DOCUMENT RESUME ED 372 612 FL 022 297 Mackey, Allison AUTHOR Targeting Morpho-Syntax in Children's ESL: An TITLE Empirical Study of the Use of Interactive Goal-Based Tasks. PUB DATE 94 NOTE 24p.; For complete volume in which this paper appears, see FL 022 292. PUB TYPE Journal Articles (080) JOURNAL CIT Working Papers in Educational Linguistics; v10 nl p67-89 Spr 1994 EDRS PRICE MF01/PC01 Plus Postage. DESCRIPTORS *Children; Elementary Education; *English (Second Language); Foreign Countries; Interlanguage; Interviews; *Language Usage; Morphology (Languages); Second Language Learning; *Sentence Structure; Spanish Speaking; *Speech Communication; Syntax; Tape Recordings; *Task Analysis IDENTIFIERS Australia ABSTRACT This paper reports on a study of the efficiency of interactive tasks at eliciting targeted morpho-syntactic structures from child speakers of English as a Second Language (ESL). Two groups of subjects between 7 and 10 years of age were examined, one group with various first language backgrounds, and a second which used Spanish as their first language. Both groups had varying degrees of exposure to English. Data for the first group were gathered through an audiotaped conversational interview, while data for the second group were gathered through a task-based elicitation procedure. The tasks included picture recognition, story completion, informal interview, picture sequencing, picture differences, and peer meeting. The results indicated that tasks were more efficient than interviews in eliciting targeted morpho-syntactic structures with child speakers of ESL and that the tasks were successful in targeting the structures for which they were designed. (MDM) *********************************************************************** * Reproductions supplied by EDRS are the best that can be made from the original document. *********************************************************************** Targeting morpho-syntax in children's ESU:An empirical study of the use of interactive goal-based tasks Alison Mackey University of Sydney This study is an investigation of the efficiency of interactive tasks at eliciting targeted morpho-syntactic structures from child speakers of English as a Second Language (ESL). The objectives of the study were to examine whether task-based elicitation procedures were more efficient than interviews at eliciting specific targeted morpho-syntactic structures with children and to ascertain how far tasks designed to target specific morpho- syntactic structures achieved their aims. Results show that the tasks were more efficient than Informal Interviews at eliciting targeted morpho-syntactic structures with child speakers of ESL and that the tasks were successful in targeting the structures for which they were designed. The study reported here was part of a wider project which was designed to provide data on the acquisition of ESL by children and the use of Rapid Profile to assess children's ESL development.2 The tasks reported on here form the elicitation procedure for Rapid Profile (Mackey, Pienemann, & Thornton, 1991; Pienemann, 1992; Pienemann & Mackey, 1993). Rapid Profile is a computer-based second language assessment device. It places language learners on a developmental schedule by screening their speech against standard patterns of acquisition of the target language. Rapid Profile assesses the learner's production of morpho-syntactic structures. For Rapid Profile to work efficiently, the structures need to be elicited from the learners in a quick and unobtrusive manner. The tasks were designed for this purpose. This study of the use of tasks to target specific morpho-syntactic structures also represents a contribution to the growing body of work on the use of tasks as research tools. Many definitions/operationalisations of the term 'task' have been offered. Pica, Kanagy and Falodun (1993) make the point that tasks are difficult to define adequately because many features can be interpreted broadly by teachers and researchers and almost any activity-generating, goal-oriented experience can be classed as a communicative task, even an interview. They characterize tasks as having two recurrent features (1993:11). The first is that they are oriented towards goals, i.e. that participants need to arrive at an outcome; ithe second is that they include some sort of work or activity that the participants themselves must carry out. Long and Crookes summarize their previous definitions, considering the dimension that tasks focus on something that is done not something that is said (1992:44). Crookes specifies that tasks can be considered as "a piece of work or an u $ DEPARTMENT CM EDUCATION "PERMISSION TO REPRODUCE THIS Once ol E.:Wootton/5. Rematch and Improvment MATERIAL HAS BEEN GRANTED P kDUCATIONAL RESOURCES INFORMATION CENTER (ERIC) czc ha document hos Mien mo/oduced