本研究探討北四縣客語的母語人士在單字詞、協同發音以及變調情境中如何產出該語言的字調。研究動機有二，首先是有鑑於過去之研究在北四縣客語字調系統的描述方面十分不一致，這是因為前人究取徑不同、或是處理聲學資料的程序不同所造成；其次是，鮮少研究探討北四縣客語鄰近字調的互動以及變調的特質。
利用六位男性與六位女性北四縣客語母語人士所產出的單字調與二字調做為聲學分析以及統計分析的語料、同時比較多人次語料正規化分析方法的適切性。本研究發現對數z-分數正規化方法最能有效降低原始基頻資料中跨語者之間的基頻變化，於是我們將此正規化方法應用到本研究所有的基頻資料，以求最佳的正規化結果、並正確了解北四縣客語的字調聲學特質。同時我們也提出多項式模型與統計模型混合的研究方法，提供多項式係數做為標示字調的平均基頻音高、線性斜率、以及字調曲率的指標，以期更清楚區辨連讀變異動因。
在字調呈現的議題方面，為利於比較，本研究利用石峰（1990）的T-值正規化方式計算北四縣客語六個單字調的調值，所得的結果和張月琴（1995）的聲學研究結果除了陽入調之外，其餘調值均一致。本研究的陰入對比陽入的調值是52對比55、比起張月琴（1995）的52對比53更具有差異識別性。
在協同發音的相關議題發現：（1）本研究語料顯示北四縣客語的協同發音有順向及逆向兩個方向，而且正如大部份前人的研究所示，順向協同發音是同化作用多於異化作用、而逆向協同發音是異化作用多於同化作用。（2）然而，有別於其他語言的地方是，北四縣客語位於後字的降調，如T2 [51]、T4 [52]、T5 [31]，傾向受到前字調T1 [35]高調尾的同化影響而使整體調高提高，但後字的T1 [35]卻是受到前字的高平調T3 [55]及T6 [55]異化影響而調高變低。基於（2）的發現，因此，（3）曲折調受到的影響大於平聲調，因此曲折/平聲調的不對稱，而非像其他語言是高調/低調上的不對稱；（4）在逆向協同發音方面，北四縣客語的模式和其他聲調語言的發現大致一致，在較佳目標調與較佳促發調的字調上也是呈現高調/低調的不對稱性，例如，具有低調特質的中降調T5 [31]位於後字、是較佳促發調、能將前字調的高調部份之音高、透過逆向異化作用而將之提高、因此前字調的高調部份是較佳目標調；（5）至於字調受到協同發音影響的程度，不必然是順向高於逆向，而是因不同字調而定；（6）為正確詮釋字調協同發音的現象，不單語音的因素、音韻的動機也該一併考量在內。
最後，我們探討構成北四縣客語變調的語音動機以及變調後的陰平T1 [31]與變調的目標調陽平T5[31]是否完全混同？本研究主張北四縣客語共時的陰平變調模式是綜合不同協同發音的歷時發展的音韻化結果。其一，前字上升陰平的高調尾透過順向協同發音的過程、提高後字調整體調高、但同時也因自身的基頻波峰延遲作用的影響（F0 peak delay effects）弱化了本身的上揚調形；其二，在另一上升陰平調T1 [35]之前、已弱化的前字上升陰平調受到後字陰平調在調形上的逆向異化作用影響、而逐漸變為下降的中降調；其三，已弱化的前字陰平調、在舒聲高平調 T3 [55]、以及入聲高平調 T6 [55]之前，受到後字兩個高平調在音高上的逆向異化作用而使得前字陰平調的整體音高大幅降至最低點。最後，變調產生的位置─詞首，是韻律上較弱的位置，出現在詞首的字調較易受到後字調的影響而變異較大，而且在詞首位的音節音長較短、無法提供足夠的時間讓陰平調完整產出調尾的高調。另外要探討的變調議題是，在變調陰平調T1 [31]與陽平調T5 [31]的調形是否完全混同的議題上，發現不論是兩種字調的基頻調形或是速率調形都顯示：兩種字調在統計上均有顯著的不同；因此北四縣客語的變調是屬於漸變式的類型，而非調類取代式的類型。.


關鍵字: 北四縣客語、字調表徵、單字調、字調協同發音、變調、字調混同、多項式模型

This dissertation investigates how native speakers of Northern Sixian (NS) Hakka produce the six lexical tones in isolation, in tonal coarticulation, and in tone sandhi, motivated by the disagreement in the descriptions of the tone system of NS Hakka among previous studies due to differences in research approaches adopted or procedures for processing acoustic data, and to the sparsity in studies of interaction between adjacent tones and the nature of tone sandhi in NS Hakka.
To achieve our research goal, methodologically, tones on mono-syllabic and disyllabic words produced by 6 male and 6 female middle-aged native speakers were collected for acoustic and statistical analysis. For the speech corpus, we addressed the issue of efficacy of F0 normalization strategies for speaker variability. The study found a log z-score normalization method optimal in reducing F0 variation across speakers from the raw data, and this normalization strategy was applied to converting all the raw F0 data for acoustic and statistical analysis into log z-score values so as to attain an accurate understanding of the acoustics of tones in NS Hakka. Meanwhile, we also proposed a mixed method of polynomial and statistical modeling, which affords acoustic indices for the mean F0, linear slope, and curvature of tonal contours, and estimate differences in these aspects between tones.
Regarding the issue of tonal representations for the tone system of NS Hakka, our data on the T-value normalized (Shi, 1990) F0 values for the isolated tones showed that the tone values calculated in this study for the six individual tones are highly consistent with those suggested by Chang (1995) to the exception of the checked Yang Ru tone. The study clarified the confusing tonal contrast between Yin Ru and Yang Ru found in Chang (1995), who annotated the contrast as [52] vs. [53], by proposing the contrast as [52] vs [55], a contrast of more differential distinctiveness.
