On the cognitive basis of contact-induced sound change: Vowel merger reversal in Shanghainese: Online Appendices

Language 92.2, June 2016 s1 ON THE COGNITIVE BASIS OF CONTACT-INDUCED SOUND CHANGE: VOWEL MERGER REVERSAL IN SHANGHAINESE: ONLINE APPENDICES YAO YAO CHARLES B. CHANG The Hong Kong Polytechnic University Boston University APPENDIX A: MATERIALS AND MODELS IN STUDY 1 ITEM LEXICAL SET SHANGHAINESE CITATION FORM* (IN STAGE II) MANDARIN FORM EMBEDDING COMPOUND (AND PART OF SPEECH) EMBEDDING COMPOUND FREQUENCY 雷 MN-[ej] le˨˧ lej˧˥ 打雷 ‘thunder strikes’ (v.) high 垒 MN-[ej] le˨˧ lej˨˦ 堡垒 ‘fortress’ (n.) low 缆 MN-[an] lɛ˨˧ lan˨˦ 光缆 ‘optical fiber’ (n.) high 澜 MN-[an] lɛ˨˧ lan˧˥ 狂澜 ‘huge wave’ (n.) low 来 MN-[aj] lɛ˨˧ laj˧˥ 上来 ‘to come up’ (v.) high 睐 MN-[aj] lɛ˨˧ laj˥˩ 青睐 ‘to favor’ (v.) low 配 MN-[ej] ph e˧˦ ph ej˥˩ 搭配 ‘to match with’ (v.) high 沛 MN-[ej] ph e˥˧ ph ej˥˩ 充沛 ‘abundant’ (adj.) low 滩 MN-[an] th ɛ˥˧ th an˥ 外滩 ‘the Bund’ (n.) high 坍 MN-[an] th ɛ˥˧ th an˥ 压坍 ‘to crash’ (v.) low 态 MN-[aj] th ɛ˧˦ th aj˥˩ 状态 ‘status’ (n.) high 胎 MN-[aj] th ɛ˥˧ th aj˥ 保胎 ‘to protect the fetus’ (v.) low 贝 MN-[ej] pe˧˦ pej˥˩ 宝贝 ‘treasure’ (n.) high 狈 MN-[ej] pe˧˦ pej˥˩ 狼狈 ‘in an extremely embarrassing state’ (adj.) low 班 MN-[an] pɛ˥˧ pan˥ 上班 ‘to go to work’ (v.) high 阪 MN-[an] pɛ˧˦ pan˨˦ 大阪 ‘Osaka (Japanese city)’ (n.) low 呆 MN-[aj] tɛ˥˧ taj˥ 痴呆 ‘retarded’ (adj.) high 歹 MN-[aj] tɛ˥˧ taj˨˦ 为非作歹 ‘to do bad things’ (v.) low * The tone of the test items will change when embedded in a compound due to tone sandhi. TABLE A1. Critical items in study 1. s2 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 634.01 8.58 73.91 < 0.0010 1530.92 13.05 117.33 < 0.001 Age = Old −54.20 3.49 −15.55 < 0.0010 26.13 3.07 8.52 < 0.001 Sex = F 37.90 2.58 14.70 < 0.0010 91.72 2.35 39.12 < 0.001 Onset = L 46.90 8.01 5.86 < 0.0010 −47.57 7.34 −6.48 < 0.001 Onset = PHTH −13.86 8.01 −1.73 0.0970 44.53 7.34 6.07 < 0.001 LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 634.72 14.88 43.26 < 0.0010 1530.40 15.62 98.05 < 0.001 LexSet = MN-[an] −5.65 15.24 −0.37 0.7200 21.87 15.89 1.38 0.150 LexSet = MN-[ej] −74.17 15.23 −4.87 < 0.0010 66.74 15.88 4.20 < 0.001 Age = Old −31.73 4.18 −7.59 < 0.0010 29.09 5.58 5.21 < 0.001 LexSet = MN-[an]: Age = Old — — — — −18.20 7.83 −2.32 0.020 LexSet = MN-[ej]: Age = Old — — — — −24.95 7.79 −3.21 0.002 Sex = F 28.12 3.15 8.93 < 0.0010 91.00 2.44 37.26 < 0.001 Onset = L 28.87 8.80 3.28 0.0042 — — — — Onset = PHTH −11.76 8.80 −1.34 0.1800 — — — — TABLE A2. Fixed-effect terms in the LM models on formant measures in the reading experiment, study 1. