Development of Core Knowledge Domains
From where does knowledge come? Scholars interested in this question have delineated three possible sources: experience, culture, and evolution. Knowledge obtained through experience is knowledge derived from one's own observation and exploration of the physical world. Knowledge obtained through culture is knowledge initially derived by someone other than oneself but acquired through the process of cultural transmission. Knowledge obtained through evolution is knowledge of a genetic origin, endowed in humans by natural selection due to its utility to our prehominid ancestors.
Although few would dispute the claim that humans acquire knowledge through experience and culture, many have disputed the claim that humans have acquired knowledge through evolution. Indeed, this claim has remained controversial since its origins in ancient Greek philosophy and its revival in eighteenth-century Enlightenment philosophy (Stich, 1975), though, in recent years, it has gained substantial support from empirical studies of infant cognition and animal cognition. Drawing on the findings of such studies, Spelke (2000) has proposed that innate knowledge, or “core knowledge,” can be identified by three defining features: (1) domain-specificity, or a restriction on the types of objects and relations the system can represent; (2) task-specificity, or a restriction on the goals and objectives the system can accomplish; and (3) encapsulation, or operational independence from other systems of knowledge.
Core knowledge is thought to guide learning in a variety of ways, from guiding the interpretation of one's early experiences to constraining the scope of one's early inferences to providing the foundations of one's future knowledge. Although there is some controversy as to what domains of knowledge are innate and what domains are not, scholars have pointed to at least five possibilities: (1) the domain of objects and their physical properties, (2) the domain of agents and their psychological properties, (3) the domain of space and its geometric properties, (4) the domain of number and its arithmetic properties, and (5) the domain of living things and their functional properties.
Evidence of early emergence, cross-species homol-ogy, and cross-cultural universality are stronger for some domains (e.g., the domain of number) than for others (e.g., the domain of living things). Nevertheless, there is ample evidence that all five domains emerge prior to formal instruction. Below are charted the development of three such domains: the domain of space, the domain of number, and the domain of living things. For each domain, characterizations are provided of (a) the domain's initial knowledge state, and (b) the domain's first major restructuring. Following these characterizations, this entry discusses general similarities and dissimilarities among the early transitions within each domain.
Spatial cognition consists of a variety of competencies, including navigation, depth perception, landmark encoding, and reorientation. Here, the focus is on reorientation, or the process of realigning one's mental representation of the environment with the environment itself, as there is much evidence that reorientation involves an evolutionarily ancient mechanism present in both human and nonhuman animals.
The earliest studies of reorientation were conducted with rats (Cheng, 1986; Margules & Gallistel, 1988). In these studies, food-deprived rats were shown food being hidden in one of four corners of a rectangular enclosure. The rats were then removed from the enclosure and disoriented via rotation. Upon their return to the enclosure, the rats searched for the hidden food in either the correct corner or the geometrically equivalent corner, that is, the corner diagonal to the correct corner, which shares with the correct corner the property of being to the left of a short wall and to the right of a long wall (or vice versa, depending on the particular hiding location). Amazingly, the disoriented rats did not use the nongeometric properties of their enclosure, like wall color or wall odor, to guide their search, even though these properties uniquely specified the food's location and were readily used as navigational cues by fully oriented rats. Similar results have since been obtained with monkeys (Gouteux, Thinus-Blanc, & Vauclair, 2001), fish (Sovrano, Bisazza, & Vallortigara, 2002), and chicks (Sovrano & Vallortigara, 2006).
Studies of reorientation in humans have revealed striking similarities between how disoriented children search for hidden toys and how disoriented rats search for hidden food (Hermer & Spelke, 1994; 1996). In these studies, children aged 18 to 24 months were shown a toy being hidden in the corner of a rectangular room and were then disoriented by being spun around with their eyes closed. Following disorientation, children tended to search for the toy in one of two locations: the correct corner or the geometrically equivalent corner. This behavior persisted even in rooms where the location of the toy was uniquely specified by a distinctive non-geometric cue: a blue wall. Thus, children, like rats, do not initially take nongeometric information into consideration when reorienting themselves, and they continue to ignore such information until around the age of 7 (Hermer-Vazquez, Moffet, & Munkholm, 2001).
