5th UNTELE conference University of Compiègne March 2004

5th UNTELE conferenceUniversity of CompiègneMarch 2004

An analysis of dyadic discourse within a learning environment designed for learner autonomy

by

David Rees Ph.D.

Institut National d’Horticulture, Angers

www.multimania.com/davidreesrees@angers.inra.fr

2. A typical language lesson

1. The learning environment

3. Negotiation

CONTENTS

4. Vygotsky’s concepts of internalisation and the ZPD

5. Discourse analysis

6. Corpus analysis

7. Results

8. Conclusions

‘Grande Ecole’ with two colleges (Bac 0+5 and Bac 2+3)Applied engineering in horticulture and landscaping

2 foreign languages (for specific purposes) with compulsory minimum levels

Compulsory foreign professional training period in Year 1

1. The learning environment

Over 50% of students on inter-university exchanges

Highly positive attitudes for learning foreign languages

All lessons in 25-post multimedia rooms, and available via intranet.

2. A typical language lesson

The Tandberg pilot

Pedagogy based on Mutual Scaffolding

1. Separate the classinto two equal groups

2. Select the sources

Video 1 Video 2

3. Diffuse the sources

4. Form intragroup dyadsMacro/micro-comprehension

4. Intragroup dyad: micro-comprehension

5. Intergroupdyads

(negociation)

3. Negotiation

The repeating, rephrasing and restructuring of phrases in L1 or L2 between two or more learners to enable them to understand the meaning of the messages they arecommunicating (Long)Interlanguage (Selinker)

1

Intragroup negotiation1

1

Student 2 Student 4

Input 1 Input 2

Student 1 Student 3

Intergroup negotiation

2

2

2

INFORMATION

GAP

NNS NNS

Negotiation of content

Negotiation of comprehension

NNS

linguisticor

semiotic‘self’

NNS

linguisticor

semiotic‘self’

ZPD ZPD

2

negotiation and the ZPD

Each learner assists theother since each hasinternalised differentsemiotic, linguistic orconceptual competences).Mutual scaffolding takes place.

5. Discourse Analysis Model

1. Problem source2. Repair type3. Discourse code

Problem Source

Ph Phonological (caused by pronunciation or accentuation)

Gr Grammatical (caused by the syntax of a word or phrase)

Lx Lexical (caused by unknown or incorrect vocabulary)

Cn Content (caused by lack of comprehension of the content or concept)

Ds Discourse (caused by pragmatic, social or cultural misunderstanding)

Ps Pause (a pause can indicate a problem and incite repair)

Problems can be due to production mistakes or comprehension difficulty

Repair Type

XL2 Explanation in L2

XL1 Explanation in L1

GT Grammatical Transformation

TL1 Translation into L1

TL2 Translation into L2

Mod Model (the repair is an attempt to provide the ‘correct’ word or form)

Syn Synonym (a synonym or alternate version is provided)

Rep Repetition

Con Confirmation

Com Completion (normally following a pause; the completion of a word or phrase)

Discourse Codes

SR Self Repair ‘she disperses, it disperses …’

RA Requested Assistance ‘how do you say disseminer’?

RR Response to Request ‘disseminer is to disperse’ / ‘I don’t know’

AC Acceptance ‘disperse, okay’

UR Unrequested Repair A. ‘who mutates ’ B. ‘that mutates’

UA Unrequested Assistance A. ‘It’s a scented fruit’ B. ‘Like the guava’

CC Confirmation Check ‘A power station, okay?’

6. Corpus Analysis

7. Results

Negotiation triggers

45% lexical

25% content problems

11% due to silence

Repair type

15% explanation in L2 15% translation in L1

10% repetition 19% confirmation 15% completion

Discourse type

44% Request for help 20% Unrequested help 23% Confirmation check

Self-repairs

43% Grammatrical transformation 50% Provision of a model

Comparison of student/teacher intervention

Mostly grammatical, phoneticand discourse triggers for the teacher

Mostly lexical and content triggersfor the students

Comparison of teacher / student repair types

Teacher: high degree explanation in L2and provision of correct model

Students: a wide-variety of repair types

Teacher: high level of non-requested aid

Students: high level of aid requests

Laughter

An average of 10 laughter ‘events’ per dyad per lesson

8. Conclusions

Dyadic, task-based pair work maximises negotiation opportunities

Negotiation leads to acquisition

Negotiation is effected by:a) task typeb) familiarity of partnersc) cultural similarity of partners

NNS-NNS negotiation appears to be more suitable than NS-NNS negotiation

Technology can enhance a dyadic learning environment

A technology-structured environment can enhance learner autonomy

5th UNTELE conferenceUniversity of CompiègneMarch 2004

An analysis of dyadic discourse within a learning environment designed for learner autonomy

by

David Rees Ph.D.

Institut National d’Horticulture, Angers

www.multimania.com/davidreesrees@angers.inra.fr