Revisiting Second Language Vocabulary Teaching: Insights from Hong Kong In-Service Teachers

In second language (L2) vocabulary research, much attention has been paid to how learners can be supported to acquire vocabulary and how teachers can be better prepared to teach vocabulary in the language classroom. As research efforts have been devoted largely to exploring effective vocabulary teaching strategies and techniques, it remains unclear what teachers know and believe regarding L2 vocabulary acquisition beyond the commonplace conception that teaching and learning are interrelated. The present study explored the epistemological and pedagogical beliefs about vocabulary development reported by four in-service English language teachers in Hong Kong through in-depth semi-structured interviews. It also examined the major factors shaping those beliefs. The results suggest that although the interviewees considered vocabulary teaching to be pivotal, they seemed to focus only on certain aspects of lexical knowledge and rely on a limited range of strategies in teaching vocabulary. Implications for L2 education and teacher development are drawn.


Introduction
Both the crucial role of vocabulary development in language mastery and its taxing nature warrant analysis of beliefs about second language (L2) vocabulary teaching and learning. The literature on L2 acquisition (SLA) frequently emphasises vocabulary's role as a fundamental building block of comprehension and communication (Nation 2013). Laufer (1998) further affirmed the importance of word building, suggesting that the major distinction between language learners and native speakers is lexical competence. Given the significance of vocabulary, it is somewhat surprising that the value of devoting teaching time to words, particularly using rich instruction, is often questioned, possibly because of the complex nature of word acquisition (Nation 2013). How language teachers should deal with various aspects of word knowledge in class, taking time constraints into consideration, however, remains unclear. Although teachers of English as a second language (ESL) are often left to their own devices in teaching vocabulary in the absence of clear guidelines, forced to rely on their own beliefs and the teaching materials provided, only a handful of studies to date have focused on what they know and believe about L2 vocabulary teaching and learning. The case study reported herein thus sought to identify the beliefs about vocabulary teaching, and the factors shaping those beliefs, held by four in-service secondary ESL teachers in Hong Kong, where the need to enhance students' vocabulary development is frequently highlighted (see, e.g. Choi and Ma 2015;Tang 2007;Tang et al. 2016). Through analysis of data gathered from interviews with these teachers, the paper identifies several issues relating to vocabulary teaching that warrant attention and discusses their implications for L2 education and teacher development.

Vocabulary Teaching in Hong Kong
In Hong Kong, vocabulary development is deemed merely ''part of the process of language skills development'' (Tang and Nesi 2003, p. 67). Research by Tang and Nesi (2003), for example, suggests that most of the territory's English language teachers confine themselves to teaching pronunciation and/or the meaning of vocabulary during lessons. One plausible explanation for teachers' failure to introduce large numbers of vocabulary items in class is the omission of a list of specific lexical items to be taught and detailed guidance on vocabulary teaching in such English language education curriculum documents as the English Language Education Key Learning Area Curriculum Guide (Primary 1-Secondary 6) published by the Curriculum Development Council (2017). Given Hong Kong's examination-oriented culture, however, it is reasonable to speculate that there is a general expectation that vocabulary items will be explicitly taught in class. For example, the assessment framework for the Hong Kong Diploma of Secondary Education Examination (HKDSE) states that students' vocabulary knowledge will be assessed in the different papers comprising the English language examination (Hong Kong Examinations and Assessment Authority 2015). To better understand vocabulary development in Hong Kong, it is necessary to obtain relevant information on teachers' beliefs, as well as the factors shaping those beliefs, particularly as it is widely acknowledged that there is a significant relationship between teaching and learning practices and that beliefs play a crucial role in guiding instructional practice (Borg 2001).

