Scholarly article on topic 'In Defense of Concordancing: An Application of Data-Driven Learning in Taiwan'

In Defense of Concordancing: An Application of Data-Driven Learning in Taiwan Academic research paper on "Languages and literature"

CC BY-NC-ND
0
0
Share paper
OECD Field of science
Keywords
{"Data-Driven Learning" / concordancing / "self-compiled corpus" / prepositions / CALL / "Hary Potter"}

Abstract of research paper on Languages and literature, author of scientific article — Hsing-chin Leel

Abstract Data-Driven Learning (DDL) is an approach in which the language learners are also research workers whose learning is driven by access to linguistic data (Johns, 1991:2). This essay aims at evaluating the potential of the DDL approach using concordance compiling software in language teaching in Taiwan. The advantages and disadvantages of using concordancing in the classroom and the values and limitations of using small corpora are discussed. A CALL program is explored in language teaching, particularly in the teaching of J.K. Rowling's uses of the preposition in Harry Potter and the Philosopher's Stone.

Academic research paper on topic "In Defense of Concordancing: An Application of Data-Driven Learning in Taiwan"

Available online at www.sciencedirect.com

\4 ScienceDirect

Procedia Social and Behavioral Sciences 12 (2011) 399-408

International Conference on Education and Education Psychology 2010

In Defense of Concordancing: An Application of Data-Driven

Learning in Taiwan

Hsing-chin Leel*

National Taipei College of Business, No.321, Sec. 1, Chi-Nan Road,Taipei, 100, Taiwan

Abstract

Data-Driven Learning (DDL) is an approach in which the language learners are also research workers whose learning is driven by access to linguistic data (Johns 1991:2). This essay aims at evaluating the potential of the DDL approach using concordance compiling software in language teaching in Taiwan. The advantages and disadvantages of using concordancing in the classroom and the values and limitations of using small corpora are discussed. A CALL program is explored in language teaching, particularly in the teaching of J.K. Rowling's uses of the preposition in Harry Potter and the Philosopher's Stone.

© 2009 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Dr. Zafer Bekirogullari of Y.B.

Keywords:Data-Driven Learning, concordancing, self-compiled corpus, prepositions, CALL, Hary Potter

1. Introduction

Johns holds the notion that "every learner is a Sherlock Holmes" and he has developed a data-driven approach to help learners become the best "language detectives" (1991, p. 1). Data-Driven Learning (DDL) is a pedagogical approach in which the language learner is also a research worker whose learning is driven by access to contextualised linguistic data (Johns, 1991; Johns, Lee & Wang, 2008). Being a research worker, the language learner will discover and analyse rules and patterns from authentic data rather than being given language descriptions by the teacher (Johns, 1988; Gavioli, 1997, 2001). The DDL approach makes use of concordance compiling software, which is able to create keyword-in-context (KWIC) concordance of lexical items. The KWIC format concordance allows the learner to "identify-classify-generalise" (Johns, 1991, p. 4). The following discussion will evaluate the potential application of the DDL approach in language teaching in Taiwan. The advantages and disadvantages of using concordancing in the classroom and the values and limitations of using small corpora are discussed. A CALL program, CONTEXTS (Johns, 1997), was chosen for this study as it enables the teacher to compose a corpus of authentic text which the students can then annotate using the concordancing software. The Author created her own materials based on a self-compiled small corpus, which could be adapted to the CONTEXTS program, to teach English prepositions. The possibility of using concordancing in a variety of different ways for

* Corresponding author. Tel.: +886-2-23226061; fax: +886-2-23226067. E-mail address: hsingchinlee@gmail.com

1877-0428 © 2011 Published by Elsevier Ltd. Selection and/or peer-review under responsibility of Dr. Zafer Bekirogullari of Y.B. doi:10.1016/j.sbspro.2011.02.049

teaching EFL college students in Taiwan, as well as its applicability as a teaching methodology to other computer software and materials, is also discussed.

Conrad (2005) informs us that there is a "need for more empirical studies on the impact of using corpus materials and techniques in the classroom" (p. 404). Corpus researchers have revolutionized English language teaching with the investigation of various language patterns as exemplified by such researchers as Holmes (1988), who examined modal verbs, Boxer and Pickering (1995) who looked at speech acts in textbook dialogues, McCarthy (1998) who examined the past perfect verb forms, and Romer (2005) who compared the use of progressive forms. Cheng and Warren (2007) suggest that textbook writers should pay more attention to the routinized everyday use of English. However, some researchers cast doubts on the usefulness of corpus-based material in the language classroom (Seidlhofer, 1999; Seidlhofer, 2003; Widdowson, 2000; Sripicharn, 2002). For example, Sripicharn (2002) concluded that there was no significant improvement between the experimental group (using DDL materials) and the controlled group in terms of text results.

