The Role of Cross-Situational Word Learning in Children's Vocabulary Acquisition: Theory, Behavior, and Mechanisms Pubblico

Abstract

Over the first 6 years of life, children are reported to have amassed a vocabulary of about 14,000 words. What are the processes that underlie such prolific learning? Based on a wealth of evidence built up over the past 40 years, one process appears to be fast mapping, children's ability to draw on a host of referential cues to infer a word's meaning from a single exposure to a new word. More recently, a growing body of evidence has suggested that another process is cross-situational word learning, word learners' ability to determine word meaning not within a single encounter but across multiple encounters by tracking the cross-situational consistency between words and their candidate referents. Although the notion that children acquire their vocabulary at least in part through cross-situational learning is neither novel nor unintuitive, theoretical treatments and empirical investigations into this learning process are scarce relative to those of fast mapping.

The overarching goal of the research presented herein is to further investigate the nature of children's cross-situational learning capacities and to better understand the role of this type of learning in vocabulary acquisition. Two behavioral experiments that examine 5- to 7-year-olds' cross-situational word learning are reported. These experiments constitute the first empirical investigations of school-aged children's ability to acquire new word-to-referent mappings when the only cues to reference are the cross-situational co-occurrence statistics between words and their referents. These studies also examine whether some of the behavioral signatures previously observed in adults' cross-situational learning are also evident in children's learning. Additionally, a series of computational simulations that explore the candidate mechanisms underlying children's cross-situational word learning are reported. Although the results of these simulations do not provide a conclusive mechanistic account of children's learning, they do specify some conditions that readily account for the observed learning patterns and accurately predict other empirical phenomena. Collectively, the research presented here contributes to our understanding of the various processes that make children's impressive word learning possible.

Table of Contents

Chapter 1. Introduction...1
Chapter 2. Fast Mapping and Cross-Situational Approaches to Children's Word Learning...4

2.1 A Fast Mapping Approach to the Mapping Problem...5
2.2 A Cross-Situational Approach to the Mapping Problem...9

2.2.1 Cross-situational word learning: Behavioral evidence...10
2.2.2 Cross-situational word learning: Computational evidence...16

2.3 Outstanding Questions...18

2.3.1 Scaling from adult word learning to children's word learning...19
2.3.2 The nature of the underlying learning mechanisms...21

Chapter 3. The Effect of Contextual Diversity on Children's Cross-Situational Word Learning...24

3.1 Background...24
3.2 Method...29

3.2.1 Participants...29
3.2.2 Stimuli...30
3.2.3 Design...31
3.2.4 Procedure...33

3.2.4.1 Familiarization phase...33
3.2.4.2 Learning phase...35
3.2.4.3 Test phase...36

3.2.5 Coding...37

3.3 Results...38
3.4 Discussion...42

3.4.1 Cross-situational word learning in children's vocabulary growth...43
3.4.2 Statistical learning in adult and child learners...45
3.4.3 Why does contextual diversity aid learning?...47
3.4.4 Conclusion...49

Chapter 4. Probing the Precision of Children's Mappings in Cross-Situational Word Learning...50

4.1 Background...50
4.2 Method...51

4.2.1 Participants...51
4.2.2 Stimuli, Design & Procedure...51

4.3 Results...52
4.4 Discussion...55

Chapter 5. Children's Cross-Situational Word Learning: Hypothesis Testing or Associative Learning?...59

5.1 Background...59

5.1.1 Hypothesis Testing...61
5.1.2 Associative Learning...63
5.1.3 Current Endeavor...65

5.2 Models...66

5.2.1 Hypothesis Testing (HT) Model...66
5.2.2 Associative Learning (AL) Model...70

5.3 Simulation 1: The role of contextual diversity in cross-situational word learning...73

5.3.1 Method...73
5.3.2 Results...75

5.3.2.1 HT model results...75
5.3.2.2 AL model results...76
5.3.2.3 Correlations between models and children's performance...76

5.3.3 Discussion...80

5.4 Simulation 2: Item selection as a function of cross-situational co-occurrence statistics...81

5.4.1 Method...81
5.4.2 Results...82

5.4.2.1 HT model results...82
5.4.2.2 AL model results...83
5.4.2.3 Correlations between models and children's performance...84

5.4.3 Discussion...87

5.5 Simulations 3 and 4: A Modified Associative Learning Model...88

5.5.1 Method...90
5.5.2 Results...91

5.5.2.1 Contextual diversity effects in MAL models...91
5.5.2.2 Item selection effects in MAL models...92
5.5.2.3 Correlations between MAL models and children's performance...93

5.5.3 Discussion...94

5.6 Simulation 5: Accounting for other empirical data...95

5.6.1 Method...96
5.6.2 Results and Discussion...97

5.7 General Discussion...98

5.7.1 Empirical evidence for the proposed parameters that improve model fit...99
5.7.2 Implications for the hypothesis testing vs. associative learning debate...102
5.7.3 Conclusion...105

Chapter 6. General Discussion...107

6.1 Contributions of a Cross-Situational Approach...108

6.1.1 Cross-situational co-occurrence statistics as an additional cue to learning...109
6.1.2 The role of cross-situational learning in the discovery of fast mapping cues...112
6.1.3 Cross-situational learning as a common theme underlying vocabulary acquisition research programs...116
6.1.4 Cross-situational learning as statistical learning: Towards a unified account of language acquisition...117

6.2 Limitations to the Cross-Situational Approach...121

6.2.1 Can cross-situational learning scale up to real-world learning environments?...121
6.2.2 Is cross-situational learning applicable to the acquisition of non-object labels?...123
6.2.3 Is cross-situational learning ability related to real-world vocabulary growth?...126

6.3 Conclusions...128

About this Dissertation

Rights statement

• Permission granted by the author to include this thesis or dissertation in this repository. All rights reserved by the author. Please contact the author for information regarding the reproduction and use of this thesis or dissertation.

School	Laney Graduate School
Department	Psychology
Degree	PhD
Submission	Dissertation
Language	English
Research Field	Psychology, Developmental Psychology, Cognitive
Parola chiave	Cross-Situational Word Learning Vocabulary Acquisition Word Learning Statistical Learning
Committee Chair / Thesis Advisor	Namy, Laura, Emory University
Committee Members	Lourenco, Stella, Emory University Nygaard, Lynne, Emory University Wolff, Phillip, Emory University Hampton, Robert, Emory University

Ultima modifica

Primary PDF

Thumbnail	Title	Date Uploaded	Actions
	The Role of Cross-Situational Word Learning in Children's Vocabulary Acquisition: Theory, Behavior, and Mechanisms ()	2018-08-28 13:25:09 -0400	Press to Select an action Download

Supplemental Files

Thumbnail	Title	Date Uploaded	Actions