Keywords

Introduction

Children who are learning to read in an L2, by default, possess knowledge of the first language, literacy, reading capability and some strategies. In other words, L2 reading has access to various other resources including the L1. As Upton and Lee-Thompson rightly state, “Reading in a second language (L2) is not a monolingual event; L2 readers have access to their first language (L1) as they read, and many use it as a strategy to help comprehend an L2 text” (2001: 469). Such children’s attempts at learning to read in a second language need to be perceived as providing opportunities to help them build on existing skills and also enable them to transfer skills from one language to another. Furthermore, for second-language learners, literacy knowledge from the first language must be perceived as something that can assist them to acquire high levels of reading comprehension in the second language (Bialystok, 2001). Reading capability in the first language has to be valued as a rich repertoire of resources that can be used to enable reading comprehension in a second language. If these resources are not exploited, it is like asking the children to keep their languages in two parts of the brain, thus creating ‘semi-linguals’. In a grassroots multilingual country like India, where individuals switch between languages with ease as and when required, this would be like placing the two languages in two separate compartments. This, unfortunately, is the reality in most language classrooms in the country.

Teachers in English-language classrooms in India do not seem to be making use of their children’s capability in L1. The ESL reading class, if it can be called that, is confined to some common, routinized classroom practices such as rote recitation of the alphabet, choral sounding out of a story word by word, repeating a sound and word a dozen times, copying the letters of the alphabet correctly over and over again, breaking down words into letters and sentences into words, ‘explaining’ a reading text followed by the dictation of answers to comprehension questions given in the text. Hence this attempt to use or rather build on the foundation of children’s L1 (Assamese) reading strategies to enable reading comprehension in an L2 (English). The argument made in this chapter is that with assistance and training, teachers can be encouraged and enabled to use children’s existing L1 reading strategies to develop L2 reading capabilities. It is hoped that eventually children will become confident, successful and proficient L2 readers.

Research on Reading Strategies Instruction

In L1 context, current research on strategy instruction is confined to the use of and training in various strategies such as summarizing, predicting, imaging, monitoring, using prior knowledge and forming questions for comprehension. However, there is a relative dearth of research focusing on L2 reading strategies instruction (Grabe, 2009: 239). There are some studies, however, that have been done; these span students from three continents, ranging in age from 15 to 40. The proficiency levels range from high to intermediate. The nature of the research has been both descriptive and interventional with some quantitative studies. Most of the studies, however, have used think-aloud protocols to capture strategy use in instructional contexts.

Carrell, Pharis, and Liberto (1989) reported a study on metacognitive strategy training for ESL reading. The research questions were: (a) Does metacognitive strategy training enhance L2 reading? (b) If so, does one type of strategy training facilitate L2 reading better than another? and (c) How is the effectiveness of metacognitive strategy training related to the learning styles of the students? The participants were 26 ESL students, with a wide range of first languages, in Level 4 of the intensive programme of the Center for English as a Second Language (CESL) at Southern Illinois University, USA. After a four-day training period, results showed that metacognitive strategy training (in semantic mapping and the experience-text relationship) was effective in developing second-language reading, and that the effectiveness of one type of training versus another may depend upon the way reading is measured. Also, the effectiveness of the training was related to differences in the learning styles of the students.

Kern (1989) also worked with university students with low to high levels of proficiency. He focused on two cognate languages, English and French, and investigated the effect of direct instruction of word, sentence and discourse analysis strategies on reading comprehension. He divided his subjects (53) into experimental and control groups. Pre- and post-tests for comprehension revealed that only the lower-ability students achieved a statistically significant gain in comprehension.

Raymond (1993) again worked with the English–French combination of languages and investigated the effects of structure strategy training on the comprehension of expository prose. The participants were 43 native English-speaking students learning French as a second language. He compared two groups of participants: a group that was taught five top-level structure strategies and a group that received no training. The selected strategies were: (1) description; (2) collection; (3) causation; (4) problem solution; and (5) comparison. The reason behind this selection was the frequency of their occurrence in the reading of prose. The participants were asked to read a text, complete a questionnaire and then do a written recall (in English). The study showed that after the treatment, the experimental group outperformed the control group by recalling more idea units from one text. This implies that training in structure strategy helped increase the amount of idea units recalled. It also showed that structure strategy use is a characteristic of skilled second-language readers. However, Raymond reiterates the need for more research that will explore the interaction of many different variables such as text content, reader interest, text difficulty and background knowledge with L2 reading comprehension.

Kitajima (1997) used a combination of word-level and metacognitive strategies for his strategy-based instruction programme with two non-cognate language. It was located in the USA and examined whether or not strategy training that orients American college students’ attention toward referential processes would help them comprehend a Japanese narrative. At the end of fifteen weeks of intervention, the experimental group of students comprehended the story at the macro level significantly better than the control group.

