The cognitive approach to creativity emphasizes the processes involved in producing effective novelty, as well as the control mechanisms that regulate novelty production, and the structures that result. Merely novel structures display surprisingness and incongruity, to be sure, but they must also be meaningful and practicable to be effective. There are no special processes or control mechanisms unique to the production of effective novelty, but metasystematic operations are particularly favorable for it. Effective novelty can be produced at lower levels of cognitive development, but children's creativity is likely to differ qualitatively from that of adults. Although the cognitive approach takes little account of motivation, personality, or the social environment, it provides an operationizable definition of some aspects of creativity, and offers insights into what needs to be fostered to promote it.
Creativity and Cognition
The modern era of creativity research began in 1950 with Guilford's address to the American Psychological Association (Guilford, 1950). He extended the then existing (and indeed, still dominant) concept of intellectual giftedness as synonymous with conventional intelligence by taking creativity into account. The importance of creativity as an element of giftedness is now so widely accepted that Cropley (1994) regarded it as essential for true giftedness. Indeed, creativity is not exclusively a cognitive phenomenon. Guilford himself gave considerable emphasis to the "creative personality," and the study of creativity and personality was prominent in early research (e.g., Barron, 1963, 1969). However, Guilford's distinction between convergent and divergent thinking established the importance of the cognitive approach to understanding creativity from the very beginning.
Cognition is defined as concerned with the ways people obtain, organize, process, store, and use information. Thinking is a process through which symbols are constructed, revised, linked up with other symbols, reorganized, and applied to abstract or concrete situations. It involves processes such as exploring, recognizing, organizing, and coding, and structures (internal representations of the external world) such as patterns, categories, networks, and systems, that result from the processes. To these may be added control mechanisms such as styles, strategies or tactics that guide the processes and affect the kinds of structures they lead to. This article reviews these processes, control mechanisms, and structures in creative thinking, as well as examines the way these aspects of cognition develop from childhood to adulthood.
Cognitive Definition of Creativity
Almost 50 years ago, Morgan (1953) reviewed a large number of definitions of creativity and showed that the single common element was novelty. Cognitive definitions of creativity usually follow this tradition and emphasize production of novelty as the crucial aspect. However, simple novelty is not on its own sufficient for a satisfactory definition of creativity, as will be emphasized in the following discussion.
Novelty can be produced in the form of mere self-expression (daubing paint on paper, writing text in any way that pleases the writer, or picking out notes at random on the piano) or of simple production of variability (doing things differently from the usual regardless of accuracy, meaning, sense, significance, or interestingness). However, it can also satisfy technical, professional, aesthetic, or scholarly criteria. The latter kind of novelty produces a shock of recognition in observers that generates "effective surprise" (Bruner, 1967). It may be regarded as involving effective novelty. Without effectiveness, novelty is more likely to be a matter of the pseudocreativity that characterizes simply being outrageous or nonconformist, or the quasicreativity that occurs in, for instance, daydreaming (Heinelt, 1974). These have elements of creativity such as fantasy, but lack task relevance and similar properties that are necessary for creative giftedness.
Finke, Ward, and Smith (1992) distinguished between two kinds of processes leading to novelty. It involves on the one hand, generating novel cognitive structures via retrieving, associating, synthesizing, transforming, and constructing analogies, and on the other, exploring the creative implications of new structures (e.g., attribute finding, interpreting, inferring, shifting context, hypothesis testing, and searching for limitations). When the first kind of process occurs, novelty is produced, to be sure, but without the second kind it is not effective (i.e., it does not lead to creativity).
Discussions of creativity frequently emphasize a product such as a work of art, a machine or a design, a production process, or a solution method. Products can also take the form of new ways of symbolizing an area of knowledge. Jackson and Messick (1965) distinguished between external criteria of the effectiveness of novel products (e.g., the appropriateness of the novelty to a problem or issue) and internal criteria such as logic, harmony among the elements of the product, and pleasingness. They thus added, what are to some extent, aesthetic criteria to the definition. Finally, social recognition or acclaim (e.g., Csikszentmihalyi, 1996) and ethicality (e.g., Grudin, 1990) are also necessary for novelty to earn the highly positive label, creativity. A detailed discussion of creative products involves issues that are partly professional/aesthetic, partly philosophical, and even partly economic in nature, and is beyond the limits of this article. For this reason, it focuses on processes, control mechanisms, and internal structures.
Limitations and Problems of the Definition
This definition is admirably straightforward, but raises the danger of reducing all creativity to a defined set of processes, regardless of the person involved, the level of that person's preexisting knowledge, the social, occupational, or educational context, or the goals and motives underlying the creativity. It is unclear whether novelty means novel in a specific time and place, in which case it is partly determined by the external context and not solely by the cognitive processes or the structures generated. It is also unclear whether a product must be novel for a specific person, in which case an ignorant person could most easily produce novelty, or novel for all societies in all history. For instance, transferring procedures from one worksetting, where they are familiar, to another where they are unknown, may produce effective novelty, but this is obviously different from inventing ideas previously unknown in any context. Case studies (e.g., Ghiselin, 1952; Weisberg, 1986) also indicate that some highly creative individuals regard the novelty they produce as obvious or even the only logical conclusion to be drawn from available data, i.e., it is conceivable that some effective novelty may not involve special forms of cognition at all.
