Corpus Analysis THEORETICAL REVIEW

According to Hutchinson and Waters 1987: 96, “evaluation is a matter of judging the fitness of something for a particular purpose”. In this case, it judges the fitness of a course book for particular learners, with their learning purpose, background knowledge, needs, and so on. It is not a matter of good or bad, but how a course book helps learners achieve the required purpose. “Evaluation is basically a matching process: matching needs to available solutions” p. 97. This matching needs to be done as objectively as possible. To do so, the needs and solutions should be looked separately. Although in the final analysis the choice made will be subjective, the subjectivity should not influence teachers’ judgment too soon without looking at possibly useful alternatives. A course book has to match with teachers and learners’ needs. Thus, subjective factors should not obscure objectivity.

5. Corpus Analysis

Yule 2010 explains corpus as a large collection of texts, spoken or written, typically stored as a database in a computer. Corpora are useful for linguist of all persuasion since they are useful resources for pursuing various researches. For example, Meyer 2004 did research in lexicography. He shows that lexicographers can effectively make dictionaries by studying word usage in very large linguistic corpora. It means that corpora help to provide the examples of usage of a word or phrase from many millions words of a text. Furthermore, McEnery and Wilson 2001: 107 say corpora that they can help to know how frequently the course book co-occurs and how many words provided in the course book . They also state that corpus being a source of ‘natural’ data; it is useful in defining evidently which forms are most frequently used in what aspect. Corpora provides rich textual information of a lexical item. With the use of corpus, it is easy to understand variability of word usage based on textual information. To summarize, Biber, Conrad and Reppen 2002 point out four basic characteristics of corpus based approach in the study of language aspects. First, corpus based analysis is empirical. The study analyses the actual patterns of use in natural texts. The focus here is on studying the use of language characteristics by considering appropriate association patterns. Second, it utilizes a large and principled collection of natural texts known as „corpus‟ as the basis for analysis. Third, it makes extensive use of computer for analysis, using both automatic and interactive techniques. Lastly, corpus based approach depends on both qualitative and quantitative analytical techniques. Quantitative involves the counts of linguistics features whereas qualitative is concerned with functional interpretation of the identified quantitative patterns. The qualitative explanation helps to explore the importance of such findings for learning about patterns of language use. Thus, the definition of corpus in this research is a collection text in written form and stored in electronic form. Then, it provides frequency data which later can be analyzed.

B. THEORETICAL FRAMEWORK