Children's sensitivity to geometric information–and only geometric information–in reorientation tasks appears to be limited to a particular kind of geometry: the geometry of extended, three-dimensional surfaces. Studies that have explored children's reorientation behavior in open space have found that children do not reorient by the geometry of moveable objects within that space (Gouteux & Spelke, 2001). Moreover, studies that have explored children's reorientation behavior in different kinds of enclosures have found that children who fail to reorient by differences in wall color will nonetheless reorient by differences in wall shape (Wang, Hermer, & Spelke, 1999). The fact that children's reorientation is sensitive to some features of the environment (i.e., wall location, wall shape) but not others (i.e., wall color, object locations) suggests that the mechanism responsible for this behavior attends only to stable features of the environment unlikely to change from day to day or from season to season.
As mentioned previously, older children (and adults) do not reorient like rats. Instead, they reorient by both geometric information (e.g., wall location) and nongeo-metric information (e.g., wall color). What allows older children, but not younger children, to use such information? One hypothesis, suggested by Hermer & Spelke (1996), is that remembering the location of an object relative to nongeometric features of the environment requires encoding such relationships in language. In support of this hypothesis, Hermer-Vazquez and colleagues (2001) have shown that children's production of the words left and right is highly correlated with their use of nongeometric information in reorientation tasks. Moreover, Hermer-Vazquez, Spelke, and Katsnelson (1999) have shown that adults' use of nongeometric information in these same tasks is significantly impaired when adults are required to listen to and repeat a tape recording of continuous speech, thereby preventing them from producing phrases like “left of the blue wall.” These data imply that the development of spatial cognition is tied to the acquisition of spatial language, though the exact nature of this relationship has yet to be determined.
Numerical cognition, like spatial cognition, appears to be ubiquitous throughout the animal kingdom (Gallistel, 1990), and evidence of numerical cognition in humans can be found as early as six months of age. For instance, habituation studies have shown that 6-month-old infants can discriminate visual arrays of 8 dots from visual arrays of 16 dots (Xu & Spelke, 2000). Infants of this age have also been shown to discriminate auditory sequences of 8 tones from auditory sequences of 16 tones (Lipton & Spelke, 2005). By 9 months of age, infants are not only able to keep track of different numerosities but are also able to add and subtract those numerosities (McCrink & Wynn, 2004). That is, if they see five objects go behind a screen followed by another five objects, they expect to see ten objects when the screen is lowered, not five, as evidenced by a difference in how long they look at each outcome.
How do infants' number representations compare to the number representations of children and adults? Although some (e.g., Gallistel & Gelman, 2003) have argued that infants' representations form the basis of all subsequent mathematical knowledge, others (e.g., Le Corre, Van de Walle, Brannon, & Carey, 2006) have argued that these representations are too imprecise to support an understanding of integers and the operations defined over them. Evidence of such imprecision comes from the finding that although 6-month-old infants can discriminate 8 dots from 16 dots and 8 tones from 16 tones, they cannot discriminate 8 dots from 12 dots or 8 tones from 12 tones. Imprecision of this nature decreases with age, but it never disappears altogether (Barth, Kanwisher, & Spelke, 2003), which has lead many to posit the existence of two distinct systems for representing number: (1) a nonverbal system capable of representing approximate numerosity, present from infancy through adulthood and shared with many nonhuman animals, and (2) a verbal system capable of representing exact numerosity, unique to humans and acquired around the age of 3 in the form of counting.
Because counting involves the mastery of a representational system not supported by core knowledge, learning how to count is not easy. In fact, studies of how children learn to count have revealed that children acquire this ability in a succession of small, discrete steps (Wynn, 1990; Carey, 2004). First, children learn their language's “count list,” or their language's list of words used to denote sets of increasing numerosity (e.g., “one,” “two,” “three”). Second, they learn how to apply this list to an array of objects by labeling each object in the array with one, and only one, word in the count list. Third, they learn that, when applying the count list to an array of objects, the last number word reached when counting corresponds to the cardinal value of the set. In other words, they learn that the word four refers not only to the fourth object encountered during the counting routine but also to the total number of objects encountered up to that point.
