Lexical Phrases and Patterns
Task 5 Embarrassing incidents
As Table 10 shows, the past perfect tense was used by nearly half the speakers. This might seem to be quite a high frequency, and that the past perfect is, therefore, well worth practising with students. Conversely, however, there is a danger in giving students lots of practice with a structure which they then think is very common, but which, in fact, was used by only a minority of participants in this study.
Table 10 Language predicted by native speakers for task 5
Structure No. of Occs % Example in recorded data prdtns in Spkrs
Max 20 data
PAST PERFECT 1 41 44 … some clown had sewn the
ticket half way.
SIMPLE PAST 17 500⫹ 100 … I got in quickly, jumped in, slammed the …
PAST CONTINUOUS 3 56 64 I was coming from
a meeting in Hong Kong.
Reflection and conclusion
The discovery that a great many predicted language items did not appear in the recorded data calls into question the appropriacy of expecting learners to use language items (like should) in contexts where fluent speakers would not naturally use them. Interactions like the fol-lowing are very unlikely to happen in the English speaking world out-side the classroom.
A: I like scuba-diving.
B: You should visit the Great Barrier Reef in Australia – it’s amazing!
I like shopping.
A: Oh, you should go to the new shopping centre.
If learners feel it necessary to use should all the time (for example at the Production stage of a PPP cycle where should has been presented), they are confined to one wording and are missing out on experimenting with other ways of expressing a whole range of similar meanings. Learners may wish to express their opinion less forcefully than should suggests, so phrases like I would say or I would recommend or Well, what you could do is would be much more appropriate. In a PPP lesson learners are being unnaturally constrained when they should be experiencing the richness of meaning potential and practising normal conversation skills.
These findings add weight to the case for a task-based framework.
While doing a speaking task in such a framework, students (like the speakers who took part in this investigation) are free to use whatever language items they like, unhindered by any perceived pressure from the teacher to use certain structures or phrases. The good sense in giv-ing students this freedom is highlighted by the fact that, for each task, the speakers in the study employed a wide variety of language items.
After being given this important freedom of approach to a task, learners would then benefit from a focus on form (see Doughty and Williams, Can we Predict Language Items for Open Tasks 179
Table 11 Language not predicted by native speakers for task 5
phrase Occs % Example in recorded data
in Spkrs data
subj⫹ didn’t know ⫹ WH- 7 14 I didn’t quite know where to put my face.
1998 for the importance of this) and from being made aware of a wide variety of useful items. This form focus could be implemented post-task, either when planning a public report of the task findings to present to the class (Willis, J. 1996a, 1996b) or through language focus activities, like those in Nunan’s Atlas series (1995a).
Further suggestions for form-focused language study
A careful read through the task transcripts in Appendix 2 allows a great many foci for language to study to be identified, in addition to those listed in the tables earlier. Here are a few examples of consciousness-raising activities, some of which focus on functions, some on lexical phrases and some on grammar.
For Task 1, ask learners to identify ways of making suggestions, ways of giving reasons and to find phrases which evaluate suggestions. Focus on the word would and ask learners to find six phrases with would, clas-sify them and practice saying them. (Ask if any of the woulds appear with an if clause.) Ask learners to identify clauses beginning with which and where (there are four) and then write sentences on the same pattern about places they would recommend.
For Task 2, ask learners to find expressions giving advice, and then to classify them along a cline of quite tentative to more definite. They could also identify examples of vague language, discourse related phrases (eg first thing, in other words), interpersonal phrases like You know?, sentences or clauses with the word If and patterns with verbs⫹ing.
You could collect further examples of these particular features from other texts or transcripts that your learners have met earlier. These texts and transcripts become a collection of useful language data that Willis (2003: 163) calls a pedagogic corpus:
Grammarians and lexicographers work with a corpus of language, a set of texts, to enable them to describe the grammar and the vocabulary of the language. In the same way, learners process a set of texts to enable them to develop their own vocabulary and to work out their own grammar of the language.
Students can constantly refer back to familiar texts and investigate whatever grammatical or lexical features are highlighted for language study in their course book. If you have to cover a structural syllabus, and have a large enough pedagogic corpus, you will probably find that all the useful items are illustrated somewhere, and there will be much more 180 Teachers Exploring Tasks in English Language Teaching
useful language besides – for example lexical phrases and chunks as identified by others in this volume. This is one way that tasks can be used as the basis for a syllabus. A course can be sequenced according to topics, each with tasks and texts, which can be used alongside a check-list of grammatical and lexical items. These can be ticked off each time they have been focused on, recycled and tested. There is no guarantee that they will have been learnt, but at least it offers some accountabil-ity, and a far richer diet of language.
For teachers preparing students for exams with an emphasis on gram-mar, this approach of exploiting a pedagogic corpus for samples of tar-get language features can be supplemented with examination style exercises, eg cloze tests and multiple-choice items, using natural exam-ples from the texts themselves. As Sheehan implies (Chapter 4) students working with natural language data learn how to learn through becom-ing text investigators. Another advantage is that students will be exposed to a far wider range of vocabulary, collocations and lexical phrases – also useful for examinations.
If you are in a similar situation to Loumpourdi (Chapter 2) obliged to teach a grammar course following a set syllabus, then you may well need to restrict yourself to using ‘closed’ tasks. Although the language in closed tasks tends to be more predictable, there will always be alternative ways of expressing similar meanings. Broadly speaking, as the present study suggests, some tasks will almost certainly generate certain features:
for example will was used by 98 per cent of speakers to predict the future, modal verb phrases were used by 100 per cent of speakers to suggest what someone worried about a family member should do, and all speakers used the simple past forms to talk about past habits and embarrassing incidents. But just as anecdote tellers switch from simple past to what is known as the dramatic present, alternative forms will also appear, like, in the prediction task I can see beer being very very expensive indeed.
I have included in Appendix 2 some sample transcripts for the tasks I used in my study, so you could try out some of these tasks with your students. After doing each task you could encourage your students to analyse the discussion that the speakers produced identifying the kinds of features listed above. The students could practise and make a note of any phrases or patterns they think they will find useful in their own spoken English.
If you need to focus on other items on your syllabus you could design some closed tasks, record some friends doing them, then find out whether they used the language items in question. If they did, great! If not, you might still decide to use the task if it generates other items that Can we Predict Language Items for Open Tasks 181
your students need to know, and alternative ways of expressing those general meanings. For advice on making task recordings see Hobbs, Chapter 12.
Finally, any data that you collect by recording your own tasks can be used not only for teaching but also for further research into different areas. And, of course, it can be used by other people. Ketko (2000), for example, used the transcripts of my recordings to analyse multi-word chunks in spoken data which she then compared with data from her learners doing the same tasks, in a process similar to that used by Baigent in Chapter 13. As Pinter (Chapter 10) commented, once the data is assembled, all kinds of avenues open themselves up for exploration.