Setting I 92
recorded; at school, about 30 hours of classroom interaction in Portuguese
were recorded by Iaicrophones worn by the teachers and others visibly
placed in the room.
The home recordings were transcribed, translated into Portuguese, and
divided into utterances and episodes (see Annex 3.6 for an example of the
original transcripts). The number of utterances transcribed was about
68.000, approximately half of which are Target Children's (TC) utterances.
The translation was made by research assistants who were native speakers
of the Ll and fluent in Portuguese, although not professional translators
or linguists, and checked by a fully bilingual member of the team.
The analysis was carried out only for the data collected in the southern
village in the District of Karracuene, and consisted of:
1. functional analysis of Target Children's utterances in spontaneous talk
at home (20 Tsonga speaking children only);
2. functional analysis of pupil-teacher interaction at school in L2;
3. analysis of TCs complex cognitive utterances and cognitive demands on
TCs in spontaneous talk at home (20 Tsonga speaking children only);
4. analysis of complex cognitive utterances and teachers’ cognitive
demands on TC at school in L2.
Classification systems were elaborated by the team on the basis of
adaptations of Tough's and Tizard's models (Tough,1976 and 1977;
Tizard,1984).
For the functional analysis of children's talk at home, the main
categories used in coding utterances were:
- social functions: self-regulative; directive; relational; routines;
- cognitive functions
. descriptive: monitoring,labelling; describing; reporting;
. analytic: sequencing; associating; recognizing
contradictions; explaining how, what for, why;
giving reasons; drawing conclusions;
generalizing;
. projective: predicting; anticipating; projecting; imagining;