Critique of Morelock's (1996) definitions of giftedness and talent.

This article discusses Morelock's (1996) attempt at differentiating giftedness from talent. In the case of the giftedness concept, she adopts the definition of the Columbus Group. A careful analysis of that definition shows many major flaws, among them the fact that it chooses precocious development instead of outstanding human abilities as the essence of giftedness, that it introduces vaguely defined concepts (e.g., heightened sensitivity, inner experiences, vulnerability), and that it affirms the qualitatively different nature and uniqueness of gifted individuals, yet uses numerous quantitative expressions to describe that attribute. Morelock's definition of talent is judged equally flawed, mainly because it maintains the confusion between the two concepts by defining both as potential, destroys the normative meaning of the concept of talent by endorsing the "talent for all" ideology, and unduly reduces the breadth of the concept by including creative productivity as a necessary component of talent. In the last part, Gagne's Differentiated Model of Giftedness and Talent is presented as a more logical and adequate solution to Morelock's attempt at "imposing order on chaos."

The aim of this text is to discuss some of the ideas presented in a Roeper Review article by Morelock (1996), titled On the nature of giftedness and talent: Imposing order on chaos. For the benefit of those who have not read it, I will first summarize its major arguments, focusing on those specific points I intend to address in my critique.

Overview of Morelock's (1996) Proposals

In the first paragraph of her text, Morelock states the main questions she will address as follows: "Who are the gifted and who are the talented? And are the two categories mutually exclusive? When we seek to meet the needs of "gifted" and "talented" children, what does that mean?" (1996, p. 4). She begins her analysis by painting a bleak picture of the semantic situation, focusing mainly on the non-differentiation between the terms giftedness and talent. That analysis closely resembles the portrait I traced when I first presented my differentiated model of giftedness and talent (see Gagne, 1985). She then shows through the work of Galton, Binet, Terman, and Hollingworth that these early leaders in the study of gifted individuals created "a germinal seed--an undifferentiated mass of conceptual understanding with potential for spawning a number of contrasting and contradictory paths for research and practice" (p. 5). She then argues that her brief historical overview shows:

a confusion of two themes in writing about "the gifted

child." The first theme identified "giftedness" as an

unusual generalized capacity for judgment and abstract

reasoning revealing itself in childhood. The primary significance

of this was thought to be that it presaged extraordinary

adult achievement across fields of endeavor.

The second theme acknowledged "giftedness" as intellectual

development surpassing that expected for a

child's chronological years. This resulted in emotional

vulnerability and educational and social needs different

from those of agemates. These two themes have, over the

years, evolved as two intermeshed strands of thought

and work." (p. 6)

Morelock describes in some detail these two strands, which she names "the gifted achiever strand" and "the gifted child strand" respectively. In her description of the evolution of the first strand, she includes the debate over the meaning of the IQ concept, not only its alleged racial and socio-economical biases or the relative importance of nature and nurture in determining individual differences in cognitive abilities, but also the debate between defenders of a "g" factor and proponents of multiple (and independent) cognitive abilities, as well as the debate over the inclusion of creativity as a necessary component of the giftedness construct. She associates this strand with the recent "talent development movement" promoted by various scholars (e.g., John Feldhusen, Don Treffinger, and Joe Renzulli). The second strand is, in her view, uniquely exemplified by what she describes as "the Columbus group vision" (p. 8), a position operationalized through the following very special definition of giftedness.

Giftedness is asynchronous development in which

advanced cognitive abilities and heightened intensity

combine to create inner experiences and awareness that

are qualitatively different from the norm. This asynchrony 

increases with higher intellectual capacity. The uniqueness

of the gifted renders them particularly vulnerable

and requires modifications in parenting, teaching and

counseling in order for them to develop optimally. (p. 8)

The appropriateness of this definition will be discussed in detail below. In the last part of her article, Morelock tries to create a rapprochement between the two strands. This is where she gives her definition of talent, differentiating it from the Columbus Group definition of giftedness.

"Giftedness" is needed to identify the particular form of

asynchronous development referenced in the Columbus

Group definition ... "talent" is needed to refer to multileveled

potential for domain-specific creative-productivity

in the world which can be fostered through appropriate

identification and environmental support." (p. 10)

Concerning the nature of the talent concept, she adds that "we need to assume that some level of talent is present in all children, and it is our responsibility to discover and foster it in all children." (p. 10) The last paragraphs are devoted to answering various questions (e.g., What is the proper place for the ideas espoused by the Talent Development Movement ... by the Columbus Group Movement? What directions of research are needed in terms of talent development?); they have no direct impact on the present discussion of the concepts of giftedness and talent.

Let us examine more closely Morelock's proposals. My comments are grouped into three major sections. In the first one, I take a close look at the Columbus Group definition of giftedness in order to show why it is in my view totally unacceptable as a definition of that construct. In the second section, I analyze Morelock's definition of talent, including its degree of differentiation from the Columbus Group definition of giftedness. In the last part, I briefly describe my own Differentiated Model of Giftedness and Talent (Gagne, 1985, 1993, 1995a), a very different--and, I believe, more logical and adequate-solution to Morelock's attempt at "imposing order on chaos."

Critique of the Columbus Group Definition of Giftedness

In reaction to various parts of the Columbus Group definition, I will try to demonstrate in this section that giftedness cannot be defined as an asynchronous (precocious) development, that some concepts (e.g., heightened intensity, inner experiences, awareness, vulnerability) are introduced without clear definitions nor clear differentiation from the concept of giftedness itself, and that there is nothing qualitative in the differences between gifted individuals and those of average abilities. Finally, I will trace the source of most of the Columbus Group's definitional errors to their improper focus on the very small subpopulation of extremely gifted (intellectually) children.

Development is not a constituent element of giftedness

We find in the first five words of the Columbus Group definition ("Giftedness is an asynchronous development") its major flaw, namely the assertion that the nature of giftedness IS a special type of development. But, first, a brief comment on the concept of asynchrony as used in that definition.