Based on the aforementioned concerns, the following questions have been addressed in this research: to what extent can students be responsible for what they learn? How much can teachers help to provide techniques for students? Can this approach to learning be combined with more traditional approaches? And finally, how can Data-Driven Learning be adapted to the needs of vocational college students learning English in Taiwan?

2. In Defense of Concordancing: New Approach to Language Teaching in Taiwan

The most beneficial aspect about using DDL is that, instead of being a top-down instructor, the teacher plays the role of "a co-ordinator of student-initiated research" (Johns, 1991, p. 3). Data-Driven Learning is claimed to make students more active learners and less dependent on teachers and textbooks. Using concordancing software is a breakthrough in language teaching which helps students to develop their ability to observe, to speculate and to identify patterns in the target language (Johns, 1988). Students can learn through the data collection to make generalisations using an inductive approach (Johns, 1991). This is especially suitable for English as a foreign language (EFL) students who have received most of their education through a medium other than English. The students can either work with the concordancer in groups or individually to interact with the texts and to develop appropriate learning strategies. In particular, the use of concordancing software offers a great contrast to the grammar teaching methodologies in Taiwan where students learn the contents of the textbook passively and mechanistically. Students need very clear instructions to cope with language structures and this is the reason why most teachers still use the traditional grammar-translation approach to teaching English in Taiwan.

As Kennedy (1991) points out, "although the most comprehensive grammar books and dictionaries have already provided the full information of grammatical functions, the DDL provides a statistical aspect of linguistic description to both support and contradict the learners' intuitions about the use of language" (p. 110). The DDL approach promotes "grammatical consciousness-raising" (Rutherford, 1987) by looking at a number of examples from the concordance lines which familiarise with lexical items. For example, by creating lists of prepositions used in context, the concordancer allows the students to see that the same language terms can be used in a variety of different manners, and by demonstrating the language in use this feeds the students' guesswork. Another result of using this approach is that students will learn better when they organise grammatical descriptions through their own engagement with the evidence of concordance lines, quite unlike the traditional pedagogical description which presents language as a set of laws (Johns, 1991).

Although grammar-based teaching preponderates in Taiwan, Willis' (1990) notion of a lexical syllabus related to DDL highlights the need for change in Taiwanese teaching strategies. Willis emphasises the importance of teaching the most common words in English. He also suggests that teachers should offer learners opportunities to experience the most useful patterns of the language time after time. Using the concordancer allows the students to observe how certain linguistic features or lexical items are signalled and emphasised by "repeated expression at sentence level" (Love & Morrison, 1991, p. 125). For instance, the study of vocabulary is often a problematic task. In Taiwan, students learn English vocabulary by rote memory. Without the chance to put the newly learned vocabulary into practice, students tend to learn and then forget most of the vocabulary. In general, they have very limited varieties of verb groups and often fall into repetitive language practices, rarely using synonyms. Students do not have the chance to realise that verbs in different contexts give different meanings. Take J. K. Rowling's (1997) use of the preposition on, from Harry Potter and the Philosopher's Stone as an example - "Term begins on 1st September / ... testing Ron

on astronomy" is a pair of uses difficult for students to distinguish. They are likely to write "Term begins in 1st September/ ... testing Ron with astronomy". The concordancing program CONTEXTS can be effectively used to reinforce an understanding of such disparate use of phrasal verbs by examining the concordance lines.

3. Advantages of Classroom Concordancing for Data-driven Learning

O'Keeffe, McCarthy, and Carter. (2007) stress the notion of using corpus in the classroom and what a corpus can tell us about vocabulary and language learning and teaching. This present study investigates the use of the text of Harry Potter and the Philosopher's Stone (Rowling, 1997). The reason why Harry Potter was chosen as supplementary reading material was so that the students could be simultaneously motivated to cultivate the habit of reading English books to improve their reading comprehension and also to enjoy and be interested in the process. The conventional course books present too few examples to help students acquire enough language. It has been suggested that the main purpose of the concordancer experiment is "to stimulate in students a wider and more flexible range of skills in relation to a subject text" (Love & Morrison, 1991, p. 130).

