Abstract Effective language-learning processes are key in multilingual societies, but past research on gesture and second-language acquisition has often focused on the relationship between gesture and cognition, but seldom on gesture as a teaching and learning tool. Although it is well established that gestures facilitate second-language learning, there is reason to think that different gestures may benefit children differentially. In the context of learning and performing a play, the experiment discussed in this article implements two English-language teaching methodologies, one with teacher gestures at the level of morphology and one with gestures at the sentence level. This experiment, with a diverse group of primary-school-age children, takes a naturalistic setting and shows that among the high and low performers there was a difference in long-term fluency development between the two experimental conditions. The data suggest that the fluency level of learners is predictive of which gesture type benefits fluency the most. Children who had a lower initial speech rate benefited more from teaching using gestures that are morphologically complex, whereas the children who had a higher initial speech rate benefited more from gestures at the sentence level. Introduction Creating a symbolic gesture in the classroom brings something concrete into being that affects ongoing thinking. Just as glancing at a written note helps one to remember, gestures provide a stable physical reference that can embody aspects of cognitive tasks. Previous gesture studies suggest their value for teaching abstract systems such as language (Macedonia and Klimesch 2014) and maths (Novack, Congdon, Hemani-Lopez, and Goldin-Meadow 2014), but research also shows that gestures are not helpful with all types of learning. In distinguishing between second-language word pairs, for example, gestures did not help when the contrast was difficult to perceive phonetically (Kelly, Hirata, Manansala, and Huang 2014). The classroom-based study outlined in this article considers gestures as a foreign-language teaching tool by implementing two language teaching methodologies, one with gestures at the level of morphology, and one with gestures at the sentence level. While learning gestures in addition to speech increases initial cognitive demands on learners, knowing meaningful gestures tied to a word or sentence has been shown to enhance learning (Macedonia and Klimesch 2014). Here we do not ask if gestures per se ‘help’. For this, the interested reader is referred to a review by Macedonia and von Kriegstein (2012). Rather, we ask whether memorizing the same play with the aid of different movement-based teaching methodologies impacts the long-term oral fluency of beginning learners in measurably different ways. In two urban schools in Germany, matched English codified gesture (CG) and scenic learning (SL) units were designed for a one-week English language theatre project for mixed groups of recent refugee and grade 6 learners. Children were placed in two experimental groups where they learnt and memorized the same text. In the CG group, the teacher provided gestures for all the words of the play, meaning words and gestures were learnt simultaneously. Consistent with SL methodology, the teacher taught the children the play supported by gestures for the most important sentences. Fluency and language learning Fluency refers to the level of oral proficiency at which a speaker is easily able to express their thoughts. Fluency has been considered in different ways. Lennon (1990) differentiates between fluency in the broad sense, meaning the ability to produce accurate and complex speech, and fluency in the narrow sense, as measured by the length and number of pauses in a given speech sample. Segalowitz (2000) observes that fluency reflects ‘cognitive fluency’, meaning the cognitive mechanisms that support performance, as well as ‘performance fluency’, which is how this ability is actually expressed. According to Ullman (2016), meanwhile, the declarative and procedural memory systems refer to the neurocognitive systems involved in learning, representing and applying relevant knowledge. Applied to language learning, declarative knowledge includes knowledge of morphology, as well as grammar and pronunciation rules. Declarative knowledge can be quickly learnt, but is slower to use, necessitates more cognitive resources than procedural knowledge, and is also more easily forgotten. Procedural knowledge requires practice but is processed fast and in parallel with other processes, putting less of a burden on working memory, and so is more suitable for fluent speech (de Jong and Perfetti 2011). The number and length of pauses influence perceptions of fluency, but alone are not reliable indicators of the proceduralization of speech, because they vary depending on task demands and planning opportunities. However, used in combination with a measure of speech, as in speech rate, pause data can provide a robust measure of proceduralization. Fluency is perceived