Asynchrony is in fact precocity. Taken literally, the concept of asynchronous development means that two (or more) things are developing at a different pace (asynchronous is Greek for "not at the same time"); thus, asynchrony implies comparisons between "things." These comparisons can take many forms. For instance, the two entities being compared can cover the whole range of living things: trees, vegetables, insects, athletes in training, children in school, and so forth. These comparisons can be made either between species, within species, but between individuals (normative asynchrony), or within individuals (ipsative asynchrony). Indeed, Terrassier (1989) himself, from whom the Columbus Group borrowed the concept of asynchrony--as Morelock herself acknowledges--used the ipsative/normative distinction to describe his six forms of asynchrony. Moreover, these comparisons can focus on any one of many different characteristics: size, ripeness, sexual maturity, abilities, temperamental predispositions, etc. Finally, asynchronism can manifest itself in either one of two directions: entity A can develop faster or slower than entity B. Two very common words express this difference in direction: precocity and retardation. My point is that the Columbus Group definition would have been less ambiguous and more user friendly if it had begun as follows: "Giftedness is a form of precocious development in which.... "But, why can't giftedness be equated with precocious development? There are at least two: a semantic one and a logical one.

A semantic incorrectness. The word giftedness has two parts: "gifted" and "ness". If one wishes to use this particular word to label an entity or phenomenon, that entity's characteristics must be congruent with the meanings conveyed by the label, each part of it. And what do these two parts suggest? In the English language, the "-ness" ending is used as "a noun-forming suffix meaning: condition, quality, or state of, as in greatness, sadness, togetherness" (Webster's, 1983, p. 1206). Stable and enduring characteristics of individuals (e.g., deafness, talkativeness, baldness, forgetfulness, inventiveness) clearly belong to that category. The same Webster's points out that words which describe processes, changes or actions usually end in "-ing" or "-ment," (e.g., improvement, learning, defining, bereavement, reacting, involvement). Thus, it is not semantically correct to use the label giftedness to designate a process, asynchronous or not.

The core of the giftedness concept. If the term giftedess suggests the idea of a stable human characteristic, what can it be? One thing is certain: the logic of the definitional process requires that it be related to the "gifted" component of the word giftedness. This component has more than one meaning. First and foremost, it means that the characteristic is given to individuals, not acquired by them. Secondly, this given attribute has to be something desirable; a gift is usually a pleasant surprise (except maybe some Christmas gifts!). Thirdly, it has to be an enduring characteristic, not something that can appear and disappear within days or weeks like a mood. Fourthly, usage and dictionaries have confirmed its normative meaning: giftedness sets apart a small group of individuals from the vast majority of the population.

Broadly speaking, stable human characteristics can be classified in two groups: abilities and personality traits. Again, usage confirms the lack of association between giftedness and personality: very rarely, if ever, do we see people described as having a gifted personality, or showing gifted optimism, gifted perseverance, or gifted concentration. Consequently, the core of the giftedness concept has to be related to human abilities. Exceptional or outstanding abilities thus satisfy three of the above conditions: they are desirable, enduring and normative. But, what about the given condition? That condition requires that in order to be qualified as gifts, outstanding human abilities must not be the result of any systematic training or learning process, as is the case with learning a school subject, playing a musical instrument, or mastering the skills of baseball. These are all examples of abilities that have been in large part acquired through formal learning, training and practice. Giftedness refers to the given or natural part of these abilities. As defined by Webster's, the term gifted means "endowed with a natural ability or aptitude" (1983, p.770). In times past, this endowment was judged to be a gift from God. Nowadays, thanks to hundreds of empirical studies (see Plomin, DeFries, & McClearn, 1990), the source of this endowment is more appropriately recognized as genetic in nature. In short, for the label giftedness to be properly used, it must designate outstanding human abilities which are, at least partly, a gift of nature. And there is no doubt in my mind (see Gagne, 1995a) that outstanding intellectual (and social, and physical) abilities deserve to be called gifts.

Being able to demonstrate logically that outstanding human abilities constitute the core component of the giftedness concept does not mean that this point of view is shared by professionals active in the field of gifted education. What evidence do we have that this is indeed the more common meaning given to the term giftedness? I could first refer the reader to Sternberg and Davidson's (1986) Conceptions of Giftedness; in this compendium of the then major definitions and models of giftedness, one can see this core component recurring again and again in most of the definitions presented. I could also point out that most dictionaries define the concept in terms very similar to those of Webster's citation above; this would show that the product of my logical analysis conforms at least with the "general" meaning given to this term. But, there is in my view no better proof than the widespread use of IQ tests to identify gifted students. Their use in most school districts and in most empirical studies for identification purposes, whether in the U.S., Canada or any other country, confirms that a high IQ is THE operational definition of giftedness. Finally, my last piece of evidence brings me back to the Columbus Group definition. Indeed, the next five words in that definition ("in which advanced cognitive abilities") mention precisely this core component. Could it be that the Columbus Group members consider both precocious development and advanced cognitive abilities to be constituent elements of the giftedness concept? If so, they err. Why can't a definition of giftedness simultaneously include advanced cognitive abilities and precocity?

Precocity is a consequence of giftedness. The answer is simple: because accelerated development is a consequence of possessing these outstanding mental abilities: giftedness causes precocity, not the inverse. And this applies not only to the domain of cognition or intellectual giftedness, but to all other giftedness domains. To be more precise, precocity is a consequence of a consequence of giftedness. Indeed, the most immediate consequence of possessing these high level natural abilities is ease of learning (Gagne, 1993; Li, 1995). Learning easily is, in my view, THE most evident manifestation of giftedness, whether the giftedness domain is intellectual, creative, socioaffective or physical. Ease of learning brings in turn speed in learning, and, viewed over time, speed in learning generates faster-paced development, or precocity. But, the rules of logic stipulate that the consequence of something cannot be at the same time a constituent of that entity. This is why I cannot equate giftedness with precocious development. What is very interesting is that the second sentence of the Columbus Group definition not only confirms what I have just stated, but, in so doing, contradicts the first one. By affirming that "asynchrony increases with higher intellectual capacities," the Columbus Group members create a causal relationship between intellectual giftedness as cause and precocious development as outcome. From their own definition, then, it can be seen that asynchrony is not giftedness, but a consequence of gifted abilities.

A limited view of giftedness. The Columbus Group's focus on precocity as the core component of their concept of giftedness poses another problem because it restricts the giftedness label to children and adolescents: precocity cannot apply to adulthood, except maybe to precocious senility! Since I believe that high natural abilities do not disappear when a person reaches adulthood, I cannot accept a definition that excludes a majority of the gifted population. This focus on children is by no means a bias specific to the Columbus Group members: it is a pervasive bias in our field. This is easily understood because most professionals work in education systems and their clients are children and adolescents. However, this situation should not affect our logic when it comes to defining our concepts. If we believe that giftedness applies to adults as well as children, our definition should clearly reflect this fact.