A concordancer searches for occurrences of a given word, part of a word, combination of words, a punctuation mark, affix, or a phrase or structure within a given text corpus to show its immediate context (Granger, 1998; Biber, Conrad & Reppen, 1998). One of the major advantages of using concordancing software, is "the automatic retrievability of extra context for individual events and repeated contexts over a wide range of data samples" (McCarthy & Carter, 2004, p. 159). As a teaching methodology which favours learning by discovery, concordancing helps learners to analyze the structure of literary writing. The students can either work with the concordancer in groups or individually to interact with the texts actively and analytically and develop appropriate learning strategies. It allows students to question, explore the word forms, word usage, vocabulary, grammatical features, syntax, agreement, stylistics and collocation.

Concordancing provides learners with a way of exploring the uses of language by offering numerous examples of genuine language. Concordances give students easy and immediate access to genuine language production using many different styles and genres (Kettemann, 1995). Through concordancing, "even students without any previous background in literature can begin to understand the way a writer works with language and the effects that can be achieved across a text" (Tribble & Jones, 1990). A concordance is a powerful hypothesis testing device on a massive amount of data. It allows controlled speculation, makes patterns of language use easy to detect and thus enhances the learners' imagination and fosters inductive thinking and exploratory learning (Kettemann, 1995). The learning process becomes more exploratory and motivating and even highly experiential. As the learners are given numerous examples regarding the syntactic-semantic forms, they inductively generalize the rules and then test their hypotheses about specific language points. As a result of this kind of text analysis, learners can use the concordance to increase their knowledge of English.

4. Creation of Small Corpora

Children's literature presents many aspects of language teaching, from things such as verbs and prepositions to morality and cultural values. Using children's books as a resource helps to attract students to longer texts and to further learn the lexical items from the texts selected by the concordancer. Kowitz and Carroll (in Johns & King, 1991, p. 137) give reasons for the use of concordancing in the literature classroom. For example, concordancing (1) "makes a text more accessible by focusing attention on specific elements"; (2) encourages group work by engaging each student to investigate the task and discuss; (3) "[concentrates] attention on a small section of text with a controlled focus" and (4) arouses attention focusing on "the interaction between the student and the text, rather than on the input from the teacher." Students can use concordancing to check their intuitions about the meaning of the words without completely relying on the dictionary which may be too small to provide enough example sentences to sensitize their use of language (Love & Morrison, 1991).

There are examples of using concordance compiling software in the classroom in literature teaching. Kowitz and Carroll (in Johns & King, 1991) provides guidance for teachers to use concordancing in the literature classroom. For example, guidance (1) of basic concordancing techniques, (2) for preparing literary texts and (3) for making lesson plans. Kowitz and Carroll used The Great Gatsby and investigated words which related to eyes, white, gold and yellow to find their symbolic meanings. Another example is to look at the word blood in concordancing lines

extracted from Julius Caesar, in which blood associates with negative feeling such as fear and pain as well as positive feeling such as family and honor. In such examples, concordancing is proven to be useful for teaching symbolism in literature (the multiple uses of the colour yellow) and for teaching ideologically and symbolically loaded vocabulary terms, like blood.

4.1 Values of creating small corpora

There is no standard or prescribed size for a corpus (Partington, 1998), as long as the data is accurate and valuable. Carter and McCarthy (1995) have argued that language corpora are "useful resource for teachers and learners" (p. 144). However, according to Tribble (2000), despite the efforts that Johns, Aston, Flowerdew and himself have made, not many teachers seem to be using corpora in their classroom.

To conduct this study, I used the Harry Potter and the Philosopher's Stone text for the purpose of compiling a syllabus to teach children's literature and also used it as a starter text to encourage students to read other full novels. The linguistic points contained in this long text deserve to be brought to the learner's attention. Using a small corpus can be more beneficial as students identify lexical items within a story they are more interested in or engaged with. Smaller specialized corpora can be designed for the needs of a certain type of learner (Wible, Kuo, Chien &Wang, 2002). There are values and limitations of using small corpora (Lee, 2006). Research has been done on using the five million-word CANCODE corpus of everyday spoken English (McCarthy, 1998a; Carter & McCarthy, 2004). Contrastively, many specialized corpora used for teaching/learning purposes have been created (see Flowerdew, 1993 and Aston, 1995). For example, Gibbs (2000; in McCarthy & Carter, 2004) bases his study of the occurrence of irony in informal talk among his 62 friends (a 100,000 word-corpus). Johns, Lee and Wang (2008) use Arthur Ransome's children's literature in a Taiwanese EFL classroom and use corpus-based CALL programs to teach English children's literature. Flowerdew (1993) used a 100,000-word corpus in examining biology lectures. Unlike large corpora, a small corpus has the advantage of not overwhelming learners with examples that are too difficult to be useful. The value of using a small corpus in literature teaching lies in the demonstration of contextual repetition and content-specific linguistic points. Aston (2002) used the analogy of "making your own corpus seems rather like making your own fruit salad" (pp. 11-12) to emphasize that in making self-built corpus, one: (1) is in control of the ingredients (materials) and can easily investigate the 'particular linguistic and conceptual characteristics' of the materials one has put in; (2) can interpret concordances or numerical data easily if you know exactly what texts a corpus consists of, since this allows a great degree of top-down processing; (3) can turn the fruit salad into a delicacy (turn a corpus into a useful resource); (4) can become a better, more critical corpus user -- 'increasing awareness of how design affects the results' rather than just reading other research papers; and (5) can 'open up a whole range of opportunities for learners to discuss how best to compile and encode corpora for particular purposes'.