Bimmel, van den Bergh, and Oostdam (2001) reported on the effects of a strategies intervention study. Strategies for reading comprehension were taught both in L1 (Dutch) and English (L2), in the hope that strategies would be successfully transferred from L1 to L2. The four reading strategies were: looking for key fragments, paying attention to structure marking elements (hinge words), making up questions (questioning) and mapping the most important information from a text (semantic mapping). Students had to work in pairs on the execution of reading tasks following a consciousness-raising method, consisting of an orientation phase in which they explored the reading tasks and selected a reading strategy, a practice/application phase in which actual performance took place and a verbalization phase, in which the steps that lead to a correct solution had to be formulated explicitly. On the basis of this programme, l5-year-old Dutch students from the third year of regular secondary education (N = 12) were trained in applying these reading strategies while reading in their first language (Dutch). The results were compared with those of a control group (N = 119). The results showed that the training programme had an effect on the mastery of the four strategic reading activities and led to a substantial improvement of the reading comprehension in the first language (Dutch). The transfer effects for L2 reading, however, could not be determined.

Dreyer and Nel (2003) outlined the format and structure of a strategic reading instruction component of an English for Professional Purposes course offered within a technology-enhanced environment. Their study addressed the following research questions: (a) What does the reading comprehension and reading strategy use profile of first-year students at Potchefstroom University, South Africa look like? (b) Did the students in the experimental group attain statistically and practically significantly higher mean scores on their end-of-semester English, Communication and TOEFL reading comprehension tests than the control group, and did they differ significantly in terms of their reading strategy use? After participating in a thirteen-week strategic reading instruction module offered in the university, it was found that the experimental group differed statistically and practically significantly from the control group on all the reading comprehension measures. This was true for both successful and at-risk students.

The above studies were carried out in non-Indian contexts: America, Canada, the Netherlands and South Africa. There is very little research on this area from India. Three such studies rooted in the Indian context are presented below.

Indian Research Studies on Strategy Instruction

The first study, by Jose (2000), aims to make learners more independent in reading by generating awareness of reading strategies. She argues for a learner-chosen strategy training programme. Twenty adult ESL learners (graduates and postgraduates) who joined an English proficiency course in the Central Institute of English and Foreign Languages, Hyderabad participated in this study. Results showed that an awareness of and exposure to alternative strategies such as active reading strategies (chiefly strategies to get the meaning of unfamiliar vocabulary), comprehension monitoring strategies and metacognitive strategies (summarizing and self-evaluation) led to enhanced reading comprehension, manifested in improved task performance. As a result of learners’ increased strategy repertoire, the readers also gained confidence and attempted independent reading.

Kaw (2006) attempted to find the role played by questioning in the teaching of reading comprehension in ESL classrooms. Her study provided an empirical justification for the use of questions in teaching second-language reading. Question-rich teaching sessions were conducted for a stipulated period of time to find out whether persistent questioning helps in making learners autonomous. The teaching sessions were based on question–answer exchanges with questions that were cognitively varied. In the feedback sessions, instead of giving away the correct answers to questions, the teacher posed further questions with the aim of simplifying the questions, helping the learners to understand the focus of the question and directing their attention to the textual clues that hold the correct answer.

The subjects for this study were 29 Class VIII students studying English as a second language in a Kendriya Vidyalaya, Hyderabad, Telangana. The findings of the study showed that the children copied answers (from the textbook) much less in the post-intervention test. It also showed that allowing learners to discover answers for themselves was beneficial as it helped them to grow autonomous and self-reliant in getting the meaning out of a text which meant that their cognitive problems in processing textual information had decreased.

Rajasekhar (2006) chose to work with three specific strategies—imagination, elaboration, prediction and confirmation (IEPC), semantic impression and mind modelling—to teach a group of Class 7, 8 and 9 students studying in a Navodaya Vidyalaya, Peddapuram, East Godavari district, Andhra Pradesh. As a result of the teaching, his students participated actively in class, became more articulate and responded to questions more confidently. The study does not document a pre- and post-intervention test but does claim that their comprehension abilities had increased. This claim is based only on teacher plausibility and in-class ongoing evaluation.

A critical look at both Indian and non-Indian studies reveals that the nature of the two languages that feature in them plays an important part. Two of the six studies dealt specifically with non-cognate languages, English and Japanese, and Afrikaans/Setswana and English. One study had a range of first languages but focused on English as a second or foreign language. The others all dealt with cognate languages. Cognitive strategies may transfer across non-cognate languages with ease, but it is possible that comprehension strategies need deliberate and focused instruction, when the distance across the two languages is vast and the script is also different.

The age group covered in these studies is another difference. The range is from eleven to 40; this implies that the focus has not been on younger children, with lower levels of proficiency. The nine research studies have either used inferential statistical measures or think-aloud protocols to capture strategy use. Also, although interventional, very few have attempted to capture developmental growth in reading comprehension. Therefore, there is a need to carry out research on strategy instruction with younger learners. Hence this attempt to work with children who are between the ages of ten and eleven, who are first-generation learners coming from homes with poor reading environments.