Evidence that children learn these three skills in stages, rather than in tandem, comes from dissociations in children's performance on simple numerical reasoning tasks. For instance, children are able to recite the count list long before they are able to apply it consistently to an array of objects. Likewise, children are able to apply the count list to an array of objects long before they realize that the last word reached when counting constitutes an answer to the question, “How many are there?” In fact, children go through a 6- to 9-month period during which time they are able to count a collection of objects but are not able to retrieve a particular number of objects from the collection. That is, when asked to retrieve a particular number of objects, they grab a handful at random and make no attempt to coordinate their knowledge of counting with their estimation of numerosity.
Of crucial importance to learning how to count is being exposed to a count list. Some cultures, such as the Pirhaha and Munduruku tribes of the Amazon rain forest, do not have count lists, and the members of those cultures cannot therefore keep track of exact numerosities (Gordon, 2004; Pica, Lemer, & Izard, 2004). For instance, when shown a collection of objects and asked to select a particular number (indicated nonverbally with fingers or sticks), Piraha adults tend to produce a close match, but not an exact match, to the requested number, implying that the only system they have for representing numerosity is the imprecise system they inherited via evolution.
Evidence for a core knowledge of living things comes not from studies of infant cognition or animal cognition but from studies of cross-cultural universals. These studies have revealed that, across cultures, children's early understanding of animals appears to be structured around three metaphysical commitments: (1) vitalism, or the belief that living things require energy in order to function; (2) essentialism, or the belief that living things possess an internal “essence” that determines their outward appearance and behavior; and (3) taxonomic relatedness, or the belief that living things can be organized into inferen-tially rich, multilevel hierarchies.
An early-developing commitment to vitalism is evident from studies of children's understanding of digestion, respiration, and circulation (e.g., Inagaki & Hatano, 1993; Morris, Taplin, & Gelman, 2000). In these studies, 5- and 6-year-old children are presented a variety of explanations for an activity like eating and are asked to select the best one. Although adults tend to prefer mechanistic explanations (e.g., we eat food “because we take the food into our body after the food is changed in our stomach”), children tend to prefer to vitalistic ones (e.g., we eat food “because our stomach takes in energy from the food”), regardless of their cultural upbringing. In line with these findings, preschool-aged children recognize that animals, but not artifacts, grow (Rosengren, Gelman, Kalish, & McCormick, 1991) and that consuming food is necessary for growth (Inagaki & Hatano, 1996).
Studies demonstrating an early-developing commitment to essentialism have focused not on children's understanding of metabolic processes but on their understanding of inheritance (Gelman & Wellman, 1991; Sousa, Atran, & Medin, 2002). In these studies, 3- and 4-year-old children are told stories about a baby animal who was raised by adult animals of a different species (e.g., a cow raised by pigs) and are asked to predict which properties it would possess as an adult: the biological properties of its birth parents (e.g., a straight tail and a diet of grass) or the biological properties of its adopted parents (e.g., a curly tail and a diet of slop). Regardless of cultural upbringing, children tend to predict that the baby animal will grow to possess the biological properties of its birth parents, not its adopted parents.
Evidence of a universal commitment to taxonomic relatedness comes from studies of how individuals from different cultures categorize the flora and fauna of their local ecologies (Atran, 1990; Berlin, 1992). These studies have found that individuals the world over classify living things into hierarchies that typically include the ranks of “kingdom” (e.g., plants, animals), “life form” (e.g., trees, birds), “generic species” (e.g., oaks, parrots), and “folk-specific species” (e.g., white oaks, African Gray parrots). These ranks constrain a variety of biological inferences— from inferences about how to extend known properties to novel organisms to inferences about how to extend novel properties to known organisms—for children and adults alike. Indeed, children as young as 2 consistently use their knowledge of taxonomic relations to constrain their extension of known properties to novel animals (Gelman & Coley, 1990).