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) −1.63 0.25 −6.42 < 0.001 LexSet = MN-[an] 0.50 0.25 2.00 0.045 LexSet = MN-[ej] 2.42 0.25 9.75 < 0.001 Age = Old −0.52 0.14 −3.63 < 0.001 Frq = H 0.33 0.10 3.38 < 0.001 Onset = L 1.14 0.14 8.38 < 0.001 Onset = PHTH −0.94 0.15 −6.35 < 0.001 TABLE A3. Fixed-effect terms in the GLM model on Diphthong in the reading experiment, study 1. Bold = p(|z|) < 0.05. s3 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 663.30 9.37 70.75 < 0.001 1550.92 13.11 118.28 < 0.001 Age = Old −53.74 3.71 −14.50 < 0.001 — — — — Sex = F 41.50 2.78 14.94 < 0.001 90.87 2.18 41.70 < 0.001 Onset = L 38.96 7.52 5.18 < 0.001 −37.91 6.47 −5.86 < 0.001 Onset = PHTH −1.50 7.52 −0.20 0.840 35.19 6.47 5.44 < 0.001 Block.L 7.91 3.19 2.48 0.014 −6.78 2.45 −2.76 0.006 Block.Q −0.96 3.19 −0.30 0.770 −1.01 2.45 −0.41 0.700 LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 675.13 18.33 36.84 < 0.001 1566.59 15.70 99.81 < 0.001 LexSet = MN-[an] −5.66 19.73 −0.29 0.770 1.39 11.38 0.12 0.890 LexSet = MN-[ej] −130.14 19.73 −6.60 < 0.001 86.36 11.38 7.79 0.001 Age = Old −37.73 7.92 −4.76 < 0.001 7.51 5.59 1.34 0.180 LexSet = MN-[an]: Age = Old 9.34 11.14 −0.84 0.400 −0.95 7.86 −0.12 0.890 LexSet = MN-[ej]: Age = Old 34.99 11.14 3.14 0.002 −43.24 7.86 −5.50 < 0.001 Sex = F 39.33 3.47 11.32 < 0.001 92.08 2.48 37.07 < 0.001 TABLE A4. Fixed-effect terms in the LM models on formant measures in the translation experiment, study 1. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) −1.38 0.26 −5.37 < 0.001 LexSet = MN-[an] 0.35 0.24 1.44 0.150 LexSet = MN-[ej] 3.12 0.27 11.60 < 0.001 Age = Old 0.10 0.24 0.42 0.680 LexSet = MN-[an]: Age = Old −0.53 0.34 −1.55 0.120 LexSet = MN-[ej]: Age = Old −0.91 0.34 −2.64 0.008 Frq = H 0.17 0.07 2.35 0.019 Onset = L 0.97 0.11 9.20 < 0.001 Onset = PHTH −0.56 0.11 −5.22 < 0.001 TABLE A5. Fixed-effect terms in the GLM model on Diphthong in the translation experiment, study 1. Bold = p(|z|) < 0.05. s4 APPENDIX B: MATERIALS AND MODELS IN STUDY 2 ITEM LEXICAL SET SHANGHAINESE CITATION FORM* (IN STAGE II) MANDARIN FORM EMBEDDING COMPOUND (AND PART OF SPEECH) EMBEDDING COMPOUND FREQUENCY 退 Structure-mismatched MN-[ej] th e˧˦ th wej˥˩ 辞退 ‘to lay off’ (v.) high 腿 Structure-mismatched MN-[ej] th e˧˦ th wej˨˦ 方腿 ‘Spam (meat)’ (n.) high 对 Structure-mismatched MN-[ej] te˧˦ twej˥˩ 不对 ‘not correct’ (adj.) high 碎 Structure-mismatched MN-[ej] se˧˦ swej˥˩ 打碎 ‘to break something’ (v.) high 配 Regular MN-[ej] ph e˧˦ ph ej˥˩ 搭配 ‘to match with’ (v.) high 贝 Regular MN-[ej] pe˧˦ pej˥˩ 宝贝 ‘treasure’ (n.) high 态 MN-[aj] th ɛ˧˦ th aj˥˩ 状态 ‘status’ (n.) high 呆 MN-[aj] tɛ˥˧ taj˥ 痴呆 ‘retarded’ (adj.) high 赛 MN-[aj] sɛ˧˦ saj˥˩ 决赛 ‘final competition’ (n.) high *The