European Journal of Psychology of Education 1994, Vol. IX, n? 1, 41-53 © 1994, I.S.P.A. Interactional Effects of the Environment on Interviews Genevieve Blanchet Alain Blanchet University of Paris VIII, France This study investigates the effect of the environment factor on discourse mode and content in an interviewsituation. Fifteen eight-year- olds were interviewed in three specific and different locations in their school (classroom, doctor's office and playground) that formed the three experimental groups. Analysis of the interview data (propositional discourse analysis) shows that each experimental group produced a specific type of speech characterized by the occurrence of particular language markers. The school interview is one of the psychologist's tools which it would be erroneous to view as a 'neutral' way of accessing children's representations. All speech in an interview situation is affected by the factors that make up the context, in particular the person conducting the interview, the type of interview (directive/non-directive), the framework and purpose. Observations have shown that psychologists work in a variety of situations in schools, and that many of these settings are inappropriate because there is no special place to see children. This physical variability may have an impact on the psychologist's practice. Bronckart's (1985) analysis of discourse functions groups the parameters which have an observable influence on discourse production in the referent and context domains. Each domain covers a specific level of discourse construction. In his theory of 'extralanguage', the 'space- time' of discourse production is comparable to a channel in information theory. The space variable is defined as a physical area where speech can be accessed. The temporal variable is defined as the physicallapse of time when speech production is accessible. The purpose of the present study is to delineate this space factor, which we assume to have a measurable impact on the referential analysis of speech. The effect of spatial environment is tested here on children's discourse construction in a school setting. School does not form a homogeneous semantic space. Rather, it is arranged in work areas, play areas, health care areas, etc., each of which has particular features. Although the interview situation in itself