2. English language education doctoral research in Singapore

The three doctoral dissertations reviewed in this section were selected from some 20 originating during this period from Singapore's two research universities, the National University of Singapore and the National Institute of Education in Nanyang Technological University. Indeed, the very fact that there are only two research universities in Singapore which produce scholarly work related to language teaching and, more generally, theoretical and applied linguistics, as well as sociolinguistics, would lead us to expect only a small number of Ph.D. dissertations. Nevertheless, as mentioned in the introduction, this must be interpreted with some caution. First, many academics whose work is broadly framed within language and education issues emanating from Singapore have prepared their doctoral dissertations in Western universities, especially the UK and Australia (e.g. Png Reference Png2007; Macknish Reference Macknish2009). Second, the dissertations nevertheless display a breadth of research topics and methodologies ranging from linguistic and corpus approaches examining technological (Wengao Reference Wengao2009), popular (Wee Reference Wee2009), oral (Soe Reference Soe2007) and educational discourses (Hong Reference Hong2009) to interpretive and critical approaches which set out to uncover mechanisms of power and literacy learning in the social and educational environments where English is part of the unequal multilingual (e.g. Lorente Reference Lorente2007; P. S. J. Tan Reference Tan2007; Salonga Reference Salonga2010) and socio-academic (Teng Reference Teng2007; Koh Reference Koh2010) landscape. And third, cutting across this more theoretical work are practice-oriented dissertations which attempt to address the pedagogical issues that confront teachers every day, especially those concerning writing and language proficiency problems in Singapore (Ong Reference Ong2010) and neighboring countries such as Malaysia and Brunei (Sharbawi Reference Sharbawi2009; Koh Reference Koh2010; S. K. Tan Reference Tan2010).

The doctoral research produced during the period under review can thus be viewed as reflecting the current intellectual landscape of sociolinguistic-educational scholarship in Singapore where linguistic, sociolinguistic and ideological theorizing on language, education and society is tied up with the pursuit of answers to pedagogical teaching and learning problems, not only among Singaporean learners of English, but among similar learners in the Southeast Asian region (e.g. Brunei, Malaysia and the Philippines) as well as among international students (e.g. from China) who study in Singapore as scholars of the Singapore government. Consequently, as will be seen later, some Ph.D. dissertations from Singapore with a regional scope (usually because the writers are themselves scholars from the region investigating language-related issues in specific educational and social environments) are couched in comparative terms with the aim of relating these issues to those relevant to Singapore (Lorente Reference Lorente2007; Sharbawi Reference Sharbawi2009; Tan Reference Tan2010).

Our decision to focus on writing, intelligibility and academic literacy thus reflects the range of recent intellectual pursuits in higher education in Singapore. It is also our intention to report on studies that are either exemplary in their approach or innovative in terms of their content and methodology. In short, the dissertations have been chosen based on the quality of the arguments put forward through well-designed methodology, as well as their ability (or at least potential) to articulate prevailing trends and tendencies in the theory and practice of English language education in Singapore.

Ong's (Reference Ong2010) dissertation investigates the effects of planning on the overall text quality of Chinese ESL learners’ drafts of argumentative essays. This experimental study involved 108 tertiary-level students from the People's Republic of China enrolled in a Communication Skills Programme in Singapore. Drawing on two recent competing frameworks for task complexity in task-based SLA research – Robinson's (Reference Robinson and Robinson2001, Reference Robinson2003) Cognition Hypothesis Model and Skehan's (Reference Skehan1998) Limited Attentional Capacity Model, and operationalizing planning in terms of the different dimensions of task complexity identified in these approaches – Ong's study is premised on the assumption that texts of a higher quality are produced with fewer generative activities during the writing process, which concomitantly reduces the cognitive load on the writer. Text quality in student writing is operationalized in terms of fluency, lexical complexity and Flesch reading ease (Flesch Reference Flesch1948).

In the study, task complexity is manipulated using three planning conditions: (a) availability of planning time, (b) provision of ideas and macro-structure and (c) draft availability. All participants were randomly assigned to the above three conditions using their writing task proficiency scores as a measure; these scores were obtained from a pilot study comprising three pre-writing tasks in the argumentative, descriptive and letter-writing genres, conducted before the main experiment. Participants were stratified into high, average and low proficiency levels and assigned to the three conditions above in a way that ensured equal numbers of participants from each writing proficiency level.

The availability of planning time was measured along four dimensions: (1) extended pre-task, (2) pre-task, (3) free writing and (4) control group. The amount of writing assistance given to the participants under provision of ideas and macro-structure involved three conditions: (1) topic, ideas and macro-structure given, (2) topic and ideas given and (3) topic given. The draft availability condition had two levels: (1) draft available and (2) draft unavailable.

The study was conducted in two separate phases. The first phase was a pilot study in which data were drawn from three pre-writing tests in the descriptive, argumentative and letter-writing genres, accompanied by pre-writing test questionnaires for the three types of pre-writing tests conducted, and a general questionnaire on the students’ language background. The second phase was the main experiment, and data were drawn from first and second drafts of argumentative writing essays and experimental writing test questionnaire responses. In analyzing the data, two measures of fluency were used. Fluency I measured the mean number of words produced per minute of transcription. Fluency II measured the mean number of words produced per minute of time on task. Both aspects of fluency were measured in the first drafts. Only fluency I was measured in the revised drafts since transcription time was the same as time on task in this case.

The results of the experimental study showed that (1) increasing task complexity by increasing planning time produced significantly greater fluency II and lexical complexity in the first drafts; (2) increasing task complexity through the provision of ideas and macro-structure produced significantly greater lexical complexity but had no effect on fluency I or fluency II in the first drafts; and (3) increasing task complexity through draft availability produced no significant differences in fluency or lexical complexity. In the case of the revised drafts, the amount of planning time given in the formulation stage of the writing process had no significant effects on fluency I or lexical complexity. With regard to the topic, ideas and macro-structure given scale, there were significant effects on lexical complexity but not on fluency I. Finally, although the dissertation predicted that task complexity with regard to draft availability would significantly reduce fluency I, no significant effects on fluency I or lexical complexity were observed.