Proposing Hypothesis

Very few studies have attempted to explore the using of reading comprehension ability in a first language (that is not similar in structure to the language under focus) as a base to enable capability in the second. Moreover, there are hardly any studies that have attempted this with young first-generation learners. As such, the research questions posited were tentative and exploratory in nature, rather than attempting confirmation of existing work. The study has attempted to answer the following research questions:

  • What are the aspects of reading capabilities of primary level Class VI bilingual children in Assam which are transferable from L1 (Assamese) to L2 (English)?

  • How can these aspects of reading be tapped to enable this transfer to happen from Assamese to English?

  • What are the reading strategies that exist in Assamese that can be tapped to enable this transfer to happen?

  • Which of these strategies will be successful and will therefore enable reading comprehension in English?

In order to determine success in reading comprehension, the following hypotheses were postulated. By the end of the intervention, the Class VI regional-medium children will be able to:

  • identify words, and infer meaning in context and sentence-level discourse;

  • use text and background knowledge, word and world knowledge to answer comprehension questions;

  • use knowledge of text structure to locate information and restate it;

  • use knowledge of text-referring words (pronominals, adjectives) and be able to indicate their referents.

It needs to be remembered, however, that this intervention is not one that is governed by strategy-based instruction (SBI) (Rubin, Chamot, Harris, & Anderson, 2007: 141). This widely used SBI model is based on the demonstration and explanation of strategies and their conscious application by learners in new settings. Although there is some instruction of ‘strategy’, it is always done in an indirect or rather inconspicuous manner, through the use of texts and tasks. Texts which were age, proficiency-level, and background and cultural knowledge appropriate were selected. Suitable tasks that were in the zone of proximal development (Vygotsky, 1978) were created. Through the ‘doing’ of these tasks it was assumed that many strategies would be tapped and transferred. In other words, this can be a variant of SBI model, which seeks to activate certain strategies for the reading of L1 texts, and then facilitate/encourage their application in dealing with L2 texts with scaffolding provided by parallel tasks rather than direct instructions.

Design of the Study

Subjects

The intervention study was two weeks long and had eight Class VI Assamese-English children as subjects. They were aged ten to eleven and had a minimum of three years of exposure to English. For purposes of confidentiality and ethics the children have all been given pseudonyms such as Karabi and Kaberi (twins), Barsha, Anjali, Manisha (5 girls), and Manash, Rahul and Rohan (3 boys). Nearly fifteen lessons had to be held over 20 days to ‘teach’ the tasks. The timings were outside school hours (7–8 a.m. and 5–6 p.m.) as the schools did not allow such teaching to happen.

Locality

Participating children were from a remote village named Garemari, Barpeta district, Assam, one of the north-eastern states of India. Barpeta district is located in the western part of Assam, nearly 150 km from the capital city, Guwahati. Garemari is situated in the eastern part of the Barpeta district, sharing borders with Nalbari and Baksha districts. The Tihu-Doomni road (also popularly known as Saahebor Aali), one of the oldest roads in the district, passes through Garemari. The Bodoland Territorial Council (BTC) has elevated the portions which pass through their area to pukka/pucca road. However, the remaining portion is still lying as kutcha road, adding woe to the remoteness of the village.

The linguistic map of Garemari is quite complex with many languages: standard varieties of Assamese, Hindi for some specific purposes, and Bodo (or Boro), a language of the Tibeto-Burmese family, whenever the need arises. However, the people of Garemari and a few other neighbouring villages use a particular dialect called ‘Bajaali dialect’, part of the Kamrupi group of dialects, in their day-to-day life, which influences school instruction as well. Officially, standard Assamese is the medium of instruction but the teacher and the students use a lot of their dialect both inside and outside the classroom.

Reading Capabilities and Strategies of Class VI Children Across Two Languages

The children in Class VI could:

  • locate print on a page very easily and had sufficient knowledge of left-to-right direction of print while reading,

  • identify the front and back side of a book and identify the starting point of reading on a page,

  • figure out the top and bottom of a page as well as a picture,

  • point out the first and last word of a particular page in both languages,

  • point out the end of the story in English after the instruction for the tasks was translated into Assamese,

  • handle juktakkhor or cluster akshara that retain most of the original symbol details,

  • read words given in the Bajaali dialect.

Furthermore, they had

  • the knowledge of print to enable them to distinguish a word from a letter or the space between words in both languages,

  • a fairly good amount of word reading ability.

