| 2 |
Perception of learner errors and non-standard features in the native and non-native language: evaluation vs. processing cost ...
|
|
|
|
Abstract:
Table of Contents 1. Introduction 1.1. Research questions and methods 1 1.2. A sociolinguistic approach to errors 4 1.3. A cognitive approach to errors 6 1.4. The structure of the thesis and its application 10 2. Error studies in Applied Linguistics 2.1. Error studies in Second Language Acquisition 2.1.1. Error Analysis 11 2.1.2. The distinction between errors and mistakes 16 2.1.3. Grammaticality and acceptability 17 2.2. Error Gravity 2.2.1. The sociolinguistic factors determining error gravity 21 2.2.2. Effects of the rater characteristics 25 2.2.3. Hierarchies of error types 28 2.3. Evaluation of native speakers’ errors 2.3.1. Writer’s identity 32 2.3.2. Error evaluation outside the academia 35 2.4. Summary 40 3. A psycholinguistic approach to error processing 3.1. Attempts to predict error gravity 41 3.2. Processing concerns underlying second language acquisition 44 3.3. Introduction to eye movement research 48 3.4. Word recognition and its implication for error processing 51 3.5. Real-time processing ... : Main Findings and Claims: 1. The approach to studying errors (or breaches of the code) has changed in the last 20 years, but the perception by lay people and the interest of non-linguists remained the same. In my study, old questions are researched with new tools. 2. Error hierarchies are useless. Instead, reactions to errors can be predicted based on the effects of several factors in individual cases. 3. Error evaluations (elicited by questionnaires) correlate with the error processing cost (elicited by eye-tracking) not more than 40%. 4. Lower evaluations are not necessarily determined by greater processing costs. Some ‘unacceptable’ errors are easily processed, and other well-acceptable items are looked at much longer. 5. The eye movement method is limited because it does not explain why people are looking longer. Eye-trackers should be synchronized at least with ERP. 6. Important findings: 1) reading times increase primarily for surface-level errors (not pragmatics or tense-aspect inconsistencies); 2) ...
|
|
URL: https://dx.doi.org/10.6094/unifr/15056
https://freidok.uni-freiburg.de/data/15056
|
|
BASE
|
|
Hide details
|
|
|
| 3 |
Perception of learner errors and non-standard features in the native and non-native language: evaluation vs. processing cost
|
|
|
|
BASE
|
|
Show details
|
|
|
|
|