Teacher Beliefs about Vocabulary Development
It is generally accepted that beliefs can be conceptualised as an aspect of cognition (see, e.g. Borg 2006). Borg (2001) defines beliefs as psychologically held understandings or propositions that are subjectively accepted as true by the individuals who espouse them despite recognising the availability of alternative beliefs held by others. Contrary to knowledge, which hinges on objective facts, beliefs are based on evaluation and judgment (Pajares 1992), although it is important to note that the two are intricately connected. In the education arena, beliefs can be broadly categorised into two types. Epistemological beliefs concern the nature and acquisition of knowledge (Schommer 1994), whereas pedagogical beliefs represent educational beliefs about teaching and learning (Ertmer 2005). Current interest in epistemological beliefs can be dated to the pioneering work of Schommer (1994), who maintained that what learners believe about the nature of knowledge and its acquisition comprises five general epistemological dimensions. I argue, however, that such beliefs are more fine-grained and can be related to specific aspects of language teaching. Here, epistemological beliefs pertinent to vocabulary development are defined as beliefs about the nature of vocabulary knowledge, which can be broadly categorised as beliefs about the role of vocabulary in SLA and the notion of a word. Pedagogical beliefs, in contrast, refer to teachers' beliefs about vocabulary teaching, for example, their beliefs about lexical instruction, class activities and resources for effective vocabulary building.
Only a handful of studies to date have investigated teacher beliefs about vocabulary teaching and learning. Based on content analysis of survey responses to openended questions collected from 250 teachers in Hong Kong and mainland China, Gao and Ma (2011) identified four categories of vocabulary teaching beliefs, concluded that such beliefs are mediated by contextual conditions and educational experiences and called for English language teachers in both locales to diversify the pedagogical activities they apply to vocabulary learning. A similar study was conducted by Macalister (2012) to determine how the vocabulary-related beliefs of Malaysian pre-service teachers differed from those of their teacher educators. After 82 respondents completed a questionnaire, 12 preservice teachers and 22 teacher educators were asked in interviews to imagine how they would conduct lessons with upper primary students based on two given texts. Although the two groups' beliefs coincided to some extent, the former devoted minimal attention to vocabulary, and none to vocabulary learning strategies, in their imagined lessons. Whilst these two studies offer insight into word learning from the teacher's perspective, caution should be exercised in interpreting their data, as their use of normative statements precludes a thorough understanding of the complex nature of beliefs about vocabulary building.
When it comes to elucidating teacher beliefs, Borg's (2006) framework seems particularly insightful, as it outlines the relationship amongst teacher cognition, teacher learning and classroom practice. His work suggests that teachers' schooling experience and professional education exert an impact throughout their careers. It also argues that teachers' cognition and practice are reciprocally informing, with contextual factors mediating the degree to which teachers adopt practices consistent with their cognition. However, the framework has relatively little to say about the extent to which its individual elements, that is, schooling, professional development, contextual factors and classroom practice, inform teacher beliefs. Figure 1 adapts Borg's framework with a specific focus on vocabulary teaching and learning for further exploration. The

Methodology
The study was framed as an instrumental case study (Stake 2005). It examined teachers' beliefs in a multifaceted manner to provide a contextualised account of how those beliefs are shaped.

Participants
The four teacher participants had all taught at the same English-medium-of-instruction secondary school in Hong Kong since beginning their careers as full-time teachers (see Table 1 for background information). Their school's ''band one'' ranking suggests that the overall academic ability of its students is above average. To prepare students to sit the HKDSE upon completion of their studies, English language lessons are primarily skills-based. The English language textbooks are used across forms, and teachers are thus required to teach vocabulary items pertaining to the themes of various textbook modules. The four participating teachers were selected because they were native Cantonese speakers who had received their education and teacher training in Hong Kong. Their educational background allowed me to examine how teachers who had learned and now taught ESL view themselves vis-à-vis English language education in an Asian context. Also, because teachers with varying amounts of experience are likely to espouse differing beliefs owing to differences in schooling and teacher training, selecting informants with a wide range of teaching experience (i.e. 5-22 years) afforded a more comprehensive picture for analysis.