However, it seemed that they did not exhibit, in an overt manner, strategies of reading comprehension. At this level, using their reading capabilities as a base, it was felt that strategies like word analysis, rephrasing a text in a simpler manner, using text structure or background knowledge to understand meaning, and being able to locate key information in the text (by underlining) are probably already being exploited by the children in Assamese. However, it is likely that these are done automatically and are therefore not being transferred to another language. The argument made in this chapter is that children’s knowledge and capabilities of using reading strategies in their L1 (Assamese) can be used to develop reading capabilities in an L2 (English).

It was hypothesized that if a ‘make them aware by giving a parallel task’ trigger was provided, the children would be able to use word analysis and other simple reading strategies like using background knowledge, locating specific information in a text, and using time and place indicators. Tasks in the language with more enabled capability, Assamese, would be used as a scaffold to spark off its use in the language with less enabled capability, English.

Procedure

Tasks to Enable Reading Capability in L2 Through L1

The intervention study was conceptualized as consisting of five tasks (Appendix 1), which focused primarily on areas such as words in context, the ability to locate key information (and then use that information to either restate it, use it where needed, or order it in a different manner), using knowledge about language(s). These tasks were graded from easy to difficult, keeping the processing demand and linguistic requirements in mind.

The tasks were conceptualized as similar across the two languages and were to be administered in that manner. As such a coding scheme has been used where the task in L1, Assamese, is mentioned as 1A, 2A etc., where ‘A’ refers to the Assamese task. This is to be followed by the task in L2, English, and is coded as 1E, 2E etc., where ‘E’ refers to the English task. It was found prudent to write and deliver all L2 task instructions also in L1 to ensure comprehension and eradicate ‘construct irrelevant variance’ (Bachman, 1990).

Task Modality

The tasks were bilingual in two ways: first in the format of the task, i.e., first the children were given the task in L1, followed by the parallel L2 task. Second, in the administration of the L2 tasks, all the instructions to these tasks (oral and written) were in the children’s L1. The deliberate attempt to use parallel tasks was to filter the declarative knowledge of the children in the best possible way to ensure that the knowledge can be put into use immediately to perform the task in L2 (Macaro, 2001). Thus, it was hoped that the children would very easily be able to perceive the strategy and that transfer would happen with ease.

Scoring Criteria

The two sets of tasks are parallel and make similar demands, except for one task (No. 5) where the amount of information required is different. The marks, however, remain the same. This alone is indicated separately (see Appendix 2).

Plausible L1 Strategies to Enable L2 Reading Comprehension

As mentioned earlier, the transfer of strategies has been rarely attempted with students who are first-generation learners. As such, the choice of strategy selection was governed by the researcher’s sense of plausibility (Prabhu, 1987) and has been grounded in an understanding of the actual contexts in which the teaching and learning part of the study would take place. As a teacher-researcher, I had to trust my own ‘grounded’ self-conceptualization of the strategies that would most effectively improve the reading comprehension of my Class VI, year three of English, Assamese-medium, first-generation learners. From the long list of strategies available, the ones selected included the exploitation of word analysis along with text- and reader-initiated strategies. Since these students are all bilingual learners who can use their knowledge about language(s) to learn another language much earlier and very easily (Mohanty, 1994), strategies that would invoke such knowledge were also added. It is observed that knowledge about own language/known language/L1 and basic cognitive skills are likely to transfer to L2 reading and are unlikely to be a source of potential interference (Grabe, 2009).

Selected Reading Strategies

From the long list of strategies deemed as teachable, the first one was word analysis, part of a strategy broadly called words in context. This strategy involves breaking bigger words into smaller meaningful chunks for an easy decode.

The second set of strategies that I felt could be exploited were text-initiated strategies (Jimenez, Garcia, & Pearson, 1996). The first of them is restating the text, which requires the student to simply re-write the information given in the text to customize the questions in hand (see tasks 2A and 2E).

The second is using text structure which could be viewed as simply using the visual representation of information in the text to comprehend it better (see tasks 3A and 3E). The third is referring to the context. Instead of just using text structure to get information, the reader is expected to determine the meaning of an unknown or hitherto unencountered word or a difficult portion of the text by referring to its L1 counterpart (provided of course that the reader has access to the parallel texts/tasks).

Although the learners are only at the primary level of education, I felt that four reader-initiated strategies (Jimenez et al. 1996) could also be exploited. The first of these is using background knowledge. It involves asking the learner/reader to consciously draw on general knowledge of the world (schema), culture-specific knowledge, formal knowledge, domain knowledge (a threshold level of knowledge about the topic under discussion) while decoding a text. In this context, it was assumed that this was ‘tapped’ in tasks 1A and 1E where their knowledge of the meaning of the terms ‘poor’ and ‘rich’ and its connotations were invoked. World/word knowledge, which is the second, is a part of the previous strategy. World/word knowledge has to be considered an essential component of reading comprehension, since every text takes for granted the readers’ familiarity with a lot of facts about the cultural and natural world and knowledge of words. Therefore, readers should be encouraged to apply such facts while reading and comprehending a text during the reading event.

