zhuyin transcription, which follows neither Korean phonology nor Mandarin phonology. Related examples follow in (22) below.
(22) Velars /k, kh, k’, h/ + vowel /jʌ/
Korean source Kong-er in Mandarin Gloss
a. nɯ.k'jʌ nə(呢).kjou ‘feeling’
b. sam.khjʌ saŋ(桑).khjou ‘swallow’
c. hjʌn.ɕil xyŋ.ɕi(西) ‘reality’
‧
allowed in either Mandarin or Korean phonotactics.55.56% (20/36) of the tokens with the labial consonants /p, ph, p’, m/ and the vowel /jʌ/ also show phonotactic violation which belongs to interlanguage phonology. Examples are shown in (23) below.
(23) Labials /p, ph, p’, m/ + Vowel /jʌ/
Korean source Kong-er in Mandarin Gloss
a. pʰjʌŋ.seŋ phjoŋ.san(三) ‘forever’
b. tɕhi.mjʌŋ tɕhi(七).mjoŋ ‘fatalness’
In (23), consonants are faithfully adopted. The rhyme part /jʌ/ is adopted into [joŋ], which is not allowed in either Korean or Mandarin. The statistics of the overall sequence of Labials and vowel /jʌ/ is shown in (24) below.
(24) Labials /p, ph, p’, m/ + vowel /jʌ/ + (CODA) Korean
onset-rhyme structure Result of adaptation Tokens Percentage /p, ph, p’/
100% of the labial stops /p, ph, p’/ are replaced with [pjou]/[phjou]/[pjo(C)], which is not allowed in either Korean or in Mandarin. 30.43% of the tokens of the labial nasal /m/ and
‧ 國
立 政 治 大 學
‧
N a tio na
l C h engchi U ni ve rs it y
vowel /jʌ/ adopt interlanguage phonology, and the rest of the tokens of the sequence adopts [mjou] in Mandarin phonotactics.
3.2.2.2 Alveolars /s, s’/ + Vowel /e/
In Mandarin phonotactics, the vowel /e/ is not allowed to solely occur in rhyme. In the previous adaptations conforming to Mandarin phonotactics, most of the occurrences of the vowel /e/ are diphthongized to [ei]. When the alveolars /s, s’/ appear before the vowel /e/, however, 77.14% (27/35) of the tokens violate Mandarin phonotactics, following interlanguage grammar. Examples are presented in (25).
(25) Alveolars /s, s’/ + Vowel /e/
Korean source Kong-er in Mandarin Gloss a. tɕu.se.jo tshu(醋).sei.jou(優桑) ‘give (me)’
d. se.saŋ sei.saŋ(桑) ‘a sprout-PAR’
The output [sei] is mostly written in English transcription, which is not allowed in present Mandarin phonology9 and Korean phonology. In the rest of the tokens, /se, s’e/ are replaced by the retroflexed [ʂei] and [se], following both Mandarin and Korean phonology.
9 There is a historical gap between Middle Mandarin and present Mandarin. The combination of
‧
onset-rhyme structure Result of adaptation Tokens Percentage
/s, s’/ e (C)
sei (27/35) 77.14%
ʂei (7/35) 20%
se (1/35) 2.86%
An overall statistics of the interlanguage grammar is shown below.
(27) Adaptation following interlanguage grammar
Korean CV sequence Kong-er Number Total Percentage
/k, kh, k’, h/ + /jʌ/ Interlanguage grammar 19
19 100%
Mandarin grammar 0 0%
/p, ph, p’, m/ + /jʌ/ Interlanguage grammar 20
36 55.56%
Mandarin grammar 16 44.44%
/s, s’/ + /e/ Interlanguage grammar 27
35 77.14%
Mandarin grammar 8 22.86%
As shown in (27), the sequences with the velars /k, kh, k’, h/ and the vowel /jʌ/ shows that result follow the interlanguage phonotactics in every case (100%), while the sequence with the labials /p, ph, p’, m/ + vowel /jʌ/ shows variants following both interlanguage and Mandarin phonotactics. Lastly, alveolar /s/ + vowel /e/ shows 77.14% of the tokens following interlanguage phonology and 22.86% of the tokens following Mandarin phonotactics.
‧
phonotactics. In the former adaptation, five apparent phonological patterns are observed.In the latter adaptation, two types of sequences following Korean grammar and following interlanguage grammar were observed.
In Korean, tensed obstruents are allowed to occur in the onset position. However, Mandarin phonotactic system only has aspiration contrast. In KM Kong-er adaptation, 100% of the tensed obstruents are mapped to unaspirated counterparts, following Mandarin phonotactics.
The Korean codas /p, k, t, m, l/ are illicit in Mandarin phonotactics, and they are dominantly deleted 48.73% (422/866), conforming to Mandarin phonotactics. Also, some are replaced by the licit Mandarin coda [n, ŋ] (20.90%, 181/866).
In Korean, the alveolo-palatal affricates /tɕ, tɕh, tɕ’/ are not allowed to occur before the glide /j/. Opposite to Korean, the Mandarin alveolo-palatal affricates /tɕ, tɕh, tɕ’/ only occur before [+high, -back] vocoids. In KM Kong-er, 125 tokens of alveolo-palatals show glide insertion after an alveolo-palatal affricate (63.45%, 125/197). The alveolo-palatal affricates are also replaced by a retroflex counterpart of Mandarin in some tokens (28.43%, 56/197).
The Korean high back unrounded vowel /ɯ/ is mostly replaced by the Mandarin mid back unrounded vowel [ɤ/ə] (80.30%, 216/269), which also mapped to the Mandarin mid back rounded vowel /u/ (17.84%, 48/269).
Lastly, a large amount number of the Korean monophthongs /o, ʌ, e/ are