tone of the test items will change when embedded in a compound due to tone sandhi. TABLE B1. Critical items in study 2. s5 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 583.56 8.72 66.91 < 0.001 1582.71 19.57 80.88 < 0.001 LexSet = MN-[aj] — — — — −21.87 21.65 −1.01 0.390 LexSet = MN-[ej] (regular) — — — — −31.14 24.54 −1.27 0.190 Age = Old −50.66 4.83 −10.50 < 0.001 13.82 4.02 3.44 < 0.001 Sex = F 36.37 3.49 10.44 < 0.001 82.82 3.82 21.71 < 0.001 Onset = PHTH 26.46 6.54 4.05 0.006 — — — — Onset = PT −2.17 6.95 −0.31 0.760 — — — — LexSet = MN-[aj]: Sex = F — — — — 12.75 4.74 2.69 0.007 LexSet = MN-[ej] (regular): Sex = F — — — — 11.64 5.33 2.18 0.029 LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 547.98 13.60 40.29 < 0.001 1594.90 20.73 76.93 < 0.001 LexSet = MN-[aj] 78.16 10.53 7.42 < 0.001 −52.35 25.90 −2.02 0.042 LexSet = MN-[ej] (regular) −31.82 12.02 −2.65 0.036 37.35 29.36 1.27 0.170 Age = Old −20.29 9.11 −2.23 0.028 −3.86 6.44 −0.60 0.540 LexSet = MN-[aj]: Age = Old −34.64 14.08 −2.46 0.015 38.70 9.99 3.87 < 0.001 LexSet = MN-[ej] (regular): Age = Old 37.65 15.77 2.39 0.016 −16.77 11.18 −1.50 0.140 Sex = F 31.43 4.58 6.86 < 0.001 81.80 4.12 19.85 < 0.001 Onset = PHTH 17.01 4.67 3.64 0.021 — — — — Onset = PT −8.42 4.97 −1.69 0.160 — — — — LexSet = MN-[aj]: Sex = F — — — — 12.60 5.16 2.44 0.015 LexSet = MN-[ej] (regular): Sex = F — — — — 7.34 5.80 1.26 0.220 TABLE B2. Fixed-effect terms in the LM models on formant measures in the reading experiment, study 2. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) −0.26 0.30 −0.87 0.380 LexSet = MN-[aj] −2.53 0.44 −5.69 < 0.001 LexSet = MN-[ej] (regular) 1.10 0.38 2.92 0.004 Age = Old −0.94 0.27 −3.46 < 0.001 LexSet = MN-[aj]: Age = Old 1.46 0.57 2.57 0.010 LexSet = MN-[ej] (regular): Age = Old −0.71 0.47 −1.49 0.140 Onset = PHTH −0.14 0.16 −0.83 0.410 Onset = PT 0.64 0.17 3.71 < 0.001 TABLE B3. Fixed-effect terms in the GLM model on Diphthong in the reading experiment, study 2. Bold = p(|z|) < 0.05. s6 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 622.65 8.44 73.81 < 0.001 1574.97 14.85 106.06 < 0.001 Age = Old −56.69 5.69 −9.96 < 0.001 — — — — Sex = F 47.37 4.04 11.72 < 0.001 81.79 2.88 28.38 < 0.001 Onset = PHTH 43.28 5.32 8.14 < 0.001 — — — — Onset = PT −8.60 5.67 −1.52 0.170 — — — — LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 539.38 16.43 32.82 < 0.001 1657.96 16.46 100.76 < 0.001 LexSet = MN-[aj] 112.90 10.70 10.55 < 0.001 −87.78 9.15 −9.60 < 0.001 LexSet = MN-[ej] (regular) 7.47 12.23 0.61 0.570 5.82 10.35 0.56 0.590 Age = Old −4.37 9.83 −0.44 0.660 −33.61 6.01 −5.60 < 0.001 LexSet = MN-[aj]: Age = Old −41.66 15.03 −2.77 0.007 47.67 9.18 5.20 < 0.001 LexSet = MN-[ej] (regular): Age = Old 1.44 16.97 0.08 0.930 −18.47 10.37 −1.78 0.079 Sex = F 42.98 4.97 8.66 < 0.001 92.30 3.09 29.88 < 0.001 Onset = PHTH 23.62 4.51 5.24 0.001 — — — — Onset = PT −9.64 4.87 −1.98 0.100 — — — — TABLE B4. Fixed-effect terms in the LM models on formant measures in the translation experiment, study 2. