Gestures in oral presentations by second language learners
Alexis Tabensky
Department of French - School of Modern Language Studies
The University of New South Wales - Australia
A.Tabensky@unsw.edu.au

Introduction

Most foreign language educators would agree that their students gesture while they learn and practice the target language in the classroom. Surprisingly, little is known about how this actually happens. This paper focuses on two issues: the use of gesture by second language learners in prepared and in improvised discourse, and the relationship between gesture and proficiency level in the target language. Regarding the first issue, research suggests that more gesturing can be expected during spontaneous than during prepared discourse as speech encoding is more likely than ‘surface’ speech, e.g. recitation and reading aloud, to activate the motor structures linked with the production of meaningful language (Cosnier as cited in Calbris 1989, de Ruiter 2000). Concerning the second, there are strong suggestions in the literature that the amount of gestures by second language speakers decreases as language ability develops (Gullberg 1998, Nobe 2001). These two hypotheses were examined by means of a study conducted on a series of prepared talks given by learners of French during their classes at university (1). Each talk was followed by a question time phase so that all speakers produced prepared discourse, involving various degrees of memorization and reading aloud, as well as improvised, spontaneous discourse. There was also a wide range of language ability levels among the learners in the group observed (2). This work is the initial stage of a research project on gesture use by learners of French studying in an Australian university.

Method

Nine students were selected as representative of three achievement levels in the class: low (Cathy, Nancy and Philippa), intermediate (Thomas, Lily and Karin), and high (Alice, Ema and Céline) (3). The speech of both talk and question time phases for all nine students was transcribed and the gestures were coded and quantified. Gesture ratios were calculated both relative to length of time and to number of clauses.

Results

Due to length limitations, only results for the whole group and per achievement level group are presented here (Tables 1 and 2).
Table 1
Gesture Production during the Talk
Time
(secs) Clauses Gestures Clause/
Minute Gesture/
Minute Gesture/
Clause
LOW 353.33 55.67 27.00 9.45 4.53 0.48
INTERMEDIATE 429.33 91.33 31.67 12.76 4.51 0.32
HIGH 527.33 105.00 160.67 11.95 16.39 1.36
AVERAGE 436.67 84 73.11 11.54 8.48 0.72
Table 2
Gesture Production during the Question Time
Time
(secs) Clauses Gestures Clause/
Minute Gesture/
Minute Gesture/
Clause
LOW 87.00 17.67 23.00 12.18 11.28 0.94
INTERMEDIATE 93.33 23.00 25.00 14.79 12.98 1.08
HIGH 241.67 69.67 63.33 17.30 13.88 0.80
AVERAGE 140.78 36.78 37.11 15.69 12.71 0.94

The talk and the question time phases

As described by the task itself, the question time phase was considerably shorter than the talk in all three groups of speakers, which resulted on average in less clauses and less gestures, if one considers absolute figures only. However, when the duration of each phase is considered (through clause/minute and gesture/minute rates, columns five and six of each table), then it appears that on average there were more clauses and more gestures in the question time phase than in the talk phase (respectively 15.69 and 12.71 against 11.54 and 8.48), and also that the gesture/clause rate increased slightly (0.94 against 0.72). These results thus go in the direction of the hypothesis, that spontaneous discourse is likely to lead to more gesturing than prepared discourse.

Achievement level

On the other hand, the hypothesis that gesture use decreases with the development of language ability was not confirmed by the results, as shown in Tables 1 and 2, where high achievement level students gesture more than students in the other two groups: they achieve the highest gesture rates during the talk phase, particularly gesture/minute rate (16.39) and also the highest gesture/minute rate during the question time phase (13.88). Admittedly, their gesture/clause rate is simultaneously the lowest (0.80 against 0.94 and 1.08) but these differences are too narrow to be significant by themselves (the implications of these results will be discussed in the next section). >From another angle, when the two phases are compared, we find that gesture ratios increased considerably in the question time phase for both low and intermediate level groups whereas they decreased also very noticeably for the high level group. These simultaneous changes in opposite directions are unquestionable, the most interesting finding of this study.

