An Electronic Magazine by Omar Villarreal and Marina Kirac
Year
6
Number 141
6700
SHARERS
are reading this issue of SHARE this week
__________________________________________________________
Thousands of
candles can be lighted from a single candle, and the life of the candle will not
be shortened. Happiness never decreases by being
SHARED
__________________________________________________________
Dear
SHARERS,
This might
probably have been for some of us our first week at work after our summer recess
and the coming week will certainly be the first week at school for most of our
SHARERS who will be busy with planning, exams and staff meetings. We hope (as we
do every year) that this will be the best year of our teaching careers and this
renewed enthusiasm after our well deserved holidays keeps us going year after
year. I am fully convinced that this desire of constant improvement is at the
heart of every true teacher. We, Argentinian teachers, know how hard teaching is
in the present circumstances but we have also learnt how to struggle hard and
survive!
Both Omar and I
hope to be by your side all this year (and if possible more frequently) as in
these past five happy years of SHARE.
Lets stick
together and make this a bright 2005 with great and modest achievements but with
the intimate conviction that we never really stop trying
hard.
Love
Omar and Marina
______________________________________________________________________
In SHARE
141
1.-
The Question of Language Use in Task Based Learning (Part
2)
2.- Constructivist Learning.
3.- Weird!
4.- II Forum
on Educating for Peace.
5.- Dcimas Jornadas de Enseanza de Lenguas
Extranjeras en Nivel Superior.
6.- Start the
2005 School Year on the right foot.
7.- CAECE
Open Day for Teachers of English.
8.-
Forthcoming Events in
9.- Positions
Vacant.
10.- Online
Courses for Teachers and Translators.
11.-
12.- Use it or Lose it.
13.- On The
Road: Previews 2005.
14.- High School Magazine Online.
------------------------------------------------------------------------
1.- THE QUESTION OF
LANGUAGE USE IN TASK BASED LEARNING (PART 2)
The following is the second part of the article by Lynne Cameron that we started publishing in SHARE 139.
The Complex Dynamics of Language Use on
Tasks
Lynne
Cameron
2000
Part
2
Analysis of the
complex dynamics of the task and language use
Analysing
complex dynamic systems
The analysis of
the talk on task as interacting complex systems follows the operationalisation
of complex systems principles established in empirical work on children learning
to reach and grasp, and learning to walk (Thelen & Smith, 1994). Thelen and Smith (T & S) discuss
extension of their methods and tools to cognition, memory and some aspects of
first language acquisition, but the analysis developed below is my own
application of their work.
Clearly, learning to use a foreign language (FL) is different from
learning to reach and grasp, but the two processes are similar in some key
respects. Firstly, neither relies on some innate component or a single cause.
They both require the interaction of many different components and processes in
specific contexts: in learning to reach, a child must co-ordinate vision, muscle
control and arm stiffness; in learning an FL, a student must co-ordinate memory,
pronunciation skills, knowledge of form and discourse skills. Secondly, they both require individual,
on-line, adaptive responses because of different intrinsic dynamics: in
reaching, different children face different problems, depending on their
agility, muscle:weight ratio etc; in FL learning, different students come with
different levels of skill, different personalities etc. On the basis of the similarities, I
apply the ideas developed by T & S to the sample of classroom
data.
An
exploratory analysis
A full analysis
in a complex systems description framework will require data collected at
frequent intervals over a period sufficient to capture the changes under
investigation. The empirical work
in this section is not such an analysis but is a preliminary, exploratory
application of ideas to classroom language learning data, carried out to assess
the viability and usefulness of the framework.
Interacting timescales
Dynamical
principles (in line with Vygotskyan methodology) require the collection of data
over several interacting timescales, particularly the microgenetic and
ontogenetic. The same complex dynamic principles are expected to apply on each
timescale, with changes at one level giving rise to changes at
another.
In T
& S's motor development studies, the timescales were in terms of (a)
minutes, during which children attempted to reach objects, and (b) months, over
which muscles and control developed. In studying classroom language learning,
three interacting timescales will typically be relevant: (a) microgenetic:
language use on a task (minutes), (b) repeated use of a task (tens of minutes),
and (c) ontogenetic: language learning as change in resources over weeks, months
(and years). The data used here
only captures the first two timescales, of use on a task and repetitions of the
task.
Variables and parameters
In order to
describe the trajectory of a complex system at a particular timescale, it is
necessary to identify its collective variable(s). A collective variable
in a complex system condenses the degrees of freedom of the system and acts as a
dependent measure of change in the system (Thelen & Smith 1994, p.251). For reaching, control of hand speed
served to index overall reaching, and so hand speed served as a collective
variable (T & S 1994, p. 272).
The trajectory and phase shifts of the system through its state space are
graphically represented by the successive values of the collective
variable. Empirical work is needed
to establish what can act as the collective variable, which is then
quantified. At the microgenetic
level of language use on task, tools from discourse analysis and conversation
analysis were used to reveal collaborative patterns of interaction (Edwards,
1997). At the next level up,
attention was paid to outcomes and output.
Time
scales are interrelated by the relation between collective variables and
control parameter. Within a particular time scale, the control parameter
is a fixed constant that affects the system, but is not affected by it. However,
if control parameters change, this can lead to a transition in the complex
system at a longer time scale. Variation in a system's behaviour around
transition points can indicate potential control parameters through
co-occurrence, and they can then be tested by manipulating them and monitoring
the impact on the system. In motor development, children's regulation of arm stiffness
proved to be a control parameter in the development of reaching (T & S 1994,
p.267). The search for control
parameters that link language use with language learning would be central to an
applied linguistics application of CST.
