by Daniel Tse
A good text is not simply a series of statements placed end to end; these statements, as sentences or clauses, must fit together so that the text as a whole stands united. We call the glue that holds a text together cohesive devices. The difference between a well-written and a poorly-written text often comes down to how cohesive that text is. How can teachers develop their students’ ability to use cohesive devices in writing? Before we explore several practical ideas, let us turn our attention to five different types of these devices.
Cohesive devices
In my article in Issue 51, we explored the reference system as one type of cohesive device. The other four that contribute to cohesion are substitution, ellipsis, conjunction, and lexical cohesion (Halliday and Hasan 1976). Although these devices are found to have distinct characteristics in their cohesive roles, the use of one type of device is by no means mutually exclusive to another.
Substitution
Akin to anaphoric reference, substitution involves the replacement of words or phrases in a preceding sentence by the pronoun one(s), the auxiliary verb do, or the adverbs so or not (Halliday and Hasan 1976: 91, 141). The example below illustrates how substitution is used to avoid unnecessary repetition of language:
There are many reasons why second language learners often struggle with writing. The most important one is their lack of knowledge of genre. Learners can increase their awareness of the stylistic features of different genres by studying sample texts. When learners do that, they should pay attention to the register, the frequently occurring grammatical structures or lexical chunks, and the way in which various ideas are organised in a text.
(Own example)
The pronoun one in the second sentence is a nominal substitute for the head noun reasons in the first; the verbal substitute do replaces the action of studying in the previous sentence.
Despite the similarity between substitution and anaphoric reference in their formal process, there is a fundamental difference in the amount of information carried over from the replaced item. In the above example, both the reference item they and the referent learners denote the exact group of people. The nominal substitute one, however, represents an incomplete whole; it does not include all reasons suggested by the preceding sentence. It is this kind of subtle difference which sets substitution apart from reference.
Ellipsis
Coursebook writers usually combine ellipsis and substitution due to their similar formal process of replacing words. Instead of using a substitute to avoid unnecessary repetition, ellipsis involves the omission of information that would otherwise be repeated. Halliday and Hasan, therefore, regard this as ‘substitution by zero’ (1976: 142).
Writing in the classroom should be treated as a continuous cycle of drafting and revising. In fact, it is already [*] in most one-to-one lessons.
(Own example)
In the above example, the lexical verb treated and its adverbial complement are omitted in the second sentence, where the asterisk [*] is placed. Nevertheless, full comprehension can still be achieved without overt repetition in this specific context.
Conjunction
Halliday and Hasan (1976) discuss conjunction in terms of how ideas relate to each other across clauses and sentences. Four broad types of conjunctive relationship are identified: additive (similar ideas), adversative (opposite or contrasting ideas), causal (cause and effect), and temporal (sequence of actions or events) (ibid.: 238-267).
These relationships are expressed linguistically by various linkers or linking words/phrases, some of which are listed below:
Additive: also, and, furthermore, moreover, etc.;
Adversative: although, but, despite, however, etc.;
Causal: hence, so, therefore, thus, etc.;
Temporal: after that, hitherto, meanwhile, then, etc.
(ibid.)
Lexical cohesion
Lexical cohesion seldom constitutes a separate topic in coursebooks as it operates in a less systematic manner than the grammatical cohesive devices. Reiteration and collocation were thought to be the major sources of cohesion in the 1970s; however, there have been more recent revisions to the classification of these two subcategories (Flowerdew 2014: 41).
Reiteration refers to the repetition of lexical items in the preceding sentence. This can be done by plain repetition of the same item or replacement by a synonym, near-synonym or general noun.
Teachers should avoid marking their students’ work with a red pen. This colour is traditionally associated with excessive error correction and negative judgements
(Own example)
In the above example, colour is a general noun as well as the superordinate of red. The cohesive elements, moreover, include collocations of varying levels of strength. For instance, the verb marking regularly co-occurs with the noun work; this is a strong verb-noun collocation in context. A looser type of collocation is the chain of lexical items throughout the example: teachers, marking, students’, work, red pen, error corrections, and judgements. When words or phrases related to the same topic appear together as a ‘chain’, they contribute to cohesion in a text.
Writing tasks
To encourage learners to use cohesion in writing, teachers can use a range of lesson activities to target the different types of cohesive devices mentioned above.
Sentence reconstruction
In sentence reconstruction tasks, learners rewrite the given sentences by changing different sentence components while maintaining the meaning of the sentence. An example of task items is provided below:
Rewrite these sentences. Use a pronoun for the underlined phrases.
- I’m selling my computer. I’ve got no space for my computer in my bedroom.
- …
(Adapted from Compact Key for Schools, 2nd ed., p. 17)
In this task item, redundant language is deliberately placed in the second sentence. To make it read more cohesively across the two sentences, learners should replace the second instance of my computer with the appropriate reference item, it.