4.2 Computing tools used for analysis

As mentioned earlier, two computing tools were used for data analysis in this article. One was MicroConcord (Scott and Johns, 1993) and the other is CONTEXTS program. There are other software tools that can similarly find and situate language terms in context. Using contemporary e-readers and document browsing software like Adobe Acrobat can also be used in the classroom to browse for specific language terms and for students to investigate differing uses of language.

5. Using the CONTEXTS Program

There are several excellent computer-mediated teaching programs, such as John Higgins' open source software. However, without a teacher to apply them and create his or her own material, the programs, at best, will be used for general practice instead of inspiring student to become language detectives. Educators should integrate their own materials using appropriate CALL programs and organise sequencing syllabuses to make the best use of these programs. This was effectively demonstrated by Johns (1991). The programs outlined in his research illustrate the importance of teaching language through 'meanings that can be realised by form' (Johns, 1991, p. 27).

5.1 Advantages of the CONTEXTS program

The use of the CONTEXTS program especially suits the learning of phrasal verbs and prepositions. Johns (1997) proposed a "one-item-multiple-contexts" exercise for the application of DDL in the teaching of prepositions.

Working out words from the context to decide which preposition goes with which verb and observe the collocations which accompany the phrasal verbs appears to be a revolutionary way to learn pedagogic grammar.

The CONTEXTS software can sort these concordance lines according to different key words, thereby encouraging the observation of patterns. During the typical CONTEXTS program, students are asked first to look at the concordance lines themselves and decide which particle goes with which verb and what it means in the specific context. After checking the students' analysis, the teacher may present another analysis. At this stage, students compare answers and gain knowledge of how to use the targeted language features. After having done the exercise, students are able to organise the rules which apply to these lexical items.

In the next stage, the teacher uses the keyword-in-context concordance, in which the keywords are arranged one below the other down the centre of the page, with a fixed number of words to provide context to the left and to the right (Johns, 1991). By focusing on the discourse, students learn English through the various contexts the language occurs in. This begins to solve the long-neglected problem of traditional grammar translation approaches in which students were required to learn the target language only by practising the grammar exercises and through rote memorization. The students can be either working with the concordancer in groups or individually interacting with the texts and developing appropriate learning strategies. As Love and Morrison report, "[this] methodology led the students to encounter more independent learning strategies than they often employ" (1991, p. 131). It is therefore plausible that each student contributes jointly to the set task in the successful acquisition of the targeted language features instead of purely by rote memory.

5.2 Disadvantages of the CONTEXTS program

One of the main arguments against the KWIC format is that the concordance lines are fragmented, which can be confusing and may not provide the appropriate length or span of context (Sripicharn, 2002). The fragmented utterances require a new way of reading; instead of reading from left to right, the learner needs to focus on the keywords both centre-out and vertically. According to Gavioli (1997), the process of reading this way, moving from the concordance to the text(s) and vice-versa on the basis of a hypothesis, is similar to the process of reading hypertext "navigation" (Nielsen, 1990) which is not linear but based on free association. Interpreting lists of concordance involves reading at various levels to confirm or contrast hypotheses (Gavioli, 1997). The lack of teachers' instructions may impair students' ability to read the concordance lines. Further, there is a question of language habits. The KWIC format, while situating the language terms in context, often removes the sentence from its paragraph context, leaving sentence fragments. This naturalizes using broken sentences and can have an adverse effect on students justifying their own use of sentence fragments.

5.3 Methods ofpreparing a data file using the authoring facility of the program

The procedure of creating a data file based on the story of Harry Potter for the CONTEXTS program was made possible by means of exercising the guidance from a Contread.txt file, so that students could look into certain lexical items from the Harry Potter story in a DDL environment. The contexts have been taken from the Harry Potter database with the help of the computer program Microconcord. The steps are as follows:

• Making concordance lines

The first step is to make concordance lines of lexical items selected from the text because CONTEXTS uses KWIC concordance lines. This was accomplished by scanning and turning the text into electronic text. Microconcord was then used to create concordance lines. Prepositions were selected, namely, about, around, away, down, from, into, off, on, out, over, through and up and concordance lines were created of these lexical items for the CONTEXTS program. The CONTEXTS program requires 10 KWIC citations for each item. To make this set of concordances representative, a hundred concordance lines were created for each item using Microconcord, and then 10 out of every one hundred lines were carefully chosen to try and find examples which were of particular use to students.

