Do teachers know how to teach grammar?- Of beliefs and misgivings, perceptions and reality

images (7)


In the prequel to this post, “Language teachers and Grammar – Should we worry?”, Steve and I partially answered the question in the title through our discussion of language teachers’ KAL (Knowledge About Language) . In that post we reviewed empirical studies which identified serious deficits in pre-service and in-service teachers’ KAL which affect their ability to teach grammar explicitly (Brumfit et al. 1996; Borg, 2001 and 2015; Swan, 2006).

In the present post we aim to widen the perspective on teachers’ ‘know-how’ in the area of explicit grammar teaching. It is important to note that our discussion of the available research will extend beyond the boundaries of what teachers do in the classroom as we believe that grammar-teaching related competence refers also to the extent to which

(a) language instructors’ beliefs and attitudes about language teaching are aligned with current theory and research;

(b) their teaching intentions match students’ learning agendas, preferences and other individual variables (e.g. age, personality, educational biographies, etc.);

(c) they are aware of their own classroom practice – this aspect of teacher competence is important, we believe, as without self-awareness one cannot identify deficits in their performance, the crucial starting point of any effective professional development.

2. Important disclaimer

The research reviewed below is by no means exhaustive or conclusive. It does, however, yield interesting and useful findings which resonate with our professional experience and should be heeded by colleagues as well as course administrators as they point to issues which do commonly hinder effective grammar teaching and learning in secondary school settings.

3. Teachers beliefs about grammar instruction

Beliefs often play an influential role in how effective the teacher will be in the classroom (Richards,1998) by acting as a filter for their instructional judgements and decision making in class and subsequently providing a “systematic justification process with which to plan, assess, judge, decide, accept, deny or act” (Ezzi, 2012).

3.1 Do teachers believe grammar should be taught?

By and large the empirical studies that we located in the core literature found that language teachers feel that Explicit Grammar Teaching is central to language learning and students need direct and explicit teaching of grammar rules for accuracy (e.g. Chandler,1998; Schultz, 1996 and 2001; Burges & Etherington, 2002; Ebsworth & Schweers, 1997; Potgieter & Conradie, 2013; Graus and Coppen, 2014).

It must be pointed out, however, that the existing empirical data are by no means consistent across the board as they are inevitably affected by socio-cultural variables, curricular constraints and students’ agendas. For instance, Eisenstein-Ebbsworth and Schweers (1996) found that, although all of the university ESL teachers they studied thought grammar should be taught, the New York- based ones were significantly less in favour of conscious grammar instruction than their Puerto Rico counterparts. The researchers attributed this phenomenon to the more traditional approach to language teaching usually adopted in Puerto Rico. As one of their informant stated, grammar learning had been such an important part of their learning experience, that they did not see any reason to abandon it. The researchers concluded that awareness of research plays a minimum role in L2 teachers’ rationale for the approach to use in grammar teaching.

Moreover, differences have been identified by some studies between second-language and foreign-language teachers. Mitchel et al. (1992) and Brumfit et al (1996) found that the second-language teachers they studied rarely did any explicit grammar teaching, wherever foreign language teachers regularly did so. This is a finding that definitely resonates with our own experience.

3.2 The main source of teachers’ beliefs about grammar teaching

Eisenstein-Ebbsworth and Schweers’ (1996) findings refer to the all-important point we made in the conclusion to our post ‘Why teachers teach the way we do’, that ultimately teachers’ beliefs about grammar pedagogy are largely shaped by their previous learning experiences much more than by their method classes on teacher training courses and subsequent CPD – which explains why L2 teachers’ grammar teaching approaches are often outdated. Borg (2015) concluded his review of the relevant literature stating that:

In reporting their beliefs about grammar teaching, teachers commonly refer to the impact of their views of their prior language learning experiences; there is evidence that these may exert a more significant impact on teachers’ views than the results of formal research into grammar teaching. This is not surprising: an apparent lack of impact of formal theory on teachers’ cognition has also been reported in mainstream education (e.g. Crawley and Salyer, 1995).

Mitchel et al. (1994) noted how the teachers they studied ‘had been influenced very little by those theories of second-language acquisition that downgrade the role of explicit form-focused instruction in the learning of a foreign language’. Their findings are echoed by a number of other studies.

In Chandler (1998), for instance, the vast majority of the respondents to a postal questionnaire stated that most of their knowledge of grammar was learnt in school and constituted the basis of their existing pedagogy. Chandler was concerned with the lack of understanding of the role of language awareness in language learning which he referred to as ‘confident ignorance’.