Generating tasks like this might be easier than it appears. Any well-written text that avoids redundant language can be re-written to put the redundancies back in.
The same task format can also be used with substitution and ellipsis. As these involve complex processes of replacing sentence components, learners can first attempt the task individually and experiment with using the two types of cohesive devices. In the subsequent pair-check, they focus on the new sentences and discuss the extent to which information can be omitted or replaced. From this post-task discussion, learners can finally infer the major grammatical patterns governing substitution and ellipsis.
Furthermore, teachers can use sentence reconstruction tasks to encourage learners to use conjunction in writing. In this context, a possible task instruction is given below:
Rewrite these sentences. Link each pair of sentences with a word or phrase.
1. I’m selling my computer. I’ve got no space for it in my bedroom.
2. …
(ibid.)
Depending on where the emphasis lies in the above sentences, learners may join the two with because (emphasis on the reason) or the more formal as/since (emphasis on the result); it is not necessary to use a comma in the new sentence. In the post-task feedback, therefore, teachers should clarify not only the differences between apparently synonymous linkers but also the use of punctuation. In my experience, certain types of punctuation errors, such as the use or non-use of commas, tend to persist in the learners’ writing due to their lack of attention - or because of L1 interference, as the rules of punctuation are not universal across languages.
Gap-fill
Gap-fills may have the status of a cliché in language classrooms, but they are a useful tool for focussing learners’ attention on cohesive devices. This task type is highly versatile; it can be used with every single type of device mentioned above. An example of task items is provided below:
Write a pronoun in each gap.
I’m selling my computer. I’ve got no space for ______ in my bedroom.
(ibid.)
When used with reference, substitution, and ellipsis, gap fills are an alternative to sentence reconstruction tasks. In the context of conjunction, however, gap-fills are less challenging, or more controlled, as the position of the linker in a sentence is already indicated for learners.
In the above example, the gap can be filled by a range of words appropriate to the context, including (this) device, gadget, thing, or even junk. Learners can, therefore, use gap-fills as a springboard for exploring various synonyms or superordinate items. McGee (2009: 215-216), however, cautions us against using synonyms indiscriminately in reiteration. While the words computer and workstation are synonymous, the latter carries a different referential meaning and thus does not fit the context of my bedroom.
Nevertheless, I usually favour a diagnostic approach to exploring lexical cohesion with more advanced students. Rather than use prescribed sentences in any random context, I collect samples from my students’ written work for the gap-fill tasks. This ensures that pair or group discussion of task sentences occurs in a meaningful environment. This approach has been found to be effective in targeting learner issues with vocabulary in writing.
Sentence continuation
In sentence continuation tasks, learners use different types of conjunction to extend the given clauses or sentences.
Continue these sentences. Add ideas with a linking word or phrase.
- I’m selling my computer…
- …
(ibid.)
In this task, learners work collaboratively in groups and continue the sentences with their free choice of linkers (for example, “I’m selling my computer because there’s a game I want to play and I need a better graphics card”). In the post-task feedback, they compare their sentences with other groups and, as a whole class, vote for the most creative ones. In my experience, not only does this task arrangement foster creativity in the learners, but the freedom of choice has also been shown to engage them in their learning. It can lighten the classroom atmosphere by making writing a less daunting skill for the learners. It also carries the benefit of showing how writing can be improved without requiring the students to write two-hundred-word texts - the usefulness of taking a microscope to their use of cohesive devices is enormous.
Conclusion
By developing learners’ ability to use different types of cohesion in writing, teachers can help them to become more fluent, more confident writers in a second language. This will enable the learners to produce cohesive texts in writing exams so that they have better chances of achieving their desired result. In the workplace, it will also enhance adult learners’ effectiveness in their written communication with their colleagues and clients.
Bibliography
Flowerdew, J. Discourse in English Language Education. New York: Routledge (2013).
Halliday, M.A.K. and Hasan, R. Cohesion in English. London: Longman (1976).
Heyderman, E. and White, S. Compact Key for Schools, 2nd ed. Cambridge: Cambridge University Press (2019).
McGee, I. ‘Traversing the lexical cohesion minefield’ in ELT Journal Vol. 63/3. Oxford: Oxford University Press (2009). 212-220.
Author Biography
Daniel Tse works as a teacher and examiner in Italy. He has taught EFL at IH Milan and San Donato since 2019. He works with Young Learners, teens, and adults across the full range of CEFR levels. He mainly teaches Cambridge/IELTS Exam Preparation and Business English at IH, as well as an increasing number of English ‘top-up’ courses in Italian schools and universities. An early-career teacher, he is currently on his journey through the DELTA. In addition to teaching, Daniel runs teacher development workshops for Macmillan/Mondadori Education in Italy. He has spoken at local conferences in Milan and Barcelona, the IATEFL Conference in the UK, and British Council TeachingEnglish webinar series as a panellist.