Abstract
Since Nelson's original (1973) study there has been a tendency in the literature to view maternal directiveness in an essentially negative way, equating it with maternal intrusiveness or insensitivity (e.g. Nelson 1973, McDonald & Pien 1982). The present study attempts to re-evaluate this negative characterization of maternal directiveness by examining the interactive behaviour of 8 mothers at the early one-word stage of development. The results show that, in contrast to McDonald & Pien's behaviour-oriented/conversation-eliciting dimension, based on the relative proportion of directives and real questions in mothers' speech, the major dimension of variation in the present sample was the relative proportion of behavioural directives and tutorial prompts or test questions. Furthermore this dimension was unrelated either to maternal sensitivity, defined in terms of responsiveness to the child's utterances, or to mothers' use of attentional directives. These findings raise doubts, not only about the tendency to equate directiveness with maternal intrusiveness or insensitivity, but also about the assumption that directiveness reflects a single unitary dimension of mother's interactional behaviour. This suggests that current views of maternal directiveness are based on a negative stereotype of the directive mother and implies the need to analyse maternal speech effects much more carefully than has been done hitherto if we are to understand the mechanisms by which they are actually mediated.