Object pronouns
Frequency
Olivia Patrick Olivia Patrick 3 x je l’ aime 1 x je le fais
( all four are failed attempts)
L’ école se spécialise Je dois vous dire
Negative forms 2 x ne-pas 1 x ne-aucune 2 x ne-pas
Relative clause 0 2 x ce que 1 x qui 1 x que
4 2
2 3
0 4
Si clause 0 2 0 2 Conjunction – que 0 2 0 2 Conjunction – other 4 0 4 0 Table 8: Complexity results: type and range of syntactical elements other than verbs
Olivia
Patrick Number of short pauses 14 28 Number of long pauses 1- False starts 5 7 Table 9: comparative analysis of dysfluencies
Assessment criteria( 10 points each) Olivia Patrick
Assessor 1 Communication 6 10 Content 6 10 Language 6 8
Overall score 30 points)
18 28
Assessor 2 Communication 7 10 Content 6 10 Language 6 9
Overall score( 30 points)
Table 10: Assessors’ grades for Olivia and Patrick’ s performance
of predictability is therefore high although some uncertainty remains since students do not know the examiners and cannot predict the exact content of the‘ conversation’. Students prepare for this section months – possibly years – in advance.
Overall, then, Olivia’ s performance tended to achieve communicative fluency at the expense of accuracy and complexity; whilst fluency and complexity took precedence in the case of Patrick. These findings seem to lend support to Skehan’ s( 1998) Trade-Off hypothesis, with some aspects of performance gaining high scores seemingly at the expense of others. However, one could argue that the relatively‘ easy’ task did not extend the learners’ linguistic repertoire, thus supporting Robinson’ s( 2001, 2005) claims that it is more complex tasks that would encourage more complex and accurate language.
19 29
Exemplar-based, rule-based language system and prescribed assessment criteria
Both Patrick and Olivia resorted to formulaic language chunks to complete the task. Olivia relied heavily on the formulation c’ est + noun / adjectives to carry the conversation forward, whilst Patrick’ s use of chunks was more subtle and diversified. Indeed, Patrick’ s text shows a relatively high frequency of unusual utterances where a logical sequence of words is repeated several times before the whole utterance comes out correctly( see extracts 3 and 4 in previous section, see also Appendix 2, line 10 and 24 for additional examples). Rather than dysfluency, this might in fact indicate a high exemplar-based system where carefully planned, rehearsed and memorized chunks were used as a strategy to cope with face-to-face interaction. In doing so, fluency in terms of the total number of words produced was boosted, along with the complexity of the speech, and accuracy but to a lesser extent, thus addressing successfully all three official VCAA assessment criteria for this section. Further evidence of this tentative conclusion may be found in the discrepancy between( 1) the highly sophisticated use of relative pronouns, and lexical knowledge and( 2) other aspects of grammar that are not yet mastered( the missed opportunity in line 5 and 32 to contract de + le into du for instance).
Furthermore, and most intriguingly, when asked what his plans were after the exam, the linguistic accuracy and complexity of Patrick’ s speech fell apart, suggesting that this relatively simple question had not been anticipated by the learner, and that he was unable to produce an accurate and / or lengthy reply on the spot( Appendix 2, lines 31-33):
Extract 7: non(…) je ne suis pas sûr(…) // [ j’ ai une(…)] // cette année, c’ est l’ année tout //, c’ est tous les choses, // donc c’ est la fin de le monde! // Alors(…) donc je n’ ai pas pensé beaucoup de projets de l’ année …( Patrick, line 31-33)
no(…) I’ m not sure(…)// [ I have a(…)] // this year, it is the year * everything //, it is * all the things, // so it is the end * from the world! // then(…) so I have not thought much of * projects * of the year
It is also interesting to note that in this instance, Patrick reverted to the same avoidance strategy as Olivia and used very short linguistic chunks consisting of the minimal unit: c’ est + adjective.
It is not possible to say whether Patrick’ s superior performance was solely due to his memorisation strategy; however, it clearly advantaged him. It is well established that formulaic chunks enable students to buy time while the next utterance is planned, leaving therefore more cognitive space for real time processing for such learners compared to learners who resort to rulebased linguistic systems or a limited set of chunks throughout the oral examination( Taguchi, 2007). The additional advantage of memorisation is that all three official assessment criteria can be addressed through a careful pre-planning of the interaction. The trade-off effect between fluency and accuracy / complexity, as seen in Olivia’ s performance, can therefore be better controlled during face-to-face interaction because of pre-planning.
CONCLUSION
Overall, as indicated by the data in this study, Olivia and Patrick managed to complete this first‘ conversation’ section of the exam to a reasonable standard with Patrick gaining very high scores. Fluency was favored by both learners although different strategies were used to achieve it. Olivia tended to avoid risk taking by resorting to a small set of relatively simple formulaic chunks easily transferable to a range of topics. Her grammatical accuracy, although relatively low, required little
10 BABEL