Data Gathering
The study's data were derived from in-depth, semi-structured interviews. The audiotaped interviews, which were conducted in English to prevent the original meaning from getting lost in translation, lasted approximately one and a half hours each. As shown in Appendix A, the interviews began with two questions inviting the teachers to discuss their epistemological beliefs about the role of vocabulary in SLA and the notion of a word, followed by questions designed to examine their pedagogical beliefs about vocabulary teaching with reference to Gao and Ma (2011), one of the few studies analysing teachers' beliefs about vocabulary teaching and learning in an Asian context. Gao and Ma's framework positing four types of vocabulary teaching beliefs-i.e. beliefs about teaching content, approaches, aims and resources-was adopted because those types were derived from teachers' responses to openended questions rather than existing categories, although I made several adaptations to it (see the discussion below). The questions on vocabulary teaching and what it involves were intended to reveal the participants' beliefs about teaching aims and content, respectively. They were followed by an exploration of beliefs about approaches to vocabulary teaching, beginning with the general question ''How should vocabulary be taught?'' and moving to specific questions pertaining to effective vocabulary instruction strategies and class activities designed to enhance vocabulary acquisition. Each interview ended with a discussion of the teacher's beliefs concerning her preferred vocabulary teaching resources and the factors shaping her beliefs.
Here, it must be noted that the selection of questions was influenced by my own assumptions. For instance, recognising that teachers are likely to espouse a range of beliefs about approaches to vocabulary teaching, I decided not to narrow the scope of analysis to whether vocabulary learning is promoted through presentation or practice, as Gao and Ma (2011) did. Instead, I examined the teachers' beliefs with reference to their use of such vocabulary instruction strategies as paraphrasing target words and presenting synonyms and antonyms. In addition, I  Nation's (2013) comprehensive taxonomy presenting nine aspects of lexical knowledge based on three broad categories, namely form, meaning and use, to elicit the teachers' epistemological beliefs about the notion of a word. To enhance the clarity and comprehensiveness of the interview protocols, I conducted a pilot study with two secondary school English language teachers prior to gathering data for the main study. The wording of some interview questions was revised based on their feedback.

Data Analysis
The interviews were transcribed verbatim and coded to identify recurring themes that illuminated the research questions. To prevent preconceptions and a priori theoretical knowledge from colouring the findings, no pre-specified codes were used to analyse the data (Flick 1998).
Owing to a lack of specific rules for interpreting qualitative data, and to ensure the validity of the research to the greatest extent possible (Punch 2005), I read the transcripts repeatedly to familiarise myself with their content before assigning codes representing the teachers' beliefs. The coding process was iterative in nature, with codes added as new ideas emerged and all transcripts reanalysed after preliminary coding. To reduce bias, a research assistant helped conduct an inter-rater reliability check, with any disagreements resolved through discussion. Member checking was also performed, with the participating teachers asked to read through the transcripts and preliminary data analysis to ensure the accuracy of my interpretations.

Results and Discussion
In this section, the salient themes identified in the interview data are reported and interpreted in light of the study's two overarching research questions to frame the in-depth discussion of issues pertinent to vocabulary teaching and learning.

RQ1: Teachers' Focus on Certain Aspects of Word Knowledge
The evidence collected in this research suggests that although the participating teachers' beliefs appeared to resonate with the view that vocabulary plays a vital role in L2 comprehension and communication (Nation 2013), their epistemological beliefs about the notion of a word and pedagogical beliefs about lexical instruction raised questions about their understanding of what constitutes word knowledge and vocabulary instruction. In response to the question ''What is meant by knowing a word?'', for example, the teachers focused only on certain aspects of word knowledge, with all four citing meaning and pronunciation and none acknowledging concept and referents. Whilst T1 and T2, who were relatively less experienced, addressed five aspects of word knowledge in the interviews, T3 mentioned only spelling, meaning and pronunciation. T4's apparent difficulty discussing the notion of a word was particularly surprising given her many years of experience. She said, ''Knowing a word … I think that means pronunciation and meaning … and … I don't know''. Interestingly, some interview responses seemed to exhibit the teachers' awareness of the multi-dimensionality of word knowledge. For example, T1 and T4 complained about their students' use of ''basic words'' such as ''happy and worried'' to express ideas and reluctance to use more complex vocabulary, whereas T2 and T3 were concerned that their students seemed not to know ''how to use words properly''. Also relevant to various aspects of word knowledge are the inconsistencies between the teachers' epistemological beliefs about the notion of a word and their pedagogical beliefs about the content of vocabulary teaching, as revealed in their interview responses. For instance, T1 mentioned word form and collocation in discussing the notion of a word but suggested that the two aspects of word knowledge do not require much teaching as they can be learned incidentally. Similarly, T2 cited constraints on use in explaining the notion of a word but said she did not find it appropriate for teachers to address the aspects of word knowledge frequently in class because learners might be ''overwhelmed with the new information''. Finally, T4 claimed that vocabulary learning involves the mastery of spelling but admitted that she rarely addresses spelling in class because she believes that the explicit treatment of pronunciation helps students to spell and that teachers should draw students' attention only to the spelling of words that are commonly confused.
Taken together, these results seem to confirm Johnson's (1992) argument that one's theoretical orientation and beliefs about teaching may be inconsistent. More importantly, they highlight the need to address Brown's (2010) concern that language teachers often equate vocabulary learning with the mastery of certain aspects of lexical knowledge such as meaning and pronunciation, resulting in scant or even no attention being devoted to other aspects. Whilst establishing a form-meaning connection for unknown words can be perceived as an initial step in vocabulary learning, Tang et al. (2016) advise ESL teachers to develop a rounded view of word knowledge to ensure adequate coverage of the various aspects of lexical knowledge.