Another reader-initiated strategy which is very important for beginners is the ability to underline key information. It includes locating key information in texts and showing it by underlining. In order to do this, it is important to teach learners how to locate key information in the text. Children can be taught to detect these by identifying the topic and main ideas, supporting details, by paying attention to connecting words during reading.

At early levels of reading (year 3 of English), it would be assumed that such learners, unless adults, would not possess any knowledge about language(s). While this may be true of monolingual learners, research has shown that bilinguals are able to use such knowledge much earlier (Mohanty, 1994). Accordingly, a few strategies that invoke the use of knowledge about language(s) were also selected as teachable. They are not very complex ones, but I saw them as essential and more importantly, plausible. The first is the use of text-referring words in tasks 5A and 5E. This strategy refers to asking learners/readers to pay attention to text-referring words such as pronominals and adjectives and figure out the function of these words in texts to aid comprehension. The second is teaching learners to form adjectival phrases such as ‘dark clouds’ and adverbial phrases, using the available knowledge in their first language, Assamese. The third was awareness of the most basic of punctuation marks, the full stop. I felt that this, along with an awareness of capital and small letters (a distinction that is not there in Assamese script) could be exploited.

Results and Discussion

Using the scoring criteria (Appendix 2), the responses were analysed in two ways. First the marks were computed and performances compared to find out whether the children were actually able to comprehend reading texts in English. It was observed that the scores of the students as a whole were similar. This may be attributed to their socio-economic background and the print-starved environment that they live in. However, their actual responses were then examined in detail to check whether any of the strategies ‘taught’ through the Assamese tasks were used to comprehend the English texts and answer the corresponding questions.

Capturing Reading Capability in English

The tasks had five broad areas as their focus: understanding the meaning of words in context, locating and restating information, using text structure, and using knowledge about language(s). Each of these aspects is discussed separately. All tasks were attempted by eight children.

Words in Context

These tasks required the children to activate background knowledge to understand the meaning of words. While observing the performance of the children in the L1, Assamese task and the parallel L2, English task (Appendix 3, Table 6.1), it was noticed that all the children had got the correct answers. This shows that the children were able to activate background knowledge while reading, to infer the meaning of words. This task has implications for the second-language reading class and also in designing classroom activities to facilitate the activation of background knowledge. Culturally familiar content could be incorporated to activate background knowledge. The explicit teaching of appropriate background information will facilitate second-language reading.

Locating Key Information

Information in a text can be stated both explicitly and implicitly. To identify and process text-explicit information, one needs to locate a key word or a phrase in a text, whereas to decipher text-implicit information one needs to go beyond mere location (Koda, 2005: 230–38). Such deciphering demands inferencing.

Locating and restating: This was the focus of tasks 2A and 2E. It required the student to locate the key information and state it as an answer to a comprehension question. The performance in the L1 Assamese task (Appendix 3, Table 6.2) shows that the children can locate specific information in texts. However, the L2 English task, 2E, created problems for all the children except one. This task required the reader to define a polyglot. But the children provided an example of a polyglot, which is stated explicitly in the statement. This does not mean that they cannot locate the specific information in the text. A post-task discussion with the children revealed that the wording of the question misled them. Once the task was clarified by the teacher-researcher, Kaberi and Rahul got it perfectly correct, which included the definition of a polyglot with an example.

It was interesting to note that those who performed poorly in the L1 Assamese task also performed poorly in the L2 English task. They randomly copied words from the given text as an answer to the question. Hence, Kaberi, Manisha, Rahul and Rohan scored 0 in the L2 task. One of the children (Manash) got the first item of the English task correct, but the second item was wrong. Although both the items focused on the same sub-skills of reading, the performance in both the tasks varied. This may be because in the first item the text matched the cultural background of the reader. However, in the second item, there was a mismatch between the text and the reader because of the unfamiliar text. Hence, the comprehension was higher in the first item than in the second one. In other words, the child made incorrect inferences and distortion such as “The Flamenco is Spain” (Grabe, 2009) appeared in the second item as an answer.

Locating and using appropriately: This task expected the children to identify information available in a text and write it in the appropriate box in a chart. The marks obtained by all eight students in task 3A (Appendix 3, Table 6.3) show that the children can locate specific information in a text, and use it appropriately in correct places. This suggests that they can separate and identify content words (noun, e.g., all kinship terms such as father, mother and grandfather) from function words (determiners, pronouns) in sentences. They can also represent/reproduce information, which is a combination of content and function words, as broken into two content words. Only Manash could not distinguish between the content word and the function word, and therefore, copied both the words (bhani dujani means two sisters, where bhani (sister) is a content word and dujani (two) is a function word) as an answer.

While reviewing the marks obtained by the students in the third English task, 3E, it was noticed that those children who scored low in the Assamese task also performed badly in the English task. They could not separate the content words from the function words and therefore copied the entire portion (e.g., “mother does work at home”, “mother does”) and included unnecessary words like ‘two’, ‘little’ from the text.