With respect to the issues associated with tonal coarticulation, the study found that: (1) both carryover and anticipatory tonal coarticulation were attested for NS Hakka with the former predominantly assimilatory and the latter dissimilatory, as reported in most previous studies for other tone languages; (2) however, unique to NS Hakka, word-final falling tones, such as T2 [51], T4 [52], T5 [31], were susceptible to assimilatory carryover effects primarily triggered by a word-initial rising T1, whereas the word-final rising T1[35] was prone to dissimilatory carryover effects induced by the two high level T3 [55] and T6 [55] in the word-initial position; based on (2) therefore, (3) a Contour/Level asymmetry was suggested regarding better effect undergors for carryover coarticulation, a pattern different from the High/Low asymmetry reported in previous studies on tonal coarticulation for other tone languages; (4) in terms of anticipatory coarticulation, patterns were quite consistent with findings from the past studies for other tone languages in that a High/Low asymmetry was found with respect to the better effect triggers, i.e. the low onset of a word-final mid-falling T5 [31] triggering a pitch raising effect on the word-initial tones, and with respect to better effect undergoers, i.e. the high pitch target of a word-initial tone that was particularly raised due to the dissimilatory anticipatory effects from the word-final T5 [31]; (5) the effect size associated with different coarticulation directions depended on the specific tones. Carryover effect size was not necessarily bigger than the anticipatory effect size; and finally (6) to make accurate interpretation of tonal coarticulation, not only phonetic groundings but also phonological motivations should be taken into account..
Finally, the acoustic study on tone sandhi in NS Hakka addressed the phonetic motivation underlying the pattern of the Yin Pin T1 sandhi in NS Hakka, and the completeness of tonal neutralization of the sandhi T1 [31] with the lexical T5 [31], to which T1 is alleged to turn in the sandhi contexts. We proposed that the synchronic phonological tone sandhi pattern in NS Hakka was a phonologized result of phonetic interactions between the word-initial Tone 1 [35] and the following sandhi context tones. First, a reduced rising form of the word-initial T1 was induced due to a high offset loss when the high offset raised the pitch level of the following sandhi context tones, and in the meantime, it was absorbed or realized in the following sandhi context tones. Second, in the T1T1 sequence, a contour dissimilatory effect was on the word-initial T1 in a reduced form by the following T1, whereby an already-reduced rising T1 in the word-initial position gradually turned a mild mid-falling tonal contour. Third, when T1 preceded the two high level T3 [55] and T6 [55], a pitch level dissimilation occurred on the realization of T1 in reduced form by the two following high level tones; in other words, a dissimilatory effect maximally lowered the pitch of the word-initial T1 to the lowest pitch range. Finally, the word-initial position, where tone sandhi occurs, is “prosodically weak,” implying that tones in this position tend to be affected by the word-final tones, and also leaving less time to fully realize the high offset target of the word-initial T1. Regarding the issue of completeness in neutralization, the study found the nature of tone sandhi in NS Hakka was gradient instead of categorical given the statistically significant differences in the realizations of tonal F0 contours and velocity contours between the sandhi T1 [31] and the lexical T5 in the same sandhi contexts.


Key words: Northern Sixian Hakka, tonal representations, isolated tones, tonal
coarticulation, tone sandhi, tonal neutralization, and polynomial modeling

Tables of Contents
Dedication ii
摘要 iii
ABSTRACT vi
Acknowledgements xi
Tables of Contents xii
List of Tables xv
List of Figures xix
Chapter I: Introduction 1
1.1 Motivation 2
1.2 The Issues 5
1.3 Introduction of Northern Sixian Hakka 6
1.3.1 Background Information 6
1.3.2 The Sound System of Northern Sixian Hakka 8
1.4 Outline of the Dissertation 12
Chapter II: Literature Review 14
2.1 Tones in Isolation 14
2.1.1 Significance of Studies in Tones in Isolation 14
2.1.2 Prior Studies on the NS Hakka Tones in Isolation 15
2.2 Tones in Context 24
2.2.1 Tonal Coarticulation 24
2.2.2 Tone Sandhi 28
2.2.3 Prior Studies on NS Hakka Tonal Coarticulation 31
2.2.4 Prior Studies on NS Hakka Tone Sandhi at Disyllabic Word Level 33
2.3 Research Questions 35
Chapter III: Methodology 38
3.1 Materials 38
3.1.1 Monosyllabic Words 38
3.1.2 Disyllabic Sequences 40
3.2 Participants 41
3.3 Instrumentation and Procedure 42
3.4 Data Processing and Measurements 46
3.5 Normalization of tonal F0 contours 47
3.6 Polynomial and Statistical Modeling of NS Hakka Tones 55
Chapter VI: Acoustic and Statistical Analyses of Isolated Tones in NS Hakka 63
4.1 Tonal Normalization and Representations 66
4.1.1 Results of F0 Normalization for the Isolated Tones 66
4.1.2 Tone Values of the Isolated Tones 72
4.2 Acoustic Features of the Isolated Tones in NS Hakka – Fundamental Frequency
77
4.2.1 Individual Growth Curve Modeling of F0 contours of the Isolated Tones in NS Hakka 77
4.2.2 Discriminant Analyses on the F0 Contours of the Isolated Tones in NS Hakka 85
4.3 Acoustic Features of Isolated Tones in NS Hakka – Duration 88