RQ1: Teachers' Preference for Particular Vocabulary Instruction Strategies
One striking feature of the interview data was the teachers' undue reliance on memory and guessing strategies in their reported beliefs about the use of vocabulary instruction strategies in class. There were no mentions of any social, cognitive and/or metacognitive strategies being adopted for vocabulary development. T1, T3 and T4 suggested that vocabulary learning can be consolidated by connecting a new word to its antonyms and/or synonyms, although none of the teachers considered word associations crucial to word acquisition. T2 and T4 stressed the importance of using the target language to teach vocabulary because of the need to maximise students' exposure to the language and school policies mandating the use of English as the medium of instruction. The former further commented that students' inability to explain the difference amongst ''apart from'', ''besides'' and ''except'' because of their reliance on Chinese translation had reinforced her belief in the importance of target language use. Other memory strategies suggested by the interviewees included memorising affixes and roots, using new words in sentences, drawing students' attention to the part of speech of a target word, studying the sound of the word and making use of pictorial representation. Discovering a new word's meaning by guessing from context was the only determination strategy cited. In the teachers' opinion, students should be encouraged to guess a word's meaning without assistance because ''nobody will tell the students what a word means during their examinations'' (T1) and ''the use of dictionaries is forbidden during assessments'' (T3). Although it is unclear whether the teachers' beliefs are manifested in their classrooms, it is worth noting that a combination of strategies has been demonstrated to contribute to effective learning (Macaro 2001). The introduction of a wide range of vocabulary learning strategies in the English language classroom is important because it allows individual language learners to choose those they find most effective.

RQ1: Discrepancies between Teachers' Reported Beliefs and Practices
Whilst observing the teachers' behaviour was beyond the scope of this research, analysis of their interview responses reveals a discrepancy between their professed beliefs and reported practices. A particular area of dissonance was their preferred instructional materials for vocabulary development. Although there was general consensus that the use of authentic materials such as newspaper articles and songs is ideal for motivating student learning, T3 and T4 both reported limited such use in class. T3 explained that the time-consuming nature of locating authentic materials suited to students' proficiency level dissuaded her from using them in class, whereas T4 stated that she rarely makes use of such materials because students often lose them even if they find them interesting. She recalled handing out newspaper articles on the popular game Pokémon Go!, noting that the students started reading them without prompting. However, she admitted that she generally relies on standardised resources instead because students' loss of such supplementary reading materials causes her to feel that the effort expended in lesson preparation has been wasted.
Another area of inconsistency lay in the teachers' stated beliefs about the use of vocabulary enhancement activities in class and their reported practices. Although all of the teachers considered such activities as games, crossword puzzles and film appreciation to be conducive to vocabulary development, only T1 reported incorporating a variety of such activities into her lessons. The other teachers identified a number of factors precluding them from the regular use of certain activities. T2, for instance, described how interesting vocabulary activities can help to cultivate an interest in learning but lamented the limited class time available for their implementation. Her usual practice, she continued, is to ask students to complete textbook gapfilling exercises focusing on meaning and grammatical functions. T3, who clearly thought highly of the school syllabus, underlined the importance of aligning all vocabulary activities with the reading texts and writing tasks therein. She stated that she regularly conducts group discussions to fulfil examination requirements although she would prefer to watch films with her students to promote incidental vocabulary learning. Finally, T4 shared her worries over in-class discipline. She described an instance in which she had asked students to post pictures of different breakfast items on the whiteboard and then write down words associated with them. The students became so overexcited that she lost control of the classroom. Accordingly, she now generally uses explicit instruction to teach vocabulary despite her belief that class activities eliciting learner production of vocabulary might be more effective.
The clear discrepancies between beliefs and practices revealed by the interview data seem to result from the constraints imposed by such contextual factors as time limitations, the school syllabus and student-related factors. The mismatch is corroborated by the findings of Lee (2008) and Underwood (2012), who both reported inconsistencies between language teachers' beliefs and behaviour.