Another bias typical of our field is to equate giftedness with high cognitive abilities. As mentioned earlier, intellectual giftedness is only one form of giftedness. We just have to leave our own field and talk with people in others to discover that the concept of giftedness can be defined very differently. In sports, for example, giftedness--or, in their own words, "natural talent"--corresponds to high level physical abilities. In visual arts, giftedness is associated much more frequently with high creativity than with manual dexterity or hand-eye coordination (S. Kay, personal communication, November 27, 1996). Consequently, we should always specify in our language what type of giftedness is under scrutiny. Not acknowledging the existence of forms of giftedness other than advanced cognitive abilities is another flaw in the Columbus Group definition. Well, it does introduce "heightened intensities" as an added characteristic; but, is that a form of giftedness? This is another ambiguity to which we will now turn.

A trio of ambiguous concepts

The central part of the first sentence in the Columbus Group definition ("in which advanced cognitive abilities and heightened intensity combine to create inner experiences and awareness") introduces three concepts that I find too ambiguous to be easily translated into measurable constructs.

Heightened intensity. What is meant by "heightened intensity"? Intensity of what? These two quantitative terms should be associated with a "something" that is more intense, but nothing is mentioned. The Columbus Group definition also asserts that this heightened intensity "combines" with the advanced cognitive abilities to produce qualitatively different inner experiences and awareness. This means (1) that this "element" is distinct from advanced cognitive abilities, (2) that it plays as important a role as these advanced cognitive abilities in creating the qualitatively different inner experiences and awareness, and (3) that both are necessary to produce this qualitatively different outcome. This wording creates a first problem because if giftedness is equated with precocious development, then that label can no longer be used to describe these two elements as forms of giftedness. Yet, as we have seen above, advanced (or outstanding) cognitive abilities is the most logical (and probable) core constituent of giftedness.

Let us assume that the Columbus Group members accept that these advanced cognitive abilities are indeed part of the nature of giftedness. Then, what happens with heightened intensity? Is it another form of giftedness? It seems to be the case since it is clearly distinguished from, but put on the same causal level as advanced cognitive abilities. The wording of that part of the Columbus Group definition creates a conjunctive definition of giftedness, namely a definition which states that the label gifted requires the simultaneous presence of these two components, that they are both essential to the presence of giftedness. In other words, persons who demonstrate advanced cognitive abilities cannot be labeled gifted unless they also show heightened intensity. This is analogous to Renzulli's (1979) three-ring definition of giftedness, in which both high abilities, and creativity, and task commitment have to be present for the label to be applied. Is this what the Columbus Group members really want to convey? Why couldn't we apply to this situation a disjunctive definition of giftedness so that advanced cognitive abilities or heightened intensity are both independent abilities and lead to distinct forms of giftedness? The wording of that part of the definition leaves open many avenues for interpretation.

Inner experiences and awareness. The terms "inner experiences" and "awareness" are also fuzzy concepts; they are not easy to translate into psychometrically valid measures. Moreover, they are not clearly defined. What kinds of inner experiences do they include: thoughts, desires, feelings, beliefs, fears, all of them? What types of awareness do the Columbus Group members have in mind; sensory, cognitive, social, affective, intrapersonal, extrasensory? These two terms are commonly used by psychologists specializing in personality development to describe affective behavior. Is this what the Columbus Group members mean by using such terms? If so, it implies that advanced cognitive abilities have an impact outside of their normal realm, namely cognitive behavior. Or, could it be that this kind of outcome results more directly from the other component, namely heightened intensity? Moreover, if this outcome is a direct manifestation of giftedness, as stated in the Columbus Group definition, then it should be observed in all gifted children without exception. As we will show later, the meager evidence in support of such a statement comes essentially from clinical observations made with a very small subgroup of the most exceptionally gifted within the total gifted population.

Is giftedness a qualitative difference?

The last part of the first sentence of the Columbus Group definition states that the differences between gifted individuals and normal or average ones are qualitative in nature. This statement is reinforced in the definition by the use of the expression "uniqueness of the gifted" further on.

Qualitative versus quantitative. Are there really qualitative differences between gifted individuals and those of average abilities? Some might argue that it depends on how one defines "qualitative." But, in research methodology and statistics, there are recognized and agreed upon definitions for these terms: qualitative data correspond to differences in kind, differences that cannot be ranked or placed on a continuum of any sort. This is the case for instance with categories like man and woman, catholic and protestant, urban and rural, ipsative and normative, intellectual and physical abilities, and so forth. If qualitative data can be identified rather straightforwardly, the situation is quite different in the case of qualitative variables. In most cases, qualitative variables are artificially constructed from quantitative measurements. The gifted/non-gifted distinction is a good case in point, since program coordinators and researchers create these two groups by arbitrarily setting a threshold on the interval scale of IQ scores (e.g., [is less than] 130 as opposed to [is greater than or equal to] 130). Using dichotomous expressions (like obese/non-obese, gifted/non-gifted, or rich/poor) does not necessarily mean that they correspond to the discrete categories of a qualitative classification system. Most of the time, they are just a practical way of simplifying for everyday use more complex (and quantitative) distinctions.

Also, employing qualitative variables does not automatically guarantee that qualitative differences will be obtained. For instance, when Kanevsky (1990) compared two groups of children, intellectually gifted (IG) and average ability (AA) pupils, on two different versions of the well-known Tower of Hanoi problem, she created these two qualitative variables, one of them (IG/AA) the artificial dichotomization of a continuous distribution. This methodological decision did not ensure that any observed differences between the two groups or between the two experimental tasks could automatically be labeled qualitative. For example, the fact that the AA group made 85% correct moves while the IG group made 90% is by no means a qualitative difference. Kanevsky also collected various qualitative data, among them whether the children accepted or refused outside help. By itself, acceptance vs refusal is a qualitative measure; but ,when she reported that a larger percentage of IG children (28%) than AA children (7%) refused a hint at least once, that difference was a quantitative difference. In other words, the difference between the two groups was not a difference in kind but in frequency. Indeed, there was much overlap between the two groups, since a large majority of the children in both groups did not refuse such outside help. A qualitative difference in such a situation would have required that all IG children refuse outside help while none of the AA children did so. When do we observe in the literature such clear-cut differences between groups of IG children and groups of AA children? The answer is "never." In other words, qualitative differences are an exceedingly rare phenomenon, not only in educational research, but in most sciences. That's why statistics are so important!