4.4 Summary and Discussion 90
Chapter V: Carryover Coarticulatory Effects in NS Hakka 94
5.1 Correlation between the Offsets of the Word-initial Tones and the F0 Data Points of the Word-final Tones 97
5.2 Dissimilatory Carryover Effects on the Rising Tone 1 [35] 100
5.3 Coarticulatory Effects on the Realizations of the High Level Tone 3 and Tone 6 109
5.4 Assimilatory Carryover Effects on the Realizations of the Falling Tones 121
5.5 Summary and Discussion 134
Chapter VI: Anticipatory Coarticulatory Effects and Tone Sandhi in NS Hakka 137
6.1 Anticipatory Coarticulatory Effects in NS Hakka 138
6.1.1 Correlation between the Onsets of the Word-final Tones and the F0 Data Points of the Word-initial Tones 138
6.1.2 Dissimilatory Anticipatory Effects on the High Level Tone 3 [55] and Tone 6 [55] 145
6.1.3 Dissimilatory Anticipatory Effects on the High Falling Tone 2 [51] and Tone 4 [52] 153
6.1.4 Assimilatory Anticipatory Effects on the Mid-Falling Tone 5 [31] 161
6.2 Tone Sandhi in NS Hakka 166
6.2.1 Word-initial Tone 1 in Sandhi and Non-sandhi Contexts 167
6.2.2 Neutralization of Sandhi Tone 1 and Lexical Tone 5. 178
6.3 Summary and Discussion 183
Chapter VII: Discussion and Conclusion 188
7.1 F0 Normalization Techniques, Tonal Representations, and Useful Mixed Method of Polynomial and Statistical Modeling on Tones 188
7.2 Tonal Coarticulation – a Process Grounded in the Interaction between Phonetic Universals and Patterns of Tonal Acoustics in NS Hakka 191
7.3 Yin Ping Tone Sandhi – Context-Induced and Gradient in Nature 194
7.4 Future Direction 196
References 198
Appendix Ａ: An Illustration for Using SPSS to Conduct Individual Growth Analysis on Tonal F0 Data 207
Appendix B: Results of Statistical Analyses on Contextual Tonal Variations due to Carryover Coarticulation Effects 215
Appendix C: Results of Statistical Analyses on Contextual Tonal Variations due to Anticipatory Coarticulation Effects 236



List of Tables
Table 1.1: The phonemic and phonetic transcriptions of consonants in NS Hakka (IPA in parenthesis represents the allophone) 10
Table 1.2: The phonemic transcriptions of finals in NS Hakka 11
Table 1.3: The six tone system in NS Hakka 12
Table 2.1a: Tonal representations of NS Hakka based on impressionistic descriptions 17
Table 2.1b: Tonal representations of NS Hakka based on acoustic descriptions 17
Table 3.1: The inventory of mono-syllabic words with the six lexical tones in NS Hakka. 39
Table 3.2: The Inventory of the disyllabic words exemplifying all the 36 ditonal sequences. 41
Table 4.1: The normalization indices (NI) of 4 different tone normalization methods 71
Table 4.2: Mean LZ normalized F0 values for the six NS Hakka isolated tones 72
Table 4.3: Tone values of the six NS Hakka isolated tones (this study vs. Chang’s) 76
Table 4.4: The mean values for the intercept, linear and quadratic slopes of the six isolated tones in NS Hakka using LZ normalized F0 79
Table 4.5: IGC results of the comparison in mean F0 height and F0 contour shape between isolated Tone 3 [55] and Tone 6 [55] 83
Table 4.6: Classification results for NS Hakka isolated tones using 2nd polynomial coefficients 87
Table 4.7: Mean duration of the 6 NS Hakka isolated tones for male, female, and all speakers 89
Table 4.8: Tone values of the six NS Hakka isolated tones 92
Table 5.1: Pearson correlation coefficients for the relationship between the offsets of the word-initial tones and the 11 F0 data points of the six word-final tones (N = no. of tokens) 98
Table 5.2: Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T1 in TxT1 sequences (WI = word-initial) 103
Table 5.3: Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T3 in TxT3 sequences (WI = word-initial) 112
Table 5.4: Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T6 in TxT6 sequences (WI = word-initial) 112
Table 5.5: Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T2 in TxT2 sequences (WI = word-initial) 124
Table 5.6 Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T4 in TxT4 sequences (WI = word-initial) 125
Table 5.7: Mean offset F0 values of the WI Tx, and the mean coefficients and duration of T5 in TxT5 sequences (WI = word-initial) 125
Table 5.8: Comparison of the amount of tonal contextual variations between the word-final T2 and T5 132
Table 6.1: Pearson correlation coefficients for the relationship between the onsets of the word-final tones and the 11 F0 data points of the six word-initial tones (N = no. of tokens) 139
Table 6.2: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T3 in the T3Tx sequences (WF = word-final) 148
Table 6.3: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T6 in the T6Tx sequences (WF = word-final) 148
Table 6.4: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T2 in the T2Tx sequences (WF = word-final) 156
Table 6.5: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T4 in the T4Tx sequences (WF = word-final) 156
Table 6.6: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T5 in the T5Tx sequences 164
Table 6.7: Mean onset F0 values of the WF Tx, and the mean coefficients and durations of T1 in the T1Tx sequences 170
Table 6.8: Polynomial coefficients for the F0 contours of the WI Tone 1 and WI Tone 5 in the same sandhi contexts (WI = word-initial; Tx = T1, T3 and T6) 179
Table 6.9: Polynomial coefficients for the velocity contours of the WI Tone 1 and WI Tone 5 in sandhi contexts (Tx = T1, T3, and T6) 182
Table 7.1: Summary of tonal coarticulation in NS Hakka 191