Nespor (1987) found the teachers can be influenced by ‘a crucial experience or some particularly influential teacher produces a richly-detailed episodic memory which later serves the student as an inspiration and a template for his or her own teaching practices”.

Farrell (1999) reported how his subjects refused to teach grammar deductively as it had not worked for them as language learners.

Other factors appear to affect teachers’ beliefs about grammar. Ebsworth and Schweers (1997) claim that, when it comes to “articulating their rationales, teachers referred to various factors shaping their views, such as student wants, and syllabus expectations”. Borg (2015) notes how in many cases teachers choose to teach grammar not because they actually believe it will enhance language acquisition, but because they believe that students expect it. Andrews (2003) notes how schools’ microcultures affect teachers beliefs (e.g. the beliefs and practices of one’s colleagues).

4. What are teachers’ instructional preferences ?

Research suggests that modern language instructors still favour the PPP (presentation, practice, production) instructional sequence marked by a deductive approach to grammar teaching with teacher-to-class interaction and drills as the preferred mode of delivery and practice. This is often the case even when they believe that language teaching must develop student communicative ability (Andrews, 2003; Wang, 2009; Borg, 2015). Despite the teachers investigated had received formal training in Communicative Language Teaching, little evidence was found of CLT in their grammar teaching classroom practices. This reflects findings by Richards and Pennington (1998), Sato and Kleinsasser (1999), Richards, Gallo, and Renandya (2001).

This type of approach is quite outdated and not aligned with the current wisdom in grammar pedagogy, which advocates a more eclectic approach (a) integrating form focused instruction (FFI) with tasks involving negotiation of meaning (CLT); (b) combining deductive and inductive learning and (c) aiming at the proceduralization (automatization) of grammar structures (Lightbown and Spada, 2008). The last point (c) is particular important in our view as it refers to the greatest shortcoming of much current grammar pedagogy; the failure, that is, to recognize that a grammar structure can only be considered as fully ‘learnt’ by a student when s/he can deploy it correctly in fluent oral/written speech and teach grammar accordingly.

A finding by Andrews (2003) is interesting in this regard as it may provide an explanation why some teachers may prefer a deductive rather an inductive approach to grammar teaching and alludes to a very common scenario in secondary British schools. Andrews found that teachers with high levels of explicit KAL feel more confident teaching grammar inductively than those with lower levels of it. The reason is obvious: deductive teaching allows the teacher total control over the pedagogic content of the grammar lesson; inductive teaching is less predictable, may require improvisation and the ability to answer students’ grammar queries on the spot.

  1. Do teachers have sufficient knowledge of their students’ grammar cognition to teach them effectively?

Research shows that the effectiveness of any instruction depends on the extent to which teacher intentions and learners’ expectations, wants and needs are aligned (Dornyei, 2005). A substantial lack of alignment often results in students’ disaffection and less effective learning. Hence, one facet of teacher competence is the ability to adapt one’s teaching to learner preferences and agendas.

Two large scale studies carried out by Schultz (1996, 2001) set out to investigate to what extent teachers’ and students’ beliefs about how languages are learnt matched.  In her 1996 study, Schultz studied teacher and student attitudes to grammar teaching and error correction. 94% of the 824 students she investigated stated they wanted to be corrected when they made errors in class. However, only 48 % of the 92 teachers on the study concurred with them. Schultz also identified ‘perturbing differences’ between teachers and students’ stances on this issue. The vast majority of the students believed that the formal study of the L2 grammar is essential to effectively master a language whereas only 64 % of the teachers concurred with this view.

In a subsequent study with Colombian students (Schultz, 2001) the gap between students and teachers’ perception was even greater. Whilst 76 % of the students said they valued grammar, only 30% of the teachers thought their students did.  Mismatches like these have been reported by numerous other studies (Cathcart & Olsen, 1976; Yorio, 1986; Spada, 1987; Wesche, 1981).

We would add that, in our experience, students’ parents’ own beliefs about language learning may play an important role, here, especially with infant and adolescent learners.

Another astonishing finding in his area comes from a study by Berry (1997), who asked ten teachers of English in Hong Kong to rate the knowledge of grammatical terminology (metalanguage) of 372 students they taught. The discrepancies were huge indicating that the teachers grossly overestimated their students’ knowledge of metalanguage. Berry stated that such discrepancies have the potential to cause serious problems in the classroom.