Based on the results of this empirical study, Ong proposes two ESL writing models that focus on cognitive processes and sub-processes in the task environment and explores their pedagogical implications, the second model serving to further emphasize the manipulation of factors in the task environment for text revision.

Viewed from a theoretical perspective, earlier studies in writing research made no predictions with regard to the manipulation of cognitive processes such as planning, transcribing and revising, or writing quality. Ong's dissertation asks to which factors of the task learners allocate their limited attentional and memory resources, and how this resource allocation may impact the resulting text in terms of writing quality. This represents an ambitious attempt to fill the research gap, in part by using insights and procedures from task-based research. It is also the first study to take into account the role of the revision process in enhancing the quality of written text.

One main limitation of this dissertation is that the research questions that had to be generated in order to test the array of planning and sub-planning conditions as well as the cognitive processes became too numerous. The complex research design, a deterrent to replicating such a study, required the researcher to devise a multiplicity of tests, involving different types of control under a number of different conditions, resulting in the production of reams of statistical analysis, as attested by the high number of statistical tables (103 in all) in the dissertation. Painstaking as the researcher was in the meticulous computing and recording of the data in these tables, the volume of data makes it difficult for the reader to get a sense of the main orientation of the question being addressed – to get the big picture, so to speak.

Nor does it help that in analyzing the data, rather than providing interpretive claims to show how the data offer evidence to support or counter the hypotheses proposed, the results are continuously compared to those obtained by various other scholars in their research studies. Besides diffusing the focus, this ultimately makes the presentation more descriptive than analytical. Much of the literature review is in the form of a chronological account of eight prominent writing models proposed by various scholars in the field, but which lack a unifying thread. A thematically focused critical synthesis presented under well categorized headings would have been more helpful. Because the writing style is somewhat indirect, all ideas being given equal importance instead of salient points being highlighted, readers are often left to work out the links for themselves. The methodological complexity of the dissertation betrays its objectivist and empiricist bias through excessive use of statistical data, which prevents it from putting forward a clear, coherent and well-reasoned argumentation vis-à-vis the theories discussed in the study.

Kaur's (Reference Kaur2009) study offers a different approach, examining the phonological intelligibility of 22 Malay and Chinese undergraduates studying at a public university in Malaysia, based on 23 hours of audio-recorded interaction between same-sex dyads elicited during the performance of four information gap tasks. The study uses the Lingua Franca Core (LFC) proposed by Jenkins (e.g. Reference Jenkins2000), in combination with insights drawn from the Communication Accommodation Theory (Beebe & Giles Reference Beebe and Giles1984; Giles & Coupland Reference Giles, Coupland, Giles, Coupland and Coupland1991), as a broad methodological framework for analyzing spoken data, and is probably the first to apply the LFC in an Asian setting. The speech data elicited during learners’ performance of the information gap tasks were drawn from interactions between same L1 background dyads (SLD) and different L1 background dyads (DLD). Arguing that intelligibility is not monolithic, but a dynamic construct that is user-derived and directed as a result of constant negotiation between speaker and listener, the study identifies specific pronunciation features that impede intelligibility in L2–L2 interactions and investigates how participants negotiate intelligibility using particular communicative strategies.

Kaur's attention is restricted to the consonantal features, aspiration of voiceless plosives and consonant simplification in word-initial, medial and final positions, and excluded all vowels and suprasegmentals. Analysis of these features was based on seven pre-determined phonological processes that commonly characterize Malaysian pronunciation patterns, namely the addition, substitution and deletion of consonants, absence of aspiration, and simplification of word-initial, word-medial and word-final consonant clusters. In addition to re-evaluating the validity of some of the concepts and issues relating to the LFC in terms of language use and intelligibility in the Malaysian context, the dissertation also sought to re-examine the notion of intelligibility itself and to problematize its investigation in the research literature.

In her study of the phonological features of bilingual English speakers, Jenkins (Reference Jenkins1995, Reference Jenkins2000) argues that instead of narrowly defining L2 speaker phonological variation, or phonological transfer, in terms of variation from a standard native speaker model, the issue should be reconsidered in the wider context of international phonological norms whose validity rests on mutual intelligibility as a basis on which to formulate a syllabus for teaching pronunciation. In similar vein, Kaur aims to use the findings of her study to re-focus English language teaching and redefine research goals in the Malaysian context. The data for the study were analyzed using a blend of descriptive and interpretive approaches, and quantitative as well as qualitative procedures. The detail with which the research design, methodology and procedures adopted for analysis are explained (such as the changes made to the main study on the basis of the two pilot studies) means that the study is eminently replicable. As in Jenkins’ study (Reference Jenkins2000), information gap tasks were used (in this case four: a ‘same or different’ picture task, a jigsaw box task, a picture sequencing task and a map task) to elicit the spoken data, supplemented by a questionnaire and interviews to gather the demographic and language history background of the participants. A post-interaction questionnaire was administered only to those participants (18 out of 22) who participated in both SLD and DLD interactions, to elicit their perceptions.

Kaur found that intelligibility was compromised to varying degrees by all of the phonological processes identified above, but that substitution of consonant segments contributed to the largest number (two-thirds) of intelligibility problems, followed by the deletion and addition of consonants and simplification of word-final consonant clusters – the latter frequently resulting in the creation of different words that did not fit the context of the interactions, or even non-words that were difficult to process. The findings revealed that, in most cases, it was only if mispronunciations occurred in words that were central or crucial to the task that problems arose, lending support to the argument that intelligibility is not a static construct: the context, speaker and hearer, as well as the importance of the word or message to the goal of the interaction, are all important factors in determining intelligibility. Furthermore, the study confirmed that variation is permissible, because hearers and speakers constantly negotiate intelligibility; indeed, variation was rarely found to lead to a complete breakdown in communication, so adherence to inner circle norms is not essential. The participants were quite skilful in negotiating intelligibility problems and most of the miscommunications were resolved fairly quickly and successfully.

