science emerged, it managed to override those early belief systems. Those accounts may well help clarify the ways in which science education today might encounter such beliefs and work with them.
There are, of course, many other features of early knowledge that influence the development of cosmology and an intuitive earth science. The large literature on how children conceptualize the spatial layout, for example, is critical to understanding how they start to build models of the arrangement of bodies in space (Newcombe et al., 1998). Similarly, the development of the ability to understand the representational meanings of maps and three-dimensional models (such as globes) undergoes dramatic changes during the preschool and early elementary school years (DeLoache, 1987, 2004; Liben, Kastens, and Stevenson, 2002) and is therefore essential to understanding how cultural artifacts may influence the teaching and learning of earth science. Moreover, the particular symbols and representations used influence the nature of the spatial representations that are constructed (Uttal, 2002), making clear the importance of understanding the ways in which children of different ages naturally conceive of spatial information and the ways in which they are and are not able to glean information from maps and models.
In short, the emergence of a folk cosmology and an intuitive earth science in the preschool and early elementary school years forms a critical skeletal structure within which more formal science curricula must function. Of all five domains considered here, some of the most dramatic changes may occur during the elementary school years with respect to cosmology; but these changes hardly occur in an intellectually empty or unformed mind. Children bring with them a substantial set of interrelated beliefs and expectations about the earth and the heavens that must be taken into account.
As we have shown above, young preschoolers can be exquisitely sensitive to abstract patterns in the world and use that sensitivity to guide how they think about the behaviors of objects, the nature of living things, and the layout of things in space, among many other problems (Keil et al., 1998). Young children and preverbal infants seem to have a strong sense of principles of cause and effect and do not merely notice spatial and temporal contiguity (Leslie, 1984). Moreover, they have reasonable expectations about how causes precede effects and how certain kinds of causes are linked to specific kinds of effects (Bullock, Gelman, and Baillargeon, 1982; Koslowski, 1996). Categorization, induction, and many other forms of reasoning seem to be guided by such abstract forms of information. At the same time, infants and young children can have enormous difficulty explicitly talking about abstract patterns, a difficulty that may well contrib-