Giftedness is a quantitative concept. Many of the terms used in the Columbus Group definition are quantitative in nature: asynchronous, advanced, heightened, intensity, higher, particularly. Note especially the second sentence: "This asynchrony increases with higher intellectual capacity." Is there a more quantitative statement? How can one defend the presence of qualitative differences and contend at the same time that the intensity of this "unique" behavior is directly related to the level of advanced cognitive abilities? How can quantitative differences suddenly transform themselves into qualitative ones? This makes no sense at all. All our measurements of human behavior are quantitative, and everything we assess as part of our identification procedures involves quantitative differences: "more" of this and "more" of that. The IQ tests and school grades which are the two major sources of information to select giftedness and talented students are perfect examples of quantitative measurements. Doesn't one get a 160 IQ by obtaining more right answers to more questions more rapidly? Aren't those questions ordered according to a quantitative index of difficulty? Indeed, the fact is that quantitative measures are the worst hurdle in trying to determine the prevalence of gifted individuals in the general population because nowhere on the normal curve is there a sign that says: "At this exact point giftedness begins." Even in sex/gender issues the situation is not clear. No doubt there are qualitative differences between men and women when we think of physical attributes. But, every psychological measure of gender differences, whatever the characteristic being assessed (e.g., desires, beliefs, drives, attitudes, abilities) show not only major individual differences within each gender, but also significant overlap between the two distributions. In the psychological realm, men and women do not differ qualitatively, but quantitatively.

Paraphrasing a comment from Roeper (1995), Morelock states that precocious development "captures the quantitative aspects of gifted development (i.e. a child's going through developmental milestones at a more rapid rate), but it fails to capture the qualitative dimensions (i.e. the complexity, depth and intensity of the inner experience accompanying those rapid changes)" (p. 9). Do the terms complexity, depth and intensity convey qualitative or quantitative differences? Is there not an infinite number of intermediate degrees or levels between the very simple and the very complex, between the very shallow and the very deep or between the very weak and the very intense? All the above arguments point in only one direction: there are NO qualitative differences whatsoever between gifted children, whatever the type of giftedness, and those with less outstanding abilities.

So, we should avoid any allusion to qualitative differences. Not only is such a statement false, but it conveys a dangerous image of general or moral superiority, of having been cast from a different, and better, mold. A colleague and friend (K. Hoekman, personal communication, December 12, 1996) helped me clarify the discomfort I felt each time I reread the Columbus Group definition. She observed that Morelock's repeated emphasis on the presence of qualitative differences implied some form of moral superiority. If we insist that gifted individuals' characteristics differ in kind rather than in degree, there is indeed a risk of conveying a false image of "natural" superiority. This false superiority/inferiority distinction plays directly into the hands of those who oppose special educational services for IG children (e.g., Margolin, 1994; Sapon-Shevin, 1994). Caution is imperative if we want to convey a correct image of the nature of the differences between IG individuals and those with average abilities.

Giftedness and vulnerability

The following comments concern the first part of the last sentence in the Columbus Group definition ("The uniqueness of the gifted renders them particularly vulnerable..."). Firstly, here is another concept whose manifestation is described quantitatively: the adverb "particularly" means that the gifted are "more" vulnerable than other individuals. Secondly, the term "uniqueness" conveys very strongly not only the idea of qualitative differences that I criticized above, but also the image of extreme exceptionality and superiority. The Webster's defines unique as "different from all others; having no like or equal" (1983, p. 1999). Thirdly, vulnerability is another concept which is most frequently encountered in the personal dynamics literature and, because of its subjective nature, is not particularly easy to operationalize. Dictionaries of psychology define it as a capacity of being wounded or injured (no doubt emotionally in the present context). The Columbus Group definition asserts that the uniquely different inner experiences and awareness of the intellectually gifted produce some form of vulnerability. How can a distinct superiority in one area create a deficiency in another?

Lastly, if such fragility exists, why is it so difficult to observe it empirically? Indeed, the quasi-totality of the research on the psychological characteristics of intellectually gifted children and adolescents contradicts this statement. Major reviews of the literature on this subject (e.g., Janos & Robinson, 1985b; Olszewski-Kubilius, Kulieke, & Krasney, 1988; Roedell, Jackson, & Robinson, 1980; Shore, Cornell, Robinson, & Ward, 1991) or large-scale studies (e.g., Terman & Oden, 1947) confirm that intellectually gifted children, adolescents and young adults obtain on any measure of motivation, social adaptation, popularity or psychological adjustment scores which are on average somewhat higher, or at least equivalent to those of average ability children. This seems to apply to most levels of giftedness or talent, even high ones. For instance, the numerous measures of personality adjustment taken as part of the Study of Mathematically Precocious Youth (e.g., Keating, 1976; Stanley, George, & Solano, 1977) have shown that these highly talented adolescents manifest indices of personal and social adjustment at least equal to intellectually average individuals whose chronological age approximates the mental age of these highly talented subjects. Similarly, various studies of radical accelerants, no doubt highly gifted and talented individuals, have shown them to be as mature psychologically as their older college peers (e.g., Fund for the Advancement of Education, 1957; Janos & Robinson, 1985a). In other words, there is little information to support this alleged heightened vulnerability of gifted individuals. The only subgroup which seems to be somewhat more at risk in terms of personal or social adjustment is comprised of exceptionally gifted individuals, those whose IQ exceeds 160 or so (Gross, 1993). But this is a very small subpopulation within the total population of intellectually gifted individuals. Could these individuals be the real target of the Columbus Group definition? I believe so, and this poses a serious problem of external validity, the subject of the last part of this first section.

A problem of external validity

Exceptionally gifted individuals, let's say those with IQs of 160 or more, are very rare. How rare? Their prevalence within the general population is approximately 1:30,000. If we similarly define the gifted population as the top 10% (IQs [is greater than or equal to] 120) of the general population, then the prevalence of exceptionally gifted individuals within the gifted population would still not exceed 1:3000. It amounts on average to just one such student in over a hundred homogeneous groups of gifted students. Even full-time teachers of the gifted would have few chances, in the course of their whole career, of teaching even one of them. In short, exceptional giftedness is a very rare phenomenon indeed.

Morelock does not hide the fact that the Columbus Group definition was strongly influenced by the characteristics of this very selective subgroup. Indeed, most of the research evidence she cites in support of that definition focuses on that particular subgroup of exceptionally gifted children. She specifically mentions the following sources of influence: Hollingworth's case studies of children above 180 IQ

(Hollingworth, 1942), Roeper' s sensitive portrayals of

gifted children (Roeper, 1995); Gross's (1993) studies of

exceptionally gifted children, the previously-cited case

study of "Jennie" (Morelock, 1991, in press) and studies

of profoundly gifted children (IQs 200+) and their families

(Morelock, 1995). Additional support has come from

clinical findings (Lovecky, 1994; Silverman, 1993; Terrassier,

1985) (p. 9).