Application
to the classroom data
Distinguishing task-as-plan from
task-as-action
What is
immediately clear from inspection of the four stretches of talk is that the task
as originally set by the teacher changes as the interaction proceeds. Although
initially asked to "tell us a little about" the animal, the talk became a series
of teacher elicitations, usually in the form of questions, that students
answered, often with minimal responses, i.e. very short or single word
responses. The choice of animals made by later students may have been affected
by listening to the struggles of earlier students. Likewise, the teacher did not
interact in exactly similar ways with each student. The task evolved or changed
dynamically from its plan to its implementation.
A
first step in a dynamic analysis is therefore to separate the task as planned
from what actually happens when individuals work on the task. Coughlan and Duff
(1994) made a similar distinction between task and activity. I prefer to use the terms
task-as-plan (similar to Breen's task as workplan, 1989), and
task-as-action. The task-as-plan is what the teacher designs or plans, at
or before the start of the activity; task-as-action is what actually happens in
the classroom when individual participants interact. Whereas the task-as-plan is
fixed at the point the activity begins, the task-in-action is dynamic. The
distinction allows us to see each instantiation of a task-as-plan as a discrete
task-as-action, and, in the data used here, there are four tasks-as-action. The task-as-plan in this data is assumed
to be the production of an oral description; further data, such as teacher's
lesson plan or interview data, would be needed to confirm this, but for my
purposes here, this is not necessary.
The collection of task-as-plan and instantiated tasks-as-action can be
seen as the "task".
Describing the dynamics of language resources in use
In each
task-as-action, a student and the teacher used their language resources in
collaborative discourse to talk about a particular animal. A microgenetic analysis of this language
use suggested an appropriate collective variable.
Appendix A shows the four tasks-as-action set out in columns, with the
teacher talk next to the students.
The collaborative discourse mostly takes the form of teacher elicitations
followed by student responses. For students A, B and C, very open teacher
questions that received no verbal response were narrowed down to more closed
questions, which elicited minimal responses of words or phrases. The trajectory
of language use can be characterised as a collaborative reduction of task
demands, from the planned extended talk to simple question-answer sequences.[i] The exchange with student E showed a
different trajectory, with an opening up of the interaction between student and
teacher.
To
capture the collaborative reduction of extended talk to question-answer, I
propose to use the Elicitation-Response Gradient as a collective variable
for this language learning event.
Elicitation-Response Gradient quantifies differences between the scope of
the teacher's question (Elicitation) and that of the student's reply (Response),
and, to do this, needs to cover lexical, grammatical, and ideational content (in
other circumstances, the affective might be more highly relevant).
Table 1 about
here
The range of
teacher elicitations in the data were extracted and ordered in terms of the
demands they placed on the students to understand and to respond with an
appropriate responses (Table 1). Demand was evaluated in terms of the
specificity of the question and the support for the answer provided in the
question. So, for example, a question that included the answer e.g. "is it big
or small?" was deemed less demanding than a question that required the student
to find a lexical item e.g. "it's white and ...?" because the question contained
the lexis needed for the answer. A score was allocated to each elicitation to
indicate its level of demand.
Table 2 about
here
Student
responses were scored in terms of their linguistic complexity, adjusted for
accuracy (using the terms in the sense of Skehan, 1996). Table 2 shows how scores were
calculated: the complexity of a response was first calculated in terms of its
length: single word (or simple formulaic phrase), phrase or clause. Additional
points were given for the introduction of each new lexical item into the
discourse, to reward student initiative in the talk. Accuracy adjustments were
made by removing a point for an inappropriate response, either grammatical
inaccuracy or inappropriate lexical choice. So when Student A responded to the
teacher's question "is it a big or a small animal?" with the response "little one", the
response was give an overall score of 2: 2 points for producing a phrase, 1
point for the new lexical item 1, and 1 point removed because "one" is not
appropriate in talking about the fox as a generic category. Since the Elicitation was scored 3, the
Elicitation-Response gradient for this exchange was 3-2 = 1. For each task-as-action, the pairs of
Elicitations and Responses were scored (see Appendix B)[ii].
Figure 1 about
here
Figure 1 shows
the Elicitations and Response scor JFIF C
$.' ",#(7),01444'9=82<.342 C 2!!22222222222222222222222222222222222222222222222222 W `"
} !1AQa"q2#BR$3br
%&'()*456789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz
w !1AQaq"2B #3Rbr
$4%&'()*56789:CDEFGHIJSTUVWXYZcdefghijstuvwxyz ? (
uwnsI4˱-֤k2schO>z7).{K/D$>5b;Y b?w?||Mm|p ZWo#Db-4fQR#JZ'{;OmLB]9 85hVYo.hXu*ϛqN)liETs *9 Կ5%bjmtG 2cul]`յk%[t9S5[ԅ`yH'=R}Ao@Hݙ@f Go
'r\kB^TdFnOpkv+ܹ+΅ I
@:k11OIؚg+>
ǽoAʹʁJ`{SIHS#AKܴNaf=Yx(ߊϾk_hjQ R CL#iAxKP(=Y^h .UaQ3}l~XTJ|F;N'ZFcsҼQBi $hb*8Ak