RQ2: Factors Accounting for Teachers' Beliefs
The interview data indicate that teachers' beliefs are primarily attributable to three factors: their own language learning experience at school, contextual factors and their professional development experience.

Language Learning Experience at School
Teachers' own schooling was found to be a major contributor to the construction of their beliefs about vocabulary building. In the interviews, all four participants acknowledged their English teachers' influence on their own beliefs and/or practice. For example, T1 reported that her teachers had provided her with detailed feedback on her writing assignments as part of vocabulary development, a practice that shaped her beliefs about the need to scaffold students' vocabulary learning based on their written work. T3 said that one of her primary school teachers had taught vocabulary by blanking out key words in song lyrics and playing the songs in class. It is possible that this experience was responsible for her belief about the importance of using authentic materials to facilitate vocabulary teaching and learning. Similarly, T4's beliefs about the efficacy of learning vocabulary through memorisation were shaped by her own ''successful'' experience with that approach. She emphasised that she asks her students to memorise particular words because her teacher had done so.
For T2, the association between her own school learning experience and her current beliefs was somewhat different. Unlike the three teachers who believed their own teachers' practices to have been effective, T2 explained that she did not consider vocabulary teaching important because her teachers had paid scant attention to it: ''My teachers influenced how I thought about the teaching of vocabu-lary…. It's [only] after I became a teacher that I gradually realised its importance because of my students' [poor] performance''. Her remarks not only highlight how her own schooling had influenced her beliefs in the early years of her career, but also show how she subsequently reconstructed those beliefs based on students' actual learning needs. Taking her students' poor test performance and limited responses in class to result from their lack of vocabulary, she began to allocate more in-class time to vocabulary teaching, she reported.

Contextual Factors
Contextual factors, such as school policies, textbooks and the English language curriculum, also played an important role in shaping the participating teachers' beliefs. Whilst they all claimed to be afforded considerable flexibility in teaching vocabulary, it became evident during the interviews that school policies influenced their beliefs about such teaching, particularly those related to lexical instruction. As already noted, the teachers found it crucial to teach vocabulary using the target language partly ''because of school policies''. Closely related is the influence of school policies on textbook use. All four teachers highlighted the role of commercial textbooks in either shaping their beliefs or determining the choice of words they teach explicitly in class or both. According to T1, consensus on vocabulary teaching is ''reached based on the content of the textbooks'' because their use ''save[s] a lot of time''. Echoing that remark, T3 stated that teachers ''basically follow the themes set in the textbooks''. She further explained that 10-12 key vocabulary items are usually highlighted in a reading passage, and then ''recycled'' throughout the module, a practice that has raised her awareness of the significance of repetition in vocabulary acquisition. The teachers also reflected on how the English language curriculum has shaped their beliefs. Noting that the curriculum places ''a lot of emphasis on task-based learning'', T2 felt that vocabulary should ''not be introduced out of context'' and that teachers should set tasks requiring students to express vocabulary items in different ways. A somewhat unexpected finding, however, was that despite the curriculum's influential role in their beliefs, none of the teachers thought highly of the curriculum guide with respect to vocabulary teaching in general. For example, T1 noted that the guide offers little insight into how vocabulary should be taught, and T4 confessed that she had read it a ''long time ago'' and forgotten almost all of its content.
Social expectations and student factors also influenced the teachers' beliefs. Hong Kong students are expected to excel in public examinations, and examination pressure was thus cited as a significant contributor to those beliefs. All four interviewees claimed that they often encourage students to guess the meaning of a word in class because they perceive it to be an effective examination strategy. ''Examination pressure'' also led to T1's belief that she should teach students how to use words with the correct degree of formality and purposefully introduce synonyms for a given word to help students achieve success in various assessment tasks. Further, it was the explanation T4 gave for her view that it is inappropriate to play games with senior secondary students (see the following section). Examination pressure, together with such learner factors as discipline and proficiency level, seems to have played a pivotal role in both shaping the teachers' beliefs about vocabulary teaching and guiding their pedagogical decisions.