The situation is also complicated by the fact that the term creativity is used in at least two ways. Data from case studies of people acclaimed as creatively gifted (e.g., Gardner, 1995; Simonton, 1994) focus on outstanding individuals with exceptional achievements in a specific field. By contrast, Nichols (1972) studied the unclaimed creativity of people who will never produce anything original or useful. In addition, both the general public and educational theorists, researchers, and practitioners are interested in creativity in children, although very few children have already achieved fame. Thus, the term creativity is used to refer both to the sublime gifts of a Michelangelo or an Einstein, but also to what Nichols (1972) called "creativity as a normally distributed trait." The latter might be called everyday creativity. The study of everyday creativity is very productive for discussions in an educational context, and is consistent with the goal of developing every student's potentials to the full.
Cognitive Approaches to Novelty Production
Divergent versus Convergent Thinking
The crucial distinction Guilford made was between convergent thinking, quickly equated with conventional intelligence, and divergent thinking, which was seen as the cognitive basis of creativity. Convergent thinking is oriented towards deriving the single best (or correct) answer to a given question from the available information. It is effective in situations where such an answer exists and needs to be recalled from stored information, or where the answer needs to be worked out from what is already known by conventional and logical search, recognition, and decision strategies. Divergent thinking, by contrast, involves producing new and possibly multiple answers from the available information. These may never have existed before, at least in the experience of the person producing the answer in question. There may be a large number of such answers and they may all be equally useful. The crucial point is that although both convergent and divergent thinking lead to knowledge production, there is a qualitative difference: Divergent thinking involves production of variability, convergent thinking production of singularity.
Other authors have also emphasized the role of thinking in creativity, often in the form of two contrasting strategies. Bartlett (1932), for instance, distinguished between "open" and "closed" thinking, Gestalt psychology between "reproductive" and "productive" thinking, and deBono (1970) between "linear" and "lateral" thinking (although this latter author was writing at a popular rather than scientific level). Rothenburg (1988) introduced the idea of "janusian" thinking, naming it after the Roman God Janus, who had two faces and could look in two directions at the same time, giving him the ability to deal simultaneously with pieces of information that would not normally occur together. Rothenburg also introduced the term "homospatial" thinking. This kind of thinking is able to unite apparently conflicting or mutually exclusive ideas, thus producing novelty.
The position adopted in the early years of creativity research was that divergent thinking (production of variability) and convergent thinking (production of singularity) are separate, more or less competing, or even mutually exclusive aspects of giftedness (see, for instance, Getzels & Jackson, 1962). However, in studies of achievement at school or university level, it has been shown that the most gifted students produce both singular and varied information (for a summary see Cropley, 1992). Research on practical creativity (e.g., Facaoaru, 1985) showed that engineers rated as creatively gifted produced a combination of facts and logic (singularity) and new ideas and unexpected suggestions (variability). Neither production of singularity nor production of variability alone leads to effective novelty. Both are needed, except under extraordinary conditions such as via the workings of chance. This raises the question of how the two combine.
Divergent and Convergent Thinking in the Production of Effective Novelty
The interaction between the two forms of information production is often explained via various threshold models. According to the best known of these, a minimum level of singularity (conventional, factual knowledge) is necessary before variability can be produced. A slight extension is the idea that. as ability to produce singularity approaches this threshold (indicated by an IQ of perhaps 130) from below, the possibility of production of variability rises (i.e. production of singularity and production of variability are positively correlated). Once IQ (regarded in this article as predominantly but not exclusively a measure of ability to produce singularity) passes the threshold, increases in IQ have no consequences for production of variability, which is then regulated by factors other than knowledge of facts. A further elaboration of this approach is the idea that production of singularity determines channel capacity (the amount and kind of information that reaches cognitive systems), the degree of production of variability then being determined by flexible and versatile handling of information delivered by the channel.
A further conceptualization of the interaction between production of variability and production of singularity is the style approach. According to this theory, singularity and variability do not directly influence each other. Both involve application of a superordinate ability to form abstract, general networks, knowledge matrices, and systems. Whether singularity or variability is produced depends upon the kinds of cognitive strategies and control mechanisms that direct knowledge production. This conceptualization of the interaction regards production of singularity (convergent thinking) and production of variability (divergent thinking) as styles for applying thinking power in dealing with information. The difference between them is qualitative rather than quantitative.
The style approach differentiates between level and style of cognitive functioning. Both gifted and nongifted individuals may prefer to apply their ability, regardless of level, to producing singularity, to producing novelty, or to some combination. An appropriate combination seems to be necessary for effective novelty. The style approach raises one of the objections to case studies of the highly gifted. By concentrating on successful individuals whose high creativity is widely acknowledged, they may confound style of thinking (divergent) with level of ability (high IQ), and as a result ignore production of novelty in people of average ability.
Coding and Networking in Novelty Production
To a new born baby, life must seem to consist of a flood of individual events occurring in a haphazard way. With increasing experience, people learn that what is going on around them has form and order, and it becomes possible for them to regulate their thinking by building inner structures according to the patterns they have perceived in the external world. This section deals with ways in which psychologists conceptualize the building of these inner representations of the external world. Strict adherence to previously learned ways of understanding experience inhibits production of novelty.
