The Problem of Metaconceptual Awareness in Theory Revision

The Problem of Metaconceptual Awareness in Theory Revision Stella Vosniadou (svosniad@phs.uoa.gr) Department of Philosophy and History of Science National Kapodistrian University of Athens University Campus, Ilisia 157 71 Athens – Greece Natassa Kyriakopoulou (tonia@aueb.gr) Department of Philosophy and History of Science National Kapodistrian University of Athens University Campus, Ilisia 157 71 Athens – Greece Abstract In this paper we report an experiment that investigated the question of whether elementary school children have metaconceptual awareness of theory revision processes. Fifty- two elementary school children (grades, 1, 3 and 5) were asked to select between phenomenological and scientific depictions of different astronomical phenomena and indicate which of these depictions were closer to “Reality” and which were closer to “Appearance”. The results showed an increase with age in the number of scientific depictions selected. They also showed that the children who selected both phenomenological and scientific representations of the astronomical phenomena were not capable of deciding which depictions best represented “Reality” and which “Appearance”. It is argued that the task requires the ability to understand that the same world situation can be represented in different ways and that children have difficulty understanding the theoretical nature of representations and thus of flexibly manipulating multiple representations of the same physical phenomenon. Introduction As children are exposed to science instruction, they gradually revise their naive physics in ways that make it more consistent with currently accepted scientific explanations. The question we investigated in this paper is the following: Are children aware of this revision process? Or more generally, are conceptual change processes in the learning of science under the full metaconceptual control of the learner? At least two alternative hypotheses can be formulated. The first is that children are like scientists who are aware of their theories and test them in an explicit fashion during the process of theory building and revising. In this case they should have full metaconceptual awareness of their theoretical views and the difference between their views and the scientific explanations to which they are exposed. The other hypothesis is that children are not like scientists in this respect. Although they are capable of interpreting new evidence to revise theories, they are neither aware of their theories nor do they explicitly evaluate them. We are not the first to claim that children can revise their theories without full metaconceptual control. Karmiloff- Smith & Inhelder (1974) argued that young children are capable of forming and revising theories without necessarily being aware of these theories. According to Kuhn, Amsel & O’Loughlin (1988) young children revise their theories as their experience increases, but lack the skillful coordination between theory and evidence of adults. According to them, the ability to think about a theory, that means to represent it as an object of cognition, is weak among young children. Klahr (Klahr, 2000. Klahr, Dunbar & Fay, 2000) investigated developmental differences in search heuristics used in scientific reasoning. They found that children are capable of distinguishing between theory (hypotheses) and evidence. However, children’s performance was inferior to that of adults when they had to distinguish between a given implausible hypothesis and a plausible hypothesis of their own creation. In contrast to adults, children did not simultaneously consider the two alternative hypotheses, but they focused on their own plausible hypothesis and tried to find evidence to support it. Possible inconsistencies were interpreted either as errors or failures to support the desired outcome. In later work, Karmiloff-Smith (1991, 1992) argued that the changes in children’s theories are connected with changes in representations. A way to revise theories is through an internal process, which Karmiloff-Smith calls “representational rediscription”. The end result of representational rediscription is the existence in the mind of multiple representations of similar knowledge at different levels of detail and explicitness, which enable the learner to appropriate this knowledge. Vosniadou (Vosniadou, 2003. Vosniadou, Skopeliti, & Ikospentaki, 2004, 2005) argued that the presence of misconceptions can be used as evidence that the process of conceptual change is not under the full metaconceptual control of the children. Many misconceptions regarding, for example, the shape of the earth are synthetic models that reveal children’s attempts to assimilate scientific information to their naive physics. The model of the “dual earth” is a clear example of a synthetic model, according to which there is a spherical earth in the sky (a planet) and a flat earth where people actually live. We believe that the formation of synthetic models is possible precisely because children are not metaconceptually aware of their own beliefs or presuppositions and of the fact that these beliefs are inconsistent with the new, scientific information to which