Early Concept Learning In Children Research Paper

1. Traditional Views Of Concept Formation

Due to the writings of the Swiss psychologist, Jean Piaget (e.g., 1936), for most of the twentieth century it was assumed that children do not begin to form concepts until late in the second year. Piaget claimed that infants know about objects only in terms of sensorimotor activity; they were thought to be able to recognize objects and to know how to interact with them, but not to be capable of classifying them or thinking about them in their absence. Because this view was so widespread little research on the foundations of concepts in infancy was conducted. Typically, research on early concept formation was carried out on newly verbal children, which meant that concept learning and language acquisition were often confounded.

The most common words for objects in daily speech are at what is called the ‘basic level’ (Mervis and Rosch 1981), that is, words such as ‘dog’ and ‘spoon’ as opposed to ‘animal’ and ‘utensil.’ Because these common words for objects are the ones that children most often hear, they are the first object words that children learn. However, it was then assumed that the language reflects fairly closely their level of conceptualization. Thus, children were said first to form basic-le el concepts such as dog and spoon and only later to form more difficult superordinate concepts such as animal and utensil. However, although young children do not often use superordinate words, they often overextend words such as ‘dog’ to refer to all animals, which does not fit well with the notion that they do not have some concept of animal.

2. Current Theory And Research

2.1 Methods Of Study

As preverbal infants have come to be studied, the traditional view of concept formation has begun to change. Infant research not only removes the confound between learning concepts and learning language, it also enables examination of concept formation at a younger age, thus providing information about the earliest foundations of the conceptual system. This research effort is still young and there is a great deal yet to be discovered about the first concepts. However, several useful techniques have been developed. These include the object-examination task, in which infants are given a series of little models of objects from one class to examine, for example a series of animals, and then an object from another class, for example, a car.

If infants examine the car longer than still another animal, it indicates the infants think that a car is different. Using this task it was found that by nine months of age infants differentiate animals, vehicles, plants, and furniture, but make relatively few distinctions within these classes. For example, they do not distinguish dogs from cats or rabbits nor tables from chairs or beds (see Mandler 1998 for a summary). Other research showed that even younger infants are sensitive to the differences in physical appearance of different kinds of land animals as well as different kinds of furniture, so the failure to categorize them on the object-examination task indicates a failure to make conceptual divisions within these domains. Thus, even though young infants are able to perceptually categorize dogs as different from cats, they do not at first treat them as if they were different kinds of things.

The object-examination task is suitable for testing infants as soon as they begin to manipulate objects (about six months). Another test, suitable from about twelve months, is the sequential-touching task. This task is an instruction-less version of the standard classification test in which children are instructed to put together objects that are the same kind of thing. In the sequential-touching task, infants are presented with a number of objects from two classes simultaneously, and encouraged to play with them. The number of times infants touch objects from the same class in a row (run length) is measured and compared to chance run lengths. This test provides data similar to that from the object-examination test, indicating that even in the second year infants are not making many subdivisions in the large classes of animals, vehicles, furniture, and utensils.

A particularly useful test of early concept formation is the generalized imitation task, suitable from about ten months of age. Infants are shown an event acted out with little models, such as a dog being given a drink from a cup. They are then encouraged to imitate what they have just seen, but are given the cup and two different objects (for example, a different dog and a cat, or a rabbit and a car). Their choice of object to use to imitate drinking is recorded. This technique, by using the generalizations that infants are willing to make, is useful to test the breadth of early concepts. It is also useful to determine how preverbal infants have interpreted the events they observe. By varying the choices provided for infants to use for their imitations, one can narrow in on what they have understood. For example, if one-year-olds are shown a little model of a dog drinking and then given a different dog and a rabbit, they are equally likely to choose the rabbit as the dog as their first choice of animal to give a drink. On the other hand if the infants are given another dog and a bird, they are more likely to use the dog first for their imitations and only later give the bird a drink. This kind of finding suggests the infants interpreted the dog as a land animal, rather than as a dog or more generally as an animal. However, their willingness to give the bird a drink as a second choice, indicates the breadth of the generalizations they have made about drinking (Mandler and McDonough 1998).