Locating and ordering: This reading comprehension task is based on the chronology of events in an individual’s life. Children have to read the passage and answer questions with the help of time-referring words to express chronology of events such as aagot (before), paasot (after), taar paasot (after that), xei xomoyot (during that time). The results (Appendix 3, Table 6.4) show that all the children got two of the five statements given in item 1 in the Assamese task 4A, wrong. These two statements (i) and (v) were not directly taken from the reading passage, unlike the other three statements. These two statements were based on a sentence in the reading passage. One was presented as a general statement about Bishnu Rabha’s higher education in Calcutta and the other stated the names of the institutions he attended in Calcutta. This probably misled the children, and led them to answer incorrectly. However, all the children scored full marks in the wh-question and the true-false one.

While reviewing the children’s performance in this reading comprehension task in English, 4E, it was noticed that all the children could rearrange the facts of Kalpana Chawla’s life correctly. The children later informed the researcher in one of the informal post-task discussions, that the knowledge of time indicators in the L1 task and numerals helped them to complete the task correctly. The children also performed well in the three other ‘literal questions’ in the text. The first question provided the scope to copy directly from the text, whereas the other two needed slight modifications. In the second question children had to use the noun Kalpana rather than its pronominal counterpart ‘she’. However, half of the children used ‘she’ or rather copied the third sentence of the passage “She liked flying and aerobatics most” instead of “Kalpana liked flying and aerobatics most” as an answer. In the third question, where there was less scope for copying from the text, two of the children wrote only the expansion of NASA ‘National Aeronautics and Space Administration’, and not the whole answer: “The full form of NASA is ‘National Aeronautics and Space Administration’.”

Using Knowledge About Language(S)

This task required the children to identify text-referring words such as pronominal antecedents (e.g., they) and adjectival phrases (e.g., blue sky) and show the link between these words and the head words by drawing a line in the text. Performance in the Assamese task (Appendix 3, Table 6.5) shows that the children have a fair knowledge of text-referring words in a text. They can identify and show the relationship between words by drawing lines connecting both the head word and the referring words. However, Rohan identified some unnecessary words like e-taa (one-classifier) and missed relevant words such as thopaa-bor (bunch of grapes), xei-bor (those). As far as the performance in the English task 5E is concerned (Appendix 3, Table 6.6), the children performed reasonably well. None of them could identify ‘two boys’ which refers to both Manash and Abhijit in the task. However, they identified both the pronominal antecedents ‘they’ and ‘them’. This indicated that the children can identify text-referring words such as pronominal antecedents, but have problem with adjectival phrases like ‘two boys’. In line with this Brunerian observation, these children were able to use a pencil to draw linking lines between pronominals and their antecedents. As with other task performances, here too it was noticed that those children who performed badly in the Assamese task also performed badly in the English task. They identified some extra words like ‘a’, and ‘the’ as text-referring words which actually do not refer to the nouns in the text.

The purpose of this chapter was to report whether strategy instruction would help the Class VI children to transfer their reading capability from Assamese to English. So, a detailed analysis of the strategies used by them, as revealed in their response sheets, had to be undertaken to capture strategy use.

Capturing Evidence of Strategy Use

Because of the small number of subjects, as the teacher-researcher, I was able to discreetly make a note of the strategies used by the children. Whenever an opportunity presented itself, or was found necessary, strategy use was explored during the intervention. The children were able to employ quite a few strategies while reading texts and comprehending them.

Underlining Key Information While Reading

This reading behaviour was noticed by the researcher during the administration of the Assamese tasks 1A and 2A. While reading the supporting sentences of the tasks, the children underlined the key words and information which gave them a clue and helped them to choose the most appropriate word. This is an early sign of the strategic behaviour of a good reader. It also indicates the children’s awareness of one of the important reading comprehension strategies in the beginning stages of reading: underlining the important parts of texts while reading. This particular behaviour was deliberately used by the researcher as a teaching aid while the children were engaged in reading the English tasks, 1E and 2E. This behaviour was again observed by the researcher while the children were engaged in tasks 4A and 4E. They underlined the key words of small paragraphs to infer the most appropriate words.

Word Analysis

This strategy involves breaking big words into smaller meaningful parts, also related to task 1E. Children had a problem with reading and understanding the meaning of the word ‘businessman’ (task 1E, item 1). They were encouraged to break it into smaller parts, i.e., business and man, citing examples such as xobdokox (dictionary) xobdo (word) and kox (a treasury) from their L1. Thus, a simple word analysis strategy can be used efficiently in the classroom to understand meaning and also encourage fluency in oral reading.