I personally found this to be an issue when I was lecturing at university and school with the language assistants I supervised. The language assistants – coming from foreign school systems with a solid tradition of formal instruction – did take for granted that their L1-English students would understand the metalanguage they used in their grammar explanations. This did put some students off.

  1. Do teachers’ beliefs and classroom practices match?

Several studies indicate that there is often a clear mismatch between what teachers says they do in the classroom when they teach grammar and what they actually do.

Andrews (2003) observed teachers whose stated beliefs were rooted in Communicative Language Teaching. He found very little evidence of CLT in their grammar teaching classroom practice.

Basturmken et al (2004) reports incongruences with regards to corrective practices whereby the teachers’ classroom behavior contradicted their stated belief that one should not interrupt students’ oral output during communicative activities to correct unless errors impede intelligibility. Basturmken and his co-workers put these incongruences down to the fact that when teachers talk about their beliefs they draw on their espoused theory and/or abstract technical knowledge of grammar pedagogy; however, when they respond to classroom contexts they draw on their practical knowledge. The two types of knowledge may not be necessarily aligned, especially in less experienced teachers (Borg, 2015).

Ng and Farrell (2003) and Farrel and Lim (2005) who found similar inconsistencies vis-à-vis error correction and other areas of grammar instruction, put them down to the fact that teachers often have to adapt their beliefs to curricular and student needs and goals. This, too, is a finding that strongly resonates with our experience and points to a common source of teacher frustration.

  1. Do teachers know how to teach grammar?

Research has identified that in many cases the problem is not so much lack of grammar knowledge but the teacher’s ability to transform that knowledge in pedagogic content; in other words, the ability to turn what they know in effective materials and lesson plans. This phenomenon seems to be partly caused by lack of knowledge of the adequate grammar terminology (Hislam and Cajkler, 2005). In our view, though, the most crucial factor is the absence of adequate training in this aspect of teaching which causes teachers to rely largely on the textbook or – as we discussed above – on their previous images of learning. Unfortunately, the kind of teacher training provided on PGCE or CELTA courses seems to fail in this respect (Borg, 2015).

  1. Concluding remarks

The research discussed above is by no means conclusive, exhaustive and consistent enough to grant any generalization of the findings discussed to the whole international language teaching population. However, the studies reviewed do identify important issues which very strongly resonate with our teaching experience (which amounts to almost 60 years between the two of us!) and have important implications for language teaching and learning.

Firstly, overall modern foreign language teachers – and ESL teachers to a lesser extent – seem to believe that grammar teaching is important and should be implemented in lessons. However, their approaches to grammar teaching appear to be rooted in very outdated views of grammar acquisition. This finding calls for much greater emphasis on professional development which models more eclectic and modern approaches to explicit grammar instruction of the likes of those recommended by Lightbown and Spada (2008), Swan (2006) and many of our blogs. The approaches envisaged would (1) integrate grammar and negotiation of meaning, (2) combine inductive and deductive teaching and (3) aim at automatization of target morphemes.

Such professional development, in order to be successful needs to involve extensive not merely intensive practice. Also, since schools’ micro-cultures seems to play an important role in re-shaping beliefs and approaches to grammar pedagogy, trusted and charismatic colleagues will play an important role in the process.

Schools administrators need to recognize that the process of re-shaping teacher beliefs and practices vis-à-vis grammar pedagogy is a very long and complex one which requires a lot of quality time and support that they must be prepared to allocate. Results cannot be expected at the end of a few weeks or even months’ training. It is likely to take much longer than that!

Another finding with huge implications for teacher training and professional development programs refers to the incongruences between teachers’ stated beliefs and actual classroom practice. Such incongruences may be due to various factors. The most important ones refer to (a) the dichotomy declarative vs procedural knowledge and (b) the inability to transform grammar knowledge into a lesson plan (in other words: the lack of a principled framework). Whereas (b) calls for CPD of the likes envisaged in the previous paragraphs, (a) calls for the need for teachers to elicit more non-judgemental student and peer feedback on their teaching practice and self-reflection supported by reflective journals, lesson-videoing, etc.

Finally, a worrying finding discussed above refers to (a) incongruences between teacher beliefs / goals and learner expectations / needs and (b) gross mismatches between teacher perceived levels of student grammar-related cognition and actual ones. The former issue can seriously hinder effective learning and it is imperative that teachers assess students’ beliefs about and attitudes to grammar to find the pedagogic fit that best suits their classroom. As for (b), an initial assessment of new students’ KAL may be desirable to better match their needs. This is rarely done in UK secondary school settings, in our experience.

