國
立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Table 4-3-1. The values of the frequencies and residuals of the responses based on the semantic relations by chi-square test
Semantic relations Paradigmatic Syntagmatic
Frequency Observed Expected Residuals Observed Expected Residuals POS of
stimuli
Nouns 131 98.7 3.3 59 91.3 -3.4
Verbs 71 103.3 -3.2 128 95.7 3.3
Responses All 202 187
Within about 51.9% (202/389) of the responses sharing the paradigmatic relation, 64.9%
(131/202) of them are responded to the noun stimuli, while 35.1% (71/202) of them are responded to the verb stimuli. On the other hand, around 48.1% (187/389) of the responses share the syntagmatic relation to the stimuli, and 31.6% (59/187) of them are responded to nouns, while 68.4% (128/187) of them are responded to verbs. The residuals indicate that the positive associations occur between nouns and the paradigmatic relation, also verbs and the syntagmatic relation, while the negative associations present in the opposite sets.
4.4 Phonological Relation
The phonological relation will be explored by phonological similarity, syllable structure, tone structure, and syllable types in the section 4.4.1, 4.4.2, and 4.4.3 respectively.
All the 318 disyllabic words collected from the association tasks are transcribed by IPA symbols for the further phonological analyses. To demonstrate the influences of the initialness and rhyme on the Taiwan Mandarin lexicon, the responses sharing the identical segments with the stimuli will be marked. The labels include the initial consonant (C), and rhyme (R) for the syllable level analysis; and the identical syllable either in the first one
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
(1S), or the second one (2S) as well as within the word pairs (W) for the word level analysis (please refer to Appendix III for the detailed information of the phonological transcriptions and labels)28. Some examples for each label are presented below with the number indicating the identical syllable (1 for the first syllable and 2 for the second syllable):
(1) 回收 ‘to recycle’ [xwej.ʂow] > 減少 ‘to reduce’ [tɕjɛn.ʂaw] > 2C
The response [tɕjɛn.ʂaw] has the identical segment to the stimuli [xwej.ʂow] on the initial consonant of the second syllable (2C) as the sound [ʂ]29.
(2) 偵辦 ‘to investigate’ [tʂən.pan] > 偵探 ‘detective’ [tʂən.tʰan] > 1S, 2R
The response [tʂən.tʰan] has the same first syllable [tʂən] (1S), and the identical segement on the second syllable to the stimuli [tʂən.pan] as the rhyme [an] (2R).
(3) 加速 ‘to speed up’ [tɕja.su] > 速度 ‘speed’ [su.tu] > W, 2R
The response [su.tu] is associated from the stimuli [tɕja.su] with the identical syllable on the different syllable within the words (W) as [su], and they also have the same rhyme without a coda on the second syllable (2R) as [u].
4.4.1 Phonological Similarity
The identical initial consonants (C), rhymes (R), and syllables (S) between stimuli and responses are included for the syllable level analysis of the distribution based on
28 The identical rhyme shares the same (G)V(N)(D) part of the segment between words, in which G is a prenucleus, V is a nucleus, N is a coda nasal, and D is a coda consonant.
29 The dot (.) between the IPA symbols indicates the syllable boundary within a word in the section 4.4 and 4.5.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
phonological similarity of the responses by a chi-square test from 3x2 tables. The visualized distribution is presented as a bar chart in Figure 4-4-1.
Figure 4-4-1. Bar plot of the distribution of the phonological similarity of the responses
The result X2 = 2.087 (df =2) does not reject the null hypothesis (p=0.35), indicating that the possibility for the identical segments to present is unrelated to the syllable position.
Only about 33% (105/318) of the disyllabic responses are found with phonological similarity to the stimuli, in which the roughly similar amounts of that are found in syllable level (58%; 61/105) and word level (56.2%; 59/105)30. The tokens of each distribution based on phonological similarity of the responses are recorded as Table 4-4-1.
