ebook img

Figure 5.4: Model predictions for identical consonants PDF

185 Pages·2013·6.26 MB·English
by  
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Figure 5.4: Model predictions for identical consonants

UC San Diego UC San Diego Electronic Theses and Dissertations Title Phonotactic probability in Amharic : : a psycholinguistic and computational investigation Permalink https://escholarship.org/uc/item/32v7w7gn Author Colavin, Rebecca Irene Victoria Publication Date 2013 Peer reviewed|Thesis/dissertation eScholarship.org Powered by the California Digital Library University of California UNIVERSITY OF CALIFORNIA, SAN DIEGO Phonotactic probability in Amharic: a psycholinguistic and computational investigation A dissertation submitted in partial satisfaction of the requirements for the degree Doctor of Philosophy in Linguistics by Rebecca Irene Victoria Colavin Committee in charge: Professor Sharon Rose, Co-Chair Professor Roger Levy, Co-Chair Professor Eric Bakovic Professor Gary Cottrell Professor Robert Malouf 2013 Copyright Rebecca Irene Victoria Colavin, 2013 All rights reserved. The dissertation of Rebecca Irene Victoria Colavin is ap- proved, and it is acceptable in quality and form for pub- lication on microfilm and electronically: Co-Chair Co-Chair University of California, San Diego 2013 iii DEDICATION To Francis, who always says “be happy” iv EPIGRAPH "With the exception of the 'explosive sounds' represented by p̣, q, ṭ, c,̣̌ ṣ and the guttural χ (frequently pronounced k by Abyssinians themselves) the pronunciation of Amharic presents little difficulty to an Englishman". —Armbruster (1908) v TABLE OF CONTENTS Signature Page . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iii Dedication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . iv Epigraph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Table of Contents . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vi List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xi Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xiii Vita . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii Abstract of the Dissertation . . . . . . . . . . . . . . . . . . . . . . . . . . . xviii 1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 2 Phonotactic models . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.1 Phonotactics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 2.2 The generative linguistic tradition . . . . . . . . . . . . . . . . . . . 9 2.3 Experimental evidence . . . . . . . . . . . . . . . . . . . . . . . . . 11 2.4 Accounting for gradient speaker judgements . . . . . . . . . . . . . 12 2.4.1 The Generalized Neighbourhood Model . . . . . . . . . . . . 15 2.4.2 Low-order n-gram models . . . . . . . . . . . . . . . . . . . 16 2.5 Accounting for unattested sequences . . . . . . . . . . . . . . . . . 19 2.5.1 The Hayes and Wilson Maxent Phonotactic learner . . . . . 23 2.5.2 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . 33 3 Overview of Amharic morphology and phonotactics . . . . . . . . . . . . 35 3.1 Morphology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 3.2 The Obligatory Contour Principle . . . . . . . . . . . . . . . . . . . 40 3.3 The Obligatory Contour Principle as applied to place of articulation 45 3.4 Dictionary study . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 3.4.1 Consonant inventory . . . . . . . . . . . . . . . . . . . . . . 49 3.4.2 OCP-Place . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 3.5 Experimental evidence . . . . . . . . . . . . . . . . . . . . . . . . . 55 3.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60 vi 4 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62 4.1 Simulation I: baseline model . . . . . . . . . . . . . . . . . . . . . . 63 4.1.1 Training data . . . . . . . . . . . . . . . . . . . . . . . . . . 64 4.1.2 Test data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67 4.1.3 Procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . 68 4.1.4 Model evaluation . . . . . . . . . . . . . . . . . . . . . . . . 69 4.1.5 Results and discussion . . . . . . . . . . . . . . . . . . . . . 72 4.2 Simulation II: modified model . . . . . . . . . . . . . . . . . . . . . 79 4.2.1 Results and discussion . . . . . . . . . . . . . . . . . . . . . 82 4.3 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86 5 Experiment I: Co-occurrence patterns . . . . . . . . . . . . . . . . . . . . 88 5.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88 5.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 89 5.2.1 Novel verb construction . . . . . . . . . . . . . . . . . . . . 90 5.2.2 Presentation . . . . . . . . . . . . . . . . . . . . . . . . . . . 99 5.2.3 Participants . . . . . . . . . . . . . . . . . . . . . . . . . . . 101 5.3 Model predictions for Experiment I . . . . . . . . . . . . . . . . . . 101 5.4 Experimental results . . . . . . . . . . . . . . . . . . . . . . . . . . 