Despite the above evidence for an early-developing conception of living things, children have been shown to experience great difficulty grasping other aspects of biological knowledge, including the very meaning of the words alive and dead (Piaget, 1929; Carey, 1985). For instance, when children aged 8 and younger are quizzed on their knowledge of what is alive and what is not, they classify many things that are alive (e.g., flowers, tress, bugs, worms) as “not alive” and many things that are not alive (e.g., the sun, the wind, clocks, fire) as “alive.” Moreover, children of this age are reluctant to extend properties true of all living things (e.g., has cells, has babies, gets sick) to plants and insects.
These misconceptions suggest that children do not initially understand life as a process of maintaining and regulating bodily functions and death as the cessation of that process. Consequently, they confuse life with ani-macy, observability, or functionality, and they confuse death with inanimacy, unobservability, or nonfunction-ality. Acquiring a correct conception of life appears to be tied to acquiring a mechanistic conception of biological functioning. Support for this claim comes from a study by Slaughter and Lyons (2003), in which children were questioned on their beliefs about death both before and after a teaching intervention designed to impart a mechanistic understanding of the internal workings of the human body. Before the teaching intervention, children revealed a number of misconceptions about the nature of death (e.g., that death can be avoided, that death can be reversed). After the teaching intervention, these same children revealed significantly fewer misconceptions, even though the intervention itself did not broach the topic of death.
Each of the developmental transitions described above share at least two commonalities. First, all three transitions involve overcoming structural limitations in the architecture of core knowledge, whether they be limitations in the information used to reorient oneself in the environment, limitations in the precision with which numerosities are represented, or limitations in the properties used to identify living things. Second, all three transitions involve the acquisition of culturally constructed knowledge, whether it be knowledge of the words left and right, knowledge of a count list, or knowledge of the inner workings of the human body.
Commonalities aside, each developmental transition exemplifies a slightly different type of knowledge acquisition. For instance, children's transition from geometry-based reorientation to landmark-based reorientation is more strategic than conceptual in nature, for this transition involves learning to attend to a spatial relationship that had previously been neglected (i.e., “left of the blue wall”) rather than learning to conceptualize space in an entirely new way. Children's transition from a vitalistic biology to a mechanistic biology, on the other hand, is more conceptual than strategic in nature, for this transition involves learning to conceptualize living things in an entirely new way (i.e., as self-sustaining, self-regulating machines) rather than learning to attend to a particular property of living things that had previously been neglected. The transition from an approximate representation of number to an integer-based representation of number also exemplifies a conceptual change, though the kinds of concepts that change within the domain of number (i.e., nominal-kinds concepts) are very different from the kinds of concepts that change within the domain of biology (i.e., natural-kinds concepts. Whether, and how, such a difference matters to the process of conceptual change itself has yet to be determined.
Atran, S. (1990). Cognitive foundations of natural history. Cambridge, England: Cambridge University Press.
Atran, S., Medin, D., Lynch, E., Vapnarsky, V., Ucan Ek', E., & Sousa, P. (2001). Folkbiology doesn't come from folkpsychology: Evidence from Yukatek Maya in cross-cultural perspective. Journal of Cognition and Culture, 1(1), 3–42.
Barth, H., Kanwisher, N., & Spelke, E. (2003). The construction of large number representations in adults. Cognition, 86, 201–221.
Berlin, B. (1992). Ethnobiological classification. Princeton, NJ: Princeton University Press.
Carey, S. (1985). Conceptual change in childhood. Cambridge, MA: MIT Press.
Carey, S. (2004). Bootstrapping and the origin of concepts. Daedalus, 59–60.
Cheng, K. (1986). A purely geometric module in the rat's spatial representation. Cognition, 23, 149–178.
Gallistel, C. R. (1990). The organization of learning. Cambridge, MA: MIT Press.
Gallistel, C. R., & Gelman, R. (2003). Nonverbal numerical cognition: From reals to integers. Trends in Cognitive Sciences, 4(2), 59–65.
Gelman, S. A., & Coley, J. D. (1990). The importance of knowing a dodo is a bird: Categories and inferences in 2-year-old children. Developmental Psychology, 26(5), 796–804.
Gelman, S. A., & Wellman, H. M. (1991). Insides and essences: Early understanding of the nonobvious. Cognition, 38, 213–244.
Gordon, P. (2004). Numerical cognition without words: Evidence from Amazonia. Science, 306, 496–499.