Associational hierarchies. Mednick (1962) stated that through experience, people learn a number of possible responses to any given stimulus. Those responses most frequently linked with a particular stimulus whenever it was encountered in the past have a high probability of being selected as appropriate when the stimulus is encountered once again. Responses seldom paired with the stimulus in the past have a low probability of being chosen. When the stimulus recurs in a new situation, most people select a response they have often made to this stimulus in the past. This means that people's reactions to familiar stimuli are consistent. However, it also means that their associations are repetitious, and that they tend to think in the same way over and over again. In other words, they do not create novelty. Mouse is a common associate to the stimulus word cheese, since these two ideas often occur together. Similarly, common associates to cheese are sandwich or milk. An uncommon associate (except to readers of Dr Suss) would be green, as in green cheese. A similarly remote associate to cheese would be moon (according to children's tales the moon is made of cheese). A person with a high preference for remote associates might create novelty by connecting both green and moon to cheese, or create even greater novelty by linking green and moon via their common (remote) association with cheese.
Category building. Bruner's (e.g., 1967) approach goes further. On the basis of experience, events that repeatedly occur together are recognized as belonging together. Initially, belonging together is understood on the basis of simple temporal or spatial proximity, or concrete resemblances such as similar color or shape. In the course of cognitive development, however, people come to understand that events have generalized abstract properties, and that these are the basis of belonging together. The abstract properties that are common to a number of individual exemplars define a category or concept such as the category of "weapon" or "tool" or "food." The process of assigning events to categories is referred to as coding. A new event is seen to have a distinctive pattern. When this pattern is judged to match the definitive properties of a particular category (i.e., pattern recognition occurs), the new event is encoded into that category. Provided the patterns match, the new event is treated as though it has all characteristics of the category, even characteristics that have not been directly observed. Thus, coding is a special form of production of novelty involving "going beyond the information given" (Bruner, 1967). Sloman (1998) identified the idea that "categories inherit properties from their superordinates" as possibly the most fundamental idea in cognitive psychology (although he went on to show that the phenomenon is not absolutely universal).
Networks. Different categories may share properties. For instance, weight, balance, a convenient length, rigidity, and portability are all properties of weapons, but also of bats and racquets used in sports such as baseball, cricket, hockey, or tennis, as well as of walking aids such as a walking stick or a crutch. The three categories overlap and form a network (for a relevant discussion of networks in thinking, see Anderson, 1976). The network makes it possible to break the boundaries of a particular category, in the present example by using a baseball bat as a walking stick, and thus producing novelty.
The concept of networks of interlocking categories was stated somewhat differently by Koestler (1964), who saw knowledge as existing in matrices. Information processing usually involves linking elements from within the same matrix and thus produces no novelty. By contrast, when two matrices are linked via "bisociation," or three are "trisociated," variability is produced. In the extreme form, it would be possible to speak of "omnisociation," where, in principle, all matrices could be linked to one another. The potential for production of novelty would then be very high.
The effect of the context. Whether coding and construction of networks produces singularity, mere variability, or effective novelty is strongly influenced by contextual factors (for instance, in a library vaguely defined paper objects are likely to be coded as books, while a hungry person is likely to code a spherical object about the size of a tennis ball, yellowish red in color, and with indentations like those on a golf ball on its surface as an orange). Contextual factors act like a key that increases a certain category's accessibility by unlocking it and leaving the door ajar. Since the codings that result from using these categories are those that have occurred in the past in the context in question, they are commonplace and entail no novelty. They may, of course, be sensible and socially acceptable, as well as being readily available, so that coding into these categories trades off production of novelty for ease of processing and avoidance of "cognitive strain" (Bruner, 1967).
The context predisposes people to code new events into certain categories. This induces a set, i.e., a tendency to see the world in fixed ways. The conventional coding of a wristwatch would be to class it as a device for telling the time. However, it is possible to break sets and code stimuli into unexpected categories such as coding the watch as an object with weight instead of as a timepiece. This recoding of the watch draws attention to previously ignored properties that it possesses, and the person is then in a position to use the watch to solve problems where a weight is required but time of day is irrelevant, thus producing novelty.
Creativity and Cognitive Development
As people gain more experience (i.e., they develop psychologically), there is a switch in thinking away from focusing on the immediate and concrete properties of real physical objects to their general properties, and then to their symbolic meaning. This is the core of the process of cognitive development. At the highest level, operations can be carried out on symbols alone-the concrete object or experience is no longer necessary. For example, at the concrete level, a dove is white, has feathers, and can be seen in trees. An eagle is brown, also has feathers, and lives high up in the crags. At a more general level, a dove and an eagle are both birds. At a symbolic level, a dove means "peace," an eagle "war." Finally, at the level of operations on symbols, an eagle wearing a World War I spiked helmet pecking a dove with the head of the Statue of Liberty means war between Germany and the USA.
When categories are based on simple, concrete information, the category boundaries are clearcut and impermeable (boundary breaking is difficult or extremely stereotyped), and the networks at best rudimentary. The result is that production of novelty is difficult or is limited to simple departures from the usual codings. Categories based on abstract, complex properties, on the other hand, offer permeable category boundaries, readily permit bi- or trisociation, and encourage the building of complex networks, thus favoring production of effective novelty. Coding based on higher levels of generalization and abstraction offers increased chances of effective novelty.
What these considerations mean that the production of effective novelty is facilitated by possession of higher order cognitive systems. The emergence of higher order cognition is the result of a systematic process of development that is linked to increasing experience with the external world. The stages in this orderly process of development are outlined below. Since adults are, as a group, more cognitively mature than children, differences are expected between the two groups in production of novelty. The nature of these differences is also discussed below.