The three clinicians mentioned at the end of the citation are well known for their work with highly and exceptionally IG children and adolescents. Such a selection bias seriously questions the external validity of the Columbus Group definition, that is the applicability to the whole gifted population (however defined and circumscribed) of any observations made from that very select subgroup. When we look at the top .1% strata within the gifted population, how justified are we to apply what we observe to the other 99.9% less exceptional group? How much do Nobel prize winners tell us about the more mundane creative endeavors of the thousands of researchers working in academic or commercial settings, researchers who are no doubt very talented, but much less so than this ultra-select group? Why is it that various phenomena, for instance vulnerability, seem to appear frequently in the clinical notes of psychologists in private practice, yet are not so easily observed in more structured data collections? Could it be that these psychologists are more prone to receive as clients children who have adjustment problems at home or at school? Those youngsters who achieve extremely well, who are active in all sorts of extracurricular pursuits, who have a pleasant relationship with parents and siblings, and who have their fair share of friends will certainly not lead their parents to seek professional advice. I would guess that the search for such advice is directly linked with the intensity of perceived problems, and this look for outside help increases if there is parallel evidence of exceptional cognitive precocity, as if high intellectual giftedness and personality problems should not cohabit. But, such a clientele constitutes a non-representative sample of intellectually gifted individuals, and professionals who help them should be well aware of that, and twice cautious about generalizing their observations to the whole population of intellectually gifted persons.

Another comment that comes to mind in view of the above discussion is to point out how much clearer the situation would be if the Columbus Group definition had included a threshold for inclusion into the gifted population, as they define it. What IQ level do Morelock and other members of the Columbus Group consider to be a minimum for a person to be labeled gifted? Because of their focus on exceptionally gifted children, I am very curious to know their answer, as well as the degree of consensus within the Columbus Group. No doubt that some professionals in the field would disagree with their chosen threshold, but the degree of selectivity of their giftedness concept would be clearly specified. Indeed, the ambiguity surrounding the Columbus Group definition is a good occasion to reaffirm the importance of including a prevalence estimate in any definition of a normative concept. When we speak of the "gifted" without specifying the subgroup being addressed (e.g., moderately, highly, exceptionally), it automatically implies that we are referring to the whole population. It would be very useful to know how much of the tail end of the normal curve is being "reserved" for gifted individuals. Estimates that can vary from 15% to 1% or 2% not only convey quite different messages but also have quite different implications for identification and servicing.

Finally, putting aside for a moment this serious problem of external validity, one might ask whether or not the Columbus Group definition could at least be an appropriate definition for the giftedness of these exceptionally gifted individuals. Unfortunately, most of the major flaws observed have nothing to do with its external validity. The problems we have mentioned remain totally relevant even when we apply that definition only to the. 1% top subgroup of exceptionally gifted individuals.

Summary

The Columbus Group definition of giftedness cannot be accepted because it suffers from at least four major flaws. Firstly, it incorrectly chooses precocious (asynchronous) development instead of outstanding human abilities as the essence of giftedness, thus disregarding the signification of both the "gifted" and "-ness" components of the term. Secondly, it introduces vaguely defined concepts (e.g., heightened sensitivity, inner experiences, vulnerability) whose relationships between them and with the core constituent are ambiguous. Thirdly, it affirms the qualitatively different nature and uniqueness of gifted individuals, yet uses numerous quantitative expressions to describe that characteristic. Fourthly, it does not specify the extension of the concept, that is the minimum degree of exceptionality which determines the threshold between those who may receive and those who many not receive the giftedness label.

Critique of Morelock's Definition of Talent

The second part of this critique addresses Morelock's definition of talent, and assesses her success in attaining her major goal, namely "imposing order on chaos" by distinguishing clearly the concepts of giftedness and talent. Let us first recall how Morelock differentiates them. She states that giftedness "is needed to identify the particular form of asynchronous development referenced in the Columbus Group definition ... "talent" is needed to refer to multi-leveled potential for domain-specific creative-productivity in the world which can be fostered through appropriate identification and environmental support" (p. 10). And she adds: "we need to assume that some level of talent is present in all children, and it is our responsibility to discover and foster it in all children" (p. 10). Before commenting on the degree of differentiation between the two concepts as defined, let us examine more closely Morelock's definition of talent. As we will see, each of its attributes (multi-leveled, potential, domain-specific creative-productivity) poses major problems.

Talent as a multi-leveled concept

By defining talent as a multi-leveled concept, Morelock obviously prepares her later statement to the effect that "some level of talent is present in all children" (p. 10). In other words, everyone is talented to some degree. A survey of numerous definitions of talent proposed over the last fifty years or so (see Gagne, 1985, for an overview) did not show a single instance of such dilution in meaning. As obscurely and ambiguously as the concept of talent was defined in the past, at least it always designated some form of above-average achievement. To my knowledge, no past scholar ever proposed that everyone was talented "to some degree." Unfortunately, this distortion of the talent concept has become more frequent within this last decade: it is not unique to the Morelock definition. In their apparent efforts to counteract accusations of elitism, to present our field in a more politically correct way, and to make our positions more palatable to a large majority of educators, a few scholars have made similar proposals, going even so far as to recommend that the giftedness concept be abandoned and replaced by this "new" definition of talent. As Feldhusen recently said: "I do believe that the term `gifted' is an albatross. I believe that we could live without it" (1994, p. 4). Similarly, Renzulli & Reis (1991) affirmed "that labeling students as `the gifted' is counter productive to the education efforts aimed at providing supplementary educational experiences" (p. 34).

The consequences of such proposals are, in my view, disastrous for our field. How can we still speak of the gifted and talented as a distinct population if the concept of talent applies to everyone? With such redefinitions, the conceptual structure of what giftedness and talent really are is being destroyed from within. I cannot express more strongly my opposition to such a major and unjustified redefinition of the concept of talent. If we want to talk about potential abilities that are exhibited by a large majority of persons, let us use the well-known term "aptitudes," which can be qualified (e.g., low, average, moderate), reserving the term gift for high aptitudes. Similarly, if we want to talk about abilities that most can develop and master, we should use the terms skills or "competencies," reserving the term talent for high level skills and competencies. As I affirmed elsewhere: "Talent is to a minority what competence is to a majority" (Gagne, 1995b, p. 354). In a similar fashion, gifts are to a few what aptitudes are to most.