The dissertation is extremely well focused and the research reported is thorough, providing ample evidence to support the argument put forward in ELF studies that negotiating intelligibility involves more than using ‘correct’ pronunciation, but also involves the use of communicative strategies that encourage information flow in a highly recursive and accommodative manner. Kaur's findings reveal that the strategies used to ensure not only that the task was completed but also that communication progressed smoothly included the ‘let it pass’ strategy, explicit checking for comprehension by speakers, indication of non-understanding by listeners, echoing or repeating the problem word, phonological anticipation, phonological adjustments in response to lexical suggestions and correction, and the use of spelling. As reported in the ELF research literature, she also found that participants displayed a high degree of mutual cooperation and tolerance, preservation of ‘face’ and collaborative communication.

As to whether participants adjusted specific phonological features to accord with the L1 of their interlocutor, Chinese L1 speakers seem to have displayed some variation in the use of two out of the eight pronunciation features identified, viz. the use of glottal stops in place of final stop consonants and the substitution of /r/ with [l] when interacting with speakers of the same L1. For the Malay participants there were no visible patterns of variation for any of the features. In fact, the results indicate that despite their differing L1s, the participants had striking similarities with respect to their phonology. Overall, the findings of the study validate some of the core features suggested in the LFC while observing some minor differences. However, in terms of the use of phonological variation to accommodate to interlocutors of same and differing L1s, the participants in the study seemed to show different patterns from those found in previous studies.

The findings and implications discussed in this dissertation are limited to some extent by the research design of the study with respect to the nature and amount of the data analyzed, the number of participants involved and their ethnicity, the assumptions regarding their L1, and the context of the research setting. For instance, unlike Jenkins’ study – which was more comprehensive in scope – only a restricted set of elements of the LFC were examined: the vowels and suprasegmentals were entirely excluded. Likewise, Jenkins’ study involved participants from a multiplicity of L1 backgrounds, whereas the participants in Kaur's study represented only two major ethnic groups in Malaysia. Another limitation, mentioned by the researcher herself, is that in the case of the Chinese group in particular there may have been discrepancies in the actual L1 and that reported by participants in the study, as their real L1 may have been one of a number of Chinese dialects associated with different ethnic groups. Similarly, the objection frequently made about data gathered from information gap tasks not being quite the same as naturally occurring talk applies. And finally, as the researcher herself admits, although the study specifically targets phonological intelligibility as its main focus, it is not always possible in analysis to clearly separate phonology from other linguistic aspects, such as syntax and lexis.

Teng Sze Mei's dissertation (Reference Teng2007) departs from those of Ong and Kaur in the sense that it sets out to capture and examine data derived not through objectivist and seemingly ‘neutral’ methodologies, but through a critical and interpretive method of inquiry. Teng's study is an ethnographic study of English language learners who are gradually being socialized into the privileged literacies of a higher academic institution in Singapore. The study attempts to capture the dialectics of literacy learning, given the various tensions and struggles that learners experience and act out in their daily academic lives. The expectations imposed upon them by social and educational institutions sometimes collide with their individual expectations and capabilities, given their differing socioeconomic, cultural, ideological and home backgrounds. In the end, the academic world is not simply a highly competitive environment, but competition creates an uneven playing field where those whose literacies (home, school, cultural, etc.) are the privileged literacies of learning in the university already have an advantage in classrooms, including English language learning classrooms.

The dissertation is firmly located within the New Literacy Studies paradigm based largely in Western institutions (Street Reference Street1984; Gee Reference Gee1996). The theoretical impetus is the notion of literacy as a social and ideological phenomenon as opposed to a purely cognitive and individual reality. Thus the focus of the study goes beyond the acquisition of skills, usually couched in quantifiable or measurable terms, and examines how literacy learning is embedded in issues of power, ideology and identity (Baynham Reference Baynham2004). For the study's three subjects, learning new literacies in the university has affected their own views not only of themselves and the world, but also of how they engage the world. Literacy has both ideological and material effects on individual learners (Luke Reference Luke2004), privileging some and disadvantaging others, and making the process of academic socialization easier for some but more painful for others.

The main source of data for the study are the written narratives of the three student-subjects. Teng convincingly explains in detail the theoretical justification for the use of narratives to study the impact of literacies on individual lives (especially their power to alter people's identities), but in general she contends that narratives as analytical material are rich in providing information about self-identification, values and ideologies as well as the motivations behind resistance and contestation (Pavlenko & Blackledge Reference Pavlenko and Blackledge2004).

It is interesting to note that Teng anchors her study largely within the work ‘outside’ Singapore and its surrounding region. Consequently, the study misses the opportunity to link research drawn on literacy practices in Singapore and in the region with research in other contexts, and thus contribute to theoretical engagements with notions of literacy using ‘local’ landscapes of literacy situated within a unique geopolitical and cultural context (e.g. Canieso-Doronila Reference Canieso-Doronila1996; Kramer-Dahl Reference Kramer-Dahl2001). Nevertheless, Teng's study leads to an understanding of learning, specifically English language learning, as a deeply social phenomenon where one's success or failure in English is better explained by one's social investment in language learning (Norton Pierce Reference Norton Pierce1995) than simply by one's individual motivation (Gardner & Lambert Reference Gardner and Lambert1972). It is a solid antidote to much work in the field of English language teaching which focuses on the study of skills rather than how these skills are tied to issues of power, identity and ideology. Whether we like it or not, the differing economies of literacy which privilege some sets of skills over others already underpin the learning of English in the classroom, and always will. In short, Teng's study clearly shows how classroom work is not and cannot be detached from the broad context of ‘globalization’.