Discussion

Delivery modes of the talk differed substantially: some students mostly read what they had prepared, others only consulted their notes or their text from time to time and recreated the speech on the spot, while some others worked out a combination of these two styles. It is this variable degree of dependency between the output and the prepared text that seems to be linked to the occurrence of gesture during the talk, that is, the more the students relied on their written text, the less they gestured, and conversely, the more they were able to distance themselves from their notes, the more they gestured. While ‘reading mode’ did not equate necessarily with low achievement level nor ‘speaking mode’ equate exclusively with high achievement level, these two modes are the most plausible explanation for the considerable difference in gesture/minute rates between the low and intermediate groups on one side (4.53 and 4.51) and the high level group (16.39) on the other. Precisely, in the question time phase where delivery mode did not intervene, the gesture/minute rates were much closer to each other (11.28, 12.98 and 13.88 respectively), increasing for both the low and intermediate groups and decreasing for the high level group, that is, individual differences in gesture/minute rate were smaller in spontaneous discourse. Given this homogeneity, the average rate of 12.71 is possibly a fairly representative measure of gesture behavior by second language learners. The other side of the coin is that homogeneity informs us very little about the relationship between gesture behavior and command of the target language. It only says that, in improvised discourse, Australian learners of French produce 12.71 gestures in a minute and, without baseline studies for Australian and French native speakers, we cannot infer any conclusion from this result. The case of the gesture/clause rates is different though, as the leveling effect brought along a proximity, for all groups, to the one gesture one clause correspondence suggested for native speakers (McNeill 1992). Moreover, in spite of these very small differences, the expectation that development of oral proficiency is accompanied by a decrease of gesture/clause ratio (Kita as cited in Gullberg 1998), seems to be confirmed by the lowest result found in the high level group (0.80 against 0.94 and 1.08). However, in the talk phase, where differences are more pronounced, we find precisely the opposite: competent learners have a ratio of more than one gesture per clause while the other two groups of students have much less than that. This suggests that gesture behavior in second language learners might not be directly dependent on their command of the target language. The use of gesture would be linked rather to the strategies adopted by the learners trying to satisfy the demands posed by the discourse genre they are dealing with. Usually a strategy is adopted in view of the linguistic resources available and of the goal to be achieved, so that the command of the target language does come into play to some degree but it is not the primary reason for a higher or lower frequency of gestures. In the present case, high achievement level students gestured considerably more during the talk than their less ccompetent counterparts because they tried to give lively presentations that were more than a reading, often resulting in a good amount of improvisation. The perceived degree of difficulty of the task, together with personality traits and also the topics themselves, may explain why students in the other two groups preferred to stick to their text. Language ability combines thus with task effect and discourse genre for the choice of strategies that lead to use of gesture. Although this study was carried out on a small sample, it can be suggested that consideration should be given to all these factors in future investigation into gesture by second language learners.

Notes

(1) I am grateful to my students who generously (and bravely) agreed to these video recordings.
(2) The class combined students from Third Year Beginner Stream and Second Year Intermediate Stream.
(3) Language gains in speaking, listening comprehension, writing and grammar had been measured by two sets of tests given at the end of each academic semester in the year. These results were used to rate all twenty-two students in the class according to their average scores. Ratings are thus internal, course-specific and have only this relative value. For this reason they are considered to show ‘achievement levels’ instead of ‘proficiency levels’.

References

Calbris G. (1989). Analyse sémiotique [Semiotic analysis]. Calbris G., Porcher L., Geste et communication [Gesture and communication] (pp.45-223). Paris: Hatier.
de Ruiter J.P. (2000). The production of gesture and speech. In D. McNeill (Ed.), Language and gesture (pp. 284-311). Cambridge: Cambridge University Press.
Gullberg M. (1998). Gesture as a communication strategy in second language discourse. A study of learners of French and Swedish. Lund: Lund University Press.
McNeill D. (1992). Hand and mind. What gestures reveal about thought. Chicago and London: The University of Chicago Press.
Nobe S. (2001). On gestures of foreign language speakers. In C. Cavé, I. Guaïtella, S. Santi (Eds.), Oralité et gestualité. Interactions et comportements multimodaux dans la communication [Orality and gestuality. Interaction and multimodal behavior in communication] (pp. 572-575). Paris: L’Harmattan.