by many as difficult to acquire and assess, but by using speech rate data (counted in syllables per time unit) researchers are able to reliably compare fluency between tasks, individuals, and over time (de Jong and Wempe 2007). Since fluency is an important goal of foreign-language education, measuring fluency, as suggested by de Jong and Perfetti (2011), can help to evaluate teaching techniques and methods. Most fluency research has focused on short-term effects, not on long-term developments. While planning and repetition may enable the speaker to benefit from priming, longer-term effects may require proceduralization (de Jong & Perfetti ibid.). In this research, since the same text was learned in both experimental conditions, if teaching gestures in one domain (learning the text of a play about animals) transfers to fluency changes in another (describing a picture of a family) and these fluency changes are sustained over time, this could indicate that the gestures contributed to proceduralization. Gesture and language learning Most teachers and learners agree on the importance of spoken fluency. While many teachers perceive their gestures as being relevant to learning, the role of gesture in teacher training has received much less emphasis than has fluency. Based on advances in neuroimaging, however, many studies have shown the supportive cognitive effects of gesture on language learning (Macedonia and von Kriegstein 2012), as has long been argued by proponents of Asher’s Total Physical Response method (Asher 1977). Neuroscientific research methods may be new, but as Mackey’s ELT Journal article from as far back as 1955 points out: ‘Objects alone are not sufficient to teach meaning. Much has to be taught through gestures like pointing and touching.’ Gestures, or hand, face and body movements made while speaking, are an important feature of human communication, and teachers using gestures in a purposeful or even scripted way is not a recent pedagogical innovation. As previously stated, gestures can support comprehension, memory and recall. When there is something about the hand shape or movement that suggests what is meant, gestures have a clear meaning. When a teacher creates a new gesture, she must produce a symbolic movement with a semantic relationship to what is meant or the gesture will not be immediately understood and must be learnt. Codified gestures refer to gestures that have a ‘dictionary meaning’ shared within a certain group (Poggi 2013). This group can be very large, as in the group of people in the world who understand the ‘thumbs up’ gesture, or as small as the students of one teacher. Codified gestures may be iconic, such as meaning ‘cat’ by showing whiskers on the side of the mouth, but may also be rather arbitrary, as when moving a closed fist back and forth to represent the word ‘why’ in German sign language. Movements indicated by the term gesture are distinct from hand movements that are a part of sign language. While sign languages can be a rich source of useful teaching gestures, the hand movements used in sign language are called signs and they follow grammar that is different from spoken speech. Codified gestures may be part of a system that shows language (such as Signing Exact English (SEE) to show spoken English) but are not a language in their own right. In this experiment, because the teacher gestures referred to fixed morphemes (‘run’ + ‘s’ in the CG condition) and fixed sentences (‘Let’s go out the window!’ in the SL condition), technically both groups of children used codified gestures to support their learning. Codified gesture condition In both experimental groups, the children had text learning phases in which they separately learnt the same text. For the CG condition, the teachers taught a set of codified gestures, one for every morpheme in the play. In this condition, most words such as ‘window’ had a single gesture, but some words such as ‘animals’ had two gestures, one for ‘animal’ and one to show the plural ‘s’. During the text learning phases, the children sat in a semicircle facing the teacher. When reading the text, the teacher initially spoke and slowly cued the play, meaning that words and gestures were learnt simultaneously. The children were instructed to speak together as soon as they recognized a gesture, but were not instructed to gesture. Once the children could recognize and speak the words, of their own accord, they began to imitate the accompanying gestures. Scenic learning condition Scenic learning is a drama-based approach to language learning that combines choral repetition and movement to learn vocabulary words, phrases or sentences (Sambanis 2013). In this condition, the children were initially told to relax, close their eyes and listen to the teacher reading through the play. During the second reading the children were to relax, listen to the teacher read, and listen for words they knew. After listening, the teachers were instructed to work through the text and help the children in ways they had had success with in the past. Techniques