Table B.1: Descriptive statistics of the mean F0 values over T1 in the TxT1 sequences 216
Table B.2: Exhaustive pairwise comparisons between each realization of T1 in TxT1 218
Table B.3: Amount of contextual variations of T1 in TxT1 220
Table B.4: Descriptive statistics of the F0 values over T3 in the TxT3 sequences 221
Table B.5: Descriptive statistics of the mean F0 values over T6 in the TxT6 sequences 222
Table B.6: Exhaustive pairwise comparisons between each realization of T3 in TxT3 223
Table B.7: Amount of contextual variations of T3 in TxT3 224
Table B.8: Exhaustive pairwise comparisons between each realization of T6 in TxT6 225
Table B.9: Amount of contextual variations of T6 in TxT6 226
Table B.10: Descriptive statistics of the mean F0 values over T2 in the TxT2 sequences 227
Table B.11: Descriptive statistics of the F0 values over T4 in the TxT4 sequences 228
Table B.12: Descriptive statistics of the F0 values over T5 in the TxT5 sequences 229
Table B.13: Exhaustive pairwise comparisons between each realization of T2 in TxT2 230
Table B.14: Amount of contextual variations of T2 in TxT2 231
Table B.15: Exhaustive pairwise comparisons between each realization of T4 in TxT4 232
Table B.16: Amount of contextual variations of T4 in TxT4 233
Table B.17: Exhaustive pairwise comparisons between each realization of T5 in TxT5 234
Table B.18: Amount of contextual variations of T5 in TxT5 235
Table C.1: Descriptive statistics of the F0 values over T3 in the T3Tx sequences 236
Table C.2: Descriptive statistics of the F0 values over T6 in the T6Tx sequences 237
Table C.3: Exhaustive pairwise comparisons between each realization of T3 in T3Tx 238
Table C.4: Amount of contextual variations of T3 in T3Tx 239
Table C.5: Exhaustive pairwise comparisons between each realization of T6 in T6Tx 240
Table C.6: Amount of contextual variations of T3 in T3Tx 241
Table C.7: Descriptive statistics of the F0 values over T2 in the T2Tx sequences 242
Table C.8: Descriptive statistics of the F0 values over T4 in the T4Tx sequences 243
Table C.9: Exhaustive pairwise comparisons between each realization of T2 in T2Tx 244
Table C.10: Amount of contextual variations of T2 in T2Tx 245
Table C.11: Exhaustive pairwise comparisons between each realization of T4 in T4Tx 246
Table C.12: Amount of contextual variations of T4 in T4Tx 247
Table C.13: Descriptive statistics of the F0 values over T5 in the T5Tx sequences 248
Table C.14: Exhaustive pairwise comparisons between each realization of T5 in T5Tx 249
Table C.15: Amount of contextual variations of T5 in T5Tx 250
Table C.16: Descriptive statistics of the F0 values over T1 in the T1Tx sequences 251
Table C.17: Exhaustive pairwise comparisons between each realization of T6 in T6Tx 252
Table C.18: Amount of contextual variations of T1 in T1Tx 253



List of Figures
Figure 1.1: Syllable structure and example words of NS Hakka 9
Figure 3.1: A display on the computer of the constituents of the recording software 43
Figure 3.2: A user interface shot with one stimulus in Chinese characters, its respective waveform, and all the function buttons 44
Figure 3.3: Photos of recording environment for this study 45
Figure 3.4: A WaveSurfer screen shot for illustrating mono-syllabic word segmentation and labeling 46
Figure 3.5: A WaveSurfer screen shot for illustrating di-syllabic words segmentation and labeling 47
Figure 3.6: An illustration for the contribution of polynomial coefficients to the formation of tonal contours 59
Figure 4.1: An invalid F0 contour truncated due to low F0 creaky voices 64
Figure 4.2: An invalid F0 contour truncated due to perturbation caused by a coda stop 64
Figure 4.3: Mean F0 contours of the six NS Hakka isolated tones produced by each speaker. 68
Figure 4.4: Distributions of all participants’ mean raw tonal F0 trajectories as functions of raw mean durations (Left = the 4 lax tones. Right = 2 checked tones) 69
Figure 4.5: The mean F0 contours of the 6 NS Hakka tones normalized by 4 different normalization methods (Left = 4 lax tones. Right = 2 checked tones.) 70
Figure 4.6: Mean T-value normalized F0 contours of the six isolated tones in NS Hakka against the percentage of the average duration of all tones (mean duration = 100) 74
Figure 4.7: Comparison of the mean intercept between the six lexical tones in NS Hakka 80
Figure 4.8: Comparison of the mean linear slopes between the six lexical tones in NS Hakka 80
Figure 4.9: Comparison of the mean quadratic slopes between the six lexical tones in NS Hakka 81
Figure 4.10: Results of discriminant analyses for F0 data of the six isolated tones using Quadratic polynomial coefficients 88
Figure 5.1: Mean F0 contours of the TxT1 ditonal sequences against the mean raw duration of each variant of the WF T1 and of each WI context tone (WI = word-initial; WF = word-final) 101
Figure 5.2: Effects of the offsets of the WI tones on the intercepts of the WF Tone 1 [35] (WI = word-initial; WF = word-final) 104
Figure 5.3: Mean F0 contours of the TxT3 ditonal sequences against the mean raw duration of each variant of the WF T3 and of each WI context tone (WI = word-initial; WF = word-final) 110
Figure 5.4: Mean F0 contours of the TxT6 ditonal sequences against the mean raw duration of each variant of the WF T6 and of each WI context tone (WI = word-initial; WF = word-final) 111
Figure 5.5: Effects of the offsets of the WI tones on the intercepts of the WF Tone 3 [55] (WI = word-initial; WF = word-final) 117