so *Cowed from the person or OrpflitilliOn egonOtong it rVe -2)C3 0 Mmor chonoes have been mode to onprove roproduCOOO Qualify 2 BEST COPY AVAILABLE Points of view of oproont stated .n th doc TO THE EDUCATIONAL RESOURCES mint do nol rCIMMInly Monment oRtmot IltrAISLAATANI 01aluT AMA...At I. WPEL , Vol. 10, No. 1 activity, usually with a specified objective, undertaken as a part of an educational course or at work" (1986:1). Long states that "by task is meant the hundred and one things people do in everyday life, at work, at play and in between" (1985:89). All these definitional features can be applied to the tasks used in this study. They are goal-oriented interactive activities which are designed to promote conversational interaction between the participants as they carry out the activity. Developing procedures for the elicitation of data has relevance for the fields of Second Language Acquisition (SLA) research and teaching. It may be the case that some researchers and teachers will view morpho-syntactic data elicited through tasks in different ways. Researchers may view the data in terms of its potential for shedding light on interlanguage/SLA processes. Teachers and those who are interested in assessing developmental level in terms of syntax and morphology may see the data as having potential for revealing the student's stage of development. They may evaluate it to determine the learner's progress towards mastery and define what they can usefully be taught next, highlighting any "gaps." This paper will focus on tasks as tools for elicitation rather than teaching. The subject of tasks and syllabus design has been well documented in the literature, especially in response to the communicative competence movement in language teaching. For a review article and further information see Long (1989) and Crookes and Long ( i992). The Effects of Tasks on Interlanguage Production Why use tasks for eliciting samples of interlanguage instead of naturally occurring data? Apart from the obvious factor of time constraints attached to collecting, transcribing, and analyzing naturalistic data there are other reasons. Previous studies have shown that gaps were present in many samples of naturally occurring data in English (Pienemann & Johnston, 1987; Pienemann, Johnston, & Brindley, 1988). Certain language features do not occur naturally in conversation very often, but are important in determining developmental level. Learners in free conversation and conversational interviews with researchers have shown avoidance strategies, such as dropping topics when structures which cause problems for those learners may need to be used. It is often these troublesome structures which need to be studied for insight into learners' interlanguage. Constantly directing a learner's attention to structures which are not being used in an interview situation may have undesirable consequences. It is obviously preferable that subjects remain as unaware as possible of the structures being studied and that the learner's performance is as naturalistic as possible. 3 68 Mackey: Targeting morpho-syntax in children's ESL A primary concern of this study was to impose constraints on learners in terms of the range of possible responses they produced. This was accomplished by providing specific contexts for morpho-syntactic structures to occur in a way that was as unobtrusive as possible, through the use of the interactive goal-based tasks. The tasks were designed to impose constraints on the learner which means contexts have to occur. These make it possible to study the interlanguage production rules. This is based on a long tradition of SLA research and data analysis (e.g., Meisel, Clahsen, & Pienemann 1981). If, for example, a learner must state what a person's actions are in order to complete a picture-based task successfully, then they will probably attempt to use either 3SGs or -ing forms. They might produce a 3rd person form without an 's' e.g., she read (-) or with an 'extra"s/copula', she is reads (>) or with an 's', she reads (+). (All of these forms are of equal interest to the SLA researcher, whereas to many teachers - and > are usually the same as both may be construed as an error which requires correction.) The tasks were designed to provide as many contexts for the structures as possible, regardless of whether the production equals that of the target language, with the intention of counteracting avoidance strategies. The aim of this study was to elicit spoken data which was as natural as possible with as many examples of filled contexts as possible. Task Design When designing the tasks a variety of methods which are in current use for eliciting morpho-syntactic structures were reviewed. A comprehensive review of such methods is contained in Larsen-Freeman and Long (1991). These include things such as reading aloud, structured