3. English language education research in Malaysia

Our choice of doctoral dissertations from Malaysia is, like our choice for Singapore, based mainly on the quality of the work they present and how well they represent the intellectual landscape of Malaysian applied linguistics. There are only three research universities in the Klang Valley that offer Ph.D. degrees related to ELT and applied linguistics, namely University Malaya (UM), University Putra Malaysia (UPM) and University Kebangsaan Malaysia (UKM). Initially, fifteen dissertations were identified from these three universities. The four that were eventually selected for review were drawn from the first two universities because they reflect areas of research we felt to be most relevant within the local context of English language teaching in Malaysia. As we have mentioned, considerations of topic and overall quality were crucial to their selection, including factors such as critical evaluation of the literature review, a sound theoretical and methodological framework and clear data analysis. The topics of these studies can be categorized under four headings: teaching of reading, vocabulary learning strategies in relation to written composition, communication strategies and language learning strategies – the areas identified as most representative of issues currently relevant to English language teaching and learning in Malaysia. The selection is limited to studies conducted by Malaysians and excludes the work of scholars from other neighboring Asian countries and the Middle East because research investigations often reference their home countries and do not reflect language teaching issues within Malaysia. This is not the case for Singapore, where the dissertations that reference cultural contexts in other countries are still deemed relevant because they attempt to engage the pertinent international literature in conceptual or methodological terms. Nor does it apply to the Philippines, as will be seen below, because the relevant dissertations are largely confined within Philippine multicultural realities.

Another point which has massive impact on the intellectual and pedagogical orientation of English language education in Malaysia is that doctoral research in English language education in the country, as mentioned earlier in the introduction, developed only very recently. The Malaysian government, after a lapse of over forty years, reintroduced English as a medium of instruction for mathematics and science in 2003 (though it has since reverted to a policy that will mandate the teaching of these subjects in Bahasa Malaysia from 2012). Although English, as in Singapore and the Philippines, has been around since colonial times, is the second most important language in the country, and is widely spoken in urban areas, it was replaced by Bahasa Malaysia as the medium of education in all government schools in 1960, when the country gained its independence. Consequently, in the absence of a well-established localized tradition of English-medium schooling, the teaching strategies, concepts, objectives and methodologies generally used in English classes in Malaysia are often based on theories and concepts researched elsewhere. In turn, research orientation in the field is derivative in that it is largely dependent on Western models and practices. Nevertheless, despite a tendency to replicate existing research in these dissertations, there is a slowly emerging trend in some research to draw upon the multilingual landscape of Malaysia, where English is both a second and a foreign language.

Given that the teaching of reading, particularly in English, has assumed an important role in developing students’ overall language proficiency, studies in the past emphasized the development of their reading comprehension skills. In this respect, Nair's (Reference Nair2008) dissertation differs from previous studies in that it focuses on the use of scaffolding by teacher trainees as a strategy for improving students’ understanding of literary texts. Scaffolding has not been widely used in the teaching of reading in Malaysia. The dissertation examines the types and characteristics of scaffolding used and the effectiveness of scaffolding strategy training (SST), as used by the teacher trainees, for improving L2 teaching and learning in general, and for reading comprehension in particular. The study was carried out using a quasi-experimental research design and a mixed method of collecting and analyzing the data, the analysis being both quantitative and qualitative.

In the first phase, 96 subjects were selected based on a proficiency test which then served to decide whether subjects were placed in experimental or control groups. SST was used to improve the comprehension of literary texts (short stories) among dyads from the experimental group. Subjects from the control group were exposed to the same short stories through normal lectures conducted by a lecturer from a teacher training institute. Pre- and post-tests comprising comprehension questions based on two short stories were conducted. In the second phase, subjects’ interactions (in dyads) from the experimental group were recorded twice: before and after they had received the SST treatment. The interviews and diary studies that were used to collect the data were analyzed to identify the types and characteristics of scaffolding utilized in mediating the meaning-making process during the reading comprehension exercise.

Nair's study provides interesting insights into the use of scaffolding as a strategy in teaching reading, such as the idea that high proficiency participants use analogy as a form of scaffolding to assist their partners’ understanding, while participants with low proficiency experience little impact from the use of scaffolding as a strategy. The interviews revealed that participants with a higher level of proficiency tend to prefer studying alone and have a higher level of confidence. The study identifies strategies that can be particularly useful when teaching students with different levels of proficiency in the target language.

Teo's dissertation (Reference Teo2009) examines how the use of lexis in the written compositions of ESL students in a selected university college in Nilai contributes to the quality of their writing. By implication the study seeks to demonstrate the existence of a continuum between reading and writing. The findings reveal Malaysian learners’ lack of ability to write in L2 effectively, one of the current issues in language teaching in Malaysia (Nayan & Jusoff Reference Nayan and Jusoff2009). However, the findings do not help to provide any conclusions about writing because effective writing includes not only knowledge of vocabulary but also writers’ ability to use appropriate rhetorical structures and sequence ideas logically. Nevertheless, the study is helpful in that it shows the importance of knowledge of lexical items to the ability to write.

Like Nair's study (Reference Nair2008), Teo's study uses an experimental method and involved 31 participants enrolled in a writing course in the American Transfer Degree Program in a selected private university college in Nilai, Negeri Sembilan. The study was conducted in two phases. In the first phase, two groups were asked to write a composition of between 300 and 450 words. A computer program, vocabProfile, was used to categorize the vocabulary produced in the compositions into several levels of word frequency. The second phase of the study included a Vocabulary Level Test and a Productive Level Test administered as pre- and post-tests to determine vocabulary size and depth. The experimental group received treatment for eight weeks in the use of vocabulary learning strategies by means of a number of vocabulary tasks, while the control group received regular instruction according to the prescribed syllabus. All in all, the data consisted of ESL composition scripts, LFP, vocabulary scores, journal-writing reports and questionnaire data.

The study found that the vocabulary learning strategies used frequently by the ESL learners were (from most to least frequent) cognitive, metacognitive, memory compensation and social. Most remarkably, the experimental group was found to have improved their vocabulary size significantly, from 2,000 to 10,000 words. The findings offer useful insights into the teaching of vocabulary and how it helps to improve students’ writing skills. Such findings may be useful to English language teachers who traditionally tend to isolate the teaching of reading from that of writing – a practice that is common in many educational institutions in the country. Teo's research highlights the potential impact of the connection between intensive reading and vocabulary learning on effective academic writing.