Figure 5.6: Mean F0 contours of the TxT2 ditonal sequences against the mean raw duration of each variant of the WF T2 and of each WI context tone (WI = word-initial; WF = word-final) 122
Figure 5.7: Mean F0 contours of the TxT4 ditonal sequences against the mean raw duration of each variant of the WF T4 and of each WI context tone (WI = word-initial; WF = word-final) 122
Figure 5.8: Mean F0 contours of the TxT5 ditonal sequences against the mean raw duration of each variant of the WF T5 and of each WI context tone (WI = word-initial; WF = word-final) 123
Figure 5.9: Effects of the offsets of the WI tones on the intercepts of the WF Tone 2 [51] (WI = word-initial; WF = word-final) 126
Figure 5.10: Effects of the offsets of the WI tones on the intercepts of the WF Tone 4 [52] (WI = word-initial; WF = word-final) 126
Figure 5.11: Effects of the offsets of the WI tones on the intercepts of the WF Tone 5 [31] (WI = word-initial; WF = word-final) 127
Figure 5.12: Comparison of the F0 realizations of (Tx)T2 and (Tx)T5 131
Figure 5.13: Mean F0 contours of the TxT5 ditonal sequences against the mean raw duration of each variant of the WF T5 and of each WI context tone (WI = word-initial; WF = word-final) 133
Figure 6.1: Effects of the onsets of the WF tones (Rise/Fall/High/Low) on mean F0 peak values of the WI tones (WF = word-final; WI = word-initial) 144
Figure 6.2: Mean F0 contours of the T3Tx ditonal sequences against the mean raw duration of each variant of the WI T3 and of each WF context tone (WI = word-initial; WF = word-final) 146
Figure 6.3: Mean F0 contours of the T6Tx ditonal sequences against the mean raw duration of each variant of the WI T6 and of each WF context tone (WI = word-initial; WF = word-final) 146
Figure 6.4: Effects of the onsets of the WF tones on the intercepts of the WI Tone 3 [55] (WI = word-initial; WF = word-final) 149
Figure 6.5: Effects of the onsets of the WF tones on the intercepts of the WI Tone 6 [55] (WI = word-initial; WF = word-final) 149
Figure 6.6: Mean F0 contours of the T2Tx ditonal sequences against the mean raw duration of each variant of the WI T2 and of each WF context tone (WI = word-initial; WF = word-final) 154
Figure 6.7: Mean F0 contours of the T4Tx ditonal sequences against the mean raw duration of each variant of the WI T4 and of each WF context tone (WI = word-initial; WF = word-final) 155
Figure 6.8: Effects of the onsets of the WF tones on the intercepts of the WI Tone 2 [51] (WI = word-initial; WF = word-final) 157
Figure 6.9: Effects of the onsets of the WF tones on the intercepts of the WI Tone 4 [52] (WI = word-initial; WF = word-final) 157
Figure 6.10: Mean F0 contours of the T5Tx ditonal sequences against the mean raw duration of each variant of the WI T5 and of each WF context tone (WI = word-initial; WF = word-final) 162
Figure 6.11: Effects of the onsets of the WF tones on the intercepts of the WI Tone 5 [31] (WI = word-initial; WF = word-final) 164
Figure 6.12: Absolute slope values of the WI Tone 5 [31] in different tonal contexts 165
Figure 6.13: Curvature values (se/sec2) of the WI Tone 5 [31] in different tonal contexts 165
Figure 6.14: Mean F0 contours of the T1Tx ditonal sequences against the mean raw duration of each variant of the WI T1 and of each WF context tone (WI = word-initial; WF = word-final) 167
Figure 6.15: Effects of the onsets of the WF tones on the intercepts of the WI Tone 1 (WI = word-initial; WF = word-final) 170
Figure 6.16: Effects of the linear slopes of the WF tones on the linear slopes of the WI Tone 1 (WI = word-initial; WF = word-final) 172
Figure 6.17: Anticipatory effects of following tones on F0 contour of the preceding rising tone in /mama / sequences in Mandarin (Xu, 1997: 75, Figure 6-b) 175
Figure 6.18: Mean F0 contours of sandhi Tone 1 vs. lexical Tone 5 in sandhi contexts estimated by IGC modeling 179
Figure 6.19: Mean velocity contours of sandhi Tone 1 [31] vs lexical Tone 5 [31] in sandhi contexts estimated by IGC modeling 181
Figure A.1: Example of vertical arrangement of the dataset for IGC modeling on the six isolated tones in NS Hakka 208
Figure A.2: Menu item “Analyze/Mixed Models/Linear” in SPSS 209
Figure A.3: Linear Mixed Models: Specify Subjects and Repeated in SPSS 210
Figure A.4: Setting dependent variable, factor, and covariates for conducting IGC modeling on the isolated F0 data in NS Hakka 210
Figure A.5: Building fixed factors in IGC modeling on the isolated F0 data in NS Hakka 212
Figure A.6: Building the random factors in IGC modeling on the isolated F0 data in NS Hakka 213
Figure A.7: Selecting items of statistics in IGC modeling on the isolated F0 data in NS Hakka 214
Figure A.8: Selecting fixed predicated values in IGC modeling on the isolated F0 data in NS Hakka 214

References
Abramson, A. S. 1979. The coarticulation of tones: an acoustic study of Thai.
Studies in Tai and Mon-Khmer phonetics and phonology in honor of Eugenie JA Henderson, 1-9.
Anderson, S.R. 1978. Tone features. In V.A. Fromkin (ed.), Tone: a linguistic survey.
New York: Academic Press, 133-176.
Andruski, J. E., & Costello, J. 2004. Using polynomial equations to model pitch contour shape in
lexical tones: An example from Green Mong. Journal of the International Phonetic Association 34(2). 125-140.
Brunelle, M. 2009. Northern and Southern Vietnamese tone coarticulation: A comparative case
study. Journal of Southeast Asian Linguistics 1. 49-62.