A very particular interpretation of a "talent for all" conception is sometimes advanced; it corresponds to an ipsative viewpoint (Messick, 1989). This viewpoint uses within-persons comparisons as opposed to the between-persons comparisons typical of a normative viewpoint. In the ipsative viewpoint, a particular child's ability profile is analyzed in order to identify his/her personal strengths and weaknesses, with the aim of building that child's competencies around his/her strengths. While this goal is very worthwhile, labeling it talent development would significantly bias the basic--normative--meaning of the concept of talent. In other words, the "personal strengths" of a child should not be called talents if, normatively speaking, they correspond to average or below average abilities; they should be called competencies, and their development should be called "competence development." It is imperative, for the sake of terminological clarity, not to confuse these two frames of reference; and since the well-recognized basic meaning of talent is normative, it is the one we should use. Moreover, a good definition of talent should specify the minimum threshold giving access to the population of talented persons or, equivalently, the approximate size of that population.

Talent as a potential

The use of the term "potential" associates the concept of talent with aptitudes rather than with achievements, and questions its differentiation from giftedness. If talent differs from giftedness, as Morelock argues, what about the advanced cognitive abilities mentioned in the Columbus Group definition of giftedness, abilities which--as the Columbus Group itself recognizes implicitly--cause the child's precocious development? Are they not also a cognitive "potential" that can be used to develop various forms of domain-specific creative-productivity, like school achievement, scientific or technical knowledge, artistic or athletic achievement, and so forth? So, if both giftedness (in the form of these advanced cognitive abilities) and talent (as defined by Morelock) are potentialities, where is the difference between them? Nowhere does she answer that question. On the other hand, if the Columbus Group concept of giftedness is not a potential, then what is this thing called giftedness, and what is that potential called talent?

Another problem ensues from Morelock's definition of talent as a potential. How will she label those who achieve best in these domains: scientists who can brilliantly synthesize the knowledge of their field, musicians who master the technical difficulties of their art, lawyers who understand the intricacies of articles of law, car mechanics or physicians who can diagnose with uncanny ability the source of a car's or patient's symptoms, all others who excel in their respective field of activity? Without the term talent, we are left without any appropriate label to designate the outstanding realizations attained through to the development of high potentialities. In everyday language, the word talent serves just this purpose: it is one of its functions (alas, not the only one!). An outstanding student is called academically talented, a very good musician is labeled talented, as will be any other person showing high level achievements in any particular field of human activity. My point is this: if there is any need for two distinct concepts, it is to differentiate between potentialities and their realizations, between aptitudes and competencies. And this is exactly how the common sense of a majority of people explains the difference between these two concepts (Gagne, Motard, & Belanger, 1991).

The problem with everyday language is that it tends to use the term talent to label both ends of the continuum that extends from natural abilities to systematically developed ones: talent often means "natural" talent. Nothing is perfect! This is indeed similar to the frequent confusion between jealousy and envy in everyday language, in which the former term is commonly used to represent both concepts. The use of the term talent to designate both natural abilities and systematically developed ones is not limited to everyday language; it can also be observed among experts in the field of gifted education. For instance, when Feldhusen proposes, as we mentioned earlier, to substitute for giftedness the term talent, he demonstrates this confusion. I might point out that the term giftedness has also been used to designate both a potential and an achievement, as in Renzulli's (1979) definition: "Gifted and talented children are those possessing or capable of developing this composite set of traits" (p. 23). Similarly, Tannenbaum (1983) affirms that "producing knowledge with great inventiveness [sic] and impact, which is a sign of giftedness, comes later in a person's growth cycle" (p. 86). But, I have never observed that tendency outside of our field; in arts and sports, indeed everywhere but in education, the term talent is by far the most popular descriptor of excellence.

Why is this double usage of talent as both potential and achievement so popular? Why is it that outside of education so few use the words gifted or giftedness? In fact, even within our field, the clients themselves are the first to shun the label. How often do we see adolescents who participate in "gifted" programs acknowledging their giftedness? Rarely. Even those who have achieved exceptionally well in school or related areas will strongly deny being gifted. I am not sure if this discomfort with the term gifted has ever been clearly assessed. My own hypothesis is that the frequent tendency by many experts in our field to idealize giftedness is a major deterrent to its generalized used by both youth and adults. Here, idealization refers to the habit of using as examples of giftedness the most exceptional behavior, the extremes of precocity, the headline-making achievements, just as if we were journalists trying to capture the attention of a blase society. This elitist tendency implicitly conveys the idea that the appropriate prototypical exemplars of giftedness are the likes of Einstein, Mozart, Picasso, Freud, and such extraordinary luminaries. Who among us would acknowledge membership in such a select group? The Columbus Group definition of giftedness is a case in point; with its hyperbolic descriptions it tends to set apart a very small and select group of highly marginal individuals who are "unique" and "qualitatively" different from the average. As mentioned above, it sounds like the description of a special caste of superior human beings. On the other hand, by defining talent as a characteristic that is present in almost everyone, as Morelock and a few others do, we facilitate its use as a personal qualifier: those who see themselves as being talented will not perceive that they are singling themselves out too much. But, in doing so, we are slowly destroying the chances for the dissemination of a very useful distinction between potentialities and realizations.

Talent as domain-specific creative-productivity

My last comment concerns the role of high creativity as a necessary component of outstanding achievements. Contrary to Morelock's positions, I believe that high creativity is not an essential building block for many fields of talent--as is the case, by the way, with any of the natural abilities. And this statement applies to all levels of talent, even exceptional talent. The importance of creativity is a very entrenched form of bias toward the gifted or talented elite on the part of many scholars and professionals in our field; creativity's stature was reinforced in no small part by Renzulli's three-ring definition (Renzulli, 1979), in which creativity is an essential constituent of giftedness. This bias is also maintained by the aforementioned tendency of many scholars to hyperbolize their exemplars of giftedness and talent, and to focus on fields of human activity in which creativity is prominently at work: architecture, scientific research, technological inventions, music composition, and so forth. At the same time, they ignore whole fields in which excellence, even exceptional excellence, does not require any special creative input; the most conspicuous of them is athletics and sports. Athletes do not need any special creative abilities to run faster, throw farther, jump higher, and break world records doing so; yet, these champions are perfect examples of exceptional talent. Outside of sports, examples are still numerous: highly talented surgeons who master their skills more than most of their colleagues, yet have not invented a new procedure, professional musicians who play with excellent technique and emotionality, yet have not composed, teachers who have this rare ability to translate difficult concepts into clear and simple language and images, yet have never invented a new pedagogy. In more mundane fields of human activity, no special creativity is needed to become talented as a translator or an interpreter, as a business secretary, a mechanic or carpenter, a financial analyst, an actor, a salesperson, etc. In short, while I do agree that high creativity is an important component of excellence in many fields of human activity, it is by no means essential, or even relevant, in many others.

