Developmental Science 7:3 (2004), pp 277–282 © Blackwell Publishing Ltd. 2004, 9600 Garsington Road, Oxford OX4 2DQ, UK and 350 Main Street, Malden, MA 02148, USA. Blackwell Publishing Ltd REPORT Abandoning a powerful error Young children who abandon error behaviourally still have to free themselves mentally: a retrospective test for inhibition in intuitive physics Norman H. Freeman, Bruce M. Hood and Caroline Meehan Department of Experimental Psychology, University of Bristol, UK Abstract When preschoolers overcome persistent error, subsequent patterns of correct choices may identify how the error had been over- come. Children who no longer misrepresented a ball rolling down a bent tube as though it could only fall vertically, were asked sometimes to approach and sometimes to avoid where the ball landed. All children showed requisite task-switching flexibility. The pattern of 4-year-olds’ correct choices among different places showed unnecessary avoidance of any place that would pre- viously have tempted them into a vertical-approach error, 5-year-olds rebounded into a reversal, and 7-year-olds were flexible. The data attest to an inhibition mechanism, ruling out competing possibilities. 1. Introduction When young children make a conceptual advance, it is good practice to devise indicator tests to examine pre- cursors and associates of the advance (see Bryant & Allegria, 1989, for methodological analysis of one domain). From such a perspective, it is understandable that chil- dren who have already just gone past the point of change are of lesser interest, except perhaps for showing where ceiling performance begins. But even fully successful behaviour may conceal mental traces that give valuable evidence on what had caused the success. Unearthing any trace needs a sort of retrospective ‘archaeological test’ that we now document as an addition to the usual test- ing. Development of the argument involves four steps. First, the general background is a functional approach to how change occurs. When a child approaches a chal- lenge, many alternative behaviours may become activated (see Siegler, 1996). The mind generates competencies that are liable ‘to compete (at the same time as they are being constructed). Hence, the dire need for a blocking mech- anism: inhibition’ (Houdé, 2000, p. 69). Thus, ‘Inhibitory processes are ubiquitous in psychological functioning’ (Leslie & Polizzi, 1998, p. 247; see also Dempster & Brainerd, 1995; Hauser, 1999; O’Reilly, 1998), and ‘Cog- nitive development can be conceived of, not only as the progressive acquisition of knowledge, but also as the enhanced inhibition of reactions that get in the way of demonstrating knowledge that is already present’ (Dia- mond, 1991, p. 67, emphasis in original). Clearly the role of inhibition in any given instance of fundamental change must be investigated in relation to competing hypotheses. As it happens, the experiment below generated evidence for an inhibitory transition mechanism in a way that ruled out four alternatives; so it is economical to defer adversarial considerations to the Discussion section. The second step is to focus on something to which the species has to adapt, whose representation inflicts per- sistent error on preschoolers: ‘Gravity is the most con- stant, pervasive and significant of all the features of man’s environment to which he orients himself. Under normal conditions it is virtually constant, both in strength and direction, and affects virtually every aspect of man’s behaviour’ (Howard & Templeton, 1966, p. 175). Directional constancy provides a functional basis for the common misconception that unsupported bodies fall straight down independently of any horizontal forces (McCloskey, Washburn & Felch, 1983). Consider the apparatus shown in Figure 1. If a ball is dropped down an opaque bent tube, preschoolers repeatedly search in Address for correspondence: Norman Freeman, Department of Experimental Psychology, University of Bristol, 8 Woodland Road, Bristol BS8 1TN, UK; e-mail: n.freeman@bristol.ac.uk