Refer to Context/Description in L1 While Reading Its L2 Equivalent

In task 1E, item 2, the children had problems with words like ‘thunder and lightning’, and ‘clouds’. The meanings of those words were unknown to them. Realizing this, I asked the children to look at the description of the Assamese item, since item 2 in the English task is a pragmatic translation of the item used in the L1 task. After a while, they themselves ‘identified’ the Assamese equivalents of the difficult words and were therefore able to work out the meaning. This suggests that a simple prompt such as ‘look at the Assamese item’ could help the children to understand the meaning of difficult words in context. In cases where a parallel task does not exist, the teacher could draw their attention to the equivalent word in their respective first/more enabled language.

Form an Adjectival Phrase

A language awareness task, by using children’s existing knowledge about L1, was deliberately tried out by the teacher-researcher. The task focused on the phrase ‘dark clouds’ (used in task 1E, item 2) where dark and cloud were used as an adjective and a noun respectively. The children provided the Assamese equivalent as kolaa megh. They were then encouraged to develop a similar combination in L1 focusing on the meaning of the expression, which resulted in phrases like nilaa megh (blue clouds), nilaa aakaax (blue sky), kolaa aakaax (dark sky), phorokaal aakaax (clear sky). Once these were available, with a little bit of help, the students were able to come up with ‘blue sky’, ‘dark sky’, etc. Children coming from an English-poor environment with low levels of proficiency in English would be intimidated by the idea of forming their own compound adjectives. Tapping such ability in the first language is a good second-language teaching and learning strategy.

Apply Word/World Knowledge to Aid Comprehension

The results of the Assamese task (4A) item I, where the children had to order events related to Bishnu Rabha’s education, showed that all the children got the sequence of the statements (i) and (v) wrong. This was therefore taken up for further discussion. The teacher-researcher explained in Assamese the chronology of events in Bishnu Rabha’s education, relating it to the children’s own life. This explanation included a discussion on the various stages of education—primary, secondary and higher education—which can be perceived as a part of shared world knowledge or ‘text-to-self/world connection’ (Nag & Sircar, 2008: 16). After the explanation, the children were able to arrange the events related to Bishnu Rabha’s education in the correct order, starting with the primary education in Dhaka, followed by the Matric examination, then the incident where he left for Kolkata and Koch Bihar for higher studies, according to the reading passage given. This kind of demonstration of connecting ideas presented in a text to one’s experience/life/world can be a starting point to teach children how to make connections in reading which will eventually help them to read between the lines.

Knowledge of Full Stop Use

This instance is associated with task 2A, item 1, which required children to locate specific information in a sentence and write answers for definitional wh-factual questions. In response to this, Karabi underlined the necessary information in the prompt sentence and copied it as an answer. She even used a comma at the end of the answer rather than a full stop (daari in Assamese). This suggests that she could copy information successfully, but was yet to develop the basic knowledge of punctuation. The non-use of daari was found in the responses of Kaberi and Manash as well. Interestingly, the non-use of full stops featured in the same set of children in the English task 2E. Therefore, the use of full stops as a punctuation mark (or, in Assamese daari) was taken up for discussion.

In a whole-class discussion session, as a response to the teacher-researcher’s question, some of the children told me that ‘|’ (daari) is used to mark the end of a sentence in Assamese. They also provided some examples to show its use in Assamese. This indicates that the children have the knowledge of the recognition as well as the use of daari in their L1. After that, they were informed that in English ‘.’ (full stop) is used to mark the end of an English sentence like the Assamese daari. The children understood that there is no difference between Assamese and English as far as the usage of ‘full stop’ is concerned, but that only the symbol which is used to mark it was different. As a result all the children used the ‘daari/full stop’ appropriately in a writing task.

Strategy Instruction Enabled L2 Comprehension

At the beginning of this chapter, research questions were posited regarding the aspects of reading capabilities of Class VI regional-medium children that may transfer from Assamese to English, and the manner in which this transfer would happen. It was also posited that these children are proficient readers in Assamese because they do use some reading strategies, and that some of these, with the help of parallel tasks and necessary instruction/awareness raising, will enable them to read and comprehend texts in English.

The hypotheses postulated are rephrased below as ‘can do’ descriptors with a qualitative judgement of partial or full success or failure, as the case may be. This is because although scoring criteria were created and marks allotted for every correct answer indicated/written for a reading comprehension question, a mark cannot be equated with the existence or absence of a sub-skill/strategy. Such ‘measurement’ is not tenable, for in a sense all testing is indirect and the validity of an inference is a matter of degree. As such, the ‘can do’ descriptors are based on the performance of the children and the marks given and the teacher-researcher’s ‘sense of plausibility’ (Prabhu, 1987). The task numbers that the ‘can do’ descriptors are based on are given in brackets.

The eight children of Class VI can

  • identify words and infer meaning in context (task 1E)

  • use text and background knowledge to answer comprehension questions (task 1E, 4E)

  • use knowledge of text structure to locate information and restate it (task 2E, 3E)

  • use knowledge of text-referring words (pronominals, adjectives) indicate their referents (task 5E).

