Introduction

A growing number of developmentalists (e.g., Keil, 1998) agree that human beings are genetically predisposed, prior to any experience, to attend to some events (e.g., an entity's movement) rather than others and to attribute a particular event to some preceding events rather than others (e.g., bodily ailment to unfamiliar food). Coherent bodies of knowledge about important aspects of the world are then built on these bases, and many researchers assume that naive physics, psychology, and biology are surely included among the core domains of thought. The acquisition of these core domains is early, easy, and almost universal in the sense that most if not all children possess domains of physics, psychology and biology before or without schooling or any other systematic teaching.

Each of the knowledge systems of these core domains supposedly includes a characteristic set of causal device. Thus, it is implied, even young children possess (1) physical or mechanical causality, by which they can predict and sometimes even explain, the motion, trajectory, and speed of solid objects in terms of 'force'; (2) psychological or intentional causality attributing (human) goal-directed behaviors to mental states including desires, beliefs, emotions and intentions; and (3) biological or teleo-vitalistic causality attributing bodily processes and properties to their functional significance and/or the working of vital power or energy within them. To put it differently, young children can make coherent and reasonable predications for representative physical, psychological, and biological phenomena. They can even offer a relevant causal explanation, or at the least, choose such an explanation out of a few alternatives. Furthermore, it is often claimed, they can apply these three different types of causalities flexibly and appropriately to specific situations.

How well are these claims about the early possession and differential application of causalities supported? How can young children recognize the proper causality for and adjust it to a variety of situations? How can they acquire, in the first place, these three causalities though adults seem not to teach them seriously? In this paper we will discuss these issues.

Young children possess three different kinds of causalities and apply them differentially

One of the prototypical situations in naïve physics is about the transmission of force from an object to another by contact. A somewhat earlier but ingenious study by Bullock, Gelman, and Baillargeon (1982), using a domino-like apparatus consisting of a rod pushing through a post, five standing wooden blocks and a rabbit doll falling from a platform, provided with us the evidence for young children's understanding of physical causality. Children aged 3-4 years, after viewing the complete "domino" sequence, were asked to predict whether the expected final event would occur when some modifications were done for the initial event (e.g., rod) or intermediary events (e.g., blocks). There were two types of modifications: one was relevant (e.g., a rod too short to hit the first block) and the other, irrelevant (e.g., the color of the rod). The children were asked to give 23 predictions in all. Results indicated that 78-91% and 70-100% for the 3- and 4-year-olds' predictions were correct.

Shulz (1980) reported that children aged 2-6 years understand physical causality even when the transmission of an effect is mediated by wind. Each child observed events such as the following: a shield was placed around a candle with the open side facing the child, the candle was lighted, one blower was switched on, and after five seconds, the shield was removed. When the child was asked which blower made the candle go out, he or she correctly identified the responsible blower.

As well known, studies on theory of mind have shown that preschool children understand that human goal-directed actions are explained in terms of intentional causality, more specifically, his or her mental states, such as intentions, desires, emotions, and beliefs. A recent meta-analysis of the false belief task performances by Wellman, Cross and Watson (2001) clearly shows that by about age four children come to understand that actions are based on the actor's representation of the world, not the world itself. In other words, intentional causality is acquired as early as mechanical causality.

Some recent studies have indicated that even infants may be sensitive to the ontological category of entities. In particular, they seem to have different expectations for humans versus nonliving solid objects. Preschool children likewise apply different causal explanations for animate entities and inanimate ones. For example, Massey and Gelman (1988), using unfamiliar animals (e.g., echidna) and artifacts (e.g., wheeled vehicles, rigid objects, etc.) as target objects, found that 3- and 4-year-olds can correctly judge whether animals and artifacts have a capacity for self-initiated movement or not. The children were correct on about 85% of their first yes/no answers, indicating that they answered that animals could go up and down a hill by themselves, while inanimate objects, even if they looked like animals, couldn't. Analyses of explanations that the children gave spontaneously or in response to a request for justification of their yes/no responses suggested that these children tended to change kinds of explanations depending on the types of the objects. When talking about an animal, children often focused on body parts that enable the target to move such as "It can move because it has feet," whereas for the wheeled vehicles or rigid objects, they referred to an agent needed to move the object down and up (e.g., "It needs a push and then it goes," or "You have to carry it down."