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) 1.02 0.37 2.74 0.006 LexSet = MN-[aj] −3.34 0.38 −8.76 < 0.001 LexSet = MN-[ej] (regular) 0.50 0.38 1.33 0.180 Age = Old −0.97 0.28 −3.50 < 0.001 LexSet = MN-[aj]: Age = Old 1.36 0.50 2.73 0.006 LexSet = MN-[ej] (regular): Age = Old −0.11 4.50 −0.23 0.820 TABLE B5. Fixed-effect terms in the GLM model on Diphthong in the translation experiment, study 2. Bold = p(|z|) < 0.05. s7 APPENDIX C: MATERIALS AND MODELS IN STUDY 3 ITEM LEXICAL SET SHANGHAINESE CITATION FORM* (IN STAGE II) MANDARIN FORM EMBEDDING COMPOUND (AND PART OF SPEECH) EMBEDDING COMPOUND FREQUENCY 赔 Onset-mismatched MN-[ej] be˨˧ ph ej˧˥ 索赔 ‘to ask for indemnification’ (v.) low 陪 Onset-mismatched MN-[ej] be˨˧ ph ej˧˥ 不陪 ‘not to accompany’ (v.) low 备 Onset-mismatched MN-[ej] be˨˧ pej˥˩ 准备 ‘to prepare’ (v.) high 倍 Onset-mismatched MN-[ej] be˨˧ pej˥˩ 两倍 ‘twice’ (adj.) high 配 Regular MN-[ej] ph e˧˦ ph ej˥˩ 搭配 ‘to match with’ (v.) high 沛 Regular MN-[ej] ph e˥˧ ph ej˥˩ 充沛 ‘abundant’ (adj.) low 贝 Regular MN-[ej] pe˧˦ pej˥˩ 宝贝 ‘treasure’ (n.) high 狈 Regular MN-[ej] pe˧˦ pej˥˩ 狼狈 ‘in an extremely embarrassing state’ (adj.) low 态 MN-[aj] th ɛ˧˦ th aj˥˩ 状态 ‘status’ (n.) high 胎 MN-[aj] th ɛ˥˧ th aj˥ 保胎 ‘to protect the fetus’ (v.) low 呆 MN-[aj] tɛ˥˧ taj˥ 痴呆 ‘retarded’ (adj.) high 歹 MN-[aj] tɛ˥˧ taj˨˦ 为非作歹 ‘to do bad things’ (v.) low * The tone of the test items will change when embedded in a compound due to tone sandhi. TABLE C1. Critical items in study 3. s8 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 630.99 10.75 58.70 < 0.001 1512.92 21.08 71.76 < 0.001 LexSet = MN-[aj] −21.17 11.10 −1.91 0.110 50.17 19.80 2.53 0.021 LexSet = MN-[ej] (regular) −20.86 11.10 −1.88 0.110 22.45 19.80 1.13 0.260 Age = Old −56.82 4.00 −14.19 < 0.001 23.93 4.01 5.97 < 0.001 Sex = F 36.15 2.93 12.33 < 0.001 82.42 4.20 19.62 < 0.001 Onset = PHTH 13.74 4.97 2.76 0.030 17.66 8.84 2.00 0.057 Frq = H 14.90 7.01 2.12 0.075 — — — — LexSet = MN-[aj]: Frq = H −18.25 9.95 −1.83 0.110 — — — — LexSet = MN-[ej] (regular): Frq = H −27.66 9.92 −2.79 0.028 — — — — LexSet = MN-[aj]: Sex = F — — — — 12.08 5.04 2.40 0.018 LexSet = MN-[ej] (regular): Sex = F — — — — 9.94 5.01 1.98 0.053 LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 583.97 16.77 34.82 < 0.001 1572.45 23.71 