From the above descriptors, it is apparent that the children were extremely successful with the tasks that required reading comprehension of texts and demanded only an indication of answers through selection or partial one-word/one-sentence responses.

Limitations of the Study

This study cannot be described as a case study, as individual performances and growth and variations between them was not its focus. At the same time, the ‘experiment’, if it could be called that, was not conducted inside the institution called ‘school’. For genuine academic reasons, it had to be conducted in the verandah of the teacher-researcher’s house and this meant that an external observer could not be brought in. Thus in a sense, the nature of the study itself, and the nature of data collected, are the two crucial limitations.

Nature of the Study

First, the study was located in a remote part of Assam, and the social and geographical location of this district (Barpeta) imposed certain limitations on it. The study was carried out with school students who study in a village which is situated in the rural area of Barpeta, a district in Lower Assam. This implies that it cannot be replicated within Assam itself without modifications, even, for instance, in the districts of Upper Assam. This is because in Lower Assam certain distinctive dialects are used as the initial mediums of instruction, while in Upper Assam standard Assamese is used. Thus the study may show different results if it is carried out in Upper Assam. On the other hand, it could be argued that the modifications are inherent to such a study and that absolute replication is not an essential condition.

Second, this kind of study is possible only in schools and contexts where both teachers and students share the same ‘vernacular language’ or dialect. In this sense, this study is narrow, but rooted or grounded in its context.

Another argument that could be made is that the period of the intervention was too short to observe any substantial growth of the children. But unless institutional support is provided, an extended (over a school term) experiment is not possible. If this were feasible then the teacher-researcher would also be the class teacher who could document their own work. Such an ‘insider intervention’ or rather ‘teaching cum observation’ would also enable a larger number of tasks for each focused sub-skill/strategy. This would possibly provide a richer picture of growth in the children.

Another problem is that the researcher was the only ‘teacher’ who tried out the intervention. If one or two more teachers could have been persuaded to also try out the experiment, the study would have benefited from it.

Nature of Data Collection

In terms of the nature of data collection and tools, information about the kind of teaching that was taking place in classes within schools was recorded only in the form of classroom observations and researcher’s notes. For practical reasons it was not possible to tape these classes. Video recordings of classroom interactions would have provided richer data. Also, the ‘documentation’ was done only by the researcher, primarily because no like-minded teacher/researcher who believed in using capability in one language to augment it in another was available. It is possible that because of this, some minute details, which may have contributed to the reliability of the study, were not taken into account.

One argument that is posited in the literature is that classroom observations do not really constitute data. In the words of Byram, “it is an open question as to whether it is at all possible to obtain data through observation that have not been influenced in some way, either by the presence of the observer, or by the data collection procedures themselves” (2004: 517). While this stance has to be accepted, one also has to remember that in the context of this study, no other documentation was possible.

The study also did not have a pre-and a post-test. Performance on the intervention tasks themselves was used as an indicator of ‘growth’ or rather, ‘small gains’ (Tharu, 1981). It was felt that in the span of twenty classes, nothing more tangible would have been possible. If institutional support had been available, a lot more could have been accomplished.

Pedagogical Implications

For Instruction

The study has serious implications for instruction, specifically for reading comprehension instruction in the ESL classroom. Strategy instruction could be a central component of reading comprehension instruction to promote strategic reading among learners (in line with Grabe, 2009: 227).

With the beginning-level learner one could think of reading instructions that focus on ‘scaffolding’ through parallel tasks (both in L1 and L2) and reading texts where children are expected to use certain strategies. With the slightly proficient learner, consistent modelling of strategy use by teachers could be useful. With the advanced learner who can articulate the transfer of strategies, consistent modelling of strategy use by students and retrospective/reflective practice of strategy use by them could promote strategic reading. This way it is hoped that strategy instruction can be embedded in everyday reading instruction rather than being taught as a separate topic/lesson in the classroom.

For Instructional Materials

To represent the societal and individual bi/multilingual nature of Indian society, one could envisage a language curriculum initiated by bilingual word cards followed by theme-based bilingual word cards (e.g., words related to family, festivals). Bilingual reading passages can then be developed around these words and themes. To create such bilingual reading passages/texts, teachers can use texts that children are reading for other subjects like science and social studies. These ‘texts’ will enable them to activate background and word/world knowledge very easily for better comprehension. Teachers who teach at the primary level in India can be of great help in the creation of such materials because one teacher handles all the subjects including the languages.

For Teachers’ Continuing Professional Development

One valuable by-product of negotiating bilingual tasks/texts with children in the classroom is the discovery of teachers’ own knowledge systems, whether in their first language or in English, resulting in a fundamental attitudinal change in teachers who then continue on the path of professional development. Thus, bilingual instructional materials and teaching procedures in language classrooms will indirectly help the teacher to reflect on the usefulness of using the first language as a resource in the English classroom.