In conclusion, the biggest lesson to be learnt from the research reviewed in this and in our previous post is that schools must place greater emphasis than it is currently done on professional development in the areas of grammar knowledge (KAL) and grammar pedagogy

-END –

PART 2 – Do teachers know grammar?

  1. The rise-fall-rise of Explicit Grammar Instruction

Explicit Grammar Instruction (henceforth EGI) was the primary mode of language instruction from the Romans to the first half of the 20th century. It fell out of favour in the 70s/80s with much of the international modern language teaching community with the advent of theories and methodologies based on the epistemological dogma that it does not significantly enhance L2 proficiency development. Such theories and methodologies, based as they were on experiential models of learning and on the attainment of communicative competence as the ultimate goal of language learning, inevitably marginalized the formal teaching of grammar. Nativist theories, such as Krashen’s, even advocated the total ban of EGI in the belief that a natural order of acquisition of L2 grammar structures exists which cannot be altered by formal instruction.

Recently, however, a substantive cohort of L2 educators has advocated that, whilst EGI, on its own, does not enhance acquisition, when integrated with approaches like CLT it may indeed be beneficial and a mounting body of empirical evidence seems to endorse this view (Ellis, 1990, Harley 1993, Ellis,2003). This has prompted a grammar revival that in the last two decades or so has started to creep into mainstream modern language education.

The effective integration of EGI into communicative language teaching or task-based learning is not without challenges. Steve and I believe that EGI can play an important role, however it (1) should not dominate modern language lessons; (2) should be carried out as part of a variable-focus curriculum concerned predominantly with the teaching of communicative functions and vocabulary; (3) should occur mainly in the context of interactional tasks aiming at developing fluency as well as grammar accuracy and syntactic complexity;  (4) should aim at creating procedural knowledge (as opposed to traditional approaches which focus mainly on declarative knowledge); and (5) should involve inductive learning as well as deductive approaches


The reader should bear in mind that the body of research we shall draw upon in our attempt to answer the above questions is by no means representative of the international teacher community. To generalize the findings of the studies reported above would be preposterous and unfair. However, the data that the studies we shall very concisely review below do yield very interesting findings which do resonate with our experience and do raise important issues which both governments and education providers must heed and address as part of their professional development programmes as they refer to important areas of teacher competence which, in our experience, are grossly neglected.

Please also note that for reasons of space we shall discuss only studies which we deem as representative of each research strand and topic.

3. Do teachers know the target language grammar?

Researchers refer to the knowledge of how language works as KAL (knowledge about language) or LA (language awareness). An important distinction must be drawn between Declarative and Procedural knowledge of the grammar of a language. Declarative knowledge refers to the explicit knowledge of the grammar rules, i.e. being able to articulate how a grammar rule works. Having the procedural knowledge of the grammar of a language means being able to use it in production, by-passing consciousness, so to speak (e.g. I can use the imperfect tense in French but I cannot explain why). Most native speakers of a language who have not been taught language explicitly, for instance, would possess procedural knowledge of their mother tongue but very little – if any – declarative knowledge of it.

As many language theorists and educators believe nowadays, there is a clear link between explicit knowledge of the formal aspects of language and performance when using that language. Hence, as Andews (2008: 1) puts it:

fostering learners’ ability to analyse and describe a language accurately is likely to help them become more effective users of that language. Arising from this is the belief that teachers of a language need an understanding of how that language works and an ability to analyze that language to function effectively as teachers.

It follows that it is paramount that teachers’ subject specific competence ought to include high levels of declarative knowledge as well as the ability to teach it effectively.

3.1 How much do university language students know?

Research investigating L2 teachers’ levels of KAL has yielded shocking results which raise serious concerns. First off, let us have a look at a set of UK-based studies which investigated how much metalinguistic knowledge modern languages university students know. Why should we be interested in this? Because (1) these students constitute the UK language learners elite, the pool from which language teachers usually come, hence, (2) studying them will tell us how much metalinguistic knowledge language teachers are exposed to at secondary school level.