While the first two important dissertations from Malaysia draw inspiration from classroom-based problems without recourse to the multicultural contexts of reading and writing in English in Malaysia, Noor's dissertation (Reference Noor2008) reframes the analytical impetus of research in the country by taking as a given the cultural and linguistic diversity of Malaysia in which everyday interaction between its multilingual speakers occurs. The study explores the specific communication strategies employed by Malaysian ESL learners when they encounter problems in spoken interaction, particularly when their L1 cultural norms interrupt and reshape their patterns of discourse and interaction.

The dissertation explores interactional strategies in a number of speech acts which include expressing agreement, disagreement, argument and negotiation between ESL learners. It differs from a number of previous studies (Lee Reference Lee1999; Ting & Lau Reference Ting and Lau2008) on communication strategies used in the Malaysian context in its adoption of an approach in which communication strategies are analyzed into the distinct categories of language production strategies and interactional strategies.

Data were obtained from the oral interactions of 53 third-year accounting students from the Faculty of Economics and Business Administration, Universiti Kebangsaan Malaysia (UKM). Audio and video recordings, transcript analysis, observations, field notes, retrospective interviews and questionnaires were used to collect the data. Observation of meetings held at two private organizations was carried out in order to triangulate the data collected and add to the data obtained in simulated discourse. Such a mixed method of data collection helped to yield a comprehensive set of input data. A slightly modified version of the Taxonomy of Communication Strategies and the categories of Interactional Communication Strategies (Celce-Murcia, Dörnyei & Thurrell Reference Celce-Murcia, Dörnyei and Thurrell1995; Dörnyei & Scott Reference Dörnyei and Scott1997; Williams & Burden Reference Williams and Burden1997) were used to analyze and classify the data.

The findings reveal that learners tend to be achievement-oriented and that communication strategies are used as independent and supportive strategies. Interaction strategies among Malaysian ESL learners, such as self-repetition, code switching, literal translation and other-repetition, are evident. These findings correlate with those of previous studies (e.g. David Reference David, Ratwatte and Herath2007; Kow Reference Kow, David, McLellan, Rafik-Gaela and Abdullah2009), but they highlight the fact that miscommunication occurs due to wrong interpretation of an utterance, unclear questions/replies, hearing problems, inappropriate choice of words and unclear pronunciation.

Some pedagogical inferences are drawn, which, it is suggested, could improve the communicative competence of the students and prevent miscommunication. These include the use of appropriate words, correct pronunciation and the teaching of a modified taxonomy of communication strategies. Given the multilingual and multiethnic composition of the country, the importance of an appropriate choice of words when communicating cannot be overemphasized. In such an environment, sensitivity to other ethnic groups when communicating is vital. Consequently, the study could have given more emphasis to the importance of strategies in communication across cultures, from which norms of inter-ethnic communication could have emerged. Nevertheless, the dissertation presents useful findings that explain the reasons for miscommunication in Malaysia.

Ismail's study (Reference Ismail2008) focuses on learning strategies, and examines factors that affect less proficient ESL learners’ use of strategies not only for language learning per se but also in content area learning. A case study approach was used. A group of seven students from the Business Management and Accounting Faculty of the Universiti Darul Iman Malaysia (UDM) in Terengganu, Malaysia, were selected for the study. The participants were in the second semester of their studies and were selected using a purposive sampling technique. The primary data was obtained from interviews. Observation and field notes were taken during lectures, and documents such as handouts and students’ written notes were also included, to make the study more comprehensive. Peer examination and member check were used to validate the study.

Ismail's findings show that there are three factors contributing to students’ learning strategies: student, home and institutional factors. These factors are deemed helpful in motivating students to learn the target language. In each category the research reported what learners considered to be the factors that motivate them. The findings further show that the language learning strategies acquired by less proficient students can be briefly categorized under two broad strategies: self-initiated and peer-based. They reveal that the teacher was an important influence on students’ motivation to learn the target language, but this did not affect the choice of language learning strategies used by the students. Moreover, when learning the content area subjects, students struggled with the content, and used strategies such as memorization, peer discussion and referring to seniors. They sought peers and seniors as sources of reference mostly to help clarify concepts in English or even in their L1, Malay.

The study also revealed that students showed little engagement in the active production of ideas in English, mainly due to low language proficiency. It may be that students have not seen the usefulness of the language classroom, its connection to content area studies and the role played by language-learning strategies to help them understand, clarify, apply and extrapolate such knowledge and skills to other tasks. Consequently, this study developed a model entitled Integrated Content Area-Language Learning Strategies (ICALLS), an integrated approach to learning which focuses on the teacher factor, relevant and meaningful curriculum, and support amongst peers. The objective of this model is not only to promote and upgrade the teaching of language but also to enhance content area subject learning.

Ismail's dissertation is a significant contribution to the development of a model to enhance teaching and learning in Malaysia, since at present, language teaching in Malaysia depends more on theoretical knowledge imported from abroad, and such knowledge and theories are sometimes inappropriate in the Asian context.

4. English language education doctoral research in the Philippines

The 25 dissertations on English language education originally selected come from four universities in Metro Manila. These are the University of the Philippines at Diliman, the country's only national university; the Philippine Normal University, the first normal school in the country; the University of Santo Tomas, the country's oldest university, run by the Dominican Order, and the De La Salle University, run by the De La Salle Brothers. The first two universities are public institutions while the latter two are private. They are all located in Metro Manila, which has historically been the seat of tertiary and advanced education in the Philippines. The studies are qualitative, making use of methods such as interviews, questionnaires and textual analysis for pedagogical more than theoretical purposes. They draw mainly, though not exclusively, on empirical studies and theory developed in the United States, which colonized the country from 1898 to 1948 and installed an educational system largely patterned on the American system. Unlike Malaysia, therefore, the Philippines has not seen any major disruption in scholarly research in language and education resulting from changes in language policy regarding the medium of instruction, since English has consistently played that role (Tupas Reference Tupas2007).