used included following in the text with a finger while the teacher reads, the teacher stopping unexpectedly and the children calling out the last word heard; reading the play in parts; and reading in small groups. The focus of the first six text-learning phases for the SL group was on understanding and fluently reading the play. For sessions 7–12, the emphasis was on using gestures at the sentence level to memorize the parts of all actors and practise speaking together. Following the SL approach, the most important and most frequent sentences of the play were practised accompanied by a simple movement. The gesture for the sentence ‘But we can’t sleep!’, for example, was enacted in the SL group by everyone simultaneously raising their open hands to shoulder height and speaking in an exasperated manner. While in the CG condition, all of the words were accompanied by gestures, in the SL condition, excluding the narrator parts, 63 per cent of the words were in sentences accompanied by gestures. For all children, it was made clear that the goal was to memorize every speaking part independent of the role they would play in the actual performance. In the SL group the children had access to the printed text but only during the text-learning phases, after which the texts were collected for safe keeping. It was emphasized that the goal for the performance was speaking and acting together in character groups, and it was not possible to practise this independently. After the text-learning phases finished, both groups (i.e. the CG and SL groups) were combined, roles were assigned and a narrator from each group was chosen. For the final five hours of teaching time, the focus moved from learning the text to presenting it on stage in an artistic way. At both schools, two children took turns reading the narrator’s part. All other children were assigned the parts of the other characters. At one school, during the final rehearsal, several children decided they did not want to perform on stage and sat for the performance in the audience. Because of the nature of the play, with several children speaking chorally for one part, this did not result in significant changes in the performance. Research question In this study, the following research question was addressed: When learning the same text, does the use of different movement-based teaching methodologies, using a gesture for every word without the written text (CG) or using a gesture for the most important sentences with access to the written text (SL), create a measurable difference in learners’ long-term oral fluency? Methodology and design It is well established that gestures facilitate second-language learning. This study contributes to the field of classroom-based research by testing the effects of gestures on fluency outcomes through a preliminary study using a randomized experimental design. Matched codified gesture and scenic learning units for beginning English learners were developed and their effects on oral fluency were tested. Following a repeated-measures experimental design, which measures changes over time, pre-, post- and delayed post-test recording analyses of oral fluency measures were carried out (including the number of syllables and morphemes and the number and length of silent pauses). The same transcriptions were also used to test for grammatical accuracy but these tests are not reported in this article. Participants Fifty-four learners from two primary schools in urban Germany (37 per cent female; ages 10–13; M = 11.65, SD = 0.78) participated in our study. At both locations, the experiment was part of a joint theatre project between members of a grade 6 class and a class of refugee children from the same school. At the first school, there were 20 grade 6 and 8 refugee learners. Of the grade 6 children, 16 (80 per cent) listed a first language other than German as their primary home language. At the second school there were 16 grade 6 and 10 refugee learners. Of the grade 6 children, all listed German as their primary home language. All the refugee children at both schools listed a first language other than German as their home language. The time the refugee children had spent in Germany at the time of the study at the first school was between 6 and 18 months. In the second school, three children had arrived between 6 and 12 months earlier, but seven had been in Germany for less than half a year. All of the children in all classes reported having previously learnt English in Germany or in their home country (grade 6 learners M = 3.94 years, SD = 0.79; refugee learners M = 2.94 years, SD = 1.98). Signed parental consent was obtained and all learners in the study agreed to participate. Procedure and materials At the outset of the study, a fluency test using a set of family pictures not required during subsequent teaching was administered to all initial participants. Students were individually recorded, and in each test they used the same single picture that they themselves had selected. The fluency tests consisted of a planning phase, where the child looked at the picture until he or she was ready to speak, followed by a 1-minute recorded monologue prompted by the sentences: ‘Tell me what you see and what you think. Start when you are ready.’ The 1-minute timer began when the child said his or her first English word. If, after beginning, children became silent and had not spoken for 10 seconds, the researcher pointed to the picture encouraging the child to speak about it. This testing format was chosen for several reasons: Describing pictures is a common task familiar to most learners. From teacher interviews, it was known that some of the refugee children had been at English-speaking schools, so the test needed to be appropriate for children of widely differing abilities. As a result of their migration, many children had experienced a break in their education and a non-threatening test was a prerequisite for permission to run the study. Typical speaking strategies used were naming items (baby, t-shirt), or constructing a story based on the family members (e.g. ‘This is the mother and the father. They love their daughter very much and she is very funny …’). All students completed a pre-test in week 1, before their gesture training, and the post-test in the week after the training, followed by the delayed post-test in week 7 at the project’s end. All training and test sessions took place during regular class hours. Taking class of origin and gender into account, learners were then randomly assigned to either the CG or SL groups. In week 2, the children received 14 hours of English instruction as part of a special week-long project. The main focus of the project week was a simple play. There were a total of 12 sessions, each 15 minutes in length, which focused on learning the text of the play. These phases took place in separate CG and SL groups and made up 3 hours of the total 14 hours of teaching time. These separate training sessions were embedded in activities that happened in the larger group. Beginning and ending the day, sport activities and creative activities, such as singing and designing costumes and props, as well as the final production, all took place in the large group with both teachers in each school present. In the text-learning sessions of the CG group, the teacher gestured the play and the students learnt the text by hearing and ‘reading’ the gestures. In the SL group, after becoming familiar with the text, the students acted out the scenes of the play using gestures at the sentence level to support their learning. During their text-learning phases, the SL learners had access to the written text, which was not the case for the CG group. To control for an influence of teacher effects, both teachers in each school taught both groups balanced for time. It is of significance that the content of the theatre piece (animals) and the family pictures (people) were not the same. For slightly more than 60 per cent of the text-learning sessions, fidelity-of-implementation observers were present in each classroom to ensure that the children were taught as intended in terms of timing, content and activities. In week 3 (post-test) and week 7 (delayed post-test), the children were again tested for fluency. Due to school changes and illness, five children were unable to complete all three recordings which resulted in a total of 49 complete sets of recordings. Transcription and pauses Following de Jong and Perfetti (2011), all monologues were transcribed by the researcher using Praat software (Boersma and Weenick 2015). To make segmenting and coding the speech and measuring pause times somewhat easier, the pause boundaries were first determined using the Praat function ‘To Textgrid (silences)’ (de Jong and Wempe 2011). We defined a pause as silence or a non-verbal filler of 300 ms (0.3 seconds) or longer. This decision follows Lennon (1990), because beginning learners speak more slowly than advanced speakers, and pauses longer than 300 ms sounded dysfluent. All pause boundaries were checked and adjusted by the researcher, if necessary. Because this study is interested in speech rate in English, silent pauses, filled pauses such as ‘uh’ and ‘mm’, and word fragments were coded and treated as pauses. First-language words which were not English, such as Kappi for ‘cap’ or lala for ‘elder brother’, were also coded as pauses and not counted as speech. (It should be noted that filled pauses such as ‘um’ do not show a lack of fluency for advanced speakers of a language; rather filled pauses are assumed to indicate lexical planning. Code-switching also does not necessarily show a lack of fluency. However, because the focus of this experiment was on fluency in English, the more simplified model of fluency outlined here was chosen.) After transcription, syllables were counted using the qdap module for the statistical software R (Rinker 2016) and speech rate (syllables/time unit) changes over time were calculated. The researcher was blind to the experimental group of all participants until transcription and coding had finished. Speech rate in