Chang, Y.-C. (張月琴).1995. Cong Shengxiao Jiaodu Lai Miaoxie Taiwan Miaoli Sixian Kejiahua
De Shengdiao Xitong從聲學角度來描寫台灣苗栗四縣的聲調系統[Tonal Representations of Sixian Hakka in Miaoli – An Acoustic Prospective]. In Tsao, F.-F. & Tsai, M.-H. (eds). The 1994 Conference on Language Teaching and Linguistics in Taiwan, 95-112. Taipei: Crane.
Chang, Y.-C., & Hsieh, F.-F. 2012. Tonal coarticulation in Malaysian Hokkien: A typological
anomaly?. The Linguistic Review 29. 37–73.
Chao Y.-R. 1930. “ə sistəm əv “toun lɛtɚz”” [A System of Tone Letters], Le Maître Phonétique
45. 24-27.
Chao, Y.-R. 1968. Language and Symbolic Systems. CUP Archive.
Chen, M. Y. 1973. Cross-dialectal comparison: a case study and some theoretical considerations.
Journal of Chinese Linguistics. 38-63.
Chen, M. Y. 2000. Tone sandhi: patterns across Chinese dialects. vol. 92. Cambridge: Cambridge
University Press.
Chen, S., Wiltshire, C., & Li, B. 2018. An updated typology of tonal coarticulation properties.
Taiwan Journal of Linguistics 16(2). To be published.
Chen, S., Zhang, C., McCollum, A. G., & Wayland, R. 2017. Statistical modelling of phonetic
and phonologised perturbation effects in tonal and non-tonal languages. Speech Communication 88. 17-38.
Chen, S.-W., Wang, B., & Xu, Y.2009. Closely related languages, different ways of realizing
focus. In Interspeech, 1007-1010.
Chen, Y., & Xu, Y. 2006. Production of weak elements in speech–evidence from F₀ patterns of
neutral tone in Standard Chinese. Phonetica 63(1). 47-75.
Cheng, Chierh, Chen, Jenn Yeu, & Gubian, M. 2013. Are Mandarin Sandhi Tone 3 and Tone 2 the
Same or Different? The Results of Functional Data Analysis. In Proceeding of Pacific Asia Conference in Language, Information, and Computation (PACLIC-27). 296-301. Taipei, Taiwan.
Cheng, Chierh, Chen, Jenn Yeu, & Xu, Yi. 2014. An acoustic analysis of Mandarin Tone 3 sandhi
elicited from an implicit priming experiment. In Fourth International Symposium on Tonal Aspects of Languages. 36-40.
Cheng, Ming-chung. 2014. Exploring the tones of Raoping Hakka in Zhuolan township: a
perspective of experimental phonetics. Journal of Language and Literature Studies, (25), 103-161.
Cheung, Yuk Man. 2011. Vowels and tones in Meixian Hakka: an acoustic and perceptual study.
Hong Kong: University of Hong Kong Dissertation.
Chung, Rung Fu 2004. Taiwan Kechia yuyin daulun台灣客家語音導論[An introduction to
Taiwan Hakka phonetics]. Taipei: Wu-nan Publisher.
Duanmu, S. 2000. The phonology of Standard Mandarin. Oxford: Oxford University Press.
Earle, M. A. 1975. An acoustic phonetic study of Northern Vietnamese tones. Monograph 11,
Speech Communications Research Laboratories Inc., Santa Barbara.
Gandour, J., Potisuk, S., Dechongkit, S. & Ponglorpisit, S. 1992. Anticipatory tonal coarticulation
in Thai noun compounds. Linguistics of the Tibeto-Burman Area 15. 111-124.
Gandour, J., Potisuk, S., & Dechongkit, S. 1994. Tonal coarticulation in Thai. Journal of
Phonetics 22(4). 477-492.
Gu, Wentao, & Lee, Tan. 2007. Effects of tonal context and focus on Cantonese F0. In
Proceedings of the 16th International Congress of Phonetic Sciences (ICPhS
2007),1033-1036. Germany.
Han, M. S., & Kim, K. 1974. Phonetic variation of Vietnamese tones in disyllabic utterances.
Journal of Phonetics 2. 223-232.
Hombert, J. M., Ohala, J. J., & Ewan, W. G. 1979. Phonetic explanations for the development of tones. Language, 37-58.
Howie, J. M. 1974. On the domain of tone in Mandarin. Phonetica 30. 129-148.
Hsiao, Yuchau E. 1991. A beat-counting theory of foot phrasing. In Proceedings of the 2nd
International Symposium on Chinese Languages and Linguistics. 265-284. Taipei: Academia Sinica
Hsiao, Yuchau E. 2015. Rethinking OCP effects on tone sandhi. Language and Linguistics, 16(6),
927-945.
Huang, Yao-huang. 2003. An acoustic study on the Hakka tones. Kaohsiung: National Kaohsiung
Normal University MA thesis.
Huang, Y., Donohue, M., Rose, P., & Sidwell, P. Normalization of Zhangzhou citation tones. 2016 In Proceeding of the 16th Australasian International Conference on Speech Science and Technology, 217-220.
Hsu, Kuei-Ping. 1996. Hakka Tone Sandhi: The Interface between Syntax and Phonology.
Hsinchu: National Tsing Hua University MA thesis.
Hyman, L.M. 2001. Tone systems. In M. Haspelmath, E. König, W. Oesterreicher, & W. Raible (eds.), Language Typology and Language Universals: An International Handbook, 1367-1380. Berlin: Walter de Gruyter
Hyman, L.M.2010. How to study a tone language, with exemplification from Oku (Grassfields
Bantu, Cameroon). UC Berkeley Phonology Lab Annual Report 2010. 179–209.
Kataoka, Reiko. 2011. Phonetic and cognitive bases of sound change. San Francisco: University
of California at Berkeley dissertation.
Kochanski, G. P., & Shih, C. 2000. Stem-ML: Language-independent prosody description. In Sixth International Conference on Spoken Language Processing.