Stages in Cognitive Development and Production of Novelty
The best known description of cognitive development is that of Piaget, who identified four stages: the Sensorimotor Stage, the Preoperational Stage, the Stage of Concrete Operations, and the Stage of Formal Operations. Case (1978) suggested that it is possible to identify a fifth stage involving recognition of second-order relations, while Sternberg and Dowing (1982) extended this to the third-order level (analogies between analogies). Commons, Richards, and Kuhn (1982) also argued for a fifth stage, the stage of Systematic Operations, at which operations are carried out on classes to build systems. These authors also recognized a sixth stage of Metasystematic Operations involving operations on systems. This differentiation of formal operational thinking offers promise for understanding the production of effective novelty and thus of creative giftedness. For instance, what Mumford and Gustafson (1988) called minor creativity (the novel application of the already known) could occur at the level of concrete operations, whereas major creativity (development of new principles) seems to imply systematic or metasystematic operations.
The five levels of creativity described by Taylor (1975) also help to clarify the relationship between cognitive development and creativity. According to him, the lowest level of creativity involves "expressive spontaneity." This involves unhindered productivity, without regard to reality. It can occur on the basis of concrete operations or even preoperational or sensorimotor thinking. In this case, the novelty is produced in the form of observable behaviors or other concrete products, is probably novel only for the producer (and perhaps adoring parents or enthusiastic teachers), and is based on concrete, physical properties of real objects. Nonetheless, novelty can be produced at this level. Indeed, as Tweney (1996) pointed out, some forms of pattern recognition may be more effective via presymbolic processes.
The next level in Taylor's model, "technical creativity," involves unusual mastery of knowledge, techniques or skills, while "inventive creativity" makes use of the already known in new ways. Both of these activities require at least concrete or formal operational thinking. However, they produce novelty based on direct application of what already exists, and are bound by the structures and rules of the existing. The final two levels are "innovative creativity" (extension of existing systems) and "emergent creativity" (development of new systems), i.e., the novelty produced is at the level of classes or systems. There may, of course, be concrete products, but the crucial point is that these forms of production of variability extend or change existing ways of symbolizing an area of knowledge. The production of variability at this level seems likely to require systematic and metasystematic operations.
Creativity in Children versus in Adults
If production of effective novelty requires formal operations or even systematic or metasystematic operations, it is obvious that people who have not yet reached these levels of cognitive development could not be creative. This is the normal state of affairs in children. On the other hand, it is customary to speak of children's creativity, even if the term is applied to rudimentary applications of fantasy such as crude drawings, simple, highly stereotyped stories or everyday pretend games (see the earlier reference to quasicreativity). In a certain sense, even a crude, childish drawing of, let us say, a human head does indeed produce something new, since prior to completion of the drawing the piece of paper on which it was drawn was blank. However, this novelty is, in the vast majority of cases, not effective.
Rosenblatt and Winner (1988) described three phases of children's creativity: the preconventional phase (up to the age of about 6-8 years), the conventional phase (from age 6-8 to about 10-12), and the postconventional phase (from about 12 years of age and extending into adulthood). Preconventional production of novelty derives from preoperational thinking. It displays spontaneity and emotional involvement and may lead to aesthetically pleasing products, but it is dominated by perception (especially visual) of the immediate concrete environment. As Runco, Okuda, and Thurston (1991) put it, it is "environmentally cued."
Conventional production of novelty involves operational thinking, but becomes increasingly rulebound as critical and evaluative skills develop, with the result that the novelty produced is stilted and conforms to external standards (i.e., it is not effective). Postconventional novelty production involves formal operational thinking and, with increasing cognitive development, operations on classes and systems, while it is enriched by, for instance, increased experience with the external world. The crucial element in postconventional production of novelty is that the individual takes account of external constraints and conventional values (necessary for effectiveness), but is able to produce novelty despite this. The difference between the preconventional phase and the postconventional can be stated rather baldly by saying that, in the main, children under about 10 produce novelty as a result of being ignorant of the constraints of the external world, a state of affairs that is not favorable to effectiveness. By contrast, people in the postconventional phase are familiar with these constraints, but are able to transcend them. For some writers (e.g., Smith & Carlsson, 1983) this means that children under about 10 cannot be creative.
However, according to Rosenblatt and Winner (1988), what is missing in the novelty production of children under about 10 is the regulatory element of (self-) evaluation. Their productions may be novel, spontaneous, uninhibited, and even aesthetically pleasing, but they often lack accuracy and adaptation to the constraints of reality (i.e., effectiveness). Vygotsky (see Ayman-Nolley, 1992) also emphasized that children lack control over the novelty they produce, although this did not lead him to conclude that children cannot be creative. Taken together with their limited knowledge base, simpler interests and motives, and less complex cognitions, this weakness of control functions leads, according to Vygotsky, to creativity that is "less rich" than that of adults. Adults' creativity does not lack the subjective elements of children's, but is enriched by adults' objectivity.
Mechanisms Guiding Cognitive Processes
Thinking, reasoning, and problem solving can be regarded as cognitive processes that use existing information to produce further information. (In the case of creativity, this further information would contain effective novelty.) This involves selecting from among the masses of information available at any moment (i.e., perception is not simply a passive acceptance of everything that impinges on the senses or is already stored in the mind); relating new information to what is already known; combining elements of new and old information; evaluating newly emerging combinations; selectively retaining successful combinations (which may then function as new information, returning the process to the phase of relating elements of information); communicating the results to others.