• Data file for the CONTEXTS program

After several failures, a data file for use in the CONTEXTS was successfully created. Following Johns' instruction (1997), the following regulations were met while creating the data file using the concordance lines: (1) the first line is the title of the text 'Prepositions in Harry Potter', which appears in the index in the CONTEXTS program (see the screenshot of the CONTEXTS); (2) the number in the second line is a 'marker number' which the current version of the program does not use, so any random number will do (e.g. 231); (3) after the 'marker number'

are 10 KWIC citations; (4) the key item starts in column 37 of the citation; (5) the concordance line is confined to 80 characters in length for every citation; (6) a bank of three 'Research tasks' following the @ character were created for each item, which may be presented to the user on request; (7) the * character indicates the end of one item; (8) there are 12 items (12 prepositions) in total in the file. Due to the length constraint, they are not listed here. A triple asterisk *** is used at the end of the last item. The following is the presentation of the data file for preposition 'about':

Prepositions in Harry Potter

this boy! -- knows nothin' abou' -n't handle.' When they told him purpose. While Ron told Hagrid all was of course nothing at all little as he could, "Their son -- he'd be in tonight. He had almost forgotten aving too good a time to think much ropped the teapot. "How do you know Gringotts,' said Ron. "Did you hear

ike an angry goose. "Don't you care @

What are the verbs concerned with thought or feeling that are followed by 'about'? Do any other verbs occur beside 'about'? Do any other noun groups follow after 'about'?

The following is a screenshot of the CONTEXTS program:

Top Menu

Prepositions in Harry Potter

Use - and ENTER to Select Guide Index Esc to Quit

about ANYTHING?' Harry thought this was

about Charlie's letter, his eyes filled wit about Charlie's work with dragons, Harry pi about Crabbe and Goyle, but as the High Tab about Dudley's age now, wouldn't he?' "

about Flamel. It didn't seem very important about Flamel. They had the dormitory to the about Fluffy?' he said. "Fluffy?' Yeah

about Gringotts? It's been all over the Dai about Gryffindor, do you only care about yo

Prepositions in Harry Potter

t its leg so he could put the money into mare, honestly. Someone knocked into

not, but he seemed to keep running into ew The feeling suddenly surged back into t back, the post had arrived, right into match and started trying to turn it into her job. Scooping the fire off him into id led them through the bar and out into ked past Quirrell's turban straight into ry's lap. Up another escalator, out into

a small leather pouch tied to it. The Harry as they hurried past him. It wa Snape wherever he went. At times, he Harry's legs. He stumbled backwards. Uncle Vernon's lap. Harry could see t a needle. By the end of the lesson, o a little jar in her pocket she scramb a small, walled courtyard, where ther Harry's eyes -- and a sharp, hot pain Paddington station; Harry only realis

about down on through

around into out up

away off over from

Mark Quiz Top Menu

6. Analysing the Data

Mindt and Weber (1989, cited in Kennedy, 1998) state that "about one word in every eight in almost any English text is a preposition" (p. 139). Prepositions make an important contribution to intra-propositional cohesion by linking lexicalized concepts in terms of place, duration, association, agency, time, cause and other relationships.

The Harry Potter corpus presents a picture of the collocational nature of certain prepositions: about, around, away, down, from, into, off, on and out, in the context of children's literature. Most of the left collocates of these prepositions are verbs. This is because the verbs in children's books tend to be more dynamic and vivid and the

prepositions emphasise these verbs. Kennedy (1998) states that a corpus can be used to show the "distribution of the various senses or semantic functions of the prepositions."