Bloor (1986), Alderson et al (1997) and Alderson et al (2010) investigated the metalinguistic knowledge of 63 students enrolled on language courses in British universities. Bloor’s (1986) findings were the most dispiriting: most students failed to meet the Department of Education and Science target that 16-year-olds should be able to identify verb, noun pronoun, adjective, adverb, article, preposition and conjunction. The only grammatical terms they could identify were ‘verb’ and ‘noun’. Alderson et al (2009) replicated Bloor’s study to see if things had improved 23 years on. Their study confirmed Bloor’s (1986) dispiriting findings although they observed slight improvements in terms of the understanding of metalinguistic terms. They recommended that an increased focus on teaching the use of the terms rather than simply presenting them to students might ensure that they are able to fully understand them, rather than just being familiar with them.

3.2 How about teacher trainees?

Wray (1993) and Williamson and Hardman (1995) investigated the KAL of pre-service teachers at the start of their teacher training programme. Their results confirmed Bloor’s (1986). Some findings were astonishing. In Wray’s study, only 30% of the subjects could identify adverbs and; 23 % pronouns and less than 10% prepositions. Williamson and Hardman (1995) found that their informants scored only 5.6 out of 10 on a question requiring them to name parts of speech. They concluded that the 99 trainees they studies had serious gaps in knowledge about grammar, misconceptions about it and a lack of metalanguage for analyzing language use. Several other studies (e.g Chandler et al., 1988) concurred with these findings.

3.3 In-service teachers’ KAL

Andrews investigated practicing teachers’ levels of KAL in a few studies (e.e. 1994, 199a and 2005). His findings confirmed Bloor’s (1986) and Alderson’s (2009). Andrews (1999a), is particularly interesting because it compared the explicit knowledge of grammar  and grammatical terminology of 4 groups:

  1. NNS (non-native speakers) of English
  2. NNS prospective teachers of English
  3. NS (native speakers) of English with a background in English studies
  4. English NS trainee teachers of modern languages

Andrews found that on average the levels of grammar knowledge were utterly inadequate, although the NNS teachers of English did much better than the other groups. The lowest scores were obtained by the prospective teachers (group 4).

Mitchell et al (1994), in an interesting project in English secondary school settings which involved, amongst others, classroom observations of several Modern Language teachers found that generally, levels of KAL were inadequate and that

There was some evidence that the limits to teachers’ own linguistic knowledge were a constraint on the development of maximally effective KAL work. This could be seen even in some KAL focused units, which at times seemed to have conveyed inaccurate messages to pupils; more generally, teachers’ tendency to avoid technical vocabulary in KAL-related talk seemed linked at times to insecurity in using grammatical or discourse terminology.

Shuib (2009) set to investigate English language teachers’ nature and level of grammatical awareness. Questionnaire and interview techniques were used to elicit data from primary school teachers who were following their B. Ed TESOL programme in Universiti Sains Malaysia in 2006 and 2007. Her data confirmed previous findings. She concluded that in terms of training, her findings suggested that more efforts need to be made at teacher training institutions to promote grammatical awareness among aspiring teachers.

Other interesting findings come from studies by Grossman, Wilson and Shulman (1989) and Beard (1999), which demonstrated that teachers tend to avoid teaching grammar due to their uncertainty about their knowledge of grammar and inadequacy of grammatical knowledge. Beard (1999) noted that besides having much ‘intuitive implicit knowledge’ about grammar, the problem for many teachers is the inability to make the implicit knowledge explicit and to use the appropriate technical terms (metalanguage).

4. Concluding remarks

Although the studies just reviewed and many other investigations carried out all over the world do paint the same bleak picture, we know that there are many excellent practitioners who do have high levels of grammar knowledge. Thus, it is important to reiterate that these research findings, whilst spotting a worrying trend, cannot and should not be generalised to the whole international language teaching community. We have, however, the ethical imperative to heed these findings and, as Borg (2015) puts it,

on the assumption that an explicit understanding of language plays a major role in the effectiveness of the work of language teachers, these findings suggest the need for language teacher preparation programmes to dedicate substantial time to the development of trainees’s declarative knowledge about the language.

We would add that practising teachers should not be afraid to recognize KAL as an area of their subject specific competence requiring development. After all, as Chandler et al.’s (1988) informants stated, most teachers’ KAL was acquired during their school days. A deficit that stems from the way an educational system is run should not be viewed as something to be ashamed of. Hence, senior teachers/professional tutors/line-managers should be encouraged to address any observed gaps in their colleagues’ KAL through professional development strategies in a non-judgemental way.

You can find more on this topic in the book ‘The language teacher toolkit’ I co-authored with Steve Smith and available for purchase at