Li, Wen-chao.1999. A Diachronically-motivated Segmental Phonology of Mandarin
Chinese. New York: Peter Lang.
Li, Qian, & Chen, Yiya. 2016. An acoustic study of contextual tonal variation in Tianjin Mandarin.
Journal of Phonetics 54. 123-150.
Lin, Maocan. 1995. A perceptual study on the domain of tones in Standard Chinese. Chinese
Journal of Acoustics 14(4). 350-357.
Lin, Y.-H. 2007. The sounds of Chinese with audio CDVol. 1. Cambridge: Cambridge University
Press.
Liu, J.-R. 2007. Tonal representations and co-articulation in Ta-pu Hakka. Taipei: National Taiwan University MA thesis.
Lo, Seu-Kim (羅肇錦). 1984. Ke Yu Yu Fa客語語法[The Grammar of Hakka]. Taipei: Student Bookstore.
Lo, Seu-Kim (羅肇錦). 1990. Taiwan De Ke Jia Hua台灣的客家話[Hakka Dialects in Taiwan],
vol. 16. Taipei: Taiyuan Publisher.
Lo, Sin-Zhi (羅新枝). 2016. An acoustic analysis of Hakka tones of Vietnam’s new immigrants in
Taiwan: a case study. Miaoli: National United University MA thesis.
More´n, Bruce & Zsiga, Elizabeth. 2006. The Lexical and Post-lexical Phonology of Thai tones.
Natural Language & Linguistic Theory 24(1). 113-178.
Myers, J., & Tsay, J. 2003. Investigating the phonetics of Mandarin tone sandhi. Taiwan Journal of
Linguistics 1(1). 29-68.
Myers, J. and J. Tsay. 2008. Neutralization in Taiwan Southern Min tone sandhi. In Y. E. Hsiao, H.-C. Hsu, L.-H. Wee and D.-A. Ho (eds.), Interfaces in Chinese Phonology: Festchrift in Honor of Matthew Y. Chen on His 70th Birthday: 47–78. Taipei: Institute of Linguistics, Academia Sinica
Norman, J. 1988. Chinese. Cambridge: Cambridge University Press.
Noss, R. B.1975. How useful are citation forms in synchronic Thai phonology.
Studies in Tai linguistics in honor of William J. Gedney, 274-84.
Ohala, John J. 1981. Articulatory constraints on the cognitive representation of speech. In Advances in Psychology , Vol. 7, 111-122. North-Holland.
Ohala, John J. 1989. Sound change is drawn from a pool of synchronic variation. Language change: Contributions to the study of its causes, 173-198.
Ohala, John J. 1993. The phonetics of sound change. In C. Jones (ed.), Historical linguistics:
problems and perspectives, 237-278. London: Longman.
Ohala, John J. 1994. Towards a universal, phonetically-based theory of vowel harmony. Proceedings of the 3rd International Conference on Spoken Language Processing (ICSLP 94), Vol. 2, 491-494. Yokohama: Acoustical Society of Japan.
Pan, Ho- Hsien. 2008. Focus and Taiwanese unchecked Tones. In C. Lee, M. Gordon and
D. Büring (eds). In Topic and Focus: Cross-linguistic Perspectives on Meaning and Intonation. Springer Netherlands, 195-213.
Pham, T. B., & Harris, R. J. 2001. Acculturation Strategies among Vietnamese-Americans.
International Journal of Intercultural Relations 25(3), 279-300.
Peng, Shu-hui.1997. Production and Perception of Taiwanese tones in Different Tonal
and Prosodic Contexts. Journal of Phonetics, 25(3), 371-400.
Pike, K. L. 1948. Tone language. Ann Arbor: University of Michigan Press.
Prom-on, Santitham, Liu, Famg, & Xu, Yi. 2012. Post-low bouncing in Mandarin Chinese: Acoustic analysis and computational modeling. The Journal of the Acoustical Society of America 132(1), 421-432.
Rose, P. 1987. Considerations in the normalization of the fundamental frequency of linguistic
tone. Speech Communication 6(4). 343-352.
Rose, P. 1993. A linguistic phonetic acoustic analysis of Shanghai tones. Australian Journal of
Linguistics 13. 185-219.
Rose, P. 2000. Hong Kong Cantonese citation tone acoustics: a linguistic tonetic
study. In Proceedings of the 8th Australian International Conference on Speech Science & Technology, 198-203.
Rose, P. 2010. The long and the short of Wencheng tones – acoustic and duditory
description of tonologically challenging phenomena in an Oujiang Wu dialect of Chinese. In Proceeding of the 10th Australasian International Conference on Speech Science & Technology, 54-57.
Rose, P. 2014. Transcribing tone – a likelihood-based quantitative evaluation of
Chao's tone letters. In Interspeech, 101-105.
Rose, P. 2016. A comparison of normalization strategies for citation tone F0 in four Chinese dialects. In Proceeding of the 16th Australasian International Conference on Speech Science and Technology, 221-224.
Shek, Daniel T.L., & Ma, Cecillia M.S. 2011. Longitudinal data analyses using linear mixed
models in SPSS: concepts, procedures and illustrations. The Scientific World Journal 11. 42-76.
Shi, Feng (石峰). 1990. Yuyinxue Tanwei語音學探微[Studies in Tones and Stops]. Beijing:
Peking University Press.
Shi, Feng (石峰) and Liao, Rung-Rung (廖榮蓉). 1994. Yuyin Chung Gao語音叢稿[Papers in
Phonetics]. Beijing: Beijing Language Institute Press.
Shih, Chilin. 1988. Tone and intonation in Mandarin . In Working Papers of the Cornell Phonetics
Laboratory 3. 83-109.