To illustrate how to draw students' attention by describing features of prepositions in CONTEXTS, the preposition on was analysed below.

ooks and equipment. Term begins 1 September We await your owl by no la

inue into the break-in at Gringotts 31 July, widely believed to be the wor

school.' 'All right, keep your hair .' A whistle sounded. 'Hurry up!' t

mething, but he didn't, so she went : "A fine thing it would be if, on the

e passageway which was the only way . All they could hear apart from th

for Filch, I bet she heard us. Come .' And Ron pulled Harry out of the

y noise, like a mouse being trodden . "Anyway -- Harry,' said the giant

ary, where Hermione was testing Ron Astronomy. Harry told them what he'd h

e high enough to see what was going . At either end of the pitch were three

10. or corridor,' Harry told Ron. "Come .' But that part of the plan didn't

The missing preposition here (which we know is on) can be difficult for the students to guess. For example, in line 2, the students are asked to find a preposition preceding a time period. There are three possible choices: in, on and at. This can be confusing for Taiwanese students. Similarly, idiomatic expression as found in the line 3, "keep your hair on" is very difficult to understand for Taiwanese English learners. Students have trouble imagining that there is a missing preposition. They are likely to fill in the gap with adjectives such as 'long' or 'short' since they think it is only possible to ask someone to keep his/her hair long or short. Line 4, "... so she went on:" is easier for Taiwanese students to guess because they have likely learned the common meaning of 'go on'. With the fifth line ".was the only way on," students may have difficulty guessing the correct item as the intended meaning is 'the only way which led anywhere' which is much more ambiguous for the students. The sixth line "come on" means 'come with me, let's keep going, don't delay'. The eighth line is also a bit confusing - to test somebody 'on' something, in this case, "Astronomy"; here something normally refers to subjects such as math and English but students might commonly write 'with' instead. These are only a few examples of the often ambiguous nature of prepositions and how concordancing can help students attend to these ambiguities in a more engaged context.

When particles used in phrasal verbs occur as prepositions, they tend to have a locative use (e.g. He turned up in the village where you were all living, during Hallowe'en ten years ago.) But if we look at the way in which 'up' is used in Harry Potter, it seems to occur mainly with collocated verbs. Kennedy (1998) notes that corpus-based studies of prepositions reveals that many of them occur frequently "in recurring collocations, and that systematic analysis of the phrases which prepositions form a part reveal a variety of different semantic roles" (p. 143).

6.1 One suggestion for the CONTEXTS program

As mentioned above, without a teacher to apply the program and create his or her own material, the program, at best, will be used only for general practice. The main problem is that there is one major limitation of this program. Because the CONTEXTS program needs only 10 KWIC citations for each item, it is not sufficient for students to discover the different patterns from the limited citations. For example, if I were to teach the pattern 'Verb + prepositional object' (Hunston & Francis, 1996), I should categorise the verbs in terms of their meaning groups because verbs belonging to the same meaning group will have the same analysis. To solve this constraint, I might need 30 instead of 10 KWIC citations from every 200 concordance lines for each item to allow my students to identify the patterns. If this were the case, there might be a need to expand the CONTEXTS program in terms of its capacity.

6.2 Further directions

There is one significant problem to tackle before Taiwanese teachers try out this approach. Is it too difficult to compile our own corpus? Or, as Love and Morrison (1991) suggest, do we need to have access to various ranges of subject texts? Again, I remain optimistic that more and more teachers and material writers will make the effort in the future. A possible solution for Technology Assisted Language Learning and concordancing is to use e-readers and

electronic text as searchable text. PDFs also have search functions and with expanded browser tools and more flexible e-texts, this form of language teaching will prove useful in future paperless classrooms.

The use of the CONTEXTS and other concordance compiling programs as language exercises is different from the traditional method of teaching grammar and literature. If the questions on the main examination papers for Taiwanese students could be derived from concordance lines, it would be a method worth trying in our teaching.

6.3 Problems for teachers

Teachers should select and compile the proper materials to the best of their ability. It is the teacher's responsibility to select suitable texts for students at different levels and with different interests. One of the latter uses of the Harry Potter corpus I have compiled is as a means of teaching the cultural aspect of children's literature. After all, language reflects culture, even though this paper has focused only on the use of language. How to integrate cultural aspects into our language classroom still awaits further study, particularly with respect to the culturally related problems it might involve.

Most language teachers who were brought up in the tradition of learning language through memorising grammar rules have difficulty in coming to terms with the use of computing tools in the classroom (Johns, 1991). However, teaching grammar can be so confusing that teachers themselves often need guidance. Teachers in Taiwan are often not inclined to show their lack of knowledge. By working through the concordance lines, students might be able to raise some questions derived from the problem they are facing and teachers would then know how to help them. Using a DDL approach offers learners direct access to the data, and allows teachers to say "I'm not sure: let's find out together" (Johns, 1991, p. 31).

6.4 Problems for students

A major problem for students who learn English in Taiwan is that they tend to learn the language as discrete linguistic units instead of in phrases or clauses. Most conventional course books are designed to teach every individual word by giving its meaning. Without building up language routines, learners lose the chance of pedagogical benefit.

