A web site for the developing language teacher

Discourse in Writing
by Emma Worrall
- 3

Some possible solutions to the problems

So, how can we address the over-use and misuse of linkers in our students' work? Crewe (1990: 321) outlines three pedagogical approaches which represent what he describes as "three stages of awareness that connectives have a textual meaning and are not just surface-level fillers". He refers to these three approaches as 'reductionist', 'expansionist' and 'deductionist'.

The reductionist approach suggests that students should be presented with a small subset of a long list of linkers and over time students should become more aware of their semantic and discoursal value.. A shorter list would have the advantage of "allowing the contrasts between the connectives to be more easily stressed" (Crewe 1990: 322). Crewe develops this further with his table of conjunctive relations (which he bases on Halliday and Hassan's table, 1976 in Crewe, 1990). He describes three problematic areas of conceptual categories: "Additive", "Adversative" and "Causal" and within those areas there are fourteen logical linkers. He suggests that advanced ESL students would be able to master these before progressing to more abstract ones (see appendix 6)

The expansionist approach encourages "Explicit markers" (Crewe 1990: 322) which are expressions which explicitly state the connection with either the preceding or following textual matter. Most of these expressions would contain the reference word 'this'. Some examples that Crewe lists are 'because of this', 'for this purpose', as a result of this'.

The criticism of the above two approaches is that they work backwards from lexis to discourse. Rather than concentrating on the discourse first they only consider the methods of controlling output. As they focus at the level of lexis the logical development of the argument or discussion is taken for granted.

To try to remedy this defect, Crewe presents a third approach: The 'deductionist' approach "begins by obliging the students before the writing process (within the essay plan, for example) to state the connection between the stages in the argument" (Crewe 1990: 323). If we do this first we can use lexical selections from a more complex list where, rather than having opaque terms like 'additive' and 'adversative' as the category titles, the sections could be labelled with discourse questions. Crewe (1990: 323) suggests some of the following:

  • Does your next section add another similar point to the argument?
    If so, is it of the same importance or of greater importance?
    Same? Use 'also', 'in addition', or 'besides'. Greater? Use 'moreover', or 'furthermore'.
  • Does your next section add an opposing point to the argument?
    If so, are both points valid or does the second one cancel out the first?
    Both valid? Use 'but', 'however', 'nevertheless' or 'on the other hand'.
    The second cancels out the first? Use 'on the contrary'.

Crewe says that the "schema" above will ultimately need to contain a full range of discourse "moves" (1990: 323), for example, listing, comparing, exemplifying, showing consequence, rephrasing, and concluding.

A further suggestion Crewe makes for teachers can help students by encouraging them to avoid using linkers in their first writing drafts and asking other students to analyse the logical progression of their argument before deciding on the suitability of possible linkers. Although this method counters the idea of looking at discourse first and then selecting the supporting words after, it is probably better than accepting students' work which has not been assessed for any logicality at all.


Crewe concludes that students need to be aware of the problems of using linkers in their writing and, therefore, they need more rigorous training which would oblige the students to think through their argument first and then decide how they might reinforce it with linkers. I would like to pursue some of the ideas that Crewe mentions in further FCE writing based classes. For example, doing a letter or essay without any linking devices and asking the students, in groups, to evaluate their work together, deciding on the types of discourse markers they should use. I also want to encourage the students to create a personal 'database' of linking devices that we use in the class so that they may refer to them in future classes and when revising for the exam. In general, I would like to do more group writing activities, whether it be peer correction or collaborative writing to give students more creative space and confidence in their work.


A. Battersby Instant Grammar Lessons, 1996, Language Teaching
Cohen et al Reading English for Specialized Purposes: Discourse Analysis and the Use of Student Informants, 1979, TESOL Quarterly13/4
W.J. Crewe ELT Journal, Vol 44/4 Oct. 1990, OUP.
J. Hadfield Classroom Dynamics, 1992, OUP.
M. McCarthy Discourse Analysis For Language Teachers, 1991, CUP.
D. Nunan Language Teaching Methodology, 1991, Prentice Hall International.
S. Thornbury About Language, 1997, CUP.
R. White & V. Arndt Process Writing, 1991, Longman Group.

To the beginning of the article

To the lesson plan

To the print friendly version

To the articles index

Back to the top

Tips & Newsletter Sign up —  Current Tip —  Past Tips 
Train with us Online Development Courses    Lesson Plan Index
 Phonology — Articles Books  LinksContact
Advertising — Web Hosting — Front page

Copyright 2000-2016© Developing