exercises, completions, elicited imitation, guided composition, Q-A stimulus reconstruction, role plays, oral interviews, and free composition. Aspects of some of these methods were included in the design of the tasks, for example guided composition was interpreted as an oral response to pictures (Story Completion task), Q-A stimulus was incorporated to a limited extent in the Informal Interview which was compared with the tasks. Space does not permit a complete review of the advantages and drawbacks of all or-these methods of elicitation. While a variety of methods were considered, many were inappropriate because of the lack of opportunity for contexts for targeted structures to occur, the likelihood that models would be provided, or because grammatical rather than pragmatic competence was the focus of this study. The tasks represent a variety of types and were designed to utilize a range of features documented in the research. Features of the task design are listed below. 4 69 WPEL , Vol. 10, No. 1 One/Two way The tasks include examples of both one and two-way information distribution. Research suggests that more useful negotiation work occurs with two way tasks (Long, 1990; Doughty & Pica, 1986). However, concerning the elicitation of morpho-syntactic structures, empirical research has not yet shown convincingly which are most effective. An example of a one-way task used in this study is a learner orally retelling a story from pictures by asking questions of a researcher who knows the story (Story Completion task). An example of a two-way task is two learners collaborating to put a set of pictures in order to create a story (Picture Sequencing task). Closed/Open Most of the tasks are closed rather than open. This reflects hypotheses that closed tasks may produce larger amounts of data and more useful negotiation work than open tasks (Long, 1990). An example of this is a "Spot the Differences" task where learners are told there are a specific number of differences (Picture Differences task). Open tasks often result in learners treating topics briefly, dropping topics when in linguistic difficulties and incorporating less feedback. One open activity, the Informal Interview, is used in this study as a comparison. Planned/Unplanned The tasks are unplanned in order to be as naturalistic as possible, and also to avoid uncertainties created by the extra time involved in carrying out planning. (It should be noted, however, that research shows that planned tasks stretch interlanguages further, e.g., Crookes, 1989). Cooperative/Competitive The tasks have learners in dyads working towards the sanieiconvergent goal in a cooperative rather than competitive manner. Learners working individually are not competing with researchers. This allows the tasks to be non-threatening. The tasks chosen were all task types which are in use in various language teaching situations and as such were likely to be tasks learners were familiar with. They appear to be of a non-serious nature and as such have some face validity in that they are in popular use and are entertaining. An additional reason for the choice of tasks as opposed to interviews for the purposes of eliciting targeted morpho-syntactic structures is that they avoid many of the unpleasant characteristics of interviews such as interrogation style questions, learner shyness, and topic control and dominance by the interviewer. 70 1 Mackey: Targeting morpho-syntax in children's ESL Task Bias An additional feature considered to be important in the task design was the attempt to avoid sexism/ethnocentric bias. This area of task design is not widely 1 discussed in the research. Many of the tasks in use in the classroom and reviewed for this study were biased in some way. In designing these tasks attempts were made to avoid bias as far as possible, for example to reflect Australia's multicultural society by including people/food/scenes etc. with a variety of origins and to reflect the status of women as active participants without stereotyping. However, more research into this aspect of task design is necessary. Resources did not allow this study to focus on that aspect as much as would have been desirable and it is inevitable that flaws remain in these tasks. Table 1 provides a summary of the task types, the structures focused on in the I. results section, the design features used, and the participants involved in the interactive situation. Table 1: Overview of Tasks Task Structure Features Participants Picture 3SG 'S' One way/closed/ 1. Subject & Recognition -ing convergent Researcher goal/unplanned 2. Story Wh questions One way/closed/ Subject & Completion convergent Researcher goal/unplanned Informal 3. General Two way/open/ Subject & Interview unplanned Researcher 4. Picture Negs One way/closed/ Subject & Subject Sequencing Cop Inv convergent & Researcher