Students in Taiwan feel safe when they do grammar exercises. It is important for them to pass the exams and, unfortunately, the exam questions are independent of and show no connection to each other. My view is that since students feel that they cannot learn English without traditional grammar practice, offering them a particular "form-focused instruction" (Johns & King, 1991) through classroom concordancing learning may be a plausible way forward.

Very often, the materials used in the classroom fail to provide the chance to look at learners' own mistakes which will outline where the difficulty lies in learning a target language. How do native speakers speak their language when they are still at the stage of learning it? Do they make the same errors as EFL learners do? If so, what are these errors? To tackle these questions, I would like to look into dialogue used in children's literature to infer patterns derived from the way people talk in real life. I am particularly interested in the analysis of the use of prepositions which appears to be extremely difficult for Taiwanese students to acquire since they do not have equivalent uses of prepositions in Chinese. For example, for 'keep running into somebody', in Chinese we say '^fs MA(lao3 zhuang4 jian4 ren2)'. Students have to guess only from intuition which particle goes with which verb.

7. Conclusion

On the basis of a pilot analysis of the verbs and particles in the Harry Potter corpus, I would argue that DDL is helpful for students both in preparing for their exams and also for their general English acquisition because they can learn English in context and are interested in doing so. At the same time, an analysis of the use of these selected items reveals their grammatical behaviour which conforms again to what is described in descriptive grammars. Although there are still some familiar patterns which do not occur in this corpus, it allows us to take a further step towards analysing more children's books in the hope of digging out more interesting features for language learning.

References

Aijmer, K., & Altenberg, B. (eds). (1991). English corpus linguistics: Studies in honour of Jan Svartvik. London: Longman.

Aston, G. (1995). Corpora in language pedagogy: Matching theory and practice. In G. Cook & B. Seidlhofer (eds.), Principle and practice in applied linguistics: Studies in honour of H. G. Widdowson (pp. 257-270). Oxford: Oxford University Press.

Biber, D., Conrad, S., & Reppen, R. (1998). Corpus linguistics: Investigating language structure and use.

Cambridge: Cambridge University Press. Boxer, D.,& Pickering, L. (1995). Problems in the presentation of speech acts in ELT material: The case of

complaints. ELT Journal, 49(1), 44-58. Carter, R., & McCarthy, M. J. (1995). Grammar and the spoken language. Applied Linguistics, 16, 141-58. Carroll, L. (1865). Alice's adventures in Wonderland. London: Penguin.

Cheng, W., & Warren, M. (2007). Checking understandings: Comparing textbooks and a corpus of spoken English

in Hong Kong. Language Awareness, 16(3), 190-207. Conrad, S. (2005). Corpus linguistics and L2 teaching. In E. Hinkel (Ed.), Handbook of research in second language

teaching and learning (pp. 393-410). Mahwah, NJ: Lawrence Erlbaum. Cook, G. (1994). Discourse and literature: The Interplay of form and mind. Oxford: Oxford University Press. Fox, G. (1991). Context dependency and transitivity in English. In T. Johns and P. King. (Eds.), Classroom Concordancing, ELR Journal, 4. Flowerdew, J. (1993). Concordancing as a tool in course design. System, 21(2), 213-229.

Flowerdew, L. (2006). Texts, tools, and contexts in corpus applications for writing. Paper presented at 40th Annual

TESOL Convention, Tampa, Florida. Gavioli, L. (1997). Exploring texts through the concordancer: Guiding the learner. In Wichmann, S. Fligelstone, A.

McEnery, & G. Knowles (Eds.) Teaching and Language Corpora, pp. 83-99. London: Longman Gavioli, L. (2001). The learner as researcher: Introducing corpus concordancing in the classroom. In G. Aston (Ed.),

Learning with Corpora, (pp. 109-137). Houston: Athelstan. Gibbs, R. (2000). Irony in talk among friends. Metaphor and Symbol, 15(1&2), 5-27.

Granger, S. (1998). Prefabricated patterns in advanced EFL writing: Collocations and formulae. In A.P.Cowie (Ed.)

Phraseology: Theory, Analysis and Applications. Oxford: Oxford University Press. Holmes, J. (1988) Doubt and certainty in ESL textbooks. Applied Linguistics, 9 (1), 21-44.

Hunston, S., & Francis, G. (1996). Pattern grammar: A corpus-driven approach to the lexical grammar of English.