66.31 < 0.001 LexSet = MN-[aj] 37.12 17.03 2.18 0.055 −33.39 29.72 −1.12 0.200 LexSet = MN-[ej] (regular) −18.35 17.01 −1.08 0.310 14.88 29.71 0.50 0.560 Age = Old −24.13 5.52 −4.37 < 0.001 −0.43 7.01 −0.06 0.970 Sex = F 27.22 4.14 6.58 < 0.001 88.05 3.14 28.03 < 0.001 LexSet = MN-[aj]: Age = Old — — — — 31.57 10.09 3.13 0.002 LexSet = MN-[ej] (regular): Age = Old — — — — 12.33 9.99 1.23 0.220 TABLE C2. Fixed-effect terms in the LM models on formant measures in the reading experiment, study 3. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) −0.44 0.45 −0.96 0.330 LexSet = MN-[aj] −2.23 0.50 −4.49 < 0.001 LexSet = MN-[ej] (regular) 0.85 0.47 1.82 0.069 Age = Old −0.79 0.18 −4.45 < 0.001 Sex = F 0.29 0.13 2.21 0.027 Onset = PHTH −0.45 0.23 −2.00 0.046 TABLE C3. Fixed-effect terms in the GLM model on Diphthong in the reading experiment, study 3. Bold = p(|z|) < 0.05. s9 LM MODEL ON F1START LM MODEL ON F2START β SE t pMCMC β SE t pMCMC (intercept) 665.84 8.22 81.00 < 0.001 1538.19 18.06 85.20 < 0.001 LexSet = MN-[aj] −32.66 6.50 −5.03 0.002 44.37 15.41 2.88 0.004 LexSet = MN-[ej] (regular) -8.00 6.47 −1.24 0.290 12.52 15.40 0.81 0.420 Age = Old −60.63 4.76 −12.73 < 0.001 — — — — Sex = F 46.89 3.41 13.75 < 0.001 83.42 2.69 31.01 < 0.001 Onset = PHTH 22.56 2.89 7.79 < 0.001 13.45 6.89 1.95 0.051 Frq = H −0.57 4.10 −0.14 0.910 — — — — LexSet = MN-[aj]: Frq = H 4.04 5.81 0.69 0.540 — — — — LexSet = MN-[ej] (regular): Frq = H −11.98 5.78 −2.07 0.077 — — — — LM MODEL ON F1END LM MODEL ON F2END β SE t pMCMC β SE t pMCMC (intercept) 574.42 17.45 32.91 < 0.001 1648.40 16.46 100.15 < 0.001 LexSet = MN-[aj] 77.97 11.79 6.61 < 0.001 −69.39 10.77 −6.44 < 0.001 LexSet = MN-[ej] (regular) −8.96 11.76 −0.76 0.450 −3.11 10.75 −0.29 0.790 Age = Old 2.44 9.83 0.25 0.820 −40.42 7.04 −5.74 < 0.001 Sex = F 39.10 4.31 9.07 < 0.001 92.10 3.12 29.56 < 0.001 Onset = PHTH 15.80 4.28 3.69 0.005 — — — — LexSet = MN-[aj]: Age = Old −44.27 13.87 −3.19 0.001 48.08 9.93 4.84 < 0.001 LexSet = MN-[ej] (regular): Age = Old −22.81 13.81 −1.65 0.100 8.86 9.89 0.90 0.380 TABLE C4. Fixed-effect terms in the LM models on formant measures in the translation experiment, study 3. Bold = pMCMC < 0.05. β SE z p(|z|) (intercept) 1.33 0.38 3.53 < 0.001 LexSet = MN-[aj] −3.61 0.35 −10.26 < 0.001 LexSet = MN-[ej] (regular) 0.14 0.31 0.46 0.640 Age = Old −1.56 0.29 −5.40 < 0.001 LexSet = MN-[aj]: Age = Old 1.99 0.46 4.36 < 0.001 LexSet = MN-[ej] (regular): Age = Old 0.68 0.41 1.66 0.096 TABLE C5. Fixed-effect terms in the GLM model on Diphthong in the translation experiment, study 3. Bold = p(|z|) < 0.05. [ctyaoyao@polyu.edu.hk] (Yao) [cc@bu.edu] (Chang) ...