104 5.5 Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 108 5.5.1 Speaker ratings . . . . . . . . . . . . . . . . . . . . . . . . . 108 5.5.2 Simulations . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 5.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 6 Experiment II: Segmental irregularities . . . . . . . . . . . . . . . . . . . 115 6.1 Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115 6.2 Methodology . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119 6.2.1 Nonce word selection . . . . . . . . . . . . . . . . . . . . . . 121 6.2.2 Presentation and speakers . . . . . . . . . . . . . . . . . . . 122 6.3 Model predictions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 6.3.1 Part A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122 6.3.2 Part B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 124 6.4 Results and discussion . . . . . . . . . . . . . . . . . . . . . . . . . 125 6.4.1 Part A . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126 6.4.2 Part B . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130 6.5 General discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 6.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139 Appendix A: Experimental items I . . . . . . . . . . . . . . . . . . . . . . . 142 Appendix B: Experimental items II . . . . . . . . . . . . . . . . . . . . . . . 150 vii Appendix C: Word rating screen . . . . . . . . . . . . . . . . . . . . . . . . 155 Appendix D: Experiment instructions . . . . . . . . . . . . . . . . . . . . . . 156 Bibliography . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 158 viii LIST OF FIGURES Figure 2.1: Structure of the (English) word space . . . . . . . . . . . . . . 10 Figure 2.2: Graphical representation of a bigram HMM . . . . . . . . . . . 21 Figure 3.1: Prohibited derivations . . . . . . . . . . . . . . . . . . . . . . . 41 Figure 3.2: Licensed derivations . . . . . . . . . . . . . . . . . . . . . . . . 42 Figure 3.3: OCP gradiency . . . . . . . . . . . . . . . . . . . . . . . . . . . 47 Figure 4.1: Log-likelihood of training data . . . . . . . . . . . . . . . . . . 73 Figure 4.2: Correlation between model predictions and speaker judgements 74 Figure 4.3: Speaker predictions and model performance for OCP-Place . . . 74 Figure 4.4: Speakers judgements for LA violations . . . . . . . . . . . . . . 78 Figure 4.5: Automatic model predictions for LA violations . . . . . . . . . 79 Figure 4.6: Log-likelihood of data: simulation II . . . . . . . . . . . . . . . 83 Figure 4.7: Speaker judgment correlation: simulation II . . . . . . . . . . . 84 Figure 4.8: Speaker predictions and model performance for OCP-Place . . . 84 Figure 4.9: Automatic model predictions for LA violations: simulation II . 85 Figure 5.1: Distribution of controls compared to real words . . . . . . . . . 95 Figure 5.2: Selection of nonce forms for the ocp-labial-left edge condition . 97 Figure 5.3: Model predictions for OCP-Place violations . . . . . . . . . . . 102 Figure 5.4: Model predictions for identical consonants . . . . . . . . . . . . 103 Figure 5.5: Model predictions for non-homorganic fricative sequences . . . . 104 Figure 5.6: Speaker ratings for OCP-Place . . . . . . . . . . . . . . . . . . 106 Figure 5.7: Speaker ratings for identical consonants . . . . . . . . . . . . . 107 Figure 5.8: Speaker ratings for non-homorganic fricatives . . . . . . . . . . 107 Figure 5.9: Comparison of speaker ratings and model predictions for OCP- Placepatterns(homorganic,non-identicalconsonants),rarepat- terns of identical consonants (ABA and AAB), frequent pat- terns of identical consonants (ABB) and non-homorganic frica- tives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112 Figure 5.10: Comparison of speaker ratings and model predictions . . . . . . 113 Figure 6.1: Model predictions for Part A . . . . . . . . . . . . . . . . . . . 123 Figure 6.2: Model predictions for Part B . . . . . . . . . . . . . . . . . . . 124 Figure 6.3: Model predictions for Part B (by segment) . . . . . . . . . . . . 125 Figure 6.4: Over-represented segment [r] . . . . . . . . . . . . . . . . . . . 126 Figure 6.5: Normally-represented segment [t] . . . . . . . . . . . . . . . . . 128 Figure 6.6: [w] in C and C . . . . . . . . . . . . . . . . . . . . . . . . . . 129 1 2 Figure 6.7: Predictions and ratings for Part A . . . . . . . . . . . . . . . . 130 Figure 6.8: Predictions and ratings for Part A (detail) . . . . . . . . . . . . 131 Figure 6.9: Nonce words with [t] (Part B) . . . . . . . . . . . . . . . . . . . 132 Figure 6.10: Nonce words with [w] . . . . . . . . . . . . . . . . . . . . . . . 133 ix

Description:
Phonotactic probability in Amharic : : a psycholinguistic and computational investigation. Permalink .. Figure 4.9: Automatic model predictions for LA violations: simulation II . 85. Figure 5.1: Distribution of controls .. Bachelor of Arts, San Diego State University, CA. 2004. San Diego State
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.