Corpus Methodologies Explained by Ji, Meng -- Read -- Imperial Library of Trantor

Index

Cover Half Title Series Information Title Page Copyright Page Contents Acknowledgements List of tables List of figures Introduction to Corpus Methodologies Explained: An empirical approach to translation studies

References

1 The need for corpora in machine translation

1. Introduction 2. Paradigms for machine translation

2.1 Rule-based machine translation 2.2 Machine translation approaches which depend on parallel corpora 2.3 Translation memories 2.4 Translation memories and machine translation 2.5 Example-based machine translation 2.6 Statistical machine translation

2.6.1 The translation model 2.6.2 The target language model 2.6.3 The decoder 2.6.4 Phrase-based models

3. Europarl: A parallel corpus for statistical machine translation 4. Matching against a database of examples

4.1 Measures used in search engine technology 4.2 Incorporation of linguistic information into matching algorithms

5. Alignment 6. Automatic generalisation of translation examples

6.1 Clustering of words with similar contexts 6.2 Paraphrasing using a pivot language

7. Bilingual term alignment for probabilistic dictionaries 8. Development of machine translation systems for less-resourced languages

8.1 Developing a Tamil to English machine translation system from scratch 8.2 Machine translation for Cebuano into English 8.3 Machine translation for Spanish and Mapudungun

9. Evaluation of machine translation output

9.1 Measures based on human judgements 9.2 Automatic measures based on reference corpora

9.2.1 BLEU 9.2.2 Word Error Rate and Translation Error Rate 9.2.3 METEOR 9.2.4 Recent measures: ParaEval and Overlap

9.3 Meta-evaluation

9.3.1 Inter-annotator agreement 9.3.2 Other qualities of a good measure

9.4 Quality estimation

10 Conclusion References

2 A multidimensional analysis of the translational Chinese genre system

1. Research question investigated – genre shifting in translation 2. Corpus data and resources used

2.1 Original and translational Chinese database

3. Principal component analysis of genres in the BNC, LCMC and ZJU corpora

3.1 PCA analysis of written genres in the BNC 3.2 Corpus findings: Distribution of genres in the four-dimensional PCA model

Dimension 1: High volume of information and lexical richness Dimension 2: Temporality of information and foreign content Dimension 3: Logical structure and the coherence of texts Dimension 4: Possibility and likelihood of the events reported

3.3 PCA analysis of textual genres in LCMC

Dimension 1 Dimension 2 Dimension 3 Dimension 4 Dimension 5 Dimension 6 Dimension 7 Dimension 8 Dimension 9 Dimension 10

3.4 PCA analysis of textual genres of translational Chinese represented by the ZJU corpus

Dimensional features Distribution of Chinese textual genres within the nine-dimensional PCA model

4. Similarity analysis of original and translational Chinese genres 5. Five types of genre shifting in translational Chinese 6. Conclusion References

3 Translator style: A corpus-assisted approach

1. What is translator style? 2. ST- vs TT-oriented studies of translator style 3. Sense-making, ‘thick description’ and innovation

3.1 Magic of numbers 3.2 Mimicking of research designs 3.3 Mystery of corpus

4. Contextualizing translator’s style

4.1 Hongloumeng and its English translations 4.2 The Chinese-English comparable/parallel corpus of Hongloumeng 4.3 Type-token ratio and sentence length in the English translations 4.4. Making sense of the statistical data

4.4.1 Two groups of translators 4.4.2 Why Hawkes used more words 4.4.3 Why the Yangs used a wider range of vocabulary 4.4.4 Why Hawkes used longer sentences

4.5 Summary

5. Conclusion Notes References

4 The translation of formal source-language lacunas: An empirical study of the Over- representation of Target-Language Specific Features and the Unique Items hypotheses

1. Introduction 2. Theoretical framework

2.1. The Spanish gerund and its Norwegian counterparts 2.2. Tense, aspect and lexical aspect: General definitions

2.2.1 The Spanish gerund – aspect 2.2.2 The Norwegian resources believed to trigger the use of Spanish gerunds in translations from Norwegian: Aspect and lexical aspect in Norwegian 2.2.3 Lexical aspect or Aktionsart in Norwegian

2.3. The hypotheses and how to test them

3. Material and method

3.1 Material

3.1.1 Presentation of the categories of analysis

3.2 Procedure for categorizing the Norwegian structures that correspond to the Spanish gerund

4. Results and discussion

Finite verbs Finite verb + og + finite verb Finite verb + å + infinitive Finite verb modified by an adverb Prepositions with complements Present participle Additions Relative sentences Idiomatic expressions Prefixes 4.1. Statistical testing of the data

5 Future studies and concluding remarks

5.1 Future studies 5.2 Concluding remarks

Appendix Notes References

5 Is there gravitational pull in translation? A corpus-based test of the Gravitational Pull Hypothesis on the language pairs Norwegian-Spanish and English-Spanish

1. Introduction 2. Theoretical background: The Gravitational Pull Hypothesis, corpus-based translation studies and contrastive linguistics

2.1 The Gravitational Pull Hypothesis

2.1.1 Predictions of the Gravitational Pull Hypothesis

2.2 The Spanish gerund and its English counterparts

2.2.1 The progressive in English and Spanish 2.2.2 Formal overlap: the Spanish estar+gerund and the English progressive 2.2.3 Functional overlap: The Spanish simple gerund and the English non-finite adverbial phrases 2.2.4 Possible grammatical overlap: The motion progressive 2.2.5 Not overlapping: Progressive with postural verbs 2.2.6 Summary of the overlap between the Spanish gerund and English resources 2.2.7 Predictions based on the overlap in English and Spanish grammatical resources

3. Material and method

3.1 Material 3.2 Method

3.2.1 Categories derived from the analysis of the corpus data 3.2.2 Procedure for categorizing the English structures that correspond to the Spanish gerund