2.2 Some Early Object Concepts

These different techniques converge in showing the global nature of the early concepts that have been studied. By nine months of age (and perhaps earlier) infants make distinctions among the domains of animals, vehicles, plants, and furniture, but within these domains make few distinctions at the basic level. For example, infants generalize any animal behavior to all other animals (including drinking and sleeping to fish), and generalize car keys as applicable to any vehicle (including forklifts and airplanes). Near the end of the first year they have learned some of the functions of ordinary household objects, but still at a global level. For example, they have learned something about the function of containers (Baillargeon 1995) but they interpret both cups and frying pans as appropriate containers from which to drink (Mandler and McDonough 1998).

Gradually during the second year infants begin to narrow down their concepts of common household objects until they begin to approximate basic-level concepts. In the studies to date, the earliest differentiation of broad concepts into subclasses occurs in the vehicle domain, followed by household artifacts, with differentiation of animals and plants occurring at a much slower rate. This ordering seems likely to be culturally determined, but this remains to be studied. In any case, by the time they are about three, children are more apt to base their generalizations on basic-level classes rather than on larger ones and in that sense they are approaching adult functioning. However, right from the beginning inductive inferences are determined by conceptual meaning, rather than physical appearance. The main difference between these early concepts and those of later years is the sophistication of the theories that underlie them (see Carey 1985).

Overall, the general principle on which the conceptual system is built in the first two years is that broad concepts gradually become subdivided into finer ones (even though there may be a few more individualized concepts from an early age). This principle explains why the conceptual system (in so far as objects are concerned) is strongly hierarchical. Every time a new distinction is learned it is done on the basis of forming a subdivision in a more inclusive class. This order of acquisition may also partly determine the order of breakdown of the conceptual system following brain damage, in which basic-level concepts disappear before more foundational ones (Patterson and Hodges 1995).

2.3 The Nature Of The Earliest Concepts

Relatively little is yet known about the kinds of information infants use to form their first concepts. It is known that infants parse the world into coherent objects, and make a distinction between animate and inanimate objects at least from two to three months of age. It seems reasonable to assume that the basis for this distinction lies in some appreciation of the different roles that animate and inanimate objects play in events. Infants are sensitive to the difference between caused and noncaused motion and to the contingent movements that characterize interactions among animates. Hence, the earliest concept of animal is likely to be something like an object that begins to move by itself and that interacts with other objects contingently and/or from a distance. Similarly, the earliest concept of artifact or inanimate thing may consist of an object that either does not move at all or, if it moves, only does so when another object comes in contact with it, and which does not interact with other objects from a distance. This distinction presumably forms the initial conceptual division of the world into two kinds of objects, from which the further division into more detailed concepts proceeds. In this sense the construction of the conceptual system lags behind the perceptual system, which is capable of making very fine distinctions from quite early in infancy. The conceptual system begins with a few crude, and probably event-related, distinctions and only gradually subdivides these into finer categories.

Bibliography:

1. Baillargeon R 1995 Physical reasoning in infancy. In: RoveeCollier C, Lipsitt L (eds.) Advances in Infancy Research. Ablex, Norwood, NJ, Vol. 9, pp. 305–71
2. Carey S 1985 Conceptual Change in Childhood. MIT Press, Cambridge, MA
3. Gelman S A, Wellman H M 1991 Insides and essences: early understandings of the nonobvious. Cognition 38: 213–44
4. Leslie A M 1988 The necessity of illusion: perception and thought in infancy. In: Weiskrantz L (ed.) Thought Without Language. Oxford University Press, Oxford, pp. 185–210
5. Mandler J M 1998 Representation. In: Damon W (ed.) Cognition, Perception, and Language, Handbook of Child Psychology. 5th edn. J Wiley, New York, Vol. 2 5th edn, pp. 255–308
6. Mandler J M, McDonough L 1998 Studies in inductive inference in infancy. Cognitive Psychology 37: 60–96
7. Mervis C B, Rosch E 1981 Categorization of natural objects. Annual Review of Psychology 32: 89–115
8. Patterson K, Hodges J R 1995 Disorders of semantic memory. In: Baddeley A D, Wilson B A, Watts F N (eds.) Handbook of Memory Disorders. Wiley & Sons, Chichester, pp. 167–86
9. Piaget J 1936 La Naissance de l’Intelligence Chez l’Enfant. [Engl. transl. 1952 The Origins of Intelligence in the Child. International Universities Press, New York]