“Grammaring with a twist” to suit the EAP classroom

by Deb Catavello

If you, like me, learnt English, or any other language in fact, using Grammar Translation (a method which now generally enjoys a bad rap) at a time when communicative language teaching hadn’t perhaps quite taken off, this is the definition of grammar you will probably be most familiar with:

“grammar (uncountable) is the rules that describe language, and a grammar (countable) is a book containing these descriptions” (Thornbury, 2001, p. 2).

Thornbury (2001) challenges this view and has argued for quite some time that, alongside the noun, there should be a verb to grammar. In claiming this, he wants to encourage teachers to combine their conceptualisation of grammar as a product with one that emphasises the process (i.e., the grammaring).

According to this approach, grammar is seen as “emerging”, and teachers are invited to reverse the traditional model by starting with lexical items and then moving on to grammar. Thornbury (2020) justifies this approach by drawing on first language (L1) acquisition research which suggests that children first learn words and then gradually add grammatical elements as their speech develops.

Another reason why grammaring might be a better alternative to the traditional model is that the latter leaves little room for creativity. Thornbury (2020) recalls his experience of teaching in Egypt in the mid-1970s where his students were growing increasingly frustrated with the heavily grammar-driven syllabus which they perceived as too slow and rigid. These, he explains, were students who didn’t have issues with comprehension (they watched films in English etc.) and were good at language learning. Although their language level was lower, these students didn’t sound that dissimilar to our IFP+ students at CALD …

… which takes me to my recent attempt to design grammaring activities that would suit an EAP context. Below, I will discuss how I adapted Thornbury’s grammaring technique to ensure the inclusion of a language focus into a course which adopts a “reactive” approach to the teaching of language.

Halfway through the first teaching block, I designed the activity in Table 1 (below). In the Accelerated Academic Language and Literacy (AALL) unit, students were working on the skill of synthesising ideas from different sources and had read four texts on the topic of Artificial Intelligence in education. By this point in the term, my students had encountered one other grammaring activity and thus knew how to go about completing this kind of task (e.g., that they could change the order of the sentences given, that grouping sentences together is a good way to start, etc.)

Table 1: Activity 1 paragraph writing

1.       In groups, combine the 10 sentences below into one body paragraph (to include in your argumentative essay on the opportunities and threats of AI in education). To do so, you need to: add cohesive devices. make the style more academic/formal (so you need to change the language). replace ‘Text 1’, ‘Text 3’ etc. with in-text citations (use Harvard). use evaluative language to show your stance. Feel free to add greater detail using the notes you took on the texts. AI could change the way students are assessed. AI could lead to academic misconduct. Text 2 says AI should be banned in exams. Text 1 says AI should not be banned. Text 1 says banning AI will increase its usage. Text 1 says that because of AI some assessments won’t work anymore. Text 1 says that because of AI some assessments need to change. Text 1 says that the new assessments should focus on criticality and analysis. Text 4 says AI could make learning more personalised. Text 4 says exam questions could be adapted to the student’s level if we used AI.

 

The first activity (Table 1) presented the students with a version of a potential synthesis that included minimal grammar. In original grammaring activities of the kind designed by Thornbury, the de-constructed text would often be completely “de-grammared” and so feature lexical items exclusively. This is thus the first “twist” to the original approach, but I will come back to this point later in this post.

The second activity (Table 2) was meant to help students practise the skill of giving feedback while also raising awareness of the different language options available to a writer. This stage of the lesson was important to ensure that students moved away from a prescriptive view of grammar as “rules” to abide by towards a perspective that sees language as “choices” made by the writer in accordance with their purpose or “voice”.

Table 2: Peer-feedback stage

2.       Look at another group’s paragraph and give them feedback (use the bullet points in Task 1).

 

Similarly, the last activity (Table 3) aimed to raise students’ awareness of some key aspects of academic writing.

Table 3: Comparison with sample paragraph

3.       Compare your paragraph with the sample below. Highlight cohesive devices, citations, formal/academic language, the writer’s voice. Sample (synthesis) – 162 words AI has the potential to significantly impact assessments in Higher Education. On the one hand, concerns have been raised over the risk of artificial intelligence (AI) to lead to academic misconduct, such as plagiarism, and this is why some advocate for a total ban of ChatGPT for homework assignments and final exams (Terwiesch, 2023). However, Lim et al. (2023) maintain that this is not an effective approach and point out that banning AI might actually increase its usage. They take a perhaps more positive approach by acknowledging the potential of AI to make some of the more traditional assignment types, such as essays, redundant and thus force educators to devise allegedly more rigorous ways to assess students’ critical and analytical skills. A further opportunity AI might provide is the ability to ensure students have a more personalised learning experience. In this regard, Southworth et al. (2023) claim AI could improve the accuracy of individual assessments by allowing exam questions to be tailored to the student’s level.