Questions goal/unplanned 5. Picture Negs One way/closed/ Subject & Subject Differences General Questions convergent & Researcher goal/unplanned Meet Partner 6. Questions One way/open/ Subject & Subject convergent & Researcher goal/unplanned 71 6 WPEL , Vol. 10, No. 1 Examples are given in Table 2 belowboth of tasks which were originally designed and tested with adults and of tasks which were designed specifically for children and tested for the first time in this study. Some tasks were appropriate for both adults and children. For tasks 1, 2, 4, and 5 at least two examples of each task-type were used with each subject. With the other two tasks (3 and 6) one example was used. Table 2: Descriptions of Redesigned Tasks Task-type Examples created for adults Examples redesigned for children Librarian- series of photographs depicting a Supplementary characters used included those Picture day in the life of a librarian. Subjects asked which would be familiar to children who have Recognition "what would she do every day?" etc. been in Australia for several months or longer e.g., their teacher, their parent/s or cartoon characters Poisoning: Series of pictures depicting a Stories depicting scenes such as zoos, picnics Story poisoning and a ransom demand. Subjects and outings and the escapades of animals. Completion instructed to find the story behind the pictures. Researcher and subject speak informally about Topics raised by interviewer for discussion Informal a number of topics. The researcher is instructed aimed at children e.g., favorite food, names of Interview to avoid dominance and topic control where friends, etc. possible. Mishaps: A series of pictures depicting a Pictures aimed at children, e.g., series of Picture number of mishaps which befall a person on mishaps happens to an animal who needs to Sequencing their way home needs to be ordered. Two relearn a skill it has lost e.g., digging/flying in subjects each have half of the pictures and order to get out of trouble. attempt to put them in order. They cannot see each other's pictures. Marriages: Pairs of wedding scenes from Scenes are changed to those which children Picture different cultures. This is a "spot the might come into contact with at school e.g., Differences differences" task where each pair has a picture animal/spaceship pictures. which the other person cannot see. They collaboratively try to work out the differences. Series of topics given to a pair of subjects to Topics aimed at children include information Meet enable them to find out information about each such as pets owned/wanted, school attended, Partner other and report it back to the researcher, and favorite TV shows. Topics include issues such as job, menu preferences, etc. Examples of the Targeted Morpho-syntactic Structures Table 33 below provides an explanation and examples of all of the targeted structures in this study. Not all of these structures are analyzed in detail in the results section due to constraints of space, but they all contribute to the total structure counts. Structures analyzed in detail in the results section are specified, together with the tasks designed to target them, in Table 1. It is important to note that many of the examples of the structures can be grammatical or ungrammatical. 72 7 Mackey: Targeting morpho-syntax in children's ESL Table 3: Examples of the Targeted Morpho-syntactic Structures Negation Neg + SVO Sentence External Negation. A negator is placed before the sentence or clause. Negator form is irrelevant Example: No me live here Neg + Verb Prevetbal Negation. A negator is placed before the main vetb in a sentence. Negator form is irrelevant Example: She not going home Neg Do-2nd In negated main clauses and wh-ci-;....ons, a negated form of the vett) 'do' is placed in second position Example: (Why ) don't he like fish Word Order Topicalisation Objects or Subordinate Clauses are placed in sentence initial position. Example: Because I love you I can't kave Particle-Shift With certain English compound verbs (e.g.,. switch off) it is possible to split the verb and the preposition Example: Have you switched the light off SVO Subject Verb Object is the basic word order for English. Canonical order is used to express a range of functions Example: I like you; No he go home Advetb Adverbs or Adverbials in sentence initial position Example: Tomorrow I'm going home Questions SVO Question Canonical word order used in question formation. Example: You live here ? Do-Fronting Direct questions with main verbs must have some form of the verb 'do' in initial position Example: Does you like fishing ? Y/N-Inversion In direct yes/no questions an auxiliary or modal verb is placed in sentence initial position Stage 4 Example: Have you got my money ? Copula Inversion Wh-questions involving the copula require that the copula and the subject change