Gouteaux, S., & Spelke, E. S. (2001). Children's use of geometry and landmarks to reorient in an open space. Cognition, 81, 119–148.
Gouteaux, S., Thinus-Blanc, C., & Vauclair, J. (2001). Rhesus monkeys use geometric and nongeometric information during a reorientation task. Journal of Experimental Psychology: General, 130, 505–519.
Hermer, L., & Spelke, E. S. (1994). A geometric process for spatial orientation in young children. Nature, 370, 57–59.
Hermer, L., & Spelke, E. S. (1996). Modularity and development: The case of spatial reorientation. Cognition, 61, 195–232.
Hermer-Vazquez, L., Moffet, A., & Munkholm, P. (2001). Language, space, and the development of cognitive flexibility in humans: The case of two spatial memory tasks. Cognition, 79, 263–299.
Hermer-Vazquez, L., Spelke, E. S., & Katsnelson, A. (1999). Source of flexibility in human cognition: Dual-task studies of space and language. Cognitive Psychology, 39, 3–36.
Inagaki, K., & Hatano, G. (1993). Young children's understanding of the mind-body distinction. Child Development, 64(5), 1534–1549.
Inagaki, K., & Hatano, G. (1996). Young children's recognition of commonalities between animals and plants. Child Development, 67, 2823–2840.
Le Corre, M., Van de Walle, G., Brannon, E. M., & Carey, S. (2006). Revisiting the competence/performance debate in the acquisition of the counting principles. Cognitive Psychology, 52(3), 130–169.
Lipton, J. S., & Spelke, E. S. (2003). Origins of number sense: Large number discriminations in human infants. Psychological Science, 14(5), 396–401.
Margules, J., & Gallistel, C. R. (1988). Heading in the rat: Determination by environmental shape. Animal Learning & Behavior, 16(4), 404–410.
McCrink, K., & Wynn, K. (2004). Large number addition and subtraction by 9-month-old infants. Psychological Science, 15(11), 776–781.
Morris, S. C., Taplin, J. E., & Gelman, S. A. (2000). Vitalism in näıve biological thinking. Developmental Psychology, 36(5), 582–595.
Piaget, J. (1929). The child's conception of the world. London: Routledge and Kegan Paul.
Pica, P., Lemer, C., & Izard, V. (2004). Exact and approximate arithmetic in an Amazonian indigene group. Science, 306, 499–503.
Rosengren, K. S., Gelman, S. A., Kalish, C., & McCormick, M. (1991). As time goes by: Children's early understanding of biological growth. Child Development, 62, 1302–1320.
Slaughter, V., & Lyons, M. (2003). Learning about life and death in early childhood. Cognitive Psychology, 46, 1–30.
Sousa, P., Atran, S., & Medin, D. (2002). Essentialism and folkbiology: Evidence from Brazil. Journal of Cognition and Culture, 2(3), 195–223.
Sovrano, V. A., Bisazza, A. & Vallortigara, G. (2002). Modularity and spatial reorientation in a simple mind: Encoding of geometric and nongeometric properties of a spatial environment by fish. Cognition, 85, B51–B59.
Sovrano, V. A., & Vallortigara, G. (2006). Dissecting the geometric module: A sense linkage for metric and landmark information in animals' spatial reorientation. Psychological Science, 17(7), 616–621.
Spelke, E. S. (2000). Core knowledge. American Psychologist, 55, 1233–1243.
Stich, S. (1975). Innate ideas. Berkeley, CA: University of California Press.
Wang, R. F., Hermer, L., & Spelke, E. S. (1999). Mechanisms of reorientation and object location by children. Behavioral Neuroscience, 113(3), 475–485.
Waxman, S., Medin, D., & Ross, N. (2007). Folkbiological reasoning from a cross-cultural developmental perspective: Early essentialist notions are shaped by cultural beliefs. Developmental Psychology, 43(2), 294–308.
Wynn, K. (1990). Children's understanding of counting. Cognition, 36, 155–193.
Xu, F., & Spelke, E. S. (2000). Large number discrimination in 6-month-old infants. Cognition, 74, B1–B11.