KarinWanrooijMA2009.doc-PhoneticSciences,Amsterdam.docVIP

Does Distributional Input Improve the Categorization of Speech Sounds? Neurobiological Aspects and Computer Simulations. Master Thesis Linguistics University of Amsterdam Supervisor: Prof. dr. Paul Boersma Second reader: Dr. David Weenink Student: Karin Wanrooij Student number: 0512133 Date: July 22, 2009 TABLE OF CONTENTS Introduction 4 PART I: SETTING 5 The relevance of studying distributional learning 6 Introduction Distributional input Categorization 11 Questions 13 Methodology 14 PART II: NEUROBIOLOGICAL ASPECTS 15 Coding mechanisms along the auditory pathways 16 Introduction Anatomical overview of the auditory pathways Sketch of coding along the pathways 17 Information transport by means of electrical activity 18 Coding by single neurons 20 Characteristic Frequency Tonotopic order and place code 21 Firing pattern, rate code and temporal code 23 Neural ensemble coding 24 Acoustic resolution at different levels for different speech sounds 26 Different pathways for auditory who, what and where information Summary 27 Auditory maps in adults 28 Introduction Anatomical sketch of the auditory cortex Tonotopicity as a central organizational principle 29 Overlapping maps for other parameters 30 The coding of speech sounds in A1 33 Animal studies Studies on human perception 35 Summary 37 Plasticity in the auditory cortex 38 Introduction Adaptive plasticity in animal infants A1 Studies on tonotopic maps in rat pups Normal development Abnormal development triggered by specific inputs 39 Conclusion 41 Adaptive plasticity in adults A1 42 Continuous plasticity Training versus passive exposure 43 Generalization to novel stimuli Conclusion Underlying mechanisms of cortical plasticity 45 Hebbian and other forms of synaptic plasticity Synaptic plasticity and synaptogenesis during development 46 S