Teacher Professional Development
The final contributor to shaping teachers' beliefs is professional development in the form of formal teacher training, the reading of scholarly work and academic exchanges with colleagues. For example, T1 stated that her beliefs about vocabulary instruction had been shaped by Revisiting Second Language Vocabulary Teaching: Insights from Hong Kong In-Service Teachers 505 reading. To illustrate, she highlighted the role of repeated exposure to vocabulary in lexical development, reporting that she ''once read an article about how students can remember a new vocabulary item'' by being exposed to it ''at least seven times''. Accordingly, she ''purposely repeat[s] words again and again in class''. The other teachers, in contrast, were not regular readers of scholarly books and articles on vocabulary teaching. T2 saw no need to read about vocabulary building because her department does ''not require teachers to come up with any new strategies for vocabulary teaching and learning'', whereas T3 considered such reading of little practical value. In her opinion, teachers are ''lone fighters in the classroom'', and should thus ''think of something practical'' rather than relying on theories. T4 reported that she reads academic books only during ''the long holidays'' despite having bought ''a lot'' of them. As most such books focus on teaching and learning vocabulary through games and activities, she considered them appropriate ''only for junior students''. Beyond reading, the teachers claimed to have limited opportunities and time to explore vocabulary teaching through teacher education, formal training or professional exchanges with colleagues. Of the four, only T2 mentioned engaging in professional exchanges with her counterparts in other schools when asked to discuss the factors that had shaped her beliefs about vocabulary teaching. She explained that she had adopted the recommendation of a friend who works at another school to ''teach junior form students more difficult words so that they would have more chances to recycle words before they take the HKDSE and remember the words better''. According to T4, the teachers in her school rarely discuss issues related to vocabulary teaching despite holding regular panel meetings, with most claiming that they simply ''[do] their own thing''. Strikingly, all four participants considered their formal teacher education to have played a minimal role in shaping their beliefs about vocabulary teaching. T2 acknowledged that ''there's not much formal training or education specifically about vocabulary teaching or learning'', whereas T3 explained that all of the courses she had taken were skillsbased, with none focusing on vocabulary teaching. T2 further reported that only one workshop on vocabulary acquisition is offered by the Education Bureau, with the instructor simply making ''general suggestions on exposing students to a wide range of vocabulary using language arts, as if the students will learn the expressions naturally''.
It is clear from these results that compared with the teachers' own schooling and contextual factors, professional development has played a relatively insignificant role in the construction of their beliefs about vocabulary development. Also noteworthy was the teachers' general lack of confidence in vocabulary teaching and acknowledgment of their own problems with vocabulary learning: ''Though I teach English, I don't know a lot of vocabulary items in English … I'm also an ESL learner'' (T1); ''Unlike some native English-speaking teachers, I don't see myself as somebody having a wide range of vocabulary and a range of techniques to teach vocabulary'' (T2); ''I don't think I … learned much about vocabulary teaching in my teacher education classes. I had to learn how to teach by teaching … I'm not good enough''. (T3); ''I did not receive any proper training in vocabulary…. My foundation isn't very good. I'm not very capable of teaching vocabulary'' (T4). These comments suggest that the teachers' low level of confidence can be attributed to their self-identification as ESL learners and lack of appropriate training.
The foregoing findings largely support Borg's (2006) schematic conceptualisation illustrating the relationship amongst teachers' beliefs, learning experience and contextual factors. Although they do not clearly demonstrate how various contextual factors are affected by teacher cognition, they imply that, for the participating teachers at least, such factors and their own schooling experience were highly influential in informing their beliefs. They also provide evidence on the interactive effects between teacher beliefs and classroom practice. As we saw above, for instance, T4 reported conducting a vocabulary activity relating to breakfast items because she believed students would benefit from putting words into active use. However, her negative experience with the activity prompted her to reformulate her beliefs about vocabulary teaching and attach greater importance to explicit instruction. Her example, together with T2's explanation of how her students' lexical deficiencies had encouraged her to alter her beliefs about the significance of vocabulary teaching and allocate more in-class time to it, illustrates how student factors can influence teachers' classroom experience, supporting Borg's (2006) argument that the classroom itself forms a part of the context. This demonstrates why contextual factors in teaching are positioned around classroom practice but not external to it in Fig. 1.
Whilst professional development in various forms seems to have contributed to shaping the participating teachers' beliefs, it appears to have played a rather minor role in moulding their belief systems about vocabulary building. Brookfield (1995) reminds us that teachers need to engage in critical reflection on their beliefs and practice to ensure they do not hold distorted or incomplete assumptions about teaching and learning. However, it can be profoundly difficult for individuals to reflect critically on their own assumptions if they fail to take different perspectives into consideration (Johns 2002). The teachers in the current study reported having few opportunities to participate in professional development activities with an explicit focus on vocabulary teaching and learning, and they considered the theoretical literature to have exerted a limited impact on the development of their beliefs. Given the significant impact of teaching on learning and growing concerns over students' lexical deficiencies in second and foreign language contexts (Choi and Ma 2015;Tang 2007;Tang et al. 2016), it is crucial to ask whether language teachers are being equipped with the professional knowledge required to construct appropriate beliefs about effective vocabulary teaching and make informed pedagogical decisions that maximise student learning.

