This book explores the concept of complementation in the adjectival domain of English grammar. Alternation between non-finite complements, especially to infinitives and gerundial complements, has been investigated intensively on the basis of large corpora in the last few years. With very few exceptions, however, such work has hitherto been based on univariate analysis methods. Using multivariate analysis, the authors present methodologically innovative case studies examining a large array of explanatory factors potentially impacting complement choice in cases of alternation. This approach yields more precise information on the impact of each factor on complement choice as well as on interactions between different explanatory factors. The book thus presents a methodologically new perspective on the study of the system of non-finite complementation in recent English and variation within that system, and will be relevant to academics and students with an interest in English grammar, predicate complementation, and statistical approaches to language.
Author(s): Juho Ruohonen, Juhani Rudanko
Publisher: Palgrave Macmillan
Year: 2020
Language: English
Pages: 161
City: Cham
Acknowledgments
Contents
List of Figures
List of Tables
1 Introduction
References
2 Statistics and Complement Selection: A Case Study of Afraid Based on Canadian and British English
2.1 Introduction
2.2 Composition of the Strathy Corpus
2.3 Complementation of Afraid in the Strathy Corpus
2.3.1 Descriptive Statistics
2.3.2 The Limitations of Univariate Statistical Tests
2.3.3 Multivariate Analysis—Preliminaries
2.3.4 The Model-Selection Procedure
2.3.5 Model Selection
2.3.6 Model Performance: Classification and Discrimination
2.3.7 Model Interpretation
2.4 Complementation of Afraid in a Sample of the BNC
2.4.1 The 50-Million Word British Counterpart to Strathy
2.4.2 Searching the Corpus
2.4.3 Model Selection
2.4.4 Model Interpretation
2.5 Concluding Remarks
References
3 Non-finite Complements of Fear Adjectives Nested Within Too Complement Constructions in Present-Day English
3.1 Introduction
3.2 Data and Method
3.3 Four Types of Non-finite Complement
3.4 Syntactic and Semantic Indeterminacy
3.5 Conclusion
Appendix
References
4 Semantics, Syntax, and Horror Aequi as Predictors of Non-finite Alternation: A Multivariate Analysis of Clausal Complements of Afraid in the NOW Corpus
4.1 Introduction
4.2 Corpus
4.3 Data Collection
4.4 Analysis of the Data
4.4.1 Model Selection
4.4.2 Model Interpretation
4.4.2.1 Complexity Effects
4.4.2.2 Aequi Effects
4.4.2.3 Semantic Effects
4.4.2.4 Other Effects
4.5 Conclusion
Appendix—Regular Expressions Used in Corpus Searches
General to Infinitive Search
General of -ing Search
Extractions, Active Voice
Too-DCCs with an Embedded Complement of Afraid
Extractions, Passive Voice
General Supplementary Search for Passive Complements
Immediately Preceding to Infinitive Search
Immediately Preceding of -ing Search
Complement Negation Search
Pre-complement Insertion Search
Intra-complement Insertion Search
References
5 Factors Bearing on Non-finite Complement Selection: A Case Study of Accustomed with Data from Hansard
5.1 Introduction
5.2 Description of the Dataset
5.3 Multivariate Analysis—Model Selection
5.4 Model Interpretation
5.5 Conclusion
References
6 Factors Bearing on Infinitival and Gerundial Complements of the Adjective Prone in Current American and British English
6.1 Introduction
6.2 Data
6.3 Explanatory Variables
6.3.1 Animacy
6.3.2 Modal Scope
6.3.3 Futurity
6.3.4 Stativity
6.3.5 Repeatability
6.3.6 Grammatical Number
6.4 Statistical Analysis
6.4.1 Descriptive Statistics
6.4.2 Multivariate Analysis—Model Selection
6.4.3 Multivariate Analysis—Model Interpretation
6.5 Discussion
6.6 Conclusion
References
7 Conclusion
References
Index