Summary

The above comments demonstrate that Morelock's definition of talent is no more acceptable than the Columbus Group definition of giftedness. Not only does it maintain the confusion between the two concepts by defining both as potential, but it also destroys the normative meaning of the concept of talent by endorsing the "talent for all" ideology, and unduly reduces the breadth of the concept by including creative productivity as a necessary component of talent. Critics are often themselves criticized because they point out the flaws in other people's productions without proposing more appropriate alternatives. The last section of this article describes such an alternative, namely my Differentiated Model of Giftedness and Talent (DMGT). It is a developmental theory of the transformation of gifts into talents, which was first presented in 1985, and regularly revised and added to since then.

Gagne's Differentiated Model of Giftedness and Talent

In the DMGT, the term giftedness designates the possession and use of untrained and spontaneously expressed natural abilities (called aptitudes or gifts), in at least one ability domain, to a degree that places an individual at least among the top 10% of age peers. Thus, gifts are essentially outstanding natural aptitudes. By contrast, the term talent designates the superior mastery of systematically developed abilities (or skills) and knowledge in at least one field of human activity to a degree that places an individual within at least the top 10% of age-peers who are or have been active in that field or fields.

The DMGT specifies five aptitude domains (see Figure 1): intellectual, creative, socioaffective, perceptual/motor, and "others." These natural abilities, which have their origin in the genetic endowment, can be observed in every task children are confronted with in the course of their schooling. Let us mention, for instance, the intellectual abilities needed to learn to read, speak a foreign language or understand new mathematical concepts; the creative abilities needed to solve different kinds of problems and produce original work in science, literature and art; the physical abilities involved in sport, music or crafts; the social abilities which children use daily in their interactions with classmates, teachers, and parents. High aptitudes or gifts can be observed more easily and directly in young children because environmental influences and systematic learning have exerted their moderating influence in a limited way only. However, they still manifest themselves in adolescents and adults through the facility and speed with which individuals acquire new skills in any given field of human activity: the easier or faster the learning process, the greater the natural abilities. It is these natural abilities or gifts that many laypersons call "talent" or, more commonly, "natural talent."

[Figure 1 ILLUSTRATION OMITTED]

As defined in the DMGT, talents are the outcome of a developmental process; they progressively emerge from the transformation of high aptitudes into the well-trained and systematically developed skills characteristic of a particular field of human activity. These fields can be extremely diverse. Figure 1 shows some of the talent fields relevant to school-aged youth. A natural ability can express itself in many different ways, depending on the field of activity adopted by the individual. For example, dexterity as a natural physical ability can be modeled into the particular manual skills of a pianist, a painter or a video game player. Similarly, intelligence as a natural ability can be modeled into the scientific reasoning of a chemist, the game analysis of a chess player or the strategic planning of an athlete. In this model, natural abilities or aptitudes act as the "raw material" or the constituent elements of talents (Gagne, 1993). It follows from this relationship that talent necessarily implies the presence of well above average natural abilities; one cannot be talented without first being gifted. However, the reverse is not true: it is possible for well above average natural abilities to remain simply as gifts, and not to be translated into talents, as is witnessed by the well-known phenomenon of academic underachievement among intellectually gifted students. The process of talent development manifests itself when the child, adolescent or adult engages in systematic learning, training and practicing. The higher the level of talent sought, the more intensive these three activities will be.

This developmental process is facilitated (or hindered) by the action of two types of catalysts; intrapersonal and environmental These catalysts are not constituent elements of talents, but they contribute actively to their development. The intrapersonal catalysts are subdivided into physical and psychological factors, all of them under the partial influence of the genetic endowment. Among the psychological catalysts, motivation and volition play a crucial role in initiating the process of talent development, guiding it, and sustaining it through obstacles, boredom and occasional failure. Hereditary predispositions to behave in certain ways (temperament), as well as acquired styles of behavior (personality traits and disorders), also contribute significantly to support and stimulate, or slow down and even block, talent development. The environment manifests its significant impact in many different ways. The surroundings exert their influence both at a macroscopic level (e.g., geographic, demographic, sociological) and at a more microscopic level (size of family, socio-economic status, etc.). Many different persons, not only parents and teachers but also siblings and peers, may exert a positive or negative influence on the process of talent development. Gifted education programs within or outside the school belong to a category labeled undertakings; they are the most systematic form of undertaking that influences the process of academic talent development. Finally, significant events (e.g., the death of a parent, winning a prize or award, suffering a major accident) can influence markedly the course of talent development. Chance could be added as a fifth causal factor associated with the environment (see Tannenbaum, 1983); but, strictly speaking, it is more appropriately a characteristic of some of the elements placed in any of the other four categories (e.g., the "chance" of being born in a particular family; the "chance" of the school in which the child is enrolled deciding to develop a program for gifted and talented students). Chance factors also exert their influence in the aptitude and intrapersonal catalyst components of the DMGT through the process of genetic transmission.

A good definition of any normative concept (e.g., poverty, obesity, deafness) must specify to what extent subjects must differ from the norm to become members of the population covered by the label; this will automatically specify the prevalence of the population. In the DMGT, the threshold for both the giftedness and talent concepts is placed at around the 90th percentile (approximately 1.3 standard deviations above the mean); in other words, those who belong to the top 10% of the relevant reference group in terms of natural ability (giftedness) or achievement (talent) belong to the gifted or talented population. It must be clearly noted, however, that this generous choice of threshold is counterbalanced by a recognition of levels of giftedness or talent within the 10% population of "basically" gifted or talented persons, the DMGT recognizes four progressively more selective subgroups. They are labeled "moderately" (the top 1%), "highly" (the top 1 in a thousand), "exceptionally" (1:10,000), and "extremely" (the top 1:100,000) respectively. In the cognitive domain, the five groups correspond to approximate IQs of 120, 135, 145, 155, and 165. The intervention programs offered to gifted or talented students should take into account the level of the students' giftedness or talent as well as the domain(s) or field(s) in which it is sited.

Conclusion

While the previous discussion has shown much disagreement between Morelock's proposals and my own, there is at least one point of agreement:, namely that the field of gifted education is fraught with conceptual chaos and in dire need of order. Order means, in my view, not only agreement over the nature of our core constructs and the relationships between them, but also on the goals and principles that should guide the choice of appropriate developmental interventions for gifted and talented individuals. Agreement on the definitions of our major constructs, giftedness and talent, would constitute not only a good starting point but a much needed foundation on which to build a sound conceptual framework. Yet, for this concerted effort to happen, scholars need to keep an open mind when given valid objections to long-held positions, and to abandon the traditional policy of "conceptual parallelism." By this I mean the habit of proposing new ideas without taking into account those already advanced by others.