In discussing the application of the processes just mentioned to form new cognitions, some contemporary authors argue that novel combinations result from random associations. Simonton (1988) introduced the chance configuration model. According to this, production of novelty requires possession of a large number of mental elements (pieces of information, memories, ideas, concepts). When these are combined randomly, a great deal of novelty is produced and, every now and then a configuration occurs (a happy combination that contains precisely what is needed, i.e., effective novelty). However, many authors argue that information processing leading to effective novelty cannot proceed by "brute force" (Simon, 1989) in a process of perceiving, blindly associating, and occasionally recognizing that a new combination is the required solution. This would lead to a "combinatorial explosion" (Simon, 1989), involving huge numbers of empty trials (Altshuller, 1984) and leading to cognitive strain (Bruner, 1967). Thus, combinatorial processes must be guided in some systematic way if they are to produce effective novelty. Several approaches to understanding these guiding mechanisms are summarized.
Cognitive styles and production of novelty. Cognitive styles are consistent and stable differences between people in the way they obtain information from the world around them, utilize their experience (sort, organize and recall information), and cope with demanding situations. (For a detailed discussion, see, for instance, Messick, 1994.) Cognitive styles are often stated in the form of bipolar dimensions such as leveling versus sharpening, focusing versus scanning, field dependence versus field independence, preference for wide versus narrow categories, or seeking cognitive complexity versus seeking simplicity. Field dependence, wide categories and preference for complexity are favorable to creativity. Although their work does not belong to the mainstream of cognitive psychology, Myers and McCaulley (1985) discerned four bipolar types that in some ways resemble cognitive styles: Extraversion versus Introversion; Sensing versus Intuiting; Thinking versus Feeling; Judging versus Perceiving. These authors reported that creativity is related to the Sensing-Intuiting dimension, with creatively gifted people very frequently being intuiters. Other researchers have shown that creativity is related to extraversion.
Thinking strategies in production of effective novelty. Thinking strategies can be conceptualized by borrowing two further concepts from Piaget: assimilation and accommodation. Assimilation involves fitting new information in with existing cognitive structures; accommodation involves recognizing that current structures are not adequate for understanding the new situation. Intuitively, production of variability is related to accommodating, production of singularity to assimilating. Martinson (1995) referred to consistent tendencies in individuals to rely on the old, on the one hand, or to see the need to generate the new, on the other, calling them "stable, cognitive, strategic dispositions." Some people consistently seek to deal with the new by reapplying existing knowledge and solution strategies (they think in an assimilative manner). This achieves cognitive economy, but is unfavorable to the production of novelty. Other people consistently deal with new situations by recognizing the need for new ways of dealing with them (they think accommodatively). This necessitates the production of novelty.
Early in the modern era, researchers attempted to identify the cognitive style or strategy that leads to effective novelty (e.g., Bloomberg, 1967). In my research on creativity and personality (e.g., Cropley, 1967), I linked it to field dependence and extraversion. Other studies emphasized preference for complexity (as against simplicity), openness to novel information, tolerance for ambiguous information, tolerance for discrepancies, and tolerance for incomplete information (e.g., Barron, 1963, 1969). More recently, however, as Puccio, Treffinger, and Talbot (1995) concluded, the relationship between cognitive processing strategies and production of effective novelty has come to be seen not as a matter of which strategies lead to it and which do not, but rather of how various styles and strategies are related to it. For instance, it could be speculated that production of variability would increase arousal by increasing uncertainty. In view of the known tendency for high arousal to narrow and rigidify thinking, production of novelty might well be easier under conditions of low initial arousal, especially in individuals with low tolerance for arousal. Issues like this suggest that there is a need for further research.
Intuition and creativity. Sometimes creative ideas seems to pop into the head of a particular person out of the blue. Ghiselin (1952) gave examples of famous creators who described their own production of effective novelty in this way. In his classical stage model of creativity, Wallas (1926) identified a stage of incubation, during which ideas seem to chum and work in the creative person's head until-apparently coming from nowhere-the required answer pops up. This is then frequently interpreted as indicating the existence of unconscious processes. The unconscious is a concept that is more characteristic of psychoanalytic discussions (e.g., Kubie, 1958) than of cognitive. Indeed, psychoanalytic creativity theory regards unconscious thinking (primary process thinking) as an important element in creativity. However, the psychoanalytic approach will not be discussed in detail here, although the idea of thought without conscious awareness is developed further in a cognitive context. Mumford and Whetzel (1996) concluded that sudden insights do occur, but explained them in terms of cognitive mechanisms in four areas: representation, constraints, structure, and selective processing.
Some cognitive writers such as Simonton (1988) and Weisberg (1986) reject the notion of unconscious processes, but others acknowledge the role of intuition (for a relevant discussion, see Policastro, 1995). The basic idea of intuition in a cognitive sense is that, early in the process of production of variability, people sometimes see the novelty for which they are striving in the form of a rough outline of the solution. Their task is then that of defining and refining this into effective novelty. The outline may be acquired via implicit learning, i.e., learning that occurs without the learner being aware of it, for instance in the course of everyday life. To take a simple example, during the course of tiding to work every day in a public transport bus and sitting just behind the driver, a person might learn a great deal about the work of a busdriver, without ever having thought of the ride to work as a learning experience.
Implicit learning leads to tacit knowledge that people do not know they possess. Such knowledge can prestructure thinking about an issue. For instance, engineers exposed to the situation above might already have acquired information about redesigning buses and have stored this in the form of tacit knowledge. Upon being hired to design a new bus, they would already possess a preliminary framework that could suggest where the required answer might be found or approximately what the eventual solution might look like. When this leads to production of effective novelty it is experienced as intuition. When it narrowly focuses information processing and leads to production of singularity it is experienced as a set or corset that blocks ideas.