A causal explanatory framework unique for biological properties and processes, however, may be acquired a little later than psychology and physics, because understanding bodily processes presupposes the awareness of the processes (Inagaki & Hatano, 2002), and because a biological mode of causal reasoning presupposes the construction of an integrated category of living things including animals and plants, which appear so different. Nevertheless, studies to date have indicated that children as young as 5 years can apply biological, or teleo-vitalistic causality to biological phenomena. When asked to choose one from three explanations presented for bodily phenomena such as digestion or respiration, young children preferred vitalistic explanations most often (Inagaki & Hatano, 1993, Morris, Taplin & Gelman, 2000). For a question, "Why do we eat food every day?", for example, the children preferred a vitalistic explanation ("Because our stomach takes in vital power from the food") to either an intentional explanation ("Because we want to eat tasty food") or a mechanical explanation ("Because we take the food into our body after its forms is changed in the stomach and bowels").

The evidence so far concerns that preschool children can apply different causalities to different entities or events. The real complexity in the enterprise of differentiating explanations lies in applying different types of causal reasoning for different behaviors of the same entities. As pointed out by Wellman, Hickling, and Schult (1997), humans are not only psychological beings but also biological organisms, and physical entities as well. In other words, human actions can be caused not only by psychological states (e.g., desires and beliefs) but also by physical forces (e.g., gravity) or biological processes (e.g., reflexes).

Thus, Wellman et al. took up young children's explanation of human actions that might induce psychological, physical, or biological reasoning, and examined whether children could apply types of causal reasoning best fit the target phenomena. In a series of their experiments, 3- and 4-year-olds heard four to nine stories about a protagonist who wanted and intended to do something and could or could not do so. For a simple, spontaneous actions (e.g., pours milk on the cereal) a protagonist can do so as he wants, but for impossible physical actions (e.g., floats in the air) or for impossible biological actions (e.g., hangs on a branch forever) a protagonist cannot do his intended actions because of physical or biological constraints. After each story, the children were asked to explain, "Why did that happen? Why did the protagonist do that?"

Wellman et al. found that the 4-year-olds consistently gave different explanations for the three kinds of human actions; they gave psychological explanations most often for the psychological actions, such as, "He thought it was milk"; biological explanations most frequently for the biologically impossible actions, for example, "His arms got hurting"; and physical explanations most often for th

The 3-year-olds consistently gave psychological explanations for the psychologically caused human actions and physical explanations for the impossible physical actions, but for the impossible biological actions they gave as many psychological explanations as biological ones. However, when asked to judge whether the desired actions were possible (before offering explanations), 3-year-olds predicted as correctly as the 4-year-olds, saying that psychological actions were possible but that the physically-constrained or biologically-constrained actions were not possible. This indicates that the 3-year-olds, like the 4-year-olds, clearly made different predictions for human actions with psychological and biological impetus.

Based on these laboratory data and further analyses of everyday conversation, Wellman et al. concluded that "children evidence at least three basic everyday reasoning systems as young as 2 years -- physical and psychological reasoning surely, and even biological reasoning in a rudimentary form," and apply them in flexible and sensible ways. However, the scrutiny of their laboratory data reveals that it is not until at 4 years of age that children understand well the psychology-biology contrast in causality.