Cluster areas are evident in the research produced between 2007 and 2010: studies on literacy and reading, technology-mediated communication, genre studies, and bilingual education, but these areas overlap, as the following review will show. They also overlap with some scholarly work from Singapore and Malaysia, especially dissertations which situate their studies within a multicultural context. Our selection has also been based on quality factors such as the broad applicability of the findings (Valdez Reference Valdez2008; Serquina Reference Serquina2010), sensitivity to culture-specific concerns (Villegas Reference Villegas2010), and relevance to pertinent issues in English language education (Bolanos Reference Bolanos2010). The contribution of a dissertation to the existing and emerging intellectual landscapes of English language education in the country was also taken into account.

Serquina's macro-study (Reference Serquina2010) investigates school factors that impact student performance in English/reading programs, the ways schools allocate resources to develop, implement and sustain reading programs, and the initiatives taken by the school and the community that influence student achievement in English/reading.

Using purposive sampling, she identifies 85 high-performing, high-poverty schools in the Philippines based on data from the National Educational and Testing Research Center, the National Census and Statistics Office and public school division superintendents. High-poverty schools are those in areas where the population forms part of the 26.5% who live below the poverty threshold in the Philippines, with resulting high rates of malnutrition and school dropout rate. The respondents to her survey were 63 school heads and 298 English teachers. From these, she chose three schools as case studies using three school performance indicators: cohort survival, retention rate and graduation rate. The respondents were the three school heads, eighteen randomly selected English teachers, fifteen parents and nine community leaders/members. Data was collected through visits, checklists, interviews, focus group discussions and observations of English classes.

Findings show that high-poverty but high-performing schools have the following attributes:

1. 1. The school head is strong both as an instructional leader and manager. Goals are shared by the teachers, parents, community leaders, education officials and heads of other schools. High expectations are established, teachers are mentored and students monitored.

2. 2. The English/reading teacher is competent, diligent, committed, knowledgeable and caring. The teacher is focused on student learning. Expectations are clear, reading methods and strategies are adapted to suit student needs and appropriate assessment tools are devised. Instructional time is maximized with enhancement and intervention sessions scheduled, even out of class.

3. 3. Strong parental and community support was provided. Resources (human, time and schedules, financial) are allocated wisely to develop, implement and sustain reading programs.

Based on these findings, Serquina proposes a model for an effective reading program in high-performing, high-poverty elementary schools, in which the core elements are the enablers (school heads, parents, community and education officials) and the implementors, among whom the teacher is the key. The missing link in the model, however, is the student. While this research is not new (see, for example, Termulo Reference Termulo2005), only Serquina's study focuses on reading programs and therefore has a direct application to teaching ESL, and ESL specific to underprivileged schools. For breadth and for its scrutiny of lived experiences, no study on reading in high-poverty schools yet matches it in the Philippines. Nevertheless, in the face of some hard facts on Philippine education (such as a perennial shortage of teachers, books, classrooms, and even of basic necessities such as electricity and running water), structural changes in the social matrix are what is required for enduring solutions to literacy problems.

Bolanos's study (Reference Bolanos2010) also deals with literacy and reading, but this time as situated in the now ubiquitous world of cyber technology, rather than the socioeconomic contexts of schoolchildren. In the Philippines, a number of dissertations have been carried out in this area (e.g. Palma Reference Palma2008; Flores Reference Flores2010), but it is Bolanos's study that has direct relevance to English language education. The dissertation takes as its starting point Foltz's argument (Reference Foltz1996) that hypertext is unique because it offers different ways of reading online text, but notes the difficulty of determining whether comprehension strategies differ in the two environments, since little investigation of this has been done. Existing studies have focused on online reading strategies, and the differences between processing online and printed texts have not been fully established (Fontanini Reference Fontanini2004), so further studies are needed to fully understand the connections between reading comprehension and hypertext. Online reading comprehension remains under-investigated (Murray, Lloyd & McPherson Reference Murray, Lloyd and McPherson2006), and it is not clear whether different cognitive processes are needed for print and digital texts.

Bolanos investigates the comprehension strategies of good L2 readers as they process expository texts in both print and hypertext environments. Specifically, the study seeks to ascertain if there are any differences in the subjects’ reading comprehension level in the two environments and from these, moves into investigating the connections between reading environments and reading strategies and the relationship between reading strategies and level of comprehension. Using the grounded theory model (e.g. Glaser Reference Glaser1992), the study identifies the comprehension strategies that emerged from the data in the verbal protocols of the respondents; these strategies are then compared to established categories. The theoretical underpinning for the study is the view that reading is an interactive, constructive, meaning-making and problem-solving process (Spivey 1987, as cited in Coiro & Dobler Reference Coiro and Dobler2007).

Sixteen college freshmen students, all female and considered good readers, were randomly assigned to read three editorials in one of the two reading environments. The research location was a private tertiary institution for women. Experienced readers were required, as the subjects needed a high degree of independence in processing both print and online texts. Each participant read the texts and as she did so, expressed her reading strategies aloud. For each student, the comprehension level and strategies employed were then compared. Three editorials from a broadsheet newspaper, the Philippine Daily Inquirer, in its print and online forms, were the primary reading materials. The think-aloud protocols were transcribed in videos. After reading each of the three texts, the participants were asked to summarize each one and respond to two questions measuring their comprehension. Additional means of eliciting information were individual interviews of the participants, observations as they read and thought aloud, focus group discussions, and a survey questionnaire on reading experiences on the internet, as adapted from Coiro & Dobler (Reference Coiro and Dobler2007).

The key findings were the following. First, the readers employed a wide repertoire of strategies for processing the texts; virtually the same strategies were used for both print and hypertext. Second, reading comprehension level was more or less the same in both environments, although the print readers had a slight advantage. This finding suggests that reading environments do not significantly affect reading comprehension either in level or quality. The comprehension level of good L2 readers remains the same, whatever the reading environment. Third, there was a significant correlation between reading environment and reading strategies, implying that an environment may trigger the use of a particular strategy. Lastly, there was also a significant correlation between reading strategies and level of comprehension; the more frequently and/or effectively the strategies are used, the better and/or more effective the reading comprehension.