Heuristics in production of effective novelty. Favorable associational hierarchies, permeable category boundaries, and loosely structured networks promote production of novelty. However, the number of individual ideas that could be produced in any given situation is very large: For instance, the total number of situations that could in theory emerge during the course of a chess game is larger than the total number of atoms in the universe. Despite this, champion players produce effective novel moves. Since it is known that human information processing capacity is limited (Miller, 1956), such players must make use of tactics for dealing with very large quantities of information very quickly (for a more detailed discussion, see Chi, Glaser, & Farr, 1988). Thus, the first step in producing effective novelty is to produce variability, to be sure, but to confine this to a number of alternatives that is large in comparison with singularity but small in comparison with the number theoretically possible through blind associating. The second step is to identify the effective novelty in this mass of material. These two steps require heuristics that consist, on the one hand, of rules or techniques for recognizing from the start blind alleys or their opposite, particularly promising lines of attack and, on the other, of evaluation skills (for example, blueprints or patterns of what an effectively novel solution might look like, and techniques for matching the present state with this blueprint).
Metacomponents of production of effective novelty: The properties of heuristics just outlined can be seen as elements of metacognition. This involves the executive processes (Resnick, 1987) in thinking which allow people to keep track of their own understanding of a situation. According to her, these processes include review of knowledge, organization of attention, and marshalling of resources. Steinberg (1988) emphasized further metacognitive processes such as defining the problem, selecting an appropriate set of processes for solving it, and combining these into a workable strategy. To these can be added evaluation of one's own progress, effective changing of course should the evaluation make this necessary, and sensing of one or more promising alternatives.
Although she was discussing higher order thinking, Resnick's (1987) analysis showed how metacognitive processes are connected with creativity. According to her, thinking favorable for production of effective novelty would be nonalgorithmic (no fixed pathway is specified in advance) and complex (oriented towards achieving multiple solutions, and guided by multiple criteria). Evaluation processes would have to be self-regulatory, tolerant of uncertainty, and based on nuanced judgment (i.e., subjective interpretation by the thinker would play an important role). The opposite of this would be thinking guided towards acquiring set knowledge that had already been laid down in advance by somebody else.
Expertise, Problem Solving, and Effective Novelty
The ideas just advanced may seem to imply that intensive knowledge of a field can block production of effective novelty. Indeed, research has shown that expertise does not always facilitate novelty production. Martinson (1995) argued that the effect is U-shaped. A very low level of familiarity with a field and consequent absence of preorganizing or constraining factors might seem, at first glance, to be favorable for novelty production. However, it can lead to a combinatorial explosion and a flood of simple variability. On the other hand, a very high level of familiarity with a field and with existing solution strategies can preorganize thinking so effectively that it leads only to production of singularity (tried and trusted, correct approaches).
One area in which expertise versus naivete may play an important role in production of effective novelty is problem solving. This is particularly noticeable in the case of problem recognition and reformulation. Experts seem to spend more time than beginners on (re)formulating problems (Rostan, 1994), and less on searching for solution strategies. Experts code (reformulate) problems on the basis of recognition of their definitive patterns. Once they have encoded the problem into a category, the process of going beyond the information given means that the solution is at hand, since this is part of the category. Thus, in a sense, experts already know the solution, and the task is to encode the problem. This is done by means of identifying its deep structure (its systematic and metasystematic properties). Beginners, on the other hand, believe that they already know the problem, which they define on the basis of its surface structure (its immediate, concrete, unique properties). As a result, it is difficult to encode and they must spend more time trying to find a solution.
However, it is also possible that the narrow preorganizing that profound experience makes possible can hinder problem solving. A concrete example is to be found in the fate of a German chemist between the two world wars. He was interested in finding what would nowadays be called an "antibiotic." Each evening before leaving his lab he set out bacteria colonies on watch glasses so that they could grow during the night and he could attempt to kill them next day. After a while he noticed that his work was being impeded by the fact that the bacteria were all dead when he arrived in the morning. He observed that the colonies of dead bacteria all had mould spores on them, whereas those that survived did not. He concluded that the spores came from mould growing in dark corners of his lab and that this mould was killing the bacteria. He arranged for the entire room to be cleaned and decontaminated. This action was successful, and he reported in a chemistry journal that the mould had been eliminated without trace and he was now able to carry on with his research on killing bacteria. Unfortunately, he had not yet had any success in finding a chemical substance that was effective! Had he had expertise as a biochemist or botanist or biologist, or been able to break the set imposed by his training as a chemist, he might have won the Nobel Prize that later went to Sir Alexander Fleming.
From the beginning of the modern era, there has been steady interest in noncognitive aspects of creativity, including motivation (e.g., Amabile, 1983), personality (e.g., Helson, 1996), and the social environment (e.g., Csikszentmihalyi, 1996). Although not implying that such factors are unimportant, this article takes little account of them. Also of lesser importance in the article are the characteristics of creative products such as works of art, literature, or music, scientific discoveries, or new conceptual systems. The focus here is on thinking processes and on strategies and tactics that guide them to produce cognitive structures displaying effective novelty. The basic cognitive phenomena are seen in people at all ages and at different levels of intellectual ability. However, they frequently involve production of mere variability rather than effective novelty. Of great interest are the processes and control mechanisms that lead to effective novelty and thus permit creative giftedness.