syllables is a measure of fluency which corresponds to expert opinions (de Jong and Wempe 2011); however, speech rate can also be measured in morphemes. While the word ‘cat’ and ‘cat’s’ both have one syllable, the word ‘cat’s’ is morphologically complex (‘cat’ + possessive ‘s’), and consists of two morphemes. Because the two gesture conditions were different at the level of the morpheme, speech rates in morpheme values were also calculated. However, the fundamental patterns observed for both measures were the same. Results As noted above, fluency was tested before the project began. An independent-samples t-test (speech rate in syllables by experimental group) compared the mean scores of the two experimental groups. The initial mean speech rate in syllables for the CG group was M = 0.63 (SD = 0.46) and for the SL group M = 0.60 (SD = 0.33), t(48.9) = 0.24, p = 0.81, indicating that the groups are comparable. The post-test happened in the week following the final presentation of gestures, followed by the delayed post-test 7 weeks after the initial fluency test and 5 weeks after the theatre project. Comparing the two experimental groups, our first analysis showed an improvement for both with no practical differences. However, when the initial speech rate was plotted against the long-term gain on the individual level, it became clear that the experimental group which the children belonged to had a different effect on high- and low-performing pupils. Figure 1 shows what is termed an X-interaction, meaning that the children who had a lower initial speech rate benefited more from the CG group, whereas the children who had a higher initial speech rate benefited more from the SL group. Any gain in fluency is a child-specific variable and measures how children compare to themselves, so cultural or first-language differences among children cannot influence our results. Because the pupils were randomly allocated to each group, we can also be confident that chance initial differences between the two groups are very unlikely to account for the difference in outcomes. Figure 1 View largeDownload slide Long-Term Gain in Speech Rate by Teaching Method Figure 1 View largeDownload slide Long-Term Gain in Speech Rate by Teaching Method Discussion This experiment compares two teaching methods. As noted above, the methods chosen for evaluation were different, most notably in the gestures the teachers used, and access to the written text. While not the only approach to classroom research, using complete methods can establish how different teaching elements, such as gesture type and access to text, work in combination and provide more ecologically valid grounds for generalization than experiments that differ in one variable only. Based on group interviews, we know the grade 6 children were in general motivated about the project and appreciated ‘doing something’ with the refugee children. However, this was not the feeling for every child at every moment. The refugee children enjoyed the theatre project, but some reported their ambiguity towards learning English when the most important language in Germany is obviously German. Based on verbal feedback we also know that several teachers had initial reservations about using the new teaching gestures. Teachers and pupils alike bring their own backgrounds, beliefs and past experiences into their classrooms; however, beliefs about learning are not everything. From our fluency data we see that teacher choices in terms of teaching methods matter. A relatively short time investment, in this case 2 hours of gesture training and 3 hours of teaching time, can make an important difference in long-term outcomes. For the teachers who were unsure if learning the gestures would be worth their effort, this is an important finding. In contrast with the teachers expressing qualms about the gestures possibly ‘not working’, for the children, using gestures to learn seemed surprisingly obvious. The understandable question some of the refugee children had about learning English referred to the language and not the teaching method. The results of this study indicate that learning a foreign language using gestures which are morphologically complex and learning using gestures at the sentence-level affect fluency differently. At the same time, because of the X-interaction in long-term gain, it is clear that no method is per se better than another. An important caveat to long-term fluency differences between the experimental groups is the high level of natural variation in the data. Learner data in general and data from children in particular tend to be highly variable. The children in this study represent many different cultures and linguistic backgrounds, with a significant number challenged by adjusting to life in a new country. Under these circumstances it was not possible to subject the learners to more extensive tests. Despite these limitations, we believe our study provides important preliminary evidence that many students, particularly struggling learners, may accrue a long-term benefit from