A common misconception among learners (and teachers) is that if they knew in advance the questions to be asked during a speaking task, they would be able to perform better. This is probably why most learners feel that the easiest part of most speaking tests is the first one, in which they are usually asked questions about themselves and their immediate environment, i.e. questions that they can easily anticipate and prepare answers for. In actual fact, though, as I will attempt to show in this article, prepared answers do not do the learner justice in a speaking test; on the contrary, having prepared an answer can easily lead to the learner actually failing the assessment.
By George Vassilakis
First published in ELT NEWS // October 2015
Consider this excerpt from the “interview” part of a spoken test at B2 level:
How much time do you spend with other members of your family?
Yes. Whereas I must say that I don’t have very much leisure time, as, in addition to going to school every day, I also attend English lessons and German lessons at a private language school in the evening, not to mention that I also keep busy by going to the gym twice a week because it is important to keep fit as well as following a healthy diet, I do occasionally have some time to myself, which I really enjoy spending with my family. It is a nuclear family with four members: two parents and two offspring. On week days it is only late in the evening that my family and I get the chance to spend a bit of time together, usually over dinner, but at the weekend we usually spend the whole day together, talking about what happened during the week, exchanging ideas and opinions and sometimes going out for dinner or to the cinema, because I like the cinema very much and I especially enjoy watching horror films and romantic films, for example Twilight, which is a combination of horror, because there are vampires in it, and also romance, because there is a love story.
What’s good about the learner’s answer
Most teachers would agree that the learner’s spoken production in this excerpt is unnatural; however, quite a few teachers also seem to believe that the learner has demonstrated that her level of spoken English is extremely high: she has, they would argue, made almost no grammar or vocabulary mistakes at all, she has used a broad range of structures, she has produced vocabulary that is quite impressive for her level, she has used a number of linking devices to connect the ideas that she talks about.
What’s not so good about the learner’s answer
On the other hand, if we examine what the learner has actually said and how she has responded to the question, it will be obvious that as a communicator, the learner is not particularly successful:
§ Her answer is unexpectedly long and can result in listener irritation
§ A lot of what she says is irrelevant to the questions she has been asked
§ The information that is pertinent to the question is quite limited
In addition, the language used by this learner does not really sound like spoken language. None of the following characteristics of conversational spoken English are present in this learner’s production:
§ Coordination
§ Simple sentence forms
§ Short chunks
§ Repetition
§ Redundancy
§ Prefabricated fillers
In short, the learner manages to demonstrate that she knows certain structures and vocabulary items in the sense that she can construct sentences incorporating them accurately, but she does not manage to use the grammar and vocabulary she knows in order to communicate a message.
How and why did the learner reach this point?
I am aware some teachers would argue that it is impossible for a learner at this level to produce, under exam conditions, such complex sentences and paragraphs in spoken interaction. I am afraid, though, that in reality quite a few learners do, in fact, produce language in the “personal interview” part of many speaking tests that sounds a lot like the excerpt above. This is because, by necessity, the kind of questions that will be asked in this part of the test are quite predictable: the learners obviously will be asked to talk about themselves, their interests, their families, and so on. And what a lot of teachers do is that they get learners to rehearse what they might say in answer to such questions, or even to write down and memorise short (or not so short!) speeches that they might make that are as packed with ‘impressive’ vocabulary and structures as possible. Some teachers even go to the extreme of giving the learners model answers to memorise and making sure that chunks from the model answers are incorporated in every conversation the learners might have. The obvious goal is, of course, to impress the examiner and, thus, earn good marks.
Will the examiners be impressed?
Examiners are not impressed so easily, though. To be more precise, examiners are not supposed to be impressed, they are supposed to mark the learners’ performance on the basis of very specific criteria and they need to find evidence in the learner’s spoken production that these criteria are actually met. In the City & Guilds Spoken ESOL Communicator Exam, for example, in addition to grammatical and lexical accuracy, which is an area where the ‘memorised paragraph’ approach might appear to be helpful by at least ensuring that there are no (serious) errors, learners are also assessed for their range, pronunciation, fluency and interaction skills.
Under the range criterion, examiners are likely to acknowledge that the ‘memorised paragraph’ learner does use a variety of structures and lexis; at the same time, however, the examiners will notice that the level of formality adopted is not appropriate to the circumstances and that many features of spoken discourse are actually entirely absent. It is, therefore, highly unlikely that the learner will be judged to have met this criterion.
Similarly, the pronunciation criterion is unlikely to have been met if the learner has produced a ‘prepared monologue,’ as the suprasegmental features of her production (rhythm, stress and intonation) tend to suffer when the focus is on remembering and reproducing stretches of language rather than on expressing meaning:.
As far as fluency is concerned, it might be claimed that because the learner has memorised whole sentences and paragraphs, she is unlikely to hesitate to search for patterns and expressions. In reality, though, the problem with reproduction of memorised chunks of language is that when hesitation does occur, it is invariably at the wrong point, as the learner is trying to either remember the exact phrase or to “fit” into her monologue a sentence that she considers worthy of inclusion on account of its complexity! This creates the impression of acute dysfluency, as the pauses and hesitations are unnatural and therefore more prominent.
Finally, with regard to discourse management and interaction skills, it is clear that the ‘memorised paragraph’ learner fails quite abysmally: no attention is paid to turn taking norms as the learner is eager to reproduce as much of the memorised chunk as possible, the learner’s contribution is of questionable relevance, the extent of their turn is inappropriately long, and, in short, the listener’s patience is tried!
The paradoxical situation that this learner finds herself in, then, is that she has spent a lot of time memorizing countless sentences and paragraphs to use in her speaking test and she has managed to use quite a few of those sentences and paragraphs, only to find that precisely because of that she has in the end failed her test. Upon leaving the examination room, she might have been asked by her teacher how well she thinks she did and she would have told her teacher that she did fine, that she managed to use “hustle and bustle” and “nuclear family” and “juvenile delinquency” and other impressive phrases, that she will definitely pass with flying colours! A month or two later, both the teacher and the learner will be bitterly disappointed, and perhaps more than a little bewildered, to see that the end result is a fail.
Does this mean learners shouldn’t prepare at all?
This is not because examiners are demented or evil. It is because the objective of speaking tests is to test spoken production and spoken interaction, mostly in an unplanned discourse setting. What learners need to do in preparation for a speaking test is not dissimilar to what they need to do in preparation for most communicative contexts: practise using the language naturally in a range of unpredictable situations, through communication tasks that will help them develop the necessary skills and strategies.
