Possessing both the declarative knowledge to know what things are and the procedural knowledge to manipulate them, infants lack only the conditional knowledge to exploit their theories and abilities. This is for the best, as infants are idiots.
Infants are declaratively and procedurally smart enough. Infants have long been recognized to have procedural knowledge, the ability to do things. These physical schemes constitute “specific, readily labelable class[es] of sensorimotor action sequences” (Flavell, Miller, & Miller, 2002, 65), and many schemes are shown from birth. Likewise, infants have both implicit and explicit declarative knowledge about the world. An example of implicit declarative knowledge are the impossible event tests, shown to infants already habituated to a possible variation (Baillargeon, 1987), Likewise, infants can be actively tested to gage their declarative knowledge. Infants can sort these facts, as well. Babies are born with biased brains that will quickly generate them (Leslie, 1988; 1995). Very young children can tell the difference between animals and objects by three months (Behl-Chadha, 1996), and can tell many types of animals apart from each other (Quinn, 1999).
Yet while researchers are repeatedly surprised about how much they seem to know, it is stunning how poorly babies can put this information together useful. With respect to conditional knowledge, they are incompetent. The text notes that “To ask… why the child bangs when provided a banging scheme and a compliant object to bang with is much like asking why she breathes when provided with lungs and air” (Flavell, Miller, & Miller, 2002, 66). Babies appear to be odd creatures, knowledgeable about the world but lacking agency.
For example, the famous A-not-B error is where a child searches for an object in a behaviorally learned location, even if the object is clearly somewhere else (Butterworth, 1977). The child has declarative knowledge of where the object is, and procedural knowledge of how to retrieve it, but does not have the conditional knowledge to tie these schemata together into a plan of action. This can be rephrased as lacking the ability to correctly inhibit behaviorally learned procedural responses even when declarative knowledge of the situation is clear(Diamond, 1991a; 1991b). A further example of this is that the A-not-B error is only made if the child has to wait an age-dependent time interval (Diamond, 1985). If the infant acts immediately, the declarative schema foremost in his mind (that the object is over there) is activated. If the infant acts after a delay, the procedural schema best learned is activated. What is missing is neither behavioral nor procedural knowledge, but the ability to conditionally tie these together.
Eventually, conditional knowledge comes about. Just as adult problem solvers use analogies (similarities between schemata) to get what they want (Henrich et al., 2001; Gardner, 2003), infants begin to show this behavior by 10 months (Chen, Sanchez, & Campbell, 1997). Logically manipulating category membership to reach logical conclusions occurs by 14 months (Mandler & McDonough, 1996).
In my introduction, I mentioned that it was good that conditional knowledge comes last. This is good because infants simply do not know too much. Operating on faulty schemes can be worse than not-operating at all. The developmental program of babies’ allow them to grow their procedural and declarative knowledge first, before allowing them to integrate these schemata into conscious actions. The “terrible twos” are terrible enough without them happening at two months, instead of two years.
Cognitive Development, a tdaxp series
1. Introduction
2. Infant Perception
3. Infant Cognition
4. Representation and Concepts
5. Reasoning and Problem Solving
6. Social Cognition
7. Memory
8. Language
9. Questions and Problems
10. Bibliography