Amsterdam: J. Benjamins. Johns, T. F. (1986). Microconcord: A language-learner's research tool. System, 14(2), 151-162. Johns, T. (1988). Whence and whither classroom concordancing? In T. Bongaerts, et al. (Eds.) Computer

Applications in Language Learning. Dordrecht: Foris. Johns T. F. (1991). Should you be persuaded-Two samples of data-driven learning materials. ELR Journal, (4), 1-16. Johns, T. F., & King, P. (eds.). (1991). Classroom concordancing. English Language Research Journal, 4, 71-103. Johns, T. F. (1997). Contexts: The background, development and trialling of a concordance-based CALL program. In Wichmann, A., Fligelstone, S., McEnery, T., & Knowles, G. (Eds.) Teaching and Language Corpora. London: Longman.

Johns, T. F., Lee, H., & Wang, L. (2008). Integrating corpus-based CALL programs in teaching English through

children's literature. Computer-Assisted Language Learning, 21(5), 483-507. Kaltenbock, G., & Mehlmauer-larcher, B. (2005). Computer corpora and the language classroom: On the potential

and limitations of computer corpora in language teaching, ReCALL, 17(1), 65-84. Kennedy, G. (1991). Between and through: The company they keep and the functions they serve. In Aijmer and

Altenberg (Eds.) English Corpus Linguistic. London: Longman. Kennedy, G. (1998). An introduction to corpus linguistics. London: Longman. Kettemann, B. (1995). On the use of concordancing in ELT. TELL&CALL, 4, 4-15.

Kowitz, J., & Carroll, D. (1991). Using computer concordances for literary analysis in the classroom. ELR Journal, 4, 135-149.

Lee, H. C. (2006). Using Children's Literature in an EFL classroom in Taiwan: An application of concordancing. Unpublished Ph.D. Thesis. The University of Birmingham.

Love, A., & Morrison A. (1991). Bringing concord to the ESP classroom. English Language Research Journal, 4, 115-133.

McCarthy, M. J. (1998a). Spoken language and applied linguistics. Cambridge: Cambridge University Press.

McCarthy, M. J., & Carter, R. (2004). Introduction to special issue on corpora, Journal of Pragmatics, 36.

Mindt, D., & Cristel, W. (1989). Prepositions in American and British English". World Englishes, 8, 229-238.

Nielsen, J. (1990). Hypertext and hypermedia. London: Academic Press.

O'Keeffe, A., McCarthy, M. J. & Carter, R. A. (2007). From corpus to classroom: Language use and language teaching. Cambridge: Cambridge University Press.

Partington, A. (1998). Patterns and meanings: Using corpora for English language research and teaching. Amsterdam and Philadelphia: John Benjamins Publishing Company.

Romer, U. (2005). Progressives, patterns, pedagogy. A corpus-driven approach to English progressive forms, functions, contexts and didactics. Amsterdam: John Benjamins.

Rowlings J. K. (1997). Harry Potter and the Philosopher's Stone. London: Bloomsbury.

Rutherford, W. E. (1987). Second language grammar: Learning and teaching. London: Longman.

Scott, M., & Johns, T. F. (1993). MicroConcord . Oxford: Oxford University Press.

Sripicharn P. (2002). Evaluating Data-driven Learning: the use of classroom concordancing by Thai learners of English. Unpublished Ph.D. Thesis. The University of Birmingham.

Seidlhofer, B. (1999). Double standards: Teacher education in the Expanding Circle. World Englishes, 18, 233-245.

Seidlhofer, B. (2003). Controversies in applied linguistics. Oxford: Oxford University Press.

Stevens, V. (1991). Concordance-based vocabulary exercises: A viable alternative to gap-fillers. English Language Research Journal, (4), 47-61. Sripicharn, P. (2002). Evaluating data-driven learning: The use of classroom concordancing by Thai learners of English. Unpublished Ph.D. Thesis. The University of Birmingham.

Tribble, C., & Jones, G. (1990). Concordances in the classroom: A resource book for teachers. London: Longman

Tribble, C. (2000). Practical uses for language corpora in ELT. In P. Brett & G. Motteram (Eds.), A Special Interest in Computers: Learning and teaching with information and communications technology (ICT) (pp. 31-42). Eynsham, UK: Information Press.

Wible, D., Kuo, C-H., Chien, F-y., & Wang, C.C. (2000). Toward automating a personalized concordancer for data-driven learning: A lexical difficulty filter for language learners. In Kettemann B. & Marko G. (eds.) (2002) Teaching and Learning by Doing Corpus Analysis. Proceedings of the Fourth International Conference on Teaching and Language Corpora, Graz, 19-24.

Widdowson, H. G. (2000). On the limitations of linguistics applied. Applied Linguistics, 21, 3-25.

Willis, D. (1990). The lexical syllabus: A new approach to language teaching. London: Collins COBUILD.