Inagaki & Hatano (2002) dealt with the biology-psychology distinction. Here children aged 4-6 years were presented with a pair of drawings of two boys, who were allegedly different in terms of biological/bodily (e.g., imbalanced diet or insufficient fresh air) or psychological/social factors (e.g., misbehaviors) in their daily activity, and asked which of the two was more likely to catch cold. Results indicated that a majority of the children in each age group chose often, as being more likely to catch cold, the boy who engaged in biologically bad activities. Although the 4- and 5-year-olds could give few reasons for their choices, about half of the 6-year-olds justified their responses to the items of eating a little or eating few vegetables, such as, "(A boy who has) little nutriment does not have energy, so germs easily enter his body" or "When this boy X eats a lot, his throat is full of nutriment. This boy Y eats little, so his throat is not full of nutriment, and so the coughing can pass through his throat." Since the choice patterns of the 4- and 5-year-olds were very similar to those of the 6-year-olds, the 4- and 5-year-olds, like the 6-year-olds, would consider that a lack of energy or vital power is likely to make children susceptible to illness. Their results also indicated that these children believed that social/psychological factors would influence susceptibility to illness. However, when forced to choose one between the biological and psychological factors, the children evaluated the former factor more important than the latter for determining susceptibility to illness.

The above experimental data strongly suggests that by about 5 years of age children come to appropriately apply different causalities to psychological, physical and biological phenomena.

Differential applications of causal devices

How can young children apply causal devices so skillfully? The knowledge systems of the core domains that include characteristic causal devices are a prerequisite, but there must be a few additional conditions. Let us propose two of them below. First, even young children have some intuition about the appropriateness of causal devices that seems similar to lay adults'. In other words, their choice of the domain to which a given phenomenon is assigned is more or less pertinent, if somewhat unstable or inaccurate. In fact, they are fairly good at clustering pieces of knowledge (Lutz & Keil, 2002). Thus, when asked to predict or explain, they can promptly activate a few causal devices as promising candidates. The living-nonliving distinction is acquired especially early, and thus they seldom explain behaviors of living and nonliving entities in similar ways. Moreover, young children can aptly switch their explanations, taking contexts into consideration. For example, for apparently the same human action of drinking, a large majority of 6-year-olds gave a biological explanation when it was "inevitable" (i.e., drinking water after running for a while) but the psychological one when it was "optional" (taking a glass of juice when beverages were offered), as illustrated as follows: one 6-year-old boy answered for the inevitable behavior, "Because the protagonist was thirsty", whereas he explained, for the optional behavior, "'Because the protagonist liked banana juice".

Second, though young children's repertory of possible causal devices must be limited, they can ingeniously process them by relying on analogy and metaphor. For example, when they see half-withered grass become greeny and healthy after a shower, they can infer that being given water produces the change. They may apply the core biological causality of "vital power taken from food/water makes a human active and lively" with a slight modification and indicate that "vital power taken from a shower makes grass healthy." Children are good analogists and are particularly good at exploiting usable partial analogical relationships, particularly often in naive biology (Inagaki & Hatano, 2003). Similarly, children may metaphorically expand the notion of vital power from the material or energy taken from food to the mental or social one such as refreshing air or sympathetic support.

Multiple origins of causalities

Preschool children have already acquired the ability to make coherent and reasonable predications for typical physical, psychological, and biological phenomena, and they can sometimes even offer a relevant causal explanation. When does this ability emerge? How is it prepared? Before concluding our paper, let us speculate about this issue, extrapolating from what we have seen for preschool children. We would like to rely on Piaget's ideas and observations rather than assuming forms of core causal knowledge to be "innate." First, we accept his general idea (1950) that intellectual devices including the core causalities are prepared during the sensorimotor period of intelligence, and are transferred to the verbal form later when children acquire the symbolic function and language. Second, we adopt his specific idea (1954) that infants' understanding of causality starts with the case in which they themselves are agents (i.e., his protocausality). As well known, Piaget was concerned with the long and gradual process of 'objectivization' of protocausality and the emergence of 'true' causality. We instead emphasize his starting point. In other words, we assume that the initial form of causality is the sense of agency that one's action consistently produces a desired change in an entity if and only if the action is executed appropriately. However, we believe that we have to expand Piaget's formulation to give a due place to the origin of psychological and biological causalities. In other words, in spite of his great insight, we are afraid, Piaget ignored the domain-specificity of causality and treated physical causality as the sole origin of forms of causalities. As Mandler (2004, p. 101) conjectures, a notion of causal force can be derived from perceptual analysis of the transfer of motion between two objects combined with "bodily experiences of pushing against resistance and being pushed," but it is only part of the whole story. Although very few episodes were reported by Piaget as to the infants' actions and reactions to caretakers and other humans, we can just assume that infants all over the world have to seek a caregiver's help to satisfy their physiological needs. They soon find that they can have the caregiver come by communication, such as moving their body, arms, hands or vocalizing. No direct contact is required. Again, their grasp of the causal role of communication in inducing the approaching behavior of the caregiver is "protocausal," that is, limited to when they are serving the causal agent.