teaching gestures that are morphologically complex. Conclusions It is known that gestures can embody speech and facilitate language learning, but gesture research from the classroom is rare. The results of this study should be of interest to English teachers of beginning learners who are in search of effective methods for improving oral fluency. In this study, using gestures at the level of morphology appears especially to benefit learners with a lower level of initial of fluency. The children in the CG condition learnt their text through interpreting their teachers’ gestures, so learners who struggle with reading and writing in a foreign language may benefit more from the opportunity to learn texts through alternative means. Thus, for students who are still developing literacy skills, codified gestures may represent a useful means of improving learning. For children who struggle with fluent speaking, ‘reading’ words from somewhere other than a page, such as their teachers’ hands, may benefit them more. Natasha Janzen Ulbricht holds an MA in Applied Linguistics and ELT from St Mary’s Twickenham, London. She has trained teachers and taught English as a foreign language in Germany, Zambia and the United States. Her research interests include gesture and foreign-language learning and teaching in difficult circumstances. She is completing her doctoral studies at the English Didactics Department of the Freie Universität Berlin.. References Asher, J. 1977. Learning another Language through the Actions . Los Gatos: Sky Oaks Productions. Boersma, P. and D. Weenick. 2015. Praat, a system for doing phonetics by computer (version 5.4.08) . University of Amsterdam, the Netherlands. de Jong, N. H. and T. Wempe. 2007. ‘ Automatic measurement of speech rate in spoken Dutch’. ACLC Working Papers 2/ 2: 51– 60. de Jong, N. and C. Perfetti. 2011. ‘ Fluency training in the ESL classroom: an experimental study of fluency development and proceduralization’. Language Learning 61/ 2: 533– 68. Google Scholar CrossRef Search ADS Kelly, S., Y. Hirata, M. Manansala, and J. Huang. 2014. ‘ Exploring the role of hand gestures in learning novel phoneme contrasts and vocabulary in a second language’. Frontiers in Psychology: Language Sciences 5/ 673: 1– 11. Lennon, P. 1990. ‘ Investigating fluency in EFL: a quantitative approach’. Language Learning 40/ 3: 387– 417. Google Scholar CrossRef Search ADS Macedonia, M. and W. Klimesch. 2014. ‘ Long-term effects of gestures on memory for foreign language words trained in the classroom’. Mind, Brain, and Education 8/ 2: 74– 88. Google Scholar CrossRef Search ADS Macedonia, M. and K. von Kriegstein. 2012. ‘ Gestures enhance foreign language learning’. Biolinguistics 6/ 3: 393– 416. Mackey, W. 1955. ‘ What to look for in a method: (III) presentation’. ELT Journal 9/ 2: 41– 57. Google Scholar CrossRef Search ADS Novack, M., E. Congdon, N. Hemani-Lopez, and S. Goldin-Meadow. 2014. ‘ From action to abstraction: using the hands to learn math’. Psychological Science 25/ 4: 903– 10. Google Scholar CrossRef Search ADS PubMed Poggi, I. 2013. ‘ Semantics and pragmatics of symbolic gestures’ in C. Müller, A. Cienki, E. Fricke, S. Ladewig, D. McNeill and S. Tessendorf (eds.). Body—Language—Communication: An International Handbook on Multimodality in Human Interaction . Berlin: De Gruyter Mouton. Google Scholar CrossRef Search ADS Rinker, T. 2016. qdap: Quantitative Discourse Analysis Package (version 2.2.5.) . University at Buffalo, Buffalo, NY. Sambanis, M. 2013. Fremdsprachenunterricht und Neurowissenschaften . Tübingen: Narr. Segalowitz, N. 2000. ‘ Automaticity and attentional skill in fluent performance’ in H. Riggenbach (ed.). Perspectives on Fluency . Ann Arbor: University of Michigan Press. Ullman, M. 2016. ‘ The declarative/procedural model: a neurobiological model of language learning, knowledge, and use’ in G. Hickok and S. Small (eds.). Neurobiology of Language . Amsterdam: Elsevier. Google Scholar CrossRef Search ADS © The Author(s) 2018. Published by Oxford University Press; all rights reserved.
ELT Journal – Oxford University Press
Published: Feb 23, 2018
It’s your single place to instantly
discover and read the research
that matters to you.
Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.
All for just $49/month
Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly
Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.
Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.
Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.
All the latest content is available, no embargo periods.
“Hi guys, I cannot tell you how much I love this resource. Incredible. I really believe you've hit the nail on the head with this site in regards to solving the research-purchase issue.”Daniel C.
“Whoa! It’s like Spotify but for academic articles.”@Phil_Robichaud
“I must say, @deepdyve is a fabulous solution to the independent researcher's problem of #access to #information.”@deepthiw
“My last article couldn't be possible without the platform @deepdyve that makes journal papers cheaper.”@JoseServera