In view of these results, the study recommends strategy-centered reading programs that cut across curriculum and academic levels to strengthen the ability of readers to process texts. Bolanos's findings are corroborated in Mante's study (Reference Mante2009), the results of which show that seven dimensions of motivation, such as social and learning environment, pressure, and mastery orientation, correlate with various types of metacognitive strategies.

The Bolanos study is significant in its use of reading tasks in both print and hypertext, thus correcting the lopsided methodology reported in similar studies. Reading strategies and processing behaviors are specified where previous studies merely describe reading behavior in general terms such as ‘active readers’. The emphasis is on the process of decoding, rather than the product, stressing the need to develop the ability to discover meanings reflexively; instead of lower-order skills such as word recognition or memorization, higher-order skills such as contextualization, comparison, translation and problem-solving are utilized to process meaning. The study uses a hypertext sample requiring only simple navigation, but in the complex and vertiginous space of the cyber world, ‘the text’ can be elusive, and only an alert reader with an arsenal of metacognitive strategies could survive.

Another area of notable doctoral research work in the Philippines is genre studies (Jawid Reference Jawid2009; Songcuan Reference Songcuan2009; Villegas Reference Villegas2010), but Villegas's dissertation is reviewed here because its socio-cultural trajectory (while similarly deployed in other works) is longer and is directed towards the relationship of genre and the writing process.

The research examines the authorial voices in academic journal articles using English from four cultures: Japanese English, Korean English, Philippine English and Singaporean English. Thirty articles from journals in education and linguistics comprised the corpus but the sampling was taken from the sections of Results and Discussion, Conclusion and Recommendations. Villegas examined the linguistic features and textual properties that realize the two related variables of authorial voice and positioning. From these, he aimed to put together a typology of authorial voices based on the types found to be dominant in each article; this was to be used as the basis for a cross-cultural comparison of authorial voices. The results would form the rubric for a model of authorial voice that could assist students both in understanding and writing academic texts.

The study locates authorial voice within the scholarship on varieties of English pioneered by Kachru (e.g. Reference Kachru1982), which soon developed into a vibrant research field. As theoretical underpinning, the study uses contrastive rhetoric (Kaplan Reference Kaplan1966; Leki Reference Leki1991; Connor Reference Connor1996), which traces its roots to the Sapir-Whorf hypothesis. However, most studies discuss the rhetorical aspect of voice: few are concerned with its grammatical aspects, to which attention must first be given, especially in the context of non-native learners of English.

The dissertation reveals the features writers use to emphasize their ideas, negotiate with readers and own the text. These linguistic features have to do with personal references, hedging, emphatics and attitudinal markers, while textual properties include reporting, voice and citing. Through these features, each writer's positioning in specific sections of the journal became evident; these were identified as asserting an agenda, pointing to details, establishing relations with the reader and evaluating claims. The dissertation concludes that different English varieties deploy culture-specific linguistic resources to express authorial voices. However, an authoritative type of voice in presenting research findings cuts across all four cultures under study.

A weakness in Villegas's study, however, is that it appears to have assumed a priori that the writing in academic journals in these countries is representative of their varieties of English, a tendency in similar earlier studies. Moreover, academic journal writing conforms to norms and procedures laid down by a particular discourse community. Villegas does not take this factor into account with respect to the journals in the sample, and he admits that some editors and/or referees of the journals in his sampling may be native speakers of English. Future studies of voice in ELT would benefit from a look at power asymmetries at work in seemingly cohesive discourse communities.

The issue of L2 or L2s shaped alongside dominant languages such as English and Filipino is taken up in Valdez's (Reference Valdez2008) study, which formulates a model for a culturally competent Filipino bilingual based on the researcher's observations of Filipino-English code-switching in the domains of business, education, law and government, religion and media. Specifically, it seeks to determine the functions of code switching, the socio-cultural factors influencing it, and the values, ideologies and norms embedded in bilingual interactions.

The research design was qualitative-descriptive, in which discourse analysis was used to examine the relationship between cultural factors that influence Filipino-English code switching. It used transcripts from a corpus of naturally-occurring data from the domains of religion, business, law and government. The corpus was collected over a period of eight months from March to November 2006. Numerous communicative situations occur in these domains, so they were considered ideal for scrutinizing bilingual interactions. Cultural factors such as ideologies, values and norms can be articulated in interactions, since the domains in question have set role relationships, locales and unique communicative situations that can illustrate the local and global contexts (Labov Reference Labov and Giglioli1970; Li Reference Li and Auer1998). Conventional role relationships between interlocutors, settings and topics influencing language are also apparent in these situations (Fishman Reference Fishman and Li1965; Sibayan Reference Sibayan1999; Pascasio 2005.)

The study yielded a number of findings. First, it identified the functions of code switching particular to each domain; not all, however, were exclusive to one domain. For instance, message qualification, lexical borrowing and personalization vs. objectivization were found to occur everywhere. Repetition and quotation were used frequently in the education and religious domains. Second, the functions of code switching were driven by certain socio-cultural factors: topic, identity (e.g. expert or non-expert, peer or subordinate, community member or Biblical scholar), social relations (e.g. asymmetrical or symmetrical relationships attributed to rank, social status or expertise) and strategy (e.g. to fill in content gaps in communication, to mark respect, or to function as utterance filter and boundary marker). Lastly, values were embedded in samples of code switching in the various domains (e.g. interactions in business, law and governance education, and media were marked by smooth interpersonal relations and personalism). ‘Ideologies of experience’ performed functions such as bringing about a teacher-centered classroom situation, agreements in business negotiations, and democratic practice in law and government. ‘Ideologies of power’ were observed to control the behavior of speakers. (The term ‘ideologies’ is loosely interpreted and seems to refer to a collective experience or a set of practices.)

A model for the Filipino Bilingual's Cultural Competence is constructed from these findings on Filipino-English code switching. Three main factors shape cultural competence, namely (1) values and ideology derived from local and collective norms, (2) social contexts such as social relations and identity and (3) elements of linguistic interactions such as topic, the structure of interactions and the functions of code switching.

