Ad. Pellegrini et al., RELATIONSHIPS, INDIVIDUAL-DIFFERENCES, AND CHILDRENS USE OF LITERATE LANGUAGE, British journal of educational psychology, 67, 1997, pp. 139-152
Background. Research in children's oral language and early literacy le
arning currently stresses the facilitative role of social context. Soc
ial context in this literature is typically treated on a macro-level,
e.g., mother-child interaction or peer interaction. We present a more
differentiated model of peer influences on children's learning one ora
l language register, 'literate language'. Literate language, which pre
dicts school-based literacy, is defined as talk about language and lit
eracy. Aims. We suggest that children's temperament and their close re
lationships, in the form of friendships, play important roles in liter
ate language learning. We present separate models for friends and nonf
riends and posit that literate language is learned more effectively be
tween friends because of the emotional tenor of this relationship. Whe
n they are with friends children, even those that might be considered
'difficult', disagree, resolve disagreements, then express emotions in
dicative of social understanding. Reflection upon emotion states, in t
urn, leads to literate language. Sample. The sample comprised 33 males
and 23 females attending American kindergarten classes, with a mean a
ge of 65 months. Methods. Dyads of same gender and race were observed
12 times across the school year during which time samples of oral lang
uage were taken. Measures of children reading and writing were also co
llected. Results. The data support our model, and the friendship model
accounting for more of the variance in literate language (R-2=.69) th
an did the nonfriend model (R-2=.43). Conclusions. Children with frien
ds engage in the sort of conceptual conflict and resolutions which max
imise use of literate language. This context seems particularly import
ant for 'difficult' children. Future research should continue to exami
ne the interface between individual and group levels variables.