Shih, Chilin. 1997. Mandarin third tone sandhi and prosodic structure. In J. Wang and N. Smith
(eds.), Studies in Chinese Phonology, 81-124. Berlin: Mouton de Gruyter
Simpson, A. P. 2009. Phonetic differences between male and female speech. Language and
Linguistics Compass 3(2). 621-640.
Sjölander, K. 2004. Snack sound toolkit. KTH Stockholm, Sweden.
http://www.speech.kth.se/snack
Sjölander, K., & Beskow, J. 2000. Wavesurfer – an open source speech tool. In Interspeech,
464-467.
Sun, Xiaoli, & Huang, Tsan. 2015. Gradience in contextual tonal realization process: an
instrumental study of Nanjin Chinese. Paper from the 18th International Congress on Phonetic Science.
Tingsabadh, K. M.1988. Loss of preeminence of citation forms in the study of tones in Thai
dialects. In Bamroongraks, Cholticha and Wilaiwan Khanittana , et. al., (eds), 224-27.
Traunmüller, H., & Eriksson, A. 1995. The frequency range of the voice fundamental in the
speech of male and female adults. Unpublished manuscript.
Trísková, H. 2011. The structure of the Mandarin syllable: why, when and how to teach it.
Archiv orientální 79(1). 99.
Tseng, Chiu-yu, Pin, Sao-Huang, Lee, Yehlin., Wang, Hsin-min, & Chen, Yong-chen. 2005. Fluent
speech prosody: framework and modeling. Speech Communication 46(3). 284-309.
Weng, Jie. 2017. Acoustic analysis of the citation tones of Miaoli Sixian Hakka. Miaoli: National
United University MA thesis
Wong, Ying Wai. 2006. Contextual tonal variations and pitch targets in Cantonese. In Proceedings
of Speech Prosody, 317-320. Germany.
Wu, Zhong-Jie. 2010. Gao Ping Liu Dui He Bei Bu Sixian Hua Yu Yan Cha Yi Zhi
Cheng Yin Tan Jiu高屏六堆和北部四縣話語言差異之成因探究 [An Inquiry
into the causes of differences between South and North Sixian Hakka]. In Proceeding of the 6th Conference on Hakka.
Xu, Yi. 1993. Contextual tonal variation in Mandarin Chinese. Storrs: University of Connecticut
dissertation.
Xu, Yi. 1997. Contextual tonal variations in Mandarin. Journal of Phonetics 25(1). 61-83.
Xu, Yi. 1998. Consistency of Tone-syllable Alignment across Different Syllable Structures and
Speaking Rates. Phonetica 55(4). 179-203.
Xu, Yi. 1999. Effects of tone and tocus on the tormation and alignment of F0 contours. Journal of
Phonetics 27(1). 55-105.
Xu, Yi. 2001. Fundamental frequency peak delay in Mandarin. Phonetica 58(1-2). 26-52.
Xu, Yi. 2004. Understanding tone from the perspective of production and perception.
Language and Linguistics 5(4). 757-797.
Xu, Yi. 2011. Post-focus compression: cross-linguistic distribution and historical origin.
In Proceedings of the 17th International Congress of Phonetic Sciences, 152-155.
Xu, Yi., & Sun, Xuejing. 2002. Maximum speed of pitch change and how it may relate to speech.
The. Journal of the Acoustical Society of America 111(3). 1399-1413.
Yang, Shi-Feng.楊時逢. 1957. Taiwan Taoyuan Ke Jia Fang Yan台灣桃園客家方言[The Hakka
Dialects of Taoyung, Taiwan]. Taipei: Academia Sinica.
Yang, Wun-Shan.楊文山. 2004. Quanguo Kejia Renkou Jichu Ziliao Diaocha Yanjiu全國客家人
口基礎資料調查研究[A Survey on the Basic Information of Hakka Population in Taiwan]. Taipei: the Council of Hakka Affairs, Taiwan.
Yip, Moira. 2002. Tone. Cambridge: Cambridge University Press.
Yu, Shiow-min. 1984. Aspects of the Phonology of Miaoli Hakka. Taipei: Fu Jen Catholic
University MA thesis.
Yuan, Jia Hua, et al.袁家驊等. 2001. Han Yu Fang Yan Gai Yao漢语方言概要[Outline of Chinese
Dialectology]. Beijing: Wenzi Gaige, Chubanshe.
Yuan, Jiahong, Shih, Chilin, & Kochanski, G. P. 2002. Comparison of declarative and interrogative
intonation in Chinese. In Speech Prosody.
Zhang, Jie. 2001. The effects of duration and sonority on contour tone distribution--typological
survey and formal analysis. Los Angeles: University of California at Los Angeles dissertation.
Zhang, Jie. 2002. The effects of duration and sonority on contour tone distribution. New York:
Routledge.
Zhang, Jie, & Lai, Yuwen. (2010). Testing the role of phonetic knowledge in Mandarin tone
sandhi. Phonology, 27(1), 153-201.
Zhang, Jie, Lai, Yuwen., & Sailor, C. 2011. Modeling Taiwanese speakers’ knowledge of tone
sandhi in reduplication. Lingua 121(2). 181-206.
Zhang, Jie & Liu, Jiang. 2011. Tone sandhi and tonal coarticulation in Tianjin Chinese. Phonetica
68(3). 161-191
Zhang, Jie. 2014. Tones, tonal phonology, and tone sandhi. In Huang, C. J., Li, Y. A., & Simpson,
A. (eds). The Handbook of Chinese Linguistics, 443-464. John Wiley & Sons.
Zhang, Jingwei. 2014. A Sociophonetic study on tonal variation of the Wúxī and Shànghǎi
dialects. Utrecht: Utrecht University dissertation.
Zhu, Xiaonong. 1999. Shanghai tonetics. Muenchen, Germany: Lincom Europa.