Conclusion and Implications
The study presented herein constitutes evidence of the value of using qualitative methods to examine the complex nature of teachers' beliefs. It also contributes to scholarly understanding of English language teachers' epistemological and pedagogical beliefs about vocabulary development and the way in which those beliefs are formed. The findings not only reveal the participating teachers' focus on certain aspects of word knowledge and certain vocabulary teaching strategies, but also the limited impact of professional development on their beliefs and practice. Prior to discussing the study's implications, however, several limitations need to be addressed. First, as data were collected from a small number of English teachers at a single secondary school in Hong Kong, the study's findings may not be directly relevant to vocabulary teaching in primary or tertiary education or other research contexts. Second, all of the participants were women, and it is thus possible that different findings would have been obtained from a male or mixed-gender sample. Third, teachers' beliefs were investigated solely through interviews. Whilst the participating teachers' professed beliefs and reported practices revealed a number of interesting issues, triangulation through lesson observations would likely have allowed the identification of tacit beliefs not expressed by the teachers.
Despite these limitations, the study offers valuable insights into L2 vocabulary teaching and has significant implications for L2 education and teacher development. First, in terms of L2 education, it provides support for the view that the importance of vocabulary development should not be downplayed in a headlong rush to adopt skills-based approaches to language teaching. Teachers and learners alike need to be made aware of various aspects of word knowledge (Nation 2013) and a range of learning strategies (see, e.g. Cohen et al. 2006;Schmitt 1997) conducive to vocabulary enhancement. Second, the study's results reveal areas of tension between teachers' professed beliefs about vocabulary teaching and their reported practice, thereby highlighting the need for teachers to make sense of the conditions in which they operate and act consistently in accordance with their expressed beliefs to avoid sending confusing messages to learners (Williams and Burden 1997). The findings also make it clear that educational policymakers need to promote a much sharper focus on vocabulary development in the English language curriculum. Greater clarity is required concerning the role of vocabulary in SLA and the most effective strategies for vocabulary building, for example.
Finally, the findings suggest that teacher development is of pivotal importance in learning to teach and teaching to learn. Regardless of their epistemological and pedagogical beliefs about vocabulary development, all four teachers in this study expressed doubts about their ability to teach vocabulary effectively. In addition, the less-experienced teachers seemed to possess a more comprehensive view of what constitutes a word than their more experienced counterparts, lending support to the argument that experience does not necessarily lead to expertise (Ericsson et al. 1993). Regardless of their number of years of teaching experience, teachers must be afforded ongoing support to identify areas for further improvement and to learn to teach in a way that enables them to make informed decisions favourable to student learning. In both their initial teacher education and ongoing teacher training, they need to be equipped with the pedagogical content knowledge required to facilitate teaching and learning and encouraged to reflect critically upon their own teaching and learning experience, as well as their students' beliefs, colleagues' opinions and diverse theoretical perspectives (Brookfield 1995). A fruitful direction for further research would be to explore how changes can be effected in teachers' beliefs and practice and the impact of those changes on students' actual learning outcomes.