The dissertation observes Filipino-English code switching in all these domains, and thus offers an authentic corpus for a model of the Filipino-English bilingual's competence. It also synthesizes the multifarious factors shaping such cultural competence. The literature review of code switching both in the international and local scholarship is comprehensive, so based on this and the findings of his study, Valdez's suggestion that the language of Filipino-English code switching is an emerging language variety becomes plausible. Wisely, he works in an area prepared by research on bilingualism and code switching in the Philippines which for the past four decades has focused on linguistic investigation (e.g. Pascasio Reference Pascasio1977; Bautista 2007).

However, the study's concept of ‘culture’ is vague and uncommitted. There is no attempt to critique given ideas of culture (and, similarly, of subculture and ideology) and how these could be reconceptualized to suit the Philippine context. Moreover, the study needs both a more nuanced discussion of the Filipino value system showing the dialectics between various Filipino theorists, and a clear definition of ideology. Nevertheless, the study is significant: by providing a macro-context for bilingualism in the Philippines, it could help inform current and future research in ELT in bilingual contexts.

A fifth area of research in the field concerns imaging and the concept of face, a common concern in recent dissertations (e.g. Gonong Reference Gonong2007; Palma Reference Palma2008). Implicit in these studies, and also true of similar studies above (Bolanos Reference Bolanos2010; Flores Reference Flores2010), is an uneasiness about the dehumanization and manipulation that come about in the shift from print to cyber literacy; there is therefore continuing research interest in the self and other in the cyber landscape (Gustilo Reference Gustilo2006). Since the self is negotiated in relation to the other, studies of this kind inevitably touch on ideological issues that a critical discourse analysis can unmask. But a broader theoretical scaffolding would be pragmatic theory, which locates these negotiations within speakers’ everyday lives and their particular milieus as shaped by various cultural variables. Mascuñana's (Reference Mascuñana2007) research on directives in the Cebuano language in Central Philippines is especially relevant to the teaching and learning of English in the Philippines as it offers a theoretical base for managing the continuing shifts in technology that have impacted on language and education. Moreover, by drawing from the multilingual base of the Brown & Levinson model (Reference Brown, Levinson and Goody1978), which was formulated from data in the languages of English, Tamil and Tzeltal, and the other vernaculars in Cebuano, her study underscores the need for locating the teaching of English not only in and among Englishes, but in the mother tongue and other languages native to the English speakers.

Mascuñana's hypothesis is that ‘certain Cebuano cultural norms and values foster a construal of self underpinning the linguistic enactment of politeness in directives’ (vi–vii). In this light, her study aimed to answer the following questions:

1. 1. Of the five major strategies identified by the Brown & Levinson model (Reference Brown, Levinson and Goody1978), what is the preferred mode in Cebuano directives?

2. 2. What types of politeness redress or face orientation (mitigation devices used to soften the threat to one's face) does Cebuano utilize for face-needs (in minimizing the threat to loss of face of interlocutors)?

3. 3. Is the choice of strategies in Cebuano directives affected by the social variables of social distance and relative power between interlocutors, and the degree of imposition of the directive, as posited by the Brown & Levinson model (Reference Brown, Levinson and Goody1978)?

Through these questions, Mascuñana sought to test the validity of some common assumptions of the Brown & Levinson model (Reference Brown, Levinson and Goody1978) concerning the pragmatics of politeness, and to point to new directions for theorizing on linguistic politeness. Brown & Levinson's starting point for analysis is the speech act, which is the basic unit of communication (Searle 1969). Speech act theories, however, have been criticized for their attention to the illocutionary force of speaker utterance rather than the hearer and the development of the interaction. To overcome this limitation, the researcher selects directives from literary texts where they are embedded in interactional contexts. Furthermore, literary texts provide a vantage point for observing society and making ‘ideological structures tangible’ (Stubbs Reference Stubbs1996). A total of 160 directives were taken from 137 short stories published in popular magazines in the year 1995.

The results of the study show that the preferred mode in Cebuano directives is the negative politeness strategy; the face orientation the speaker shows for the face needs of the hearer is the negative face orientation. Face, as defined by Brown and Levinson (61) is the ‘public self-image’ that each person wishes to claim for himself or herself. It is of two types: positive face or the desire to be appreciated, and negative face, the desire to have freedom to do what one wants unimpeded by any imposition. The choice of strategy appears to be influenced by the contextual variables of social distance, degree of imposition, and the relative power relationship between hearer and speaker. The findings reveal a link between the Cebuano bilateral kinship system and the linguistic enactment of politeness in directives, as demonstrated by the absence of conventional indirect forms of directives in Cebuano. These indirect forms have been observed to be prevalent among English-speaking communities. The notion of face has been characterized as individual (Goffman Reference Goffman1955) or public (Brown & Levinson Reference Brown, Levinson and Goody1978) but the Cebuano concept of face actually has two faces – the individual and the public. The element of the ‘public’ in face situations leads to a culture-based interpretation of contextual variables. In Cebuano, where a premium is placed on relatedness in the community, social distance is minimized (for example, through in-group markers). Power is infused with the favorable attributes of benevolence, kindness and nurture while imposition is underplayed through reciprocal exchange, linguistically performed through hedges and imposition-minimizers. In light of these findings, Mascuñana proposes a modification of the Brown & Levinson model (Reference Brown, Levinson and Goody1978).

Surface changes in language, medium and mode can be positioned properly within the base of pragmatic theory – its core principles of interaction performed within the cultural variables of social distance, imposition and power (Brown & Levinson Reference Brown, Levinson and Goody1978: 76). Thus, technology-mediated communication in English will not be just a practical concern for strategies of manipulation and the development of new types of literacy skills, but a pragmatic concern that takes into account these three cultural variables. Mascuñana's study offers a theoretical grounding for ELT and multilingualism in the Philippines, particularly in light of the recent push for the mother tongue as medium of instruction in the first four or five years of elementary education before the shift to English and Filipino. However, since its data is derived from a literary corpus in just one regional language, there is a need to broaden its applicability by testing the model in natural and spontaneous conversations across a number of regional languages.