Overall, I feel these materials worked well in my EAP classroom as they did lead to a much-needed language focus which included, among other things, input on:

• Reporting verbs and referencing
• Cohesion
• Paragraph structure
• Style (avoiding repetition and vague language, hedging)

 

More specifically, I think Activity 1 lent itself well to the teaching of the pattern “source+ reporting verb + claim” (e.g., Lim et al. (2023) claim that …). “Facilitating pattern detection” is an important element of language teaching (Thornbury, 2001, p. 65) and is behind my persistent use of “Text # says” in Activity 1. By doing this, I was trying to provide the conditions for this reporting pattern to emerge.

This activity also helped me, the teacher, identify language points that needed reviewing. For example, although cohesive devices were something we had discussed only a couple of weeks before, some of the paragraphs my students produced included problematic instances of reference words as shown in Table 4, where it is unclear what some of the “it”s (in bold, Table 4) are supposed to refer to. Another language point we ended up revising was hedging language, as some of the claims that the students presented did not include an appropriate amount of caution (in bold, Table 5).

Table 4: Student sample

The role of AI, in the realm of education, has numerous different perspectives and it could change the way students are being assessed. On one hand, Terwiesch (2023) voiced that there are concerns it could pave the way for academic misconduct, advocating for it to be banned in exams. However, the opposing viewpoints by Lim et al. (2023), presents an alternative perspective, suggesting for AI not being banned, as it may inadvertently increase its usage. According to them, because of AI, it could consequently necessitate a shift towards new assessment methods. Instead, it should focus on criticality and analysis. Adding on, Southworth et al. (2023), introduces the idea that AI could enhance personalization in learning and adapt exam questions to individual students’ levels. This multifaceted topic underscores the complexity of AI in education, encompassing its challenges, and the need for thoughtful implementation.

 

Table 5: Student sample

The use of AI in education could cause long-term effects such as changing how students are assessed. While some believe that AI can increase academic misconduct rates and should be prohibited during the exam process (Terwiesch, 2023), others argue that AI should not be banned and that banning the technology could result in More students using it (Lim et al., 2023). Nevertheless, it is almost unanimously agreed amongst researchers that AI could nullify the effectiveness of certain exam styles (Lim et al., 2023) and that adjusting the traditional methods of examination could soon be necessary to accommodate different learning capabilities (Southworth et al., 2023). To assess students thoroughly and accurately in the age of AI, exams will need to focus on criticality and analysis a lot more than they may have in the past (Lim et al., 2023).

 

Overall, the most valuable benefit that grammaring offers is its ability to incorporate language work without compromising on a communicative approach to teaching and learning. An added benefit is that not much preparation is needed on the part of the teachers who, as Thornbury (2001, p. vi) puts it, should “make discourse [their] resource: nothing more is needed in order to free grammar”. It didn’t take me long to design Activity 1 since the students and I had been working on those four texts for quite a while.

However, there’s something I hope to do differently next time. Trusting students is seen as a premise on which grammaring tasks are based (Thornbury, 2020). In this respect, my only regret is perhaps not having trusted my students sufficiently; this lack of trust explains my choice to have the sentences in Table 1 as not completely “de-grammared”. Thornbury (2020) points out that grammar compensates for the lack of context, so the more context we have, the less grammar we need (Fig. 1). In this sense, I suppose I did not trust that my students and I shared enough knowledge/context and so felt compelled to include some grammar (Table 1).

While writing this blog post, I worked on a different version of this activity where I provide students with lexical items only and leave out as much grammar as possible (Table 6). Is this one step too far? I look forward to testing this adaptation next year…

Table 6: Latest version of Activity 1

AI      change     assessments     misconduct     ban Usage    criticality      analysis       personalisation      student’s level     questions Terwiesch (2023)    Lim et al. (2023)   Southworth et al. (2023)

 

References

Lim, W.M., Gunasekara, A., Pallant, J.L., Pallant, J.I. and Pechenkina, E., 2023. Generative AI and the future of education: Ragnarök or reformation? A paradoxical perspective from management educators. The International Journal of Management Education, 21(2), p.100790.

Southworth, J., Migliaccio, K., Glover, J., Glover, J., Reed, D., McCarty, C., Brendemuhl, J. and Thomas, A., 2023. Developing a model for AI Across the curriculum: Transforming the higher education landscape via innovation in AI literacy. Computers and Education: Artificial Intelligence, 4, p.100127.

Terwiesch, C., 2023. Would chat GPT3 get a Wharton MBA. A prediction based on its performance in the operations management course. Wharton: Mack Institute for Innovation Management/University of Pennsylvania/School Wharton.

Thornbury, S. (2020) ‘Grammaring’ activities by Scott Thornbury video master class. Available at: https://www.youtube.com/watch?v=HWWE1Jt4OGE (Accessed 08 January 2024).

Thornbury, S. (2001) Uncovering grammar. Edited by A. Underhill. Oxford: Macmillan Heinemann English Language Teaching.