In a fairly recent review, Finke, Ward, and Smith (1992) argued that, after the initial pioneering period, there has been insufficient attention to this issue, especially from the standpoint of practical procedures to foster creativity. Indeed, a brief examination of leading journals specializing in cognitive psychology (e.g., Cognition, Cognitive Psychology, and Memory and Cognition) indicates that the mainstream cognitive research and theory reported in them is dominated by detailed, relatively narrow, basic research. Thus, there are grounds for agreeing with Finke, Ward, and Smith that practically oriented, applied research on "creative cognition" needs to be carried out in order to study creative giftedness.
Despite this conclusion, the limitations of the cognitive approach already mentioned raise the question of its usefulness, especially within an educational orientation. In fact, the approach has a number of strengths. Cropley (1992) reported that some teachers interested in fostering creativity often have difficulty distinguishing between it and mere deviation from the conventional or even sheer misbehavior. As a result, they may set their sights too low and tend to accept unquestioningly any unconventional or surprising behavior for fear of suppressing creativity. Fostering creativity can then degenerate into pointless busywork. On the other hand, some teachers assume that fostering creativity involves producing little Einsteins or Michelangelos and set their sights too high, seeing no real hope of fostering it in school. Still others reject fostering creativity as elitism. Creative cognition provides a framework for clearing up some of these problems, for instance through analysis of the nature of the novelty involved in creative giftedness.
The cognitive approach also makes it possible to break creativity down into its constituent elements: processes, guiding strategies, and structures. This offers insights into the kinds of activity needed to foster it. The cognitive elements can also be operationalized and even measured using appropriate instruments, as in the recent program of Mumford and associates, who have designed tests of processes such as problem construction, information encoding, category selection, category reorganization, and category combination (see, for instance, Mumford, Baughman, Maher, Costanza, & Supinski, 1997).
The practical educational importance of such definition and operationalization has been reemphasized in recent research by Eisenberger and Cameron (1996). They showed that it is possible to develop creativity in schoolchildren, despite the pessimistic conclusions summarized by Cropley (1992), and to do this in such a way that the processes and properties fostered via special procedures endure over time and generalize to situations other than the training program. To achieve this, however, it is vital that the children know precisely what they are supposed to acquire, and that their teachers are able to recognize and reinforce the required behaviors when they see them (Eisenberger & Armeli, 1997). Simply, let us say, reducing the level of formality in the classroom is insufficient and, in any case, is too vague. The analysis presented in this article helps define and operationalize creativity in the required way, and thus makes a contribution to developing a rigorous foundation for procedures aimed at fostering the development of creative giftedness. The cognitive approach also offers theoretical insights that suggest possibilities for further research (such as the speculations on arousal and creativity outlined earlier). Thus, it has a place in discussions on cognition of the gifted.
Altshuller, G. S. (1984). Creativity as an exact science. New York: Gordon and Breach.
Amabile, T. M. (1983). The social psychology of creativity. New York: Springer.
Anderson, J. R. (1976). Language, memory and thought. Hillsdale, NJ: Erlbaum.
Arieti, S. (1976). Creativity: The magic synthesis. New York: Basic Books.
Ayman-Nolley, S. (1992). Vygotsky's perspective on the development of imagination and creativity. Creativity Research Journal, 5, 101-109.
Barron, F. X. (1963). Creativity and psychological health. New York: Van Nostrand.
Barron, F. X. (1969). Creative person and creative process. New York: Holt, Rinehart and Winston.
Bartlett, F. C. (1932). Remembering. Cambridge: Cambridge University Press.
Bloomberg, M. (1967). An inquiry into the relationship between field independence-dependence and creativity. Journal of Psychology, 67, 127-140.
Bruner, J. S. (1967). On knowing. Essays for the left hand. New York: Athaneum.
Case, R. (1978). Intellectual development from birth to adulthood: A neo-Piagetian interpretation. In R. S. Siegler (Ed.), Children's thinking: What develops? (pp. 37-71). Hillsdale, NJ: Erlbaum.
Chi, M. T. H., Glaser, R., & Farr, M. J. (Eds.). (1988). The nature of expertise. Hillsdale, NJ: Erlbaum.
Commons, M. L., Richards, F. A., & Kuhn, D. (1982). Systematic and metasystematic reasoning: A case for levels of reasoning beyond Piaget's stage of formal operations. Child Development, 53, 1058-1069.
Cropley, A. J. (1967). Creativity. London: Longman.
Cropley, A. J. (1992). Fostering creativity in the classroom. Norwood, NJ: Ablex.
Cropley, A. J. (1994). Creative intelligence: A concept of "true" giftedness. European Journal for High Ability, 5, 6-23.
Csikszentmihalyi, M. (1996). Creativity: Flow and the psychology of discover and invention. New York: Harper Collins.
deBono, E. (1970). Lateral thinking. New York: Harper and Row.
Eisenberger, R., & Armeli, S. (1997). Can salient reward increase creative performance without reducing intrinsic creative interest? Journal of Personality and Social Psychology, 72, 652-663.
Eisenberger, R., & Cameron, J. (1996). Detrimental effects of reward: Reality or myth? American Psychologist, 51, 1153-1166.
Facaoaru, C. (1985). Kreativitat in Wissenschaft und Technik [Creativity in science and technology]. Bern: Huber.
Finke, R. A., Ward, T. B., & Smith, S. M. (1992). Creative cognition. Boston, MA: MIT.