30 Fifteen responses are encoded with phonological similarity in bothsyllable level (1S or 2S) and
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Table 4-4-1. The distribution of the phonological similarity of the responses in syllable level and word level
Syllable level Word level Initial consonant Rhyme
1st syllable 12 17 14
2nd syllable 8 24 19
Within words 26
Total 20 41 59
In the analysis of syllable level, more identical initial consonants present on the first syllable (60%; 12/20) than on the second syllable; while more identical rhymes present on the second syllable (58.5%; 24/41) than the first one. In general, more identical rhymes (67.2%; 41/61) than initial consonants (32.7%; 20/61) are found between word pairs of the association tasks. On the other hand, in the analysis of word level, more identical syllables shared on the same syllabic positions (55.9%; 33/59) than on the different ones.
4.4.2 Syllable Structure and Tone Structure
Three-hundred and eighteen disyllabic responses are remained for the analysis of syllable structure and tone structure. For the former analysis, the responses are transcribed as the initial consonant (C), prenucleus (G), nucleus (V), coda nasal (N), and coda consonant (G) according to Wan’s (1999) syllable structure with a dot (.) between syllables.
Identical syllable structures between the stimuli and responses are marked as occurring in the first syllable (1S) and/or the second syllable (2S). As for the analysis of tone structure, high-level (55), high-rising (35), low-falling (21), and high-falling (51) tones are encoded to the responses; the responses with tone sandhi (TS; N=17) as underlying representation
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
of low-falling tone (21) on both syllables are excluded for the further analysis. Identical tone structures between the stimuli and responses are labeled as occurring in the first syllable (1S) and/or the second syllable (2S) as well, and analyzed only in syllable level.
(for the detailed labels for syllable structure and tone structure of the word pairs, please refer to the IPA transcription in Appendix III). An example is provided below to illustrated the labels used to mark the shared syllable and tone structure between word pairs from association tasks.
宣導 ‘to propagate’ [ɕɥɛn55 taw21] > 遵守 ‘follow’ [tswən55 ʂow21] >1S2S
The stimuli [ɕɥɛn55taw21] has the same syllable structure as the response [tswən55 ʂow21], i.e., CGVN in the first syllable (1S) and CVG in the second syllable (2S). They also share the same tone structure, i.e., high-level tone (55) on the first syllable (1S) and low-falling (21) on the second syllable (2S).
To investigate the dependency between syllable structure and tone structure in syllable level, a chi-square test with 2x3 tables is conducted. The visualized distribution is presented as a bar chart in Figure 4-4-2.
Figure 4-4-2. Bar plot of the distribution of the shared syllable structure and tone structure
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
The result X2 = 0.41786 (df =2) does not reject the null hypothesis (p=0.81), indicating that the possibility for the syllable structure to share between stimuli and responses is unrelated to the tone structure in syllable level.The numeric distributions of the syllable structure and tone structure are shown as Table 4-4-2.
Table 4-4-2. The distribution of syllable structure and tone structure of the responses Syllable structure Tone structure
1st syllable 51 57
2nd syllable 57 71
Both syllables 11 16
Total 119 144
Only around 37.4% (119/318) of the responses are encoded with shared syllable structure with the stimuli; 42.9% (51/119) of them present in the first syllable, 47.9% (57/119) of them occur in the second syllable, and 9.2% (11/119) of them are in both syllables. On another hand, around 45.2% (144/318) of the responses share the tone structure with the stimuli; 39.6% (57/144) of them present on the first syllable, 49.3% (71/144) of them occur on the second syllable, and 14% (16/144) of them are on both syllables. The similarity of syllable structure as well as tone structure present slightly more on the second syllable than the first one, and more on the same syllable than the different ones.
4.4.3 Syllable Type
Forty-two disyllabic words are primed as the stimuli in the free association tasks, and 323 disyllabic words, 15 monosyllabic words, and 68 words with more than two syllables are associated by the ten participants. Each distribution of the syllable types is shown in Table 4-4-3, and visualized as Figure 4-4-3.
‧
國立 政 治 大 學
‧
N a
tio na
l C h engchi U ni ve rs it y
Table 4-4-3. The statistics of the distributions of the syllable types Nouns Verbs All responses Stimuli
Monosyllable 7 7 14 0
Disyllable 156 162 318 42
More than disyllable 38 36 74 0
Total 201 205 406 42
Figure 4-4-3. Bar plot of the distribution of syllable types of the responses
Though there are only disyllabic stimuli, disyllabic words, monosyllabic words, and words with more than two syllables are all responded by the participants. Responses with different syllable types are distributed roughly balanced to stimuli between nouns and verbs.