The case of biological causality is a little different, primarily because biological processes are slower to proceed than the physical or mental ones. The situation in which infants are likely to recognize the effect of his attempt to operate regularly is probably the removal of hunger and lassitude by being fed, and this is consistent with the eating-centered nature of young children's naive biology. Infants may not suffer from severe hunger in the technologically advanced society, but they become hungry surprisingly fast. Although feeding may not be at infants' disposal, we assume their persistent request is usually met promptly by the caregiver. Thus they learn that something from milk or another food, which might later be conceptualized as 'vital power,' makes them full of energy. In conclusion, the core causalities are prepared during the sensorimotor period of intelligence as hypothesized by Piaget. However, unlike his assumption, there must be multiple origins and multiple pathways of development for physical, psychological, and biological causalities.

References

Bullock, M., Gelman, R., & Baillargeon, R. (1982). The development of causal reasoning. In W. J. Friedman (Ed.), The developmental psychology of time. New York: Academic Press.

Inagaki, K., & Hatano, G. (1993). Young children’s understanding of the mind-body distinction. Child Development, 64, 1534-1549.

Inagaki, K., & Hatano, G. (2002). Young children's naive thinking about the biological world. New York: Psychology Press.

Inagaki, K. & Hatano, G. (2003). Conceptual and linguistic factors in inductive projection: How do young children recognize commonalities between animals and plants? In D. Gentner, & S. Godin-Meadow (Eds.), Language in mind. The MIT Press.

Keil, F. C. (1998). Cognitive science and the origins of thought and knowledge. In W. Damon (Ed.), Handbook of child psychology, 5th ed., Vol. 1: R. M. Lerner (Ed.), Theoretical models of human development. New York: Wiley.

Lutz, D. J. & Keil, F. C. (2002). Early understanding of the division of cognitive labor. Child Development, 73, 1073-1084.

Mandler, J.M. (2004) The foundations of mind. New York: Oxford University Press.

Massey, C. M., & Gelman, R. (1988). Preschooler's ability to decide whether a photographed unfamiliar object can move itself. Developmental Psychology, 24, 307-317.

Morris, S. C., Taplin, J. E., & Gelman,< S. A. (2000). Vitalism in naive biological thinking. Developmental Psychology, 36, 582-613.

Piaget, J. (1950). The psychology of intelligence. Routledge and Kegan Paul.

Piaget, J. (1954). The construction of reality in the child. New York: Ballantine books.

Shultz, T. R. (1980). Rules of causal attribution. Monographs of the Society for Research in Child Development, Vol. 47, No. 1.

Wellman, H. M., Hickling, A. K., & Schult, C. A. (1997). Young children's psychological, physical and biological explanations. In H. M. Wellman & K. Inagaki (Eds.), The emergence of core domains of thought: Children's reasoning about physical, psychological, and biological phenomena . San Francisco: Jossey-Bass.

Wellman, H. M., Cross, D. & Watson, J. (2001). Meta-analysis of theory-of-mind develop