Gardner, H. (1995). Leading minds: An anatomy of leadership. New York: Basic Books.
Getzels, J. A., & Jackson, P. W. (1962). Creativity and intelligence. New York: Wiley. Ghiselin, B. (1952). The creative process. Los Angeles: University of California Press.
Grudin, R. (1990). The grace of great things: Creativity and innovation. New York: Ticknor and Fields.
Guilford, J. P. (1950). Creativity. American Psychologist, 5, 444-454.
Heinelt, G. (1974). Kreative Lehrer / kreative Schuller [Creative teachers/creative students]. Freiburg: Herder.
Helson, R. (1996). In search of the creative personality. Creativity Research Journal, 9, 295-306.
Jackson, P. W., & Messick, S. (1965). The person, the product, and the response: Conceptual problems in the assessment of creativity. Journal of Personality, 33, 309-329.
Koestler, A. (1964). The act of creation. New York: Macmillan.
Kubie, L. (1958). Neurotic distortion of the creative process. Lawrence, KS: University of Kansas Press.
Martinson, O. (1995). Cognitive styles and experience in solving insight problems: Replication and extension. Creativity Research Journal, 8, 291-298.
Mednick, S. A. (1962). The associative basis of creativity. Psychological Review, 69, 220-232.
Messick, S. (1994). The matter of style: Manifestations of personality in cognition, learning, and teaching. Educational Psychologist, 29, 121-136.
Miller, G. A. (1956). The magical number seven, plus or minus two: Some limits on our capacity for processing information. Psychological Review, 63, 81-97.
Morgan, D. N. (1953). Creativity today. Journal of Aesthetics, 12, 1-24.
Mumford, M. D., & Gustafson, S. B. (1988). Creativity syndrome: Integration, application, and innovation. Psychological Bulletin, 103, 27-43.
Mumford, M. D., & Whetzel, D. L. (1996). Insight, creativity, and cognition: On Sternberg and Davidson's The nature of insight. Creativity Research Journal, 9, 103-107.
Mumford, M. D., Baughman, W. A., Maher, M. A., Costanza, D. P., & Supinski, E. V. (1997). Process-based measures of creative problem-solving skills: IV. Category Combination. Creativity Research Journal, 10, 59-71.
Myers, I. B., & McCaulley, M. H. (1985). Manual: A guide to the development and use of the Myers-Briggs Type Indicator. Palo Alto, CA: Consulting Psychologists Press.
Nichols, J. G. (1972). Creativity in the person who will never produce anything original and useful. The concept of creativity as a normally distributed trait. American Psyschologist, 27, 717-727.
Policastro, E. (1995). Creative intuition: An integrative review. Creativity Research Journal, 8, 99-114.
Puccio, G. J., Treffinger, D. J., & Talbot, R. J. (1995). Exploratory examination of relationships between creativity styles and creative products. Creativity Research Journal, 8, 157-172.
Resnick, L. B. (1987). Education and learning to think. Washington, DC: National Academy Press.
Rosenblatt, E., & Winner, E. (1988). The art of children's drawing. Journal of Aesthetic Education, 22, 3-15.
Rostan, S. M. (1994). Problem finding, problem solving and cognitive controls: An empirical investigation of critically acclaimed productivity. Creativity Research Journal, 7, 97-110.
Rothenburg, A. (1988). Creativity and the homospatial process: Experimental studies. Psychiatric Clinics of North America, 11, 443-460.
Runco, M. A., Okuda, S. M., & Thurston, B. J. (1991). Environmental cues and divergent thinking. In M. A. Runco (Ed.), Divergent thinking (pp. 79-85). Norwood, NJ: Ablex.
Simon, H. A. (1989). The scientist as a problem solver. In D. Klahr & K. Kotovsky (Eds.), Complex information processing (pp. 375-398). Hillsdale, NJ: Erlbaum.
Simonton, D. K. (1988). Scientific genius: A psychology of science. New York: Cambridge University Press.
Simonton, D. K. (1994). Greatness: Who makes history and why? New York: Guilford.
Sloman, S. A. (1998). Categorial inference is not a tree: The myth of inheritance hierarchies. Cognitive Psychology, 35, 1-33.
Smith, G. J. W., & Carlsson, 1. (1983). Creativity in early and middle school years. International Journal of Behavioral Development, 6, 167-195.
Sternberg, R. J., (1988). A three-facet model of creativity. In R. J. Sternberg (Ed.), The nature of creativity (pp. 125-147). New York: Cambridge University Press.
Sternberg, R. J. &Dowing, D. C. (1982). The development of higher order reasoning in adolescence. Child Development, 53,209-221.
Taylor, I. A. (1975). An emerging view of creative actions. In I. A. Taylor, & J. W. Getzels (Eds.), Perspectives in creativity (pp. 297-325). Chicago: Aldine.
Tweney, R. D. (1996), Presymbolic process in scientific creativity. Creativity, Research Journal, 9, 163-172.
Wallas, G. (1926). The art of thought. New York: Harcourt Brace.
Weisberg, R. W. (1986). Creativity. New York: Freeman.
Manuscript submitted December, 1997.
Revision accepted December, 1998.
Arthur Cropley completed his undergraduate studies in Australia and was a school teacher there and in England and Canada for seven years. After graduate study in Canada, he taught at universities in Australia, Canada, and Germany until retiring in 1998. He is the author of 20 books and a large number of chapters and articles, and received the 1997 Creativity Award of the World Council for Gifted and Talented Children.…