places. Example: Who is she? Do/Aux-2nd In English main clauses, the auxiliary and modal verbs are placed in second position in affirmative and wh-questions i.e. they change places with the subject Example: Why have you kft home ? Cancel Inv. In relative clauses wh-question inversions do not apply Example: I wonder what she's doing ? Verb Morphology -ed Regular past tense marking on main vain Example: He walked (+) He goed (>) He eat (-) -Irr past Past marking with irregular main verbs Example: She came (+) -ing Any use of the 'ing' postfix. Example: She going home - 3sg -s Third person singular 'S' marking Example: He eats (+) She is eats (>) She eat (-) 73 WPEL , Vol. 10, No. 1 Noun Morphology Plural -s postfix to nouns Addition of plural Example: dogs (+) breads(>) The three dog (-) Poss -s Possessive 's* marldng on nouns Example: Pat's cat (+) Pat cat (-) Pronoun Morphology Possessive Use of possessive pmnouns Example: Their (+) Peter's his (>) they dog (-) Object Use of object pronouns. ...called her (+) -John him (>) -called she (-) Use of -ly to construct adverbs out of adjectives. Adverb Example: run slowly (+) run fastly (>) run slow (-) General Single Words Use of single words to express complex intentions Example: Central -> "I am going to Central" Formulae Learners may use quite complex 'unanalysed' chunks of language which they have simply memorized Example: How do you do Omission The subject of a sentence is missing Subject Example: go home The main verb is omitted Verb Example: she home Copula The copula is omitted Example: That dog big The article is missing from a noun phrase Article Example dog is big As mentioned earlier, Rapid Profile provided the motivation for why these particular morpho-syntactic structures were targeted. Pienemann, Johnston, and Brindley (1988) provide a full discussion and explanation of the placement of these structures into their proposed developmental schedule in terms of the processing constraints which they claim apply to these structures. For further information about the way Rapid Profile incorporates these structures and the acquisition constraints see Pienemann (1992). Research Questions and Hypotheses The two main objectives of this study were, firstly, to investigate whether the tasks used were more efficient than the Informal Interviews at eliciting targeted morpho- syntactic structures with children and, secondly, to test whether tasks elicited the structures which they were meant to target when used with children. Related questions were concerned with whether there was a difference between subject and researcher, and subject and subject situations and whether there was a difference between the two groups of subjects. 74 Mackey: Targeting morpho-syntax in children's ESL Specific research questions and hypotheses formulated were: 1. Did the tasks or the interview produce a higher density of total structures? Hypothesis 1: The density of total structures in the interview would not be higher than in the tasks. The interview was targeted at total structures whereas the tasks were targeted at specific structures. However, it was expected that the tasks and the interview would be similar in terms of the total structures they generated. 2. Did each of the tasks produce the structures at which they were targeted: Hypothesis 2: 3SGs would be produced in greatest quantities by the Picture Recognition task. Hypothesis 3: -ing would be produced in greatest quantities by the Picture Differences task. Hypothesis 4: Question forms would be produced in greatest quantities by the Picture Differences task. Hypothesis 5: None of the structures targeted by the tasks would be produced in the greatest quantities by the Informal Interview. The targeted structures which are the focus of this question are the six types of question formations and the 3SGs and -ing morphological forths. Time constraints did not allow for analysis of each of the 28 structures, so these structures were selected as interesting on the basis that they were either high level structures, found to be difficult to both produce and elicit, or because previous studies (Mackey, Pienemann, & Doughty, 1992) had found them to be more difficult to elicit. 3. Was a range of questions produced ? Hypothesis 6: The Picture Differences task would result in each subject producing at least three different question types from a range of five. Although other tasks were designed to produce specific question types (for example the Meet Partner task and Copula inversion/yes/no inversion questions) the Picture Difference task was targeted at a wide range of questions. 4. Did the elicitation situation affect the production of structures? Hypothesis 7: The researcher and subject situation (ns & nns) would be equally as productive in terms of total structures as the subject and subject (nns & nns) situation. 75 1 0