How can we expect to create a common conceptual edifice if we keep erecting little unconnected cabins one beside the other? Would it not be better to choose one of them as a promising candidate, discuss and agree on what needs to be corrected, then continue adding to it agreed upon pieces? I believe this to be the appropriate way to end up with a major housing complex instead of haphazardly built little bungalows. Is it not how science advances: current views are discussed, new hypotheses are put forward and tested, untenable positions are abandoned. It takes time, but little by little progress is made. No doubt that we should adopt such an approach.

Because of that conviction, I was very disappointed that Morelock chose to ignore my own model and build instead her own little bungalow. Not only would I have appreciated a detailed and thoughtful critique of those aspects she didn't like, but it would have been much more productive in the long run for our field. This is why I decided to write this article and discuss our divergent, sometimes opposite viewpoints. It is my conviction that if we want to become a respected subdomain of educational research, we have no choice but to start building our knowledge base in a more united and interconnected way.

(*) The preparation of this article was made possible thanks to a grant from the Social Sciences and Humanities Research Council of Canada. The author sincerely thanks Larry Coleman, Katherine Hoekman, Lannie Kanevsky, Sandra I. Kay, and Rena Subotnik for their precious comments to an earlier draft; these brought about major changes in content and tone. Yet, the opinions expressed remain the sole responsibility of the author.

REFERENCES

Feldhusen, J. F. (1994). No, it is more than terminology that needs changing: A rejoinder to Barbara Clark. NAGC Communique, VI (2), p. 4-5.

Fund for the Advancement of Education (1957). They went to college early. New York: Ford Foundation.

Gagne, F. (1985). Giftedness and talent: Reexamining a reexamination of the definitions. Gifted Child Quarterly, 29, 103-112.

Gagne, F. (1993). Constructs and models pertaining to exceptional human abilities. In K. A. Heller, F. J. Monks, & A. H. Passow (Eds.), International Handbook of Research and Development of Giftedness and Talent, (pp. 69-87). Oxford: Pergamon Press.

Gagne, F. (1995a). The differentiated nature of giftedness and talent: a model and its impact on the technical vocabulary of gifted and talented education. Roeper Review, 18, 103-111.

Gagne, F. (1995b). Hidden meanings of the "talent development" concept. Educational Forum, 59, 350-362.

Gagne, F., Motard, D., & Belanger, J. (1991). Popular conceptions of the differential characteristics and prevalence of gifted and talented persons. Paper presented at the Ninth World Conference on Gifted and Talented Children, held in August in The Hague, Netherlands.

Gross, M. (1993). Exceptionally gifted children. New York: Routledge.

Janos, P. M., & Robinson, N. M. (1985a). The performance of students in a program of radical acceleration at the university level. Gifted Child Quarterly, 29, 175-179.

Janos, P. M., & Robinson, N. M. (1985b). Psychosocial development in intellectually gifted children. In F. D. Horowitz, and M. O'Brien (Eds.) The gifted and talented: developmental perspectives (pp. 149-195). Washington, D. C.: American Psychological Association.

Kanevsky, L. S. (1990). Pursuing qualitative differences in the flexible use of a problem solving strategy by young children. Journal for the Education of the Gifted, 13, 115-140.

Keating, D. P. (Ed.) (1976). Intellectual talent: Research and development. Baltimore, MD: Johns Hopkins University Press.

Li, R. (1996). A theory of conceptual intelligence: Thinking, Learning, Creativity, and Giftedness. Westport, CT: Preager.

Margolin, L. (1994). Goodness personified: The emergence of gifted children. New York: Aldine De Gruyter.

Messick, S. (1989). Validity. In R. L. Linn (Ed.) Educational Measurement (3rd ed), pp. 13103. New York: Macmillan.

Morelock, M. (1996). On the nature of giftedness and talent: Imposing order on chaos. Roeper Review, 19, 4-12.

Olszewski-Kubilius, P. M., Kulieke, M. J., & Krasney, N. (1988). Personality dimensions of gifted adolescents: A review of the empirical literature. Gifted Child Quarterly, 32,347-352.

Plomin, R., DeFries, J. C., & McClearn, G. E. (1990). Behavioral genetics: A primer (2nd ed.). New York: Freeman.

Renzulli, J. S. (1979). What makes giftedness: A reexamination of the definition off the gifted and talented. Ventura, CA: Ventura County Superintendent of Schools Office.

Renzulli, J. S., & Reis, S. M. (1991). The reform movement and the quiet crisis in gifted education. Gifted Child Quarterly, 35, 26-35.

Roedell, W. C., Jackson, N. E., & Robinson, H. B. (1980). Gifted young children. New York: Teachers College Press.

Roeper, A. (1995). Annemarie Roeper : Selected writings and speeches. Minneapolis, MN: Free Spirit.

Sapon-Shevin, M. (1994). Playing favorites: Gifted education and the disruption of community. Albany, NY: State University of New York Press.

Shore, B. M., Cornell, D. G., Robinson, A., & Ward, V. S. (1991). Recommended practices in gifted education: A critical analysis. New York: Teachers College Press.

Stanley, J. C., George, W. C., & Solano, C. H. (Eds.) (1976). The gifted and the creative: A fifty-year perspective. Baltimore, MD: Johns Hopkins University Press.

Sternberg, R. J., & Davidson, J. E. (Eds.) (1986). Conceptions of giftedness. New York: Cambridge University Press.

Tannenbaum, A. J. (1983). Gifted children: Psychological and educational perspectives. New York: Macmillan.

Terman, L. M., & Oden, M. H. (1947). The gifted child grows up (Genetic studies of genius, vol. IV). Stanford, CA: Stanford University Press.

Terrassier, J.-C. (1989). Les enfants surdoues ou <<la precocite embarrassante>> (2e ed.) [Gifted children or "an embarrassing precocity" (2nd ed.)]. Paris: ESF.

Webster's. (1983). Webster's New Universal Unabridged Dictionary (2nd ed.). Cleveland, OH: New World Dictionaries.

Francoys Gagne is Professor, Department of Psychology at the University of Quebec at Montreal, Montreal, Canada and a member of the Editorial Advisory Board of the Roeper Review.

Manuscript submitted June, 1996. Revision accepted January, 1997.